Debating Dispositions: Issues in Metaphysics, Epistemology and Philosophy of Mind 9783110211825, 9783110184037

Citation preview

Debating Dispositions

Debating Dispositions Issues in Metaphysics, Epistemology and Philosophy of Mind

Edited by Gregor Damschen, Robert Schnepf and Karsten R. Stüber

Walter de Gruyter · Berlin · New York

Printed on acid-free paper which falls within the guidelines of the ANSI to ensure permanence and durability. Library of Congress Cataloging in Publication Data A CIP catalogue record for this book is available from the Library of Congress. ISBN 978-3-11-018403-7 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. © Copyright 2009 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printed in Germany Cover design: Martin Zech, Bremen Printing and binding: Hubert & Co GmbH & Co KG, Göttingen

V

Acknowledgements The anthology grew out of the spirited discussion and philosophical collegiality among the participants of a conference on dispositions that took place at the Leucorea in Wittenberg in summer 2006. With a few exceptions, all of the articles in this anthology were first presented as talks and then later revised in light of the debate at this conference. The conference was funded by the Deutsche Forschungsgemeinschaft (DFG), the Association of Friends and Supporters of the Martin-Luther-University Halle-Wittenberg and Walter de Gruyter publisher. We are grateful for their generous support. We also would like to express our gratitude to the Böhringer-Ingelheim Foundation and the College of the Holy Cross, USA for grants supporting the publication of this anthology. We would particularly like to thank Gunnar Schumann who energetically helped with the organization of the Conference in Wittenberg and who, together with Jens Gillessen, formatted the texts according to the publisher’s guidelines. In this context we also have to acknowledge appreciatively Rainer Enskat, who supported this project from the very beginning with his professional advice and effort. The cooperation with the publishing house has been friendly, efficient, and exemplary. Sabine Vogt and Christoph Schirmer have been in charge of this project in the first phase, while Gertrud Grünkorn and her team have supervised the project on the editorial level later on in a friendly and cooperative manner. Last but not least, we would like to thank the contributors to this anthology: Without their willingness to participate in the conference and to contribute to the anthology, without the intensive and stimulating discussions after each and every talk, and without their latent disposition toward collegiality that instantly manifested itself in the context of this conference this anthology on disposition could not have been realized. The Editors

Table of Contents Introduction ..................................................................................................IX Damschen/Schnepf/Stüber I. The Discovery of Dispositions: Ancient Foundations Knowledge and Virtue as Dispositions in Plato’s Theatetus..................... 3 Francisco J. Gonzalez Aristotle’s Theory of Dispositions: From the Principle of Movement to the Unmoved Mover............................................................................... 24 Ludger Jansen Dispositions in Greek Historiography...................................................... 47 Burkhard Meißner II. The Debate about Dispositions from the Beginning of Modern Sciences to the 20th century The Dispositions of Descartes................................................................... 71 Peter Machamer Explicable Explainers: The Problem of Mental Dispositions in Spinoza’s Ethics............................................................................................. 79 Ursula Renz Harmonizing Modern Physics with Aristotelian Metaphysics: Leibniz’s Theory of Force............................................................................................ 99 Michael-Thomas Liske From Ordinary Language to the Metaphysics of Dispositions – Gilbert Ryle on Disposition Talk and Dispositions ............................. 127 Oliver R. Scholz

VIII

Table of Contents

III. Contemporary Philosophical Analyses of Dispositions Hic Rhodos, Hic Salta: From Reductionist Semantics to a Realist Ontology of Forceful Dispositions ......................................................... 145 Markus Schrenk Ascribing Dispositions .............................................................................. 168 Stephen Mumford Dispositional Pluralism ............................................................................. 186 Jennifer McKitrick Dispositions and Their Intentions........................................................... 204 Andrea Borghini IV. The Role of Dispositions in Scientific and Philosophical Contexts Dispositions in Physics.............................................................................. 223 Andreas Hüttemann The Role of Dispositions in Historical Explanations........................... 238 Robert Schnepf Empathy, Mental Dispositions, and the Physicalist Challenge........... 257 Karsten Stueber Dispositional Knowledge-How versus Propositional Knowledge-That......................................................................................... 278 Gregor Damschen Epistemic Virtues and Cognitive Dispositions...................................... 296 David Henderson and Terry Horgan The Epistemic Function of Virtuous Dispositions .............................. 320 Elke Brendel List of Contributors ................................................................................... 341

Introduction GREGOR DAMSCHEN, ROBERT SCHNEPF, KARSTEN STUEBER Ordinary language and scientific discourse are filled with linguistic expressions for dispositional properties such as “soluble,” “elastic,” “reliable,” and “humorous.” We characterize objects in all domains – physical objects as well as human persons – with the help of dispositional expressions. Hence, the concept of a disposition has historically and systematically played a central role in different areas of philosophy, ranging from metaphysics to ethics. In this context one only needs to think of the important function that the concept of potentiality has in Aristotle’s metaphysics and the central role that the concept of a habitus plays in Aristotelian ethics. Yet, according to the orthodox view, ever since modern times the status of dispositions has been ontologically and epistemologically suspect. From the perspective of the mechanistic sciences of the 17th and 18th century, Aristotelian potentialities were generally regarded as occult qualities being of no explanatory help for our understanding of how the world works. Philosophically equally influential has been Humean empiricism and its epistemological skepticism regarding the existence of causal powers. Within the context of 20th century philosophy, particularly due to the influence of logical positivism, those Humean inclinations have persisted in that one generally felt that dispositional properties cannot be regarded as being ontologically autonomous. Moreover, one felt that dispositional talk could be regarded as cognitively significant only if one could show it to be analyzable in terms of semantically less objectionable notions. No wonder then that in the century of logical analysis the history of the concept of disposition is to a large extent characterized by a discussion about various attempts to semantically analyze dispositional language. So far, none of the proposed analyses seems to be without its shortcomings. Accordingly, from the perspective of semantic analysis, the status of dispositional language appears to be anything but settled. Yet in recent years, various philosophers have started to question the negative attitude towards dispositions and have begun to argue for a serious reevaluation of the philosophical presuppositions responsible for the modern suspicions about dispositions and dispositional terminology. Some philosophers have maintained that dispositions and dispositional terminology are on par with non-dispositional properties and predicates. Some have even

X

Introduction

more strongly suggested that dispositions are ontologically more basic than non-dispositional properties. In short, the current philosophical climate is again well disposed towards dispositions. Consequently, most of the articles in this anthology reflect this friendlier attitude toward dispositions. Indeed various authors argue explicitly against the modern and Humean prejudice regarding dispositions and claim that dispositions have to be regarded as part of the basic furniture of the universe. Three interrelated systematic topics are at the foreground of the contemporary discussion about dispositions; that is, semantic, epistemic, and ontological considerations. The semantic question concerns the issue of whether or not statements like “the glass is fragile” can be completely analyzed in terms of notions that are both epistemically and metaphysically innocuous. Within the context of 20th century philosophy that meant that they have to be analyzed in terms of notions that are acceptable to Humeans, who abhor the postulation of necessary connections and causal efficacy in nature. Particularly important in this context has been the discussion about various proposals of analyzing dispositional statements in terms of counterfactual conditionals (e.g. David Lewis) and whether or not such analyses can be shown to be immune to counter-examples. As one of the authors in this anthology suggests the discussion of such counterexamples has reached “folkloric status” within the context of analytic metaphysics. The epistemic problem considered in this context concerns the question of whether or not one is justified in ascribing dispositions even though they are empirically not directly accessible. From a Humean perspective, only observable properties can justify the ascription of dispositional terms. Normally however the fragility of a glass is ascribed before it manifests its disposition. Finally, from an ontological perspective, one has focused on the question of whether – and if so, how – dispositional properties depend or supervene on non-dispositional or categorical properties. More specifically, one has been interested in debating of whether or not the existence of bare dispositions, that is, dispositions whose existence does not depend on categorical dispositions is possible. One also has investigated of whether or not the distinction between dispositional and non-dispositional categorical properties can be made in a principled manner, or whether all properties contain a dispositional element; such as, that the property of a triangle has the disposition to make us count up to three if we count a triangle’s corners. A number of articles in this anthology address and document the 20th century discussion about dispositions within these three dimensions; particularly the articles by Schrenk, Mumford, McKitrick, and Borghini in the third part. Yet in contrast to some of the recent books and anthologies that are primarily concerned with addressing and collecting specific semantic, epistemic, and ontological arguments for or against the existence of

Introduction

XI

dispositions, the invitation to debating dispositions – the title of this anthology – is more broadly conceived. It is the hope of the editors that the more welcoming attitude towards dispositions in the recent philosophical climate allows for a much more wide-ranging reflection on the nature of disposition by documenting in detail the importance of this concept in the history of philosophy from ancient to contemporary times. In this manner, we hope to open the contemporary discourse to insights gained in the history of philosophy. As some of the articles reveal, the result of such historical research is at times rather surprising. Moreover, the anthology broadens the perspective on dispositions not merely by including a historical dimension but by also addressing the issue of disposition in more localized contexts: Contexts, in which dispositional terminology play a central role, but which have been neglected in the current debate on dispositions. This negligence can be explained by the aforementioned general skepticism about dispositions and dispositional terminology. If one is generally skeptical about the validity of dispositional terminology, then differences among contexts, where dispositional terminology is used, do not seem to matter much. Yet, once the philosophical dominance of the general skeptical attitude towards disposition is alleviated, as is the case in the recent philosophical climate, a more detailed and localized discussion of dispositions becomes necessary. Even if dispositions can be shown on a very general level to be part of the furniture of the universe, this insight does not automatically imply that all dispositional terminology is immune from extinction. Certainly one is inclined to say that the solubility of salt in water is due to its internal physical structure. The solubility of salt might therefore be regarded as being reducible to some categorical lower order physical properties. Yet such reducibility might not be in the offing for mental dispositions such as belief, desires and so on. For that very reason and in order to recognize differences between types of dispositions, in this anthology, dispositions are not merely discussed on the most general level but the topic of disposition is also addressed in more localized circumstances. They are addressed in the context of epistemology, where dispositions have lately been much talked about by so called virtue epistemologists. They are also discussed in articles focusing on the philosophy of mind and and in articles addressing the nature of dispositions delineated by our folk psychological vocabulary. Moreover, questions that were already centrally important for Plato of whether the human mind and human knowledge has a propositional structure – as is assumed within the contemporary cognitive model of the mind – or whether the structure of mind and knowledge is fundamentally non-propositional and irreducibly “dispositional” are newly addressed in this anthology. Finally, the status of dispositions is illuminated by discussing their role in a natural science like physics – a discipline that has been skeptical about dispositions since the

XII

Introduction

foundations of modern science – and a human science such as history, since the attribution of “mentalities” and “character traits” to individual or collective agents such as nations seem to play a central role in historical explanations. The anthology is divided into four main sections. The contributions of the first part analyze the ancient foundations of the discussion about dispositions. In his “Knowledge and Virtue as Dispositions in Plato’s Theaetetus,” FRANCISCO GONZALEZ argues that already Plato acknowledges the reality of dispositions. As he shows in his illuminating interpretation of the Theaetetus, Plato conceives of knowledge essentially in dispositional terms. Accordingly, without counting dispositions among the things that fully are, as Plato explicitly does in his Sophist, he would not be able to make ontological sense of the possibility of knowledge. As Gonzalez concludes, “the epistemology of the Theaetetus can be said to require the ontology of the Sophist.” LUDGER JANSEN further expands the exploration of the ancient reflection on dispositions by analyzing in detail the most comprehensive account by an ancient philosopher in his “Aristotle’s Theory of Dispositions: From the Principle of Movement to the Unmoved Mover.” He situates Aristotle’s theory within its linguistic and philosophical contexts and delineates its wide-ranging conceptual framework for analyzing the nature of dispositions. The precise nature of Aristotle’s ontological commitments regarding the existence of possibilia or mere potentialities is further explained in comparison to the Megarian position that denies that such entities exist and in comparison to the influential interpretation by Nicolai Hartmann and Jaakko Hintikka. For Jansen, Aristotle succeeds in providing a “consistent ontology of causal properties with an enormous explanatory appeal.” The final essay in this part, BURKHARD MEIßNER’S contribution “Dispositions in Greek Historiography,” closely analyzes the philosophical foundations and the use of dispositional terminology in the texts of ancient historians, particularly Plutarch, Herodotus, Thucydides, Xenophon, and Polybius. As he argues and concretely illustrates, ancient historians utilize the attribution of dispositional character traits as their central explanatory strategy in their historical narratives. Yet, the use of such dispositional terminology is not merely determined by their explanatory interests, but is also colored by the educational and moral interests that motivated ancient historians to write their historical narratives in the first place. The second part of the anthology examines the problem of disposition within the context of the foundation of modern science and analyzes this dispute up to the 20th century. The view is rather widespread that modern science with its mechanist paradigm simply had no use for the scholastic talk of dispositions, faculties, capacities, essences, and natures.Corresponding to

Introduction

XIII

the friendly attitude towards dispositions dominant in contemporary debates, this picture of early-modern philosophy can be slightly modified. PETER MACHAMER argues in his “The Dispositions of Descartes” that Descartes, despite his otherwise sceptical attitude towards scholastic notions, had a manifest need for dispositions, especially in his natural philosophy. Machamer illustrates this point with the help of Descartes’ reflection on states of equilibrium in his mechanics. Moreover, Machamer argues that Descartes’ conception of God’s activity as recreating the whole nature in every instant constitutes a conceptual scheme that requires reference to dispositions. Only in this manner can we coherently understand the world we live in. URSULA RENZ provides in her “Explicable Explainers: The Problem of Mental Dispositions in Spinoza’s Ethics” a thorough analysis of the role of dispositions in Spinoza’s philosophy of mind. She starts with the observation that on an ontological level Spinoza’s necessitarianism leaves no room for the reality of possibility and, consequently, the reality of dispositions. On the other hand, Spinoza uses dispositional terms within his philosophy of mind in order to account for mental states and actions. According to Renz, Spinoza’s view can be made sense of if one understands dispositions as “explicable explainers.” They are “properties” that serve an epistemic function in explanatory contexts, but they can be explained by other more basic properties of an individual. In this manner, Spinoza’s rejection of the mind as a cause behind its activities and as bearer of mental properties can be reconciled with his use of dispositions in explanatory contexts. The contribution “Harmonizing Modern Physics with Aristotelian Metaphysics: Leibniz’s Theory of Force” by MICHAEL-THOMAS LISKE discusses Leibniz’s philosophical reasons for an excessive use of dispositional terminology in an intellectual climate determined by modern science. Liske’s paper provides an overview of the different types of dispositions and their several functions within the physics and metaphysics of Leibniz. As Liske explains, Leibniz made the concept of force a central category of his metaphysical system in order to provide an answer to questions that in principle could not be answered within the mechanistic framework. Whereas the quantitative principles of modern science certainly are far superior in their predictive power than the qualitative principles of ancient and medieval science, they do not allow us to answer the question of why nature obeys one mechanistic law rather than another. It is in the context of such questions that Leibniz makes use of the dispositional concept of a force. He conceives of it in Aristotelian terms as an entelechy; a goal directed principle that is immanent in nature. OLIVER R. SCHOLZ’S article “From Ordinary Language to the Metaphysics of Dispositions: Gilbert Ryle on Disposition Talk and Dispositions” concludes the historical sections of this anthology by discussing

XIV

Introduction

the relevant claims of the philosopher who was one of the main figures most responsible for the revival of interest in the concept of disposition in the 20th century. Scholz shows how Ryle intended the Concept of Mind to be a prime example for a new philosophical method of linguistic analysis and provides a comprehensive account of Ryle’s analysis of dispositions. Specifically he shows how, for Ryle, statements about the meaning of disposition talk are intertwined with ontological claims about the nature of dispositions itself. This however is Ryle’s fundamental mistake; a mistake that confuses the meaning or sense of linguistic expressions with their reference. Accordingly, even though Ryle has been influential in rehabilitating dispositions as a topic of philosophical conversation, his own account shows severe deficits. As Scholz concludes, “more promising accounts had to await the return of scientific realism.” The articles of the third section of this anthology are focusing on issues that are the main topics of the current discussion about dispositions. MARKUS SCHRENK’S contribution “Hic Rhodos, Hic Salta: From Reductionist Semantics to a Realist Ontology of Forceful Dispositions” offers a detailed discussion of the several attempts of semantic reduction of dispositions documenting how every new analysis has provoked novel counterexamples. Instead of endlessly prolonging the debate about the proper analysis of dispositional statements, Schrenk suggests to challenge the Humean framework that has motivated the search for a semantic analysis in the first place. He opts for a version of dispositional realism. The difficult task for a dispositional realist consists in explicating the nature of dispositional powers. Schrenk argues that this notion can not be cashed out in terms of metaphysical necessity, as some philosophers have recently claimed. Rather, a different anti-Humean connection in nature has to do that job. Schrenk provides some tentative suggestions of how one could conceptualize this different connection and proposes a return to a Leibnizian notion of force. STEPHEN MUMFORD’S essay “Ascribing Disposition” continues the argument for dispositional realism. Yet in contrast to Schrenk, Mumford’s argument focuses on the epistemic problems traditionally associated with ascribing dispositions to objects or persons. Mumford argues that, pace Humean empiricism, we have good reasons for accepting an ontology which contains dispositions and powers that are seen as basic entities and as grounding necessary connections in nature. Moreover, for Mumford the concept of a dispositional power is primary, since it allows us to analyze concepts such as causation, laws of nature, modality, and properties. In order to support his position Mumford develops a “transcendental argument” emphasizing the fruitfulness and explanatory power of an ontology that contains the assumption of the existence of unverifiable dispositions.

Introduction

XV

In her essay “Dispositional Pluralism,” JENNIFER MCKITRICK argues explicitly against the philosophical tendency of making all-or-nothing claims about dispositions such as that all properties are dispositions, or that all properties are non-dispositions, that all dispositions are intrinsic, and so on. For McKitrick, this philosophical tendency is overlooking the fact that there is a plurality of different disposition types. She argues for her position by suggesting that a semantic analysis of our ordinary way of talking and of ascribing dispositions is more consistent with dispositional pluralism rather than dispositional absolutism. Moreover since we are also ordinarily justified in thinking that our ordinary ascriptions are true, we have reasons for claiming that there is a plurality of different types of dispositional properties; and not merely a plurality of different disposition concepts. ANDREA BORGHINI’S “Dispositions and Their Intentions” addresses the question of how exactly to analyze the nature of dispositions within the context of dispositionalism according to which dispositions are primitive denizens of reality with an irreducibly modal character. Among dispositionalists, Charlie Martin, Ullin Place, and George Molnar have argued that the modal character of dispositions should be understood in terms of their intentionality. Other dispositionalists, most notably Stephen Mumford, have challenged this understanding of the modal character of dispositions. Borghini defends a fresh version of the intentional understanding of dispositions. The core of the proposed view consists in treating a disposition as a primitive entity whose understanding depends on a metaphorical specification of its intention. In the final section, the role of dispositions in different areas of scientific and philosophical research are analyzed. As mentioned above, the contributions in this section are intended to broaden the current framework of the discussion by addressing the subject of disposition in more localized contexts. The first two articles address the topic of dispositions from the perspective of a natural and a human science. ANDREAS HÜTTEMANN defends a version of dispositional realism in his article “Dispositions in Physics” by arguing for the following three theses: First, in contrast to Armstrong, he argues that law-statements should be understood as attributing dispositional properties. In this context, dispositions are, however, not understood as causes of their manifestations. Rather Hüttemann conceives of them as contributors to the behavior of compound systems. It is in this sense, that he defends his third claim that within physics dispositional properties have to be regarded as irreducible properties that have no need for an additional categorical basis. ROBERT SCHNEPF, on the other hand, tackles the issue of disposition by looking more closely at the nature of historical explanations. In his essay “The Role of Dispositions in Historical Explanations,” he analyzes dispositional

XVI

Introduction

explanations such as the explanation of Caesar’s behavior during the Roman Civil War in terms of a so called “Clementia Caesaris” or Max Weber’s appeal to the “protestant spirit” in his account of the rise of modern capitalism. Schnepf focuses on the epistemological problem of ascribing dispositions to historical actors. He shows that especially Max Weber’s methodological reflections on this issue fit very well with an analysis of dispositions in terms of counterfactuals. For Schnepf, this implies that dispositions should be understood as theoretical terms. More substantial metaphysical assumptions of forces, faculties, or capacities are therefore of no use in a historian’s explanatory work. KARSTEN STUEBER addresses the question of disposition within the context of philosophy of mind; the conceptual domain that Ryle first hoped to fully analyze with the help of the concept of disposition. In his contribution “Empathy, Mental Dispositions, and the Physicalist Challenge,” he is particularly interested in investigating the ontological status of higher order dispositions. Stueber argues for the special status of mental dispositions such as beliefs and desires because of their doubly dispositional character. Folk psychological predicates ascribe dispositional properties to other agents. Yet, as Stueber shows, in contrast to ascriptions of properties and dispositions in the physical sciences, the ascription of mental dispositions is epistemically special, because it depends essentially on the use of the first person perspective and our empathic ability to put ourselves in the shoes of another. It is exactly for this reason that Stueber regards our folk psychological practices as constituting an ontologically relatively autonomous and epistemically special explanatory domain. Stueber also shows that his position is fully compatible with the assumption of ontological physicalism. GREGOR DAMSCHEN’S contribution “Dispositional Knowledge-How versus Propositional Knowledge-That” relates to issues in the philosophy of mind and epistemology; that is questions about the structure of the mind and the nature of knowledge. In particular, Damschen deals with the question of the structure of knowledge and the precise relationship between propositional “knowledge-that” and dispositional “knowledge-how.” In the first part of his essay, he provides an analysis of the term ‘knowing how’ and argues that the usual alternatives in the recent epistemological debate – knowing how is either a form of propositional or dispositional knowledge – are misleading. In fact it depends on the semantic and pragmatic context of the usage of this term whether ‘knowing how’ refers to a type of dispositional knowledge, to propositional knowledge, or to a hybrid form of both. Only in the first case, can one say that dispositional know how cannot be reduced to any form of propositional knowledge. Yet for Damschen, this case is the most interesting one to consider in the investigation of the nature of knowledge, if one assumes that knowing that p presupposes “having found out that p.” This

Introduction

XVII

assumption, as Damschen argues, seems to be implied in an internalist conception of knowledge. Having found something out, however, presupposes certain acts of epistemic inquiry and corresponding epistemic abilities. Accordingly, dispositional knowledge has to be understood as being at the very core of our notion of knowledge, including propositional knowledge. The last two articles in this anthology presuppose the ontological relaxed attitude towards dispositions manifested in the prior articles. They do not fundamentally question the reality of dispositions but take them for granted. Their purpose is rather to discuss and elaborate on the use of the concept of disposition in recent epistemology, particularly virtue epistemology. The central concern of DAVID HENDERSON and TERENCE HORGAN in their paper “Epistemic Virtue and Cognitive Dispositions” is not to explicate the metaphysical status of cognitive dispositions. Rather, they are interested in making a point about the range of dispositions that are epistemically important. From their point of view, epistemology in the modern period has understood only a narrowly restricted range of cognitive dispositions as epistemically relevant; what they refer to as classically inferential processes (or dispositions to classical inference). But for Henderson and Horgan, it is important to recognize that the useful epistemic chores are not all implemented by classical inference, but by dispositions keyed to richer sets of information. In the very last essay of this anthology “The Epistemic Function of Virtuous Dispositions,” ELKE BRENDEL takes a critical look at virtue epistemology. While she acknowledges that thinking of intellectual virtues as dispositions provides important epistemological insights, Brendel is rather skeptical about the attempt to define knowledge in terms of virtuous dispositions. As she argues, this could be a feasible and promising epistemological project if and only if some of the central concepts and ideas of virtue epistemology are revised or at least refined. The major problem for defining knowledge in terms of intellectual virtues and dispositions consists in the fact that many intellectual virtues are not strictly truth-conducive.

I. The Discovery of Dispositions: Ancient Foundations

Knowledge and Virtue as Dispositions in Plato’s Theaetetus FRANCISCO J. GONZALEZ If we turn to Plato with the modern question of whether or not ‘dispositions’ are to be numbered among the things that are, we find an answer that is not only clear but startling. In the work by Plato most explicitly devoted to questions of ontology, the work in which is to be found the most explicit reflection on the meaning of being, a definition of being is suggested, which not only includes dispositions, capabilities and powers among what is, but even identifies being with having a certain disposition, capability or power. Thus in the Sophist the Stranger suggests that “beings are nothing other than dunamis”( µ , 247e4), where dunamis means here not mere logical possibility, but a real disposition to act or be acted upon. The definition of course does not suggest that beings cease to be when their dunamis is actualized, i.e., when they are actually acting upon, or being acted upon by, something else. The point instead is that because beings do not need to be actually acting or reacting at any point in time in order to be, what it means for them to be is nothing more than having the power to act or react. As opposed to an ontology that begins with the assumption that to be is to be actual and then does not know what to do with dispositions, the suggestion in the Sophist is that to be is to be a power, with the result that the actual exercise of this power is ontologically derivative and not what defines a thing’s being and makes it be. But what recommends the latter view? The Stranger first introduces the definition as an improvement to an ontology (that of the so-called ‘Giants’) that identifies what is with what is tangible and material. What the Stranger cites as ‘immaterial’ and thus as a challenge to the Giants’ ontology, are none other than certain dispositions of the soul, including virtue and knowledge. What therefore recommends the above characterization of beings is that it can include among what is both material objects and immaterial dispositions because both can be said to be by virtue of having a certain power to act or be acted upon. But the Stranger also uses this characterization of beings to challenge the ontology of the ‘Gods’ or ‘Friends of the Forms’ which identifies what is with what is eternal and unchanging. The problem with this ontology is that it excludes all change and motion, including our coming-to-know the Forms and their coming-to-be-known. In being made to accept the above characterization of beings the ‘Friends of the Forms’ avoid an ontology that

4

Francisco J. Gonzalez

excludes its own possibility as ontology while also avoiding a reduction of being to motion: our being is not our knowing but our having the power to know; the being of the Forms is not their being-known but their having the power to be known. In short, an ontology premised on the characterization of beings as dunamis can include among beings the immaterial as well as the material, motion and change as well as permanence and stability. What is worth emphasizing for the purposes of the present paper, however, is that what necessitates the definition of beings as dunamis in both cases, i.e., in the critique of the ‘Giants’ as well as in the critique of the ‘Gods’, is the introduction of the dispositions of virtue and knowledge. These dispositions appear to serve here as a criterion for what counts as an adequate ontology. Consequently, a better understanding of these dispositions promises to shed some light on this ontology and to clarify what exactly is meant by the identification of beings with dunamis. What exactly does it mean to characterize knowledge or virtue as a dunamis? And how do we determine what kind of a dunamis we are speaking of here? For answers to these questions we must turn to the dialogue that, in terms of dramatic chronology, immediately precedes the Sophist: the Theaetetus. What is at issue in this dialogue is precisely the question, “What is knowledge?”, a question that proves inseparable from the question of what constitutes a virtuous disposition.

1. Virtue, Irrational Powers, and Socratic Midwifery There is a strong temptation to begin an analysis of the account of knowledge in the Theaetetus with the first definition of knowledge as perception and thus skip all the preliminaries of the investigation. This would, however, be a serious mistake since the prologue provides crucial clues and guidelines as to the nature of the inquiry to follow. In the present context, three points in particular deserve to be highlighted. First, the question, “What is episteme?” arises in the context of Socrates’ wish to examine Theaetetus’ soul with regard to virtue ) and wisdom ( , 145b1-2). The connection is made by Socrates’ ( µ and are the same thing (145e6). Therefore in suggestion that what follows we must never lose sight of the fact that the ‘knowledge’ under discussion is understood as the arguably more practical disposition of beingwise and therefore in close connection, if not identity, with that disposition of the soul called ‘virtue.’ The second point worth highlighting is the example the dialogue provides of the kind of definition being sought. After Theaetetus responds to Socrates’ question with a list of examples of knowledge, a list which significantly includes not only geometry but also the technical skills of the craftsmen such as shoemaking, Socrates explains what he wants instead by giving the provoca-

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

5

tively and presumably ironically inappropriate example of defining mud as earth mixed with water (147c4-6). When Socrates characterizes this as being a simple and commonplace definition of a simple and commonplace thing, we must wonder if knowledge is also a simple and commonplace thing that admits of such a simple and commonplace definition. Fortunately, however, Theaetetus is inspired to provide another and very different example: that of defining incommensurable lines. These are characterized as lines which when squared produce an oblong number, i.e., a number that cannot be produced by the multiplication of equal integers. For example, a line with the square root of 3, or what the Greeks would call the power of three, becomes when squared a number, i.e., 3, that is ‘oblong’ in the sense that it is not the product of any whole number multiplied by itself. The same would be true of the square root of 5, the square root of 7, etc. Theaetetus describes how he and the young Socrates classified all these incommensurable lines as powers ( µ ). The justification is that while these lines are incommensurable ( µ µ µµ! ) with the lines which when squared produce an even number, they have the power of producing plane figures that are themselves ! $ % , 148b1-2). But what does this commensurable ( " ’ example suggest regarding the attempt to define knowledge? Because an incommensurable line cannot be defined in itself, i.e., its length cannot be stated, it can be defined, and made commensurable, only by being characterized as a power to produce plane figures. If this case is truly parallel with the case of knowledge, then can we not infer that if knowledge turns out to be indefinable in itself, i.e., incommensurable with any of the things in terms of which Socrates and Theaetetus will try to define it, it too should be understood as a power and thus in terms of what it is able to do? In this case, we would need to move away from thinking of knowledge as some product or outcome we can possess and define, thinking of it instead as something we do. If in the Sophist the example of knowledge requires a definition of beings as dunamis, we here begin to see how this definition can in turn clarify the problem the Theaetetus confronts in the attempt to define knowledge. The third point that deserves highlighting is Socrates’ comparison of his method to the art of midwifery: a comparison that suggests two things relevant for our purpose. The first is that Socrates himself cannot give birth to knowledge; he is “barren of wisdom” ( & ' (µ , 150c4); he cannot himself reveal anything from himself on account of not possessing anything ) µ * + , 150c6). Therefore, Socrates cannot teach his wise ( interlocutors anything but can only deliver them of the beautiful things they have within themselves (150d6-e1, see also 157c-d). But the second important point made by Socrates’ analogy is that even those pregnant with knowledge cannot give birth to it without him. Those who give themselves all the credit for their offspring and leave Socrates’ company in the belief that he is no good

6

Francisco J. Gonzalez

(150e1-2) become ignorant fools ( µ , " , 151a1). The clear suggestion, then, is that knowledge can be brought to light, and even properly reared (see 150e5), neither by Socrates on his own nor by the interlocutor on his own, , 150d2, d4, 151a3-4). This is why the but only in their association ( other skill Socrates claims for himself, one whose connection to midwifery is sufficiently unclear that Socrates must argue for it at length (149d5-150a7), is that of matchmaking, of being able to determine with whom a person should associate (151a2-b6). It would be wrong, however, to conclude from the analogy that Socrates can only help deliver knowledge and therefore plays no role in its procreation. While this indeed would be the implication of a strict analogy with the midwife, Socrates himself emphasizes a disanalogy here: what Socrates delivers his interlocutors of can be either false phantoms or something genuine and true, where the difference between the two is very hard to discern (150a9-b2). Therefore, what Socrates claims to be the most important thing in his art (µ!& , 150b9) is something that plays no role whatsoever in the midwife’s art: i.e., the ability to determine through every kind of test whether the offspring of his interlocutor’s mind is true or false. In this case, Socrates is not delivering knowledge, but only beliefs that can be determined to be true only in the testing and examination undertaken in the discussion. In other words, knowledge does not precede the association between Socrates and the interlocutor but is obtained, if at all, only in and through this association. While helping an interlocutor to give birth to an idea can involve a laborious and lengthy elaboration of this idea,1 as we see in Socrates’ lengthy explanation of the thesis that knowledge is perception in terms of Protagorean relativism and Heraclitean flux, this carefully delivered child of the interlocutor’s mind can still prove a mere phantom. This is why Socrates sees the testing of the interlocutor’s newborn idea for its truth or falsity as a more important part of his art of midwifery. Here we immediately confront the perplexing circularity of the discussion in the Theaetetus. The discussion puts into practice Socrates’ self-proclaimed ability to distinguish Theaetetus’ true beliefs about knowledge from his false beliefs about knowledge by means of examination. But is not this ability to “discriminate the true from the false” ( ) ) ,! - µ , 150b3-4) itself a kind of knowing? Does not this ability which Socrates identifies as the most important part of his art hold the answer to the question “What is knowledge?” If answering this question requires, in the words of the second century anonymous commentary on the dialogue, identifying the “criterion” ( ), i.e., “that through which we judge as through an organ” ( ) ’ . _____________ 1

This is a point on which my commentator for an earlier version of this paper presented to the Arizona Colloquium in Ancient Philosophy, George Rudebusch, rightly insisted.

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

7

µ / & , 2, 24-26),2 then is not such a criterion already invoked and assumed by Socrates’ art of midwifery? Is not Socrates here doing what he ' ” is searching for? Does perhaps his claim at 150d1 to be “ mean not that he is not at all wise, according to the usual translation, but rather that he is not entirely wise, according to the interpretation defended long ago by the anonymous commentator; 3 a claim that would allow that he is in a certain sense wise despite his inability to give birth to wisdom? And is not the wisdom Socrates has, as knowledge of how to discriminate true from false propositions, irreducible, under threat of an infinite regress, to the wisdom he does not have, i.e., knowledge of propositions or knowledge-that? Do we not find in Socrates’ practice a fundamentally and irreducibly dispositional knowledge? These are the questions that will need to be kept in mind as we proceed.

2. Perceiving Knowledge as Perception In the case of the first definition to which Theaetetus gives birth, it is not hard to see the contradiction between how knowledge is being perceived, i.e., as perception, and what Socrates professes to be doing. Note first the reflexivity in how this first definition is introduced, a reflexivity that will be seen to characterize the other definitions as well. Theaetetus suggests defining knowledge as perception because that is “how it now appears” ( & 0 , 151e2). Since Socrates soon afterwards identifies “appears” ( ) with “is , , 152b11), Theaetetus is in effect saying that he now perceived” ( ( , perceives knowledge to be perception: the way in which he attempts to know _____________ 2

3

The word is used by Socrates in the Theaetetus when he asks if man has within himself the of future events as he is said by Protagoras to have within himself the of what is white, heavy or light (178b6, c1). Here Socrates is clearly identifying having the within oneself and being the measure (µ! ). See Diehls and Schubart 1905, 53, 40-41; also 55, 42-45. The commentator defends this claim by seeing in the midwife analogy the implication that Socrates, like the ordinary midwife, at one time was pregnant and gave birth (54, 2-9) and insisting that even now Socrates is capable of giving birth, being & only in relation to those whose innate opinions he tries to deliver (54, 913; see also 55, 19-33). But if Socrates has some sort of wisdom, it is to be identified not with some knowledge he is hiding within himself while conversing with the young, but rather with that very art of midwifery that enables him to converse with the young. Thus Sedley, while defending the anonymous commentator’s interpretation of 150d1, rightly sees the wisdom Socrates possesses as “consisting in the insights that enable him to practice midwifery itself” (2004, 3137). However, Sedley does make Plato pregnant. The central thesis of his book is that the Theaetetus throughout alludes to Platonic doctrines for which Socrates was the midwife but avoids making the character Socrates explicitly defend these doctrines because Socrates did not himself give birth to them. The conclusions of this paper will show, I believe, that the conceptions of both knowledge and midwifery suggested by the Theaetetus cannot be reconciled with such an interpretation.

8

Francisco J. Gonzalez

what knowledge is already assumes the conception of knowledge at which he arrives.4 How, indeed, can this circularity be avoided when we are speaking of a knowledge of knowledge? But then it is important to note that Socrates’ way of attempting a knowledge of knowledge already in itself contradicts the manner and conception of knowing introduced by Theaetetus. As already noted, Socrates’ goal as a midwife is not only to help his interlocutors give birth to their offspring, but also to test their offspring to see if they are false or true. This is what he claims to be doing now with Theaetetus (160e-161a). Yet what this testing shows is that the definition of knowledge as perception undermines the very maieutic method by which Socrates seeks to deliver and test it by implying that all of our offspring are true and none are false. If to perceive something is already to know it, then whatever we perceive must be true; if knowledge is always true, knowledge cannot be identified with perception unless perception is always true. This is how Socrates derives from the first definition of knowledge what he calls ‘the Truth of Protagoras’: that whatever appears to me is truly as it appears. The incompatibility of this ‘truth’, and thus of the first definition, with Socrates’ maieutic method is made explicit in the following passage: I say nothing of my own case or of the ludicrous predicament to which my art of midwifery is brought, and, for that matter, this whole business of philosophical conversation ( %µ 1 2 !& , &µ , 161e6-7), for to set about overhauling and testing one another’s notions and opinions when those of each and every one are right, is a tedious and monstrous display of folly, if the Truth of Protagoras is really truthful and not amusing herself with oracles delivered from the unapproachable shrine of this book (161e-162a).

And it is of course Socrates who sets about determining whether the Truth of Protagoras is really truthful.

3. Dialectic versus Expertise: Taking Measure of Theodorus But what the discussion of the first definition of knowledge shows to be opposed to Socratic examination is not only Protagorean relativism but also, and perhaps surprisingly, the claim to expertise embodied by the mathematician _____________ 4

The anonymous commentator already insisted long ago that 3 , in Theaetetus’ first definition of knowledge is to be understood not as the sense organs ( ( , ) but rather as any kind of apprehension ( 4 ) (59, 46-49). For why, he rightly asks, would Theaetetus, after offering geometry and other sciences as cases of knowledge, retreat to perception? (59, 49 - 60, 8). The anonymous commentator’s distinction is clarified later when, in commenting on the identification of “appears” and “is perceived” at 152b-c and claiming that this is Plato’s addition to what Protagoras said, he writes that “they called every apprehension ( 4 ), whether through the sense organ ( ( , ) or through something else, perception ( 3 , )” (66, 39-43). See also Polansky 1992, 67-8 and 103-4.

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

9

Theodorus, who repeatedly refuses to participate in the discussion with the excuse that he is not dispositionally inclined to Socrates’ kind of ‘dialectic’ % ! , 146b3; see also 162a-b, 165a). It is no accident ( , 5 that Socrates makes a point of characterizing Protagoras and Theodorus as friends, a characterization Theodorus accepts (161b8, 162a4, 168c3, 168e7). The source of their kinship, I suggest, is that both Protagoras and Theodorus must ultimately make knowledge the result of some kind of intuition, since even Theodorus can appeal to nothing but intuition to secure the ultimate and unquestioned starting points of his mathematical deductions.5 This is why it should be no surprise that Theodorus’ student, Theaetetus, first seeks the answer to the question, “What is knowledge?” in perception, both as the content of the answer and as the way in which he arrives at the answer. At one point Theodorus revealingly characterizes the dialectical discussions he is disinclined to join as being, in contrast to geometry, “abstract” or “bare” (4 - +& , 165a2): what can this mean except that Socratic arguments do not appeal to intuition in the way geometry does? While both Socrates and Theodorus demand argument and proof (162e-163a), Theodorus apparently finds the ‘measure’ for his proofs in a ‘perception’ independent of any dialectical examination. When Socrates finally succeeds in dragging Theodorus into the discussion, it is with the purpose of determining precisely whether Theodorus is a measure in geometrical demonstrations or everyone is as competent as he is (169a). But are these the only possible alternatives? That they are not is suggested by the way in which Socrates brings into question as much Theodorus’ expertise as Protagoras’ relativism. This occurs already at the very beginning of the dialogue. There Socrates describes Theodorus as an expert in the liberal arts (6 * , 145a8), claiming that his judgment regarding Theaetetus’ virtue and knowledge must therefore be taken seriously. Yet in direct contradiction to his own example of simply believing what a musician says about harmony ( ,+µ ,’ , 144e5), Socrates recommends that he and Theaetetus take Theodorus’ judgment seriously not by believing it, but by eagerly examining ( , µ " , !4 , , 145b2-3) and being examined ( % , 145b4). Thus begins the battle between a conception of knowledge as authoritative expertise and a conception of knowledge as inescapably dialectical.6 _____________ 5 6

See Polansky 1992, 72; also 111, 117-8 Burnyeat observes: “But we should not fail to think about the dramatic emphasis which Plato has contrived to place on the notion of expertise” (1990, 3). But what Burnyeat does not note is the dramatic emphasis placed on the tension between expertise, as represented by Theodorus, and Socrates’ peculiar art of dialectic, or between knowledge and the knowledge of knowledge. Burnyeat’s commentary assumes throughout that knowledge is identified in the dialogue with expertise and certainty (see e. g. 15, 19, 44 [where “the quest for certainty” is attributed to both ancients and moderns], 47; at one point he even suggests that “the idea of expertise about man”

10

Francisco J. Gonzalez

Accordingly, at 169a Theodorus can prove himself to be the measure only by submitting to the common argument with Socrates and thus ceasing to be the measure! The paradox is that dialectical discussion is the only means by which Theodorus can justify his being a measure against those who would deny him this. In general, experts who claim to be the measure of what counts for knowledge can justify their claim only by submitting to a discussion of what knowledge is, but in so doing they cease to be the measure they claim to be. Indeed, how can anyone be a measure in the discussion of the nature of knowledge, when what is sought is precisely such a measure? Do we not again find ourselves in a vicious circle? How can we seek to know the measure of knowledge without already invoking such a measure? A solution to this problem could be found in Socratic discussion itself if, while recognizing no one as the measure, it could itself provide some kind of measure. This is precisely what is suggested by a crucial passage at 179a-b. Here Socrates asserts that it is necessary for Protagoras to agree that one person is wiser than another and that the wise person is the measure (µ! ). But who is Socrates to assert that this is necessary? Socrates himself immediately complains that his earlier defense of Protagoras forced him into the 7 position of being a measure even though he lacks knowledge ( µ µ , 179b2-3). But if Socrates himself is not a measure, both because Protagoras is wrong about everyone being a measure and because Socrates is not wise, then by what measure can Socrates judge it necessary that only the wise man is the measure? If only the wise man is the measure, then presumably only the wise man could be the measure of the truth that only the wise man is the measure. Yet the passage at 169a begins by invoking a different kind of measure. Levitt’s translation reads: “Then we shall be giving your master fair measure [µ 8 ] if we tell him that he has now got to admit that one man is wiser than another . . .” How can Socrates give fair measure in asserting that only someone with the wisdom he lacks could be a measure? A crucial word in the cited sentence is the ‘we’ (µ 8 1µ" , 179a10); what provides the fair measure for what must be admitted as true is not Socrates, but rather his discussion with Theodorus. The measure invoked by Socrates is neither every individual nor the expert: Socrates himself lacks expertise and he also, despite what he suggests to the contrary, does not allow Theodorus the expert to be the measure: instead of appealing to Theodorus’ expert opinion, he _____________ was Plato’s “most persistent dream” [186]). The only textual evidence for this appears to be 152c5-6 and 160d1-3 where Socrates says that perception is both always the perception of being and 4 , which Levitt translates as “unerring.” But is it correct to identify being 4 with being certain? And even if knowledge is initially understood as expertise and certainty, the arguments against the identification of knowledge with perception are that it undermines not expertise but language and any relation to being and truth. How, furthermore, can one be certain that knowledge is certainty? To the extent that a knowledge of knowledge is possible at all, it is not certainty, but rather what is exhibited in the dialogue: dialectic.

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

11

forces him, very much against his inclination, to join the discussion. Thus Socrates at one point describes his midwife’s art as a knowledge of “nothing more than” how to receive “in measure” (µ 8 , 161b2-5) the logoi provided by his interlocutor. While Protagoras might claim that everyone is the measure and Theodorus might claim that he is the measure, it is only Socrates’ dialectic that can provide the measure for these claims. That is precisely the power that defines it.7

4. Striving after Being After having reduced to absurdity the definition of knowledge as perception by associating it with Protagorean relativism and Heraclitean flux, Socrates shows more positively and directly why perception cannot be knowledge and in a way that leads to the next definition: he argues that knowledge is impossible without an apprehension of being and truth and that the soul apprehends being and truth through itself and not through the senses (186c-e). This does not mean that the soul apprehends being/truth ‘automatically’ and directly: Socrates stresses the laborious process of education involved (186c35). Indeed, what we have here is a characterization of knowing that corresponds much more closely to that implied by Socrates’ midwifery: knowledge of what is and what is true is not passively received from without but rather is engendered from within the soul itself through its own laborious activity. Furthermore, the apprehension of being and the other ‘commons’ ( , 185e) is at one point characterized as a striving ( !& , , 186a4). This ‘striving after being,’ as the characteristic of knowing that proves incompatible with the definition of knowledge as perception, appears to be precisely what characterizes Socrates’ dialectic. And Socrates’ argument here implies that any account of knowledge that does not take into account this disposition cannot succeed. As striving, this disposition is clearly more than a mere ability, but rather a positive tendency towards being and truth. If knowledge is a power, it is therefore one in a stronger sense than a mere technical ability or know-how.

_____________ 7

This idea of dialogue itself, rather than expertise, being the measure for what is true is not unique to this passage of the Theaetetus, but is articulated even more clearly in the Republic (348a-b) and the Protagoras (338a-b).

12

Francisco J. Gonzalez

5. Believing Falsely that Knowledge is True Belief Theaetetus now suggests that knowledge “might be” ( % ) “true belief”, while granting that as they proceed it might appear differently (187b5-7). In this way the definition of knowledge as true belief is introduced as a belief that might be either true or false. But then it seems that knowledge should be sought not in true belief, but in that which is capable of distinguishing between true and false belief. This is indeed why Socrates proceeds to show that the identification of knowledge with true belief cannot account for the possibility of false belief, a possibility assumed and exhibited by Socrates’ own method. Thus again the definition is contradicted by the very method with which it is considered and tested. In this context, Socrates identifies thinking with a dialogue, a give-andtake of question and answer that the soul carries out with itself, and doxa with a silent logos that concludes and brings to an end this thinking (189e-190a; see also Sophist 263e). This characterization of thinking clearly picks up from that earlier characterization of the soul’s relation to being and truth as a striving: the soul does not ‘intuit’ being, but rather approaches it through a process of continually interrogating itself. But then is the doxa that concludes this thinking/striving knowledge? A negative answer is suggested not only by the eventual refutation of the definition of knowledge as doxa, but also by the words with which Socrates prefaces his characterization of thinking: “I am saying this as someone who does not know” (9 & µ ( : , 189e7). Socrates thus makes clear that the outcome of his soul’s conversation with itself is not knowledge.8 Perhaps the positive and most important contribution of this passage, however, is its implied distinction between thinking, on the one hand, and judging or believing, on the other.9 Doxa is explicitly characterized as what terminates the soul’s dialogue with itself, what concludes the dialectical backand-forth of question and answer (190a2-4). Socrates describes the soul as having a doxa, as believing or judging ( ; < ), only when it is no longer wavering or hesitating (µ ), only when ), either slowly or quickly rushing it arrives at something definite (? upon it ( @; ). Socrates thereby distinguishes doxa from the thinking that he and Theaetetus are presently carrying out both in their own souls and in conversation with each other. Socrates and Theaetetus at no point ‘judge’ or ‘believe’ if this means arriving at something definite and univocal and ceasing _____________ 8 9

Cf. Dixsaut 1997, 10. As Sedley notes, a distinction is being made here between thinking and judgment: “Thinking is interpreted as replicating within the soul the form of Socratic dialectic, with judgment identified, not with thought as a whole, but with its final stage or outcome” (2004, 130).

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

13

to waver: they never come to a judgment or belief (even their apparent judgments are only questions) because they never desist from that conversation that is thinking. But if knowledge cannot be identified with doxa of any kind, does not Socrates’ distinction between doxa and the thinking/conversation he himself practices suggest that perhaps knowledge is somehow instantiated in the latter? As we have been led to ask already, is not Socrates’ art of thinking, in the form of a conversation with himself and with Theaetetus, itself a kind of knowing?

6. Thinking of Knowledge without Knowing Knowledge The preceding has repeatedly characterized Socratic conversation as itself instantiating a certain kind of knowing, and thus being a certain kind of ‘measure,’ while at the same time acknowledging that Socrates repeatedly disclaims the possession of knowledge. What can this mean? How can Socrates’ method exemplify the knowledge it claims not to possess? Knowledge is either possessed or not possessed. This is precisely the difficulty Socrates runs up against in attempting to explain the possibility of falsehood. The reason for the failure of the attempts to explain false belief as the exchanging of something we know for something else we either know or do not know is, in short, that we can mis-take neither what we possess nor what we do not possess. If I genuinely possess two things, how could I think that the one is the other? If I possess only one of them, how could I mistake this thing which I have ‘in hand’ for something else I do not have at all? Though the aviary analogy is meant to address this problem, it leaves us with it: even if I can mistakenly grab one bird while seeking another, how can I mistake this bird for the other one once I have it in hand (199d)?10 Significantly, it is in the context of this recurring problem that Socrates objects (196d-197a) that the discussion is ‘tainted’ (µ , A !& , , 196e1-2) by the fact that he and Theaetetus have been constantly using different forms of the word ‘know’ and its opposite despite their claim not yet to know what knowledge is. Theaetetus appropriately responds that no discussion can avoid these words (196e8-9) and Socrates must acknowledge that his very objection uses them (196e5-7)!11 Yet we still must ask if the account of falsehood does not fail on account of the inadequacy of the presupposed conception of knowledge as the direct and complete possession of an object _____________ 10 11

For an account of the inadequacies of the wax tablet and aviary analogies and the argument that what they fail to capture are precisely the powers that define the soul, see Gonzalez 2007. This circularity is of course no problem for a ‘nominalist’ position that simply stipulates what is to count as knowledge, but Plato is not a nominalist and instead believes that knowledge has a true essence we must seek to discover.

14

Francisco J. Gonzalez

for oneself. Such a conception seems quite different from the one presupposed by Socrates’ own practice of dialectic: if we take the midwife analogy seriously, neither Socrates nor Theaetetus possess bits of knowledge inside their souls like birds in an aviary waiting to be seized.12 Even if Theaetetus is pregnant, he might be pregnant with mere wind-eggs ( µ " , 151e6, 157d3, 210b9) as opposed to fertile eggs capable of hatching into birds. Only in the dialogue between Socrates and Theaetetus do beliefs prove false and therefore only in dialogue, if anywhere, can knowledge be born and spread its wings. Knowledge is not a possession. If we look at what is enacted in the discussion itself for an explanation of the possibility of falsehood, we see that what makes possible Theaetetus’ own false beliefs about knowledge is not the exchange of one possessed bit of knowledge for another, but rather, to pick up on the language at 186a4, a stretching-out-towards the truth that makes it possible to mis-take it.13 Miles Burnyeat, correctly observing that “What is at stake in the discussion of false judgment is nothing less than the mind’s relation to its objects” (68), has found the suggestion here of what he calls “a third epistemic route,” distinct from both perception and knowledge, and which he characterizes, with reference to the example of arithmetic at 198c, as “unknowingly thinking of twelve, i.e., thinking of twelve without recognizing it as twelve” (112). What Burnyeat is drawing our attention to here is the crucial point that questioning is itself a way of encountering something, that the question, rather than being the absence of a relation between the mind and its objects, is itself such a relation.14 _____________ 12

13

14

Yet Polansky notes that the sciences, e.g., arithmetic, appear to be identified in the aviary analogy not with the birds, but with the ability to chase the birds (1992, 197). The ambiguity then is significant: is knowledge like the bird we seek to possess or like the search itself? In Socrates’ account of believing falsely illustrated by his example of a person approaching from a distance and being inadequately seen as someone else, the eagerness ( , µ ,A, 193c2) to recognize the person clearly plays a central role. As Burnyeat observes with regard to the example in the text, “in asking the question [“What is five and seven?”] he is unknowingly thinking of the number twelve” (1990, 114). Yet Burnyeat believes that only 198c is “the textual basis for my talk of a third epistemic route which completes the parallel with the Wax Block. But is it enough to build an interpretation on?” (113). What he thus fails to see is that the dialogue as a whole is the textual basis for this third epistemic route, since Socrates’ question, “What is knowledge?” is evidently the kind of relation to knowledge that enables Socrates and Theaetetus to mis-take what knowledge is. In asking this question they are already “going after” ( ’ B *+µ , , 187c1) what is in question. Furthermore, the earlier identification of thinking with both the soul’s “reaching out” or “striving” towards being and the soul’s questioning in conversation with itself supports this suggestion of a “third epistemic route.” Indeed, in the claim at 186a that the soul “strives after” being and the other “commons,” the verb “strives after” ( !& ) is substituting for the word “considers” ( " ) in the immediately preceding descriptions of the soul’s relation to the “commons” (185e2, e7). There is no reason, then, to agree with Burnyeat that Plato “does not yet recognize [the third epistemtic route] clearly and distinctly as a further way, besides knowledge and perception, to get hold of something with the soul” (115). Later in the text Burnyeat suggests identifying the third epistemic route with “a capacity for intermediate analysis of eleven. He [the dunce] has learned, per-

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

15

But this “unknowledgeable thinking”, though putting us in some relation to being and truth, is presumably not knowledge, even when it is true. As already noted, in judging that knowledge is true judgment, Socrates and Theaetetus do not claim to know that knowledge is true judgment, so that their very relation to the definition contradicts its content. But then what else is needed for knowledge? Need it be something that brings thinking, understood as the exchange of question and answer, to an end in an infallible belief? Or can there be a “knowledgeable thinking” that is not that possession of knowledge that appears to be the prerogative of the gods? The answer to these questions begins to emerge with Socrates’ direct demonstration that true judgment is not knowledge.

7. The Jury Example: Seeing versus Logos Socrates’ argument appeals to the fact that members of a jury can have true judgment regarding a crime that has been committed without having knowledge. Socrates’ explanation suggests that what the jury members lack, and what must therefore be added to true judgment to convert it into knowledge, is some kind of first-hand experience of the matter (( + µ+ ( ! , contrasted with ; 5 …/ , ,! , 201b8, 201c2), though what this means when we are talking about virtue or knowledge, and how this is not a return to a conception of knowledge as perception, is initially unclear. Furthermore, what Socrates says immediately before suggests that the reason why the jurors lack knowledge is that the clock in the courtroom makes impossible the kind of discussion that could alone teach them the truth of the crime, rather than merely persuade them (201b3). Burnyeat has drawn attention to the seeming contradiction between these two explanations of the jury’s lack of knowledge (124). But it is not hard to see that these two explanations are reconciled in Socrates’ practice and that this practice therefore itself exhibits what is required for knowledge. Socrates’ description of his art of midwifery shows that knowledge is to be attained neither through instruction nor through intuition, but rather through some third route between the two: Socrates’ interlocutors must indeed come to see the truth for themselves ’ C A D + + , 150d7-8), but can ( do so only by appealing to Socrates and the god for the proper ‘delivery’ and testing of their insights (150d8-e1). In short, philosophy, as Socrates under_____________ haps, that eleven is nine and two, and this partial grasp of eleven enables him to think of eleven and judge that it is five and seven. No breach of K’ occurs, because what he is using is a capacity for true judgment concerning eleven, not knowledge” (179). But is this adequate? The dunce must exist in some relation to eleven that goes beyond his analysis of it as nine and two if he is to mistake it for the sum of five and seven.

16

Francisco J. Gonzalez

stands and practices it, is neither being taught without ‘seeing’ nor ‘seeing’ without being taught: instead, leisurely dialogue is in philosophy a way, and the only way, of being an eyewitness to the truth.15 Yet if the conception of knowledge presupposed by Socrates’ practice resolves the tension in the jury example by reconciling logos with ‘seeing something for onself,’ the latter is immediately suppressed in the discussion not only by Theaetetus’ identification of knowledge with an account – an identification that retains only one side of Socrates’ example – but also by the very way in which this definition is introduced: Theaetetus admits that it is based on what he has heard from others ( ( + % , 201c8); as for explaining the definition, he claims to be able at best to follow what someone else says ( !& D ! , 201d6-7). Socrates then proceeds himself to give an account of the definition based on a dream he has had; and in claiming to be thus offering Theaetetus a ‘dream in exchange for a dream’ ( - / , 201d8), he significantly identifies Theaetetus’ second-hand knowledge itself with a dream. On one level this is all in keeping with the reflexive character of the dialogue: just as knowledge was ‘perceived’ to be perception and was merely ‘believed’ to be true belief, now it is believed on the basis of a logos to be true belief with a logos. But then what Socrates’ comments suggest is that logos, as something that can be heard from others, is so far from providing the knowledge which the jury example showed to involve some kind of ‘eyewitnessing’ that it can be likened to a mere dream. In defining knowledge as true belief with logos on the basis of a logos they have heard, Socrates and Theaetetus are dreaming of knowledge rather than knowing knowledge. In this case, is not the awakening and first-hand experience required for knowledge precisely the testing and examination to which Socrates proceeds to subject this dream of knowledge? And is not the specific dream from which the ensuing dialogical examination must awaken us precisely the dream that there is some logos which by itself could convert our beliefs into knowledge? In this case, does not the absence of a logos of knowledge at the end of the dialogue represent, instead of a failure, an awakening to a genuine knowledge of knowledge?

8. Awakening from the Dream Though any interpretation of Socrates’ objections to the ‘dream’ must be controversial, I will here only suggest that the problem with the characteriza_____________ 15

In this way, the conception of knowledge assumed by Socrates’ midwifery is neither strictly ‘internalist’ nor strictly ‘externalist’ but instead refuses the either/or represented by these two positions.

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

17

tion of knowledge as an analysis of a complex whole into its elements is that neither each element itself nor even the whole itself, to the extent that it is more than the mere sum of its parts, can be known through such analysis. Both the endpoint and the starting point of analysis lie beyond analysis, i.e., & (202b6, 205e2-3). Analysis presupposes a relation to the object that are cannot itself be identified with analysis. What then is the nature of this relation? To identify it with some sort of immediate intuition would be to fall back into a conception of knowledge as perception. The relation at issue here seems indeed to be precisely the one to which the dialogue has repeatedly drawn our attention: i.e., the ‘striving’ or ‘going after’ enacted by the shared search of dialogue. Here we must again attend not only to the content of the definition but also to the way in which it is posited. In analyzing knowledge into ‘true belief’ and ‘account,’ how do Socrates and Theaetetus ‘know’ the elements ‘true belief’ and ‘account’ as well as the whole ‘knowledge’ to the extent that it is more than the mere ‘sum’ of these elements?

9. Failing to give an Account of Account But here we seem simply to return to that relation of the mind to its objects that Burnyeat calls “unknowledgeable thinking.” How does such thinking, both as the thinking of the definition and the thinking defined in the definition, become knowledge? Socrates proceeds to show that if logos converts judgment into knowledge, this cannot be logos understood as the mere expression of judgment, since in that case anyone capable of speech could transform a belief into knowledge by merely expressing it; we would thus be back to Protagorean relativism (206d-e). Nor can logos be understood as the enumeration of a thing’s elements provide knowledge, since such enumeration is perfectly compatible with the lack of knowledge when, for example, I can correctly state a things’ elements without recognizing those same elements in something else (207a-208b). Socrates’ use of the wagon example to explain this second understanding of logos is significant for what it suggests about what knowledge would re, 207c3) by simply quire: if one does not know ‘the being of a wagon’ ( going through its hundred timbers, as one clearly does not, how then does one know what it is? The obvious answer is: by using it (see Polansky, p. 227). As Socrates himself suggests in other dialogues, it is the user of something who has knowledge of what it is (Cratylus 390b-d, Euthydemus 290b-d, Republic 602a1-2). The knowledge of how to use the wagon does not require knowledge of every piece of wood in the wagon while, on the other hand, the ability to enumerate the one hundred timbers does not imply knowledge of how to use the wagon. In terms of Socrates’ other example, the person who can cor-

18

Francisco J. Gonzalez

rectly write out all of the syllables in Theaetetus’ name nevertheless does not, strictly speaking, know how to spell even “Theaetetus” when he is unable to spell other names containing the same syllables. In other words, knowing how to spell the name of Theaetetus is not the same as have a correct judgment regarding the elements of this name. If we ask, therefore, what sense of knowledge is missed by the mere enumeration of elements, the answer appears to be a certain kind of ‘know-how.’16 Socrates’ midwifery, I would suggest, with its capability of testing the truth or falsity of those beliefs with which others present it, is precisely such a know-how. It is significant in this regard that one of the examples of knowing-how-to-use in the cited passage from the Cratylus is knowing how to use words in the given-and-take of question and answer, i.e., dialectic. The inadequacy of the conception of logos as the enumeration of elements can be seen as pointing to a conception of logos as that know-how called dialegesthai. Yet there is one more meaning of logos which Socrates rejects as inadequate for knowledge: logos as the statement of a difference that distinguishes the object of judgment from other things (208c7-8). Socrates’ objection is that simply thinking of one thing rather than another requires having in mind what distinguishes that thing from the other (209d1-2). A statement of the distinguishing difference therefore cannot transform belief into knowledge because it adds nothing to what is already in belief itself, unless one specifies that what is added is knowledge of the distinguishing difference. But in this case we explicitly fall into the circularity that has been seen to ‘taint’ the entire discussion: I must already be in a relation of knowing towards whatever I propose to add to true belief to make it knowledge.17 The circularity becomes especially head-spinning when we consider that the account of knowledge as “true belief with an account” can not only be interpreted as the mere verbal expression of a belief about knowledge or as an enumeration of the elements of knowledge, but can also be interpreted as the statement of the difference that distinguishes knowledge from belief (where the difference is ‘account’or logos). _____________ 16

17

Burnyeat goes too far, however, in seeing (1990, 212-213) here an identification of knowledge with some practically unattainable expertise and mastery of a whole domain. Buryneat tries to make such an identification more palatable by suggesting that Plato’s claims about knowledge can be reexpressed as claims about understanding (216-218). I agree, but only with the qualification that “understanding” here means a know-how or capability, rather than the systematic and exhaustive explanation of an entire domain. As Sedley rightly observes: “It is not hard to work out that this same objection is equally threatening to all definitions of knowledge as true judgment plus something. Whatever that something may be – whether justification, analysis, warrant, differentiation, or, as in the Meno, ‘calculation of the cause’ – the same problem will threaten to arise. It is not enough to have mere true judgment about the extra something, or simply to assert it, or for it merely to exist. The knowing subject can stand to it in no cognitive relation weaker than that of knowing it. And as soon as this is recognized, circularity sets in. Knowledge will be defined as true judgment plus knowledge of something” (2004, 176).

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

19

The implication of Socrates’ concluding argument is this: the judgment that knowledge is true belief with an account is mere belief unless we know that having an account is the distinguishing trait of knowledge. Yet by what account can we know this? And if we attempt to demonstrate by giving an account that giving an account is the distinctive trait of knowledge, are we not begging the interpretation of knowledge we are trying to demonstrate? Here again the circle must draw our attention to what is happening in the dialogue. The final account of knowledge as true judgment with an account is delivered and assessed by that peculiar know-how that Socrates characterizes as his midwifery. Now we can ask: does not this know-how provide a sense of logos different from the ones explicitly considered but also manifest in their consideration, one indicated and enacted by the dialogue, one that we can call ‘dialectic’ and that Socrates himself characterizes as having the ability to distinguish true from false beliefs? If no one logos can justify a belief, then perhaps the only justification possible is that of which Socrates is capable: what one could call dialectical justification, i.e., the kind of justification produced by the give-and-take of question and answer. This dialectical ability would then be what makes thinking knowledgeable without making it the final possession of knowledge. Socrates hints at this at 202c2-3 where he describes the person who has true judgement without a logos as unable to give and receive a logos (µ µ 2 - !; , +& ). But then Socrates immediately proceeds to describe the contrasting person, not as some who can give and receive a logos, but rather as someone who gets hold of a logos and thereby becomes ‘perfect’ with regard to knowledge ( 8 ) µ * , 202c45). What Socrates is describing here is his ‘dream’ (202c5) and there is a sense in which the rest of the discussion pursues this dream of possessing a logos that would make one perfect with regard to knowledge. There are indeed passages later in the text, as Burnyeat points out (178), where having a logos is identified with the ability to give a logos, defined in either of the three ways considered (206d-e, 207b-c, 208b, 208c). Yet what is significant is that all of these passages strictly separate giving a logos from receiving a logos, thus making logos something the person with knowledge possesses and can exhibit on demand rather than something existing in an exchange between two people. The ability to utter a statement, the ability to list a thing’s elements, the ability to offer a distinguishing mark: can any of these interpretations capture that peculiar way of dealing with and using logoi which Plato calls the % µ of 2 - !; , +& ? Do we not have in this phrase a meaning of logos that cannot be captured in a logos, but can only be exhibited in deed? In showing that knowledge cannot be the expression of a true belief nor the enumeration of elements nor the statement of a distinguishing difference, and in therefore also showing that we cannot know knowledge by expressing

20

Francisco J. Gonzalez

a belief about it or enumerating its elements or stating what distinguishes it from belief, 18 Socrates forces us to look for what knowledge is in what he and Theaeetetus are doing in asking the question “What is knowledge?” Are they merely expressing beliefs, or enumerating elements, or stating a thing’s distinguishing difference? All of this can of course be found in what they do but all of this is subordinated to a dialectical know-how that cannot be reduced to any of these senses of giving a logos. This know-how is clearly not the possession of a logos that will make one perfect in knowledge, but neither is it mere belief or ignorance. The dialogue between Socrates and Theaetetus is knowledgeable about knowledge without possessing a final logos of knowledge, a logos that would itself somehow need to be known through a logos. To seek to know by means of a logos that knowledge is true belief with a logos is to be caught unawares in a vicious circle.19 To suggest that knowledge can arise only in the give and take of logoi, only in dialogue, i.e., that it can never be captured in a logos, is to recognize the circle without becoming its victim.

10. The Dialogue’s Outcome: Knowledge of Knowledge as a Way of Being with Others Only these reflections can prepare us for Socrates’ description of the dialogue’s positive outcome. After suggesting that Theaetetus’s present barrenness might be only temporary, a suggestion that cannot inspire much confidence given Socrates’ own permanent barrenness, Socrates reassures Theaetetus that if he never gives birth to another child he has still gained something positive from the discussion, i.e., a gentler disposition in being with others (E F G " 2 - 1µ : , 210c2-3) as well as a certain kind of virtue and knowledge: namely, exhibiting temperance in not thinking he knows what (+µ ( ! $ µ H , , 210c3-4). But in he does not know ( 8 + 8 _____________ 18

19

Burnyeat rightly maintains that no “fourth sense” that could solve the problems raised by the Theaetetus is to be found in other dialogues (1990, 236-7). “The very indeterminacy of ‘account’ makes us concentrate on the general form of a problem that nearly all epistemologies must face. Plato’s own thoughts about knowledge are no more immune to the difficulty than anyone else’s” (238). Burnyeat also argues that an appeal to Forms provides no solution (238). Polansky goes even further in his defense of what he calls “the completeness of the Theaetetus” (1992, 242). He too argues that Plato has no other account of account or knowledge than those offered in the Theaetetus (235-6). He also denies that an explicit appeal to the Forms would add anything to the discussion (16). But Polansky suggests that the dialogue exhausts not only Plato’s possibilities, but also our own. The conception of knowledge as justified true belief does not add anything to the conceptions Plato reviews under the heading of opinion with an account (211, 236-7). Even modern idealism and historicism fall within the range of possibilities considered in the dialogue (242-3). Cf. Dixsaut: “La différence entre l’opinion droite et le savoir ne peut constituer le contenu d’une opinion droite ou d’un savoir, mais le savoir est cette difference” (1998, 304)

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

21

this case a certain knowledge of knowledge, i.e., knowing what you know and do not know, has been achieved despite, or rather because of, the failure to give an account of knowledge. Furthermore, Socrates associates this knowledge of knowledge with a certain way of being with others. This outcome of a knowledge of knowledge exhibited in dialogue with others should not surprise us because it is precisely the outcome Socrates predicted at the start of the dialogue. First, recall that the declared purpose of the discussion is to examine Theaetetus’ virtue and knowledge (145b). Secondly, when Socrates’ question “What is knowledge?” meets with silence, Socrates reveals as his motives in asking the question his love of logos ( & ) and his desire to “make us converse and become friendly and talkative with one another” (1µI 5 !& , &+ & & , , 146a78).20 To claim that philosophical knowledge is a dialectical disposition, a power of giving and receiving logoi in the constant striving after the truth, is not to provide an answer to the question “What is knowledge?” but rather to insist that the question must always remain open. On the other hand, this is not to say that there can be no answer of any kind and that therefore philosophical conversation is condemned to being like that practiced, according to Socrates, by the Heracliteans who “never reach any conclusion with each other, they are so careful not to allow anything to be stable, either in an argument or in their own souls” (180a7-b1).21 A kind of ‘answer,’ though not one that closes the question, is to be found in the performance of philosophical conversation itself.22 In the Theaetetus, Socrates shows both that a knowledge of knowledge _____________ 20

21

22

An interesting parallel is to be found in Xenophon’s Memorabilia IV. v. 12. Socrates here is % and & , described as arguing, at the end of a discussion with Euthydemus on 8 that by sorting things according to kinds ( !& &! ), choosing the good and rejecting the bad, one becomes better and happier and more able to pursue dialectic ( ). Even if Kahn is correct in arguing that this µ … !& , passage in Xenophon is derived from Plato and thus offers no independent testimony on the historical Socrates (1996, 76-79), it at least shows that even Xenophon, with or without Plato’s help, saw becoming “dialectical,” not least in the sense of an ability to “collect and divide,” both as essential to being good and as the goal of a Socratic discussion. Thus Xenophon’s stated goal in proceeding to report another discussion is to show how Socrates “ 8 ! .” G + George Rudebusch, my commentator for an earlier version of this paper, suggested that my interpretation characterized Socratic conversation in the way that Socrates characterizes the conversation of the Heracliteans. The above comments make clear, I hope, that this is not at all the case. In this way my interpretation differs from the one Burnyeat characterizes as follows: “Plato has no answer to the question ‘What is knowledge?’ because he is actually arguing that no answer is possible: he does not believe that knowledge exists” (1990, 235). Polansky’s interpretation is similar to my own when he asserts that “The entirety of the dialogue depicts just what knowledge is. … philosophical understanding is present in deed (ergon) in the dialogue. That deed, the construction of the dialogue, is Plato’s account of knowledge” (1992, 239); see also 244-5, where

22

Francisco J. Gonzalez

in the sense of expertise on knowledge is not to be attained and that the resulting lack of closure itself exhibits knowledge of knowledge in the sense of % of knowing what one does not know and thus continually the 8 striving for what one does not know. In other words, dialectic can provide an alternative to both the expertise of a Theodorus and Protagorean relativism. As a thinking that strives towards being and truth in the give-and-take of question and answer, it both can mis-take what it seeks and come to recognize this mistake as such. It is thus knowledgeable thinking without being the kind of possession of knowledge that characterizes expertise. In the words of the Philebus, “the power of dialectic” (1 2 !& , % µ , 57e6-7) is “the power of loving the truth” ( % µ I 2 , 2 , 58d4-5). To say that knowledge is inherently dialectical is therefore to say that it is a power of loving wisdom, rather than being the possession of wisdom. The infallibility that characterizes the possession of wisdom is indeed what Socrates desires, and therefore what he envisions in his construction of the perfect state in the Republic, but the only wisdom he claims to have and considers human is the disposition of having, properly directing, and acting upon this desire. In the Theaetetus Plato is not, as some have suggested,23 abandoning Socrates for his failure to define knowledge, or for the barrenness with which he afflicts others and is himself afflicted, but rather is holding him up as a measure of what knowledge can and should be, as the closest approximation to that divine measure that will always remain beyond our reach (?µ 8 , 7, 176b1). If the above reading of the Theaetetus is anywhere near the mark, then, as the Sophist suggests, only a definition of beings as power can include knowledge among the things that are. Indeed, to risk the use of terms already judged to be anachronistic, the epistemology of the Theaetetus can be said to require the ontology of the Sophist. Since knowledge can never be for us a stable and final possession – a static state of the soul to mirror a static reality – but a potency that remains a potency in being exercised and enacted, those who, like the ‘friends of the Forms’, identify what is with what is actual, must exclude from reality their own knowing relation to reality. Both the Theaetetus and the Sophist together show that, rather than trying to make dispositions fit an ontology modeled on what is actual, we should proceed in the opposite direction: allow our reflection on dispositions such as virtue and knowledge _____________

23

Polansky adds: “The entire dialogue, the commentary establishes, acts out what it is about” (245; see also 25). But Polansky appears to equate this knowledge exhibited in the deed of the dialogue with a systematic understanding of all the elements (239); he thus appears to miss the central dialectical and dispositional dimension of this knowledge. Long, for example, asserts that Plato drops Socrates in favor of “the detachment of theoretical from practical inquiry, the depersonalization of dialectic, and a new standard of rigour in the method of doing what we would call logic and metaphysics” (1998, 134).

Knowledge and Virtue as Dispositions in Plato’s Theaetetus

23

to bring into question and transform our ontology. Both materialist Giants and idealist Gods have thus much to learn from these two dialogues.

Literature Burnyeat, Miles. 1990. The Theaetetus of Plato. Indianapolis: Hackett. Diehls, Hermann and Wilhelm Schubart, eds. 1905. Anonymer Kommentar zu Platons Theaetet. Berlin: Weidmannsche Buchhandlung. Dixsaut, Monique. 1998. Le Naturel Philosophe. Paris: J. Vrin. Dixsaut, Monique. 1999. “What is it Plato Calls Thinking?” In Boston Area Colloquium in Ancient Philosophy, 13, ed. John J. Cleary and Gary M. Gurtler, 1-27. Leiden, Boston, and Cologne: Brill. Gonzalez, Francisco J. 2007. “Wax Tablets, Aviaries, or Imaginary Pregnancies? On the Powers in Theaetetus’ Soul.” Études Platoniciennes IV: 272-293. Kahn, Charles. 1996. Plato and the Socratic Dialogue. Cambridge: Cambridge University Press. Long, Anthony A. 1998. “Plato’s Apologies and Socrates in the Theaetetus.” In Method in Ancient Philosophy, ed. Jyl Gentzler, 113-136. Oxford: Clarendon Press. Polansky, Ronald. 1992. Philosophy and Knowledge: A Commentary on Plato’s Theaetetus. Lewisburg: Bucknell University Press. Sedley, David. 2004. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus. Oxford: Clarendon Press.

Aristotle’s Theory of Dispositions From the Principle of Movement to the Unmoved Mover LUDGER JANSEN It could well be argued that no one influenced and shaped our thinking about dispositions and other causal properties more than Aristotle. What he wrote about power and capacity (dynamis), nature (physis), and habit (hexis) has been read, systematized, and criticized again and again during the history of philosophy. In what follows, I will sketch his thoughts about dispositions and argue that it can still be regarded as a good theory.1

1. It’s all Greek to Me If asked to explicate the thoughts of an ancient thinker about some modern concept, the first problem to be solved is: Which word do I have to browse for in the index? The origin of the problems discussed in contemporary theories of dispositions – be it of dispositional predicates or of dispositional properties – dates back to the heyday of logical empiricism. The problem of dispositions arose from the quest for an intimate connection between experimental observations and the explanatory language used in scientific theories. This quest is very much a project of the twentieth century and it is, thus, no trivial matter that ancient thinkers had any thoughts about this particular topic at all. Nevertheless, the word “disposition” itself has a Latin origin and the Latin word dispositio has, in turn, a Greek equivalent, diathesis. But taken in this way, “disposition” means something like “orderly arrangement”, be it of things, of speeches, or of soldiers in an attacking army. Aristotle, of course, has a theory about the correct arrangements of the parts of a speech or of a drama, and he outlines it in his writings on rhetoric and poetics. But when we are asked for Aristotle’s theory of dispositions, “disposition” means some causal property. There is, of course, ample material on causal properties in the writings of Aristotle. Yet in these contexts, Aristotle uses words like dynamis (“power” or “capacity”), physis (“nature”), or hexis (“habit”). Accordingly, I _____________ 1

This article is a précis of my book on Aristotle’s theory of dispositions (Jansen 2002). I leave it to the reader to judge about how I deviate from other recent interpretations like Witt 2003 and Makin 2006.

Aristotle’s Theory of Dispositions

25

will start my discussion of Aristotle’s account of dispositional causal properties with presenting what Aristotle says about dynamis and will later contrast it with his statements about physis and hexis.2

2. From Homer to Aristotle When expounding his theory of dispositions, the key word for Aristotle is dynamis. In Aristotle’s time, this word was in common usage, and it can already be found in Homer. Here are four quotes featuring this word:3 [Odysseus:] but bring ye healing, my friends, for with you is the dynamis. (Odyssey X 69; transl. Murray) [Telemachos to Nestor:] O that the gods would clothe me with such dynamis, that I might take vengeance on the wooers for their grievous sin (Odyssey III 205-206; transl. Murray) [Alexandros to Hector:] we will follow with thee eagerly, nor, methinks, shall we be anywise wanting in valour, so far as we have dynamis; but beyond his dynamis may no man fight, how eager soever he be (Ilias XIII 785-787; transl. Murray) [Achilles to Apollo:] Verily I would avenge me on thee, had I but the dynamis. (Ilias XXII 20; transl. Murray)

In Homer, the dynamis is something with or within a man that allows him to fulfill a certain task or to defeat his enemy, and sometimes the dynamis is thought of as being given by a god. Afterwards, the word acquired a wide range of possible meanings. It can even refer to the riches of a wealthy man (cf. Plato, Republic 423a: chrêmata te kai dynameis) or the army of a kingdom (cf. Plato, Menexenos 240d: hê Persôn dynamis, the army of the Persians), and even the phonetic quality of a letter (cf. Plato, Cratylus 412c: tên tou kappa dynamin) or the meaning of letters and syllables (cf. Plato, Hippias maior 285d).4 From the sixth century BC onwards, the word dynamis is also used in philosophical and medical contexts.5 For example, Alcmaeon of Croton (ca. 570500) uses the term to define health (hygieia) as the balance of powerful things _____________ 2

3

4 5

That Aristotle’s theory of dynamis is a theory of dispositional properties has also be seen (among others) by Liske 1996. Already Wolf 1979 discusses both Aristotle’s theory and modern theories of dispositions, even though she discusses it under the name of “possibility” (“Möglichkeit”). The translation is Murray’s; I modified it by replacing Murray’s terms “power” and “strength” by the original dynamis. There are six more occurrences of the word in Homer: Ilias VIII 294, XIII 786 and Odyssey II 62, XX 237, XXI 202 and XXIII 128. Though the noun is quite rare, there are in all about 140 occurrences of words (including verbs and adjectives) containing the root dyna-. It would be worth to check our findings against this much broader basis. All occurrences of dynamis in Plato (and many in earlier authors) are collected and discussed in Souilhé 1919. For a survey of dynamis in the Hippocratic texts cf. Plamböck 1964.

26

Ludger Jansen

(isonomia tôn dynameis), that is, the equal presence “of moist and of dry, of cold and of hot, of bitter and of sweet” (DK 24 B 4). It is, however, not clear whether Alcmaeon uses dynamis to denote an abstract power or the powerful thing itself, i.e., whether dryness or the dry is the dynamis. In a quotation from Democritus (ca. 460-370), it is clear that the dynamis to be healthy is not some concrete thing but some property that resides in the human body (DK 68 B 234). It is exactly for this reason that people should care for their health by adjusting their diet rather than praying to the gods. This ambiguity is, perhaps, also reflected in Anaximenes’ (ca. 580-520) remark that neither the hot nor the cold are substances, but properties of an underlying matter (DK 13 B 1 = KRS 143: pathê koina tês hylês epigignomena tais metabolais). For Anaximenes, powers “interpenetrate the elements or bodies” that are their bearers (DK 13 A 10 = KRS 145: tas endiêkousas tois stoicheiois ê tois sômasi dynameis).6

3. Active Powers Defined In his theorizing about dispositions, Aristotle could, thus, draw on ample material from various philosophical and non-philosophical sources. There was an established linguistic usage of the word dynamis, at least since Homeric times. In addition, the word had already entered medical thinking and natural philosophy – and one can find beginnings of a more systematic treatment of the concept of dynamis in various authors. Yet, the first comprehensive treatise on dynamis, which we know of, is the one by Aristotle, i.e., the ninth book of his Metaphysics.7 Considering the by then quite respectable history of the word, it should not come as a surprise that Aristotle, in his well-known manner, treats dynamis as a word with many different meanings, as a polachôs legomenon, as something that is spoken of in many different ways. Although the word dynamis has many different meanings, Aristotle thinks that nearly all of them are related to one another, that they make up a sophistically knit web of meanings. At the center of this web is a meaning quite close to the Homeric use of the term: it is dynamis as an active power. For dynamis used in this way, Aristotle gives the following definition: _____________ 6

7

There is also a special use of dynamis and dynaton in geometry, which Aristotle explicitly mentions as a metaphorical use of the term (Metaphysics V 12, 1019b 33-34; IX 1 1046a 6-9). On this cf. Jansen 2002, 58-63 with further references. Smeets 1952 carves up Metaphysics IX 1-9 in many different passages by different hands, distinguishing bits written by Aristotle at different times in his life, his students or even later Aristotelians. Without doubt the text has its history and developed over same time. However, I show in Jansen 2002 that such a dissection of the text is not necessary and that, on the contrary, the whole text can be read as a contribution to one single theory.

Aristotle’s Theory of Dispositions

27

Dynamis means a source (archê) of movement (kinêsis) or change (metabolê), which is in something else or in itself as something else. (Metaphysics V 12, 1019a 15-16)

The words featuring in this definition are all widely used Greek words, but in Aristotle’s terminology they function as technical terms that are in need of an explanation. I will, in turn, explain what Aristotle means by the terms “principle”, “change,” and “movement”, and what he wants to express by the strange phrase “in something else or in itself as something else”. To begin with, a principle (an archê) is defined by Aristotle as “a first thing [...] from which movement and change take their inception” (Metaphysics V 1, 1013a18). In this vein, he calls the father and the mother the principles of the child (1013a9), because the coming-to-be of a child starts with an interaction between father and mother. “Change and movement” (kinesis and metabolê) are probably mentioned as a pair in the definition in order to indicate that an active power can be related to any of the different kinds of changes that Aristotle distinguishes at other places (notably in Categories 14, Physics V 2 and VII 2). According to Aristotle, one can distinguish between two fundamental types of changes. The first kind is substantial change; a coming-to-be or a passing-away of a substance, which is an entity that exists on its own, like a man, a dog, or a tree. Thus, birth is the beginning of a man’s existence and death the end of his existence; both are substantial changes. The other kind is the change of some accident, which can be further differentiated according to the category the changing accident belongs to. Aristotle acknowledges that there are three accidental categories with irreducible changes: quality, quantity and place. A change in quantity can either be growth or diminution.

4. The Location of Active Powers The strange phrase “in something else or in itself as something else” still needs to be explained. I will follow Aristotle’s own strategy and explain its meaning through the discussion of two examples, that is, architecture and medicine; or the art of building and the art of healing. Now, where is the art of building located? It is not in the house to be built, because this does not yet exist and non-existing things cannot be bearers of any properties. Nor is it in the building material: logs and stones know no art. It is, of course, in the builder (Metaphysics V 12, 1019a 16-17): He has the active disposition to bring about a change “in something else,” i.e., in the building material, from being mere logs and stones to being a new house. Thus the point of the first part of our strange phrase (“in something else”) is that an active power causes changes in something that is distinct from the thing that is the bearer of that power. The other part of Aristotle’s strange phrase can be illuminated with the

28

Ludger Jansen

help of his second example, the art of healing. Where can we find the art of healing? It is, obviously, in the practitioner, for example in Hippocrates. But what happens if Hippocrates becomes ill himself? In many cases, Hippocrates will be able to heal himself. It is the same ability that allows a person to heal other people when they have the flu and to heal himself when he has it – there is no necessity for Hippocrates to learn something new. But when he does indeed heal himself, Hippocrates is at the same time the bearer of the art of healing and the object undergoing the change of becoming healthy. This fact notwithstanding, Aristotle wants to classify the art of healing as an active power. Yet even though it is true that Hippocrates does not heal someone else, Aristotle would say that he heals himself “as another.” Aristotle explains this formulation in the context of his treatment of the difference between accidental and non-accidental happenings: [...] it may happen that someone becomes his own cause (aitia) of health, if he is a healer; but he has the art of healing not insofar as he is being healed, but it just happens (symbebêken), that the same person is a healer and is being healed. Therefore, [being a healer and being healed] are at times separated from each other. (Physics II 1, 192b 23-27)

Hippocrates’ ability to heal is independent from his being able to become healthy: His ability to heal is due to his study of medicine, his ability to become healthy is due to his being a human with a certain bodily constitution. There is no intimate connection between these two properties of Hippocrates – he can have the one without the other. It is only by accident that Hippocrates can heal himself. For this reason, Aristotle says that a practitioner may be able to heal himself, but if he does so, he heals himself as another, i.e., not as a practitioner, but as a human being with a certain bodily constitution. The art of healing is within the healed, but not as healed (Metaphysics V 12, 1019a 18).

5. Extending the Conceptual Network According to Aristotle, the word dynamis has many meanings. Most of them, or so Aristotle says, are systematically connected with one another, and the concept of an active power is at the core of this conceptual network. Intimately connected with it is the concept of a passive disposition. To have a passive disposition allows its bearer to undergo a change. A passive disposition is a principle of change in the bearer of the disposition, caused by something else or by itself as something else. Thus, in order to be realized or manifested a passive disposition requires a corresponding active power, and vice versa. Aristotle also talks about qualified dispositions, which are principles to do

Aristotle’s Theory of Dispositions

29

something well or to act after a decision to do so, as opposed to do something somehow or by accident. Aristotle illustrates this concept by contrasting a drunkard’s ability to walk with the ability to walk of a sober person. It should be clear that both can walk somehow. Yet only the sober person can walk well, i.e., without staggering and without pausing. Finally, Aristotle mentions resistance dispositions, which allow their bearers to resist changes and stay unchanged. If, for example, a rod is flexible, it can resist breaking when being bent. Thus, a resistance disposition is a principle for not being changed by something else. All of these different dynameis are ultimately related to an active power: Having a passive disposition means to have the disposition to be changed by something with a matching active power. Having a resistance disposition means to have the disposition not to be changed by something with a matching active power, and having a qualified disposition means to have any disposition in a qualified way, where this disposition is itself an active power or, again, related to an active power. Accordingly, Aristotle says that the concept of an active power is the core concept of dynamis, its kyrios horos (Metaphysics V 12, 1020a4). So far, the different varieties of dynamis are tied together by a so called pros hen structure: they all share an (implicit) reference to one and the same core concept of active power. However, in extending the conceptual network of dynamis, Aristotle also uses his second tool for enlarging conceptual networks; that is, analogy. In this manner, he introduces a second family of dynameis or dispositions: Our meaning [...] is as that which is building is to that which is capable of building, and the waking to the sleeping, and that which is seeing to that which has its eyes shut but has sight, and that which has been shaped out of the matter to the matter, and that which has been wrought up to the un-wrought. [...] some [of these] are as movement to dynamis, and the others as substance to some sort of matter. (Metaphysics IX 6, 1048a35-b9; transl. Ross)

This second family is introduced by a set of examples, and the reader is invited to recognize the similarity between these examples by considering analogous cases (tô analogon synhoran, 1048a 37). Those cases that are “as substances to some sort of matter” are said to stand in an analogy to those cases that are “as movement to dynamis :” Aristotle’s claim is that, in a way, a substance relates to its matter like a change relates to the respective dynamis. The new members of the conceptual network are no longer principles for change, like those varieties of dynamis we discussed before. Rather, they are principles for being something.8 Instead of principles for healing or for becoming healthy, we now deal with principles for being healthy, or for being red or _____________ 8

For the distinction between principles of change and principles of being cf. Berti 1990.

30

Ludger Jansen

round or a sword. Principles for change are relevant for dynamic causal explanations. If we want to explain how it comes about that this iron is a sword, we refer to the dynamis of a blacksmith to mold the iron into sword shape and to the matching dynamis of the iron to be molded in this manner. If, however, we want to explain how the sword’s iron is now, at this very moment of time, related to the sword, we are in search for a static ontological explanation. And Aristotle’s answer is, obviously, that the iron is realizing its dynamis to be shaped like a sword. Here we see how Aristotle’s theory of dispositions is relevant to the very heart of his ontology, the hylomorphic composition of substances of form and matter.

6. The Syntactical Structure of a Dynamis Ascription It is revealing to have a closer look at the Greek phrases that Aristotle uses to ascribe dispositions or dynameis.9 Most directly Aristotle ascribes a disposition by saying that something has a dynamis for something (echei tên dynamin tou … = “has the disposition to …”), but he also uses the verb dynasthai (“to be capable”) for this purpose; either a finite form of this verb like dynatai (“it is capable”) or the participle dynameon (“being capable”). He also employs the adjective dynaton (“capable”), of which Aristotle explicitly says that something is dynaton to do something, if it has the dynamis to do this (Metaphysics IX 1, 1046a20-21). To express that someone has the disposition to walk (badizein), the following Greek phrases can thus be used: echei tên dynamin tou badizein – dynatai badizein – dynameos badizein estin –dynaton esti badizein. In the context of Aristotle’s metaphysics, there is another phrase that is important here: dynamei badizontos estin. This phrase uses the dative case dynamei to express a certain respect (i.e. in its function as dativus respectus), saying that with respect to his dynamis, someone is a walker, traditionally translated as “someone is a potential walker.” The adjective dynaton can, however, also indicate that something is possible and in these cases dynaton estin means the same as “It is possible that” – and thus it is sometimes used synonymously with endechestai which means “It may happen that.” Aristotle himself discusses this use of dynaton and he explicitly says that this use of dynaton is ou kata dynamin (1019b 34), that it is not based on dispositions. It belongs to the talk about possibility, not to the talk about dispositions.10 To be sure, there are intimate connections between disposition talk and possibility talk. But there are important differences be_____________ 9 10

For textual references cf. Jansen 2002, 20-26. Cf. Jansen 2002, 21-24 on the use of dynaton in the context of modal logic and van Rijen 1989 on Aristotle’s overall theory of possibility.

Aristotle’s Theory of Dispositions

31

tween them and thus they have to be kept apart.11 To begin with, there is an intriguing syntactical difference that reveals, or so I will argue, a crucial ontological difference. Syntactically, “It is possible that …” is a sentence operator: It combines with a sentence and forms a sentence again. The phrases that are used to ascribe dispositions, on the other hand, are predicate modifiers,12 both in ancient Greek and in modern languages. Phrases like “… has the disposition to …” or “… is able to …” combine with predicates and form new predicates. They combine with, say, actualization predicates in order to yield disposition predicates.

7. The Ontological Structure of Having a Dynamis In the following, I will defend the claim that the above syntactical difference mirrors a crucial ontological difference. This will be obvious if we have a look at the usual possible worlds semantics for modal operators like “It is possible that …”.13 According to this approach, a sentence of the form It is possible that p is true in the actual world if and only if there is a possible world w such that w is accessible from the actual world and the sentence p is true in this possible world w. The truthmaker of such a sentence is not to be found in the actual world, but is located in some possible world. A dynamis, on the other hand, i.e. an ability or disposition, is something that can be encountered in the actual world. It is me in the actual world that has or does not have the ability to speak Chinese. Such an ability is a quality token of which I am the bearer. Thus a disposition ascription of the form x has the disposition to do (or to be) F is true if and only if there is a quality token d such that (1) x is the bearer of d and (2) d allows x to do (or to be) F. An Aristotelian dynamis is part of the furniture of the actual world, and dynamis ascriptions are about the actual world. They ascribe actual properties to actual things. By no means do they constitute a “ghost world” of mere possibilia.14 We can sum up Aristotle’s stand in this regard, by formulating two principles, the Bearer Principle and the Principle of Actuality. The Bearer Principle says that dispositions, like all other properties, have always a bearer. No dis_____________ 11

12 13 14

I argued for this in Jansen 2000. Buchheim/Kneepkens/Lorenz 2000 is a collection of essays that discuss the contrast between disposition talk and possibility talk from Aristotle to Heidegger. Cf. also Jacobi 1997. Cf. Clark 1970. For more references cf. Jansen 2002, 28-34. Cf. Weidemann 1984, Hughes/Cresswell 1996. Against Hartmann 1938, 5 (“Gespensterdasein”).

32

Ludger Jansen

position can exist without a bearer: a disposition exists if and only if there is a bearer having that disposition. The Principle of Actuality says that nothing has only potential properties or dispositions. If x has at time t the disposition to be or to do F, then there is at least one G, such that x is actually realizing G at t.15 The Principle of Actuality has a somewhat trivial instantiation, because for Aristotle the dichotomy between actuality and potentiality (or between categorical and dispositional properties) does not make up distinct classes of things but is meant to clear up ambiguities in language. One and the same property like mathematical knowledge is both a disposition and a realization. It is the disposition to solve mathematical problems, but at the same time it is the realization of the disposition to learn about mathematics (cf. De anima II 5, 417a 22-b 2 with Physics VIII 4, 255a 33-b 5). Consequently, a disposition is itself the realization of another disposition, and a potentiality is something that is actual. In this manner, the Principle of Actuality is trivially satisfied, if we choose “the disposition to be or to do F” as an instantiation for G.

8. Hartmann and Hintikka: Two Influential Interpretations We are now prepared to review the two interpretations of Aristotle’s teachings about dynamis that were probably most influential in the twentieth century: those by Nicolai Hartmann and Jaakko Hintikka. In his ontology of modality, Hartmann distinguishes between two kinds of possibility: total possibility and mere partial possibility.16 In Hartmann’s eyes, it is total possibility that is the only serious candidate for a rigorous treatment in an ontology of modality: A state of affairs s is called totally possible, if and only if all necessary conditions for s are given. As a consequence of his conception of determinism, necessary conditions are jointly sufficient. For this reason, various modalities collapse in Hartmann’s theory: Contrary to intuition, one can no longer extensionally differentiate between possibility and necessity: All and only totally possible states of affairs are necessary. Hartmann accepts this consequence, which is a rather unfortunate result in my eyes. But more important for his interpretation of Aristotle is Hartmann’s concept of partial possibility: A state of affairs s is partially possible if and _____________ 15

16

Cf. Kosman 1969, 43: “[...] for anything which is potentially A, there is some B which at the same time that thing is actually.” Menn 1994, 94 neglects the principle of actuality, although he seems to be conscious about it (cf. 95 n. 32), and thus ascribes Aristotle a theory of possibilia, i.e. a theory about non-being but possible things. Cf. also Stallmach 1959, 79, arguing against Hartmann 1938: “Auch bei Aristoteles kommt keine Möglichkeit vor ohne eine Wirklichkeit, die sie trägt, nur ist diese nicht – wie die Megariker wollen – schon die Wirklichkeit dessen, dessen Möglichkeit sie erst ist.” Cf. Hartmann 1938. On Hartmann’s modal ontology cf. Hüntelmann 2000, on his interpretation of Aristotle cf. Seel 1982 and Liske 1995.

Aristotle’s Theory of Dispositions

33

only if at least one necessary condition for s is given. Hartmann now accuses Aristotle that he has only dealt with the inferior concept of partial possibility and rejected the Megarian concept of dynamis (to be discussed in the next section), which Hartmann sees as a precursor of his own views.17 But of course there are many different kinds of necessary conditions for s, even if we take only those necessary conditions into account for which it is a contingent matter whether or not they obtain.18 Thus it is clear that Hartmann’s interpretation is far too unspecific as an interpretation of dynamis – while having a dynamis for F certainly is a necessary condition to do F, we do not do justice to Aristotle’s account of dynamis if we treat it as being on a par with the obtaining of just any necessary condition. While Hartmann interprets Aristotle in terms of his concept of partial possibility, Jaakko Hintikka’s interpretation draws on the so called Principle of Plenitude. In Hintikka’s wording, the Principle of Plenitude says that “[n]o unqualified possibility remains unrealised through an infinity of time.”19 The Principle of Plenitude is closely related to a temporal interpretation of the alethic modalities, i.e., of possibility and necessity. According to such a temporal interpretation, a proposition p is necessary, if and only if it is always the case that p, and it is possible, if and only if it is at least at one time the case that p. Now it is normally not disputed that it is always the case that p if p is necessary and that whatever is the case at some point in time must be possible.20 It is, however, not that clear that all possibilities will or even could be realized at some point of time. It is both possible that I sit at noon and that I stand at that time, but of course I can realize only one of these possibilities at noon. Even if we skip the reference to a certain time, there remain problems: It is possible that, in the future, my son will marry and start a family, but it is as well possible that he remains a bachelor for all his life. But, of course, not both possibilities can be realized. To discard such obvious counter-examples to the Principle of Plenitude, Hintikka talks about “unqualified possibilities:” Unqualified possibilities are possibilities that can, in principle, be realized at _____________ 17 18

19 20

Cf. Hartmann 1937. Hartmann’s interpretation of Aristotle is influenced by the – different – position of Zeller 1882. As any necessary proposition is implied by any statement, a necessary statement like “1 + 1 = 2” may be seen as expressing a condition that is necessary for any other statement. If seen thus, there are no states of affairs that are not partially possible, even impossible states of affairs are partially possible when we take “necessary condition” in the logical sense and allow necessary propositions to be included within the set of conditions. Hintikka 1973, 96. These two rules correspond to the rules of medieval logic that (a) it is valid to conclude actuality from necessity (ab necesse ad esse valet consequentia) and (b) to conclude possibility from actuality (ab esse ad posse valet consequentia). The scope of the following rule, however, is left too vague and can give rise for criticism. For there are necessary propositions like “1 + 1 = 2” or “At twelve o’clock it is twelve o’clock” that may be said to be true, but are maybe not true at any point of time but rather in some timeless manner.

