Wittgenstein and Naturalism

This is the first collection of essays to focus explicitly on the relation between
Wittgenstein and Naturalism. The volume brings together 14 previously
unpublished essays by internationally renowned philosophers and emerging
research leaders. It is philosophical rather than exegetical in intent and
explores how Wittgenstein’s ideas can be used to advance ongoing debates
about metaphysical and methodological naturalism which are central to
current analytic philosophy.
This volume seeks to apply and examine Wittgensteinian thought in relation
to ongoing debates about naturalism. It brings together philosophers who
combine an interest in Wittgenstein with substantial research interests
outside of Wittgenstein scholarship. All of the chapters are fundamentally
concerned with addressing questions and topics that are a focus of current
philosophical debates, and with using Wittgensteinian ideas and insights as
needed in order to do so.
Kevin M. Cahill is a Professor of Philosophy at the University of Bergen. He
works mainly on Wittgenstein’s Philosophy and the Philosophy of the Social
Sciences. His publications include The Fate of Wonder: Wittgenstein’s
Critique of Metaphysics and Modernity (2011).
Thomas Raleigh is a Post-Doctoral Fellow at Ruhr-University, Bochum.
His research is primarily in the Philosophy of Mind and Epistemology,
with particular interest in the work of Wittgenstein. As well as the present
volume, he is also the co-editor, together with Jonathan Knowles, of
Acquaintance: New Essays (forthcoming).

Wittgenstein’s Thought and Legacy
Edited by Eugen Fischer, University of East Anglia, UK and
Severin Schroeder, University of Reading, UK

1 Wittgenstein on Thought and Will
Roger Teichmann
2 Wittgenstein on Sensation and Perception
Michael Hymers
3 Wittgenstein and Naturalism
Edited by Kevin M. Cahill and Thomas Raleigh

Wittgenstein and Naturalism

Edited by Kevin M. Cahill
and Thomas Raleigh

First published 2018
by Routledge
711 Third Avenue, New York, NY 10017
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2018 Taylor & Francis
The right of the editors to be identified as the authors of the
editorial material, and of the authors for their individual chapters,
has been asserted in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted
or reproduced or utilised in any form or by any electronic,
mechanical, or other means, now known or hereafter invented,
including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from
the publishers.
Trademark notice: Product or corporate names may be
trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
ISBN: 978-1-138-23686-8 (hbk)
ISBN: 978-1-315-30159-4 (ebk)
Typeset in Sabon
by Apex CoVantage, LLC

Contents

Acknowledgments
List of Abbreviations
List of Figures
Introduction

vii
viii
ix
1

TH O M A S R A L E IGH AN D KE VIN M. CAH IL L

PART I

Varieties of Naturalism
1 Wittgenstein and Naturalism

13
15

PAU L F. S N OWDO N

2 Wittgenstein’s Liberal Naturalism of Human Nature

33

DAV I D M ACARTH UR

3 Naturalism in the Goldilocks Zone: Wittgenstein’s
Delicate Balancing Act

56

DA N I E L D. HUTTO AN D GL E N DA SATN E

PART II

Language: Self, Truth, and Mathematics
4 Sensations, Natural Properties, and the Private
Language Argument

77

79

W I L L I A M C HIL D

5 Wittgenstein, Self-Knowledge, and Nature
A N N A L I SA CO L IVA

96

vi

Contents

6 The End of an Affair

119

C H A R L E S TR AVIS

7 Later Wittgenstein and the Genealogy of Mathematical
Necessity

151

S O R I N B A N GU

PART III

Animal Minds, Human Psychology
8 Minding the Gap: In Defense of Mind-Mind Continuity

175
177

D O R I T B A R- O N

9 Rational Animals

204

J U L I A TA N N EY

10 Modes of a “Complicated Form of Life”: Expression
and Human-Animal Continuity

223

S TI N A B ÄC K S TRÖ M

PART IV

Naturalism and Meta-Philosophy

241

11 Wittgenstein, Hume, and Naturalism

243

B E N E D I C T S MITH

12 Wittgensteinian ‘Therapy’, Experimental Philosophy,
and Metaphilosophical Naturalism

260

EUGEN FISCHER

13 Representationalism, Metaphysics, Naturalism:
Price, Horwich, and Beyond

287

J O N ATH A N KN OWL E S

14 Do Pragmatic Naturalists Have Souls? Should Anyone
Be Paid to Worry About It?

309

B J Ø R N TO R GRIM RA MB E RG

List of Contributors
Index

329
335

Acknowledgments

Several of the essays in this volume were first presented at the conference
“Wittgenstein, Philosophy of Mind, and Naturalism” held at the University of Bergen on June 12–13, 2015. The editors were co-organizers
of this conference, and they would like to thank the Universitetsfond at
the University of Bergen as well as the “Representationalism or AntiRepresentationalism?” project at the Norwegian University of Science
and Technology for the generous support that made this event possible.
In addition, we would like to express our gratitude to Andrew Weckenmann and to co-editor Alexandra Simmons from Routledge for their
help and patience throughout the process of putting this volume together.
Thanks are also due to series co-editor Eugen Fischer for much guidance
along the way. Finally, we would like to thank our contributors, whose
hard work and philosophical acumen made this volume possible.

Abbreviations

BB
BT
CV
LWVC

OC
PG
PI
RFM

RPP I

RPP II

TLP
WLC II

The Blue and the Brown Books. New York: Harper
Torchbooks, 1958.
The Big Typescript TS 213. Translated by C. Grant Luckhardt
and Maximilian A. E. Aue. Oxford: Blackwell, 2005.
Culture and Value (rev. 2nd ed.). Edited by G.H. von Wright.
Translated by Peter Winch. Oxford: Blackwell, 1998.
Ludwig Wittgenstein and the Vienna Circle. Edited by Brian
McGuinness. Translated by Joachim Schulte and Brian
McGuinness. Oxford: Blackwell, 1979.
On Certainty. Translated by Denis Paul and G. E. M.
Anscombe. New York: Oxford: Blackwell, 1969
Philosophical Grammar. Edited by R. Rhees. Translated by
Anthony Kenny. Oxford: Basil Blackwell, 1974
Philosophical Investigations. (See individual contributions for
relevant edition)
Remarks on the Foundations of Mathematics. Edited by G. H.
von Wright, R. Rhees, G. E. M. Anscombe. Revised editon.
Oxford: Blackwell, 1978.
Remarks on the Philosophy of Psychology. Edited by G. E.
M. Anscombe and G. H. von Wright. Translated by G. E. M.
Anscombe. Oxford: Basil Blackwell, 1980.
Remarks on the Philosophy of Psychology. Edited by G. H.
von Wright and Heikki Nyman. Translated by C. G.
Luckhardt and M. A. E. Aue. Oxford: Basil Blackwell, 1980.
Tractatus Logico-Philosophicus. (See individual contributions
for relevant edition.)
Wittgenstein’s Lectures Cambridge 1932–35. Edited by Alice
Ambrose. Chicago: University of Chicago Press, 1982.

Figures

7.1 The Queen’s pebbles
7.2 An arrangement of 25 batches of 25 pebbles each,
i.e., 24 horizontal batches and one vertical.
8.1 “Pure” Triangulation
8.2 “Reflective” Triangulation
8.3 “Intermediate” Triangulation

161
162
183
184
194

Introduction
Thomas Raleigh and Kevin M. Cahill

Wittgenstein is still routinely hailed as one of, perhaps the most, important
and influential of 20th-century philosophers, and his work continues to
generate a great deal of scholarly debate and literature. For example,
John Searle, interviewed recently in the magazine New Philosopher,
though characterizing his own work as ‘profoundly anti-Wittgensteinian’
nevertheless stated that: “Wittgenstein was the greatest philosopher
of the 20th century”.1 Moreover, many of the most prestigious living
philosophers, currently working in some of the most highly-respected
departments—e.g., Robert Brandom, John McDowell, Crispin Wright,
Daniel Dennett, Simon Blackburn, Paul Horwich, Penelope Maddy, Richard Moran, Barry Stroud, Huw Price, and others—explicitly acknowledge their Wittgensteinian sympathies. Yet while substantial numbers
of articles and monographs on his philosophy continue to be produced,
Wittgenstein’s thought (or Wittgensteinian philosophy, more broadly)
often seems absent from philosophical debates where the agenda has
been set by various forms of naturalism.
Wittgenstein was centrally concerned with the puzzling nature of the
mind, mathematics, morality, and modality. He also developed innovative
views about the status and methodology of philosophy and was explicitly
opposed to crudely ‘scientistic’ world-views. His later thought has thus
often been understood as elaborating a nuanced form of naturalism (e.g.,
Cavell, 1979, McDowell, 1994, Pears 1995, Crary 2007, McGinn 2013,
Macarthur 2015, Livingston 2015),2 appealing as it does to such notions
1 www.newphilosopher.com/articles/john-searle-it-upsets-me-when-i-read-the-nonsensewritten-by-my-contemporaries/
2 See Stanley Cavell, The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy
(Oxford: Clarendon Press, 1979); John McDowell, Mind and World (Cambridge: Harvard University Press, 1994); David Pears, “Wittgenstein’s Naturalism,” The Monist 78:4
(1995): 411–424; Alice Crary, “Wittgenstein and Ethical Naturalism,” in Wittgenstein and
His Interpreters: Essays in Memory of Gordon Baker, eds. Guy Kahane, Edward Kanterian, and Oskari Kuusela (Oxford: Blackwell, 2007), 295–319; Marie Mcginn, “Liberal
Naturalism: Wittgenstein and McDowell,” in Philosophical Methodology: The Armchair
or the Laboratory?, ed. Matthew C. Haug (Oxon and New York: Routledge, 2013),

2 Thomas Raleigh and Kevin M. Cahill
as “form of life”, “primitive reactions”, “natural history”, “general facts
of nature”, and “common behavior of mankind”.3 Nevertheless, as indicated above, Wittgenstein’s work is often bypassed from much of the
contemporary debates on naturalism and naturalizing projects.4 This
is especially the case with those debates in recent analytic philosophy
concerned with what Frank Jackson has labelled ‘placement problems’.5
These problems involve questions concerning how or whether various
important phenomena—consciousness, ethical and rational norms, moral
and aesthetic values, logical and mathematical truths, free will, familiar
everyday objects, social institutions and practices etc.—can be explained,
somehow or other, as being part of, or supervening on, the one same
natural/physical domain that is studied by the natural sciences.
There are a number of possible explanations one might speculatively
advance for Wittgenstein’s comparative absence from these discussions:
(i) The inevitability that in the era that comes a full generation after the
death of a philosopher as famous and influential as Wittgenstein,
there will be a ‘swing of the pendulum’ in philosophical fashion back
away from his work as his most famous and influential pupils die or
retire and their pupils in turn come to prominence.
(ii) The highly original, idiosyncratic and at times difficult nature of his
prose-style, together with the relative lack of references to other philosophers, can make it hard to see how Wittgenstein’s writings fit
into the dialectic of standard contemporary debates.
(iii) The great successes of the physical and biological sciences and their
prestige and prominence in academia in the era since Wittgenstein’s
death. Though Wittgenstein himself had a background in engineering before he came to philosophy, he was openly hostile towards the
trend of venerating the empirical sciences as the only source of truth
62–85; David Macarthur, “Liberal Naturalism and Second-Personal Space: A NeoPragmatist Response to the ‘Natural Origins of Content’,” Philosophia 43 (2015): 565–578;
Paul M. Livingston, “Naturalism, Conventionalism, and Forms of Life: Wittgenstein and
the ‘Cratylus’,” Nordic Wittgenstein Review 4:2 (2015): 7–38.
3 See Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe, 3rd ed.
(Oxford: Blackwell, 2001), §19, §23, §241, p. 218, §25, §415, p. 230, and §206.
4 As an important exception, it is worth nothing that the recent recoil from ‘orthodox’
metaphysics has seen a surge of interest in neo-pragmatist and expressivist theorizing
in the philosophy of language, a general approach which bears obvious affinities with
Wittgenstein’s work—e.g., in a recent volume on pragmatism and expressivism (Price
et al. 2013) Huw Price, Simon Blackburn, Robert Brandom, and Paul Horwich all explicitly avow their indebtedness to Wittgenstein. See Huw Price, Simon Blackburn, Robert
Brandom, Paul Horwich, and Michael Williams, Expressivism, Pragmatism and Representationalism (Cambridge: Cambridge University Press, 2013). See also many of the
papers in the recent volume Steven Gross, Nicholas Tebben, and Michael Williams, eds.,
Meaning Without Representation (Oxford: Oxford University Press, 2014).
5 Frank Jackson, From Metaphysics to Ethics (Oxford: Clarendon, 1998).

Introduction

3

or insight, especially concerning the nature of our human minds and
culture. This strand in his writings might give the impression of a
hostility to science itself.
(iv) Wittgenstein’s meta-philosophical claims—about the nature of philosophical problems and their (dis)solution—seem to be in tension
with the revival, post-Kripke, of traditional metaphysical notions
and topics, and the current respectability of a priori analysis and
appeals to intuition to reach conclusions about fundamental metaphysical principles.
Whatever the explanation may be, there are very strong reasons for
thinking not only that Wittgenstein’s later thought in particular evinces
some kind of naturalism, but that it is also a potentially rich and fecund
source of insight and challenge for many important contemporary discussions. Among other things, the essays in this volume exemplify what we
hope will become a newly prevalent attitude in mainstream philosophy
(see, e.g., Maddy 2014)6—that one can draw on, use, and be in dialogue
with Wittgenstein without necessarily being a card-carrying ‘Wittgensteinian’. (And, conversely, that one can respond to and criticize aspects of
Wittgenstein’s work without being a card-carrying anti-Wittgensteinian.)
This volume thus seeks to explore the significance of Wittgenstein and
Wittgensteinian thought more generally to the ongoing conversation
about naturalism. It brings together philosophers who have a direct
scholarly interest in Wittgenstein’s work with others who have substantial research interests outside of Wittgenstein scholarship but whose work
nevertheless draws on Wittgenstein’s philosophical legacy in important
ways. All of the chapters are primarily concerned with addressing questions and topics that are the focus of current philosophical debates and
with using Wittgensteinian ideas and insights as needed in order to do so.
We hope that this collection of essays will make clear, at both the topical and the methodological level, how Wittgenstein’s work and thought
remain relevant to these contemporary philosophical issues and capable
of giving fresh impulses to the debates surrounding them.
The collection is divided into four sections, each of which addresses a
different aspect of Naturalism and its relation to Wittgenstein’s thought.
More detailed descriptions are given below. The first section considers
how ‘Naturalism’ could or should be understood. The second section deals
with some of the main problematic domains—consciousness, meaning,
mathematics—that philosophers have typically sought to naturalize. The
third section focuses on the now much-discussed topic of animal minds
and their relation to the human mind. The final section is concerned with

6 Penelope Maddy, The Logical Must: Wittgenstein on Logic (Oxford: Oxford University
Press, 2014).

4 Thomas Raleigh and Kevin M. Cahill
the naturalistic status and methodology of philosophy itself. The volume
casts a fresh, Wittgensteinian light on many classical philosophical issues
and brings Wittgensteinian ideas to bear on a number of newer topics
that are a focus of current debates—e.g., experimental philosophy, neopragmatism, animal cognition/ethics—in which naturalism is playing a
central role.

