Gregor Damschen. (2009) - Debating Dispositions, Issues On Metaphysics, Epistemology and Philosophy of Mind

Debating Dispositions
Debating Dispositions

Issues in Metaphysics, Epistemology
and Philosophy of Mind

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

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, with-
out permission in writing from the publisher.
Printed in Germany
Cover design: Martin Zech, Bremen
Printing and binding: Hubert & Co GmbH & Co KG, Gttingen
V
Acknowledgements

The anthology grew out of the spirited discussion and philosophical collegial-
ity 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 Bhringer-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 ener-
getically helped with the organization of the Conference in Wittenberg and
who, together with Jens Gillessen, formatted the texts according to the pub-
lishers 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 Grnkorn
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/Stber

I. The Discovery of Dispositions: Ancient Foundations

Knowledge and Virtue as Dispositions in Platos Theatetus..................... 3
Francisco J. Gonzalez
Aristotles Theory of Dispositions: From the Principle of Movement
to the Unmoved Mover............................................................................... 24
Ludger Jansen
Dispositions in Greek Historiography...................................................... 47
Burkhard Meiner

II. The Debate about Dispositions from the Beginning of Modern Sci-
ences to the 20
th
century

The Dispositions of Descartes................................................................... 71
Peter Machamer
Explicable Explainers: The Problem of Mental Dispositions in
Spinozas Ethics ............................................................................................. 79
Ursula Renz
Harmonizing Modern Physics with Aristotelian Metaphysics: Leibnizs
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
Table of Contents
VIII
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 Httemann

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 Aristotles 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 17
th
and 18
th
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 20
th
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
Introduction
X
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 20
th
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 triangles corners.
A number of articles in this anthology address and document the 20
th

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
Introduction
XII
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 Platos
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 Aristotles Theory of Dispositions:
From the Principle of Movement to the Unmoved Mover. He situates
Aristotles theory within its linguistic and philosophical contexts and
delineates its wide-ranging conceptual framework for analyzing the nature of
dispositions. The precise nature of Aristotles 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 MEINERS 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 20
th
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 Gods 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 Spinozas Ethics a thorough analysis of the role of
dispositions in Spinozas philosophy of mind. She starts with the observation
that on an ontological level Spinozas 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, Spinozas
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, Spinozas 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: Leibnizs Theory of Force by MICHAEL-THOMAS LISKE
discusses Leibnizs philosophical reasons for an excessive use of dispositional
terminology in an intellectual climate determined by modern science. Liskes
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. SCHOLZS 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
Introduction
XIV
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 20
th

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 Ryles 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 Ryles 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
SCHRENKS 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 MUMFORDS essay Ascribing Disposition continues the
argument for dispositional realism. Yet in contrast to Schrenk, Mumfords
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 BORGHINIS 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 HTTEMANN
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 Httemann 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
Introduction
XVI
explanations such as the explanation of Caesars behavior during the Roman
Civil War in terms of a so called Clementia Caesaris or Max Webers 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 Webers 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 historians
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 DAMSCHENS 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 Platos 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 ques-
tions of ontology, the work in which is to be found the most explicit reflec-
tion 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 duna-
mis(to vto e rotiv oux oiio ti aiqv ouvoi, 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 sug-
gestion 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
things 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 identi-
fies what is with what is eternal and unchanging. The problem with this on-
tology 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
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 ma-
terial, motion and change as well as permanence and stability.
What is worth emphasizing for the purposes of the present paper, how-
ever, 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 identi-
fication 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 ques-
tion, What is knowledge?, a question that proves inseparable from the ques-
tion 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 seri-
ous 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 particu-
lar 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
(oprt\) and wisdom (ooo, 145b1-2). The connection is made by Socrates
suggestion that raiot\q and ooo are the same thing (145e6). Therefore in
what follows we must never lose sight of the fact that the knowledge under
discussion is understood as the arguably more practical disposition of being-
wise 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 in-
cludes not only geometry but also the technical skills of the craftsmen such as
shoemaking, Socrates explains what he wants instead by giving the provoca-
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
(ouvori). The justification is that while these lines are incommensurable (e
\xri rv ou ourtpou) with the lines which when squared produce an even
number, they have the power of producing plane figures that are themselves
commensurable (toi o raiarooi o ouvovtoi, 148b1-2). But what does this
example suggest regarding the attempt to define knowledge? Because an in-
commensurable line cannot be defined in itself, i.e., its length cannot be
stated, it can be defined, and made commensurable, only by being character-
ized 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 under-
stood 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 prob-
lem 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 rele-
vant for our purpose. The first is that Socrates himself cannot give birth to
knowledge; he is barren of wisdom (oyovo rii ooo, 150c4); he cannot
himself reveal anything from himself on account of not possessing anything
wise (oio to qorv rriv ooov, 150c6). Therefore, Socrates cannot teach his
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
(150e1-2) become ignorant fools (ooUri, 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,
but only in their association (ouvouoo, 150d2, d4, 151a3-4). This is why the
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 mid-
wife, 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
(ryiotov, 150b9) is something that plays no role whatsoever in the midwifes
art: i.e., the ability to determine through every kind of test whether the off-
spring of his interlocutors 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, knowl-
edge 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 elabo-
ration of this idea,
1
as we see in Socrates lengthy explanation of the thesis
that knowledge is perception in terms of Protagorean relativism and Hera-
clitean flux, this carefully delivered child of the interlocutors mind can still
prove a mere phantom. This is why Socrates sees the testing of the interlocu-
tors 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 (to xpvriv to oiqUr tr xo \, 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
(xpit\piov), i.e., that through which we judge as through an organ (to oiou
_____________
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.
7
xpvorv e opyovou, 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 ou ti aovu ooo
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 wis-
dom 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 reflexiv-
ity in how this first definition is introduced, a reflexivity that will be seen to
characterize the other definitions as well. Theaetetus suggests defining knowl-
edge as perception because that is how it now appears (e yr vuv ovrtoi,
151e2). Since Socrates soon afterwards identifies appears (ovrtoi) with is
perceived (oioUovroUo, 152b11), Theaetetus is in effect saying that he now
perceives knowledge to be perception: the way in which he attempts to know
_____________
2 The word xpit\piov is used by Socrates in the Theaetetus when he asks if man has within himself
the xpit\piov of future events as he is said by Protagoras to have within himself the xpit\piov of
what is white, heavy or light (178b6, c1). Here Socrates is clearly identifying having the xpit\piov
within oneself and being the measure (rtpov).
3 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 giv-
ing birth, being oyovo only in relation to those whose innate opinions he tries to deliver (54, 9-
13; 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 de-
fending the anonymous commentators interpretation of 150d1, rightly sees the wisdom Socrates
possesses as consisting in the insights that enable him to practice midwifery itself (2004, 31-
37). 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 concep-
tions of both knowledge and midwifery suggested by the Theaetetus cannot be reconciled with
such an interpretation.
8
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 un-
dermines 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 percep-
tion 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 con-
versation (ouaooo \ tou oioiryroUoi apoyotro, 161e6-7), for to set about over-
hauling and testing one anothers 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 unapproach-
able 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 op-
posed 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 ooUqoi in Theaetetus first defini-
tion of knowledge is to be understood not as the sense organs (oioUqt\piov) but rather as any
kind of apprehension (ovtiqi) (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 commentators distinction is clarified later when, in commenting on the
identification of appears and is perceived at 152b-c and claiming that this is Platos addition
to what Protagoras said, he writes that they called every apprehension (ovtiqi), whether
through the sense organ (oioUqt\piov) or through something else, perception (ooUqoi) (66,
39-43). See also Polansky 1992, 67-8 and 103-4.
9
Theodorus, who repeatedly refuses to participate in the discussion with the
excuse that he is not dispositionally inclined to Socrates kind of dialectic
(o\Uq t[ toioutq oioirxtou, 146b3; see also 162a-b, 165a). It is no accident
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
(iio ioyoi, 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 dialec-
tical examination.
When Socrates finally succeeds in dragging Theodorus into the discus-
sion, 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 sug-
gested 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 (oo aoioro rrtoi, 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 (rariUorU ov, 144e5), Socrates recommends that he and
Theaetetus take Theodorus judgment seriously not by believing it, but by ea-
gerly examining (apoUurioUoi ovooxrooUoi, 145b2-3) and being examined
(raiorixvuvoi, 145b4). Thus begins the battle between a conception of knowl-
edge as authoritative expertise and a conception of knowledge as inescapably
dialectical.
6
_____________
5 See Polansky 1992, 72; also 111, 117-8
6 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.
Burnyeats 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
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 discus-
sion 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 (rtpov).
But who is Socrates to assert that this is necessary? Socrates himself immedi-
ately complains that his earlier defense of Protagoras forced him into the
position of being a measure even though he lacks knowledge (ro or t
ovraiot\ovi, 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 presuma-
bly 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. Levitts translation reads: Then we shall be giving your
master fair measure [rtpe] 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 (rtpe \iv, 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 Platos 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 oruor, which Levitt translates as unerring. But is it correct to identify being oruor
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 ex-
pertise 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 cer-
tainty, but rather what is exhibited in the dialogue: dialectic.
11
forces him, very much against his inclination, to join the discussion. Thus
Socrates at one point describes his midwifes art as a knowledge of nothing
more than how to receive in measure (rtpe, 161b2-5) the logoi provided
by his interlocutor. While Protagoras might claim that everyone is the meas-
ure 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 impos-
sible without an apprehension of being and truth and that the soul appre-
hends 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 (186c3-
5). Indeed, what we have here is a characterization of knowing that corre-
sponds 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 (to xoivo,
185e) is at one point characterized as a striving (raopryroUoi, 186a4). This
striving after being, as the characteristic of knowing that proves incompati-
ble 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 disposi-
tion 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
5. Believing Falsely that Knowledge is True Belief
Theaetetus now suggests that knowledge might be (xivouvruri) true be-
lief, 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 be-
tween true and false belief. This is indeed why Socrates proceeds to show that
the identification of knowledge with true belief cannot account for the possi-
bility 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 con-
sidered and tested.
In this context, Socrates identifies thinking with a dialogue, a give-and-
take 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 souls 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 think-
ing/striving knowledge? A negative answer is suggested not only by the even-
tual 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 (e yr q rioe, 189e7). Socrates thus
makes clear that the outcome of his souls 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 souls dialogue with itself, what concludes the dialectical back-
and-forth of question and answer (190a2-4). Socrates describes the soul as
having a doxa, as believing or judging (oooriv), only when it is no longer
wavering or hesitating (q oiotop) but says the same (to outo p), only when
it arrives at something definite (opoooo), either slowly or quickly rushing
upon it (raooo). 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 Cf. Dixsaut 1997, 10.
9 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).
13
to waver: they never come to a judgment or belief (even their apparent judg-
ments 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 Socra-
tes 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 pos-
sess. 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 (q xoUope oioiryroUoi,
196e1-2) by the fact that he and Theaetetus have been constantly using differ-
ent forms of the word know and its opposite despite their claim not yet to
know what knowledge is. Theaetetus appropriately responds that no discus-
sion 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 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.
11 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
for oneself. Such a conception seems quite different from the one presup-
posed 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 (ovrioio, 151e6, 157d-
3, 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 minds 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 refer-
ence 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 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 sig-
nificant: is knowledge like the bird we seek to possess or like the search itself?
13 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 (apoUuqUe, 193c2) to rec-
ognize the person clearly plays a central role.
14 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 com-
pletes 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 knowl-
edge that enables Socrates and Theaetetus to mis-take what knowledge is. In asking this question
they are already going after (r o rporUo, 187c1) what is in question. Furthermore, the ear-
lier identification of thinking with both the souls reaching out or striving towards being and
the souls 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 (raopryrtoi) is substituting for the word considers (raioxoariv) in the
immediately preceding descriptions of the souls 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 epis-
temic route with a capacity for intermediate analysis of eleven. He [the dunce] has learned, per-
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, under-
stood as the exchange of question and answer, to an end in an infallible be-
lief? 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 knowl-
edge. 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 (ioovti ovov rotiv riorvoi,
contrasted with r oxo[ xpvovtr opUo arioUrvtr, 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 im-
possible the kind of discussion that could alone teach them the truth of the
crime, rather than merely persuade them (201b3). Burnyeat has drawn atten-
tion to the seeming contradiction between these two explanations of the
jurys lack of knowledge (124). But it is not hard to see that these two expla-
nations 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
(outo aop outev aoiio xo xoio rupovtr tr xo trxovtr, 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
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 re-
solves the tension in the jury example by reconciling logos with seeing some-
thing for onself, the latter is immediately suppressed in the discussion not
only by Theaetetus identification of knowledge with an account an identifi-
cation 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 (riaovto tou oxouoo, 201c8); as for explaining
the definition, he claims to be able at best to follow what someone else says
(iryovto rtrpou, 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 of-
fering Theaetetus a dream in exchange for a dream (vop ovt ovrpoto,
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 eyewit-
nessing 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 ensu-
ing 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 know-
ledge?
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.
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.,
are oioyo (202b6, 205e2-3). Analysis presupposes a relation to the object that
cannot itself be identified with analysis. What then is the nature of this rela-
tion? 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 defini-
tion, become knowledge? Socrates proceeds to show that if logos converts judg-
ment into knowledge, this cannot be logos understood as the mere expression of
judgment, since in that case anyone capable of speech could transform a be-
lief into knowledge by merely expressing it; we would thus be back to Prota-
gorean relativism (206d-e). Nor can logos be understood as the enumeration of
a things 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 some-
thing 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-
quire: if one does not know the being of a wagon (ouoo, 207c3) by simply
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 knowl-
edge 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
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 ap-
pears to be a certain kind of know-how.
16
Socrates midwifery, I would sug-
gest, 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 ques-
tion 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 inade-
quate 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 distin-
guishing 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 ex-
plicitly fall into the circularity that has been seen to taint the entire discus-
sion: 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 be-
lief with an account can not only be interpreted as the mere verbal expres-
sion 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 accountor logos).
_____________
16 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 Platos claims about knowledge
can be reexpressed as claims about understanding (216-218). I agree, but only with the qualifica-
tion that understanding here means a know-how or capability, rather than the systematic and
exhaustive explanation of an entire domain.
17 As Sedley rightly observes: It is not hard to work out that this same objection is equally threat-
ening 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 rec-
ognized, circularity sets in. Knowledge will be defined as true judgment plus knowledge of some-
thing (2004, 176).
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 ac-
count can we know this? And if we attempt to demonstrate by giving an ac-
count 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 dis-
tinguish true from false beliefs? If no one logos can justify a belief, then per-
haps 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 (q
ouvorvov oouvo tr xo orooUoi ioyov). But then Socrates immediately pro-
ceeds to describe the contrasting person, not as some who can give and re-
ceive a logos, but rather as someone who gets hold of a logos and thereby be-
comes perfect with regard to knowledge (trire apo raiot\qv rriv, 202c4-
5). 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 mak-
ing 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 things 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 ouvoi of
oouvo tr xo orooUoi ioyov? 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
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 things
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 re-
duced 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 knowl-
edge, 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 Dialogues Outcome:
Knowledge of Knowledge as a Way of Being with Others
Only these reflections can prepare us for Socrates description of the dia-
logues positive outcome. After suggesting that Theaetetuss 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 ([ttov
opu toi ouvouoi xo \rpetrpo, 210c2-3) as well as a certain kind of virtue
and knowledge: namely, exhibiting temperance in not thinking he knows what
he does not know (oepove oux oiorvo riorvoi o q oioUo, 210c3-4). But in
_____________
18 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.
Platos own thoughts about knowledge are no more immune to the difficulty than anyone elses
(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 Platos 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).
19 Cf. Dixsaut: La diffrence entre lopinion droite et le savoir ne peut constituer le contenu dune
opinion droite ou dun savoir, mais le savoir est cette difference (1998, 304)
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 knowl-
edge of knowledge with a certain way of being with others. This outcome of a
knowledge of knowledge exhibited in dialogue with others should not sur-
prise 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 (iioioyo) and his desire to
make us converse and become friendly and talkative with one another (\o
aoi[ooi oioiryroUoi xo iou tr xo apooqyopou oii\iou yyvroUoi, 146a7-
8).
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 An interesting parallel is to be found in Xenophons Memorabilia IV. v. 12. Socrates here is
described as arguing, at the end of a discussion with Euthydemus on oepoouvq and ryxpotrio,
that by sorting things according to kinds (oioiryovto xoto yrvq), choosing the good and reject-
ing the bad, one becomes better and happier and more able to pursue dialectic (opotou tr xo
ruooiovrototou xo oioiryroUoi ouvotototou). Even if Kahn is correct in arguing that this
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 Platos
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 Xenophons stated goal
in proceeding to report another discussion is to show how Socrates oioirxtixetrpou raori
tou ouvovto.
21 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 con-
versation of the Heracliteans. The above comments make clear, I hope, that this is not at all the
case.
22 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). Polanskys interpretation is
similar to my own when he asserts that The entirety of the dialogue depicts just what knowl-
edge is. philosophical understanding is present in deed (ergon) in the dialogue. That deed, the
construction of the dialogue, is Platos account of knowledge (1992, 239); see also 244-5, where
22
in the sense of expertise on knowledge is not to be attained and that the re-
sulting lack of closure itself exhibits knowledge of knowledge in the sense of
the oepoouvq of knowing what one does not know and thus continually
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 (\ tou oioiryroUoi ouvoi, 57e6-7) is the
power of loving the truth (ouvoi rpov tou oiqUou, 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 oth-
ers 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 (ooeoi Ur, 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 knowl-
edge 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
_____________
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 dia-
lectical and dispositional dimension of this knowledge.
23 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).
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. tudes Platoniciennes IV: 272-293.
Kahn, Charles. 1996. Plato and the Socratic Dialogue. Cambridge: Cambridge University Press.
Long, Anthony A. 1998. Platos 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 Platos Theaetetus. Lewis-
burg: Bucknell University Press.
Sedley, David. 2004. The Midwife of Platonism: Text and Subtext in Platos Theaetetus. Oxford:
Clarendon Press.
Aristotles 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 phi-
losophy. In what follows, I will sketch his thoughts about dispositions and
argue that it can still be regarded as a good theory.
1
1. Its 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 theo-
ries of dispositions be it of dispositional predicates or of dispositional prop-
erties dates back to the heyday of logical empiricism. The problem of dispo-
sitions 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 Aristotles 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 prcis of my book on Aristotles 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.
25
will start my discussion of Aristotles account of dispositional causal proper-
ties 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 Aristotles time, this word was in common usage, and it can al-
ready 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: chrmata te kai dynameis) or the army of a kingdom (cf.
Plato, Menexenos 240d: h Persn dynamis, the army of the Persians), and even
the phonetic quality of a letter (cf. Plato, Cratylus 412c: tn 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 phi-
losophical and medical contexts.
5
For example, Alcmaeon of Croton (ca. 570-
500) uses the term to define health (hygieia) as the balance of powerful things
_____________
2 That Aristotles theory of dynamis is a theory of dispositional properties has also be seen (among
others) by Liske 1996. Already Wolf 1979 discusses both Aristotles theory and modern theories
of dispositions, even though she discusses it under the name of possibility (Mglichkeit).
3 The translation is Murrays; I modified it by replacing Murrays 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.
4 All occurrences of dynamis in Plato (and many in earlier authors) are collected and discussed in
Souilh 1919.
5 For a survey of dynamis in the Hippocratic texts cf. Plambck 1964.
Ludger Jansen
26
(isonomia tn 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: path koina ts hyls epigignomena tais metabolais). For Anaximenes,
powers interpenetrate the elements or bodies that are their bearers (DK 13
A 10 = KRS 145: tas endikousas tois stoicheiois tois smasi 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 polachs 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 dy-
namis as an active power. For dynamis used in this way, Aristotle gives the
following definition:
_____________
6 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.
7 Smeets 1952 carves up Metaphysics IX 1-9 in many different passages by different hands, distin-
guishing bits written by Aristotle at different times in his life, his students or even later Aristote-
lians. 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.
27
Dynamis means a source (arch) of movement (kinsis) or change (metabol), 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
Aristotles terminology they function as technical terms that are in need of an
explanation. I will, in turn, explain what Aristotle means by the terms princi-
ple, 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 arch) 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 metabol) 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 Aris-
totle 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 mans 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 Aristotles 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 Aristotles strange phrase can be illuminated with the
Ludger Jansen
28
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 hap-
pens (symbebken), that the same person is a healer and is being healed. Therefore, [be-
ing 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 be-
come 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 Hippo-
crates 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. Inti-
mately connected with it is the concept of a passive disposition. To have a
passive disposition allows its bearer to undergo a change. A passive disposi-
tion is a principle of change in the bearer of the disposition, caused by some-
thing else or by itself as something else. Thus, in order to be realized or mani-
fested a passive disposition requires a corresponding active power, and vice
versa.
Aristotle also talks about qualified dispositions, which are principles to do
29
something well or to act after a decision to do so, as opposed to do some-
thing somehow or by accident. Aristotle illustrates this concept by contrasting
a drunkards 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 bear-
ers 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 princi-
ple 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 match-
ing active power, and having a qualified disposition means to have any dispo-
sition 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. (Meta-
physics IX 6, 1048a35-b9; transl. Ross)
This second family is introduced by a set of examples, and the reader is in-
vited to recognize the similarity between these examples by considering
analogous cases (t analogon synhoran, 1048a 37). Those cases that are as sub-
stances to some sort of matter are said to stand in an analogy to those cases
that are as movement to dynamis : Aristotles claim is that, in a way, a sub-
stance 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.
Ludger Jansen
30
round or a sword. Principles for change are relevant for dynamic causal ex-
planations. 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 swords iron is now, at this very moment of time,
related to the sword, we are in search for a static ontological explanation. And
Aristotles answer is, obviously, that the iron is realizing its dynamis to be
shaped like a sword. Here we see how Aristotles 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 tn 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 tn dynamin tou badizein
dynatai badizein dynameos badizein estin dynaton esti badizein. In the context of
Aristotles 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 poten-
tial walker.
The adjective dynaton can, however, also indicate that something is possi-
ble 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 ex-
plicitly 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 For textual references cf. Jansen 2002, 20-26.
10 Cf. Jansen 2002, 21-24 on the use of dynaton in the context of modal logic and van Rijen 1989 on
Aristotles overall theory of possibility.
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 onto-
logical 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 dy-
namis ascriptions are about the actual world. They ascribe actual properties to
actual things. By no means do they constitute a ghost world of mere possi-
bilia.
14
We can sum up Aristotles 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 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 Heideg-
ger. Cf. also Jacobi 1997.
12 Cf. Clark 1970. For more references cf. Jansen 2002, 28-34.
13 Cf. Weidemann 1984, Hughes/Cresswell 1996.
14 Against Hartmann 1938, 5 (Gespensterdasein).
Ludger Jansen
32
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 cate-
gorical 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 Aristotles teach-
ings about dynamis that were probably most influential in the twentieth cen-
tury: 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 Hartmanns
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 pos-
sible, 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 Hartmanns theory: Contrary to
intuition, one can no longer extensionally differentiate between possibility and
necessity: All and only totally possible states of affairs are necessary. Hart-
mann accepts this consequence, which is a rather unfortunate result in my
eyes. But more important for his interpretation of Aristotle is Hartmanns
concept of partial possibility: A state of affairs s is partially possible if and
_____________
15 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 Hart-
mann 1938: Auch bei Aristoteles kommt keine Mglichkeit vor ohne eine Wirklichkeit, die sie
trgt, nur ist diese nicht wie die Megariker wollen schon die Wirklichkeit dessen, dessen
Mglichkeit sie erst ist.
16 Cf. Hartmann 1938. On Hartmanns modal ontology cf. Hntelmann 2000, on his interpretation
of Aristotle cf. Seel 1982 and Liske 1995.
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 Hartmanns interpre-
tation 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 Aristotles account of dynamis if we treat it as being on a par with the ob-
taining of just any necessary condition.
While Hartmann interprets Aristotle in terms of his concept of partial
possibility, Jaakko Hintikkas interpretation draws on the so called Principle of
Plenitude. In Hintikkas wording, the Principle of Plenitude says that [n]o un-
qualified 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 neces-
sary 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 Cf. Hartmann 1937. Hartmanns interpretation of Aristotle is influenced by the different
position of Zeller 1882.
18 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.
19 Hintikka 1973, 96.
20 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
oclock it is twelve oclock that may be said to be true, but are maybe not true at any point of
time but rather in some timeless manner.
Ludger Jansen
34
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, Hin-
tikka 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 falla-
cious) 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 as-
sumption 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., dispo-
sitions that do not have all of eternity at their disposal for realizing them-
selves. 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 Aristotles theory of
dynamis.
As different as Hartmanns and Hintikkas 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 Aristotles theory of
dispositions, has to interpret dynaton as a predicate modifier. Such an interpre-
tation 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.
35
9. The Megarian Challenge
Aristotle himself had to defend his theory of dispositions against an alterna-
tive position put forward by a group of philosophers called Megarians, who
argue for a position very similar to Hartmanns 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 act-
ing, 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 differ-
ence 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 Par-
menides rejoices), because if there is no principle of change to become some-
thing 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 opin-
ions 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 two-
sided: It is possible to have a disposition and not to realize it at the same
time.
_____________
22 On the attempts to identify these philosophers cf. Jansen 2002, 139-143.
23 For a formal account of this last argument cf. Jansen 2002, 146-149.
Ludger Jansen
36
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 con-
tain any reference to a for-time. And it would indeed be ontologically ques-
tionable to say that someone who can stand at day and at night has two dis-
tinct 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 pa-
rameter without falling back into the Megarian mess. The idea is the follow-
ing: 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 be-
tween. Hence if something has at t a disposition to do or to be F, this disposi-
tion 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 ref-
erence to the for-time has been in our previous formulations into a bound
variable (bound by the existential quantifier some):
25
We started with as-
cribing 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-long-
as-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 ap-
pearance of disposition ascriptions in natural language.
10. Dispositions, Realizations, and Their Conditions
In the Megarian picture, there was an intimate interconnection between hav-
ing 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 Waterlow 1982, 40.
25 For a formalization of this ideas cf. Jansen 2002, 152-154 and 194-196.
37
manifesting a disposition is both necessary and sufficient for having the re-
spective dynamis. Aristotle struggled hard to argue against the Megarian posi-
tion, 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 suf-
ficient 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 neces-
sary 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 On this principle cf. Weidemann 1999.
27 The proof is to be found in Euclid, Elements X 117; it is alluded to in Analytica Priora I 23, 41a 26-
7 and I 44, 50a 35-38. For the details of the argument cf. e.g. Jansen 2002, 159-162.
Ludger Jansen
38
this reason, Aristotle does not need to include a if nothing external inter-
feres 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 ex-
pected to realize themselves and that no hindrances like antidotes are pre-
sent. 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 analy-
sis 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 hav-
ing a disposition or not. That is, are being F according to the disposition
and being F according to the realization compatible or incompatible predi-
cates? 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 hav-
ing 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 disposi-
tions 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 Aristotles paradigmatic example for this kind of
disposition. The rational formula that accompanies the art of medicine is
_____________
28 On this cf. Moline 1975.
29 On finkish (i.e. disappearing in a deceitful way) dispositions cf. Martin 1994; on antidotes cf.
Bird 1998.
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 realiza-
tions. 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 practitioners healing disposition. First, the practitioner
has to decide to activate his medical knowledge. But even this is not suffi-
cient: 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 appropri-
ate manner, which may bring about the patients health or the patients
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 pas-
sage:
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 dyna-
mis; 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 For a more detailed account cf. Jansen 2002, 78-92.
31 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.
Ludger Jansen
40
Thus whereas an active power is a principle of change in another or as an-
other, 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 condi-
tions, 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 dy-
namis, hexis is a word with many different meanings, to which Aristotle dedi-
cates a chapter in his dictionary of ambiguous philosophical terms (Metaphysics
V 20). The noun hexis derives from the verb echein, to have. As this etymol-
ogy 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 (kathhauto) 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 addi-
tion of a goal of action from the outside.
41
13. Does the Unmoved Mover Possess Dispositions?
Finally, I want to turn to one of the most prominent elements of Aristotles
metaphysics, the godly unmoved mover, who keeps the heavens in circula-
tion. 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 can-
not be present,
(Z3) everything that is capable (dynaton) of being may possibly (endechetai) not be ac-
tual.
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 chal-
lenge 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 hapls), 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 re-
alized, as when a man [is said] to be able to walk because he may walk. The latter be-
longs only to that which is changeable; the former can also belong to the unchange-
able things. [...]. Now, while the one way to be dynaton cannot truly be said of things
Ludger Jansen
42
being necessary in the unqualified sense, the other [way to be dynaton can be predi-
cated] truly. (De Interpretatione 13, 23a7-16; my translation)
The author here clearly distinguishes between inclusive and exclusive predica-
tion 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 un-
moved 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 pas-
sage 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 Aristotles
last word on the matter, then the unmoved mover cannot possibly have any
dynameis.
14. Is Aristotles Account of Dispositions a Good Theory?
Aristotles 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
Aristotles Metaphysics.
32
In particular, Aristotles theory of dynamis has been
the object of many disputes. There are three standard objections against it: (1)
Aristotles 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 objec-
tions 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 not-
being
34
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 half-
being: It is a full-fledged property of a full-fledged thing. It is, however, a full-
fledged 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 Hobbes, Leviathan, ed. Tuck, 461.
33 Cf. also Jansen 2001, 276-278 and Jansen 2004.
34 Hartmann 1938, 5 (Gespensterdasein).
35 Tegtmeier 1997, 36-40 (Halbexistenz).
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 disposi-
tion 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 worlds micro-
structure. This objection is often put forward in connection with Molires
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 exami-
nation:
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 [can-
didate] 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 perspec-
tive 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 be-
cause the question already presupposes that it is the opium which is the rele-
vant 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 satisfac-
tory, 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 respec-
tive 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 disposi-
tion. 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 On this scene and its background in the philosophical and theological discussions of Molires
time cf. Hutchinson 1991.
37 Cf. Molire 1926, VIII 328; the translation is Hutchinsons (1991, 245).
Ludger Jansen
44
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 Latto aristotelico e
le sue ermeneutiche, ed. Marcello Snchez Sorondo, 43-61. Rome: Herder 1990; German
transl.: Der Begriff der Wirklichkeit in der Metaphysik des Aristoteles In Aristoteles.
Metaphysik. Die Substanzbcher (Klassiker auslegen 4), ed. Christof Rapp, 289-311. Ber-
lin: Akademie-Verlag 1996.
Bird, Alexander. 1998. Dispositions and Antidotes. Philosophical Quarterly 48: 227-234.
Buchheim, Thomas, and Corneille Henri Kneepkens, Kuno Lorenz, eds. 2001. Potentialitt
und Possibilitt. 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 Mglichkeitsbegriff.
Sitzungsbericht der Preuischen Akademie 10 (1937), repr. in: Kleinere Schriften, vol. II. Ber-
lin: de Gruyter 1957, 85-100.
Hartmann, Nicolai. 1938. Mglichkeit und Wirklichkeit, Berlin: de Gruyter.
Hintikka, Jaakko. 1973. Time and Necessity. Studies in Aristotles 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.
45
Hntelmann, Rafael. 2000. Mglich ist nur das Wirkliche. Nicolai Hartmanns Modalontologie des
realen Seins. Dettelbach: Rll.
Hutchinson, Keith. 1991. Dormitive Virtues, Scholastic Qualities, and the new philoso-
phies. History of Science 29: 245-278.
Jacobi, Klaus. 1997. Das Knnen und die Mglichkeiten. Potentialitt und Possibilitt.
In Cognitio humana: Dynamik des Wissens und der Werte. Akten des XVII. Kongresses fr
Philosophie. ed. Christoph Hubig, 451-465. Berlin: Akademie-Verlag; repr. in: Buch-
heim/Kneepkens/ Lorenz 2001, 9-23.
Jansen, Ludger. 2000. Sind Vermgensprdikationen Modalaussagen? Metaphysica, special
edition 1: 179-193.
Jansen, Ludger. 2002. Tun und Knnen. Ein systematischer Kommentar zu Aristoteles Theorie der
Vermgen im neunten Buch der Metaphysik. Frankfurt: Hnsel-Hohenhausen.
Jansen, Ludger. 2004. Dispositionen und ihre Realitt. In Was ist wirklich? Neuere Beitrge
zu Realismusdebatten in der Philosophie, ed. Christoph Halbig and Christian Suhm, 117-
137. Frankfurt-Lancaster: Ontos.
Kosman, L. A. 1969. Aristotles definition of motion. Phronesis 14: 40-62.
Liske, Michael-Thomas. 1995. In welcher Weise hngen Modalbegriffe und Zeitbegriffe
bei Aristoteles zusammen? Zeitschrift fr philosophische Forschung 49: 351-377.
Liske, Michael-Thomas. 1996. Inwieweit sind Vermgen intrinsische dispositionelle
Eigenschaften? In Aristoteles. Metaphysik. Die Substanzbcher, ed. Christof Rapp, 253-
287. Berlin: Akademie-Verlag.
Lovejoy, Arthur O. 1936. The Great Chain of Being. A Study of the History of an Idea. Cam-
bridge 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 Aristotles Concept of Tvrpyrio: Tvrpyrio and
^uvoi. Ancient Philosophy 14: 73-114.
Molire, Jean Baptiste. The Plays of Molire, transl. A. R. Waller, 8 vols. Edinburgh: John
Grant 1926.
Moline, Jon. 1975. Provided Nothing External Interferes. Mind 84: 244-254.
Plambck, 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.
Souilh, Joseph. 1919. tude sur le terme ^YNAMI2 dans les dialogues de Platon. Paris: Felix
Alcan.
Stallmach, Josef. 1959. Dynamis und Energeia. Untersuchungen am Werk des Aristoteles zur Pro-
blemgeschichte von Mglichkeit und Wirklichkeit. Meisenheim am Glan: Hain.
Tegtmeier, Erwin. 1997. Zeit und Existenz. Parmenideische Meditationen. Tbingen: Mohr-
Siebeck.
van Rijen, Jeroen. 1989. Aspects of Aristotles Logic of Modalities. Dordrecht-Boston-London:
Kluwer.
Waterlow, Sarah. 1982. Passage and Possibility. A Study of Aristotles Modal Concepts. Oxford:
Clarendon Press.
Weidemann, Hermann. 1999. Aus etwas Mglichem folgt nichts Unmgliches. Zum
Verstndnis der zweiten Prmisse von Diodors Meisterargument. Philosophiegeschichte
und logische Analyse 2: 189-202.
Ludger Jansen
46
Witt, Charlotte. 2003. Ways of Being. Potentiality and Actuality in Aristotles Metaphysics, Ithaca,
NY: Cornell University Press.
Wolf, Ursula. 1979. Mglichkeit und Notwendigkeit bei Aristoteles und heute. Mnchen: Fink.
Zeller, Eduard. 1882. ber den xupiruev des Megarikers Diodorus. Sitzungsberichte der
Preuischen 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 Gefhrten des Odysseus.
Man kommt nicht weit, lauscht man dem christlichen Sirenengesang, der seit Jahrhun-
derten ablenkt von den klassischen Texten.
1