34

Ludger Jansen

any point of a potentially eternal history, like the possibility that something red is round or the possibility that there exists an animal that is able to fly. It has been a matter of debate whether Aristotle does or does not accept the Principle of Plenitude. While Lovejoy, in his great study on the Principle of Plenitude,21 claims that Plato accepted the principle but Aristotle did not, Hintikka takes the opposite stand and attributes the principle to Aristotle, but not to Plato. I will not argue for any of these alternatives here, but rather draw attention to two important observations: (a) If Aristotle subscribed to the principle, it was nothing he took for granted. For in his De Caelo I 12 he presents a rather lengthy (and maybe fallacious) proof of this principle for the very special case of eternal entities. There he argues for the following claim: If it is possible for something to exist eternally, it will exist eternally, which in turn implies that all eternal beings are necessary beings. If the Principle of Plenitude would be a tacit background assumption of the semantics of dynaton or dynamis, he would not have needed to construct such an elaborated argument for this claim. Thus, for Aristotle, the Principle of Plenitude cannot be a trivial element of the semantics of dynaton. (b) Even if it were such an element, the “unqualified possibilities” that feature in the Principle of Plenitude are not the topic of Metaphysics IX, but rather the dispositions of finite things and people. In Metaphysics IX, Aristotle talks about architects and people skilled in other arts and sciences, about blind and seeing animals, about sitting and standing men, about fluitplayers, sperms and wooden boxes. These are all finite things having finite dispositions, i.e., dispositions that do not have all of eternity at their disposal for realizing themselves. Accordingly, a principle about “unqualified possibilities” would be of no help in explaining the teaching of Metaphysics IX. It is neither a plausible nor a helpful starting point when we try to make sense of Aristotle’s theory of dynamis. As different as Hartmann’s and Hintikka’s interpretations are, they do have something in common. Both Hartmann and Hintikka analyze Aristotle’ dynaton solely in terms of modal operators, i.e., as being the Greek equivalent of something like “It is possible that ...” or, in logical notation, “ p.” As I have argued in the last two sections, such a translation is both syntactically and ontologically misleading, if we care about the dynaton that is related to a disposition. Whoever, like me in this paper, cares about Aristotle’s theory of dispositions, has to interpret dynaton as a predicate modifier. Such an interpretation is both truer to the Greek syntactical constructions that Aristotle uses to ascribe dispositions and more appropriate for representing the ontological structure underlying these ascriptions. _____________ 21

Lovejoy 1936.

Aristotle’s Theory of Dispositions

35

9. The Megarian Challenge Aristotle himself had to defend his theory of dispositions against an alternative position put forward by a group of philosophers called “Megarians,” who argue for a position very similar to Hartmann’s account of total possibility.22 Aristotle describes this position as follows: There are some who say, as the Megarians do, that a thing can act only when it is acting, and when it is not acting it cannot act, e.g. that he who is not building cannot build, but only he who is building, when he is building [...]. (Metaphysics IX 3, 1046a 29-32)

The Megarians regard the actual realization of a property as a necessary and sufficient condition for having the disposition to manifest this property: x has a disposition to do or to be F at t if and only if x is actually F at t. Aristotle formulates no less than four arguments against this position, outlining the strange consequences (atopa, 1046a 33) that such a position would have: (1) Learning a craft is different from (and more difficult than) merely switching from non-employing to employing a craft. If the builder would not have any building disposition when not building, there would be no difference between a non-building builder and someone who is not a builder at all. (2) Also, there would be no difference between a thing being perceivable and that thing being perceived (and Protagoras would be right). For then a thing would be perceivable if and only if it would actually be perceived. (3) Also, people would many times become blind and deaf when closing their eyes or entering a silent room. (4) Finally, Megarians do away with change and becoming (and Parmenides rejoices), because if there is no principle of change to become something not yet existing, nothing can ever come into existence that is not yet present.23 To be sure, none of the above points constitute a knock down argument against the Megarian position. The Megarians could very well (and maybe they did) embrace the Parmenidean and Protagorean implications. However, any philosopher who, like Aristotle, sees some value in common-sense opinions and rejects positions that are more revisionary than necessary has plenty of reasons to reject the Megarian claim. This is the lesson Aristotle learns from the discussion of the Megarian position: Contrary to what the Megarians say, terms for the possession of a disposition and terms for their respective realization usually have different extensions. As a rule, dispositions are “twosided:” It is possible to have a disposition and not to realize it at the same time. _____________ 22 23

On the attempts to identify these philosophers cf. Jansen 2002, 139-143. For a formal account of this last argument cf. Jansen 2002, 146-149.

36

Ludger Jansen

One therefore has to distinguish between the time at which something has a disposition and the time for which this disposition allows a realization. An owl does already at daytime possess the disposition to realize sophisticated night-time vision when it is dark. Here, daytime is the at-time, i.e., the time at which the owl has that disposition, whereas night-time is the for-time, i.e., the time for which that disposition allows a realization. Disposition ascriptions in natural language contexts normally do not contain any reference to a for-time. And it would indeed be ontologically questionable to say that someone who can stand at day and at night has two distinct dispositions: one for standing at day and one for standing at night. Thus it should come as no surprise that some criticize such an analysis because “it does not make sense to speak of a capacity for standing-at-t, but only for standing.”24 But there is help on the way: We can get rid of the for-time parameter without falling back into the Megarian mess. The idea is the following: As a relevant causal factor for its realization, a disposition precedes its effect. Thus, the realization of a hitherto unrealized disposition could happen at some time in the future, given that the disposition does not get lost in between. Hence if something has at t a disposition to do or to be F, this disposition at least allows its bearer to realize F at some t* immediately after t. Logically speaking, what I suggest is to turn the free variable that the reference to the for-time has been in our previous formulations into a bound variable (bound by the existential quantifier “some”):25 We started with ascribing a dynamis for a realization for a specific time in the future; but now we ended up ascribing a dynamis for a realization at some time in the future – lest it be that the disposition gets lost and thus ceases to exist. Thus to say that someone has now a disposition for standing therefore is to say that he has now a disposition for standing-at-any-point-of-time-t-in-the-future-as-longas-the-disposition-does-not-get-lost. This means that we interpret a dynamis as a causal factor that precedes its effect and that may (but need not) be co-present with its realization. This preserves a real distinction between at-time and for-time as well as the appearance of disposition ascriptions in natural language.

10. Dispositions, Realizations, and Their Conditions In the Megarian picture, there was an intimate interconnection between having a disposition or dynamis and realizing it: According to the Megarians, something has a dynamis when and only when realizing it. In this picture, _____________ 24 25

Waterlow 1982, 40. For a formalization of this ideas cf. Jansen 2002, 152-154 and 194-196.

Aristotle’s Theory of Dispositions

37

manifesting a disposition is both necessary and sufficient for having the respective dynamis. Aristotle struggled hard to argue against the Megarian position, and to establish the possibility of unrealized dispositions. Consequently, realizing or manifesting a disposition can no longer be regarded as being a necessary condition for having a disposition. Nor can it be regarded as a sufficient condition for having a disposition, if co-presence with its realization is only a contingent and not a necessary feature of a dynamis. As he disposed of the Megarian position, Aristotle presents a new necessary condition for having a disposition: For x to have a disposition to do or to be F, it must be logically possible to assume that x actually does or is F.26 Such an assumption will lead to contradictions if we, for example, assume that the diagonal of the square has the disposition to be measured with the same unit as the length of one side.27 Now, when does a disposition become realized? This question does not arise in the Megarian picture, because there a dynamis does not exist at all before it is realized. Within the Megarian picture it may, however, be asked how and when a dynamis or its realization can come into existence. We do not know how the Megarians answered these questions, nor do we know whether the Megarians bothered to address them, in the first place. But since Aristotle allows for unrealized dispositions, there is a real question for him. He answers it by referring to the conditions that have to be met in order for a disposition to be realized: […] as regards dynameis of the latter kind [of the non-rational dynameis], when the agent and the patient meet in the way appropriate to the disposition in question, the one must act and the other be acted on […]. (Metaphysics IX 5, 1048a 5-7)

In this passage, Aristotle presupposes the contrast between rational and non-rational dispositions. I will discuss this distinction and its relevance in the next section. Here I will focus on what this passage tells us about non-rational dispositions, i.e., such dispositions that can also be had by non-living things, plants, or beasts. Such dispositions are realized, when the bearer of the active power (the “agent”) and the bearer of the complementary passive disposition (the “patient”) meet in an “appropriate way.” This implies normally that the bearers of complementary active and passive dispositions are contiguous, but it may also include further appropriate marginal conditions. Note that these conditions are conditions for the realization of a disposition, not for having the disposition. Otherwise Aristotle would not have managed to evade the Megarian problems. Moreover, the realization conditions of a dynamis belong to the definition of the dynamis in question: If we talk about dynameis with different realization conditions, we talk about different kinds of dynameis. For _____________ 26 27

On this principle cf. Weidemann 1999. The proof is to be found in Euclid, Elements X 117; it is alluded to in Analytica Priora I 23, 41a 267 and I 44, 50a 35-38. For the details of the argument cf. e.g. Jansen 2002, 159-162.

38

Ludger Jansen

this reason, Aristotle does not need to include a “if nothing external interferes” phrase into his account when a dynamis gets realized.28 Two standard realization conditions are that the dynamis does not cease to exist – which excludes that they are “finkish”, i.e. that they disappear when they are expected to realize themselves – and that no hindrances like antidotes are present. In this way, Aristotle has a plausible answer to two infamous problems of the contemporary debate of dispositions.29 To be sure, Aristotle has no formalized account of the contrafactual conditional made up out of all these realization conditions. Nor does he – contrary to many modern theorists about dispositions – think that such a conditional provides a reductive “analysis” of dynamis-predicates. Rather, he puts it forward as a claim about certain entities in the world. Finally, we may wonder whether the non-realization is necessary for having a disposition or not. That is, are “being F according to the disposition” and “being F according to the realization” compatible or incompatible predicates? There are certainly incompatible cases, like having a disposition for automatic self-destruction: Having such a disposition surely is not compatible with its realization, for if it is realized, there no longer is a bearer that could be the bearer of this disposition. On the other hand, there are cases where having a disposition clearly is compatible with realizing it. A medical practitioner, for example, does not loose his power to heal his patients when he actually does so. Otherwise he would be constantly loosing and re-gaining his power when beginning or ending the treatment of his patients.

11. Rational Dispositions A very special variety of dispositions are the so-called rational dispositions (dynameis meta logou, cf. Metaphysics IX 2, 1046b 2). There are several reasons for calling them rational dispositions. First, Aristotle describes these dispositions by saying that they are present in the rational part of the soul. They cannot be possessed by inanimate things, plants, or mere beasts. Second, these dispositions are accompanied by a logos, a rational formula; normally, this is the definition of the realization that the disposition can bring about. Third, these dispositions are realized through ratiocination, i.e. by means of practical syllogisms. What this means can be illustrated with the help of the art of medicine, which is Aristotle’s paradigmatic example for this kind of disposition. The “rational formula” that accompanies the art of medicine is _____________ 28 29

On this cf. Moline 1975. On “finkish” (i.e. disappearing in a deceitful way) dispositions cf. Martin 1994; on antidotes cf. Bird 1998.

Aristotle’s Theory of Dispositions

39

the logos or definition of health. Starting from a definition like “Health is XYZ,” the medical practitioner can deliberate about how and whether to heal his patients: Health is XYZ. XYZ will come about if I do F. I can do F. Thus I will do F.

A special feature of rational dispositions is that they can have contrary realizations. Medical knowledge is normally used to heal patients, but an evil doctor can use the very same knowledge to kill people. Consequently, the art of medicine can have effects as distinct as health and death. Moreover, rational dispositions cannot be realized in as simple a manner as the non-rational dispositions discussed in the preceding section. It is clear that spatial vicinity between a medical practitioner and an ill patient does not automatically lead to a realization of the practitioner’s healing disposition. First, the practitioner has to decide to activate his medical knowledge. But even this is not sufficient: The practitioner has also to decide on his goal: Does he want his patient to be healthy or dead? Only then is he able to consider and decide on possible means for the end chosen by him. And only then will he act in the appropriate manner, which may bring about the patient’s health or the patient’s death.30

12. Natures and Habits As already mentioned before, the different kinds of dynameis that I discussed up to now are not the only dispositional causal properties that Aristotle knows of. Other such properties are natures and habits (physeis and hexeis). But what are natures for Aristotle? Aristotle often remarks that a nature, a physis, is a principle of movement.31 Physis thus has the same genus as dynamis. But what is its specific difference? Aristotle spells this out in the following passage: And I mean by dynamis not only that definite kind which is said to be a principle of change in another thing or in the thing itself regarded as other, but in general every principle of movement or of rest. For nature (physis) also is in the same genus as dynamis; for it is a principle of movement – not, however, in something else but in the thing itself qua itself. (Metaphysics IX 8, 1049b 5-10, transl. Ross, italics mine; cf. De Caelo III 2, 301b 17-19)

_____________ 30 31

For a more detailed account cf. Jansen 2002, 78-92. Cf. Physics II 1, 193a 28-30; III 1, 200b 12f; De Anima II 1, 412b 17; Metaphysics V 4, 1015a 15-19; XI 1, 1059b 17-18.

40

Ludger Jansen

Thus whereas an active power is a principle of change “in another or as another,” a physis is a principle of change in a thing “in itself qua itself.” And whereas an active power needs a complementary passive disposition in order to be realized, there is no such need for a physis. If something has a physis to do or to be F, its realization depends only on the appropriate marginal conditions, but it does not require the spatial vicinity of the bearers of other causal properties. Another kind of causal properties goes under the name of hexis. Like dynamis, hexis is a word with many different meanings, to which Aristotle dedicates a chapter in his dictionary of ambiguous philosophical terms (Metaphysics V 20). The noun hexis derives from the verb echein, “to have.” As this etymology indicates, a hexis is in general either the having of something or that what is had by something. As a further possible meaning, Aristotle proposes the following definition: Hexis means a disposition (diathesis) according to which that which is disposed is either well or ill disposed, and this either in itself (kath’hauto) or with reference to something else (pros allo). (Metaphysics V 20, 1022b 10-12)

What is of particular interest for us, are the hexeis of the non-rational faculties of the soul, which determine both our emotional reactions and many of our actions. Traditionally, these hexeis are called virtues and vices: Virtues, if they dispose for good acting; vices, if they dispose for bad acting. On first sight, a virtue like justice has a structure similar to a dynamis. At a given time, someone can have the virtue without acting justly, e.g., when sleeping. And when the just person is acting justly, the virtue of justice is thought to have a causal influence. Virtues (and vices) are realizable and causal properties, but Aristotle takes great pains in distinguishing non-rational virtues from rational dynameis. For we have seen that in the case of a rational dynamis, like the art of medicine, one and the same dynamis can be the cause of contrary realizations, i.e. of health and death. The art of calculating just prices is such a rational dynamis – but just as medicine can also be used to kill people, this art can be used to calculate and to charge unjust prices (cf. Nicomachean Ethics V 1; Plato, Hippias minor). He who has the virtue of justice does not only know what is just, he is also inclined to act justly. Whereas a rational dynamis allows for contrary realizations, a virtue is directed to one realization only. And while a rational dynamis needs an appropriate will and goal in order to be realized, a virtue informs the will by itself and does not need the addition of a goal of action from the outside.

Aristotle’s Theory of Dispositions

41

13. Does the Unmoved Mover Possess Dispositions? Finally, I want to turn to one of the most prominent elements of Aristotle’s metaphysics, the godly unmoved mover, who keeps the heavens in circulation. In this context, one is inclined to inquire whether the unmoved mover possesses any dispositions, any dynameis. In Metaphysics IX 8, where Aristotle argues for the priority of realizations over dispositions, we find contradictory evidence on this matter. There (in 1050b 8-11), Aristotle says the following: (Z1) “Every dynamis is at the same time [a dynamis] for the opposite.” (Z2) “For, while that which is not capable (dynaton) of being present in a subject cannot be present,” (Z3) “everything that is capable (dynaton) of being may possibly (endechetai) not be actual.”

Taken together, (Z1) and (Z3) suggest that what is eternal has no dynamis, because for him everything that is eternal is necessary and cannot not be otherwise (De Caelo I 12). But if we accept this, then we are forced to say that, whatever eternal things do, is not based on a dynamis to do this. But (Z2) seems to contradict this conclusion in articulating the following principle of enabling: Everything that happens, happens because there have been dynameis that enabled this happening. Otherwise it would not have happened.

If this is universally valid, everything that eternal entities are or do is based on dynameis, too. We are obviously faced with a trilemma: (A1) What is eternally F, is necessarily F. (A2) What is eternally F, has the dynamis to be F. (A3) All dynameis are two-sided.

These three propositions are jointly incompatible. Now (A1) is not a topic in Metaphysics IX, but it is defended in De Caelo I 12 and Aristotle does not challenge this principle anywhere else. We may thus reject (A2) or (A3). To reject (A2) is to reject the Principle of Enabling. To reject (A3) is to admit “one-sided” dispositions, that is, dispositions that are necessarily realized. That we do indeed have these options is confirmed through a passage in De Interpretatione 13: For the term dynaton is not said with one meaning only (ouk haplôs), but at one time it is true that it is realized, as when someone [is said] to be able (dynaton) to walk because he walks, and generally when something is able [to be something] because that which it is said to be able of is already realized; but sometimes because something may be realized, as when a man [is said] to be able to walk because he may walk. The latter belongs only to that which is changeable; the former can also belong to the unchangeable things. [...]. Now, while the one way to be dynaton cannot truly be said of things

42

Ludger Jansen

being necessary in the unqualified sense, the other [way to be dynaton can be predicated] truly. (De Interpretatione 13, 23a7-16; my translation)

The author here clearly distinguishes between inclusive and exclusive predication of being dynaton to do or to be something. In an inclusive manner, it is said, even unchangeable and necessary things (like the unmoved mover) can be said to be dynaton to do or to be something. Thus whoever wants to ascribe dynameis to the unmoved mover has to accept that these dynameis are never unrealizsed. Otherwise we should refrain from ascribing dynameis to the unmoved mover. This would still not imply that what the unmoved mover does is inexplicable, for, as we have seen, Aristotle knows principles of change and being like natures that go beyond the sphere of dynamis. And while the passage in De interpretatione 13 leaves open which horn of the dilemma is to be preferred, the passages in Metaphysics IX 8 suggest that Aristotle wants to restrict the term dynamis to two-sided dispositions. If this is indeed Aristotle’s last word on the matter, then the unmoved mover cannot possibly have any dynameis.

14. Is Aristotle’s Account of Dispositions a Good Theory? Aristotle’s philosophy has often been criticized. Notably Hobbes dismissed Aristotelian thinking as “vain philosophy” and claimed “that scarce any thing can be more absurdly said in natural philosophy than that which is called Aristotle’s Metaphysics.”32 In particular, Aristotle’s theory of dynamis has been the object of many disputes. There are three standard objections against it: (1) Aristotle’s powers, dispositions and potentialities create a ghostly world of possiblilia, (2) they are explanatory idle (the virtus dormitiva objection), and (3) they are empirically inaccessible. I will discuss and reject each of these objections in turn.33 The first objection attacks the suppossedly dubious ontological status of dynameis. They are said to form a “ghost world” in between being and notbeing34 or to be a kind of “half-being.”35 In fact, I have already answered this objection when explaining the Bearer Principle and the Principle of Actuality. A power or disposition is nothing ghostly nor something that has only halfbeing: It is a full-fledged property of a full-fledged thing. It is, however, a fullfledged property with a certain peculiarity: It is related to some action, passion or another property, which it enables or causes, and which is called the reali_____________ 32 33 34 35

Hobbes, Leviathan, ed. Tuck, 461. Cf. also Jansen 2001, 276-278 and Jansen 2004. Hartmann 1938, 5 (“Gespensterdasein”). Tegtmeier 1997, 36-40 (“Halbexistenz”).

Aristotle’s Theory of Dispositions

43

zation of the disposition. Now it is possible, that a disposition occurs without being realized, but this does not diminish the ontological status of the disposition itself (but concerns only the non-occurring of the realization at this time). The second objection says that referring to dispositions does not explain anything, but rephrases in new words the problem in question. Instead, it is claimed, science has to explain phenomena by describing the world’s microstructure. This objection is often put forward in connection with Molière’s joke at the expense of the medical profession in his Le Malade Imaginaire. There, a to-be doctor of medicine answers during his doctoral viva voce examination:36 BACHELERIUS: “I am asked by the learned doctor for the cause and reason that opium makes one sleep. To this I reply that there is a dormitive virtue in it, whose nature it is to make the senses drowsy.” – CHORUS: “Very, very, well answered. The worthy [candidate] deserved to join our learned body.”37

Though in the play the examination board is full of praise for this answer, it is not apt to increase the reputation of the medical profession from the perspective of the audience of the play. Obviously, this answer does indeed only rephrase the problem. It is not at all informative. Yet, this does not imply that science can do without dispositions. First, the answer is not informative because the question already presupposes that it is the opium which is the relevant causal factor. If asked, why someone fell asleep, it would actually be informative to point out that the job had been done by the opium and not by some other thing around in this situation. Second, how could an informative answer to the original question look like? We could point out that opium consists out of 37 alkaloids, among which is morphine. But this would only be a satisfactory explanation if we know that morphine has a virtus dormitiva. Of course, we can also ask why morphine has such a dormitive virtue. And we could refer to some molecular structures in our nervous system and to the molecular structure of the morphine. Again, this answer can only be satisfactory, if we know something about the dispositions of the molecular structures in question, e.g., that the morphine molecules have the disposition to bind to and to activate certain receptors in our nervous system, and that the respective parts of our nervous system have the matching passive disposition. Again, we do not have totally eliminated the talk about dispositions, but only replaced the talk about one disposition with the talk about another disposition. This shows that we do not explain certain events merely by referring to properties of microstructures and merely by using categorical property terms. We also need dispositional property terms. The third objection claims that dispositions are empirically inaccessible, _____________ 36 37

On this scene and its background in the philosophical and theological discussions of Molière’s time cf. Hutchinson 1991. Cf. Molière 1926, VIII 328; the translation is Hutchinson’s (1991, 245).

44

Ludger Jansen

because we perceive only their realizations. Accordingly, dispositions have to be regarded as monsters of bad metaphysics. Obviously, we should be careful with this kind of argument: Otherwise one could argue analogously that the whole ‘external world’ is empirically inaccessible and thus a monster of bad metaphysics, because we are acquainted with ‘internal’ sense data only. The natural reaction to such an argument would consist in saying that we perceive the world through our senses and sense data. In a similar way, dispositions are not only described in terms of their realizations, but also recognised through them. Along such lines Aristotle admits the epistemological priority of the realization, through which the dynamis can be recognized (Metaphysics IX 8, 1049b 13-17). But although the realization is epistemically prior, the dynamis can nevertheless be recognized: By showing his students calculating, a teacher of mathematics can provide evidence for the claim that his students have acquired the dynamis for calculations and thus prove the efficiency of his teaching (1050a 17-19). Hence Aristotle needs not to be impressed by these three objections. His account of dispositions can still be regarded as a consistent ontology of causal properties with an enormous explanatory appeal.

Literature Berti, Enrico. 1990. “Il Concetto di Atto nella Metafisica di Aristotele.” In L’atto aristotelico e le sue ermeneutiche, ed. Marcello Sánchez Sorondo, 43-61. Rome: Herder 1990; German transl.: “Der Begriff der Wirklichkeit in der Metaphysik des Aristoteles” In Aristoteles. Metaphysik. Die Substanzbücher (Klassiker auslegen 4), ed. Christof Rapp, 289-311. Berlin: Akademie-Verlag 1996. Bird, Alexander. 1998. “Dispositions and Antidotes.” Philosophical Quarterly 48: 227-234. Buchheim, Thomas, and Corneille Henri Kneepkens, Kuno Lorenz, eds. 2001. Potentialität und Possibilität. Modalaussagen in der Geschichte der Metaphysik. Stuttgart-Bad Cannstatt: Frommann-Holzboog. Clark, Romane. 1970. “Concerning the Logic of Predicate Modifiers.” Nous 4: 311-335. Hartmann, Nicolai. 1937. “Der Megarische und der Aristotelische Möglichkeitsbegriff.” Sitzungsbericht der Preußischen Akademie 10 (1937), repr. in: Kleinere Schriften, vol. II. Berlin: de Gruyter 1957, 85-100. Hartmann, Nicolai. 1938. Möglichkeit und Wirklichkeit, Berlin: de Gruyter. Hintikka, Jaakko. 1973. Time and Necessity. Studies in Aristotle’s Theory of Modality, Oxford: Clarendon Press. Hobbes, Thomas. 1651. Leviathan, ed. Richard Tuck. Cambridge: Cambridge University Press 1991. Homer. The Iliad (Loeb Classical Library), transl. A. T. Murray. Cambridge, MA: Harvard University Press 1924. Homer. The Odyssey (Loeb Classical Library), transl. A. T. Murray. Cambridge MA: Harvard University Press 1919. Hughes, G. E., and Max J. Cresswell. 1996. A New Introduction to Modal Logic. London-New York: Routledge.

Aristotle’s Theory of Dispositions

45

Hüntelmann, Rafael. 2000. Möglich ist nur das Wirkliche. Nicolai Hartmanns Modalontologie des realen Seins. Dettelbach: Röll. Hutchinson, Keith. 1991. “Dormitive Virtues, Scholastic Qualities, and the new philosophies.” History of Science 29: 245-278. Jacobi, Klaus. 1997. “Das Können und die Möglichkeiten. Potentialität und Possibilität.” In Cognitio humana: Dynamik des Wissens und der Werte. Akten des XVII. Kongresses für Philosophie. ed. Christoph Hubig, 451-465. Berlin: Akademie-Verlag; repr. in: Buchheim/Kneepkens/ Lorenz 2001, 9-23. Jansen, Ludger. 2000. “Sind Vermögensprädikationen Modalaussagen?” Metaphysica, special edition 1: 179-193. Jansen, Ludger. 2002. Tun und Können. Ein systematischer Kommentar zu Aristoteles’ Theorie der Vermögen im neunten Buch der Metaphysik. Frankfurt: Hänsel-Hohenhausen. Jansen, Ludger. 2004. “Dispositionen und ihre Realität.” In Was ist wirklich? Neuere Beiträge zu Realismusdebatten in der Philosophie, ed. Christoph Halbig and Christian Suhm, 117137. Frankfurt-Lancaster: Ontos. Kosman, L. A. 1969. “Aristotle’s definition of motion.” Phronesis 14: 40-62. Liske, Michael-Thomas. 1995. “In welcher Weise hängen Modalbegriffe und Zeitbegriffe bei Aristoteles zusammen?” Zeitschrift für philosophische Forschung 49: 351-377. Liske, Michael-Thomas. 1996. “Inwieweit sind Vermögen intrinsische dispositionelle Eigenschaften?” In Aristoteles. Metaphysik. Die Substanzbücher, ed. Christof Rapp, 253287. Berlin: Akademie-Verlag. Lovejoy, Arthur O. 1936. The Great Chain of Being. A Study of the History of an Idea. Cambridge MA: Harvard University Press. Makin, Stephen. 2006. Aristotle. Metaphysics Book Theta, transl. with an introduction and commentary (= Clarendon Aristotle Series). Oxford: Clarendon Press. Martin, C. B. 1994. “Dispositions and Conditionals.” Philosophical Quarterly 44: 1-8. Menn, Stephen. 1999. “The Origins of Aristotle’s Concept of : and µ .” Ancient Philosophy 14: 73-114. Molière, Jean Baptiste. The Plays of Molière, transl. A. R. Waller, 8 vols. Edinburgh: John Grant 1926. Moline, Jon. 1975. “Provided Nothing External Interferes.” Mind 84: 244-254. Plamböck, Gert. 1964. “Dynamis im Corpus Hippocraticum.” Akademie der Wissenschaften und der Literatur. Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse. Nr. 2/1964. Seel, Gerhard. 1982. Die Aristotelische Modaltheorie. Berlin-New York: de Gruyter. Smeets, Albert. 1952. Act en potentie in de Metaphysica van Aristoteles. Historisch-philologisch onderzoek van boek IX en boek V der Metaphysica. Leuven: Universiteitsbibliotheek. Souilhé, Joseph. 1919. Étude sur le terme dans les dialogues de Platon. Paris: Felix Alcan. Stallmach, Josef. 1959. Dynamis und Energeia. Untersuchungen am Werk des Aristoteles zur Problemgeschichte von Möglichkeit und Wirklichkeit. Meisenheim am Glan: Hain. Tegtmeier, Erwin. 1997. Zeit und Existenz. Parmenideische Meditationen. Tübingen: MohrSiebeck. van Rijen, Jeroen. 1989. Aspects of Aristotle’s Logic of Modalities. Dordrecht-Boston-London: Kluwer. Waterlow, Sarah. 1982. Passage and Possibility. A Study of Aristotle’s Modal Concepts. Oxford: Clarendon Press. Weidemann, Hermann. 1999. “Aus etwas Möglichem folgt nichts Unmögliches. Zum Verständnis der zweiten Prämisse von Diodors Meisterargument.” Philosophiegeschichte und logische Analyse 2: 189-202.

46

Ludger Jansen

Witt, Charlotte. 2003. Ways of Being. Potentiality and Actuality in Aristotle’s Metaphysics, Ithaca, NY: Cornell University Press. Wolf, Ursula. 1979. Möglichkeit und Notwendigkeit bei Aristoteles und heute. München: Fink. Zeller, Eduard. 1882. “Über den des Megarikers Diodorus.” Sitzungsberichte der Preußischen Akademie der Wissenschaften, Berlin 1882, 151-159; repr. in: Eduard Zellers kleine Schriften, vol. I, ed. Otto Leuze, 252-262. Berlin: Reimer 1910.

Dispositions in Greek Historiography BURKHARD MEISSNER “Um von der Antike sich Rechenschaft zu geben, ihrem so weit reichenden Einfluss, gilt es zuallererst, beide Ohren fest zu verstopfen wie die Gefährten des Odysseus. Man kommt nicht weit, lauscht man dem christlichen Sirenengesang, der seit Jahrhunderten ablenkt von den klassischen Texten”. 1

In the above passage, the German poet Durs Grünbein uses Odysseus as a metaphor for what lies between us and the classical tradition: The dangerous sirens represent what distiguishes the Christian Occident from the core of ancient paganism and European cultural identity. As a whole, Grünbein’s essays in his book are concerned with identifying this ancient core as a necessary, though not sufficient condition for the existence of modern Europe: as Antike Dispositionen (ancient Dispositions), as if the ancient world had a disposition to modern development, which, in turn, would have been triggered somehow by historical circumstances and contingent factors. Ancient literature having but a disposition to framing notions of dispositions: Grünbein’s metaphor provides us with a sufficiently complex picture of ancient historiography and its use of dispositional explanations as a precursor to the modern debate about dispositions in history, sociology and psychology. In looking at their ancient precursors, we should, however, neither expect much familiarity, nor complete difference. What I am going to do is presenting a few examples of dispositional explanations and notions for dispositions from Greek historians between Herodotus and Polybius, which I think are representative not only of classical approaches to human behavior, but also of some ways we still speak about it in the social and historical disciplines. Today, people talk about dispositions referring to a variety of things. The concept is especially prominent in teachers’ training research, where educating 1

Grünbein 2005, 395. The text continues: “Mehr noch, man müßte zuerst die Stimme des eigenen Ichs unterdrücken lernen, denn die Beschwichtigungen kommen von innen, aus dem eigenen Echoraum.” („In order to render account to oneself for antiquity and its far-reaching influence, it is first of all necessary to plug one’s ears tightly as the companions of Odysseus did. One does not come a long way listening to the siren song of Christianity, which has been a distraction from the classical texts for centuries. Moreover, at first one should learn to suppress the inner voice of one’s self, because all the appeasements come from the inner side, from one’s own echosphere.“)

48

Burkhard Meißner

teachers is held to consist of developing and strengthening some more or less pre-existent dispositions to being a teacher into a habit which is typical of a master teacher. Evaluating teachers consists in profiling their behavior according to a set of moral values that define the core of what one expects from a teacher or thinker.2 In this context, dispositions are necessary conditions for being a master; they can, more or less, be present in people’s behavior, giving some leeway for quantitative or qualitative differentiations. In metaphysical contexts, this deviation from the determinist model has raised the question, whether or not dispositions can be used in “causal” explanations at all.3 As we shall see, this educational dimension is also prominently present in the ancient concept of disposition, if, for the time being, we allow for some lack of precision as to what dispositions are.

Dispositions in Ancient Biographies My first ancient example is taken from Plutarch, who, around the turn from the first to the second century AD, did not write history in the proper sense of the word, but biographies. More appropriately, they should be termed ethographies, descriptions of manners, since (1) these texts are seldom ordered along the chronological lines of a life-span and since (2) they do not deal primarily with the details of their protagonists’ lives, but rather with their manners, their attitudes and principles, as much as they can be made transparent by merely narrating their deeds. In most cases, Plutarch juxtaposes the life of one Greek with that of a Roman of similar kind, adding some comparative remarks after each pair of biographies. Within the corpus of his works, the mythical figures Theseus and Romulus figure prominently as the first couple of biographies. Plutarch’s syncrisis of both begins with the words “this is, what we found worth mentioning about Romulus and Theseus. ) ....”4 The most The first of these two, out of his own design ( important term here is . It means a quality of one’s behavior that can be morally evaluated as bad or good. It manifests itself in a sense in one’s deeds. It is also, in virtue of its being a stable personal quality, an efficient cause of those deeds, a necessary, but not sufficient condition. The notion of is very similar to the concept of disposition. And it seems to be 2

3 4

For a review of this debate in German cf. A. Meißner 2004, 9ff.; 30ff.; 55ff. Cf. also Tishman/Jay/Perkins; Richard; Tishman/Andrade; Johnson and Newman 1996; Katz 1993; The Symposium on Educator Dispositions Proceedings 2003. Negative: Raymont, Does Anything Break Because it is Fragile?; Raymont 2001. Positive: Mumford 1998; Vanderbeeken and Weber 2002. Plutarch, Comp. Rom. et Thes. 1,1.

Dispositions in Greek Historiography

49

this concept that Plutarch wants to convey to his readers, because he talks about it so prominently, especially in his comparative chapters. It is, therefore, with the ancient biographers, Plutarch and Cornelius Nepos, that we find most of the material, which is related to notions of dispositions. I am just giving a few examples to illustrate how, in ancient biographies, actions are conceived of as displaying dispositions that in turn can be used as positive or negative paradigms for modeling one’s own behavior. In the case of Demetrius Poliorcetes and Marcus Antonius, Plutarch’s comparison first extends to the amount of rapid change that both figures brought about. Afterwards, in its second chapter, he turns to the , the disposition, which their deeds seem to make transparent, and which are praised in the case of Demetrius: Demetrius, Plutarch holds, cannot be censured for having established himself as a ruler over subjects who actually wanted to be governed and accepted him out of their own accord. In contrast, Antonius tried to set up a kind of tyrannical régime over people who had already resisted a similar attempt on Caesar’s part.5 Yet, what is the role of dispositions in explaining rather than evaluating human behavior? Let us look at an example from Plutarch’s Life of Demetrius, in which Demetrius is depicted confronting Pyrrhus about the domination over heartland Macedonia in 289 BC. Demetrius was invading Aetolia, where Pyrrhus had left his general Pantauchus with a garrison. Plutarch explains Demetrius’ motives by alluding to his disposition: Demetrius was not disposed to keeping quiet, and he realized how much making war upon external enemies helped him secure the loyalty of his subjects.6 Thus, Plutarch explains Demetrius’ action using two principles, a disposition (Demetrius’ adventurousness) and Demetrius’ rational calculation as to how effectively to maintain his position in Macedonia, given the situation as it is. This is a general feature of Plutarch’s and the biographers’ mode of explaining facts and actions. In the same historical context, but in a different biography, in his Life of Pyrrhus, Plutarch describes how Demetrius’ general Pantauchus provoked king Pyrrhus to a hand-to-hand fight:7 Pantauchos who in regard to his courage, his practical skills with weapons and his corporeal fitness was undoubtedly the best among Demetrius’ generals, and who was 5

6

7

Plutarch, Comparatio Demetrii et Antonii 2,1: The disposition ( ) out of which they gained their rulership, is, with regard to Demetrius, not to be criticized, because Demetrius attempted at ruling over people and being their king, who were accustomed to being ruled and living under a monarchic system; Antony’s disposition, however, was bad and tyrannical, for after the Roman people had just been freed from Caesar’s monarchy, Antony attempted at making them slaves anew. Plutarch, Demetrius 41,1: Hence he marched back to Macedonia, and since he was not disposed to keeping quiet and realized that his people remained loyal in wartime, but stirred up rebellion and revolution while at home, he undertook an operation against the Aetolians. Plutarch, Pyrrh. 7, esp. 7,4(7)N5(10).

50

Burkhard Meißner

self-confident and ambitious enough, on this occasion provoked Pyrrhus for a fight man-to-man; Pyrrhus, on the other hand, who wanted to be second to none of the kings in courage and audaciousness, and who expected to enjoy the fame of one Achilleus more for his deeds than because of his birth, went through the first line of battle in the direction of the enemy to Pantauchos.

Pantauchus’ courage ( ), his corporeal fitness ( µ ), self-confidence ( ) and disposition to sound judgment ( µ ), as well as Pyrrhus’ own courage and adventurousness are cited as the guiding dispositions towards this kind of behavior, which, as a matter of fact, was nearly as exceptional in the Hellenistic World as it would be today. What Plutarch’s explanation actually amounts to is something like: Both individuals were disposed to behaving as they did under the prevailing conditions, and this is why they behaved the way they did. It is easily conceivable that such a line of argument is well suited to a kind of literature which, like Plutarch’s Lives, aims at producing positive or negative paradigms for human behavior, since, after all, such literature is less about explaining behavior in detail as it is about describing general dispositions and praising or blaming them. It is, therefore, no accident that we find notions of dispositions employed ordinarily in the ancient biographical literature, where they are used to evaluate rather than to explain people’s behavior. Biographies are a rather late development in ancient literature. Specifically the genre of political biographies is, as Joseph Geiger has shown conclusively, a late offshoot from the tree of historico-political literature, which has to be seen in the context of the intellectual climate of the first century B.C. and of the work of Cornelius Nepos.8 Dispositions, however, can be found much earlier, and Aristotle in his Ethics dwells upon what virtues are (in the end, he claims, they are dispositions, ), in which he seems to have been followed by the 2nd century Greek historian Polybius. Moreover, both use the or as a term for dispositions, the difference being that the former is just a disposition, whereas the latter, according to Aristotle, is a disposition, which in addition to having developed has done so on the basis of conscious choice: The consciously chosen disposition (prohairesis) is, on the one hand, a disposition; however: It is not simply a disposition, but one that prefers the one over the other. This, in turn, is not possible without empirical observation and rational choice. Therefore, the consciously chosen disposition (prohairesis) presupposes an opinion based on practical wisdom.9 8 9

J. Geiger, Cornelius Nepos and Ancient Political Biography, Stuttgart, Wiesbaden (1985) Cf. N. Horsfall, Cornelius Nepos, Oxford (1989). Aristoteles, Ethica Eudemia 1226b6-9: µ , ! " #$ %, ##' ' %

(

) ' % (· + ! , - ( .%/ 0 . 1 2 (#3 . ) . 2 (# ( .3 . Cf. B. Meißner, 567897:;:?6;7. Polybios über den Zweck pragmatischer

Geschichtsschreibung, Saeculum XXXVII (1986) 313-351.

Dispositions in Greek Historiography

51

The notions of hexis and proheiresis can also be found already in Plato, although the term prohairesis is rather uncommon in Plato’s dialogues: We find it once in the Parmenides used for a consciously chosen alternative.10 In the spurious Definitions it is used for a more constant disposition in the Aristotelian sense: There, friendship ( # ) is defined as a of the life of the other or as a community of and @ 11, and (µ as a manifestation of a practical .12 is, of course, much more common in Plato. In any case: The notion is prominently present, and it refers to a disposition, which can be consciously chosen, and which can be rationally debated about. Our first look, and it should not be more than a first look, shows: There was an abstract terminology for dispositions since the 4th century B.C. Moreover, the notion of disposition played a prominent role in the moralistic ancient biographical literature, with probable precursors in those philosophers who prominently dealt with moral issues.

Dispositions in Herodotus Let us now turn to historiography proper, and to the first pragmatic historian and historian of war, Herodotus.13 Notions which come close to dispositions are used by Herodotus to describe how the Persian Harpagos intended to dethrone the Median king Astyages with the help of the latter’s grandchild, Cyrus. According to Herodotus, Harpagos’ intention was directed towards (µ A 14. Harpagus revenging himself upon Astyages: Herodotus calls this himself, for his rank was insufficient, tried to use Cyrus as an instrument for his own design. After notifying and convincing some of the nobles about his plans, Harpagos had to find a way to directly influence Cyrus’ will. This he did using a secret messenger. Generally: In Herodotus’ narrative, Harpagus’ (µ functions like a kind of background process or continuous condition for what is realized under more specific conditions as an action, i.e.: revolution in Persia. For the success of Cyrus’ plan guiding this revolt, a disposition is again one of the necessary and somehow sufficient conditions: In this 10

11

12 13 14

Plato, Parmenides 143c: If we choose from them whatever you like, i.e. the essence and the difference, or the % ), essence and the singularity, or the singularity and the difference, do we not, in any of these choices ( choose what could rightly be called “both”? Ps.-Plato, Definitiones 413a-b: Friendship: a kind of consensus as to what is good and just. A preference ( ) for the same way of life. Shared views about dispositions and behaviour. Consensus about the way of life. Partnership, based on goodwill. Partnership in mutually doing well. Ps.-Plato, Definitiones 413e. Cf. Meißner 2004. I 123.

52

Burkhard Meißner

situation, the Persians were delighted to free themselves, having found a leader, and since long they had been impatient with Median rule.15 Here, a Persian disposition towards not bearing Median rule is presupposed; to this add the situational circumstances of a person capable to rule over them and Harpagus as a general who wants to get away with his king; and as a result you get revolution. A disposition that figures prominently in Herodotus is the agent’s state of mind, especially if it is “non-standard.” The most prominent non-standard state of mind that is exemplified and triggered under specific circumstances is madness, and the most prominent madman in Herodotus is the Persian king Cambyses, whom Herodotus describes at length in his third book (III 25): When they had reported this to him, Cambyses was stirred up by anger and immediately set out on his expedition against the Ethiopians without having ordered any provisions of food for the army and without realizing that he was about to lead an operation to the most extreme parts of the world. He was so mad and without rational control that when he heard about the fish-eaters he instantly started his campaign...

There are the more constant features of Cambyses’ nature: his µ , his insanity and his lack of self-restraint, which structurally make the king exceptionally prone to immediately reacting to a situation on the basis of a mere impulse rather than rational planning. As a rule, military operations may go wrong, despite all planning, because of their inherently contingent nature. They will, in all likelihood, however, be unsuccessful if there is no planning at all. Cambyses’ lack of control made it all too likely that his operations went wrong, as, indeed, his expedition to Meroë did. Thus Herodotus, as we see, explains what was going on under Cambyses referring to the latter’s madness as a cause. Being disposed to madness in Herodotus means: being unable to overcome one’s own emotions and to plan one’s actions rationally. It is a deviation from a norm which allows for differences not only in kind but in degree as well. Cambyses grows even more mad than he already was after he had killed the Egyptian Apis’ calf, and his own brother and sister.16 15 16

Herodotus I 127. Herodotus III 30-38: But, as the Egyptians say, Cambyses, who had even before not been a man of rational control, went instantly mad because of this crime. The first of his outrages was the killing of Smerdis, although he had the same father and mother as Cambyses, whom he had sent back to Persia from Egypt out of envy... The second outrage was, as the Egyptians say, the killing of his sister, who had accompanied Cambyses into Egypt and lived with him there as his wife, although she was his sister having the same father and mother... Thus mad Cambyses behaved towards his own kinsfolk, either because of the Apis incident or because of anything else because of which it normally happens that people get into bad situations. There are rumours that Cambyses since his birth was stricken with a severe illness, which some call the “sacred illness”. Given that his body was affected by such a severe malady, it would not be uncommon if his mind were not in a healthy state either. In the same way he behaved madly towards the other Persians as well... And many similar outrages he committed against the Persians and the allies while he was still staying at Memphis; so he entered the old graves and looked at the dead bodies; similarly, he went into the sanctuary of Hephaestus and ridiculed the sacred images... He also went into the temple of the Cabyri into which no one is allowed to enter except the priests, played with the sacred images and

Dispositions in Greek Historiography

53

Dispositions can change: They can be altered, they can grow and diminish. Certain dispositions, however, develop unidirectionally. Cambyses is mad. He cannot control his emotions, and consequently, he breaks human and divine law. This makes two things unavoidable: Becoming ever more mad and not being successful. Cambyses’ madness in a way continuously worsens by being repeatedly applied to similar situations. Cambyses’ madness thus resembles what Aristotles calls a hexis: a disposition which can diminish and develop into the right direction, giving place to a virtue, or, as in this case, running into the extremes it may develop into outright vice. Other central dispositional categories, to which Herodotus appeals, are the notions of desire, purpose, and intention.17 A third phenomenon that he regularly mentions in this context are the µ , the customs and laws of groups or peoples.18 This notion lies somewhat between dispositions in the narrow sense of the word and prescriptive norms: It refers to positive laws, since what has been enacted as a law is called µ . Especially prominent in Herodotus with his special interest in ethnographic material is, however, the notion of µ in the plural, that is, µ , meaning customs: The customs of the Persians, for example (I 131). They play a role as principles in explanations, although these explanations are seldom made explicit. Customs are by their very nature dispositions to act in a specific way under specific circumstances, though described under a general perspective. What Herodotus descibes has always to be seen before the background of Greek realities: Neither do the Persians produce images of their gods, nor do they use temples or altars (I 131) (as the Greeks do), and they offer sacrifice to their gods in a peculiar manner (I 132) (different from that of the Greeks). Among the customs of the Babylonians, Herodotus describes the auctioneering of maidens, temple prostitution and burial customs (I 196-200). One Indian tribe is described as follows: Another tribe who lives eastward from these Indians are nomads, they live on raw meat and are called Padaeans. They are said to use the following customs: Whenever anyone from their village is ill, man or woman, the men who are most acquainted with the person kill him, and they say, this is, because, if he was dissolved by the illness, his flesh would be spoiled for them. The patient denies to be ill at all, but the attackers would not agree, killing him and eating his flesh in a ceremony. If a

17 18

burnt some of them... On the basis of multiple evidence it seems pretty obvious to me that Cambyses was completely mad. Otherwise he would not have undertaken to make a mock of sacred objects and well established customs. If one asked all mankind to choose the best customs out of the whole multitude of institutions, everyman would, after carefully comparing them, select his own ones. Thus convinced is everybody that his own customs are the best. It is therefore not to be expected that anybody but a madman would ridicule these institutions... Herodotus I 96: Deioces was the son of Phraortes. This Deioces was a lover of tyranny, and therefore he undertook the following actions. E.d.: Herodotus I 94,1: The Lydians observe the same laws as the Greeks, except for ... For Herodotus’ perspective as an ethnographer and scientific writer see Bichler 2001; Raaflaub 2002; Bichler 2004; Meißner 2005, esp. 90ff.; 107ff.

54

Burkhard Meißner

woman falls sick, similarly the women most near to her socially do to her the same as the men19. As in this case, Herodotus emphasizes quite often the contrast between a mere utilitarian attitude towards each other and the Greek mode of social life, health care and humanitarian standards. Projecting a lack of such ethical values onto peoples as remote as these Indians helps define the more clearly what lies at the heart of the Greek mode of life. That is, a set of dispositions which, contrary to what Herodotus alleges about the Indians, would not trigger killing the weak and ill ones, but would trigger caring for them, if need be. According to Herodotus, political institutions and the dispositions of people living with them mutually influence each other. The idea that “people’s laws are the consequences of people’s actions, as peoples actions are the consequences of people’s laws” is widespread in Greek literature. This theory of an intimate relation between institutions, usual occupations and habits lies at the heart of how Greek political theory evaluates constitutional systems in Plato, Aristotle and Polybius. Therefore, where habits are bad, the constitutions must be bad, and vice versa. We find this line of thought already in Herodotus in a fictitious debate about the virtues and flaws of different types of constitutions. The protagonists of this debate are members of a group of Persian nobles who had overturned a false pretender to the Persian throne. Eventually they made Dareius the legitimate ruler, but before deciding thus they would allegedly have debated whether or not to set up the rule of the many (democracy), of a minority (oligarchy) or of one monarch.20 Notions of dispositions play a key role in the arguments about the advantages and disadvantages of different political systems: The interlocutors repeatedly refer to dispositions, which the rulers develop necessarily under the condition of their ruling. Dispositions function as a causal link between the political balance of power, the behavior of the rulers and the legitimacy of their rule. Speaking in favor of B µ (democracy), one of the interlocutors quotes as a rule that monarchs, who can do what they want without being liable to anybody, will probably develop intentions outside the usual; C2 and , pride and envy, arise, and as a 19 20

Herodotus III 99. Thus, we already find in Herodotus the formal classification of political systems according to how many people share the power over them: everyone, a limited number of citizens or one single person. What we do not yet find, however, is the formal distinction, introduced by Aristotle, as to whether or not the rulers use their power in their own interest or for the benefit of those governed. Implicitly, however, we find this notion of legitimate and illegitimate rule, and the mechanisms of political deterioration, much discussed from Plato’s idea of a cyclic schedule of revolutions to Aristotle’s denial of its existence and Polybius’ restatement of the theory, can be seen in nuce in Herodotus already. Plato, Respublica VIII 543a-569c; Leges IV 5,712b-713a; Politicus 31,291c-292d; Aristotle, Ethica Nicomachea 1160a31-b23; Politica III 68,1278b6-1280a6; Polybius VI 5,1-9,11. Cf. Meißner 1986, 322ff.

Dispositions in Greek Historiography

55

result, violence and lawlessness. Absolute power causes a king’s behavior to deteriorate.21 Another interlocutor, being a partisan of oligarchic rule, refers to the fact that the common crowd’s understanding and knowledge is necessarily limited and to the hybris by which the masses subject everything to their jurisdiction.22 He, particularly, stresses their emotional and irresponsible rather than rational mode of decision making. To make a pro-monarchic point, the third interlocutor holds that the competitiveness of aristocratic elites leads to a disposition for civil strife and violence.23 All the arguments in this fictitious debate presuppose that on the political level there are various anthropological constants. Yet, there is also change. Change is explained, at least in part, by factors that are intrinsic to the political systems with their specific distribution of power. More concretely, it is explained in terms of the effects the distribution of power has on the behavior of those who speak and decide. The arguments rest upon the assumption that change on the political level is brought about by a mutual causal interrelation between the distribution of political power and the habits of those who are in power. The latter’s dispositions towards breaking the laws and enriching themselves is what is more or less generally strengthened by the political system, which in turn leads to the latter’s instability. Elsewhere, too, Herodotus employs dispositions like envy ( ) to explain, for example, the incapability of the Greeks to decide who was best at strategic counseling during the Xerxes war.24 Herodotus, as we have seen, combines situational conditions with dispositions to give explanations of peoples’ actions; frequently, he narrates and explains decisions to go to war using this pattern. For example, Dareius is said to have been angry with the Athenians after the Ionian revolt and even more so after the battle of Marathon25. Xerxes, however, in the beginning of his reign, was not at all disposed towards fighting the war against Greece, and he pursued the project of 21 22 23

24 25

A king behaves most inconsistently with himself (D7 µ 0 , III 80). Herodotus III 81: ... E?µ # ( ! ( % ( F2 ... Contrary to the single tyrant the crowd does not even know that it is acting when it is acting (Herodotus III 81). The crowd is uneducated and does neither know the standards (. #) ) nor what is usual and fitting ( B.G ): Does it not shuffle the affairs of the state, upon which it falls as a river does which is swallen by the winter rainfall? (Herodotus I 81) A precursor to Plato’s cyclic interpretation of revolutions is given by Dareius, allegedly the advocate of monarchy: In an oligarchy there arises competition between many people for distinction in performing public services, and consequently fierce antagonisms use to arise. Everybody wants to be the leader and to prevail with his projects; this leads to intense enmity among the people, from which civil strife arises, and from civil strife violent murder, and from violent murder monarchy. (Herodotus III 82). In a democracy, too, factions emerge and strife arises until again a monarch stands up to secure peace and freedom. Herodotus VIII 124; cf. 125. Herodotus VII 1: When news about the battle which had been fought at Marathon reached king Dareius, son of Hystaspes, the anger which he felt against the Athenians because of the attack on Sardeis, grew even more powerful, and he wanted even more to wage war against Greece.

56

Burkhard Meißner

collecting an army against Egypt.26 Xerxes’ µ # H , his pride, the strategic purpose to show his force and the aim to leave a memorial to posterity, are then presented as dispositional motives for his digging a canal through the Athos.27 A similarly striking example can be seen in how Herodotus describes Mardonius’ attack on Athens before the battle of Plataeae. Instead of bribing the Greek elites separately, as the Thebans had advised him to do, Mardonius wanted to be recognized as the satrap of the Great King’s European realm,28 although Herodotus does not use a proper term for this disposition. In the first comprehensive Greek prose work of history, we find quite a few examples of principles which are used in explanations, and which are at the same time dispositional properties of human beings. The explanatory status of these principles is, however, complex and difficult to pin down: On the one hand, they are constant factors, as opposed to the circumstances of situations, practical planning and reasoning, and they are different from the conditions of the physical environment. On the other hand, continuously having a disposition or showing the appropriate behavior as well as environmental factors may lead to a more virulent manifestation of the disposition or even a change in it (as was the case with Xerxes). Therefore, the determining principles underlying Herodotus’ usage of dispositional explanations remain somewhat ambiguous.

Dispositions in Thucydides Thucydides, who was more clearly interested in what is constant and what 0 as he calls it29, was consequently changes in human behavior, the more inclined than Herodotus to referring human actions to dispositions and similar notions. He explains, for example, the actions of the Athenian general Demosthenes who in 426 B.C. attacked the Aetolians out of a personal regard ) for the Messenians. Thus, character constants, peculiarities of the (! situations and utilitarian calculations explain actions of war in Thucydides.30

26 27

28

29 30

Herodotus VII 5. Herodotus VII 24: As I found out, Xerxes ordered to dig the canal because of his pride, that is: because he wanted to demonstrate his power and to leave a memorial for posterity. For, although he could have drawn his ships across the isthmus without much labour, he ordered a canal be dug for the sea with a width so that two triremes could pass each other with the oars actually driving them. Those whom he had made overseers of the dig he also ordered to build a bridge over the river Strymon. Herodotus IX 3: This was the advice of the Thebans. Mardonius, however, would not follow it; instead, he was possessed of a strong desire to take Athens a second time, partly because of his lack of judgement, partly probably in order to tell the Great King at Sardeis by fire-signals across the islands that he had taken Athens. Thucydides I 22,4. Thucydides III 94-95.

Dispositions in Greek Historiography

57

The Thebans had always wanted to take Thespiae, and when the occasion arose in 423 B.C. they did it.31 What is the role of dispositions in Thucydides? Describing the customs and especially the modes of fighting wars among the Greeks in the unspecified olden days, Thucydides claims that the islanders had then been more inclined to piracy than the people on the mainland, essentially because they were allegedly of Carian and Phoenician origin. Like Carians and Phoenicians, they are, therefore, supposedly more disposed to robbery32. Thucydides does not use an abstract noun for the disposition that his explanation presupposes. Notionally, this disposition consists of: being a robber in an especially intense way. As we can see in this example, there is an element of generalization in Thucydides’ explanatory usage of dispositions: There is the assumption of a continuity or essence of behavior. There is as we see a quantitative or qualitative aspect, and in most cases there is a moral element as well, because after all being a robber is not just like having the one or the other disposition, but one which is normally regarded illegitimate. Generalizations like the one we observe in this example are abundant in Thucydides; most commonly they can be found in his many speeches interpreting the issues and interests at stake. Describing the war council at Sparta in 431 B.C., Thucydides lets the Corinthians dwell at length upon what they think is peculiar with Athenian dispositions towards war, gain, change and progress, as opposed to how the Lacedaemonians usually behave: Confronted with such a city as an enemy, Lacedaemonians, you remain passive, and you seem to think that peace and stability are not guaranteed by those men who in their actions combine strength and preparedness with righteousness and who are determined to making it clear that they are not giving in once they are wronged, and not to harm others and defending yourselves from being harmed you regard as being of the same moral value. Hardly would you succeed with such an attitude, even if you lived in the neighbourhood of a city like your own. But now, as we have several times declared at length, your customs seem old-fashioned in comparison with theirs. It is, however, unavoidable that as in the arts and crafts so always the newer prevails over the older. In times of peace and stability observing the traditional customs and institutions is probably the best thing for a city; for cities that are under much pressure to proceed in various directions, however, it is necessary that they develop new ideas. Because of this the city of the Athenians on the basis of the manifold experiences which they have made has met with more innovation than yours. But let your slow-motion thinking come to an end now!33

31

32 33

Thucydides IV 133,1: In the same summer the Thebans destroyed the walls of the Thespians whom they charged with pro-Athenian leanings; they had always had this in mind, but it turned out to be an easy task right now, because the Thespian youth had been killed in the battle at Delium against the Athenians. Thucydides I 7,1N8,2. Thucydides I 71,1-4. About the innovative mood inherent in the alleged argument of the

58

Burkhard Meißner

Attitudes and dispositions are realized when they are applied to specific situations; they govern how people react to these situations; they can be changed; they may be inappropriate to a given situation: 2 H , intellectual slowness and resistance to change, is what the Corinthians call the Lacedaemonian disposition towards innovation in order to get away with it. In principle, dispositions are used not only to explain actions as instances of a more general rule governing people’s behavior. Rather, they are more commonly: used to explicate why actions in a specific situation are successful or unsuccessful. In a rhetorical context, arguments about the appropriateness of certain actions in a given situation can be derived from considerations about their success or failure, as Thucydides’ example shows. In 427 B.C., when the Athenians were debating about how to treat the Mytileneans after they had revolted against Athens, Cleon spoke in favor of a fierce reaction to protect the Athenian empire by inflicting terror upon those who tried to disrupt it. Thus Cleon, according to Thucydides, showed himself as the most brutal (2 ) of all the citizens and as the rhetorically most 34). In his speech, Cleon interpreted the influential one ( dispositions of the Athenians in dealing with their “allies:” Trusting each other mutually in daily life, they applied the same attitude towards their subjects rather than holding them down by force.35 Thucydides makes the Athenian rhetor employ a similar line of argument as the Corinthians had used to convince the Lacemaemonians about the necessity to go to war: What Cleon tries to show is the inappropriateness of the alleged disposition of the Athenians towards their allies in the Delian League; he charges his fellow Athenians with treating their subjects too friendly, given that they are really dominating the allies against their will. Eventually, however, the contrary opinion prevailed in the Athenian assembly: The Athenians despatched a trireme to put a halt to the ongoing execution of the Mytileneans.36 In a later chapter, concerning the same year 427 B.C., Thucydides describes what amounts to a pathology of the Greek world during the war. His diagnosis is led by the anthropological interest guiding his work, and consequently constants and dispositions figure prominently in his account of what went wrong in Greece during the last third of the 5th century. Thucydides stresses that the one great war between Sparta and Athens and the many petty civil wars, which were so characteristic of political life in the Greek city states, got increasingly interconnected, because each party in any of the Greek states attached itself to one of the great contenders.

34 35 36

Corinthians, cf. Mehl 1991, esp. 45; Meier 1995. Thucydides III 36,6. Thucydides III 37,1-3. Thucydides III 42-49.

Dispositions in Greek Historiography

59

And revolutions brought about many bad things for the cities in Greece; it has always been like this and always will be as long as human nature remains the same, but these were more or less detrimental and different in kind according to the peculiarities and circumstances of the situations... There were revolutions in the cities, and generally those, who continued making revolution let the thrust of revolution increase and by far exceeded those who preceded them by being extremely innovative in their measures and in how absurd the reproaches were which they eventually inflicted. The usual relations between words and things were reversed according to their ideologies. Daring without rational restraint was called courage and social skills, prudence was seen as cowardice, rational judgement as a pretext for weakness and broad knowledge for inactivity.37

What Thucydides describes here is a rapid increase of lawlessness and violence: A fundamental shift in how people were generally disposed toward each other, triggered by the circumstances. The basis of mutual faith was not divine law, but fellowship in crime.38 , I # µ ), from The efficient cause of all this was greed and ambition ( # which stems an extremely competitive mood. The political leaders in the cities used well-established values as terms for their behaviour, acting in the name of political equality or the higher wisdom of an aristocratic élite, and they made these established notions and values the prize in their fight.39

What is described here is the dissolution of law and humanity, which for Thucydides is exemplified for the first time in the war which was fought on the island of Corcyra. In Corcyra most of these crimes were committed for the first time... The Cercyraeans applied these passions first of all Greeks to each other... 40 Thucydides calls what happened an J G, an extreme emotion. What Thucydides describes in his account of the Greek pathology are underlying dispositions that develop and express themselves much more clearly than ever during the war. Similar to Aristotle, Thucydides looks at these dispositions as single entries in a complicated tableau of positive and negative dispositions, i.e.: as virtues and vices.41 Ideologically, the virtues were all too often used as mere pretexts for the vices, Thucydides argues, and in doing so he assumes a structural similarity between these vices and the virtues, which were used to camouflage them. This kind of behavior developed continuously in a process of innovation, addition and outdoing each other, which resembles what the Corinthians in their alleged speech presuppose to be the general condition of competition and development. Within the context of this line of thought, dispositions are understood as 37 38 39 40 41

Thucydides III 82,2-4. Thucydides III 82,6. Thucydides III 82,8. Thucydides III 84,1N85,1. Cf. Aristotle, Ethica Nicomachea 1105b19N1107a33.

60

Burkhard Meißner

being part of a complex system of mutual causality, in which actions, political systems, and general dispositions reciprocally explain each other. In the debate about whether or not to send an expedition to Sicily in 415 B.C., Nicias opposed the project as much as he could, being one of the selected commanders himself, while Alcibiades spoke in favor of the idea. The debate, far from being faithfully recorded, has been depicted by Thucydides as representing two opposing mentalities which prevailed in the Athenian public. Nicias, in his speech, opposed the young, inexperienced and ambitious Alcibiades and his followers, whom Nicias denounced as a kind of jeunesse dorée with a disposition to extravagant life ( #( # ), risking their own private fortunes and the well-being of the state.42 Thus, Thucydides uses generalizations about dispositions of this kind, much like Herodotus, not only to explain, but also to evaluate human behavior, and to clarify its underlying motives, its consequences, its concomitant factors and its moral quality in a dialectical way; usually by contrasting two opposing views, as in this case, in the shape of two contradictory speeches.

Dispositions in Xenophon In Xenophon’s Hellenica, dispositions are used mainly in assessing individual people’s ambitions and intentions. Alcibiades operated around Samos in 408407; he is said to have wanted to return to Athens with his soldiers 43). The fact that the (2 (# µ µ $ 0 $ # A K. Thessalian tyrant Lycophron of Pherae made war upon the city of Larisa in Thessaly in 405/404 B.C., is explained by his wish to rule over Thessaly as a L# 3 M # 44): To explain the protagonists’ whole (2 (# µ actions, Xenophon refers to relatively constant wishes or desires, which, given specific circumstances, occasions and some strategic, utilitarian calculation on part of the protagonists, lead to their actions. It is interesting, though, to look at how these comparatively constant dispositions change, if at all they change. There is one good example for such change: In 404 B.C., 30 aristocrats who later became known as tyrants established themselves as rulers of Athens, the most prominent of whom were Critias and Theramenes. Being initially both aristocrats with a ruthless, hands-on, anti-democratic attitude, they eventually found themselves on opposing sides when it came to putting their disposition into practice. As Xenophon describes it: 42 43 44

Thucydides, VI 12,2N13,1. Xenophon, Hellenica I 4,8. Xenophon, Hellenica II 3,4.

Dispositions in Greek Historiography

61

Initially Critias and Theramenes had much the same mind, and they were friends. When, however, the former was more and more up to killing many citizens, since he had been banished under the democracy, Theramenes intervened and expressed his opinion that putting people to death is not fitting if it is only because someone had held offices under the democracy without doing any harm to members of the aristocratic élite whatsoever, given that you and I both have been saying and doing things merely for the sake of being popular. The other who was still using the language of personal intimacy towards Theramenes, retorted that there is no alternative for those who want to get total control but to get rid of those who are able to hinder them.45

What Xenophon puts into the mouths of the two aristocrats is the same issue which Thucydides lets his protagonists debate at length on the occasion of the Sicilian expedition and during the revolt of the Mytileneans. As far as our topic is concerned, Xenophon explains the motives of the protagonists with their basic decisions (cf. 2 (# µ ), which they follow consequently under the impulse of other factors. Both aristocrats, the extreme and the more restricted one, share the same basic disposition: Keeping the populace under oligarchic control. However: While Critias is disposed to ruling ruthlessly and violently, Theramenes prefers to remain faithful to elementary values of law and humanity. Thus, dispositions are not only used to explain behavior, but also to debate the right line of action: What distinguishes Theramenes and Critias in Xenophon’s account is their different disposition as rulers with regard to violence. Juxtaposing two contradictory or complementary positions as Xenophon does in this case, aims at delineating a framework of practical possibilities: In expressing those two positions, the protagonists of Xenophon’s Greek Histories express the two opposing extremes in a practical dilemma. In this case, the dilemma consists of the principle of legitimacy (which requires the rulers to gain acceptance by observing well-established moral and legal norms) and the principle of total rule and control (which requires the ruler to get away with all kinds of opposition movements). What Xenophon underlines with this literary means of imagining a virtual debate is the practical necessity to make a choice between opposing alternatives. Presenting these alternatives in terms of dispositions (cf. G : being inclined to) and moral choice (cf.: B.N : fair), however, presupposes that there is some kind of choice between the two, making the tow opposing dispositions somehow open to choice, dispute and rational argument. Besides dispositions which the agents somehow have, there are also contingent, external factors present in Xenophon’s explanations of actions and human behavior (as indeed in the explanations of earlier historians): Dispositional explanations are a means not only to explain change and continuity in individuals or societies, but also to delineate to which extent 45

Xenophon, Hellenica II 3,15N16.

62

Burkhard Meißner

their actions are governed by rules and principles and how much influence contingencies have. Xenophon, for example, describes the situation in and around 369 B.C., when many Greek states revolted from Sparta, and Messenia was eventually freed from Spartan domination. In so doing, Xenophon emphasizes the extreme self-assertiveness of the Greek protagonists to point out the importance of the unforeseen intervention on part of the Persian satrap: When such a self-esteem was prevailing among all the Greek allies, Philiscus of Abydos came from Ariobarzanes with a huge amount of money. His first action was to call the Thebans with their allies and the Spartans to an assembly at Delphi to debate about peace. When they got together they would not consult the god about how to achieve peace, but rather negotiated by themselves. And when the Thebans did not give in as far as Messene’s independence from Sparta was concerned, Philiscus collected a large army of mercenaries to take the side of the Spartans46.

Eventually, a reinforcement army from Sicily came in to make up for this move,47 but what is important for us here, is the following: Besides the wishes and more constant intentions of the protagonists there are contingent factors which are caused by the complexity of their interactions and by external interventions. Among the more stable dispositions there are some that , decision. This line of thought is especially important Xenophon labels in Xenophon’s more fictitious works with an educational purpose, especially in the Cyropaedia.48 This is so, because within the educational context of the Cyropaedia the protreptic prospective is prevailing. Given the author’s educational aims, the decision aspect of Xenophon’s notion of disposition (they can be chosen, pursued and successfully developed) can be understood as the first intended stage in a multi stage-process: The literary work intends at making the reader decide in the right manner in matters practical, and by continuously doing so develop certain dispositions.

Polybius It is therefore probably no accident that we find the notion of dispositions particularly well developed when educational purposes govern the historiographic work. The description of Rome’s expansion by the second century Achaean historian Polybius, while explicitly aiming at a kind of 46 47 48

Xenophon, Hellenica VII 1,27. Xenophon, Hellenica VII 1,28. Xenophon, Cyropaedia VIII 3,8. If you let it be known that I have given you the choice, you will find me a different sort of servant the next time I shall serve you. When Pheraulas had distributed everything as he was instructed to do, he immediately started to arrange everything for the procession in order that it might be as well prepared as it might be in every respect.

Dispositions in Greek Historiography

63

practical and moral improvement of its readers ( 0 ), dwells at length , on dispositions in the Aristotelian sense, and on % , on consciously selected and developed habits. These qualities of peoples’ habits develop or emerge, according to Polybius, depending upon social influences, natural endowments, regular exercise, contingent factors and, above all: the political constitution.49 Repeatedly, Polybius uses his protagonists and their behavior as examples of what good or bad habits are and how they develop. For him, Philipp V. of Macedon is of special importance as such an example ( µ ): Being gifted and a good king at the beginning of his career, his manners turned worse when he took up licentiousness and a bad ; following this course consequently, he developed bad habits.50 I think, for all those who are interested in political affairs and want to derive practical improvement from history, Philipp is an extremely clear example.51 The example consists of the change in Philipp’s and the consequential development of bad habits. There are numerous similar stories in Polybius: Positive ones like Scipio and Philopoemen52 and negative ones like the Cynaethians in Arcadia, who, because of a lack of musical education, were eventually reduced to a merely animalistic life53. In all these examples, there are natural conditions ( H ), social structure and political constitution, contingent factors and decisions which, by processes of gradual emergence, produce dispositions that intensify. Dispositions may also change into other dispositions.

Summary In all contexts where dispositions and dispositional explanations are appealed to by ancient historians, there is a strong element of moral evaluation present. In Herodotus, dispositions not only explain how political systems change, but also decide about their legitimacy. Herodotus explains, but also evaluates the behavior of his protagonists, using notions of virtues and dispositions. The same holds true of Thucydides, who in most cases moves the evaluative aspect from his narrative into the fictitious speeches, which he puts into the mouths of his protagonists. In Xenophon and Polybius, the moralistic and didactic purpose, in addition to explaining historically what happened, is even more prominent than in the works of their predecessors. For that very reason, various notions of disposition play such a prominent role in their 49 50 51 52 53

Cf. Meißner 1986, esp. 332ff.; Eckstein 1995; Petzold 1999, 182f. Cf. Walbank 1938. Polybius VII 11,3. Polybius X 21f.; XXXI 25-30. Polybius IV 20f.

64

Burkhard Meißner

works, since dispositions are the primary object of their didactic effort. In all these authors, dispositions are formed in a complicated dialectical process within a context of repeated activities, social conditions, contingent factors, but also rational choice and conscious training. Although dispositions may be used to explain specific human actions, their role can be more generally be described as making behavior calculable (if not predictable) and practically manageable. If, what someone does, is, what he always does, knowing what he always does contributes to being not surprised. This kind of knowledge is probably not of the strictly nomological kind; it may be termed a preconception or even a prejudice. In the classical Greek historiography, statements about dispositions are results of generalizations, and they may prove wrong under specific circumstances. Under a pragmatic perspective, however, knowing one’s dispositions can be as useful as a nomological explanation in the proper sense of the word, and the pragmatic perspective is prominently present in most practical concepts which the ancient historians whom we have been looking at in this paper explicitly or implicitly hold. In nuce these historians convey notions of dispositions much like those which lie at the heart of the present volume on dispositions in the social and historical sciences.

Literature Bichler, R. 2004. “Herodotus’ Ethnography. Examples and Principles.” In: The World of Herodotus, ed. V. Karageorghis and I. Taifacos, 91-112. Nicosia. Bichler, R. 2001. Herodots Welt. Berlin. Eckstein, A.M. 1995. Moral Vision in the Histories of Polybius, Berkeley / Los Angeles / London. Fortenbaugh, W. 1975 (repr. 2003). Aristotle on Emotion. A Contribution to Philosophical Psychology, Rhetoric, Poetics, Politics and Ethics. London Geiger, J. 1985. Cornelius Nepos and Ancient Political Biography, Stuttgart / Wiesbaden. Grünbein, D. 2005. “Zwischen Antike und X.” In D. Grünbein, Antike Dispositionen. Aufsätze, 393-398. Frankfurt/M. Horsfall, N. 1989. Cornelius Nepos. Oxford. Johnson, D. and Newman, S. 1996. Student Dispositions, http://www.ncrel.org/sdrs/areas/issues/methods/instrctn/in5lk13-1.htm. Kaster, R.A. 2005. Emotion, Restraint, and Community in Ancient Rome. New York. Katz, ERIC Development Team. 1993. Dispositions as Educational Goals, http://www.eric.ed.gov/contentdelivery/servlet/ERICServlet?accno=ED363454. Konstan, D. 2006. The Emotions of the Ancient Greeks: Studies in Aristotle and Classical Literature. Toronto. Meier, C. 1995. “Ein antikes Äquivalent des Fortschrittsgedankens: das ‘Könnensbewusstsein’ des 5. Jahrhunderts v. Chr.” In Die Entstehung des Politischen bei den Griechen, C. Meier, 435-499. Frankfurt/M. Meissner, A. 2004. Lehrerbildung zwischen Reform und Revolution. Hohengehren.

Dispositions in Greek Historiography

65

Meissner, B. 1986. “WXYZ[Y\]^_ ]`\aX]Y. Polybios über den Zweck pragmatischer Geschichtsschreibung.” Saeculum 37: 313-351. Meissner, B. 2004. “Strategies in Herodotus.” In The World of Herodotus, ed. V. Karageorghis and I. Taifacos. Nicosia. Meissner, B. 2005. “Anfänge und frühe Entwicklungen der griechischen Historiographie.” In Die antike Historiographie und die Anfänge der christlichen Geschichtsschreibung, ed. E.-M. Becker, 83-109. Berlin / New York. Mehl, A. 1991. “Vorstellungen vom Erfinden im Athen des 5. Jh.s v.Chr.” In Das andere Wahrnehmen. Beiträge zur europäischen Geschichte, August Nitschke zum 65. Geburtstag gewidmet, ed. M. Kintzinger, W. Stürner, and J. Zahlten, 41-46. Köln/Weimar/Wien. Mumford, S. 1998 (repr. 2003). Dispositions. Oxford. Nussbaum, M. 1994. The Therapy of Desire: Theory and Practice in Hellenistic Ethics. Princeton. Petzold, K.E. 1999. “Rev. of Eckstein, Moral Vision.” In Geschichtsdenken und Geschichtsschreibung, Petzold, K.E., 182f. Stuttgart. Raaflaub, K.A. 1985. Die Entdeckung der Freiheit. Zur historischen Semantik und Gesellschaftsgeschichte eines politischen Grundbegriffes der Griechen. München. Raaflaub, K.A. 2002. “Philosophy, Science, Politics: Herodotus and the Intellectual Trends of his Time.” In Brill’s Companion to Herodotus, ed. E. J. Bakker, J. F. de Jong and H. van Wees, 149-186. Leiden/Boston/Köln. Raymont, P. 2001. “Are Mental Properties Causally Relevant?” Dialogue 40: 509-528. Raymont, P. Does Anything Break Because it is Fragile?, http://www.bu.edu/wcp/Papers/Meta/MetaRaym.htm. Richard, R. Of Dispositions, Attitudes, and Habits: Exploring how emotions shape our thinking, http://learnweb.harvard.edu/ALPS/thinking/docs/article1.html. The Symposium on Educator Dispositions Proceedings. 2003. Richmond, Kentucky, Carl Perkins Center, The College of Education, November 19, 20, 23 (2003), http://educatordispositions.org/moodle/ Tishman, S. and Andrade, A., Thinking Dispositions, http://learnweb.harvard.edu/alps/thinking/docs/Dispositions.htm. Tishman, S., Jay, E. and Perkins, D.N. Teaching Thinking Dispositions: From Transmission to Enculturation, http://learnweb.harvard.edu/ALPS/thinking/docs/article2.html. Tishman, S., Jay, E. and Perkins, D. N. 1992. Teaching Thinking Dispositions: From Transmission to Enculturation, http://learnweb.harvard.edu/ALPS/thinking/docs/article2.html, August 1, 1992. Walbank, F. W. 1938. “Philippos tragodoumenos.” JHS 58: 55-68; repr. in: Stiewe, K. and Holzberg, N., eds. 1982. Polybios, 15-29. Darmstadt. Vanderbeeken, R. and Weber, E. 2002. “Dispositional Explanations of Behavior.” Behavior and Philosophy 30: 43-49.

Further Reading References to literature are to be found in the footnotes to this articles; earlier literature can be found in: Lendle, O. 1992. Einführung in die griechische Geschichtsschreibung. Darmstadt. Meister, K. 1990. Die griechische Geschichtsschreibung. Stuttgart/Berlin/Köln. Herodotus:

66

Burkhard Meißner

Bakker, E. J., de Jong, J. F. and van Wees, H. eds. 2002. Brill’s Companion to Herodotus. Leiden/Boston/Köln. Bichler, R. and Rollinger, R. 2000. Herodot. Olms Studienbücher Antike III. Hildesheim (et al.). Boedeker, D. 2000. “Herodotus’ Genre(s).” In Matrices of Genre. Authors, Canons, and Society, ed. M. Depew, and D. Obbink, 97-114. Cambridge, Mass./London. Boter, G. 2002. “Over hybris, phthonos en aitie en in Herodotus.” Lampas. Tijdschrift voor nederlandse classici 35: 105-23. Harrison, T. 2003. “The cause of things: envy and the emotions in Herodotus’ Histories.” In Envy, Spite and Jealousy. The Rivalrous Emotions in Ancient Greece, Edinburgh Leventis Studies II, ed. D. Konstan and N. Keith Rutter, 143-164. Edinburgh. Luraghi, N. ed. 2001. The Historian’s Craft in the Age of Herodotus. Oxford. Vignolo Munson, R. 2001. “Ananke in Herodotus.” The Journal of Hellenic Studies 121: 30-49. Thomas, R. 2000. Herodotus in Context. Ethnography, Science and the Art of Persuasion. Cambridge. Thucydides: Classen, C. J. 2000. “Le virtù nelle Storie di Tucidide.” In Actas del X congreso español de estudios clásicos I: Sesiones de inauguración y clausura, Lingüística griega, Literatura griega, ed. E. Crespo and M. J. Barrios Castro, 351-55. Madrid: Sociedad española de estudios clásicos. Gaertner, M. 2000. Denken und Tun. Versuch ihrer Neubestimmung aufgrund moderner ästhetischer Sichtweisen. Diss. phil. München. Heitsch, E. 1996. Geschichte und Situationen bei Thukydides (Beiträge zur Altertumskunde LXXI). Stuttgart/Leipzig. Murari Pires, F. 1998. “The rhetoric of method (Thucydides 1.22 and 2.35).” The ancient history bulletin 12: 106-112. Raaflaub, K. A. 2001. “Father of All, Destroyer of All: War in Late fifth-Century Athenian Discourse and Ideology.” In War and Democracy. A Comparative Study of the Korean War and the Peloponnesian War, ed. D. McCann and B. S. Strauss, 307-56. Armonk, N.Y./London. Roberts, J. T. 2001. “Warfare, Democracy, and the Cult of Personality.” In War and Democracy. A Comparative Study of the Korean War and the Peloponnesian War, ed. D. McCann and B. S. Strauss, 238-269. Armonk, N.Y./London. Rood, T. C. B. 1998. Thucydides. Narrative and explanation. Oxford. Yunis, H. E. 2002. “Narrative, rhetoric, and ethical instruction in Thucydides.” In Papers on rhetoric IV (Università degli studi di Bologna. Dipartimento di filologia classica e medioevale. Papers on rhetoric VI), ed. L. Calboli Montefusco, 275-86. Roma. Xenophon: Dillery, J. 1995. Xenophon and the History of His Times. London/New York. Hogg, G. 1996. The self-education of Cyrus: a literary commentary of Book 1 of Xenophon’s Cyropaedia. Edinburgh. Howie, J. G. 1996. “The major aristeia in Homer and Xenophon.” In Roman poetry and prose, Greek poetry, etymology, historiography (ARCA. Classical and medieval texts, papers and monographs XXXIV, Papers of the Leeds international Latin seminar IX 1996), ed. F. Cairns and M. Heath, 197-217. Leeds. Humble, N. 1999. “Sophrosyne and the Spartans in Xenophon.” In Sparta. New perspectives, ed. S. Hodkinson and S. A. Powell, 339-53. London.

Dispositions in Greek Historiography

67

Musti, D. 2000. “Il tema dell’ autonomia nelle Elleniche di Senofonte.” Rivista di filologia e di istruzione classica 128: 170-81. Patzer, A. “Der Xenophontische Sokrates als Dialektiker.” In Der fragende Sokrates (Colloquia Raurica VI), ed. K. Pestalozzi, 50-76. Stuttgart/Leipzig. Tuplin, C. J. 1996. “Xenophon’s Cyropaedia: education and fiction.” In Education in Greek fiction (Nottingham classical literature studies IV, ed. A. H. Sommerstein, and C. Atherton, 65-162. Bari. Wilms, H. 1995. Techne und Paideia bei Xenophon und Isokrates (Beiträge zur Altertumskunde LXVIII), Stuttgart/Leipzig. Polybius: Beister, H. “Pragmatische Geschichtsschreibung und zeitliche Dimension.” In Rom und der Griechische Osten, Festschrift für Hatto H. Schmitt zum 65. Geburtstag dargebracht von Schülern, Freunden und Münchener Kollegen, ed. C. Schubert, K. Brodersen and U. Huttner, 329-49. Stuttgart. Champion, C. B. 1993. The indirect historian. The depiction of group character in Polybius’ “Histories”, 1-6 (Univ. Microfilms Internat.), Diss. Ann Arbor, Mich.: Princeton Univ. Derow, P. S. 1994. “Historical Explanation: Polybius and his Predecessors.” In Greek Historiography, ed. S. Hornblower, 73-90. Oxford. Eckstein, A. M. 1997. “Physis and nomos: Polybius, the Romans, and Cato the Elder.” In Hellenistic constructs. Essays in culture, history, and historiography (Hellenistic culture and society XXVI), ed. P. Cartledge, P. Garnsey and E. S.Gruen, 175-98. Berkeley/Los Angeles/London. Hahm, D. E. 1995. “Polybius’ applied political theory.” In Justice and generosity. Studies in Hellenistic social and political philosophy (Proceedings of the sixth symposium Hellenisticum), ed. A. Laks, M. Schofield, 7-47. Cambridge. Martinez Lacy, F. and Ricardo, J. 1991. “Polybius and his Concept of Culture.” Klio 73: 8392. Podes, S. 1990. “Handlungserklärung bei Polybios: Intellectualisme Historique? Ein Beitrag zur hellenistischen Historiographie.” Ancient Society 21: 215-40. Schepens, G. 1990. “Polemic and Methodology in Polybius’ Book XII.” In Purposes of History. Studies in Greek Historiography from the 4th to the 2nd Centuries B.C. (Proceedings of the International Colloquium Leuven, 24-26 May 1988), Studia Hellenistica XXX, ed. H. Verdin, G. Schepens and E. de Keyser, 39-61. Leuven. Vimercati, E. 2001. “Polibio come filosofo della storia.” Invigilata lucernis [Bari, Istituto di Latino] 23: 239-62.

II. The Debate about Dispositions from the Beginning of Modern Sciences to the 20th century

The Dispositions of Descartes1 PETER MACHAMER In Descartes’ natural philosophy he has a great need for dispositions, or as he sometimes calls them, tendencies or inclinations. This need arises for anyone who employs an equilibrium model in the science of motion. In a balance at rest, both weights are pressing downwards, yet there is no motion, no change. But the weights are active. They are forces ready to go into action when the situation changes (when the system enters a state of unstabilized dis-equilibrium.) For another example, consider an equation: when the expressions flanking both sides of the equality-sign (=) are complete, there exists a stability reflecting the equilibrium of all the components, now seen as relations among the variables composing each side of the equation. The variables may be conceived as being disposition terms. However, if one changes the value of one variable, then some of the others must be changed too, or a state of disequilibrium, an inequality, results. If the equation is descriptive of a physical situation, e.g., a simple machine or the path of some motion, then, as with the balance, the forces, powers and their dispositions, represented by the variables should be conceived as acting or being ready to act in order to bring about a state of equilibrium or restoring the equality. This essay will concentrate on this kind of physical disposition in Descartes, and try to lay out the way that Descartes discusses such matters, even though no force or activity belongs to matter per se. Before turning to the physical dispositions of matter, there is another form of disposition in Descartes’ work that needs to be mentioned, though I shall only outline how it works here. These are the dispositions of the mind to construct sensations when presented with a particular motion of the pineal gland that has been produced by seeing a material object. In the later Descartes (from 1643 onwards), these dispositions are present in the mind as innate general ideas that will be activated when the appropriate physical mo-

_____________ 1

Much of this essay is based on a larger work, by myself and J. E. McGuire, Descartes’s Changing Mind. This book is under contract to Princeton University Press. In it the positions attributed here to Descartes in outline are discussed in much more detail. An earlier, somewhat different version of some of this material appeared in an essay by myself, J. E. McGuire and Justin Sytsma, “Descartes on Causation in the World of Matter.” Philosophica 76 (2005) 11-44.

72

Peter Machamer

tions in the brain cause the gland to move in a specific pattern of motions. The pattern of motions is specific to the material object, which serves as a cause for all subsequent bodily motions (through the medium of transmission and in the perceiver’s body). It is this specificity of motion, being maintained throughout, that provides the content for a particular sensation, i.e., the sensation of this triangle. The activation of general innate ideas is how the mind conceptualizes particular inputs. Exactly how the mind is active and the forms of activity governing this conceptualization and conscious awareness Descartes leaves unclear. But what is clear is that it is not the active power of the will, which is under our control, which is involved. Certainly Descartes holds that the will is involved in a different mental tendency, in our tendency to make judgments about the existence of objects and their properties on the basis of these sensations. Descartes holds we have natural tendencies to believe that the objects we see have the properties that we experience, but that we need to learn to ignore these tendencies (see Principles of Philosophy I. 71 and 72.) The objects that we see need not have the properties that we are inclined to attribute to them. As Descartes says, in the Principles of Philosophy, II: 3: “…sensory perceptions are related exclusively to this combination of the human body and mind. They normally tell us of the benefit or harm that external bodies may do to this combination, and do not, except occasionally and accidentally, show us what external bodies are like in themselves.” (AT VIIIA 41-42; CSM 224). But my main point here is the role of innate ideas, functioning as the genus of a species and of a particular. These have to be active dispositions that are brought into play when the mind actively forms a general idea and also when it forms a particular sense impression on the basis of physical motions in the brain. Maybe there is a third major kind of disposition in Descartes. The inclination of the individual human ego to love God and understand Him as best a finite being is able. Through the training of the ego’s will, humans can learn to direct their understanding rightly and realize the limits of understanding by learning science. This is how we come to know, as much as we may, the manners in which God has harmonized His creation. This last I shall not directly discuss in this essay, but it is nonetheless a most important and often neglected theme (cf. Davenport 2006). But let us return to the first kind of disposition, that ascribed to bodies as related to motion. Descartes is well aware of the equilibrium conception of how to treat motion, though unlike Galileo before him and Euler after, he does not express himself in these terms. We can see how his concepts work by examining his late position on motion. This is a coherent position, if one accepts that Descartes is treating dispositions as epistemic concepts and not trying to reify them into fundamental ontology. Most commentators on Descartes have assumed that the late works, like the early work, are primarily an

The Dispositions of Descartes

73

attempt to describe the ontology of the world. Ted McGuire and I have argued at great length that this is a misreading of Descartes. We argue that Descartes’ later philosophy exhibits an epistemic stance towards what is in the world. But I cannot argue for this in detail in this paper. All I will do here is try to show how Descartes’ constructs a coherent position, even though an epistemic position, on physical dispositions. It is clear that motion, for the later Descartes of the Principles, consists in change of relational place, i.e., in the transference of a body away from those bodies surrounding it. It is the re-positioning of a part of matter, or individual body, relative to the bodies that immediately surround it. These surrounding bodies are considered by a perceiver to be at rest. As he says later: “I have stated that this transference is affected from the vicinity, not of any contiguous bodies, but only those which we consider to be at rest. For transference is reciprocal. And that exactly the same force and action is required for one transference as for the other” (II.29, italics mine; AT VIIIA 55-56; M 53). Descartes does not conceive motion as an action or force that causes transference. To attribute force to matter would be to ascribe active properties to matter as he had done earlier in The World, the Discourse, and the Meditations. The later Descartes says, as we saw above, motion, for us, refers in fact only to transference, not to force or action. So why does Descartes make a reference to “force and action”? His reference can only be to God’s action. God’s immutability or constancy is the genus of motion through which we have to regard the moving parts of matter as species.2 God has endowed us with a concept of action with which to think and talk about matter, even though matter per se has no active properties. This is how we humans may conceive of motion in order to do science. But Descartes needs a way to legitimate, in a non-ontological sense, the use of terms like “force”, “impulse”, “tendency”, or “inclination,” because this is how ordinary people think and speak about such matters. So he wishes to find uses for these terms in a way that departs not too far from the ways ordinary speech would explain how the material world works. In Principles II, Article 30, he writes: …if we wished to characterize motion strictly in terms of its own nature, without reference to anything else, then in the case of two contiguous bodies being transferred in opposite directions, and thus separated, we should say that there is just as much motion in the one body as in the other. But this would clash too much with our ordinary way of speaking. (AT VIIIA 56-57; M 54)

This social epistemic qualifier concerning what is considered to be in motion or at rest is important, because in Descartes’ plenum, in the actual world, everything is in motion, and nothing is at rest. So we cannot require bodies to move relative to things that are really at rest as a condition for thinking or

_____________ 2

Of course, mind can move matter, but this is not relevant to our concern here.

74

Peter Machamer

talking coherently about motion. We attribute motion to one body rather than to another, even though this is not really a correct attribution. According to our standard manner of speaking “movement…as customarily interpreted, is nothing other than the action by which some body travels from one place to another” (II.24, AT VIII-1 53; M 50). To be able to speak according to accepted usage is important to Descartes, for he uses this commitment, along with the principle of relational motion, in an attempt to avoid Galileo’s “error” about Copernicanism. In Principles III, Article 28, he argues: 28. That the Earth properly speaking, is not moved, nor are any of the Planets; although they are carried along by the heaven. And it is important to remember here what was said earlier concerning the nature of movement; i.e., that (if we are speaking properly and in accordance with the truth of the matter) it is only the transference of the body from the vicinity of those bodies which are immediately contiguous to it and considered to be at rest, into the vicinity of others. However, in common usage, all action by which any body travels from one place to another is also called movement; and in this sense of the term it can be said that the same thing is simultaneously moved and not moved, according to the way in which we diversely determine its location, From this it follows that no movement, in the strict sense, is found in the Earth or even in the other planets; because they are not transported from the parts of the heaven immediately contiguous to them. (AT VIIIA 90; M 94)

Saving himself from the Copernican heresy, for which Galileo was condemned in 1633, is a serious motivation for Descartes. I noted above that in Principles Part II, Article 29 Descartes introduces the notions of force and action in reference to motion understood as transference: “And … exactly the same force and action is required for one transference as for the other” (AT VIIIA 55-56; M 53). I suggested there that this new reference to force could only be to the action of God’s creative power. But Descartes, we also noted, makes extensive use of the vocabulary of force and power when speaking of causal relations among bodies. Moreover, in the passages we have been discussing, he invokes the notion of tendency to movement in contexts in which he employs the terms “force” and “power”. This needs to be understood in the light of Descartes’ theory of continual recreationism that he first presented in Meditation III, and again in the Principles I.21 (AT VIIIA 13; CSM I 200). This recreationism doctrine holds that according to the nature of time each moment of creation or conservation is independent of every other moment. Descartes writes in the Meditations: For a lifespan can be divided into countless parts, each completely independent of the others, so that it does not follow from the fact that I existed a little while ago that I must exist now, unless there is some cause which as it were creates me a fresh at each moment – that is preserves me. For it is quite clear to anyone who attentively considers the nature of time that the same power and action are needed to preserve anything

The Dispositions of Descartes

75

at each individual moment of its duration as would be required to create that thing anew if it were not yet in existence. (AT VII 49; CSM 33)

We must reconcile this theological doctrine with his late commitment to a view of material bodies as only extended. What I shall outline below is how this account of a human lifespan as requiring God’s constant creation from instant to instant to preserve the human’s existence and identity is parallel to God’s work in the instantaneous conserving the quantity of motion in the world in a manner that allows humans to make the world intelligible. This reconciliation is important, since all contributors to the debate concerning the nature of Cartesian body-body causation have to contend with the idea of instantaneous dispositions as found in Principles, Part II, Article 43. There Descartes explicitly considers “in what the force (vis) of each body to act on another (ad agendum in aliud) or to resist the action (actione) of that other consists: namely, in the single fact that each thing tends (tendat), so far as in itself, (quantum in se est), to remain in the same state, in accordance with our first law” (AT VIIIA 66; CSM 1 243). Without question Descartes links bodily force with the notion that each body possesses a tendency to remain in whatever state it is. Nevertheless, how the equation of ‘tendency’ with ‘force’ is to be understood is not at all straightforward since we may not think of forces being inherent in matter. This conundrum has led to many divergent interpretations. In what follows I argue for an interpretation of Descartes in the Principles that entails Descartes’ commitment to extended substance, force, tendency, and recreationism cannot be readily understood in terms of the standard doctrines of occasionalism, conservationism, or concurrentism. According to the Principles, God stands to the world in two compatible ways: (1) as the primary causal agency of everything that exists and happens; and (2) as the conserver of equilibrium conditions (constancy) across time distributed among created bodies. Descartes’ conception of creation seems to presuppose a doctrine of causal harmony that has a triadic structure. God creates bodies by willing (but recall, that in God willing-understanding-creating-conserving are identical) them to exist in the world. Yet at one and the same time bodies, as products of God’s creative power, are a causal source of the created mind’s ability to form ideas of them, but we know these bodies only to an extent that allows us to construct the science we need for preserving the mnd body union. This is the teleological note that Descartes first introduces at the end of Meditation VI. Yet, as noted, bodies are simply extended, quantifiable things such that given their passive nature their motion may only be viewed as a transference from the vicinity of one group to another. This tells us nothing about how they will behave specifically in relation to one another since, conceived as such, they are nothing more than ‘blank’ quantities constituted by such and such extensions. It is God that must fix and conserve the various sorts of

76

Peter Machamer

transferences, conceived by us as exchanges, among them by filling in the ‘blanks’ in terms of determinate sizes, speeds, and directions of motion. In other words, God plays an essential role as the conserver of determinate equilibrium conditions among bodies, and this provides the basis that allows us to calculate their exchanges with one another. Thus, in the Principles, Descartes elaborates his collision laws in terms of calculational rules for determining “how much motion of a given body is altered by collision with other bodies” (II.45, AT VIIIA 67; CSM I 244). It is by means of these calculatory concepts that we find the world to be intelligible. Now we need to consider directly how Descartes’ alliance of ‘tendency’ talk with ‘force’ talk (Principles, Part II, Article 43) can be interpreted. The first thing to appreciate is that Descartes’ laws are defined for a ‘tendency’ to move rather than for actual motion, actual transference. This view, and its consequences, can best be seen in terms of the second law of nature in the Principles, (third law in The World). Descartes tells us that “the cause (causa) of this rule, like the preceding one is the same, namely, the immutability and simplicity of the operation by which God conserves (conservat) motion in matter. For he always conserves the motion precisely as it is at the very moment in which he conserves it, and not according to how it might have been at some earlier time” (Italics mine, AT VIII-A 64; CSM I 242). Descartes illustrates this by returning to an example he had used in The World, that of a stone’s motion constrained in a sling and directed from a center of rotation. In The World the sling example is an illustration; but in the plenistically constrained world of Principles III it becomes the model for setting out his theory of whirling vortical globules through which he will explain falling bodies, planetary motion, and other phenomena. When the stone moves in a closed circle of motion in the sling, according to the second law, it is disposed to move tangentially. But owing to the external constraint of the sling, it is diverted continuously along a curved path. Descartes illustrates this by a subjunctive claim: If the stone were released from the sling, it would straightway move along a tangent to its circle of motion. This endeavor or tendency manifests itself only at the instant the stone is released; it is present, nevertheless, at each instant of its revolution, even though the sling impedes the stone from actually moving along the tangent. But this tendency falls under another description. As the stone turns in the sling, the string is drawn taut. Consequently, although the whirling stone tends to move in a straight line, the sling resists that tendency insofar as it restrains the stone. This interaction provides a basis for describing a second effect, namely that the stone can be viewed by us as receding radially from the center of rotation, the direction of which is altered at each successive instant. It is clear that these effects are considered in isolation from the stone’s complete and actual motion in a circle. This point is emphasized by Descartes in

The Dispositions of Descartes

77

Article 57 of Part III, where he argues that “there can be strivings (conatus) toward diverse movements in the same body at the same time” (AT VIIIA 108). Clearly, the only pattern to which Descartes appeals is that of the second law (i.e., that every part of matter always in fact tends to recede along a straight line). Nevertheless, he brings this tendency or endeavor under two different descriptions: an endeavor to move tangentially from the circle of revolution and an endeavor to recede radially from the center of revolution. It is this second description that Descartes uses to characterize the globules of the first and second elements when he claims in Part III that light consists solely in an endeavor to recede from the center of rotation of the sun’s vortex (Article 60, AT VIIIA 112). Just as the stone can be said to be constrained by the sling in its endeavor to recede from the center, so the effort of each of the globules to recede from the center of the vortex can be said to be restrained by those that are beyond it. In both cases, the endeavor to recede is balanced by a resistance. In sum, there are three factors involved in Descartes’ account of circular motion: the resistance due to an external constraint, and an internal ‘force of motion’ which can be described under two aspects. The tendency to recede radially from the center is opposed by an external constraint, so that their effects are balanced. But no actual motion occurs along the string, because there is no real tendency from the center. However, the tangential tendency is not so affected by the sling, as must be the case if Descartes’ second law is to hold. This way of thinking of phenomena (i.e., as manifesting diverse tendencies simultaneously) is consistent with by Descartes’ earlier analysis of the path of light through a refracting surface. He divides the surface into a perpendicular and a horizontal component. After the light strikes the surface, only the alteration of its perpendicular determination need be considered, because its determination parallel to the surface is in no way affected. Similarily, the sling acts like a refracting surface. It resists the stone’s radial tendency to recede, whereas the parallel component, its tangential tendency, is unimpeded. What is important to see is that Descartes’ tactic of invoking distinct but simultaneous tendencies works because each description ties back to the more basic tendencies bodies have to conserve their states. Thus, the only ‘real’ tendency is the tendency to move along the tangent. This tendency is always constrained in the world, yet we can also describe it as a tendency to move toward the center and away from the center. This allows Descartes to interchange these descriptors, and they well illustrate the perpectivalism inherent in his late epistemic stance, where he develops the idea that science is a human enterprise undertaken for the good health of human beings. This epistemic orientation differs decisively from his use of the sling example in The World.