I. Varieties of Naturalism
The first essay of the volume deals with the large question as to whether
Wittgenstein should be read as a naturalist at all. Paul Snowdon argues
that although the later Wittgenstein’s explicitly meta-philosophical statements avow a thoroughly anti-theoretical and negative approach to philosophy, in fact Wittgenstein’s later work provides a number of positive
philosophical views that can rightly be labelled ‘naturalistic’. In particular, this label can be illuminatingly applied to the picture Wittgenstein
provides of language and of the foundations of meaning and intentionality. Moreover, much of Wittgenstein’s later philosophy of language is animated by his opposition to the sort of non-naturalistic theorizing about
meaning which appeals to something like Frege’s ‘third realm’7—i.e. an
ontology of abstract propositions/thoughts to which we are somehow
psychologically related.8 However, whilst Snowdon takes the general picture of the nature of language and meaning that the later Wittgenstein
provides to be not only naturalistic but broadly correct, his paper concludes by identifying some specific elements and claims within this naturalistic picture which are unpersuasive or overly restrictive.
The essays by David Macarthur and by Dan Hutto & Glenda Satne
then take up the questions of how exactly one should develop a Wittgensteinian form of naturalism and what exactly the place and the role of the
natural sciences should be within such a worldview.
Macarthur’s paper—like Snowdon’s—begins by recognizing that there is
a prima facie tension between the explicitly anti-theoretical methodology
that Wittgenstein espouses and the idea that he is nevertheless committed
to some form of naturalism. Macarthur also emphasizes how, throughout his life, in both the earlier and later periods, Wittgenstein very clearly
stated his opposition to the idea, influentially championed by Quine, that
philosophy is continuous with science. It is clear then that Wittgenstein
would have no sympathy for the sort of ‘scientific naturalism’ that takes
metaphysics to be in the business of providing ‘super’ explanations modeled on the explanations that physics provides. Rather, Macarthur argues,
7 See Gottlob Frege, “Der Gedanke,” Beiträge zur Philosophie des deutschen Idealismus
1:2 (1918): 58–77.
8 Though compare Charles Travis’s contribution to the present volume, which argues that
there is in fact a deep affinity between the later Wittgenstein and the views of Frege in
Der Gedanke.

Introduction

5

building on important earlier work,9 that we should treat Wittgenstein
as a ‘liberal naturalist’—with the emphasis falling on method as opposed
to doctrine and human nature rather than nature as such. An important
aspect of the method is not to countenance super-natural entities nor admit
the explanatory exhaustiveness of the sciences—thus allowing for a liberal
plurality of ways of knowing and understanding the world, particularly the
human world. Macarthur thus interprets Wittgenstein’s form of naturalism
as ‘non-constructive and dialectical’ rather than a substantive metaphysical doctrine. Wittgenstein aims to remind us of certain facts about natural human behavior and the natural history of human beings in order to
prevent us from being tempted by the chimerical explanations associated
with metaphysical pictures/theories about, paradigmatically, the nature of
mind, language, and meaning. Wittgenstein aims to replace the philosophical demand for a substantial theory of mind, meaning and language with
an unfamiliar (or rather, unfamiliarly familiar, hence uncanny) description
of how we use the terms “mind”, “meaning” and “language”.
However, Hutto & Satne, in their contribution, argue that this kind
of Liberal Naturalism is too liberal insofar as it fails to offer any positive account of the relation between the many different possible modes
or domains of explanation—i.e. between the different special sciences,
or between the ‘scientific image’ and the ‘manifest image’ more generally.
Moreover, they argue that the specific version of Liberal Naturalism that
we find in John McDowell’s influential writings risks re-introducing the
kind of super-natural, platonic entities that Wittgenstein was concerned
to reject. Hutto & Satne suggest that the correct philosophical framework to be drawn from the later Wittgenstein is what they call ‘Relaxed
Naturalism’,10 a form of Naturalism that insists, contra McDowell, that
the nature of meaning and intentionality can and should be illuminated by
a synthesis of the various human sciences. Relaxed naturalism thus adopts
an integrative approach to explanation that endorses central neo-pragmatist ideas and is akin to the Sydney, as opposed to the Canberra Plan,
in adopting “a somewhat more permissive understanding of what it is for
some everyday concept to ‘find a place’ in a scientific vision of reality”.11

II. Language: Self, Truth, and Mathematics
One of the most important recent influences on how ‘naturalism’ is
understood within philosophy is David Lewis’s development of the idea
that certain properties are inherently more natural than others, quite
9 See David Macarthur, “Taking the Human Sciences Seriously,” in Naturalism and Normativity, eds. M. De Caro and D. Macarthur (New York: Columbia University Press,
2010), 123–141 and also Macarthur, “Second-Personal Space”.
10 See Daniel D. Hutto and Glenda Satne, “The Natural Origins of Content,” Philosophia
43:3 (2015): 521–536.
11 Jennan Ismael, “Naturalism on the Sydney Plan,” in Philosophical Methodology, 87.

6 Thomas Raleigh and Kevin M. Cahill
independently of our particular human practices of classifying and
categorizing things.12 Bill Child’s paper considers the extent to which
Wittgenstein’s later philosophy is opposed to such a doctrine of natural
properties, and in particular the extent to which Wittgenstein’s (so-called)
‘Private Language Argument’ relies on denying that sensational/phenomenal properties are ‘natural’ in something like Lewis’s sense. Child accepts
that any plausible account of our sensations and sensation concepts will
have to acknowledge some notion of ‘natural’ phenomenal similarity that
goes beyond the subject’s recognitional and classificatory capacities—
contra some of Wittgenstein’s more anti-Platonistic statements. However,
Child argues that the notion of naturalness alone cannot determine the
reference of our sensation/phenomenal concepts—we will still have to
appeal to our classificatory practices as a crucial part of the referential
story. Thus we arrive at a kind of middle-ground which, Child argues, not
only avoids the problems which beset the more extreme versions of both
anti-Platonism and the natural properties view, but is also still recognizably a Wittgensteinian position.
Annalisa Coliva’s contribution is concerned not only with how we
self-ascribe sensations, but also with our avowals of our own intentions
and other propositional attitudes. Coliva explores how Wittgenstein’s
remarks on self-knowledge can be developed into an expressivist and
constitutivist theory of first-personal psychological self-ascriptions. She
shows how such an account is able to respect three of the most distinctive
features of our psychological self-ascriptions: groundlessness (our avowals are not based on either inner observation or inference), transparency
(for many mental states, being in the state guarantees being in a position
to know that one is in that state) and authority (unless we suspect insincerity, a subject’s first-personal avowals are not open to being doubted or
challenged by others). Moreover, by emphasizing the continuity between
first-personal psychological ascriptions and naturally or instinctively
expressive non-linguistic behaviors, such a Wittgensteinian theory can
provide a ‘soft-naturalistic’ explanation of these distinctive features of
self-knowledge which avoids appealing to any kind of special epistemic
access to some mysterious private phenomenal realm.
In contrast to a number of the other papers in this volume, Charles
Travis’s contribution discerns a non-naturalistic strand in Wittgenstein’s
thought, interpreting the transition from his earlier to his later work as
turning back towards Frege by accepting that representational contents
cannot be reduced to any kind of concrete, particular elements. Travis
traces how Wittgenstein came to realize that the failures of the Tractarian account of judgment and meaning were ultimately due to a failure
to respect a fundamental and categorical distinction between the general
and the particular. The Tractatus is thus vulnerable to a general Fregean
12 See David Lewis, “New Work for a Theory of Universals,” Australasian Journal of
Philosophy 61:4 (1983): 343–377.

Introduction

7

critique of any form of correspondence theory of truth. On Travis’s reading then the later Wittgenstein emerges as an anti-naturalist thinker, one
who recognized that ‘naturalism is not the price of objectivity’.13
When asked in 1944 to provide a short description of his work, Wittgenstein claimed that his ‘chief contribution has been in the philosophy of
mathematics’.14 And mathematics, along with the mental, has provided
one of the perennial problem cases for naturalizing projects in philosophy.
Sorin Bangu, in his paper, defends a Wittgensteinian account of mathematical necessity according to which mathematical claims are strictly
speaking neither true nor false but rather are statements of rules—rules
which govern what it is to count or to calculate correctly. Bangu provides
a “genealogical” account of how empirical regularities can be ‘hardened’
so that they come to function as rules and are thus removed from possible
empirical confirmation or disconfirmation. Bangu suggests that such an
account not only allows for a naturalistic way to avoid Platonist mythologizing of mathematics, but it also offers a new way to answer skepticism
about the objectivity of mathematical truth.

III. Animal Minds, Human Psychology
Wittgenstein wrote at PI §25 “Commanding, questioning, recounting,
chatting, are as much a part of our natural history as walking, eating,
drinking, playing”. This clearly suggests that he believed our biological
heritage had much relevance for our view of philosophy. Yet, while everyone agrees that Wittgenstein rejected the idea that philosophy could actually be one of the natural sciences, the nature of the relation between the
activity of philosophy and other kinds of activities whereby we obtain
knowledge about the world is much disputed, both quite generally and as
it pertains to how we understand Wittgenstein’s philosophy. This relation
is especially acute when it comes to the question of animal minds.
In her “Minding the Gap”, Dorit Bar-on aims to engage what she
describes as a long tradition of “continuity skeptics” who have sought to
establish on conceptual grounds that the minds of non-human animals
(and possibly even of very young humans) are separated from our minds
by an unbridgeable gap. This, it is thought, undermines the possibility
of an intelligible philosophical explanation of the natural emergence of
mind; and it renders futile any search for natural precursors of our own
minds in the mental capacities of ‘simpler minds’. Drawing on resources
from both everyday life and natural science, Bar-on presents a form of
non-reflective communication that we share with non-linguistic and
pre-linguistic creatures: expressive communication. She argues against
continuity skeptics such as Donald Davidson that proper appreciation
13 See section 9 of Travis’s contribution to this volume.
14 See Ray Monk, Ludwig Wittgenstein: The Duty of Genius (New York: The Free Press,
1990), 466.

8 Thomas Raleigh and Kevin M. Cahill
of the role expressive capacities play in the lives of creatures possessing
them points to a sensible intermediate stage in (what Wittgenstein would
describe as) a natural history that could connect us with our pre-human
ancestors (as well as connecting adult language users with their younger,
preverbal selves). She concludes with some reflections on the implications
of the existence of such a natural history for a philosophical understanding of the relationship between human and non-human mindedness.
Julia Tanney’s “Rational Animals” intersects Bar-on’s in several interesting respects. Bar-on draws upon both common sense and considerations from evolutionary biology to make her case against the continuity
skeptic. When it comes to how to think about animal minds, Tanney
wants to stress the importance of common sense as well. But, unlike Baron, she also brings out how common sense both infuses emerging fields
such as dog science and to a certain extent stands in tension with some
of the remaining Cartesian dogmas about the mind that those fields have
inherited. In focusing on animals who share our homes and thus are participants in some of our most treasured (and intimate) normative practices, her paper asks us to give philosophical weight to what we might
think of as the everyday phenomenology of our lives with these creatures.
In particular, drawing on Ryle and Wittgenstein, Tanney argues that mental predicates are internally related to the thick descriptions by which
we describe the performances that puzzle us. This position, which is as
robustly anti-behaviorist as it is anti-Cartesian, invites a re-examination
of the similarities and dissimilarities between species of rational animals.
If continuity skepticism as Bar-on depicts it shows a marked, perhaps
unwarranted, tendency to downplay the extent to which animals are genuinely minded, Stina Bäckström’s “Expression and Human-Animal Continuity” counters that defenders such as Bar-on and Alice Crary of what
Bäckström terms the “continuity view” nevertheless overlook crucial distinctions between animal and paradigmatically human minds. After recasting the central commitments of the continuity view, Bäckström adduces
certain key passages in Wittgenstein that form the core of an argument
for that view. In particular, these serve to bring out how certain temporal
aspects of language together with connections between non-verbal and verbal expression suggest a kind of mentality that goes beyond anything we can
reasonably ascribe to creatures without language. Finally, drawing on different lines of thought, one in John McDowell the other in Wittgenstein, Bäckström suggests that the concept of “expression”, a concept central to many
continuity view arguments, cannot necessarily do the work required of it.