In the above passage, the German poet Durs Grnbein 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, Grnbeins
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:
Grnbeins metaphor provides us with a sufficiently complex picture of an-
cient 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 ex-
planations 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 Grnbein 2005, 395. The text continues: Mehr noch, man mte zuerst die Stimme des eigenen
Ichs unterdrcken 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 ones 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 ones self, because all the appeasements come from the inner side, from ones own
echosphere.)
Burkhard Meiner
48
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 peoples
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 etho-
graphies, 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. Plutarchs syncrisis of both begins with the
words this is, what we found worth mentioning about Romulus and Theseus.
The first of these two, out of his own design (apooproi) ....
4
The most
important term here is apooproi. It means a quality of ones behavior that
can be morally evaluated as bad or good. It manifests itself in a sense in ones
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
apooproi is very similar to the concept of disposition. And it seems to be

2 For a review of this debate in German cf. A. Meiner 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.
3 Negative: Raymont, Does Anything Break Because it is Fragile?; Raymont 2001. Positive: Mumford
1998; Vanderbeeken and Weber 2002.
4 Plutarch, Comp. Rom. et Thes. 1,1.
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 ones own
behavior.
In the case of Demetrius Poliorcetes and Marcus Antonius, Plutarchs
comparison first extends to the amount of rapid change that both figures
brought about. Afterwards, in its second chapter, he turns to the apooproi,
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 rgime over people
who had already resisted a similar attempt on Caesars part.
5
Yet, what is the role of dispositions in explaining rather than evaluating
human behavior? Let us look at an example from Plutarchs 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, Plu-
tarch 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 Plutarchs and the biographers mode of ex-
plaining 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 Plutarch, Comparatio Demetrii et Antonii 2,1: The disposition (apooproi) 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; Antonys disposition,
however, was bad and tyrannical, for after the Roman people had just been freed from Caesars monarchy, Antony
attempted at making them slaves anew.
6 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.
7 Plutarch, Pyrrh. 7, esp. 7,4(7)N5(10).
Burkhard Meiner
50
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 (ovopro), his corporeal fitness (peq), self-confidence (Uopoo)
and disposition to sound judgment (|povqo), 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 Plutarchs 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 Plutarchs 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 peoples 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, rri), in which he seems to have been followed
by the 2
nd
century Greek historian Polybius. Moreover, both use the oproi
or apooproi 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 J. Geiger, Cornelius Nepos and Ancient Political Biography, Stuttgart, Wiesbaden (1985) Cf. N.
Horsfall, Cornelius Nepos, Oxford (1989).
9 Aristoteles, Ethica Eudemia 1226b6-9: \ yop apooproi oproi rv rotv, ou oaie or, oii'
rtrpou apo rtrpou touto or ou oiov tr ovru oxrre xo oui[. oio rx ooq ouirutix[ rotiv \
apooproi. Cf. B. Meiner, HPAIMATIKB I2TOPIA. Polybios ber den Zweck pragmatischer
Geschichtsschreibung, Saeculum XXXVII (1986) 313-351.
51
The notions of hexis and proheiresis can also be found already in Plato,
although the term prohairesis is rather uncommon in Platos 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 (|iio) is defined as a apooproi of the
life of the other or as a community of apooproi and apoi
11
, and apoUuo
as a manifestation of a practical apooproi.
12
oproi 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 4
th
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 de-
throne the Median king Astyages with the help of the latters grandchild,
Cyrus. According to Herodotus, Harpagos intention was directed towards
revenging himself upon Astyages: Herodotus calls this raiUuriv
14
. Harpagus
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
raiUuo functions like a kind of background process or continuous condition
for what is realized under more specific conditions as an action, i.e.: revo-
lution in Persia. For the success of Cyrus plan guiding this revolt, a disposi-
tion is again one of the necessary and somehow sufficient conditions: In this

10 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 (apooiprtri),
choose what could rightly be called both?
11 Ps.-Plato, Definitiones 413a-b: Friendship: a kind of consensus as to what is good and just. A preference
(apooproi) 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.
12 Ps.-Plato, Definitiones 413e.
13 Cf. Meiner 2004.
14 I 123.
Burkhard Meiner
52
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 circum-
stances 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 agents 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 ovo, 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 Mero did. Thus Herodotus, as we see,
explains what was going on under Cambyses referring to the latters madness
as a cause. Being disposed to madness in Herodotus means: being unable to
overcome ones own emotions and to plan ones 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 Herodotus I 127.
16 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 be-
cause 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
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 vooi, 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 voo. Especially prominent in
Herodotus with his special interest in ethnographic material is, however, the
notion of voo in the plural, that is, vooi, 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 auctionee-
ring of maidens, temple prostitution and burial customs (I 196-200). One
Indian tribe is described as follows: Another tribe who lives eastward from these Indi-
ans 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

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...
17 Herodotus I 96: Deioces was the son of Phraortes. This Deioces was a lover of tyranny, and therefore he
undertook the following actions.
18 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; Meiner 2005, esp. 90ff.; 107ff.
Burkhard Meiner
54
woman falls sick, similarly the women most near to her socially do to her the same as the
men
19
. 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 peoples
laws are the consequences of peoples actions, as peoples actions are the
consequences of peoples 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 ioovoq
(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; upi and |Uovo, pride and envy, arise, and as a

19 Herodotus III 99.
20 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 Platos idea of a cyclic
schedule of revolutions to Aristotles 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 6-
8,1278b6-1280a6; Polybius VI 5,1-9,11. Cf. Meiner 1986, 322ff.
55
result, violence and lawlessness. Absolute power causes a kings behavior to
deteriorate.
21
Another interlocutor, being a partisan of oligarchic rule, refers
to the fact that the common crowds 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 beha-
vior 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 inter-
relation between the distribution of political power and the habits of those
who are in power. The latters 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 latters instability. Elsewhere, too,
Herodotus employs dispositions like envy (|Uovo) 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 dispo-
sitions 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 Marathon
25
. 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 A king behaves most inconsistently with himself (`Avopootototov or aovtev, III 80).
22 Herodotus III 81: ... Oiou yop opqou ouorv roti oouvrtetrpov ouor upiototrpov ...
23 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 (xoiov) nor
what is usual and fitting (oix\iov): 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 Platos cyclic inter-
pretation 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.
24 Herodotus VIII 124; cf. 125.
25 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.
Burkhard Meiner
56
collecting an army against Egypt.
26
Xerxes ryoio|poouvq, 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 Kings 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
changes in human behavior, the ovUpeaivov as he calls it
29
, was consequently
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
(opi) for the Messenians. Thus, character constants, peculiarities of the
situations and utilitarian calculations explain actions of war in Thucydides.
30

26 Herodotus VII 5.
27 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.
28 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 proba-
bly in order to tell the Great King at Sardeis by fire-signals across the islands that he had taken Athens.
29 Thucydides I 22,4.
30 Thucydides III 94-95.
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 un-
specified 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 robbery
32
. 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 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.
32 Thucydides I 7,1N8,2.
33 Thucydides I 71,1-4. About the innovative mood inherent in the alleged argument of the
Burkhard Meiner
58
Attitudes and dispositions are realized when they are applied to specific situa-
tions; they govern how people react to these situations; they can be changed;
they may be inappropriate to a given situation: pooutq, 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 peoples 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 (ioiototo) of all the citizens and as the rhetorically most
influential one (aiUovetoto
34
). In his speech, Cleon interpreted the
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 5
th
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.

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

The efficient cause of all this was greed and ambition (airovro, iiotio), from
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 lite, 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
Thu-
cydides calls what happened an opy\, 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 develo-
ped 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 Thucydides III 82,2-4.
40 Thucydides III 84,1N85,1.
41 Cf. Aristotle, Ethica Nicomachea 1105b19N1107a33.
Burkhard Meiner
60
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 dore with a disposition to extravagant life (aoiutriro), 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 Xenophons Hellenica, dispositions are used mainly in assessing individual
peoples ambitions and intentions. Alcibiades operated around Samos in 408-
407; he is said to have wanted to return to Athens with his soldiers
(ouiorvo rto tev otpotietev oaoairiv oxoor
43
). The fact that the
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
whole (ouiorvo opoi iq t[ Orttoio
44
): To explain the protagonists
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 di-
spositions 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 Thucydides, VI 12,2N13,1.
43 Xenophon, Hellenica I 4,8.
44 Xenophon, Hellenica II 3,4.
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 lite 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 alternati-
ve 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. ouiorvoi), 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 Xenophons 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
Xenophons 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. apoart\: being
inclined to) and moral choice (cf.: rixo: 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 Xenophons 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.
Burkhard Meiner
62
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 Messenes independence from Sparta was concerned, Philiscus
collected a large army of mercenaries to take the side of the Spartans
46
.
Eventually, a reinforcement army from Sicily came in to make up for this mo-
ve,
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
Xenophon labels oproi, decision. This line of thought is especially important
in Xenophons 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 authors
educational aims, the decision aspect of Xenophons 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 Romes expansion by the second
century Achaean historian Polybius, while explicitly aiming at a kind of

46 Xenophon, Hellenica VII 1,27.
47 Xenophon, Hellenica VII 1,28.
48 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.
63
practical and moral improvement of its readers (oiopUeoi), dwells at length
on rri, on dispositions in the Aristotelian sense, and on apooiprori,
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
(aopooriyo): Being gifted and a good king at the beginning of his career, his
manners turned worse when he took up licentiousness and a bad apooproi;
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 Philipps apooproi and the consequential development of bad habits.
There are numerous similar stories in Polybius: Positive ones like Scipio and
Philopoemen
52
and negative ones like the Cynaethians in Arcadia, who,
because of a lack of musical education, were eventually reduced to a merely
animalistic life
53
. In all these examples, there are natural conditions (|uoi),
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 di-
dactic 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 Cf. Meiner 1986, esp. 332ff.; Eckstein 1995; Petzold 1999, 182f.
50 Cf. Walbank 1938.
51 Polybius VII 11,3.
52 Polybius X 21f.; XXXI 25-30.
53 Polybius IV 20f.
Burkhard Meiner
64
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 ones dispositions can be as useful as a nomological ex-
planation 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 He-
rodotus, 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.
Grnbein, D. 2005. Zwischen Antike und X. In D. Grnbein, Antike Dispositionen. Auf-
stze, 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 quivalent des Fortschrittsgedankens: das Knnensbewusst-
sein 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.
65
Meissner, B. 1986. WXYZ[Y\]^_ ]`\aX]Y. Polybios ber 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. Anfnge und frhe Entwicklungen der griechischen Historiographie.
In Die antike Historiographie und die Anfnge 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. Beitrge zur europischen Geschichte, August Nitschke zum 65. Geburtstag
gewidmet, ed. M. Kintzinger, W. Strner, and J. Zahlten, 41-46. Kln/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 Ge-
schichtsschreibung, Petzold, K.E., 182f. Stuttgart.
Raaflaub, K.A. 1985. Die Entdeckung der Freiheit. Zur historischen Semantik und Gesellschafts-
geschichte eines politischen Grundbegriffes der Griechen. Mnchen.
Raaflaub, K.A. 2002. Philosophy, Science, Politics: Herodotus and the Intellectual Trends
of his Time. In Brills Companion to Herodotus, ed. E. J. Bakker, J. F. de Jong and H.
van Wees, 149-186. Leiden/Boston/Kln.
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. Einfhrung in die griechische Geschichtsschreibung. Darmstadt.
Meister, K. 1990. Die griechische Geschichtsschreibung. Stuttgart/Berlin/Kln.

Herodotus:

Burkhard Meiner
66
Bakker, E. J., de Jong, J. F. and van Wees, H. eds. 2002. Brills Companion to Herodotus.
Leiden/Boston/Kln.
Bichler, R. and Rollinger, R. 2000. Herodot. Olms Studienbcher 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 Historians 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 virt nelle Storie di Tucidide. In Actas del X congreso espaol de
estudios clsicos I: Sesiones de inauguracin y clausura, Lingstica griega, Literatura griega, ed.
E. Crespo and M. J. Barrios Castro, 351-55. Madrid: Sociedad espaola de estudios
clsicos.
Gaertner, M. 2000. Denken und Tun. Versuch ihrer Neubestimmung aufgrund moderner sthetischer
Sichtweisen. Diss. phil. Mnchen.
Heitsch, E. 1996. Geschichte und Situationen bei Thukydides (Beitrge 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 (Universit 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 Xenophons
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 mono-
graphs 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.
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. Xenophons 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 (Beitrge zur Altertumskunde
LXVIII), Stuttgart/Leipzig.

Polybius:

Beister, H. Pragmatische Geschichtsschreibung und zeitliche Dimension. In Rom und der
Griechische Osten, Festschrift fr Hatto H. Schmitt zum 65. Geburtstag dargebracht von
Schlern, Freunden und Mnchener 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: 83-
92.
Podes, S. 1990. Handlungserklrung 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 Descartes
1
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 situa-
tion 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 reflect-
ing the equilibrium of all the components, now seen as relations among the
variables composing each side of the equation. The variables may be con-
ceived 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 dis-
equilibrium, 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 vari-
ables 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 concen-
trate 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 Des-
cartes (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, Descartess 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.
Peter Machamer
72
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 perceivers body). It is this specificity of motion, being maintained
throughout, that provides the content for a particular sensation, i.e., the sen-
sation 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 Des-
cartes 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 be-
lieve 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 inclina-
tion of the individual human ego to love God and understand Him as best a
finite being is able. Through the training of the egos 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 man-
ners 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 ne-
glected 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 Des-
cartes have assumed that the late works, like the early work, are primarily an
73
attempt to describe the ontology of the world. Ted McGuire and I have ar-
gued at great length that this is a misreading of Descartes. We argue that Des-
cartes 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 contigu-
ous 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 proper-
ties to matter as he had done earlier in The World, the Discourse, and the Medita-
tions. 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 Gods action.
Gods 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, im-
pulse, 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 ref-
erence 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 mo-
tion 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.
Peter Machamer
74
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 movementas 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 Des-
cartes, for he uses this commitment, along with the principle of relational
motion, in an attempt to avoid Galileos 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; al-
though 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 con-
demned 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 transfer-
ence: And exactly the same force and action is required for one transfer-
ence 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 Gods 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 mo-
vement in contexts in which he employs the terms force and power. This
needs to be understood in the light of Descartes theory of continual recrea-
tionism 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 consid-
ers the nature of time that the same power and action are needed to preserve anything
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 Gods constant creation from
instant to instant to preserve the humans existence and identity is parallel to
Gods 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 con-
sists: 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 what-
ever 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 interpre-
tations.
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 doc-
trines 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 Gods creative power, are a causal source of the created minds 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
Peter Machamer
76
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 equi-
librium 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 mat-
ter. 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 ear-
lier 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 stones 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 vorti-
cal 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 exter-
nal 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 mo-
tion. 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 stones com-
plete and actual motion in a circle. This point is emphasized by Descartes in
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 sec-
ond 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 suns 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 tenden-
cies simultaneously) is consistent with by Descartes earlier analysis of the
path of light through a refracting surface. He divides the surface into a per-
pendicular 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. Simila-
rily, the sling acts like a refracting surface. It resists the stones radial tendency
to recede, whereas the parallel component, its tangential tendency, is unim-
peded. 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 in-
herent 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.
Peter Machamer
78
What can we conclude from the second law and Descartes treatment of
the nature of tendencies? First, they are always directed toward the conserva-
tion of the future states of a body just as in Gods recreating the human body
from instant to instant. Descartes affirms, rightly, that movement cant 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 bodys tendency to move in a straight line
and the conservation of whatever state it will turn out to be in. Now, although
each bodys tendency is conserved by God in each independent part of time
in accordance with the doctrine of recreationism, nevertheless, we must as-
cribe 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 Gods harmonious plan. God con-
serves the existence of material substances by recreating them in each inde-
pendent 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 Spinozas Ethics