78

Peter Machamer

What can we conclude from the second law and Descartes’ treatment of the nature of tendencies? First, they are always directed toward the conservation of the future states of a body just as in God’s recreating the human body from instant to instant. Descartes affirms, rightly, that movement can’t take place in an instant. Nevertheless “every moving body, at any given moment in the course of its movement, is inclined to continue that movement in some direction in a straight line, and never in a curved one” (AT VIIIA 64; CSM I 242). In other words, at each moment in which it is moving, a body may be said to be endowed with a tendency to preserve whichever of its future states results from intelligible exchanges with other bodies. Thus, there is a strong and intrinsic connection between a body’s tendency to move in a straight line and the conservation of whatever state it will turn out to be in. Now, although each body’s tendency is conserved by God in each independent part of time in accordance with the doctrine of recreationism, nevertheless, we must ascribe a tendency to each body independently of all other bodies. In other words, these bodily tendencies explain how bodies conserve their states at successive moments of time according to God’s harmonious plan. God conserves the existence of material substances by recreating them in each independent moment of time, and by the same action he recreates their natures independently of each prior moment of their existence. Such independent recreation requires that we humans make use of tendency and equilibrium talk in order to make sense of the motion and interaction of bodies. So in summary, dispositions for Descartes are really only ways in which we humans have to think and talk so that we may make sense of the world as best we can. God has given this ability to think coherently and make sense of the world as best we may, if we learn to avoid error (Principles I, 35-38).

Literature The following abbreviations are used in the text: AT = Charles Adam and Paul Tannery, eds. 1964. Oeuvres de Descartes. Paris: J Vrin. CSM = John Cottingham, Robert Stoothof and Dugald Murdoch. 1985. The Philosophical Writings of Descartes, Volumes I and II. Cambridge University Press. CSMK = John Cottingham, Robert Stoothof, Dugald Murdoch, and Anthony Kenny. 1985. The Philosophical Writings of Descartes, Volume III. Cambridge University Press. M = Rene Descartes. 1983. Principles of Philosophy, translated by Valentine Rodger Miller and Reese P. Miller. Dordrecht, Holland: D. Reidel. O = Paul J. Olscamp. ed. 2001. Rene Descartes, Discourse on Method, Optics, Geometry, and Meteorology. Hackett. Davenport, Anne Ashley. 2006. Descartes’ Theory of Action. Brill. Machamer, Peter, J. E. McGuire, and Justin Sytsma. 2005. “Descartes on Causation in the World of Matter.” Philosophica 76: 11-44.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics URSULA RENZ

0. Introduction Spinoza is often considered to be the Megarian among the early modern philosophers: The ontological arsenal of his metaphysics, so it is widely believed, is reduced to one singular entity, whose being is mere and eternal actuality, whereas singular things, timely events, and dispositional properties are regarded as being merely illusory.1 This view, it is further assumed, undermines our common sense view of mental dispositions. We usually think of ourselves as entities that are endowed with certain non-actualized mental properties like, for example, the ability to find the solution to mathematical problems or the inclination to get angry when someone disturbs us at work. And we often conceive of these traits as full-blown properties, which are just as real as our actual properties such as the property of my being female. But according to the standard conception of Spinoza’s so called necessitarianism this cannot be the case. For necessitarianism unlike determinism not only rejects the idea of free will, but precludes the idea of individual subjects who are endowed with non-actualized mental capacities. In this paper, I would like to challenge this view. Spinoza’s metaphysics of modality is more moderate, and less absurd, than the Megarian picture that many textbooks depict. As I see it, Spinoza’s intention is not to deny the reality of particular things nor their dispositional properties, but to show that they are conditioned entities that are completely explicable in terms of their properties and of the modifications they have undergone Understood in this manner, his approach is a suitable example of what I would call ‘the third way’ in the philosophy of dispositions. He neither endorses an empiricist or semantic reductionism, nor is he a metaphysical realist, who considers dispo-

_____________ 1

This view was put forward in the Enlightenment by early modern philosophers who maintained in some respects similar views, but wanted to disassociate themselves from the Dutch heretics, that is, in particular Nicolas Malebranche and Gottfried Wilhelm Leibniz. For the Aristotelian Critique of the Megarian view of dispositional properties as well as for a defense of our common sense view, cf. Jansen in this volume.

80

Ursula Renz

sitions as fundamental forces or irreducible properties of things.2 In contrast to these well known accounts, Spinoza’s approach is more appropriately characterized as a combination of a version of radical rationalism and some kind of common sense realism. He assumes that every thing, event, or state can be explained in terms of its actual properties and its actual causal relations. Yet this is not equivalent to saying that dispositional properties do not really exist or, that it is pointless to refer to them. For Spinoza, dispositional properties have an important explanatory function in our account of certain phenomena, even though they can be further analyzed. This approach to dispositional properties is of particular interest in regard to mental capacities. Spinoza’s philosophy of mind is best understood as following from his rejection of the conceptual framework underlying the Cartesian conception of mind. As explicitly stated in the correspondence with Hobbes on the Meditationes, Descartes maintains that the existence of the human mind as a substance cannot be directly known, but it can be only indirectly inferred, that is, only insofar as it is conceived of as the ontological substratum of our mental acts, faculties and properties.3 The assumption that the mind is endowed with specific faculties in turn suggests that there is some part of human mental behavior that cannot be completely explained in terms of its actual mental features. Spinoza in contrast conceives of the human mind in a manner that denies the admission of such an unintelligible rest. Neither is the mind itself a substratum behind or bearer of mental acts and properties, nor are there any specific faculties it is endowed with. Unlike Descartes, whose rationalist claims in regard to physical objects are quite radical, while being quite restricted in his view of the human mind, Spinoza assumes that both, mental and bodily entities, are completely intelligible. In the first section of this paper, I will expound this approach in more detail. I will give a short sketch of Spinoza’s theory of modality and explicate the explanatory function of dispositional concepts in the Ethics. I will show that they are useful only in specific contexts; when we want to account for the existence and the actions of those things that are not determined by their own essence to be or to act in a certain way. I maintain that they explain why those things, which are not determined by their essence to do or not to do certain things, behave the way they do. Finally, I will show how Spinoza’s so called necessitarianism reduces to his rationalist conviction that all there is, including contingent particulars and possibilities, can, in principle, be completely conceived of in terms of necessary causal connections (section 1). The other three sections of the paper are concerned with the problem of mental dispositions in the Ethics. It is a remarkable feature of Spinoza’s philoso-

_____________ 2 3

Cf. also Schrenk and Mumford in this volume. Cf. AT VII, 176.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

81

phy of mind that he rejects the idea that the mind is essentially endowed with mental faculties that are ontologically distinct from its actions. This is a fundamental challenge to our ordinary concept of the human mind. What is the human mind, so one might ask, if one cannot understand it as some kind of substance endowed with certain faculties prompting a certain type of mental output? Before addressing the question of the ascription of particular mental capacities, I would, therefore, like to discuss the implications of Spinoza’s concept of the human mind (section 2). The third section is concerned with Spinoza’s concept of idea. Due to the historical influence empiricist approaches had in contemporary philosophy of mind, we often tend to think of ideas in terms of episodic representational mental states. This, however, does not correspond to the way Spinoza makes use of this term. In the Ethics, ideas certainly are considered to be the basic units of the mental. Unlike Lockean ideas, however, these basic units are not merely individuated by their representational content, but also by their inferential relations to other mental states or ideas. This conception has enormous implications: It suggests that in the Ethics the term ‘idea’ refers at least sometimes to a disposition to think of particular things in a certain way and not to our actual mental state. Therefore, given this understanding of ‘idea’, we do indeed have mental dispositions according to Spinoza. They are nothing unreal or illusory. They do however not consist in some ultimate and undetermined faculty such as a free will, but in the ideas connected by association or inference with other ideas we actually have (section 3). In the fourth section, I will discuss the passage of the Ethics where Spinoza most explicitly makes use of a dispositional concept, namely the Scholium of 2p13: He says that minds are to the same extend capable, ‘aptus’, of perceiving many things at once as the corresponding bodies are capable of doing many things at once. I will argue that Spinoza thereby establishes a conceptual device that allows him to make sense of central intuitions that we usually rely on in trying to understand the experiences of other persons or other organisms. Once again, we encounter the combination that I have described above as radical rationalism mixed with common sense realism (section 4).

1. A Type of Transcendental Philosophy: Interpreting Spinoza’s Theory of Modality The intention to understand a phenomenon is often driven by the following expectation: We assume that the object in question (a particular event, the quality of an experience, or the disposition to act in a certain way) can in principle be explained in terms of its properties and of the modifications to

82

Ursula Renz

which it has been subjected and which, taken together, completely describe it. If we had a complete description of an object, we could, therefore, derive necessarily true statements about its properties including existence or occurrence at a given moment. If we lack a complete description however, we are required to look for further explanations. It is the intuition that our demand for knowledge is fully satisfied only by a complete explanatory description of the object that motivates Spinoza’s metaphysics and, in particular, his theory of modality. He thereby endorses a very strong notion of what it is to fully know or understand something. We fully know a thing, if and only if we grasp what necessitated it. In contemporary analytic Spinoza scholarship, this quest for explanatory completeness is often identified with a metaphysical necessitarianism. Spinoza is claimed to establish a modal metaphysics that denies the reality of any possibilia. Don Garrett for instance ascribes the view to Spinoza that “every actual state of affairs is logically or metaphysically necessary, so that the world could not have been in any way different than it is.”4 Such a necessitarian view of modality contradicts the common sense belief that things can have dispositions to act in a certain manner. If all things in the universe are metaphysically necessary, conceptually it does not seem to make sense to assume that dispositional properties exist.5 I think however that this necessitarian reconstruction of Spinoza’s metaphysics is rather shortsighted. It is true that in many places Spinoza denies the existence of mere possible things. So he says, for instance, in 1p29: In nature there is nothing contingent, but all things have been determined from the necessity of the divine nature to exist and produce an effect in a certain way.6

And in 1p33, he asserts: Things could have been produced by God in no other way, and in no other order than they have been produced.7

_____________ 4 5

6 7

Garrett 1991, 191f. At this point, I would like to address a distinction made by Ludger Jansen in his discussion of Aristotle’s theory of dispositions (in this volume). According to Aristotle, dispositions are not merely possibilities that one had to spell out in terms of sentence operators, but some kind of predicate qualifiers. I think this is a convincing and fair description of our common sense intuitions. In a necessitarian approach, however, this distinction is challenged. The claim that every thing can be completely described implies that every property, which a thing has, can be explicated in terms of necessary and, hence, unqualified predicates. The sentence that Ludger Jansen has the non-actualized ability to speak Chinese is therefore not a wrong, but an incomplete description of the actual state of Ludger Jansen’s mind. 1p29, cf. for the citation: Spinoza, Collected Works, ed. Curley, 433. In the following all English citations are taken without change from Curley’s translation. 1p33, Collected Works, 436.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

83

According to these two propositions, Spinoza seems to put forward a radical necessitarianism that precludes the existence of unrealized entities and, hence, also dispositional properties. This interpretation is further supported by Spinoza’s theological position motivating the above claims. There is considerable evidence that he developed his views on modality in full awareness of its theological implications. As his letter to Oldenburg from February 1676 reveals, he was willing to affirm all the dangerous claims, which follow from the necessity of God and which Oldenburg advised him to withdraw.8 He knew that the assumption of divine necessity ruled out the traditional notion of God’s personhood as well as of divine teleology, and he, of course, also knew that identifying God’s necessity with the necessity of being undermines the idea of free will.9 As far as theology is concerned, he obviously was willing to embrace necessitarianism. The picture, however, changes when it comes to the explanation of the existence and actions of particulars.10 Unlike the divine substance, whose existence and properties follow from its essence, particular things are determined to exist and to act by other particulars. In 1p28 Spinoza claims: Every singular thing, or any thing which is finite and has a determinate existence, can neither exist nor be determined to produce an effect [operari] by another cause, which is also finite and has a determinate existence; and again, this cause also can neither exist nor be determined to produce an effect unless it is determined to exist and to produce an effect by another, which also is finite and has a determinate existence, and so on, to infinity.11

He obviously thinks that particulars, too, are subject to necessary determination, otherwise he could not rule out unconditioned existence or action. In contrast to the above cited claims however, the necessity involved in this case is not defined in terms of God’s necessary creation, but in terms of some necessary antecedent and the infinite causal chain of particular things. Moreover, it has to be emphasized that he does not deny that, given different antecedents, things could have turned out differently. This suggests that Spinoza’s claims on modality are motivated by the epistemological intuitions sketched above, rather than by some metaphysical ideas concerning the reality or non-reality of things. Spinoza’s necessitarianism, as I see it, is not a type of descriptive metaphysics delineating which things exist and which do not. Instead it is better understood as a version of transcendental argument that analyzes the necessary presuppositions of a radical rational-

_____________ 8 9 10 11

Cf. Letter 78. In: Spinoza Opera, vol. 4 (=G IV), 326f. For the theological background cf. Carriero 1991 and Perler 2006. It has to be mentioned that Spinoza does not regard God as a particular thing, as particulars are merely affections or modifications of God’s attribute. Cf. also 1p25c, Collected Works, 431. 1p28, Collected Works, 432.

84

Ursula Renz

ism. I shall not discuss here whether or not this weakens Spinoza’s necessitarianism.12 But I think we can, without going into further discussion, ascribe to Spinoza the following claim: Even if necessitarianism is true, it does not help us in explaining either the existence or the action of any particulars. If, however, we accept this claim, we cannot directly appeal to metaphysical necessity in order to explain the behaviour of a particular thing. Instead, we have to seek a causal analysis of its concrete determination. Considered in this way, Spinoza does not maintain that we can deduce a complete description of things from the conceptual claims grounding his metaphysics. In this context, a comparison of Spinoza with Leibniz might be helpful: Leibniz, in some tension with (if not to say in contradiction to) the metaphysical premises of his concept of complete notions,13 rejects necessitarianism for theological reasons. Just as in Leibniz, Spinoza’s views on God’s freedom and on the particulars belong to different and separate domains of philosophical concerns, i.e. theology on the one hand, ontology on the other. In contrast to Leibniz, however, Spinoza combines a necessitarian picture of God – ruling out traditional theology – with a determinist view about the ontology of particulars that makes sure that complete explicability holds. In light of our discussion of Spinoza’s modal metaphysics so far, the following claims seem to hold true in the Spinozistic framework in regard to dispositional properties: (1) Contrary to what one might assume at first glance, there is conceptual space for dispositional properties in Spinoza’s modal metaphysics. The ascription of dispositional predicates is not pointless, although they have a restricted scope. It makes no sense to ascribe dispositional properties to God (including the capacity to create or not to create certain particular things), but we can reasonably argue about dispositions of particular things. (2) Due to the necessity of causal determination, and provided that causal influences are always actual, the ascription of a dispositional concepts has however only some kind of provisional epistemic legitimacy. Dispositions can in principle be explained in terms of the actual properties of its bearer, or in terms of things affecting its bearer. That someone is inclined to do x rather than y has thus nothing to do with an original power, but with her being the focal point of several (internal and external, direct and indirect) causal relations, which together amount to a certain disposition to do x. As humans we

_____________ 12

13

In order to answer this question, we would have to determine how the two necessities involved, the divine necessity and, the case of necessary causal determination of particulars, are conceptually related. This cannot be done without at the same time defending a particular interpretation of other issues in Spinoza’s metaphysics. For further discussion cf. Schütt 1985, Garrett 1991, and Curley/Walski 1991. Complete notions include not only the hypothetical conditions, but also the knowledge when they are given. I would like to thank Robert Schnepf for a very helpful discussion on this issue.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

85

can, however, never eliminate all of our dispositional concepts, nor can we reasonably deny the reality of the dispositional properties denominated by them. Certainly, if we could account for the whole sum of all actual causal influences on an object, we could dismiss dispositional concepts. But as human beings, we (almost) never arrive at this point. Meanwhile our common sense language as well as many provisional scientific explanations make use of dispositional concepts and reasonably assume that they refer to some real causal interaction between real things. Dispositional concepts are thus best characterized as some kind of explicable explainers. They help us to conceive of the properties of things in a provisional common-sense-like manner, when we lack a complete causal analysis. (3) This points to another, rather meta-theoretic consequence of Spinoza’s modal metaphysics. The restricted scope and the provisional epistemic legitimacy of dispositional concepts lead to an important shift in conceiving of the main focus of a philosophical investigation of dispositional terms. What philosophy primarily has to discuss is not whether or not dispositional properties really exist. The interesting question is rather how we should conceive of them in order to account for the explanatory function they perform in science. Philosophy has to provide the conceptual framework that facilitates the rational use of dispositional terms while it insists on further explanation of the dispositional properties that they denote. These three points have an enormous impact on Spinoza’s philosophy of mind. Being a particular, the human mind belongs to those entities to which dispositional properties can be ascribed. The question arises: What kind of causal influence lies behind our mental dispositions? Another important issue is how certain specific mental capacities are to be conceived of in order to be subject to further analysis. Before addressing these two problems in detail, some general assumptions underlying Spinoza’s concept of the human mind need to be exposed.

2. Re-categorizing Human Thought: Spinoza’s Deflationary Concept of Mind Spinoza’s general approach to the mental is probably best characterized by a close analysis of his first definition of the human mind.14 It says:

_____________ 14

The Ethics provides two definitions, one in 2p11 and another in 2p13 (both cited in the main text). Even though the second one appears to be merely a more precise version of the first, they address slightly different problems. The first determines what minds consist in, the second answers the question of what individuates particular minds. This difference has always been neglected in Spinoza scholarship. For a detailed discussion cf. also Renz 2006.

86

Ursula Renz

The first thing that constitutes the actual being of a human Mind is nothing but the idea of a singular thing which actually exists.15

At first glance, this definition sounds rather odd. Usually, we would expect the human mind to have ideas, and not to consist of or to be an idea. A closer look at the following passages, however, leaves no doubt, that Spinoza really wanted to say that the mind is an idea. Since already on the next page, he gives a second definition of the human mind where he states more precisely: The object of the idea constituting the human Mind is the Body, or a certain mode of Extension which actually exists, and nothing else.16

Why does Spinoza claim that the mind is an idea, instead of merely pointing to the fact that he has ideas? What motivates this move? And what insights does it provide? In order to answer these questions, we have to pay attention to the following two points: (1) The quoted passage relies on an ontological assumption about humanity in general. The essence of man is to be conceived of in terms of modes, and not in terms of substance or of substantial form, as has been common in the philosophical tradition before Spinoza.17 Accordingly, neither the existence, nor the actions, or passions of a person can be understood with the help of the concept of his essence. If we want to explain one of these properties, we have to refer to the causal interactions of a person with other particular things. These insights also apply, of course, to the human mind. Our mental life has also to be accounted for in terms of an analysis of the causal interactions between several mental states, or in Spinoza’s terminology: ideas; and not by reference to the essence of the human mind. When we want to explain why someone has a certain idea, we have to examine how this idea is caused by another mental state. I will address the question of what it means that an idea is caused by another idea later on. Here it is only to be emphasized that, by conceiving of the mind as a particular idea, Spinoza satisfies first of all the demand underlying the ontological assumption that man consists of modes. If all explanations of mental states is based on causal interactions among different ideas, and if the mind has to be understood as a cause of our mental states, then the mind itself must be conceived of as an idea. (2) The claim that the mind is a particular idea undermines one of the implicit tenets of our ordinary concept of the human mind, namely the assumption that it is the bearer of mental states or properties, which in turn presupposes a categorical distinction between the ontological status of the

_____________ 15 16 17

2p11, Collected Works, 456. 2p13, Collected Works, 457. Cf. 2p10 and 2p10c, Collected Works, 454. As has been emphasized by Gueroult 1974, 111, this denial is directed at the same time against Aristotelianism as well as against Cartesianism.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

87

mind and the mental. By defining the mind as an idea, Spinoza rejects the notion of the human mind as some kind of bearer. One can assume that this is an essential part of his program. Spinoza wants to get rid of one of the basic conceptual constraints grounding many approaches in the philosophy of mind. By maintaining that the human mind is nothing but a particular idea, he flatly refuses to consider minds as some kind of unintelligible substrata of mental states. Instead, the mind is conceived of as belonging to the same ontological category as our other ideas. It thus cannot be a principle beside, behind, or beyond our actual mental life. And this, in turn, implies a radical denial of any essential mental faculty.18 Although puzzling at first glance, Spinoza’s claim that the human mind is or consists of a particular idea makes quite a lot of sense. It presents a fundamental critique of any concept of the human mind that presupposes a categorical distinction between the mind and the mental. In a way, his critique resembles the one put forward by Gilbert Ryle in The Concept of Mind against the Cartesian model of the mental.19 Like Ryle, Spinoza attacks the conceptual framework of Cartesianism and seeks to undermine the widespread belief that the human mind is something behind the mental. But unlike Ryle’s, Spinoza’s approach is not based on an analysis of language, nor is it driven by an empiricist conception of causality according to which the causal relations are nothing but the regular association between events or properties. Instead, his arguments rely on the rationalist assumption that the mental life of a person, though we often lack knowledge of it, could in principle be exhaustingly analyzed in terms of the ideas he has. Given this rather deflationary approach to the human mind, one might ask how Spinoza can account for our phenomenological intuition that we ourselves are epistemic subjects, who have ideas or mental states? I will answer this question in the next section. It is, however, worth to emphasize the following two points: (1) A closer analysis of the argument by which Spinoza demonstrates that the human mind consists of a particular idea shows that he does not preclude the intuition that we can have ideas or mental states, nor does he deny that some ideas can be prior to others. On the contrary, he claims that the idea which is the human mind must in some sense be prior to the mental states we have.20 Spinoza does not want to deny all the distinctions we usually make when reflecting on our mental states. His goal is rather to suggest a conceptual framework that does not arbitrarily end the theoretical analysis of

_____________ 18 19 20

Cf. also Bartuschat 1992, 96 and 101, and Yoshida 2004, 63f. Cf. Ryle 1949, 16ff. I will not discuss here whether or not Ryle’s exposition of the Cartesian myth provides a fair analysis of Descartes’ own approach. Cf. the first of the two references to 2ax3 in 2p11dem; Collected Works, 456.

88

Ursula Renz

the mental at a certain point. Thus, Spinoza does not think that anything is fundamentally wrong with our phenomenological intuitions, but he does question whether sceptical conclusions should be drawn from them. There is no unintelligible rest in conscious experience according to Spinoza. It is for this reason that he puts forward a concept of the human mind that is completely different from the one maintained either by the Aristotelian or Cartesian approaches to the mind. Instead of considering the mind as a substance, or as a part of a substance that has the faculty to have or to produce ideas, he suggests that we should conceive of it in terms of the same ontological category as the mental, that is, in terms of a mode of thought or of a particular idea. (2) The claim that the human mind is a particular idea only gives a provisional definition of the human mind, as it does not yet provide a criterion to distinguish it from other ideas, nor does it show how single minds are individuated. As the second of the two claims quoted above suggests (2p13), Spinoza considers only those ideas that are about ourselves, insofar as we are a certain body, to be human minds. Consequently, the idea that constitutes the human mind involves necessarily some kind of basic self-awareness.21 Spinoza asserts that thought and extension are generally conceived of as distinct features of one and the same thing, and that mental and bodily items of particular things are only conceptually distinct, and not ontologically different things.22 We can assume that the idea, which constitutes our mind, is about some aspect of ourselves, that is, of ourselves insofar as we are bodily things. The mind is thus distinguished from other ideas by the fact that it consists in some kind of self-awareness. One can, of course, speculate whether or not stones have such knowledge according to Spinoza.23 Yet it is at least obvious that our ideas of stones are not items of self-awareness, and that they are not minds. It thus follows that the concept of the human mind suggested in the Ethics, though deflationary in its general approach, does not reject the intuition that we, insofar as we are thinking things, can have ideas. Spinoza does not deny the assumption that humans have the mental capacity to have ideas, he only suggests a different theoretical framework for thinking about it. Up to now, I have not discussed precisely what ideas are. They are sometimes described as ‘mental states’, and I have used this notion occasionally in this section. But, in some way, this expression is problematic. Not only is the term ‘mental state’ often used as a dummy term, which does not explicate

_____________ 21 22 23

I argued for this in more depth in Renz 2006 and Renz 2009. Cf. for these claims 2p7s, where Spinoza’s so called identity theory is defended. Collected Works, 451f. Spinoza himself does so in a hypothetic manner in a letter to Schuller. However, he never does assert that stones have a mind or consciousness. Cf. also G IV, 266.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

89

what it means to be in a mental state, it also tacitly presupposes that mental states are primarily occurring entities. Spinoza however does not understand the mental in this manner as a closer analysis of some of his psychological claims will show: Not only are our actual and conscious thoughts conceived of as ideas, but also our unconscious mental states, that are connected by associative or inferential relations to our conscious ideas. In the following section therefore, I would like to discuss in more detail how Spinoza understands the concept of ideas and how he uses it to address the problem of mental capacities.

3. Explaining Dispositions in Terms of Content: Spinoza’s Concept of Idea and Mental Causation In order to understand Spinoza’s notion of idea, one has to draw attention to the following three points: (1) Ideas are mental and not bodily items. Spinoza emphasizes this in his definition of the term ‘idea,’ where he maintains that ideas are concepts “of the Mind that the Mind forms because it is a thinking thing.”24 This shows that, in some respects, his position is closer to Descartes’ substance dualism than one might think. Although he does assume that mind and body express one and the same thing, and hence belong to one and the same reality, he does not ascribe ideas to bodies or reduce them to bodily states. On the contrary, although the mind and the body of a person also refer to one and the same thing, they are conceptually distinct, and to neglect their conceptual difference is no less problematic than to ignore their ontological identity.25 (2) Ideas always involve an implicit knowledge claim, or in Spinoza’s words: an affirmation.26 This assertion is best clarified with the help of a short

_____________ 24

25 26

2def3, Collected Works, 447. Here Spinoza opposes Hobbes’ critique of Descartes according to which thoughts, although they are mental, are to be ascribed to the body cf. AT VII, 172f. For the impact the Objections against the Meditations in general had on Spinoza, cf. also Rousset 1992. For a further discussion of the implications of conceptual cf. also Della Rocca 1996. Della Rocca showed that the physical and the mental constitute two opaque semantic contexts. Curley 1969, 123-127, suggested a logical reading of Spinoza’s concept of ideas. This has been criticized by Wilson 1999, 153, and Della Rocca 1996, 8, who put forward psychological interpretations. I think that the logical and the psychological interpretations are not necessarily alternatives to each other. Ideas are ascribed in the Ethics to minds just as today we ascribe mental states to mind, the term ‘idea’ hence applies to psychological entities. The content of particular ideas, on the other hand, is to a large degree determined by inferential relations to other ideas, which suggests that they are logical entities. Taking these points together one can say that Spinoza’s approach is similar to the one put forward by Wilfrid Sellars in Empiricism and the Philosophy of Mind.

90

Ursula Renz

examination of his discussion of the problem of free will. Just as in the case of the concept of mind, Spinoza’s first step is to deny the existence of something behind our volitions.27 There is no faculty by means of which we either will or will not do something. One might however object that this denial makes any volition unintelligible, for how can we exert particular volitions if no faculty of willing is activated? Spinoza responds to this objection by arguing that all ideas involve some kind of affirmation.28 He illustrates this claim with the same example Descartes uses to defend his notion of innate ideas: Given that we have the geometrical concept of a triangle, we know and affirm by the very same concept that its three angles equal 180°.29 The necessary connection between the conceptual knowledge of a thing and the propositional knowledge of its having certain properties functions in this manner as the model for Spinoza’s concept of affirmation. In contrast to Descartes, for whom only innate ideas necessarily involve the affirmation of some proposition, for Spinoza, it characterizes all of our ideas, including obscure and inadequate ones. This in turn sheds some light on the underlying psychology of the process of affirmation. According to Descartes we can withdraw from our affirmation of ideas, for it is a voluntary act. Not so in Spinoza. For him, the affirmation of an idea is required by its particular content. Accordingly, the will of a person is merely a function of the content of the ideas he has. If he has the idea of x and if this idea bears a relation to the idea of y, he necessarily will affirm the idea of y whenever he thinks of x. (3) In the introduction, I said that the term ‘idea’ is used in the Ethics to refer to the basic mental units. It would however be wrong to take ideas as some kind of psychological atoms. On the contrary, they are always and necessarily related to other ideas by different kinds of connections, inferential as well as associative ones, and they can never be isolated from those connections. Conceiving of ideas in this manner has enormous implications for the question of the determination of mental content as well as for the problem of epistemic justification. The relation of an idea to other ideas determines at least in part what particular content it has, and it has to be assumed, therefore,

_____________ 27

28 29

2p48: In the mind there is no absolute, or free, will, but the Mind is determined to will this or that by a cause which is also determined by another, and this again by another, and so to infinity. Spinoza generalizes this point in 2p48s: In this same way it is also demonstrated that there is in the Mind no absolute faculty of understanding, desiring, loving etc. From this it follows that these and similar faculties are either complete fictions or nothing but Metaphysical beings, or universals, which we are used to forming from particulars. So intellect and will are to this or that idea, or to this or that volition as ‘stone-ness’ is to this or that stone, or man to Peter or Paul. Collected Works, 483. 2p49: In the Mind there is no volition, or affirmation and negation, except that which the idea involves insofar as it is an idea. Collected Works, 484. 2p49dem, Collected Works 484. Descartes uses this example in the fifth meditation, cf. AT VII, 68.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

91

that Spinoza asserts some kind of ideational holism.30 Moreover, as every idea involves some knowledge claim, he also maintains an epistemic coherentism. Although Spinoza assumes that the truth of an idea ultimately consists of its complete correspondence with the object,31 he nevertheless asserts that the adequacy of an idea, that is, the knowledge it provides us, does not depend on its relation to the represented object, but on its relation to all other ideas.32 One might wonder whether all of these beliefs form a coherent whole. Let me remind you that the principal goal of Spinoza’s metaphysics was to ensure complete intelligibility of being. One can argue that complete intelligibility can be maintained if and only if, one singular concept of reality is affirmed and ideally represented by one singular true idea. This idea does not only correspond to reality, but also involves a whole coherent system of ideas by means of which we can adequately think of all particular that there are. It is, I claim, this intuition that stands behind Spinoza’s notion of the ‘idea of God’, and not the assumption of some divine self-consciousness which continuously generates our minds and its mental states, as it sometimes suggested in pantheistic readings.33 Instead, the question arises how Spinoza’s notion of idea contributes to the explanation of mental dispositions. One has to distinguish between two different layers of the problem. On a general level, we must clarify what it means, in a Spinozistic framework, that the human mind, which itself is an idea, has other ideas. The intuition that human beings are capable of having mental dispositions is to be accounted for within Spinoza’s framework by explicating the relation between the idea constituting a mind and the ideas it has. On a more specific level, explanations are also required for certain characteristic mental dispositions that we ascribe to particular persons, for example, when we say someone is irascible. How can Spinoza account for these kinds of properties? In other words: How can he explain that someone behaves, under certain condition, rather in this than in that way? Concerning the first question, we can say that if there is an idea constituting a mind, it provides us at the same time with other ideas. For every single idea is necessarily connected with other ideas, and, given that ideas involve affirmations, in having one idea we at the same time have many other ideas.

_____________ 30 31 32

33

For Spinoza’s holism, cf. also Brandom 1976, and Della Rocca 1996. Cf. 1ax6, Collected Works, 410. Cf. 2def4, Collected Works. Spinoza equates adequacy of ideas with the intrinsic denomination of their truth. This has sometimes been taken as a psychological feature, so that adequacy corresponds to the certainty an idea involves. The problem of this interpretation is that it suggests a psychological understanding of adequacy. But this gives rise to serious problems, as I expound elsewhere, cf. Renz 2009. Instead I take “intrinsic” to hint at the assumption, that by mere coherence with all our other ideas, a particular idea can be epistemologically justified. This view has been expounded and criticized by Wilson 1999, 126ff.

92

Ursula Renz

What accounts for the existence of our mental capacities is therefore the fact that our mind, conceived of as an idea, relates to other ideas we affirm, if not prevented by another, opposite idea. I will not discuss this answer in depth. I would, however, like to emphasize one point: Whether this approach can satisfy us, or not, depends, among other things, on the answer to the following question: Who is the subject that has the idea constituting the human mind and who is thereby endowed with the disposition to affirm other ideas? God? Any other ideal observer? Or the person whose mind we are talking about? As mentioned above, the idea which constitutes our mind is distinguished from other ideas that are not minds by the fact that it possesses some kind of self-awareness. It must therefore be the person herself, the subject who has the idea constituting her own mind. One can speculate that the subject is provided with other ideas due to the fact of her self-awareness. The mental capacity to have ideas depends on the existence of self-awareness. In a way, this comes strikingly close to the fundamental insight of Kant’s transcendental deduction, though the latter is conceived of as a transcendental argument, which is not the case in Spinoza. Both, Kant and Spinoza, seem to hold that it is due to self-awareness that we can know other things as well. The second question is even more important in this context: How can we account for the fact that persons have a tendency or an inclination to behave in a certain way rather than in another? As we have shown, Spinoza denies that the essence of man can be conceived in terms of substances or substantial forms, and therefore, as claimed above, our mental life also has to be explained by an analysis of the causal interaction between mental items. We however have not yet explained what it means for an idea to cause another idea. Knowing that Spinoza maintains some kind of ideational holism, we can presume that an idea is caused by another one if and only if, the content of the latter is partly determined by the content of the former. Understood in this manner, the question of causation of ideas is not concerned with the generation of mental states, but with those semantic processes that determine the contents of our thought. To cause an idea is not to bring it into existence via a psychological process, but to determine its particular content. But how can this notion of mental causation help to explain why someone behaves in this way rather than in that way? Well, this kind of mental causation works also in the case of the relation between a person’s mind (which consists in a particular idea) and the ideas this person has. The content of the idea constituting his mind is one, though, not exclusively, the cause for what he is thinking or feeling in certain circumstances. If we take into consideration that ideas constituting minds are always concrete ideas that differ in extension as well as in intension, we can use this notion of mental causation as a conceptual model for the explanation of our own as well as of other people’s

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

93

mental dispositions. Let’s for example assume that I perceive myself as a powerful personality. I will tend, in that case, to perceive other strong personalities as peers, whereas, if the idea constituting my mind represented me as a weak and fragile person, I will think of strong personalities as a threat to me. This is of course a very rudimentary model, but not in itself inconsistent. Moreover, it shows why the acquisition of knowledge and the reflection on one’s own behavior can, at least in principle, change one’s mental dispositions. Since the mind is itself an idea and is not categorically distinct from the ideas we have, it is not only the case that the mind determines what we think, but what we think determines what mind we have. And this, of course, is no less plausible. Once I have learned that other individuals who I was inclined to perceive as heroes are no less dependent on other human beings than I am, and make mistakes just as I do, I will tend to perceive myself as stronger than before. To summarize, we can say that Spinoza does not deny the reality of mental dispositions but he suggests a conceptual shift in our understanding of them. Instead of referring to general and irreducible faculties which, when activated, produce a wide range of effects, we have to focus on the more fine grained mental capacities involved in our ideas. The intention of this conceptual shift is clear: Ideas can be analyzed by their relations to other ideas, and in this manner mental dispositions become completely explicable. One could object here that, even if we could in principle account for the mental capacities of a person in terms of the system of ideas he has, to look for a complete analysis of this system is a highly ambitious goal. It is indeed extremely improbable that we will ever acquire full knowledge of his mental dispositions. But what is worse: Spinoza’s concept of idea cannot provide a criterion for the ascription of those provisional dispositional concepts we need as long as we lack a complete analysis. Such a criterion however, is, even for Spinoza’s rather deflationary view, quite an important requirement, for we cannot get rid of all of our provisional dispositional concepts at one and the same time. The next section, therefore, will discuss the principle Spinoza introduces in order to account for the specific mental capacities of specific types of beings. I will show that this principle also provides a heuristic criterion for the rational ascription of provisional dispositional concepts.

4. Conceptualizing Types of Mind: The Proportionate Correlation Between Mental and Physical Capacities It is one of the striking aspects of Spinoza’s philosophy of mind that, once he has given his definition of the human mind, he interrupts his discussion with a short exposition of the basic principle of mechanistic physics and of his

94

Ursula Renz

natural philosophy in general. But why do we need physical knowledge if the mental is conceptually distinguished from the body, so that not only causal interaction, but even explanatory reduction is precluded?34 Spinoza justifies this move by appealing to the idea of specific differences between human minds and the minds of other things.35 One might wonder whether this is consistent with the dismissal of the notion of essential mental faculties discussed above. We cannot, however, deny that minds differ enormously in what they can or cannot do. Even if we reject natural kinds, or categorical differences between types of minds, we must, from a commonsense point of view, be able to differentiate between certain types of minds. We must however not merely posit these differences, but they must be somehow accounted for. Spinoza therefore puts forward the idea of a correlation between an individual’s mental and physical capacities: I say this in general, that in proportion as a Body is more capable [aptius] than others of doing many things at once, or being acted on in many ways at once, so its Mind is more capable than others of perceiving many things at once. And in proportion as the actions of a body depend more on itself alone, and as other bodies concur with it less in acting, so its Mind is more capable of understanding distinctly.36

The question arises whether or not Spinoza thereby embraces some version of physicalism implying an explanatory reduction of mental capacities to physical capacities. I don’t think that is the case, since he merely states proportionality between the amount of certain bodily and the amount of certain mental capacities. This, moreover, has to be seen in the light of 17th century psychology. Spinoza hereby rejects one of the underlying ideas of Descartes’ Passions de l’âme according to which the passions of the body are correlated with actions of the mind and vice versa.37 Whereas Descartes seems to assume some kind of inverse proportion, Spinoza maintains that the amount of certain capacities of a mind is proportionally the same as certain capacities of the body. But how, one might wonder, does this assumption of proportionate correlation help to account for differences between types of minds, especially if it does not allow for a reduction to types of bodies? In order to answer this question, three points have to be observed: (1) The assertion of a correlation between mental capacities and physical constitution is not inferred from empirical observation. Instead, it is implied by one of the central ontological claims of the Ethics asserting that there is a pervasive correspondence of the causal order of ideas with the causal order of

_____________ 34 35 36 37

For a detailed analysis of this point cf. Della Rocca 1996. Cf. Collected Works, 458. Collected Works, 458. Cf. AT XI, 327f. and 354f.

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

95

things.38 By assuming a correlation between mental and physical capacities of individuals, Spinoza hence relies on a conceptual constraint of his metaphysics that at the same time precludes any ontological reduction of one type of attribute to another. It does not, however, preclude the possibility that empirical knowledge provided by natural sciences can be used to account for dispositional properties. On the contrary, given this correlation and given certain mechanistic explanations of physical capacities we can, at least in principle, justify the ascription of mental capacities to certain types of things in this manner. (2) The correlation holds universally, since it applies to all individuals. According to the single definition given in the physical excursus, a thing is an individual if and only if a certain number of particulars constitute either one moving thing or one homeostatic system in which a fixed proportion of motion and rest is maintained.39 In other words: An individual is a thing whose physical existence and actions can be explained in terms of the causal role of its parts. But this definition not only covers all bodies, and hence, all finite things, but it also applies to the universe, as far as it is conceived of in terms of the causal role of its parts. The assumed correlation between mental and bodily capacities can thus be used to analyze all kind of things and their ideas, though, as I will argue below, not all kinds of things have minds. (3) Both kinds of dispositional concepts involved in the correlation are described in a manner which allows for degrees. The proportion of a body’s capacity to perform certain causal roles corresponds to the proportion of its mind’s perceptions. It would, however, be wrong to conclude that Spinoza wanted to provide a quantitative analysis of the mental. The Ethics not only lacks any reliable system of measuring the causal roles of bodies. It also suggests, looking at the rather sketchy way his mechanistic physics is exposed just after the above cited statement, that Spinoza did not even want to develop one. The correlation between bodily and mental capacities, though it could be elaborated in a more sophisticated manner, is not primarily thought of as leading to results in scientific measuring, but to provide a ground for a rational ascription of mental dispositions. Spinoza’s claim of a correlation between mental and physical capacities does not have the status of a descriptive hypothesis that can be empirically confirmed or falsified.I It explicates some kind conceptual device which he considers to be fundamental for his analysis of the mental. The resulting position is much closer to our ordinary intuitions than it appears to be at first glance. To illustrate this point, remember for a moment Thomas Nagel’s

_____________ 38 39

Cf. 2p7: The order and connection of ideas is the same as the order and connection of things. Collected Works, 451. Collected Works, 460.

96

Ursula Renz

famous question What’s like to be a bat?. I will however utilize it differently from the manner in which it is commonly used in the literature.40 In recent discussions of contemporary philosophy of mind this question has mainly been brought up in order to argue for the inexplicability of conscious experience from a third person perspective. Nagel himself uses it in this way. Part of the game of Nagel’s thought experiment, however, lies in the fact that it makes use of a highly suggestive question we frequently raise in ordinary speech in order to appeal to the imagination of others. We often question what’s like to be an x or a y. It makes sense to ask what it is like to be this or that creature, though we might often lack the answer. In everyday life, we often raise similar questions in regard to other persons. We, for instance, ask what it is like to be an inhabitant of one of these regions which, for the fourth time within a few years, have been inundated. And we sometimes even appeal to others by questioning what’s like to be in this or that position. Spinoza’s assumption that mental and physical capacities are correlated will of course not contribute to a solution of the problem of consciousness. (Anyway, I think the problem of consciousness in the modern sense of the word is beyond Spinoza’s historical horizon, but that’s another issue I will not address here.) But this correlation does help clarify another point which is of an even greater importance. There is a striking tension in our intuitions about whether or not we can really know what’s like to be someone or something else. It is often assumed that we cannot know what it is like to be something or someone else “unless we put on his shoes and walk around a little bit.” On the other hand, we tend to think that other entities experience certain affections more or less in the manner that we do, if we had a similar constitution and if we were in similar circumstances. We therefore presuppose that the experiences of different subjects are systematically comparable with our own experience. Contemporary discussions in the philosophy of consciousness suggest that we have to reject one of these two intuitions. Spinoza’s conceptual device, in contrast, takes them to be quite compatible. Due to the universality of the assumed correlation, we only need one single psychology for all kind of beings. On the other hand, given that the difference between our mental capacities and the capacities of other beings is a matter of degree, it can easily happen that, from a certain degree onward, the differences are so big that they go beyond our imaginative capacities to envisage what it is like to be x or y. This, however, is not to say that an explanation of other minds is completely impossible.

_____________ 40

Nagel indirectly suggests my unorthodox reading of his example, by the justification he gives for his choice of bats instead of mice, pigeons or whales: “if one travels too far down the phylogenetic tree, people gradually shed their faith that there is experience at all.” Nagel 1991, 423 (emphasis U.R.).

Explicable Explainers: the Problem of Mental Dispositions in Spinoza’s Ethics

97

To summarize the results of this section, Spinoza does not reduce the mental to the physical by claiming that an overall proportionate correlation between mental and physical capacities exists. Nor does he forcibly maintain the idea that all things are endowed with minds. Instead, he establishes a conceptual device which provides us with a structure that enables us to account systematically for all types of mental capacities. The assumption of a proportionate correlation between mental and physical capacities is an important corroboration of the rationalist claim that all there is and all that happens in the world is completely conceivable. It places rationalism in an ontological region that is often thought to be epistemically inaccessible, the realm of the subjective experience of other things.

5. Conclusion: Explicable Explainers In this paper, I have discussed Spinoza’s views on those issues which are systematically related to the question of dispositional properties; in particular to the topic of mental dispositions. Although his metaphysical views appear to be rather odd at first glance, they can be understood as expressing a very simple, rationalist expectation, that is, the idea that every thing and every phenomenon, if only analyzed carefully enough and by means of the right concepts, can be fully understood. Spinoza emphatically embraces this idea, but as I have argued, that is all he wants to say when he denies in 1p33 that “[t]hings could have been produced by God in no other way (…) than they have been produced.”41 He does not, however, reject the idea of something possibly being the case nor does he deny that dispositional properties are real. There remains however a puzzling ambiguity. According to Spinoza’s rationalism it is possible to analyze all of our notions of dispositional properties in terms of actual causal relations. On the other hand, since all finite things are causally linked to each other, a complete analysis would require so much knowledge that it seems to surpass our cognitive capacities. As finite human beings we, thus, may never be able to completely describe any property of a thing. Therefore, the complete description of dispositional concepts through concepts of actual terms is at the same time both possible and impossible. It is ontologically conceivable. But historically, it can never be achieved. All of this, however, does not have to vex Spinoza, since his claim regarding a complete analysis of dispositional properties does not amount to an ontological denial of their reality. He can confidently accept to live and work in a provisional state of knowing. His combination of radical rationalism with common sense realism is a bet whereby he gains a lot without losing anything.

_____________ 41

Collected Works, p. 436.

98

Ursula Renz

Literature Descartes, René. 1996. Oeuvres de Descartes. Ed. Charles Adam/Paul Tannery. Paris: Vrin (=AT). Bartuschat, Wolfgang. 1992. Spinozas Theorie des Menschen. Hamburg: Meiner. Brandom, Robert. 1976. “Adequacy and the Individuation of Ideas in Spinoza’s Ethics.” Journal of the History of Philosophy 147-162. Carriero, John. 1991. “Spinoza’s Views on Necessity in Historical Perspective.” Philosophical Topics 19: 47-96. Curley, Edwin. 1969. Spinoza’s Metaphysics. Cambridge, Mass.: Harvard University Press. Curley, Edwin and Walski, Gregory. 1991. “Spinoza’s Necessitarianism Reconsidered.” In New Essays on the Rationalis, ed. Gennaro and Huenemann, 241-262. Della Rocca, Michael. 1996. Representation and the Mind-Body Problem in Spinoza. Oxford: Oxford University Press. Garrett, Don. 1991. “Spinoza’s Necessitarianism.” In God and Nature: Spinoza’s Metaphysics, ed. Yirmiyahu Yovel. Gueroult, Martial. 1974. Spinoza II. L'âme. Paris. Hampe, Michael and Schnepf, Robert, eds. 2006. Ethik in geometrischer Ordnung dargestellt. Berlin: Akademie-Verlag. Nagel, Thomas. “What Is It Like to Be a Bat?” Reprinted in: Rosenthal, David M. 1991. The Nature of Mind. New York / Oxford: Oxford University Press. Perler, Dominik. 2006. “Das Problem des Nezessitarismus.” In Hampe/Schnepf (eds.), 5980. Renz, Ursula. 2006. “Die Definition des menschlichen Geistes und die numerische Differenz von Subjekten (2p11-2p13s).” In Hampe/Schnepf (eds.), 101-121. Renz, Ursula. 2007. Die Erklärbarkeit von Erfahrung. Realismus und Subjektivität in Spinozas Theorie des menschlichen Geistes. Zürich (manuscript: Habilitationsschrift). Rousset, Bernard. 1996. Spinoza. Lecteur des Objections faites aux Méditations de Descartes et de ses Réponses. Paris: Éditions Kimé. Ryle, Gilbert. 1949. The Concept of Mind. London: Hutchinson. Schütt, Hans-Peter. 1985. “Spinozas Konzeption der Modalitäten.” Neue Hefte für Philosophie 24/25: 165-183. Spinoza, Baruch de. 1988. The Collected Works of Spinoza. Ed. by Edwin Curley, Princeton: Princeton University Press (= Collected Works). Spinoza, Baruch de. 1925. Opera. Ed. by Carl Gebhardt. Heidelberg: Winter (= G I-IV). Wilson, Margaret D. 1999. Ideas and Mechanism. Essays on early modern philosophy. Princeton: Princeton University Press. Yoshida, Kazuhiko. 2004. Vernunft und Affektivität. Untersuchungen zu Spinozas Theorie der Politik. Würzburg: Königshausen & Neumann.

Harmonizing Modern Physics with Aristotelian Metaphysics. Leibniz’s Theory of Force MICHAEL-THOMAS LISKE

0. Introduction Common wisdom has it that the philosophy at the time when Leibniz began his philosophical career was divided between two opposing schools: Continental rationalism and Anglo-Saxon empiricism. But there is another division that goes deeper, and which is the proper opposition within early modern philosophy: the Aristotelian approach, which aims at understanding natural processes in terms of qualities and powers, versus the quantitative analysis of natural functions undertaken by modern mechanistic science. German universities, even the Protestant ones, still taught a conservative philosophy that was widely determined by Aristotelian-Scholastic concepts and patterns of thought. Leibniz was introduced to this philosophy at the University of Leipzig under the guidance of his teacher Jacob Thomasius. Beside this philosophical tradition, a new movement in philosophy and science was coming up, originating from Descartes, which completely disapproved of the Aristotelian approach as being sterile. It regarded the Aristotelian practice of explaining natural processes in terms of ad hoc postulated potencies as having no scientific explanatory value, such as, for example, trying to understand a natural process like warming in qualitative terms by means of an underlying disposition to warming. According to his own account of his philosophical development, Leibniz, having weighed one against the other, came to a decision in favor of modern philosophy against his Aristotelian beginnings. However, when he realized that modern philosophy, too, was deficient, he reached out for a synthesis of both. Historically, Leibniz may have contributed to a legend about himself. Systematically, there were good reasons for Leibniz to regard the modern approach, if it remained unmodified, as equally insufficient. When empiricists and rationalists, following Descartes, try to eliminate all qualitative and metaphysical aspects from natu-

_____________ I should like to thank Mr. Markus Geisler for assisting me in preparing the English version of this paper.

100

Michael-Thomas Liske

ral science by explaining all natural processes in purely quantitative terms of the extension of bodies, the modifications of extension, which are size and shape, and the changes extension suffers by the effect of the bodies’ motions, they are able to explain precisely through mathematical functions how nature works. By means of such functional dependencies, one can predict natural events and so control and influence them. Surely, explaining all natural processes mechanistically by the same all-encompassing quantitative principles is more convincing than explaining them by postulating in an ad hoc manner for each event a different quality or potency. But the mechanistic approach does not answer the more fundamental question of why nature is the way it is. It does not explicate why it does obey these mechanistic laws rather than different ones. For instance, in the case of a catastrophe such as a tsunami, mechanistic science makes clear the precise functional dependencies leading up to the final event. But it does not at all answer the existential question of why such a disaster takes place in the first place and why it is me who falls victim to it. These types of question become urgent in the context of a conception of possible worlds, which, originating in Duns Scotus and handed down by baroque scholastics, took on a determining role for Leibniz’s thought. If one assumes with Leibniz that entirely different histories of the world, governed by totally different laws, could have been realized it must be explained why nature is determined by these laws rather than others. According to Leibniz, this question, however, can only be answered in qualitative and teleological terms like the concepts of that which is the best and most suitable or in terms of a directedness toward a goal. If such a teleological view is not to be scientifically worthless from the very beginning, it must not be defined in light of a notion of external suitability such as that an object or an animal of one kind serves the purposes of that of another kind. Rather, the teleological view must be based on a directedness toward an immanent goal: Something develops in virtue of an inner striving force toward the state which represents the perfection of this development and thus is its goal. (For instance, according to Aristotle, the state of a mature, adult specimen of the relevant kind is the goal of the individual process of growing and maturing.) Having said this, we have reached Leibniz’s concept of force, which he understands as an Aristotelian entelechy, that is, as a goal-directed principle that is immanent to the object and constitutes it. Understood this way, the concept of force suits Leibniz in his intention to combine an exact mechanistic description and explanation of natural processes with a metaphysical understanding of why things happen the way they do. On the one hand, in mechanistic physics force is an important natural constant. This becomes obvious when Leibniz seeks to prove against Descartes that it is not the quantity of motion (mass multiplied by velocity) but force (mass multiplied by velocity squared) that the physical laws of conserva-

Leibniz’s Theory of Force

101

tion refer to. On the other hand, the notion of force seems promising when one tries to find a concept leading from a mechanistic explanation of the phenomena of movement to a metaphysical understanding of the foundations of motion; for instance, when we face the basic question: How can actual motion originate from a state of rest characterized by pure potentiality for motion? Leibniz realized: if one wants to avoid the assumption that the actualization of every potential for motion is the effect of some other object that actually possesses the feature to be actualized – an assumption that would lead to an infinite regress – one must assume an inner principle in virtue of which something can make itself move. This inner principle is what Leibniz refers to as force. It is not understood as a mere potency but as a potency that is amplified by a striving directed toward a goal; it causes motion by itself if nothing external hinders it. Such an immanent principle is, however, not primarily the cause of physical movements. In its primary sense, it is the origin of an object’s proper activity, which constitutes it as a unitary substantial object. In this manner Leibniz equates primary force with the AristotelianScholastic substantial form. The human soul or mind is the principle of the proper human activity in that it disposes a human being to a way of living and acting determined by rational decisions. This way of existing constitutes an individual’s being a man. From the point of view of modal logic, three stages must be distinguished: A mere faculty or natural constitution (as the natural faculty of a newborn child to speak) does not immediately allow for the corresponding activity but that natural faculty must first be further developed. An already developed potency immediately allows its realization, but not necessarily from itself; rather it needs to be activated by an external stimulus (as for instance an adult’s developed faculty of speech when she or he sleeps: he begins to speak only after he has been woken up and has been spoken to or asked something by someone else). Only a potency amplified by a striving can transfer itself directly into an activity, for instance, when an adult capable of speech has the desire to talk with somebody about something, given the right occasion. Because Leibniz takes substance to be an autonomous system bringing forth all its activities from an immanent principle, force, especially primitive active force as the principle of the activity constitutive of a certain substance, must be not only a developed potency, but one amplified by a striving which can transfer itself into activity (cf. s. I). How is all that has been said connected to the general theme of this volume, disposition? Can Leibniz’s forces be interpreted as dispositions? Not in a straightforward manner: For dispositions are, according to the established usage, only developed potencies which immediately allow for their actualization, but need to be triggered by appropriate external situations. Just in this sense Leibniz, too, uses the term ‘disposition’ (not referring to forces but

102

Michael-Thomas Liske

mostly to innate ideas). As a determinist, Leibniz assumes that in reality there cannot be mere faculties or natural constitutions that are open for different actualizations. Such dipositions are for him mere abstractions of reason. He emphasizes that even if a disposition is not an amplified potency including (like force) a certain striving as its integrative moment, it must always be connected with a certain striving and so directed at a certain realization (cf. s. II). The fact that force, understood as an amplified disposition, is re-introduced to modern science clearly shows that science is related in an important way to the ontology of substance. For, being dispositional properties, forces are attributes of substantial subjects. Leibniz thus disavows a merely relational view of motion implied by an analysis of motion in terms of purely quantitative functional dependencies. If Leibniz assumes that, in reality, dispositions always are amplified by a striving and therefore are forces, his concept of disposition seems to differ in an important way from the common one: Many familiar dispositions like solubility and fragility are passive dispositions to suffer a change. A striving, however, is always directed forward and thus toward an activity. There seem to be no forces to suffer a change. When Leibniz nevertheless assumes passive forces this is because he understands them as those forces whose effects are weaker than the effects of the other acting forces (within a certain process) and so are inferior to them. Their passivity consists in their having to yield to the activity of the superior forces. Still, they cause an opposing activity, or reaction, insofar as they limit and modify the prevailing activity through their resistance (in the form of impenetrability and inertia). Thus, they prove to be real forces. (cf. s. III). The first instances of forces qua dispositions (which are amplified by a striving toward the state to be effected) that come to mind are physical effects. What reasons did Leibniz have to conceive of these physical forces as being merely derived modifications of underlying metaphysical forces? Above all, there were two reasons: 1.) Only metaphysical forces are absolute attributes of the individuals they constitute, whereas physical forces are always relative properties because they are related to movements. 2.) Leibniz defines physical force as the present moment in the course of a movement; this moment is directed by striving to the following one, therefore includes it by anticipating it and so brings about the transition to it. In this manner he can regard physical force as a modification, that is, as a transitory or momentary particular state of an underlying permanent subject. The primitive metaphysical force is such a permanent subject insofar as it is conceived of as the basic tendency of the striving which determines the whole succession of an individual’s perceptions according to a law characteristic for this individual and thus constitutes its individuality (cf. s. IV).

Leibniz’s Theory of Force

103

For Leibniz, the momentary striving for movement (as an infinitesimal quantity) is but one type of physical force, the so called dead force (vis mortua) or potential energy, in contrast to living force (vis viva) or kinetic energy, which arises when infinitely many solicitations of vis mortua in the course of a movement continuously sum up to result in a force which brings about an effect of a measurable quantity (cf. s. V). In contrast to this reasonable differentiation of physical forces, Leibniz’s treatment of primitive metaphysical force suffers from an ambiguity: On the one hand, he takes bodily substances (like living beings) to be genuine substantial unities whose unity is constituted by a primitive active force as its immanent principle, which is the disposition to the activity characteristic for the living being as a whole. According to the phenomenalistic approach, on the other hand, the exclusive domain of primitive force is the mental activity of monads. The individuality of a monad is constituted by the interaction of active force, which strives forward toward ever clearer and more distinct perceptions, and passive force, which hinders active force and so leads to obscure and confused perceptions. The whole realm of bodies is a mere phenomenon (cf. s. VI). After these detailed analyses, the physical and the metaphysical aspects of Leibniz’s concept of force can be treated in more depth. Even if Leibniz has discovered the concept of force in the context of physics, from a systematic point of view, the metaphysical understanding of the concept is primary. This is so because the metaphysical concept is capable of explaining what for Leibniz needed to be explained in the context of his deterministic ontology of substance. As a determinist, he was looking for some factor sufficient to effect a future change directly by itself (according to the determining laws). Since Leibniz, as a substance ontologist, endorses agent causality, he has to answer the question in virtue of which attribute a substance is determined to effect a change in something else. Leibniz answers by pointing to force as a substance’s disposition amplified by a striving. Because of this striving, the substance is directed and determined to a certain activity which by itself explains a change in another thing. It is in virtue of such a disposition of a substance that the change in the other thing immediately follows if nothing hinders it. Accordingly, the determining laws themselves have to be understood as forces or dispositions directed by a striving to a certain sequence of events and determining them in this manner (cf. s. VII).

104

Michael-Thomas Liske

1. Primitive Active Force as an inner Principle of Activity which constitutes Substance and the amplified Notion of a Potency that actualizes itself by virtue of an inherent Striving Active force, which is commonly called force in an unrestricted sense, must not be conceived of as a mere potency (as usually understood by the scholastics) or as receptivity to action; rather, it involves a directed momentary velocity (conatus1), or tendency toward action, so that action follows if nothing else hinders. This is what properly consists in – too little understood by the scholastics. For such a potency involves the act and does not consist in a bare faculty. Yet, it does not always entirely proceed to the action toward which it tends, namely when an obstacle occurs. Furthermore, active force is twofold, primitive and derivative, i.e. either substantial or accidental. Primitive active force, which Aristotle calls , commonly called substantial form, is the one natural principle that together with matter, or passive force [as the other principle], completes corporeal substance which by itself is a unity and not a mere aggregate of several substances. For there is an important difference – to give an example – between an animal and a flock. Therefore, this entelechy is either a soul or something analogous to soul and it always naturally actualizes some organic body that, taken by itself (if the soul is set apart or removed), is not a unitary substance, but an aggregate of many, for instance a natural mechanism. (GP IV 395 sq.)2

This passage from a treatise (dated May 1702) on the philosophical basis of dynamics3 illustrates clearly the two main features of Leibniz’s conception of force. From the viewpoint of modal logic, force, being an amplified potency, is an intermediary stage between the mere faculty to act and the action itself (atque inter facultatem agendi actionemque ipsam media), as Leibniz says in the treatise On the Emendation of First Philosophy and the Notion of Substance (GP IV 469). From the viewpoint of a metaphysics of substance, primitive active force fulfils those ontological tasks in constituting substance that scholastic Aristotelianism ascribed to substantial form, which is to make a corporeal substance

_____________ 1

2 3

See SD I p. 10, l. 117-121 (GM VI 237). Leibniz’s reform of traditional Cartesian-Newtonian mechanics, which is manifest especially in the notion of conatus, consists in his interpreting the instantaneous velocity as being not merely the average velocity, but rather a measure of the body's activity; see Gale 1973, 200. All translations of quotations from Leibniz are my own. By introducing the term ‘dynamics’ for a new science, Leibniz claims to have discovered a new approach to explaining motion that reforms traditional mechanics as it was presented especially by Descartes, from which Leibniz tried to distance himself in important aspects, although he was deeply influenced by it. Certainly, Descartes, Galileo, and Newton treat, among other things, forces causing motion, but not under the title of dynamics. Newton makes a rather pragmatic use of the force concepts without enquiring into their ontological status. So, Leibniz is entitled to consider it his innovation when he claims: The principles of the physical universe and thus of mechanics and its laws of motion are to be found not in geometry, but in the indivisible entities of metaphysics, that is in the substances constituted by primitive forces. On the whole problem, see Gabbey 1971. On the opposition dynamics-mechanics, see also Gale 1973, 184-188.

Leibniz’s Theory of Force

105

that is essentially one (unum per se) from an assemblage or mere aggregate of disconnected material parts. In Aristotelian terminology, form is the principle for a thing’s being this or that, that is, the principle constituting it as the entity it is. Accordingly, the concept of substantial form implies the postulate that there is a principle that constitutes something as a substantial entity. But Leibniz faults the scholastics for being unable to explain how such a constitution is possible. This can be done, in Leibniz’s view, with the help of the concept of force. Being borrowed from physics, it can (in contrast to the abstract fictions of the Scholastics) be given concrete meaning through observations of nature; on the other hand, the concept of force allows for metaphysical deepening. For this purpose, Leibniz employs the conception of a principle of activity. In Theod. I § 87, he interprets form (in Aristotle’s as well as scholastic usage) as a principle of activity, inherent in the agent itself. (This surely meets Aristotle’s intentions.) The fact that something as a whole can perform a characteristic activity by all its parts’ acting together makes it an entity that is essentially one from a mere aggregate of parts. Such an activity is grounded in force. For the primary force to be able to serve as the inner principle of a thing’s characteristic activity and thus constituting the thing as a substantial unity, Leibniz must conceive of force as an amplified potency, which, by containing an inner striving, brings about the act by itself. Why this force has to be a kind of potency can be seen from the above mentioned paragraph (I 87) from Theodicy where Leibniz presents his theory of force in the context of the Aristotelian scholastic tradition of form and act. Following Aristotle, Leibniz distinguishes between two types of act or actualization of potency: the permanent and the successive act. The actualization that consists in the action is something transitory. Since only a permanent and durable act qualifies as form, in this case the act cannot be the action itself, but only a principle disposing to an action: be it the substantial form that disposes to the essential way of action constitutive of an object of this kind, be it an accidental form that disposes to an action further qualifying the substance.4 While the actions themselves take place in a succession of different phases (l’Acte successif), being disposed to such an activity is a permanent state. It can be considered an act because it is an already developed potency that actualizes the respective natural constitution or aptitude. A developed potency that, given proper circumstances, immediately turns into the related activity still does not satisfy Leibniz’s demands of the concept

_____________ 4

Or le Philosophe Stagirite conçoit qu’il y a deux especes d’Acte, l’Acte permanent et l’Acte successif. L’Acte permanent ou durable n’est autre chose que la Forme Substantielle ou Accidentelle: la forme substantielle (comme l’Ame par example) est permanente tout à fait, au moins selon moy, et accidentelle ne l’est que pour un temps. Mais l’acte entierement passager dont la nature est transitoire, consiste dans l’action même. (GP VI 150)

106

Michael-Thomas Liske

of force. Such a potency would be the active potency of scholasticism, with which Leibniz, in the Emendation treatise, sharply contrasts his newly introduced concept of force. For the active potency or faculty of the schoolmen is merely the immediate possibility to act, which nevertheless needs an external stimulus in order to be transferred into action.5 The possibility immediately preceding the actualization (propinqua possibilitas) completely includes all preconditions of activity. But according to Aristotle’s and the scholastic conception, it cannot transfer itself into activity. It must be activated from the outside by something that actually possesses the feature to be actualized, ultimately by the unmoved mover. As Leibniz takes the substantial form to be an inherent principle of activity sufficient to produce by itself the activity in question, it cannot be a mere potency in need of external actualization. Let us return to the passage quoted at the beginning. Here too, Leibniz rejects the scholastic understanding of active potency or active force as mere potency (simplex potentia) which, taken by itself, would be at rest. Leibniz regards such a conception as an internally inconsistent one. He expresses the inconsistency of this conception by characterizing it as receptivitas actionis. Receptivity, as a thing’s faculty to receive a qualification from the outside or to be formed in a certain manner by an external source, lacks the specificity of action because of its essentially passive nature. Activity, characterized by spontaneity or the faculty to turn itself into action, is guaranteed only by an amplified potency, which is force. An amplified potency is not only an already developed potency6, which fulfills all preconditions of immediate actualization, but it also includes a directed momentary velocity or a tendency toward action (involvit conatum seu tendentiam ad actionem). Because this tendency, or striving, is the transition from potency to activity, a substance constituted by a striving force as the fundamental property is essentially, and therefore always, active. This makes force apt to serve as constituting principle of a Leibnizian substance that, being an entirely autonomous system, cannot receive any energy enabling it to act from outside of itself, but can only be active due to its own inner sources. All of the above certainly does not imply that a substance always exercises the activity toward which it tends completely and unimpeded, for it can be hindered by external obstacles (etsi non semper integre procedat ad actionem ad quam tendit quoties scilicet objicitur impedimentum, 395). An external obstacle, however, can only modify, but never completely prevent, the activity arising from within the substance. Leibniz’s characterization of substance as essentially

_____________ 5

6

Differt enim vis activa a potentia nuda vulgo scholis cognita, quod potentia activa Scholasticorum, seu facultas, nihil aliud est quam propinqua agendi possibilitas, quae tamen aliena excitatione et velut stimulo indiget, ut in actum transferatur. (GP IV 469) Although Leibniz does not use the terminology: developed versus amplified potency (which is mine), he clearly makes the distinction between these two stages of potency.

Leibniz’s Theory of Force

107

active by itself has two implications pointing in opposite directions: Activity cannot be conferred to a substance from the outside, nor can it – on the other hand – entirely be taken away by something external. In regard to material entities, Leibniz can reconcile his metaphysical account of a body as always acting and as always being in motion with the observed phenomena by distinguishing two partial forces (SD I p.14, l. 182-194; GM VI 238f.): The force proper to each of the several parts of an aggregate is the one by which they act on one another, thus changing their respective position or moving in relation to one another (vis respectiva seu propria). The force common to all the parts is the one by which the aggregate as a whole has external effects and moves in a certain direction (vis directiva seu communis). Thus, the movement of a bodily complex can be hindered to such a degree that it seems to be entirely at rest and without any activity in regard to its manifest external effects. But the hidden inner movements of its parts can never completely cease. At any time, therefore, they can give rise to an observable external effect. Developed potencies, even forces as amplified potencies, are not to be identified with Nicolai Hartmann’s total possibilities (Totalmöglichkeiten) as is clearly indicated by the qualification ‘if nothing hinders.’ Leibniz regularly uses this phrase in order to mark the condition for the potency’s being entirely transferred into the action toward which it strives. By applying the three stages of potency – underlying natural constitution, disposition as a developed potency, force as an amplified potency – also to physical phenomena, Leibniz does not presuppose his metaphysical concept of an entirely autonomous subject bringing forth every action from its own sources without any real influence from the outside. Rather, his concept of potency is based on a distinction of external and internal conditions of realization. In contrast to a mere natural faculty, a potency is developed insofar as its internal conditions of realization are given in totality. Consequently, it can be immediately active if the additional necessary condition is given, which is that suitable external circumstances obtain. This also means that it cannot transfer itself into action, but its activity must be triggered by external stimuli. For his concept of a potency to be applicable to the metaphysical definition of substance as an entity spontaneously active from itself, Leibniz introduces amplified potency because amplified potency transfers itself into action in virtue of its inner striving. That which possesses such a force is always active at least in form of an inner activity, which leads to a motion of its parts in relation to one another. For the intended external effect of the whole to come about entirely, external hindrances must have been removed. Although such hindrances cannot prevent the activity of an amplified potency, they can modify its external efficacy. In this sense, Leibniz can conceive of force as an intermediary stage between mere faculty, which is at rest, and the intended external activity.

108

Michael-Thomas Liske

2. In Reality, there are no mere Faculties but only Dispositions or developed Potencies predetermined to certain Effects or tending toward them In dealing with force, Leibniz (as far as I know) does not use the term ‘disposition.’ In passages where he does speak of ‘disposition,’ especially when (discussing Locke’s position) he argues for his own theory of innate ideas and principles, he regularly contrasts disposition as developed potency with mere or bare faculty. In his letter to Burnett (dated Dec. 3, 1703), he expresses the view that Locke’s empiricism fails in that it cannot explain unrestricted universality that is characteristic of necessary truths of reason, for instance of logic and mathematics. The empiricist method of induction allows for certainty only in cases that have been already observed; in cases not yet observed, only greater or lesser probability is possible (GP III 291). The alternative view of Leibniz’s innatism can be illustrated with the help of an example that is not explicitly mentioned by Leibniz himself: The faculty to argue according to a certain logical law, or to know how to use it correctly in an argument, is not of the same kind as the faculty to speak a certain language like English as possessed by newborn infants. The latter is a mere faculty and must be further developed through instructions, that is, external information. In the case of truths of reason, on the other hand, we possess from our birth a disposition to knowledge (disposition à la connoissance) which enables us, given proper circumstances, to immediately utilize logical laws in constructing our argument. The knowledge of a logical law is only occasioned by an experienced situation.7 As a necessary condition for successful logical thinking, these structures of thought have always been rooted in our mind. Accordingly, Leibniz in Nouveaux Essais (NE I 1, § 11, A VI 6, 80) concludes from the fact that we readily assent to certain truths that we must be endowed with more than a bare faculty (faculté nue) or mere possibility of understanding those truths (possibilité de les entendre). Rather, a disposition, an aptitude, or a preformed internal structure must determine our mind in such a way that these truths are derivable from it. A parallel can be drawn between determinism and the view that our mind has a certain preformed structure: Just as for determinists future events are, if not explicitly given in the present situation, in any case implicitly

_____________ 7

Il n’y a pas seulement dans nostre esprit une faculté, mais encor une disposition à la connoissance, dont les connoissances innées peuvent estre tirées. Car toutes les verités necessaires tirent leur preuve de cette lumiere interne et non pas des experiences des sens qui ne font que donner occasion de penser à ces verités necessaires et ne sauroient jamais prouver une necessité universelle, faisant connoistre seulement l’induction de quelques exemples et de la probabilité pour les autres qu’on n’a point encor essayés. (GP III 291)

Leibniz’s Theory of Force

109

included in such a way that they necessarily will evolve from that situation according to the relevant laws, our mind is preformed according to necessary truths of reason in such a way that it is disposed to recognize them as true immediately by itself, given proper circumstances, and to use them correctly. Nearer to the theory of force is another usage of ‘disposition,’ again in an epistemological context (NE II 1, § 2, A VI 6, 110), where Leibniz contrasts the notion of disposition with the pure potencies of the scholastics (les pures puissances de l’école) in a similar way as he contrasts the concept of force with the concept of a mere potency. He rejects the notion of a mere potency because, in reality, every potency implies a striving or tendency. Leibniz considers a pure potency, which, without any action, is completely at rest, an imaginary construct of abstract thought, since conceiving of a potency as being a pure potency to every relevant act abstracts from the directedness towards concrete actions. (Take, for example, the faculty to speak any language whatsoever.) Real dispositions, on the contrary, always imply some striving, directed not to any arbitrary, but to some specific activity. Therefore, dispositions always are specific dispositions to one activity rather than another (une disposition particuliere … à une action plustost qu’ à l’autre), for instance the disposition which enables me to speak a certain language. Because of these tendencies, a disposition always brings about some at least minimal action, as Leibniz also stresses when he treats force. Yet, when he says: “Beyond disposition (outre la disposition), there is a tendency toward action,” he means that disposition itself as a developed potency still does not include a tendency, but nevertheless is always connected with it. In contexts other than innatism, too, Leibniz uses the term ‘disposition’ to refer to a developed potency already directed to a certain activity. In Theod. I § 46 (GP VI 128), he speaks of a disposition to the corresponding action (disposition à l’action), which contains a predetermination such that the effect is completely preformed in nuce. For the determinist, such dispositions or developed potencies, which create determination, are the conditio sine qua non of causal efficacy (une cause ne sauroit agir, sans avoir une disposition à l’action). Bare potencies, which are possibilities of alternative courses of events, according to Leibniz, can only be abstractions of thought. This understanding is also implied in the preface to his Theodicy (GP VI 40), where Leibniz calls the disposition of matter, conferred to it by God, organism, that is, those organizational structures which dispose to a characteristic way of acting.

110

Michael-Thomas Liske

3. The Reinterpretation of Aristotle’s Entelechy as active striving Force and of Matter as passive Force which, gradually inferior, hinders and modifies active Force Let us again have a look at the theory of force from the viewpoint of a metaphysics of substance. According to Aristotle, a particular substance is constituted by form and matter as two complementary principles. In Leibniz this corresponds to the view that primary active and passive forces are interwoven. (For those primitive forces are responsible for the constitution of a substantial individual as basic ontological unit.) Aristotle, too, took form to be active because it is the determining principle; matter, on the other hand, which is to be determined or formed, is the passive principle. In a certain way, therefore, Leibniz is right in claiming that his conception of primitive active force reinterprets Aristotle’s first entelechy ( ) and that his conception of primitive passive force reinterprets the Aristotelian-scholastic prime matter, each on the level of modern science. Yet, by being integrated in Leibnizian thought, both notions undergo important transformations. The concept of entelechy is not only widened in extension, but also modified in its sense. Aristotle defines soul as first entelechy, and thereby restricts the scope of ‘first entelechy’ to living beings. By saying (in the passage quoted at the beginning) “Therefore, this entelechy is either soul or something analogous to soul,” Leibniz widens the extension of the concept. As Leibniz tries to overcome all unbridgeable qualitative contraries in favor of degrees continually succeeding one another, any absolute insurmountable opposition between the inorganic and inanimate, on the one hand, and the living, on the other , disappears. Rather, everything is animated by its being dominated by a principle at least analogous to the soul (if not by a soul in the proper sense), which is an entelechy, or a sort of striving force. This interpretation as a striving force is what the modification of Aristotle’s term consists in. For Aristotle, on the other hand, the concept of qua completion (to bring to an end = ) and the concept of qua actualization, activity, are, if not synonymous, certainly identical in extension. The first entelechy is thus the elementary actualization, i.e. its first or basic stage. The modal status of the soul is characterized as an intermediate stage between a mere faculty (for instance a little child’s faculty of commanding an army, see de anima II 5, 417b 30f.) which has yet to be developed before enabling to the corresponding activity, and the action itself. Because it is not essential to man to always exercise his characteristic act of thinking (i.e. the second and final act), his soul as the constitutive principle or essence can only be understood as the disposition (first act) enabling him to exercise the relevant act as soon as proper circumstances are given. This understanding of

Leibniz’s Theory of Force

111

disposition, though, does not satisfy Leibniz, because it does not allow for an activity arising solely from the active subject itself. It is for this reason that Leibniz adopts an amplified understanding of entelechy as striving force, as is confirmed by the passage quoted above. After speaking of a tendency toward action which (if not hindered) leads to action, he adds: Exactly this is what entelechy consists in. Presumably, the idea of striving toward a goal, or teleology,8 may have suggested itself to Leibniz because of the word ‘ ’ which is part of ‘ ’ (although, I have to admit, he never explicitly explains the etymology of this artificial term in this manner). By interpreting entelechy as striving force, Leibniz can treat entelechy as equivalent to active force. Because striving always includes a moment of being directed forward, which by itself leads to progress, it is opposed to a hindering passive force, which is backward directed. Analogously to the distinction between primitive and derivative active forces, he distinguishes between primitive entelechy (which is his reinterpretation of Aristotle’s first entelechy) and derivative entelechy. This complex conception of ‘entelechy’ is documented in NE II 22, §11, A VI 6, 216 where Leibniz tries to show that by taking power as the source of action, that is as its origin sufficient by itself to bring about the action, Locke has overcome the concept of power (developed in the preceding chapter 21) as a mere precondition which enables, or makes its bearer apt, to an action (aptitude). Leibniz thinks that, in fact, Locke supports his own concept of an amplified power which, by including a tendency, becomes a striving power (entelechy) leading (if not prevented) to action and thus making its bearer a cause. Entelechy, therefore, is presupposed by the concept of cause. In this context Leibniz differentiates furthermore between primitive entelechy, or soul, as the principle constituting substance and derivative entelechy, which is relevant to physical movement9 and its elements, for instance conatus and impetus. Leibniz’s interpretation of Aristotle’s concept of matter as passive force is another important modification. Certainly, for Aristotle too, matter is characterized by a passive potency to be formed or shaped as something definite. Such receptivity might be understood as a thing’s potency to receive a certain form from an external agent although it hardly can be regarded as a force because force is a principle that acts by itself. Yet, Leibniz does not distinguish between potency and force (because he regards a potency conceived of as a pure possibility merely as an abstraction of thought). In his 1702 treatise on dynamics, Leibniz first draws a distinction between passive and active potency, and soon refers to them as forces: passive force – which constitutes

_____________ 8 9

Kangro 1969, 135 stresses the close connection of final cause and entelechy in Leibniz. On entelechy as principle of physical movements, see Bergmann 2002, 158-163.

112

Michael-Thomas Liske

matter, or mass – and active force that constitutes entelechy, or form.10 By passing over so readily from potency to force (vis), Leibniz shows: From the very beginning, he understands even passive potency in the amplified sense of ‘potency’ as ‘force.’ ‘Passive force,’ however, seems to be an inconsistent notion. Force obviously is an active disposition to act. Striving, as the additional element by which force is distinguished from mere potency, is essentially directed toward activity or toward a progression to be brought about by the activity and is not directed to being acted upon. – For Leibniz, it is only possible to speak of passive force because he abandons the qualitative opposition of activity and passivity in favor of a graduation of activity. Ultimately, passive force turns out to be an acting force, although in its grade of efficiency it is inferior to another force so that it is acted upon by the superior force. Yet, it hinders and modifies the action of the prevailing force, thus remaining active. In physics, Leibniz illustrates his understanding of passive force in more concrete terms as resistance.11 The passive force of resistance manifests itself in two ways: On the one hand, as impenetrability (antitypia) it prevents that the space occupied by one body can be taken by another without the first body’s being removed. On the other hand, it is responsible for the fact that a body cannot be removed without exerting force. In general: As a body, because of inertia12, tends to remain in its state of rest as well as in its state of steady linear motion, any change in some body’s state of motion requires that the body causing the change exerts force and so is hindered in its own motion. From these physical observations, Leibniz draws the following metaphysical conclusions: The essence of matter cannot be defined in the statical framework of geometry, which is presupposed by Descartes’ definition of matter as res extensa, but is to be sought in a dynamic element: in the passive force as resistance. If it is to be the basis of materiality, passive force must be evenly distributed through all matter, and so be proportional to matter’s bulk. Observed phenomena only seemingly contradict this claim. According to Leibniz, if a body is lighter and so possesses lesser inertia, this is because it has lesser bulk, possessing big pores, filled with other matter.

_____________ 10 11

12

Porro µ seu potentia in corpore duplex est, Passiva et Activa. Vis passiva proprie constituit Materiam seu Massam, Activa seu formam. (GP IV 395) It is not by chance that Leibniz, in his 1702 treatise on dynamics (GP IV 395, the passage interpreted in the following), does not so much treat primitive passive force (materia prima), but rather explains the impact of derivative passive force on physical motion. While, from an ontological point of view (ordo expositionis), physical force is derivative, from an epistemological point of view (ordo inventionis), it comes first: The physical analysis of motion allows a meaningful distinction between active and passive force, and so Leibniz can differentiate force in general in this manner. On Leibniz’s notion of inertia, also on the background of Kepler’s and Newton’s notion of inertia, see Bernstein 1981.

Leibniz’s Theory of Force

113

4. The derivative Forces are the Tending of the momentary State of Motion toward the next One. The primitive Force is the basic Direction of the Tending constitutive of an Individual The two main distinctions between forces can be combined with one another (see SD I p. 6/8, l. 69-110; GM VI 236 sq.) and are meant to reinterpret an Aristotelian dichotomy each. The distinction between active and passive forces corresponds to Aristotle’s form and matter; the distinction between primitive and derivative forces corresponds to Aristotle’s substance and accident. According to a letter to Jaquelot (dated March 22, 1703) (GP III 457), primitive (active) force is essential to the particular substance (being, like Aristotle’s first entelechy, a principle constituting substance). It belongs to the particular substance taken by itself or inheres in it as an absolute characteristic. Derivative accidental force, on the other hand, also depends on other bodies, being something relative. This relativity results from the fact that derivative forces underlie bodily movement. Because space is a system of relations and interactions of bodies, any movement in space results in a communication of motion. In these interactions, the underlying forces modify and limit each other. (According to Leibniz, however, this does not happen by real influence, but in virtue of an ideal harmony.) Accordingly, Leibniz conceives of the derivative forces as modifications of or as transitory limited manifestations of the primitive force. Analogously, the different shapes of bodies are limitations of extension. However, even derivative forces as causes or principles are real to a higher degree than the observable movements they cause.13 Taken by themselves, these movements would be purely relative and thus phenomenal: When several bodies change their location in relation to one another, it can be only arbitrarily decided which body is the center at rest and which bodies move relative to it. Only if one gives up Descartes’ mechanical view explaining motion in terms of extension and its modifications, and relates motion to force as its cause, is one able to regard a body as being in motion in a nonarbitrary way. Although Leibniz presents this argument in the Discourse on Metaphysics §1814 with respect to the general concept of force, he clearly has in mind what he later, in his more differentiated terminology, calls ‘derivative active force.’ In his correspondence with de Volder, Leibniz specifies the type of modification derivative force is identified with. He takes derivative force to be the momentary state, or the present moment, in the action of moving (letter

_____________ 13 14

Hacking 1985 tries to justify by means of a physical criterion Leibniz’s assumption that motion is a non-real (phenomenal) mechanical quantity and that vis viva is a real one. See Garber 1985, 81.

114

Michael-Thomas Liske

dated June 30, 1704; GP II 270). This action is to be defined as force multiplied by the duration of its activity. The momentary state seems to be something statical and therefore it seems not apt to regard it as a sort of force, which is essentially dynamic. To avoid this conclusion, it is important to conceive of the derivative force as the momentary state in relation to the following one (momentaneum est, sed cum relatione ad statum sequentem). It is the momentary state insofar as it tends to the future one. This very striving constitutes force as a disposition to change that (without hindrance) by itself causes this change. Derivative force, however, being a momentary state directed to change, does not qualify as a primitive entity subsisting by itself, neither does physical action because it is the temporal succession of these changing states. Rather, derivative force, being a modification, that is a transitory, particular manifestation, must be related to a permanent underlying force. The letter to de Volder dated Jan. 21, 1704 contains a remarkable proposal for how to understand this primitive force. The derivative force is the present state itself insofar as it tends toward the succeeding one or involves it in advance (ipse status paesens dum tendit ad sequentem seu sequentem praeinvolvit) as everything present is pregnant with the future development. But the persistent itself, insofar as it includes all cases, possesses a primitive force. Therefore, primitive force is, as it were, a law of series; derivative force is, so to speak, a determination marking a limit (terminus) in the series. (GP II 262)

When Leibniz shortly before speaks of forces which “are active and nevertheless (et tamen) modifications,” he wants to say that between these two conceptions there is, at least superficially regarded, a contradiction. For a modification15 is a transitory state within a process of change or even (as Leibniz puts it) a limit (terminus), that is an infinitesimal moment of motion, or (more generally) of some change. This limiting moment within the course of motion is represented by the present moment. Already Augustine conceived of the presence as the unextended limit of past and future, memory and expectation. But how can an unextended moment be active in view of the fact that causal efficacy, as well as the change to be effected, obviously presuppose duration? Leibniz’s solution comprises two thoughts: First, the modification is the transitory particular manifestation of a permanently underlying active principle. Even more important, second, is the notion of striving. As striving always means a tending toward a future state (tendere ad), the present (unextended) moment of striving can anticipate the intended future state (prae-involvere). It, thus, includes the (temporally extended) phase of the transition from one state to another and it can in this manner cause the temporal process of change from one to the other.

_____________ 15

Lodge 2001 thinks that derivative forces (qua phenomena, that is the intentional objects of perceptual states) can be regarded as modifications of primitive forces (the internal tendencies of perceiving substances) only if a modification is understood as not necessarily inhering its subject.

Leibniz’s Theory of Force

115

The phrase “The present state is pregnant with the future one” (praesens gravidum futuro) is characteristic of the manner in which Leibniz conceives of an individual which, in the unextended present moment, contains already its whole history in form of traces of everything that has happened and marks of everything that will happen (DM §§ 8 and 13). A striving, however, is directed only toward a single future state and so belongs only to the level of derivative forces. The concept of an individual, on the other hand, which determines the entire future development, corresponds to primary force. Primary force (as we have seen) is a metaphysical factor constituting substance, which is exactly the role of the concept of an individual. By characterizing primitive force as a law of series, Leibniz expresses the thesis that it determines and rules the whole succession of states. For that reason, knowledge of primitive force allows us to deduce from the present state the entire future development, as does the concept of an individual. If we want to express the dynamic moment of striving missing in the statical conception of a law of series, we can interpret primitive force as the basic tendency of the striving characterizing individuality. Just this conception is articulated in a later non-dated letter to de Volder (GP II 275) where the primitive forces are considered as internal tendencies of simple substances by means of which these substances pass from perception to perception according to a certain law of their (individual) nature. This basic direction of the striving confers to the succession of states a characteristic peculiar to this individual. It thus constitutes an individual substance in its individuality.16 Derivative force, as a momentary striving toward the succeeding state, then can plausibly be seen as modification, that is, a particular transitory manifestation of this primitive force in its permanent basic direction. The interpretation of primitive active force as a law indicates a deterministic view.17 For a determinist, it is a central question of how it is possible

_____________ 16

17

This interpretation shows how the following thesis by Bobro & Clatterbaugh 1996 can be avoided. The monadic agency view (that an active force, inhering each monad, causes its perceptual changes) is inconsistent with each of the two other views, for which there is textual evidence in Leibniz: the conceptual unfolding view and the efficacious perception view (that perceptual change is caused by the individual concept or the preceding perception respectively). It is more to the point to regard these three aspects as complementary to each other: The individual concept is a substance’s basic characteristic from which not only all its predicates can be derived, but which as a basic law inheres primitive force, so that this force leads to a striving according to a certain inner tendency. The preceding perceptual state’s being directed to the following one in form of a striving is a moment in the whole process of determination which consists in primitive active force’s determining the whole sequence of perceptions according to a certain fundamental tendency. Leibniz treats forces and dispositions in the context of determination. This is not only implied by the fact that he conceives of primitive force as the law of series which determines the whole succession of states; in NE I 1, § 11 (A VI 6, 80) and Theod. I § 46 (GP VI 128) (see sec. II above), he explicitly speaks of determination. Determination is brought about by the system of laws of

116

Michael-Thomas Liske

that the whole future development is already contained in the present moment. As a means to tackle this problem, the concept of force as a disposition amplified by a striving suggests itself. Already derivative force provides an explanation of how the present state contains the subsequent one, namely by containing a striving and so being directed toward the future state. Primitive active force corresponds to the deterministic laws ruling the whole sequence of states. One aspect of this is that the course of events is ruled by unchanging laws, the other aspect is that it is ruled dynamically by a striving that tends not only to the subsequent state, but entails a whole sequence of states characterized by a certain basic direction or tendency. A further aspect of the equation of laws with forces is contained in De ipsa natura (1698), which Leibniz wrote as a defence of his dynamics against the occasionalist theory of motion.18 If one wants to prevent independent created substances from vanishing into mere Spinozistic modifications of the one divine substance (GP IV 508sq.), then the laws of motion must be understood as being immanent to the corporeal substances in motion, that is their internal natures, or efficient forces (efficacia, vis) that cause the whole series of phenomena of motion in an ordered way or according to a law (507).

5. Vis Mortua is the infinitesimal potential Energy of the Moment; Living Force (Vis Viva) is kinetic Energy arising in the course of Movement and enabling an Effect of a measurable Quantity We already encountered the typical hypothesis associated with Leibniz’s theory of force: Even in an infinitely short period of time, when movement in no measurable quantity takes place, a force can be present, namely in the form of

_____________

18

one world (for instance the real one). Concerning the modal status of dispositions and forces, this means: They have to be considered only within the actual world and do not include relations between several possible worlds. By means of these transworld relations, Leibniz on the contrary tries to account for counterfactual alternatives and so to weaken determination in order to avoid necessity. Leibniz himself expresses this in his letter to Remond, dated June 22, 1715: His dynamics serves as an important basis of his entire system, for it makes clear the difference between truths whose necessity is brute and geometrical, and contingent truths originating from fitness (convenance) and final causes (GP III 645). This also means: If Leibniz had stuck to the geometrical approach of Descartes’ pure rational mechanics, the basic laws of motion would be derivable a priori from extension as the essence of matter and so would hold necessarily in all possible worlds. Only by founding them on dynamics, that is, on the notion of force, is he able to guarantee their contingency. See D. M. Wilson 1976. This is relevant for the following reason: In order to overcome Hume’s thesis that there are no connections other than relations of ideas, it must be acknowledged that there are not only de re necessary connections but also other types of real connection, for which a Leibnizian force constituted by a striving is a remarkable candidate. See Schrenk in this volume. See C. Wilson 1987, esp. 165sq.

Leibniz’s Theory of Force

117

a striving disposing to an activity, provided all hindrances are removed. Leibniz makes use of this idea especially in order to define vis mortua in contrast to vis viva. In contemporary physical terminology, these two sorts of derivative forces would be called potential versus kinetic energy. The concept of potential energy includes the philosophical conception of potency, or power to act. Here, Leibniz’s amplified conception of potency appears again: Although this force is called ‘dead’ insofar as there is not any actual movement in it yet (quam et mortuam appello quia in ea nondum existit motus), Leibniz does not take it to be entirely at rest. Rather, he supposes it to include a solicitation to motion, that is, a sort of striving (see SD I p. 12/14, l. 166-177, esp. 166-168; GM VI 238). Examples of vis mortua are: centrifugal force, when a stone is thrown around in a sling, or centripetal force, like gravitation, when a body is attracted by a bigger one, for instance by the earth, or the elastic force of a stretched body; if motion in each case is still hindered, that is, if the stone cannot yet start flying away, if the heavy body cannot yet fall to earth or the stretched elastic object (for instance the stretched bow) cannot yet restore its original shape (l. 171-173). When this motion actually takes place, the energy that for instance the stone gains while falling down is living force (l. 170 sq., 173-176). Presumably, one of the reasons why it is called ‘living’ is that it can bring about a certain effect of a measurable quantity, for instance to elevate another body to a certain height (aut agitur de absoluto quodam effectu producto, qualis est grave elevare ad datam altitudinem, GP II 154).19 Vis viva itself must possess a measurable finite quantity. Vis mortua, in contrast, is momentary and so has an infinitely small quantity: be it that we take it as solicitation, that is, a body’s striving for motion while its motion is still impeded; be it that we take it as conatus, that is, the momentary velocity either in the first moment of a heavy body’s fall or in another phase of its motion whose duration tends to the limit zero (vires mortuae quales habet conatus primus gravis descendentis aut qui quovis momento acquiritur, GP II 154). Consequently, vis viva, or its impetus, comes about by infinitely many solicitations of vis mortua summing up in a continuous succession (vis est viva, ex infinitis vis mortuae impressionibus continuatis nata, SD I l. 175-177; nam impetus continuato solicitationum cremento formatur, GP II 154). Yet, it is not possible to uphold without qualification the distinction between resting (dead) potential energy, being a momentary infinitesimal striving for motion, and kinetic energy, being the force accumulated in the course of motion and bringing about an effect of a measurable quantity. This can be seen from Leibniz’s controversy with the Cartesians on vis viva where he tries to show: The decisive physical quantity that is conserved in motion is not the quantity of motion (mass multiplied by velocity), but force (mass multiplied

_____________ 19

First letter to de Volder delivered 16th Jan. 1699, a text which I use as a parallel to SD.

118

Michael-Thomas Liske

by velocity squared).20 Obviously, this law of conservation refers to vis viva, which is measured by the quantity of the effect the bearer of this force is able to bring about. One of Leibniz’s premises explicitly refers to the quantity of effect: The force required to elevate a body of four pounds to the height of one foot equals the force required to elevate a body of one pound to the height of four feet. This force surely can be taken to be the potency to bring about an effect of a certain quantity, or the kinetic energy which a body acquires in the course of its motion (for instance the downwards motion of the pendulum). However, this also can be described as potential energy, or the disposition the body possesses because of its mass and its height. When this potential energy is being transformed into kinetic energy during the fall, Galileo’s laws of free fall become relevant. According to these, not velocity but velocity squared increases in proportion to the distance the body falls. For the first body (of one pound) to acquire the same force as the second body (of four pounds), it has to fall four times the distance of the second body (which is the consequence of the above premise concerning the height of elevating when applied to the height of fall). Falling the fourfold distance, however, the body acquires, according the laws of free fall, only the double velocity. If the first body possesses the same force as the second (1x4 = 4x1), it has a lesser quantity of motion (1x2 < 4x1). If the force is to be conserved, the quantity of motion cannot be conserved.

6. Realism versus Phenomenalism. Are primitive Forces metaphysical Principles of corporeal Substances or do they constitute Monads: mental Entities which are the only real Entities? Leibniz is undecided about the question whether there are genuine corporeal substances or whether only the perceiving and striving monads, or mental entities, are properly real substantial unities and everything else is a mere phenomenon. He adopts a similarly ambivalent attitude toward the question of

_____________ 20

See Brevis Demonstratio erroris memorabilis Cartesii … circa legem naturalem … A VI 4, 2027-2030; DM § 17, A VI 4, 1556-1558; letter to Bayle (GP III 42-49). On what Leibniz intended to prove in his Brevis Demonstratio, see Brown 1984. Gale 1979 tries to integrate this argumentation into the development of Leibniz’s doctrine of dynamics, which is deeply influenced by metaphysical assumptions. Papineau 1977 discusses the argument of Brevis Demonstratio on the background of the contemporary vis viva controversy. Already in Brevis Demonstratio the opposition of statics and dynamics becomes relevant when Leibniz identifies as the source of Descartes’ error that he mistakenly generalizes what is true in statics and applies it to dynamics. See Freudenthal 2002, 582 sq. Freudenthal’s main subject is the controversy between Leibniz and Papin, who critizises the Brevis Demonstratio. According to Kvasz 2001, Leibniz’s demonstration is false from a scientific point of view because he neglects the motion of the earth.