IV. Naturalism and Meta-Philosophy
Benedict Smith’s essay takes on the essential task of attempting to place
the subject of Wittgenstein and naturalism into an historical context.
Commentators such as Bernard Williams have portrayed Wittgenstein’s

Introduction

9

conception of philosophy as ‘exclusively a priori’. Yet this cannot account
for the central role he accorded to observing and describing the variety
of our practices in the context of everyday life. To provide an alternative
view Smith compares Wittgenstein’s practice of philosophy with that of
Hume. Humean naturalism is sometimes assumed to have paved the way
for scientism. But, in stark contrast, Hume’s work is also cited as inspiration for ‘subject naturalism’ as it has figured in Price’s important work on
naturalism. There are of course important differences between Hume and
Wittgenstein (and between each of them and Price). But recognizing the
elements arguably shared by Hume and Wittgenstein can provide a better way to characterize the latter’s naturalistic approach to understanding the nature of philosophy and ourselves. This helps clarify the role
Wittgenstein gave to human practice and experience and demonstrates
that his attitude to the method and subject matter of philosophy was not
‘exclusively a priori’.
Eugen Fischer examines how the ‘Warrant Project’ in current experimental philosophy is promoting a new kind of methodological naturalism that he argues is fundamentally consistent with the so-called
“therapeutic” thrust of Wittgenstein’s meta-philosophy. While traditional
or first-order methodological naturalists seek to address philosophical
problems about a topic X (say, the mind, perception, or knowledge)
by building on scientific findings about X, Fischer argues that the new
meta-philosophical naturalism by contrast invites us to contribute to the
resolution of philosophical problems or debates about X by turning to
scientific (psychological) findings about the way we think about X—in
general or when doing philosophy. Fischer claims that the resources this
new naturalism encourages us to use can provide an empirical vindication of key aspects of Wittgenstein’s conception of philosophy and give
us an empirically successful diagnostic theory that can guide a ‘therapeutic’ philosophical practice. The paper develops this hypothesis by
showing how concepts and findings from psycholinguistic research on
routine comprehension processe scan be used to facilitate a ‘diagnostic’
resolution and ‘therapeutic’ treatment of philosophical problems which
arise when those processes misfire.
Jonathan Knowles examines some much discussed recent work by Huw
Price and Paul Horwich. In different ways, each of these philosophers has
portrayed themselves as anti-representationalists and anti-metaphysicians,
views that they claim to be inspired by or as having affinities with Wittgenstein’s later philosophy. They differ, or would appear to differ, however, with respect to the question of naturalism. As is well known, Price
subscribes to a form of naturalism that he terms “subject naturalism”,
which purports to break with more traditional, and metaphysical, forms
of “object naturalism”. Prices’s outlook is closely tied to a global expressivist view of language he finds in the later Wittgenstein. Horwich, on
the other hand, apparently rejects naturalism and instead advocates what

10 Thomas Raleigh and Kevin M. Cahill
he takes as a Wittgenstein inspired rejection of theory in philosophy. In
his paper, Knowles examines what these commitments amount to for the
respective philosophers, critically evaluates the overall views they articulate, and presents his own assessment of (the interrelationships between)
representationalism, metaphysics, naturalism, and science.
Wittgenstein’s attitude to the discipline of philosophy was notoriously
ambivalent. He left (or tried to leave) the profession on more than one
occasion and was not averse to advising students of the subject to do
something “useful” instead. This was partly a result of his understanding
of just what philosophy is; put a bit crudely, Wittgenstein saw philosophy
as either metaphysics or critique of metaphysics. And once the Western
metaphysical tradition had been shown up as so many multifarious illusions, what is left for the philosopher, let alone the profession as a whole,
to do? This is the question that Bjørn Ramberg takes up in his contribution. At present, perhaps the single most pressing problem in mainstream
metaphysics would appear to be the question of the place of mind in the
natural world: work must be done to mark out or make up appropriate
space. By contrast, thinkers such as Richard Rorty, Huw Price, and Philip
Kitcher, who fall under the umbrella of what Ramberg terms “Philosophy Critical Pragmatic Naturalism” (or PCPN) think the centrality of the
“placement problem” itself is highly problematic. Yet Ramberg wonders,
what, then, is it that pragmatic naturalists actually do, once the debunking lessons are learned? Once naturalism does its work on representational metaphysics and philosophy of mind, can a pragmatist concern
with practice, with the real problems of real human beings, provide disciplinary direction? Or does it amount simply to a wide-open job description for scientifically informed intellectuals-at-large? By comparing the
overlapping but distinct responses of Rorty, Price, and Kitcher, Ramberg
provides us with at least a partial sketch of a post-Wittgensteinian philosophical landscape deflated of its traditional metaphysical ambitions.

References
Cavell, Stanley. The Claim of Reason: Wittgenstein, Skepticism, Morality, and
Tragedy. Oxford: Clarendon Press, 1979.
Crary, Alice. “Wittgenstein and Ethical Naturalism.” In Wittgenstein and His
Interpreters: Essays in Memory of Gordon Baker, edited by Guy Kahane, Edward
Kanterian, and Oskari Kuusela, 295–319. Oxford: Blackwell, 2007.
Frege, Gottlob. “Der Gedanke.” Beiträge zur Philosophie des deutschen Idealismus 1:2 (1918): 58–77.
Gross, Steven, Tebben, Nicholas, and Williams, Michael, eds. Meaning Without
Representation. Oxford: Oxford University Press, 2014.
Hutto, Daniel D. and Satne, Glenda. “The Natural Origins of Content.” Philosophia 43:3 (2015): 521–536.
Ismael, Jennan. “Naturalism on the Sydney Plan.” In Philosophical Methodology:
The Armchair or the Laboratory? edited by Matthew C. Haug, 86–103. Oxon
and New York: Routledge, 2013.

Introduction

11

Jackson, Frank. From Metaphysics to Ethics. Oxford: Clarendon, 1998.
Lewis, David. “New Work for a Theory of Universals.” Australasian Journal of
Philosophy 61:4 (1983): 343–377.
Livingston, Paul M. “Naturalism, Conventionalism, and Forms of Life: Wittgenstein and the ‘Cratylus.’ ” Nordic Wittgenstein Review 4:2 (2015): 7–38.
Macarthur, David. “Liberal Naturalism and Second-Personal Space: A NeoPragmatist Response to the ‘Natural Origins of Content.’ ” Philosophia 43 (2015):
565–578.
———. “Taking the Human Sciences Seriously.” In Naturalism and Normativity, edited by M. De Caro and D. Macarthur, 123–141. New York: Columbia
University Press, 2010.
Maddy, Penelope. The Logical Must: Wittgenstein on Logic. Oxford: Oxford
University Press, 2014.
McDowell, John. Mind and World. Cambridge: Harvard University Press, 1994.
McGinn, Marie. “Liberal Naturalism: Wittgenstein and McDowell.” In Philosophical Methodology: The Armchair or the Laboratory? edited by Matthew
C. Haug, 62–85. Oxon and New York: Routledge, 2013.
Monk, Ray. Ludwig Wittgenstein: The Duty of Genius. New York: The Free Press,
1990.
Pears, David. “Wittgenstein’s Naturalism.” The Monist 78:4 (1995): 411–424.
Price, Huw, Blackburn, Simon, Brandom, Robert, and Horwich, Paul. Expressivism, Pragmatism and Representationalism. Cambridge: Cambridge University
Press, 2013.
Wittgenstein, Ludwig. Philosophical Investigations. Translated by G.E.M. Anscombe. 3rd ed. Oxford: Blackwell, 2001.

Part I

Varieties of Naturalism

1

Wittgenstein and Naturalism
Paul F. Snowdon

1
In considering to what extent, if any, Wittgenstein can be described as a
naturalist, we face two very difficult tasks. One is fixing the significance
of ‘naturalism’; the other, far more difficult, is determining what the
later Wittgenstein is telling us. Anyone engaging with this issue must be
extremely hesitant. But such difficulty should not silence us, and I want
to propose three claims. The first (I) is that it is true, and illuminating, to
describe Wittgenstein as a naturalist. ‘Naturalism’, that is, is an appropriate label to apply to his later philosophy. Second, (II) we should look with
approval on what can be seen as some of the basic negative elements in
his naturalism. Third, (III) Wittgenstein’s naturalism is, in a sense, overly
restrictive. What is the relation between these proposals? (I) is for present
purposes the most basic claim, combining an interpretation of Wittgenstein’s philosophy, together with an interpretation of what ‘naturalism’
can be taken to be, so that the latter term applies to what we can read
Wittgenstein as offering us. (II) and (III) are evaluative comments, part
favorable, part critical, on some details in Wittgenstein’s supposed view.
These evaluations can be considered even if (I), as I understand it, is
rejected.
Now, my discussion is not going to be quite as tidy as the summary I
have just given might lead you to expect. I want to start, though, by making some remarks about the employment of ‘ism’s, in philosophy. ‘Naturalism’ is just one of many ‘ism’s that philosophers use. It is quite obvious
that there are risks to employing these terms. For some reason or other,
in philosophy such terms acquire different interpretations as their use
spreads. Employing them, therefore, risks misunderstanding. It is, then,
necessary to say, fairly soon, what the term ‘naturalism’ as employed here,
means. However, one point of view in this debate is that any such ascription has to be wrong, since whatever exactly ‘naturalism’ means, it must
stand for some philosophical theory, and, as we supposedly know, Wittgenstein is against philosophical theories. Part of what needs scrutinizing
is how solid this popular line of thought is.

16

Paul F. Snowdon

2
I want to start by clarifying talk of ‘naturalism’. There are, I want to
suggest, three basic aspects that are involved in clarifying talk of ‘naturalism’. The first is to specify the domain D, about which theories
in some sense are being proposed. The second is the range of features
which are appealed to in the theoretical proposals and which merit
being called ‘natural’. We have then something that can be called a ‘naturalistic’ account of D. But there is a third aspect. I have talked so far of
theories of D, without specifying what sort of theory is being proposed.
One sort of theory, and perhaps the standard one when ‘naturalism’ is
talked of in the present, is what philosophers would call ‘reductive’.
The idea is that facts and features of the D-sort (ethical facts, or psychological facts, are standard cases) can be reduced to, come down to,
are nothing over and above, facts or features in the specified natural
domain.1 Now, the term ‘reductive’, despite its popularity, is, in a way,
misleading. It suggests that some theory or account ‘reduces’, which is
to say, in some way diminishes or subtracts from the analysed phenomenon. But no theory can be both true to a phenomenon and somehow
or other count as reducing it. ‘Reductive’ here means—displaying the
type of elements which make up, or constitute, the phenomenon. (In
fact, ‘constitutive’ is a better label than ‘reductive’.) Second, a successful ‘reductive’ approach here need not proceed by way of proposing
equivalents in the ‘natural’ vocabulary to sentences in the language of
the domain D, or by specifying in detail the natural grounds for the
facts being analyzed. Rather, the theory simply needs to make it plausible that the natural range specified in a general way is what grounds
or constitutes the truths of the domain. A theory of this sort I shall call
‘naturalismR’. So the question we face is whether Wittgenstein can be
called a ‘naturalistR’.
There is, though, another element that is relevant to understanding
this interpretative issue. To describe Wittgenstein as a naturalist is not
to say that we can find in the text an explicit endorsement of that idea
(though something close to that might be available in the case of meaning), but, rather, to say that the direction of his discussion, with its hints
and examples, can be fitted onto or around this general idea. The debate
has, therefore, to be acknowledged to be a somewhat messy one.
Although I am trying to make a case for saying Wittgenstein can be
counted as a naturalistR it would, in a more comprehensive treatment,
be worth exploring whether there are other possible (and non-reductive)
senses in which Wittgenstein might be called a naturalist. But I cannot
explore that here.
1 I here simply employ expressions which are standardly assumed to pick out well enough
what reduction is.

Naturalism

17

3
I would like though, before I develop my own interpretation, to engage
briefly with one distinguished Wittgenstein interpreter, namely David
Pears, who unhesitatingly applies the term ‘naturalism’ to Wittgenstein’s
later philosophy.2 Pears says; “There are several kinds of philosophical
naturalism and one of their leading ideas is that the right method in philosophy is not to theorize about things but to describe them as we find
them in daily life. Wittgenstein’s later philosophy is evidently a naturalism
inspired by this idea”.3 What Pears seems to be suggesting here is that
in Wittgenstein’s thought his anti-theoretical conception of philosophy is
the ground or inspiration for his naturalism. What that thought, which
certainly has its point, does not do is to illuminate the sense in which it
is legitimate to call Wittgenstein a naturalist. One might also add that if
his naturalism flows from his anti-theoretical philosophy, it can then be
asked; what does his anti-theoretical conception flow from? What Pears
then proceeds to do is (in section II) to give an account of Wittgenstein’s
treatment of rule following, and (very briefly) in section (III) of sensation
language, with the overall critical aim of proposing that Wittgenstein’s
naturalism “should not have been restricted to the therapeutic treatment
of myths that we generate when we reflect on our own language and
thought”. Rather, he asks, “Ought it not to have ranged more freely across
the border between philosophy, conceived in that way, and science?”4
Why does Pears say this? He develops his criticism against an account
of Wittgenstein’s discussion of rule-following. Pears reads this as attacking two views, which he calls ‘realism’ and ‘conventionalism’. It does not
matter for our purposes here what these two views are. Against the background according to which Wittgenstein has shown they are wrong Pears
says; “However, it might be doubted whether this is the point, or kind of
point at which the inquiry should terminate . . . But a complete treatment
of his illusion would surely require some account of the way in which
the word ‘intention’ was introduced into language . . . The interlocutor’s
mythical inner object may be eliminated by Wittgenstein’s treatment, but
not, I think, his feeling that there is something here which he still does

2 See David Pears, ‘Wittgenstein’s Naturalism’, The Monist Vol 84 No 4 (1995): 411–424.
I am very grateful to Bill Child for drawing my attention to this interesting paper.
3 Ibid., 411.
4 Ibid., 412. This quotation raises a number of questions which I shall not take up. Is Pears
envisaging that the direction Wittgenstein should have gone is to do philosophy as he
wanted to do but also to have crossed over into science? This is doubly puzzling. Pears
seems to neglect the possibility of philosophical theories. Also, it is hard to see how it is
possible to combine philosophy conceived of by Wittgenstein with crossing the border
to science. But also puzzling is that if the source of Wittgenstein’s naturalism is his antitheoretical approach to philosophy, would not adopting a more theoretical approach
undermine the motivation for his naturalism?

18

Paul F. Snowdon

not understand”.5 Pears mentions intentions at this point because he likens understanding an instruction that one gives to possessing an intention. According to him, after Wittgenstein’s critical discussion, we are left
finding intentions mysterious and so need something more. Pears then
sketches a positive theoretical account of what an intention is, or rests
on.6 The point of Pears’s discussion seems to be to oppose Wittgenstein’s
attitude to philosophical theorizing.
What should we say, briefly, about the critical direction of Pears’s argument? Let us assume to begin with that Pears is right that the rulefollowing discussion successfully attacks two views (of rule following),
those of realism and conventionalism. Now, it is obvious, although this
is not quite how Pears puts it, that in the absence of any proof that there
can be only two possible theories of rule-following it cannot be a consequence of what Wittgenstein has argued for that we should abandon the
search for a theory of what rule-following is. And, surely, Wittgenstein
does not show there can be only two theories.
This completely sound general point, however, will not convince Wittgenstein sympathizers. Thus, did Wittgenstein really think that the antitheoretical approach to rule-following is primarily supported by refuting
two theories of rule-following? Or does he think that it is supported by
more general considerations about philosophy adduced elsewhere? Further, is Pears even right in supposing that the rule-following considerations amount to a refutation of two theories? His criticism depends on
that reading, which is far from obviously correct.
Pears points out three other things. The first is that reflective people
engaging with Wittgenstein’s discussion will remain puzzled about the nature
of rule-following and still want some positive theoretical account.7 Again,
this is true. However, in the context of evaluating Wittgenstein’s antitheoretical conception it is not really clear what weight can be attached
to such theoretical yearnings. They are probably very hard to extirpate,
but that need not make them respectable. The second is that Wittgenstein
himself does appeal in some of his discussion to the postulated existence
of certain pre-reflective dispositions.8 The point of this for Pears is perhaps twofold. He could quite rightly say that appealing to such postulated dispositions is clearly an appeal to theory. He also himself thinks
that it introduces a theoretical model or template which can be usefully
appealed to—it is part of a good theory. Now, I think that it is hard not
to agree with both the points that Pears seems to be making. Wittgenstein
appears not to be consistently anti-theoretical, and he, perhaps, does not
appreciate the importance in a general way of dispositions. Third, Pears
5
6
7
8

Ibid., 418.
See Ibid., 418–420.
See Ibid., 418.
See Ibid.