URSULA RENZ
0. Introduction
Spinoza is often considered to be the Megarian among the early modern phi-
losophers: 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 re-
garded 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 Spinozas 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. Spinozas metaphysics
of modality is more moderate, and less absurd, than the Megarian picture that
many textbooks depict. As I see it, Spinozas 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.
Ursula Renz
80
sitions as fundamental forces or irreducible properties of things.
2
In contrast
to these well known accounts, Spinozas approach is more appropriately char-
acterized 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 ex-
plained 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. Spinozas 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 indi-
rectly 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 Des-
cartes, 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 de-
tail. I will give a short sketch of Spinozas 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 Spinozas so called
necessitarianism reduces to his rationalist conviction that all there is, including
contingent particulars and possibilities, can, in principle, be completely con-
ceived 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 Spinozas philoso-
_____________
2 Cf. also Schrenk and Mumford in this volume.
3 Cf. AT VII, 176.
Explicable Explainers: the Problem of Mental Dispositions in Spinozas 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 fun-
damental 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 Spinozas
concept of the human mind (section 2).
The third section is concerned with Spinozas 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 infer-
ential relations to other mental states or ideas. This conception has enormous
implications: It suggests that in the Ethics the term idea refers at least some-
times 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 un-
real or illusory. They do however not consist in some ultimate and undeter-
mined 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 de-
scribed above as radical rationalism mixed with common sense realism (sec-
tion 4).
1. A Type of Transcendental Philosophy:
Interpreting Spinozas 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
Ursula Renz
82
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 occur-
rence 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 Spinozas
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 possi-
bilia. 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 disposi-
tions 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 dispo-
sitional properties exist.
5
I think however that this necessitarian reconstruction of Spinozas meta-
physics 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 Garrett 1991, 191f.
5 At this point, I would like to address a distinction made by Ludger Jansen in his discussion of
Aristotles 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 intui-
tions. 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 expli-
cated 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 de-
scription of the actual state of Ludger Jansens mind.
6 1p29, cf. for the citation: Spinoza, Collected Works, ed. Curley, 433. In the following all English
citations are taken without change from Curleys translation.
7 1p33, Collected Works, 436.
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 Spinozas theological position
motivating the above claims. There is considerable evidence that he devel-
oped 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 Gods personhood as well
as of divine teleology, and he, of course, also knew that identifying Gods
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 necessitarian-
ism.
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 deter-
mined 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 ex-
ist nor be determined to produce an effect unless it is determined to exist and to pro-
duce 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 determina-
tion, 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 Gods necessary creation, but in terms of some
necessary antecedent and the infinite causal chain of particular things. More-
over, it has to be emphasized that he does not deny that, given different ante-
cedents, things could have turned out differently.
This suggests that Spinozas claims on modality are motivated by the epis-
temological intuitions sketched above, rather than by some metaphysical ideas
concerning the reality or non-reality of things. Spinozas 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 transcenden-
tal argument that analyzes the necessary presuppositions of a radical rational-
_____________
8 Cf. Letter 78. In: Spinoza Opera, vol. 4 (=G IV), 326f.
9 For the theological background cf. Carriero 1991 and Perler 2006.
10 It has to be mentioned that Spinoza does not regard God as a particular thing, as particulars are
merely affections or modifications of Gods attribute. Cf. also 1p25c, Collected Works, 431.
Ursula Renz
84
ism. I shall not discuss here whether or not this weakens Spinozas necessi-
tarianism.
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 meta-
physical premises of his concept of complete notions,
13
rejects necessitarian-
ism for theological reasons. Just as in Leibniz, Spinozas views on Gods free-
dom 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 Spinozas modal metaphysics so far, the fol-
lowing 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 concep-
tual space for dispositional properties in Spinozas 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 rela-
tions, which together amount to a certain disposition to do x. As humans we
_____________
12 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 conceptu-
ally related. This cannot be done without at the same time defending a particular interpretation
of other issues in Spinozas metaphysics. For further discussion cf. Schtt 1985, Garrett 1991,
and Curley/Walski 1991.
13 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.
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 hu-
man 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
Spinozas modal metaphysics. The restricted scope and the provisional epis-
temic legitimacy of dispositional concepts lead to an important shift in con-
ceiving of the main focus of a philosophical investigation of dispositional
terms. What philosophy primarily has to discuss is not whether or not dispo-
sitional properties really exist. The interesting question is rather how we should
conceive of them in order to account for the explanatory function they per-
form 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 Spinozas 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 Spinozas concept of the human mind
need to be exposed.
2. Re-categorizing Human Thought:
Spinozas Deflationary Concept of Mind
Spinozas 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 an-
swers the question of what individuates particular minds. This difference has always been ne-
glected in Spinoza scholarship. For a detailed discussion cf. also Renz 2006.
Ursula Renz
86
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 hu-
manity 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 Spinozas 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 differ-
ent 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 as-
sumption that it is the bearer of mental states or properties, which in turn
presupposes a categorical distinction between the ontological status of the
_____________
17 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.
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 de-
nial of any essential mental faculty.
18

Although puzzling at first glance, Spinozas claim that the human mind is
or consists of a particular idea makes quite a lot of sense. It presents a funda-
mental critique of any concept of the human mind that presupposes a cate-
gorical 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 Ryles, Spinozas
approach is not based on an analysis of language, nor is it driven by an em-
piricist 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 ana-
lyzed 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 an-
swer 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 Cf. also Bartuschat 1992, 96 and 101, and Yoshida 2004, 63f.
19 Cf. Ryle 1949, 16ff. I will not discuss here whether or not Ryles exposition of the Cartesian
myth provides a fair analysis of Descartes own approach.
20 Cf. the first of the two references to 2ax3 in 2p11dem; Collected Works, 456.
Ursula Renz
88
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 com-
pletely different from the one maintained either by the Aristotelian or Carte-
sian 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 cate-
gory 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 pro-
visional 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 indi-
viduated. 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 dis-
tinct 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 Eth-
ics, 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 some-
times 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 I argued for this in more depth in Renz 2006 and Renz 2009.
22 Cf. for these claims 2p7s, where Spinozas so called identity theory is defended. Collected Works,
451f.
23 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.
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 under-
stands the concept of ideas and how he uses it to address the problem of
mental capacities.
3. Explaining Dispositions in Terms of Content:
Spinozas Concept of Idea and Mental Causation
In order to understand Spinozas 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 con-
trary, 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 Spinozas
words: an affirmation.
26
This assertion is best clarified with the help of a short
_____________
24 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.
25 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.
26 Curley 1969, 123-127, suggested a logical reading of Spinozas concept of ideas. This has been
criticized by Wilson 1999, 153, and Della Rocca 1996, 8, who put forward psychological inter-
pretations. I think that the logical and the psychological interpretations are not necessarily alter-
natives 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
Spinozas approach is similar to the one put forward by Wilfrid Sellars in Empiricism and the Phi-
losophy of Mind.
Ursula Renz
90
examination of his discussion of the problem of free will. Just as in the case
of the concept of mind, Spinozas first step is to deny the existence of some-
thing 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 argu-
ing 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 proposi-
tional knowledge of its having certain properties functions in this manner as
the model for Spinozas concept of affirmation. In contrast to Descartes, for
whom only innate ideas necessarily involve the affirmation of some proposi-
tion, for Spinoza, it characterizes all of our ideas, including obscure and in-
adequate 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 nec-
essarily related to other ideas by different kinds of connections, inferential as
well as associative ones, and they can never be isolated from those connec-
tions. 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 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 under-
standing, 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.
28 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.
29 2p49dem, Collected Works 484. Descartes uses this example in the fifth meditation, cf. AT VII,
68.
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 Spinozas metaphysics was to
ensure complete intelligibility of being. One can argue that complete intelligi-
bility can be maintained if and only if, one singular concept of reality is af-
firmed 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 Spinozas notion of the idea of
God, and not the assumption of some divine self-consciousness which con-
tinuously generates our minds and its mental states, as it sometimes suggested
in pantheistic readings.
33

Instead, the question arises how Spinozas 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 Spinozas 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 char-
acteristic mental dispositions that we ascribe to particular persons, for exam-
ple, when we say someone is irascible. How can Spinoza account for these
kinds of properties? In other words: How can he explain that someone be-
haves, under certain condition, rather in this than in that way?
Concerning the first question, we can say that if there is an idea constitut-
ing 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 For Spinozas holism, cf. also Brandom 1976, and Della Rocca 1996.
31 Cf. 1ax6, Collected Works, 410.
32 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 corre-
sponds 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 co-
herence with all our other ideas, a particular idea can be epistemologically justified.
33 This view has been expounded and criticized by Wilson 1999, 126ff.
Ursula Renz
92
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 empha-
size 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 Kants transcenden-
tal 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 substan-
tial 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 some-
one behaves in this way rather than in that way? Well, this kind of mental cau-
sation works also in the case of the relation between a persons 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 con-
ceptual model for the explanation of our own as well as of other peoples
93
mental dispositions. Lets for example assume that I perceive myself as a
powerful personality. I will tend, in that case, to perceive other strong person-
alities 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
ones own behavior can, at least in principle, change ones mental disposi-
tions. 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 men-
tal 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 concep-
tual 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: Spinozas 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 Spinozas 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 crite-
rion 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 Spinozas 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
Ursula Renz
94
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 enor-
mously in what they can or cannot do. Even if we reject natural kinds, or
categorical differences between types of minds, we must, from a common-
sense point of view, be able to differentiate between certain types of minds.
We must however not merely posit these differences, but they must be some-
how accounted for. Spinoza therefore puts forward the idea of a correlation
between an individuals 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 dont think that is the case, since he merely states pro-
portionality between the amount of certain bodily and the amount of certain
mental capacities. This, moreover, has to be seen in the light of 17
th
century
psychology. Spinoza hereby rejects one of the underlying ideas of Descartes
Passions de lme 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 cor-
relation 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 For a detailed analysis of this point cf. Della Rocca 1996.
35 Cf. Collected Works, 458.
36 Collected Works, 458.
37 Cf. AT XI, 327f. and 354f.
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 at-
tribute to another. It does not, however, preclude the possibility that empiri-
cal knowledge provided by natural sciences can be used to account for dispo-
sitional 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. Ac-
cording 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 mo-
tion 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 bodys
capacity to perform certain causal roles corresponds to the proportion of its
minds 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 sug-
gests, 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 ra-
tional ascription of mental dispositions.
Spinozas 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 posi-
tion is much closer to our ordinary intuitions than it appears to be at first
glance. To illustrate this point, remember for a moment Thomas Nagels
_____________
38 Cf. 2p7: The order and connection of ideas is the same as the order and connection of things. Collected Works,
451.
39 Collected Works, 460.
Ursula Renz
96
famous question Whats 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 experi-
ence from a third person perspective. Nagel himself uses it in this way. Part
of the game of Nagels 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
whats 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 whats like to be in this or that position.
Spinozas 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 Spinozas historical horizon, but thats 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 whats 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 affec-
tions 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 experi-
ence.
Contemporary discussions in the philosophy of consciousness suggest
that we have to reject one of these two intuitions. Spinozas conceptual de-
vice, 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 com-
pletely 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 phyloge-
netic tree, people gradually shed their faith that there is experience at all. Nagel 1991, 423 (emphasis
U.R.).
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 con-
ceptual device which provides us with a structure that enables us to account
systematically for all types of mental capacities. The assumption of a propor-
tionate 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 Spinozas 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 Spinozas ra-
tionalism 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 regard-
ing 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.
Ursula Renz
98
Literature
Descartes, Ren. 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 Spinozas Ethics.
Journal of the History of Philosophy 147-162.
Carriero, John. 1991. Spinozas Views on Necessity in Historical Perspective. Philosophical
Topics 19: 47-96.
Curley, Edwin. 1969. Spinozas Metaphysics. Cambridge, Mass.: Harvard University Press.
Curley, Edwin and Walski, Gregory. 1991. Spinozas 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. Spinozas Necessitarianism. In God and Nature: Spinozas Metaphysics,
ed. Yirmiyahu Yovel.
Gueroult, Martial. 1974. Spinoza II. L'me. 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.), 59-
80.
Renz, Ursula. 2006. Die Definition des menschlichen Geistes und die numerische Diffe-
renz von Subjekten (2p11-2p13s). In Hampe/Schnepf (eds.), 101-121.
Renz, Ursula. 2007. Die Erklrbarkeit von Erfahrung. Realismus und Subjektivitt in Spinozas
Theorie des menschlichen Geistes. Zrich (manuscript: Habilitationsschrift).
Rousset, Bernard. 1996. Spinoza. Lecteur des Objections faites aux Mditations de Descartes et de ses
Rponses. Paris: ditions Kim.
Ryle, Gilbert. 1949. The Concept of Mind. London: Hutchinson.
Schtt, Hans-Peter. 1985. Spinozas Konzeption der Modalitten. Neue Hefte fr 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 Affektivitt. Untersuchungen zu Spinozas Theorie der
Politik. Wrzburg: Knigshausen & Neumann.
Harmonizing Modern Physics with Aristotelian
Metaphysics. Leibnizs 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: Conti-
nental 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 sci-
ence was coming up, originating from Descartes, which completely disap-
proved of the Aristotelian approach as being sterile. It regarded the Aristote-
lian 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 ac-
count of his philosophical development, Leibniz, having weighed one against
the other, came to a decision in favor of modern philosophy against his Aris-
totelian 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 un-
modified, 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
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 proc-
esses 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 con-
ception of possible worlds, which, originating in Duns Scotus and handed
down by baroque scholastics, took on a determining role for Leibnizs
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 teleo-
logical 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 im-
manent 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 speci-
men of the relevant kind is the goal of the individual process of growing and
maturing.) Having said this, we have reached Leibnizs 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 proc-
esses 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 Des-
cartes 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-
Leibnizs 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 actu-
alization 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 pri-
marily the cause of physical movements. In its primary sense, it is the origin
of an objects proper activity, which constitutes it as a unitary substantial
object. In this manner Leibniz equates primary force with the Aristotelian-
Scholastic 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
individuals being a man.
From the point of view of modal logic, three stages must be distin-
guished: 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
adults 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. Be-
cause 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 vol-
ume, disposition? Can Leibnizs 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 actualiza-
tion, 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
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 con-
nected 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 quantita-
tive 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 pas-
sive 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 proc-
ess) 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 activ-
ity, 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 ef-
fects. 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 attrib-
utes 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 mo-
ment 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 metaphysi-
cal 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 indi-
viduals perceptions according to a law characteristic for this individual and
thus constitutes its individuality (cf. s. IV).
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 differ-
entiation of physical forces, Leibnizs 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 primi-
tive 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
Leibnizs 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 ef-
fect 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
substances disposition amplified by a striving. Because of this striving, the
substance is directed and determined to a certain activity which by itself ex-
plains a change in another thing. It is in virtue of such a disposition of a sub-
stance that the change in the other thing immediately follows if nothing hin-
ders 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
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 recep-
tivity to action; rather, it involves a directed momentary velocity (conatus
1
), or tendency
toward action, so that action follows if nothing else hinders. This is what rvtrirrio
properly consists in too little understood by the scholastics. For such a potency in-
volves 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. Fur-
thermore, active force is twofold, primitive and derivative, i.e. either substantial or ac-
cidental. Primitive active force, which Aristotle calls rvtrirrio \ apetq, commonly
called substantial form, is the one natural principle that together with matter, or pas-
sive 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 differ-
ence 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
dynamics
3
illustrates clearly the two main features of Leibnizs 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 trea-
tise 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 Aristo-
telianism ascribed to substantial form, which is to make a corporeal substance
_____________
1 See SD I p. 10, l. 117-121 (GM VI 237). Leibnizs 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 bo-
dy's activity; see Gale 1973, 200.
2 All translations of quotations from Leibniz are my own.
3 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.
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 things 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 constitu-
tion is possible. This can be done, in Leibnizs 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 ob-
servations of nature; on the other hand, the concept of force allows for meta-
physical deepening. For this purpose, Leibniz employs the conception of a
principle of activity. In Theod. I 87, he interprets form (in Aristotles as well
as scholastic usage) as a principle of activity, inherent in the agent itself. (This
surely meets Aristotles 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
things 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 quali-
fies as form, in this case the act cannot be the action itself, but only a princi-
ple disposing to an action: be it the substantial form that disposes to the es-
sential 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 (lActe succes-
sif), being disposed to such an activity is a permanent state. It can be consid-
ered an act because it is an already developed potency that actualizes the re-
spective natural constitution or aptitude.
A developed potency that, given proper circumstances, immediately turns
into the related activity still does not satisfy Leibnizs demands of the concept
_____________
4 Or le Philosophe Stagirite conoit quil y a deux especes dActe, lActe permanent et lActe
successif. LActe permanent ou durable nest autre chose que la Forme Substantielle ou Acciden-
telle: la forme substantielle (comme lAme par example) est permanente tout fait, au moins se-
lon moy, et accidentelle ne lest que pour un temps. Mais lacte entierement passager dont la na-
ture est transitoire, consiste dans laction mme. (GP VI 150)
106
of force. Such a potency would be the active potency of scholasticism, with
which Leibniz, in the Emendation treatise, sharply contrasts his newly intro-
duced 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 pre-
conditions of activity. But according to Aristotles and the scholastic concep-
tion, it cannot transfer itself into activity. It must be activated from the out-
side 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 re-
gards such a conception as an internally inconsistent one. He expresses the
inconsistency of this conception by characterizing it as receptivitas actionis. Re-
ceptivity, as a things 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 potency
6
, which fulfills all preconditions of immediate actualiza-
tion, 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. Leibnizs characterization of substance as essentially
_____________
5 Differt enim vis activa a potentia nuda vulgo scholis cognita, quod potentia activa Scholastico-
rum, seu facultas, nihil aliud est quam propinqua agendi possibilitas, quae tamen aliena excita-
tione et velut stimulo indiget, ut in actum transferatur. (GP IV 469)
6 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.
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 mate-
rial entities, Leibniz can reconcile his metaphysical account of a body as al-
ways 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 Hartmanns total possibilities (Totalmglichkeiten) as
is clearly indicated by the qualification if nothing hinders. Leibniz regularly
uses this phrase in order to mark the condition for the potencys being en-
tirely 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 dis-
tinction 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 an-
other. 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 exter-
nal 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
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 dispo-
sition. In passages where he does speak of disposition, especially when (dis-
cussing Lockes 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 Lockes empiricism fails in that it cannot explain unrestricted uni-
versality that is characteristic of necessary truths of reason, for instance of
logic and mathematics. The empiricist method of induction allows for cer-
tainty only in cases that have been already observed; in cases not yet ob-
served, only greater or lesser probability is possible (GP III 291). The alterna-
tive view of Leibnizs innatism can be illustrated with the help of an example
that is not explicitly mentioned by Leibniz himself: The faculty to argue ac-
cording to a certain logical law, or to know how to use it correctly in an ar-
gument, 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 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 experi-
enced 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 (facult nue) or mere possibility of understanding those truths (pos-
sibilit 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 deriv-
able 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 ny a pas seulement dans nostre esprit une facult, mais encor une disposition la connois-
sance, dont les connoissances innes peuvent estre tires. Car toutes les verits necessaires tirent
leur preuve de cette lumiere interne et non pas des experiences des sens qui ne font que donner
occasion de penser ces verits necessaires et ne sauroient jamais prouver une necessit univer-
selle, faisant connoistre seulement linduction de quelques exemples et de la probabilit pour les
autres quon na point encor essays. (GP III 291)
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 lcole) 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 be-
cause, in reality, every potency implies a striving or tendency. Leibniz consid-
ers a pure potency, which, without any action, is completely at rest, an imagi-
nary 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 what-
soever.) Real dispositions, on the contrary, always imply some striving, di-
rected not to any arbitrary, but to some specific activity. Therefore, disposi-
tions always are specific dispositions to one activity rather than another (une
disposition particuliere une action plustost qu lautre), for instance the disposi-
tion which enables me to speak a certain language. Because of these tenden-
cies, a disposition always brings about some at least minimal action, as Leib-
niz also stresses when he treats force. Yet, when he says: Beyond disposition
(outre la disposition), there is a tendency toward action, he means that disposi-
tion itself as a developed potency still does not include a tendency, but never-
theless 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 laction), which contains a predetermination such that the effect is
completely preformed in nuce. For the determinist, such dispositions or de-
veloped potencies, which create determination, are the conditio sine qua non of
causal efficacy (une cause ne sauroit agir, sans avoir une disposition laction). Bare
potencies, which are possibilities of alternative courses of events, according to
Leibniz, can only be abstractions of thought. This understanding is also im-
plied in the preface to his Theodicy (GP VI 40), where Leibniz calls the disposi-
tion of matter, conferred to it by God, organism, that is, those organizational
structures which dispose to a characteristic way of acting.
110
3. The Reinterpretation of Aristotles 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 meta-
physics of substance. According to Aristotle, a particular substance is consti-
tuted by form and matter as two complementary principles. In Leibniz this
corresponds to the view that primary active and passive forces are inter-
woven. (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 Aristotles first entelechy (rvtrirrio \ apetq) 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 im-
portant transformations. The concept of entelechy is not only widened in
extension, but also modified in its sense. Aristotle defines soul as first entel-
echy, 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 contrar-
ies in favor of degrees continually succeeding one another, any absolute in-
surmountable 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 Aristotles
term consists in. For Aristotle, on the other hand, the concept of rvtrirrio
qua completion (to bring to an end = trio) and the concept of rvrpyrio 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 childs faculty of commanding an
army, see de anima II 5, 417b 30f.) which has yet to be developed before ena-
bling 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 rele-
vant act as soon as proper circumstances are given. This understanding of
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 teleol-
ogy,
8
may have suggested itself to Leibniz because of the word trio which
is part of rvtrirrio (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 hinder-
ing 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 Aristotles first entelechy)
and derivative entelechy. This complex conception of entelechy is docu-
mented 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 sup-
ports 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 de-
rivative entelechy, which is relevant to physical movement
9
and its elements,
for instance conatus and impetus.
Leibnizs interpretation of Aristotles concept of matter as passive force is
another important modification. Certainly, for Aristotle too, matter is charac-
terized by a passive potency to be formed or shaped as something definite.
Such receptivity might be understood as a things 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 distin-
guish 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 Kangro 1969, 135 stresses the close connection of final cause and entelechy in Leibniz.
9 On entelechy as principle of physical movements, see Bergmann 2002, 158-163.
112
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 addi-
tional element by which force is distinguished from mere potency, is essen-
tially 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 opposi-
tion 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 effi-
ciency 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 with-
out the first bodys 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 inertia
12
, tends to remain in its state of rest as well as in its
state of steady linear motion, any change in some bodys state of motion re-
quires 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 defini-
tion 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 mat-
ters bulk. Observed phenomena only seemingly contradict this claim. Ac-
cording 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 Porro to ouvoixov seu potentia in corpore duplex est, Passiva et Activa. Vis passiva proprie
constituit Materiam seu Massam, Activa rvtrirriov seu formam. (GP IV 395)
11 It is not by chance that Leibniz, in his 1702 treatise on dynamics (GP IV 395, the passage inter-
preted 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 man-
ner.
12 On Leibnizs notion of inertia, also on the background of Keplers and Newtons notion of
inertia, see Bernstein 1981.
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 Aristotles form and matter; the distinction between
primitive and derivative forces corresponds to Aristotles substance and acci-
dent. According to a letter to Jaquelot (dated March 22, 1703) (GP III 457),
primitive (active) force is essential to the particular substance (being, like
Aristotles first entelechy, a principle constituting substance). It belongs to the
particular substance taken by itself or inheres in it as an absolute characteris-
tic. 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 them-
selves, 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 explain-
ing 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 non-
arbitrary way. Although Leibniz presents this argument in the Discourse on
Metaphysics 18
14
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 modi-
fication 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 Hacking 1985 tries to justify by means of a physical criterion Leibnizs assumption that motion is
a non-real (phenomenal) mechanical quantity and that vis viva is a real one.
14 See Garber 1985, 81.
114
dated June 30, 1704; GP II 270). This action is to be defined as force multi-
plied by the duration of its activity. The momentary state seems to be some-
thing 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 con-
ceive of the derivative force as the momentary state in relation to the follow-
ing one (momentaneum est, sed cum relatione ad statum sequentem). It is the momen-
tary 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 pro-
posal 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 prae-
involvit) as everything present is pregnant with the future development. But the persis-
tent itself, insofar as it includes all cases, possesses a primitive force. Therefore, primi-
tive 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 neverthe-
less (et tamen) modifications, he wants to say that between these two concep-
tions there is, at least superficially regarded, a contradiction. For a modifica-
tion
15
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 gen-
erally) 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?
Leibnizs solution comprises two thoughts: First, the modification is the tran-
sitory 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 chan-
ge 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.
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 inter-
pret primitive force as the basic tendency of the striving characterizing indi-
viduality. 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 ten-
dencies of simple substances by means of which these substances pass from
perception to perception according to a certain law of their (individual) na-
ture. This basic direction of the striving confers to the succession of states a
characteristic peculiar to this individual. It thus constitutes an individual sub-
stance 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 determi-
nistic view.
17
For a determinist, it is a central question of how it is possible
_____________
16 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 percep-
tual changes) is inconsistent with each of the two other views, for which there is textual evi-
dence in Leibniz: the conceptual unfolding view and the efficacious perception view (that per-
ceptual 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 substances 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 states 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 forces determining the whole sequence of perceptions according to a certain fundamental
tendency.
17 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 suc-
cession 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
that the whole future development is already contained in the present mo-
ment. 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 unchang-
ing 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 char-
acterized 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 under-
stood 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 Leibnizs the-
ory 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
_____________
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 dy-
namics serves as an important basis of his entire system, for it makes clear the difference be-
tween 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 Humes 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 candi-
date. See Schrenk in this volume.
18 See C. Wilson 1987, esp. 165sq.
117
a striving disposing to an activity, provided all hindrances are removed. Leib-
niz 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 poten-
tial energy includes the philosophical conception of potency, or power to act.
Here, Leibnizs 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 at-
tracted 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 bodys 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 bodys 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). Conse-
quently, 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 be-
tween resting (dead) potential energy, being a momentary infinitesimal striv-
ing 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 Leibnizs 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
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 Leibnizs 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 ac-
quires 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, Gali-
leos 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 phe-
nomenon. 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 Leibnizs doctrine of dynamics, which is deeply influenced by metaphysical as-
sumptions. 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 dy-
namics becomes relevant when Leibniz identifies as the source of Descartes error that he mis-
takenly generalizes what is true in statics and applies it to dynamics. See Freudenthal 2002, 582
sq. Freudenthals main subject is the controversy between Leibniz and Papin, who critizises the
Brevis Demonstratio. According to Kvasz 2001, Leibnizs demonstration is false from a scientific
point of view because he neglects the motion of the earth.
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 acknowl-
edging 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 per-
ceptions 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 can-
not 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 clear-
ly: 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: ac-
tive 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
Philarte 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
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
monads 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 percep-
tions, 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 charac-
teristic 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 sub-
stances, in the case of passive forces, or matter, the phenomenalistic view
proves to be more conclusive. It can separate unequivocally the primary pas-
sive forces, which hinder and modify the activity of the monads, (materia pri-
ma) 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 im-
penetrability and inertia of bodies, one is left with an unsatisfactory distinc-
tion: Impenetrability and inertia, taken as materia prima, constitute the material-
ity 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
Leibnizs 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.
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 subjects 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 sub-
stance must, if it is to turn itself into actualization (as is required by its auton-
omy) 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 bodys 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 Leibnizs 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 cer-
tain height. This means according to Galileos laws of free fall: it is not the
product of mass and velocity (Descartes) but the product of mass and veloc-
ity squared.
24
The concept of force implies that it is directed toward produc-
ing 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 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 spontane-
ously accomplish itself; for a tending is presupposed by the powers spontaneously actualizing it-
self rather than being actualized by external conditions). This understanding of Leibnizian force
certainly is essentially correct. From a terminological point of view, however, Duchesneaus use
of disposition is inconsistent both with the present use, according to which a disposition must
be actualized by suitable external conditions, and with Leibnizs own use in NE II 1, 2 (as we
have examined in II above). One should better say: The tendency to transfer itself into actualiza-
tion, 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 Leib-
niz, in reality there are never mere dispostitions but always amplified dispositions.
24 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 meas-
ured by the effect) (see Gueroult 1962, ch. V, esp. 117), but also empirical generalizations such as
Galileos laws of free fall (see Duchesneau 1998, esp. 79-81).
122
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) under-
stood force only as active force and as being directed to acting. It took Leib-
niz 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 perma-
nent 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 individ-
ual to be constituted.
Regarded from a slightly different point of view: The systematic place of
the concept of force within Leibnizs 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
Leibnizs philosophical thought, can be found related to other concepts al-
ready 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 On Leibnizs initial conception of force and the ensuing differentiation, see Boudri 2002, ch. 3.2
and 3.3, 75-91.
26 Gale 1970 makes a threefold distinction of entities and their properties: those on the observable
level, those on the explanatory level of Leibnizs 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 ex-
planation of observable facts. This epistemological tripartition is discussed in Duchesneau 1994,
215-217.
123
from April 1679 (A VI 4, 1410-1441), for example, he conceives of determi-
nation 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 immedi-
ately follows (A VI 4, 28 l. 11, 308 l. 5) and which can be considered the im-
mediate 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 fol-
lows is later in the order of things (natura posterius); its existence can be ascer-
tained 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 ex-
plained 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 fac-
tor 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 sub-
stance ontology, decidedly endorses the concept of what is nowadays called
agent causality. Even if we immediately observe events or actions as the pre-
ceding state after which a change follows, these events are nothing isolated,
free-floating, but must be related to an acting subject, an agens, which pro-
duces 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 On Leibnizs conception of actio and agens in the period leading to the Discours de Mtaphysique, see
Schneider 2001.
28 In the Discourse of metaphysics, Leibniz expresses the thesis that actions properly are to be attrib-
uted to individual substances, according to the scholastic tenet actiones sunt suppositorum. On this
tenet and Leibnizs general theory of activity, see McGuire 1976, s. II.
124
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) de-
termined 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.
Leibnizs 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 (devel-
oped 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 observa-
tions.
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 phi-
losophical 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 sys-
tems property. Taking into account the fact that laws of nature are idealiza-
tions, one has to say: What is ascribed to the system according to this under-
standing 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 dispo-
sition in general, not force as a disposition amplified by a striving. If the con-
cept 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 condi-
tions 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 See also Httemanns paper in this volume.
30 See also Schrenks paper in this volume.
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 be-
comes active in form of a striving. Rather, a disposition must always be striv-
ing 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 de-
grees.
Literature
A VI 4 Leibniz, Gottfried Wilhelm. 1999. Philosophische Schriften 1677-Juni 90.
In Smtliche Schriften und Briefe. Sechste Reihe: Philosophische Schriften, Bd.
4, ed. Berlin-Brandenburgische Akademie der Wissenschaften zu Berlin
und Akademie der Wissenschaften in Gttingen. Berlin: Akademie Verlag.
A VI 6 Leibniz, Gottfried Wilhelm. 1962. Nouveaux essais. In Smtliche Schriften
und Briefe. Sechste Reihe: Philosophische Schriften, Bd. 6, ed. Deutsche
Akademie der Wissenschaften zu Berlin: Akademie Verlag.
GM Leibniz, Gottfried Wilhelm. 1962. Leibnizens mathematische Schriften.
Bd. I-VII, ed. Carl Immanuel Gerhardt. Hildesheim: Georg Olms.
GP Leibniz, Gottfried Wilhelm. 1978. Die philosophischen Schriften von G.
W. Leibniz. Bd. I-VII, ed. Carl Immanuel Gerhardt. Hildesheim: Georg
Olms.
SD 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-Universitt
Mainz.
Bernstein, Howard R. 1981. Passivity and Inertia in Leibnizs Dynamics. Studia Leibnitiana
13: 97-113.
Bobro, M. & Clatterbaugh, K. 1996. Unpacking the Monad: Leibnizs Theory of Causal-
ity. 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, Franois. 1994. La dynamique de Leibniz. Paris: J. Vrin
Duchesneau, Franois. 1998. Leibnizs 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
Gabbey, Alan. 1971. Force and Inertia in Seventeenth-Century Dynamics. Studies in the
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 Leibnizs
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.
2
1962. Leibniz. Dynamique et mtaphysique. 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 Gre, insbesondere der action
motrice, bei Leibniz. Studia Leibnitiana Supplementa 2: 133-149.
Kvasz, Ladislav. 2001. Leibnizs criticism of the Cartesian physics. In Nihil sine ratione.
VII. Internationaler Leibniz-Kongre. Vortrge 2. Teil, ed. H. Poser, 669-676.
Lodge, Paul. 2001. Primitive and Derivative Forces in Leibnizian Bodies. In Nihil sine
ratione. VII. Internationaler Leibniz-Kongre. Vortrge 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 Univer-
sity Press.
Papineau, David. 1977. The Vis Viva Controversy: Do Meanings Matter? Studies in the
Schneider, Martin. 2001. Leibniz Theorie der Aktion im Jahrzehnt vor dem Discours de
mtaphysique 1677-1686. Studia Leibnitiana 33: 98-121.
Wilson, Catherine. 1987. De Ipsa Natura. Sources of Leibnizs Doctrines of Force, Activ-
ity and Natural Law. Studia Leibnitiana 19: 148-172.
Wilson, Margaret D. 1976. Leibnizs 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 Dispositions
1
OLIVER R. SCHOLZ
1. Context: When, Where and Why
In order to do justice to Gilbert Ryles contribution to the topic of this an-
thology, it is important to be clear about the historical and argumentative con-
text in which he discusses dispositions. The locus classicus of Ryles treatment
is The Concept of Mind, first published in 1949. To be sure, there are adumbra-
tions 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 mani-
festo, exemplification and test of a new philosophical method. It is the first
book-length application of the special brand of ordinary language philoso-
phy
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 phi-
losophy? or, as he puts it in his autobiography, the question What consti-