Leibniz’s Theory of Force

119

which sort of entities primitive forces are to be ascribed to. He is fairly sure that the unity of an individual substance is constituted by the interaction of primitive forces: one active, striving forward and one passive, hindering. What, however, can be a candidate for such a substance in the proper sense that is to be constituted by those primitive forces? In Specimen Dynamicum, Leibniz ascribes primitive metaphysical force to corporeal substance. In contrast to the derivative forces, which act in motions and thus interactions of bodies, he defines it as inner principle essentially inhering every corporeal substance as such (primitiva quae in omni substantia corporea per se inest, SD I p. 6, l. 70 sq.; GM VI 236). In virtue of this inner principle of activity, each corporeal substance is essentially and always active (at least in the form of inner motion). It never is totally at rest. By acknowledging them as real, he can ascribe a primitive active force or an inner striving to corporeal substances as metaphysical ground of their reality (although their appearance as bodies is a phenomenon). We encountered this position already in the passage from Leibniz quoted at the beginning of section I: If, in this passage, he acknowledges a corporeal entity like an animal to be a genuine substance, this means: It is essentially one, its original unity stemming from an inner principle (unum per se, GP IV 395). For this reason, he ascribes an active force to it, which is an inner principle insofar as it is the disposition to an activity characteristic and essential to the animal as a whole. This activity constitutes the unitary way of existing of such a substance.21 Things are different with the strictly phenomenalistic view to be found in the correspondence with de Volder. In the draft of a letter dated June 19, 1706, Leibniz expresses clearly his phenomenalistic position regarding bodies. There is nothing properly real besides the perceiving simple substances, their perceptions and the relations between them: the striving to ever clearer perceptions by a subject remaining identical, and the harmony obtaining between the perceptions of different perceivers (GP II 281). As phenomena result from an underlying reality and, therefore, are derivative, primitive forces cannot belong to corporeal substances but only to mental entities, the perceiving monads. So Leibniz, in the letter dated 1703 (GP II esp. 250-252), says clearly: Entelechy as primitive active force cannot properly impel the mass of the body; rather, it is joined with the primitive passive potency, which is the force complementary to it, and in this manner constitutes the monad (250), which is a perceiving, or purely mental entity. Correspondingly, the later draft says: Active force and passive force are to be found in the perceiving subject: active force consists in the passage toward that which is more perfect, passive

_____________ 21

This is the prevailing view which can be found in many texts, for instance in Conversation of Philarète and Ariste. Here, corporeal substance is composed of two natures: primitive active force (first entelechy) and matter as primitive passive force, which manifests itself as impenetrability (GP VI 588).

120

Michael-Thomas Liske

force consists in the contrary. Primitive active force and primitive passive force have their domain in the mental activity of perceiving. Active force, striving forward, aims at the passage toward more perfect perceptions, that is, perceptions becoming ever clearer and more distinct. The passive force, which hinders and modifies the superior active force, has the effect that many perceptions remain unclear and confused. Both taken together constitute the monad’s individuality. Hindering passive force is indispensable because the differences between individual monads are grounded in the fact that each monad perceives only few things clearly and distinctly and has access to the rest by means of what it clearly and distinctly perceives. The latter constitutes its individual point of view and its individuality. Of the five ontological stages (252) only the first three, which belong to the mental, constitute reality in the proper sense, namely: primary active force (soul, entelechy), primary passive force (materia prima), and the monad constituted by them. When innumerable subordinate monads, whose passive forces hinder clear and distinct perceptions, come together, the phenomenon of body results (stage 4). Body is nothing real, but only the way an aggregate of subordinate monads appear to us. Even if they are dominated by a higher monad, or soul, the animal thus constituted nevertheless is a phenomenon (stage 5). Acordingly, in corporeal substances there can be no primitive forces disposing the bodies to a characteristic unifying activity and thus constituting them as substances. While active forces convincingly can be understood both as belonging to the activity of monads and as being the principle of unity of corporeal substances, in the case of passive forces, or matter, the phenomenalistic view proves to be more conclusive. It can separate unequivocally the primary passive forces, which hinder and modify the activity of the monads, (materia prima) from the derivative ones, which account for impenetrability and inertia of the corporeal mass (materia secunda). If, however, (as in SD I 8, l. 94-100, GM VI 236 sq.) primitive passive force, or materia prima, is seen as being impenetrability and inertia of bodies, one is left with an unsatisfactory distinction: Impenetrability and inertia, taken as materia prima, constitute the materiality of the body as such; taken as materia secunda, they determine the concrete movements of particular bodies.22

7. Conclusion: The Concept of Force shows how physical and metaphysical Enquiries gain from each other Leibniz’s theory of force is a good example of how explorations of issues in different domains of investigations that he pursued at the same time – an

_____________ 22

See Allen 1984, 60.

Leibniz’s Theory of Force

121

expression of his encyclopaedic interests – became fruitful for each other. The rationalist conception of substance as an entirely autonomous system subsisting solely by itself as well as the logical assumption that every predicate true of a subject is analytically contained in the subject’s concept, lead to the following metaphysical postulate: An inner principle is by itself sufficient to bring about the characteristic activity of a substance. It is the disposition to a certain way of acting that performs the function of a principle persistently underlying the changing actions and being the essence of the substance. However, a disposition, being a developed potency, is not sufficient, but substance must, if it is to turn itself into actualization (as is required by its autonomy) involve a striving.23 This type of potency, amplified by encompassing a striving, is most aptly identified as force: a concept Leibniz discovered at the same time in the context of his physical enquiries through critically discussing Descartes’ laws of motion. If one follows Descartes in defining the essence of body as extension and so tries to describe motion in exclusively geometrical categories as change in a body’s location, one is led to formulate empirically inadequate laws of motion. Leibniz and Descartes agree that it is force that is conserved during mechanical motion and proves to be the central physical quantity. But the controversial question is: What is the true measure of force? In this context, it is Leibniz’s central conviction that force is to be measured by its effect. Therefore, it must be defined as the disposition to bring forth an effect of a certain quantity, that is, to raise a body of a certain mass to a certain height. This means according to Galileo’s laws of free fall: it is not the product of mass and velocity (Descartes) but the product of mass and velocity squared.24 The concept of force implies that it is directed toward producing a certain effect, which means that it is a potency amplified by a tending or striving for efficacy. When Leibniz, in his Brevis Demonstratio (1686), arrived at the concept of force in a physical context through his critique of Cartesian

_____________ 23

24

Duchesneau 1994, 214 speaks of dynamic dispositions insofar as he understands the power to act as a disposition spontaneously tending to accomplish itself (or rather: tending to spontaneously accomplish itself; for a tending is presupposed by the power’s spontaneously actualizing itself rather than being actualized by external conditions). This understanding of Leibnizian force certainly is essentially correct. From a terminological point of view, however, Duchesneau’s use of ‘disposition’ is inconsistent both with the present use, according to which a disposition must be actualized by suitable external conditions, and with Leibniz’s own use in NE II 1, § 2 (as we have examined in II above). One should better say: The tendency to transfer itself into actualization, which (if nothing hinders) leads to a spontaneous actualization, amplifies disposition by something which is not yet implied in the concept of ‘disposition,’ although according to Leibniz, in reality there are never mere dispostitions but always amplified dispositions. On the development of this theory see Gale 1988. That this argumentation, often repeated since the Brevis Demonstratio, is indeed physical in nature can be seen from the fact that Leibniz does not only use a priori principles (equipollency of entire cause and total effect; force is to be measured by the effect) (see Gueroult 1962, ch. V, esp. 117), but also empirical generalizations such as Galileo’s laws of free fall (see Duchesneau 1998, esp. 79-81).

122

Michael-Thomas Liske

mechanics, he was not yet aware of what can be achieved with this concept in the context of metaphysics (as the brevity of the exposition shows). All the differentiations of the concept of force we discussed in the previous sections in the context of his fully developed dynamics, e. g. in Specimen dynamicum, are still absent.25 In his first approach, Leibniz (following common usage) understood force only as active force and as being directed to acting. It took Leibniz a further step to become aware that the passivity of matter, too, can be understood as force insofar as it manifests itself in resisting the prevailing force, thus limiting it. Further, he was confronted with the need to distinguish the type of force assumed for explaining observed physical phenomena from the metaphysical type of force.26 For this purpose, Leibniz made use of the opposition of ‘instantaneous’ versus ‘permanent’: The physical force is the momentary state insofar as it tends to the following one, and so explains the transition from one state to the other, that is the observable change. The metaphysical force, on the other hand, must constitute substance as a permanent entity. Therefore, Leibniz takes metaphysical force to be the basic law that confers to the whole sequence of states (which should be understood in this context as mental states of perception) a character peculiar to the individual to be constituted. Regarded from a slightly different point of view: The systematic place of the concept of force within Leibniz’s philosophy is the intersection between physics and metaphysics. We probably are inclined to relate force rather to physical motions. If Leibniz, on the other hand, takes metaphysical force to be the primary type of force, this is because the concept of force was capable of performing an explanatory function seriously required for his metaphysical approach. He was searching for a candidate to fulfill this function for quite a long time until the physical concept of force proved to be apt. This can be seen from the following: Phrases that occur in the context of the problem of force after 1686, when the concept of force begins to gain importance in Leibniz’s philosophical thought, can be found related to other concepts already in the decade before. At that time, Leibniz was able to express precisely what needed to be explained without having yet developed the conceptual framework with the help of which it could be explained. In his De affectibus

_____________ 25 26

On Leibniz’s initial conception of force and the ensuing differentiation, see Boudri 2002, ch. 3.2 and 3.3, 75-91. Gale 1970 makes a threefold distinction of entities and their properties: those on the observable level, those on the explanatory level of Leibniz’s physical theory and those on the metaphysical level. However, the derivative forces should not be viewed as properties of observable entities. For even the forces acting in observable movements cannot themselves be immediately observed and their existence is denied by strict empiricists of the Humean tradition. Forces are postulated and can only be justified as theoretical entities that provide the most complete and coherent explanation of observable facts. This epistemological tripartition is discussed in Duchesneau 1994, 215-217.

Leibniz’s Theory of Force

123

from April 1679 (A VI 4, 1410-1441), for example, he conceives of determination in terms of a formula that later became a central characterization of force (as an amplified potency), namely that the effect immediately follows from it if nothing hinders. He defines determination as a state from which, taken by itself, something follows if nothing else occurs which impedes it. (A VI 4, 1429 l. 20f., see also 1426 l. 20, 1427 l. 15 f., 1428 l. 8, 1429 l. 9, 1430 l. 14, 1436 l. 7). For Leibniz as a determinist, it is important to find factors that determine events to the course that afterwards becomes real. What exactly is it that brings about the determination? In the period before 1686, Leibniz considers actio.27 He relates the phrase ‘state from which [something] follows’ (status ex quo sequitur) not only to determination, but also to action by defining action as a state of something from which a change in another thing immediately follows (A VI 4, 28 l. 11, 308 l. 5) and which can be considered the immediate cause (causa proxima) of this change (A VI 4, 1411 l. 8). This sequence should not be understood as a temporal relation, but is to be conceived of as being based on logical priorities of inference and ontological priorities of dependence which also can be given in simultaneous events: that which follows is later in the order of things (natura posterius); its existence can be ascertained and explained from the existence of that which is earlier in the order of things (A VI 4, 1427 l. 17 sq., 1436 l. 9 f.). Accordingly, Leibniz explicitly says that the action is called cause insofar as it is the state from which a present, i.e. simultaneously occurring change (mutationis alicujuis praesentis), can be explained and accounted for (A VI 4, 1412 l. 14f.). The action which, as a cause, accounts for a simultaneously occurring change certainly is an important factor of determination but as a mere part of it (pars determinationis, A VI 4, 1428 l. 11), it requires also other moments, namely temporal ones. This is true especially because an essential aspect of determination is predetermination, which means that events are already preformed in previous states and follow from them according to the laws of nature. Leibniz, as a proponent of substance ontology, decidedly endorses the concept of what is nowadays called agent causality. Even if we immediately observe events or actions as the preceding state after which a change follows, these events are nothing isolated, free-floating, but must be related to an acting subject, an agens, which produces these effects and is their principle.28 Accordingly, the causally relevant specifications of actio are, properly understood, to be ascribed to the agens. It is the agens from whose state a change in another thing follows (A VI 4, 153 l. 21, 305 l. 22) and which therefore is its proper cause (A VI 4, 393 l. 16). The

_____________ 27 28

On Leibniz’s conception of actio and agens in the period leading to the Discours de Métaphysique, see Schneider 2001. In the Discourse of metaphysics, Leibniz expresses the thesis that actions properly are to be attributed to individual substances, according to the scholastic tenet “actiones sunt suppositorum.” On this tenet and Leibniz’s general theory of activity, see McGuire 1976, s. II.

124

Michael-Thomas Liske

consequential question, which Leibniz, however, did not yet pose at the time, is: In virtue of which kind of property is an acting substance (or agent) determined to produce by its agency a change of state in another thing? This question consequentially leads to the concept of active force as a disposition amplified by a striving. Only such amplification determines the disposition to a certain activity, namely to the one the striving is directed to; the activity follows if nothing external impedes it. Leibniz’s metaphysical assumption that a substance is an entirely autonomous system, which produces all of its activities spontaneously and solely from itself, may be quite problematic. But the concept of force (developed in this context) as a disposition that in virtue of a striving is directed toward a certain activity and thus turns itself into the corresponding action (if nothing hinders) proves to be fruitful in explaining even physical observations. It is remarkable that the determining laws are taken to be forces, that is, dispositions that by virtue of a striving are directed toward a certain sequence of events and that determine in this manner the actual course of events. The conception of laws of nature as dispositional properties deserves serious philosophical consideration – far beyond the specific assumptions of Leibnizian metaphysics, even apart form the striving implied in the concept of force – provided that any lawful behavior of a system can be interpreted as the system’s property. Taking into account the fact that laws of nature are idealizations, one has to say: What is ascribed to the system according to this understanding is not the categorical property to actually behave in a certain way, but the disposition to react in such and such a way under ideal circumstances. Such a disposition cannot be immediately observed but has to be introduced by extrapolation: for, if the actual conditions get closer to the ideal ones, the behavior of the system, too, approaches the type of behavior hypothetically assumed in the law.29 So far, talk of dispositions in the context of natural laws concerns disposition in general, not force as a disposition amplified by a striving. If the concept of disposition is, however, to achieve its full potential for explaining natural phenomena, one must be able to explain the following distinction with the help of this concept: the difference between situations where the disposition exists but remains inactive because the triggering conditions are not given, and situations where the disposition is activated by adequate conditions but where the manifestation of its activity is prevented or at least limited by antidotes. For us to be able to give a satisfying account of this difference, we must postulate forces.30 An activated disposition (as opposed to an inac-

_____________ 29 30

See also Hüttemann’s paper in this volume. See also Schrenk’s paper in this volume.

Leibniz’s Theory of Force

125

tive one) would be characterized by a striving for a certain activity; the effect of this striving, however, can be prevented or limited. Leibniz interprets this situation somewhat differently: According to his principle of autonomy of substances, it must not depend on external conditions that a disposition becomes active in form of a striving. Rather, a disposition must always be striving from an inner source. An entirely inactive disposition is a pure fiction. Accordingly, a potency always strives and always brings about some form of activity. The striving and the activity, however, occur in highly different degrees.

Literature A VI 4

A VI 6 GM GP SD

Leibniz, Gottfried Wilhelm. 1999. “Philosophische Schriften 1677-Juni 90.” In Sämtliche Schriften und Briefe. Sechste Reihe: Philosophische Schriften, Bd. 4, ed. Berlin-Brandenburgische Akademie der Wissenschaften zu Berlin und Akademie der Wissenschaften in Göttingen. Berlin: Akademie Verlag. Leibniz, Gottfried Wilhelm. 1962. “Nouveaux essais.” In Sämtliche Schriften und Briefe. Sechste Reihe: Philosophische Schriften, Bd. 6, ed. Deutsche Akademie der Wissenschaften zu Berlin: Akademie Verlag. Leibniz, Gottfried Wilhelm. 1962. “Leibnizens mathematische Schriften.” Bd. I-VII, ed. Carl Immanuel Gerhardt. Hildesheim: Georg Olms. Leibniz, Gottfried Wilhelm. 1978. “Die philosophischen Schriften von G. W. Leibniz.” Bd. I-VII, ed. Carl Immanuel Gerhardt. Hildesheim: Georg Olms. Leibniz, Gottfried Wilhelm. 1982. Specimen Dynamicum. Hamburg: Felix Meiner.

Allen, Diogenes. 1984. “From vis viva to Primary Force in Matter.” In Leibniz’ Dynamica (= Studia Leibnitiana Sonderheft 13), ed. A. Heinekamp, 55-61. Stuttgart: Steiner. Bergmann, Markus. 2002. Unendlicher Panpsychismus. Kraft und Substanz in der Philosophie des Individuums von Leibniz. Inauguraldissertation der Johannes Gutenberg-Universität Mainz. Bernstein, Howard R. 1981. “Passivity and Inertia in Leibniz’s Dynamics.” Studia Leibnitiana 13: 97-113. Bobro, M. & Clatterbaugh, K. 1996. “Unpacking the Monad: Leibniz’s Theory of Causality.” The Monist 79: 408-425. Boudri, J. Christiaan. 2002. What was Mechanical about Mechanics. The Concept of Force between Metaphysics and Mechanics from Newton to Lagrange (= Boston Studies in the Philosophy of Science 224). Dordrecht/Boston/London: Kluwer. Brown, Gregory. 1984. “‘Quod ostendendum susceperamus’. What did Leibniz undertake to show in the Brevis Demonstratio?” Studia Leibnitiana, Sonderheft 13: 122-137. Duchesneau, François. 1994. La dynamique de Leibniz. Paris: J. Vrin Duchesneau, François. 1998. “Leibniz’s Theoretical Shift in the Phoranomus and Dynamica de Potentia.” Perspectives on Science 6: 77-109. Freudenthal, Gideon. 2002. “Perpetuum mobile: the Leibniz-Papin controvery.” Studies in the History and Philosophy of Science 33: 573-637.

126

Michael-Thomas Liske

Gabbey, Alan. 1971. “Force and Inertia in Seventeenth-Century Dynamics.” Studies in the History and Philosophy of Science 2: 1-67. Gale, George. 1970. “The Physical Theory of Leibniz.” Studia Leibnitiana 2: 114-127. Gale, George. 1973. “Leibniz’ Dynamical Metaphysics and the Origins of the vis viva Controvery.” Systematics 11: 184-207. Gale, George. 1988. “The Concept of ‘Force’ and Its Role in the Genesis of Leibniz’s Dynamical Viewpoint.” Journal of the History of Philosophy 26: 45-67. Garber, Daniel. 1985. “Leibniz and the Foundation of Physics: The Middle Years.” In The Natural Philosophy of Leibniz, ed. K. Okruhlik and J. R. Brown, 27-130. Dordrecht: D. Reidel Publishing Company. Gueroult, Martial. 21962. Leibniz. Dynamique et métaphysique. Paris: Aubier-Montaigne. Hacking, Ian. 1985. “Why Motion is Only a Well-Founded Phenomenon.” In The Natural Philosophy of Leibniz, ed. K. Okruhlik and J. R. Brown, 131-150. Dordrecht: D. Reidel Publishing Company. Kangro, Hans. 1969. “Der Begriff der physikalischen Größe, insbesondere der action motrice, bei Leibniz.” Studia Leibnitiana Supplementa 2: 133-149. Kvasz, Ladislav. 2001. “Leibniz’s criticism of the Cartesian physics.” In Nihil sine ratione. VII. Internationaler Leibniz-Kongreß. Vorträge 2. Teil, ed. H. Poser, 669-676. Lodge, Paul. 2001. “Primitive and Derivative Forces in Leibnizian Bodies.” In Nihil sine ratione. VII. Internationaler Leibniz-Kongreß. Vorträge 2. Teil, ed. H. Poser, 720-727. McGuire, J. E. 1976. “‘Labyrinthus Continui’: Leibniz on Substance, Activity, and Matter.” In Motion and Time, Space and Matter: Interrelations in the History of Philosophy and Science, ed. Peter K. Machamer and Robert G. Turnbull, 290-326. Ohio: Ohio State University Press. Papineau, David. 1977. “The Vis Viva Controversy: Do Meanings Matter?” Studies in the History and Philosophy of Science 8: 111-141. Schneider, Martin. 2001. “Leibniz’ Theorie der Aktion im Jahrzehnt vor dem Discours de métaphysique 1677-1686.” Studia Leibnitiana 33: 98-121. Wilson, Catherine. 1987. “De Ipsa Natura. Sources of Leibniz’s Doctrines of Force, Activity and Natural Law.” Studia Leibnitiana 19: 148-172. Wilson, Margaret D. 1976. “Leibniz’s Dynamics and Contingency in Nature.” In Motion and Time, Space and Matter: Interrelations in the History of Philosophy and Science, ed. Peter K. Machamer and Robert G. Turnbull, 264-289. Ohio: Ohio State University Press.

From Ordinary Language to the Metaphysics of Dispositions Gilbert Ryle on Disposition Talk and Dispositions1 OLIVER R. SCHOLZ

1. Context: When, Where and Why In order to do justice to Gilbert Ryle’s contribution to the topic of this anthology, it is important to be clear about the historical and argumentative context in which he discusses dispositions. The locus classicus of Ryle’s treatment is The Concept of Mind, first published in 1949. To be sure, there are adumbrations of the basic idea in the paper “Knowing How and Knowing That” read some years before to the Aristotelian Society. But The Concept of Mind has a whole chapter on “Dispositions and Occurrences” (Chapter V) along with an improved version of the material on “Knowing How” (Chapter II). What is more, the disposition-occurrence opposition permeates the whole book. The Concept of Mind is well-known as a sustained attack on the Cartesian Myth, also called: the dogma of the Ghost in the Machine, or in more literal terms: mind-body substance dualism. The disposition-occurrence opposition is one of the major weapons in this fight. It is less well-known that The Concept of Mind is also intended as the manifesto, exemplification and test of a new philosophical method. It is the first book-length application of the special brand of “ordinary language philosophy”2 which was developed by Ryle and some of his Oxford comrades in the 1930s. (The greatest intellectual debt is, of course, to his Cambridge colleague Ludwig Wittgenstein.3) The question “What is the nature and method of philosophy?” or, as he puts it in his autobiography, the question “What consti-

1

2 3

I would like to express my gratitude to the organizers and participants of the Wittenberg conference for many stimulating discussions. In addition, I want to thank Rosemarie Rheinwald, Richard Schantz and Ludger Jansen for their comments on one of the final versions. Special thanks are due to Rudolf Owen Müllan for correcting my English. The best book on this movement, its philosophical methods and achievements is von Savigny 1969 (on Ryle see especially Chapter 2). Cf. Hacker 1996, 168-72; Place 1999, 364-69.

128

Oliver R. Scholz

tutes a philosophical problem; and what is the way to solve it?”4 occupied Ryle from the beginning to the end of his career, from the early papers “Systematically Misleading Expressions” (1931-32) and “Categories” (1938) up to the Tarner Lectures of 1954 (published under the title Dilemmas) and beyond. In his short autobiography, he gives a lively description of the historical background: “In the 1920s and the 1930s there was welling up the problem ‘What, if anything, is philosophy?’ No longer could we pretend that philosophy differed from physics, chemistry and biology by studying mental as opposed to material phenomena. We could no longer boast or confess that we were unexperimental psychologists. Hence we were beset by the temptation to look for non-mental, non-material objects – or Objects – which should be, for philosophy, what beetles and butterflies were for entomology. Platonic Forms, Propositions, Intentional Objects, Logical Objects, perhaps, sometimes, even Sense-Data were recruited to appease our philosophical hankerings to have a subject-matter of our own./ I had learned, chiefly from the Tractatus logico-philosophicus, that no specifications of a proprietary subject-matter could yield the right answer, or even the right sort of answer to the original question ‘What is philosophy?’ [...] Philosophical problems are problems of a special sort; they are not problems of an ordinary sort about special entities.”5

The philosopher’s task is to detect and avoid linguistic confusion: “My interest was in the theory of Meanings – horrid substantive! – and quite soon, I am glad to say in the theory of its senior partner, Nonsense.”6 In contrast with the lexicographer, “the philosopher’s proprietary question is not ‘What does this or that expression mean?’ but ‘Why does this or that expression make nonsense? and what sort of nonsense does it make?’”7

Thus, Ryle decided to become a No-nonsense philosopher. In his search for allies, he read Bertrand Russell, Ludwig Wittgenstein, Edmund Husserl as well as other phenomenologists and the logicial empiricists, especially Rudolf Carnap.8 The young Alfred Ayer was sent to Vienna as a spy and returned with his notorious anti-metaphysical manifesto Language, Truth, and Logic, first published in 1936. In “Systematically Misleading Expressions” (1931-32), one of the first manifestos of Oxford-style ordinary language philosophy, Ryle tried to show that there are many expressions “which are [...] systematically misleading, that is to say, that they are couched in a syntactical form improper to the facts recorded and proper to facts of quite another logical form than the facts re4 5 6 7 8

Ryle 1970, 12. Ryle 1971b, vii. Ryle 1970, 7. Ryle 1970, 6sq. In the 1940s, it came to a parting of the ways. In a review of Meaning and Necessity, Carnap’s philosophical method and his theory of meaning are criticized in an unusually harsh tone. See Ryle 1945b.

From Ordinary Language to the Metaphysics of Dispositions

129

corded.”9 The main task of philosophy is to detect, correct and avoid confusions caused by such systematically misleading expressions. In “Categories” (1938), Ryle claimed, in a similar vein, that “we are in the dark about the nature of philosophical problems and methods if we are in the dark about types or categories.”10 In regard to his magnum opus The Concept of Mind, Ryle reveals that “[...] by the later 1940s it was time, I thought, to exhibit a sustained piece of analytical hatchet-work being directed upon some notorious and large-sized Gordian knot. After a long spell of enlightened methodological talk, what was needed now was an example of the method really working, in breadth and depth and where it was really needed./ For a time I thought of the problem of the Freedom of the Will as the most suitable Gordian knot; but in the end I opted for the Concept of Mind [...].” Ryle emphasized: “The Concept of Mind was a philosophical book with a meta-philosophical purpose.”11 In a nutshell: in The Concept of Mind Ryle brought together two ideas: (i) the suspicion that philosophical problems are the result of mistakes of a special sort, namely category mistakes, (ii) and the idea, already in the air for some time, that some (maybe all?) mental phenomena can be construed as dispositions, rather than as “ghostly” inner occurrences.12 To sum up, Ryle’s magnum opus The Concept of Mind is first and foremost a meta-philosophical journeyman’s work intended to show the strengths of the new philosophical method with a prominent example, one of the grand old philosophical problems. The choice of the example was of secondary importance, though by no means completely arbitrary; first Ryle thought of the problem of free will, but in the end he decided on the mind-body problem. Thus, in the second place, The Concept of Mind is intended as a reductio ad absurdum of mind-body substance dualism, the disposition-occurrence opposition being one of the major weapons against the Cartesian Myth. As we shall see, Ryle’s obsession with questions of philosophical method did not, of course, prevent him from making controversial claims about substantial subject-matter, e.g., claims about disposition talk and even claims about dispositions. 9 10 11 12

Ryle 1931-32, 143. Ryle 1938, 189. Ryle 1970, 12. In academic psychology, William Stern and others began to prefer the term “dispositions” rather than “faculties” (Vermögen). His book Die menschliche Persönlichkeit (1918) has a long chapter on dispositions (“Die Dispositionen”), where he explicitly recognized the mutability of psychological dispositions (see also Stern 1911; cf. Pongratz 1967, 71sq. and Pongratz 1972). Among the logical empiricists, Carnap 1928, §§ 24, 150 and Hempel 1935 may be mentioned for dispositional analyses of mental phenomena. In addition, Wittgenstein had suggested a dispositional account of understanding; Braithwaite and others construed beliefs as dispositions (Braithwaite 1932-33).

130

Oliver R. Scholz

2. What Ryle Said What did Ryle have to say about disposition words, disposition sentences and last, but not least about dispositions? What is his account of disposition talk and what is his account of dispositions? The clumsiness of these formulations is deliberate. In what follows I will distinguish carefully between two sorts of claims: (1) claims about disposition talk, i.e. disposition words and disposition ascriptions (let’s call these claims meaning claims13 or conceptual claims); (2) claims about dispositions per se in contrast to other ontological categories such as occurrences (for obvious reasons, I shall call these claims ontological claims). As we have already seen, Ryle’s main interest in The Concept of Mind is in the words and sentences we use for describing “specifically human behaviour”; let’s call them psychological ascriptions.14 In addition, Ryle makes claims about dispositions in general. Taken together, we may – and should – distinguish four sorts of claims: (M.1) meaning claims about disposition talk in general; (M.2) meaning claims, specifically, about mental disposition talk; (O.1) ontological claims about dispositions in general; (O.2) ontological claims, specifically, about mental dispositions.

Let’s begin with some representative quotes from different chapters of The Concept of Mind: (Quote 1) “To possess a dispositional property is not to be in a particular state, or to undergo a particular change; it is to be bound or liable to be in a particular state, or to undergo a particular change, when a particular condition is realised.”15 (Quote 2) “It is being maintained throughout this book that when we characterize people by mental predicates, we are not making untestable inferences to any ghostly processes occurring in streams of consciousness which we are debarred from visiting; we are describing the ways in which those people conduct parts of their predominantly public behaviour. True, we go beyond what we see them do and hear them say, but this going beyond is not a going behind, in the sense of making inferences to oc-

13

14

15

I avoid the term “semantic claims”, since the difference between questions of meaning, on the one hand, and questions of reference (and truth), on the other hand, will be important in what follows. For this reason I chose the more specific term “meaning claims”. Most of the “mental-conduct concepts” whose logical geography Ryle examines in his book, are everyday concepts like “wanting”, “enjoying”, “feeling” or “imagining”. On occasion, he also examines technical concepts which are part of philosophical or scientific jargon, e.g. the concept “volition” (Ryle 1949a, Chapter III). Ryle 1949a, 43.

From Ordinary Language to the Metaphysics of Dispositions

131

cult causes; it is going beyond in the sense of considering, in the first instance, the powers and propensities of which their actions are exercises.”16 (Quote 3) “Dispositional statements are neither reports of observed or observable states of affairs, nor yet reports of unobserved or unobservable states of affairs. They narrate no incidents.”17 (Quote 4) “To talk of a person’s mind is not to talk of a repository which is permitted to house objects that something called ‘the physical world’ is forbidden to house; it is to talk of the person’s abilities, liabilities and inclinations to do and undergo certain sorts of things, and of the doing and undergoing these things in the ordinary world.”18

When we extract Ryle’s major claims from these and related quotations and systematize them a bit, we get the following picture: (M.1) Meaning claims about disposition talk in general: (M.1.1) The logico-linguistic category of disposition talk is (qualitatively) different from the logico-linguistic category of occurrence talk. (M.1.2) Disposition ascriptions are equivalent in meaning to hypothetical sentences.19 (M.1.3) The hypothetical analyses of some disposition ascriptions describe not what the property bearer can do, but what it probably would do, given that the conditions specified in the antecedent are fulfilled. (M.1.4) Disposition ascriptions are law-like. Natural laws and disposition ascriptions are inference tickets that warrant an inference to the state of affairs of the type described in the consequent.20 (M.1.5) Disposition talk is not causal talk. (M.1.6) Disposition ascriptions are not fact-stating reports. (M.1.7) Disposition ascriptions can be tensed.21 (M.2) Meaning claims about mental disposition talk: (M.2.1) Most psychological ascriptions are purely dispositional (i.e., most psychological ascriptions are equivalent in meaning to hypothetical sentences).

According to Ryle, one finds among them ascriptions of knowing, believing, understanding, remembering, wanting, intending, heeding etc. 16 17 18 19

20 21

Ryle 1949a, 51. Ryle 1949a, 125. Ryle 1949a, 199. The modal sentence (containing ‘can’, ‘could’ or ‘would’) that provides the analysis for a disposition ascription is a hypothetical or conditional sentence, the antecedent of which specifies the conditions under which a manifestation of the disposition is expected, and the consequent specifies the nature of the manifestation. See e.g. Ryle 1949a, 121-125, 127. A similar view of laws had already been suggested in Schlick 1931. (Schlick credits Ludwig Wittgenstein with the idea.) See e.g. Ryle 1949a, 125.

132

Oliver R. Scholz

(M.2.2) Some psychological ascriptions are not purely dispositional, but only partly so22 (i.e., some psychological ascriptions are equivalent in meaning to semihypothetical or mongrel categorical-hypothetical sentences). (M.2.3) Psychological ascriptions are not reports of inner occurrences. (M.2.4) A purely dispositional predicate cannot at the same time be understood as an occurrent predicate. (A given predicate, F, is purely dispositional if and only if a statement ascribing F to someone is equivalent in meaning to a subjunctive conditional or a conjunction of subjunctive conditionals where the antecedent of each conditional specifies some situation, and the consequent specifies a reaction of the subject of ascription to that situation). (M.2.5) Most psychological verbs are dispositional verbs. (They take habitual aspect: “Ryle used to swim” – “Ryle swims” – “Ryle will swim regularly” [a disposition to do something intermittently from time to time].) (M.2.6) Some psychological verbs are activity verbs. (They take continuous aspect: “Ryle was swimming” – “Ryle is swimming” – “Ryle will be swimming” [an ongoing activity or process].) (M.2.7) Some psychological verbs are achievement verbs.23 (They take punctual aspect: “Ryle struck his fist on the table” - “Ryle strikes his fist on the table – now!” – “Ryle will strike his fist on the table” [an isolated instantaneous event].)

Now, it should be clear that Ryle does not and cannot leave it at that; he does not confine himself to grammatical observations and conceptual claims. To be sure, it would have been in better accordance with the tenets of Oxford analysis, if he had restricted himself to remarks on grammar and other aspects of language use. But frequently, he makes claims about what the mind is and what it is not, as well as claims about what sort of thing a particular sort of mental phenomenon is and what it is not. His methodological manifestos notwithstanding, Ryle does not only offer antidotes to talking nonsense about the mind, but he makes substantial claims about mental phenomena. Here are some examples: (Quote 5) “Inclinations and moods, including agitations, are not occurrences and do not therefore take place either publicly or privately. They are propensities, not acts or states.”24 (Quote 6) “Feelings, on the other hand, are occurrences [...].”25

22 23

24 25

See e.g. Ryle 1949a, 97, 117, 141. As has been noted, Ryle’s concept of achievement verbs confounds two distinctions: a) the distinction between verbs signifying instantaneous events (stops and starts) and verbs signifying temporally extended situations (processes and states), b.) the distinction between success verbs and verbs that are neutral in this respect (cf. Place 1999, 367). Ryle 1949a, 83. Ryle 1949a, 83.

From Ordinary Language to the Metaphysics of Dispositions

133

(Quote 7) “The radical objection to the theory that minds must know what they are about because mental happenings are by definition conscious, or metaphorically selfluminous, is that there are no such happenings; there are no occurrences taking place in a second-status world, since there is no such status and no such world and consequently no need for special modes of acquainting ourselves with the denizens of such a world.”26

What these and other passages also show is that Ryle presupposes a relation of strict exclusion between occurrences and dispositions on the ontological level, not only on the conceptual level. Regardless of whether or not he is aware of the difference between these two levels, he is, in any case, committed to claims on both levels. If anyone should think that I am not doing Ryle justice, since he did not really intend to go beyond conceptual or meaning claims,27 he has to be reminded of the argumentative context of the claims. Remember that Ryle wanted to refute Cartesian dualism.28 Substance dualism is, however, not a point about grammar. It is an ontological thesis with far-reaching consequences. (Think, e.g., of life after death!) Furthermore, in The Concept of Mind and in later writings, Ryle seems to suggest a prima facie attractive alternative to Cartesian substance dualism according to which the human mind is, at least in the main, a complex of acquired and typically multi-track dispositions. This also strikes me as an ontological thesis about the nature of mind, i.e., about what the human mind really is. Accordingly, let us now turn to Ryle’s most prominent ontological claims: (O.1) Ontological claims about dispositions in general: (O.1.1) The ontological category of dispositions is (qualitatively) different from the ontological category of occurrences; dispositions are not and cannot be occurrences. More precisely, dispositions are contrasted with (a) occurrences, episodes, events, happenings, incidents; (b) processes; (c) acts and activities; (d) states. (O.1.2) Dispositions are not and cannot be processes. (O.1.3) Dispositions are not and cannot be acts or activities. (O.1.4) Dispositions are not and cannot be states. (O.1.5) Dispositions are not events; therefore dispositions are not and cannot be causes. (O.1.6) Dispositions are not “real properties” of substances.

26 27 28

Ryle 1949a, 161. Maybe the sentences in the material mode are only slips of the pen or chosen for the sake of stylistic variation. But would this be a plausible interpretation? See especially Ryle 1949a, 63; cf. Alston 1971, 363.

134

Oliver R. Scholz

Although Ryle casually speaks of dispositions as properties,29 it is hard to see how he could acknowledge them as “real properties” given his view that disposition ascriptions are not fact-stating.30 (O.1.7) There are single-track dispositions and multi-track dispositions. (O.1.8) Dispositions can change. (O.2) Ontological claims about mental dispositions: (O.2.1) Mental dispositions are not inner occurrences. (O.2.2) Mental dispositions are not inner processes. (O.2.3) Mental dispositions are not inner acts or activities. (O.2.4) Mental dispositions are not inner states. (O.2.5) Mental dispositions are typically multi-track dispositions. (O.2.6) The human mind is, in the main, a complex of acquired multi-track dispositions. (O.2.7) If psychological ascriptions imply the existence of occurrences, these are occurrences that take place in the ordinary public physical world. (O.2.8) Beliefs, desires, motives etc. are dispositions to behave in certain ways, but are not causes of action.

3. Critique In this part of my talk, I will comment on some of the major difficulties in Ryle’s account. Naturally, I shall focus on questions related to the main theme of this anthology. 3.1 Critique of Ryle’s Applications in Philosophical Psychology Let me mention, in passing, some rather moderate criticisms. Many philosophers have contested Ryle’s claims about this or that particular mental phenomenon. These criticisms are moderate not only in the sense of being restricted to the phenomenon in question, but also in the sense that they do not contest Ryle’s disposition-occurrence opposition. Terence Penelhum, e.g., defended an episode-view of pleasure against Ryle’s dispositional account.31 Ullin T. Place rejected the dispositional account 29 30

See, e.g., Ryle 1949a, 43, 116. Cp. Mellor 1974, 161.

From Ordinary Language to the Metaphysics of Dispositions

135

of heeding, attention and consciousness.32 It seemed to many commentators that Ryle’s view is “a nonstarter as an analysis of our phenomenal concepts, such as sensation and consciousness itself.”33 3.2 Critique of Ryle’s Account of Dispositions: The Main Objection The most fundamental problem with Ryle’s various claims in The Concept of Mind may be put thus: Ryle does not clearly distinguish between his conceptual or meaning claims on the one hand and his ontological claims about the mind or some specific mental phenomenon on the other hand.34 Worse yet, he often begins with claims of the first type (M) and ends up with a conclusion, which is clearly of the second type (O). However the corresponding pairs of conceptual and ontological claims may be logically and epistemologically related, it seems clear that the ontological claims do not simply follow from the meaning claims. To make this more concrete, let’s look at some of these pairs: (M.1.1) The logico-linguistic category of disposition talk is (qualitatively) different from the logico-linguistic category of occurrence talk. (O.1.1) The ontological category of dispositions is (qualitatively) different from the ontological category of occurrences; dispositions are not and cannot be occurrences. (M.1.5) Disposition talk is not causal talk. (O.1.5) Dispositions are not events; therefore dispositions are not and cannot be causes. (M.1.7) Disposition ascriptions can be tensed. (O.1.8) Dispositions can change.

Let’s begin with a general line of objection. Here is a dilemma: Either you are not afraid of the analytic-synthetic distinction or you are afraid of the analytic-synthetic distinction. If you are not afraid of the analytic-synthetic distinction, I can be brief. Meaning claims are analytic judgments. Ontological claims are synthetic judgments. You cannot, and should not, infer synthetic judgments from analytic judgments. If you are afraid of the analytic-synthetic distinction, then you probably think that there are no purely analytic truths. Even if you were right in this respect, that won’t help Ryle. On the contrary,

31 32 33 34

Penelhum 1956-57. Place 1954 and 1956. Chalmers 1996, 14. Medlin 1967; Alston 1971; see also Place’s comments on Medlin in Place 1999, 386-91. (Place doesn’t mention Alston’s similar critique.)

136

Oliver R. Scholz

if there are no purely analytic judgments, then the whole project of Oxford analysis is doomed from the start. After considering this general line of objection, let’s look briefly at some more specific objections. Consider Rylean inferences such as “Since a skill is a disposition, it cannot be an occurrence” or “Since believing is a disposition, it cannot be a state”. Our meaning claim only tells us: (M) A purely dispositional predicate cannot also be interpreted as an occurrent predicate.

Of course, we can formulate a corresponding claim on the ontological level: (O) A pure disposition cannot be a pure occurrence (where by “being a pure disposition” we mean “being a disposition and nothing else” and by “being a pure occurrence” we mean “being an occurrence and nothing else”).

So far, so good. As has been argued above, Ryle is committed to both the meaning claim and the ontological claim. Moreover, he is committed to a strong principle of exclusion between dispositions and occurrences on both levels. From these considerations, we can get an idea of the immense burden of proof Ryle has to carry. He not only has to show that our ordinary concept of, e.g., a skill is a purely dispositional concept, i.e., that it can be fully analysed into a conjunction of subjunctive conditionals, but that the skill itself is purely dispositional in the requisite sense, i.e., a disposition and nothing else. But Ryle has done nothing to prove that the mental phenomenon is purely dispositional in the requisite sense, except for his argument that our ordinary concept of it is a purely dispositional concept. He has not argued for the ontological claim; he not even attempted to argue for this, since he is simply sliding from claims of the (M)-variety to claims of the (O)-variety without being aware of the difficulties involved. What Ryle would have needed to bridge the gulf between the above-mentioned (M)- and (O)-claims is a very strong principle of the following sort: (EP) Exclusiveness Principle: A phenomenon which can be conceptually identified by the use of a purely dispositional predicate cannot (also) be an occurrence.35

Yet, things get even more complicated: Up to now, I have assumed that Ryle contrasts dispositions only with occurrences; but in various places he also contrasts dispositions with episodes, events, happenings, incidents, processes, doings and undergoings, acts, activities and states.36 It has to be emphasized that such a list is a metaphysician’s nightmare. One could try to soften it by breaking down Ryle’s ontological hotchpotch into four classes of categories: (a) occurrences, events, happenings, incidents; (b) processes (something involving duration and internal change); (c) acts (an initiation of a state attrib35 36

This was pointed out by Alston 1971; throughout this whole section I am indebted to his sharpsighted critique. In some cases, I have nevertheless tried to improve on his formulations. Ryle 1949a; cf. Alston 1971, 368.

From Ordinary Language to the Metaphysics of Dispositions

137

uted to an agent) and activities; (d) states. Thus, strictly speaking, what Ryle needs is a principle like the following: (EP+) Exclusiveness Principle: A phenomenon which can be conceptually identified by the use of a purely dispositional predicate cannot (also) be (a) an occurrence, an episode, an event, a happening, an incident; (b) a process; (c) an act or an activity; (d) a state.

In other words, Ryle has to presuppose that if a phenomenon, e.g. a mental phenomenon, is – so to speak – at least a disposition, that same phenomenon cannot also possess properties that would qualify it as an occurrence or a process or an act or a state. Nothing weaker than this sort of presupposition could bridge the logical gap between a premise saying that our concept of a phenomenon is purely dispositional to the conclusion that the phenomenon cannot be an occurrence, a process, an act or a state. How can principles such as (EP) or (EP+) be justified? First, it should be clear that assumptions of this sort are not generally warranted. It is a very important insight that the properties of the things that we conceive extend far beyond the ordinary concepts we have at our disposal to grasp them. In most cases, there is a chance of discovering some of these properties. This is what science is good for. Consider as an example the concept of electric current most people use: Electric current is whatever comes out of the wall socket and gets the kitchen motors and the TV set going. This definition is practically adequate; but it does not and should not preclude attempts to find out more about electric current, about its many properties and its intrinsic nature. Note, incidentally, that someone who uses the folk concept of electric current may be totally agnostic as to the ontological category to which electric current belongs. Alternatively, imagine that he is more of the dogmatic sort, as philosophers sometimes tend to be. Then, he may self-assuredly claim that electric current is a kind of stuff. If he is a good old Rylean, he might even argue as follows: (M) Electric current talk is stuff talk.

Therefore: (O) Electric current cannot be such-and-such (where “being such-and-such” is something on the ontological level deemed incompatible with “being some kind of stuff”).

Obviously, such an argument is quite preposterous; but, sadly, many of the arguments Ryle offers in The Concept of Mind are essentially of the same kind and therefore suffer from the same defect. To take stock: First and foremost, it is not in general true that whenever a certain concept C is limited to certain kinds of features F, the phenomenon of which C is a concept will also be limited to those features.37 This important 37

Cf. Alston 1971, 368.

138

Oliver R. Scholz

general consideration leaves open the possibility that the justification of (EP) or (EP+) stems from specific considerations of the concepts “disposition” and “occurrence”. Remember the complication mentioned above: Ryle contrasts dispositions not only with occurrences, but also with processes, acts and states. This immediately raises the question: How are these entities to be individuated? Since I do not have the space to discuss the principles of individuation38 that have been suggested for all the categories mentioned, I confine myself to an informal discussion. Consider only two items from our list – processes and states – and ask yourself the following question in light of your intuitive understanding of these notions: Is it to be rationally expected that there is an a priori argument (proceeding via Old School Oxford analysis) that some phenomenon, e.g. a mental phenomenon, picked out by a dispositional concept (from our rich resources of disposition talk) might not on the ontological level turn out to be a process or a state? I suspect, the answer has to be: No! 3.3 Critique of Ryle’s Account of Dispositions: Further Objections To this main objection, let me add a few short remarks on related matters: a.) Meaning and Reference. The assimilation of conceptual questions to questions about the nature of something has been encouraged by the confusion of meaning and reference.39 Consider the following triads: (M-R-O I) (M) Most psychological ascriptions are purely dispositional (i.e., most psychological ascriptions are equivalent in meaning to hypothetical sentences). (R) Most psychological ascriptions refer to pure dispositions. (O) Most psychological phenomena are pure dispositions.

And: (M-R-O II) (M) Disposition talk is not causal talk. (R) Disposition terms do not refer to causes. (O) Dispositions cannot be causes.

Philosophers like Ryle easily slide from statements of type (M) to statements of type (R) and (O). Nevertheless, the meaning of an expression leaves unre-

38 39

At least, it should be clear by now that predicate identity is not adequate. Something stronger seems to be needed, e.g. causal relevance. Cf. Alston 1971. Alston 1971, 381sqq.

From Ordinary Language to the Metaphysics of Dispositions

139

solved most questions as to the existence and the nature of the entity to which the expression refers.40 First, it leaves open the possibility of reference failure, i.e., the possibility that the expression fails to refer to anything at all. Second, if reference does not fail, the entity referred to by the expression may possess endless properties not anticipated in the expression’s meaning. b.) Predicates and Properties. The assimilation of conceptual questions to questions about the nature of something may also have been encouraged by the assimilation or identification of predicates and properties. But, whereas predicates are language- and mind-dependent entities, most properties of objects are not.41 c.) Logic. Without an adequate logic of counterfactual conditionals, the simple hypothetical analysis of dispositional statements, as offered by Ryle, is a rather blatant case of obscurum per obscurius. Worse still, Ryle did not make the tiniest effort or show the slightest interest in developing such a logic. As Peter Geach put it already in 1957: “It ought to be, but plainly is not, generally known to philosophers that the logic of counterfactual conditionals is a very ill-explored territory; no adequate formal logic for them has yet been devised, and there is an extensive literature on the thorny problems that crop up. It is really a scandal that people should count it a philosophical advance to adopt a programme of analysing ostensible categoricals into unfulfilled conditionals, like the programmes of phenomenalists with regard to ‘physical-object’ statements and of neo-behaviourists with regard to psychological statements.”42

d.) The “Inference Ticket” View. Ryle’s “inference ticket” view of laws of nature and of disposition ascriptions has also come in for criticism. One problem is that this view seems unable to explain that dispositions can change. It is true that Ryle allows that disposition ascriptions can be tensed (a meaning claim!) and that dispositions can change (an ontological claim!), but he fails to give the ontological claim any real content. As Peter Geach derisively remarks, “on Ryle’s view [...] ‘the rubber has begun to lose its elasticity’ has to do not with a change in the rubber but with the (incipient?) expiry of an inference-ticket.”43

4. Ryle’s Legacy Let’s take stock, then. After all this indolent censoriousness, I shall start with the positive. Gilbert Ryle’s The Concept of Mind still offers one of the most comprehensive and detailed surveys of what is nowadays called folk psychol-

40 41 42 43

Reference is more closely related to truth than meaning is. Martin in: Armstrong, Martin and Place 1996, 71; cf. Place 1999, 387sq. Geach 1957, 6sq. Geach 1957, 7.

140

Oliver R. Scholz

ogy. The distinction between “knowing how” and “knowing that” still has its defenders as well as detractors.44 Ryle’s warnings about category mistakes and fallacies such as the homunculus fallacy are very much relevant in the current situation of psychology and the neurosciences.45 His positive ideas prompted philosophers like Wilfrid Sellars, David Armstrong, David Lewis and Daniel C. Dennett to develop functionalist theories of mind.46 Ryle, also quite correctly, pointed out that very many (but not all) ordinary psychological concepts are dispositional concepts, typically denoting multi-track dispositions. But, as we have seen, his general account of dispositions suffers from severe lacunae and defects. More promising accounts had to await the return of scientific realism.47

Literature Alston, W. P. 1971. “Dispositions, Occurrences, and Ontology.” In Canadian Journal of Philosophy 1: 125-154; reprinted in: Tuomela, Raimo. ed. 1978. Dispositions, 359-388. Dordrecht: Reidel. Armstrong, D. M., Martin, C.B. and Place, U.T. 1996. Dispositions 0 A Debate, ed. T. Crane, London and New York. Bennett, M. R. and Hacker, P. M. S. 2003. Philosophical Foundations of Neuroscience. Oxford: Blackwell. Braithwaite, R. 1932-33. “The Nature of Believing.” Proceedings of the of the Aristotelian Society 33: 129-146. Carnap, R. 1928. Der logische Aufbau der Welt. Hamburg: Meiner. Carnap, R. 1936-37. “Testability and Meaning.” In Philosophy of Science 3: 419-471 and 4: 140. Chalmers, D. J. 1996. The Conscious Mind. In Search of a Fundamental Theory. New York and Oxford: Oxford University Press. Geach, P. 1957. Mental Acts. London: Routledge and Kegan Paul. Hacker, P. M. S. 1996. Wittgenstein’s Place in Twentieth Century Philosophy. Oxford: Blackwell. Hempel, C. G. 1935. “Analyse logique de la psychologie”. Revue de Synthèse 10: 27-42; transl. W. Sellars as: Hempel, C. G. 1949. “The Logical Analysis of Psychology.” In Readings in Philosophical Analysis, ed. H. Feigl and W. Sellars, 373-84. New York: AppletonCentry-Crofts. Medlin, B. 1967. “Ryle and the mechanical hypothesis.” In The Identity Theory of Mind, ed. C. F. Presley, 94-150. St. Lucia: University of Queensland Press. Mellor, D. H. 1974. “In Defense of Dispositions.” In Philosophical Review 83: 157-181.

44 45 46 47

See Gregor Damschen’s paper in this volume. Cf. e.g. the diagnosis of the endemic mereological fallacy in psychology and neuroscience offered by Bennett and Hacker 2003, 68-85 and passim. Hilary Putnam’s and Jerry Fodor’s functionalism is different; they take their inspiration from the idea of a Turing machine. See the contributions of Andreas Hüttemann, Stephen Mumford and Markus Schrenk to this volume.

From Ordinary Language to the Metaphysics of Dispositions

141

Place, U. T. 1954. “The Concept of Heed.” In British Journal of Psychology 45: 243-255. Place, U. T., 1956. “Is Consciousness a Brain Process?” In British Journal of Psychology 47: 4450. Place, U. T. 1999. “Ryle’s Behaviorism.” In Handbook of Behaviorism, ed. W. O’Donohue and R. F. Kitchener, 361-98. San Diego: Academic Press. Place, U. T. 2004. Identifying the Mind. Selected Papers, ed. G. Graham and E. R. Valentine. Oxford: Oxford University Press. Pongratz, L. J. 1967. Problemgeschichte der Psychologie. Bern and Munich: Francke. Pongratz, L. J. 1972. “Disposition II.” In Historisches Wörterbuch der Philosophie, vol. 2, ed. J. Ritter and K. Gründer, 265-66. Basel: Schwabe. Ryle, G. 1932. “Systematically Misleading Expressions.” Proceedings of the Aristotelian Society 32: 139-170; reprinted in: Ryle 1971b, 39-62. Ryle, G. 1938. “Categories.” Proceedings of the Aristotelian Society 38: 189-206; reprinted in: Ryle 1971b, 170-184. Ryle, G. 1945-46. “Knowing How and Knowing That.” Proceedings of the Aristotelian Society 46: 1-16; reprinted in: Ryle 1971b, 212-225. Ryle, G. 1949a. The Concept of Mind. London: Hutchinson. Ryle, G. 1949b. “Discussion of Rudolf Carnap: Meaning and Necessity.” Philosophy 24: 6976; reprinted in: Ryle 1971a, 170-184. Ryle, G. 1950. “If”, “So” and “Because.” In Philosophical Analysis, ed. M. Black, 323-40, Ithaca and London: Cornell University Press; reprinted in: Ryle 1971b, 234-49. Ryle, G. 1954. Dilemmas. Cambridge: Cambridge University Press. Ryle, G. 1970. “Autobiographical.” In Ryle, ed. O. P. Wood and G. Pitcher, 1-15. New York: Doubleday & Co. Ryle, G. 1971a. Collected Papers, Volume 1: Critical Essays. London: Hutchinson. Ryle, G. 1971b. Collected Papers, Volume 2: Collected Essays 1929-1968. London: Hutchinson. von Savigny, E. 1969. Die Philosophie der normalen Sprache. Eine kritische Einführung in die “ordinary language philosophy”. Frankfurt am Main: Suhrkamp. Schlick, M. 1931. “Die Kausalität in der gegenwärtigen Physik.” In Die Naturwissenschaften 19: 145-162. Soames, S. 2003. Philosophical Analysis in the Twentieth Century. Volume 2: The Age of Meaning. Princeton and Oxford: Princeton University Press. Stern, W. 1911. Die differentielle Psychologie in ihren methodischen Grundlagen. Leipzig: Johann Ambrosius Barth. Stern, W. 1918. Die menschliche Persönlichkeit. Leipzig: Johann Ambrosius Barth. Tuomela, R., ed. 1978. Dispositions. Dordrecht: Reidel.

III. Contemporary Philosophical Analyses of Dispositions

Hic Rhodos, Hic Salta: From Reductionist Semantics to a Realist Ontology of Forceful Dispositions1 MARKUS SCHRENK

0. Abstract If dispositions were real properties they would bestow the world with a connection in nature a Humean could not accept (1.). The nature of that dispositional connection or power is the target of my paper’s main argument. I will argue that it can not be cashed out in terms of metaphysical necessity. Metaphysical necessity might connect synchronically co-existent properties – kinds and their essential features, for example – but it cannot serve as the binding force for successions of events. That is, metaphysical necessity is not fit for diachronic, causal affairs which causal laws, causation, and, especially, dispositions are involved with (3.&4.). A different anti-Humean connection in nature has to do that job. I will present a tentative suggestion how we could start to conceptualise this different connection (5.&6.). My core argument is embedded in a debate which has been the battleground for Humean vs. anti-Humean intuitions for many decades – the conditional analysis of dispositional predicates (2.) – but I believe (yet cannot prove here) that the arguments generalise to causation and causal laws straightforwardly.

_____________ 1

Thanks are due first and foremost to Gregor Damschen, Robert Schnepf and Karsten Stueber for the great conference in Wittenberg for which this paper has been written. I am also very grateful to the other participants and contributors to this book for their valuable comments and questions. Further thanks go to Stephen Williams, Stephen Mumford, and Charlotte Mattheson who have read and criticised an earlier version of the paper. I specially thank Matthew Tugby for his many helpful comments on the earlier draft. I have started the paper as Junior Research Fellow at Worcester College, Oxford, and finished it for publication in Nottingham as a Postdoctoral Research Fellow within the AHRC Metaphysics of Science project.

146

Markus Schrenk

1. The Humean Legacy Hume’s Followers: There are two major strands in 20th century analytic philosophy which both share the Humean sentiment that necessary connections in nature2 do not exist and that, consequently, dispositional properties (powers, capacities, potencies, etc.) are not real properties and have to be analyzed in terms of non-dispositional categories. The two schools I have in mind are early twentieth century logical empiricism and the neo-Humean supervenience program David Lewis has famously given the name and credo for. One of the logical empiricists’ main aims – known under the heading ‘verificationism’ – was to give analyses of all notions in terms of observational vocabulary. Only if this was possible, they believed, would those terms belong to a meaningful language. Dispositional predicates – not belonging to observational language themselves – had, therefore, either to be analyzed or to be deleted from scientific language. As we know now, verificationism is untenable and it saw its downfall partially because dispositional predicates proved to be indefinable by the means available to the empiricist. One brief part of this paper is the story of the difficulties empiricists had to face. As mentioned, the second incentive to reduce dispositions comes from the metaphysical position David Lewis has dubbed Neo-Humean Supervenience “in honor of the great denier of necessary connections:” It is the doctrine that all there is to the world is a vast mosaic of local matters of particular fact, just one little thing and then another. […] We have geometry: a system of external relations of spatio-temporal distance between points. […] And at those points we have local qualities: perfectly natural intrinsic properties. […] For short: we have an arrangement of qualities. And that is all there is. There is no difference without a difference in the arrangement of qualities. All else supervenes on that. (Lewis 1986, IX-X)

What a follower of this creed wants to fight are philosophical arguments against Humean supervenience. When philosophers claim that one or another commonplace feature of the world cannot supervene on the arrangement of qualities, I make it my business to resist. (Lewis 1986, XI)

Needless to say, dispositions are supposed to belong to the things supervening on local matters of fact. The history of the conditional analysis of dispositional predicates which starts with the empiricists will lead us smoothly to Lewis’s reductive analysis. Note the following superficial difference between these two groups of opponents of dispositions: logical empiricism was occupied with the task of confirming an epistemic credo by the means of a semantic dogma: all factual knowledge comes from sense experience and it can be shown that this is true by

_____________ 2

Please note that I will use ‘natural necessity’, ‘de re necessity’, ‘metaphysical necessity’, and ‘necessity in nature’ synonymously. The same holds for the group ‘power’, ‘capacity’, ‘disposition’, etc.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

147

proving that the verificationist theory of meaning is adequate. Humeansupervenience is, on the other hand, a metaphysical doctrine: all there is, is a huge pattern of point size and separate property instantiations; a kind of ‘metaphysical pointillism’3. However, deep down both neo-Humean supervenience and logical empiricism share the same root and this is the aforementioned hostility to (hidden) links in nature. That is, they both subscribe to Hume’s arguments against any such connection. I will briefly rehearse Hume’s well known line of reasoning. According to Hume “all objects of human reason and enquiry may naturally be divided into two kinds […] Relations of Ideas, and Matters of Fact.” (Hume 1777, 25) The first, Relations of Ideas, can be revealed a priori by pure thought, whereas Matters of Fact have to be discovered by experience, that is, by sense perception. Whether or not a sharp line can be drawn between a priori and a posteriori knowledge is a matter of much debate. However, that there is no entirely different third way to obtain knowledge (via dream interpretation, for example) is relatively uncontroversial. Consequently, the reasons to support necessary connections in nature – here especially those between a certain trigger event, an object being disposed, and a certain reaction – have to arise on either a priori or a posteriori grounds. Yet, neither, says Hume, is tenable. There cannot be a priori grounds for a necessary relation in nature, for as far as we know or as far as we are able to imagine anything may produce anything. No pure thought can reveal that, for example, water has the power to suffocate or fire the capacity to consume. Hence, our reason, unassisted by experience, [cannot] ever draw any inference concerning real existence and matter of fact. (Hume 1777, 27; my italics)

Why is there no possibility of experiencing necessary connections with our senses either? We might repeatedly observe one kind of event followed by another kind of event but, so Hume continues, what we cannot observe is the alleged necessary connection between them and, even less so, an alleged power of the first to bring about the second: The scenes of the universe are continually shifting, and one object follows another in an uninterrupted succession; but the power or force, which actuates the whole machine, is entirely concealed from us, and never discovers itself in any of the sensible qualities of body. […] External objects as they appear to our senses, give us no idea of power or necessary connection. (Hume 1777, 63-4)

Hume’s first argument, that we have no a priori way of discovering necessary connections in nature, is widely accepted amongst both Hume’s supporters and his opponents such as dispositionalists.

_____________ 3

A term I borrow from Jeremy Butterfield.

148

Markus Schrenk

Hume’s Opponents: However, a fast-growing community of philosophers does not any longer accept Hume’s second argument. Some people state that some links in nature – certain instances of causation as, for example, forces on our body or successful acts of the will – can be directly and non-inferentially experienced.4 In fact, I believe that a version of this is correct and that it is the most promising move against Hume (but I cannot argue for this claim here). Yet, many of those who accept anti-Humean de re connections do not pursue this line of argument but support ideas put forward by Kripke and Putnam. That is, they conceive of those connections as a posteriori, conceptually contingent, yet, metaphysically necessary links which are partially argued for on the grounds of semantic considerations about direct reference and rigid designation but also on grounds of scientific discovery. Many people have been impressed by these arguments and especially antiHumeans, such as dispositionalists, saw the chance of a revival of antiHumean de re connections. Stathis Psillos, for example, comments (here speaking of laws rather than dispositions): It was Kripke’s liberating views in the early 1970s that changed the scene radically. By defending the case of necessary statements, which are known a posteriori, Kripke [1972] made it possible to think of the existence of necessity in nature which is weaker than logical necessity, and yet strong enough to warrant the label necessity. […] As a result of this, the then dominant view of laws as mere regularities started to be seriously challenged. (Psillos 2002, 161; my italics)5

Yet, I am convinced that Kripke’s influence has to be thought of as merely psychological rather than philosophical: Kripke has opened people’s minds to connections in nature that have been banned from (some) philosophy since Hume, but Kripke has not yet come up with the kind of link in nature dispositionalists need. This will be the main subject of sections 3 and 4. In sections 5 and 6 I will suggest what, instead of Kripkean metaphysical necessity, could do the job of dispositional powers.6 In the following section, however, I will first

_____________ 4

5

6

See, for example, David Armstrong: “causation is given in experience” and “the dyadic predicate ‘causes’ is as much an observational predicate as any other predicate in our language, especially in such cases as our awareness of pressure on our own body” (Armstrong 1997, 228). Particularly interesting for my later arguments are passages in Evan Fales’s Causation and Universals, especially chapter 1, p. 48. I am not here saying that Psillos is an anti-Humean or dispositionalist. I quote him because he gives expression to the view that Kripke was involved essentially in the anti-Humean revolution of the defenders of strong laws. I believe something even more heretic, namely that alleged metaphysical necessities like “water = H2O” should be treated as normative methodological claims or recommendations with the aim to enhance good scientific conduct rather than as literal truths. That is, I reject the metaphysical doctrine of natural necessity in its fact-stating form while I subscribe to the prescriptive version of it (cf. Watkins 1958, 356-7). Yet, this has only little to do with the main subject of this paper. As I said above, here I only aim to show that even if we accept necessity in its fact-stating form, it is not fit for the purposes of the dispositionalist.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

149

present a brief history of the conditional analysis of dispositional predicates into which my main argument will be embedded.

2. A Brief History of Semantic Reduction The history of the attempts to analyze dispositional predicates is well documented and some counterexamples to it have reached folkloric status in analytic metaphysics. Therefore, only a brief rehearsal of the ups and downs of the analysis should be sufficient here. The Material Conditional Analysis: The very first and most straightforward attempt to analyze dispositional predicates, like “x is soluble”, and its rejection can be found in Carnap’s Testability and Meaning. “ ” is meant to symbolize the material implication: S(x) iff T(x)

R(x) (cf. Carnap’s Testability and Meaning 1936, 440)

That is, x is soluble in water, S(x), iff if x is put into water, T(x), then x dissolves, R(x). This minimalist suggestion does not work for various reasons: the two most famous difficulties concern (i) the so-called “Void Satisfaction” and the (ii) “Random Coincidence.” (i) Void Satisfaction: Imagine a match m, which has never been and will never be put into water. The definiens, ‘T(m) R(m)’, comes out true because its antecedent, T(m), is false – this is one of the well known paradoxes of the material implication. As a consequence, we have to attribute solubility to the match. In other words, the match voidly satisfies the criterion for being soluble. An untenable result. (ii) Random Coincidence:: A coincidence occurs randomly when both the antecedent, T(x), and the consequent, R(x), happen, yet by sheer accident and not because an object has the disposition to react with R when T-ed. To my knowledge it was Jan Berg who coined the term random coincidence in the context of dispositional predicates (Berg 1960). An example modelled after Berg’s is this: suppose we define the disposition “x is explosive” as “if you knock on x, it bursts into pieces” (I guess Berg had something like a landmine in mind). Imagine you knock on a table and it happens to burst into pieces for some weird reason such as an elephant stepping on it a second later. We would, then, not want to say that it is explosive although, unfortunately, the definiens conditions for explosiveness are fulfilled. Enhancing the Naïve Conditional Analysis: Even readers who are not familiar with the dispositions debate will immediately have a multitude of promising

150

Markus Schrenk

ideas concerning how one could avoid either the void satisfaction or the random coincidence or both. Here are some such ideas: (i) We could try to reformulate the antecedent of our conditional so that elephants and other such influences are forbidden as unwanted external factors. (ii) We could specify the consequent in more detail so that bursting due to elephants is not the right kind of bursting. (iii) We could add a clause demanding that the bursting comes about merely due to the intrinsic setup of the object. (iv) We could insist that the object disposed to explode should belong to a kind of object or material which usually shows explosive behaviour (the wood the table was made of would be excluded). (v) We could abandon the explicit definition of dispositional predicates and revert to implicit definitions. (vi) A time variable (t for the antecedent, t + t for the consequent) might be cleverly inserted into the definition. (vii) Finally, we could give up the material conditional and use a stronger logical connective. All these possibilities have been suggested in one form or another (and I do not claim that this list is exhaustive). Unfortunately, there is no space to focus on all of them here. I will leave aside Carnap’s suggestion to define dispositional predicates with the help of reduction sentences (cf. suggestion (v)) and also the various further attempts of Trapp and Essler to overcome certain follow-up problems reduction sentences had to face (including redefinitions with time variables: (vi)).7 I especially regret having to leave aside Eino Kaila’s forward-looking definition which demands of the disposed object not only that a conditional is true of it but also that it is a member of an appropriate class of objects for which the dispositional behaviour is a law-like regularity (cf. suggestion (iv)).8 I also neglect early attempts to define the relation ‘c causes e’ as a stronger-than-material conditional and to use it for the definition of disposi-

_____________ 7

8

Carnap 1936/7, Trapp 1975, Essler 1975, Essler & Trapp 1977, Pap 1963. Someone who takes dispositions to be indefinable real properties could actually appreciate reduction sentences as describing the symptoms of dispositions. The fact that reduction sentences are mere implicit definitions (or “conditioned definition” as Carnap calls them (Carnap 1936, 443)) should suit the dispositionalist well. Spohn and Mellor both mention reduction sentences with sympathy (cf. Spohn 1997, Mellor 1999). At the time Kaila offered his definition, natural kinds were not yet en vogue in philosophy so that no successful characterisation of the ominous class of objects could be found. I believe it could be worth rethinking Kaila in the light of essentialism and modern formulations of lawhood (cf. Kaila 1939). Wedberg in (Wedberg 1944, 237) has, independently, developed a very similar if not equivalent definition (see Storer 1951 for the equivalence). For a discussion of Kaila see also Berg 1960 and Pap 1955.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

151

tions (this is one variant of (vii)).9 Instead, I will turn straight away toward modern attempts to rescue the conditional analysis with the help of counterfactual conditionals (again suggestion (vii)). We will see that the other above mentioned attempts to save the analysis (suggestions (i), (ii), (iii), (vi)) will also find their way into the analysis. The Bare Counterfactual Conditional Analysis: The first promising semantics for counterfactual conditionals were introduced independently by Stalnaker and Lewis as late as the 60s or early 70s. We can, for our purposes at least, treat the point at which Lewis enters the stage as the turning point from empiricism to neo-Humeanism. The bare counterfactual analysis x has the disposition D if x were exposed to the test T, x would show the reaction R has, interestingly, never been explicitly suggested in the literature. By the time Lewis and Stalnaker had published their theories of counterfactuals people seemed to have temporarily lost interest in the dispositions debate. On the other hand, everyone seemed to have tacitly assumed that the problems of the original analysis would be solved easily with the rise of counterfactuals.10 This hope is certainly justified concerning the void satisfaction difficulty. Remember that this difficulty is merely a consequence of the paradoxes of material implication: as soon as the antecedent is false the whole conditional is true regardless of the truth value of the consequent. A counterfactual conditional cures this disease: there has to be a close possible world in which the antecedent and also the consequent are true. If not, the counterfactual conditional is false. Hence, an empty or void satisfaction cannot occur. Other possible worlds take care of that. However, people should have been suspicious when it comes to the random coincidence difficulty for this is not merely based upon the deficiencies of truth functional logic. The reason is that the random coincidence case is one in which both the antecedent event and the reaction happen: yet, not for the right reasons. And these wrong reasons can be operative even in nearby possible worlds. To see this, take again our example, the non-explosive table. We just have to assume that the elephant is trained to step on the table whenever someone knocks on it.11 In that way we transport the elephant’s interference into nearby possible worlds. That is, it becomes true of the table that if we were to knock on it (= if it were exposed to the test T), it would burst into pieces (= it

_____________ 9 10

11

Burks 1951, 1955, Sellars 1958. Lewis, for example, writes: “All of us used to think, and many of us still think, that statements about how a thing is disposed to respond to stimuli can be analysed straightforwardly in terms of counterfactual conditionals” (Lewis 1997, 143). It is true that the earlier cases of random coincidences have to be turned into a regulated coincidence in order to attack also the counterfactual analysis, but this is a relatively small alteration.

152

Markus Schrenk

would show the reaction R). Yet, it is still not the table that is explosive but it bursts because of the elephant that is heavy. Martin’s Finks: Charles Martin revived the dispositions debate in the mid 90s with his seminal text “Dispositions and Conditionals” (Martin 1994). He did not refer to random coincidences but he had a similar idea in mind. Martin has convincingly shown that the bare counterfactual conditional is neither sufficient nor necessary for an object to be disposed to do something. His example is of a live wire (live being the disposition in question)12 to which a machine – Martin calls it an electro-fink – is connected. This machine is built in such a way that it stops the power supply immediately if the wire is touched by a conductor. The conditional analysis of ‘x is live’, taken to be ‘if x were touched by a conductor, then electric current would flow from x to the conductor,’ is inadequate, since the wire is live ex hypothesi, but the conditional is not true due to the fink. The conditional is not necessary. We can rephrase the story mutatis mutandis such that a “reverse-electro-fink” is operating on a non-live wire and thereby show that the conditional is not sufficient either (“We turn a switch on our electro-fink so as to make it operate on a reverse cycle” (Martin 1994, 3)). Our immediate reply is likely to be that the peculiar intervention of the fink does not really belong to what normally happens (compare: the trained elephant). This intuition suggests that we can upgrade the analysans in a straighforward way: “Conditionals which give the sense of power ascriptions are always understood to carry a saving clause (the full details of which are commonly not known)” (Martin 1994, 5). Yet, Martin shows (or maybe only seems to show, as we shall see in a moment) that no appropriate definition can be given for a ceteris paribus or ‘all else being equal’ clause. So, it is not an option, after all, to define, for example: “If the wire is touched by a conductor and other things are equal, then electrical current flows from the wire to the conductor” (Martin 1994, 5). Martin’s conclusion regarding this negative result is radical and pessimistic: counterfactual conditionals are not the appropriate tools to use in defining dispositional predicates. For Martin, dispositions are irreducible (cf. Martin 1994, 8). Martin’s article in the Philosophical Quaterly was the starting point for a rejuvenated debate about dispositions. The prevention of finks (and other counteracting devices; see below) became the main concern in this discussion for over a decade. The goal has been either to enhance the conditional analysis with tenable proviso clauses or otherwise to show that this is not possible and that dispositions are, thus, irreducible real properties.

_____________ 12

Other examples are to be found easily if it is in doubt whether being live should count as a disposition.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

153

Lewis’s Analysis: I turn first to David Lewis’s answer to Martin. In “Finkish Dispositions” (1997), Lewis did not want to follow Martin’s bold step. Instead, he offers a sophisticated reformulation of the primitive counterfactual analysis adding both a certain ontological assumption about dispositions and time variables. The ontological assumption is that there is a causally active, possibly categorical basis B underlying each disposition. In the case of solubility, for example, basis B may be a molecular structure XYZ with which water-dipoles can interact. Lewis argues (with Prior et al. (1982)) in favor of such a basis B. Lewis’s basic way of accounting for Martin’s fink cases is expressed thus: “A finkish disposition is a disposition with a finkish base” (Lewis 1997, 149). Here’s his complete analysis: Something x is disposed at time t to give response r to stimulus s, iff, for some intrinsic property B that x has at t, for some time t' after t, if x were to undergo stimulus s at time t and retain property B until t', s and x’s having of B would jointly be an xcomplete cause of x’s giving response r. (Lewis 1997, 157)

“An unlovely mouthful” as Lewis himself concedes (Lewis 1997, 157). Let me try to explain the most important feature of his definition. The crucial phrase to make the definition immune to finks is “retain property B from t to t.'” Note that since B is not the dispositional predicate itself there is no circularity in Lewis’s account. Remember that the fink is, as much as the disposition itself, supposed to be activated by the trigger conditions of the disposition. Once they occur the fink destroys the object’s basis B for the disposition. Yet, Lewis’s definition takes care of such an unfortunate event in that it includes the condition “if x were to […] retain property B until t.'” Lewis seems to have saved the conditional analysis of dispositional predicates. The void satisfaction difficulty is taken care of by the mere fact that counterfactual conditionals are employed and the random coincidence objection, here in the guise of finks and reverse-finks, is ineffective too. Unfortunately, there is a whole zoo of little machines, interferers, and unwanted processes – antidotes, prodotes, masking, mimicking, to name but a few – which can make life difficult for those who aim to provide a counterfactual analysis, even, it will be shown, for Lewis’s analysis. I will focus on antidotes as introduced by Alexander Bird in his “Dispositions and Antidotes” (Bird 1998). Bird’s Antidotes: Bird’s derivative of finks, namely antidotes, do not destroy the intrinsic basis, B, of a disposed object but interfere with the causal process extrinsic to the object. One of his examples is of an uranium pile that is above critical mass (cf. Bird 1998, 229). The pile has the disposition to chain-react catastrophically. Yet, there is a safety mechanism that lets boron moderating rods penetrate the pile in case radioactivity increases. The boron rods absorb the radiation and prevent a chain reaction. Although the intrinsic structure of the uranium pile is not altered, its disposition will not be manifested. Hence, Lewis’s demand for saving the intrinsic basis is futile.

154

Markus Schrenk

Again, the counterfactual analysis of dispositional predicates seems to fail. Yet, Bird also offers an interesting suggestion about how to save it once again. He shows how it might be possible, in contrast to Martin’s arguments, to define proviso clauses that exclude finks, antidotes, and everything else that could interfere with the disposition-manifestation process. For that purpose, Bird employs insights of externalist semantics. In short: the class C of normal conditions is the class of all circumstances which crucially resemble those archetypal situations in which we tested objects of the same kind positively. Consequently, if one makes a disposition ascription in a certain context, one assumes that the conditions of the context of utterance are similar to or are the same as those archetypal situations. I know that this summary does not do justice to Bird’s ideas. Yet, for reasons of space I can only further highlight the fact that, in his final analysis of dispositions, Bird keeps Lewis’s reference to a basis property on top of his own insertion of normal conditions (cf. Bird 1998, 233). Latest developments: The next step in the analysis of dispositional predicates suggests itself. Once a successful way to define normal, or ideal, or ceteris paribus conditions is found there might not be the need, anymore, to refer to basis properties. I believe that the application of the tools known from externalist semantics (reference to paradigm cases in order to define normal conditions demonstratively) is a promising route. In Schrenk 2000 (unpublished)13 I have defended such a thin version of the conditional analysis, which characterizes communicative-relevant conditions as context sensitive elements that are partially fixed by paradigm cases. Yet, since the discussion of conditional analyses is, in this paper, only a vehicle to reach the actual subject, I have to leave things at this superficial stage. What we have to conclude from the whole discussion remains, in any case, open: the conditional analysis might or might not succeed depending, amongst other things, upon how many counterexamples along the lines of finks and antidotes will still be found.14

_____________ 13

14

I have learned only shortly before the deadline for this paper that Sungho Choi seems to propose a similarly slim analysis in his “The Simple vs. Reformed Conditional Analysis of Dispositions.” (Choi 2006). – Note aside, although I believe that such a semantic analysis is possible, I do not think that dispositions are thereby ontologically reduced to non-dispositional properties. For further discussions see, for example, Molnar 1999, Malzkorn 2000, Gunderson 2000 with a reply by Bird (Bird 2000), Mumford 2001, and Choi 2003.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

155

3. Hic Rhodos, hic salta! How does the Dispositionalist deal with Antidotes? We have heard about the attempts to analyze dispositional predicates in terms of counterfactual conditionals (2.). In the eyes of the dispositionalists, these attempts have failed. However, which consequence we should and can draw from the alleged failure is uncertain.15 In any case, the counterexamples dispositionalists have launched against semantic reductions – finks, antidotes, etc. – are undoubtedly a hard nut to crack. How hard this nut really is I hope to reveal in this core section of my paper, where I will turn the tables and ask what the dispositionalists themselves have to say about the counterexamples. At this point, the choice of my pompous paper title ‘Hic Rhodos, hic salta!’ gets some justification. The origin of this odd saying is a punch line in Aesop’s fable The Braggart where an athlete boasts that he once performed a stunning jump in Rhodes. Addressing him with the words ‘Hic Rhodos, hic salta!’ a bystander challenges the athlete to demonstrate his capacities here and now. Antidotes for Dispositionalists: I am especially interested in what the dispositionalists make of certain antidote cases where the trigger of the disposition is pulled, yet, its manifestation still does not occur. Here is roughly where the problem lies. Peter Lipton described a tripartite distinction with respect to the status of an object’s dispositions. The tripartite distinction is, I think, usually taken to be exhaustive: “For dispositions [there is] […] a tripartite distinction: displaying, present-but-not-displaying, or absent”(Lipton 1999, 163). However, I will show that these three elements are not enough to capture antidote cases: the realist about dispositions needs one further element in order to account for these cases. I will provisionally call it the pushing, trying, aiming, or attempting to manifest element. The need for it emerges if one runs quickly through Lipton’s tripartite distinction.

_____________ 15

Inspired by logical empiricists’ metaphysics and epistemology people have assumed that if Humeans should be able to provide a watertight analysis that answers to all possible counterexamples then dispositions are not real. Yet, outside the constraints of empiricism/verificationism conclusions on these matters are not so straightforward. The same holds also for the opposite position: if no counterfactual analysis ever succeeds can we conclude that dispositions are real? Sceptical about easy solutions to these questions John Heil recently wrote: “Even if you could concoct a conditional analysis of dispositionality impervious to counter-examples, it is not clear what you would have accomplished. You would still be faced with the question, What are the truth makers for dispositional claims? Suppose you decide that ‘object o is fragile’ implies and is implied by ‘o would shatter if struck in circumstances C’. You are not excused from the task of saying what the truth maker might be for this conditional. Presumably, if the conditional is an analysis, its truth maker will be whatever the truth maker is for the original dispositional assertion. This is progress?” (Heil 2005, 345)

156

Markus Schrenk

- Absence: An object might not have a disposition at all (or it might have lost it due to a fink). Hence, even if sufficient trigger conditions are fulfilled, the object does not display the disposition’s manifestation. - Displaying: An object has the disposition, the relevant trigger conditions are fulfilled and, no surprise, it displays the disposition’s manifestation. - Present-but-not-displaying: This case is the difficult one. I will explain what I have in mind with the help of two imaginary scenarios. First, think of sugar’s solubility. Clearly, sugar does not display its solubility when it is not in contact with any water. (Likewise, a rubber band does not display its elasticity as long as it is not pulled; a match does not display its flammability as long as it is not struck, etc.) This, then, is one kind of being present-but-not-displaying: the trigger is entirely absent. Bird’s antidote case, however, is quite different: the nuclear power station’s uranium pile is moderated by boron rods. That is, while the uranium rods are still disposed to melt and, moreover, also triggered to do so, they do not. It is important to note that this is not simply a case where there are no activation circumstances: the mass of the uranium is not reduced to less-than-critical. The uranium is still very much poised to chain react. We have, hence, a second kind of being present-but-not-displaying: being present plus triggered, yet, not manifesting. To expose the principal difference between the sugar and the uranium story yet again consider two analogous cases: think, as another antidote (uranium-type) case, of an electron in an electric and a gravitational field. According to its Coulomb capacity it should accelerate in one direction, according to its mass in another. Let us imagine, however, that the two fields’ strengths and directions are such that the electron remains stably in its initial position. Again, both dispositions are present-but-not-displaying. Yet, they are triggered. They do not display because the gravitational force is an antidote to the Coulomb capacity and vice versa. Now, contrast the latter case of a force equilibrium with the case of a stationary electron that is not surrounded by any field at all. It is charged and massive. However, while these dispositions are present they are untriggered and so not displaying. This is similar to the sugar case. A dispositional stage merely called “presence”, is, because of the antidote cases, not enough, for we could not distinguish between (i) being present plus being not triggered and (ii) being present plus triggered, yet, not manifesting. Yet, what is wrong with further differentiating Lipton’s present-but-notdisplaying into two subgroups, one of which accommodates cases where the disposition is activated and hence pushing towards its manifestation? After all, isn’t this what dispositionalism is all about? Dispositions, we are told, are powerful properties, which bring about their manifestation; or at least they try. It is precisely this power or link towards their manifestation that the dispositionalists, but not the Humeans, should have available in their ontology. Therefore,

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

157

there should be no problem for dispositionalists in accepting a fourth stage, where the disposition is pushing, trying, or aiming to manifest.16 What’s the push? Actually, I have no reason to question this fundamental insight about dispositions. As will become clear later, I endorse this antiHumean intuition. Yet, I will show that it is not so clear how dispositionalists should conceptualize the pushing we have extracted from the antidote case. Can we grasp its nature further? No doubt, to capture its nature in terms of counterfactual conditionals is not possible for the dispositionalists. Claiming that the uranium aims to chain react means that, if the uranium were on its own (without the boron), it would melt, is the reductionists’, not the dispositionalists’, story. In fact, this story is what the reductionists have painstakingly tried to elaborate in their conditional analyses of dispositions, but this is what the dispositionalists believe to have proven impossible with cases like the antidote cases. Clearly, then, counterfactuals are not the tools to use when conceptualizing the aiming or trying to bring about the manifestation of a disposition. The obvious second place to look for a possible conceptual background to those pushes is, of course, the initial place of departure for both the antiHumean dispositionalists and the Humean reductionists. The major metaphysical difference between, on the one hand, the anti-Humean and, on the other hand, the empiricist or neo-Humean, is their belief or, respectively, disbelief in necessary connections in nature. Naturally, one might think, de re necessity has to play its role for the dispositionalists at some point and the place is precisely where reductionism (allegedly) fails; that is, in antidote cases. In other words, the disposition’s pushing, trying, attempting to manifest, which the Humean accounts do not accommodate, should be cashed out in terms of natural necessity. However, what I aim to show next and what constitutes a main element of this paper, is that metaphysical necessity is the wrong kind of “secret connection.” This type of necessity, even though it seems to have been such a boost for dispositionalists, because it promises to be a successful weapon against the Humeans, is of no use. In order to prove my claim I will look in detail at Brian Ellis’s theory of the dispositional pushes, because he does explicitly analyze them in terms of Kripkean metaphysical necessity.

4. Necessity cannot do the Job It is not unusual to think of powers in association with necessity. Even Harré and Madden, who wrote too early to be under the psychological influence of

_____________ 16

In a successful display of its manifestation the disposition will also be said to have pushed for its display – effectively.

158

Markus Schrenk

Kripke, subtitled their book Causal Powers with “A Theory of Natural Necessity.” However, I want to focus on a theory Brian Ellis has put forward. He relies heavily on Kripkean metaphysical necessity in order to capture the link between a trigger event and a manifestation event mediated by a disposition. Here is, first, his general claim regarding necessity: Essentialists have their own special brand of necessity. This kind of necessity has traditionally been called “metaphysical necessity.” (Ellis 2002, 110)

When Ellis talks about necessity (and he uses all of “physical necessity”, “natural necessity”, “de re necessity”, and “real necessity” as synonyms), he has Kripkean metaphysical necessity in mind, which is strongly associated with truth in all possible worlds: Real necessity is no less strict than any other kind of necessity. […] If essentialists are right, and the laws of nature are really necessary, then they must be counted as necessary in the very strong sense of being true in all possible worlds. Truth in all possible worlds is the defining characteristic of all forms of strict necessity. (Ellis 2002, 110; my emphasis)

Synchronic versus Diachronic Affairs: Yet, Kripkean necessity relates first of all natural kinds (the elite amongst the properties) to features (further properties) they possess essentially. That is, the relation of metaphysical necessity is typically a link between one property and another (or things and their properties) – for example, an electron necessarily having unit charge, protons necessarily having rest mass 1.6726 × 10-27 kg, water necessarily being H2O. Yet, if metaphysical necessity is typically attributed with respect to one property having another property, then metaphysical necessity is normally a synchronic business. In the case of dispositions, however, we are confronted with a different affair. There, we have characteristically a diachronic case of one property instance (or event) at t, namely the trigger (plus other activation conditions if needed), and another property instance (or event) at t+Vt, namely the disposition’s manifestation. Consequently, Ellis (and anyone else who thinks necessity can be of help in an account of dispositions) has, in a first step, to explain how the normally synchronic Kripkean necessity can be extended to diachronic trigger-manifestation events. Ellis has, in fact, a story to tell: not only are there natural kinds of objects that have certain properties necessarily (mostly powers in his view), there are also natural kinds of processes. And, here comes the crucial point, in the case of natural kinds of processes, two event types are indeed linked by metaphysical necessity. The disposition’s trigger event leads with natural necessity to the disposition’s manifestation event because this process is a natural kind itself: Suppose, for example, that p is a natural dispositional property that would be triggered in circumstances of the kind C to produce an effect of the kind E. Then the processes of this kind will themselves constitute a natural kind, the essence of which is that it is a display of P. (Ellis 2002, 158; my emphasis)

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

159

Therefore, […] for all x, necessarily, if x has p, and x is in circumstances of the kind C, then x will display an effect of the kind E. (Ellis 2002, 158; my emphasis)

Although these steps remain elliptical – is, for example, the move from synchronic to diachronic links warranted that straightforwardly? Is the step from natural kinds of objects to natural kinds of processes legitimate? – I will, for the sake of the argument, accept their tenability. Enough problems arise when we apply Ellis’s idea to antidote and similar cases. Necessity’s Failure: Remember that this detour has been taken in order to find out whether one specific conceptualization of the dispositional push that is needed for antidote cases is tenable. Presently, we aim to challenge Ellis’s position that conceives of the pushes as metaphysical necessity. I will now present two counterarguments against this view. Suppose there is a disposition P to react with E when in circumstances C. As I read Ellis, this is to say that there is a natural kind of process: the process from C events to E events. Further, if I interpret him correctly, C events and E events are joined as a matter of metaphysical necessity. Our problem is now that in antidote cases, E does not come about although C occurs. Yet, how can that at all be possible if C and E are linked by metaphysical necessity? Not even an antidote should be able to interfere with metaphysical necessity, should it? (The uranium pile is triggered, C, to chain react, E, for it has critical mass.) There are only three possible answers I can imagine but all lead into severe difficulties. (1) In antidote cases, not C but C* is realized, that is, not those sure-fire circumstances that, if realized, would definitely bring about E, but only those similar, that is, diluted, antidote riddled circumstances. Yet, even if this is so, remember that antidote cases must differ from cases where the trigger is not at all pulled. (Clearly, when C does not occur there is no problem in accommodating E’s non-occurrence.) Therefore, while C* are diluted circumstances they are still circumstances where the trigger, C, is pulled. So, C* has to be imagined after all as a case of C plus A (uranium above critical mass plus boron rods), say. However, and this is the crucial argument, necessity is monotonic: if C necessarily leads to E, so must C plus A. In fact, my argument is trivial and it is well known in a different disguise: necessities – in the following historical example of the analytic or de dicto kind – cannot handle cascading if-then sentences. Remember Goodman’s match: if match m had been scratched it would have lighted, but if match m had been wet and scratched it would not have lighted. Moreover, if match m had been wet and scratched and the surrounding temperature had been extremely high it would have lighted, and so on. Surely, none of the links in those conditionals can be of de dicto necessity. This is a message which has been frequently acknowledged and which was once a reason for Humeans, like David Lewis, to develop semantics for counterfactuals. But why should, now, metaphysical

160

Markus Schrenk

necessity be able to handle the very same sort of difficulty (antidotes)? It is not clear that it could, yet, this is what anti-Humeans tacitly assume. (2) One might have the following intuitive reaction towards the first argument: C has never been the correct first relatum of the necessary relation under concern. Rather, what is linked necessarily to E is C and the absence of any interfering factor. Then C* poses no problem because C* (=C+A) simply is C with an interfering factor. Yet, this move does not do the trick either.17 First, we get the problem of late preventions: suppose Bird’s uranium pile was well above critical mass at time t and free of any interfering factor. However, at time t* boron rods are inserted. We can assume that this was still well in time to prevent E at t+ t. Was the necessity between C and the absence of any interfering factor and E at t interrupted later at t*? The curse of diachronicity (as opposed to synchronicity) strikes back. The second problem is that metaphysical necessity is, next to being monotonic, discrete: that some two specific properties or event types, an uninterfered with C and an E, are necessarily linked (and hence conjoined in all possible worlds) has no bearing whatsoever on the instantiations and correlations of any other properties or event types. As a consequence, the natural kind of process from uninterfered with C to E with its internal metaphysical link cannot help to explain a disposition’s power in non ideal cases. Yet, this needs to be explained for we know many cases of partial displays in impure, interfered with C cases: there can be partially dissolved sugar (because of supersaturated water); smouldering, yet not burning inflammables (in case of low oxygen levels); lower than expected accelerations (because of counteracting forces); etc. However, a partial E* cannot be explained by a partial, impure C* because of necessity’s discreteness: only pure (uninterfered with) Cs and pure Es are linked by metaphysical necessity. Yet, C* bringing about of E* remains unexplained.18 Spelled out in anti-Humean terms (and we have accepted an anti-Humean picture as our working hypothesis), the problem is that if there’s only a necessary link between C and the absence of any interfering factor and E then, because of necessity’s discreteness, C*’s (the impure case’s) push or oomph to-

_____________ 17

18

To be fair to Ellis I have to confess that I omitted a line from one of his quotes where he already explicitly excludes interferences in the fashion of attempt (2): “Therefore, […] for all x, necessarily, if x has P, and x is in circumstances of the kind C, then x will display an effect of the kind E, unless there are defeating conditions that would mask this display.” (Ellis 2002, 158; my emphasis) As I will show now, this additional line still does not resolve our problems. Andreas Hüttemann has convincing arguments for dispositional realism which revolve around what he calls “CMDs”: continuously manifesting dispositions. The upshot here is that these continuous displays cannot be captured by metaphysical necessity.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

161

wards an impure E* remains mysterious. Y (Cx Ex) does not explain C* bringing about E*.19 (3) You might want to put forward the following idea at this point: metaphysical necessity holds between an infinite number of pairs of ultra fine grained event-types: every single member of C, C’, C’’, etc. ad infinitum is linked with metaphysical necessity to the respective succeeding fine grained event-types E, E’, E’’, and so on. C-events, conceived in this way, are not just the instantiation of a single property but are instantiations of ultra fine grained situation types encompassing fairly large space-time areas.20 Any minimal alteration to a particular situation would qualify it as the instantiation of a different situation type. In this picture, metaphysical necessity could be ubiquitous and it could, hence, explain the bringing about of each E (E’, E’’, …) by each C (C’, C’’, …). However, this suggestion is unsatisfying in many respects. First, it is almost certain that any of these ultra fine grained events occurs only once in the whole of the world’s history. Therefore, second, what happens regularly and what happens with necessity falls apart because nothing ever happens repeatedly. (Or only trivially so: for each fine grained C and E, “all Cs are Es” is only a trivial universal truth if C and E happen only once). Third, properties, dispositions, or particulars having those features lose their impact. It is always only the whole situation which necessitates the next situation so that properties or individuals are not, after all, powerful themselves (or only in a derivative way). Yet, it is an essential part of most anti-Humean or dispositionalist theories that individual objects and their various powers are responsible for the goings on in the world (cf. Mumford 2004, Ellis 2002, Molnar 2003). Finally, fourth, I have the feeling that the resulting picture would be more a kind of über-Humeanism rather than an anti-Human view. What we create if we endorse this picture is a kind of hyper-mosaic: co-instantiations of event-types that are the same in each and every possible world. Yet, just by multiplying these event pairs in an infinity of possible worlds they do not thereby gain any intrinsic link, push, or production character: they somehow still remain co-instantiations of unrelated facts. These are what I believe to be decisive reasons against metaphysical necessity being the appropriate fuel for the power of dispositions. Necessity is the wrong kind of “secret connection” because it is in a dilemma. Either it is too strong (monotonicity): C to E although A, or it has no power (discreteness): partial or impure or too short instantiations of C. Metaphysical necessity is

_____________ 19

20

In yet other words, the necessary relation between C and E cannot have any influence on a situation in which its first relatum, C, is not realised. Over-exaggerating the affair a little, one can make the following parallel: that water is necessarily H2O has nothing to do with alcohol being C2H5OH (and even muddy water or tea might already lose the (necessary) link to H2O). These could even be states of the whole world at specific times.

162

Markus Schrenk

modelled too much after logical or conceptual necessity. The Kripkean antiHumean move is, hence, of no use for the dispositionalists. It might have been psychologically important for anti-Humeans to gain courage to stand up against the Humean creed but philosophically it has, at least for dispositions (or causation or causal laws or any diachronic business), no impact.

5. Other Suggestions on how to Conceptualize the Dispositional Push Some dispositionalists have conceptualized dispositions along lines other than metaphysical necessity. I will discuss only one such theory from the recent literature: Stephen Mumford’s “dispositional possibility” and “dispositional necessity.” We need some kind of dynamic anti-Humean de re link between events that can explain the pushes we need for a description of dispositions in antidote cases. This dynamic de re link cannot be metaphysical necessity. It has to be some intra-world relation. Now, Mumford does explicitly underline the difference between synchronic and diachronic necessities. After having given two examples of the known synchronic species he introduces a novel “necessary connection that dispositions or causal powers bring to the world” (Mumford 2004, 168). He underlines that dispositional properties “are typically dynamic”, i.e., that they are “responsible for, or productive of, changes in those and other particulars.” Mumford denies that synchronic necessities (including metaphysical necessity) can have this dynamic aspect (Mumford 2004, 168). Yet, in his closer portrayal of this very promising new dynamic de re link, Mumford, unfortunately, in my opininon, comes close to adopting the traditional necessity view. He first distinguishes two subspecies of his new dynamic de re link: Dispositional possibility: The having of one property may dispositionally make possible the having of another property. For example, being fragile makes possible being broken. Dispositional necessity: The having of one property may dispositionally make necessary the having of another property. For example, having gravitational mass necessitates attraction of other objects. (Mumford 2004, 177)

The intuitions behind these characterizations are clearly akin to mine. Somehow a dispositional push needs to be captured. Yet, I believe that Mumford lapses back onto the old static necessity in his characterization of the second part of the distinction. What he labels “dispositional necessity” seems to be nothing but the old metaphysical necessity. That two objects have gravitational mass necessitates attraction between them, fair enough, but not as a matter of

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

163

any new dynamic dispositional necessity. Rather, if we also have metaphysical necessity still available, mass and gravitational attraction could be seen as being linked by metaphysical necessity. Gravitational mass metaphysically necessitates attraction, that is, in all possible worlds, where there are gravitational masses they attract each other. Mumford’s valuable dynamic force has, however, not entirely disappeared. It can be found in the attraction itself! More radically, I believe that the attraction has to be identified with the dynamic de re link he is after. Attraction is a sample of this novel de re link. I come back to this idea shortly (6.). First, I want to turn towards Mumford’s first kind of dispositional link: dispositional possibility. Another quote concerning this connection reveals again that Mumford has seen the need for an anti-Humean connection in nature that is different from any form of necessity: There is a connection between these two properties [being fragile and being broken; MAS] that is more than bare compatibility although it is less than necessitation, as being fragile does not necessitate being broken. […] Fragility has a causal connection with being broken. […] Thus we need a relation that represents connections in nature that are less than necessity […] but more than mere unconnected compatibility […] [Dispositional possibility] is the connection. (Mumford 2004, 177)

I believe that Mumford’s dispositional possibility comes as close as possible to capturing the pushes we need in antidote cases. He might very well succeed in distinguishing a present, yet, not triggered and a present, triggered, yet, not manifesting case: in the latter dispositional possibility could be seen to be active. Against this background, it should also be clear why I have earlier on subsumed attraction under dispositional possibility. Attraction might or might not lead to movements, deformations, or holding other forces in check. Yet, clearly, attractions, like pushes, are a kind of anti-Humean glue in nature that are less than necessity but that link otherwise unconnected entities.