Naturalism

19

himself sketches a theory of intention.9 It would take us in the wrong
direction to analyze Pears’s proposal. All I shall say in relation to Pears’s
project is that Wittgenstein himself regards talk of intentions as deeply
puzzling and worth considerable attention. Wittgenstein’s approach,
though, conforms to his own conception of what can, as it were, be
achieved by philosophy.
My main comment, though, on Pears’s article is that he is more concerned
to say what is dubious about what he calls Wittgenstein’s naturalism than
explaining why or showing that Wittgenstein should count as a naturalist.
He is, we might say, more relevant to my claim (III) than to claim (I). This
means that we have to pursue the central tasks unaided by Pears.

4
At this point, the central objection to calling Wittgenstein a naturalistR
deserves expression. It will be said that Wittgenstein can hardly count as
a naturalist since he is against philosophical theories (of any character),
and naturalismR is a philosophical theory.10 Now, I do not want at this
stage to try to respond to this objection. Rather, I shall try initially to
make the best positive case that I can (at least in this context) in favor of
describing Wittgenstein as a naturalist.
What, then, is the domain for which Wittgenstein is supposed to be
giving a ‘naturalistic’ account? The answer, I think, is that Wittgenstein is
fundamentally a philosopher of language; his focus is on language, what
underpins it, and the psychological elements that come with language.
Of course, Wittgenstein does not simply propose a theory of language.
What is clear about his later writings is that he turns his attention to
different subject matters and what he sees as their distinctive languages.
For example, he scrutinizes our talk about sensations, and our talk about
knowledge and certainty, and our talk of mathematics, and so on, grappling with all of them. In consequence he counts as a philosopher of
experience, of knowledge, and of mathematics, etc. However, in his later
writings, he also and centrally writes about language itself and about the
understanding of language, and I think it is appropriate, therefore, to
describe him as, in some sense, developing an account of language (and of
understanding). I am thinking here particularly of the early sections in the
Investigations (§ 1–43), concerned with meaning, and the sections about
understanding and rule-following (§ 138–242).11

9 Ibid., 418–419.
10 As Thomas Raleigh pointed out to me, this line of thought infers from Wittgenstein’s
anti-theoretical stance that he is not a naturalist, whereas Pears seems to explain Wittgenstein’s naturalism as arising out of his anti-theoretical stance.
11 No doubt these citations reveal that my own reading of Wittgenstein does not range
over all the highways and byways of his vast corpus, but is rather focused on the central

20

Paul F. Snowdon

If we assume for the moment that we should interpret Wittgenstein as
aiming to locate the factors which are constitutive of the phenomena he
is investigating, roughly meaning and understanding, what, in brief, is the
reason for thinking that his constitutive account is naturalistic? The brief
answer is that Wittgenstein appeals to elements which clearly deserve to
be called ‘naturalistic’. They are manifestly elements in the natural world.
Thus, Wittgenstein famously summarizes his general conclusion about
meaning in section 43; “For a large class of cases—though not for all—in
which we employ the word ‘meaning’ it can be defined thus: the meaning
of a word is its use in the language”. Surely, Wittgenstein means by ‘use’
the kinds of interpersonal interactions which he brings to the fore right
in section 1, to contrast with the so-called Augustinian conception. I have
in mind the famous shopkeeper case, and the subsequent variations on
it. Wittgenstein is claiming that totally natural occurrences of this sort
should be the core of an analysis of meaning. We can add that a typical Wittgensteinian remark is; “Commanding, questioning, storytelling,
chatting, are as much a part of our natural history as walking, eating,
drinking, playing”.12
In connexion with so-called rule-following here is a much quoted
remark. “Let me ask this: what has the expression of a rule—say a
sign-post—got to do with my actions? What sort of connexion is there
here?—Well, perhaps this one: I have been trained to react to this sign in
a particular way, and now I do so react to it” (§198). On the face of it,
here Wittgenstein is picking out a clearly natural occurrence—training
and reactions—as indicating how to answer the initial question.13
Is it right though to regard Wittgenstein as engaged in what I have
called reductionR, locating the phenomena which constitute or ground
meaning and understanding? This is a central question, but here I cannot
do more than select some evidence that relates to it. In the case of meaning the most obvious point is that Wittgenstein seems to say that meaning
is use. What meaning comes down to is use. This looks like a reductiveR
proposal.14
Now, there is a difference, it seems to me, between the meaning cases
and the understanding (and rule following) case, which is that Wittgenstein does not provide in the latter case anything close to the general
and standard parts of his writings. It is, though a generalization that seems reasonable
that writers usually put what they see as most important first, and in the Investigations
that is language (and perhaps the nature of philosophy).
12 Ludwig Wittgenstein, Philosophical Investigations (Oxford: Blackwell, 2001), §25.
13 One thing to note about this remark is that Wittgenstein alludes to training, which is a
natural interactive process, and one that he regards as an important element amongst
the grounds of understanding.
14 If it is felt that this neglects Wittgenstein’s own restriction of the use proposal to a ‘large
class of cases’, the concessive response would be to say that Wittgenstein’s proposal
looks reductiveR for a large class of cases.

Naturalism

21

claim about meaning and use in sec. 43. So the reason to claim that his
approach is reductiveR in relation to rule-following is harder to support. However, the basic question of the passage is, surely; what is understanding? And his discussion considers possible answers to that question.
One possible answer that he rejects is, of course, that understanding is
an experience. He says “Try not to think of understanding as a ‘mental
process’ at all”.15 Again, when he considers reading his interest is in the
issue what reading is.16 So, the focus is throughout a constitutive one.
Further, when Wittgenstein in his discussion engages with the question
why we say someone understands something what he tends to mention
are that there has been training in the background and earlier uses of
the terms.17 This is close to saying that the correctness of the ascription
of understanding is grounded in such elements, which seem to be purely
natural ones. Another feature of his discussion is to emphasize that we
simply act when told to do things. He says, “How can he know how he is
to continue a pattern by himself—whatever instruction you give him?—
Well, how do I know?—If that means ‘Have I reasons?’ the answer is : my
reasons will soon give out. And then I shall act, without reasons”.18 Now,
it is unclear what is meant by this remark, but it is not unreasonable to
hear in it the idea that understanding rests on and presupposes a level
of operations that get carried out which are not themselves guided by
understanding, but which must be there for understanding to be present.
On this view, understanding involves operations which are pre-cognitive.
This is, surely, close to a naturalistic conception. It is true that by the end
there is no clear statement of the constituents, but that is, I am inclined
to think, because Wittgenstein thinks there is with understanding, given
its complexity, no easy way to specify the relevant constituents. However,
the elements that he tends to mention are clearly natural ones.
So, what is indicated is that Wittgenstein locates the real basis for
meaning and understanding in such obviously natural phenomena. He
seems to me to be offering a naturalistic account, or theory, of language.
This is a very simple argument for claim (I)—not of course for claims (II)
or (III).
Although this is not a theme being talked about here, it seems to me
that as well as being a naturalist we can also attach the label ‘externalism’
to Wittgenstein’s approach to meaning. He is what I would call an ‘it goes
without saying externalist’. Giving examples of what grounds meaning
Wittgenstein simply and unhesitatingly brings external things—such as
apples, colors, bricks, etc.—into his account. Language gets it meaning
15 Ibid., § 154.
16 The reading section is of central importance in understanding Wittgenstein’s approach
to understanding.
17 See, for example, ibid., § 179.
18 Ibid., § 211.

22

Paul F. Snowdon

in relation to the environmental items that its use and users engage with.
By calling this an ‘it goes without saying’ version, I intend to contrast it
with the type of externalism defended by Putnam (and perhaps Kripke)
which attempts to derive externalism from certain theoretical premises.
Wittgenstein, by contrast, takes it as obvious.19

5
The above argument is, of course, very simple and sketchy, indeed quite
clearly too simple to generate conviction. But I want now to set out and
engage with an argument the other way—an argument designed to show
that Wittgenstein is not a naturalist. This argument starts from a way
of reading him that has seemed plausible to me.20 The basic idea is that
Wittgenstein is first and foremost a negative philosopher. This means that
his primary aim is to establish conclusions of the form Not (P), or, perhaps, do not believe that P. Usually, he is saying something like; “Here is
a conception of this phenomenon that we should reject”. It may be that
the term ‘negative’ has itself a rather negative tone, and another, perhaps
more positive, way of expressing the same point is to say that Wittgenstein is engaged in demythologizing. He is uprooting myths.
There are two main general reasons to interpret him this way. The first
is that it is the obvious way to read him given the manifest content of his
discussions. One example will have to suffice. In section 243, Wittgenstein raises the famous question about whether someone can introduce
their own private terms for their sensations. In the next sections (to about
§315) Wittgenstein is clearly attacking a conception of private experience
according to which the answer to this question would be ‘yes’. He aims to
uproot that conception by arguing (amongst other things) that actually
the correct answer is ‘no’. The discussion is primarily negative, to such
an extent that most readers would be hard pressed by the end of it to
say what positive conception of sensations and our way of talking about
them Wittgenstein favors.
The second reason for regarding Wittgenstein as a negative philosopher is that he advances a conception of (good) philosophy which seems
to represent it as highly negative. Consider his remark “The philosopher’s
treatment of a question is like the treatment of an illness”.21 This is one of
his most famous metaphors. What does it mean? The obvious suggestion
19 Anyone sympathetic to this characterization cannot object to applying a philosophical
‘ism’ to Wittgenstein.
20 I endorsed this reading in: Paul Snowdon, ‘Private Experience and Sense Data’, in The
Oxford Handbook of Wittgenstein, eds., O. Kuusela and M. McGinn (Oxford: Oxford
University Press, 2011), 402–428.
21 PI §255.

Naturalism

23

is that it represents the goal of proper philosophy as eliminating something bad, an ‘illness’, health here not representing having a better theory,
but rather solely being clear of the confusion, that is of the bad health.
The strong suggestion is that returning to health is not improving one’s
theory of the world, nor is it making a (true) addition to one’s cognitive
understanding, but is rather eliminating the activity which amounts to
being ill. Consider also §309; “What is your aim in philosophy?—To
show the fly the way out of the fly bottle”. This compares bad philosophy to being trapped like a fly and the aim is to release the fly from the
trap. The idea of a trap here is that of accepting a bad view, and the aim
is to release one from that view. Here we again have the idea that good
philosophy acts to free you from traps, simply releasing you. (We do,
of course, in considering these remarks of Wittgenstein face one major
difficulty posed by his work, that of determining the significance of his
omnipresent metaphors and similes.)
In the light of this ‘negative’ reading of Wittgenstein, some philosophers who accept it are tempted to propose an argument which might be
formulated as follows. 1] Wittgenstein does not mean to alter or criticize
or really add to any of our ordinary thinking. 2] If Wittgenstein had
proposed that naturalism about meaning and rule-following were correct
he would be recommending an alteration to our thinking; therefore, 3]
Wittgenstein cannot be proposing naturalismR.22
I want to set up some resistance to this argument. First, the characterization of Wittgenstein’s practice of philosophy needs some modification if it is to be true to his actual practice. Wittgenstein clearly offers
general hypotheses for consideration. Thus, for example, he links meaning to use, he makes conjectures about how sensation words gain their
meaning, he stresses the role of context in judgment, and so on. Another
example of what is clearly positive theorizing is his idea that something
about language is what makes us do bad philosophy. A related example
is his evident claim that there is something inadequate about traditional
philosophy. That is obviously theoretical. So Wittgenstein invites us to
consider and presumably accept some manifestly theoretical proposals.
22 Here are, perhaps, two examples of more or less this structure of argument. In John
McDowell’s, ‘Wittgenstein on Following a Rule’, in Mind, Value and Reality (London:
Harvard University Press, 1998), 221–262. McDowell opposes an interpretation of the
rule following passages that Wright proposed, which, according to McDowell, committed Wittgenstein to what McDowell calls ‘idealism’, by saying; “We may well hesitate to
attribute such a doctrine to the philosopher who wrote: ‘If one tried to advance theses
in philosophy, it would never be possible to debate them, because everyone would agree
to them” (PI §128) Again, William Child says that “the deflationary interpretation of
Wittgenstein’s discussion of rules is . . . in many ways more faithful to his philosophy
than the constructivist reading” (Child, Wittgenstein, 133). In fairness to Child, it needs
remarking that he himself is not entirely happy with the ‘quietist’ reading.

24

Paul F. Snowdon

Further, although he compares good philosophy to curing a condition,
his therapy in fact works by persuading you of some unobvious truth,
e.g. that something is a consequence of a claim, or that there are difficulties with a claim, acceptance of which contributes to one’s overall
theory. There is clearly a level in Wittgenstein’s practice where unobvious and so (in one sense) theoretical claims are being proposed for our
acceptance.
I am suggesting, then, that there are at least two levels in W’s actual
practice of philosophy where there are propositions and claims advanced
for assessment and possibly belief. He advances what are clearly general
theories about certain things. Second, in trying to achieve his negative
purpose he advances arguments containing (unobvious and new) claims
that are also offered as true.
There is another observation I want to make at this point. It is all very
well to hang the label ‘negative’ on Wittgenstein’s practice and also his
conception (or philosophy) of philosophy, but we should acknowledge a
contrast. The impression that we get from his philosophy of philosophy is
that it does not yield any theories (or theoretical claims) at all but rather
issues in the relief of not playing the philosophical theorizing game. We
can certainly call that a negative conception. But in his practice what is
negative are the claims that he is persuading us to accept. That is not
negative in the same way; it is, rather, theorizing to negative conclusions.
Theoretical claims can be negative.
Against this background of a more accurate characterization of Wittgenstein’s practice, I think we can weaken the above argument. There are,
as I see it, at least three reasons for not placing too much reliance on it.
The first, already brought out, is that despite his own characterization of
what he is doing as non-theoretical, leaving everything as it is, merely a
treatments of an illness, etc., it has to be acknowledged that in practice
Wittgenstein proposes unobvious new claims (theories of greater or lesser
generality) for our consideration. He is simply not totally ‘quiet’. Indeed,
offering negative conclusions is not being quiet. Second, if Wittgenstein
characterizes himself as leaving everything as it was, interpreting what
that amounts to entirely depends on Wittgenstein’s own view of how
everything was. When I declare on looking over my room that nothing
has been changed that clearly depends on my sense of how it was before.
So, Wittgenstein might think that what he is saying amounts to changing
nothing, but it is quite possible that he has completely failed to realize
how ‘things’ were. Wittgenstein’s assurance that he is changing nothing is
not self-certifying. Third, when Wittgenstein says that philosophy leaves
everything as it is he has to exclude philosophy itself. The practice of
(bad) philosophy and the convictions generated by it are to be excised.
The question then arises as to what Wittgenstein thinks the extent of
old bad philosophy is, since that he certainly does not aim to leave it as
it has been. Indeed, once raised this represents a very hard question for

Naturalism

25

Wittgenstein, since how does he know how widespread bad philosophy
(on his conception) has been? It remains quite possible for him to think
that it has been widespread, and so, maybe, many relatively common or
even quite usual opinions are the products of bad philosophy. The fact
that the extent of so-called ‘philosophy’ has no really discernible boundaries means we have no idea what Wittgenstein thinks it is legitimate to
challenge and what it is not legitimate to challenge.23
I hope that these remarks have to some extent taken the wind out of
the sails of the general argument.
If I were summarizing these remarks about Wittgenstein’s approach to
philosophy I would say, by way of explanation, that he espoused for reasons that I find hard to discern a deeply negative conception of philosophy, while being at the same time an outstandingly creative philosopher
or thinker and this creativity made it impossible for him to be true in his
actual practice to his preferred conception.