1 I would like to express my gratitude to the organizers and participants of the Wittenberg confer-
ence for many stimulating discussions. In addition, I want to thank Rosemarie Rheinwald, Rich-
ard Schantz and Ludger Jansen for their comments on one of the final versions. Special thanks
are due to Rudolf Owen Mllan for correcting my English.
2 The best book on this movement, its philosophical methods and achievements is von Savigny
1969 (on Ryle see especially Chapter 2).
3 Cf. Hacker 1996, 168-72; Place 1999, 364-69.
Oliver R. Scholz
128
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 Sys-
tematically 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 back-
ground:
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 en-
tomology. Platonic Forms, Propositions, Intentional Objects, Logical Objects, per-
haps, sometimes, even Sense-Data were recruited to appease our philosophical han-
kerings 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 phi-
losophy? [...] Philosophical problems are problems of a special sort; they are not
problems of an ordinary sort about special entities.
5
The philosophers 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 philosophers 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 Car-
nap.
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 pub-
lished 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 re-
corded and proper to facts of quite another logical form than the facts re-

4 Ryle 1970, 12.
5 Ryle 1971b, vii.
6 Ryle 1970, 7.
7 Ryle 1970, 6sq.
8 In the 1940s, it came to a parting of the ways. In a review of Meaning and Necessity, Carnaps phi-
losophical method and his theory of meaning are criticized in an unusually harsh tone. See Ryle
1945b.
129
corded.
9
The main task of philosophy is to detect, correct and avoid confu-
sions caused by such systematically misleading expressions. In Categories
(1938), Ryle claimed, in a similar vein, that we are in the dark about the na-
ture 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 ana-
lytical 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, Ryles magnum opus The Concept of Mind is first and foremost
a meta-philosophical journeymans 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 im-
portance, 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 ab-
surdum of mind-body substance dualism, the disposition-occurrence opposi-
tion being one of the major weapons against the Cartesian Myth.
As we shall see, Ryles obsession with questions of philosophical method
did not, of course, prevent him from making controversial claims about sub-
stantial subject-matter, e.g., claims about disposition talk and even claims
about dispositions.

9 Ryle 1931-32, 143.
10 Ryle 1938, 189.
11 Ryle 1970, 12.
12 In academic psychology, William Stern and others began to prefer the term dispositions rather
than faculties (Vermgen). His book Die menschliche Persnlichkeit (1918) has a long chapter on
dispositions (Die Dispositionen), where he explicitly recognized the mutability of psychologi-
cal 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 disposi-
tional analyses of mental phenomena. In addition, Wittgenstein had suggested a dispositional ac-
count of understanding; Braithwaite and others construed beliefs as dispositions (Braithwaite
1932-33).
Oliver R. Scholz
130
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 (lets call these claims
meaning claims
13
or conceptual claims); (2) claims about dispositions per se in con-
trast to other ontological categories such as occurrences (for obvious reasons,
I shall call these claims ontological claims).
As we have already seen, Ryles main interest in The Concept of Mind is in
the words and sentences we use for describing specifically human behav-
iour; lets 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.
Lets 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 predomi-
nantly 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 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.
14 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 ex-
amines technical concepts which are part of philosophical or scientific jargon, e.g. the concept
volition (Ryle 1949a, Chapter III).
15 Ryle 1949a, 43.
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 sta-
tes of affairs, nor yet reports of unobserved or unobservable states of affairs. They
narrate no incidents.
17

(Quote 4) To talk of a persons 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 persons 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 Ryles 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 de-
scribed 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 psychologi-
cal 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 Ryle 1949a, 51.
17 Ryle 1949a, 125.
18 Ryle 1949a, 199.
19 The modal sentence (containing can, could or would) that provides the analysis for a disposi-
tion 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 speci-
fies the nature of the manifestation.
20 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.)
21 See e.g. Ryle 1949a, 125.
Oliver R. Scholz
132
(M.2.2) Some psychological ascriptions are not purely dispositional, but only partly
so
22
(i.e., some psychological ascriptions are equivalent in meaning to semi-
hypothetical 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 condi-
tional or a conjunction of subjunctive conditionals where the antecedent of each con-
ditional specifies some situation, and the consequent specifies a reaction of the sub-
ject 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 as-
pect: 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 See e.g. Ryle 1949a, 97, 117, 141.
23 As has been noted, Ryles 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).
24 Ryle 1949a, 83.
25 Ryle 1949a, 83.
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 self-
luminous, 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 conse-
quently 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 rela-
tion of strict exclusion between occurrences and dispositions on the ontologi-
cal 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, commit-
ted 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 re-
minded 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 conse-
quences. (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 Ryles 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 cau-
ses.
(O.1.6) Dispositions are not real properties of substances.

26 Ryle 1949a, 161.
27 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?
28 See especially Ryle 1949a, 63; cf. Alston 1971, 363.
Oliver R. Scholz
134
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 disposi-
tions.
(O.2.7) If psychological ascriptions imply the existence of occurrences, these are oc-
currences 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
Ryles account. Naturally, I shall focus on questions related to the main theme
of this anthology.
3.1 Critique of Ryles Applications in Philosophical Psychology
Let me mention, in passing, some rather moderate criticisms. Many philoso-
phers have contested Ryles 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 Ryles
disposition-occurrence opposition.
Terence Penelhum, e.g., defended an episode-view of pleasure against
Ryles dispositional account.
31
Ullin T. Place rejected the dispositional account

29 See, e.g., Ryle 1949a, 43, 116.
30 Cp. Mellor 1974, 161.
135
of heeding, attention and consciousness.
32
It seemed to many commentators
that Ryles view is a nonstarter as an analysis of our phenomenal concepts,
such as sensation and consciousness itself.
33
3.2 Critique of Ryles Account of Dispositions: The Main Objection
The most fundamental problem with Ryles various claims in The Concept of
Mind may be put thus: Ryle does not clearly distinguish between his concep-
tual 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 conclu-
sion, which is clearly of the second type (O). However the corresponding
pairs of conceptual and ontological claims may be logically and epistemologi-
cally related, it seems clear that the ontological claims do not simply follow
from the meaning claims. To make this more concrete, lets 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 cau-
ses.
(M.1.7) Disposition ascriptions can be tensed.
(O.1.8) Dispositions can change.
Lets 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 ana-
lytic-synthetic distinction. If you are not afraid of the analytic-synthetic dis-
tinction, 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 wont help Ryle. On the contrary,

31 Penelhum 1956-57.
32 Place 1954 and 1956.
33 Chalmers 1996, 14.
34 Medlin 1967; Alston 1971; see also Places comments on Medlin in Place 1999, 386-91. (Place
doesnt mention Alstons similar critique.)
Oliver R. Scholz
136
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, lets 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 predi-
cate.
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 disposi-
tion we mean being a disposition and nothing else and by being a pure occur-
rence 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 ana-
lysed 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 disposi-
tional 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 be-
tween 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 metaphysicians nightmare. One could try to soften it by
breaking down Ryles ontological hotchpotch into four classes of categories:
(a) occurrences, events, happenings, incidents; (b) processes (something in-
volving duration and internal change); (c) acts (an initiation of a state attrib-

35 This was pointed out by Alston 1971; throughout this whole section I am indebted to his sharp-
sighted critique. In some cases, I have nevertheless tried to improve on his formulations.
36 Ryle 1949a; cf. Alston 1971, 368.
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. Al-
ternatively, 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 some-
thing 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.
Oliver R. Scholz
138
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 indi-
viduated? Since I do not have the space to discuss the principles of individua-
tion
38
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 onto-
logical level turn out to be a process or a state? I suspect, the answer has to
be: No!
3.3 Critique of Ryles 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 ques-
tions 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 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.
39 Alston 1971, 381sqq.
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 expressions 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 ob-
jects 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 condition-
als, 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. Ryles 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
Ryles 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. Ryles Legacy
Lets take stock, then. After all this indolent censoriousness, I shall start with
the positive. Gilbert Ryles The Concept of Mind still offers one of the most
comprehensive and detailed surveys of what is nowadays called folk psychol-

40 Reference is more closely related to truth than meaning is.
41 Martin in: Armstrong, Martin and Place 1996, 71; cf. Place 1999, 387sq.
42 Geach 1957, 6sq.
43 Geach 1957, 7.
Oliver R. Scholz
140
ogy. The distinction between knowing how and knowing that still has its
defenders as well as detractors.
44
Ryles 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) ordi-
nary psychological concepts are dispositional concepts, typically denoting
multi-track dispositions. But, as we have seen, his general account of disposi-
tions 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: 1-
40.
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. Wittgensteins Place in Twentieth Century Philosophy. Oxford: Blackwell.
Hempel, C. G. 1935. Analyse logique de la psychologie. Revue de Synthse 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: Appleton-
Centry-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 See Gregor Damschens paper in this volume.
45 Cf. e.g. the diagnosis of the endemic mereological fallacy in psychology and neuroscience offered
by Bennett and Hacker 2003, 68-85 and passim.
46 Hilary Putnams and Jerry Fodors functionalism is different; they take their inspiration from the
idea of a Turing machine.
47 See the contributions of Andreas Httemann, Stephen Mumford and Markus Schrenk to this
volume.
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: 44-
50.
Place, U. T. 1999. Ryles Behaviorism. In Handbook of Behaviorism, ed. W. ODonohue
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 Wrterbuch der Philosophie, vol. 2, ed. J.
Ritter and K. Grnder, 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: 69-
76; 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 Einfhrung in die ordi-
nary language philosophy. Frankfurt am Main: Suhrkamp.
Schlick, M. 1931. Die Kausalitt in der gegenwrtigen 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 Persnlichkeit. 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 Dispositions
1
MARKUS SCHRENK
0. Abstract
If dispositions were real properties they would bestow the world with a con-
nection in nature a Humean could not accept (1.). The nature of that disposi-
tional connection or power is the target of my papers main argument. I will
argue that it can not be cashed out in terms of metaphysical necessity. Meta-
physical 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, disposi-
tions 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 battle-
ground for Humean vs. anti-Humean intuitions for many decades the condi-
tional analysis of dispositional predicates (2.) but I believe (yet cannot prove
here) that the arguments generalise to causation and causal laws straightfor-
wardly.
_____________
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 Re-
search Fellow within the AHRC Metaphysics of Science project.
Markus Schrenk
146
1. The Humean Legacy
Humes Followers: There are two major strands in 20
th
century analytic philoso-
phy which both share the Humean sentiment that necessary connections in
nature
2
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 veri-
ficationism 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 observa-
tional 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 par-
ticular 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 dif-
ference 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 ar-
rangement 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 Lewiss
reductive analysis.
Note the following superficial difference between these two groups of op-
ponents of dispositions: logical empiricism was occupied with the task of con-
firming an epistemic credo by the means of a semantic dogma: all factual knowl-
edge 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 neces-
sity 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. Humean-
supervenience 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 metaphysi-
cal pointillism
3
.
However, deep down both neo-Humean supervenience and logical em-
piricism share the same root and this is the aforementioned hostility to (hid-
den) links in nature. That is, they both subscribe to Humes arguments against
any such connection. I will briefly rehearse Humes well known line of reason-
ing.
According to Hume all objects of human reason and enquiry may natu-
rally 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 en-
tirely 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 exis-
tence 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 ma-
chine, 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)
Humes first argument, that we have no a priori way of discovering necessary
connections in nature, is widely accepted amongst both Humes supporters
and his opponents such as dispositionalists.
_____________
3 A term I borrow from Jeremy Butterfield.
Markus Schrenk
148
Humes Opponents: However, a fast-growing community of philosophers
does not any longer accept Humes 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 contin-
gent, yet, metaphysically necessary links which are partially argued for on the
grounds of semantic considerations about direct reference and rigid designa-
tion but also on grounds of scientific discovery.
Many people have been impressed by these arguments and especially anti-
Humeans, such as dispositionalists, saw the chance of a revival of anti-
Humean de re connections. Stathis Psillos, for example, comments (here speak-
ing of laws rather than dispositions):
It was Kripkes 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 Kripkes influence has to be thought of as merely
psychological rather than philosophical: Kripke has opened peoples 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 disposi-
tionalists 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 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 in-
teresting for my later arguments are passages in Evan Faless Causation and Universals, especially
chapter 1, p. 48.
5 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.
6 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 doc-
trine 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.
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 docu-
mented and some counterexamples to it have reached folkloric status in ana-
lytic 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 at-
tempt to analyze dispositional predicates, like x is soluble, and its rejection
can be found in Carnaps Testability and Meaning. is meant to symbolize the
material implication:
S(x) iff T(x) R(x) (cf. Carnaps 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 dis-
solves, 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 antece-
dent, 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 knowl-
edge it was Jan Berg who coined the term random coincidence in the context of
dispositional predicates (Berg 1960). An example modelled after Bergs 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 Nave Conditional Analysis: Even readers who are not familiar
with the dispositions debate will immediately have a multitude of promising
Markus Schrenk
150
ideas concerning how one could avoid either the void satisfaction or the ran-
dom 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 un-
wanted 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 be-
long to a kind of object or material which usually shows explo-
sive behaviour (the wood the table was made of would be ex-
cluded).
(v) We could abandon the explicit definition of dispositional predi-
cates 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 Carnaps suggestion to define disposi-
tional 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 Kailas
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. sugges-
tion (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 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 de-
scribing the symptoms of dispositions. The fact that reduction sentences are mere implicit defini-
tions (or conditioned definition as Carnap calls them (Carnap 1936, 443)) should suit the dispo-
sitionalist well. Spohn and Mellor both mention reduction sentences with sympathy (cf. Spohn
1997, Mellor 1999).
8 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.
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 counter-
factual 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 impli-
cation: as soon as the antecedent is false the whole conditional is true regard-
less 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 ran-
dom 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 elephants 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 Burks 1951, 1955, Sellars 1958.
10 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).
11 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.
Markus Schrenk
152
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.
Martins 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 op-
tion, 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 conduc-
tor (Martin 1994, 5). Martins 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).
Martins 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 disposi-
tion.
153
Lewiss Analysis: I turn first to David Lewiss answer to Martin. In Finkish
Dispositions (1997), Lewis did not want to follow Martins bold step. Instead,
he offers a sophisticated reformulation of the primitive counterfactual analysis
adding both a certain ontological assumption about dispositions and time vari-
ables. 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.
Lewiss basic way of accounting for Martins fink cases is expressed thus: A
finkish disposition is a disposition with a finkish base (Lewis 1997, 149).
Heres 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 xs having of B would jointly be an x-
complete cause of xs 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 Lewiss account. Remember that the fink is, as much as the disposition it-
self, supposed to be activated by the trigger conditions of the disposition.
Once they occur the fink destroys the objects basis B for the disposition. Yet,
Lewiss 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 predi-
cates. The void satisfaction difficulty is taken care of by the mere fact that
counterfactual conditionals are employed and the random coincidence objec-
tion, here in the guise of finks and reverse-finks, is ineffective too. Unfortu-
nately, there is a whole zoo of little machines, interferers, and unwanted proc-
esses 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 Lewiss analysis. I will focus on antidotes as intro-
duced by Alexander Bird in his Dispositions and Antidotes (Bird 1998).
Birds Antidotes: Birds 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,
Lewiss demand for saving the intrinsic basis is futile.
Markus Schrenk
154
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 Martins arguments, to de-
fine 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 ar-
chetypal 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 Birds ideas. Yet, for reasons of space I can only further highlight the
fact that, in his final analysis of dispositions, Bird keeps Lewiss 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 pari-
bus 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 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.
14 For further discussions see, for example, Molnar 1999, Malzkorn 2000, Gunderson 2000 with a
reply by Bird (Bird 2000), Mumford 2001, and Choi 2003.
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 dispo-
sitionalists 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 Aesops
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 by-
stander challenges the athlete to demonstrate his capacities here and now.
Antidotes for Dispositionalists: I am especially interested in what the disposi-
tionalists 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 objects 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 Liptons
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 counterex-
amples 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 po-
sition: if no counterfactual analysis ever succeeds can we conclude that dispositions are real? Scep-
tical 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 pro-
gress? (Heil 2005, 345)
Markus Schrenk
156
- 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 dispositions manifestation.
- Displaying: An object has the disposition, the relevant trigger conditions
are fulfilled and, no surprise, it displays the dispositions 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 sugars 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.
Birds antidote case, however, is quite different: the nuclear power stations
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 (ura-
nium-type) case, of an electron in an electric and a gravitational field. Accord-
ing 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 ca-
pacity and vice versa.
Now, contrast the latter case of a force equilibrium with the case of a sta-
tionary 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 Liptons present-but-not-
displaying into two subgroups, one of which accommodates cases where the
disposition is activated and hence pushing towards its manifestation? After all,
isnt this what dispositionalism is all about? Dispositions, we are told, are pow-
erful properties, which bring about their manifestation; or at least they try. It is
precisely this power or link towards their manifestation that the dispositional-
ists, but not the Humeans, should have available in their ontology. Therefore,
157
there should be no problem for dispositionalists in accepting a fourth stage,
where the disposition is pushing, trying, or aiming to manifest.
16

Whats the push? Actually, I have no reason to question this fundamental
insight about dispositions. As will become clear later, I endorse this anti-
Humean 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 coun-
terfactual 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 dispositional-
ists, 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 conceptualiz-
ing 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 anti-
Humean dispositionalists and the Humean reductionists. The major meta-
physical difference between, on the one hand, the anti-Humean and, on the
other hand, the empiricist or neo-Humean, is their belief or, respectively, dis-
belief in necessary connections in nature. Naturally, one might think, de re ne-
cessity 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 dispositions 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 connec-
tion. 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
Elliss 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 Harr
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.
Markus Schrenk
158
Kripke, subtitled their book Causal Powers with A Theory of Natural Neces-
sity. 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 tradi-
tionally been called metaphysical necessity. (Ellis 2002, 110)
When Ellis talks about necessity (and he uses all of physical necessity, natu-
ral 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 neces-
sary 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 typi-
cally 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 H
2
O. Yet, if meta-
physical necessity is typically attributed with respect to one property having
another property, then metaphysical necessity is normally a synchronic busi-
ness. In the case of dispositions, however, we are confronted with a different
affair. There, we have characteristically a diachronic case of one property in-
stance (or event) at t, namely the trigger (plus other activation conditions if
needed), and another property instance (or event) at t+Vt, namely the disposi-
tions 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 neces-
sity. The dispositions trigger event leads with natural necessity to the disposi-
tions 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)
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 syn-
chronic 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 Elliss idea to antidote and similar cases.
Necessitys 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 Elliss
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 al-
though 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 accommo-
dating Es 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 Goodmans 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 ac-
knowledged and which was once a reason for Humeans, like David Lewis, to
develop semantics for counterfactuals. But why should, now, metaphysical
Markus Schrenk
160
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 argu-
ment: C has never been the correct first relatum of the necessary relation un-
der 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 Birds uranium pile
was well above critical mass at time t and free of any interfering factor. How-
ever, 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 dia-
chronicity (as opposed to synchronicity) strikes back.
The second problem is that metaphysical necessity is, next to being mono-
tonic, discrete: that some two specific properties or event types, an uninter-
fered with C and an E, are necessarily linked (and hence conjoined in all possi-
ble 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 can-
not help to explain a dispositions 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 necessitys discreteness: only pure (uninterfered with) Cs and pure Es are
linked by metaphysical necessity. Yet, C* bringing about of E* remains unex-
plained.
18

Spelled out in anti-Humean terms (and we have accepted an anti-Humean
picture as our working hypothesis), the problem is that if theres only a neces-
sary link between C and the absence of any interfering factor and E then, be-
cause of necessitys discreteness, C*s (the impure cases) push or oomph to-
_____________
17

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, necessar-
ily, 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.
18

Andreas Httemann has convincing arguments for dispositional realism which revolve around
what he calls CMDs: continuously manifesting dispositions. The upshot here is that these con-
tinuous displays cannot be captured by metaphysical necessity.
161
wards an impure E* remains mysterious. Y (CxEx) does not explain C*
bringing about E*.
19

(3) You might want to put forward the following idea at this point: meta-
physical 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 al-
teration to a particular situation would qualify it as the instantiation of a differ-
ent 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 al-
most certain that any of these ultra fine grained events occurs only once in the
whole of the worlds history. Therefore, second, what happens regularly and
what happens with necessity falls apart because nothing ever happens repeat-
edly. (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, dispo-
sitions, 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 ber-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 neces-
sity 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