6. Why Forces might be the Right Kind of Entities For me, too, Rhodos is here and it is time for my own jump. My plan is to make the proverb of my paper’s pretentious title the motto for a possible solution. What we must find is a proper conceptualization of the pushes dispositions afford when they are triggered in antidote circumstances. What we need are non-modal pushes without any connotation of necessity or direct reference to other possible worlds; pushes that have their power to jump here and now. A special antidote case I have already described can serve as a model for a possible solution: the electron in an electric and a gravitational field held static in a force equilibrium. The electron’s Coulomb capacity faces an antidote: a

164

Markus Schrenk

gravitational force. Coulomb’s disposition is present, triggered but not displaying – its display would be the electron’s acceleration. What, then, is its push? A Newtonian force!21 And forces seem to be exactly the kind of thing we need in order to conceptualize the pushes dispositions afford when triggered: intuitively, forces have the relevant ‘oomph’ while not extending their power to possible worlds. However, I want to be very cautious and my aim is only to explore in a preliminary manner the philosophical potential of an account of forces. I am afraid I do not have a full-blown account to offer yet. For example, I am not saying that dispositional pushes are Newtonian forces. For a start, my claim is much weaker: the idea of Newtonian forces and the intuitions we have about forces in everyday life form the right conceptual background to think of dispositional pushes. There are many open questions a forces account would bring with it. A crucial question is, for example, how the insights we might gain from the electron case carry over to other cases that do not obviously involve forces: a wire’s being live, the uranium pile, or even mental dispositions. Are forces and the intuitions we attach to them mere metaphors for those dispositions? Or do other dispositions have to be reduced to more fundamental ones, which, in turn, can be analysed in terms of Coulomb’s capacity, gravitational mass, etc., i.e., capacities that do involve forces? After all, solubility, for example, is a matter of molecular structures, chemical bonds, and forces between molecules: water dipoles tear, qua Coulomb force, the Na+ and Cl- ions apart. The possibility of analyzing everyday dispositions in terms of molecular goings-on that involve forces could be taken as a warranty for metaphorical talk about dispositional forces even at the macroscopic level. However, even if this is the route to take there is a multitude of further problems and unsolved questions. The most pressing one is this: forces are no longer respectable entities in current physical science. On the contrary, it seems physics has nowadays abandoned talk about forces entirely. In macroscopic physics energy-based accounts (Lagrangians and Hamiltonians) replace forces or, in the General Theory of Relativity, geometry replaces forces, and quantum phenomena are best described in terms of probabilistic functions of initial conditions. Forces, one might (radically) conclude, do not exist. Like phlogiston they have been deleted from scientific ontology. If so, the rug is pulled out from under my account. Luckily, there is a growing community of philosophers who defend forces against reduction: for example, Bigelow, Ellis, and Pargetter in “Forces” (Bige-

_____________ 21

Extreme anti-realists or instrumentalists about component forces (such as Cartwright claims to be) will not be happy with what I am going to say here. Naturally, I have to be a realist about component forces but I cannot defend this realism here.

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

165

low et al. 1988) and, very recently, Jessica Wilson in “Newtonian Forces” (Wilson 2007). There is no space to discuss their arguments here, but I hope that they can rescue forces from phlogiston’s fate. A further oddity about dispositional forces is that an account of them seems to presuppose that dispositions rest inactive (asleep) until their force is triggered. They are, in some sense, constantly poised to make things happen. Only if certain conditions are met do they start pushing for their manifestation (i.e., until they are woken up). Take inflammability as an example: a match is not constantly on the verge of burning; the push to burn only occurs once certain sufficient conditions (scratching) are met. One can almost say that dispositions have two conditionals attached to them: if sufficient trigger conditions are met they push to their manifestation and, furthermore, if their push to manifestation is unchallenged by antidotes they manifest themselves. Actually, this activation idea is not at all unfamiliar. A system might have certain potential energy that can only be released when certain activation energy is put into the system. Thus, a dispositional ascription is an ascription of potential energy to an object and the trigger specifies the activation circumstances.22 To end on a more positive note, forces do not only come with problems; they might also have the potential to solve an infamous counterargument against pan-dispositionalism – the view that all properties are dispositional in nature. The argument often referred to as the “always packing, never travelling” argument is this: if the manifestation of all dispositions is yet another disposition then no manifestation will ever really be manifested. The world would be in a state of constant flux. A forces account has the potential to solve this riddle: force equilibria bring things to a halt. Note also, that we find historical theories of dispositions which my forces account resembles. Compare, for example, Leibniz’s active force and Aristotle’s dynamis as presented in this volume by Michael-Thomas Liske and Ludger Jansen respectively.

7. Conclusion Hume’s arguments against connections in nature have predominantly been read as a rejection of necessity in nature. Yet, necessity, especially when formulated in possible worlds talk, is not the only anti-Humean connection possible. Reconsider Hume: The scenes of the universe are continually shifting, and one object follows another in an uninterrupted succession; but the power or force, which actuates the whole machine, is entirely concealed from us, and never discovers itself in any of the sensible

_____________ 22

I owe this suggestion to Stephen Williams.

166

Markus Schrenk

qualities of body. […] External objects as they appear to our senses, give us no idea of power or necessary connection. (Hume 1777, 63-4)

Even if this might be the historically correct text exegesis we should not think that Hume is giving synonyms when he says “power or necessary connection” but alternatives! Power and necessary connection should not be identified with each other. Events happening due to forceful dispositional pushes should be thought of as more than mere coincidences but as less than being necessitated.

Literature Armstrong, David. 1997. A World of States of Affairs. Cambridge: Cambridge University Press. Berg, Jan. 1955. “On Defining Disposition Predicates.” Analysis 15: 85-89. Berg, Jan. 1958. “A Note on Reduction Sentences.” Theoria 24: 1-8. Berg, Jan. 1960. “Some problems concerning disposition concepts.” Theoria 26: 3-16. Bigelow, John, Brian Ellis, and Robert Pargetter. 1988. “Forces.” Philosophy of Science 55: 614–630. Bird, Alexander. 1998. “Dispositions and Antidotes.” Philosophical Quarterly 48: 227-234. Bird, Alexander. 2000. “Further Antidotes: A reply to Gundersen.” Philosophical Quarterly 50: 229-233. Burks, Arthur. 1951. “The Logic of Causal Propositions.” Mind 60: 363-382. Burks, Arthur. 1955. “Dispositional Statements.” Philosophy of Science 22: 175-193. Carnap, Rudolph. 1936. “Testability and Meaning I” Philosophy of Science 3: 419-471. Carnap, Rudolph. 1937. “Testability and Meaning II.” Philosophy of Science 4: 1-40. Choi, Sungho. 2003. “Improving Bird’s Antidotes.” Australasian Journal of Philosophy 81: 573580. Choi, Sungho. 2006. “The Simple vs. Reformed Conditional Analysis of Dispositions.” Synthese 148: 369-379. Ellis, Brian: 2001. Scientific Essentialism. Cambridge: Cambridge University Press. Ellis, Brian: 2002. The Philosophy of Nature. Chesham: Acumen. Essler, Wilhelm 1975. “Die Kreativität der bilateralen Reduktionssätze.” Erkenntnis 9: 383392. Essler, Wilhelm and Rainer Trapp. 1977. “Some Ways of Operationally introducing Disposition Predicates with Regard to Scientific and Ordinary Practice.” Synthese 34: 371396. Fales, Evan. 1990. Causation and Universals. London/New York: Routledge. Gunderson, Lars. 2000. “Bird on Dispositions and Antidotes.” Philosophical Quarterly 50: 227229. Harré, Rom and E. H. Madden. 1975. Causal Powers. A Theory of Natural Necessity. Oxford: Basil Blackwell. Heil, John 2005. “Dispositions.” Synthese 144: 343–356. Hume, David. 1777. Enquiries concerning Human Understanding and concerning the Principles of Morals. Ed. L. A. Selby-Bigge and P.H. Nidditch. Oxford: Oxford University Press. Kaila, Eino. 1939. Den mänskliga kunskapen. Stockholm. Kaila, Eino. 1942. “Der physikalische Realitätsbegriff.” Acta philosophica fennica 4: 33-34. Kaila, Eino. 1945. “Wenn ... So ..” Theoria 11: 88-98. Kaplan, David. 1989. “Demonstratives.” In Themes from Kaplan, eds. J. Almog, J. Perry, and

From Reductionist Semantics to a Realist Ontology of Forceful Dispositions

167

H. Wettstein. Oxford: Oxford University Press. Lewis, David. 1973. Counterfactuals. Oxford: Oxford University Press. Lewis, David. 1986. Philosophical Papers, Volume II. Oxford: Oxford University Press. Lewis, David. 1997. “Finkish Dispositions.” Philosophical Quarterly 47: 143-158. Lipton, Peter. 1999. “All Else Being Equal.” Philosophy 74: 155-68. Malzkorn, Wolfgang. 1999. “On the Conditional Analysis of Dispositions.” In Akten des 22. Internationalen Wittgenstein-Symposiums, eds. Uwe Meixner and Peter Simons. Kirchberg/W. Malzkorn, Wolfgang. 2000. “Realism, Functionalism and the Conditional Analysis of Dispositions.” Philosophical Quarterly 50: 452-469. Martin, Charles B. 1994. “Dispositions and Conditionals.” Philosophical Quarterly 44: 1-8. Mellor, Hugh. 1999. “Dispositions.” Handout for a talk given at King’s College London, 07th December 1999. Mellor, Hugh. 2000. “The Semantics and Ontology of Dispositions.” Mind 109: 757-780. Molnar, George. 1999. “Are Dispositions Reducible?.” Philosophical Quarterly 49: 1-17. Molnar, George. 2003. Powers: a Study in Metaphysics. Ed. Stephen Mumford. Oxford: Oxford University Press. Mumford, Stephen. 2001. “Realism and The Conditional Analysis of Dispositions: Reply to Malzkorn.” Philosophical Quarterly 51: 375-8. Mumford, Stephen. 2004. Laws in Nature. London: Routledge. Pap, Arthur. 1955. Analytische Erkenntnistheorie. Wien. Pap, Arthur. 1963. “Reduction Sentences and Disposition Concepts.” In The Library of Living Philosophers, Vol. XI: Rudolf Carnap. Ed. P. A. Schilpp. La Salle, Il. Prior, Elisabeth, Frank Jackson, and Robert Pargetter. 1982. “Three Theses about Dispositions.” American Philosophical Quarterly 19: 251-257. Psillos, Stathis. 2002. Causation and Explanation. Chesham: Acumen. Salmon, Nathan. 1981. Reference and Essence. New York: Prometheus Books. Schrenk, Markus. 2000. The Conditional Analysis of Dispositional Predicates. Unpublished Master’s Thesis, Bonn University. Available in German on Schrenk’s website: http://www.nottingham.ac.uk/philosophy/staff/schrenk/MarkusSchrenk/Home. html Sellars, Wilfrid. 1958. “Counterfactuals, Dispositions, and the Causal Modalities.” In Minnesota Studies in The Philosophy of Science, Vol. II. Eds. Herbert Feigl, Michael Scriven, and Grover Maxwell. Minneapolis: University of Minnesota Press: 225-308. Spohn, Wolfgang. 1997. “Begründungen apriori – oder: Ein frischer Blick auf Dispositionsprädikate.” In Das weite Spektrum der analytischen Philosophie – Festschrift für Franz von Kutschera. Ed. Wolfgang Lenzen, 323-345. Berlin. As “A Priori Reasons: A Fresh Look at Disposition Predicates” available in English on Spohn’s website: http://www.unikonstanz.de/FuF/Philo/Philosophie/Spohn/spohn_papers.shtml Stalnaker, Robert. 1968. “A Theory of Conditionals.” Studies in Logical Theory, American Philosophical Quarterly 2: 98-112. Storer, Thomas. 1951. “On Defining ‘Soluble’”. Analysis 11: 134-137. Storer, Thomas. 1954. “On Defining ‘Soluble’ – Reply to Bergmann.” Analysis 14: 123-126. Trapp, Rainer. 1975. “Eine Verfeinerung des Reduktionssatzverfahrens zur Einführung von Dispositionsprädikaten.” Erkenntnis 9: 355-382. Watkins, J. 1958. “Influential and confirmable metaphysics.” Mind 67: 344-365. Wedberg, Anders. 1942. “Review of Kaila 1939.” Journal of Symbolic Logic 7: 43-44. Wedberg, Anders. 1944. “The logical construction of the world.” Theoria 10: 216-246. Wilson, Jessica. 2007. “Newtonian Forces.” The British Journal for Philosophy of Science 58: 173205.

Ascribing Dispositions STEPHEN MUMFORD … upon the whole, there appears not, throughout all nature, any one instance of connexion which is conceivable by us. All events seem entirely loose and separate. One event follows another; but we never can observe any tie between them. They seem conjoined, but never connected. And as we can have no idea of any thing which never appeared to our outward sense or inward sentiment, the necessary conclusion seems to be that we have no idea of connexion or power at all, and that these words are absolutely without any meaning, when employed either in philosophical reasonings or common life. (Hume 1748, 74)

1. Introduction Contrary to what some Humeans would have us think, dispositions are ascribed with relative ease, suggesting that there is at least some idea of a disposition that is in ordinary use. Where this idea comes from, and how warranted it is, is not clear. Certainly we have some sense of the world around us containing threats and promises but, if certain forms of empiricism are to be believed, there is only a limited degree of empirical justification for our dispositional vocabulary. If we are just inferring from past regularities in the behaviour of things – in objects, substances and persons – to properties of those things that necessitate that kind of behaviour, then the inference is unsafe and more likely the product of ‘custom and habit’ than of reason or experience. An empiricist might say that in ascribing dispositions we are merely projecting our own expectations of future behavior on to the objects themselves. But neither experience, nor reason, will vindicate any such ascription. In this paper, I attempt an answer to this kind of empiricist challenge to disposition ascription. The answer has a number of dimensions. In the first place, I will argue that the classical renunciation of the idea of necessary connection is far from safe.1 There are at least some empirical sources of the idea of power. But I will also concede that experience alone may be inadequate to

_____________ 1

I will not be tackling the detail of Hume’s argument, however. For that see Mumford 2004, chap. 4.

Ascribing Dispositions

169

account for all our disposition ascriptions. Indeed, it is characteristic of dispositions that their ascriptions are often warranted when they are empirically inaccessible and unverifiable. The mark of a dispositional ontology is the acceptance of verification transcendent power ascriptions. But I will then argue that the ascription of unverifiable dispositions can be warranted even in such circumstances if there are good reasons to accept a dispositional ontology. I proceed to present an argument for the acceptance of such an ontology. It can be characterised as a transcendental justification. The dispositional ontology is a fruitful and explanatory one. Given that metaphysics is on the whole a non-empirical study of the nature of reality, then such factors as productivity and explanatory unity are the best grounds on which to accept a metaphysics of dispositions or causal powers. If, for such reasons, there are good grounds for the dispositional ontology, then disposition ascriptions should be seen as justifiable even in those cases where they are empirically inaccessible. It should be clear from what has been said that dispositions and powers are being treated as equivalent things. It should also be clear that such dispositions are being treated as mind-independent features of reality that are possessed by or instantiated in particulars. Much more can be said about how they relate to the other categories such as that of a property but that will be among the things that a correct theory will clarify. I will skip further preliminaries, therefore, and instead consider the non-empirical nature of disposition ascriptions.

2. Empirically Inaccessible Dispositions The realist about powers is willing to ascribe various classes of empirically inaccessible dispositions or powers, ranging from those that are never tested, to those that are tested and fail to manifest, to those that cannot be tested. Before looking at such cases, however, it is worth considering the general issue of what counts as an appropriate way to empirically access a power. One reason that the conditional analysis of powers seemed so appealing to those of a broadly empiricist bent was that it appeared to cash out power ascriptions in empirically acceptable ways. The idea was that the presence of powers could not be verified if they were supposed to have modal properties. But we could, instead of that, test for the presence of the power by testing whether it manifests in the appropriate stimulus conditions. A conditional analysis naturally results from this, where the antecedent of the conditional names the appropriate stimulus conditions and the consequent names the appropriate manifestation. Ryle (1949) is the most obvious exponent of such an analysis. A particular x thus has disposition D if and only if it gives the

170

Stephen Mumford

appropriate manifestation M upon occurrence of stimulus S. The relevant conditional is thus: x (Dx

(Sx

Mx))

Where disposition ascriptions are understood this way then the required test will be obvious: subject the relevant particular to the requisite stimulus conditions. It will display the appropriate manifestation if and only if it has the disposition. But if this kind of testing really were to be the only way of gaining empirical evidence for the correctness of disposition ascriptions then a number of cases become problematic. In the first place, there are those disposition ascriptions that, as a matter of contingent fact, have not been tested. We would be happy, for instance, with an ascription of fragility to an old vase. This vase, perhaps for the very reason that it is believed to be fragile, is never allowed to be in the appropriate stimulus conditions for fragility. Hence it never manifests its fragility. This creates more problems than just the mere one of how we are able to ascribe such dispositions in cases that have not been tested. Sx Mx is true whenever Sx is false, on a material reading of the conditional. But that would make everything fragile that has not been tested for fragility. One might then say in response that the conditional would have to be stronger than material.2 But that looks like a conditional with some modal power and it might then be wondered how that itself can be empirically known. What, in other words, would be an empirically acceptable truthmaker for any such stronger-thanmaterial conditional? The realist about powers has an answer: the power is the truthmaker for any such conditional. But this is precisely what the empiricist was seeking to avoid. The second kind of example makes the case for empirically inaccessible dispositions even stronger. This is the kind of instance where the appropriate test for the disposition does occur but the manifestation still does not occur because of some interfering factor that prevents the manifestation. There is a range of such factors that might prevent manifestation. Some would be purely contingent, as when a match does not light when struck because it is damp. But some such preventions might be quite deliberate. Bird’s (1998) antidotes are a case where a disposition is possessed, tested and retained but still does not manifest. I take arsenic, for instance, but it does not manifest its power to poison me because I have also taken the antidote, British anti-Lewisite. Wright (1991) had already discussed many cases of this ilk. A soldier is brave, let us assume, but on the few occasion where his bravery is put to the test, he

_____________ 2

Carnap’s (1936/7) reduction sentences were, however, an attempt at an account that is restricted to material conditionals.

Ascribing Dispositions

171

is drunk or otherwise unable to produce the appropriate response in brave actions. Next we come to the finkish cases, from Martin (1994). These are distinguished from the cases of ‘mere’ prevention by the factor that it is the very stimulus conditions themselves that prevent the manifestation, for instance, by removing the disposition. Mellor’s example illustrates this nicely (1974, 173). A nuclear power station is disposed to explode but it never does so. If all works well, safety mechanisms will cut in if the reactor is about to go critical, so it never does. Does that mean that it is not disposed to explode? That would be absurd because the safety mechanisms would then serve no purpose, which they clearly do. One final kind of case is worthy of note. It might be regarded as the opposite of the untested disposition. What should we make of continuously manifested dispositions? Is there a meaningful way of empirically testing for them? Gravitational mass is usually understood to be dispositional in nature, yet every material thing must manifest it all the time: it attracts things constantly. That would suggest that there was no non-trivial conditional that would set up test conditions for it. It might of course be thought that there was no special problem here, because the manifestation can always be detected, but how could one test the credentials of such a power as a power? How could such a thing be correctly regarded as a power if its nonmanifestation is counterfactual? In all these cases, it seems that the disposition ascription is, to a degree, evidence transcendent. Certainly the truth of the ascription cannot be verified in a narrowly empirical way. There is, of course, some empirical basis for such ascriptions, as I will outline in the next section. But ultimately, I argue, it will be for theoretical and metaphysical reasons that we should accept an ontology of causal powers.

3. Sources of the Idea of Power Nevertheless, there are some legitimate empirical sources of the general idea of power. One obvious source is analogy. While this particular vase may never have been tested for fragility, others have been so tested and have broken. If our untested vase is enough like the broken past vases, then we can very easily form the idea of untested powers that are ready to be manifested if all the conditions are right. Other cases give us the idea that there are unmanifested powers and also that there are appropriate stimulus conditions for their manifestation. The importance of this latter point is almost entirely overlooked as the focus remains on conditional statements: on antecedent and stimulus conditions. The insurmountable problem of the naïve conditional

172

Stephen Mumford

analysis of dispositions is that we cannot specify non-trivially all the conditions that would have to be right for the manifestation to occur. We cannot rule out a potential infinity of preventers of manifestation. What we must add, therefore, is some existential quantification to form a Ramsey sentence. Experience shows us that, despite all the potential preventers, there are appropriate ‘ideal’ conditions for the manifestation to occur and this claim must be added to any conditional analysis if it is to work (Mumford 1998, 87-91). Hence, experience delivers the following idea to us: C x (Dx

(Cx & Sx

Mx)).

Added to simple analogy, however, there are many other empirical cues, sometimes theoretically based, upon which we make disposition ascriptions. As we investigate the way that the world works, we will note patterns of behavior. The periodic table, for instance, groups elements with like constitution and is able to explain their behavior in a way that becomes projectable. Elements in the same column will have the same number of free spaces in their outer shells of electrons, thus permitting similar kinds of chemical reaction. This allows us to understand the behavior of possible elements that perhaps exist only theoretically. As another example, I may see the insides of a machine and, if my understanding is developed enough, I can see what powers this machine has. I need not have put it directly to the test or know whether it has actually performed such a function, but I can nevertheless understand that it has the power to do, just from my mechanical knowledge. Of course, in such cases our knowledge of unmanifested powers may just be based on our beliefs about the constituent powers or of the underlying laws of nature. But why should these not be legitimate empirical sources of the idea of unmanifested power? There is a further source of the idea of power, of which I have been persuaded recently.3 While Hume may have been right that we do not see any necessary connection between distinct existences (1739-40, 88), his mistake, and the mistake of those who followed him, might have been his concentration on visual experience.4 While we cannot see powers, we might be able to feel them. As an example of this, consider two teams in a tug-of-war contest.

_____________ 3 4

By Markus Schrenk, in conversation. Exactly what Hume thought, whether he was a causal sceptic or causal realist, remains a matter of extreme controversy among Hume scholars (see Strawson 1989). The position for which I am offering an alternative is a position that until recent years was taken as an orthodox interpretation of Hume. If this, previously standard, reading of Hume is wrong then I will be happy to concede that Hume himself was not a Humean in the traditional sense. My own understanding of Hume is, however, that the traditional reading is correct.

Ascribing Dispositions

173

The rope between the teams is taut and the teams are straining with all their might. But the contest is evenly balanced and neither team moves forward. The contest reaches an equilibrium state where nothing at all appears, visually, to be happening. But the team members can feel the force of the pull exerted by the opposition. They can also feel the power in their own legs and arms. While the teams are stationary, this power remains a power for only a possible manifestation. Both the power exerted on them, and the power that they exert, has merely possible manifestations. Armstrong (1997, 211-19) made much of such bodily experience in the case of causation and thought that we could acquire direct experience of causes. But the tug-of-war example shows that if we can gain the idea of causation through the sense of touch (though I acknowledge that Armstrong’s claim is controversial), we can just as easily gain the idea of a power or disposition through touch. Indeed, I would venture that the idea of power is prior to the idea of cause because, as I will argue below (§6), the best theory of causation is one based on powers. Hume professes that he has no idea of necessary connection but there is some plausibility that really he does have such an idea. While he may not see the necessary connection between two billiard balls that collide, he might nevertheless gain the idea of necessary connection in his own bodily interactions with the world around him. Hume speaks as if he sits above the causal order of things, as an inert and passive observer. If he put himself in the billiard ball’s position, interacting with other objects and feeling the force, then he would have found it much easier to form the idea of power. Hume, as an embodied person, was of course in exactly that kind of condition so we should conclude that, despite his sincere confession, he no doubt did have an idea of power and of necessary connection after all. Overall, the empirical legitimacy of the idea of power is strong. Yet it still might not be strong enough for some of the most extreme forms of empiricism, such as verificationism or its modern forms in anti-realism. The verificationist still has trouble, for instance, with the idea of finkish powers. And what of a case of a unique power, the like of which has never been manifested? In such cases one cannot say that one has experience of the power in similar cases. There could also be powers which have yet to be discovered and for which we as yet have no concept. The history of technology suggests a quest to discover and harness new powers, as we did in the case of electricity. Isn’t it plausible that there are still such powers lying in wait for us? I have conceded that by strict empiricist standards there are powers that are untested or untestable, unknown and perhaps even unknowable. The justification for this view has to be found, I argue, in the overall metaphysical case for a powers ontology. If we find good philosophical reasons for why the world is a world of powers, then we can find sense in untestable and unknowable in-

174

Stephen Mumford

stances. In the remaining sections of this paper, therefore, I will explain the attractions of an ontology of real dispositions.

4. The Case for a Dispositional Ontology I have tried to establish thus far that although there are some sources of the idea of power in our experience, the case for the existence of powers or dispositions is not an immediately and conclusively empirical one. There is at least some credibility to the thought that we could have all the worldly experiences that we actually have and yet the world would still not be a world that contains real dispositions. The disagreement between dispositionalists and Humean anti-dispositionalists is very unlikely to be over the nature of our experience.5 Rather, the disagreement will be over the metaphysical facts underlying that experience. Hence it would be unreasonable to think that Lewis’s position of Humean Supervenience (1986a, xi), for example, would be a disagreement with the dispositionalist over any of the empirical facts. How then can we settle this issue? The ultimate judgement, I contend, is a metaphysical one. Whether or not the world is a world containing dispositions is not a matter we will settle empirically but only in the court of metaphysics. If a metaphysics of causal powers is superior to its rivals, then that will be good grounds for saying that there are causal powers. This claim requires at least some comment on theory choice in metaphysics. We need to know what it is for one metaphysical position to be superior to another and how the quality of superiority relates to the matter of truth. Such issues are difficult and complicated. I can only state rather than argue for my view. My position is a realist one in which, although the truths of metaphysics cannot be accessed by empirical means alone, it is nevertheless the case that there is metaphysical truth. This truth should be understood broadly as correspondence, for example, it is true that universals exist if and only if universals do indeed exist. If we have an attraction to a correspondence theory of truth generally then we should not be put off by it in metaphysics simply because we cannot know the truths empirically. We have, therefore, to employ at least some non-empirical investigation if we want to discover the truths of metaphysics. We have to consider arguments and we also have to look at the strength and explanatory power of a metaphysical position. These would be pragmatic considerations in favor of a view and they alone do not guarantee its truth. If they did so, that would suggest a

_____________ 5

Though this is not to claim that there are no interesting discussions to be had about whether we experience causal powers, as I indicated in the previous section.

Ascribing Dispositions

175

coherence theory of truth where the theory that came out best on pragmatic criteria was pronounced true on those very grounds. Rather, I take such pragmatic factors to be the best, though fallible, indicators of where the real truth – truth as correspondence – lies. Although much more could be said to justify the claim that pragmatic factors, such as explanatory power, are the best, though fallible, indicators of truth in metaphysics, my aim in the rest of this paper is to argue that an ontology of causal powers will come out best on such criteria. The argument is that if we accept that the world contains real dispositions or causal powers then we will be able to generate a plausible and unified account of various outstanding metaphysical problems. If there were real dispositions in the world, they would explain what causes are, what laws are, events, properties, the de re modal features of the world, and so on. If dispositions are able to explain all these things, then we would have a good, though fallible, transcendental argument for their existence. My remaining task, therefore, is to show how an ontology of dispositions might be able to deliver all this.

5. Properties A particular is disposed to do certain things in virtue of its properties. When something is red, it is able to play a certain causal role in virtue of being red. Included in the role is the power to cause certain sensations in perceivers. When something is circular, it is able to roll smoothly on a flat surface. Something that is square will not have this power but it will have others. This suggests a theory of what properties are in themselves. One theory that suggests itself is that the identity of the property can be given in terms of its causal role, where the role is understood in dispositional terms. A circular thing need never actually roll on a flat surface but while ever it is circular it is able to do so. Identity conditions for properties are delivered if one thinks it plausible that properties F and G are identical if and only if they have all the same causal powers. Reflection on this principle shows its plausibility. Could something have the property of circularity that did not have the disposition to roll smoothly on a flat surface? If it cannot roll smoothly is it still really circular? If it rolls with a bump, but is on a flat surface, then doesn’t it have a corner? And have we no choice but to say that if two properties have different causal powers then they are different properties? To say that the identity of a property is dictated by its causal powers is one thing, but a more radical view is to say that a property is constituted by those powers. This was a view that Shoemaker (1980) once held: that a prop-

176

Stephen Mumford

erty is just a cluster of causal powers.6 There is something more for the powers theorist in this account because it means that an ontological reduction is in the offing. If one accepts powers as one’s fundamental ontological category, then properties – and ultimately relations also – are no addition of being. They would be an ontological free-lunch. One would not need properties, as a separate and distinct kind of thing in the world, above the powers. Simple properties would be clusters of single, simple powers. Complex properties would be clusters of many powers. And some properties would resemble in virtue of having some shared powers in their respective clusters (Mumford 2004, 170-4). If one were not to go for such an account of properties, then they would have to be something extra to their causal role. But this creates difficulties. Armstrong, for instance, allows that the different properties are just primitively numerically different (1983, 160 and 1989, 59). Lewis (forthcoming) has a similar line, accepting that the identity of a property is a primitive, irrespective of its causal role. But as Black (2000) argues, this means that two properties F and G could swap their powers in another world and yet still be F and G. Squareness, for example, could take on the causal role of circularity. Something that was square could roll smoothly on a flat surface – and yet still be square. This looks absurd, it should be countered. Anything that behaves like a circle, is circular. What other facts could there be in virtue of which something was a circle? The idea that properties are powers, or at least clusters thereof, seems a plausible view, therefore. But it has to be noted that Shoemaker (1998) abandoned this view. Why did he do so? There seem to have been two reasons. One seemed to be a concern that powers were endowed onto a property by the laws of nature. The property and its powers had therefore to be distinct even if one could still employ the powers in the identity conditions for properties. I will argue against such a view of laws, below. His second concern is more serious. A power is always a power for something else. Solubility, for example, is a power to dissolve. Gravitation is a power to attract. Powers are powers for some further property. One could say here, as Shoemaker did (1998, 412), that one must use the notion of a property to explain the notion of a power, so one cannot provide a reductive analysis of properties in terms of powers. Armstrong (2005) takes the problem in a different direction. If all properties are powers, then a power can only be a power for a further power. So nothing ever passes from potency to act, he says. Powers get passed

_____________ 6

Shoemaker said that these powers were conditional upon what other properties were coinstantiated. Hence, being knife-shaped means that something has the power to cut only conditional upon it being also hard. A knife-shaped cloud would not be able to cut.

Ascribing Dispositions

177

around but are never realised in something actual. All remains pure potentiality. This is a real problem for the account of properties but, ultimately, it is a consequence of the view that can bravely be upheld. Powers will indeed always be powers for further powers and, as I argue below, this will provide a neat account of causation, but it is a mistake to say that nothing is actual. Properties do involve potentiality but this is an ontology in which potentialities are treated as real. Indeed, being potent is a mark of reality – the so-called Eleatic mark of reality – to which realists about powers are likely to subscribe. Being potent is not, therefore, inconsistent with being actual, as Armstrong suggests. It is instead the mark of being actual. Armstrong’s assumption is that a disposition becomes actual only in its manifestation. And if this manifestation is ‘only’ a further power, then it still has not passed from potency to act. But the realist about powers is one who accepts their reality even before they are manifested. They will not accept, therefore, the characterisation of the situation as powers failing to attain fully-blown actuality. This objection can be answered, therefore, which is just as well because Shoemaker’s change of heart left him with a view that has the same problems as the Lewis and Armstrong view. The properties on his new position will be some unknown and primitive component underlying the powers. Shoemaker’s attraction to the original theory was that it was only through a property’s causal powers that we could know of it. By pulling properties and powers apart, he has lost that feature and the property takes on the role of some mysterious underlying substrata to the cluster of causal powers. Best, I suggest, to stick to the original theory and thus claim that a theory of properties is the first useful task for which we have put powers to work.

6. Causation Another way in which Armstrong presented his objection to the original Shoemaker theory of properties was to say that if every property were a power, or cluster of powers, then causation would become the mere passing around of powers. The world would be a world of shifting potencies (Armstrong 2005, 314). If it is agreed that the theory of properties can survive this criticism, then one is likely to think what an attractive account of causation this could turn out to be. Causation might indeed be best characterised as passing powers around. Fire has the property of being hot. In virtue of that, it has the causal power to heat another thing, such as my body. I sit by the fire and it causes me to become hot. The fire’s power to heat things has now been passed to me. Because my body is hot, I now have the power to heat other things, and when

178

Stephen Mumford

when they are heated, they have the power. That the world is a world of shifting potencies sounds plausible. A scientific account of the world might characterise it as a history of redistribution of energies, which sounds much the same as shifting potencies. It should be noted, however, that it need not be always the very same power that is passed by the cause to the effect. When heat causes me to become hot, I acquire the same causal power that the cause had, albeit in a lesser degree due to the fire’s heat also passing elsewhere. But when I manifest my power to break something, that which I have broken does not acquire the same power to break something merely in virtue of the fact that its new state was caused by that power. Rather, in causing something to happen, I am often causing a change – though not always, as I shall explain shortly. If one thing has a power to change another thing, that means that it has a power to change the second thing’s properties. But we have already seen in the previous section that properties can be understood as clusters of causal powers. To change a thing’s properties is, therefore, to give it a new set of causal powers. But this need not be the acquisition of the same powers as the cause unless it is a special case of the cause passing on one of its own properties. A hot thing can cause another thing to be hot but an explosive thing, while being able to affect other things in a dramatic way, does not automatically make them in turn explosive. A notable exception to this kind of model of causation is so-called immanent causation (Armstrong 1997, 73), where one stage or state of a thing causes a qualitatively identical latter stage or state of the same thing. Like sometimes causes like, which is cause without change. It may be controversial whether we need a notion of immanent causation but as it could be easily accommodated within the powers theory, there is no need yet to rule it out. Hume’s paradigm – the ‘perfect instance’ – of causation was the billiard balls crashing around the billiard table. Events were the relata of causal relations. But he could see only the events and never any necessary connection between them. Any kind of causation there would then have to be a contingent relation only. What is more, Hume thought that the idea of cause came from constant conjunction, from which we mistakenly infer the necessary connection. C. B. Martin was said to have asked the question whether, in a world that contains only a bang followed by a flash, the bang caused the flash. There is a constant conjunction alright – a regularity – but, we think, it should be an open question whether there was a real causal connection between these constantly conjoined events. A causal realist is able to say that there is some further fact of causal connection that the constant conjunction theorist is denying. Realists accept necessary connections between distinct existences, which orthodox Humeans deny. Such a necessary connection is what must be there for the bang to have caused the flash.

Ascribing Dispositions

179

The realist about dispositions has a theory of causation at their fingertips. Causation becomes the manifestation of power where a power manifests itself in either a new property of something or the sustaining of the same property. This simple and basic statement will have to be refined somewhat, however. As Molnar (2003, 194) noted, events are polygenous. Typically, they result from many powers working with each other or sometimes working against each other. Different voices in a choir can add a note to create a chord, which none of the individual choir members could have produced alone. Horses on either side of a canal are able together to pull the barge safely down the centre of the waterway, where one alone would have pulled it against the wall. Events occur, most typically, as the end result of many powers working together. But another advantage of a theory of causation based on powers is that it need not always be events, qua occurrences, that are the relata of causal relations. A fridge magnet may sit motionless on a fridge. That it does so is surely a causal matter, yet the example does not suit Hume’s paradigm well. Though there are some very broad conceptions of events, what is very interesting about this particular case is that, because of causation, nothing is happening. The causal relata are uneventful facts in this case. But power explains why this is a case of causation. That the magnet sits there, rather than falls to the ground, is a result of the magnetic power involved. The power manifests itself in the magnet not moving but remaining in a stationary state. A world that contains real dispositions has the resources to account for what Roy Bhaskar calls generative mechanisms (1975, 229). A Humean world is a world populated only by events but the realist judges that such a world is not a rich enough description of reality. To truly understand the world we have to posit a layer of powers underlying that world of events. Without this further layer, we do not understand all the polygenous contributions to events and we do not understand all the contributions to various equilibrium states, in which very little is happening. But if we accept a world of powers, we can understand all of this with relative ease.

7. Events On the subject of causation, it was noted that when things change, they change their causal powers. This suggests a theory of events: namely, an event is a change of causal powers. There are two main theories of events. Kim and Lewis opt for a property exemplification view (see, for example, Lewis 1986b). An event for them is a particular bearing a property at a time. My own view is that such a property exemplification is best understood as a state of affairs, rather than an event. But there would be, in any case, no problem for a theory of powers in accounting for property exemplifications. Properties

180

Stephen Mumford

as powers have already been discussed. The theory of events I favor, however, is one where an event essentially is a change. As Lombard (1986) describes the case, this means either a particular moving within a ‘quality space,’ for example, moving from green to red, or a particular instantiating a dynamic property, such as the property of rotting or growing. Both these kinds of event involve change and thus a realist theory of dispositions would be able to account for events as changes of the powers of a particular. This is not, of course, an analysis of events solely in terms of powers because the crucial notion of a change (of powers) has yet to be explained and such an account of change is unlikely to be in terms or powers alone.7 But events can quite simply be understood in terms of difference of property over time, which for the powers theory comes down to a difference of dispositions over time.

8. Laws of Nature For a Humean, laws, if they are anything at all, can amount to little more than regularities. There is a sophisticated regularity theory, presented by David Lewis (1973, 72-77), in which the only regularities that are laws are those that would be axioms or theorems in the best possible systematization of the world’s history. But this remains a regularity theory, in which there is no necessity in any of the regularities that would come out as the laws. For a long time, the deficiencies of such a view have been known. A clear conceptual distinction can be drawn between accidental regularities (all gold spheres are less than a mile in diameter) and genuinely lawlike ones (all uranium spheres are less than a mile in diameter). And if everything is loose and separate, then how would any explanation work in science and how would any prediction be rational? Because of such problems, Armstrong (1983) proposed a theory of laws in which they were understood not as simple universally quantified conditionals but as higher-order relations of necessitation between universals, such as the law N(F,G), which entails that x (Fx Gx) but is not entailed by it. In Armstrong’s metaphysics, however, every property is categorical and being F would not necessitate being G but for the law N(F,G). What is more, it is a matter of contingency that N(F,G) is a law. It could have been that this was not a law or that F and G participated is some other laws that were not laws of our world. Yet still it would be F and G that were involved in these different laws. It is clearly the case, then, that this un-

_____________ 7

I have a preference for the ancient criterion of change rather than a ‘Cambridge’ criterion. Thus, I say that a change occurs when some a particular x has some causal power P at time t1 that it does not have at time t2 (or it does not have P at t1 and then has it at t2).

Ascribing Dispositions

181

derstanding of the properties F and G is open to the charge of quidditism, which we have already encountered. A realist about dispositions would be able to understand laws in dispositional terms. It would be causal powers that were doing all the work in the dynamic operations of the world. Where would this leave laws? It could be argued that all laws-talk would be reducible to dispositions-talk (Mumford 1998a). In place of every law, we would find dispositions. Hence the law of gravitational attraction would be explained in terms of the disposition of massive things to attract. If we couple this with a mild form of essentialism, which has already been implicit, in which the causal role of a property is essential to it, then we get the regularity that laws were supposed to explain. Hence, the property would not be gravitational mass unless it attracted other objects in a certain way (as described in the gravitation law). Every massive thing must behave in this way; otherwise it wouldn’t be massive. Initial skepticism about how a dispositional theory would deliver the requisite regularity of laws (Everitt 1991) is thus dispelled. If a property is identified and constituted by its causal powers, then the same property will always have the same dispositional behaviour. The above account suggests that laws are reducible to the causal powers that constitute properties. But there are also reasons to think that laws should be eliminated rather than reduced (Mumford 2004). The foregoing arguments show that laws of nature do no metaphysical work at all. They are completely dispensable. Yet the concept of a law of nature was supposed to be the concept of something that governed or necessitated the events of the world. What sense would there then be in having laws reducible? How could they govern that to which they were reducible? And how could the world’s properties ever behave differently to the way that they do? It seems they can’t and so laws could never have done any work in the world in any circumstances. What this would show is that if one has a realist theory of powers, then one does not need any separate account of laws showing how one kind of state of affairs would necessitate another kind. Armstrong did need an account of laws, because he thought all properties were categorical. This account ran aground on the problem of quidditism. But now we see that an account of laws was never needed and that difficulty is entirely avoidable. A powers ontology shows that laws are entirely superfluous in our metaphysics.

9. Modality For a naïve Humean theory, all is contingent. Anything could happen and everything is in principle recombinable with everything else. The only sense that can be made of necessity is analyticity. But even modern-day Humeans,

182

Stephen Mumford

post-logical positivism, realize that this is a useless theory of modality. There have been two attempts at a more sophisticated account, though Humean in spirit, of what necessity and possibility are. One is Lewis’s (1986) modal realism, which is a claim that there are many concrete possible worlds, more or less like ours. For something to be possible, it is true at at least one world. For something to be necessary, it is true at all worlds. But this provides us only with inter-world necessity. There is still no intra-world necessity because all of these worlds are Hume-worlds, with no necessary connections between distinct existences. The theory of counterfactuals is similarly blighted from being built on a Humean metaphysical basis. Humeans must have non-intensional logics because they have only extensional truth at their disposal, with no additional modal properties. But the material conditional would be hopeless for counterfactuals as they all have false antecedents so would be true trivially. Lewis’s ingenious solution had us considering the closest possible worlds to ours in which the antecedent was true. Again these would be Hume-worlds, however, so the counterfactual would be deemed true merely if the consequent was true in all the closest possible worlds to ours in which the antecedent was true. In other words, the truth conditions of such counterfactuals remains material: if the consequent is true in those worlds, the counterfactual is true. If it is false, the counterfactual is false. The truth of such counterfactuals is therefore not grounded at all in the way things must be. And because it is not, we honestly have very little idea, while ever we treat these worlds as authentically Hume-worlds, whether the consequent will be true in the worlds where the antecedent is true. How could we know that if there is no necessity in such worlds? The second broadly Humean approach to the theory of possibility is Armstrong’s (1989) combinatorial theory. All wholly distinct particulars are, on this view, freely recombinable. This theory is better for a powers theorist, if nothing else because it is an attempt at a this-worldly theory of possibility. If we can account for modal truths with reference to this world alone then why should we use the possible worlds instead? But Armstrong’s theory has unconstrained recombinability of wholly distinct elements. Is this too permissive? A theory of powers is a theory of how certain properties make other properties possible and exclude still others. This would place restraints on Armstrong’s combinatorial freedom. Yet it might be objected that this provides us only with a theory of physical possibility and necessity? Surely there could be a sense in which something is logically possible while being physically impossible. I agree that there is such a sense, but that it is incredibly weak. Logical possibility and necessity is founded merely on logical form, and analytic possibility and necessity is founded only on the meaning of words. Neither is any guide as to what is really possible and necessary in the world.

Ascribing Dispositions

183

De re – some say metaphysical – possibility and necessity is founded on the way the worlds is (following Ellis 2001). Thus if we want to know what is really possible in our world, we need to understand the way that the world works. We need to understand the causal roles of properties, in other words, to understand the powers of things. And, as we have seen already, there is no worldly contingency to these. Something that is circular must have a certain causal role, or it would not be circular. Thus there are no possible worlds in which a circular thing can have a different causal role to the one it has in our world. And there is no permissible recombination that separates circularity from its causal role. I hold what I have called a restricted combinatorialism (Mumford 2004, ch.10), which is based on Armstrong’s theory but with all the limitations on possibility that powers give the world. But this is good news. If we no longer have a Hume-world, in which any combination goes, but a powerful world instead, then we get a this-worldly and appropriately restricted theory of possibility and necessity. Above all, we can retain a commitment to naturalism because we will not require reference to concrete possible worlds in order to explain modality, even if such a commitment could, contrary to what I believe, explain modality.

10. The Balance Sheet I have indicated how a theory of powers might generate theories of properties, causation, events, laws and modality. I do not rule out that other metaphysical categories might be explicated in similar terms. These accounts show, I maintain, how a powers theory does better than broadly Humean theories. Indeed, Humeanism can make such categories problematic in the first place, as is almost certainly the case with causation. And in other cases, Humeanism is retained only at the expense of some pretty extravagant and implausible claims: modal realism, quidditism, regularity theory, and the like. I submit, therefore, that when the final balance sheet is drawn up, the powers theory comes out ahead. We may yet have concerns about methodological issues in metaphysics. We may not be yet sure whether we can say with certainty that powers exist or that it is merely the most coherent account of the world that there are powers. Either way, I claim that powers come out on top by all the standard pragmatic criteria, in terms of costs and benefits of the ontological assumptions. Humeans begin with a relatively sparse ontology, usually understood in terms of the world being nothing more than a patchwork of unconnected events or facts. Everything else must supervene on that. But sometimes the cheapest deal is not the best and can leave us impoverished further down the

184

Stephen Mumford

line. In accepting real causal powers, we are accepting something with a bit more metaphysical clout and yet, I contend, the higher price is worth paying because it leaves us with good accounts of all the things in this world for which we think we need accounts.

11. Ascribing Dispositions Such metaphysical considerations as I have raised in the second half of the paper are what justify the ascriptions of dispositions. Even where their presence cannot be completely empirically warranted, because they are untested or untestable, we have good reasons still to believe that powers are there because we have good metaphysical reasons to believe that the world is a world of powers. I have described the argument for powers as a transcendental argument. It is not intended to be transcendental in the full-blooded Kantian sense, however, as I certainly do not want to claim merely that powers are a feature of our own understanding and not of the world. Rather, I want to say that the assumption of powers is justified because the world must be this way for us to make sense of it. Metaphysical knowledge is fallible, however. We might be extremely unlucky and live in a world in which everything would seem to make sense if it were a world of powers, and yet it still not be a world of powers. The world would have been very cruel to us if that were the case. We have some empirical sources of the idea of power or disposition, though they are not conclusive. We also have independent metaphysical grounds for accepting an ontology of powers, though they are not infallible. When we have a convergence of ideas from experience and theory, suggesting the same kind of account of the world, then that is about as good as it gets.

Literature Armstrong, D. 1983. What is a Law of Nature? Cambridge: Cambridge University Press. Armstrong, D. 1989. A Combinatorial Theory of Possibility. Cambridge: Cambridge University Press Armstrong, D. 1997. A World of States of Affairs. Cambridge: Cambridge University Press. Armstrong, D. 2005. “Four Disputes About Properties.” Synthese 144: 309-20. Bird, A. 1998. “Dispositions and Antidotes.” Philosophical Quarterly 48: 227-34. Bhaskar, R. 1975. A Realist Theory of Science. Leeds: Leeds Books Limited. Black, R. 2000. “Against Quidditism.” Australasian Journal of Philosophy 78: 87-104. Carnap, R. 1936/7. “Testability and Meaning.” Philosophy of Science 3: 420-71 and 4: 1-40. Ellis, B. 2001. Scientific Essentialism. Cambridge: Cambridge University Press. Everitt, N. 1991. “Strawson on Laws and Regularities.” Analysis 50: 206-8. Hume, D. 1739-40. A Treatise of Human Nature, L. A. Selby-Bigge (ed.), Oxford: Clarendon Press 1888.

Ascribing Dispositions

185

Hume, D. 1748. An Enquiry Concerning Human Understanding In Enquiries Concerning Human Understanding and Concerning the Principles of Morals, L. A. Selby-Bigge (ed.), 3rd edn rev. P. H. Nidditch. Oxford: Clarendon Press 1975. Lewis, D. 1973. Counterfactuals. Oxford: Blackwell. Lewis, D. 1986. On the Plurality of Worlds. Oxford: Blackwell. Lewis, D. 1986a. Philosophical Papers II. Oxford: Oxford University Press. Lewis, D. 1986b. “Events.” In Philosophical Papers II: 241-69. Lewis, D. Forthcoming. “Ramseyan Humility.” In D. Braddon Mitchell and R. Nola, The Canberra Programme. Oxford: Oxford University Press. Lombard, L. 1986. Events. London: Routledge Kegan Paul. Martin, C. B. 1994. “Dispositions and Conditionals.” Philosophical Quarterly 44: 1-8. Mellor, D. H. 1974. “In Defense of Dispositions.” Philosophical Review 83: 157-81. Molnar, G. 2003. Powers: a study in metaphysics. S. Mumford (ed.). Oxford: Oxford University Press. Mumford, S. 1998. Dispositions. Oxford: Oxford University Press. Mumford, S. 1998a. “Laws of Nature Outlawed.” Dialectica 52: 83-101. Mumford, S. 2004. Laws in Nature. Abingdon: Routledge. Ryle, G. 1949. The Concept of Mind. London: Hutchinson. Shoemaker, S. 1980. “Causality and Properties.” In Identity, Cause and Mind, expanded edition. Oxford: Oxford University Press 2003: 206-33. Shoemaker, S , 1998. “Causal and Metaphysical Necessity.” In Identity, Cause and Mind, expanded edition. Oxford: Oxford University Press 2003, 1984: 407-26. Strawson, G. 1989. The Secret Connexion: Causation, Realism, and David Hume. Oxford: Clarendon Press. Wright, A. 1991. “Dispositions, Anti-Realism and Empiricism.” Proceedings of the Aristotelian Society 91: 39-59.

Dispositional Pluralism JENNIFER MCKITRICK In this paper, I make the case for the view that there are many different kinds of dispositions, a view I call dispositional pluralism. The reason I think that this case needs to be made is to temper the tendency to make sweeping generalization about the nature of dispositions that go beyond conceptual truths. Examples of such generalizations include claims that all dispositions are intrinsic, essential, fundamental, or natural.1 In order to counter this tendency, I will start by noting the extent to which it is at odds with the semantics of dispositions, according to which there are many kinds of disposition ascriptions. From there, I will try to support a metaphysical claim that there are different kinds of dispositions. To bridge the gap between semantics and metaphysics, I appeal to epistemology. I’ll consider the question “when do we have good reason to believe that a disposition ascription is true?” If our disposition ascriptions are true, then we are right about what kinds of dispositions things have, and what kinds of dispositions there are. I claim that our evidence for different kinds of dispositions is on a par; we have reason to believe that various kinds of dispositions are instantiated. There might be good reasons to want to distinguish between different kinds of dispositions. One might want to focus on the intrinsic dispositions, or the fundamental dispositions. One might want to go further and say that other properties aren’t really dispositions at all. However, the question “which are the real dispositions?” becomes a terminological issue. One can decide not to call certain properties “dispositions,” even when they otherwise seem like dispositions, because they fail to satisfy certain conditions. This might be a harmless terminological decision. However, I think it is preferable to keep the concept “disposition” closer to its ordinary English usage as a more general concept. Defending dispositional pluralism involves arguing against extremist or absolutist positions about dispositions – views according to which all dispositions are necessarily this or that. Debates about dispositions are often presented as dichotomies between two extremes.2 These dichotomies include intrinsic versus extrinsic, grounded versus ungrounded, reducible versus irre-

_____________ 1 2

For example Ellis 2002; Molnar 2002; Heil 2005. For example, see Ellis 2002, 59-60.

Dispositional Pluralism

187

ducible, fundamental versus derivative, essential versus non-essential, firstorder versus higher-order, natural versus unnatural, causally inert versus causally efficacious, and non-existent versus universal. However, there’s room in logical space for mixed or moderate positions along each of these dimensions.

1. Dispositions Talk A number of English words are roughly synonymous with “disposition”: tendency, power, force, predisposition, liability, susceptibility, propensity, potentiality, proclivity, inclination, ability, capability, faculty, and aptitude. There are different connotations and shades of meaning, between, for example, a power to act and a liability to be acted upon. Some terms, such as “proclivity” suggest rational agency, while others such as “force” suggest fundamental properties of matter. Many terms in English are dispositional. They cover a broad range of qualities, from fragility to courage. They include: terms from various branches of science, such as charge, energy, reactivity, conductivity, malleability, solubility, elasticity, fitness, and fertility; character traits, such as integrity, or being punctual, neat, kind, considerate, and shy; common qualities of physical objects, such as being elastic, comfortable, flammable, intoxicating, or hazardous; and complex social concepts, such as marketable, redeemable, taxdeductible, collectible, humorous, provocative, titillating, recognizable, and enviable. Obviously, dispositional terms are quite diverse. At this point, the reader may wonder, on what grounds do I classify these terms as dispositional? I suggest that a term is dispositional if it has the following marks of dispositionality: 1. The term is associated with an event type – the manifestation of the disposition; 2. The term is associated with an event type in which the manifestation occurs – the circumstances of manifestation; 3. The term is ascribable to an object when the manifestation is absent; 4. If a dispositional term is ascribable to an object, then a certain subjunctive conditional to the effect that “if the circumstances of manifestation were to occur, then the manifestation would occur” is true; 5. The term is semantically equivalent to an overtly dispositional locution – “the disposition to so and so.”3 I offer these marks of dispositionality as rules of thumb, not as an analysis. They do strike me as jointly sufficient, since I cannot imagine a term bearing all five marks and yet failing to be a disposition term. However, I hesitate to

_____________ 3

This is an adaptation of the marks of dispositionality for properties (McKitrick 2003, 156-158).

188

Jennifer McKitrick

assert that each is necessary. Counter-examples to conditional analyses suggest that the fourth mark is not always true of disposition terms. The carefully packed glass is still fragile, but the counterfactual “if it were struck, it would break” is not true of it.4 Also, if there are some disposition terms which are not applicable in the absence of the associated manifestations, then the necessity of the third mark is called into question as well. For example, we call a structure “stable” as long as it stays intact. When the structure falls down, it no longer manifests stability, nor is it stable. If “stability” is a disposition term, it is one that is not attributable in the absence of its manifestation. But putting aside odd exceptions, a wide variety of terms bear these marks of dispositionality.

2. Philosophical Distinctions Philosophers have made a number of philosophical distinctions between different kinds of dispositions. Aristotle makes a distinction between rational and non-rational capacities. Non-rational dispositions of objects, including human bodies, manifest in circumstances of manifestation due to physical necessity. Examples of non-rational dispositions include the disposition of a rock to fall, and the disposition of human skin to tan in sunlight. Rational dispositions, on the other hand, are dispositions of rational agents to perform certain actions if they so choose, such as playing a musical instrument, or speaking a language.5 Ryle distinguishes single-track or specific dispositions on the one hand from multi-track or generic (general) dispositions on the other.6 Some dispositions are triggered in just one kind of circumstance and manifest themselves in only one way. For example, the only manifestation of solubility is dissolving, and being immersed in liquid is its only circumstances of manifestation. Perhaps there are some differences between particular immersions or dissolvings, but they must be similar enough to all count as the same kinds of events. However, other dispositions, such as bravery, have different kinds of manifestations which occur in different kinds of circumstances. Various circumstances, such as fires, battles, amusement park rides, medical procedures, and intimidating social situations can trigger acts of bravery. The manifestations are as diverse as rushing into a burning building, or taking an unpopular po-

_____________ 4 5 6

For a discussion of counterexamples to conditional analyses of dispositions, see Smith 1977, Johnston 1992, Martin 1994; Bird 1998. See Ludger Jansen in the volume for more on Aristotle’s many distinctions among dispositional properties and the like. Ryle 1949, 118.

Dispositional Pluralism

189

litical stand. Perhaps the circumstances and manifestations can be given a unifying abstract characterization such as fear-inducing situations which involve danger or substantial personal risk and confronting that fear or danger. However, the types of events that count as manifestations and circumstances of manifestation of multi-track dispositions are diverse and only count as the same kind of event at some very high level of abstraction. C. D. Broad distinguishes hierarchies of dispositions, of first, second, or higher order.7 A higher-order disposition is a disposition to acquire or lose a disposition. Magnetizability is a higher-order disposition, on Broad’s terminology. Some pieces of metal are magnetic – are disposed to attract or repel other metals and magnets. However, other pieces of metal aren’t magnetic, but can become magnetic if they are subject to a strong enough magnetic field. These non-magnetic metal pieces are thus magnetizable. They have a second-order disposition – the disposition to acquire the disposition of being magnetic. Rom Harre, following Aristotle and Locke, distinguishes powers and liabilities, also known as active powers and passive powers.8 If something has power to A, it will do A in certain circumstances. On the other hand, if something has a liability, it has a disposition to suffer change. In the case of liabilities, the thing that changes is the disposed object, and situation that produces change is external to the object. Fragility is supposed to be a liability. The fragile thing is subject to an external force and changes internally. External circumstances, such as the strike of a hammer, trigger the manifestation. Poisonousness, on the other hand, is supposed to be a power. When someone is poisoned, the cause of poisoning is in the disposed object – the poison. The locus of manifestation is not in the disposed object, but external to it. The poison has an effect on something else, and the manifestation is that something else changing. One may argue that this distinction does not hold up to scrutiny as metaphysical, but is merely pragmatic. The manifestations of dispositions have multiple causal factors. Whether the salient causes are internal or external to the disposed object is determined by the relevant interests and background assumptions. In the example above, the solidity of the striking hammer is considered to be causally relevant to the shattering, but the fragility of the glass isn’t. But it’s not clear on what grounds we make that distinction. Consider explosiveness, which intuitively seems like a power. But the manifestation of explosiveness is the explosion, which typically destroys the explosive object. The cause of the explosion has to do with the nature of the bomb, but also with being ignited – a cause external to the bomb itself. When explosive-

_____________ 7 8

Broad 1925, 432. Harre 1970, 84; Aristotle 1941; Locke 1990, 105-107.

190

Jennifer McKitrick

ness is manifest, the bomb ends up in little scattered pieces as a result of an external cause. Structurally, that sounds a lot like fragility, which was supposed to be a liability. One may argue that, in the case of explosiveness, things external to the bomb are affected as well. However, this doesn’t clearly distinguish it from fragility, since things external to the struck glass may be affected when its shards go flying. Despite the suspicion that the active/passive distinction is merely pragmatic, perhaps other distinctions are more robust. Philosophers often distinguish sure-fire or deterministic dispositions on the one hand, and tendencies or probabilistic dispositions on the other. In the case of a sure-fire disposition, when the disposed object is in the circumstances of manifestation, the occurrence of the manifestation is physically necessary or exceptionless. In the case of tendencies, when the disposed object is in the circumstances of manifestation, the manifestation might occur, but it might not. Examples of probabilistic dispositions include the disposition of enriched uranium to decay and probably most behavioral dispositions. A sociable person typically engages in conversation but may neglect to on occasion. One might want to say that a thing has a probabilistic disposition if it manifests that disposition in the circumstances most of the time. However, something may be a tendency even if it doesn’t usually happen in the circumstances of manifestation. The recovering alcoholic has a tendency to drink, but resists it.9 So, a disposition is probabilistic if there are occasions, or even if it is merely possible that the disposed object is in the circumstances of manifestation and yet the manifestation does not occur. However, if that were all that it took for a disposition to be a tendency, then it is not clear whether there would be there any sure-fire dispositions. Many counterexamples to conditional analyses consist of imagining a possible scenario in which a disposed object is the circumstances of manifestation yet fails to manifest the disposition. They seem to show that virtually all dispositions are such that, if objects which instantiate them are placed in the circumstances of manifestation, it’s possible that the manifestation does not occur. This is because the disposition might be masked, an antidote to the disposition might be delivered, or the disposition might be finkish.10 However, there still seems to be a difference between probabilistic and sure-fire dispositions even if we can’t perfectly articulate it. Perhaps a sure-fire disposition is such that, if the disposed object were in the circumstances of manifestation under ideal conditions, the manifestation would necessarily occur, while the probabilistic disposition is such that, even if the disposed object were in the circumstances of manifestation under ideal conditions, the

_____________ 9 10

See Jansen on tendencies (2006). See Johnston 1992; Martin 1994; Bird 1998.

Dispositional Pluralism

191

manifestation only has a certain probability of occurring. Even if there is no funny business such as masks or finks, the manifestation of a probabilistic disposition might not occur in the circumstances of manifestation. Perhaps we cannot specify a way to determine if we are dealing with a probabilistic disposition or a thwarted deterministic disposition, but that’s the conceptual distinction at least.

3. Semantic Support for Pluralism Because philosophers have distinguished so many different dispositional concepts, and the terms that bear the marks dispositionality are so diverse, dispositional semantics does not support absolutist claims. Dispositional terms attribute a wide variety of kinds of properties to objects: intrinsic and extrinsic properties, reducible and irreducible properties, essential and non-essential properties, natural and unnatural properties, and so on. Absolutist claims do not fall out of an analysis of the concept of a disposition, nor do they follow from particular dispositional concepts, such as fragility. Probably the most sweeping absolutist claims about dispositions are that they are non-existent, or on the other extreme, that they are universal. In other words, either all properties are dispositions, or all properties are nondispositional. However, it seems clear that some terms bear marks of dispositionality and others do not. Even absolutists who say there are no dispositions and others who say that all properties are dispositions acknowledge a distinction between dispositional and non-dispositional predicates.11 A commonly made generalization about dispositions is that they are all intrinsic properties. One could also claim that they are all extrinsic properties, a view that Brian Ellis attributes to Humeans.12 Dispositions might appear to be extrinsic if what dispositions a thing has depends upon the prevailing laws of nature. Another reason one might have for thinking dispositions are extrinsic is that they are relations an object has to possible events – the manifestations. However, if the third mark of dispositionality obtains even sometimes, then it is possible for a thing to have a disposition and not exhibit its manifestation. If the particular manifestation event does not occur, and dyadic relations require the existence of both relata, then the disposition cannot be a relation. The best way I can see of saving the idea that dispositions are relations and thus extrinsic is to make the hard case that there is a sense in

_____________ 11

12

For example see Armstrong 1973, 15; Shoemaker 1980, 211. However, others suggest that there’s no clear way to demarcate a dispositional/non-dispositional distinction even at the conceptual level (Mellor 1974, 171; Goodman 1983, 41). Ellis 2002, 60.

192

Jennifer McKitrick

which mere possibilia, such as merely possible events, exist.13 So, I will leave aside this reason for thinking that all dispositions are extrinsic. The idea that all dispositions are intrinsic, or that they are all extrinsic, has no semantic support either. Dispositional concepts would be concepts of intrinsic properties if they were necessarily equally applicable to perfect duplicates. However, while some dispositional predicates are equally applicable to perfect duplicates, others are not. For example, ‘fragility’ applies equally to perfect duplicates: One would be hard-pressed to justify calling one glass fragile but not its perfect duplicate. ‘Vulnerability,’ on the other hand, is a dispositional term that can apply differentially to perfect duplicates. For example, a newborn infant left alone in the woods is more vulnerable than his perfect twin at home in his crib. ‘Visibility’ also differs in its applicability to perfect dublicates. While duplicate glasses are equally fragile, the glass that’s hidden in a dark room is not visible, but its well-light, out-in-the-open duplicate is. The applicability of the dispositional predicate is not merely a matter of the object being in the circumstances of manifestation of the disposition, but of it being in the circumstances of possession of the disposition.14 Another dimension of absolutism concerns the natural/unnatural distinction. One might argue that all dispositions are natural properties, or, perhaps that all dispositions are unnatural properties. The primary reason for thinking that all dispositions are natural properties is the view that the only real properties, and hence the only real dispositions, are the natural ones. This view naturally goes along with the idea that not every predicate corresponds to a property.15 While I agree that some predicates, such as ‘non-self-instantiating’ do not correspond to any property, suffice it to say that dispositional pluralism fits best with a fairly liberal ontology of properties.16 But like the disputes about “real” dispositions, the dispute about the sparseness or abundance of properties might have a terminological interpretation as well. While the sparse property theorist distinguishes between the predicates that refer to properties and those that don’t, the liberal property theorist allows that most predicates refer to properties, and distinguishes between natural and unnatural properties. The liberal property theorist need not have a bloated ontology if she can allow that unnatural properties are reducible to natural properties, or that property claims are true in virtue of actual features of particulars. In fact, the fundamental ontologies of the sparse and liberal theorists could be the same, and they differ only in what they choose to call a property.

_____________ 13 14 15 16

Rosenberg 2004, 211. See McKitrick 2003 for an extended defense of extrinsic dispositions. Armstrong 1996, 18. An example of a liberal, or abundant view of properties is that of David Lewis, according to which there is a property for every possible set of possibilia (1983).

Dispositional Pluralism

193

To arrive at the opposite conclusion that dispositions are unnatural, one might reason as follows. A disposition is a secondary property (a property which consists in having some property or other) and is thus multiply realizable. Multiply realizable properties can be shared by objects with different realizer properties, so they are equivalent to disjunctive properties, and disjunctive properties are unnatural.17 But once again, our linguistic practices do not support either extreme. Some dispositional predicates are applicable only to things that are similar to one another in some important way, while other dispositional predicates seem applicable to a diverse, gerrymandered group. For example, things which are ‘electrically conductive’ probably have certain compositional and structural similarities. However, things which are ‘provocative’ form a diverse group with no relevant respect of intrinsic similarity. So, it seems that some dispositional predicates seem to pick out natural properties, while others pick out unnatural properties. Dispositional essentialists claim that all dispositions are essential properties of the objects which instantiate them.18 An object with a certain disposition cannot lose that disposition and still be the same object. A plausible example of a disposition which is essential to its possessor is the electrical charge of an electron. Arguably, a particle without the disposition to repel negatively charged particles could not be an electron. On Ellis’ view, all genuine dispositions are similarly definitive of the objects which instantiate them. On the opposite extreme is the view that dispositions are non-essential properties. On such a view, an object could lose any of its dispositions if, for example, it were subject to different laws of nature. However, attributions of dispositional predicates follow not such strictures. On the one hand, some disposition ascriptions seem contingent. The courageous person might have been otherwise, given a different upbringing. The fragile doll house could have been put together with a stronger adhesive. On the other hand, some disposition ascriptions seem necessarily true. In other words, there are some disposition predicates such as ‘having negative charge’ that apply to certain objects in every circumstance in which that object exists. Furthermore, disposition ascriptions do not distinguish between grounded and ungrounded properties, reducible and irreducible properties, fundamental and derivative properties, or first-order and higher-order properties. If I attribute a disposition to an object, learning that the disposition ascription was true in virtue of the fact that the object had some distinct property would not give me a reason to withdraw my disposition attribution. For

_____________ 17 18

Lewis suggests this sort of picture (1986, 224). Ellis 2002, 59.

194

Jennifer McKitrick

example, suppose I claim that Joe is irritable. Then I’m told that Joe is irritable in virtue of some of his neurological features – his irritability derives from or is based on these neurological features. I withdraw neither my claim that Joe is irritable nor my belief that his irritability is a disposition. Even if my claim that Joe is irritable were reducible to a claim about some distinct property, reduction is not elimination, and I have no reason to withdraw my disposition ascription. On the other hand, if I attribute a disposition to an object, learning that that object had no distinct property in virtue of which that claim was true would give me no reason to withdraw my claim. For example, suppose I claim that a massive object is disposed to attract other massive objects. If I learned that the object has no distinct property in virtue of which this is true, that this was a fundamental, irreducible feature of the object, I would not withdraw my disposition claim. So, there seem to be disposition terms that attribute derivative, grounded, and perhaps reducible properties to things, and others that could attribute fundamental, ungrounded, irreducible properties to things. The semantics are consistent with there being all of these kinds of dispositions. Furthermore, whether any or all dispositions are causally inert or causally efficacious cannot be determined by examining language. The causal power, or lack-there-of, is not always part of the dispositional concept. (Notable exceptions are dispositional kin concepts such as ‘power’ and ‘capability.’) Finding out that a property is causally inert or efficacious does not necessarily lead one to withdraw a disposition claim. When the doctor attributes a ‘dormitive virtue’ to the sleeping pill, many seem convinced that he has not revealed any causally efficacious property of the pill. If this is right, then some dispositional ascriptions are not attributions of causally efficacious properties. Could one consistently maintain that other disposition ascriptions do attribute causally efficacious properties? That depends on one’s reasons for thinking that dormitivity is inert. Those reasons may or may not apply to all disposition terms. For example, if you think that dormitivity is a second-order property – a property of having some property or other that causes sleep upon ingestion, then you might think all the causal work is done by the lower-order property, and so the dormitivity is inert.19 However, that is consistent with there being some first-order dispositions that do not lose out in a causal exclusion argument, and so they can be considered causally efficacious. To summarize what I have said so far, philosophers have distinguished several different disposition concepts. Natural language, English anyway, presents a wide variety of disposition terms and concepts. We can and do

_____________ 19

Such causal exclusion arguments are put forth by Kim 1993, 353; Prior, Pargetter, and Jackson 1982, 255.

Dispositional Pluralism

195

attribute a variety of dispositional predicates: intrinsic and extrinsic, natural and unnatural, essential and non-essential, higher-order and fundamental. Our disposition ascriptions are neutral with respect to whether the dispositions are reducible or irreducible, bare or grounded, essential or non-essential, inert or efficacious.

4. Beyond Conceptual Analysis? I’ve argued that we employ numerous and diverse dispositional concepts. What does that tell us about the world? I hope it is not too naïve to think that a long entrenched tradition of employing certain concepts with apparent success gives some reason for thinking that those concepts are related to the world in a meaningful way. I am supposing that, if our disposition ascriptions are true, then the dispositions we ascribe to things exist, in whatever sense properties exist (as universals, tropes, natural kinds, genuine similarities, etc.). I am well aware that this assumption stands in opposition to a major project of the last century, to semantically reduce disposition ascriptions. If that project were successful, one could say that disposition ascriptions are true, but not because dispositions exist, but because the ascriptions are merely ways of asserting something that is consistent with the non-existence of dispositions, such as a conditional, or a claim about non-dispositions. Just as the claim that “The average American woman has 1.5 children” doesn’t commit one to the existence of the average American woman nor half-children, it is thought that claims such as “x has a disposition to so and so” doesn’t commit one to the existence of dispositions. However, it is not easy to say what disposition ascriptions mean if there are no dispositions.20 The denier of dispositions is in the uncomfortable position of either claiming that all of our numerous and varied disposition ascriptions are false, or explaining how they could be true if there are no dispositions. If we accept that the truth of disposition claims gives us evidence for the existence of dispositions, then we know that dispositions exist to the extent that we know that disposition ascriptions are true. So, what is our evidence for truth of disposition claims? Dispositions are not directly observable; however, their manifestations often are. Simply put, when we observe that an object regularly exhibits a certain manifestation in certain circumstances, then we have reason to believe that the object has a disposition to exhibit that manifestation in those circumstances. We are also sometimes justified in be-

_____________ 20

See Markus Schrenk’s paper in the volume for a recap of the unhappy history of that project. However, even some non-reductionist, such as George Molnar, reject the view that ascribable dispositional predicates correspond to genuine dispositions (Molnar 2003, 27).

196

Jennifer McKitrick

lieving that an object has a disposition even if we have never observed that particular object manifesting that disposition. In that case, we have reason to believe that it is relevantly similar to objects which have regularly exhibited that manifestation in those circumstances. I think that something like this is all the reason we ever have for believing that something has a particular disposition. Of course, we are fallible, and we might over-generalize or misidentify the relevant respect of similarity. The process is more detailed and controlled in a scientific experiment, and of course, the case is much more complicated when the manifestation is itself unobservable, such as the manifestation of an electron to repel other electrons. (The question of how to determine when unobservable manifestations occur is answered by however we determine that any unobservable state of affairs obtains.) It seems to me that the evidence we have for different disposition claims does not discriminate between different kinds of dispositions. The evidence for the assertibility of our ascriptions of different kinds of dispositions seems to be on a par. For example, though flammability may be a more natural property than provocativeness, our evidence that a red cape is provocative is not unlike our evidence that it is flammable: When it, or capes like it, are waved in front of a bull, the bull charges; when they are ignited, they burn. Favorable evidence is not exclusive to natural properties. In a similar vein, I could argue that we have evidence for dispositions on both sides of each distinction. However, I’m going to concentrate on what strikes me as the hardest case – ungrounded or bare dispositions. Is the evidence for bare dispositions on a par with our evidence for grounded dispositions? One might think that, in the case of bare dispositions, even if an object has exhibited a certain manifestation in the past, since the disposition is not grounded in any other property of the object, the object might have mysteriously lost the disposition in the mean time. However, if we are going to be that skeptical about the stability of dispositions, this skepticism would not single out bare dispositions as possibly fleeting. Suppose I pick up a rubber band, stretch it, then put it down. It resumes its former shape, I figure that it has the disposition of elasticity, and I presume there’s something about its structure and composition that accounts for this. But a minute later, if I want to adopt a skeptical attitude, for all I know, its underlying structure might have changed, and it’s no longer elastic. In fact, it has happened that I opened my drawer and picked up a rubber band that appears to be the same as it did the last time I used it, but when I go to use it again, I find that it has lost its elasticity and has become brittle. Of course, this loss of elasticity takes much longer than a minute, and my practice of assuming that, other things being equal, things retain their dispositions, serves me pretty well. I don’t see where the ground of the disposition, or lack-

Dispositional Pluralism

197

there-of, makes a difference in my confidence that things retain their dispositions when they are not manifesting them. What about when a particular instance of a bare disposition has never been manifest? How do we know it is there at all? As mentioned above, in general, when an object has never manifested a particular disposition, our evidence that it has that disposition is its similarity to other objects which we have observed to manifest that disposition. But in the case of bare dispositions, one may argue, it’s not clear what the relevant respect of similarity is. Since the disposition is bare, there is no observable property that objects share, which grounds the disposition in question. This suggests that our evidence for unmanifested bare dispositions would be inferior to our evidence for unmanifested grounded dispositions. While this argument seems plausible, it assumes that our evidence for unmanifested grounded dispositions is the observation of a causal basis shared with manifested dispositions. However, this is unlikely. Consider your reasons for believing that something has an unmanifested grounded disposition, such as a tablet that is water-soluble. Your evidence for the tablet’s water-solubility is the observable properties it shares with things that have dissolved in the past. But the observable properties are unlikely to be the causal basis of the tablet’s water-solubility. You do not observe a particular molecular structure, or anything that is a plausible candidate for being a causal basis of solubility. Similar points can be made about unmanifested elasticity, fragility, inflammability, etc. So, either we are not justified in believing such ascriptions of unmanifested grounded dispositions, or the claim that we are only justified in believing an ascription of an unmanifested disposition when we observe its causal basis is false. I think we are justified in making ascriptions of unmanifested dispositions in the absence of any observation or knowledge of a causal basis. So, our evidence for dispositions which may happen to be bare is on a par with grounded dispositions. That is not to give evidence for the bareness of those dispositions. For that, we have the (defeasible) evidence that the property in question is fundamental.21

5. Property Dualism Dispositional pluralism, the view that there are many different kinds of dispositions, is obviously inconsistent with the denying that dispositions exist. An absolutist who claims that all properties are non-dispositional is unlikely to be convinced by my claims that we have evidence for the truth of disposition

_____________ 21

See arguments for the ungrounded disposition in Mumford 2006; Molnar 2003, 131-132.

198

Jennifer McKitrick

ascriptions and thus evidence for the existence of dispositions. Though we seem to have both dispositional and non-dispositional predicates, and true sentences ascribing dispositional predicates to objects, one might argue that the fact that a predicate bears the marks of dispositionality doesn’t show that the property it picks out is a disposition. It is thought that the dispositional concept, tied as it is to a causal role, is just an oblique way to referring to what is in fact a non-dispositional property. On such a view, when we are better acquainted with the occupant of this causal role, we can jettison the dispositonal talk if we choose.22 So, while “magnetic” for example may bear the marks of dispositionality, if it picks out a property, on this view, it nevertheless picks out a non-dispositional property. Hence one can recognize both dispositional and non-dispositional predicates and yet deny the existence of dispositions. So, a full defense of dispositional pluralism should include an argument against anti-dispositionalism, if you want to call it that.23 What about pandispositionalism – the view that all properties are dispositions? Pandispositionalism is consistent with the view that there are different kinds of dispositions, so I need not rule out pandispositionalism in order to defend dispositional pluralism. However, property dualism (in this context, the view that there are both dispositional and non-dispositional properties) is more in the spirit of pluralism, and happens to be the view that I favor. But like the anti-dispositionalist, the pandispositionalist is unlikely to be convinced by my semantic/epistemic arguments. The pandispositionalist can acknowledge ascribable non-dispositional predicates, and yet maintain that all properties are dispositions. Such a theorist might point out that bearing the marks of dispositionality is not necessary for a term to pick out a disposition. For example, one may point out that the term ‘red,’ as ordinarily understood, does not bear the marks of dispositionality: it has not strong conceptual association with triggering events, manifestations, or conditionals, nor is it normally thought of as equivalent in meaning to an overtly dispositional locution. Nevertheless, one might maintain that to be red is nevertheless to have a disposition to cause certain types of visual experiences. So, a defense of property dualism cannot rest on the evidence for the truth of both disposition ascriptions and non-disposition ascriptions. Even when the truth of those ascriptions is not at issue, whether the predicates they employ refer to disposition is. Finding plausible examples of nondispositional properties is good reason to reject pandispositionalism. While the pandispositionalist might talk you

_____________ 22 23

Proponents of this “promissory note” conception of disposition terms include Quine (1969, 20) and Armstrong (1973, 15). I will not give that argument here, for lack of space and of anything new to say. For a defense of dispositions over Humean views, see for example, Molnar 2003, 111-121; Ellis 2002, 60-65; Mumford 1998, 170-191.

Dispositional Pluralism

199

into thinking that shape and mass are dispositional,24 other properties seem immune to such strategies. Molnar makes a plausible argument for the view that spatio-temporal relations are irreducible, non-dispositional properties.25 Also, arguably, we have evidence for non-dispositional qualities of our own experiences. Even if redness is a disposition to produce certain visual experiences, some properties of those visual experiences seem non-dispositional. Whether or not one finds such examples plausible, pandispositionalism is called into question by the so-called “always packing, never traveling” (APNT) objection.26 If all properties are dispositions, when objects manifest their dispositions, they merely acquire new dispositions. Each object packs and repacks its trunk full of properties, but it never takes off. One might not think that this regress is vicious. Instead, one might think a disposition to produce a disposition to produce a disposition is no worse than a cause which produces an effect which is also a cause for a further effect, onward into the future. If there’s anything to the objection, there must be an important difference between APNT and a simple causal chain. To make the objection clearer, we should take it beyond the metaphorical level. When an object manifests a disposition, some object acquires new properties, either the disposed object, some other object(s), or both: the elastic band takes on a new shape; the provocative cape changes the bull’s mood; the soluble table dissolves and the surrounding liquid approaches saturation, etc. Furthermore, it seems reasonable to grant that sometimes the manifestation of a disposition involves the acquisition of new dispositions. Dispositions to acquire dispositions are the second-order dispositions discussed by Broad, with magnetizability being a plausible example. But could it be the case that all dispositions are like that – merely dispositions for further dispositions? Could every manifestation of a disposition involve nothing more than the acquisition of new dispositions? Here’s one way of formalizing APNT in an attempt to clarify just what’s wrong with this picture: 1. A manifestation of a disposition is constituted by a particular acquiring some properties: If some particular, a, manifests a disposition, then some particular b acquires some properties. (Possibly a = b, throughout.) 2. If all properties are dispositions, and if a manifests a disposition, then some b acquires some dispositions (and does not acquire any non-dispositions). 3. A disposition is either manifest or latent (producing no manifestation).

_____________ 24 25 26

That’s one interpretation of what’s going on in Mellor (1974, 171) and Goodman (1983, 41). Molnar 2003, 159. This objection appears in many places. A nice discussion appears in Molnar 2002, 173-181.

200

Jennifer McKitrick

4.

If all properties are dispositions and b acquires some dispositions, then either b’s new dispositions remain latent or some c acquires some dispositions. 5. If all properties are dispositions and c acquires some dispositions, then either c’s new dispositions remain latent or some d acquires some dispositions. 6. If all properties are dispositions and d acquires some dispositions, then either d’s new dispositions remain latent or some e acquires some dispositions. 7. etc. 8. Therefore, if all properties are dispositions, every manifestation of a disposition is constituted by either a. a particular having a disposition that produces no manifestation, or by b. a particular having a disposition that gives something a disposition, which produces no manifestation, or by c. a particular having a disposition that gives something a disposition, which gives something a disposition, which has no manifestation, or by … d. a particular having a disposition that gives something a disposition, which gives something a disposition, ad infinitum. 9. It is not plausible that every manifestation of a disposition is constituted by a, b, c, or … d. 10. Not all properties are dispositions. The strength of this argument depends on premise 9. The disjunctions enumerated in premise 8 are supposed to bring out the sense in which nothing really happens in the pandispositionalist world. One response that’s often made against the APNT argument is that it illegitimately assumes that dispositions are not fully real, and so that when something gains a disposition, nothing really happens. However, I acknowledge that something gaining a disposition is an event, or something happening. However, I find it odd to suppose that all that ever happens is that things loose and acquire dispositions. Maybe what’s wrong with the scenario described in premise 8 above is that it’s not clear how one could ever observe a manifestation of a disposition. This thought inspires the following variation on the APNT argument: 1. A disposition is either latent or manifest. 2. If a disposition is latent, it is not observable. 3. If a disposition is manifest, the disposition itself is at most indirectly observable. 4. Therefore, dispositions are never directly observable.

Dispositional Pluralism

201

5.

Therefore, if all properties are dispositions, no properties are directly observable. 6. The only way to indirectly observe a property is to directly observe some other property. 7. Therefore, if all properties are dispositions, all properties are unobservable. 8. Some properties are observable. 9. Therefore, not all properties are dispositions. This argument relies on the contingent claim that some properties are observable and perhaps the argument would not be sound in a world where observation is impossible. If one wants to oppose pandispositionalism in such a world, perhaps a dispositional account of “observable” could make it work. Never the less, the 8th premise above seems obviously true in the actual world, which is probably good enough to show that pandispositionalism is not true in the actual world. To summarize this section, one of the contrasting pair of absolutist positions concerning dispositions warrant further discussion, anti-dispositionalism (the view that no properties are dispositions) and pandispositionalism (the view that all properties are dispositions). I have argued that we have reasons to think that disposition claims are true, which gives us reason to think that dispositions exist. If the anti-dispositionalist remains unconvinced, my case needs to be supplemented with other arguments for the existence of dispositions. While dispositional pluralism is consistent with pandispositionalism, property dualism (the view that both dispositional and non-dispositional properties exist) is more in line with the spirit of pluralism, and is a more plausible view in its own right. The APNT objection shows pandispositionalism to be an unattractive metaphysical picture.

6. Conclusion Semantic and empirical evidence are consistent with there being different kinds of dispositions. Dispositions are a broad and heterogeneous kind of property. You may be interested in a certain kind of disposition – the natural, intrinsic, or fundamental. However, that focus should not be presented as an absolute claim that all dispositions are such and such. This may be a personal, linguistic preference on my part, but it is not an arbitrary preference. For one thing, dispositional pluralism is closer to ordinary English – both the term ‘disposition,’ its synonyms, and other recognizably dispositional predicates. I think it is better, other things being equal, to use terms in readily recognizable ways. Otherwise, one should clarify with a précising definition – stipulating that this loose, umbrella term is going to be used in a more specific

202

Jennifer McKitrick

way. Mentioning that you are using ‘disposition’ in the philosopher’s and not the lay person’s sense is insufficient.27 Even the “philosopher’s sense” is quite pluralistic, as the philosophical distinctions at the outset show. A second, pragmatic issue is the usefulness of theorizing about dispositions. Often, dispositional theorists try to generate interest in dispositions by pointing out how pervasive dispositional concepts are. This is false advertising if a scant minority of these concepts actually corresponds to real dispositions, on the theorists view. Many of the things theorists say about dispositions can shed light on “so-called” dispositional theories of value, belief, colors, beauty, knowledge, fitness and others. Theorists would be prudent not to undercut the importance of their work by using the term “disposition” so narrowly that the concept is not relevant to these other philosophical concerns. Thirdly, dispositional pluralism facilitates rather than restricts discussion among dispositional theorists. Often, those with sparse theories of dispositions express their willingness to yield to science the last word about fundamental properties. Therefore, on such views, whether any property, such as charge or mass, is a disposition, might be an unresolved scientific question. If so, we cannot confidently give any examples of genuine dispositions. Then we’ve traveled a long way from the useful concept that pervades our daily lives to one which we scarcely know how to apply. What are those things that we used to think of as dispositions? Are they non-dispositional properties, or nothing at all? How are we supposed to make sense of the ways we have been speaking? Speaking of numerous and various dispositions, and then considering, if we wish, whether any of these are ungrounded, extrinsic, derivative, or what have you, seems more helpful to metaphysical inquiry than starting off by stipulating away the existence of such dispositions.

Literature Armstrong, D. 1973. Belief, Truth, and Knowledge. London: Cambridge University Press. Armstrong, D., Martin and Place. 1996. Dispositions: A Debate. New York: Routledge. Bird, A. 1998. “Dispositions and Antidotes.” The Philosophical Quarterly 48: 227-234. Broad, C. D. 1925. The Mind and its Place in Nature. Harcourt Brace. Ellis, B. 2002. The Philosophy of Nature: A Guide to the New Essentialism. Montreal: McGillQueen’s University Press. Goodman, N. 1983. Fact, Fiction, and Forecast. Cambridge MA: Harvard University Press. Harre, R. 1970. “Powers.” The British Journal for the Philosophy of Science 21: 81-101. Heil, J. 2005. “Dispositions.” Synthese 144: 343-356. Jansen, Ludger. 2006. “The Ontology of Tendencies and Medical Information Sciences.”

_____________ 27

Prior 1985, 1.

Dispositional Pluralism

203

In WSPI 2006: Contributions to the Third International Workshop on Philosophy and Informatics, eds. E. Johansson, B. Klein, and T. Roth-Berghofer. Johnston, M. 1992. “How to Speak of the Colors.” Philosophical Studies 68: 221-263. Kim, J. 1993. Supervenience and Mind. New York: Cambridge University Press. Lewis, D. 1983. “New Work for a Theory of Universals.” Australasian Journal of Philosophy 61: 343-377. Lewis, D. 1986. Philosophical Papers, Volume II. New York: Oxford University Press. Locke, J. 1990. An Essay Concerning Human Understanding. London: Dent. Martin, C. B. 1994. “Dispositions and Conditionals.” The Philosophical Quarterly 44: 1-8. McKitrick, J. 2003. “A Case of Extrinsic Dispositions.” Australasian Journal of Philosophy 81: 155-174. Mellor, D. H. 1974. “In Defense of Dispositions.” The Philosophical Review 83: 157-181. Molnar, G. 2003. Powers. Oxford: Oxford University Press. Mumford, S. 1998. Dispositions. Oxford: Oxford University Press. Mumford, S. 2006. “The Ungrounded Argument.” Synthese 149: 471-489. Prior, E. 1985. Dispositions. Aberdeen: Aberdeen University Press. Prior, E., Pargetter, and Jackson. 1982. “Three Theses about Dispositions.” American Philosophical Quarterly 19: 251-257. Quine, V. W. O. 1969. “Natural Kinds.” In Essays in Honor of Carl G. Hempel, ed. N. Rescher, Dordrecht. Rosenberg, Gregg. 2004. A Place for Consciousness. Oxford: Oxford University Press. Ryle, G. 1949. The Concept of Mind. New York: Barnes & Noble. Shoemaker, S. 1980. “Causality and Properties.” In Time and Cause: Essays Presented to Richard Taylor, ed. P. van Inwagen, 109-136. Dordrecht: Reidel. Smith, A. S. 1977. “Dispositional Properties.” Mind 86: 439-445.

Dispositions and Their Intentions ANDREA BORGHINI Abstract: Dispositional Realism is the view according to which some denizens of reality – i.e., dispositions – are properties, may exist in the natural world, and have an irreducible modal character. Among Dispositional Realists, Charlie Martin, Ullin Place and George Molnar most notably argued that the modal character of dispositions should be understood in terms of their intentionality. Other Dispositional Realists, most notably Stephen Mumford, challenged this understanding of the modal character of dispositions. In this paper, I defend a fresh version of the intentional understanding of dispositions. I start by distinguishing between two questions about properties, respectively addressing their identity conditions and their individuation conditions. I, then, define categorical and dispositional properties in terms of their qualitative character, and examine their identity and individuation conditions. I conclude that the attribution of intentions is a conceptual tool introduced in order to alleviate the burdensome task of specifying the conditions of individuation of a disposition; however, such attribution does not affect the identity of a disposition.

Nominalists believe that there are no properties, but only particulars. Realists, instead, believe that some denizens of reality are properties. Among the Realists, some are Dispositional Realists: they believe that some properties are dispositions. Roughly, the latter are those entities with an irreducibly modal character that may be instantiated by objects in the natural (i.e., spatiotemporal) world. Other Realists, deny the existence of dispositional properties, thereby denying Dispositional Realism. In this paper, I shall assume Realism about properties. My aim will be to provide a fresh understanding of Dispositional Realism, which gives some merit to the view according to which dispositions have intentions.

1. Singling Out Properties As the debate on the ontological status of properties grew consistently over the past few decades, it was enriched by the addition of an increasingly technical vocabulary; it is thereby convenient to start off by introducing two distinctions that we shall employ during the discussion.

Dispositions and Their Intentions

205

First distinction. There are two kinds of theories of properties: abundant and sparse. Abundant theories incorporate the following principle: AB: For any predicate within the language of the theory, there is a property.1 Sparse theories, on the other hand, reject AB. For a sparse theorist, what properties there are cannot be established just by looking at the predicates within the language of the theory, as only a few predicates single out a property. In what follows, I shall endorse a sparse theory of properties.2 Second distinction. Some properties are said to be pure and some to be impure. Impure properties are those whose identity is fixed via reference to some particular. For example, Dustin Hoffman’s being cheerful is identified via reference to Dustin Hoffman. Pure properties are those that are identified without reference to any particular. For example, Being cheerful.3 The discussion that follows is concerned with pure properties. Derivatively, it might be applied to impure ones as well, although I shall not attempt to do so. One of the problems for a theory according to which properties are sparse and pure is to devise a criterion (or: some criteria) through which properties can be singled out. For example, suppose that, in the language of the theory, you have the predicate: ‘To be an electron’; does such a predicate, indeed, single out a property – Being an electron – or, rather, does it refer to an array of properties, perhaps a gerrymandered one? I shall label this the Singling Out Question (SOQ): SOQ:

Under what circumstances ought one to commit to the existence of a property?

As I see it, SOQ is composed of two sub-problems: the Identity Question (IDQ) and the Individuation Question (INQ): IDQ: INQ:

What makes a property the property that it is? Under what conditions is the individuation of a property achieved?

IDQ is a metaphysical question: to tell what fixes the identity of a property is to tell what makes it a unity (one property) and what distinguishes it from all other properties. INQ is an epistemic problem: it concerns the sort of evidence that is necessary to single out a property.

_____________ 1 2 3

Contradictory predicates as well as predicates that give rise to logical paradoxes ought to be excluded. For a recent presentation of AB, see Field 2004. See Armstrong 1979, Lewis 1986, 59-69, Swoyer 1996, Mellor and Oliver 1997, 1-33, and Shaffer 2004. See Khamara 1988, Humberstone 1996, Langton 1998 and Langton and Lewis 1998.

206

Andrea Borghini

In what follows, I shall first address IDQ and then INQ. My main argument will purport to show that the attribution of intentionality to dispositional properties is done in connection with INQ and not with IDQ. And, because of this, there is no need to attribute intentions to dispositions; yet, in order to individuate dispositions, and to single them out, talk of intentions comes in handy.

2. On the Identity of Properties I: Categorical and Dispositional Entities It is fairly ordinary to distinguish between two kinds of entities: categorical and dispositional. Yet, it is a major point of controversy how the distinction ought to be understood. The vast majority of the contestants focused on a certain purported difference between dispositional and categorical ascriptions: the first would entail conditionals, while the latter would not.4 Although this distinction enlightens a relevant side of the debate, to the eye of the Dispositional Realist it proceeds from a methodological vice: it is not through a linguistic distinction that we establish a metaphysical one. In other words, it is not methodologically sound to invoke certain features of ascriptions to substantiate differences among the ascribed entities. Linguistic facts can be hints or guides to ontological facts; yet, the latter will need an independent justification to be established. Here is the one I wish to propose for telling apart categorical from dispositional entities. We start by saying that all properties have a qualitative character, sometimes labelled also nomic role.5 This includes all aspects that each instance of a property entertains. Aspects are divided in two kinds, intrinsic and relational, defined as follows: Intrinsic aspect:

a feature that each instance of a property entertains regardless its environment.6 Relational aspect: a relation that each instance of a property entertains with instances of other properties.7

_____________ 4 5 6

7

See the discussion in Mellor 1974, Prior 1982, Prior 1985, Mumford 1998, 64-92, and, for an up to date overview, the article by Schrenk in this volume. See, for example, Robinson 1993 and Shaffer 2005. I reject the existence of intrinsic aspects, as they cannot be individuated – see Lewis 200+, Langton 2004 and Langton 1998. However, I shall leave this point on a side here, as it does not affect our discussion. Some prefer to define the qualitative character in a way that renders the subjects of the relations the particulars instantiating the properties, rather than the properties themselves. In the sequel, I shall speak as if the properties themselves are related. However, what I will say shall not depend

Dispositions and Their Intentions

207

Thus, the qualitative character will be defined as follows: Qualitative character of property P: all the intrinsic and relational aspects of P. For example, the nomic role of Being a molecule of oxygen will include, among others, that its instances will be related to instances of Being a molecule of hydrogen and Being water. Both intrinsic and relational aspects may be causally efficacious: that is, they do, or may, bring about changes in reality. Thus, identifying a property with its qualitative character is to tie it necessarily with (some of) the changes that it does and may bring about. For some, however, this proposal is not adequate, as a property may partially or completely change its qualitative character and still retain its identity. These postulate the existence of non-causally efficacious aspects, called quiddities, such that each property – purportedly – has one and only one quiddity and each quiddity belongs to one and only one property. Thus, the quiddity of Being a molecule of oxygen will be peculiar to such a property and will make up, at least partially, its identity.8 Although I am suspicious of the theoretical plausibility of quiddities, what I will say about the identity of properties will not hinge on whether properties have quiddities. I shall, therefore, remain neutral with respect to this point. Now, some aspects seem to be an always-or-never affair: when an instance of a property possesses it, it is manifest at all times. For example, instances of Being round are, at any time, round; instances of Being a father are, at all times, related to those of Being a child. In this sense, such a kind of aspects can be labeled as categorical. On the other hand, some aspects seem to be such that they can lay latent at a time and be manifested at another time. Instances of Being fragile, for example, are related only occasionally with those of Being broken. In this sense, some aspects can be labeled as dispositional.9 We, thus, have a categorical vs dispositional distinction at the level of properties’ aspects:

_____________ 8 9

on this choice and, with some efforts, it is possible to rewrite it so that the qualitative character will involve relations among bearers of the properties rather than instances of the properties. See Armstrong 1989, Lewis 200+, and Shaffer 2005. It is the conviction of the majority that all dispositional aspects are relational. And, in the sequel, I shall employ ‘dispositional relation’ as synonymous with ‘dispositional aspect’. But, this will be no more than a terminological choice: I shall not attempt to undermine the thesis that there are non-relational dispositions. For a recent discussion, see McKitrick 2003.

208

Andrea Borghini

Categorical aspect: Dispositional aspect:

an aspect such that, when included in the qualitative character of an instance of a property, is manifest at all times. an aspect such that, when included in the qualitative character of an instance of a property, may manifest at a time and not manifest at another time.