6
If these remarks do weaken the appeal of the general argument, there is
really nothing for it but to scrutinize and try to understand Wittgenstein’s
text, and to critically evaluate it. Obviously I cannot do that here, so, I am
simply going, rather abruptly, to propose a few unqualified things which
seem to me to fit my reading.
If you have a language within a group, it would seem that you must
have linguistic elements which have meaning, but you also must have
members of the group, the speakers and hearers, who employ the language with understanding. I have some inclination to say that in the earliest sections of the Investigations Wittgenstein is focusing, primarily, on
what meaning is, and in the rule-following sections is focusing, primarily,
on what understanding is. At least, I shall divide things up that way. This
goes along with the fact that Wittgenstein, in the Investigations, shows
far more interest in opposing imagistic accounts of understanding in the

23 An aspect of Wittgenstein’s attitude to allowing that philosophy can be challenged but
nothing else is his thinking about religion. In relation to that it seems that Wittgenstein
is opposed to what we might call philosophical criticisms of religion. But why is that?
Wittgenstein must think that religious faith cannot be regarded as having a philosophical origin, since if it does then it is not beyond criticism. However, even if we accept
that, Wittgenstein’s slogan that philosophy leaves everything as it is, does not imply that
faith is beyond criticism, unless we have to regard any criticism of faith as amounting
to philosophy, which is not allowed to alter anything. And we might very well wonder
why we should describe any general criticism of faith as ‘philosophy’. Why would the
criticism of religion amount to philosophy, but, say, the criticism of astrology, which, I
assume, Wittgenstein would not rule out in advance, not amount to that? Sorting out
these issues would take us too far away from naturalism, but they illustrate just how
obscure in many central ways Wittgenstein’s attitude to philosophy is.

26

Paul F. Snowdon

early parts of the rule-following section, rather than at the beginning of
his discussion of meaning.
I want now to fill out a little what it means to call Wittgenstein’s treatment of meaning and understanding ‘naturalistic’. There is for both main
topics a positive side and a negative side. On the positive side what has
be said is two things. The first is that the elements that Wittgenstein picks
out in his account merit being called ‘naturalistic’. The second positive
component is that Wittgenstein is picking them out not as features that
merely happen to be associated with language (meaning and understanding), but as elements that ground or constitute them. I have tried to make
a case for these claims. But there is a negative aspect to Wittgenstein’s
naturalism, which is that there are no traces in his account, and indeed
there is evidence of opposition to, the appeal to what I think we could
describe as one main non-naturalistic feature. The non-naturalistic feature that is conspicuous by its absence is what I shall call the proposition
or the thought, conceived as many philosophers, including for example
Frege, conceive of it. How are propositions/thoughts brought in and how
are they thought of? In ‘The Thought’, Frege regards thoughts as real
things which ‘belong(s) neither to [the] my inner world as an idea, nor yet
to the outer world of material, perceptible things’.24 They exist in their
own unchanging realm, what we might call a third realm. But, since they
have a fundamental role in language they have to relate in some way to
our own minds. As Frege puts it, “. . . It is advisable to choose a special
expression and the word ‘apprehend’ offers itself for the purpose”. So on
this conception, along with the realm of thoughts, there is a psychological
relation to that realm, that of apprehending (some of its) denizens. Now,
about this psychological relation, Frege says, “although the thought does
not belong to the contents of the thinker’s consciousness yet something in
his consciousness must be aimed at the thought”.
I want, in response to this, to claim three things. First, the ontology
of thoughts (and propositions), and their components, plus the hypothesized psychological relation to them is quite reasonably called ‘nonnaturalistic’. There is an obvious sense in which a theory of language and
of cognition which postulates such things and relations is postulating
things which are not part of the natural realm. Second, Wittgenstein’s
complete avoidance of them is deliberate, and should be counted as an
element in his naturalism. Third, we can read parts of his discussion of
understanding and rule-following as attacks by him on what such ‘nonnaturalistic’ models involve.
I want to develop this last remark. One major theme and direction of
focus in Wittgenstein’s discussion of rule following is ‘what comes before

24 Gottlob Frege, ‘The Thought’, in Philosophical Logic, ed., P. F. Strawson (Oxford:
Oxford University Press, 1967), 35. All other Frege quotations from the same page.

Naturalism

27

the mind of the understander’. Thus in §139, as his exploration is beginning, he says, “When someone says the word ‘cube’ to me, for example,
I know what it means. But can the whole use of the word come before
my mind, when I understand it in this way? Well, but on the other hand
isn’t the meaning of the word also determined by this use? And can’t
these ways of determining meaning conflict? Can what we grasp in a
flash accord with a use, fit or fail to fit it? And how can what is present
to us in an instant, what comes before the mind in an instant, fit a use?
What really comes before our mind when we understand a word?—Isn’t
it something like a picture? Can’t it be a picture?”25
Two things stand out about this passage. The first is that Wittgenstein
starts from the link between meaning and use which the earlier discussion of meaning aims to establish. So it can be said that what he is trying
to do is to think out a defensible account of understanding within the
framework of a ‘use’ theory of meaning. The second is that his initial, and
in fact enduring, focus is on what ‘comes before the mind’ of the understander. Now, that surely means what figures in the consciousness of the
understander. We can hear this discussion as impinging on Frege’s talk of
the apprehending of thoughts, which is what understanding is supposed
to be in that model, as registering in the consciousness of the understander. Wittgenstein is asking what actually is present in consciousness
when there is understanding.
What does Wittgenstein do with the question? Famously he considers the suggestion that what comes before the mind (consciousness) is a
picture—this is how we would describe an image—of the signified thing.
Putting it rather generally, Wittgenstein’s response is to argue that a picture before the mind does not fix a use; and if it does not fix a use its
occurrence cannot constitute understanding. As he says in §140, “What
is essential to see is that the same thing may come before our minds when
we hear the word and the application be different. Has it the same meaning both times? I think we shall say not”. But Wittgenstein also points out
that it is no better to model understanding as the occurrence before the
mind of a formula, because the same problem arises. As he says in §152,
“ ‘B understands the principle of the series’ surely doesn’t mean simply:
the formula ‘a = . . . ’ occurs to B. For it is perfectly imaginable that the
formula should occur to him and that he should nevertheless not understand”. In effect, both of these points are anti-imagistic. In one case, the
point relates to images of the signified thing or feature, and in the second
to images of a formula (or a bit of language). In the famous section 201
where Wittgenstein talks of the paradox, he is, as I read it, repeating
this last point employing the terminology of ‘interpretation’, by which he
means having a formula or a rule in mind. The general conclusion that

25 I have compressed this layout of this passage to save space.

28

Paul F. Snowdon

Wittgenstein draws at this point is this, “what this shows is that there is a
way of grasping a rule which is not an interpretation, but which is exhibited in what we call ‘obeying the rule’ and ‘going against the rule’ ”. A part
of this general conclusion is in effect that understanding is independent
of the contents of consciousness, and has to do with practice, and, as we
might say, our operations. So we can perhaps read him as saying that
there is no way to articulate the Fregean idea that this relation of apprehending an abstract thought registers in consciousness.
If this is not implausible how might the supporter of Frege react to
the problem? One way would be to retain the idea of apprehending a
thought, but deny that it needs to register in or involve consciousness. But
this seems to amount to postulating a relation to an abstract entity that
does not integrate into an account of the cognitive lives of people who
understand. We, as physically structured animals, have lives in which language plays a crucial role, and how can an apprehending of an abstract
thought impinge on that? A second way to respond is to suggest that
Wittgenstein has overlooked a mode of consciousness that is different
from any he considers, which is the conscious grasping of a thought.
Now, there is no simple response here, but something that Wittgenstein
himself says would look like a reasonable first response. In §192 he says
to someone, “you have no model of this superlative fact, but you are
seduced into using a super-expression”. So you can say that you consciously grasp a thought, but given our approach to consciousness, what
kind of thing is it, what model for it can we offer?
There is also another theme in Wittgenstein that relates to this dialectic, though I do not feel confident that I can expound it properly. What
Wittgenstein stresses is that there are things that we simply do when we
follow rules—operations that we perform. One famous remark is this;
‘How am I able to obey a rule?’—if this is not a question about
causes, then it is about the justification for my following the rule in
the way I do.
If I have exhausted the justifications I have reached bedrock, and my
spade is turned. Then I am inclined to say: ‘this is simply what I do’.
(§ 217)
He also says;
When I obey a rule, I do not choose. I obey blindly.
(§ 219)
Now, there are puzzling passages, and I do not feel confident as to their
significance, but a natural way to interpret them is that Wittgenstein is trying to convey that we need to recognize that in our cognitive operations,

Naturalism

29

there is a level where we cannot regard those operations as guided by or
responsive to understanding. They are operations we perform, but which
are below the level of operations of the understanding, but which constitute them as operations of understanding. Someone who is inclined
to say this will oppose the postulation of unconscious apprehendings of
thoughts as the basis of understanding. Rather, at the basic level, there
are operations which cannot be intellectually explicated, but which are
presupposed by intellectual operations.
I have been trying to substantiate the claim that Wittgenstein appeals
in his theories about language to what are legitimately called natural
elements—interpersonal interactions, and levels of operation—and he
avoids and denigrates the invocation of non-natural elements such as
Fregean thoughts and propositions. This looks like naturalism to me.
I have been rather relaxed in my employment of the natural/nonnatural distinction. In effect I have been assuming that without laying
down an explicit criterion demarcating them we can agree on how to
classify candidate elements which are proposed in theorizing about language into the natural and the non-natural. In particular, social interactions and non-intellectually guided operations are natural features,
whereas apprehensions of abstract thoughts and propositions are not.
Further, I tend to assume that we share a commitment to keeping the elements in our theories of ourselves, who are natural physical and social
creatures, to natural elements. We surely regard ourselves as the products of evolutionary processes in which natural complexities emerge
which generate our nature and capacities. And we cannot really include
the emergence of a capacity to apprehend thoughts of the philosophically postulated variety in that way. Naturalism in relation to language
and cognition is, I think, simply a basic commitment of how we must
understand ourselves.

7
Finally, I want to argue that we should be wary of some of the elements
in Wittgenstein’s version of naturalism.
What should our attitude be to Wittgenstein’s ‘use’ slogan? It is very
difficult to assess his remark—the reason being, I suggest, that the term
‘use’ is very indefinite. What exactly is ‘the use of a word’? There is one
use of ‘use’ which is very general and it is clear that that is not meaning—
thus a hammer has a use—but it does not have a meaning. Similarly—a
word might have a use—eg to open a gate—but that use is not having a
meaning. Further, the idea of use is so general that the slogan can cover
accounts of meaning that W is opposed to. E.g. one can describe the
role of triggering images as a use for words—but Wittgenstein would not
think that use is the meaning constituting use. There is another reading

30

Paul F. Snowdon

of talk of meaning as use that is trivial. Someone might say that meaning
is determined by use, since they are reading use as ‘what you use the term
to mean’. Clearly, meaning is determined by use in that sense. What that
does not illuminate at all though is what using an expression to mean
something actually amounts to, or, comes down to. The most obvious way
to interpret Wittgenstein’s suggestion is, as I have been doing, to allow
‘use’ to stand for the features that he presents in his examples. These are
predominantly the use of language is ones dealing with others in ways
which influence their behaviour. ‘Use’ stands for something interpersonal
and social. This is true of the example in §1, the five red apples example,
and in §2 the block, pillar, slab, etc. example, and in §8, where, as it were,
talk of numbers is added and so are demonstratives. It is very natural to
read Wittgenstein’s ‘use’ slogan as intending by ‘use’ such aspects. This
leads to us notice an aspect of Wittgenstein’s uses that stands out. They
are mainly, as one might say, imperatival. The use amounts to a speaker
getting a hearer to do something, such as hand him an apple or a slab etc.
Now, in thinking of imperatives, interpersonal responses seem a reasonable aspect to bring in, (I do not mean correct, but at least reasonable),
but it is far harder to make this seem plausible as a model of what we
might call descriptive meaning. If I say, ‘Bergen is very busy today’, there
is nothing the audience is expected to do in response to my remark. Nothing, that is, that is clearly related to the meaning of the remark. So, if ‘use’
means something like interpersonal responses there is no obvious application for the slogan to large central parts of language.26
One might point out something very odd about the five red apples
game. The hearer is triggered by the use of ‘red’ to look at a color chart
to determine the color needed, but this in effect means that he does not
understand the word ‘red’. Curiously, Wittgenstein gives a model of
meaning here that side-lines understanding. Another curious aspect of
Wittgenstein’s game is that it is one sided. In language, the speaker role
and the hearer role are present in the same individual. But in Wittgenstein’s model there is no such element. It is as if the shop keeper has
internalized the provider role and the shopper the demander role. Wittgenstein’s models are also over-automatic. Understanders of language
need not cooperate. There might be multiple demands on them, or they
might have run out of the required things. One cannot model meaning
on the presence of invariable routines, as Wittgenstein’s games involve.
Another source of things not running smoothly are mistakes of various
kinds. One cannot equate the shop keeper understanding the word ‘five’
with his giving five so and so’s. Maybe he mishears, maybe he miscounts,
maybe he simply picks up the wrong number by mistake. Such irregularities do not disturb meaning.