In yet other words, the necessary relation between C and E cannot have any influence on a situa-
tion 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).
20 These could even be states of the whole world at specific times.
Markus Schrenk
162
modelled too much after logical or conceptual necessity. The Kripkean anti-
Humean 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 Mumfords 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 differ-
ence 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 metaphysi-
cal 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 tradi-
tional 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 bro-
ken.
Dispositional necessity: The having of one property may dispositionally make necessary the
having of another property. For example, having gravitational mass necessitates attrac-
tion of other objects. (Mumford 2004, 177)
The intuitions behind these characterizations are clearly akin to mine. Some-
how 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
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 necessi-
tates attraction, that is, in all possible worlds, where there are gravitational
masses they attract each other. Mumfords valuable dynamic force has, how-
ever, 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 Mumfords 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 na-
ture 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 be-
ing 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 [] [Disposi-
tional possibility] is the connection. (Mumford 2004, 177)
I believe that Mumfords 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 ac-
tive.
Against this background, it should also be clear why I have earlier on sub-
sumed 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 papers pretentious title the motto for a possible solu-
tion. What we must find is a proper conceptualization of the pushes disposi-
tions 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 electrons Coulomb capacity faces an antidote: a
Markus Schrenk
164
gravitational force. Coulombs disposition is present, triggered but not display-
ing its display would be the electrons 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: intui-
tively, 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 disposi-
tional 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 elec-
tron case carry over to other cases that do not obviously involve forces: a
wires 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 Coulombs 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 possi-
bility of analyzing everyday dispositions in terms of molecular goings-on that
involve forces could be taken as a warranty for metaphorical talk about dispo-
sitional 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 macro-
scopic 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.
165
low et al. 1988) and, very recently, Jessica Wilson in Newtonian Forces (Wil-
son 2007). There is no space to discuss their arguments here, but I hope that
they can rescue forces from phlogistons 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 condi-
tions 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 travel-
ling 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, Leibnizs active force and Aristotles
dynamis as presented in this volume by Michael-Thomas Liske and Ludger
Jansen respectively.
7. Conclusion
Humes arguments against connections in nature have predominantly been
read as a rejection of necessity in nature. Yet, necessity, especially when formu-
lated 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 ma-
chine, is entirely concealed from us, and never discovers itself in any of the sensible
_____________
22 I owe this suggestion to Stephen Williams.
Markus Schrenk
166
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:
614630.
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 Birds Antidotes. Australasian Journal of Philosophy 81: 573-
580.
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 Kreativitt der bilateralen Reduktionsstze. Erkenntnis 9: 383-
392.
Essler, Wilhelm and Rainer Trapp. 1977. Some Ways of Operationally introducing Dispo-
sition Predicates with Regard to Scientific and Ordinary Practice. Synthese 34: 371-
396.
Fales, Evan. 1990. Causation and Universals. London/New York: Routledge.
Gunderson, Lars. 2000. Bird on Dispositions and Antidotes. Philosophical Quarterly 50: 227-
229.
Harr, Rom and E. H. Madden. 1975. Causal Powers. A Theory of Natural Necessity. Oxford:
Basil Blackwell.
Heil, John 2005. Dispositions. Synthese 144: 343356.
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 mnskliga kunskapen. Stockholm.
Kaila, Eino. 1942. Der physikalische Realittsbegriff. 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
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. Kirch-
berg/W.
Malzkorn, Wolfgang. 2000. Realism, Functionalism and the Conditional Analysis of Dispo-
sitions. 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 Kings 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 Disposi-
tions. 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 Mas-
ters Thesis, Bonn University. Available in German on Schrenks website:
http://www.nottingham.ac.uk/philosophy/staff/schrenk/MarkusSchrenk/Home.
html
Sellars, Wilfrid. 1958. Counterfactuals, Dispositions, and the Causal Modalities. In Minne-
sota 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. Begrndungen apriori oder: Ein frischer Blick auf Dispositions-
prdikate. In Das weite Spektrum der analytischen Philosophie Festschrift fr Franz von Kut-
schera. Ed. Wolfgang Lenzen, 323-345. Berlin. As A Priori Reasons: A Fresh Look at
Disposition Predicates available in English on Spohns website: http://www.uni-
konstanz.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 Einfhrung von
Dispositionsprdikaten. 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: 173-
205.
Ascribing Dispositions

STEPHEN MUMFORD
upon the whole, there appears not, throughout all nature, any one instance of con-
nexion 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 ab-
solutely 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 as-
cribed with relative ease, suggesting that there is at least some idea of a dispo-
sition 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 con-
taining threats and promises but, if certain forms of empiricism are to be
believed, there is only a limited degree of empirical justification for our dispo-
sitional vocabulary. If we are just inferring from past regularities in the behav-
iour 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 con-
nection 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 Humes argument, however. For that see Mumford 2004, chap.
4.
169
account for all our disposition ascriptions. Indeed, it is characteristic of dispo-
sitions 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 ontol-
ogy. I proceed to present an argument for the acceptance of such an ontol-
ogy. 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 dispo-
sitions are being treated as mind-independent features of reality that are pos-
sessed 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 prelimi-
naries, 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
Stephen Mumford
170
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 condi-
tions. 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 num-
ber 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 appropri-
ate 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 when-
ever 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-than-
material 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. Birds (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 Carnaps (1936/7) reduction sentences were, however, an attempt at an account that is restricted
to material conditionals.
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 distin-
guished 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. Mellors 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 criti-
cal, 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 pur-
pose, which they clearly do.
One final kind of case is worthy of note. It might be regarded as the op-
posite 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 con-
stantly. 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 de-
tected, 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 non-
manifestation 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 bro-
ken. 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 unmani-
fested powers and also that there are appropriate stimulus conditions for their
manifestation. The importance of this latter point is almost entirely over-
looked as the focus remains on conditional statements: on antecedent and
stimulus conditions. The insurmountable problem of the nave conditional
Stephen Mumford
172
analysis of dispositions is that we cannot specify non-trivially all the condi-
tions 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. Ex-
perience shows us that, despite all the potential preventers, there are appro-
priate 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, so-
metimes theoretically based, upon which we make disposition ascriptions. As
we investigate the way that the world works, we will note patterns of behav-
ior. The periodic table, for instance, groups elements with like constitution
and is able to explain their behavior in a way that becomes projectable. Ele-
ments 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 ma-
chine 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 un-
manifested power?
There is a further source of the idea of power, of which I have been per-
suaded 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 concentra-
tion 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 By Markus Schrenk, in conversation.
4 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 interpreta-
tion 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.
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 Armstrongs claim is controversial), we can just as easily
gain the idea of a power or disposition through touch. Indeed, I would ven-
ture 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 pro-
fesses that he has no idea of necessary connection but there is some plausibil-
ity 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 balls 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 neces-
sary 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 empiri-
cism, such as verificationism or its modern forms in anti-realism. The verifica-
tionist 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 mani-
fested? 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.
Isnt 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 pow-
ers 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-
Stephen Mumford
174
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 dis-
positions is not an immediately and conclusively empirical one. There is at
least some credibility to the thought that we could have all the worldly experi-
ences 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
Lewiss 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 em-
pirically but only in the court of metaphysics. If a metaphysics of causal pow-
ers 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 metaphys-
ics. 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 correspon-
dence theory of truth generally then we should not be put off by it in meta-
physics 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.
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 on-
tology 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, transcen-
dental 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. Some-
thing that is square will not have this power but it will have others. This sug-
gests 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 some-
thing 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 doesnt 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-
Stephen Mumford
176
erty is just a cluster of causal powers.
6
There is something more for the pow-
ers theorist in this account because it means that an ontological reduction is in
the offing. If one accepts powers as ones 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 primi-
tively numerically different (1983, 160 and 1989, 59). Lewis (forthcoming) has
a similar line, accepting that the identity of a property is a primitive, irrespec-
tive of its causal role. But as Black (2000) argues, this means that two proper-
ties 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. Some-
thing 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 some-
thing 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) aban-
doned 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 prop-
erties. 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 coinstanti-
ated. 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.
177
around but are never realised in something actual. All remains pure potential-
ity.
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 al-
ways 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 potentiali-
ties 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. Armstrongs assumption is
that a disposition becomes actual only in its manifestation. And if this mani-
festation 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
Shoemakers 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. Shoe-
makers attraction to the original theory was that it was only through a prop-
ertys causal powers that we could know of it. By pulling properties and pow-
ers 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 sug-
gest, 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 po-
wer, or cluster of powers, then causation would become the mere passing
around of powers. The world would be a world of shifting potencies (Arm-
strong 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 po-
wer to heat another thing, such as my body. I sit by the fire and it causes me
to become hot. The fires 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
Stephen Mumford
178
when they are heated, they have the power. That the world is a world of shift-
ing potencies sounds plausible. A scientific account of the world might char-
acterise 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 fires 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 things properties. But we have already seen
in the previous section that properties can be understood as clusters of causal
powers. To change a things 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 proper-
ties. 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 automati-
cally 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.
Humes paradigm the perfect instance of causation was the billiard
balls crashing around the billiard table. Events were the relata of causal rela-
tions. 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 contin-
gent 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.
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 to-
gether. 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 Humes paradigm well.
Though there are some very broad conceptions of events, what is very inter-
esting about this particular case is that, because of causation, nothing is hap-
pening. 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 chan-
ge 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
Stephen Mumford
180
as powers have already been discussed. The theory of events I favor, how-
ever, is one where an event essentially is a change. As Lombard (1986) de-
scribes 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
worlds history. But this remains a regularity theory, in which there is no ne-
cessity 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 ura-
nium 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 uni-
versally 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 Armstrongs 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).
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 disposi-
tional 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 es-
sential 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 wouldnt be massive. Initial skep-
ticism about how a dispositional theory would deliver the requisite regularity
of laws (Everitt 1991) is thus dispelled. If a property is identified and consti-
tuted 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 con-
cept 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 worlds proper-
ties ever behave differently to the way that they do? It seems they cant 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 ac-
count of laws, because he thought all properties were categorical. This ac-
count 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 nave 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,
Stephen Mumford
182
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 Lewiss (1986) modal real-
ism, 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 dis-
tinct 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 addi-
tional modal properties. But the material conditional would be hopeless for
counterfactuals as they all have false antecedents so would be true trivially.
Lewiss 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 conse-
quent was true in all the closest possible worlds to ours in which the antece-
dent 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 counterfac-
tuals 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
Armstrongs (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 Armstrongs theory has
unconstrained recombinability of wholly distinct elements. Is this too permis-
sive?
A theory of powers is a theory of how certain properties make other
properties possible and exclude still others. This would place restraints on
Armstrongs combinatorial freedom. Yet it might be objected that this pro-
vides us only with a theory of physical possibility and necessity? Surely there
could be a sense in which something is logically possible while being physi-
cally 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.
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 Armstrongs 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 pos-
sibility 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 be-
lieve, explain modality.
10. The Balance Sheet
I have indicated how a theory of powers might generate theories of proper-
ties, causation, events, laws and modality. I do not rule out that other meta-
physical 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 assump-
tions. 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
Stephen Mumford
184
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 pres-
ence 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 transcenden-
tal argument. It is not intended to be transcendental in the full-blooded Kant-
ian 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.
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: Clar-
endon 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 gen-
eralization about the nature of dispositions that go beyond conceptual truths.
Examples of such generalizations include claims that all dispositions are in-
trinsic, 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 ascrip-
tions. 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. Ill consider the question when do we
have good reason to believe that a disposition ascription is true? If our dis-
position ascriptions are true, then we are right about what kinds of disposi-
tions 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 arent 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 disposi-
tions are necessarily this or that. Debates about dispositions are often pre-
sented as dichotomies between two extremes.
2
These dichotomies include
intrinsic versus extrinsic, grounded versus ungrounded, reducible versus irre-
_____________
1 For example Ellis 2002; Molnar 2002; Heil 2005.
2 For example, see Ellis 2002, 59-60.
187
ducible, fundamental versus derivative, essential versus non-essential, first-
order versus higher-order, natural versus unnatural, causally inert versus caus-
ally efficacious, and non-existent versus universal. However, theres 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 exam-
ple, a power to act and a liability to be acted upon. Some terms, such as pro-
clivity suggest rational agency, while others such as force suggest funda-
mental 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, solu-
bility, elasticity, fitness, and fertility; character traits, such as integrity, or being
punctual, neat, kind, considerate, and shy; common qualities of physical ob-
jects, such as being elastic, comfortable, flammable, intoxicating, or hazard-
ous; and complex social concepts, such as marketable, redeemable, tax-
deductible, 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 fol-
lowing 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 sub-
junctive conditional to the effect that if the circumstances of mani-
festation were to occur, then the manifestation would occur is true;
5. The term is semantically equivalent to an overtly dispositional locu-
tion 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).
Jennifer McKitrick
188
assert that each is necessary. Counter-examples to conditional analyses sug-
gest 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 neces-
sity 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 dispo-
sitions are triggered in just one kind of circumstance and manifest themselves
in only one way. For example, the only manifestation of solubility is dissolv-
ing, and being immersed in liquid is its only circumstances of manifestation.
Perhaps there are some differences between particular immersions or dissolv-
ings, 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 manifes-
tations which occur in different kinds of circumstances. Various circum-
stances, 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 For a discussion of counterexamples to conditional analyses of dispositions, see Smith 1977,
Johnston 1992, Martin 1994; Bird 1998.
5 See Ludger Jansen in the volume for more on Aristotles many distinctions among dispositional
properties and the like.
6 Ryle 1949, 118.
189
litical stand. Perhaps the circumstances and manifestations can be given a
unifying abstract characterization such as fear-inducing situations which in-
volve 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 Broads termi-
nology. Some pieces of metal are magnetic are disposed to attract or repel
other metals and magnets. However, other pieces of metal arent 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 li-
abilities, 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 some-
thing has a liability, it has a disposition to suffer change. In the case of liabili-
ties, 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. Poi-
sonousness, 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 some-
thing else changing.
One may argue that this distinction does not hold up to scrutiny as meta-
physical, 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 isnt. But its not clear on what grounds we make that distinction. Con-
sider explosiveness, which intuitively seems like a power. But the manifesta-
tion 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 Broad 1925, 432.
8 Harre 1970, 84; Aristotle 1941; Locke 1990, 105-107.
Jennifer McKitrick
190
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 sup-
posed to be a liability. One may argue that, in the case of explosiveness,
things external to the bomb are affected as well. However, this doesnt 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 prag-
matic, perhaps other distinctions are more robust. Philosophers often distin-
guish sure-fire or deterministic dispositions on the one hand, and tendencies
or probabilistic dispositions on the other. In the case of a sure-fire disposi-
tion, 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 de-
cay 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 ten-
dency even if it doesnt 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 manifesta-
tion 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 condi-
tional analyses consist of imagining a possible scenario in which a disposed
object is the circumstances of manifestation yet fails to manifest the disposi-
tion. They seem to show that virtually all dispositions are such that, if objects
which instantiate them are placed in the circumstances of manifestation, its
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 cant 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 See Jansen on tendencies (2006).
10 See Johnston 1992; Martin 1994; Bird 1998.
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 thats the conceptual
distinction at least.
3. Semantic Support for Pluralism
Because philosophers have distinguished so many different dispositional con-
cepts, and the terms that bear the marks dispositionality are so diverse, dispo-
sitional 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 non-
dispositional. However, it seems clear that some terms bear marks of disposi-
tionality and others do not. Even absolutists who say there are no dispositions
and others who say that all properties are dispositions acknowledge a distinc-
tion 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 ex-
trinsic is that they are relations an object has to possible events the manifes-
tations. However, if the third mark of dispositionality obtains even some-
times, 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 dy-
adic 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 For example see Armstrong 1973, 15; Shoemaker 1980, 211. However, others suggest that
theres no clear way to demarcate a dispositional/non-dispositional distinction even at the con-
ceptual level (Mellor 1974, 171; Goodman 1983, 41).
12 Ellis 2002, 60.
Jennifer McKitrick
192
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 dupli-
cates. 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 ex-
ample, 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 thats
hidden in a dark room is not visible, but its well-light, out-in-the-open dupli-
cate 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 distinc-
tion. 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 prop-
erties, 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 plural-
ism 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 dont, the liberal property theorist allows that most predicates
refer to properties, and distinguishes between natural and unnatural proper-
ties. 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 Rosenberg 2004, 211.
14 See McKitrick 2003 for an extended defense of extrinsic dispositions.
15 Armstrong 1996, 18.
16 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).
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 realiz-
able. Multiply realizable properties can be shared by objects with different
realizer properties, so they are equivalent to disjunctive properties, and dis-
junctive 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 disposi-
tional predicates seem to pick out natural properties, while others pick out
unnatural properties.
Dispositional essentialists claim that all dispositions are essential proper-
ties of the objects which instantiate them.
18
An object with a certain disposi-
tion 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 genu-
ine dispositions are similarly definitive of the objects which instantiate them.
On the opposite extreme is the view that dispositions are non-essential prop-
erties. On such a view, an object could lose any of its dispositions if, for ex-
ample, it were subject to different laws of nature.
However, attributions of dispositional predicates follow not such stric-
tures. 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 ob-
ject 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 proper-
ties. If I attribute a disposition to an object, learning that the disposition as-
cription was true in virtue of the fact that the object had some distinct prop-
erty would not give me a reason to withdraw my disposition attribution. For
_____________
17 Lewis suggests this sort of picture (1986, 224).
18 Ellis 2002, 59.
Jennifer McKitrick
194
example, suppose I claim that Joe is irritable. Then Im told that Joe is irrita-
ble 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 prop-
erty, reduction is not elimination, and I have no reason to withdraw my dis-
position 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 deriva-
tive, 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 disposi-
tions.
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 dor-
mitive virtue to the sleeping pill, many seem convinced that he has not re-
vealed 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 ones reasons for thinking
that dormitivity is inert. Those reasons may or may not apply to all disposi-
tion terms. For example, if you think that dormitivity is a second-order prop-
erty 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 ex-
clusion 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.
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?
Ive argued that we employ numerous and diverse dispositional concepts.
What does that tell us about the world? I hope it is not too nave 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 pro-
ject 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 doesnt 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 doesnt commit one to the
existence of dispositions. However, it is not easy to say what disposition as-
criptions 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; how-
ever, 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 Schrenks 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).
Jennifer McKitrick
196
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 dis-
position. Of course, we are fallible, and we might over-generalize or misiden-
tify the relevant respect of similarity. The process is more detailed and con-
trolled in a scientific experiment, and of course, the case is much more
complicated when the manifestation is itself unobservable, such as the mani-
festation 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, Im going to concentrate on what
strikes me as the hardest case ungrounded or bare dispositions. Is the evi-
dence for bare dispositions on a par with our evidence for grounded disposi-
tions? 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 mysteri-
ously lost the disposition in the mean time.
However, if we are going to be that skeptical about the stability of dispo-
sitions, 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 re-
sumes its former shape, I figure that it has the disposition of elasticity, and I
presume theres 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 its 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 prac-
tice of assuming that, other things being equal, things retain their dispositions,
serves me pretty well. I dont see where the ground of the disposition, or lack-
197
there-of, makes a difference in my confidence that things retain their disposi-
tions 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 disposi-
tions, one may argue, its 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 evi-
dence 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 disposi-
tion, such as a tablet that is water-soluble. Your evidence for the tablets wa-
ter-solubility is the observable properties it shares with things that have dis-
solved in the past. But the observable properties are unlikely to be the causal
basis of the tablets water-solubility. You do not observe a particular molecu-
lar structure, or anything that is a plausible candidate for being a causal basis
of solubility. Similar points can be made about unmanifested elasticity, fragil-
ity, inflammability, etc. So, either we are not justified in believing such ascrip-
tions 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 dispo-
sitions, 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.
Jennifer McKitrick
198
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 doesnt 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 disposi-
tonal 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 neverthe-
less 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 disposi-
tions? 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 con-
vinced 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 asso-
ciation with triggering events, manifestations, or conditionals, nor is it nor-
mally 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 prop-
erty 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 disposi-
tion is.
Finding plausible examples of nondispositional properties is good reason
to reject pandispositionalism. While the pandispositionalist might talk you
_____________
22 Proponents of this promissory note conception of disposition terms include Quine (1969, 20)
and Armstrong (1973, 15).
23 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.
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 experi-
ences, 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 ef-
fect, onward into the future. If theres 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 prop-
erties, either the disposed object, some other object(s), or both: the elastic
band takes on a new shape; the provocative cape changes the bulls 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? Heres one way of formalizing APNT in an
attempt to clarify just whats wrong with this picture:
1. A manifestation of a disposition is constituted by a particular acquir-
ing 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 manifesta-
tion).
_____________
24 Thats one interpretation of whats going on in Mellor (1974, 171) and Goodman (1983, 41).
25 Molnar 2003, 159.
26 This objection appears in many places. A nice discussion appears in Molnar 2002, 173-181.
Jennifer McKitrick
200
4. If all properties are dispositions and b acquires some dispositions,
then either bs new dispositions remain latent or some c acquires
some dispositions.
5. If all properties are dispositions and c acquires some dispositions,
then either cs new dispositions remain latent or some d acquires
some dispositions.
6. If all properties are dispositions and d acquires some dispositions,
then either ds 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 manifesta-
tion, or by
b. a particular having a disposition that gives something a dis-
position, which produces no manifestation, or by
c. a particular having a disposition that gives something a dis-
position, which gives something a disposition, which has no
manifestation, or by
d. a particular having a disposition that gives something a dis-

position, which gives something a disposition, ad infinitum.
9. It is not plausible that every manifestation of a disposition is consti-
tuted by a, b, c, or d.
10. Not all properties are dispositions.
The strength of this argument depends on premise 9. The disjunctions enu-
merated in premise 8 are supposed to bring out the sense in which nothing
really happens in the pandispositionalist world.
One response thats often made against the APNT argument is that it il-
legitimately assumes that dispositions are not fully real, and so that when
something gains a disposition, nothing really happens. However, I acknowl-
edge that something gaining a disposition is an event, or something happen-
ing. However, I find it odd to suppose that all that ever happens is that
things loose and acquire dispositions.
Maybe whats wrong with the scenario described in premise 8 above is
that its not clear how one could ever observe a manifestation of a disposi-
tion. 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.
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 unob-
servable.
8. Some properties are observable.
9. Therefore, not all properties are dispositions.
This argument relies on the contingent claim that some properties are observ-
able and perhaps the argument would not be sound in a world where observa-
tion 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 8
th
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 posi-
tions 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 disposi-
tions. 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 pandispositional-
ism 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 prcising definition
stipulating that this loose, umbrella term is going to be used in a more specific
Jennifer McKitrick
202
way. Mentioning that you are using disposition in the philosophers and not
the lay persons sense is insufficient.
27
Even the philosophers sense is quite
pluralistic, as the philosophical distinctions at the outset show.
A second, pragmatic issue is the usefulness of theorizing about disposi-
tions. Often, dispositional theorists try to generate interest in dispositions by
pointing out how pervasive dispositional concepts are. This is false advertis-
ing if a scant minority of these concepts actually corresponds to real disposi-
tions, on the theorists view. Many of the things theorists say about disposi-
tions 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 con-
cerns.
Thirdly, dispositional pluralism facilitates rather than restricts discussion
among dispositional theorists. Often, those with sparse theories of disposi-
tions express their willingness to yield to science the last word about funda-
mental 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
weve 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 consider-
ing, 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: McGill-
Queens 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.
203
In WSPI 2006: Contributions to the Third International Workshop on Philosophy and Informa-
tics, 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. 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
Quine, V. W. O. 1969. Natural Kinds. In Essays in Honor of Carl G. Hempel, ed. N. Re-
scher, 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 Mar-
tin, Ullin Place and George Molnar most notably argued that the modal character
of dispositions should be understood in terms of their intentionality. Other Dispo-
sitional 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 in-
dividuation conditions. I conclude that the attribution of intentions is a concep-
tual 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 Real-
ists, 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., spatio-
temporal) world. Other Realists, deny the existence of dispositional proper-
ties, 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 tech-
nical vocabulary; it is thereby convenient to start off by introducing two dis-
tinctions that we shall employ during the discussion.
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 prop-
erty. 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 Hoffmans 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 ap-
plied 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 spar-
se and pure is to devise a criterion (or: some criteria) through which proper-
ties 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: What makes a property the property that it is?
INQ: 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 evi-
dence that is necessary to single out a property.
_____________
1 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.
2 See Armstrong 1979, Lewis 1986, 59-69, Swoyer 1996, Mellor and Oliver 1997, 1-33, and Shaffer
2004.
3 See Khamara 1988, Humberstone 1996, Langton 1998 and Langton and Lewis 1998.
Andrea Borghini
206
In what follows, I shall first address IDQ and then INQ. My main argu-
ment will purport to show that the attribution of intentionality to disposi-
tional 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 regard-
less its environment.
6
Relational aspect: a relation that each instance of a property entertains with
instances of other properties.
7
_____________
4 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.
5 See, for example, Robinson 1993 and Shaffer 2005.
6 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.
7 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
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 hydro-
gen 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 par-
tially or completely change its qualitative character and still retain its identity.
These postulate the existence of non-causally efficacious aspects, called quiddi-
ties, 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 in-
stance of a property possesses it, it is manifest at all times. For example, in-
stances 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 ex-
ample, 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:

_____________
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.
8 See Armstrong 1989, Lewis 200+, and Shaffer 2005.
9 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.
Andrea Borghini
208
Categorical aspect: an aspect such that, when included in the qualita-
tive character of an instance of a property, is mani-
fest at all times.
Dispositional aspect: an aspect such that, when included in the qualita-
tive 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? The-
re 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: a property whose qualitative character includes
some and only categorical aspects.
Dispositional property 1: 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 Martins contribu-
tion to Armstrong et al. 1996.
10

Categorical property 2: a property whose qualitative character includes
some categorical aspects.
Dispositional property 2: 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 fur-
ther 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 clai-
med that modal and non-modal entities are identical.
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 dis-
tributed among properties in a mutually exclusive way.
Now, to clarify the distinction, an advocate of the second way might in-
troduce 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*: a property whose qualitative character includes
some essential aspects and these are all categorical.
Dispositional property 2*: a property whose qualitative character includes
some essential aspects and these are all disposi-
tional.