It should be noted that these definitions are compatible with an aspect not being shared by all instances of a property. Thus, it might be that different instances of Being uranium do not share certain dispositional aspects. Whether or not to accept such properties will depend on how strictly one defines the identity of a property. I shall leave this issue open. Can we derive, from this, a distinction at the level of properties too? There are two ways of doing it. First way. Define categorical properties as those whose qualitative character includes some and only categorical aspects; define dispositional properties as those whose qualitative character includes some and only dispositional relations. An advocate of this view is Molnar 2003: Categorical property 1: Dispositional property 1:

a property whose qualitative character includes some and only categorical aspects. a property whose qualitative character includes some and only dispositional aspects.

Second way. Define categorical properties as those whose qualitative character includes some (but, perhaps, not only) categorical aspects; define dispositional properties as those whose qualitative character includes some (but, perhaps, not only) dispositional aspects. An advocate of this view is Martin’s contribution to Armstrong et al. 1996.10 Categorical property 2: Dispositional property 2:

a property whose qualitative character includes some categorical aspects. a property whose qualitative character includes some dispositional aspects.

_____________ 10

A third way sees dispositional and categorical properties as identical. That is, according to this view there is no metaphysical distinction between categorical and dispositional aspects; still, both dispositional and categorical ascriptions may have truth-makers and these will be one and the same kind of entity. See for example Mumford 1998 and Heil 2003. As it shall become clear further on, I shall not consider this way, as I believe that it cannot properly accommodate for the primitive modal character of dispositions. If (some) modal sentences are true in virtue of some primitive modal entities, and reality includes entities that are non-modal, then it cannot be claimed that modal and non-modal entities are identical.

Dispositions and Their Intentions

209

Only the first way can be properly said to define two kinds of properties; the second way, just makes the categorical and the dispositional two aspects of properties. Indeed, all properties that entertain some, but not only, categorical aspects will entertain some dispositional relations; and, vice versa, all properties that entertain some, but not only, dispositional relations will entertain some categorical aspects. So, the categorical and the dispositional will not be distributed among properties in a mutually exclusive way. Now, to clarify the distinction, an advocate of the second way might introduce a distinction between essential and accidental aspects – call this second way*. Even if a property has both categorical and dispositional aspects, only one or the other kind can be essential to the property. Thus, the distinction will be as follows: Categorical property 2*: Dispositional property 2*:

a property whose qualitative character includes some essential aspects and these are all categorical. a property whose qualitative character includes some essential aspects and these are all dispositional.

Furthermore, the advocate of the second way* might also introduce a milder distinction among dispositional and categorical properties as follows: Aspect categorical property:

a property whose qualitative character includes some categorical aspects, but it is not categorical 2*. Aspect dispositional property: a property whose qualitative character includes some dispositional aspects, but it is not dispositional 2*.

More below, we shall draw further considerations regarding the two ways. We can for now conclude that, whichever we choose, we have two kinds of aspects and, derivatively, two kinds of properties.

3. On the Identity of Properties II: Primitive Modalities We shall now consider IDQ with respect to the two kinds of properties: what makes a categorical property the property it is? And, what makes a dispositional property the property that it is?

210

Andrea Borghini

Presumably11, the identity of a categorical property will be fixed by its aspects as well as by its quiddity (if it has one). The identity of a dispositional property, on the other hand, will be fixed by the relations that it is disposed to entertain with other properties as well as by its quiddity (if it has one). While many are ready to accept categorical aspects of properties, the number of those who are ready to accept dispositional relations among properties is smaller. Talk of dispositions is suspicious. Dispositions are but promises; and dispositional relations are but promises of relations. For this reason, many attempted to reduce the dispositional to the categorical. The debate, over the last few decades, has been extensive. Slowly, the number of arguments against the possibility of a reduction has grown larger, and with it the number of the Dispositional Realists.12 Still, there seems to be some disagreement regarding primitive modalities. It is opportune, at this point, to understand where a Dispositional Realist may locate herself within the contemporary debate on the metaphysics of modality. Contemporary theories of modality recognize that some sentences that contain modal terms (briefly: modal sentences) are irreducible. Thus, if we believe that modal sentences have a truth-value, we need to make room for the existence of some irreducibly modal entities. For many years it was believed that David Lewis’s modal realism was an exception to this; however, a number of criticisms has now shown that modal realism might fail to define modal facts in terms of non-modal ones.13 The purpose of a metaphysical theory of modality can, indeed, be seen as that of providing a satisfactory account of the irreducibly modal entities. Among the proposals, the so-called modalist position, as it has been defended by Forbes 1985, Forbes 1989 and Chihara 1998 is perhaps the most liberal: it accepts all sorts of modal propositions – without regard to their constituents – as primitively modal. Each proposition, whose translation into the language of a modal semantics (a Kripke-style possible worlds semantics) contains some occurrences of the symbols ‘ ’ or ‘I’, is a primitive modal proposition, and it carries a reference to primitive modal entities (presumably: primitive modal propositions themselves, or primitive modal facts or states of affairs). Among the constituents of the primitive modal propositions there are entities of all sorts: members of the natural world, abstract mathematical entities, fictional entities, and what else you may have to add to the list.

_____________ 11 12

13

For some, the identity of a property will be entirely fixed by the quiddity, no matter what its qualitative character be. Among the many contributions arguing in favor of this side of the debate, or acknowledging its achievements, see: Manley and Wasserman 2008, Molnar 2003, Molnar 1999, Ellis 2002, Ellis 2001, Mumford 1998, Mellor 2000, Mellor 1974, Martin 1994, Popper 1990, Harré 1970 and Fara 2005. See Shalkowski 1994, Melia 2003, Divers 2002, Cameron 2008, Denby 200+ and Borghini 2007.

Dispositions and Their Intentions

211

On the other hand, some believe that not all apparently modal propositions need to be regarded as primitive. Rather, we ought to select certain propositions as irreducibly modal or, perhaps, some entities embedded in those propositions; all other modal propositions will, then, be explained in terms of those. Thus, for example, for the linguistic ersatzist, primitive modal propositions involve linguistic entities; for the combinatorialist, primitive modal propositions are re-combinations of actual entities; for the fictionalist, they are all entities of a fiction. The Dispositional Realist has, fundamentally, two options: being a modalist or endorsing a dispositional theory of possibility – a recently advanced theory, according to which all modal sentences are interpreted as attributions of primitive dispositions (see Borghini and Williams 2008). The other options are not open. Indeed, the Dispositional Realist claims that (at least some) dispositions – that is: certain modal entities – belong to the (actual) natural world; however, all other theories deny this: according to them, modal entities are – for example – linguistic entities, or fictions, or re-combinations of actual entities, or concrete worlds other than the actual. But, the extreme liberality of the modalist might be unpalatable to most Dispositional Realists, for two reasons. The first has to do with the business of linguistic reduction. Most Dispositional Realists include dispositions among the primitive modal entities on account that you cannot apparently reduce dispositional talk to talk of categorical entities. But, for the probabilistic, mathematical, or counterfactual propositions we might find a reduction. Indeed, that is what the dispositional theory of possibility claims: that all modal talk can ultimately be interpreted as dispositional talk; that is, as attributing some dispositional properties (or: relations) to certain entities. The second reason is metaphysical. The Dispositional Realist will not (or: ought not to) accept a proliferation of the kinds of modal primitives. Dispositions are properties; hence, metaphysical simplicity suggests seeking for an explanation according to which all modal primitives are properties. The modalist, on the other hand, will typically endorse a view according to which modal primitives are facts or situations or states of affairs rather than properties. Whether the Dispositional Realist be a modalist or a dispositionalist, the following moral can be drawn for present purposes: if you accept that dispositions are real, then you need to look no further for your modal primitives when it comes to disposition ascriptions. The modalist will not attempt to reduce dispositions to some other kinds of entities; and the dispositionalist – by definition – will do the same. From this follows that the so-called identity theory of dispositions, according to which dispositional and categorical properties are identical, falls

212

Andrea Borghini

short of an adequate explanation of the nature of possibility.14 This is a remark that has gone unnoticed so far in the disposition debate. If you take modal sentences at face value, possibility and actuality cannot coincide. It is obvious that there are possibilities that are not actualized. Thus, it is obvious that there are modal entities that are distinct from actual ones. But, dispositions are a kind of modal entities; and many dispositions are never manifested; hence, dispositions cannot be identical to some non-modal entities. In other words: if you take modal sentences at face value, the realm of the possible and the one of the actual are distinct. For the Dispositional Realist, dispositions represent part of (if not all) the realm of the possible; hence, they cannot be identical with the realm of the actual. Indeed, by definition, a Dispositional Realist cannot attempt to reduce dispositional talk to talk of recombinations, or of linguistic entities, or of concrete worlds other than our own. What the Dispositional Realist accepts is that the realm of the possible and the one of the actual may both be part of the (actual) natural world, as dispositions may belong to objects in the natural world. Let us now explore a little further the peculiarity of dispositions as modal primitives. A modal primitive is an entity which expresses a possibility: that a certain situation can, could have, might … obtain. If we say that dispositions are modal primitives, we are accepting that certain situations are disposed to obtain. This, roughly, means that they will obtain, if certain conditions will also obtain; but, this is no definition of the entity in question: it is just an illustration, a tool that is useful for us to talk about the entity in question. Dispositions are a ductile modal primitive. They belong to a certain ontological category – namely, properties; yet, it is left open to what sort of individuals (if any) they are ascribed. Thus, you may find dispositions ascribed to entities in the natural world as well as to mathematical entities, social institutions, fictional entities, or any other realm of being one might envisage. This gives a great explanatory power to the Dispositional Realist, at the cost of admitting one kind of properties. To accept that some modal primitives thrive in the natural world poses obvious epistemic worries: it amounts to giving plausibility to the hypothesis that there might be infinite features of our environment that lie hidden to us. Still, we also have strong reasons to swallow this pill. And, we ought to resist the temptation to make the swallowing less unpleasant by trying to further explain the ontological structure of our modal primitives. Dispositions are just that: primitive properties, perhaps always lying hidden to our sensory perceptions. Joe enjoys the property Being a father; along the same lines, perhaps he enjoys the property Being brave in wartime; but, (hopefully) we might never find that out. Both entities are properties: if you can,

_____________ 14

For an exposition of the identity theory, see Mumford 1998 and Heil 2003.

Dispositions and Their Intentions

213

can, ontologically speaking, make sense of the first, you ought to be able to make sense also of the latter. The only difference between categorical and dispositional entities consists in the fact that the latter, sometimes, lie hidden to our senses. Yet, this ought to be no scandal: if you accept that modal propositions have truth-values, then you ought to accept that there are some modal entities that lie hidden to your senses. Perhaps, it is a scandal that such entities belong to the natural world: but this calls for an epistemic justification, rather than a further analysis of the metaphysical nature of those entities. Metaphysically speaking, they are just properties. However, some Dispositional Realists have given in to the temptation. For example, Martin and Pfeifer 1986, Place 1996, Place’s contribution to Armstrong et al. (1996, 19-33) and Molnar 2003 invoked an additional notion to explain the metaphysical nature of dispositions: intentions. According to their proposal, intentionality is the mark of the dispositional. By ‘intention’ here it is not meant that plan to carry out a certain action; rather, that feature of a mind to be in a state, which is about something without being that thing. Joe can ‘intend’ the apple in front of him, without being that apple or having the apple as a part of himself. Along the same lines, if the glass is fragile, it means that it ‘intends’ breaking, whose state would be revealed where the right circumstances to obtain. From this, a debate ensued on whether dispositions ought to be understood – at the metaphysical level – in terms of intentions or, rather, in terms of other kinds of entities, such as functions (for the latter suggestion, see Mumford 1999; for a reply, Place 1999). But, in light of what we have said, it should be clear that dispositions ought not to be understood in terms of other modal notions. Still, invoking intentions can be useful; not to answer IDQ, rather, to answer INQ. To this task we shall now turn.

4. On The Individuation Of Properties I: Categorical Properties Suppose you agree that there are both categorical and dispositional entities. You ought to wonder, next, how the individuation of these entities is achieved. At first, let us briefly consider categorical entities. Presumably, those that inhabit the natural world will be individuated in terms of the way they manifest themselves to our senses. Those that do not inhabit the natural world (if there are any) will be individuated in terms of the concepts through which they are expressed to us, be those concepts abstractions from sensory experiences or a priori.

214

Andrea Borghini

Here we find an argument in favor of the second way (and the second way*) of defining categorical and dispositional properties: the way that allows for a property of a kind to entertain aspects of the other kind. Indeed, in order to individuate a categorical property, we need to postulate that it is capable of dispositional relations, namely, certain ways in which the property interacts with an observer. Thus, it is natural to define a categorical property as that entity which, under normal conditions, is disposed to bring about certain sensory or conceptual experiences in a subject. Being red will bring about, under certain standard conditions, the visual experience of red; Being an equilateral triangle will bring about, under certain standard conditions, the conceptual experience of an equilateral triangle.15 The disposition to bring about a sensory or conceptual experience will not be the defining aspect of a categorical property. However, if we would not allow the property to have such an aspect, we could not account for its capacity of being individuated. This capacity need not be an essential aspect of the property; that is, the property might exist, even if unknowable. Still, it seems to be an essential aspect to the individuation of the property. We should – at this point – mention an alternative explanation, which does not compel the acceptance of the second way of defining categorical and relational properties. Perhaps, individuation is not a dispositional relation between an individuating subject and an individuated entity; rather, it is just a state of the subject. In other words: epistemic attitudes are not part of the qualitative aspects of a property. So, there is no dispositional relation involved and no need to attribute dispositional aspects to categorical properties.16 I shall, for the time being, leave open the choice as to which explanation has more merit. Either way, the individuation of a categorical property will involve some form of evidence remarking the existence of categorical aspects. And, for present purposes, it is relevant to stress that such evidence may be achieved through the senses. We cannot make a similar claim with respect to dispositional aspects. This is what puzzles their detractors.

5. On The Individuation Of Properties II: Dispositional Properties If dispositional aspects cannot be individuated in terms of the way they reveal to our senses – as, indeed, they might never be revealed – how then shall we account for their individuation? This is where invoking the intentional character of a disposition comes in handy.

_____________ 15 16

Different, but related arguments are offered also in Franklin 1988 and Blackburn 1990. For this view, see Ducasse 1942, Chisholm 1957, Sellars 1967, Tye 1984, and Thomas 2003.

Dispositions and Their Intentions

215

The intention of a disposition is, simply, the situation that the property is disposed to bring about, if the right conditions obtain. Fragility has the intention of breaking, were the right conditions to obtain. It is important to stress that no self-consciousness is typically attributed to the intentional character of dispositions.17 Indeed, the defenders of the intentional character of dispositions reject the so-called Brentano thesis, according to which intentionality is both necessary and sufficient for telling apart the psychological and the nonpsychological. (For a detailed discussion, see Molnar 2003, 60-81.) I believe that Martin and Pfeifer, Place, and Molnar’s aim is noble. Having to defend the relatively novel Dispositional Realism, they attempted to clarify the concept of a disposition in order to clarify the conditions under which dispositions may be singled out. However, it is my conviction that a mistake lies at the foundation of their explanation. The problem with singling out dispositions does not call Dispositional Realists to revise the primitive modal character of dispositions. It is not a metaphysical clarification that is called for. As primitive modal entities, dispositions are no more mysterious than re-combinations, ersatz linguistic entities or worlds other than our own. Each of those categories is supposed to possess an irreducibly modal character: they are combinations, propositions, worlds that could have been actual. Rather, the problem with singling out dispositions has to do with the difficulty in individuating them. Dispositions may lie hidden for their entire existence: how can we even start talking about them? It is here that one can appeal to the situation that a disposition intends to bring about. Intentional talk is a conceptual ladder used to individuate dispositions. In connection with the difficulties associated with the individuation of dispositions, we shall now consider a problem affecting the definition of the qualitative character of a property. Thus far, we have taken for granted that the qualitative character is made out by a multiplicity of aspects. But, when we move to consider INQ, this claim reveals to be not as innocuous as it might at first appear. Suppose that each instance of a property has multiple aspects (be them dispositional or categorical). Suppose also that the aspects are such that they are shared among different properties. Then, in order to individuate a property one would have to individuate all of its aspects; to individuate some but not all, might leave indeterminate which property is under consideration. But, aspects might be infinite. Hence, individuation might never be achieved. We could, then, suppose that properties can share aspects only to a limited extent. This, however, burdens the theory with the task of finding out those aspects of a property that can be shared and those that are specific. The

_____________ 17

See Martin and Pfeifer 1986, Place 1999, Place 1996 and Molnar 2003.

216

Andrea Borghini

risk is that, in order to ease the task of individuation, we end up positing ad hoc aspects for each property. Finally, we could suppose that aspects are exclusive: they cannot be shared at all among different properties. This would render the theory even more unpalatable, though. Even assuming that intrinsic aspects can be regarded as exclusive to each property, it seems to run against evidence to say that relations among properties are exclusive. For example Being a mother and Being a father seem to share a relationship with Being a child. Then, why not to consider each aspect a property on its own? For example, ‘Being a mother’ would not count, under this view, as a genuine property; rather, it would be the name of two properties: Being a parent and Being a female; each of those properties would have just one aspect: Being a parent would have a relationship to Being a child; Being a female would have a relationship to – say – the property of being a certain reproductive organ. This proposal implies a bizarre ontology of properties: a property would be entirely constituted either by an intrinsic or by a relational aspect. No property could entertain more than one relation with another. This seems to run counter to our evidence too. Consider, for example, a dispositional property; on the face of it, such a property seems to require a relation to a vast number of other properties in order to be manifested. So, which option should we choose? I shall make a plea for the first one. To individuate a property is a hard task and, perhaps, it cannot ever be fully achieved. Dispositional properties offer a handy illustration. Determining the properties to which a dispositional property is related proved to be hopeless, in most cases. It is for this reason that disposition ascriptions cannot be analyzed in terms of conditionals. And, it is for this reason that the conceptual ladder of the intentional character comes in handy when we want to individuate a disposition. Invoking the intentional character is a tool to focus the task of individuation on a small number of properties: those embedded in the situation for which the dispositional property has an intention. Talk of intentions, however, does not affect the identity conditions of a dispositional property; and it does not solve the impasse of spelling out all the aspects included in the qualitative character of a property; more modestly, it alleviates the impasse, by focusing only on those aspects that the property seems to have an active role in bringing about.

6. Conclusions A debate has spanned on whether Dispositional Realists’ attribution of an intentional character to dispositions made of them panpsychists or mei-

Dispositions and Their Intentions

217

nongianists. I believe both of those allegations can be rejected. Dispositions are modal primitives. As such, they do not compel us to the literal existence of intentions or to entities that exist in a different sense than actual entities. More simply, dispositions compel us to the existence of entities (i.e., dispositional aspects) that bring about changes in reality, but not in a permanent way. However, we invoke intentions in order to individuate dispositions. This has to do with the general business of singling out the entities we are dealing with, not with the business of defining the identity of such entities. Talk of intentions helps to alleviate the difficulties we face in individuating a dispositional property.

Literature Armstrong, D. M. 1989. A Combinatorial Theory of Possibility. Cambridge: Cambridge University Press. Armstrong, D. M., C. B. Martin, and U. T. Place. 1996. Dispositions. A Debate. New York: Routledge. Bird, A. 1998. “Dispositions and Antidotes.” Philosophical Quarterly 48: 227-234. Blackburn, S. 1990. “Filling in Space.” Analysis 50: 62-65. Borghini, A. 2007. “Vizi e virtù del concretismo.” Annali del Dipartimento di Filosofia dell’Università di Firenze 13: 183-195. Borghini, A. and Williams, N.E. 2008. “A Dispositional Theory of Possibility.” Dialectica, 62: 21-41. Cameron, R. 2008. “Recombination and Intrinsicality.” Ratio 21: 1-12. Chihara, C. 1998. The Worlds of Possibility. Oxford: Oxford University Press. Chisholm, R. 1957. Perceiving. Ithaca (N.Y.): Cornell University Press. Denby, D. A. 200+. “Generating Possibilities.” Philosophical Studies, forthcoming. Divers, J. 2002. Possible Worlds. New York: Routledge. Ducasse, C. J. 1942. “Moore’s Refutation of Idealism.” In P. Schilpp, ed. The Philosophy of G.E. Moore. Chicago: Northwestern University Press. Ellis, B. 2001. Scientific Essentialism. New York: Cambridge University Press. Ellis, B. 2002. The Philosophy of Nature – A Guide to the New Essentialism. Montreal: McGillQueen’s University Press. Fara, M. 2005. “Dispositions and Habituals.” Noûs 39: 43-82. Field, H. 2004. “The Consistency of the Naïve Theory of Properties.” The Philosophical Quarterly 54: 78-104. Forbes, G. 1985. The Metaphysics of Modality. Oxford: Clarendon Press. Forbes, G. 1989. The Languages of Possibility. Oxford: Blackwell. Franklin, J. 1988. “Are Dispositions Reducible to Categorical Properties?” Philosophical Quarterly 38: 62-64. Harré, R. 1970. “Powers.” British Journal for the Philosophy of Science 21: 81-101. Heil, J. 2003. From an Ontological Point of View. New York: Oxford University Press. Humberstone, I. L. 1996. “Intrinsic/Extrinsic.” Synthese 108: 205-267. Jackson, F., Prior, E., and Pargetter, R. 1982. “Three Theses about Dispositions.” American

218

Andrea Borghini

Philosophical Quarterly 19: 251-256. Khamara, E. J. 1988. “Indiscernibles and the Absolute Theory of Space and Time.” Studia Leibnitiana 20: 140-159. Langton, R. 1998. Kantian Humility. Our Ignorance of Things in Themselves. Oxford: Clarendon Press. Langton, R. 2004. “Elusive Knowledge of Things in Themselves.” Australasian Journal of Philosophy 82: 129-136. Langton, R. and Lewis, D. K. 1998. “Defining ‘Intrinsic’.” Philosophy and Phenomenological Research 58: 333-345. Lewis, D. K. 1997. “Finkish Dispositions.” Reprinted in: D. K. Lewis. 1999. Papers in Metaphysics and Epistemology, 133-151. New York: Cambridge University Press. Lewis, D. K. 200+. “Ramseyan Humility.” In D. Braddon-Mitchell and R. Nola, eds. The Canberra Plan. forthcoming. Manley D. and Wasserman R. 2008. “On Linking Dispositions with Conditionals.” Mind 117: 59-84. Martin, C. B. 1994. “Dispositions and Conditionals.” The Philosophical Quarterly 44: 1-8. Martin, C. B. and Pfeifer K. 1986. “Intentionality and the Non-Psychological.” Philosophy and Phenomenological Research 46: 531-554. Mackie, J. L. 1977. “Dispositions, Grounds, and Causes.” Reprinted in R. Tuomela, ed. 1978. Dispositions, 99-107. Dordrecht: D. Reidel Publishing. McKitrick, J. 2003. “A Case for Extrinsic Dispositions.” Australasian Journal of Philosophy 81: 155-174. Melia, J. 2003. Modality. Montreal & Kingston: McGill-Queen’s University Press. Mellor, D. H. 1974. “In Defense of Dispositions.” The Philosophical Review 83: 157-181. Mellor, D. H. 2000. “The Semantics and Ontology of Dispositions.” Mind 109: 757-780. Mellor, D. H. and Oliver, A., eds. 1997. Properties. Oxford: Oxford University Press. Molnar, G. 1999. “Are Dispositions Reducible?” Philosophical Quarterly 49: 1-17. Molnar, G. 2003. Powers. New York: Oxford University Press. Mondadori, F. and Morton, A. 1976. “Modal Realism: The Poisoned Pawn.” The Philosophical Review 85: 3-20. Mumford, S. 1998. Dispositions. New York: Oxford University Press. Mumford S. 1999. “Intentionality and the Physical: A New Theory of Disposition Ascription.” The Philosophical Quarterly 49: 215-225. Mumford, S. 2006. “The Ungrounded Argument.” Synthese 149: 471-489. Place, U. T. 1996. “Intentionality as the Mark of the Dispositional.” Dialectica 50: 91-120. Place, U. T. 1999. “Intentionality and the Physical: A Reply to Mumford.” The Philosophical Quarterly 49: 225-231. Popper, K. 1990. A World of Propensities. Bristol: Thoemmes. Prior, E. 1982. “The Dispositional/Categorical Distinction.” Analysis 42: 93-96. Prior, E. 1985. Dispositions. Aberdeen: Aberdeen University Press. Robinson, D. 1993. “Epiphenomenalism, Laws, and Properties.” Philosophical Studies 69: 134. Sellars, W. 1967. Science and Metaphysics. London: Routledge and Kegan Paul. Shaffer, J. 2004. “Two Conceptions of Sparse Properties.” The Philosophical Quarterly 85: 92102. Shaffer, J. 2005. “Quiddistic Knowledge.” Philosophical Studies 123: 1-32. Shalkowski, S. 1994. “The Ontological Ground of the Alethic Modality.” Philosophical Review 103: 669-688. Swoyer, C. 1996. “Theories of Properties: From Plenitude to Paucity.” Philosophical Perspectives 10: 243-264. Thomas, A. 2003. “An Adverbial Theory of Consciousness.” Phenomenology and the Cognitive

Dispositions and Their Intentions

219

Sciences 2: 161-185. Tye, M. 1984. “The Adverbial Theory of Visual Experience.” The Philosophical Review 93: 195-225.

IV. The Role of Dispositions in Scientific and Philosophical Contexts

Dispositions in Physics1 ANDREAS HÜTTEMANN

1. Introduction According to a well-known story, physics no longer appeals to dispositional properties since early modern times. Descartes, for instance, writes in a letter to Morin (13.7.1638): Compare my assumptions with the assumptions of others. Compare all their real qualities, their substantial forms, their elements and countless other such things with my single assumption that all bodies are composed of parts. […]. Compare the deductions I have made from my assumption – about vision, salt, winds, clouds, snow, thunder, the rainbow, and so on – with what the others have derived from their assumptions on the same topics.2

Real qualities and substantial forms were conceived of as dispositional properties. Descartes rejected them because he took them to be explanatorily useless. And indeed Descartes was not the only one to reject dispositions.3 The point, however, is that a certain kind of disposition was rejected – dispositions as causal powers. In what follows, I will argue there is a different sense of dispositions that is presupposed in physical practice: Dispositions as contributors. What I will claim is that dispositions as contributors are presupposed in physics precisely because of the competing explanatory strategy that Descartes mentions: explaining the behavior of bodies in terms of the parts. I will argue firstly that law-statements should be understood as attributing dispositional properties. Second, the dispositions I am talking about should not be conceived as causes of their manifestations but rather as contributors to the behavior of compound systems. And finally I will defend the claim that dispositional properties cannot be reduced in any straightforward sense to

1 2 3

I would like to thank Rosemarie Rheinwald and the participants at the Wittenberg conference on dispositions for helpful comments on earlier versions of the paper. Descartes 1991, 107 (AT II, 200). Newton, for instance, writes: “…since the moderns – rejecting substantial forms and occult qualities – have undertaken to reduce the phenomena of nature to mathematical laws …” (Newton 1999, 381).

224

Andreas Hüttemann

non-dispositional (categorical) properties and that they need no categorical bases in the first place.

2. Laws attribute Dispositional Behavior to Physical Systems Let me start by giving some examples of the attribution of dispositional properties in physics. Newton in the third part of his Principia argues that certain well-known phenomena – among them Kepler’s laws – can be deduced from his theory. Proposition 13 of Book III of the Principia captures Kepler’s first and second law: The planets move in ellipses that have a focus in the center of the sun, and by radii drawn to that center they describe areas proportional to the times. (Newton 1999, 817)

Newton argues for this theorem as follows: Now that the principles of motions have been found, we deduce the celestial motions from these principles a priori. Since the weights of the planets toward the sun are inversely as the squares of the distances from the centers of the sun, it follows (from book 1, props, 1 and 11, and prop. 13, corol. 13) that if the sun were at rest and the remaining planets did not act upon one another, their orbits would be elliptical having the sun in their common focus, and they would describe areas proportional to the times. (italics mine) (Newton 1999, 817/8).

The point is that the behavior that Newton attributes to the planetary system and that he deduces from his principles is not one that the planetary system in fact displays. It would display or manifest the behavior if there were no external influences. As a matter of fact, Newton argues, these influences are small and thus we get something very close to what Kepler attributed to the planets. For our purposes, however, it is not important whether or not the external influences are small. The important point is that certain behavior is attributed to the planetary system provided certain circumstances obtain, which as a matter of fact fail to obtain. The behavior that Newton attributes to the planetary system is not manifest. It would be manifest if the influences were absent. Newton’s first law is another and a much simpler example for the attribution of conditional behavior: Every body continues in its state of rest or of uniform motion in a straight line, unless it is compelled to change that state by forces impressed upon it. (Newton 1999 416)

Again, the law attributes behavior to bodies. It says that the bodies would manifest certain behavior in the absence of disturbing factors or forces. On the view that laws should be read as attributing properties to systems, the laws we have just discussed attribute a property to bodies that becomes manifest given the right circumstances. In other words, Newton’s first law attributes a

Dispositions in Physics

225

dispositional property to the body. The bodies have the property whether or not the forces are present. The property becomes manifest if the forces are absent. This observation on Newton’s first law seems to be true of most laws in physics. They attribute behavior to physical systems provided there are no external influences, no disturbing factors. Take the following example: Hydrogen atoms behave in accordance with the Schrödinger equation with a Coulomb potential.

This law is true provided there are no external electric or magnetic fields. Thus it attributes a behavior to hydrogen atoms that becomes manifest only given certain conditions are fulfilled. That is, the law statement attributes a dispositional property to the atoms. I employ the categorical/dispositional distinction as follows: A dispositional property is a property that, if instantiated by an object, is manifest under specific conditions only. A categorical property by contrast is a property that, if instantiated by an object, is manifest under all conditions. So, according to this distinction categorical properties are limiting cases of dispositional properties. Two remarks: First, there is of course the problem of exactly specifying under which conditions a particular disposition becomes manifest. The way I draw the distinction does not imply that this can be finitely done. Second, in the limiting case of a categorical property the distinction between a property and its manifestation doesn’t do any work. This is certainly not the orthodox way to draw the distinction, but the usual suspects fall on the right sides. Solubility and fragility, if instantiated by an object, become manifest under specific conditions. On the other hand, triangularity or massiveness, if instantiated by an object, are manifest under all conditions. Given this distinction we can sum up what has been said about laws as follows: The behavior that laws of nature typically attribute to physical systems is a behavior that becomes manifest given certain circumstances obtain – usually the absence of external disturbances. Law statements thus attribute dispositions to physical systems.

3. Missing Instances – a Problem for the Categorical View Let me now illustrate how the assumption of dispositional properties solves some problems that arise on the competing view that laws attribute categorical properties to physical systems. Here is the problem: If laws attribute categorical properties to physical systems, they attribute properties that, if instantiated – i.e. if the system in question has the property –, are manifest under all

226

Andreas Hüttemann

conditions. A fortiori, according to this view, if a categorical property is not manifest, it is not instantiated – the system does not have the property. But clearly, the behavior that our above law statements describe is hardly ever – if at all – manifest. Thus the laws that I have cited above seem to have no instances. There seem to be no systems that display the behavior in question. There simply are no bodies that are not compelled to change their state due to forces impressed on them. So Newton’s first law is never instantiated (with the possible exception of the universe as a whole). What then is the point of affirming rather than denying Newton’s first law? Similarly, there simply are no hydrogen-atoms that live in field-free surroundings, so what is the point of the claim that hydrogen atoms behave according to the Schrödinger equation with the coulomb-potential? In fact we have three problems here: Laws often appear to have no instances. Thus: • • •

How can laws that attribute properties to physical systems, which appear to have no instances, be tested? How can laws that attribute properties to physical systems, which appear to have no instances, be explanatorily relevant for actual phenomena? Why are we interested in laws that attribute properties to physical systems, which appear to have no instances? What’s the point of postulating these laws that describe what is going to happen when there are no forces – even though we know that these are always present? (Cartwright 1989, 189)

4. Testing Laws that appear to have no Instances I will start by having a look at how laws that attribute properties to physical systems, which appear to have no instances, can be tested. An analysis of what goes on in testing will provide answers for the other questions as well. Let me start with an example. According to Galileo, all bodies fall with the same speed in a vacuum. This is how Salviati, Galileo’s spokesman, argued for this claim: We have already seen that the difference of speed between bodies of different specific gravities is most marked in those media which are the most resistant: thus, in a medium of quicksilver, gold not merely sinks more rapidly than lead but it is the only substance that will descend at all; all other metals and stones rise to the surface and float. On the other hand the variation of speed in air between balls of gold, lead, copper, porphyry, and other heavy materials is so slight that in a fall of 100 cubits a ball of gold would surely not outstrip one of copper by as much as four fingers. Having

Dispositions in Physics

227

observed this I came to the conclusion that in a medium totally devoid of resistance all bodies would fall with the same speed.4

What is the law that has been tested? It is certainly not the law that all bodies fall with the same speed. That would be false. A tomato and a table tennis ball, when released from the top of staircase, reach the ground floor in different times. The law that has been tested is not a law about falling bodies in a medium. Rather the law in question pertains to the vacuum. All bodies fall with the same speed in a vacuum. The law concerns situations in which no disturbing factor, such as a resisting medium, intervenes. The role of the experimental results concerning falling bodies in the media is to provide evidence for the law in the vacuum. Salviati is not interested in the fact or the regularity that gold sinks in quicksilver as such. But rather because it provides evidence (in the context of the other results) for what would happen in the vacuum. On the view that law-statement describe categorical properties this procedure generates a puzzle. In the actual experimental situation the alleged categorical property is not manifest and therefore not instantiated. How can Salviati gain evidence for a counterfactual situation in which the categorical property would be manifest, but which is, after all, counterfactual? I will now try to explain how on the dispositionalist view this puzzle disappears. The remarkable thing is that the law concerns a situation (vacuum) that is not realized and maybe even non-realizable. However, I assume that nobody will deny that it can nevertheless be tested along the lines Salviati has outlined. So how does it work? The following explanation of the test procedure seems reasonable: The first point is that, if Saliviati’s argument is to work, the ideal (the counterfactual) and the less than ideal (the real) situation must be connected. The ideal situation is the situation in which no disturbing factors are present, i.e., the situation in which the manifestation conditions obtain. The relevant systems, i.e., the falling body in the medium and the falling body in the vacuum, must have something in common. It is plausible to assume that there is something that is present in the less than ideal situation and would be present in the ideal situation. Salviati’s argument works because the system, the falling object, has some kind of feature, some kind of property, which is present in the less than ideal situation and would also be present if the ideal were realised. But, secondly, it clearly cannot be a categorical property. A categorical property is a property that is manifest whenever the system has the property. The behavior that Salviati attributes to the falling bodies (falling with a certain velocity) cannot be considered to be such a property because it is manifest in the vacuum only. The property is not always manifest. What we have is:

4

Galilei 1954, 71-2.

228

Andreas Hüttemann

Something carries over from the ideal to the less than ideal.5 It would be present and manifest under ideal circumstances. It is present but fails to be (completely) manifest in less than ideal circumstances. Properties of systems that become manifest under certain conditions only are dispositional properties as opposed to categorical properties that are manifest in all circumstances. Thus, given that the behavior of the falling body is a dispositional property, we understand how our knowledge of what is going on in the less than ideal situation carries over to the ideal situation. By contrast, Salviati’s argument would generate a puzzle, if law statements were understood as attributing categorical properties to physical systems. Thirdly, even though the falling bodies do have dispositions that are not (completely) manifest in the less than ideal situations these situations provide evidence for the ideal situations. That couldn’t be the case if manifestation were an all or nothing affair. The point is that the disposition is partially manifest in the less than ideal situation and this provides the basis for an extrapolation to the ideal situation. In this context it is helpful to introduce a distinction I have drawn elsewhere: Fragility is an example of a discontinuously manifestable disposition (DMD). A thing is either broken or it is not; fragility cannot be partially manifest. It is an all or nothing affair.6 Not all dispositions are discontinuously manifest. Continuously manifest dispositions (CMDs) allow for partial manifestations. If the partial manifestations are continuously ordered they allow for extrapolation. Take, for example, properties that we attribute to ideal crystals: conductivity, specific heat, etc. There are laws that attribute behavior to the crystal in case there are no impurities (disturbing factors). Even though what the laws attribute are dispositional properties, which – presumably – will never be completely manifest, we can nevertheless get empirical evidence for the disposition’s obtaining. We can order different samples of the crystal according to the degree to which the manifestation condition is realized. The fewer the impurities, the more the manifestation condition is realized. If we measure the quantities in question with respect to the different samples, we are able to extrapolate to the behavior of the pure system as the limiting case. The disposition is partially manifest in the non-ideal situation. The transition from the less than ideal to the ideal is continuous so as to allow for extrapolation.

5 6

This is essentially Cartwright’s argument for capacities (Cartwright 1989, 189). Apparently intuitions are divided on whether a thing can be partially broken or not. Anyway, for the purposes of this paper I am not committed to the view that there are examples for DMDs. All I am claiming is that the dispositions that law-statements ascribe to physical systems are CMDs.

Dispositions in Physics

229

The falling objects in a vacuum as discussed by Salviati provide another example of a CMD. The thinner the medium, the closer we approach the ideal condition, the more the behavior in question becomes manifest. Given such continuity we can accumulate evidence for what would happen if the ideal circumstances were realized even if they actually never are. The partial manifestations of the disposition allows for an extrapolation to the ideal situation under the assumption of continuity between partial and complete manifestation. I have used such notions as ‘partial manifestation’ and ‘continuity’. Prima facie these appear to be rather vague concepts. But as a matter of fact they can be made precise in physics. The central idea is this: A system in a less than ideal situation, i.e. in the presence of disturbing forces or other influences, is part of a compound system. The behavior of such a compound system can be explained in terms of what the parts contribute to the compound behavior. Thus, the system’s behavior is partially manifest if the system contributes to the behavior of the compound system of which it is a part and of which the disturbing factors are parts as well. These contributions can be made quantitatively precise. Let me illustrate this through a simple example. Take a compound massive system consisting of three subsystems. We are only interested in mass. Leaving out relativistic effects we know that the mass of the compound (M) adds up as follows: m1 + m2 + m3 = M.

Thus, what we have is a law of composition for our three masses. There are similar laws of composition for vectors (such as velocities and forces) or tensors. The law tells us how the different subsystems contribute to the mass (behavior) of the compound. It also tells us what would happen in the limiting case of, say, m1, approaching zero.7 In the case Salviati describes, we can consider the medium plus the falling body as a compound system and provide – at least in principle – a complete account of its behavior. There are laws of composition that tell us how the different contributions such as the disturbing factors affect the behavior of the compound system. Thus, a system in the presence of disturbing factors can be conceived of as a compound consisting of the system in question plus the other factors. The laws of composition will also tell us what is going to happen if the medium (a disturbing factor) is replaced by a thinner medium. So it is furthermore possible to make precise the notion of continuity between the less than ideal and the ideal. 7

More on laws of composition in Hüttemann 2004, chapter 3.

230

Andreas Hüttemann

So, if we have a complete description of all the contributing factors, it is the laws of composition on the basis of which we can give a quantitative account of partial manifestation and of continuity between the non-ideal and the ideal situation.

5. Solutions Those problems that are generated on the view that laws attribute categorical properties to physical systems are absent on the rival view that they attribute dispositional properties: •

How can laws that attribute properties to physical systems, which appear to have no instances, be tested?

The problem disappears if it is assumed that law statements attribute continually manifestable dispositional properties to physical systems. The properties in question are instantiated in the less than ideal situations though not (completely) manifest. But partial manifestation suffices for extrapolation and therefore for testing laws that are not manifest. They can be tested because their partial manifestation provides the basis for an extrapolation to what would happen in the ideal situation – to what would happen if the manifestation conditions were realized. •

How can laws that attribute properties to physical systems, which appear to have no instances, be explanatorily relevant for actual phenomena?

Even though the dispositional properties are not completely manifest, they are instantiated, i.e. the system has the dispositional property in question and in this sense the property is actual. So we explain something actual in terms of something that is actual. They are explanatorily relevant for actual phenomena because we can explain actual phenomena in terms of contributions of various factors (all of which are partially manifest dispositional properties which are instantiated by the systems in question). •

Why are we interested in laws that attribute properties to physical systems, which appear to have no instances? What’s the point of postulating these laws that describe what is going to happen when there are no forces – even though we know that these are always present?

We are interested in laws that attribute dispositional behavior to physical systems because that allows us to explain the behavior of compound systems in

Dispositions in Physics

231

terms of contributions. It enables us to explain the behavior of compound systems in terms of the behavior of the parts.

6. Rejoinder by the Categoricist I argued that understanding law-statements as ascriptions of dispositional properties of a certain kind makes rational how we test laws and how we explain the behavior of compound systems in terms of contributions of the parts (micro-explanation). But maybe the categoricist has an explanation as well. According to the categoricist (or the ‘humean’ metaphysician) Salviati is merely confronted with certain laws that can be understood as attributing categorical properties to physical systems. One law concerning gold in quicksilver, another one concerning wood in quicksilver, one law concerning gold in air and another one concerning wood in air. These various regularities can be understood as attributing behavior to certain compound systems such as gold in quicksilver or wood in the air. The behavior in question is manifest (gold actually falls with a determinate velocity and acceleration in quicksilver). We do not have to appeal to non-manifest dispositional behavior. The Humean or categoricist can provide a consistent description of all that happens. In reply it has to be admitted that the categoricist can indeed provide a consistent and complete account of all that happens (of all events). The question, however, is: What do we expect of a metaphysical theory? Its job is not just to give a complete and consistent description of events. Even a naïve regularity account of laws could achieve that. The problem with a naïve regularity view of laws of nature was not that it failed on this count. The problem was that it was unable to give an account of the role laws of nature play in scientific practice, such as in explanation, confirmation and induction.8 Similarly, I want to argue that the categoricist cannot explain the role of laws of nature in scientific practice. More particularly, two things need to be mentioned in this context: •

•

8

On the categoricist’s view it remains unclear why the cases that Salviati mentions have anything in common. What is it that connects the different laws? Why can Salviati legitimately infer from the observed phenomena to the ideal case? The law that Salviati is interested in and is testing concerns falling bodies in a vacuum, not in a medium. Similarly we are very often interested in Armstrong 1983, 39-59.

232

Andreas Hüttemann

ideal crystals etc. (Thus we try to shield off external influences in experiments.) On the categoricist’s view it remains unclear why we are interested in these ideal cases. These problems stay with the categoricist even if higher order laws are introduced.9 These are meant to systematize the empirical laws I have mentioned. We can think of Newton’s second law or the Schrödinger-equation as higher order laws that allow for lower-level laws with missing values. Thus Newton’s second law tells us (given the right force-functions) how bodies gravitate towards one another and thus how bodies fall. Such higher order laws would allow for missing values. Newton’s second law tells us what would happen, if certain forces, that are in fact not actual, were to obtain. It also tells us what would happen if no forces were present (the ideal case I mentioned). Even if we assume that we introduced these higher order laws for reasons of simplicity and systematization, the problems of the categoricist remain. The account still does not explain why we are in particular interested in the ideal cases. Why is Salviati interested in the vacuum case rather than in the other cases? Why do ideal cases serve as a basis for explanations of compounds’ behavior? The categoricists or Humeans may be able to give a complete and consistent account of all that happens, however, they cannot give a rationale for the role laws of nature play in scientific practice.

7. The Explanatory Role of Dispositions I will now turn to the explanatory role of dispositional properties that laws of nature ascribe to physical systems. Two aspects can be distinguished. First, there is the argument for dispositions that has just been outlined. The assumption that physical systems have CMDs is the best explanation for the success of part-whole-explanations (micro-explanations) and for extrapolations to ideal cases. Second, we have to look at the role of dispositions within such explanations. In an explanation, in which we appeal to CMDs, the explanandum is usually not the manifestation of the dispositions in question. In fact, it is usually the case that these dispositions fail to be perfectly realized. Bodies usually fail to continue in their state of rest or of uniform motion in a straight line, because disturbing forces are present. In these cases the alleged explanandum, the manifestation does not even exist. But even if we would come across a case in which a body manifests the disposition that Newton’s first law men9

Armstrong 1983, 111-116.

Dispositions in Physics

233

tions, we would not explain the fact that the body continues in its state in terms of the disposition. The fact that a body continues in its state of rest or of uniform motion in a straight line in the absence of disturbing factors is a brute fact and remains a brute fact, even with the assumption of an underlying dispositional property. So, I will argue for the claim that CMDs are not introduced as causes of their manifestations, but rather as contributors to the behavior of compound systems. (That makes CMDs different from Cartwright’s capacities.) In an explanation, in which we appeal to CMDs, the explanandum is usually the behavior of some compound system (the falling stone in the medium, the dissolved sugar in the water, the moving body in an external force field) and we appeal to the dispositional non-manifest behavior of the parts. If we explain the behavior of a compound system in terms of parts we want to attribute certain properties to the parts in terms of which the behavior or property of the compound can be explained. We want to know what the different parts contribute. Thus, my claim is that in physics we appeal to dispositions as contributors to the behavior of compounds – not as causes. But aren’t the dispositions causally relevant for the manifestations of the compound’s behavior? Clearly the dispositions are relevant for the compound’s behavior, but they are not causally relevant. The masses of the parts contribute to the mass of the compound – they don’t cause the compound to have a certain mass. According to the physical part-whole relation the behavior of the parts and the laws of composition determine the behavior of the compound. The compound’s behavior is thus counterfactually dependent on the behavior of the parts. This, however, does not imply that the relation is causal. If one assumes that causal relations are temporal, there is an easy way to see this. The part-whole relation is a synchronic relation, not a temporal relation. Therefore, the part-whole relation and thus the relation of the dispositions of the parts to the behavior of the compound cannot be a causal relation. It is due to the part-whole relation that we have to introduce dispositions in physics. The part-whole relation in physics is a non-causal relation. Therefore, dispositions in physics should not be considered as causes of their manifestations.10

10

In Hüttemann 2004, 83ff. I argue in more detail why the part whole relation is non-causal.

234

Andreas Hüttemann

8. The Irreducibility of Dispositional Properties to Categorical Properties In this final section, I will turn to the third claim I advertised at the outset. Dispositional properties cannot be reduced in any straightforward sense to non-dispositional properties. For a start, let me recapitulate the distinction between dispositional and categorical properties: A dispositional property is a property that, if instantiated by an object, is manifest under specific conditions only. A categorical property by contrast is a property that, if instantiated by an object, is manifest under all conditions. What would a reduction of dispositional property look like? Armstrong argues as follows: A good model for the identity of brittleness with a certain microstructure of a brittle thing is the identity of genes with (sections of) DNA-molecules. Genes are, by definitions those entities, which play the primary causal role in the transmission and reproduction of hereditary characteristics. […] in fact sections of DNA play that role. So genes are (identical with) sections of DNA. (Armstrong 1996, 39)

This model of reduction is standard in the philosophy of mind and often called the functional model of reduction. It can easily be transferred to the case of dispositional and categorical properties. Slightly modifying Stephen Mumford’s terminology we have the following argument: The argument from the identity of functional role: 1. Disposition D = the occupant of role R. 2. Categorical Base C = the occupant of role R. Therefore: Disposition D = categorical base C. (Mumford 1998, 146)

Let us, for instance, assume that fragility can be functionally characterized as the property in virtue of which an object breaks, when suitably dropped. As it turns out the microstructure of the object is responsible for the breaking of the object if it had been suitably dropped. So fragility is identical with some micro-based property of the object in question. It turns out that even though we have two predicates, namely fragility on the one hand and the microdescription on the other, there is only one property that both refer to. I think an argument of this kind is a good argument for the identification of fragility or solubility with some kind of micro-based property. In this sense I have no objection to the reductionist claim. It is important to realize that this reductionist claim does not imply that the micro-structural property in question is categorical. Such a claim would require an additional argument. In fact it is hard to see how the micro-based property in question could be categorical. If it were a categorical property there would not have been a reduction in the first place. Fragility’s functional

Dispositions in Physics

235

role is the role of a dispositional property: to break if suitably dropped, i.e. to break given the appropriate circumstances. If the micro-based property realizes this role, then it is in virtue of the micro-based property that the object breaks, given the very same circumstances. If the property to be reduced is one that manifests a certain behavior under specific conditions then the reducing property, i.e. the one that is the occupant of the reducing properties functional role, must manifest the behavior in question under the very same circumstances. The micro-based property can be an occupant of the disposition’s functional role only if it a dispositional property itself. For the reduction to work it has to be a reduction of a dispositional property to a dispositional property. It might be objected that what Armstrong and others have in mind is reduction to categorical properties plus laws.11 But as we have seen, an appeal to laws is an appeal to dispositions. So this move is of no help for the categoricist. Second Objection: Couldn’t the defender of reduction to categorical properties argue that in such a reduction appeal is typically made to mass, shape etc, i.e., to categorical properties. I agree that in such reductions one might appeal to properties such as the shape or the mass of a subsystem, which are probably the best candidates for categorical properties. But shape or mass on its own do not explain why a fragile glass breaks when suitably dropped. It is in virtue of the fact that the shape and mass (of the constituents of the fragile body) feature in laws that they are appealed to in the reduction of fragility and solubility. It is in virtue of the laws that we understand why the micro-property fills the relevant functional role. As I argued before, these laws are best understood as attributions of dispositions to physical systems. So the appeal to mass and shape does not undermine the claim that the reducing micro-structural property has to be a dispositional property. The standard examples for dispositions in the relevant literature are solubility and fragility. It seems obvious that these need to be reduced to proper physical properties if there is a place for them in a physical description of the world. It is, however, important to separate two issues. One may very well agree that the above-mentioned dispositions stand in need of reduction qua being macro-properties that need to be connected to micro-physics. It is a separate issue whether properties need to be reduced to categorical properties because they are dispositional. If something is a macro-property of a compound system it might be reasonable to expect a micro-explanation (microreduction). As we have seen there is no reason to assume that dispositional properties as such should be reducible to categorical properties.

11

Armstrong 1996, 41.

236

Andreas Hüttemann

There seems to be a widespread assumption that on a fundamental level there can only be categorical properties. But this is most likely not the case. Newton’s first law: Every body continues in its state of rest or of uniform motion in a straight line, unless it is compelled to change that state by forces impressed upon it.

provides a perfect example of a law that attributes a dispositional property to physical systems. Nevertheless it cannot be reduced to anything more fundamental. It has been argued that irreducible fundamental dispositions are incoherent. Simon Blackburn considers the case that science finds dispositional properties all the way down. The problem he sees is this: The problem is very clear if we use a possible world analysis of counterfactuals. To conceive of all truths about the world as dispositional, is to suppose that a world is entirely described by what is true at neighbouring worlds. (Blackburn 1990, 64)

On our conception of dispositions it would be wrong to claim that our world is described in terms of what is true in other worlds. Newton’s first law, for example, is true of systems in our world because these systems have the relevant disposition in our world. Furthermore we have epistemic access to the non-manifest dispositions due to their partial manifestation. If the dispositions in question are continually manifestable dispositions the problem Blackburn envisages does not arise.

9. Conclusion In this paper I have argued for the following claims: Contrary to what is often supposed, dispositions do play a significant role in understanding the scientific practice that has been established since the 17th century. I argued that given our interest in ideal cases, in what would happen if no disturbing factors were present, law-statements should be understood as attributing continually manifestable dispositional properties. Second, the dispositions I am talking about should not be conceived as causes of their manifestations but rather as contributors. And finally, I argued dispositional properties cannot be reduced in any straightforward sense to categorical properties.

Literature Armstrong, David M. 1983. What is a Law of Nature? Cambridge: Cambridge University Press. Armstrong, David M., Martin C. B. and Place, U. T. 1996. Dispositions – A Debate, ed. by T. Crane. London: Routledge.

Dispositions in Physics

237

Blackburn, Simon. 1990. “Filling in Space.” Analysis 50: 62-65. Cartwright, Nancy. 1989. Nature’s Capacities and their Measurement. Oxford: Clarendon Press. Descartes, Rene. 1991. Philosophical Writings, Vol. III, ed. J. Cottingham, R. Stroothoff, D. Murdoch and A. Kenny. Cambridge: Cambridge University Press. Galilei, Galileo. 1954. Dialogues Concerning Two New Sciences, transl. H. Crew and A. de Salvio. New York: Dover Publ. Inc. Hüttemann, Andreas. 2004. What’s wrong with Microphysicalism? London: Routledge. Mumford, Stephen. 1998. Dispositions. Oxford. Newton, Isaac. 1999. The Principia, transl. I. B. Cohen and A. Whitman. Berkeley.

The Role of Dispositions in Historical Explanations ROBERT SCHNEPF

1. The epistemological and ontological problems of dispositions are well known: It is not possible to observe them. Knowledge of them can only be attained by observing their manifestations. But an observable categorical property can be the result of different underlying dispositions. Moreover, it often is uncertain if the object in question has acquired the underlying dispositional property some time before the corresponding manifestation or at the very moment of its realization. Decisive criteria are mostly not at hand. So it seems difficult to verify dispositional explanations and it is nearly impossible to answer pertinent questions concerning the acquisition and the loss of dispositions. Last but not least, the ontological status of dispositions seems to be unclear. It is not only debated if disposition terms can be interpreted as mere theoretical terms along the lines once proposed by Rudolf Carnap in his Testability and Meaning,1 but it is also an open question, if – assumed the reality of dispositions in our ontology – dispositional properties can be reduced in some way or other on categorical properties. Most, if not all of these problems, are related to the simple fact that we assume certain things to possess a dispositional property even if there is no manifestation of the corresponding categorical property. A piece of salt is assumed to be water-soluble even if it never was put into water. Nevertheless dispositional explanations are commonly used not only in the natural sciences and everyday life but also in the historical sciences. Aggressive behavior is explained by an aggressive character, Chamberlain’s politics in light of his “weak character” (compared with Churchill’s). According to a widespread theory of explanation something is explained if it is shown that it was to be

_____________ 1

Carnap 1936, 1937; Carrier 2007, 58 et seqq. Kamlah 2002, 253, who develops further the theories of Kaila 1945 and Lewis 1979 and 1997, surprisingly conclude that disposition terms are in need of a “realistic” interpretation. In what follows I will assume that there are no definitions of disposition terms in the strict extensionalist sense. From the point of view of a historian who is in search of explanations, there is – so I will argue – no sense in assuming or searching for a strict definition in this sense for the terms that he uses.

The Role of Dispositions in Historical Explanations

239

expected.2 Certainly, if salt is water-soluble than it is to be expected that this piece of salt dissolves if put into water. But if the solution of salt in water is explained by the corresponding disposition of the salt’s solubility, this seems rather odd. For the question of why the salt dissolves in water is not sufficiently answered by pointing to salt’s water-solubility. The desire for explanation will come to an end only if it is explained why salt is water-soluble. The same would hold for an explanation of Chamberlain’s politics in terms of his “weak character”: Chamberlain’s politics was to be expected, given his character and the circumstances of his political decisions. In the historical sciences, dispositional explanations are widespread. For example: A certain behavior, single acts, or a way of reacting are explained by referring to a character trait or a mentality. Cicero explained the well known clementia caesaris in terms of Caesar’s clemency. The cruel decisions of Philip II of Spain are explained by his cruel temperament. Well known is also Max Weber’s analysis of the capitalist’s spirit in order to explain why people behave in a capitalistic manner. Some historians explain the cruelty of German soldiers in the World War II in terms of a murderer’s mentality that they acquired before the war. A person can possess such a mentality even if he does not find himself in situations that trigger a corresponding manifestation. But if the manifestation takes place it can be explained by referring to the disposition. If Caesar’s character contains clemency his merciful treatment of prisoners of war during the civil war would have been to be expected. And the cruel character of Philip II of Spain justified the nervous expectations of the Dutch people. Nevertheless, like the explanation for the salts dissolution in water by its water-solubility, the explanation of the clementia caesaris in light of Caesar’s clemency and of Philip’s cruel decisions in terms of his cruelty demands for further explanation. Some might challenge the above explanation of Caesar’s acts by arguing that they were part of a calculated political strategy while others – accepting the explanation – might try to further explain Caesar’s clemency pointing to his upbringing. John W. N. Watkins maintained that “the ultimate premises of social science are human dispositions, i. e. something familiar to us. They ‘are so much the stuff of our everyday experience that they have only to be stated to be recognized as obvious’”. In his eyes dispositions “provide social sciences with a natural stopping-place in the search for explanation”. The historian explains “the familiar in terms of the familiar”.3 In my paper I want to challenge this position. I will argue, that while dispositional explanations are familiar in the historical science and everyday life, they are nonetheless epistemically dubious

_____________ 2

3

Hempel 1970, 336. In my view showing that the explanandum is or was to be expected, given certain circumstances, is sufficient for rendering it an adequate explanation – Schnepf 2006, 87 et seqq. J. W. N. Watkins 1952, 32-33, citing L. Robins 1935, 79.

240

Robert Schnepf

and often worthless, if not taken exclusively as a first step in the search for better explanations (and in this function, they might be sometimes indispensable): • In the historical sciences as in everyday life dispositions should be ascribed with caution and methodological reflection; it is familiar to us that ascribing a disposition evokes grave problems and can not be done beyond doubt. We discuss if our friend is timid as we discuss if Philip II is cruel. No interesting disposition ascription is beyond reasonable doubt; different interpretations of the same occurrences are always possible. • There seems to be a difference between explaining the clementia caesaris (his concrete acts against the prisoners of war) by referring to Caesar’s clemency or by referring to his general political strategy. While the first explanation seems to be odd, the second seems to be informative and instructive. The disposition term contained in the explanation most often only repeats the explanandum’s description in another way. • It seems obvious that dispositions can be made a proper object of further explanations in everyday life, in natural sciences and in historical sciences. Reference to a certain disposition does not stop the process of asking for explanation. So we may ask why our friend is timid or why Philip II was cruel. I want to argue that the explanation of a disposition can often be given without referring to further dispositions assumed to be more basic.4 Dispositions may play a crucial role in searching and forming historical explanations without being a necessary part of the ultimate explanation reached at the end of the scientist’s efforts. Dispositions can do this job if they are informative and not repetitive. Informative dispositional explanations in historiography are the first step to discover non-dispositional causes which are beyond the scope of the explanandum’s first description and to develop more interesting explanations which opens room for fruitful further questions. For this purpose dispositional predicates do not have to be interpreted in a realistic manner. In my essay, I want to show that the ontological questions are of little interest for the actual job of historians and that even a realistic way of interpreting disposition terms solves none of the methodological problems in the historical sciences. On the contrary, the methodological and epistemological problems can be understood and analyzed better, if we work along the lines originally drawn by Carnap – enriched with the tools of condition analy-

_____________ 4

This does not imply that dispositions are in the end always reducible to categorical properties; but it does imply that in the analysis of dispositions, categorical properties play a indispensable role.

The Role of Dispositions in Historical Explanations

241

sis, nomological knowledge5 and possible worlds semantics by Kaila and others.6 Nomological knowledge seems to be a necessary condition for solving the problem of unobserved non-manifested dispositions by appeal to other more or less observable properties (for example the molecular structure of salt for its solubility in water or the letters of Philip II for his cruelty in religious matters). Nomological knowledge in this sense makes it possible to connect disposition terms with other parts of the theory (or sketch of theory) in order to explain certain occurrences.

2. In what follows, I want to argue for my position – and to dispute Watkins’ (and others’) claims – by analyzing the epistemological and methodological problems posed by a very special kind of dispositions in history – namely by mentalities. For that purpose, I will analyze some examples from historical writings. First, however, it needs to be shown that mentalities are correctly understood as dispositions. Unfortunately, historians use different concepts of mentality.7 But they all have certain features in common: Mentalities neither can be directly observed nor can they be experienced, nevertheless they are used in historical explanations. Ascribing a mentality allows to predict or explain certain observable behavior with a certain degree of plausibility. It allows us to understand a specific act as an observable manifestation of an underlying disposition. Accordingly, mentalities can be the proper object of further explanatory effort. They can play a role as part of theories or theory fragments (for example in psychology or sociology but even in everyday folk theories). Therefore, mentalities can be conceived of as dispositions. Appeal to such mentalities can play a necessary role in the search of historical explanations without necessarily being part of the ultimate explanation asked for. Nevertheless, as I will argue, ascribing a certain mentality to a person, who lived centuries ago, poses grave – in the end unsolvable – methodological problems.8 In this context, I will take it for granted what a great theorist of historiography once pointed out in his drastic manner: We cannot feel what it

_____________ 5

6 7 8

By “nomological knowledge” I only mean knowledge of more or less well confirmed hypotheses about observable regularities. As far as I can see this is the same as Max Weber’s conception of nomological knowledge – and this might be sufficient for the context of historiography, notwithstanding the extensive debate about the nature of natural laws. Kamlah 2002, 271 et seqq. Graus 1987, Gilcher-Holthey 1998 – for further discussion Schnepf (forthcoming 2010). In fact, even the ascription of mentalities to our contemporaries, even to our neighbors and friends, poses these grave methodological problems.

242

Robert Schnepf

is like to be an ancient Greek as we cannot know what it is like to be a dog.9 In the same way we can not feel what it is like to be Caesar or Philip II of Spain. We do not have any direct epistemic access to the mentality of other persons by empathy or sympathy or something like that as there is no direct epistemic access to the feelings or thoughts of a dog. This holds in history as in every-day life. Fortunately one must not be Caesar to understand Caesar and some historians made a good job in understanding and explaining his behavior and acts – obviously without being Caesar. In what follows, I combine methodological and epistemological reflections with observations on how some historians have worked in order to get some information about the use and the value of dispositional concepts in historical explanations. A further restriction of my argument is necessary, since the expression “mentality” is very broad and different schools of historiography understand this concept very differently. Since the beginning of the 20th Century a special branch of history called the historiography of mentalities has been developed by Marc Bloch, Ferdinand Braudel, Philippe Aries and others who worked together in the famous historiographic journal Annales. They use the expression “mentality” to refer to different ways of perceiving the world, different ways of thinking, having convictions, different styles of living and so on as it is exemplified in phrases such as “the mentality of a certain person”, “the mentality of the working class in England before the 1914”, and the “folk mentality of the middle ages”. Some authors take mentalities to be collective entities being invariant during long periods of time (Braudels “long durée”), whereas others take mentalities to be able of change within a couple of years. Notwithstanding this broad and often confusing use of the term “mentality,” the historians know quite well how to work with it. Fortunately the conception of a “History of Mentalities” was developed by integrating sociological concepts into the new approach. In his methodological writings, Braudel refers often to Werner Sombart and sometimes to Max Weber.10 Weber tried to reconstruct a very special type of mentality (the Calvinist mentality) in order to explain a very special kind of behavior (namely the inner worldly ascetism of the early capitalists) without claiming to have direct epistemic access to the individual minds of the historical agents.11 It is Weber who cited the aforementioned dictum “One need not to be Caesar to understand Caesar” in one of his methodological essays.12 In the following I will be only concerned with mentalities of individual persons and not with dispositions of collectives, structures or institutions.

_____________ 9 10 11 12

G. W. F. Hegel 1970, 552. Cf. Braudel 1992a and 1992b. M. Weber 1988 – for a discussion of Weber’s approach compare Lehmann/Roth 1987. Cf. M. Weber, WL p. 428.

The Role of Dispositions in Historical Explanations

243

Consequently, I will focus on Weber’s and Sombart’s explanatory use of concepts of “Geist” (spirit, mentality); in contrast to Lujo Brentano’s explanations of the same historical phenomena.13 For Sombart, the spirit of capitalism is a special mental disposition that causes the specific way of acting in capitalistic societies. To be sure, in order to fully explain this behavior other more common dispositions are necessary as well (for example, prudence or the rationality of the individual persons). But to explain the special way of economic behavior in capitalism additional mental dispositions have to be appealed to according to his view.14 Weber’s explanation of the rise of modern capitalism by reference to the protestant spirit will be analyzed in opposition to Sombart’s and Brentano’s approaches. My analysis of this debate will support the thesis that the ascription of dispositions in historiography is nothing beyond reasonable doubt and that it requires careful methodological considerations. In what follows I will restrict my analysis of dispositions in historiography to this example. To illustrate my points I will particularly make use of Max Weber’s methodological reflections. In this context another assumption should be mentioned. Max Weber is committed to methodological individualism. This is evident in Max Weber’s essay on the categories of sociology (“Über einige Kategorien der verstehenden Soziologie”)15 and in Economy and Society (“Wirtschaft und Gesellschaft”).16 As John W. N. Watkins puts it: “According to this principle, the ultimate constituents of the social world are individual people who act more or less appropriately in the light of their dispositions and understanding of their situation. Every complex social situation, institution, or event is the result of a particular configuration of individuals, their dispositions, situations, beliefs, and physical resources and environment.”17 Similarly, Sombart’s explication of the term “economical system” illustrate the same methodological commitment: “Under this term I understand a peculiarly ordered form of economical activity, a particular organization of economical life within which a particular mental attitude predominates and a particular technique is ap-

_____________ 13 14 15

16 17

Cf. L. Brentano 1923. Cf. Sombart 1916, 2 et seq. Cf. M. Weber 1985, 439: “Das Ziel der Betrachtung: „Verstehen“, ist schließlich auch der Grund, weshalb die verstehene Soziologie (in unserem Sinne) das Einzelindividuum und sein Handeln als unterste, als ihr „Atom“ – wenn der an sich bedenkliche Vergleich hier einmal erlaubt ist – behandelt. (…) Begriffe wie „Staat“, „Genossenschaft“, „Feudalismus“ und ähnliche bezeichnen für die Soziologie, allgemein gesagt, Kategorien für bestimmte Arten menschlichen Zusammenhandelns, und es ist also ihre Aufgabe, sie auf „verständliches“ Handeln, und das heißt ausnahmslos: auf Handeln der beteiligten Einzelmenschen, zu reduzieren.” Cf. M. Weber 1976, 6: “Handeln im Sinn sinnhaft verständlicher Orientierung des eigenen Verhaltens gibt es für uns nur als Verhalten von einer oder mehreren einzelnen Personen.” J. W. N. Watkins 1957, 105 et seq.; methodological individualism is challenged, for example, by Dray 1990, 47 et seqq.

244

Robert Schnepf

plied.”18 At first sight, methodological individualism seems to minimize ontological commitments. We seem to have epistemic access to the intentions of individual actors, whereas we are not willing to attribute “intentions” to collectives, organizations, systems, or societies as wholes (and if we do so, we do it only in a metaphorical sense). Watkins agrees with Weber that all historical descriptions and explanations containing reference to supra-individual entities should and could be reformulated as descriptions and explanations containing only individual persons and their intentions. Here, I do not want to question whether this kind of reductionism is in principle possible but take methodological individualism as a clear and plausible methodological advice.19 But while Watkins combines the principle of methodological individualism with a theory of dispositional explanations in history according to which dispositions are explanation stoppers, I will argue – based on a closer analysis of the debates between Max Weber, Lujo Brentano, and Werner Sombart – that attributions of dispositional properties should be treated only as first steps in the search and defense of historical explanations, even if one wants to hold onto the principle of methodological individualism.

3. Let’s start with a closer look at Weber’s famous Protestantism thesis: The object of explanation is the rise and growth of capitalism. Given methodological individualism, this way of putting the question has to be regarded as a mere abbreviation. “Capitalism” denotes nothing more than the fact that people behave in a very special manner. Accordingly, the rise of capitalism has to be understood as a change of the behavior of individuals. For example: Instead of consuming earned money instantly or amass hidden treasures, money now is reinvested into economic ventures. New ways of banking correspond to the needs of these new kinds of behavior. According to methodological individualism, the rise of capitalism is nothing more than the sum of changes in the behavior of individuals. To explain the rise and growth of capitalism would then be to explain this change of individual behavior.20

_____________ 18 19

20

W. Sombart 1916 I, 21 et seq. – translation by T. Parsons. In fact, intentions, volitions, and even the will itself can be interpreted as dispositions and consequently as mere theoretical terms. Weber’s methodological approach to explain an individual’s behavior using ideal types provides some support for my thesis. To solve epistemological problems, we should not trust our evidences and prejudices of our every day life as manifest in our natural language – I argue for this point at length in Schnepf (2010, forthcoming) using arguments of Martin 1977 against his intentions. Cf. M. Weber 1988, 4 et seqq.

The Role of Dispositions in Historical Explanations

245

For Weber, this change in individual behavior can at least partly be explained by a change in an ethics of economic behavior. The rise of Calvinism changed an agent’s inner convictions so that a different way of dealing with economic matters became rational and appropriate. Immanent ascetism and success in economic matters now counted as a sign of being elected by God. Change of religious beliefs seems thus to be an explanation of the necessary change of spirit and behavior to account for a change of observed economic behavior. To sum up: The single acts constituting the capitalistic economic system are explained by agents’ new spirit or mentality, that is, by a certain disposition to act. This disposition in turn is explained in light of their religious convictions (which might by itself be regarded as a special kind of disposition). The change of religious convictions explains in this manner the change of the dispositions to act at the beginning of the capitalistic epoch. However it should be stressed that this explanatory scheme cannot be generalized: It is not Weber’s aim to explain the whole of modern capitalism but only a very special form of it at its very beginning. Moreover, the differentiation between private household and office, new ways of banking and technical developments are necessary conditions for the rise of capitalism. Weber does not think that a Calvinist’s ethics of economic behavior is the only relevant factor in the rise of capitalism but that it is the decisive one. And as he emphasizes numerous times, the need for explanation does not come to an end with his analysis, since Weber insists that it is still an open question of why religious believes have changed. For Weber, his explanation of early capitalism is only a first step in the search for a more global theory. Nevertheless, Weber tries to reconstruct one side of a complex causal history by describing or reconstructing a very special disposition that accounts for the selfrationalization necessary for a style of life appropriate to early modern capitalism.21 Three methodological problems can be brought up at this point: First, it is clear that different descriptions of the change from pre-capitalistic to capitalistic societies are possible. Even the single acts of the individual persons can be described in different ways. For example, the economist and friend of Weber Lujo Brentano denied that there is a principal change of behavior.22 According to his view, individual persons try always and under all circumstances to acquire the greatest amount of goods that is possible. What has changed in capitalist societies, compared to earlier ones, however, are the means by and the circumstances under which the individual persons try to achieve their constant goals. To Brentano, it is something like a natural im-

_____________ 21

22

Cf. M. Weber 1988, 12: “Denn wie von rationaler Technik und rationalem Recht, so ist der ökonomische Rationalismus in seiner Entstehung auch von der Fähigkeit und Disposition der Menschen zu bestimmten Arten praktisch-rationaler Lebensführung überhaupt abhängig.” Cf. L. Brentano, Puritanismus und Kapitalismus, in: Brentano 1923, 363-425.

246

Robert Schnepf

pulse (“Naturtrieb”) that is invariant during the whole history and that lies at the bottom of every individual behavior. During medieval ages this impulse was moderated by a Catholic ethics of economy (which was contemptuous of the acquisition of treasures and luxury goods and didn’t allow charging interest for loans), while in modern times political theories like Machiavelism and new ethics like a Calvinist ethics set this impulse free. Brentano – as a good Aristotelian – would deny that any change of basic human dispositions occurred. To explain the raise of capitalism does not require explaining a change in dispositions but it demands an explanation for why the same dispositions yielded different manifestations in different historical circumstances.23 Werner Sombart, the other great historian of capitalism, on the other hand agreed with Weber that the rise of capitalism implied a change in the basic dispositions to act. Yet he construed the basic dispositions differently than Weber. Sombart holds that a pre-capitalist economy is governed by the principle of self-sufficiency and that an individual person’s aim consisted in receiving only the support necessary for retaining his social status (“Bedarfsdeckungsprinzip”). This principle has to be strictly distinguished from the capitalist principle of unlimited acquisition (“Erwerbsprinzip”).24 Whereas Weber holds that the principle of unlimited acquisition might have governed even ancient and medieval economic activities and that the only but decisive change in behavior can be attributed to a new combination of religious beliefs and new techniques of self-control; Sombart assumes a much stronger difference between the two economic systems. For Sombart, explaining the rise of capitalism does not mean to explain the new way of rationalization given a constant impulse to acquire goods. He explains the change in economic systems in light of a fundamental change in an individual’s basic dispositions; from a disposition to conserve one’s own status to a disposition to act according to the principle of unlimited acquisition. Weber and Brentano share the assumption of a common basic disposition to acquire goods. Brentano wants to explain modern capitalism by different factors with the result of setting this impulse free. Weber holds that this impulse in itself is by no means specific to modern capitalism but that a new form of its moderation by permanent self-control due to new religious beliefs and practices is at the root of the whole drama. To speak more technically: Weber does not assume a change of the common basic disposition but the acquisition of a new one to moderate and control the effects of the first. Sombart opposes both Weber and Brentano by assuming a change in those dispositions that are adequate to understand the

_____________ 23 24

So the basic disposition assumed by Brentano can be interpreted as “finkish” – as most dispositions attributed by historians. Cf. Sombart 1916, 11 et sqq.

The Role of Dispositions in Historical Explanations

247

economic behavior and by denying a constant disposition to unlimited acquisition. So Brentano, Sombart, and Weber give different descriptions of the change to be explained and consequently assume different dispositions in order to explain the so described change. To Weber and Sombart, the explanation entails a change of dispositions possessed by individual persons while Brentano does not assume a change in disposition but only in the circumstances and techniques of their actualization. Taking into account that Weber, Brentano, and Sombart are dealing more or less with identical historical sources, I take these differences to strongly support my first point, namely, that the ascription of dispositions in history or historiography is nothing beyond reasonable epistemic doubt and that appealing to dispositions in historical explanation can not mean to explain the familiar with the familiar in a familiar way. The ascription of dispositions seems to be dependent on the theory presupposed in order to develop an explanation. Even the interpretation of historical sources in the historical sciences is guided by the assumption of basic dispositions of individual agents, which are in no way proved or provable. They have the status of more or less well confirmed hypotheses. There is thus no such thing as a simple and innocent description in historiography. A second point must be mentioned. At first glance, the three different explanations or sketches of explanations of the rise of capitalism seem to violate the principle of methodological individualism. The three presumed dispositions are described only abstractly and there are no individual persons analyzed by the three historians. Weber starts with an analysis of the Calvinist economic ethics and ends with showing why self-moderation and methodological self-control demanded by a Calvinist ethics fits with a capitalist’s life style. To be sure, there is some discussion of individual persons. For example, Benjamin Franklin is discussed but only in order to reconstruct his ideas and not his individual dispositions that are responsible for his actual behavior. The concrete dispositions of single persons are not within the scope of Weber’s research. Weber deals with the problem of explaining the behavior only on the level of so called “ideal types.” Yet Weber’s remarks on ideal types are not at all clear. One forms the concept of an ideal type – so he seems to assume – by abstracting from certain features of their impure realizations in individual cases and by combining them into one coherent concept, such as the concept of the “capitalist spirit”. For an ideal type to be a fruitful concept in research demands in no way that any individual exists, who performs according to the features combined in this concept. On the contrary, observing the deviance from the ideal type opens up the possibility for a closer description of the single case and the search for the causes of the observed deviance. Ideal types thus do not establish a direct access to the individuals’ disposition. For that reason, Weber’s explanation of the origin of capitalism in terms of an

248

Robert Schnepf

ideal type does not involve an analysis of the individuals’ behavior and its real causes. Rather, the explanation is “accomplished” on a very abstract level. On the one hand, this seems to be a grave violation of the principle of methodological individualism. On the other hand, the gap between abstract ideal-typical descriptions and individual persons can not be bridged.25 This problem is present in Brentano and Sombart as well. In some sense we may agree with Brentano what it means to desire goods or to have a natural inclination toward well-being and luxury. But we do not know exactly in what way this abstract concept characterizes the individual disposition of an individual person. The same holds for Sombart’s conception of the change of disposition responsible for the origin of capitalism. We can reconstruct what it is to have a disposition underlying the rationalized way of life on a general level. But we do not know exactly which disposition or psychological causes actually underlie the actions of an individual person. Weber is quite explicit about this point: The behavior explained on the level of ideal types can be the result of different psychological characteristics in concrete individual persons. Even on an abstract level the use of disposition terms, conceived of as ideal types, leaves room for controversy. For example: Between Brentano, Sombart and Weber there is some debate on how to interpret and explain the theories and the life of Benjamin Franklin. Weber uses some utterances of Franklin to support his conception of the capitalist spirit while Brentano not only challenges the correctness of Weber’s interpretation of Franklin’s writings but also their impact on his life. This does not necessarily falsify Weber’s attribution of a corresponding disposition to Franklin. Franklin may have possessed the disposition in question but may not have actualized it for additional reasons. Again, Weber is quite explicit about this consequence of his way of thinking: The correct ascription of a disposition like capitalist spirit does not imply that any individual person acts in a corresponding manner. Using more or less familiar disposition terms in historical explanations is therefore not sufficient for fulfilling the promises of methodological individualism. If we want to combine dispositional explanations with methodological individualism we have to accept that those dispositions play only a heuristic role and that dispositional explanations can be only first steps in searching for adequate explanations. A third point to be mentioned is that the ascriptions of dispositions in history are in need of a method of verification or confirmation. To be more precise, two different problems of verification should be distinguished. First we have to decide which of the three accounts, that is, Brentano, Sombart or Weber’s versions of dispositional explanation for the origin of capitalism, is correct. This question is a question on the level of ideal types. Yet even if we

_____________ 25

I take this to be a grave problem for all theories of methodological individualism.

The Role of Dispositions in Historical Explanations

249

decided among these different general approaches to the problem, we still have to confront the question of whether we can attribute those ideal typical dispositions to an individual person. That is to say, we have to ask whether the assumed disposition is causally efficacious on the level of an individual’s mind. For example, if we accept Weber’s approach and start to analyze Benjamin Franklin, we may attribute a disposition of methodic self-control to him (supposedly acquired by accepting certain religious beliefs) but we may deny that religious beliefs were relevant in the case of Franklin. So we have to search the causal factors that are in his case relevant for explaining why he has the ascribed disposition. If we accept that dispositional concepts used in history are constructed on a very general level, which might be described as the level of ideal types, we have to accept that they play a mere heuristic role and that we cannot interpret them in a realistic manner. We have no direct epistemic access to the thought and motives of Benjamin Franklin and there are no experimenta crucis to falsify the general approach or the ascription of a certain disposition. Every discrepancy between the disposition ascribed and the actual behavior of Benjamin Franklin can be explained by assuming further causal factors. So we have to ask what type of arguments is used to justify a certain general approach (for example Weber’s or Sombart’s or Brentano’s) and what are the criteria to justify historical explanations of individual behavior. To be sure, Weber argues that one has to use what he calls “nomological knowledge”. This knowledge might be acquired by everyday experience or by science. In both cases it is – in the eyes of Weber – nothing more than an universalization of observed regularities under certain descriptions and by consequence only a more or less confirmed hypothesis. Against Weber one has also to take into account that different presupposed ideal types yield different “laws” or law-like sentences and different causal factors. For if we attribute to Franklin the disposition of a Calvinist’s mentality one has to expect all those manifestations of this disposition, which one associates with it by “nomological knowledge” attained by the observation of more pure instantiations of this mentality. Franklin’s deviant behavior has to be explained by further circumstances, preventing the pure manifestation of the attributed disposition. Yet, if we were to attributing a disposition to Franklin that is constructed much more along his actual behavior, there would be no need for the search of preventing circumstances so that he is regarded as falling under different “laws”. Note however that both law-like sentences might be equally confirmed by observations. Weber’s nomological knowledge may play a methodological role but does in no way solve the problem of ascribing dispositions in the historian’s work.

250

Robert Schnepf

4. For Brentano, Sombart, and Weber, disposition terms are somehow theoretical terms. Weber is quite explicit on this issue. There is a given manifold of historical data or phenomena that can be described as sequences of actions of individuals. Without selecting and putting this manifold in order neither description nor explanation are possible.26 Dispositions like characters or the “capitalist’s spirit” are not given phenomena but constructed theoretical concepts in order to explain certain regularities to be observed within the “chaotic realm” of historical phenomena. The different explanations given by Brentano, Sombart, and Weber use different theoretical concepts to order and to explain the manifold of given historical phenomena. More or less the same set of historical data can be explained in terms of a constant disposition and changing background conditions Brentano proposes, in light of a change of basic dispositions (and an corresponding change of background conditions) Sombart suggests, or by reference to the acquisition of a further disposition modifying the basic first one (and an corresponding change of background conditions) as Weber argues. Ascribing a disposition and dispositional explanations are therefore mere hypotheses in need of additional confirmation. They cannot be directly confirmed by the given data since all three explanations cover the same set of phenomena. Each approach is able to reject counter-examples by claiming exceptions or further specific circumstances. Let’s now return to the question of whether dispositions can function as natural explanation stoppers Watkins has argued and – if they can not – what other methodological function they could have in the work of the historian. Brentano, Sombart, and Weber argue for their favorite disposition ascription very differently. These differences can be explained in view of their different conception of dispositional terms and the understanding of their explanatory role. Brentano conceives of dispositions in a realistic manner. The basic dispositions are constituents of human nature. All human beings tend to acquire goods. Brentano argues therefore on Aristotelian grounds. Against Sombart’s theory of a change of basic dispositions Brentano argues with Aristotle and his distinction between an economy concerning one’s own house and an

_____________ 26

Cf. M. Weber 1985, 177. Stephen Mumford objected to my thesis in a discussion that this approach presupposes a Humean picture of sense data or something like this, while he prefers a non-Humean world in which causal relations for example could be seen (cf. Mumford, in this volume). His observation is – to be sure – quite right and Weber influenced by Neo-Kantians like Heinrich Rickert would certainly agree. In this essay I do not have the space to enter into these discussions. But as far as I see it, the historian is in fact confronted with a prima facie chaotic manifold of sources that are open to different interpretations. Even if one assumes that causal relations are visible in our perception of the world (which I in fact would deny) it seems rather strange to hold that causal relations in the past are in some sense visible or observable.

The Role of Dispositions in Historical Explanations

251

economy concerning further trade. Assuming basic and unchangeable dispositions, every change in behavior must be explained by changed background conditions. The gap between the constant nature of man and his real behavior can be bridged by specific causal hypotheses. The changing background conditions are the relevant causal factors in historical explanation. In Brentano, dispositional explanations seem to be informative, since the gap between the disposition ascribed and the actual behavior is bridged with information about relevant causal factors The central problem in his approach consists in the fact there is no way of verifying or confirming the basic hypothesis, i.e. the disposition ascription. For it seems possible to assume another disposition as ingredient of the human nature and constructing another causal story. Leaving this problem aside, it is obvious, that using disposition terms in this manner does not provide us with explanation stoppers, because the gap between the ascribed disposition and the observed behavior is now in need of explanation and demands information about further causally relevant factors. Taking disposition terms in a realistic manner is therefore of no help in searching for the causes of the observed behavior. It is only a very special way of posing the question. Sombart’s account of disposition ascription seems to avoid this difficulty. Sombart takes his dispositions as part of a theory or as theoretical terms. These terms are constructed. In a first step one has to collect the data which are observations of acting individual persons. In a second step one idealizes the given descriptions of the phenomena by excluding irregularities and subsumes them under more abstract concepts (for example “rationalization”). In a third step one interprets these idealized phenomena as the results of actions of individual persons (i.e. a person acting completely rational), harmonizes them with others and ascribes to individual persons a certain type or character which is the adequate cause of his behavior or acts (insofar they are rational – for example).27 The concept of a disposition is always formed in order to causally explain a special type of behavior. So rationality is a disposition to explain the rational features of actions. If we observe that human beings changed their behavior in some fundamental respects we can explain this shift by constructing the concepts of two dispositions, the one explaining the first type of behavior and the second explaining the latter type of acting. The change of behavior can now be described as a change of basic dispositions. According to Sombart’s way of forming disposition concepts the gap between the dispositions ascribed and the observed behavior can not be as wide as in the case of Brentano’s dispositional explanations. For his dispositional concepts are constructed to explain a specific type of behavior. On the other hand, they dispositional explanations of this type seem to be odd and repeti-

_____________ 27

Cf. W. Sombart 2003, 7 et seq.

252

Robert Schnepf

tive. Explaining a change of behavior by describing it as a change of dispositions each explaining a special kind of behavior explains nothing. But this way of forming disposition terms can nevertheless play a very fruitful role in the historian’s search for explanations: The acquisition and change of disposition itself has now to be explained. Forming disposition terms in this manner opens a new field of scientific research and does in no way function as explanation stopper. I take Weber’s account of ideal types as one of the most promising theories of dispositional explanations in historiography. What is it to ascribe a disposition to an individual person? According to Weber in attributing a disposition – e.g. a character or spirit – one attributes non-dispositional categorical properties and postulates law-like connections between having those categorical properties and acting in a certain manner. In describing a person as a Calvinist one simultaneously maintains that he would act in this special way (for example, that he would control his time-management) if certain conditions are fulfilled. The criteria of being a Calvinist might be non-dispositional (for example the membership in a church or being baptized in a Calvinist church or being the offspring of a Calvinist family). Being Calvinist is thus not interpreted as a disposition but as a categorical property or as an observable datum (provable by checking the registers of the church). The law-like connection can however not be verified by observable regularities. Most Calvinists may behave in a way that does not correspond to the behavior commonly associated with the attributed disposition. To explain their actual behavior one would thus appeal to circumstances that prevent them from being “ideal” Calvinists. Here is how the explanations of such actual behavior would proceed: Within the Calvinists system of beliefs it would be rational and consequent to perform acts of this kind. From the point of view of a constructed ideal Calvinist, derived from Calvinist theology, it is necessary to control one’s own time-management. This is the only way to confirm the law-like connection between being Calvinist and performing certain acts. It is no surprise that this kind of regularity cannot be observed. According to Weber, dispositions imply therefore counterfactual conditionals (what he calls “objektive Möglichkeit”, i.e. objective possibilities). If I have to explain the observed behavior of an individual person using such an ideal type two cases can occur. First, the person acts according to the ascribed disposition or the ideal type used by the historian. The actions are explained by the person’s being of this type. Second, the person does not act in the manner that was expected. In this case the historian has to search for additional non-dispositional properties of the person that are causally relevant to explain this deviation. The hypothesis that claims that a special factor is responsible for a deviation in behavior has also to be justified by nomological knowledge. In neither case, can the real mo-

The Role of Dispositions in Historical Explanations

253

tives, thoughts or psychological characteristics of the person be known. The explanation conforms to the dictum of methodological individualism not because it appeals to a person’s intimate thoughts but by understanding the actions performed by the individual as rational actions given that she is a Calvinist (for example) or Calvinist under special conditions. Dispositional explanations of this kind are informative for the dispositional concept is constructed from two sides: It is a theoretical term constructed as implying law-like sentences that can be used in deductive explanations. In this sense, they can be interpreted as Carnap interpreted dispositions in “Meaning and Testability”. According to this approach, dispositions and disposition ascriptions are mere hypotheses unless they are tested. On the other side, the law-like connection is sustained not by observation but by showing that the behavior to be explained would be rational given that the agent is to be subsumed under the ideal type used by the historian. This kind of reasoning confirms counterfactual conditionals in the case of not being tested or testable. Explaining is not showing that the explanandum was to be expected given the explanans but to connect the explanans with the explanandum in a way that we can answer further questions, for example how the explanandum was brought about. We want to connect the explanandum with the explanans by something like a connecting story. In the case of Weber this is done by showing that the action to be explained was not only to be expected if the agent can be subsumed under the ideal type but that the action can be conceived as reasonable from his point of view. This presupposes that his point of view can be constructed as a point of view of someone who can be subsumed under this ideal type. This is possible even if it is not possible to know what it is like to be a Calvinist of the 16th century. If the properties attributed to single persons described as an ideal type are categorical properties, then nothing prevents that this fact can be further explained in a second step. For example, the fact that a person was baptized in a Calvinist church can be the object of a further explanation. This explanation may be of a different type – for example connecting the economical situation of the family with their religious attitudes.

5. I do not want to defend all aspects of Weber’s approach. To me the debate between Weber, Sombart and Brentano allows me to argue for the following points: • Dispositional concepts in historiography need to be reducible to observable non-dispositional properties, if we want to justify their ascription. There is no need for and no use of interpreting them in a

254

Robert Schnepf

•

•

realistic manner, since there is no epistemic access to the “real” properties and mental states of individuals and dispositional explanations can only be defended in reference to observable facts. To use one of my other examples: Caesar’s way of dealing with his prisoners of war can be explained by his clemency but this explanation cannot be defended by reference to Caesar’s disposition. This would be circular: Explaining Caesar’s acts in terms of a disposition can only be defended by these very observable acts by Caesar – that is by the explanandum. No methodological problem of a historian’s research is solved by interpreting disposition-term in a realistic way. A dispositional explanation is informative only if there are two or more different approaches possible. On the one hand, the disposition must be conceivable as connecting the ascription of categorical properties and counterfactual conditionals. On the other hand, the same disposition may establish a way of understanding the actions actually performed as rational under the ascribed ideal type. Or it may open up the possibility to connect the explanation given by disposition ascription with other parts of our historical knowledge or with other theories widely accepted. One might try to integrate an explanation within a sociological theory, for example. Explaining Caesar’s way of dealing with his enemies in terms of his clemency is not informative because it does not allow us to understand those acts as rational acts from the point of view of – let us say – a master of Realpolitik or to understand his behavior in terms of our knowledge of the Roman society. Disposition terms play a very special role in historical explanations. They are not explanation stoppers (as Watkins and others assumed) but we can they have important heuristic value insofar as we can use them to provide fruitful descriptions of the explanandum and to formulate new questions. If we subsume Caesar under the type of a cold but ingenious politician (which one might interpret as a disposition) we know where we have to look for relevant factors that might explain his behavior. If we accept this approach, we are searching for factors that Caesar himself might have regarded as relevant for his decision making. If we subsume him under the type of a mild benefactor we would attribute to him a completely different way of thinking and consequently would have to search for other factors relevant for his decision making. If one accepts the ascription of clemency to Caesar one may ask how he acquired this disposition especially if we notice his cruelty during the Gallic war. Disposition terms do not explain by themselves but they direct our search for explanations.

The Role of Dispositions in Historical Explanations

255

Literature Braudel, Fernand. 1992a. “Geschichte und Sozialwissenschaften. Die lange Dauer.” In F. Braudel. 1992. Schriften zur Geschichte 1, Gesellschaft und Zeitstrukturen. Stuttgart: Klett–Cotta: 49-87. Braudel, Fernand. 1992b. “Geschichte und Soziologie.” In F. Braudel. 1992. Schriften zur Geschichte 1, Gesellschaft und Zeitstrukturen. Stuttgart: Klett–Cotta: 99-121. Brentano, Lujo. 1923. Der wirtschaftende Mensch. Gesammelte Rede und Aufsätze. Leipzig: Felix Meiner Verlag. Carnap, Rudolf. 1936. “Testability and Meaning (part I).” Philosophy of Science 3: 420-471. Carnap, Rudolf. 1937. “Testability and Meaning (part II).” Philosophy of Science 4: 1-40. Carrier, Martin. 2006. Wissenschaftstheorie zur Einführung. Hamburg: Junius. Dray, William. 1980. Perspectives on History. London et. al.: Routledge & Kegan Paul. Gellner, E. A. 1956. “Explanations in History.” Aristotelian Society, Supplementary Volume 30: 157-176. Gilcher-Holtey, Ingrid. 1998. “Plädoyer für eine dynamische Mentalitätsgeschichte.” Geschichte und Gesellschaft 24: 476-497. Graus, František. 1987. “Mentalität – Versuch einer Begriffsbestimmung und Methoden der Untersuchung.” In F. Graus, ed. Mentalitäten im Mittelalter. Methodische und inhaltliche Probleme. Sigmaringen: Jan Thorbecke Verlag: 9-48. Hegel, Georg Friedrich Wilhelm. 1970. Vorlesungen über die Philosophie der Geschichte, Werke Bd. 12. Frankfurt a.M.: Suhrkamp. Hempel, Carl. 1970. Aspects of Scientific Explanation and other Essays in the Philosophy of Science. New York: The Free Press. Hintze, Otto. 1929. “Der Moderne Kapitalismus als historisches Individuum. Ein kritischer Bericht über Sombarts Werk.” In Bernhard von Brocke, ed. 1987. Sombarts „Moderner Kapitalismus“. Materialien zur Kritik und Rezeption. Munich: DTB: 322-377 (orig. published in: Historische Zeitschrift 139 (1929) 457-509). Kamlah, Andres. 2002. Der Griff der Sprache nach der Natur. Paderborn: mentis. Kaila, Eino. 1945. “wenn … so …” Theoria 11: 88-98. Kalberg, Stephen. 1994. Max Weber’s Comparative Historical Sociology. Chicago: The University of Chicago Press. Lehmann, Hartmut, and Roth, Guenther, eds. 1987. Weber’s Protestant Ethics. Origins, Evidence, Contexts. Cambridge: Cambridge University Press. Lewis, David. 1979. “Counterfactual Dependence and Time’s Arrow.” Nous 13: 455-476. Lewis, David. 1997. “Finkish Dispositions.” The Philosophical Quarterly 47: 143-158. Martin, Rex. 1977. Historical Explanation. Re-enactment and Practical Inference. Ithaca, London: Cornell University Press. Parsons, Talcott. 1928. “‘Capitalism’ in Recent German Literature: Sombart and Weber.” The Journal of Political Economy 36: 641-61. Ringer, Fritz. 1997. Max Weber’s Methodology. The Unification of the Cultural and Social Sciences. Cambridge Mass./London: Harvard Univ. Press. Robins, Lionel. 1935. The Nature and Significance of Economical Sciences. London. Schnepf, Robert. 2006. Die Frage nach der Ursache. Systematische und problemgeschichtliche Untersuchungen zum Kausalitäts- und zum Schöpfungsbegriff. Göttingen: Vandenhoeck & Ruprecht. Schnepf, Robert. 2009. “Von der Seele zum Geist: Diltheys Grundlegung der Geisteswissenschaften.” In K. Crone, R. Schnepf, and J. Stolzenberg, eds. Über die Seele. Frankfurt a.M.: Suhrkamp (forthcoming 2009).

256

Robert Schnepf

Schnepf, Robert. 2010. Geschichte erklären. Göttingen: Vandenhoeck & Ruprecht (forthcoming 2010). Scriven, Michael. 1966. “Causes, Connections and Conditions in History.” In William H. Dray, ed. Philosophical Analysis and History. New York/London: Harper & Row Publisher: 238-264. Sombart, Werner. 1916. Der moderne Kapitalismus. 2nd edition. Munich/Leipzig: Duncker & Humblot (reprinted Munich 1969). Sombart, Werner. 1916. Der Bourgeois. Zur Geistesgeschichte des modernen Wirschaftsmenschen. 6. edition. Berlin: Duncker & Humblot 2003. Watkins, John W. N. 1952. “Ideal Types and Historical Explanation.” The British Journal of the Philosophy of Science 3: 22-43. Watkins, John W. N. 1957. “Historical Explanation in the Social Sciences.” The British Journal for the Philosophy of Science 8: 104-117. Watkins, John W. N. 1992. Wissenschaft und Skeptizismus. Tübingen: Mohr. Weber, Max. 1985. Gesammelte Aufsätze zur Wissenschaftslehre. Tübingen: J.C.B. Mohr. Weber, Max. 1988. Gesammelte Aufsätze zur Religionssoziologie I. Tübingen: J.C.B. Mohr.

Empathy, Mental Dispositions, and the Physicalist Challenge KARSTEN STUEBER In recent years, the subject of dispositions has become a topic of renewed philosophical interest leading to a sophisticated and wide-ranging debate. The philosophical environment has been much more receptive to the idea that dispositional terminology refers to real and irreducible properties that form part of the furniture of the universe; another sign that philosophy in the analytic tradition is no longer beholden to the empiricist framework from the beginning of the 20th century. Yet the debate about dispositions tends to stay on a very general level; trying to answer, for example, questions such as whether all dispositional terminology can be fully analyzed in terms of counterfactual conditionals and whether all dispositional properties are in need of a causal basis or whether they have an autonomous causal explanatory role to play. Given the general suspicion of dispositional terminology and dispositional properties within the empiricist framework taking such generalized perspective is certainly appropriate. Yet from this viewpoint, specific distinctions between different classes of dispositions are easily overlooked.1 Friends of dispositions as irreducible elements of the universe have to respond to arguments intended to show that all dispositions have to be regarded as reducible, that they are nothing but lower order categorical properties, or that all dispositional statements are cognitively insignificant. Having in general saved dispositional properties from ontological extinction, however, does not necessarily allow us to save each and every type of a prima facie dispositional property from such fate. Particularly within the contemporary framework of physicalism, the ontological status of higher order dispositional properties, such as fragility, that objects possess because of a certain micro-physical structure is anything but certain. In this article, I will discuss the physicalist challenge in regard to the particular domain of higher order dispositions that is delineated by our folk psychological vocabulary. As I will show, the general discussion about dispositions and their ontological legitimacy tends to take place firmly within the context of a physicalist conception of reality and a detached conception of

_____________ 1

See also McKitrick’s contribution to the anthology in this regard.