26 This paper was first presented in Bergen!

Naturalism

31

In the light of these worries I think that it is reasonable to be cautious about accepting the Wittgensteinian model of meaning, naturalistic
though it is.27
Now for some equally brief remarks about rule-following. It is far
harder to understand this discussion, which is complex and allusive,
but one move that Wittgenstein seems to make it to set himself against
employing the notion of a disposition in any central way in his account.
I want to query the apparent grounds he offers for this.
In §149 Wittgenstein is considering the quite natural idea that knowing the ABC should be thought of as a disposition, which explains the
so-called manifestations on occasions of that knowledge. Against this
idea Wittgenstein brings the point that to attribute a disposition requires
fulfilment of two criteria, one being what it does, but the other being
what aspect of a things construction grounds the disposition. His point is
that when we talk about knowing the ABC, we completely lack the latter
information. This argument invites the following response. We seem to
attribute dispositions to things without having any knowledge of structure. People said that sugar is water-soluble without knowing anything
about the structure of sugar (or indeed water). Indeed, part of the point
of talking of dispositions is, surely, to convey the powers of things where
we are ignorant of the composition of the things themselves.
I read another passage as opposed to the same suggestion. In §153
Wittgenstein is opposing an idea which he formulates as amounting to
the claim that understanding is something “hidden behind those coarser
and therefore more readily visible accompaniments”. Now, that way of
speaking fits the disposition idea. Against this Wittgenstein raises what
seems to be an epistemological problem. Wittgenstein says, “And how
can the process of understanding have been hidden, when I said ‘Now I
understand’ because I understood?!” In the next sentence Wittgenstein
insinuates that the idea that the state of understanding is ‘hidden’ is problematic. At least two points suggest themselves here. The first is that Wittgenstein’s own puzzles about our self-knowledge of understanding count
against simply saying that we know we understand simply because we
understand. How we know we are in that state is quite unclear. Maybe it
is like knowing we have a disposition. The second is that if there is some
sense in which a disposition is per se hidden, there is no obvious reason
to deny that the presence of understanding is not in the same way hidden. Whatever that notion of being hidden is it is simply not obvious that

27 It has to be acknowledged that these problems do not refute Wittgenstein’s approach,
but it would not be enough to respond by pointing out that his examples are intended to
be very simple, and hence will be over-simple. The question is whether we should allow
that they indicate where the core basis of meaning is to be found, and the difficulties
seem to me to indicate that they are dubious in that role.

32

Paul F. Snowdon

aspects of ourselves are not hidden from us. Clearly that dispositions are
hidden does not mean that we cannot easily get to know about them.
My response then is to suggest that Wittgenstein does not produce any
strong evidence that the standing condition of, for example, knowing the
ABC is not a disposition, or something disposition like. Nor does he provide any evidence that the theory of understanding should not attribute a
central role to the notion of a disposition.
It is worries along these lines that lead me to endorse (III).
I hope, then, that I have assembled some sort of defense of my three
claims.28

References
Child, W. Wittgenstein, London: Routledge, 2011.
Frege, G. ‘The Thought’ in Philosophical Logic, edited by P. F. Strawson, 17–38,
Oxford: Oxford University Press, 1967.
Kuusela, O. and McGinn, M. The Oxford Handbook of Wittgenstein, Oxford:
Oxford University Press, 2011.
McDowell, J. ‘Wittgenstein on Following a Rule’ in Mind, Value and Reality,
221–262, London: Harvard University Press, 1998.
Pears, D. ‘Wittgenstein’s Naturalism’, The Monist Vol 84 No 4 (1995), 411–424.
Snowdon, P. F. ‘Private Experience and Sense Data’ in The Oxford Handbook of
Wittgenstein, 402–428, edited by O. Kuusela and M. McGinn, Oxford: Oxford
University Press, 2011.
Wittgenstein, L. Philosophical Investigations, Oxford: Blackwell, 2001.

28 I am very grateful to Kevin Cahill for the invitation to speak at the workshop, and to
both him and especially Thomas Raleigh for their contribution to the experience and to
the discussion. I gained a lot from the general discussions there, and in connection with
my own paper I wish especially to thank Sorin Bangu, Bill Child, Harald Johannessen,
Alois Pichler, and Charles Travis. I am also very grateful for comments to Brian Garrett
and Arthur Schipper.

2

Wittgenstein’s Liberal Naturalism
of Human Nature
David Macarthur

science . . . brings confusions if we think that its results reveal to us what
the world is like.
—Rush Rhees1

Introduction
A recent commentator on Wittgenstein claims that “naturalism is fundamental to Wittgenstein’s philosophical method”.2 What is meant by
“naturalism” in this context? Like many of the key terms by which
philosophers define themselves, “naturalism” has no stable meaning in
the history or historiography of philosophy but is subject to a dizzying
degree of semantic flux. It is less the solution to a well-defined problem
than the site of philosophical controversy most especially concerning the
relation between philosophy and science; but it is also associated with
philosophical myths about science such as that all sciences posit causal
laws or that, in principle, they all ultimately reduce to physics.3 The difficulty of interpreting Wittgenstein as some kind of naturalist is further
complicated if we suppose, not unreasonably, that the term “naturalism”
refers to a philosophical doctrine or set of such, a claim which stands in
stark contrast to Wittgenstein’s unambiguous intention to philosophize
without philosophical doctrine.4
I shall assume at the outset that the prima facie tension between the idea
that Wittgenstein is committed to some form of naturalism (conceived as
doctrine) and his aim to philosophize without theses or a priori dogmas is
to be relieved by denying that Wittgenstein’s naturalism is to be thought of
1 Rush Rhees, Without Answers (London: Routledge, 1969), 8.
2 Marie McGinn, “Wittgenstein and Naturalism”, in Naturalism and Normativity, eds.
Mario de Caro and David Macarthur (New York: Columbia University Press, 2010),
347.
3 I have discussed the relation between metaphysical myths and science in “Taking the
Human Sciences Seriously”, in Naturalism and Normativity.
4 See, e.g., Ludwig Wittgenstein, Philosophical Investigations, 4th ed., trans. G. E. M. Anscombe, Peter Hacker, and Joachim Schulte (Oxford: Blackwell, 2009), §109, §126.

34

David Macarthur

as doctrinal. But, then, how is it to be characterized? Given his uniqueness
and originality, saying whether he is a naturalist in some interesting sense
will be no less difficult than characterizing his philosophy as a whole. The
main focus of the discussion will be on Wittgenstein’s post-Tractatarian
outlook but it should be noted that he held a consistent line on the relation
between science and philosophy, in his sense, throughout his life.

Scientific Naturalism and Philosophy
Though it is employed in a wide variety of ways, the term “naturalism” as
it is most commonly used in contemporary philosophy refers to the popular program of scientific naturalism. According to this general outlook,
science provides, or at least promises to provide, a complete metaphysics
and/or epistemology—and what counts as “science” for these purposes
can vary considerably along a spectrum from physics at one extreme to
a broad range of natural and social sciences at the other. Ontology is
provided for by the claim that a complete account of all that exists is provided by the composite ‘scientific image’ of the world which comes from
combining the images associated with physics, chemistry, biology, and so
on; and epistemology is provided for by the plural ‘scientific method’, the
loose assortment of broadly empirical methods of the explanatorily successful sciences. Here, knowledge is identified with scientific knowledge.5
Let us compare this outlook with Wittgenstein’s methodological remarks
about his own philosophy (which I shall henceforth, for convenience,
italicize as philosophy) such as:
it is . . . essential to our investigation that we do not seek to learn anything new by it. We want to understand something that is already in
plain view. For this is what we seem in some sense not to understand.
It was correct that our considerations must not be scientific ones . . .
And we may not advance any kind of theory. There must not be anything hypothetical in our considerations. All explanation must disappear, and description alone must take its place . . . The problems are
solved, not by coming up with new discoveries, but by assembling
what we have long been familiar with . . .
Philosophy just puts everything before us, and neither explains nor
deduces anything.—Since everything lies open to view, there is nothing to explain. For whatever may be hidden is of no interest to us.
5 In the Tractatus, Wittgenstein seems to have adopted this scientistic conception of
knowledge himself whilst at the same time denying that philosophy is a matter of knowledge or science. The ground of this view is his identification of “what can be said”
with “propositions of natural science”. No doubt this is one of the sources of his influence on the development of logical positivism. Ludwig Wittgenstein, Tractatus LogicoPhilosophicus, trans. C. K. Ogden (London: Kegan Paul, 1922), 6.53.

Human Nature

35

The name “philosophy” might also be given to what is possible
before all new discoveries and inventions.
The work of the philosopher consists in marshaling recollections for
a particular purpose.6
It is at least clear that Wittgenstein’s naturalism, if we should call it that,
is not at all the common scientific kind. Indeed, his philosophy seems
designed to combat any attempt to answer philosophical questions with
scientific materials or methods. More than that, it provides one with critical tools for the strongest opposition to scientific forms of philosophy—
scientific naturalism included—in the literature.
Philosophy concerns “what is in plain view”, “what we have long been
familiar with”, “what is possible before all new discoveries”, and various “recollections”—by which he means our ‘knowledge’ of language,
or, more specifically, our ‘knowledge’ of the logic of language. However
we are supposed to understand this unprecedented kind of knowledge—
which cannot be assimilated to either traditional a priori or empirical
knowledge—we can at least say that it is available, or it makes itself
manifest, without performing any observations or experiments.7 Wittgenstein distinguishes philosophy sharply from the empirical sciences,
which are precisely attempting to discover new knowledge about objective causal relations between things by observation and experiment in
order to explain, predict and control natural phenomena.8 The paradigm
is natural science, which typically involves hypothesizing ‘unobservable’
entities (e.g. electrons, quarks, gravity waves, chemical bonds, genes, etc.)
to causally explain correlations in observable phenomena. Consequently,
‘science’ in the relevant sense concerns a desire for new discoveries about
the hidden, the strange or remarkable—precisely the opposite direction
of orientation of philosophy which aims to return us to the manifest
world of our everyday lives, but with newly opened eyes.
Scientific naturalism treats the success of modern science as having metaphysical significance, so that science becomes a general theory of being
and knowing (e.g. it has been thought to imply: (1) strict physicalism—the
view that there is nothing over and above the entities posited by physics;
(2) causal fundamentalism—the view that there is a single, causal order;
and (3) experiential foundationalism—the view that the foundation of
empirical knowledge is subjective experience). Wittgenstein sets himself
against the very idea of scientific philosophy, seeing it as his mission to
“bring words back from their metaphysical to their everyday use”—the

6 PI, §89, §109, §126, §127.
7 See Stanley Cavell, “Must We Mean What We Say”, in Must We Mean What We Say
(Cambridge: Cambridge University Press, 1969), ch. 1.
8 That Wittgenstein held this particular conception of empirical science is plausible;
although it has to be admitted that the texts are inconclusive on the matter.

36

David Macarthur

bearing of which on naturalism depends on the fact that in its orthodox
form it is a scientific metaphysics that models itself on science.9
Since Wittgenstein is very far from scientific naturalism we must ask:
What kind of naturalism, if any, does Wittgenstein espouse? And what
is its relation to his new method of philosophizing without a priori doctrines? In this paper I want to consider these questions by way of first
examining Wittgenstein’s conception of the relation between science and
philosophy. I shall then discuss the appearance in his work of such ideas
as “the natural history of human beings”, “very general facts of nature”,
“forms of life”, and “primitive” reactions—all of which seem to turn on
some idea of nature or, rather, human nature. Only at that point will we
be in a position to consider the question of naturalism, what we might
call that, in Wittgenstein’s work.

Scientific Metaphysics: The Case of Jackson’s Physicalism
Wittgenstein is notorious for utterly rejecting the idea that philosophy is
one of the sciences; a view in stark contrast to the Quine’s highly influential form of naturalism which states that “philosophy . . . [is] continuous
with the rest of science”.10 In the Tractatus, Wittgenstein is unequivocal
in stating:
Philosophy is not one of the natural sciences.
(The word ‘philosophy’ must mean something which stands above
or below but not beside the natural sciences.)
The object of philosophy is the logical clarification of thoughts.
Philosophy is not a theory but an activity.
A philosophical work consists essentially in elucidations.11
Philosophical problems are distinct from the empirical questions dealt
with in natural science: in philosophy, “It is not problems of natural science which have to be solved”.12 Indeed, we can generalize and say philosophy is autonomous from the natural and social sciences—at least if
we follow Wittgenstein and understand “science” not in terms of what
Germans call Wissenschaft—a system of putative knowledge and schol9 PI, §116.
10 W. V. Quine, “Reply to Putnam”, in The Philosophy of W. V. Quine, eds. L. E. Hahn
and P. A. Schillp (La Salle: Open Court Press, 1986), 430–431. It is not enough to say
that since philosophy is a different subject than philosophy there is no tension between
Wittgenstein and Quine here. Wittgenstein thought that philosophy was a one way of
continuing traditional philosophy. Quine thought the same about his naturalist outlook. For further discussion of Quine’s naturalism, see my paper “Quinean Naturalism
in Question”, Philo, vol. 11, no. 1 (2008): 1–14.
11 TLP, 4.111–4.112.
12 Ibid., 6.4312.

Human Nature

37

arship which includes mathematics, literary studies, theology, etc.—but
as empirical studies that attempt to provide objective causal explanations
of observable phenomena.13 This unqualified attitude to the empirical
sciences as a whole is expressed in the Investigations, “It was correct that
our considerations must not be scientific ones . . . philosophical problems . . . are, of course, not empirical problems”.14
Whereas the sciences typically posit hidden ‘unobservables’ to causally explain phenomena, philosophy is an activity, “the logical clarification of thoughts,” that depends on the idea that nothing is hidden.
Science explains phenomena by subsuming them under local causal
patterns (e.g. social sciences) or universal causal laws (e.g. physics);
or, alternatively, by appeal to the causal propensities of things.15 Philosophy relies on no more than our prior knowledge of the language
acquired when we learnt our mother tongue, and manifested as our
capacity to recount “criteria”, in recollecting or imagining concrete
communicative contexts, for the application of concepts of ‘objects’ to
the world.
In the Blue Book, Wittgenstein observes that traditional philosophy—
by which he means, primarily, metaphysics16—tends to use scientific
models of explanation. He traces the metaphysical urge to explain phenomena in general to:
our preoccupation with the method of science. I mean the method
of reducing the explanation of natural phenomena to the smallest
possible number of primitive natural laws; and, in mathematics, of
unifying the treatment of different topics by using a generalization.
Philosophers constantly see the method of science before their
eyes, and are irresistibly tempted to ask and answer questions in the
way science does. This tendency is the real source of metaphysics,
and leads the philosopher into complete darkness.17

13 Wissenschaft is a much broader category than “empirical science”, since it refers to a
system of learning, scholarship, and putative knowledge, which applies equally to the
empirical sciences and abstract fields (e.g. mathematics, logic) as well as to research into
evaluative matters (e.g. political and cultural studies) and systems of metaphysics (e.g.
theology, ontology).
14 PI, §109.
15 See, e.g., Nancy Cartwright, The Dappled World: A Study of the Boundaries of Science
(Cambridge: Cambridge University Press, 1999).
16 Wittgenstein’s conception of traditional philosophy as fundamentally metaphysics has
more going for it than we might initially suppose. Metaphysics is at the heart of the
beginnings of philosophy in Plato and Aristotle; and if we consider the modern turn to
epistemology in Descartes and Kant then these can be translated into metaphysics in so
far as we are dealing with what can be known a priori. We might say Descartes is concerned with the metaphysics of physical science; and Kant with “critical” metaphysics.
17 Ludwig Wittgenstein, The Blue and Brown Books (Oxford: Blackwell, 1969), 18.