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 as-
pects 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 disposi-
tional property the property that it is?
Andrea Borghini
210
Presumably
11
, the identity of a categorical property will be fixed by its as-
pects 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 prop-
erties is smaller. Talk of dispositions is suspicious. Dispositions are but prom-
ises; 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 be-
lieved that David Lewiss 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 de-
fended 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 enti-
ties, fictional entities, and what else you may have to add to the list.
_____________
11 For some, the identity of a property will be entirely fixed by the quiddity, no matter what its
qualitative character be.
12 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, Harr 1970 and
Fara 2005.
13 See Shalkowski 1994, Melia 2003, Divers 2002, Cameron 2008, Denby 200+ and Borghini 2007.
211
On the other hand, some believe that not all apparently modal proposi-
tions 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 mo-
dalist or endorsing a dispositional theory of possibility a recently advanced theory,
according to which all modal sentences are interpreted as attributions of pri-
mitive dispositions (see Borghini and Williams 2008). The other options are
not open. Indeed, the Dispositional Realist claims that (at least some) disposi-
tions 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 enti-
ties, 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 a-
mong the primitive modal entities on account that you cannot apparently
reduce dispositional talk to talk of categorical entities. But, for the probabilis-
tic, mathematical, or counterfactual propositions we might find a reduction.
Indeed, that is what the dispositional theory of possibility claims: that all mo-
dal 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. Disposi-
tions are properties; hence, metaphysical simplicity suggests seeking for an
explanation according to which all modal primitives are properties. The mo-
dalist, on the other hand, will typically endorse a view according to which
modal primitives are facts or situations or states of affairs rather than proper-
ties.
Whether the Dispositional Realist be a modalist or a dispositionalist, the fol-
lowing moral can be drawn for present purposes: if you accept that disposi-
tions 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, ac-
cording to which dispositional and categorical properties are identical, falls
Andrea Borghini
212
short of an adequate explanation of the nature of possibility.
14
This is a re-
mark 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, disposi-
tions are a kind of modal entities; and many dispositions are never mani-
fested; 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 Dis-
positional Realist cannot attempt to reduce dispositional talk to talk of re-
combinations, 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 ob-
tain; 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 onto-
logical category namely, properties; yet, it is left open to what sort of indi-
viduals (if any) they are ascribed. Thus, you may find dispositions ascribed to
entities in the natural world as well as to mathematical entities, social institu-
tions, 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 hid-
den 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, (ho-
pefully) 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.
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 analy-
sis 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, Places 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 under-
stood 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 an-
swer 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 achie-
ved.
At first, let us briefly consider categorical entities. Presumably, those that
inhabit the natural world will be individuated in terms of the way they mani-
fest 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 experi-
ences or a priori.
Andrea Borghini
214
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 con-
ceptual experiences in a subject. Being red will bring about, under certain stan-
dard 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 Different, but related arguments are offered also in Franklin 1988 and Blackburn 1990.
16 For this view, see Ducasse 1942, Chisholm 1957, Sellars 1967, Tye 1984, and Thomas 2003.
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 inten-
tion 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 disposi-
tions reject the so-called Brentano thesis, according to which intentionality is
both necessary and sufficient for telling apart the psychological and the non-
psychological. (For a detailed discussion, see Molnar 2003, 60-81.)
I believe that Martin and Pfeifer, Place, and Molnars aim is noble. Hav-
ing 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, dispo-
sitions are no more mysterious than re-combinations, ersatz linguistic entities
or worlds other than our own. Each of those categories is supposed to pos-
sess 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 dif-
ficulty 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 prop-
erty 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 lim-
ited 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.
Andrea Borghini
216
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 re-
garded 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 exam-
ple, 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 prop-
erty; 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 ana-
lyzed 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 individu-
ate a disposition.
Invoking the intentional character is a tool to focus the task of individua-
tion on a small number of properties: those embedded in the situation for
which the dispositional property has an intention. Talk of intentions, how-
ever, 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 quali-
tative character of a property; more modestly, it alleviates the impasse, by fo-
cusing 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-
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., disposi-
tional 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 disposi-
tional property.

Literature

Armstrong, D. M. 1989. A Combinatorial Theory of Possibility. Cambridge: Cambridge Univer-
sity 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 virt del concretismo. Annali del Dipartimento di Filosofia
dellUniversit 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. Moores 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: McGill-
Queens University Press.
Fara, M. 2005. Dispositions and Habituals. Nos 39: 43-82.
Field, H. 2004. The Consistency of the Nave 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
Harr, 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
Andrea Borghini
218
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-Queens 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 Philosophi-
cal 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 Ascrip-
tion. 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
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: 1-
34.
Sellars, W. 1967. Science and Metaphysics. London: Routledge and Kegan Paul.
Shaffer, J. 2004. Two Conceptions of Sparse Properties. The Philosophical Quarterly 85: 92-
102.
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 Perspec-
tives 10: 243-264.
Thomas, A. 2003. An Adverbial Theory of Consciousness. Phenomenology and the Cognitive
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 Physics
1
ANDREAS HTTEMANN
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 quali-
ties, 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 prop-
erties. Descartes rejected them because he took them to be explanatorily use-
less.
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 contribu-
tors. What I will claim is that dispositions as contributors are presupposed in
physics precisely because of the competing explanatory strategy that Des-
cartes 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 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.
2 Descartes 1991, 107 (AT II, 200).
3 Newton, for instance, writes: since the moderns rejecting substantial forms and occult
qualities have undertaken to reduce the phenomena of nature to mathematical laws (New-
ton 1999, 381).
Andreas Httemann
224
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 prop-
erties in physics. Newton in the third part of his Principia argues that certain
well-known phenomena among them Keplers laws can be deduced from
his theory. Proposition 13 of Book III of the Principia captures Keplers 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 in-
versely 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 remain-
ing planets did not act upon one another, their orbits would be elliptical having the sun in their com-
mon 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 exter-
nal 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 plan-
ets. For our purposes, however, it is not important whether or not the exter-
nal influences are small. The important point is that certain behavior is attrib-
uted 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.
Newtons first law is another and a much simpler example for the attribu-
tion 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, Newtons first law attributes a
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 Newtons 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 Schrdinger equation with a Cou-
lomb 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 disposi-
tional property is a property that, if instantiated by an object, is manifest un-
der specific conditions only. A categorical property by contrast is a property
that, if instantiated by an object, is manifest under all conditions. So, accord-
ing 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 doesnt 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 sys-
tems 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 categori-
cal properties to physical systems. Here is the problem: If laws attribute cate-
gorical properties to physical systems, they attribute properties that, if instan-
tiated i.e. if the system in question has the property , are manifest under all
Andreas Httemann
226
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 ques-
tion. There simply are no bodies that are not compelled to change their state
due to forces impressed on them. So Newtons 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 Newtons 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
Schrdinger equation with the coulomb-potential?
In fact we have three problems here: Laws often appear to have no in-
stances. Thus:

How can laws that attribute properties to physical systems, which appear
to have no instances, be tested?
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? Whats 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? (Cart-
wright 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, Galileos spokesman, ar-
gued 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 me-
dium 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, cop-
per, 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
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 differ-
ent 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 ex-
perimental 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 pro-
cedure 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 dis-
appears. 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 Saliviatis argument is to work, the ideal (the counterfac-
tual) 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. Salviatis 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.
Andreas Httemann
228
Something carries over from the ideal to the less than ideal.
5
It would be pre-
sent 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 proper-
ties 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, Salviatis 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 couldnt be the case if manifestation
were an all or nothing affair. The point is that the disposition is partially mani-
fest in the less than ideal situation and this provides the basis for an extrapola-
tion to the ideal situation.
In this context it is helpful to introduce a distinction I have drawn else-
where: Fragility is an example of a discontinuously manifestable disposition
(DMD). A thing is either broken or it is not; fragility cannot be partially mani-
fest. It is an all or nothing affair.
6
Not all dispositions are discontinuously
manifest. Continuously manifest dispositions (CMDs) allow for partial mani-
festations. If the partial manifestations are continuously ordered they allow
for extrapolation.
Take, for example, properties that we attribute to ideal crystals: conduc-
tivity, 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 dis-
positions 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 par-
tially 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 This is essentially Cartwrights argument for capacities (Cartwright 1989, 189).
6 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.
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 situa-
tion under the assumption of continuity between partial and complete mani-
festation.
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 systems 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 dis-
turbing factors are parts as well. These contributions can be made quantita-
tively precise.
Let me illustrate this through a simple example. Take a compound mas-
sive 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:

m
1
+ m
2
+ m
3
= 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 ten-
sors. 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 limit-
ing case of, say, m
1
, 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 be-
tween the less than ideal and the ideal.

7 More on laws of composition in Httemann 2004, chapter 3.
Andreas Httemann
230
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 ac-
count 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:

to have no instances, be tested?

The problem disappears if it is assumed that law statements attribute continu-
ally manifestable dispositional properties to physical systems. The properties
in question are instantiated in the less than ideal situations though not (com-
pletely) 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.

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 vari-
ous 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 sys-
tems, which appear to have no instances? Whats the point of postulat-
ing 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 sys-
tems because that allows us to explain the behavior of compound systems in
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 ex-
plain 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 quick-
silver, 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 hap-
pens.
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 ques-
tion, 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 nave
regularity account of laws could achieve that. The problem with a nave regu-
larity 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:

On the categoricists view it remains unclear why the cases that Salviati
mentions have anything in common. What is it that connects the differ-
ent laws? Why can Salviati legitimately infer from the observed phenom-
ena 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

8 Armstrong 1983, 39-59.
Andreas Httemann
232
ideal crystals etc. (Thus we try to shield off external influences in ex-
periments.) On the categoricists view it remains unclear why we are in-
terested in these ideal cases.

These problems stay with the categoricist even if higher order laws are intro-
duced.
9
These are meant to systematize the empirical laws I have mentioned.
We can think of Newtons second law or the Schrdinger-equation as higher
order laws that allow for lower-level laws with missing values. Thus Newtons
second law tells us (given the right force-functions) how bodies gravitate to-
wards one another and thus how bodies fall. Such higher order laws would
allow for missing values. Newtons 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 com-
pounds behavior?
The categoricists or Humeans may be able to give a complete and consis-
tent 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 as-
sumption that physical systems have CMDs is the best explanation for the
success of part-whole-explanations (micro-explanations) and for extrapola-
tions to ideal cases.
Second, we have to look at the role of dispositions within such explana-
tions. 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 usu-
ally 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 Newtons first law men-

9 Armstrong 1983, 111-116.
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 underly-
ing 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 Cart-
wrights capacities.)
In an explanation, in which we appeal to CMDs, the explanandum is usu-
ally 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 behav-
ior 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 arent the dispositions causally relevant for the manifestations of the
compounds behavior? Clearly the dispositions are relevant for the com-
pounds behavior, but they are not causally relevant. The masses of the parts
contribute to the mass of the compound they dont 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
compounds behavior is thus counterfactually dependent on the behavior of
the parts. This, however, does not imply that the relation is causal. If one as-
sumes 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. There-
fore, 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 Httemann 2004, 83ff. I argue in more detail why the part whole relation is non-causal.
Andreas Httemann
234
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 instanti-
ated 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 defini-
tions those entities, which play the primary causal role in the transmission and repro-
duction 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 Mumfords terminology we have the follow-
ing 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 micro-
description 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 re-
quire 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. Fragilitys functional
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 real-
izes 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 func-
tional role, must manifest the behavior in question under the very same circum-
stances. The micro-based property can be an occupant of the dispositions
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 re-
duction 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 cate-
goricist.
Second Objection: Couldnt 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 re-
ducing micro-structural property has to be a dispositional property.
The standard examples for dispositions in the relevant literature are solu-
bility 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 com-
pound system it might be reasonable to expect a micro-explanation (micro-
reduction). 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.
Andreas Httemann
236
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.
Newtons 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 funda-
mental.
It has been argued that irreducible fundamental dispositions are incoher-
ent. 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. Newtons first law, for
example, is true of systems in our world because these systems have the rele-
vant disposition in our world. Furthermore we have epistemic access to the
non-manifest dispositions due to their partial manifestation. If the disposi-
tions in question are continually manifestable dispositions the problem Black-
burn 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 scien-
tific practice that has been established since the 17
th
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.
237
Blackburn, Simon. 1990. Filling in Space. Analysis 50: 62-65.
Cartwright, Nancy. 1989. Natures 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.
Httemann, Andreas. 2004. Whats 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 uncer-
tain if the object in question has acquired the underlying dispositional prop-
erty 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 perti-
nent 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 disposi-
tions 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, Chamberlains politics in light of his
weak character (compared with Churchills). 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.
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 salts solubility, this seems
rather odd. For the question of why the salt dissolves in water is not suffi-
ciently answered by pointing to salts water-solubility. The desire for explana-
tion will come to an end only if it is explained why salt is water-soluble. The
same would hold for an explanation of Chamberlains politics in terms of his
weak character: Chamberlains politics was to be expected, given his charac-
ter 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 Caesars clemency. The cruel decisions of Philip II
of Spain are explained by his cruel temperament. Well known is also Max
Webers analysis of the capitalists spirit in order to explain why people be-
have in a capitalistic manner. Some historians explain the cruelty of German
soldiers in the World War II in terms of a murderers 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 Caesars 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 Caesars clemency and of Philips cruel decisions in terms of his cruelty
demands for further explanation. Some might challenge the above explana-
tion of Caesars acts by arguing that they were part of a calculated political
strategy while others accepting the explanation might try to further ex-
plain Caesars clemency pointing to his upbringing.
John W. N. Watkins maintained that the ultimate premises of social sci-
ence 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 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.
3 J. W. N. Watkins 1952, 32-33, citing L. Robins 1935, 79.
Robert Schnepf
240
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 indispen-
sable):
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 rea-
sonable doubt; different interpretations of the same occurrences are
always possible.
There seems to be a difference between explaining the clementia cae-
saris (his concrete acts against the prisoners of war) by referring to
Caesars 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 explanandums 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 his-
torical 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 dis-
positions assumed to be more basic.
4
Dispositions may play a crucial role in searching and forming historical expla-
nations without being a necessary part of the ultimate explanation reached at
the end of the scientists efforts. Dispositions can do this job if they are in-
formative and not repetitive. Informative dispositional explanations in histo-
riography are the first step to discover non-dispositional causes which are
beyond the scope of the explanandums 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 realis-
tic 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 epistemo-
logical 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.
241
sis, nomological knowledge
5
and possible worlds semantics by Kaila and oth-
ers.
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 reli-
gious 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 under-
stood 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 expla-
nations 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 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 Webers conception of
nomological knowledge and this might be sufficient for the context of historiography, not-
withstanding the extensive debate about the nature of natural laws.
6 Kamlah 2002, 271 et seqq.
7 Graus 1987, Gilcher-Holthey 1998 for further discussion Schnepf (forthcoming 2010).
8 In fact, even the ascription of mentalities to our contemporaries, even to our neighbors and
friends, poses these grave methodological problems.
Robert Schnepf
242
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 com-
bine 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 20
th
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 expres-
sion 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 dure),
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 concep-
tion 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 Cae-
sar 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 G. W. F. Hegel 1970, 552.
10 Cf. Braudel 1992a and 1992b.
11 M. Weber 1988 for a discussion of Webers approach compare Lehmann/Roth 1987.
12 Cf. M. Weber, WL p. 428.
243
Consequently, I will focus on Webers and Sombarts explanatory use of
concepts of Geist (spirit, mentality); in contrast to Lujo Brentanos explana-
tions of the same historical phenomena.
13
For Sombart, the spirit of capital-
ism 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
Webers explanation of the rise of mod-
ern capitalism by reference to the protestant spirit will be analyzed in opposi-
tion to Sombarts and Brentanos approaches. My analysis of this debate will
support the thesis that the ascription of dispositions in historiography is noth-
ing beyond reasonable doubt and that it requires careful methodological con-
siderations. In what follows I will restrict my analysis of dispositions in histo-
riography to this example. To illustrate my points I will particularly make use
of Max Webers methodological reflections.
In this context another assumption should be mentioned. Max Weber is
committed to methodological individualism. This is evident in Max Webers
essay on the categories of sociology (ber einige Kategorien der verste-
henden Soziologie)
15
and in Economy and Society (Wirtschaft und Gesell-
schaft).
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, Sombarts expli-
cation 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 Cf. L. Brentano 1923.
14 Cf. Sombart 1916, 2 et seq.
15 Cf. M. Weber 1985, 439: Das Ziel der Betrachtung: Verstehen, ist schlielich 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 er-
laubt ist behandelt. () Begriffe wie Staat, Genossenschaft, Feudalismus und hnliche
bezeichnen fr die Soziologie, allgemein gesagt, Kategorien fr bestimmte Arten menschlichen
Zusammenhandelns, und es ist also ihre Aufgabe, sie auf verstndliches Handeln, und das
heit ausnahmslos: auf Handeln der beteiligten Einzelmenschen, zu reduzieren.
16 Cf. M. Weber 1976, 6: Handeln im Sinn sinnhaft verstndlicher Orientierung des eigenen
Verhaltens gibt es fr uns nur als Verhalten von einer oder mehreren einzelnen Personen.
17 J. W. N. Watkins 1957, 105 et seq.; methodological individualism is challenged, for example, by
Dray 1990, 47 et seqq.
Robert Schnepf
244
plied.
18
At first sight, methodological individualism seems to minimize onto-
logical commitments. We seem to have epistemic access to the intentions of
individual actors, whereas we are not willing to attribute intentions to col-
lectives, 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 methodo-
logical 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 de-
bates between Max Weber, Lujo Brentano, and Werner Sombart that attri-
butions 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.
Lets start with a closer look at Webers famous Protestantism thesis: The
object of explanation is the rise and growth of capitalism. Given methodo-
logical 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 cor-
respond to the needs of these new kinds of behavior. According to methodo-
logical 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 W. Sombart 1916 I, 21 et seq. translation by T. Parsons.
19 In fact, intentions, volitions, and even the will itself can be interpreted as dispositions and con-
sequently as mere theoretical terms. Webers methodological approach to explain an individuals
behavior using ideal types provides some support for my thesis. To solve epistemological prob-
lems, 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 argu-
ments of Martin 1977 against his intentions.
20 Cf. M. Weber 1988, 4 et seqq.
245
For Weber, this change in individual behavior can at least partly be ex-
plained by a change in an ethics of economic behavior. The rise of Calvinism
changed an agents 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 reli-
gious convictions (which might by itself be regarded as a special kind of dis-
position). 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 gener-
alized: It is not Webers aim to explain the whole of modern capitalism but
only a very special form of it at its very beginning. Moreover, the differentia-
tion between private household and office, new ways of banking and techni-
cal developments are necessary conditions for the rise of capitalism. Weber
does not think that a Calvinists ethics of economic behavior is the only rele-
vant 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. Neverthe-
less, Weber tries to reconstruct one side of a complex causal history by de-
scribing or reconstructing a very special disposition that accounts for the self-
rationalization necessary for a style of life appropriate to early modern capital-
ism.
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 capi-
talistic 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 circum-
stances 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 Cf. M. Weber 1988, 12: Denn wie von rationaler Technik und rationalem Recht, so ist der
konomische Rationalismus in seiner Entstehung auch von der Fhigkeit und Disposition der
Menschen zu bestimmten Arten praktisch-rationaler Lebensfhrung berhaupt abhngig.
22 Cf. L. Brentano, Puritanismus und Kapitalismus, in: Brentano 1923, 363-425.
Robert Schnepf
246
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 didnt allow charging inter-
est 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 oc-
curred. 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 persons aim consisted in
receiving only the support necessary for retaining his social status (Be-
darfsdeckungsprinzip). 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 differ-
ence 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 sys-
tems in light of a fundamental change in an individuals basic dispositions;
from a disposition to conserve ones own status to a disposition to act ac-
cording 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 So the basic disposition assumed by Brentano can be interpreted as finkish as most disposi-
tions attributed by historians.
24 Cf. Sombart 1916, 11 et sqq.
247
economic behavior and by denying a constant disposition to unlimited acqui-
sition. 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 expla-
nation entails a change of dispositions possessed by individual persons while
Brentano does not assume a change in disposition but only in the circum-
stances 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 be-
yond reasonable epistemic doubt and that appealing to dispositions in histori-
cal 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 interpreta-
tion 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 historio-
graphy.
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 methodo-
logical self-control demanded by a Calvinist ethics fits with a capitalists 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 We-
bers research. Weber deals with the problem of explaining the behavior only
on the level of so called ideal types. Yet Webers remarks on ideal types are
not at all clear. One forms the concept of an ideal type so he seems to as-
sume 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 ac-
cording to the features combined in this concept. On the contrary, observing
the deviance from the ideal type opens up the possibility for a closer descrip-
tion 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, Webers explanation of the origin of capitalism in terms of an
Robert Schnepf
248
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 incli-
nation 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 Sombarts conception of the change of disposi-
tion 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 actu-
ally 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 Webers interpretation of Franklins writ-
ings but also their impact on his life. This does not necessarily falsify Webers
attribution of a corresponding disposition to Franklin. Franklin may have
possessed the disposition in question but may not have actualized it for addi-
tional 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 individu-
alism. 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
Webers 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.
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 individuals
mind. For example, if we accept Webers approach and start to analyze Ben-
jamin 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 his-
tory 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 epis-
temic 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 fur-
ther causal factors. So we have to ask what type of arguments is used to jus-
tify a certain general approach (for example Webers or Sombarts or Bren-
tanos) 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 dif-
ferent laws or law-like sentences and different causal factors. For if we
attribute to Franklin the disposition of a Calvinists mentality one has to ex-
pect all those manifestations of this disposition, which one associates with it
by nomological knowledge attained by the observation of more pure in-
stantiations of this mentality. Franklins 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. Webers nomological knowledge may play a
methodological role but does in no way solve the problem of ascribing dispo-
sitions in the historians work.
Robert Schnepf
250
4.
For Brentano, Sombart, and Weber, disposition terms are somehow theoreti-
cal 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 de-
scription nor explanation are possible.
26
Dispositions like characters or the
capitalists spirit are not given phenomena but constructed theoretical con-
cepts in order to explain certain regularities to be observed within the cha-
otic 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 expla-
nations are therefore mere hypotheses in need of additional confirmation.
They cannot be directly confirmed by the given data since all three explana-
tions cover the same set of phenomena. Each approach is able to reject
counter-examples by claiming exceptions or further specific circumstances.
Lets 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 dis-
positions are constituents of human nature. All human beings tend to acquire
goods. Brentano argues therefore on Aristotelian grounds. Against Sombarts
theory of a change of basic dispositions Brentano argues with Aristotle and
his distinction between an economy concerning ones own house and an
_____________
26 Cf. M. Weber 1985, 177. Stephen Mumford objected to my thesis in a discussion that this ap-
proach 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 cha-
otic 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.
251
economy concerning further trade. Assuming basic and unchangeable disposi-
tions, 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 con-
ditions 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 be-
tween 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.
Sombarts 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 sub-
sumes 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 Sombarts 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 Brentanos dispositional explanations. For his dispositional con-
cepts 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.
Robert Schnepf
252
tive. Explaining a change of behavior by describing it as a change of disposi-
tions 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
historians 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 expla-
nation stopper.
I take Webers account of ideal types as one of the most promising theo-
ries of dispositional explanations in historiography. What is it to ascribe a
disposition to an individual person? According to Weber in attributing a dis-
position e.g. a character or spirit one attributes non-dispositional categori-
cal properties and postulates law-like connections between having those cate-
gorical 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 condi-
tions 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 observ-
able datum (provable by checking the registers of the church). The law-like
connection can however not be verified by observable regularities. Most Cal-
vinists may behave in a way that does not correspond to the behavior com-
monly associated with the attributed disposition. To explain their actual be-
havior 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 Cal-
vinist, derived from Calvinist theology, it is necessary to control ones 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 im-
ply therefore counterfactual conditionals (what he calls objektive Mglich-
keit, 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 persons being of this type. Sec-
ond, 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 per-
son 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-
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 persons 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 disposi-
tional concept is constructed from two sides: It is a theoretical term con-
structed as implying law-like sentences that can be used in deductive explana-
tions. 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 explan-
andum 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 ex-
pected 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 proper-
ties, 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 dif-
ferent type for example connecting the economical situation of the family
with their religious attitudes.
5.
I do not want to defend all aspects of Webers 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 ob-
servable non-dispositional properties, if we want to justify their as-
cription. There is no need for and no use of interpreting them in a
Robert Schnepf
254
realistic manner, since there is no epistemic access to the real
properties and mental states of individuals and dispositional explana-
tions can only be defended in reference to observable facts. To use
one of my other examples: Caesars way of dealing with his prisoners
of war can be explained by his clemency but this explanation cannot
be defended by reference to Caesars disposition. This would be cir-
cular: Explaining Caesars acts in terms of a disposition can only be
defended by these very observable acts by Caesar that is by the ex-
planandum. No methodological problem of a historians 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 disposi-
tion 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 dis-
position 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
Caesars 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 disposi-
tion) 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 bene-
factor we would attribute to him a completely different way of think-
ing 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.
255
Literature
Braudel, Fernand. 1992a. Geschichte und Sozialwissenschaften. Die lange Dauer. In F.
Braudel. 1992. Schriften zur Geschichte 1, Gesellschaft und Zeitstrukturen. Stuttgart:
KlettCotta: 49-87.
Braudel, Fernand. 1992b. Geschichte und Soziologie. In F. Braudel. 1992. Schriften zur
Geschichte 1, Gesellschaft und Zeitstrukturen. Stuttgart: KlettCotta: 99-121.
Brentano, Lujo. 1923. Der wirtschaftende Mensch. Gesammelte Rede und Aufstze. 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 Einfhrung. 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. Pldoyer fr eine dynamische Mentalittsgeschichte. Ge-
schichte und Gesellschaft 24: 476-497.
Graus, Frantiek. 1987. Mentalitt Versuch einer Begriffsbestimmung und Methoden
der Untersuchung. In F. Graus, ed. Mentalitten im Mittelalter. Methodische und inhaltli-
che Probleme. Sigmaringen: Jan Thorbecke Verlag: 9-48.
Hegel, Georg Friedrich Wilhelm. 1970. Vorlesungen ber 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 kriti-
scher Bericht ber 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 Webers Comparative Historical Sociology. Chicago: The University
of Chicago Press.
Lehmann, Hartmut, and Roth, Guenther, eds. 1987. Webers Protestant Ethics. Origins, Evi-
dence, Contexts. Cambridge: Cambridge University Press.
Lewis, David. 1979. Counterfactual Dependence and Times 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 Webers 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 Untersu-
chungen zum Kausalitts- und zum Schpfungsbegriff. Gttingen: Vandenhoeck & Rup-
recht.
Schnepf, Robert. 2009. Von der Seele zum Geist: Diltheys Grundlegung der Geisteswis-
senschaften. In K. Crone, R. Schnepf, and J. Stolzenberg, eds. ber die Seele. Frank-
furt a.M.: Suhrkamp (forthcoming 2009).
Robert Schnepf
256
Schnepf, Robert. 2010. Geschichte erklren. Gttingen: Vandenhoeck & Ruprecht (forthco-
ming 2010).
Scriven, Michael. 1966. Causes, Connections and Conditions in History. In William H.
Dray, ed. Philosophical Analysis and History. New York/London: Harper & Row Pub-
lisher: 238-264.
Sombart, Werner. 1916. Der moderne Kapitalismus. 2
nd
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. Tbingen: Mohr.
Weber, Max. 1985. Gesammelte Aufstze zur Wissenschaftslehre. Tbingen: J.C.B. Mohr.
Weber, Max. 1988. Gesammelte Aufstze zur Religionssoziologie I. Tbingen: 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 ana-
lytic 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 coun-
terfactual 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 disposi-
tional properties within the empiricist framework taking such generalized
perspective is certainly appropriate. Yet from this viewpoint, specific distinc-
tions 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 re-
ducible, 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 par-
ticular domain of higher order dispositions that is delineated by our folk psy-
chological vocabulary. As I will show, the general discussion about disposi-
tions 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 McKitricks contribution to the anthology in this regard.
Karsten Stueber
258
objectivity within the physical sciences. The domain of higher order disposi-
tions referred to within the context of folk psychology is however signifi-
cantly 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 rea-
son, 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 com-
monly associated with characteristic manifestation events in a variety of cir-
cumstances. Yet in contrast to ascriptions of physical dispositions such as
fragility, the ascription of mental dispositions depends essentially on our em-
pathic ability, that is, our ability for mental imitation and our capacity to re-
enact another persons 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 persons 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 psycho-
logical 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 con-
text 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 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 introduc-
tion of my 2006.
3 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 unob-
servable 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 Ryles 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 men-
tal 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.
Ryles 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 dis-
solve 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 Princi-
ple, 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. Ascrib-
ing 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 re-
garding the furniture of the universe are decided not by philosophical arm-
chair considerations but in terms of whether or not a property has to be ap-
pealed 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 Ryle 1949, 125.
5 See the contribution by Scholz in this anthology and Mumford 1998, 22ff.
6 In this respect see also Heil 2003, chaps. 3 and 9.
Karsten Stueber
260
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 theo-
ries and our understanding of the historical progress of science.
For this very reason, Carl Hempels 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 occur-
rence of a certain event, given the initial conditions. Dispositional explana-
tions accomplish this task by telling us that a certain object has the tendency
to manifest a certain type of behavior in certain circumstances. The short-
comings of Hempels conception of explanation are well known.
8
Yet inde-
pendent of these well known concerns, for our topic more important is the
fact that nothing in Hempels 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 de-
pendent 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 Hempel 1965, 462.
8 See Salmon 1989.
9 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.
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 com-
mitment seems to entail a corresponding epistemic postulate that it is in prin-
ciple 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 proper-
ties gives rise to higher order properties. Accepting brute correlations or iden-
tities 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 presup-
position of a reductive scientific practice. It would seem to undermine the
philosophical justification of the physicalist framework and reveal its meta-
physics 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 effi-
cacy of the micro-physical properties they depend on. If higher order proper-
ties 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 be-
liefs 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 Jackson
10
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 knowl-
edge 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 furni-
ture 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.
Karsten Stueber
262
ample, of the fragility of various objects. In reality, however, no natural prop-
erty 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 con-
temporary 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 suc-
cess of the various arguments to save dispositions from ontological extinc-
tion. For my purposes, it is more important to point out that the above strate-
gies 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 ontologi-
cally 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 un-
derstood without concepts like beliefs and desires. Certainly no notion cur-
rently used in the physical sciences seems to come close to play that role. It is
_____________
11 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 dis-
cussion and more precise articulation of this argument.
12 McKittrick 2003.
13 Heil 2003.
14 Mumford, this anthology.
15 See for example Htttemann in this anthology.
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 epis-
temic 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 psychologi-
cal perspective are concerned, even if one admits the ontological priority of
the physical. I would like to suggest that the ontological and epistemic com-
mitments 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 accessi-
ble 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 organ-
isms with a specific physical and psychological constitution. For other crea-
tures 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 persons
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 every-
thing 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-
Karsten Stueber
264
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 psychologi-
cal framework is adequate for a rigorous scientific investigation of human
behavior. Scientific behaviorists like Watson and Skinner adopted eliminativ-
ist 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 psy-
chology have again become prominent in the context of connectionist ac-
counts of mental capacities and have been forcefully articulated by Paul
Churchland.
17