258

Karsten Stueber

objectivity within the physical sciences. The domain of higher order dispositions referred to within the context of folk psychology is however significantly different because it, in contrast to the conceptual framework of the physical sciences, involves not merely the detached third person perspective but involves essentially the use of the first person perspective. For this reason, the ascriptions of mental states such as beliefs and desires should be conceived of as having a doubly dispositional character. Folk psychological predicates ascribe complex multi-track dispositional properties to other agents, since they ascribe higher order properties to people that are commonly associated with characteristic manifestation events in a variety of circumstances. Yet in contrast to ascriptions of physical dispositions such as fragility, the ascription of mental dispositions depends essentially on our empathic ability, that is, our ability for mental imitation and our capacity to reenact another person’s thought processes in our own mind.2 And it is also for this reason that one is not epistemically justified in expecting that our folk psychological practice of explaining another person’s behavior in terms of his reasons can be accounted for from the physicalist perspective. Once we reject the epistemic legitimacy of requiring a reductive account for our folk psychological practices, we also alleviate the eliminativist pressure on those very same practices. More positively expressed, I will attempt to argue that we should philosophically reckon with the possibility that certain features of reality are only accessible to creatures that share our empathic dispositions and our psychological constitution. Accordingly, we should regard our folk psychological practices as constituting an ontologically relatively autonomous and an epistemically special explanatory domain and practice.

1. Physicalism and its Ontological Challenge to Dispositions For most of the 20th century the philosophical discussion about dispositions and the more specific discussion about mental properties have been closely linked. Arguments for and against the validity of dispositional terminology show at times a structural similarity to arguments first developed in the context of the philosophy of mind.3 As it is well known, it was Gilbert Ryle, with the publication of his seminal The Concept of Mind, who established the close connection between these topics of philosophical concerns. Ryle suggested a dispositional conception of mental terminology as a philosophical corrective

_____________ 2

3

This is also the manner in which I will understand the concept of empathy in this article. For an overview of the history of the empathy concept and the plurality of its definition in the various philosophical and psychological disciplines and sub-disciplines see my 2008a and the introduction of my 2006. See the later chapters in Mumford 1998.

Empathy, Mental Dispositions, and the Physicalist Challenge

259

to the “linguistically confused” Cartesian framework of thinking about the mind as the immaterial analogue to the body or as thinking about the mind as a ghost in the machine. Yet in conceiving of mentality in dispositional terms and in conceiving of dispositional statements in a certain manner, Ryle left mentality and dispositions in an ontologically precarious situation. For Ryle, dispositional statements are not fact-reporting statements. They “are neither reports of observed or observable, nor yet reports of unobserved or unobservable state of affairs.”4 Rather, they are mere inference tickets that justify our inferring one state of affair from another. If one would like to succinctly summarize Ryle’s position one is tempted to do so by saying that Ryle was doing away with the Ghost in the Machine via a dispositional analysis of mental terminology but that he was getting nothing in return. Minds do not seem to be anything in the world: Like causation within a Humean framework, minds have only to do with our ways of talking about the world. Ryle’s claim that dispositions belong to non-factual discourse is hard to swallow.5 It seems to be a factual matter that some objects will actually dissolve in water and that others will not, as much as it is a factual difference between objects that some have a specific shape and others do not. Certain facts in the world do seem to make the claim that glass is fragile or that Peter is angry true or false. Prima facie, objects that have certain dispositions possess different properties than objects that do not have such dispositions. The real question then is whether there is a real and irreducible difference between so called dispositional properties and non-dispositional properties or whether it is merely a difference that concerns linguistic entities; that is, a distinction between dispositional and non-dispositional predicates.6 Traditionally, according to what is commonly referred to as Eleatic Principle, the test for the reality of a property has been whether or not its existence makes some kind of causal difference. If there is no difference in this regard between an object having a property P (like it having a specific mass) and it not having the property, then P cannot be regarded as objectively real. Ascribing it to an object does not depend on facts in the world, but its ascription is merely subjective and depends on the eyes of the beholder. Moreover, to address questions about the reality of specific properties, philosophers are well advised to analyze existing scientific practices. Ultimately, questions regarding the furniture of the universe are decided not by philosophical armchair considerations but in terms of whether or not a property has to be appealed to in our scientific investigation of the causal structure of the world, that is, whether or not it is appealed to in our causal explanatory practices

_____________ 4 5 6

Ryle 1949, 125. See the contribution by Scholz in this anthology and Mumford 1998, 22ff. In this respect see also Heil 2003, chaps. 3 and 9.

260

Karsten Stueber

based on our “final” theories of the world. Educated guesses about how such practices will look like are only possible in light of our existing scientific theories and our understanding of the historical progress of science. For this very reason, Carl Hempel’s account of dispositional explanations falls short of saving the causal-explanatory role of dispositional properties. According to Hempel, dispositional explanations have the following form: i was in a situation of kind S3. i has property M. (L) Any x with property M, will in a situation of kind S, behave in a manner R3. i behaved in manner R3.7

Accordingly, dispositional explanations share most features that Hempel regards as crucial for an argument to constitute an explanation. Like any other explanation, they address certain kinds of epistemic desiderata in that they provide information that supplies a justification for our expecting the occurrence of a certain event, given the initial conditions. Dispositional explanations accomplish this task by telling us that a certain object has the tendency to manifest a certain type of behavior in certain circumstances. The shortcomings of Hempel’s conception of explanation are well known.8 Yet independent of these well known concerns, for our topic more important is the fact that nothing in Hempel’s defense of dispositional explanations allows us to answer the question of whether such explanatory schemes referring to dispositional properties are indeed sanctioned by our best scientific theories of the world or whether they are merely “folk” or “commonsensical” ways of talking about the world without any real ontological significance. Even worse, serious doubts about the ontological status of higher order properties including dispositional properties such as beliefs and desires arise within the perspective of contemporary physicalism; a position that is best understood as the philosophical articulation of the underlying assumptions of a reductive scientific research program that was so successful in the last few centuries and promises to be so in the future. Physicalism as the philosophical articulation of our ongoing scientific practice is commonly understood as containing both ontological and epistemic commitments.9 Ontologically speaking, the physicalist views higher order macroscopic facts as being dependent or supervenient on basic micro-physical facts. As contemporary metaphysicians like to express it in a slogan: God only had to fix all the mi-

_____________ 7 8 9

Hempel 1965, 462. See Salmon 1989. See for example J. Poland 1994. It is also part of Chalmers’ conception of physicalism and his argument for the claim that phenomenal consciousness cannot be accounted for within the physicalist framework. Hence physicalism must be false. See Chalmers 1997.

Empathy, Mental Dispositions, and the Physicalist Challenge

261

cro-physical facts in order to fix all the facts. Yet, since physicalism is best understood as articulating a reductive scientific practice its ontological commitment seems to entail a corresponding epistemic postulate that it is in principle possible to know the nature of this ontological dependency. Hence, physicalism commits us epistemically to accept the claim that it is in principle possible to explain how the interaction of lower order micro-physical properties gives rise to higher order properties. Accepting brute correlations or identities between macroscopic and micro-physical facts without any hope of ever finding out how and why certain microscopic facts give rise to macroscopic phenomena would not be consistent with viewing physicalism as the presupposition of a reductive scientific practice. It would seem to undermine the philosophical justification of the physicalist framework and reveal its metaphysics as mere materialistic dogma. To express it in another slogan: God had not only to fix all the physical or micro-physical facts in order to fix all fact, but he and anybody else would only have to know all the physical facts in order to know all the facts. Given physicalism, the following worry arises, which has been the subject of an intense discussion in contemporary philosophy: If the existence of higher order properties depends on physical or micro-physical properties, then one could argue that their causal efficacy is nothing but the causal efficacy of the micro-physical properties they depend on. If higher order properties could be easily identified with lower order properties one would not have to be overly concerned about the question of whether such higher order properties really exist. Yet as it is well-known, mental properties such as beliefs and desires (and dispositional properties such as solubility in water or fragility) are multiply realizable and for that reason can not be identified with lower order properties. More generally, if one assumes with Prior, Pargetter, and Jackson10 that all dispositional properties are in need of a categorical basis, but claims (i) that dispositional properties cannot be identified with their basis, and (ii) that all the causal work is done on the lower level, then it seems that dispositional properties do not have any causal role to play. To express this worry in a slightly different fashion one could argue that knowledge that a particular object is fragile does not add anything over and above our knowledge of its micro-physical structure in allowing us to manipulate and intervene in the natural processes occurring in the world. Dispositional properties could then not be understood as being part of the ultimate furniture of the world. They have to be conceived of as linguistic fictions; as being linguistic categories created because it is convenient for us to speak, for ex-

_____________ 10

E. Prior, R. Pargetter, and F. Jackson 1982.

262

Karsten Stueber

ample, of the fragility of various objects. In reality, however, no natural property corresponds to our dispositional way of speaking.11 Within the philosophical discussion about dispositions there have been a variety of argumentative strategies utilized in order to respond to the general ontological challenge to dispositional properties. It has been argued that not all dispositions are in need of a categorical basis;12 that the strict distinction between categorical and dispositional properties cannot be philosophically maintained;13 that dispositional categories such as the category of a power is ontologically basic;14 and that the fundamental level as it is described by contemporary physics itself contains reference to dispositional properties.15 Taken together, these various strategies undermine the physicalist challenge to dispositions by suggesting that the basic ontological level of the universe cannot be rationally conceived of without reference to dispositions, as at least one of its irreducible elements. Here I will not comment on the ultimate success of the various arguments to save dispositions from ontological extinction. For my purposes, it is more important to point out that the above strategies do not save higher order dispositional properties, particularly higher order mental properties. Having such higher order dispositional properties requires a specific physical basis. As we all know, we need a well-functioning brain to have mental states and capacities. Consequently, one would still have to worry about the objective reality of such properties, since their supposedly causal efficacy seems to be nothing but the causal efficacy of lower order physical properties. In the end, the above mentioned strategies to ontologically safeguard dispositions allow us merely to infer that some dispositional properties are not higher-order and that they have to be accepted among the basic elements of the universe. Whereas one can live easily with the recognition that no real property might correspond to a dispositional predicate such as fragility, higher order mental predicates that are at the core of our folk psychological practices, like the concepts of beliefs and desires, are not so easily done away with. Our ethical and legal practices, for instance, depend centrally on the notion of responsible agency and it is hard to conceive of how such notion can be understood without concepts like beliefs and desires. Certainly no notion currently used in the physical sciences seems to come close to play that role. It is

_____________ 11

12 13 14 15

In the philosophy of mind this argument is well known as the causal explanatory exclusion argument best articulated by Jaegwon Kim. See Kim 1998a and Stueber 2005 for a critical discussion and more precise articulation of this argument. McKittrick 2003. Heil 2003. Mumford, this anthology. See for example Hütttemann in this anthology.

Empathy, Mental Dispositions, and the Physicalist Challenge

263

for this reason that the debate about the ontological status of mental states has been so intense in current philosophy of mind. However, merely wanting to save mental properties does not constitute a sufficient reason for holding onto the folk psychological framework and for regarding it as ontologically sound, especially if one is inclined, as I am, to accept the ontological framework of physicalism. In the following, I will try to alleviate the physicalist challenge to mental dispositions by providing a more fine-grained analysis of the relationship between the ontological and the epistemic commitments associated with the physicalist framework. I would like to suggest that there is a specific reason to reject the epistemic postulate of physicalism insofar as mental properties accepted within the folk psychological perspective are concerned, even if one admits the ontological priority of the physical. I would like to suggest that the ontological and epistemic commitments of physicalism come apart for properties that are identifiable only within a particular conceptual framework that is constituted by or centrally involves at least some epistemic capacities that do not have their equivalent in the physical sciences. Physicalism as an ontological thesis is compatible with the view that certain conceptual repertoires used for explanatory purposes and certain properties identified with the help of these repertoires are accessible only to creatures that share the physical and psychological constitution of normal functioning human beings. This does not imply that there are any non-physical facts; it just means that certain facts are only accessible to organisms with a specific physical and psychological constitution. For other creatures such facts are epistemically inaccessible. In this respect, there exist an essential epistemic contrast between the perspective of the physical sciences and the perspective of folk psychology. In contrast to the physical sciences, our folk psychological practice is centrally tied to the first person perspective and our empathic abilities; that is, our ability for mental imitation and our capacity to reenact another person’s thought processes in our own mind. For that very reason, one is also not justified in expecting that properties, identified within the conceptual domain of such a practice, can be accounted for from the perspective of the physical sciences that abstracts from such genuine human capacities, even if everything depends ontologically on the interaction among physical entities and properties identified within those theoretical practices.

2. Mental Dispositions and the Function of Empathy Within the context of analytic philosophy, one has mostly assumed that our folk psychological practices are on par with other scientific practices in that the folk psychological vocabulary is used in order to predict and explain hu-

264

Karsten Stueber

man behavior. Consequently, the respect that philosophers and scientist have shown toward folk psychology and its ontological assumptions has been closely associated with their judgments of whether or not the folk psychological framework is adequate for a rigorous scientific investigation of human behavior. Scientific behaviorists like Watson and Skinner adopted eliminativist attitudes towards folk psychology because the behaviorist framework promised to be able to account for human behavior without appealing to inner mental states. Since the cognitive revolution within psychology, the fate of mental states assumed to exist within folk psychology appears to be much brighter.16 More recently, however, eliminativist attitudes towards folk psychology have again become prominent in the context of connectionist accounts of mental capacities and have been forcefully articulated by Paul Churchland.17 One has therefore assumed that folk psychology, like the physical sciences, has to be understood as an epistemic enterprise that is ideally adopted from the “perspective of nowhere,” to use Nagel’s term, or from the detached perspective. Science in general is conceived of as being committed to standards of objectivity that are allowing us to overcome the subjective limitations of the human mind leading to well-known biases and other inferential shortcomings. Only in this manner can we expect that we are able to describe the world as it really is and not merely as we think it to be. The scientific perspective is a perspective that does not depend on us humans sharing a common psychological structure of the mind. Rather, it is a perspective that could be shared even by Martian scientists interested in exploring the structure of the world: It is a perspective accessible to any reasoner committed to the normative ideals of rationality regardless of how such reasoning is actually realized and embodied in the physical world. I have called the philosophical conception of folk psychology that conceives of folk psychology as being committed to or adopted from such a detached perspective the detached conception of folk psychology. It is a conception of folk psychology that has dominated philosophy of mind in the 20th century. It has been most memorably articulated by Wilfried Sellars with his myth of Genius Jones, who from a behaviorist basis develops an explanatory framework postulating inner mental states in order to provide a systematic and explanatory account of observed human behavior.18

_____________ 16 17 18

See however the skeptical voice of Stephen Stich 1983 in this context. For example Churchland 1989. W. Sellars 1963. For a discussion see Stueber 2006, chap. 1. While the detached conception of folk psychology certainly has been dominant throughout the 20th century, this is not to say that other positions were not articulated. Particularly the phenomenological and hermeneutic traditions have to be mentioned in this context. For a brief survey see the introduction of my 2006 and the first sections of my 2008a.

Empathy, Mental Dispositions, and the Physicalist Challenge

265

More recently, voices have grown stronger that challenge the assumption that folk psychology should be compared to standard scientific practices. These voices challenge the claim that prediction and explanation are the primary purposes of folk psychology. They suggest that the primary purpose of folk psychology is instead related to our practical and ethical concerns.19 Here, I do not have the space to sufficiently address all of the arguments articulated in favor of this revisionary view of folk psychology. Suffice it to say that I regard our folk psychological conception of other agents as acting because of their mental states – their beliefs, desires, and intentions – as being central for our viewing them as agents who can be blamed, praised, and with whom we can socially bond. The fact that folk psychology is central for our ethical practices does not contradict the view that folk psychology has to be understood as a causal explanatory practice of accounting for a person’s behavior. Indeed only under the presupposition that beliefs and desires and so on play a role in our causal-explanatory account of human behavior does it also make sense to view an agent as a person who is responsible for his behavior. That is, only as long as we can understand an agent as having acted because of his reasons does it make sense to view his external behavior as something that originates in him in a manner that allows us to praise or blame him for his actions. Accordingly, I tend to agree with the view that folk psychology is on par with other scientific practices in being a causal explanatory practice. Yet I disagree with the orthodox account of folk psychology according to which it is an explanatory practice that is adopted from the detached perspective of the physical sciences. Rather, I view folk psychology as an explanatory practice that is adopted from what I call the engaged perspective; from a perspective that does not abstract from our uniquely human capacities, particularly the human capacity for empathy. It is an explanatory practice that can only be accessed from a human point of view and not from the point of nowhere. Organisms that do not share our psychology will not fully grasp the point of our folk psychological practices. Unlike the physical sciences folk psychology is a causal explanatory framework with a particular human touch.20

_____________ 19

20

For some recent voices in this regard see Hutto and Ratcliffe 2007, particularly the articles by Andrews, Knobe, Ratcliffe, and Morton. I discuss this claim as argued for by Knobe in my “The Ethical Dimension of Folk Psychology?,” Inquiry (forthcoming 2009). This does not imply that we cannot sensibly extend our folk psychological scheme to account for animal behavior. I will not address this complex issue in this article. Ordinarily we do seem to extend out folk psychological framework to animals and it would count against my conception of folk psychology if I would deny the possibility of such extension in an a priori manner. I however would suggest that we can apply folk psychology in this manner because we do share basic psychological mechanisms with some species in the animal kingdom, particularly primates. Some primates do seem to share some basic capacity for empathy, for example. For an interesting discussion see DeWaal 2006.

266

Karsten Stueber

My understanding of folk psychology as an engaged practice in the above sense has two components: First, our perceptual encounter with the world is not one-dimensional in that we do not primarily perceive the world as consisting of one type of entities that differ only in their physical characteristics and in the complexity of behavior that they show. Indeed this has been the common assumption among various philosophical accounts of how we acquire knowledge of other minds. According to this assumption, we access mental states of other people always indirectly. Our access is mediated by various inferential mechanisms using either analogical reasoning as the Cartesians suggest or theoretical principles, as theory theorists suggest. Yet the assumption that we encounter the external world in a perceptually onedimensional manner has to be regarded as a philosophical fiction. Human beings do not encounter the world in this manner and infer only indirectly that some physical things also have minds because of the complexity of their observed behavior. Rather, we perceptually encounter the world in already distinguishing between mere physical objects and creatures that are more like us. As recent neuro-scientific research establishing the existence of so called mirror neurons has shown, our perception of the observation of other human beings engages very different neurobiological systems than the observation of inanimate objects. In observing another person’s bodily movements and his facial expressions associated with emotions such as sadness, fear, or disgust similar neurons are activated in us that are normally are activated when we execute the same bodily movement or when we feel those emotions. As neuroscientists like to express it, there exists a significant overlap in the activation of neurons underlying the execution and the observation of bodily movements and of facial expressions associated with specific emotions. From the neurobiological perspective, the perceptual encounter of other human beings has to be understood as an inner resonance phenomenon, as a form of inner imitation. In light of Theodor Lipps’s view of empathy as a form of inner imitation, I refer to those resonance mechanisms as mechanisms of basic empathy. They allow us to understand another human being’s movement as a goal directed movement (as being directed towards the cup) and they allow us to read another person’s facial features as expressing certain emotions and feelings. From the very beginning, human beings relate to other people in a very different fashion than to physical objects. This does not imply that infants already understand other minds and other persons in a conceptually sophisticated manner. Rather one should conceive of the level of basic empathy as allowing the infant to grasp the movement of another as goal directed in the same manner that his own action could be directed towards that cup, that is as something that can practically imitated by him; that he is disposed to imitate in certain circumstances, or that he himself could do. In providing us with mechanisms of basic empathy, nature in my opinion has endowed us

Empathy, Mental Dispositions, and the Physicalist Challenge

267

with “perceptual similarity spaces” that make the development of the conceptual framework of folk psychology possible in that it allows us to practically grasp that the other person is “minded” like us; a similarity that then can be further explored and specified in a conceptually more fine-grained manner by acquiring the conceptual scheme of folk psychology. It provides us with an intersubjective basis for our natural exploration of the mind in that it allows us to grasp that the other person is feeling something that we ourselves can feel when we are taught that he is feeling sad. On the most basic level then the attribution of mental states to another person – such as attribution of emotions based on the observation of facial expressions or the attributions of goal-directedness based on the observations of movements of the other bodies – is not merely an attribution of mental dispositions from the detached perspective. Rather, it is an attribution from the engaged perspective in that, when I perceptually encounter the other person face to face, it directly resonates with me.21 The above observations do not imply that mental states do not ontologically depend on certain physical properties or the physical structure of the brain. Yet they should make us more careful of how exactly we conceive of the relationship between the ontological and epistemic components of physicalism. It should make us suspicious of claims articulated particularly in the contemporary discussion of consciousness that physicalism as an ontological thesis also implies that the application conditions of all concepts can be fully defined from the perspective of the physical sciences as Chalmers and Jackson claim.22 Based on this assumption, Chalmers argues that physicalism has to be false because it is conceivable that creatures that are physically like us in all respects lack phenomenal states. Certainly, if one is given all the information about a person’s brain described from the detached third person perspective without encountering that person face to face, one is without doubt able to conceive of that person as not feeling anything. Yet if I am right, this fact does not prove the falsity of physicalism as an ontological thesis. Rather, the fact that folk psychological concepts – including phenomenal concepts – cannot be fully accounted for from the perspective of the physical sciences can be explicated in light of the fact that only folk psychology is an engaged practice in the above sense. Accordingly, some folk psychological concepts are primarily learned in the face to face encounter with another human being; not by theoretically registering the exact contraction of facial or bodily muscles but by activating mechanisms of basic empathy. Asking us to conceive of

_____________ 21

22

For my philosophical interpretation of mirror neurons in regard to the debate between simulation theory and theory theory see also my 2006 chaps. 3 and 4. For a summary of the scientific results on mirror neurons see particularly the recent books by Rizzolatti and Sinigaglia 2008 and Iacoboni 2008, a book written in a more popular style. Chalmers 1997 and Chalmers and Jackson 2001.

268

Karsten Stueber

the possibility that a person might be a zombie, while providing us only with the information that he cannot be distinguished from us within the perspective of the physical sciences, is asking us to adopt a detached perspective. Hence, the so called conceivability argument against physicalism asks us to take a perspective that abstracts from our own capacities of basic empathy that are normally central for the application of some of our folk-psychological concepts. From that perspective, we indeed have to wonder whether other people are zombies; that is, people without any feelings and emotions. Yet, given the above considerations, the detached view does not reflect normal linguistic competency of every folk-psychological concept. Assuming such competency, the right question to ask is not whether we can, from a detached perspective, conceive of somebody as lacking certain psychological states. We have to ask instead whether we can imagine him or her lacking such states assuming that we were confronted with him or her in a perceptual encounter in the physical world. If we confront our physical replica in this manner, we do not have any reason not to ascribe normal mentality including phenomenal states to our physical replica, since it is to be assumed that the observation of their physical expressions and movements resonates with us. All of this is compatible with physicalism as an ontological thesis. Physicalism understood in this manner does not imply that all of the concepts, which a human being as a specifically structured physical organism uses, have application conditions that can be fully accounted from the detached perspective of the physical sciences. Physicalism also does not imply that all of our concepts are accessible to all creatures capable of comprehending the physical sciences, even if they lack a human being’s psychological capacity of empathy.23 Nevertheless, I would admit that not all of our folk psychological concepts are closely tied to mirror neurons and considerations about mirror neurons do not easily generalize to folk psychology as a whole.24 Nothing that I have said so far implies that the fully developed explanatory practice of folk psychology is not committed to the same standards of objectivity as the physical sciences, even if its perceptual basis has to be regarded as being dif-

_____________ 23

24

It seems to me that Wittgenstein expresses a very similar sentiment in the following aphorism: “But can’t I imagine that the people around me are automata, lack consciousness, even though they behave in the same way as usual? – If I can imagine it now – alone in my room – I see people with fixed looks (as in trance) going about their business – the idea is perhaps a little uncanny. But just try to keep hold of this idea in the midst of your ordinary intercourse with others, in the street, say! Say to yourself for example: “The children over there are mere automata; all their liveliness is mere automatism.” And you will either find these words becoming quite meaningless; or you will produce in yourself some kind of uncanny feeling, or something of the sort.” (Wittgenstein, Philosophical Investigations § 420) In this respect I would disagree with Iacoboni, who gives the impression in his 2008 that mirror neurons are the only mechanisms for human intersubjectivity and our understanding of other minds.

Empathy, Mental Dispositions, and the Physicalist Challenge

269

ferent than the perceptual basis of the physical sciences. Furthermore nothing I have suggested so far seems to indicate that Genius Jones lacking our system of mirror neurons could not learn to use most of our folk psychological system for the explanation of human agency, particularly insofar as explanations in terms of beliefs and desires are concerned.25 Given his knowledge of human neurobiology he might even be able to explain how our perceptual basis of the application of some of our folk psychological concepts differs from his own. The argument so far has thus only limited value because the attribution of mental states like beliefs and more complex emotions like guilt do not seem to activate our mirror neuron systems. Yet even those more complex forms of mental state attributions work, or so I am prepared to argue, only as long as the attribution to other agents “resonates” with the understanding of our own agency from the first person perspective. The argument for this claim, however, does not depend on knowledge of neurobiological mechanisms underlying our capacity for “reading” other minds. It depends on the fundamental fact that in our folk psychological practices we conceive of beliefs and desires not merely as inner mental causes but as an agent’s reasons for acting. Attributions of mental states to an agent are regarded to describe the parameters within which an agent conceives of his relation to the world and in light of which he deliberates about his actions. In this sense, the notion of rational agency is at the center of our fully developed folk psychological practice. Rational agents do not necessarily act for good reasons, or reasons that are above reproach, but they act in light of beliefs and desires that they regard as reasons for doing what they doing; that is, they act because something can be said for their action from their perspective. Understanding beliefs and desires as reasons for which a person acts does require more than mere knowledge of general principles of how beliefs and desires in general interact in order to produce a certain kind of behavior. It is exactly in this respect that our folk psychological practices differ from the physical sciences. This feature of folk psychological explanations is best illustrated by belief-desire explanations that somehow fail to have their normal explanatory potential. Take for example the case of a rather unusual behavior of a professor talking about the irreducibility of mental disposition: He suddenly takes out a gun and shoots the person who has just entered the room. Pointing out that he just had a desire to kill the next person coming into the room and that he believes that this could be accomplished by shooting him, does not sufficiently explain why he did what he did. In contrast, pointing to the belief that there is a beer in the fridge and to the desire to drink a beer

_____________ 25

I would however argue that Genius Jones would not be able to fully use folk psychological concepts for self reports, at least not in a manner that satisfies what is commonly called the univocality constraint; that is, mental terms in first person reports and third person reports have the same meaning. For an argument in this respect see my 2006, chap.4.

270

Karsten Stueber

normally provides an explanation for that person’s behavior of going to the fridge. Significantly, the difference in the explanatory potential between these two accounts cannot be explicated in terms of general principles that according to the theory theory position are at the foundation of all of our explanatory practices: Both explanations seem to make use of the same principle; that is, that if somebody desires x and believes that A-ing is a means of achieving x then, ceteris paribus, he will do A.26 Rather, the difference between those cases has to do with the fact that only in the second case can we fully understand why such beliefs and desires could be somebody’s reasons for acting in this particular manner. They could be his reasons because I can easily understand how they could be my reasons for getting a beer from the fridge. But we are puzzled in the first case because we are unable to understand analogously the cited beliefs and desires as reasons for acting and are therefore inclined to doubt the explanatory value of the offered account. The difficulty has to do with the fact that we do not know anything about the agent’s other beliefs and desires. Thoughts can be understood as reasons for acting only in the context of other relevant thoughts. Normally we presuppose (rightly or wrongly) that other people share our beliefs and so on. In understanding the belief that a class with a certain professor provides an easy A as a reason for a person to take the class we presuppose that the person has similar beliefs about the adopted grading system (that an A is the highest grade); that it is very important to get good grades, that it is more important for example than having to work hard and so on. For that very reason, we also understand easily how a belief about beer can be part of a reason to get the beer (we would however have difficulty understanding that behavior if we know that somebody thinks that the use of alcohol is a mortal sin). In the first case, however, given our moral values and our knowledge about the legal repercussions of such action, we think that the agent had overwhelming reasons for acting very differently. This does not mean that we never will be able to understand how those beliefs and desires could be a person’s reasons for his rather unusual behavior. Maybe if we are told that the person grew up in a very different culture, that this killing was an honor killing, that he knows which person would enter the lecture hall and so on, then we might be able to understand how somebody might act in this manner for those reasons.

_____________ 26

To clarify my position briefly: I do object to the theory theory position as an account of the underlying psychological mechanisms of our folk psychological practices. I do not think that theory theorists have succeeded in showing that theoretical inferences utilizing knowledge of folk psychological principles are centrally involved on the causal level. Yet I accept that implicit reference to generalizations is central for the epistemic justification of explanations. Nevertheless, I argue that such use of generalizations do not make reenactive empathy superfluous even in the epistemic context of justification. See my 2006, chap.5 and my 2008b.

Empathy, Mental Dispositions, and the Physicalist Challenge

271

It is worth emphasizing that we are able to do so only because, provided the information about relevant differences between us and the subject, we are able to put ourselves in his shoes and consider the situation from his perspective. His beliefs and desires are thus grasped as his reasons because given those beliefs and desires (and given his other thoughts about the world) I would have also acted in this manner in this situation. I thus can understand his beliefs and desires and so on as his reasons because I can reenact those beliefs and desires through my imaginative capacities as mental states that could be my reasons for acting; that is through the use of what I call reenactive empathy. Most importantly, such reenactment of another person’s thoughts as his reasons in my own mind, or what I also call the use of reenactive empathy, is absolutely essential for understanding thoughts as reasons. As I have pointed out, understanding thoughts as reasons requires us to conceive of them as holistically “fitting” in with an agent’s other beliefs and desires in the relevant manner in a specific context. In particular, it requires figuring out which of his many other thoughts are relevant for consideration in a specific situation in order for a specific belief-desire pair to be a reason for acting. Obviously, in trying to account for beer-getting-behavior all alcohol related thoughts would be prima facie relevant to consider; less relevant seem to be thoughts about drugs and their medical efficacy. Yet one could imagine that even those thoughts are relevant in a different context where one’s doctor has described beer as a medical cure for an ailment. What is more, as the so called frame problem in cognitive science has demonstrated, no general theory of relevance does seem to exist that allows us to decide which thoughts are relevant to consider in specific contexts.27 Our only option for practically solving this problem is by imaginatively putting ourselves in a particular situation, by engaging those thoughts in our own mind and by using them as a basis for our own deliberative encounter with the world. One last caveat, I claim only that reenactive empathy is essential for understanding how certain thoughts could function as reasons for a certain action. I do not maintain that empathy is on its own capable of deciding among a variety of plausible interpretive hypotheses, which account for an agent’s behavior in terms of very different reasons. Empathy on its own is not able to decisively say whether the agent acted for those or other thoughts as his reasons. Such decisions have to be made in light of further evidence regarding which belief attribution and so on is more plausible given the social context and the personal biography of the agents. Yet empathy is always involved in establishing interpretive hypotheses as hypotheses that have to be

_____________ 27

Within the context of the contemporary theory of mind debate, it was particularly Jane Heal who argued for simulation theory in this manner. It constitutes what I call the argument from the contextuality of thoughts as reasons. I also argue for empathy using the argument from the essential indexicality of thoughts as reasons. See Stueber 2006, 152 ff.

272

Karsten Stueber

taken seriously because only in this manner can we understand an agent’s thoughts as potential reasons for acting.28 To summarize this section, I take the above considerations to have established the following points: Within the context of folk psychology, explanations in terms of beliefs and desires play a central role, even though not the only role.29 Explanations in terms of beliefs and desires can develop their full explanatory potential only if they can be understood as reasons for acting. Understanding reasons for acting requires the use of our empathic capacities to reenact thoughts in our own mind and imaginatively take another person’s point of view. Given the epistemic centrality of empathy, particularly reenactive empathy, for folk psychology, our practice of attributing mental dispositions to other agents is very different from the attribution of other properties within the physical sciences. In contrast to the physical sciences, the full explanatory value of the attribution of mental dispositions can be grasped only in the context of our empathic capacities; or our disposition to engage with, resonate with, and imitate the mental life of other agents. For that very reason, I am tempted to speak of the doubly dispositional character of the attribution of mental properties to other agents. It attributes a tendency towards action to the other person in light of a uniquely human disposition of empathy. More pointedly, I am tempted to speak of beliefs and desire as empathic dispositions.

3. Folk Psychology and Physicalism: Some Suggestions for Thinking about the Autonomy of Folk Psychology Yet, what does the above argument about the centrality of empathy in folk psychology mean for the physicalist challenge toward higher order mental dispositions? I have already indicated in the last section – focusing on the folk psychological concepts that are tied to mechanisms of basic empathy – that it should lead us to be very careful in understanding the relationship between the ontological and epistemic components of physicalism. One should not expect that one can fully account for the application conditions of folk psychological concepts from the perspective of the detached perspective of the physical sciences, even if one accepts physicalism as an ontological thesis. In addition, the recognition of the centrality of reenactive empathy for folk psychology has implications for conceiving of folk psychology as an autonomous explanatory practice and for alleviating the eliminativist challenge. Given the

_____________ 28 29

See Stueber 2008b in this regard. If one feels the need for an empirical justification of this claim see Malle 2004. For an emphasis on other explanatory strategies besides belief/desire explanations see Andrews 2008.

Empathy, Mental Dispositions, and the Physicalist Challenge

273

fact that our physical theorizing is adopted from a detached perspective and does not engage our empathic capacities, there is no reason to assume that the explanatory game played in the physical sciences is in competition with our folk psychological practices or that it is able to fully absorb our folk psychological practice.30 Given its epistemic uniqueness it is thus to be expected that folk psychology will survive the advance of knowledge in the physical sciences since the knowledge game of the physical sciences is defined by very different rules. Since, as I said before, the existence of properties depends on their being appealed to in our final explanatory practices, one is tempted to conclude that dispositional mental properties exist, even if higher order dispositional properties such as fragility do not. Nevertheless, contemporary philosophers of mind might not be easily persuaded that the above suggestions have any bearing on the ontological significance of our folk psychological practices. They might be tempted to conclude that my argument at most proves that we are likely to continue to engage in folk psychological talk for the purpose of bonding with each other, not that it has any independent causal explanatory value in comparison with the explanations of the physical sciences. In particular, they are probably tempted to make that point by emphasizing some similarities between my strategy of countering physicalism and Davidson’s suggestion that “mental and physical predicates are not made for each other.”31 Like Davidson I also emphasize the centrality of rational agency in our folk psychological practices, even though in contrast to Davidson I argue for this essential difference by pointing to the central involvement of empathy in our understanding of rational agency, rather than by emphasizing its normative character. Davidson in my opinion does not sufficiently recognize the first personal or egocentric character of our interpretive practices and conceives of interpretation too much as a form of theory construction from the third person perspective.32 Despite those differences, one could however argue that my suggestion about the explanatory autonomy of folk psychology shares the same philosophical problems associated with Davidson’s position of anomalous monism. It is the general consensus in the philosophical community that Davidson is unable to save mental causation from the physicalist challenge because he is unable to explain how an event causes another event qua mental event. It seems that

_____________ 30

31 32

Furthermore, if one assumes that no law-like correlation between intentional states like belief and desires and properties that play an explanatory role in the physical sciences is to be found, given content externalism; it can also not be assumed that our folk psychological practice can be adopted merely in terminology sanctioned by the physical sciences. Davidson 1980, 218. Moreover, Davidson conceives of rationality as a notion that is defined by the principles of our best theories of rational inferences such as logic and probability theory. Any violation from those norms regardless of the reasons or causes of such violation would then count as a form of irrationality. For a different and contextualist understanding of rationality see Stueber 2006, chap.2.

274

Karsten Stueber

within the Davidsonian framework all the causal work is being done on the physical level.33 Given these worries, the physicalist challenge to the ontological status of higher order mental properties does not seem to be met. One still could be convinced that folk psychology is not a genuine causal explanatory practice, even if, for whatever pragmatic, aesthetic, or ethical/social reasons, we continue to talk about each other within its conceptual framework. From a causal-explanatory point of view, folk psychology thus would share the fate of theories like the humor theory of disease, which was eliminated by the advance of the medical sciences. Yet the physicalist challenge presupposes that folk psychological explanations explain the same type of events as physical explanations and it presupposes that they explain it causally in the same manner as physical explanations. Only under this assumption does it make sense to presuppose that various explanatory practices are in competition with each other and that one can compare their explanatory value. At a minimum, however the considerations in the second section of this paper imply that at least one of these assumptions is false. Folk psychology is a very different explanatory practice, or so one could argue, than the explanatory practices in the physical sciences because they are epistemically very differently structured. What counts as an explanation within this framework depends on our empathic capacities. It is for reasons like these that one might be tempted to follow Kim and Dray who suggest that folk psychological practices make use of notions of rational necessity and rational causation rather than notions of natural causation and “natural necessity” as the physical sciences.34 They do not causally explain behavior in the same sense of causation as the physical sciences. Rather, they show it to be caused in the sense of it being rationally compelling behavior, given an agent’s reasons for acting. Moreover, Kim suggests that folk psychological explanation explain “intentions and decisions” rather than physical actions.35 For this very reason, folk psychology is not an explanatory practice that is in epistemic competition with the explanatory practices of the physical sciences in the same manner that we do not conceive of the New England Patriots to be in competition with the New York Yankees or Manchester United. To regard x to be in be in competition with y, requires minimally that x and y are playing according to same rules. Consequently, we should not regard our folk psychological practices to be reductively explicable by the physical sciences.

_____________ 33 34 35

In this respect see the various articles (particularly the one by Kim) in Heil and Mele 1993. For my take on Davidson’s anomalous monism see the first sections of my 2005 and 1997. Dray 1957 and Kim 1984 and 1998b. Kim 1984, 319.

Empathy, Mental Dispositions, and the Physicalist Challenge

275

Even though I have sympathies for the claim that our folk psychological practices are causal explanatory in a different sense of causation than the physical sciences, I am not in favor of bifurcating our notion of causation for the following reasons: i.) Rational agents are embodied agents who interact in and with the physical world because of their decisions and actions. ii.) The notion causation is in my opinion closely tied to our ability to intervene with the world.36 Mental states like beliefs and desires are parameters that we know how to manipulate in order to change the behavior of various agents. Our educational institutions depend on such knowledge. In the same manner, knowledge of the physical sciences provides us with causal knowledge; that is, knowledge that allows us to intervene in the world by manipulating some of its parameters. iii.) An agent’s ability to make sense of his behavior in terms of reasons is an ability that seems to be central to his normal functioning as a human being. The inability to give at least some account of what we are doing and why we are doing it (why I am for example sitting in front of the computer writing those lines) would have to count as a severe case of mental deficiency.37 Given (i.) – (iii.) we do not seem to be justified in suggesting that folk psychological practices are using a notion of causation that is radically different from the notion appealed to in the physical sciences. I thus do not interpret the fact that our folk psychological practice differs epistemically from the physical sciences as suggesting that we use a different notion of causation in both explanatory practices. Rather I interpret it as suggesting that only folk psychology allows us to access some causally relevant properties of natural organism that are structured like us; properties that are however not accessible and definable from the detached perspective of the physical sciences. This view is in my opinion even more plausible if one recognizes – as I have also argued for in more detail elsewhere38 – that folk psychological practices are practices that have a slightly different explanandum than the physical movements of agents as described by the physical sciences. They describe the interaction of physical organisms in wider historical and social environments, which can only be described in terms of categories that radically criss-cross the categories of the physical sciences and that makes essential use of our folk psychological notions. Indeed the conceptual repertoire of folk psychology allows us to situate an agent within his wider environment and allows us to

_____________ 36 37 38

See also Woodward 2003 in this respect. See Stueber 2006, 45-46. Stueber 2005.

276

Karsten Stueber

understand his actions as rational responses to the demands of the environment as he conceives of it and that he finds himself in. As the argument in this article has revealed, the explanatory domain of folk psychology is, for this reason, circumscribed by our epistemic capacity of empathy. It is an explanatory domain that contains an irreducibly ego-centric moment; in which each investigator has to treat the object of investigation as a creature like himself; as a creature whose feelings, emotions, and reasons resonate in him. Rather than excluding this moment of subjectivity from our investigation of the world as is the case in the detached conception of the physical sciences, we should embrace it and philosophically recognize it as an essential element for the investigation of the world, or at least some aspects of it. In this sense one should follow an old advice of Aristotle and choose the ideal of objectivity according to its appropriateness for the specific domain of investigation. Otherwise one falls into the trap of a new idol, the ideal of false objectivity and the claim that the detached perspective of objectivity is adequate for all pursuits of knowledge in all domains of inquiry. Granted all of the above, the fact that the central categories of our folk psychological practices refer to what I have termed empathic dispositions, also allows us to philosophically safeguard those higher order dispositions from ontological extinction.

Literature Andrews, K. 2008. “It’s in your nature: A pluralistic folk psychology.” Synthese 165: 13-29. Chalmers, D. 1997. The Conscious Mind. Oxford: Oxford University Press. Chalmers, D. and F. Jackson. 2001. “Conceptual Analysis and Reductive Explanation.” Philosophical Review 110: 315-361. Churchland, P.M. 1989. “Eliminativism and the Propositional Attitudes.” In A Neurocomputational Perspective, 1-22. Cambridge, MA: MIT Press. Davidson, D. 1980. Essays on Action and Events. Oxford: Clarendon Press. DeWaal, Fr. 2006. Primates and Philosophers. Princeton: Princeton University Press. Dray, W. 1957. Laws and Explanation in History. Oxford: Clarendon Press. Heil, J. 2003. From an Ontological Point of View. Oxford: Oxford University Press. Heil, J. and Mele, A., eds. 1993: Mental Causation. Oxford: Clarendon Press. Hempel, C. 1965. Aspects of Scientific Explanations. New York: Free Press. Hutto, D. and M. Ratcliffe, eds. 2007. Folk Psychology Re-Assessed. Dordrecht: Springer. Iacoboni, M. 2008. Mirroring People. New York: Farrar, Straus and Giroux. Kim, J. 1984. “Self-Understanding and Rationalizing Explanations.” Philosophia Naturalis 21: 309-320. Kim, J. 1998a. Mind in a Physical World. Cambridge, MA: MIT Press. Kim, J. 1998b. “Reasons and the First Person.” In Human Action, Deliberation, and Causation, ed. J. Bransen and St. Cuypers, 67-87. Dordrecht: Kluwer Academic Publishers. Malle, B. F. 2004. How the Mind Explains Behavior: Folk Explanations, Meaning and Social Interaction. Cambridge, MA: MIT Press.

Empathy, Mental Dispositions, and the Physicalist Challenge

277

McKitrick, J. 2003. “The Bare Metaphysical Possibility of Bare Dispositions.” Philosophy and Phenomenological Research 66: 349-369 Mumford, S. 1998. Dispositions. Oxford: Oxford University Press. Poland, J. 1994: Physicalism: The Philosophical Foundations. Oxford: Clarendon Press. Prior, E., Pargetter, R., and Jackson, F. 1982. “Three Theses about Dispositions.” American Philosophical Quarterly 19: 251-257. Rizzolatti, G. and C. Sinigaglia. 2008. Mirrors in the Brain: How Our Minds Share Actions and Emotions. Oxford: Oxford University Press. Ryle, G. 1949. The Concept of Mind. London: Hutchinson. Sellars, W. 1963. Science, Perception, and Reality. New York: The Humanities Press. Salmon. W. 1989. “Four Decades of Scientific Explanation.” In Scientific Explanation. Minnesota Studies in the Philosophy of Science, vol. 13, ed. P. Kitcher and W. Salmon, 3-219. Minneapolis: University of Minnesota Press. Stich, S. 1983. From Folk-Psychology To Cognitive Science. Cambridge: MIT Press. Stueber, K. 1997. “Psychologische Erklärungen im Spannungsfeld des Interpretationismus und Reduktionismus.“ Die Philosophische Rundschau 44: 304-328. Stueber, K. 2005. “Mental Causation and the Paradox of Explanation.” Philosophical Studies: 243-277. Stueber, K. 2006: Rediscovering Empathy: Agency, Folk Psychology, and the Human Sciences. Cambridge, Mass.: MIT Press. Stueber, K. 2008a. “Empathy.” The Stanford Encyclopedia of Philosophy (Summer 2008 Edition), Edward N. Zalta, ed., URL = . Stueber, K. 2008b. “Reasons, Generalizations, Empathy, and Narratives: The Epistemic Structure of Action Explanation.” History and Theory 47: 31-43. Watson, J.B. 1924. Behaviorism. Chicago: University of Chicago Press.

Dispositional Knowledge-How versus Propositional Knowledge-That1 GREGOR DAMSCHEN

1. Introduction Pelé knows how to play soccer; Anne-Sophie Mutter knows how to play the violin; and Michael Schumacher knows how to drive a car. All three have performed these activities successfully, over a long span of time, on a professional level. These kinds of human activity give credibility to the widely held idea that persons who know how to perform an action possess a stable disposition that enables them to successfully perform this action under certain, suitable conditions. Such a stable disposition to perform an intentional action is a practical ability or skill of a person. If knowledge-how is an ability, and therefore a dispositional property, then there is a form of knowledge that consists of a relation between a person and a practical ability. There are, however, other forms of knowledge than knowledge-how; there is, for instance, knowledge-that, a form much investigated by philosophers since Plato. Knowledgethat, however, does not express any relation between a person and a practical ability, but rather a relation between an epistemic subject and a proposition. In regard to the above distinctions, two questions arise, which epistemologists have discussed for quite some time, at least ever since Gilbert Ryle attempted to sort out these issues. Question 1: Is it at all true that someone who knows how to do something is disposed to perform an action under favorable conditions? Is knowledge-how a practical ability? Or is knowledge-how rather a hidden knowledge-that, and therefore also a relation between an epistemic subject and a proposition? Within the last few years, an influential article by Stanley and Williamson, who defend the intellectualist thesis that every type of knowledge

_____________ 1

First and foremost I would like to thank Rainer Enskat and Eli Trautwein for extensive conversations and valuable comments on earlier drafts of this paper. Further thanks go to Dirk Effertz, Vittorio Hösle, Robert Schnepf, Dieter Schönecker, Karsten Stüber, and audiences in Wittenberg, Halle, Iowa, Notre Dame, Cologne, and Lucerne for helpful questions.

Dispositional Knowledge-How versus Propositional Knowledge-That

279

is a knowledge-that,2 has stimulated a debate. In order to answer the first question complex appropriately one has to take into account the semanticpragmatic aspects of the use and meanings of the term ‘knowledge-how’ and its grammatical variations as well as the nature of knowledge-how. Question 2: What is the connection between knowledge-how and knowledge-that? Is knowledge-how a form of knowledge in its own right, independent of knowledge-that, is it subordinate to knowledge-that, or is knowledge-that subordinate to it? Thus, the second question complex is aimed at determining the correlation between knowledge-how and knowledge-that; as well as the possible function that dispositional knowledge can have in the sphere of propositional knowledge. I will deal with both questions in the course of my paper. In the first part, I argue that the term ‘knowledge-how’ is an ambiguous term in a semanticpragmatic sense, blending two distinct meanings: ‘knowledge-how’ in the sense of knowledge-that, and ‘knowledge-how’ in the sense of an ability. In the second part of my paper, I construe five alternative ways of correlating knowledge-that and knowledge-how in the sense of an ability. I will discuss each of these alternatives and will then argue in favor of one of them. For this purpose, I will develop a reductio ad absurdum argument which is very different from the one that Ryle constructed.3 This argument will show that knowledge-how is not a species of knowledge-that but rather that knowledgethat is a species of knowledge-how. More specifically, dispositional knowledge-how is at the core of propositional knowledge-that and accordingly should be understood to be at the center of epistemology. The general intellectualist assumption that all knowledge-how is a knowledge-that is as false as its opposite, that is, the anti-intellectualist assumption that no knowledge-how is a knowledge-that. The truth lies rather in the middle: many forms of knowledge-how are propositional knowledge-that, but some forms of knowledge-how are purely dispositional abilities. Since the main point of my paper will be the proof that every knowledge-that presupposes a certain dispositional knowledge-how, I primarily understand my thesis as a solution to the problem of construing the functional role of dispositional knowledge-how in relation to propositional knowledge-that.

_____________ 2

3

Jason Stanley and Timothy Williamson (Stanley/Williamson 2001) have recently maintained, (taking their cue from Gilbert Ryle’s reflections on the ‘intellectualist legend’ [Ryle 1945/6 and 1949, 30-31]) that knowledge-how has been incorrectly understood as an independent form of knowing, more specifically as an ability, when it is, on the contrary, nothing but a species of knowledge-that. Ryle 1945/6 and Ryle 1949, 30-31.

280

Gregor Damschen

2. Two Kinds of Knowledge Syntactically the verb ‘to know’ in English permits very different sentential complements. One can, for example, know that such-and-such is the case. But one can also know where something is, when something occurred, who has something, what someone has, where someone comes from, where someone is going to, why, whatever … for, and how come something happens, how high or how long something is, how fast someone runs, and so on. These examples of knowledge sentences containing embedded questions seem to be very intimately related to knowledge-that, in that they seem, like knowledge-that, to express a relation between an epistemic agent and a proposition or fact.4 To briefly clarify this point, take the following two examples: If someone says, she knows where St. Peter’s Cathedral is located, she in fact is saying nothing more than that she knows that St. Peter’s is located at such-and-such a location. And if someone says, he knows, how tall the Eiffel Tower is, he is saying nothing more than he knows that the Eiffel Tower has this or that height. These two small transformations of a knowing-where- and a knowing-how-statement into a knowing-that-statement can be performed analogously with the just mentioned types of knowledge sentences. Has the common core of knowledge statements thus been found? Is every knowing a hidden knowledge-that? Things get more complicated if another sentential complement of ‘to know’ is taken into account: knowing how to do something. Although this expression, looked at syntactically, also contains the knowledge predicate and an embedded question, which begins with ‘how’ – like the expression ‘know how tall the Eiffel Tower is’ – there are cases in English where, in sentences of the type ‘I know how such-and-such is done,’ we do not want to express a relation between an epistemic agent and a proposition, but a relation of a conscious agent to a set of personal dispositions or abilities to successfully complete acts.5 If this non-propositional form of knowledge is admitted, a search for a uniform definition of the term ‘knowledge,’ which should be the goal of the theory of knowledge, has first to answer the above questions about the nature of knowledge-how. Thus, every search for a definition of knowledge must address the following two questions: first, what is knowledge-how and second, what is the connection between knowledge-how and knowledge-that?6 We

_____________ 4 5

6

Cf. Karttunen 1977. The Greek noun episteme, that gives epistemology its name ‘theory of knowledge,’ has a meaning in the sense of knowledge, how something is done, and in the sense of knowledge, that such-andsuch is the case. Cf. Liddell/Scott 1996, 660: “acquaintance with a matter, understanding, skill… 2. professional skill… II. generally knowledge… 2. scientific knowledge, science…” For Plato’s types of knowledge see Damschen 2003. Only recently has interest in these questions returned more intensively subsequent to Gilbert Ryles exemplary reflections in his “Knowing How and Knowing That” (Ryle 1945/6) and in his

Dispositional Knowledge-How versus Propositional Knowledge-That

281

will see that we can answer the second question without already having a comprehensive answer to the first question.

3. An Ambiguity in the Term ‘Knowledge-How’ There is a distinction between knowing how to do something and knowing that something is the case. This distinction is not merely syntactic; thus one cannot analyze it by analyzing the syntax of the embedded indirect questions in knowledge statements. The distinction I have in mind is a semanticpragmatic one. For ‘knowing how’ is specifically used to express the relation between an epistemic agent and a set of dispositions or abilities, and not the relation between an epistemic agent and a proposition. In order to decide whether a speaker’s use of a ‘knowing how’ phrase corresponds to our canonical explanation of the phrase (that is, expressing the relation of an agent to an ability) one should substitute the phrase “knowing how to do such-andsuch” with expressions such as “able to do such-and-such” or “disposed to do such-and-such.” If the speaker accepts the substitution, we can assume that he does not equate knowing how, here, with knowing that. The truth conditions for a sentence like “I know how one should play the violin” consist in the fact that the speaker, who makes this claim, possesses in fact the practical skill to play the violin and, under suitable conditions, plays the violin successfully. I call this type of non-propositional knowledge-how in the following ‘dispositional knowledge-how’. French, German, or Latin speakers have their own native syntactic construction available to form equivalents to the English dispositional knowledge-how phrase, ‘I know how to do this.’ In these three languages, the verb ‘to know’ is joined to an expanded infinitive clause: ‘Je sais faire quelque chose,’ ‘Ich weiß, das-und-das zu tun,’ ‘aliquid facere scio.’7 German speakers explicitly refer to a ‘Wissen-zu,’ [knowledge-to] when they mean knowledge-how in the sense of a practical ability. Since the aim of epistemology is to study knowledge in general and not merely knowledge constructions capable of being expressed in the knowledge predicate in English syntax, we must take syntactical information obtained from other languages very seriously. The above substitution test shows very clearly that, in English too, the expression ‘knowledge-how’ can refer to something which is not a knowledge-that.

_____________

7

The Concept of Mind (Ryle 1949). For example the former question was most recently analyzed by Hawley 2003 and Enskat 2005, and the latter by Stanley/Williamson 2001, Koethe 2002, Enskat 1998, 2003, 2005, Damschen 2005, Noë 2005, Hetherington 2006, Bengson/Moffett 2007, Lihoreau 2008, Williams 2008, and Adams 2009. Another approach which remains valuable is that of Polanyi 1958. Cf. Rumfitt 2003.

282

Gregor Damschen

Accordingly, there are modes of use of the verb ‘knowing-how’ by means of which the speaker expresses that he possesses a practical ability. The opponents of the view that there is knowledge-how which is a practical skill, however, do not yet admit defeat. They often propose a counter-argument like the following:8 Let us assume that a master violinist like Anne-Sophie Mutter knows how one should play the violin. She has sufficiently often played the violin in the past, so that we are entitled to make this ascription. Let us assume furthermore that she lost both hands in a bad accident. Then we have the case of a violinist who does know how one should play the violin but is actually unable to play the violin. According to the opponents of dispositional knowledge-how, this case clearly shows that to know how one performs an action is not the same as to have an ability to perform the action, for there are cases in which someone does know how one performs a certain action, but does not possess the corresponding ability. I doubt that this example or similar examples of this type support the view that knowledge-how is not the same as having an ability. Of course, there may be the case where Anne-Sophie-Mutter knows how to play the violin and nevertheless is not able to play the violin at the moment, because she is sleeping at the moment or because no violin is available, although she would like to play it. Nobody, however, would therefore claim that AnneSophie Mutter, in these periods of time, lost her ability to play the violin, and then regains it suddenly by means of a miraculous act when she wakes up or a violin appears. Neither of these two cases shows in any way that to know how one should play the violin is not the ability to play the violin. The same considerations also apply to the case where Anne-Sophie Mutter loses both hands: a situation that is logically comparable to the case where no violin is available at the moment. In both cases does she know how to play the violin and possesses an ability to play the violin without being able to play the violin at that very moment. The realization of an ability always presupposes certain conditions that must be satisfied so that the ability can be carried out successfully. These conditions can be external conditions that do not have to do with the bearer of the ability, e.g. the lack of a violin when one wants to play the violin. However, these conditions also can be conditions that are connected with the body of the bearer of an ability, e.g. if a violinist has lost both hands or, to assume a less drastic case, has perhaps such a strong influenza that she cannot hold the violin.9 Practical knowledge and practical abilities have, therefore, four important properties:10

_____________ 8 9 10

Cf. Stanley/Williamson 2001, 416. Cf. Noë 2005, 282-283. Cf. Noë 2005, 284-286.

Dispositional Knowledge-How versus Propositional Knowledge-That

283

1. They are always the knowledge of someone, i. e., they are the knowledge of a person. In this regard, one can call them personal knowledge. For practical abilities do not exist independently, but require a bearer who is able to perform intentional actions. 2. The second property of practical knowledge is connected directly with its first property. Normal abilities are embodied. I presume that persons who we know have a body. I do not want to commit myself to the view that there are no persons without bodies. But it is true anyway that the cases of knowledgehow, which we look at, are always cases where the bearer of a knowledge-how is a person who has a body. The acquisition of practical skills changes our body and we perform our skills by means of our body. 3. The third property of practical knowledge is that it is situated. Dispositional knowledge can only be realized if certain external conditions are fulfilled. Pelé cannot play soccer if no football is available, and Michael Schumacher cannot drive any formula-one car if its tires are slashed. Basically, every proposition like ‘Somebody S knows how one X-s’ has to be completed in the following way: ‘Somebody S knows how to X under circumstances C.’ 4. The fourth property of practical knowledge is to be disposed to perform an action successfully. If Anne-Sophie Mutter cannot successfully play the violin or if Michael Schumacher cannot successfully compete in a car race, one could neither say of Mutter that she knows how one should play the violin nor of Schumacher that he knows how one should race cars. So, knowledge-how is a personal, embodied and situated ability to successfully realize an action. Our analysis thus has shown that there are manifestations of ‘knowledge-how’ that are instantiated when someone possesses a practical ability. The above canonical instances of ‘knowing how’ (that is, knowing how as an ability) are, however, not the whole story. For, confusingly, in English and in the other languages mentioned, there is a second, completely different use of the expression ‘knowledge-how’ with which a relation between an epistemic agent and a proposition can, indeed, be expressed. Again, our semantic test is quite simple. One must ask the speaker who says ‘I know how one does such-and-such’ whether the statement he makes can be rendered straightforwardly and without distortion as ‘I know, that one must do suchand-such in order to act in this or that manner.’ If he affirms this, then he wants to say that he is the carrier of a propositional knowledge-how. Furthermore, if the speaker is able to list a sequence of appropriate steps to act in this or that manner he is indeed a carrier of such propositional knowledge-how. Thus, depending on the intention of the speaker, sentences such as (1) I know how to drive a car

284

Gregor Damschen

can have firstly (1a) a non-propositional and dispositional meaning, e.g. as the answer of an active driver to the question of whether he has the ability to drive a car successfully. Yet, secondly the sentence can also have (1b) a propositional meaning, e.g. as the answer of a driving instructor to the question whether he can articulate in propositional form the actions necessary to control a car, the relevant theoretical knowledge, and the most important rules of the road. Finally the sentence can thirdly (1c) have both meanings simultaneously, i.e. the dispositional and the propositional, e.g. as the answer of a driving instructor to the question whether he himself is able to drive a car and whether he also can communicate to his students in propositional form everything that one has to do in the process.11 The last sense of knowledge-how in the sense of a hybrid between dispositional and propositional knowledge-how is the meaning that one comes across most frequently as David Lewis accurately observed: “It would be feeble, I think, just to say that we’re fooled by the ambiguity of the word ‘know’: we confuse ability with information because we confuse knowledge in the sense of knowing-how with knowledge in the sense of knowing-that. There may be two senses of the word ‘know,’ but they are well and truly entangled. They mark the two pure endpoints of a range of mixed cases. The usual thing is that we gain information and ability together. If so, it should be no surprise if we apply to pure cases of gaining ability, or to pure cases of gaining information, the same word ‘know’ that we apply to all the mixed cases.”12 Thus, we have three instances of knowledge-how. To keep them easily apart in the following, I will use subscripts to distinguish between them: (a) knowledge-howd, that is dispositional and non-propositional knowing how; (b) knowledge-howp, that is knowing how that can be articulated in terms of propositions; (c) knowledge-howh, that is knowing how as hybrid. It was no doubt the propositional use of knowledge-howp or knowledge-howh which led philosophers to their general claim that every knowledge-how is a

_____________ 11

12

To formulate this insight more precisely I give the relevant truth conditions for each of the three cases: (1a) (1) is true iff the speaker of the sentence (1) has the personal ability to successfully drive a car. (1b) (1) is true iff the speaker of the sentence (1) explicitly knows that one must do such-andsuch to drive a car. (1c) (1) is true iff both (1a) and (1b) are met. Lewis 1999, here p. 289.

Dispositional Knowledge-How versus Propositional Knowledge-That

285

knowledge-that.13 For the purpose of this paper it is only important to recognize that the dispositional form of knowledge-how in the sense of a practical ability (knowledge-howd) exists.14

4. Five Alternative Answers Asking whether and how knowledge-how in the dispositional sense (knowlege-howd) and knowlege-that in the propositional sense are semantically related yields the following five alternatives: (A1)Knowledge-howd is a species of knowledge-that, i.e.: knowledge-that is a necessary condition for knowledge-howd. (A2)Knowledge-that is a species of knowledge-howd, i.e.: knowledge-howd is a necessary condition for knowledge-that. (A3)Knowledge-howd and knowledge-that are extensionally equivalent. (A4)Knowledge-howd and knowledge-that have a common intersection. (A5)Knowledge-howd and knowledge-that are completely distinct. These five alternative ways of thinking about the relation between knowledgehow and knowledge-that correspond to alternative strategies in looking for a definition of knowledge.15 (S1) If we think of all forms of knowledge as propositional knowledge, we must be convinced of having found a good argument for alternative A1.16 (S2) On the other hand, if we opt for alternative A2, we must place the search for a definition of knowledge-how at the center of our epistemological reflections; knowledge-that would then only be a spe-

_____________ 13 14 15

16

E.g. Hintikka 1975, or Stanley/Williamson 2001. Cf. on this also Ryle 1949, Rumfitt 2003, Rosefeldt 2004, Noë 2005, 284, Lihoreau 2008, and Williams 2008. Asking whether and how knowledge-how in the propositional sense (knowledge-howp) and knowledge-that in the propositional sense are semantically related yields at least the following alternatives: (B1) Knowledge-howp is a species of knowledge-that. (B2) Knowledge-that is a species of knowledge-howp. (B3) Knowledge-howp and knowledge-that are extensionally equivalent. (B4) Knowledge-howp and knowledge-that have a common intersection. (B5) Knowledge-howp and knowledge-that are completely distinct. Although alternative B1 seems to be the right answer, deciding this claim is not the subject of this paper. Cf. e.g. Hintikka 1975, 3: “… all of the different constructions in terms of the verb ‘know’ can be reduced … to the sense in which the nature of knowledge as a propositional attitude is most explicit … the ‘knowing-that’.”

286

Gregor Damschen

cies or subcategory of knowledge-how.17 (S3) If we choose alternative A3, we must stipulate that an extensional definition of knowledge-that be identical with an extensional definition of knowledge-how. (S4) Alternative A4 presents us with a non-uniform definition for the term ‘knowledge’ encompassing three equally valid and distinct definitions of a form of pure knowledgethat, a form of pure knowledge-how and a mixed form of the two. (S5) Finally, alternative A5 like A4 presents us with a non-uniform definition of the term ‘knowledge’ entailing two, rather than three, equally valid and distinct definitions of knowledge-how and knowledge-that. Most epistemologists, without always stating it explicitly or perhaps having consciously thought about it, follow strategy S1 or S5. But no matter, our reflections up to this point have shown that those who would like to provide a definition of knowledge can only begin their work in a methodologically controlled manner if they have found at least one good argument for one of the five listed alternatives and, at the same time, good arguments against the remaining four alternatives. Unfortunately there have been only few arguments which could make the decision between these five alternatives easier. The most familiar argument against alternative A1 is Gilbert Ryle’s argument against what he calls the ‘intellectualist legend’.18 As we have seen, Ryle’s “intellectualist legend” argument against A1 has recently been disputed by Stanley and Williamson.19 However, their argument for A1 is not satisfactory.20 For it is mainly based on unconvincing linguistic reflections regarding the syntactic relatedness of knowledge-how and knowledge-that sentence constructions in the English language,21 which cannot support a broader metaphysical project of finding a general definition of knowledge.22 For alternative A5 there have been weakly plausible arguments, genealogically rooted in Ryle’s original argument.23 There have also been a few arguments for saying – vide alternative A2 – that some-

_____________ 17

18 19 20 21 22 23

Cf. e.g. Colin McGinn [McGinn 1984], here p. 529: “And once we take on the responsibility of confronting the whole family of knowledge locutions, it is by no means guaranteed that propositional knowledge will emerge as fundamental: perhaps the core notion will attach most directly to some other locution, so that knowledge-that comes out as a species of some more basic type of knowledge.” Ryle 1945/6 and Ryle 1949, 30-31. Stanley/Williamson 2001, 412-417. Whether their reconstruction in fact grasps Ryle’s original argument will have to remain untreated here. Cf. on this Rosefeldt 2004 and Noë 2005. See Stanley/Williamson 2001, 417-444. See Rumfitt 2003. Furthermore their concept of knowledge-how as a relation between an epistemic agent and practical propositions leads to unacceptable consequences. See Koethe 2002 and Schiffer 2002. Cf. e.g. Carr 1979; Carr 1981; Devitt 1996, 52; Putnam 1996, xvi; Lewis 1999, 288.

Dispositional Knowledge-How versus Propositional Knowledge-That

287

one who knows that p has the capacity to state correctly that p.24 However, these claims fail since we can imagine someone who, stating that p, is really just guessing that p is the case. If, by chance, p is actually the case, he correctly states that p and a fortiori he has the capacity to state correctly that p, but we would not say that he knows that p. Moreover, if having a capacity to state p correctly involves reasons for p, these reasons will contain propositional knowledge-that. Thus, any reducing of knowledge-that to knowledge-how will prove to be circular.25 I have found no explicit arguments for assumptions A3 or A4 in the philosophical literature. I will now introduce a new argument against A1 which, in a very broad sense, follows in the tradition of Ryle’s original argument. I, however, will make use of very different premises. I try to accomplish two objectives with my argument: First, as a reductio ad absurdum it shows that the claim that dispositional knowledge-howd is a species of knowledge-that is untenable; second, using the first and the last of its premises we can construct a hypothetical syllogism that shows that knowledge-that is a subcategory of dispositional knowledge-howd.

5. The Species-of-Knowledge Argument The argument against alternative A1, which I call the species-of-knowledge argument, leads formally speaking to a reductio ad absurdum. It thus includes premise A1, which will be refuted, and two other premises P1 and P2:

(P1) For all epistemic agents S, for all intentional acts F: If S regularly and successfully completes the intentional act F (which S completed earlier sufficiently often, regularly and successfully), then S knows how to complete act F. (A1)For all epistemic agents S, for all intentional acts F, there is at least one proposition T: If S knows how to complete act F, S knows that T(F). (P2) For all epistemic agents S, for all propositions T, there is at least one intentional analysis act U: If S knows that T, then S completed the intentional analysis act U sufficiently often, regularly, and successfully, whether T is the case, or whether not-T is the case.

_____________ 24 25

Hartland-Swann 1956 and 1957, Roland 1958. For a new argument for assumption A2 see now also Hetherington 2006. See the critique on Hartland-Swann by Ammerman 1957.

288

Gregor Damschen

Let us start with premise A1 that in the end will be refuted with the help of the species-of-knowledge argument. According to A1, any knowledge how act F is completed can be traced back to the attitude of knowledge in relation to the contents of a specific practical proposition T(F). This practical proposition T(F) states, that such-and-such must be done to complete act F. Accordingly, knowing how to do F would be nothing but to know that such-andsuch must be done in order to do F. But this would imply that every knowledge-how would be nothing but a kind of knowledge-that. Premise P1 states a sufficient, not a necessary condition for dispositional knowledge-howd. Even though a comprehensive definition of knowledgehowd continues to be a desideratum,26 for the purpose of the current argument it is not necessary to have such a definition at hand. Premise P1 can be made plausible as follows: it can be straightforwardly admitted that someone who was sufficiently often engaged regularly and successfully in an intentional act F, also has the ability to complete act F. The practice of a regular and successful act is a sufficient condition for a person having the corresponding ability. The condition that it must involve an intentional, i.e. intentional and consciously planned act, rules out activities performed by robots or nonpersonal entities. For none of us would say that a robot in a strict sense ‘knows’d how to do such-and-such. It may certainly have a disposition towards doing something very specific, but a knowledge-howd seems to be more than a mere disposition to do something. It seems also to involve at least a relation of the agent to himself, his disposition, and his conscious act. For our earlier analysis has suggested that knowledge-how is a personal, embodied and situated ability to successfully realize an action. However, the reason premise P1 contains the explicit condition that only an adequately large number and only a successful realization of intentional actions can be sufficient for the ascription of a knowledge-how is a response to the objection of some philosophers (most notably, Stanley and Williamson) that the fact that someone performs a non-intentional action does not imply that he has a knowledge-how. Stanley and Williamson discuss the following example: (2) If Hannah digests food, she knows how to digest food.27 And they assert that the proposition is false for the following reasons: “Digesting food is not the sort of action one knows how to do.” I agree in a certain way: Indeed, Hannah cannot know how to digest food. But the reason for that is not (as Stanley and Williamson assert) that digesting food is a spe-

_____________ 26 27

See Hawley 2003 and Williams 2008. Stanley/Williamson 2001, 414.

Dispositional Knowledge-How versus Propositional Knowledge-That

289

cial and mysterious sort of non-intentional action in Hannah for which a corresponding knowledge-how does not exist. The reason is simply the fact that digesting food in Hannah is not an action that Hannah performs or could perform at all: neither a ‘non-intentional action’ (whatever this may be) nor an intentional action. It is not Hannah who digests food but her digestive system, and in fact in her and for her.28 Neither digesting nor breathing nor the process of metabolism is something a person can actively do. These processes take place in and for persons, but they are not performed by persons as actions. In this regard, the funny thing about the proposition “If Hannah digests food, she knows how to digest food” seems to be that it is true, for the antecedent of the proposition is false, because it is not Hannah who digests the food. There are additional examples of this type, such as the one in which Hannah buys lottery tickets and wins. The proposition that is supposed to show that there are actions that do not imply any knowledge-how is as follows: (3) If Hannah wins a fair lottery, she still does not know how to win the lottery, since it was by sheer chance that she did so. Admittedly, there is probably no knowledge how one wins the lottery, but Hannah here does not perform an action which is sufficient for a knowledgehow. For to win the lottery is not an action Hannah could perform. Hannah is able to perform many actions: She can go to the lottery shop, she can buy a lottery ticket. However, winning the lottery is not an action Hannah can actively perform, but rather an event that happens to Hannah. In this respect, this example does not show that there are actions that do not presuppose any knowledge-how, because it fails to meet the condition that the actions should be (intentional) actions of a person.29 Of course, there are actions which do not imply knowledge-how. These are frequently actions that are performed in acquiring a new ability. For example, the sentence “Arma virumque cano Troiae qui primus ab oris Italiam fato profugus Laviniaque venit litora” could be correctly pronounced purely by chance by the beginning Latin student. The action is thus successfully performed from the perspective of a third person, but perhaps the lucky Latin student cannot consistently reproduce his correct pronunciation. But if the speaker successfully and correctly pronounces the sentence “Arma virumque cano Troiae qui primus ab oris…” – not only on a single day, but for a longer period of time, not only in a single situation, but in various situations, not only this sentence of the language, but also additional sentences – then we

_____________ 28 29

Cf. Noë 2005, 279. Cf. Noë 2005, 280.

290

Gregor Damschen

can assume that he has acquired the ability of correctly pronouncing this sentence. Given a correspondingly great number of successful performances of correctly pronouncing the sentence, we will be inclined to ascribe a corresponding ability to the speaker. Premise P1 does not say more; but it also does not say less. Premise P1 should, however, not be understood as saying that there are only abilities if they are performed regularly. It is certainly imaginable that there are certain abilities which under favorable conditions, if at all, can only be performed once in a lifetime (for example, being courageous when saving someone’s life). The premise P1, moreover, is implicitly accepted by Stanley and Williamson.30 They accept that performing an action F implies that the person who performs it also knows how one performs the action, but only if the actions in question are intentional actions. So Stanley and Williamson accept premise P1, although, unlike myself, they accept the truth of the premise A1. The acceptance of P1 thus appears to be independent of the acceptance or rejection of A1.

6. Premise P2: Knowledge-That presupposes some Kind of Analysis Premise P2 formulates a necessary condition for knowledge-that: whoever knows that p is the case has also analyzed whether p or not-p is the case. I think that this premise states succinctly our epistemic basic intuition expressed by the Gettier cases that we would not consider a belief that was merely accidentally true to be knowledge.31 But how does one advance from merely accidentally true belief to a non-accidentally true belief? To arrive at a non-accidentally true belief that p is the case, we must firstly analyze in a specific manner whether p is true or not-p is true.32 This means that we must complete some kind of analysis program. This analysis program contains a set of analysis acts. But if we want to complete this analysis program successfully we must be able to distinguish truth from falsehood at each step of the program. For this purpose we must have an ability, a “truth-discriminating capacity.”33

_____________ 30 31 32 33

Stanley/Williamson 2001, 414-415. Cf. Gettier 1963. Cf. McGinn 1984, 536 ff. McGinn 1984, 538. Cf. Lehrer 1990, 5: “This kind of knowledge (sc. knowledge-that; author) rests on our capacity to distinguish truth from error” (my italics). For a concept of a truth discriminating capacity in Plato’s Theaetetus see Gonzalez’ paper in this volume.

Dispositional Knowledge-How versus Propositional Knowledge-That

291

Colin McGinn, who (in accord with Austin and Goldman) represents a special version of a reliability theory of knowledge, articulates the idea of this type of truth-discriminating capacity as follows: “The guiding idea of the theory is simple: to say that a person S is globally reliable with respect to a range of propositions is to say that S can discriminate truth from falsehood within that range of propositions; global reliability is a capacity to tell the difference between true propositions and false ones within some given class of propositions. We then say that S knows that p just if his (true) belief that p is acquired by the exercise of a capacity to discriminate truth from falsehood within some relevant class R of propositions.”34 This ability has to be understood as being realized in very many forms, since its object ranges over the spectrum of all propositions. If we examine this ability more closely, we can see that it comprises some or all of the following five sub-abilities: 1. the ability to pose questions and to recognize possible answers to these questions as appropriate while others are rejected as inappropriate (question-based knowledge), 2. the ability to use categories appropriately (knowledge of categories), 3. the ability to make judgments (judgment knowledge), 4. the ability to complete thinking acts of a more complex nature (ability of connecting judgments), and 5. being able to complete physical acts (practical knowledge). The mere fact of being able to distinguish true propositions from false propositions, however, does not suffice to ascribe a specific knowledge-that to someone. It also involves him/her not only having this ability in the relevant case but also him/her having in fact successfully realized it. To know that p is the case one, therefore, must in fact have found out that p is the case.35 Before we can know that p is true, we must first successfully complete the truth analysis program with our attention on p. This successfully completed analysis program which produces the result that either p or not-p is true, is thus the smallest common denominator of what must at least be added to a true belief if it is not to be merely accidental. The exact nature of the analysis program and the knowledge-howd that is required for it depends, however, on the situation: the truth analysis can consist in personally traveling to Rome and going to St. Peter’s Square if one wants to know whether it is really true that St. Peter’s is in Rome, or alternatively – to take Goldman’s well-known but extremely artificial example – examining the object that looks like a barn from a distance from the inside if one wants to distinguish fake barns from real barns.36 But such intentional

_____________ 34 35

36

McGinn 1984, 536-537. Cf. Ryle 1971 (1945/6), 224: “To know a truth I must have discovered … it”; Clark 1963, 48: “Knowing implies having found out.” Enskat 1998 and 2005 discusses extensively the idea that propositional knowing-that rests on a dispositional knowing-how. Cf. Goldman 1992, 86 ff.

292

Gregor Damschen

analysis acts are necessary as part of an analysis program not only in the field of empirical knowledge, but also in regard to non-empirical knowledge. For example, if one wants to know whether 5 + 7 = 12, the relevant analysis program consists in calculating the mathematical equation sufficiently often and successfully oneself in an appropriate way.37

7. The Reductio ad absurdum How and why do the three premises A1, P1, and P2 together yield a reductio ad absurdum? If premise A1 is plausible and true then there should be no case in which A1 together with other plausible and true premises leads to absurd consequences. Let us take as an example the special case of an act F, e.g. that Albert Einstein walks from his house to his office in Princeton, implying that he also successfully arrives and has walked sufficiently often and successfully in the past from his house to his office. In addition, the following two assumptions are made: Z1: Einstein’s walk from his house to his office is an intentional act F1, Z2: For all x: If x is an analysis act U, with whose help one can analyze whether T or not-T, then x is an intentional act F. Then, assuming the truth of the three premises A1, P1 and P2, the following vicious regress results: (1) Einstein walks from his house to his office. (2) Einstein F1-s. (3) Einstein knows how to F1. (4 Einstein knows that T(F1). (5) Einstein completed analysis act U1. (6) Einstein F2-s. (7) Einstein knows how to F2. (8) Einstein knows that T(F2). (9) Einstein completed analysis act U2. (10) Einstein F3-s. and so on and so forth.

[assumption] [1, Z1] [2, P1] [3, A1] [4, P2] [5, Z2] [6, P1] [7, A1] [8, P2] [9, Z2]

Thus if premises P1 and P2 are accepted and it is also assumed that A1 is true, the result is a vicious regress so that the act F1 never could have oc-

_____________ 37

The premise P2 is at least implicitly contained in internalist interpretations of the classic third condition of knowledge, the justification condition.

Dispositional Knowledge-How versus Propositional Knowledge-That

293

curred. But that is absurd, for according to our assumption Einstein in fact does walk from his house to his office. Since we presuppose that premises P1 and P2 are true, premise A1 must therefore be false. Consequently, it is not the case that dispositional knowledge-howd is a species of knowledge-that.

8. Knowledge-That: a Species of Knowledge-Howd If A1 is false, A2, A4 or A5 can still be true (except A3, which is false if A1 is false). But someone, who holds premises P1 and P2 to be true, must also hold A2 to be true. For P2 states that someone who knows that p is the case has also analyzed whether p or not-p is the case sufficiently often, regularly, successfully and intentionally. And P1 states that someone who regularly and successfully completes the intentional act F (which he earlier has sufficiently often, regularly and successfully completed) also knowsd how act F is completed. If the analysis of whether p or not-p is the case is itself an intentional act F (let’s call this premise G), then this act presupposes a knowledge-howd. All of this leads to the following hypothetical syllogism: (P2) If S knows that p, then S has completed sufficiently often, regularly and successfully the intentional analysis act U, whether T is the case, or whether not-T is the case. [A B] (G) If S has completed sufficiently often, regularly and successfully the intentional analysis act U, whether T is the case, or whether not-T is the case, then S has completed sufficiently often, regularly and successfully an intentional act F(p). [B C] (P1) If S has completed sufficiently often, regularly and successfully an intentional act F(p), then S knowsd how to complete an intentional act F(p). [C D] Therefore, (A2) If S knows that p, then S knowsd how to complete an intentional act F(p). [A D] Hence, someone who holds P1, P2 and G true must also hold A2 true. But if A2 is true, dispositional knowledge-howd and propositional knowledge-that have neither an intersection nor are they completely distinct. As a consequence, A4 and A5 are false. Hence A1, A3, A4 and A5 are false, A2 is true. Dispositional knowledge-howd is a necessary condition for propositional knowledge-that, and propositional knowledge-that is a species or subcategory of dispositional knowledge-howd.

294

Gregor Damschen

9. Conclusion Let us summarize the results of my considerations: It has been shown that there are two types of knowledge-how: the first kind depends on propositional knowledge-that (knowledge-howp), and the second kind is a nonpropositional form of knowledge-how in the sense of a disposition or ability (knowledge-howd). Moreover, the species-of-knowledge argument helped to demonstrate that dispositional and non-propositional knowledge-howd, and not knowledge-that, is basic for our concept of knowledge. Examined more carefully, it was also shown that the dispositional knowledge-howd is a necessary condition for knowledge-that, hence knowledge-that is a species of the dispositional knowledge-howd. And finally, it has been shown that the relation between both specified forms of knowledge can be determined without presupposing in advance a complete and thoroughly accepted definition of propositional knowledge-that or dispositional knowledge-howd.

Literature Adams, Marcus P. 2009. “Empirical Evidence and the Knowledge-That/Knowledge-How Distinction.” Synthese 170: 97-114. Ammerman, Robert. 1957. “A Note on ‘Knowing That’.” Analysis 17: 30-32. Bengson, John, and Moffett, Marc A. 2007. “Know How and Concept Possession.” Philosophical Studies 136: 31-57. Carr, David. 1979. “The Logic of Knowing How and Ability.” Mind 88: 394-409. Carr, David. 1981. “Knowledge in Practice.” American Philosophical Quarterly 18: 53-61. Clark, Michael. 1963. “Knowledge and Grounds: A Comment on Mr. Gettier’s Paper.” Analysis 24: 46-48. Damschen, Gregor. 2003. “Grenzen des Gesprächs über Ideen. Die Formen des Wissens und die Notwendigkeit der Ideen in Platons Parmenides.” In Platon und Aristoteles – sub ratione veritatis. Festschrift für Wolfgang Wieland zum 70. Geburtstag, ed. Gregor Damschen, Rainer Enskat, and Alejandro G. Vigo, 31-75. Vandenhoeck & Ruprecht: Göttingen. Damschen, Gregor. 2005. “Ist Wissen-dass eine Unterart von Wissen-wie?” [Is Knowledge-That a Species of Knowledge-How?]. In Philosophische Perspektiven. Beiträge zum VII. Internationalen Kongress der Österreichischen Gesellschaft für Philosophie, ed. O. Neumaier, C. Sedmak, and M. Zichy, 290-295. Frankfurt am Main / Lancaster: ontos. Devitt, Michael. 1996. Coming to Our Senses. New York: Cambridge University Press. Enskat, Rainer. 1998. “Authentisches Wissen. Was die Erkenntnistheorie beim Platonischen Sokrates lernen kann.” In Amicus Plato magis amica veritas. Festschrift für Wolfgang Wieland zum 65. Geburtstag, ed. R. Enskat, 101-143. Berlin: de Gruyter. Enskat, Rainer. 2003. “Ist Wissen der paradoxe epistemische Fall von Wahrheit ohne Wissen? Platon, Gettier, Sartwell und die Folgen.” Zeitschrift für philosophische Forschung 57: 431-445. Enskat, Rainer. 2005. Authentisches Wissen. Prolegomena zur Erkenntnistheorie in praktischer Hinsicht. Göttingen: Vandenhoeck & Ruprecht.

Dispositional Knowledge-How versus Propositional Knowledge-That

295

Gettier, Edmund. 1963. “Is Justified True Belief Knowledge?” Analysis 23: 121-123. Goldman, Alvin. 1992. Liaisons. Philosophy Meets the Cognitive and Social Sciences. Cambridge/Mass. Hartland-Swann, John. 1956. “The Logical Status of ‘Knowing That’.” Analysis 16: 111115. Hartland-Swann, John. 1957. “Knowing That – A Reply to Mr Ammermann.” Analysis 17: 69-71. Hawley, Katherine. 2003. “Success and Knowledge-How.” American Philosophical Quarterly 40: 19-31. Hetherington, Stephen. 2006. “How to Know (that Knowledge-That is KnowledgeHow).” In Epistemology Futures, ed. S. Hetherington, 71-94. Oxford: Oxford University Press. Hintikka, Jaakko. 1975. “Different Constructions in Terms of the Basic Epistemological Verbs. A Survey of Some Problems and Proposals.” In J. Hintikka, The Intentions of Intentionality and Other New Models for Modalities. Dordrecht: Reidel: 1-25. Karttunen, L. 1977. “Syntax and Semantics of Questions.” Linguistics and Philosophy 1: 3-44. Koethe, John. 2002. “Stanley and Williamson on Knowing How.” The Journal of Philosophy 99: 325-328. Lehrer, Keith. 1990. Theory of Knowledge. Boulder: Westview Press. Lewis, David. 1999. “What Experience Teaches”, In D. Lewis, Papers in Metaphysics and Epistemology. Cambridge: Cambridge University Press: 262-290. Liddell, Henry George, and Robert Scott. 1996. A Greek-English Lexicon, with a revised supplement. Oxford: Clarendon Press. Lihoreau, Franck. 2008. “Knowledge-How and Ability.” Grazer Philosophische Studien 77 (1): 263-305. McGinn, Colin. 1984. “The Concept of Knowledge.” Midwest Studies in Philosophy 9: 529554. Noë, Alva. 2005. “Against Intellectualism.” Analysis 65: 278-290. Polanyi, Michael. 1958. Personal Knowledge. Towards a Post-Critical Philosophy (corrected edition 1962). Chicago: University Press. Putnam, Hilary. 1996. “Introduction.” In The Twin Earth Chronicles, ed. A. Pessin, and S. Goldberg. Armonk, NY: Sharpe. Roland, Jane. 1958. “On ‘Knowing How’ and ‘Knowing That’.” Philosophical Review 67: 379388. Rosefeldt, Tobias. 2004. “Is Knowing-how Simply a Case of Knowing-that?” Philosophical Investigations 27: 370-379. Rumfitt, Ian. 2003. “Savoir faire.” The Journal of Philosophy 100: 158-164. Ryle, Gilbert. 1945/6. “Knowing How and Knowing That.” Proceedings of the Aristotelian Society 46, 1945-46: 1-16 (also in: G. Ryle. Collected Papers, vol. 2, Collected Essays 19291968. London 1971: 212-225). Ryle, Gilbert. 1949. The Concept of Mind. Chicago: University Press. Schiffer, Steven. 2002. “Amazing Knowledge.” The Journal of Philosophy 99: 200-202. Stanley, Jason, and Williamson, Timothy. 2001. “Knowing How.” The Journal of Philosophy 98: 411-444. Williams, John N. 2008. “Propositional Knowledge and Know-How.” Synthese 165: 107125.

Epistemic Virtues and Cognitive Dispositions DAVID HENDERSON AND TERRY HORGAN

1. Orientation The preponderance of the discussions in this volume make a case for positing dispositions, either by attention to the needs of some contemporary discipline, or by way of reflection on some historical thinker’s insight, or by unfiltered metaphysical argument. The present piece does not argue for positing dispositions, and rather flatly presupposes that there are dispositions. Our concern is with an epistemological point – one concerning the range of dispositions (or cognitive processes understood dispositionally) that are significant for epistemology. Most contemporary epistemologists are committed to the epistemic significance of cognitive dispositions. They may write instead of beliefs being the result of appropriate processes – but the processes in question are understood functionally, as dispositions to certain kinds of transitions. They are concerned with what are appropriate processes – functionally understood – by which to fix beliefs, and are thereby committed to dispositions. However, they tend to be concerned with a narrowly circumscribed set of dispositions – too narrowly circumscribed, we argue. To appreciate this very standard epistemological commitment to dispositions, reflect on the distinction between an agent’s being objectively justified in a belief and that agent’s merely having justification for that belief. For an agent to be objectively justified in a belief requires that the processes by which that belief was formed be suitably responsive to the reasons (or information) possessed – it requires that, prompted by possessed information, the belief in question be brought about “in the right way.” One can then have justification for a belief (having adequate reasons or supporting information) and yet not thereby be justified in holding that belief. For example, it is possible that an agent have adequate justification (possess information with a content significantly supportive of the content of the agent’s belief) and yet would have held the belief even were that information not to have been possessed. To be justified, a belief must be dependent on the contentfully supporting information that the agent possessed – so that, were the agent not to have possessed that in-

Epistemic Virtues and Cognitive Dispositions

297

formation, the agent would not have formed or retained the belief in question. Yet, even this does not quite ensure that the agent is objectively epistemically justified in the belief. For example, an agent might have significant justification (supporting information) for a belief that P, and this information may serve to raise the question of whether of not P. This might then trigger wishful thinking yielding the belief that P – wishful thinking that would (once the question was somehow raised) yield the belief that P even were the agent not to possess the justification in question. The agent’s belief is then counterfactually dependent on this “trigger,” but is not thereby objectively justified. Being objectively justified requires a richer array of counterfactual dependencies. The agent’s cognitive processes must exhibit patterns of counterfactual dependencies to ranges of relevant information. They must be describable in terms of “invariant generalizations”, so that the resultant belief is caused in the right way – in a way that would be sensitive to various information that the agent has or might have.1 In sum, being justified in some belief turns not just on the information possessed, but also on the cognitive processes in play. The cognitive processes in question are understood largely in terms of stable dispositions to transitions between representations or informational states. Thus, drawing the important distinction between being justified versus merely having justification leads epistemologists to focus on stable cognitive dispositions and to their manifestation in episodes of belief formation and maintenance. Now, our central concern in this paper is not to examine the metaphysical status of cognitive dispositions, but rather to make a point about the range of dispositions that are epistemically important. Our suggestion is that much epistemology, particularly since the modern period, has treated only a narrowly restricted range of cognitive dispositions as epistemically relevant to being objectively epistemically justified. Epistemologists have been fixated on what we will call classically inferential processes (or dispositions to classical inference). We argue that a wider range of processes and dispositions is epistemically crucial. The core idea of a virtue is that of a stable, good or excellent, disposition. This much has not changed all that much since Aristotle. Intellectual virtues were stable dispositions to reason well, when input and time were afforded the agent. Epistemic virtues are stable cognitive dispositions that are good (even excellent) from an epistemic point of view. Virtue epistemology is an approach to epistemology in which epistemic virtues are accorded an important and central place.2 This is correct, so far as it goes. But, as it stands, it does not

_____________ 1 2

Woodward’s (2003) discussion of causes and invariance provides a useful framework for approaching the question of what it is to be caused in the right way. This somewhat generic understanding of epistemic virtues and virtue epistemology aligns with Greco’s (2000, chapter 7-8; 2002a) understanding. It is largely in keeping with Sosa’s (1991a,

298

David Henderson and Terry Horgan

capture what is distinctive about virtue epistemology. For the traditional epistemology for which virtue epistemology is to provide an alternative is itself committed to the epistemic importance of dispositions. We suggest that virtue epistemology provides an important alternative insofar as it is open to the epistemic significance of a wider range of dispositions than has been much epistemology since the modern period. (This contrast will be developed in the next section.) Insofar as virtues are stable dispositions systematically contributing to epistemically desirable or productive results, and insofar as agents value true systems of belief, one does and should care about being virtuous. Evaluative concepts such as objective epistemic justification have coalesced around concerns arising within the individual and joint project of producing true systems of belief. They were developed because, both in one’s ongoing self-regulation of one’s cognitive life, and in people’s joint regulation of their joint or community’s epistemic life, one must evaluate beliefs and the processes by which they arose and were maintained. Thus, a core idea informing virtue epistemology is that fruitful and revealing epistemic norms, standard, or models will commonly make reference to stable cognitive dispositions of agents. The idea is that such stable-disposition-featuring standards (a) reflect central aspects of our epistemically central evaluative concepts, and (b) describe (and facilitate becoming) the sort of epistemic agent who, individually or in community, tends to have satisfactory or optimum success at fulfilling that central epistemic goal of possessing systems of true beliefs. What is optimal or satisfactory must be understood as relative to the kinds of creatures whose belief fixation is being evaluated. Typically, epistemologists have had in view adult humans with normal ranges of cognitive capacities and possibilities. (Such, after all, are the agents who developed the evaluative concepts in question, and who regulate their individual and joint project thereby.) What dispositions do human agent’s tend to develop? What plasticity is found here? What stable dispositions can they develop with training? The virtue epistemologist seeks to understand which constellation of stable dispositions among those that humans can with training come to possess, would have the optimum or at least satisfactory tendency to produce and maintain systems of true belief. They insist that objectively justified beliefs are those that are produced by such constellations of stable dispositions – such virtues. Now, in determining what stable dispositions humans can come to possess, with what training, epistemologists must draw on empirical results. It is thus fitting that most virtue epistemologists have been naturalized episte-

_____________ 1991b) influential epistemology. More restricted understandings of the epistemic virtues (such as Zagzebski’s, 1996) are to be found, but we do not believe they provide the best understanding – although they certainly can be instructive at points. The central points made here would seem to carry over to several of the more restrictive virtue epistemologies.

Epistemic Virtues and Cognitive Dispositions

299

mologists. (Much of what we say in what follows will be informed by contemporary cognitive science.) A rather diverse set of dispositions would qualify as virtues, given the core notion just set out. Very general traits of cognitive character having to do with belief fixation are sometimes mentioned as epistemic virtues. Openmindedness and intellectual courage, for example, are sometimes mentioned – by Zagzebski 1996, for example. Also commonly mentioned are dispositions to employ general cognitive processes characterized in terms of classes of inference: the disposition to reason in ways that are deductively sound, for example, or the sensitivity of inductive strength of evidence might be cited as epistemic virtues. These general inferential competences are yet very general dispositions. Stable dispositions of a more fine grained sort contribute to general inferential competences. One who is sensitive to the inductive force of evidence must be sensitive to certain sample characteristics – possible biasing factors, for example, or size with respect to the population at issue, for another. General inductive inferential competence requires such more specific stable dispositions or capacities – without which inductive inference from samples (in particular) would be subject to unacceptable pitfalls. (One might say the same thing for the simpler matter of deductive inferential competence.) Here one is reminded of Robert Cummins’ (1975, 1983) treatment of componential and functional analyses as an explanatory strategy. Sophisticated capacities or dispositions of some system are typically explained in terms of the organized working of simpler capacities or dispositions possessed by that system or its components – the latter being required for the realization of the former. Thus, the virtue of inductive inferential competence turns in part on various simpler capacities – virtues – such as sensitivity to sample biases, and sensitivity to counter-evidence, to competing explanation, to possible common causes, and the like. Not all cognitive competences are so clearly inferential in character. Consider a kind of general perceptual competence. To begin with, people develop domain specific perceptual competences: one might have developed a competence in identifying paintings and painters of the neo-romantic period, for example, and not in identifying beetle species. Agents can be more or less sensitive to when they have attained such a trained domain specific competence and when they have not. One who has such general stable perceptual sensitivities is perceptually competent in a particularly general way: such an agent readily arrives at perceptual beliefs only on matters on which they have reasonable domain specific competence – inhibiting judgments both where conditions do not afford a basis for confident judgment, and where training has not afforded a domain specific competence. This amounts to a general perceptual competence – and it seems on a par (with respect to both generality and epis-

300

David Henderson and Terry Horgan

temic usefulness) with the general inferential competences mentioned already. Both general and domain specific perceptual competences seem to qualify as epistemic virtues.3 The virtue of general perceptual competence also depends on a range of sensitivities – including (often inarticulate) sensitivities to one’s training or qualifications or track record in the relevant domain, and sensitivities to the quality of the particular presentation (light levels or background noise levels, degrees of obstruction, and the like). One’s perceptual processes should be conditioned by relevant background information – which must be recognized and brought to bear. With sensitivities to such matters in place, one should freely form perceptual beliefs on matters of refined competence, unless inhibited by background, or sensitivity to degraded presentations, and one should inhibit judgments when training is sparse. Perception and deductive, inductive, and abductive inference are not the only sources of belief, nor the only domains of general competence. Memory and testimony are also epistemically important. With respect to each, there would seem to be more specific competences or virtues.4

2. Why Virtue Epistemology A. The Pivotal Contrast and Claim Virtue epistemology is contrasted with what has become the traditional epistemological approach. Much epistemology from the last four hundred years has supposed that a rather limited range of cognitive processes is epistemically significant and crucial: (what we will here term) classically inferential processes. Intuitively, the idea has been that the epistemically relevant processes by which beliefs are formed and maintained are processes in the neighborhood of articulate (or at least articulatable) argument – processes in which informa-

_____________ 3

4

It does not seem useful to extend the concept of epistemic virtue (and perceptual virtue or competence in particular) to apply to capacities within the retina. Visual acuity is not usefully thought of as a virtue. Rather, as a general rule, epistemic virtues will have to do with cognitive processes (although not necessarily articulate processes, as we will soon see) – processes determining what is done with the information one does receive. A perceptually competent agent learns to deal with, be sensitive to, the acuity of eye sight, hearing, and so on. A virtuous perceptual agent can have poor eye sight, provided that agent has developed dispositions that compensate for the degraded character of the visual input received. One’s competence with testimony would also seem to be dependent on a range of more particular sensitivities or dispositions (Henderson, forthcoming). Even one’s general competence with memory might also yield to useful analyses (Goldberg and Henderson, forthcoming).