38

David Macarthur

For example, the early C20th method of logical analysis, which ultimately
analyzed propositions of natural language into “atomic propositions”,
was modeled on the chemical or physical analysis of a material compound into its component parts.18 In Russell’s work, at least, that was
taken to show that the world consisted of atomic facts corresponding
to true atomic propositions all of whose elements we must be individually acquainted with. However, let us consider in greater detail the more
recent example of Frank Jackson to illustrate the power of Wittgenstein’s
diagnosis of the source of metaphysics in the temptation to give science
the wrong kind of significance in our thinking.
Frank Jackson combines an ontological form of scientific naturalism
with an epistemological commitment to a priori analysis.19 He explicitly models metaphysical explanations on the explanations of physics,
remarking:
Metaphysicians seek a comprehensive account of some subject
matter—the mind, the semantic, or, most ambitiously, everything—in
terms of a limited number of more or less basic notions. In doing this
they are following the good example of physicists. The methodology
is not that of letting a thousand flowers bloom but rather that of
making do with as meager a diet as possible . . . But if metaphysics
seeks comprehension in terms of limited ingredients, it is continually
going to be faced with the problem of location. Because the ingredients are limited, some putative features of the world are not going
to appear explicitly in the story. The question then will be whether
they, nevertheless, figure implicitly in the story. Serious metaphysics
is simultaneously discriminatory and putatively complete, and the
combination of these two facts means that there is bound to be a
whole range of putative features of our world up for either elimination or location.20
Rather than “draw[ing] up big lists” of everything that appears in the
manifest image, the job of the metaphysician is to follow the physicist’s
mode of explanation by comprehensively explaining the highly complex
appearances of things in terms of the smallest possible number of “basic
18 This is the basis of a criticism Wittgenstein makes of the Tractatus: “My notion in the
Tractatus Logico-Philosophicus was wrong . . . because I too thought that logical analysis had to bring to light what was hidden (as chemical and physical analysis does)”.
Philosophical Grammar, ed. Rush Rhees, trans. Anthony Kenny (Oxford: Blackwell,
1974), 210.
19 There is a tension between scientific naturalism and the commitment to a substantial
notion of the a priori, as Quine shows in “Two Dogmas of Empiricism”, From a Logical
Point of View (Cambridge, MA: Harvard University Press, 1953), 20–46.
20 Frank Jackson, Mind, Method and Conditionals: Selected Essays (London: Routledge,
1998), 156.

Human Nature

39

notions”.21 In this spirit, Jackson defends physicalism, which claims that
“a complete account of what our world is like, its nature” can in principle be obtained from a relatively small set of physical properties and
relations.22
Part of Wittgenstein’s teaching here is to carefully observe the distinction between science and metaphysics; since it is a failure to respect this
difference that leads to the confused idea of a “scientific” metaphysics,
e.g. scientific naturalism. It is worth observing that science alone does
not explain Jackson’s commitment to the metaphysical thesis of physicalism. For a start, science is really a plurality of distinct sciences, there
being no wholesale semantic reductions of, say, psychology to biology, or
biology to chemistry, or chemistry to physics; not to mention, of physics
to physics—say, general relativity to quantum mechanics. The “unity”
of the sciences is a metaphysical dream that has not come close to being
realized. Even if we were committed only to the entities posited by the
successful sciences we would be ontological pluralists acknowledging all
the many kinds of things not recognized by physics, say, chemical or biological or psychological kinds.
But why should we be only committed to recognizing the entities posited by the successful sciences? What of, say, persons, a second-person
concept whose deployment requires acknowledgement by others, or common artifacts like tables and chairs, whose identity conditions involve
certain interests and purposes? Persons and artifacts are too subjective to
figure in the objective explanations of science. So we would be mistaken
to say that science investigates the world as such. It investigates certain
objectively specifiable natural phenomena with a view to discovering
objective causal capacities or causal patterns and laws. John Dewey said
that “the naturalist is one who has respect for the conclusions of natural
science”.23 But one can have respect for the conclusions of natural science, indeed for the conclusions of the successful sciences in general, and
still hold that there is more in the world than is dreamt of in scientific
theorizing; and by “more” I do not mean anything supernatural. I mean
persons, chairs, buildings, clothing, and the like.
It is also worth asking: What does “physical” mean in a context where
it is being put to this metaphysical explanatory work? Jackson claims
that we can give an ostensive definition of “physical” by pointing to common objects like tables and chairs and then picking out the physical as
“the kinds of properties and relations needed to give a complete account
of things like them”.24 If we want to know what kinds these are we are
21 Ibid.
22 Frank Jackson, From Metaphysics to Ethics (Oxford: Clarendon, 1998), 6.
23 John Dewey, “Anti-Naturalism in Extremis”, in Naturalism and the Human Spirit, ed.
Y. Krikorian (New York: Columbia University Press, 1944), 2.
24 Jackson, From Metaphysics to Ethics, 7.

40

David Macarthur

told to look to physics or the physical sciences and their explanations
of non-sentient reality, especially appealing to “the success of microexplanations of macroscopic phenomena”.25
Jackson argues that the properties and relations posited by current
physics give us a reasonable confidence that we at least know the kinds of
properties and relations required to explain things like tables and chairs.
But given the many astonishing conceptual innovations in the history
of science (e.g. the special and general theories of relativity, quantum
mechanics), this is really an unwarranted hope which has more to do
with metaphysical ambition than serious reflection on the relation of science to its own past and possible future.
Furthermore, a chair, say, cannot be completely explained in terms of
the properties and relations posited by current physics or the physical sciences more broadly conceived. A chair is not a posit of any explanation
in the physical sciences. The physical sciences might be expected to be
able to explain the physical stuff that the chair is made of—say, atoms of
various kinds, organized in complex ways, etc. But, as Aristotle famously
noted with regard to a sculpture and the bronze that composes it, there
is a difference in identity conditions between an artifact (a sculpture, a
chair, etc.) and the physical stuff that it is made of—which is clear if we
consider counterfactual circumstances in which, e.g. a chair loses some
paint or sustains a scratch or has its fabric or padding replaced without
ceasing to be the same chair.
Physical science cannot explain what a chair is or why it exists for
the simple reason that any such explanation must involve an appeal to
human interests and intentions. A chair is an artifact typically made to be
sat upon by people. One cannot fully explain a chair without adverting
to the human interests it serves (e.g. relaxing, conversing, reading, listening to a lecture, watching TV or a film, eating together at a table) and the
intentions of the designers and carpenters or metallurgists or fabricators
who made it (e.g. to make it comfortable, or luxurious, or tailored to a
specific function like dentistry or space flight).
Jackson defines physicalism as the view that a complete account of
the physical facts entails a complete account of psychological (and other
globally supervenient) facts including facts about interests and intentions. The only reason we are given to accept this metaphysical thesis
is the proposal that we can give an a priori analysis of folk psychological notions like “intention” in causal-functional terms and then empirically discover the physical states or events that best realize the relevant
causal-functional role(s). But why suppose there is an a priori analysis of
“intention” or other content-involving psychological notions in causalfunctional terms? A major stumbling block is that intention—like other
content-involving states in this respect—is a rationally normative notion
25 Ibid.

Human Nature

41

and, as has often been noted, there is no reducing or explaining rational
normativity in merely causal terms.26
If someone has an intention, then one can appropriately ask a “Why?”—
question which is understandable as the request for reasons in the standard normative sense. If appropriate reasons for one’s intention, or the
actions it gives rise to, are not forthcoming then one (or one’s intentions,
or actions) can be criticized as thoughtless, stupid, unjustified, pointless
and so on. Such rational normativity cannot be captured in terms of what
simply happens, whether by chance or physical necessity, since the very
idea of normative assessment involves ideals which may not be instantiated but which are, despite that, still operative in the standard-setting
sense. For example, what are in fact ill-considered intentions (i.e. based
on inadequate or poor reasons) may still aim at being good ones; just as
a false belief still aims at being true. In other words, the standard or ideal
retains its normative force, even when it is not instantiated by the states
that are bound or beholden to it. Causal notions, or functions understood
in causal terms, which account for what actually happens or perhaps
what had to happen, are precisely unable to capture this distinctive normative force.
If the problem of normativity cannot be answered, then there simply is
no “complete account” of human interests and intentions (etc.) in terms
of micro-properties and micro-relations of the sort posited by the physical sciences. And this point is unaffected by whether we consider mental
types or tokens. Consequently, Jackson’s attempt to define the physical
in terms of examples of ordinary objects like tables and chairs fails to
achieve its larger metaphysical purpose. The suggestion that physical
properties and relations at the micro level can explain everything in the
manifest image, the world of our everyday lives, fails. Persons and the
realm of human artifacts go largely unexplained by the physical sciences
given that a “complete account” of them must advert to the rationally
normative. So, adopting the scientific model of explanation in metaphysics, indeed treating physics as the model of one’s method of metaphysics,
leads Jackson into an aporia in the attempt to explain the term “physical” and a blind spot as regards the problem of normativity. This is a
good example of what Wittgenstein means by saying that treating science as one’s model of metaphysics leads the philosopher into complete
darkness.27
26 See, e.g.: Donald Davidson, “Could There be a Science of Rationality,” in Naturalism
in Question, eds. Mario De Caro and David Macathur (Cambridge, MA: Harvard
University Press, 2004), ch. 8; John McDowell, Mind and World, 2nd ed. (Cambridge,
MA: Harvard University Press, 1996); and Hilary Putnam, “The Content and Appeal
of Naturalism”, in Naturalism in Question, ch. 4.
27 Apart from not explaining what he presumes to explain, Jackson “obliterates the
distinction between [the] factual and conceptual”, as Wittgenstein puts it. The facts
discovered by physics are treated as playing a foundational role in our thinking on the

42

David Macarthur

Nature and the Natural in the Investigations
We have seen that Wittgenstein has an antipathy to scientific naturalism
and the influence of science on philosophy (e.g. our thralldom to the scientific method of thinking)—which also includes the myths surrounding
scientific progress that permeate the culture at large (e.g. that scientific
progress is an advancement in civilization). But, despite that, there are
aspects of Wittgenstein’s work that we might want to count as elements
of something that deserves the name of naturalism. Stanley Cavell writes:
Wittgenstein’s Philosophical Investigations . . . may be said to propose Wittgenstein’s work there as a naturalizing of philosophy . . .28
What might this mean? We get some sense of it by contrast with traditional philosophy understood as the metaphysical quest for “something
extraordinary” or “strange” or “astonishing”, say, Forms or universals
or substances or mindless (living) bodies or a supernatural God—all
of which Wittgenstein thinks of in terms of the “pursuit of chimeras”.29
Traditional philosophy of language also sees language as “some nonspatial, atemporal non-entity” and imagines “a super-order between—so to
speak—super-concepts”.30 The task of naturalistic philosophy is to root
out, diagnose and treat metaphysics, a main source of which is our “tendency to sublimate the logic of our language”.31
Part of the point of calling Wittgenstein’s method a form of naturalism
is the importance it attaches to reminding us of the ordinary use of words
as they function in the everyday lives of human beings considered as
natural beings in a natural world—the relevant contrast being supernatural beings in a supernatural world. The problem is that our ordinary use
of words is something that, for the most part, tends to escape our notice.
It is in order to get some kind of grip on the prodigious complexity and
diversity of how words function that Wittgenstein calls our attention to
what he describes as “the natural history of the human”, for example:
Giving orders, asking questions, telling stories, having a chat, are as much
a part of our natural history as walking, eating, drinking, playing.32

28
29
30
31
32

grounds that physical truths entail truths of every other kind. And knowing this entailment relation is treated as a condition of understanding for the concepts expressed by
non-physical sentences. Ludwig Wittgenstein, Zettel, 2nd ed., trans. G. E. M. Anscombe
(London: Blackwell, 1967), §458.
Stanley Cavell, “Postscript (2002) to The Investigations Everyday Aesthetics of Itself”,
in Naturalism in Question.
PI, §93, §195, §197, §94.
Ibid., §108, §97.
Ibid., §38.
Ibid., §25.

Human Nature

43

And in another remark, he adds:
What we are supplying are really remarks on the natural history of
human beings; not curiosities, however, but facts that no one has
doubted, which have escaped notice only because they are always
before our eyes.33
The paradoxicality of Wittgenstein’s procedure is evident in this latter
remark since it seems that we need to be reminded of something that we,
in a sense, already know, but that we are, perversely, blind to precisely
because it is so familiar. But then it seems that perversity, too, is part of
our natural history; as is the temptation to pursue chimerical explanations of things in general from the armchair without careful observation or
experiment. To use a later metaphor, we have the urge to build philosophical air-castles (Luftgebaude). The image of the human is, then, of something ordinary, familiar and common and, at the same time, of something
extraordinary, strange and astonishing. As Cavell puts it, “we might also
wish to say that what is natural to the human is precisely the unnatural”.34
So when Wittgenstein writes, “what we do is to bring words back from
their metaphysical to their everyday use”, what we are in effect doing is
bringing ourselves back from unnaturalness to naturalness.35 But since
the everyday use to which we are “returning” is something that we have
hitherto failed to notice, we can say that what is natural to us is no more
a matter of recovery than of creation. Wittgenstein’s philosophy, then,
aims to forge a conception of human naturalness against a vision of ourselves as perennially tempted to unnaturalness.
Another dimension of Wittgenstein’s naturalizing of philosophy concerns his rejection of the idea that reason or the language that expresses
conceptual connections (from a “logical” or grammatical point of view)
is presuppositionless. Wittgenstein remarks, “concepts . . . are the expression of our interest and direct our interest”.36 The crystalline purity that
the Wittgenstein of the Tractatus supposed characterized concepts—
expressed in the vision of “a super-order between . . . super-concepts”37—
is a metaphysical illusion concerning language. Just as our interests can
be looked at from the point of view of culture and society or from the
point of view of human animality, so, too, our concepts can be looked at
from the point of view of two different conceptions of form of life which
Cavell has called the anthropological and the biological: the former being
a matter of differences at the social or cultural level (say, knowing the

33
34
35
36
37

Ibid., §415.
Naturalism in Question, 276–277.
PI, §116.
Ibid., §569.
Ibid., §97.