One has therefore assumed that folk psychology, like the physical sci-
ences, has to be understood as an epistemic enterprise that is ideally adopted
from the perspective of nowhere, to use Nagels 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 shortcom-
ings. 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 psycho-
logical 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 embod-
ied in the physical world.
I have called the philosophical conception of folk psychology that con-
ceives of folk psychology as being committed to or adopted from such a
detached perspective the detached conception of folk psychology. It is a con-
ception of folk psychology that has dominated philosophy of mind in the 20
th

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 See however the skeptical voice of Stephen Stich 1983 in this context.
17 For example Churchland 1989.
18 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 tradi-
tions have to be mentioned in this context. For a brief survey see the introduction of my 2006
and the first sections of my 2008a.
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 pri-
mary 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 persons be-
havior. 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 be-
havior. 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 prac-
tice 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 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).
20 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 how-
ever 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 dis-
cussion see DeWaal 2006.
Karsten Stueber
266
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 con-
sisting 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 ac-
quire 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 Carte-
sians suggest or theoretical principles, as theory theorists suggest. Yet the
assumption that we encounter the external world in a perceptually one-
dimensional 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 persons 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 neu-
roscientists like to express it, there exists a significant overlap in the activation
of neurons underlying the execution and the observation of bodily move-
ments 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 Lippss 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 beings movement as a
goal directed movement (as being directed towards the cup) and they allow us
to read another persons 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 in-
fants already understand other minds and other persons in a conceptually
sophisticated manner. Rather one should conceive of the level of basic empa-
thy 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
267
with perceptual similarity spaces that make the development of the concep-
tual 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 bod-
ies 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 reso-
nates with me.
21

The above observations do not imply that mental states do not ontologi-
cally 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 physi-
calism. 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 Jack-
son 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 informa-
tion about a persons brain described from the detached third person perspec-
tive 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 mus-
cles but by activating mechanisms of basic empathy. Asking us to conceive of
_____________
21 For my philosophical interpretation of mirror neurons in regard to the debate between simula-
tion 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.
22 Chalmers 1997 and Chalmers and Jackson 2001.
Karsten Stueber
268
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 perspec-
tive 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 accessi-
ble to all creatures capable of comprehending the physical sciences, even if
they lack a human beings psychological capacity of empathy.
23

Nevertheless, I would admit that not all of our folk psychological con-
cepts are closely tied to mirror neurons and considerations about mirror neu-
rons 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 It seems to me that Wittgenstein expresses a very similar sentiment in the following aphorism:
But cant 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 peo-
ple with fixed looks (as in trance) going about their business the idea is perhaps a little un-
canny. 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 mean-
ingless; or you will produce in yourself some kind of uncanny feeling, or something of the sort.
(Wittgenstein, Philosophical Investigations 420)
24 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.
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 sys-
tem of mirror neurons could not learn to use most of our folk psychological
system for the explanation of human agency, particularly insofar as explana-
tions 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 ar-
gument for this claim, however, does not depend on knowledge of neurobio-
logical 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
agents reasons for acting. Attributions of mental states to an agent are re-
garded 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 illus-
trated 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 sud-
denly 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 uni-
vocality 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.
Karsten Stueber
270
normally provides an explanation for that persons 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 accord-
ing to the theory theory position are at the foundation of all of our explana-
tory 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 under-
stand why such beliefs and desires could be somebodys reasons for acting in
this particular manner. They could be his reasons because I can easily under-
stand 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 analo-
gously 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 agents other beliefs and desires. Thoughts can be understood as reasons
for acting only in the context of other relevant thoughts. Normally we pre-
suppose (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 rea-
sons for acting very differently. This does not mean that we never will be able
to understand how those beliefs and desires could be a persons 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. Neverthe-
less, 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.
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 perspec-
tive. 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 reenac-
tive empathy. Most importantly, such reenactment of another persons
thoughts as his reasons in my own mind, or what I also call the use of reenac-
tive 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 agents 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 ones 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 rele-
vant 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 un-
derstanding 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
agents 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 re-
garding which belief attribution and so on is more plausible given the social
context and the personal biography of the agents. Yet empathy is always in-
volved 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 es-
sential indexicality of thoughts as reasons. See Stueber 2006, 152 ff.
Karsten Stueber
272
taken seriously because only in this manner can we understand an agents
thoughts as potential reasons for acting.
28

To summarize this section, I take the above considerations to have estab-
lished the following points: Within the context of folk psychology, explana-
tions 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 persons
point of view. Given the epistemic centrality of empathy, particularly reenac-
tive empathy, for folk psychology, our practice of attributing mental disposi-
tions to other agents is very different from the attribution of other properties
within the physical sciences. In contrast to the physical sciences, the full ex-
planatory 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 rea-
son, I am tempted to speak of the doubly dispositional character of the attri-
bution of mental properties to other agents. It attributes a tendency towards
action to the other person in light of a uniquely human disposition of empa-
thy. 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 psy-
chological 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 psy-
chology has implications for conceiving of folk psychology as an autonomous
explanatory practice and for alleviating the eliminativist challenge. Given the
_____________
28 See Stueber 2008b in this regard.
29 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.
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 psy-
chological 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 dis-
positional 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 Davidsons 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 ra-
tional 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 Davidsons 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 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.
31 Davidson 1980, 218.
32 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 irra-
tionality. For a different and contextualist understanding of rationality see Stueber 2006, chap.2.
Karsten Stueber
274
within the Davidsonian framework all the causal work is being done on the
physical level.
33
Given these worries, the physicalist challenge to the ontologi-
cal 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 ad-
vance of the medical sciences.
Yet the physicalist challenge presupposes that folk psychological expla-
nations explain the same type of events as physical explanations and it pre-
supposes that they explain it causally in the same manner as physical explana-
tions. 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 considera-
tions in the second section of this paper imply that at least one of these as-
sumptions 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 agents reasons for acting. Moreover, Kim suggests that folk psycho-
logical 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 In this respect see the various articles (particularly the one by Kim) in Heil and Mele 1993. For
my take on Davidsons anomalous monism see the first sections of my 2005 and 1997.
34 Dray 1957 and Kim 1984 and 1998b.
35 Kim 1984, 319.
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 de-
sires are parameters that we know how to manipulate in order to
change the behavior of various agents. Our educational institu-
tions depend on such knowledge. In the same manner, knowl-
edge 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 agents ability to make sense of his behavior in terms of rea-
sons is an ability that seems to be central to his normal function-
ing 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 ex-
ample 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 psy-
chological 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 accessi-
ble 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 elsewhere
38
that folk psychological practices are
practices that have a slightly different explanandum than the physical move-
ments of agents as described by the physical sciences. They describe the inter-
action 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 See also Woodward 2003 in this respect.
37 See Stueber 2006, 45-46.
38 Stueber 2005.
Karsten Stueber
276
understand his actions as rational responses to the demands of the environ-
ment 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 explana-
tory 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. Its 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 Neurocompu-
tational 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.
277
McKitrick, J. 2003. The Bare Metaphysical Possibility of Bare Dispositions. Philosophy and
Phenomenological Research 66: 349-369
Poland, J. 1994: Physicalism: The Philosophical Foundations. Oxford: Clarendon Press.
Prior, E., Pargetter, R., and Jackson, F. 1982. Three Theses about Dispositions. Ameri-
can 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. Minne-
sota 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 Erklrungen 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. Cam-
bridge, Mass.: MIT Press.
Stueber, K. 2008a. Empathy. The Stanford Encyclopedia of Philosophy (Summer 2008 Edition),
Edward N. Zalta, ed., URL =
<http://plato.stanford.edu/archives/sum2008/entries/empathy/>.
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-That
1
GREGOR DAMSCHEN
1. Introduction
Pel 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 profes-
sional 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 disposi-
tion that enables them to successfully perform this action under certain, suit-
able 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 there-
fore 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, knowl-
edge-that, a form much investigated by philosophers since Plato. Knowledge-
that, 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 episte-
mologists 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 some-
thing is disposed to perform an action under favorable conditions? Is knowl-
edge-how a practical ability? Or is knowledge-how rather a hidden knowl-
edge-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 conver-
sations and valuable comments on earlier drafts of this paper. Further thanks go to Dirk Effertz,
Vittorio Hsle, Robert Schnepf, Dieter Schnecker, Karsten Stber, and audiences in Witten-
berg, 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 semantic-
pragmatic 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 knowl-
edge-that? Is knowledge-how a form of knowledge in its own right, inde-
pendent of knowledge-that, is it subordinate to knowledge-that, or is knowl-
edge-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 semantic-
pragmatic 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 dif-
ferent from the one that Ryle constructed.
3
This argument will show that
knowledge-how is not a species of knowledge-that but rather that knowledge-
that is a species of knowledge-how. More specifically, dispositional knowl-
edge-how is at the core of propositional knowledge-that and accordingly
should be understood to be at the center of epistemology. The general intel-
lectualist 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 knowl-
edge-how are purely dispositional abilities. Since the main point of my paper
will be the proof that every knowledge-that presupposes a certain disposi-
tional 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 Jason Stanley and Timothy Williamson (Stanley/Williamson 2001) have recently maintained,
(taking their cue from Gilbert Ryles 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.
3 Ryle 1945/6 and Ryle 1949, 30-31.
Gregor Damschen
280
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 inti-
mately 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. Peters Cathedral is located, she in fact is saying nothing
more than that she knows that St. Peters is located at such-and-such a location.
And if someone says, he knows, how tall the Eiffel Tower is, he is saying noth-
ing 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 knowl-
edge 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 expres-
sion, 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 defini-
tion 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 ad-
dress the following two questions: first, what is knowledge-how and second,
what is the connection between knowledge-how and knowledge-that?
6
We
_____________
4 Cf. Karttunen 1977.
5 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-and-
such 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 Platos
types of knowledge see Damschen 2003.
6 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
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 semantic-
pragmatic 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 speakers use of a knowing how phrase corresponds to our ca-
nonical 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-and-
such 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 condi-
tions 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 success-
fully. 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 wei, 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 epistemol-
ogy 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 seri-
ously. The above substitution test shows very clearly that, in English too, the
expression knowledge-how can refer to something which is not a knowl-
edge-that.
_____________
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, No 2005, Hetherington 2006, Bengson/Moffett 2007, Liho-
reau 2008, Williams 2008, and Adams 2009. Another approach which remains valuable is that of
Polanyi 1958.
7 Cf. Rumfitt 2003.
Gregor Damschen
282
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 op-
ponents 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 dispo-
sitional knowledge-how, this case clearly shows that to know how one per-
forms 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 Anne-
Sophie 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 con-
siderations 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 success-
fully. 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 Cf. Stanley/Williamson 2001, 416.
9 Cf. No 2005, 282-283.
10 Cf. No 2005, 284-286.
283
1. They are always the knowledge of someone, i. e., they are the knowl-
edge 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 knowledge-
how, 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. Disposi-
tional knowledge can only be realized if certain external conditions are ful-
filled. Pel 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 per-
form 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 suc-
cessfully realize an action. Our analysis thus has shown that there are manifes-
tations 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 epis-
temic 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 such-
and-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. Further-
more, 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

Gregor Damschen
284
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 pro-
positional meaning, e.g. as the answer of a driving instructor to the question
whether he can articulate in propositional form the actions necessary to con-
trol a car, the relevant theoretical knowledge, and the most important rules of
the road. Finally the sentence can thirdly (1c) have both meanings simultane-
ously, i.e. the dispositional and the propositional, e.g. as the answer of a driv-
ing 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 every-
thing 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 accu-
rately observed: It would be feeble, I think, just to say that were fooled by
the ambiguity of the word know: we confuse ability with information be-
cause 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-how
d
, that is dispositional and non-propositional
knowing how;
(b) knowledge-how
p
, that is knowing how that can be articulated in
terms of propositions;
(c) knowledge-how
h
, that is knowing how as hybrid.
It was no doubt the propositional use of knowledge-how
p
or knowledge-how
h
which led philosophers to their general claim that every knowledge-how is a
_____________
11 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-and-
such to drive a car.
(1c) (1) is true iff both (1a) and (1b) are met.
12 Lewis 1999, here p. 289.
285
knowledge-that.
13
For the purpose of this paper it is only important to recog-
nize that the dispositional form of knowledge-how in the sense of a practical
ability (knowledge-how
d
) exists.
14
4. Five Alternative Answers
Asking whether and how knowledge-how in the dispositional sense (know-
lege-how
d
) and knowlege-that in the propositional sense are semantically
related yields the following five alternatives:

(A1) Knowledge-how
d
is a species of knowledge-that, i.e.:
knowledge-that is a necessary condition for knowledge-how
d
.
(A2) Knowledge-that is a species of knowledge-how
d
, i.e.:
knowledge-how
d
is a necessary condition for knowledge-that.
(A3) Knowledge-how
d
and knowledge-that are extensionally equivalent.
(A4) Knowledge-how
d
and knowledge-that have a common intersection.
(A5) Knowledge-how
d
and knowledge-that are completely distinct.

These five alternative ways of thinking about the relation between knowledge-
how 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 pro-
positional knowledge, we must be convinced of having found a good argu-
ment 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 E.g. Hintikka 1975, or Stanley/Williamson 2001.
14 Cf. on this also Ryle 1949, Rumfitt 2003, Rosefeldt 2004, No 2005, 284, Lihoreau 2008, and
Williams 2008.
15 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 al-
ternatives:
(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.
16 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.
Gregor Damschen
286
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 pre-
sents us with a non-uniform definition for the term knowledge encompass-
ing three equally valid and distinct definitions of a form of pure knowledge-
that, a form of pure knowledge-how and a mixed form of the two. (S5) Fi-
nally, 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 hav-
ing 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 Ryles argument against what he calls the
intellectualist legend.
18
As we have seen, Ryles intellectualist legend argu-
ment 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 Ryles original argument.
23
There
have also been a few arguments for saying vide alternative A2 that some-
_____________
17 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 proposi-
tional 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.
18 Ryle 1945/6 and Ryle 1949, 30-31.
19 Stanley/Williamson 2001, 412-417. Whether their reconstruction in fact grasps Ryles original
argument will have to remain untreated here. Cf. on this Rosefeldt 2004 and No 2005.
20 See Stanley/Williamson 2001, 417-444.
21 See Rumfitt 2003.
22 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.
23 Cf. e.g. Carr 1979; Carr 1981; Devitt 1996, 52; Putnam 1996, xvi; Lewis 1999, 288.
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 Ryles 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-how
d
is a species of knowledge-that is untenable;
second, using the first and the last of its premises we can construct a hypo-
thetical syllogism that shows that knowledge-that is a subcategory of disposi-
tional knowledge-how
d
.
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 ear-
lier 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 in-
tentional analysis act U sufficiently often, regularly, and successfully,
whether T is the case, or whether not-T is the case.

_____________
24 Hartland-Swann 1956 and 1957, Roland 1958. For a new argument for assumption A2 see now
also Hetherington 2006.
25 See the critique on Hartland-Swann by Ammerman 1957.
Gregor Damschen
288
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 proposi-
tion T(F) states, that such-and-such must be done to complete act F. Accord-
ingly, knowing how to do F would be nothing but to know that such-and-
such must be done in order to do F. But this would imply that every knowl-
edge-how would be nothing but a kind of knowledge-that.
Premise P1 states a sufficient, not a necessary condition for dispositional
knowledge-how
d
. Even though a comprehensive definition of knowledge-
how
d
continues to be a desideratum,
26
for the purpose of the current argu-
ment 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 non-
personal 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 to-
wards doing something very specific, but a knowledge-how
d
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, em-
bodied 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: Di-
gesting 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 See Hawley 2003 and Williams 2008.
27 Stanley/Williamson 2001, 414.
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 sys-
tem, 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 ac-
tions. In this regard, the funny thing about the proposition If Hannah di-
gests 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 fol-
lows:

(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 knowledge-
how. 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 ac-
tively 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 ex-
ample, 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 Cf. No 2005, 279.
29 Cf. No 2005, 280.
Gregor Damschen
290
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 corre-
sponding 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
someones 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 prem-
ise A1. The acceptance of P1 thus appears to be independent of the accep-
tance 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 ex-
pressed 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 spe-
cific 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 pro-
gram. For this purpose we must have an ability, a truth-discriminating capac-
ity.
33

_____________
30 Stanley/Williamson 2001, 414-415.
31 Cf. Gettier 1963.
32 Cf. McGinn 1984, 536 ff.
33 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 dis-
criminating capacity in Platos Theaetetus see Gonzalez paper in this volume.
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 differ-
ence 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 fol-
lowing 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 com-
plete 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 rele-
vant 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-how
d
that is
required for it depends, however, on the situation: the truth analysis can con-
sist in personally traveling to Rome and going to St. Peters Square if one
wants to know whether it is really true that St. Peters is in Rome, or alterna-
tively to take Goldmans 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 McGinn 1984, 536-537.
35 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.
36 Cf. Goldman 1992, 86 ff.
Gregor Damschen
292
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 pro-
gram 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, im-
plying 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: Einsteins walk from his house to his office is an intentional act F
1
,
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. [assumption]
(2) Einstein F
1
-s. [1, Z1]
(3) Einstein knows how to F
1
. [2, P1]
(4 Einstein knows that T(F
1
). [3, A1]
(5) Einstein completed analysis act U
1
. [4, P2]
(6) Einstein F
2
-s. [5, Z2]
(7) Einstein knows how to F
2
. [6, P1]
(8) Einstein knows that T(F
2
). [7, A1]
(9) Einstein completed analysis act U
2
. [8, P2]
(10) Einstein F
3
-s. [9, Z2]
and so on and so forth.

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 F
1
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.
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-how
d
is a species of knowledge-that.
8. Knowledge-That: a Species of Knowledge-How
d
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, suc-
cessfully 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 knows
d
how act F is com-
pleted. If the analysis of whether p or not-p is the case is itself an intentional
act F (lets call this premise G), then this act presupposes a knowledge-how
d
.
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 suc-
cessfully an intentional act F(p). [B C]
(P1) If S has completed sufficiently often, regularly and successfully an in-
tentional act F(p), then S knows
d
how to complete an intentional act
F(p). [C D]
Therefore, (A2) If S knows that p, then S knows
d
how to complete an in-
tentional 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-how
d
and propositional knowledge-that
have neither an intersection nor are they completely distinct. As a conse-
quence, A4 and A5 are false. Hence A1, A3, A4 and A5 are false, A2 is true.
Dispositional knowledge-how
d
is a necessary condition for propositional
knowledge-that, and propositional knowledge-that is a species or subcategory
of dispositional knowledge-how
d
.
Gregor Damschen
294
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 proposi-
tional knowledge-that (knowledge-how
p
), and the second kind is a non-
propositional form of knowledge-how in the sense of a disposition or ability
(knowledge-how
d
). Moreover, the species-of-knowledge argument helped to
demonstrate that dispositional and non-propositional knowledge-how
d
, and
not knowledge-that, is basic for our concept of knowledge. Examined more
carefully, it was also shown that the dispositional knowledge-how
d
is a neces-
sary condition for knowledge-that, hence knowledge-that is a species of the
dispositional knowledge-how
d
. And finally, it has been shown that the relation
between both specified forms of knowledge can be determined without pre-
supposing in advance a complete and thoroughly accepted definition of pro-
positional knowledge-that or dispositional knowledge-how
d
.
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. Phi-
losophical 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. Gettiers Paper.
Analysis 24: 46-48.
Damschen, Gregor. 2003. Grenzen des Gesprchs ber Ideen. Die Formen des Wissens
und die Notwendigkeit der Ideen in Platons Parmenides. In Platon und Aristoteles sub
ratione veritatis. Festschrift fr Wolfgang Wieland zum 70. Geburtstag, ed. Gregor Dam-
schen, Rainer Enskat, and Alejandro G. Vigo, 31-75. Vandenhoeck & Ruprecht:
Gttingen.
Damschen, Gregor. 2005. Ist Wissen-dass eine Unterart von Wissen-wie? [Is Knowled-
ge-That a Species of Knowledge-How?]. In Philosophische Perspektiven. Beitrge zum VII.
Internationalen Kongress der sterreichischen Gesellschaft fr 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 Platoni-
schen Sokrates lernen kann. In Amicus Plato magis amica veritas. Festschrift fr 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 fr philosophische Forschung
57: 431-445.
Enskat, Rainer. 2005. Authentisches Wissen. Prolegomena zur Erkenntnistheorie in praktischer
Hinsicht. Gttingen: Vandenhoeck & Ruprecht.
295
Gettier, Edmund. 1963. Is Justified True Belief Knowledge? Analysis 23: 121-123.
Goldman, Alvin. 1992. Liaisons. Philosophy Meets the Cognitive and Social Sciences. Cam-
bridge/Mass.
Hartland-Swann, John. 1956. The Logical Status of Knowing That. Analysis 16: 111-
115.
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 Knowledge-
How). In Epistemology Futures, ed. S. Hetherington, 71-94. Oxford: Oxford Univer-
sity 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: 529-
554.
No, 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: 379-
388.
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 1929-
1968. 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: 107-
125.
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 disci-
pline, or by way of reflection on some historical thinkers insight, or by unfil-
tered 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 dispo-
sitions (or cognitive processes understood dispositionally) that are significant
for epistemology.
Most contemporary epistemologists are committed to the epistemic sig-
nificance of cognitive dispositions. They may write instead of beliefs being
the result of appropriate processes but the processes in question are under-
stood 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 disposi-
tions, reflect on the distinction between an agents being objectively justified in a
belief and that agents 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) pos-
sessed 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 agents 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-
297
formation, the agent would not have formed or retained the belief in ques-
tion. Yet, even this does not quite ensure that the agent is objectively epis-
temically 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 agents belief is then counter-
factually dependent on this trigger, but is not thereby objectively justified.
Being objectively justified requires a richer array of counterfactual dependen-
cies. The agents 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 disposi-
tions and to their manifestation in episodes of belief formation and mainte-
nance.
Now, our central concern in this paper is not to examine the metaphysi-
cal 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 nar-
rowly 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 infer-
ence). We argue that a wider range of processes and dispositions is epistemi-
cally 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 Woodwards (2003) discussion of causes and invariance provides a useful framework for ap-
proaching the question of what it is to be caused in the right way.
2 This somewhat generic understanding of epistemic virtues and virtue epistemology aligns with
Grecos (2000, chapter 7-8; 2002a) understanding. It is largely in keeping with Sosas (1991a,
298
capture what is distinctive about virtue epistemology. For the traditional epis-
temology for which virtue epistemology is to provide an alternative is itself
committed to the epistemic importance of dispositions. We suggest that vir-
tue 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 ones ongoing self-regulation of
ones cognitive life, and in peoples joint regulation of their joint or commu-
nitys epistemic life, one must evaluate beliefs and the processes by which
they arose and were maintained. Thus, a core idea informing virtue episte-
mology 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 epis-
temic 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, episte-
mologists 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 agents tend to develop? What
plasticity is found here? What stable dispositions can they develop with train-
ing? The virtue epistemologist seeks to understand which constellation of
stable dispositions among those that humans can with training come to pos-
sess, 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
Zagzebskis, 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.
299
mologists. (Much of what we say in what follows will be informed by con-
temporary 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. Open-
mindedness 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 dispo-
sitions.
Stable dispositions of a more fine grained sort contribute to general infer-
ential 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. Gen-
eral inductive inferential competence requires such more specific stable dis-
positions 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 sensitiv-
ity to counter-evidence, to competing explanation, to possible common
causes, and the like.