Epistemic Virtues and Cognitive Dispositions

301

tion is represented (the premises) and in which these representations, and just these, figure causally and decisively in the transition to some new belief (or in the retentive re-evaluation and vindication of some antecedent belief). The intuitive idea can be unpacked by distinguishing three notions of inferential processes. In each, agents are taken to possess “information,” where information is understood so as not to suppose truth (apparent information, but we will drop the qualifier ‘apparent’). The information possessed provides some contentful (more or less strong) support for various claims or beliefs that the agent might form, abandon, retain, or revise. Inferential processes, broadly understood, are simply those cognitive processes in which beliefs are formed or maintained on the basis of the information possessed. Being based on information is a causal notion, pointing to arrays of counterfactual dependencies and to dispositions. This is the broadest and most tolerant notion of an inferential process.5 For an inferential process, broadly understood, to be epistemically appropriate, it must yield beliefs that are necessarily or probably true, given the correctness of the information on which it relies. To qualify as a classically inferential process, a much more restricted class of processes, two additional things must be true of a cognitive process. First, in a classical inferential process, the information figuring in the inference is explicitly represented in the cognitive system that is the agent. This is not to say that the information need be consciously represented – just that, in taking its causal role in the cognitive system, that information is occurrently represented.6 Second, the causal processes whereby beliefs are fixed (formed, revised, or retained) must be occurrently isomorphic with the deductive and inductive support relations obtaining between the information that the agent possesses. This is to say that the contentful support relations between the information possessed is mirrored by, or isomorphic with, causal dependencies between the representations in the agent’s cognitive system. This obtains when the relevant psychological states (representations) are occurrent, and as such they causally conspire to generate or sustain the belief in question in a way that mirrors the way in which the pieces of information represented contentfully conspire to “support” the belief in question. A careful reading of a diversity of modern and contemporary texts suggests a wide partisanship for

_____________ 5

6

Indeed, as noted later, this notion probably strains the notion beyond common usage, insofar as common usage associated inference with argument, and insofar as argument is understood as involving transitions between representations (premises and conclusions). We employ this broad usage here because, once one recognizes the possibility of the sorts of informed transitions we envision here, one will find it reasonably natural to refer to them as “inferential after a fashion.” This much does not require that the inference rules to which the system conforms are themselves represented.

302

David Henderson and Terry Horgan

the presupposition that empirically relevant processes are classical inferential processes.7 Cognitive processes might be broadly inferential without being classically inferential. For example, the agent might possess information that is not represented (as representations are standardly understood) and yet causally conditions the agent’s belief fixing processes. (If this is difficult to envision, bear with us, and we will explain more fully.) If this obtains, then the agent’s broadly inferential processes will not be occurrently isomorphic with the possessed information and contentfully appropriate transitions.

_____________ 7

The epistemological project has commonly been conceived of as indicating by what arguments one can attain various forms of justified belief and knowledge regarding the wider world. This much is a fairly pervasive heritage of modern philosophy. Descartes’ first philosophy sought a classically inferential path from self-evident truths to what is needed for modern science. Hume’s more skeptical first philosophy sought to bring home the limited power of “reason” – by making evident the limits of classical argument to take us beyond certain starting points. Similarly, the justification of induction was thought to depend on deductive demonstrations of the likely reliability of inductive inference. Here, a priori argument, by tracking just the contentful relations between certain representational states, was to vindicate rules and arguments (or to fail to vindicate the inferences in question). Parallel demands were placed on the epistemic vindication of testimony. Kant reframes the epistemological project, not by repudiating the epistemic focus on classical inference, but by narrowing the ground that it is supposed to cover. The beginning of the epistemic chore – the starting position – is a set of perceptual judgments that are essentially more substantive than Humean appearances. The epistemic task is to do justice to the phenomenal world – the world as it appears to us. In effect, this takes us from the phenomena (the phenomenal objects, objects as they appear to us) to some suitable systemization of them (to accounts intended to be true of the phenomenal world). While much may have gone on in us to get us to the Kantian starting places – to the objects as they appear to us – the epistemological chore is the classically inferential chore of getting to a general account of the phenomenal world. More recent epistemology has retained this focus on classical inference – witness Carnap’s attempt to develop an inductive logic in terms of observation reports and ranges of state descriptions. More recently, BonJour’s (1985) coherentism, with its roots in Davidson 1983, provides a prominent example. Foundationalist approaches have generally faced an interesting problematic – that some finding regress stopping accessible states that can be understood as inferential bases (premises, reasons to think other claims true) and which yet do not themselves require inferential justification. Thus, the coherentist BonJour 1985, thought it unlikely that the foundationalist would succeed, while the foundationalist BonJour 1998 has decided that, on pain of skepticism, there must be the resources for such a structure of argument. The recurrent theme: epistemic agents must argue their way about on the basis of representations – that classical inference is the epistemic engine. This focus on inference from articulate reasons as premises has certainly been encouraged by access internalist epistemology. In Henderson and Horgan 2000 we develop a few examples of the contemporary focus on classical inference (BonJour 1985, Pollock 1986, Moser 1989, Audi 1993). It is clear that any epistemology that holds that epistemic agents must base their beliefs on representations, the content of which is internalistically accessible (any epistemology that conforms to access internalist constraints at least this far) will be an epistemology that is focused on classically inferential processes as the processes that make for objective epistemic justification (or its absence). It is worth noting that even epistemologists who incline towards a kind of reliabilism may think that the relevant (reliable or not) processes are those that make use of just such representations. Audi 1993 might serve as an example.

Epistemic Virtues and Cognitive Dispositions

303

Our suggestion is straightforward: while some epistemically important cognitive dispositions may be dispositions to classical inference, not all are. Not all our epistemically important, productive, or needful chores are, or can be, well managed only at the level of classical inference. Insofar as some such chores must be managed in significant measure by way of dispositions that are broadly inferential but not classically inferential, a superior epistemological perspective will give significant attention to virtues – to epistemically good dispositions – beyond what has been the standard epistemological focus. It will be helpful to have in play a third class of inferences – intermediate between broad and classical. Classical inferential processes are transitions between representations and representations, and are occurrently isomorphic with the contentful support relations obtaining between the featured representations. Broadly inferential processes move from “information,” that may, but need not be represented, to some belief. The dynamics of the transitions may be conditioned by informational states that are not representations. Now, talk of “broadly inferential processes” is a little strained, for inference suggests argument, and argument suggests propositional representational states (premises and conclusions). Broadly inferential processes as we understand them need not begin with propositional representational states – and so need not begin with premises in the standard sense. Perceptual processes, for example may begin with much information that is not propositionally represented, and so are ill understood in terms of inference as a matter of argument.8 Now, to get to our intermediate category of inference, one can envision cognitive processes that make for transitions between propositional representations while also being conditioned in part by some or much information that is possessed by the system and yet not represented in that processing. Such processes would not be occurrently isomorphic with the contentful support relations obtaining between the representational states – as they would be conditioned by information that is possessed but not represented. Such processes, involving transitions between propositional representational states, which do not satisfy the occurrent isomorphism requirement for classical inference, constitute a kind of inference intermediate to broad and classical. Because they are marked by having premises and conclusions (proposi-

_____________ 8

These point are related to the power of virtue theory to move beyond an overly constraining concern with reasons in epistemology – a strength early appreciated by Sosa 1980. We need not here commit to whether perceptual processes begin with nonpropositional representations of some sort (whether the initial states in the process – the initial total activation space in a given perceptual episode, perhaps – need be a representation) – and the matter will depend on just how representation is understood. In any case, the present point reflects at least one sense in which perceptual processes may be thought of as direct – as we suggest that they commonly are neither classically or argumentatively inferential.

304

David Henderson and Terry Horgan

tional representational beginnings and endings), we can write of argumentative inference. Some of the stable dispositions that interest us here often mediate transitions between propositional representations – and thus are dispositions to argumentative inference. But, we will argue that the epistemically important dynamic of these transitions often cannot be understood wholly at the level of representations – articulate or not. When this is so, the processes are broadly and argumentatively inferential, but not classically inferential. (All classical inference is argumentative inference, but not all argumentative inference need be classical inference.) To the extent that processes crucial to human epistemic successes fail to qualify as classically inferential, and are rather either argumentatively or broadly inferential, epistemology must abandon the fixation on classical inference and must attend to a wider set of dispositions than is common. It must become a distinctively virtue epistemology. Prominent in our argument for virtue epistemology will be our discussion of central processes of belief formation, processes of inductive reasoning and abductive reasoning. We argue that the epistemic chores an agent faces here cannot be managed wholly at the level of classical inference. Such argumentative inference draws on information that is essentially richer than what is or can be represented. The epistemically virtuous dispositions are dispositions to accommodate such information, and thus are not classically inferential. We now flesh out the suggestion that certain subtle dispositions (commonly acquired or refined and developed by training), (a) are crucial to systematic human epistemic successes on a range of cognitive chores (chores that are commonly recognized by epistemologists to be important and necessary), and (b) are not readily understood fully in terms of classical inference. The virtue theorist sees the epistemological fixation on classical inference as analogous to the contemporary automotive industry’s overwhelming reliance on the internal combustion engine – it is time we recognize the promise of hybrid technology. B. The Argument Human epistemic competence, that combination of processes by which humans do and should manage their epistemic chores, cannot be understood solely in terms of classical inference. Among those epistemic chores widely recognized as epistemically needful are ones managed, not by classical inference, but by dispositions keyed to essentially richer sets of information. By way of example, we now discuss two epistemic chores: (1) chores within central processes of belief fixation requiring the holistic accommodation of vast ranges of potentially relevant information, and (2) chores involved in fast,

Epistemic Virtues and Cognitive Dispositions

305

sensitive, formation of perceptual beliefs – an activity or process that is itself permeable to, or conditioned by, much background information. Both reflect an essential place for dispositions essentially richer than what can be managed by classical inference. Our discussion of the first of these chores develops the case most fully, while our discussion of the second will need to be more suggestive. The Chores, Part 1: Holistic Processes of Central Belief Fixation a. Holism, the Frame Problem for Computationalism, and the Need for morphological Content Certain problems and developments in cognitive science provide reason to believe that the human cognitive processes that typically produce and sustain beliefs are not of a sort that is well understood when preoccupied with classically inferential processes.9 There is reason to believe that human belief generation must, and epistemically ought, proceed in an alternative manner. The alternative understanding emerges in the wake of a family of recalcitrant difficulties within the classical, computational, conception of mind in cognitive science – difficulties often classified under the rubric “the frame problem.” In this section we first describe those difficulties, drawing upon an influential discussion by Fodor (1983), and recommend the alternative conception suggested by frame-type problems. We then sketch one way of elaborating this conception, inspired by connectionism and by the form of mathematics that goes naturally with it, dynamical systems theory. The frame problem is a direct challenge to computational cognitive science – but the alternative understanding of epistemically virtuous belief generation to which it points is incompatible with the standing epistemological fixation with classically inferential processes. In the closing pages of The Modularity of Mind, Fodor 1983 argues that certain problems in classical cognitive science look to be in-principle problems, and hence that the prospects for understanding human belief fixation wholly within the framework of classical cognitive science are very bleak. These problems continue to plague the computational approach. The main claim of Fodor’s influential book is that the human cognitive system possesses a number of important subsystems that are modular: domain specific, mandatory, limited in their access to other parts of the larger cognitive system, fast, and informationally encapsulated. Where the classical com-

_____________ 9

Access internalist epistemology has encouraged this problematic preoccupation with classical inference, and it is also called into question by these results. The points made in the present section are closely related to discussions in Henderson 1994 and Henderson and Horgan 2000. We here draw on material in Horgan and Tienson 1994, 1996 and Horgan 1997a.

306

David Henderson and Terry Horgan

putational approach has gotten somewhere, he says, is in understanding such modular subsystems, which by their nature delimit the class of relevant information. However, classical computationalism has made very little progress in understanding what he terms central processes. Belief fixation on the basis of current input together with other beliefs – is a paradigmatic example. These processes are non-modular: they need to have access to a wide range of cognitive subsystems, and to information on an indefinitely wide range of topics. And the very considerations that point to non-modularity, he maintains, also constitute grounds for extreme pessimism about the prospects of explaining central processes within the framework of classical computational cognitive science. For our purposes here, this is reason for pessimism about the prospects for understanding central processes of belief fixation wholly in terms of classical inference. Fodor articulates these considerations in terms of the analogy between belief fixation in human cognition and scientific confirmation. This is to garner lessons about central cognitive processes like belief fixation from what we know about “empirical inference in science” (Fodor 1983, 104). Scientific confirmation, “the nondemonstrative fixation of belief in science,” has two crucial features. It is (in Fodor’s terminology) isotropic and Quineian: By saying that confirmation is isotropic, I mean that the facts relevant to the confirmation of a scientific hypothesis may be drawn from anywhere in the field of previously established empirical (or, of course, demonstrative) truths. Crudely: everything that the scientist knows is, in principle, relevant to determining what else he ought to believe… (1983, p. 105) By saying that scientific confirmation is Quineian, I mean that the degree of confirmation assigned to any given hypothesis is sensitive to properties of the entire belief system; as it were, the shape of our whole science bears on the epistemic status of each scientific hypothesis (1983, p. 107).

Isotropy brings in the whole of current theory: any bit of apparent information from any portion of one’s belief system might, in some circumstances, be evidentially relevant to any other. Being Quineian makes confirmation holistic in a deeper way: confirmation depends upon “such considerations as simplicity, plausibility, and conservatism” (Fodor 1983, 108), which are determined by the global structure of the whole of the current belief system and of potential successor systems. Since belief fixation in human cognition is commonly a matter of inductive inference from the information provided by input systems and the information antecedently possessed, evidently it too must be isotropic and Quineian. Fodor concludes that it must be non-modular. He also stresses that these global aspects of belief fixation look to be at the very heart of the problems that classicism has encountered in attempting to understand such central processes:

Epistemic Virtues and Cognitive Dispositions

307

The difficulties we encounter when we try to construct theories of central processes are just the sort we would expect to encounter if such processes are, in essential respects, Quineian/isotropic… The crux in the construction of such theories is that there seems to be no way to delimit the sorts of informational resources which may affect, or be affected by, central processes of problem-solving. We can’t, that is to say, plausibly view the fixation of belief as effected by computations over bounded, local information structures. A graphic example of this sort of difficulty arises in AI, where it has come to be known as the “frame problem” (i.e., the problem of putting a “frame” around the set of beliefs that may need to be revised in light of specified newly available information) (Fodor, 1983, pp. 112-3).

Let us take a closer look at the challenges posed by isotropy in particular. The distinction between being justified versus merely having justification turns on the idea that, when one is justified in one’s belief, the justification that one has for that belief is not causally inert, but rather “comes into play” or is “appropriately operative” in generating or sustaining that belief. Whatever information one might possess, if one does not systematically use it, and use it in an appropriate way, then, even if one somehow arrives at true beliefs, one does so by accident, and not by justificatory processes. For such distinctions to make sense, it must be possible for one to have justifications for a belief which do not, and did not, “come into play” in the fixation of that belief. In such cases, something else must then be relevantly operative in belief-fixation. Perhaps one has “other justification” for the belief in question, and these might be there operative in the relevant belief-fixation (in which case one may yet be justified). Or perhaps what is in play in fixing the belief in question is not justificatory information – not something that one should talk of as justification that one has (in this case one has justification, but is not justified). Such possibilities are presupposed by the distinction between being justified in believing and merely having justification for a belief. The distinction requires the selective efficacy of beliefs and information states representing the justification that the agent has. One might think that holism is incompatible with such distinctions. But, a closer look at the holistic dimension that Fodor terms “isotropy” should serve to dispel this suggestion. When Fodor writes of confirmation as isotropic he insists that “the facts relevant to the confirmation of a scientific hypothesis may be drawn from anywhere in the field of previously established empirical (or, of course, demonstrative) truths. Crudely: everything that the scientist knows is, in principle, relevant to determining what else he ought to believe” (p. 105, emphasis added). Fodor’s first formulation reflects a point that is lost in the crude reformulation. Isotropism is the potential relevance of everything (all the beliefs and information one has) to everything else one believes – but this is not to say flatly that everything one believes is actually relevant to everything else one believes. One must be careful in conceiving of relevance here, and it will be helpful to distinguish a strict sense, actual relevance, and an extended

308

David Henderson and Terry Horgan

sense that might best be thought of as relevance to relevance. In the extended and attenuated sense it is correct that everything that one believes (and more generally, all the information that one in some sense possesses) is relevant to any episode of belief fixation. Take any two beliefs that are mutually irrelevant on their face – that is, when considered of themselves, without taking into account what else one might believe. Then notice that, with various sets of possible auxiliary information or beliefs serving as intermediaries, the beliefs that were of themselves irrelevant would become mutually relevant. What is actually relevant to what depends then on what else one actually happens to believe – or on what other information one actually has (and how it “adds up”). To then be sensitive to what is actually relevant to what one does or might believe, one must be sensitive to how the set of what else one believes makes for various stripes of mediated actual relevance. Much actual relevance is mediated relevance, so this is important. To be sensitive to mediated relevance, and thus to much actual relevance within one’s doxastic and informational set, one must be sensitive to the full set of one’s information or beliefs. Thus, in view of the potential relevance of everything to everything, everything that one believes, all one’s information, becomes relevant in the extended or attenuated sense – it is relevant to relevance – and one must be appropriately sensitive to this in belief fixation. Only then can the actually relevant beliefs and information “come into play,” as required if the agent is to be justified in the resulting belief as opposed to merely having justification. The prospects for understanding central processing within the classical computational paradigm look very discouraging indeed. Not only do we have no computational formalisms that show us how to manage the epistemic chores associated with the Quinean and isotropic aspects of central processes of belief fixation; it’s a highly credible hypothesis that a (tractable) computational system with these features is just impossible, for belief systems on the scale possessed by human beings. Human central processing evidently does not operate via any kinds of computation we currently know about or can even contemplate. Something else is needed. What might it be? These frame-type problems arise largely because of the apparent computational intractability of managing all relevant information, insofar as that information gets explicitly represented in the course of cognitive processing. What this suggests is that belief fixation and related cognitive processes operate in a way that accommodates much relevant information automatically and implicitly. The suggestion is that the holistic aspects of belief fixation involve not the finding and fetching of relevant representations from memory-banks where they are stored in explicit form (to accommodate isotropy), and not the overt representation and comparative evaluation of large-scale alternative belief-systems (to accommodate the Quinean dimension). Rather, these holistic aspects are somehow implicit in the structure of the cognitive system, in

Epistemic Virtues and Cognitive Dispositions

309

such a way that temporal transitions from one occurrent cognitive state to another accommodate the holistic aspects automatically. In the terminology of Horgan and Tienson (1995, 1996), important holistic informational content is morphological, rather than occurrent. Morphological content is information that: (i) (ii)

is implicit in the standing structure of the cognitive system (rather than explicitly represented in the system’s occurrent cognitive states or explicitly stored in memory), and gets accommodated in cognitive processing without getting explicitly represented in occurrent cognitive states, either conscious or unconscious.

The apparent moral of the frame problem is that, in general, human belief fixation must operate in a way that draws heavily upon morphological content, in order to avoid computational intractability. As we will put it, these processes are essentially morphological. Belief fixation is not accomplished (simply) by computationally manipulating explicit, occurrent, representations of all relevant information. Nor are they accomplished by “proceduralized” computational processes that are mere shorthand algorithms for computations that could instead have been carried out in a way that renders all relevant information explicitly. Essentially morphological processing is a fundamentally different way of accommodating the holistic aspects of belief fixation. Recognition of the holistic elements of belief-fixation, of its Quineian and isotropic character, is not, at least not primarily, a product of cognitive science. Rather, it arises out of fairly traditional epistemological reflection – for example, out of reflection on the role of ancillary hypotheses and webs of belief that can be found in writings of philosophers such as Hempel and Quine. Reflecting on how the implications of one sentence commonly turn on what other sentences one combines with it highlights the information to which a successful epistemic agent must be sensitive – revealing the informational tasks facing an epistemic agent. (One could point to a vast range of philosophical work – to name a few examples, to BonJour 1985, to Shapere’s (1982) discussions of the range of information and understandings that inform “observations” in physics, to Goodman’s (1973) work on induction and the projectibility of certain predicates – itself gauged in terms of much background information.) Central processes of belief fixation then must be handled essentially morphologically – so that much possessed information is accommodated automatically, by way of morphological content. Not only does this doom classical computational cognitive science, it also indicates that these epistemic chores cannot be managed by way of processes wholly at the level of classical inference. Significant ranges in information must condition belief fixation without being represented – and thus these processes cannot be occurrently isomorphic with contentful relevance relations obtaining between

310

David Henderson and Terry Horgan

the information deployed here. Dispositions subtler and richer than those evinced at the level of classical inference are epistemically essential. b. A Way to think about morphological Content How might the daunting task of essentially morphological processing get accomplished in human cognition? To our knowledge, there certainly are no models in cognitive science that come close to achieving such processing for cognitive tasks even remotely comparable in complexity to those handled in real human thought. Nor is there any good reason, as far as we know, to think that any extant models are likely to “scale up” or be straightforwardly extended to provide an adequate account of real human cognition. This is no less true for connectionist models than it is for classical computational models. Nonetheless, a general conception of human cognition has begun to emerge within cognitive science that is potentially more powerful than the classical computational conception of mind, and that provides the broad outlines of an answer to the question, “How is essentially morphological processing possible?” This alternative conception draws cautiously upon connectionist modeling, in a way that eschews unduly optimistic assumptions about the scale-up potential of extant models. It also draws upon a form of mathematics that is natural for describing connectionist models – dynamical systems theory. Here we offer a very brief sketch of connectionism, dynamical systems theory, and the nonclassical framework – with emphasis on features that are especially germane to morphological content.10 In a connectionist system, information is actively represented as a pattern of activation. When the information is not in use, that pattern is nowhere present in the system; it is not stored as a data structure. The only representations ever present are the active ones. On the other hand, information can be said to be implicitly present in a connectionist system – or “in the weights” – if the weighted connections subserve representation-level dispositions that are appropriate to that information. Such information constitutes morphological content in the system, rather than explicitly-represented content. Among the apparent advantages of connectionist systems, by contrast with classical computational systems, is that morphological information “in the weights” gets accommodated automatically during processing, without any need for a central processing unit to find and fetch task-relevant information from some separate memory banks where it gets stored in explicit form while not in use.

_____________ 10

This nonclassical framework for cognitive science is described at length in Horgan and Tienson 1996.

Epistemic Virtues and Cognitive Dispositions

311

Learning is conceived quite differently within connectionism than it is within the classical approach, since connectionist systems do not store representations. Learning is the acquisition, “in the weights,” of new morphological content. The branch of mathematics called dynamical systems theory is often applied to connectionist models. To describe some physical system (e.g., a planetary system, or a connectionist network) mathematically as a dynamical system is to specify in a certain way its temporal evolution, both actual and potential. The set of all possible states of the physical system – so characterized – is the mathematical system’s abstract, high-dimensional state space. Each magnitude or parameter of the physical system is assigned a separate dimension of this mathematical space, and each possible state of the physical system, as determined by the values of these magnitudes, corresponds to a point in state space. A dynamical system, as such, is essentially a complete mathematical description of how the physical system would evolve temporally from any possible initial state; it is a collection of trajectories through state-space, with a trajectory emanating from each point in state space. In essence, it is a description of a complex disposition. A useful geometrical metaphor for dynamical systems is the notion of a landscape. A dynamical system describing a physical system involving n distinct magnitudes is the n-dimensional analog of a two dimensional, non-Euclidean, contoured surface: i.e., a topological molding of the ndimensional state space such that, were this surface oriented “horizontally” in an (n+1)-dimensional space, a ball would “roll along the landscape,” from any initial point p, in a way that corresponds to the way the physical system would evolve from the physical state corresponding to p. Connectionist systems are naturally describable, mathematically, as dynamical systems. The state space of a network is its “activation space” (which has as many dimensions as the network has nodes), and the dynamical system associated with the network is its “activation landscape.” In connectionist models, cognitive processing is typically construed as evolution along the activation landscape from one point in activation space to another – where at least the beginning and end points are interpreted as realizing intentional states. So in terms of the mathematics of dynamics, occurrent cognitive states are realized mathematically as points on the activation landscape, which are then realized physically as distributed patterns of activation in the nodes of the network. Morphological content – the information implicit “in the weights” – is embodied in the topographical contours of the network’s high-dimensional activation landscape. Thus, the various superimposed slopes on the activation landscape subserve trajectories from one occurrent cognitive state to another that automatically accommodate the morphological content.

312

David Henderson and Terry Horgan

From this mathematical perspective, training a network is a matter of (i) molding the activation landscape, thereby inducing new topological contours embodying new morphological content, while simultaneously (ii) refining the cognitive/mathematical realization relation whereby intentional states get realized mathematically as points on the landscape. (The weight-change training procedures employed in connectionist modeling bring about a coevolution of these two factors.) Once “trained up,” the system’s temporal trajectories from one occurrent intentional state to another will automatically accommodate the relevant morphological content. Horgan and Tienson 1994 and 1996 describe a non-classical framework for cognitive science that they call the dynamical cognition framework (the DC framework). This alternative approach offers an answer, in principle, to the question, “How could the holistic, Quineian/isotropic, aspects of cognitive processes be accommodated automatically and morphologically?” The answer is this: In principle, Quineian/isotropic information could be embodied morphologically in the complex and subtle topography of a high-dimensional activation landscape subserved by the human central nervous system. Given a sufficiently nuanced realization relation from cognitive states to points on this landscape, the landscape’s multifarious, superimposed, topographical contours guarantee that the cognitive system’s transitions from one occurrent cognitive state to another are automatically appropriate not only to the explicit content of these occurrent states themselves, but also to very large amounts of implicit Quineian/isotropic information. We maintain that the frame-type problems encountered in classical cognitive science provide a strong reason to maintain that the holistic, Quineian/isotropic aspects of belief-fixation are accommodated in an essentially morphological way in human cognition. We also maintain that this conclusion is further reinforced by the in-principle account of such essentially morphological processing that is provided by the DC framework. But it should be noted that the appeal to the DC framework is not essential here. Even if one were thoroughly dubious about connectionism and about the usefulness of dynamical-systems ideas in cognitive science, one still could accept the argument from frame-type problems to the conclusion that processes like belief-fixation must somehow be essentially morphological. c. Morphological Content, Dispositions, and Virtue Epistemology The central claims in the above subsections are these: First, we have argued that there are holistic aspects of belief fixation. Isotropy means that the central processes must accommodate the potential relevance of each piece of possessed information to any given belief, and thus that all information pos-

Epistemic Virtues and Cognitive Dispositions

313

sessed is relevant to what is relevant. The processes by which an individual manages to fix beliefs must be sensitive to what is relevant – and this cannot be managed computationally. Being Quineian means that global features of belief systems are themselves relevant, and it seems likely that this itself also cannot be gauged computationally. Second, we have suggested that, in managing the indicated epistemic chores, the competent human agent must accommodate much possessed information automatically without representation, and thus without it featuring as a premise in a classical inference. The competent human cognitive agent must make use of morphological content. Third, while there are certainly broadly or argumentatively inferential moments in the processes here envisioned, the whole cannot be fully understood as classically inferential. Here, it is good to reflect on the dynamical system framework we have recommended. One might think of some of the transitions between representational states as inferences on the order of those that have been the common epistemological fare – as transitions between representations (premises and conclusions). But remaining only at that level, one cannot fully understand the dynamics of belief fixation by which crucial holistic chores are managed. Instead of remaining at the level of what is classically inferential, it is best not to loose sight of the cognitive system’s dynamics and the morphological content on which it relies. This is to focus on epistemically crucial dispositions beyond those that have been the common epistemological fare. Such virtues are broadly inferential, and are sometimes argumentatively inferential, but are not restricted to the sort of classically inferential transitions that have too commonly marked the limits of epistemic concern. Only with this less constrained focus can one appreciate the full range of information in play within central processes of belief fixation, and how such information might have an impact on such processes. Consider the dynamical system as a topography of the system’s weight space. This is a matter of its tendencies to move from one total state (or pattern of activation) to others – to descend in state (or activation) space. Only some of the patterns of activation through which it would descend, from a given point in activation space, are themselves representations – the sole stuff of classical inferences. But the whole dynamic is a more subtle disposition, one that reflects the capacity of the human cognitive system to make sensitive use of vast ranges of information. The capacities or dispositions that the system acquires through courses of experience, through training, are highly desirable epistemically – are epistemic virtues. Some of the patterns of activation through which the system passes are representations, so some of the transitions between such representations will count as argumentative inferences. But, just as not all the information that the system possesses is represented, or even could all be represented, and just as not all the transitions are determined or conditioned only by representations, not all the epistemically desir-

314

David Henderson and Terry Horgan

able dispositions of the system are to be understood as at the level of classical inference. We thus arrive at the general theme or claim on which this paper pivots: some epistemically important cognitive dispositions are dispositions to inference broadly conceived that cannot be fully understood at the level of classical inference. Insofar as not all our epistemically important, productive, or needful chores are or can be well managed only at the level of classical inference, insofar as they must be managed in some significant measure by way of dispositions that are not, strictly speaking, classically inferential, then a superior epistemological perspective will give significant attention to virtues – to epistemically good dispositions – beyond the classically inferential sorts that have been common epistemological fare. In his virtue ethics, Aristotle was notably reticent regarding precise principles or rules by which the morally appropriate action could be determined. He seems to suggest that no set of precise exceptionless rules can ultimately capture what it is to live a morally good or correct life – stated somewhat anachronistically: one cannot write a program for the moral life. Rather than rules – he urges kinds of training, kinds of sensibilities or capacities to foster. Famously, he insists that the correct action is the action that the morally virtuous person would undertake – and that this is a matter of a rational judgment drawing on trained sensibilities.11 The suspicion of precise rules (moral rules or rules of inference on the order of a program) has been a recurring feature of a number of virtue theories, moral or epistemic. In light of our discussion of central processes of belief fixation, and of the place for morphological content in particular, we can at least suggest why there might be something importantly right about this suspicion of rules as the end all and be all of epistemology – at least if one thinks of rules as something on the order of a computational program for belief-fixing processes. Belief fixation is probably too complex and subtle to conform to programmable rules at all.12 In this paper, the central processes of belief fixation serve as our central exhibit of the epistemological importance of dispositions to broadly inferential transitions that cannot be understood in terms of classical inference. We have developed the case at some length. The general picture of epistemically important processes and dispositions that has emerged can be applied to other epistemological topics – to the epistemic generation or use of percep-

_____________ 11

12

In the Nichomachean Ethics, one reads: I mean moral virtue; for it is this that is concerned with passions and actions, and in these there is excess, defect, and the intermediate. For instance, both fear and confidence and appetite and anger and pity and in general pleasure and pain may be felt both too much and too little, and in both cases not well; but to feel them at the right times, with reference to the right objects, towards the right people, with the right motive, and in the right way, is what is both intermediate and best, and this is characteristic of virtue (Aristotle, Nicomachean Ethics, II 6, 1106b, 15-20. ). See Horgan and Tienson 1996 for a fairly extended discussion of this point.

Epistemic Virtues and Cognitive Dispositions

315

tion, memory, or testimony, for example. In each case, one would find that belief formation can and should be conditioned by, or permeable to, indeterminate ranges of possessed information that must be accommodeated quickly and it seems largely automatically. The central processes of belief fixation we have discussed presented us with transitions that are broadly and argumentatively inferential in character. Before closing, we want to discuss (in a more abbreviated fashion) an epistemic context where the transitions are broadly inferential, less clearly argumentatively inferential, and clearly not classically inferential. Consider perceptual processes. Note that perceptual processes are akin to central processes of belief fixation in a very important way – they are permeable to, and necessarily sensitive to, wide ranges of background information. Again, this is managed very quickly – and automatically. Again, it is a highly plausible hypothesis that this is managed by reliance on morphological content within a kind of cognitive dynamical system. Again, virtue epistemology seems vindicated. The Chores, Part 2: Sensitive Perceptual Processes What is it to be perceptually competent with respect to some limited matter? What is it to be perceptually competent generally? When one is a competent perceptual judge of the wildlife in one’s environment, for example, one can with reasonable sensitivity respond to common episodic encounters (and to more or less enduring traces) in one’s environment. Due to one’s sensitivity, one can produce reasonably accurate identifications of the species of wildlife involved in the episodes (or leaving the traces). One who is perceptually competent with respect to a delimited domain is able to make reliable judgments in a range of common environmental conditions. Thus, one who is perceptually competent with respect to local wildlife can make reliable identifications when various creatures are partially obscured, quickly glanced, or imperfectly illuminated (common conditions). Importantly, such capacities themselves require reasonable sensitivity to when conditions do not allow the agent to make a reliable identification. A competent perceptual judge on a given subject matter is thus one who has been trained up to be reasonably reliable on the matter in question – one who can consequently render verdicts that are likely true – and this requires not rendering verdicts in certain ranges of difficult cases.13

_____________ 13

For reasons having to do with the new evil demon problem, and pointing to the epistemic significance of certain nonstandard forms of reliability, these remarks need to be qualified. They hold true of perceptual competence for agents in reasonably hospitable epistemically possible global environments – such as that provided by the actual global environment. However, from

316

David Henderson and Terry Horgan

It is important to pause and notice the range of information that the agent – as cognitive system – must possess and use, if the agent is to exercise a domain specific perceptual competence in a given episode. The agent will need to sensitively possess much information particular to the episode. This information need not, and commonly will not be represented in the perceptual processes. For example, there need be no representation of the degree of occlusion, nor of just what “parts” are shielded from view, in the course of the perceptual process by which the agent generates or refrains from generating a judgment in response to a fleeting glimpse of some member of the local fauna. Yet, a competent judge of local fauna will have acquired a sensitivity to such matters. Similarly for light levels, shadows, and the like. Auditory input may condition the visual reception – and again in ways that are not fully susceptible to argumentative reconstruction. In competent and epistemically highly laudable perceptual processing, much such information, as morphological content, is likely accommodated automatically and very quickly. The information accommodated in a competent perception is typically not limited to information about the particulars of the episode. The perception of the competent judge will be conditioned by much background information – in ways that seem not at all dissimilar to the way in which wide ranges of relevant information can be accommodated in central processes of belief fixation. Consider the hesitance that a competent ornithologist would experience when, in the United States, confronted with a passing glimpse of a large woodpecker having what appears to be a white bill and white on both leading and trailing edges of its underwings. Perhaps the location is one in which Pileated Woodpeckers are common. An ornithologist would have some reasonable feel for the relevant ranges and characteristic population densities, of Pileated Woodpeckers and other candidates – of base rates. But the Pileated Woodpecker has a relatively dark bill and a black trailing edge of the underwings. What of the Ivory Billed Woodpecker? Obviously very different base rates (even given some uncertainty regarding whether it is extinct or just exceedingly rare). The perceptual process of a competent agent is permeable to such information, and much more of relevance. One might think of the processes as giving rise first to a perception or perceptual seeming that serves at the basis for a perceptual belief. The substantive perceptual seeming might be something on the order of the appearance of a large bird with a light colored beak and wing patches on a black background, moving through the dappled

_____________ the epistemic point of view, one can be perceptually competent in a demon-infested global environment, where all ones training has been deceptive, and apparent successes and failures have resulted in an exquisitely sensitive perceptual system that is nevertheless not globally reliable. For more on the relevant forms of reliability, see Henderson and Horgan (in press). Montmarquet 1987 rejects the association of virtue with reliable processes, but we believe that the better move is to refine the understanding of the relevant form of reliability.

Epistemic Virtues and Cognitive Dispositions

317

sunlight of deep forest cover. Perceptual seemings, given or taken, involve what seem to be salient objects standing forth – for example, an animal of a more or less familiar kind situated with respect to the agent and certain familiar things in the agent’s immediate environment. Whatever goes on in the processes yielding such perceptions must be quite sophisticated and complex, being conditioned by the ranges of information already suggested. Typically, the competent agent proceeds without hesitation or reflection to a perceptual belief. Occasionally, however, the perception or perceptual seeming will arise with an uneasiness. It may arise with what we might call a “warning flag,” commonly associated with an issue demanding further attention. Further observation may seem needed because of light levels or angles, or occlusion. Or there may be the thought that the seeming is very unlikely to be true. Here, background information will have occasioned restraining thought– much as, in central processes of belief fixation, background information may be accommodated so as to make certain information stand out as relevant. In such cases, further thought may have an argumentatively inferential character. But, in the more straightforward episodes, it is plausible that the agent’s perceptual processes need not pass through propositional representations on the way to the forming of a perceptual belief – and thus may be broadly inferential without being argumentatively or classically inferential. We believe that perceptual processes – with their conditioning by, or permeability to, much background information that is accommodated automatically – require dispositions that are much like those involved in central processes; they are dependent on complex and subtle dispositions that are rightly thought of as epistemic virtues. Since the starting places – the appearances – are not comfortably seen as premises, it seems best to categorize the rudimentary perceptual processes making for the transitions here as neither classically nor argumentatively inferential. These broadly inferential perceptual processes can then occasion argumentatively inferential processes in some cases. What then of an agent who is perceptually competent generally? This would seem to be an agent who, by virtue of training, has acquired a reasonable range of domain specific perceptual competences, and who tends to refrain from forming perceptual judgments with respect to domains about which such competence is yet to be acquired. Such sensitivity to one’s sensitivities is itself an epistemic virtue. We have argued that human epistemic competence cannot be understood solely in terms of classical inference. Among the epistemic chores widely recognized as epistemically needful are ones managed, not by classical inference, but by dispositions keyed to essentially richer sets of information. Central processes of belief fixation turn on holistic sensibilities that require the use of morphological content. So also do competent human perceptual proc-

318

David Henderson and Terry Horgan

esses. These have served to illustrate the epistemological importance of cognitive dispositions beyond those commonly of concern in epistemology.

Literature Aristotle. 1941. Nicomachean Ethics. W. D. Ross, trans. In The Basic Works of Aristotle. Ed. R. McKeon. New York: Random House.. Audi, Robert. 1993. The Structure of Justification. Cambridge: Cambridge University Press. BonJour, Laurence. 1985. The Structure of Empirical Knowledge. Cambridge: MA: Harvard University Press. BonJour, Laurence. 1998. In Defense of Pure Reason. Cambridge: Cambridge University Press. BonJour, Laurence. 2002. Epistemology: Classic Problems and Contemporary Responses. Oxford: Rowman and Littlefield. Burge, Tyler. 1993. “Content Preservation.” Philosophical Review 102: 457-88. Burge, Tyler. 1997. “Interlocution, Perception, and Memory.” Philosophical Studies 86: 21-47. Coady, C. 1992. Testimony: A Philosophical Study. Oxford: Clarendon Press. Coady, C. 1994. “Testimony, Observation and Autonomous Knowledge.” in Knowing from Words. Eds. B. K. Matilal and A. Chakrabarti, 225-250. Dordrecht: Kluwer Academic Publishers. Cummins, Robert. 1975. “Functional Analysis.” Journal of Philosophy 72: 741-60. Cummins, Robert. 1983. The Nature of Psychological Explanation. Cambridge, MA: M.I.T. Press. Davidson, Donald. 1983: “A Coherence Theory of Truth and Knowledge.” In Kant oder Hegel. Ed. Dieter Henrich. Stuttgart: Klett-Cotta. Fodor, Jerry. 1983. The Modularity of Mind: An Essay in Faculty Psychology. Cambridge, MA: M.I.T. Press. Goldberg, Sanford and Henderson, David. Forthcoming. “Monitoring and Reductionism in the Epistemology of Testimony.” Philosophy and Phenomenological Review. Goodman, Nelson. 1973. Fact, Fiction, and Forecast. 3rd edition. Indianapolis: Bobbs-Merrill. Greco, John. 2000. Putting Skeptics in Their Place. Cambridge: Cambridge University Press. Henderson, David.1994. “Epistemic Competence and Contextualist Epistemology.” Journal of Philosophy 91: 627-49. Henderson, David. Forthcoming. “Testimonial Belief and Epistemic Competence.” Nous. Henderson, David and Horgan, Terrence. 2000. “Iceberg Epistemology.” Philosophy and Phenomenological Research 61: 497-535. Henderson, David and Horgan, Terrence. Forthcoming. “The Ins and Outs of Transglobal Reliabilism.” In Internalism and Externalism in Semantics and Epistemology. Ed. Sanford Goldberg. Oxford: Oxford University Press. Horgan, Terrence and Tienson, John. 1994. “A Nonclassical Framework for Cognitive Science.” Synthese 101: 305-45. Horgan, Terrence and Tienson, John. 1995. “Connectionism and the Commitments of Folk Psychology.” Philosophical Perspectives 9: 127-52. Horgan, Terrence and Tienson, John. 1996. Connectionism and the Philosophy of Psychology. Cambridge, MA: M.I.T. Press. Kitcher, Philip. 1992. “The Naturalist Returns.” Philosophical Review 101: 53-114. Montmarquet, James. 1987. “Epistemic Virtue.” Mind 96: 482-97. Moser, Paul. 1989. Knowledge and Evidence. Cambridge: Cambridge University Press.

Epistemic Virtues and Cognitive Dispositions

319

Pollock, John. 1986. Contemporary Theories of Knowledge. Lanham, Md: Rowman and Littlefield. Sosa, Ernest. 1980. “The Raft and the Pyramid: Coherence Versus Foundationalism in the Theory of Knowledge.” Midwest Studies in Philosophy 5: 3-25. (Reprinted in Sosa 1991.) Sosa, Ernest. 1991a. Knowledge in Perspective. Cambridge: Cambridge Univ. Press. Sosa, Ernest. 1991b. “Reliabilism and Intellectual Virtue.” In Knowledge in Perspective. E. Sosa, 131-45. Cambridge: Cambridge University Press. Shapere, Dudley. 1982. “The Concept of Observation in Science and Philosophy.” Philosophy of Science 49: 485-525. Woodward, James. 2003. Making Things Happen. Oxford: Oxford University Press. Zagzebski, Linda. 1996. Virtues of the Mind. Cambridge: Cambridge University Press.

The Epistemic Function of Virtuous Dispositions ELKE BRENDEL In most accounts of virtue epistemology, knowledge and other epistemic goods are defined in relation to intellectual virtues. Intellectual virtues are either cognitive faculties or character traits of various kinds. It is argued that it is appropriate to regard intellectual virtues as dispositions of a certain kind. Thinking of intellectual virtues as dispositions provides important insights into the nature and epistemological significance of intellectual virtues. But the attempt to define knowledge in terms of virtuous dispositions is only a feasible and promising epistemological project if some of the central concepts and ideas of virtue epistemology are revised or at least refined.

1. The Main Claims and Tenets of Virtue Epistemology Virtue epistemology is a recent movement in epistemology whose central thesis is that major epistemological concepts, like knowledge or epistemic justification should be understood with reference to certain cognitive or intellectual virtues. According to virtue epistemologists, epistemology is fundamentally normative in character. Just as in virtue ethics, where we praise moral agents for acting out of moral virtues, in virtue epistemology we commend the beliefs of an epistemic subject in light of virtuous intellectual character traits that the epistemic subject possesses. Virtue epistemology is therefore often characterized as being “subject based” rather than “belief based,” since properties of beliefs are not the main focus of their epistemic evaluation. Rather, the intellectual properties of the epistemic agent are relevant in this context. There is a plethora of different approaches to virtue epistemology. Virtue epistemologists disagree about the nature of intellectual virtues, about the exact epistemic role intellectual virtues play, and about the actual tenets of virtue epistemology. Some epistemologists maintain that the concept of an intellectual virtue helps us to resolve traditional disputes in epistemology, such as the foundationalism-coherentism dispute,1 or provides us with adequate solutions to some notorious problems of knowledge, like skepticism,

_____________ 1

See Sosa 1980.

The Epistemic Function of Virtuous Dispositions

321

the Gettier problem or the lottery paradox.2 Others think of virtue epistemology as shifting the focus of traditional epistemology. Whereas traditional epistemology focuses on the individual knower and on what it is for such a knower to possess propositional knowledge, some virtue epistemologists are concerned with the social aspects of knowledge and with notions like understanding and wisdom.3

2. Dispositions in Virtue Epistemology Even if virtue epistemologists disagree about the precise nature of intellectual virtues, and, in particular, about the nature of the intellectual virtues that give rise to justified beliefs or knowledge, in most accounts the notion of a certain kind of disposition plays a crucial role in characterizing intellectual virtues. For example, Roberts and Wood understand virtue epistemology as exploring “dispositional properties of persons that bear on the acquisition, maintenance, transmission, or application of knowledge and allied epistemic goods such as truth, justification, warrant, coherence and interpretative fineness.”4 A virtue reliabilist such as Ernest Sosa regards intellectual virtues as “stable dispositions for belief acquisition.”5 According to Sosa these virtuous dispositions are acquired or innate faculties, like perception (eyesight, hearing), memory, introspection, rational intuition, and logical reasoning, the proper exercise of which leads to the formation of true beliefs in a reliable way.6 This notion of an intellectual virtue is used to provide an account of the nature of knowledge and epistemic justification: Knowledge, according to Sosa is “true belief out of intellectual virtue, belief that turns out right by reason of the virtue and not just by coincidence.”7 Following Lorraine Code and James Montmarquet’s accounts of virtue responsibility,8 Linda Zagzebski has developed a detailed theory of virtue epistemology in which intellectual virtues are conceived of as analogous to moral virtues in a neo-Aristotelian sense. In particular, she stresses the point that virtues in general consist of two elements: an element of reliable success and a motivational element: A virtue, then, can be defined as a deep and enduring acquired excellence of a person, involving a characteristic motivation to produce a certain desired end and reliable suc-

_____________ 2 3 4 5 6 7 8

See, for example, Greco 2000a, Greco 2003, and Greco 2005. See, for example, Zagzebski 1996, 43ff., Code 1987, Montmarquet 1993, and Riggs 2003. Roberts/Wood 2003, 257. Sosa 1980, 189 – my emphasis. See Sosa 1991, 277ff. Sosa 1991, 277. See Code 1987, Montmarquet 1993.

322

Elke Brendel

cess in bringing about that end. What I mean by a motivation is a disposition to have a motive; a motive is an action-guiding emotion with a certain end, either internal or external.9

An intellectual virtue thus consists in part of the disposition to have a desire for true beliefs and in part of a reliable mechanism for attaining true beliefs. Zagzebski does not conceive of intellectual virtues as natural faculties or cognitive abilities in Sosa’s sense. She thinks that faculties such as eyesight and memory “are not virtues at all in traditional virtue theory.”10 For her, like for Montmarquet, intellectual virtues are certain character traits of the epistemic subject such as open-mindedness and fair-mindedness, intellectual integrity, intellectual carefulness and humility, perseverance, impartiality, and flexibility. Zagzebski then defines knowledge as “a state of true belief arising out of acts of intellectual virtue.”11 In his version of what he calls agent reliabilism,12 John Greco combines a subjective condition of epistemic responsibility with a condition of objective reliability in order to give necessary conditions for knowledge. The notion of a disposition plays a crucial role in his analysis of knowledge: in order for S to know that p, S must be, according to Greco, subjectively justified in believing p, which means that “S’s believing p is the result of dispositions that S manifests when S is thinking conscientiously, or is motivated to believe the truth.”13 In addition, S must also be objectively justified in order to know that p, that is, the dispositions that result in S’s believing p make S reliable in believing p (in the present conditions, with respect to p.)14 Greco also thinks of intellectual virtues as certain kinds of cognitive dispositions – dispositions that manifest themselves when thinking conscientiously. Only if these intellectual virtues are the actual causes of the belief that p, can it count as knowledge.15

_____________ 9 10

11 12 13 14 15

Zagzebski 1996, 137 – my emphasis. Zagzebski 1996, 9. Greco objects to this. With Sosa he points out that it is perfectly in accordance with the Greek tradition to understand virtues as innate or acquired faculties. In Plato’s Republic (Book I, 342-352), vision is called “the virtue of the eyes and hearing the virtue of the ears.” – see Greco 2000b, 180. Goldman also contends that Sosa’s (and Greco’s) account is based on an Aristotelian conception of excellences since an excellence in the Aristotelian tradition “is some kind of ability, disposition, power, faculty, or habit.” (Goldman 2001, 30). Zagzebski 1996, 271. See Greco 2000a, 164ff. See Greco 2000a, 218, Greco 2002, 303 or Greco 2004, 6 – my emphasis. See Greco 2000a, 218. In Greco 2003, 128, Greco added a third condition that together with the other two conditions should result in necessary and sufficient conditions for knowledge: “S believes the truth regarding p because S is reliable in believing p. Alternatively: the intellectual abilities (i.e., powers or virtues) that result in S’s believing the truth regarding p are an important necessary part of the total set of causal factors that give rise to S’s believing the truth regarding p.” Greco claims that this added third condition allows us to handle the Gettier problem. In Gettier cases this third condi-

The Epistemic Function of Virtuous Dispositions

323

For the moment, I want to postpone consideration of the adequacy of these different approaches to virtue epistemology or their success (or otherwise) in giving appropriate definitions of knowledge and other central epistemic concepts. Instead I want to focus on the notion of a disposition that many virtue epistemologists use in order to characterize their accounts of the nature of intellectual virtues. So far we have seen that there are various kinds of intellectual virtue and that the concept of a disposition is used to describe the epistemic theories that are based on these intellectual virtues: First, intellectual virtues are seen as general (innate or acquired) mental faculties and cognitive abilities or competences – such as good eyesight or hearing, good memory, introspection and logical reasoning – or they are conceived of as more specific cognitive skills, capacities or faculties, such as the faculty of grasping certain patterns, of seeing coherent structures or interconnections between things, or of keeping track of complex argument structures.16 These mental faculties and abilities are epistemic virtues, since they are considered to be stable and reliable dispositions for forming accurate beliefs about the world. The second kind of intellectual virtues are character traits, such as openmindedness, conscientiousness, perseverance, intellectual curiosity and creativity, impartiality before evidence, unbiased examination of arguments. Roberts and Wood also regard character traits such as honesty, charity, fairness and humility as epistemic virtues. They are “dispositional properties that tend to the moral but make people better or worse epistemic agents.”17 Other character traits that Sherman and White call “emotional dispositions” are also considered to be intellectual virtues, such as “a passion for the truth, a delight in learning, excitement in discovery, pride in one’s accomplishments, respect for good arguments, repugnance at intellectual dishonesty, and in the case of empirical science, surprise at the disconfirmation of one’s theory and joy at its verification.”18 Taking these forms of intellectual virtues as certain kinds of dispositions, having such dispositions seems to be a necessary condition for the acquisition of knowledge or for the achievement of other epistemic ends in virtue epistemology. Virtue epistemologists either use the term “disposition” in a rather vague manner or they understand it as a non-analyzed pre-theoretical term, whose meaning is supposed to be precise enough for the given purpose. But it is not at all clear whether it is really legitimate to call all of the intellectual virtues listed above “dispositions.” In order to figure out which intellectual

_____________ 16 17 18

tion seems to be violated since the epistemic subject believes the truth regarding p in such cases out of luck and not because of S’s intellectual abilities. See Hookway 2003, 187. Roberts/Wood 2003, 257. Sherman/White 2003, 38.

324

Elke Brendel

virtues can be regarded as dispositions, and what kind of epistemic function they have in a virtue-based epistemology, it is first necessary to analyze the notion of a disposition as applied to intellectual virtues in more detail and to develop an account of dispositions that satisfies important requirements that virtue epistemologists (implicitly) assume when talking about intellectual virtues as certain kinds of disposition.

3. Some Claims about Dispositions The vast number of different accounts of dispositions in the philosophical literature shows that there is a disagreement on almost all crucial questions about the ontology, semantics, epistemology and scientific function of dispositions. Of course, I do not claim that I can give necessary and sufficient conditions for possession of a disposition or that doing so would provide us with satisfactory answers to all the important questions in this area. These questions include the following: what is the proper analysis of disposition ascriptions? What is the ontological status of dispositions? How is the dispositionalcategorical distinction to be understood? Are dispositions intrinsic or extrinsic? What is their causal or explanatory function? Instead, I will provide a list of criteria for a property being a disposition that, to my mind, any plausible explication of the nature of dispositions has to account for. I shall then turn to the question of whether intellectual virtues fulfill these criteria. Let us first start with some uncontroversial claims about dispositions: 1. Dispositions are powers or capacities that are ascribed to objects, such as individuals, artefacts, substances, and living beings (animals, persons). Fragility, solubility, combustibility, being poisonous are typical examples of (physical) dispositions that we attribute to (tokens or types of) glass vases, sugar, pieces of wood, and fly agarics and so on. Dispositions can also be mental properties that we normally ascribe to humans and/or non-human animals, such as generosity, open-heartedness, tameness etc. 2. Dispositions have specific manifestations and manifest themselves under certain circumstances. The typical manifestation of fragility is breakage. Typically, the fragility of a glass vase manifests itself when the vase is dropped to a solid floor or smashed with a hammer etc. A fragile glass vase is thus disposed to break when exposed to some of these circumstances, i.e., when the vase is exposed to conditions for manifesting the disposition of fragility. In a situation in which a friend asks a generous person for money, this person is normally disposed to give him money freely and more than this friend expected. Trying to characterize dispositions in terms of typical circumstances in which the disposition in question manifests itself gives rise to serious prob-

The Epistemic Function of Virtuous Dispositions

325

lems. How can these manifesting situations be described in an analysis of disposition ascriptions? What are the normal or typical circumstances under which a disposition manifests itself? Is the actual manifestation of a disposition necessary for truly ascribing a disposition to an object? Do we still call a property a disposition if it is manifested all the time or most of the time? In answering these questions I have to give more specific (and, I guess, more disputable) criteria of dispositions: 3. An object possesses a disposition in virtue of certain intrinsic features of the object. These intrinsic features are among the causes that explain why an object manifests a disposition under certain circumstances. A glass vase possesses the disposition of fragility, i.e., normally breaks when struck, because of its specific molecular structure. A piece of wood is disposed to float on water in virtue of its buoyancy, i.e., its specific density compared to the density of the water. If an object o shows some reaction r under certain circumstances c, r can only be regarded as a manifestation of a disposition d, if o possesses d in virtue of intrinsic features i and i is causally relevant for r. The intrinsic features of objects in virtue of which they possess a certain common disposition can, of course, be different. A certain glass vase, a certain tea cup, a certain marble statue etc. are all fragile, but they are fragile in virtue of different intrinsic features, i.e., different molecular structures.19 4. Dispositions are fairly stable features of an object. We would not call a chemical substance soluble that changes its molecular structure permanently so that it is sometimes soluble in water and sometimes not. We would not call a person generous if he is generous one day and miserly the other day. 5. A disposition can come in different degrees. A thin glass vase is more fragile than a solid coffee-mug. Some people are more or less generous or more or less open-hearted than others. 6. A disposition can manifest itself in multiple ways.

_____________ 19

I will leave it open whether calling all these different things fragile means that they really possess the very same disposition or whether it is just a matter of linguistic convenience to subsume them under the same label. John Heil, for example, objects to the view that dispositions are multiply realized “higher-level properties”: “We find it convenient to say that a teacup, a piece of slate, a pocket watch, and a gramophone record all possess the same disposition: being fragile. These items are examples of things that typically shatter when struck or dropped. But do they, on that account, possess the very same disposition? That seems unlikely: the objects shatter in different ways. To be sure, the shatterings are similar enough to fall under a single predicate. But the similarity in question is far from precise. I take it as uncontroversial that, if distinct objects possess the very same property, F, they must be precisely similar F-wise. To assume that “is fragile” must name a higher-level property is to let the linguistic tale wag the ontological dog.” (Heil 2005, 347).

326

Elke Brendel

The fragility of a glass, for example, can manifest itself differently under different circumstances: dropping the glass to the floor, smashing it with the hammer etc. Stephen Mumford gives another example: “The multiple manifestations of elasticity include stretching, contracting, bouncing, deforming, and reforming.”20 7. Dispositions normally don’t manifest themselves in isolation, but in cooperation with other dispositional properties. A glass vase which is dropped to a floor breaks because of its molecular structure and the molecular structure of the floor. John Heil stresses this point by claiming that: “The manifestation of a disposition is a manifestation of reciprocal disposition partners.” Here are his examples: A salt crystal manifests its disposition to dissolve in water by dissolving in water. But this manifestation is a manifestation of both the salt crystal’s disposition to dissolve in water and the water’s reciprocal disposition to dissolve salt. A match bursts into flame when it is scratched across the abrasive surface of a matchbox. The match’s bursting into flame is a manifestation of dispositions possessed by the match, the surface of the matchbox, and the surrounding air.21

Since the manifestation of an object’s disposition also depends on the object’s “reciprocal disposition partners” it is not just in virtue of the intrinsic features of the object that a disposition manifests itself. This gives rise to the following claim: 8. Even though a disposition of an object o manifests itself at least partly in virtue of some of o’s intrinsic features, dispositions are not entirely intrinsic in nature, i.e. an object does not possess its dispositions regardless of what is going on outside of itself. An object o can also change its dispositions or cease to possess some of its dispositions if some extrinsic features are changed. A prominent example is Sydney Shoemaker’s “mere-Cambridge power.”22 Shoemaker’s key to his house not only possesses the disposition to open locks of a certain shape, but also possesses the disposition to open his front door. Although the key possesses these dispositions in virtue of some of its intrinsic properties, the latter disposition can be lost without undergoing any changes of the key’s intrinsic properties. The key can lose the disposition to open Shoemaker’s front door simply by replacing the lock of his front door by a lock with a completely different shape. A similar example is Jennifer McKitrick’s “disposition to dissolve the contents of my pocket”:23 Whether a bucket of water possesses the disposition to dissolve the content of her pocket does not only depend on

_____________ 20 21 22 23

Mumford 1998, 6. Heil 2005, 350. See Shoemaker 1984, 221. See McKitrick 2003, 160.

The Epistemic Function of Virtuous Dispositions

327

the water’s intrinsic properties, but also on the actual content of McKitrick’s pocket. Changing or destroying an object’s dispositional properties by extrinsic interferences is a quite common phenomenon.24 Another common feature of dispositions discussed in the literature is the phenomenon of “masked” manifestations of dispositions: 9. The manifestation of a disposition can be “masked”, i.e. inhibited by some extrinsic interference. In these cases the disposition is not changed or destroyed by extrinsic interference. The object still possesses its disposition, but the disposition is prevented from manifesting itself by some extrinsic measures. A well-known example of such a “masking” condition is Mark Johnston’s fragile glass cup. which is prevented from breaking if struck by protecting packing material. The glass cup is still fragile, it still has the disposition to break when dropped, even though it would not – under the given circumstances – break when dropped.25 Another famous example is Alexander Bird’s “antidote” case where the manifestation of a poisonous substance, i.e., to kill a person after ingesting the substance, is inhibited by administering an effective antidote.26 Although the manifestation of the disposition is masked, the substance is still poisonous and still possesses the disposition to kill people when ingested. If it is true that objects possess dispositions in virtue of some of their intrinsic properties, the actual manifestation of a disposition is therefore not necessary for an object to possess that disposition. We can still truly ascribe a disposition to an object even if the disposition has never manifested itself. Since the fragility of a glass vase is explained by its specific molecular structure, we can truly ascribe fragility to a glass vase with such a specific molecular structure even if this glass vase has never encountered an actual breakage. Similarly, a sugar cube is soluble even if it has never been dissolved in water, and a fly agaric is poisonous even if nobody has ever ingested it. In some cases the actual manifestation of a disposition would lead to the destruction or damage of the object (as on the above cases). In other cases the manifestation does no harm to the object – as in the case of the manifestation of elasticity of an elastic band by moderately stretching it. In cases of mental dispositions this claim seems to be much more disputable. Can we still truly ascribe generosity to a person who never showed any signs of generosity in his whole life? Can we say a person is open-hearted even if she never or hardly ever displays her open-heartedness? If a person’s generosity or open-heartedness

_____________ 24

25 26

See, for example, McKitrick’s discussion of other examples, such as “vulnerability”, “visibility” or “recognizability” (McKitrick 2003, 161ff.), which show that in certain contexts dispositions can be extrinsic. See Johnston 1992, 223. See Bird 1998.

328

Elke Brendel

consists in some intrinsic properties of that person it seems that, as in the cases of physical dispositions like fragility, solubility, being poisonous, elasticity etc., the actual manifestation of some mental disposition is not necessary for truly ascribing this disposition to the person. That is why Mumford writes: […] a case could be made for saying that someone is brave even though they have never acted bravely. There may have been no appropriate circumstances in which they could have acted bravely, or if they were in such a situation they may have been drunk or affected by food additives.27

Since open-heartedness, for example, is a much more complex property than say fragility of an object and since open-heartedness does clearly not consist in one single neurological state of a person, a person’s behavior is often our only indication of ascribing dispositional properties to a person. Furthermore, it seems to be very unlikely that a nevertheless open-hearted person has never or almost never been in appropriate situations in which she could actually be open-hearted. That is why we are reluctant to believe that a person possesses some mental dispositional property if a person never or almost never acted in a way in which this disposition is manifested. But this doesn’t mean that the actual manifestation of a disposition is a necessary condition for defining a person’s possession of a disposition. Dispositions are hidden powers or capacities of objects. They are, as Nelson Goodman put it, “rather ethereal”. They are the “threats and promises” of a thing that a thing possesses besides its observable properties.28 They are as Mumford describes them with regard to Goodman, “properties that somehow are not always manifest but which seem to lurk in a mysterious realm intermediate between potentiality and actuality.”29 The way I characterized dispositions so far does not necessarily banish them to a “mysterious realm intermediate between potentiality and actuality.” Dispositions are fairly stable properties that objects possess in virtue of some of their intrinsic features, and they manifest themselves under certain appropriate circumstances. But even if dispositions are seen as hidden powers of an object, this does not mean that an object does not possess a disposition any longer just because the disposition is always or most of the time manifest. The disposition does not cease to exist just because the disposition reveals itself by being permanently manifested. A piece of wood that has been floating on water for the last ten years still possesses the disposition to float on water. After all this piece of wood still has its special intrinsic property (its specific density) in virtue of which it is able to float on water. Similarly, a person who acts in an aggressive way almost all of his life still has the disposition to be aggressive, and a person

_____________ 27 28 29

Mumford 1998, 8. See Goodman 1954, 40. Mumford 1998, 4.

The Epistemic Function of Virtuous Dispositions

329

who acts in an open-hearted way most of the time still has the disposition to be open-hearted. But I think most people would nevertheless be intuitively reluctant to say that this piece of wood is disposed to float on water, just as most people are reluctant to say that a person is disposed to be aggressive or is disposed to be open-minded if this person acts mostly aggressive or mostly openhearted. We simply say that this piece of would floats on water, or that this person is aggressive, or that that person is open-minded. We can thus formulate our last claim about dispositions: 10. It is part of the pragmatics of disposition ascriptions that an assertion of the form “Object o is disposed to x under appropriate circumstances c” tends to imply pragmatically that the disposition d that manifests itself by x under circumstances c has not always been actually manifested, i.e., that the appropriate circumstances under which this disposition normally manifests itself are not always present. But this does not semantically imply that the assertion is literally false. If a dispositional property d of an object o is manifested all or most of the time by the outcome x, it seems to be inappropriate to say that o is disposed to be x. In such a case we rather tend to simply say that o is x.30 It is nevertheless important to see that in cases in which the disposition has been manifested all or almost all the time by the outcome x, a sentence of the form “o is disposed to x under appropriate circumstances c” is not false. It seems to be a kind of a violation of Grice’s conversational maxim of quantity if a person says “o is disposed to x” instead of simply saying “o is x” if o is in fact always or most of the time x. In such cases, saying “o is disposed to x” would not be as informative as is possible – although the sentence would not turn out to be literally false.

4. Are Intellectual Virtues Dispositions? With these claims about dispositions at hand (which, of course, describe only a few of the necessary criteria for a property being a disposition), we can now try to answer our initial question of whether the intellectual virtues that are the main focus of virtue epistemology can be regarded as dispositions. We have seen that there are two main categories of intellectual virtues: 1. certain kinds of mental faculties and cognitive abilities, and 2. certain kinds of character traits. Both types of intellectual virtue can be seen as capacities that are ascribed to epistemic subjects. An epistemic subject possesses these capacities in virtue of some innate or acquired intrinsic properties of the subject. They are called “intellectual

_____________ 30

… unless we use this disposition ascription to explain o’s behaviour.

330

Elke Brendel

virtues” since the proper exercise of these capacities aims at achieving epistemic goods, such as justified beliefs, knowledge, understanding or wisdom. In order to count as an intellectual virtue that is a means of acquiring such epistemic goods, it is necessary that these capacities are fairly stable and reliable faculties or rather permanent personal excellences of the epistemic subject. S can only acquire perceptual knowledge via good eyesight if good eyesight is a fairly stable and reliable faculty of S. If S’s visual organ only occasionally functions properly, S’s beliefs that are formed on the basis of S’s visual experiences would not be considered as knowledge – even if these beliefs happen to be true. Furthermore, we would not call fair-mindedness, intellectual integrity, perseverance, impartiality etc. a “character-trait” of a person S if these traits are not rather permanent excellences of S’s character. In the same vein, Douglas Butler remarks: “Character traits must be fairly permanent features of one’s personality. One is not usually said to have the trait of kindness on Monday, to lack it on Tuesday, to have it again on Wednesday, and so on.”31 In addition, mental faculties, cognitive abilities and character traits clearly come in degrees and can manifest themselves in multiple ways. The quality of mental faculties, such as good eyesight or hearing, can differ among people. One person’s logical reasoning abilities can be better or worse than those of another person. One person can be more persevering in her intellectual inquiries and deliberations than another person etc. Many character traits that virtue epistemologists regard as intellectual virtues are rather abstract and complex excellences of a person. Accordingly, these character traits have a broad variety of possible manifestations in the intellectual activities of epistemic subjects. Consider, for instance, the many different ways in which a person can express his or her intellectual creativity. Furthermore, an intellectual virtue as a means of acquiring and maintaining knowledge and other epistemic goods typically manifests itself in intellectual inquiries and deliberations in coordination with other cognitive excellences. In order to acquire knowledge in a specific scientific field, a complex interaction between the exercise of different intellectual virtues is normally necessary, such as perceptual faculties, logical reasoning, the competent application of specific cognitive skills and expert knowledge in this field, character traits, such as curiosity, intellectual creativity, the capacity to see relevant interconnections and coherent structures, the rigorous and unbiased examination of evidential data and a motivation for truth and scientific success. Sosa contends that even in cases of alleged spontaneous perceptual beliefs (such as the belief that there is a laptop in front of me); it is not by the exercise of our visual organs alone that we form a reliable true perceptual belief:

_____________ 31

See Butler 1988, 231.

The Epistemic Function of Virtuous Dispositions

331

For even when perceptual belief derives as directly as it ever does from sensory stimuli, it is still relevant that one has not perceived the signs of contrary testimony. A reason-endowed being automatically monitors his background information and his sensory input for contrary evidence and automatically opts for the most coherent hypothesis even when he responds most directly to sensory stimuli.32

So, the reliability even of our most spontaneously formed perceptual beliefs by the exercise of our visual faculties, is, according to Sosa, implicitly being monitored by the exercise of other intellectual virtues, such as a coherence check on the basis of background information. Abrol Fairweather goes even further and claims that another intellectual virtue, namely the desire for truth, is a necessary condition for knowledge. According to Fairweather, evidence can only play a justifying role in our belief formation if we have an appropriate epistemic motivation, i.e., if the desire for truth guides our belief formation. Fairweather contends that the mere possession of justifying evidence for a belief that p is not itself sufficient for the epistemic subject to be justified in believing that p. In Fairweather’s example of “Conrad, the Doxastic Conformist”,33 Conrad’s primary goal is to bring his belief system into conformity with the belief system of Mr. Cool, his role model. One day Mr. Cool expressed this belief about the outcome of the election in November and his justifying grounds for his belief in conversation with Conrad. Directed by his desire to be in conformity with Mr. Cool’s beliefs, Conrad adopts this belief and the justifying grounds for it. Although Conrad possesses good evidence for his belief, the belief is not based on the good evidence. Fairweather claims that since Conrad’s belief is not mobilized by his desire for truth (but rather by his desire to conform), his belief cannot count as justified. According to Fairweather, the desire for truth also needs to be properly exercised in order for a perceptual belief to have a positive epistemic status: Sosa makes the important point that the credibility of sensory reports can always be overridden by background information, and even when we accept the reports of our senses our monitoring system is a silent partner in producing belief. But our monitoring system only makes a positive contribution to the epistemic status of our beliefs when it is controlled by a desire for truth. I think this shows that epistemic motivations play a significant justificatory role even in the case of perceptual beliefs.34

The proper exercise of intellectual virtues in our intellectual and cognitive endeavours is of the utmost importance for the achievement of our epistemic goals. In order to analyze epistemic justification or knowledge in terms of intellectual virtues, virtue epistemologists must explain what the proper exercise of intellectual virtues consists in. What are, for example, “acts of intellectual virtue” that give rise to true belief? In my general analysis of dispositions

_____________ 32 33 34

Sosa 1991, 240. Fairweather 2001, 74. Fairweather 2001, 78.

332

Elke Brendel

I pointed out that dispositions normally manifest themselves in cooperation with “reciprocal disposition partners.” We will now see that the exercise of intellectual virtues that aim at achieving epistemic goods is particularly determined by their proper cooperation with the exercise of other virtues. Only in concert with their “reciprocal disposition partners” can intellectual virtues contribute to knowledge or to other epistemic ends. I shall further argue that pathological manifestations of intellectual virtue can occur if intellectual virtues are not properly “in tune” with their reciprocal partners. We have already seen that only through the proper interaction of the two kinds of intellectual virtue can we successfully and reliably reach epistemic goods. Christopher Hookway also strongly emphasises this point: […] we would not be reliable seekers after the truth or effective solvers of theoretical problems if we did not possess specific skills and capacities: good eyesight and hearing, a reliable memory, good knowledge and specific subject matters and so on; but our success also requires us to possess traits of character which enable us to use our skills and capacities effectively when inquiring and deliberating. Both kinds of ‘virtues’ can contribute to our reliability or to our cognitive success.35

So, in order to achieve our desired epistemic ends a complex interaction between natural faculties, cognitive abilities, specific cognitive skills, character traits as well as epistemic motivation is necessary. The grade of complexity of the interaction between these different intellectual virtues is, of course, dependent on the specific subject of our cognitive enterprises. In order to achieve a true perceptual belief (like the belief that there is a laptop in front of me) the exercise of much less intellectual virtue is needed than for acquiring knowledge about some advanced scientific topic. Clearly, some intellectual virtues are more important for reaching our epistemic ends than others. Some intellectual virtues only play a minor role in our intellectual inquiries, some only have a supportive function in cooperation with other intellectual virtues enabling us to reach our epistemic goals. So, for example, humility, an intellectual virtue that Roberts and Wood examine in great detail, is certainly a praiseworthy character trait of an epistemic subject that can support the formation of true beliefs. In contrast to a vain and arrogant person an intellectually humble person can better overcome obstacles to the acquisition of knowledge “in the long run”. But it only has a “narrowly epistemic advantage,” one that can only be displayed in conjunction with other epistemic virtues: The thesis is that intellectual humility fosters certain intellectual ends when it is conjoined, in a personality, with other epistemic virtues. Our claim is not that all people who lack humility will be in all respects epistemic failures; we even think that vanity, arrogance, and other anti-humility vices can on occasion contribute to the acquisition, refinement, and communication of knowledge. Rather, we claim that over the long

_____________ 35

Hookway 2003, 188.

The Epistemic Function of Virtuous Dispositions

333

run, just about everybody will be epistemically better off for having, and having associates who have, epistemic humility.36

Furthermore, there is sometimes a gradual transition from virtues to vices. The exercise of alleged intellectual virtues in isolation without a proper balancing with other intellectual virtues can result in various pathological forms of “virtues.” In such cases, the virtuous character of these “intellectual excellences” can swing to the other extreme. Pathological exercises of alleged intellectual virtues can even prevent epistemic subjects from reaching their epistemic aims. A person who is too thorough, too critical and too scrupulous and who always challenges even the best evidence for a hypothesis cannot be regarded as an intellectually responsible person. Exaggerated prudence can become an obstacle to a person’s epistemic advancements.37 As with other dispositions, the manifestation of intellectual virtues can be disrupted or inhibited. Many different factors can result in the disruption or “masking” of the proper exercise of an intellectual virtue. The exercise of natural and cognitive faculties such as good eyesight or reliable memory can be “masked”, for example, by hallucinogenic drugs. The manifestation of logical or good analytic reasoning skills can be inhibited by a momentary tiredness and exhaustion of an epistemic subject, who normally possesses excellent skills in logical reasoning. A person, who normally bases her beliefs on a careful and impartial analysis of given empirical data, can act carelessly or come to hasty conclusions as a result of a momentary outburst of anger or sorrow. We have seen so far that both kinds of intellectual virtue fulfill the criteria for dispositions given in claims 1-9: They are fairly stable capacities ascribed to epistemic subjects, and they have specific manifestations. Epistemic subjects possess these capacities in virtue of some intrinsic features. Intellectual virtues come in different degrees and manifest themselves in multiple ways. More importantly, they manifest themselves in an often complex cooperation with other intellectual virtues in the formation of true beliefs. Manifestations of intellectual virtues can also be changed, disrupted or “masked” by some external interference. Whereas intellectually virtuous character traits are intuitively rather clear-cut cases of (mental) dispositions, it seems to be somehow odd to say that the exercise of a natural faculty such as good eyesight manifests a certain disposition. A person with good eyesight simply has this faculty. We would not normally say that this person is disposed to certain manifestations of good eyesight

_____________ 36 37

Roberts/Wood 2003, 271f. Zagzebski also mentions other kinds of virtue the improper exercise of which can turn into vices, like William James’ “questioning mania,” or virtues that are exercised in a way such that they are “not so much vices as a waste of intellectual energy that could be put to better use.” – See Zagzebski 1996, 153.

334

Elke Brendel

under appropriate circumstances. But since natural faculties and character traits share very important features of properties normally regarded as dispositions (see claim 1-9), it seems to me theoretically unjustified and premature to exclude natural faculties from the extension of the concept of a disposition. I would rather argue that our tendency not to use the term “disposition” with regard to such faculties can be explained by the pragmatics of disposition ascriptions as outlined in claim 10. For a person with good eyesight, the appropriate circumstances under which this faculty manifests itself are frequent and almost always actual during waking hours. So, saying that a person exercising the faculty of good eyesight most of the time is disposed to accurate visual perceptions of her environment would lead to a false pragmatic implicature – namely, that with regard to this person this disposition is not mostly manifested. But this does not mean that the person does not actually possess a disposition for accurate visual perception of her environment.

5. On the Prospects of Defining Knowledge in Terms of Virtuous Dispositions In the preceding paragraph, I argued that it is appropriate to regard intellectual virtues that function as means of acquiring knowledge (and other epistemic goods) in different accounts of virtue epistemology as virtuous dispositions. In particular, in applying certain crucial features of dispositions to intellectual virtues we gained some important insights into the nature of intellectual virtues and their epistemic role in our intellectual inquiries and deliberations. I shall now finally turn to the question of whether defining knowledge in terms of these virtuous dispositions is a feasible and promising epistemological project. An internalist deontological account of virtue epistemology according to which a person S is only in a position to know that p, if S is epistemically responsible for holding her belief that p, or if we can praise S for holding the belief that p because of a conscious and well-considered act of intellectual virtues, faces serious problems. Intellectual virtues differ greatly in the degree of conscious and voluntary control we have over them. The extent to which we have control over our cognitive facilities is, of course, an empirical matter. But if a person S has no control over the manifestation of a certain cognitive faculty in a certain situation (or only to a very low degree), it seems inappropriate to say that we praise S for forming a true belief on the basis of the manifestation of such a cognitive faculty or that S is epistemically responsible for forming that belief. But this does not necessarily exclude S from having knowledge. Greco stresses this point as well. For him praiseworthiness and epistemic responsibility is therefore not a necessary condition on knowledge:

The Epistemic Function of Virtuous Dispositions

335

Suppose it turns out that human perception is not within our strong control, even indirectly. Suppose also that, nevertheless, our perception is as reliable as we ordinarily assume it is. Would we hesitate to say that we can know what we clearly see to be the case? For example, would we say that I cannot know that there is a truck bearing down on me because, as it turns out, I cannot help but believe that there is, given the truck that I see? Or suppose that we come across alien beings who have far more accurate and reliable perception than we do, but who do not have the power to believe otherwise than how they perceive. It seems wrong to say that, for lack of control in this strong sense, such beings do not have perceptual knowledge. 38

In some cases the reliability of our perceptual faculties resulting in the formation of a true belief is sufficient for having knowledge – whether or not we have control over these faculties. Yet, I do not think that this insight necessarily undermines the general idea behind the attempt of defining knowledge in virtue epistemology. In many cases, merely making proper use of a reliable perceptual faculty is not sufficient for having knowledge. More abstract, less trivial and more profound forms of knowledge than the perceptual knowledge that there is a laptop in front of me or that there is a truck bearing down on me, such as expert knowledge in certain scientific fields, clearly call for a conscious and well-considered exercise and complex cooperation of different intellectual virtues. But even in cases of perceptual knowledge, ascribing knowledge to a person depends on a given context. The salience of an errorpossibility can cast doubt on someone’s possession of knowledge, even if the person has a true belief obtained by normally reliable means such as sense perception. The preceding considerations suggest that defining knowledge as true belief that arises in a certain way out of intellectual virtue implies a contextualist account of knowledge.39 There are different dimensions of context-sensitivity of the truth-conditions of knowledge-ascriptions of the form “S knows that p” in accounts of knowledge in virtue epistemology: First, the degree of the reliability of S’s intellectual virtues, the exercise of which result in S’s formation of the true belief that p, depends on various context-dependent standards operative in the specific situation of knowledge-ascription, like the salience of error-possibilities of p, speaker’s intentions, questions of what is at stake in the given context etc.40 Second, what kinds of intellectual virtue and the extent

_____________ 38 39 40

Greco 2000a, 201f. See, for example, Greco 2000a, who claims that his “agent realism is consistent with contextualism” (252). There is a wide variety of contextualist accounts of knowledge. Contextualists disagree over the question of whether it is the context of the speaker or the context of the epistemic subject that determines the truth-conditions of knowledge-ascriptions. They also disagree over the question of what factors are responsible for raising or lowering the standards of knowledge-ascriptions and over the exact nature of the mechanisms that induce context changes. For an overview of different contextualist approaches to knowledge, see, for example, Brendel/Jäger 2005. In this

336

Elke Brendel

to which intellectual virtues have to be exercised in order to gain knowledge is also a context-sensitive matter. In a normal situation of a true perceptual belief, such as S’s belief that there is a lake in the distance, the proper functioning of S’s reliable visual faculties seems to be a sufficient condition for S having knowledge that there is a lake in the distance. If S is walking around in a hot desert for many hours and is almost fainting because of severe dehydration, the possibility that she perceives a Fata Morgana becomes a salient errorpossibility in this context. Just trusting her visual faculties seems to be insufficient for claiming that she knows that there is a lake there, even if her belief happens to be true. An adequate account of knowledge along the lines of virtue epistemology thus has to be an agent-based account in which knowledge is analyzed as true belief that is the result of the proper exercise of certain intellectual virtues. Among these virtues are reliable cognitive faculties as well as different kinds of character traits. These intellectual virtues are dispositions of the epistemic subject that manifest themselves in a sometimes complex interaction with other cognitive dispositions in order to achieve knowledge. Whether and to what extent these intellectual virtues have to be consciously controlled by the epistemic subject in order to gain knowledge, and what kinds of epistemic virtues have to be exercised is context-dependent. But it is still questionable whether such a definition in the form of necessary and sufficient conditions for knowledge is feasible. Even if a definition of knowledge in virtue epistemology succeeds in solving or circumventing some of the most notorious problems of knowledge, such as the Gettier problem, the lottery paradox and the sceptical challenge, there are further questions that remain to be answered: First, as we have seen, for some virtue epistemologists, like Zagzebski and Fairweather, intellectual motivation plays a significant role in the formation of a true belief that properly aims at the truth. According to Zagzebski, an act that does not arise out of the right motive, i.e. an act that is not properly motivated by a love or passion for truth, deprives the epistemic subject of knowledge.41 It seems to me that such a demanding requirement can lead to a counter-intuitive and intellectually over-loaded concept of knowledge. Even if it is true that sensory stimuli can be overridden by background information controlled by a “desire for truth,” most perceptual beliefs are spontaneously formed without an explicit aid or control of an epistemic motivation. That is why it normally seems to be odd to say that a person should get credit or should be praised for getting the truth via such a perceptual belief. Only if a

_____________ 41

paper I do not want to embrace a particular contextualist account of knowledge, since the definitions of knowledge in virtue epistemology are consistent with different contextualist theories. See Zagzebski 2003, 151f.

The Epistemic Function of Virtuous Dispositions

337

person really has to make an actual intellectual effort in her belief forming process does it seem to be appropriate to praise her for getting the truth. Zagzebski points out that “what is noble about the desire for truth” is that “we often have to give something up.”42 So, in particular, we praise a person for ridding herself of old conventional beliefs that are no longer tenable in the light of convincing new evidence. But in the case of a normal perceptual belief, the “monitoring system” and the “desire for truth” are indeed completely “silent partners” in the belief forming process, so that we can hardly claim that believing truly is really an intellectual act motivated by the valuing of truth. Thus, an account of virtue epistemology in which praiseworthiness, credit, and intellectual motivation are regarded as significant defining elements of knowledge faces serious problems, in particular in relation to perceptual knowledge. It seems that we either have to weaken the concept of praiseworthiness and credit so that we can “praise” a person if she comes to believe the truth as a result of reliable sense perception - the rationale for that could be that she at least bases her belief on her reliable sense perception and does not come to believe something false etc. – or virtue epistemologists have to admit that praiseworthiness and credit might be crucial concepts for other epistemological projects but are not suitable as necessary conditions for all kinds of knowledge. Second, many intellectual virtues are not strictly truth-conducive. It therefore seems to be premature to define intellectual virtues in terms of their truth-conduciveness, i.e. as cognitive dispositions the proper exercise of which results in the formation of true beliefs. As we have already seen, most intellectual virtues in isolation are not truth-conducive. Fair-mindedness alone without the proper aid of other cognitive faculties and skills does not lead to knowledge. Pathological forms of intellectual virtue or the wrong combination of intellectual virtues with other faculties or character traits can be intellectual blameworthy or can be an obstacle in the search for truth, as in the case of a biased but persevering researcher. Furthermore, intellectual virtues normally contribute to knowledge only in a very indirect way. As we have seen, intellectual humility, for example, only fosters knowledge when combined with other intellectual virtues, since in the long run an intellectually humble person is “epistemically better off.” The history of science also shows that even the most intellectually virtuous scientists were wrong about very many things, but were nevertheless of utmost importance for the development of the scientific progress. Montmarquet therefore contends that defining intellectual virtues in terms of their truth-conduciveness does not adequately give expression to the fact that we

_____________ 42

Zagzebski 2003, 153.

338

Elke Brendel

very often ascribe intellectual virtues to a person even if her intellectual inquiries do not immediately aim at a correct theory: [I]f we are to appraise the relative worth or ‘virtue’ of epistemic agents by the truthconduciveness of their intellectual dispositions, then how are we to accommodate the approximate equality of epistemic virtue we find in such diverse agents as Aristotle, Ptolemy, Albertus Magnus, Galileo, Newton, and Einstein? From our current vantage point, we recognize these thinkers as differing greatly in the truth of their respective beliefs and systems of belief.43

In order to explain intellectual virtues in terms of their truth-conduciveness in a definition of knowledge, it is therefore necessary to account for the fact that many intellectual virtues only contribute to truth in a very indirect way and that many acts out of intellectual virtues can even immediately result in false beliefs. Zagzebski admits that “there is more than one sense in which a virtue can be truth conducive.”44 She therefore widens the concept of truthconduciveness in the following way: I suggest that we may legitimately call a trait or procedure truth conducive if it is a necessary condition for advancing knowledge in some area even though it generates very few beliefs and even if a high percentage of the beliefs formed as the result of this trait or procedure are false.

Furthermore, the mere truth-conduciveness of some intellectual dispositions does not necessarily qualify them as virtues. If I manifest my dispositions of intellectual curiosity, desire for truth and perseverance by carefully counting all chairs in the building of the philosophy department in Mainz and come up with a true belief about the number of these chairs as a result of my “inquiry”, I would hardly be described as an “intellectually virtuous” person. We normally do not evaluate a person as intellectually virtuous just on the basis of how she aims at the truth of one single, isolated belief. Hookway stresses the point that: “Virtues regulate inquiries and deliberations and only indirectly regulate beliefs.”45 Whether a person can be regarded as intellectually virtuous also depends on the procedure of how she succeeds in finding out new and interesting things that are explanatorily relevant for some inquiries and on how she contributes to the understanding of some of our research projects. The last considerations suggest that the project of defining knowledge in terms of intellectual virtue needs at least some fundamental refinements of some of its central concepts. First, the relationship between credit/praiseworthiness and knowledge has to be explained in a way that accounts for the fact that spontaneously acquired true perceptual beliefs can count as knowledge. Otherwise it seems that knowledge in virtue epistemology is an intellectually too demanding epistemic concept. Second, the truth-conduciveness of

_____________ 43 44 45

Montmarquet 1993, 21. Zagzebski 1996, 181. Hookway 2003, 197.

The Epistemic Function of Virtuous Dispositions

339

intellectual virtues has to be explained in a manner that takes into account the fact that many intellectual virtues contribute to the success of our epistemic aims only in a very indirect way. Furthermore, the discussion of intellectual virtues has shown that our epistemic aims are not restricted to the mere collection of true beliefs and the avoidance of false beliefs. Many intellectual virtues are directed at understanding relevant and important subjects rather than at mere truth. They derive their epistemic value not primarily from their contribution to epistemically justified beliefs and knowledge. They rather contribute to some other of our epistemic goods, such as understanding and wisdom.46 Therefore, in an adequate account of epistemic virtues a shift in focus from propositional knowledge and epistemic justification to wider epistemic concepts, such as understanding and wisdom, is required.47

Literature Alcoff, Linda M., ed. 1998. Epistemology: The Big Questions. Oxford: Blackwell Publishers. Bird, Alexander. 1998. “Dispositions and Antidotes.” The Philosophical Quarterly 48: 227-234. Brendel, E., and Ch. Jäger, eds. 2005. Contextualisms in Epistemology. Dordrecht: Springer. Brendel, E., and Ch. Jäger. 2005. “Contextualist Approaches to Epistemology: Problems and Prospects.” In Contextualisms in Epistemology, ed. E. Brendel, and Ch. Jäger, 1-30. Dordrecht: Springer. Butler, Douglas. 1988. “Character Traits in Explanation.” Philosophy and Phenomenological Research 49: 215-238. Code, Lorraine. 1987. Epistemic Responsibility. Hanover: University Press of New England and Brown University. DePaul, M., and L. Zagzebski, eds. 2003. Intellectual Virtue. Oxford: Oxford University Press. Fairweather, Abrol. 2001. “Epistemic Motivation.” In Virtue Epistemology. Essays on Epistemic Virtue and Responsibility, ed. A. Fairweather, and L. Zagzebski, 63-81. Oxford: Oxford University Press. Fairweather, A., and L. Zagzebski, eds. 2001. Virtue Epistemology. Essays on Epistemic Virtue and Responsibility. Oxford: Oxford University Press. Goldman, Alvin. 2001. “The Unity of the Epistemic Virtues.” In Virtue Epistemology. Essays on Epistemic Virtue and Responsibility, ed. A. Fairweather, and L. Zagzebski, 30-48. Oxford: Oxford University Press. Goodman, Nelson. 1954. Fact, Fiction, and Forecast. Cambridge/Mass.: Harvard University Press. Greco, John. 2000a. Putting Skeptics in Their Place. Cambridge: Cambridge University Press. Greco, John. 2000b. “Two Kinds of Intellectual Virtue.” Philosophy and Phenomenological Research 60: 179-184. Greco, John. 2002. “Virtues in Epistemology.” In The Oxford Handbook of Epistemology, ed. P. Moser, 287-315. Oxford: Oxford University Press. Greco, John. 2003. “Knowledge as Credit for True Belief.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 111-134. Oxford: Oxford University Press.

_____________ 46 47

See, for example Riggs 2003. I am very grateful to Quassim Cassam for helpful comments on an earlier version of this paper.

340

Elke Brendel

Greco, John, 2005. “A Different Sort of Contextualism.” In Contextualisms in Epistemology, ed. E. Brendel, and Ch. Jäger, 241-258, Dordrecht: Springer. Heil, John. 2005. “Dispositions.” Synthese 144: 343-356. Hookway, Christopher. 2003. “How to be a Virtue Epistemologist.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 183-202. Oxford: Oxford University Press. McKitrick, Jennifer. 2003. “A Case for Extrinsic Dispositions.” Australasian Journal of Philosophy 81: 155-174. Montmarquet, James. 1993. Epistemic Virtue and Doxastic Responsibility. Lanham: Rowman and Littlefield. Johnston, Mark. 1992. “How to Speak of the Colors.” Philosophical Studies 68: 221-263. Moser, Paul, ed. 2002. The Oxford Handbook of Epistemology. Oxford: Oxford University Press. Mumford, Stephen. 1998. Dispositions. Oxford: Oxford University Press. Riggs, Wayne D. 2003. “Understanding ‘Virtue’ and the Virtue of Understanding.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 203-226. Oxford: Oxford University Press. Roberts, R. C., and W. J. Wood. 2003. “Humility and Epistemic Goods.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 257-279. Oxford: Oxford University Press. Sherman, N., and H. White. 2003. “Intellectual Virtue: Emotions, Luck, and the Ancients.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 34-53. Oxford: Oxford University Press. Shoemaker, Sydney. 1984. Identity, Cause, and Mind. Cambridge: Cambridge University Press. Sosa, Ernest. 1980. “The Raft and the Pyramid: Coherence versus Foundations in the Theory of Knowledge.” Midwest Studies in Philosophy 5: 3-25 – quotes taken from the reprint in Sosa, Ernest. 1991. Sosa, Ernest. 1991. Knowledge in Perspective. Cambridge: Cambridge University Press. Zagzebski, Linda. 1996. Virtues of the Mind. Cambridge: Cambridge University Press. Zagzebski, Linda. 2003. “Intellectual Motivation and the Good of Truth.” In Intellectual Virtue, ed. M. DePaul, and L. Zagzebski, 135-154. Oxford: Oxford University Press.

List of Contributors Andrea Borghini is Assistant Professor in the Philosophy Department at the College of the Holy Cross. He specializes in Metaphysics, Philosophy of Science (especially Philosophy of Biology), and Early Modern Philosophy. He has authored an introductory volume to contemporary theories of possibility (Carocci 2008, in Italian) and a variety of articles, among them “A Dispositional Theory of Possibility” (with Neil E. Williams) in Dialectica (2008).

Elke Brendel is Professor of Philosophy at the University of Bonn. Her research interests are in logic, epistemology, and the philosophy of Language. She is the author of Wahrheit und Wissen, Paderborn 1999; Contextualisms in Epistemology, Dordrecht 2005 (ed. with Chr. Jäger), Zitat und Bedeutung, Hamburg 2007 (ed. with J. Meibauer and M. Steinbach), Grundthemen Philosophie: Wissen, Berlin-New York 2010; Understanding Quotation. Linguistic and Philosophical Analyses, Berlin-New York 2010 (ed. with M. Meibauer and M. Steinbach).

Gregor Damschen holds a research fellowship of the Swiss National Science Foundation (SNF) at the University of Lucerne, and is a lecturer in philosophy at the Universities of Lucerne and Halle. His research focuses on topics in epistemology (knowing how and knowing that, certainty, and ultimate justification), bioethics, and ancient philosophy. He has published widely in these areas and is the co-author of a forthcoming book about methods in philosophy: Selbst Philosophieren. Argumentieren und Interpretieren (Reclam 2010).

Francisco J. Gonzalez is Associate Professor of Philosophy at the University of Ottawa. He is the author of Dialectic and Dialogue: Plato’s Practice of Philosophical Inquiry (Northwestern 1998) and Plato and Heidegger: A Question of Dialogue (Penn State Press 2009).

David Henderson is the Robert R. Chambers Professor of Philosophy at the University of Nebraska, Lincoln. His work has focused on both the philosophy of social science and on epistemology, where he has sought to garner from contemporary cognitive science lessons regarding venerable epistemological positions. In addition to his “Epistemic Competence and Contextualist

342

List of Contributors

Epistemology,” Journal of Philosophy (1994), work with Terry Horgan, notably, “Iceberg Epistemology,” Philosophy and Phenomenological Research (2000), provides the basis for their contribution to the present volume.

Terence Horgan is Professor of Philosophy at the University of Arizona. He very much enjoys collaborative work and has published very widely and extensively in metaphysics, philosophy of mind, philosophy of psychology, philosophy of language, meta-ethics and epistemology. He is co-author (with John Tienson) of Connectionism and the Philosophy of Mind (MIT Press 1996) and (with Matjaz Potrc) of Austere Realism (MIT Press 2008).

Andreas Hüttemann is Professor of Philosophy at Westfälische WilhelmsUniversität Münster. His main areas of research are philosophy of science and early modern philosophy. He is the author of What’s Wrong With Microphysicalism (Routledge 2005), and has edited Kausalität und Naturgesetz in der frühen Neuzeit (Wiesbaden 2001) and, together with Gerhard Ernst, Time, Chance and Reduction (Cambridge 2010).

Ludger Jansen teaches Philosophy at the University of Rostock and is executive director of the Centre for Logic, Philosophy and History of Science. He is the author of a book on Aristotle’s theory of dispositions (Tun und Können, Frankfurt 2002) and co-editor of an introduction to biomedical ontology (Biomedizinische Ontologie, Zürich 2008).

Michael-Thomas Liske is Professor of Philosophy at the University of Passau. He is interested in a systematic interpretation of philosophical classics (Aristotle, Scholastics, Leibniz) from the viewpoint of analytic philosophy. Liske is the author of Leibniz’ Freiheitslehre (Hamburg: Meiner 1993) and Gottfried Wilhelm Leibniz (München: Beck 2000).

Peter Machamer is Professor of History and Philosophy of Science and Associate Director of the Center for Philosophy of Science at the University of Pittsburgh. He and J.E. McGuire have just published Descartes’s Changing Mind with Princeton University Press.

List of Contributors

343

Jennifer McKitrick is Associate Professor at the University of NebraskaLincoln. Her work focuses on issues in the metaphysics of science. She has edited two anthologies – Dispositions and Laws of Nature (Springer 2005) and Establishing Medical Reality: Essays in the Metaphysics and Epistemology of Biomedical Science (Springer, forthcoming) – and is the author of several papers on dispositions, including “A Case for Extrinsic Dispositions” (Australasian Journal of Philosophy 2003).

Burkhard Meißner is Professor of Ancient History at the Helmut Schmidt Universität (University of the Armed Forces of the Federal Republic of Germany) in Hamburg. His research is focused on Ancient Greek history. He is particularly interested in topics related to technology, warfare, and to the conceptions of history among Ancient Greek historians. He is the author of three books: Historiker zwischen Polis und Königshof (Göttingen 1992); Die technologische Fachliteratur der Antike (Berlin 1999) and Hellenismus (Darmstadt 2007).

Stephen Mumford is Professor of Metaphysics at the University of Nottingham. He is the author of Dispositions (Oxford 1998), Laws in Nature (Routledge 2004) and David Armstrong (Acumen 2007) and editor of Russell on Metaphysics (Routledge 2003) and Powers by the late George Molnar (Oxford 2003). He is currently working on a book on causation, Getting Causes from Powers, to be coauthored with Rani Lill Anjum.

Ursula Renz is Professor of Philosophy at the University of Klagenfurt. She is the author of Die Rationalität der Kultur: Kulturphilosophie und ihre transzendentale Begründung bei Hermann Cohen, Paul Natorp und Ernst Cassirer (2002), Die Erklärbarkeit von Erfahrung: Realismus und Subjektivität in Spinozas Theorie des menschlichen Geistes (2009) and has written several articles on German Philosophy and Early Modern Philosophy. Recently, together with Hilge Landweer as coeditor, she has published Klassische Emotionstheorien (2008).

Robert Schnepf is a Senior Lecturer (Privatdozent) at Martin-LutherUniversität Halle-Wittenberg. He is interested in Early Modern Philosophy, Epistemology, Metaphyscis, and the Philosophy of History. He has published widely in these areas. Schnepf is also the author of three books – Die Metaphysik im ersten Teil der Ethik Spinozas (Würzburg 1996); Die Frage nach der Ursache – systematische und problemgeschichtliche Untersuchungen zum Kausalitäts- und zum Schöp-

344

List of Contributors

fungsbegriff (Göttingen 2006); and Geschichte erklären (Göttingen 2010) – and the editor of a variety of anthologies.

Oliver R. Scholz is Professor of Philosophy at the Westfälische WilhelmsUniversität Münster, Germany. His research interests include metaphysics, epistemology and philosophy of language. He is the author of Bild, Darstellung, Zeichen (1991, 3rd edition 2009) and Verstehen und Rationalität (1999, 2nd edition 2001).

Markus Schrenk held a junior research fellowship at Worcester College, Oxford, and a postdoctoral research fellowship within the AHRC funded Metaphysics of Science Project that was jointly undertaken by the Universities of Birmingham, Bristol, and Nottingham. He specializes in laws of nature, causation, dispositions, and necessity. Markus Schrenk is author of the book The Metaphysics of Ceteris Paribus Laws.

Karsten Stueber is Professor of Philosophy at the College of the Holy Cross. He has published widely in the areas of philosophy of language, philosophy of mind/psychology, and the philosophy of the social sciences. He is the author of Donald Davidsons Theorie sprachlichen Verstehens (Anton Hain 1993) and of Rediscovering Empathy: Agency, Folk Psychology, and the Human Sciences (MIT Press 2006). In addition, he is the co-editor of Philosophie der Skepsis (utb 1996) and Empathy and Agency (Westview Press 2000).