Not all cognitive competences are so clearly inferential in character. Con-
sider a kind of general perceptual competence. To begin with, people develop do-
main 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 af-
forded 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
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 ones 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). Ones 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 inhib-
ited 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 epis-
temological approach. Much epistemology from the last four hundred years
has supposed that a rather limited range of cognitive processes is epistemi-
cally significant and crucial: (what we will here term) classically inferential proc-
esses. 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 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 deter-
mining 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 percep-
tual agent can have poor eye sight, provided that agent has developed dispositions that compen-
sate for the degraded character of the visual input received.
4 Ones competence with testimony would also seem to be dependent on a range of more particu-
lar sensitivities or dispositions (Henderson, forthcoming). Even ones general competence with
memory might also yield to useful analyses (Goldberg and Henderson, forthcoming).
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 informa-
tion 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 pos-
sessed. 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 explic-
itly 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 repre-
sented.
6
Second, the causal processes whereby beliefs are fixed (formed, re-
vised, or retained) must be occurrently isomorphic with the deductive and induc-
tive 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 dependen-
cies between the representations in the agents 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 con-
tentfully conspire to support the belief in question. A careful reading of a
diversity of modern and contemporary texts suggests a wide partisanship for
_____________
5 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 in-
volving 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.
6 This much does not require that the inference rules to which the system conforms are them-
selves represented.
302
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 re-
presented (as representations are standardly understood) and yet causally
conditions the agents belief fixing processes. (If this is difficult to envision,
bear with us, and we will explain more fully.) If this obtains, then the agents
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. Humes
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 reli-
ability 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 vindi-
cate 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 phe-
nomenal 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 ac-
counts 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 Carnaps
attempt to develop an inductive logic in terms of observation reports and ranges of state descrip-
tions. More recently, BonJours (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 episte-
mology 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 epis-
temic justification (or its absence). It is worth noting that even epistemologists who incline to-
wards 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.
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 repre-
sentations. 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 sug-
gests 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 ex-
ample may begin with much information that is not propositionally repre-
sented, and so are ill understood in terms of inference as a matter of argu-
ment.
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 infor-
mation that is possessed by the system and yet not represented in that proc-
essing. 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 classi-
cal inference, constitute a kind of inference intermediate to broad and classi-
cal. 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 repre-
sentations 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 com-
monly are neither classically or argumentatively inferential.
304
tional representational beginnings and endings), we can write of argumentative
inference.
Some of the stable dispositions that interest us here often mediate transi-
tions 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 infer-
ence need be classical inference.) To the extent that processes crucial to hu-
man 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 argumenta-
tive 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 (com-
monly acquired or refined and developed by training), (a) are crucial to sys-
tematic human epistemic successes on a range of cognitive chores (chores
that are commonly recognized by epistemologists to be important and neces-
sary), 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 industrys 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 hu-
mans 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 infer-
ence, but by dispositions keyed to essentially richer sets of information. By
way of example, we now discuss two epistemic chores: (1) chores within cen-
tral processes of belief fixation requiring the holistic accommodation of vast
ranges of potentially relevant information, and (2) chores involved in fast,
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 sug-
gestive.
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 classi-
cally inferential processes.
9
There is reason to believe that human belief gen-
eration must, and epistemically ought, proceed in an alternative manner. The al-
ternative understanding emerges in the wake of a family of recalcitrant
difficulties within the classical, computational, conception of mind in cogni-
tive science difficulties often classified under the rubric the frame prob-
lem. In this section we first describe those difficulties, drawing upon an
influential discussion by Fodor (1983), and recommend the alternative con-
ception suggested by frame-type problems. We then sketch one way of elabo-
rating this conception, inspired by connectionism and by the form of mathe-
matics 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 clas-
sically inferential processes.
In the closing pages of The Modularity of Mind, Fodor 1983 argues that cer-
tain 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 Fodors 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 cogni-
tive 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 sec-
tion 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
putational approach has gotten somewhere, he says, is in understanding such
modular subsystems, which by their nature delimit the class of relevant in-
formation. 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 main-
tains, 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 gar-
ner 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 Fodors terminology) isotropic and Quineian:
By saying that confirmation is isotropic, I mean that the facts relevant to the confir-
mation of a scientific hypothesis may be drawn from anywhere in the field of previ-
ously 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 confirma-
tion assigned to any given hypothesis is sensitive to properties of the entire belief sys-
tem; 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 informa-
tion from any portion of ones 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 simplic-
ity, plausibility, and conservatism (Fodor 1983, 108), which are determined
by the global structure of the whole of the current belief system and of poten-
tial successor systems.
Since belief fixation in human cognition is commonly a matter of induc-
tive inference from the information provided by input systems and the infor-
mation 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 prob-
lems that classicism has encountered in attempting to understand such central
processes:
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 re-
spects, 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 cant, 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 ones belief, the justification that one
has for that belief is not causally inert, but rather comes into play or is ap-
propriately operative in generating or sustaining that belief. Whatever infor-
mation 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 justi-
fication 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 re-
quires 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). Fodors 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
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 gen-
erally, 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 ac-
count 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 rele-
vance, and thus to much actual relevance within ones doxastic and informa-
tional set, one must be sensitive to the full set of ones information or beliefs.
Thus, in view of the potential relevance of everything to everything, everything
that one believes, all ones 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; its a highly credible hypothesis that a (tractable) computa-
tional 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 compu-
tational 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 oper-
ate 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 holis-
tic aspects are somehow implicit in the structure of the cognitive system, in
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) is implicit in the standing structure of the cognitive system
(rather than explicitly represented in the systems occurrent
cognitive states or explicitly stored in memory), and
(ii) 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 proc-
esses 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 compu-
tational processes that are mere shorthand algorithms for computations that
could instead have been carried out in a way that renders all relevant informa-
tion explicitly. Essentially morphological processing is a fundamentally differ-
ent 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 sci-
ence. 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 informa-
tional tasks facing an epistemic agent. (One could point to a vast range of
philosophical work to name a few examples, to BonJour 1985, to Shaperes
(1982) discussions of the range of information and understandings that in-
form observations in physics, to Goodmans (1973) work on induction and
the projectibility of certain predicates itself gauged in terms of much back-
ground information.)
Central processes of belief fixation then must be handled essentially mor-
phologically so that much possessed information is accommodated auto-
matically, 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 proc-
esses cannot be occurrently isomorphic with contentful relevance relations obtaining between
310
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 ex-
tended to provide an adequate account of real human cognition. This is no
less true for connectionist models than it is for classical computational mod-
els.
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, dynami-
cal systems theory, and the nonclassical framework with emphasis on fea-
tures 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 representa-
tions 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 appro-
priate to that information. Such information constitutes morphological con-
tent in the system, rather than explicitly-represented content. Among the
apparent advantages of connectionist systems, by contrast with classical com-
putational systems, is that morphological information in the weights gets
accommodated automatically during processing, without any need for a cen-
tral 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.
311
Learning is conceived quite differently within connectionism than it is within
the classical approach, since connectionist systems do not store representa-
tions. Learning is the acquisition, in the weights, of new morphological
content.
The branch of mathematics called dynamical systems theory is often ap-
plied 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 systems 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 deter-
mined 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 dimen-
sional, non-Euclidean, contoured surface: i.e., a topological molding of the n-
dimensional 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 dy-
namical 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 networks 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
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 train-
ing procedures employed in connectionist modeling bring about a co-
evolution of these two factors.) Once trained up, the systems 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 mor-
phologically 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 landscapes multifarious, superimposed, topographical con-
tours guarantee that the cognitive systems transitions from one occurrent
cognitive state to another are automatically appropriate not only to the ex-
plicit 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 cogni-
tive science provide a strong reason to maintain that the holistic,
Quineian/isotropic aspects of belief-fixation are accommodated in an essen-
tially morphological way in human cognition. We also maintain that this con-
clusion 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 proc-
esses 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 cen-
tral processes must accommodate the potential relevance of each piece of
possessed information to any given belief, and thus that all information pos-
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 manag-
ing the indicated epistemic chores, the competent human agent must accom-
modate much possessed information automatically without representation,
and thus without it featuring as a premise in a classical inference. The compe-
tent 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 classi-
cally inferential. Here, it is good to reflect on the dynamical system frame-
work we have recommended. One might think of some of the transitions be-
tween representational states as inferences on the order of those that have
been the common epistemological fare as transitions between representa-
tions (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 systems dynamics and the
morphological content on which it relies. This is to focus on epistemically crucial disposi-
tions 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 systems weight
space. This is a matter of its tendencies to move from one total state (or pat-
tern 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 sys-
tem acquires through courses of experience, through training, are highly de-
sirable epistemically are epistemic virtues. Some of the patterns of activation
through which the system passes are representations, so some of the transi-
tions 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 deter-
mined or conditioned only by representations, not all the epistemically desir-
314
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 prin-
ciples 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 vir-
tuous person would undertake and that this is a matter of a rational judg-
ment 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 mor-
phological 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 inferen-
tial 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 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, to-
wards 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. ).
12 See Horgan and Tienson 1996 for a fairly extended discussion of this point.
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, indeter-
minate 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 argumenta-
tively 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 proc-
esses 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 hy-
pothesis that this is managed by reliance on morphological content within a
kind of cognitive dynamical system. Again, virtue epistemology seems vindi-
cated.
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 ones environment, for example, one can
with reasonable sensitivity respond to common episodic encounters (and to
more or less enduring traces) in ones environment. Due to ones 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 judg-
ments in a range of common environmental conditions. Thus, one who is
perceptually competent with respect to local wildlife can make reliable identi-
fications 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
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 percep-
tual 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 generat-
ing 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 sus-
ceptible to argumentative reconstruction. In competent and epistemically
highly laudable perceptual processing, much such information, as morpho-
logical 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 rele-
vant 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 rea-
sonable 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 under-
wings. What of the Ivory Billed Woodpecker? Obviously very different base
rates (even given some uncertainty regarding whether it is extinct or just ex-
ceedingly rare). The perceptual process of a competent agent is permeable to
such information, and much more of relevance. One might think of the proc-
esses 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 envi-
ronment, 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.
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 famil-
iar things in the agents 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 agents per-
ceptual processes need not pass through propositional representations on the
way to the forming of a perceptual belief and thus may be broadly inferen-
tial without being argumentatively or classically inferential.
We believe that perceptual processes with their conditioning by, or
permeability to, much background information that is accommodated auto-
matically 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 appear-
ances 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 ones 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 infer-
ence, but by dispositions keyed to essentially richer sets of information. Cen-
tral processes of belief fixation turn on holistic sensibilities that require the
use of morphological content. So also do competent human perceptual proc-
318
esses. These have served to illustrate the epistemological importance of cogni-
tive 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 Aca-
demic 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. 3
rd
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.
319
Pollock, John. 1986. Contemporary Theories of Knowledge. Lanham, Md: Rowman and Little-
field.
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. Philoso-
phy 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 cogni-
tive faculties or character traits of various kinds. It is argued that it is appro-
priate 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 justifica-
tion 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 epis-
temic 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 ade-
quate solutions to some notorious problems of knowledge, like skepticism,
_____________
1 See Sosa 1980.
321
the Gettier problem or the lottery paradox.
2
Others think of virtue epistemol-
ogy 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 under-
standing 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 justi-
fication: Knowledge, according to Sosa is true belief out of intellectual vir-
tue, belief that turns out right by reason of the virtue and not just by coinci-
dence.
7
Following Lorraine Code and James Montmarquets accounts of virtue re-
sponsibility,
8
Linda Zagzebski has developed a detailed theory of virtue episte-
mology 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 See, for example, Greco 2000a, Greco 2003, and Greco 2005.
3 See, for example, Zagzebski 1996, 43ff., Code 1987, Montmarquet 1993, and Riggs 2003.
4 Roberts/Wood 2003, 257.
5 Sosa 1980, 189 my emphasis.
6 See Sosa 1991, 277ff.
7 Sosa 1991, 277.
8 See Code 1987, Montmarquet 1993.
Elke Brendel
322
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 ex-
ternal.
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 cog-
nitive abilities in Sosas 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 Ss 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 Ss believing p make S reliable in believing p (in the
present conditions, with respect to p.)
14
Greco also thinks of intellectual vir-
tues 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 Zagzebski 1996, 137 my emphasis.
10 Zagzebski 1996, 9. Greco objects to this. With Sosa he points out that it is perfectly in accor-
dance with the Greek tradition to understand virtues as innate or acquired faculties. In Platos
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 Sosas (and Grecos) account is
based on an Aristotelian conception of excellences since an excellence in the Aristotelian tradi-
tion is some kind of ability, disposition, power, faculty, or habit. (Goldman 2001, 30).
11 Zagzebski 1996, 271.
12 See Greco 2000a, 164ff.
13 See Greco 2000a, 218, Greco 2002, 303 or Greco 2004, 6 my emphasis.
14 See Greco 2000a, 218.
15 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 regard-
ing p because S is reliable in believing p. Alternatively: the intellectual abilities (i.e., powers or vir-
tues) that result in Ss believing the truth regarding p are an important necessary part of the total
set of causal factors that give rise to Ss 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-
323
For the moment, I want to postpone consideration of the adequacy of
these different approaches to virtue epistemology or their success (or other-
wise) in giving appropriate definitions of knowledge and other central epis-
temic 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 vir-
tues, since they are considered to be stable and reliable dispositions for form-
ing accurate beliefs about the world.
The second kind of intellectual virtues are character traits, such as open-
mindedness, conscientiousness, perseverance, intellectual curiosity and crea-
tivity, impartiality before evidence, unbiased examination of arguments. Rob-
erts 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 ones accomplishments, respect
for good arguments, repugnance at intellectual dishonesty, and in the case of
empirical science, surprise at the disconfirmation of ones 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 epis-
temology. 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
_____________
tion seems to be violated since the epistemic subject believes the truth regarding p in such cases
out of luck and not because of Ss intellectual abilities.
16 See Hookway 2003, 187.
17 Roberts/Wood 2003, 257.
18 Sherman/White 2003, 38.
Elke Brendel
324
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 vir-
tues 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 dispo-
sitions. Of course, I do not claim that I can give necessary and sufficient con-
ditions for possession of a disposition or that doing so would provide us with
satisfactory answers to all the important questions in this area. These ques-
tions include the following: what is the proper analysis of disposition ascrip-
tions? What is the ontological status of dispositions? How is the dispositional-
categorical distinction to be understood? Are dispositions intrinsic or extrin-
sic? 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 un-
der 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 dis-
posed 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-
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 disposi-
tion 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 cer-
tain 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 simi-
larity 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).
Elke Brendel
326
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 mani-
festations of elasticity include stretching, contracting, bouncing, deforming,
and reforming.
20

7. Dispositions normally dont manifest themselves in isolation, but in co-
operation 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 crystals disposition to dissolve in
water and the waters reciprocal disposition to dissolve salt. A match bursts into flame
when it is scratched across the abrasive surface of a matchbox. The matchs 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 objects disposition also depends on the objects
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 follow-
ing claim:
8. Even though a disposition of an object o manifests itself at least partly
in virtue of some of os 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 Shoemakers mere-Cambridge power.
22
Shoemakers 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 pos-
sesses these dispositions in virtue of some of its intrinsic properties, the latter
disposition can be lost without undergoing any changes of the keys intrinsic
properties. The key can lose the disposition to open Shoemakers front door
simply by replacing the lock of his front door by a lock with a completely
different shape. A similar example is Jennifer McKitricks 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 Mumford 1998, 6.
21 Heil 2005, 350.
22 See Shoemaker 1984, 221.
23 See McKitrick 2003, 160.
327
the waters intrinsic properties, but also on the actual content of McKitricks
pocket.
Changing or destroying an objects 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 mani-
festations 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 in-
terference. 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 Johnstons 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 Birds 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 in-
trinsic 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 struc-
ture, we can truly ascribe fragility to a glass vase with such a specific molecu-
lar 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 manifesta-
tion does no harm to the object as in the case of the manifestation of elas-
ticity of an elastic band by moderately stretching it. In cases of mental disposi-
tions 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 persons generosity or open-heartedness
_____________
24 See, for example, McKitricks discussion of other examples, such as vulnerability, visibility
or recognizability (McKitrick 2003, 161ff.), which show that in certain contexts dispositions
can be extrinsic.
25 See Johnston 1992, 223.
26 See Bird 1998.
Elke Brendel
328
consists in some intrinsic properties of that person it seems that, as in the
cases of physical dispositions like fragility, solubility, being poisonous, elastic-
ity 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 persons 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 doesnt mean that the
actual manifestation of a disposition is a necessary condition for defining a
persons 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 some-
how 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 Mumford 1998, 8.
28 See Goodman 1954, 40.
29 Mumford 1998, 4.
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 open-
hearted. 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 cir-
cumstances 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 liter-
ally 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 neverthe-
less important to see that in cases in which the disposition has been mani-
fested 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 Grices 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 in-
nate or acquired intrinsic properties of the subject. They are called intellectual
_____________
30 unless we use this disposition ascription to explain os behaviour.
Elke Brendel
330
virtues since the proper exercise of these capacities aims at achieving epis-
temic 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 Ss visual organ only occasionally func-
tions properly, Ss beliefs that are formed on the basis of Ss visual experi-
ences 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 Ss character. In the same vein, Douglas
Butler remarks: Character traits must be fairly permanent features of ones
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
persons logical reasoning abilities can be better or worse than those of an-
other 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 vari-
ety of possible manifestations in the intellectual activities of epistemic sub-
jects. 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 maintain-
ing knowledge and other epistemic goods typically manifests itself in intellec-
tual 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 per-
ceptual faculties, logical reasoning, the competent application of specific cog-
nitive skills and expert knowledge in this field, character traits, such as curios-
ity, 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.
331
For even when perceptual belief derives as directly as it ever does from sensory stim-
uli, it is still relevant that one has not perceived the signs of contrary testimony. A rea-
son-endowed being automatically monitors his background information and his sen-
sory 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 Fairweathers example of Conrad, the Doxastic Con-
formist,
33
Conrads primary goal is to bring his belief system into conformity
with the belief system of Mr. Cool, his role model. One day Mr. Cool ex-
pressed 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. Cools 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 Conrads 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 monitor-
ing 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 motiva-
tions play a significant justificatory role even in the case of perceptual beliefs.
34

The proper exercise of intellectual virtues in our intellectual and cognitive en-
deavours 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 exer-
cise of intellectual virtues consists in. What are, for example, acts of intellec-
tual virtue that give rise to true belief? In my general analysis of dispositions
_____________
32 Sosa 1991, 240.
33 Fairweather 2001, 74.
34 Fairweather 2001, 78.
Elke Brendel
332
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 deter-
mined 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 hear-
ing, 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, de-
pendent 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 intellec-
tual virtue that Roberts and Wood examine in great detail, is certainly a
praiseworthy character trait of an epistemic subject that can support the for-
mation of true beliefs. In contrast to a vain and arrogant person an intellectu-
ally humble person can better overcome obstacles to the acquisition of knowl-
edge 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 con-
joined, 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.
333
run, just about everybody will be epistemically better off for having, and having asso-
ciates 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 vir-
tues. 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 persons 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 sub-
jects 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 Roberts/Wood 2003, 271f.
37 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.
Elke Brendel
334
under appropriate circumstances. But since natural faculties and character
traits share very important features of properties normally regarded as dispo-
sitions (see claim 1-9), it seems to me theoretically unjustified and premature
to exclude natural faculties from the extension of the concept of a disposi-
tion. 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 ap-
propriate circumstances under which this faculty manifests itself are frequent
and almost always actual during waking hours. So, saying that a person exer-
cising the faculty of good eyesight most of the time is disposed to accurate
visual perceptions of her environment would lead to a false pragmatic impli-
cature 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 intellec-
tual virtues that function as means of acquiring knowledge (and other epis-
temic goods) in different accounts of virtue epistemology as virtuous disposi-
tions. In particular, in applying certain crucial features of dispositions to
intellectual virtues we gained some important insights into the nature of intel-
lectual virtues and their epistemic role in our intellectual inquiries and delib-
erations. I shall now finally turn to the question of whether defining knowledge
in terms of these virtuous dispositions is a feasible and promising epistemo-
logical 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:
335
Suppose it turns out that human perception is not within our strong control, even in-
directly. 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 ac-
curate 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 knowl-
edge 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 error-
possibility can cast doubt on someones 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 be-
lief 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 reli-
ability of Ss intellectual virtues, the exercise of which result in Ss 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, speakers intentions, questions of what is at stake in
the given context etc.
40
Second, what kinds of intellectual virtue and the extent
_____________
38 Greco 2000a, 201f.
39 See, for example, Greco 2000a, who claims that his agent realism is consistent with contextual-
ism (252).
40 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/Jger 2005. In this
Elke Brendel
336
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 Ss belief that there is a lake in the distance, the proper functioning of
Ss 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 error-
possibility in this context. Just trusting her visual faculties seems to be insuffi-
cient 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 episte-
mology 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 an-
swered:
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 mo-
tivated 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
_____________
paper I do not want to embrace a particular contextualist account of knowledge, since the defini-
tions of knowledge in virtue epistemology are consistent with different contextualist theories.
41 See Zagzebski 2003, 151f.
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 be-
lief, 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 praisewor-
thiness 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 episte-
mological projects but are not suitable as necessary conditions for all kinds of
knowledge.
Second, many intellectual virtues are not strictly truth-conducive. It there-
fore 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 combina-
tion of intellectual virtues with other faculties or character traits can be intel-
lectual 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 virtu-
ous scientists were wrong about very many things, but were nevertheless of
utmost importance for the development of the scientific progress. Montmar-
quet 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.
Elke Brendel
338
very often ascribe intellectual virtues to a person even if her intellectual in-
quiries do not immediately aim at a correct theory:
[I]f we are to appraise the relative worth or virtue of epistemic agents by the truth-
conduciveness 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 truth-
conduciveness 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 nor-
mally 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/praise-
worthiness and knowledge has to be explained in a way that accounts for the
fact that spontaneously acquired true perceptual beliefs can count as knowl-
edge. Otherwise it seems that knowledge in virtue epistemology is an intellec-
tually too demanding epistemic concept. Second, the truth-conduciveness of
_____________
43 Montmarquet 1993, 21.
44 Zagzebski 1996, 181.
45 Hookway 2003, 197.
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 col-
lection 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 contri-
bution 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 con-
cepts, 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. Jger, eds. 2005. Contextualisms in Epistemology. Dordrecht: Springer.
Brendel, E., and Ch. Jger. 2005. Contextualist Approaches to Epistemology: Problems
and Prospects. In Contextualisms in Epistemology, ed. E. Brendel, and Ch. Jger, 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. Ox-
ford: 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 See, for example Riggs 2003.
47 I am very grateful to Quassim Cassam for helpful comments on an earlier version of this paper.
Elke Brendel
340
Greco, John, 2005. A Different Sort of Contextualism. In Contextualisms in Epistemology,
ed. E. Brendel, and Ch. Jger, 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 Phi-
losophy 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 Univer-
sity 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 Uni-
versity 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 Sci-
ence (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 Disposi-
tional 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. Jger), Zitat und Bedeutung, Ham-
burg 2007 (ed. with J. Meibauer and M. Steinbach), Grundthemen Philosophie:
Wissen, Berlin-New York 2010; Understanding Quotation. Linguistic and Philosophi-
cal 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 philoso-
phy 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 Univer-
sity of Ottawa. He is the author of Dialectic and Dialogue: Platos Practice of Phi-
losophical 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 philoso-
phy of social science and on epistemology, where he has sought to garner
from contemporary cognitive science lessons regarding venerable epistemo-
logical positions. In addition to his Epistemic Competence and Contextualist
342
Epistemology, Journal of Philosophy (1994), work with Terry Horgan, notably,
Iceberg Epistemology, Philosophy and Phenomenological Research (2000), pro-
vides 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 ex-
tensively 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 Httemann is Professor of Philosophy at Westflische Wilhelms-
Universitt Mnster. His main areas of research are philosophy of science and
early modern philosophy. He is the author of Whats Wrong With Microphysical-
ism (Routledge 2005), and has edited Kausalitt und Naturgesetz in der frhen
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 ex-
ecutive director of the Centre for Logic, Philosophy and History of Science.
He is the author of a book on Aristotles theory of dispositions (Tun und Kn-
nen, Frankfurt 2002) and co-editor of an introduction to biomedical ontology
(Biomedizinische Ontologie, Zrich 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 Gott-
fried Wilhelm Leibniz (Mnchen: 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 Descartess Changing
Mind with Princeton University Press.

343
Jennifer McKitrick is Associate Professor at the University of Nebraska-
Lincoln. 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 dispo-
sitions, including A Case for Extrinsic Dispositions (Australasian Journal of
Philosophy 2003).

Burkhard Meiner is Professor of Ancient History at the Helmut Schmidt
Universitt (University of the Armed Forces of the Federal Republic of Ger-
many) 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 Knigshof (Gttingen 1992); Die techno-
logische Fachliteratur der Antike (Berlin 1999) and Hellenismus (Darmstadt 2007).

Stephen Mumford is Professor of Metaphysics at the University of Notting-
ham. 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 co-
authored with Rani Lill Anjum.

Ursula Renz is Professor of Philosophy at the University of Klagenfurt. She
is the author of Die Rationalitt der Kultur: Kulturphilosophie und ihre transzendentale
Begrndung bei Hermann Cohen, Paul Natorp und Ernst Cassirer (2002), Die Er-
klrbarkeit von Erfahrung: Realismus und Subjektivitt in Spinozas Theorie des mensch-
lichen Geistes (2009) and has written several articles on German Philosophy and
Early Modern Philosophy. Recently, together with Hilge Landweer as co-
editor, she has published Klassische Emotionstheorien (2008).

Robert Schnepf is a Senior Lecturer (Privatdozent) at Martin-Luther-
Universitt 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 Metaphy-
sik im ersten Teil der Ethik Spinozas (Wrzburg 1996); Die Frage nach der Ursache
systematische und problemgeschichtliche Untersuchungen zum Kausalitts- und zum Schp-
344
fungsbegriff (Gttingen 2006); and Geschichte erklren (Gttingen 2010) and the
editor of a variety of anthologies.

Oliver R. Scholz is Professor of Philosophy at the Westflische Wilhelms-
Universitt Mnster, 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 Rationalitt (1999, 2nd edi-
tion 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).

Gregor Damschen. (2009) - Debating Dispositions, Issues On Metaphysics, Epistemology and Philosophy of Mind

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Gregor Damschen. (2009) - Debating Dispositions, Issues On Metaphysics, Epistemology and Philosophy of Mind

Transféré par

Droits d'auteur :

Formats disponibles

Debating Dispositions

x has the disposition to do (or to be) F

d. a particular having a disposition that gives something a dis-

Vous aimerez peut-être aussi