44

David Macarthur

difference between a wave and a salute, or between an endearment and
an insult); and the latter involving the idea that we share a common
human nature involving knowing how we see, hear, touch, smell, and
taste, as well as our how we tend to move, and eat, wash, dress, have sex,
and so on. In contrast to the form of life of a dog or cat or horse are such
differences as greeting versus barking, drinking versus lapping, jogging
vs cantering, being aroused vs being on heat. That we greet each other,
or drink, or jog, or can find another arousing, are as much as part of
our nature as that we issue orders, ask questions, and talk about things.
Human nature is one aspect of the “general facts of nature”, an indeterminate and diverse array of more or less stable contingencies, upon which
the point of our concepts depends, as Wittgenstein notes:
What we have to mention in order to explain the significance, I mean
the importance, of a concept are often extremely general facts of
nature: such facts as are hardly ever mentioned because of their great
generality.38
That, of course, does not mean that Wittgenstein is doing a kind of
anthropology. As we have seen, Wittgenstein strongly denies that philosophy is any kind of empirical science. His quasi-anthropological reflections
are either so obvious that they do not require empirical confirmation
(e.g. the trivial claim that humans walk, and chat, and play); or they are
offered not as plausible empirical hypotheses but simply as possible (i.e.
intelligible) causes of how our concepts are formed. An example of this
latter appeal to an imagined anthropology occurs when he asks, “How
does a human being learn the meaning of names of sensations?” and he
responds:
Here is one possibility: words are connected with the primitive, the
natural, expressions of the sensation and used in their place. A child
hurts himself and he cries; and then adults talk to him and teach
him exclamations and, later, sentences. They teach the child new pain
behaviour.39
This should not be read as a hypothesis about how we in fact learn the
meaning of sensation words. In this instance, Wittgenstein is not concerned with empirical accuracy; even if it seems plausible, it is only
intended as a possibility, an intelligible way in which our concepts might
have been formed. As he explains:
We are, indeed, also interested in the correspondence between concepts and very general facts of nature. (Such facts as mostly do not
38 PI, §142. Emphasis added.
39 Ibid., §244. Emphasis added.

Human Nature

45

strike us because of their generality.) But our interest is not thereby
thrown back on to these possible causes of concept formation; we are
not doing natural science; nor yet natural history—since we can also
invent fictitious natural history for our purposes.40
The point is to communicate the idea of a “correspondence” between concepts and general facts of nature. Such a correspondence might be that
between the concept of chair and the fact that humans normally like to sit
down to relax, talk and eat. For Wittgenstein’s purposes it is enough to get
across the idea of such a “correspondence” to imagine the facts of nature to
be quite different from what they, in fact, are. So, e.g. in a world where the
weights of things fluctuated unexpectedly and often due to frequent manifestly detectable gravity waves then we would not weigh meat on a butcher
shop scales in order to determine its cost. Perhaps under those circumstances,
we would determine the price of meat by measuring its volume instead.

Wittgenstein’s Non-Explanatory Liberal Naturalism
A fundamental feature of Wittgenstein’s philosophy is metaphysical quietism, which finds expression in the methodological remark, “All explanation must disappear and description alone must take its place”.41 That is,
we must—through skeptical diagnosis and a kind of imaginative therapy
that involves replacing good pictures (e.g. ‘meaning is use’) for misleading or bad pictures (e.g. ‘meaning is naming objects’)—overcome the urge
to provide metaphysical explanations of phenomena in general from the
armchair. We are to replace that chimerical activity with descriptions of
particular aspects of the conceptual landscape informing our form of life,
by way of recounting or recalling ordinary criteria for the application of
concepts of familiar things.
It is important to take note of the fact that scientific naturalism is a
metaphysical explanatory program; so if we want to say that Wittgenstein is some kind of naturalist then it must be understood in such a way
that it is not at all in the same line of explanatory work as orthodox
naturalism. It will have to be understood as playing an important role not
in the framing or justification of some metaphysical system but in helping
to overcome metaphysics—including the ubiquitous human temptation
to use words with a metaphysical emphasis (i.e. something that does not
require the background of a fully worked out metaphysical system of the
kind typical within academic philosophy).
Thus, the role of Wittgenstein’s naturalism is non-constructive and dialectical. When we are tempted to talk or think in a metaphysical manner—
and Wittgenstein’s focus is our tendency to metaphysicalize (say, by
inquiring into the essence of) mind and language—its point is to remind
40 Ibid., Part ii, sect. 11, §365. Emphasis added.
41 Ibid., §109.

46

David Macarthur

ourselves of the ordinary uses of words in human linguistic transactions
and to get us to see how little these uses depend upon, and how much
they are obscured by, the kinds of a priori explanations traditional philosophy tends to offer.
With this important caveat regarding explanatoriness, we might say
that Wittgenstein is a liberal naturalist in the sense that Mario De Caro
and I have defined it, which we might characterize thus:
1. a denial that reality is, or contains, any supernatural beings—where,
let us say, a supernatural being is an unobservable super-human being
posited for causal explanatory purposes but which lacks the requisite
scientific support to be credible e.g., the theological commitment to
a supermundane God or Devil;42
2. a denial that reality is exhausted by the scientific image (or, better, the
composite scientific images) of the world;43
3. the affirmation that there are plurality of non-scientific and scientific
forms of knowing and/or understanding. The sciences do not provide
a single unified and complete account of reality.44
Liberal naturalism, so understood, is a negative discipline that avoids the
two extremes of supernaturalism and scientism in one’s philosophical
reflections as well as an acknowledging that there is a plurality of different kinds of understanding or knowing both within and outside the
sciences. This is a metaphysical quietist version of liberal naturalism, one
that avoids supernatural theological commitments in philosophy as well
as refusing to give science an unwarranted ontological significance.45
As a trained engineer, Wittgenstein was well versed in a considerable
amount of modern natural science and aware of its importance in many
aspects of our lives.46 Indeed, he took himself to be living through what he
called “the scientific and technological age”.47 Part of what this means is

42 For further discussion of the concept of the supernatural, see my paper “Liberal Naturalism and the Philosophy of the Manifest Image”, in For a New Naturalism, eds. Arran
Gare and Wayne Hudson (Candor, NY: Telos, 2017).
43 For convenience, I will use the term “scientific image” to refer not to the image of any
given science or branch of such but the composite image which combines all the images
of all the explanatorily successful sciences.
44 See, e.g. John Dupre, The Disorder of Things: Metaphysical Foundations for the Disunity of the Sciences (Cambridge, MA: Harvard University Press, 1995).
45 It is, then, a non-scientific naturalism, but this way of putting it is misleading. I prefer to
call it a liberal naturalism since Wittgenstein is not denying the reality of the scientific
image.
46 Wittgenstein discusses the theoretical physicist Heinrich Hertz in the Tractatus, 4.04,
6.361; and he is known to have studied the writings of other notable physicists such as
Ernst Mach and Ludwig Boltzmann.
47 Ludwig Wittgenstein, Culture and Value, revised 2nd ed. (Oxford: Blackwell, 1998), 64.

Human Nature

47

that we learn basic science through our education system48; and established
scientific knowledge is at the heart of our epistemic practices which involve
the appeal to scientific expertise in many areas such as manufacturing,
medicine, pharmaceuticals, mining, law, engineering, agriculture and so on.
Nonetheless, Wittgenstein was deeply out of sympathy with the scientific spirit of his times. He felt that his way of thinking was of an entirely
different kind than that of “the typical western scientist”: “I am aiming
at something different than are the scientists & my thoughts move differently than do theirs”.49 He was much more sympathetic to religion,
although it is clear that Wittgenstein did not accept an interpretation of
it in terms of an intelligent being with supernatural powers. He remarks:
Christianity is not based on a historical truth, but presents us with a
(historical) narrative & says: now believe! But not believe this report
with the belief that is appropriate to a historical report—but rather:
believe, through thick & thin & you can do this only as the outcome of a life. Here you have a message!—don’t treat it as you would
another historical message! Make a quite different place for it in your
life.—There is no paradox about that!50
Although nature or reality is not equivalent to the scientific image of
the world, it is important to see that science plays an crucial role in ruling
out certain supernatural explanations and posits as otiose (say, because
they do not have sufficient or strong enough connections to the empirical). Wittgenstein can be read as demystifying the notion of nature by
being prepared to recognize the reality of non-scientific things (i.e. things
that are not the posits of any successful explanatory science) that we
experience and mark in our language e.g. persons, artworks, and artifacts, like chairs, houses, and cars.
We can further refine this conception by noting that Wittgenstein does
not put any explanatory weight on a notion of nature in philosophy.
Indeed, we might say that we only come by our idea of nature by excluding
various supernatural causes of phenomena e.g. miracles, ghosts, fairies,
angels, demons, God. Natural phenomena fit Wittgenstein’s description
of normal circumstances: “we recognize normal circumstances [natural
phenomena] but cannot precisely describe them. At most, we can describe
a range of abnormal [supernatural] ones”.51 This goes with the thought
48 Scientific discoveries as well as such things as geographical and historical facts are
transmitted to us by our system of education. On Certainty, eds. G. E. M. Anscombe
and G. H. von Wright, trans. Denis Paul and G. E. M. Anscombe (Oxford: Blackwell,
1969), §162–163.
49 CV, 9.
50 Ibid., 37.
51 OC, §27.

48

David Macarthur

that we only recognize the familiar in retrospect when confronted with
what seems unfamiliar or strange.52 Part of the power of philosophy (and,
we might add, art) is to be able to transfigure the familiar or natural in
order to make it newly available to us.
Notwithstanding his approach to nature, Wittgenstein does appeal to
a relatively substantial notion of human nature in his descriptions of
grammar—the logical dimension of our use of concepts. That is not to
say human nature is a theoretical notion in Wittgenstein’s philosophy.
We only need to be reminded of “the primitive, natural, expressions”
of pain—say, screaming, howling, moaning, groaning, not touching the
painful area, limping, or keeping the damaged part from bearing weight
or doing work, applying bandages, or slings, or ointments, etc.—to
acknowledge that we know quite a lot about how humans experience
and give expression to pain.53 And the same goes for other sensations,
e.g. a sour taste, giddiness, an itch. Some of these aspects of the human
are common to the animal world; for instance, pets and livestock, for
example, display similar pain behavior. The connection with animals also
animates Wittgenstein’s view of the expression of intention. He remarks,
What is the natural expression of an intention?—Look at a cat when
it stalks a bird; or a beast when it wants to escape.
((Connection with propositions about sensations.))54
Cats express pain in a paw by mewling, holding the paw off the ground
and delicately licking it. They need not be able to say, “I have pain in my
paw” in order to express it. So, too, they express the intention to catch
a bird through the intelligibility of their behavior, by creeping up on it
without notice and then making a sudden leap for the bird.
The dialectical reactive character of Wittgenstein’s naturalism is particularly evident in his treatment of how we learn language. His reflections on human nature are motivated, in large part, by a desire to combat
our tendency to over-intellectualize language when we reflect upon it
philosophically. As we have seen, one technique is to present possible
non-intellectualist ways in which we might have learnt language. For
example, in the case of learning the names for sensations Wittgenstein
52 Wittgenstein remarks, “The aspects of things that are most important for us are hidden
because of their simplicity and familiarity. (One is unable to notice something—because
it is always before one’s eyes.)” PI, §129. It is also important to see that the supernatural
is not simply the unfamiliar. Abnormal, strange, and unfamiliar phenomena are aspects
of the natural world, too. That is one reason I prefer to think of the supernatural in
terms of unobservable (and typically superhuman) agencies posited to figure in causal
explanations of observable phenomena. Thus supernatural explanations mimic those
of science, as Wittgenstein thought.
53 Ibid., §244.
54 Ibid., §647.

Human Nature

49

says that perhaps “the verbal expression of pain replaces crying”.55 That
this is intelligible is enough to break the grip of the idea that the only way
of learning the name for pain is the intellectualist view: for example, that
defended by John McDowell in Mind and World according to which in
learning a language we come to appreciate the way in which “impressions of ‘inner sense’ must be, like the impressions of ‘outer sense’, passive
occurrences in which conceptual capacities are drawn into operation”.56
The intellectualist thus holds that pain impressions must come to be seen
as already possessed of conceptual content if they are to play the right
role in justifying judgments, e.g. of “I am aware that I am in pain now”.
Wittgenstein’s alternative possibility at least shows that the argument for
this “must” is broken-backed.57
An additional element of Wittgenstein’s naturalist critique of intellectualism is to attempt to expose and replace its imaginative bedding in some
spell-binding picture of how language operates. Consider this remark,
I want to regard man here as an animal; as a primitive being to which
one grants instinct but not ratiocination. As a creature in a primitive
state. Any logic good enough for a primitive means of communication needs no apology from us. Language did not emerge from some
kind of ratiocination.58
Here Wittgenstein is replacing the picture of ‘man as intellect’—associated
with our tendency to see the mind and its thoughts as something supernatural—with the natural picture of ‘man as animal’. Pictures are not
empirical hypotheses; rather they are simplified imaginative schemas or
models that we use as heuristics to help us understand the highly complex
and diverse phenomena that confronts us in everyday life and in the sciences. Like an architectural model of a building, they are, in themselves,
neither true nor false; that is, unless and until we apply them to the world
in some way, such as by an act of comparison.
At the base of language is not rational insight into the unboundedness
of the conceptual (contra McDowell) but “primitive” (spontaneous, unreflective, ungrounded) forms of acting and reacting. Consider this remark:
The origin & the primitive form of the language game is a reaction;
only from this can the more complicated forms grow. Language—I
want to say—is a refinement, ‘in the beginning was the deed’.59

55 Ibid., §244.
56 McDowell, Mind and World, 22.
57 Of course, to discover where and how the argument breaks down is a matter for
investigation.
58 OC, §475.
59 CV, 36.

50

David Macarthur

This, too, is a picture as opposed to an armchair contribution to evolutionary anthropology. It is enough to break the grip of the picture that language
must have originated in reason and argument to imagine an alternative
possibility. It need make no claim to historical accuracy or empirical verification. In order to properly understand its bearing, to put it in the right
light, it is important that we make a connection between it and the important Investigations theme of explanations and reasons coming to an end:60
. . . As if giving grounds did not come to an end sometime. But the
end is not an ungrounded presupposition: it is an ungrounded way
of acting.61
According to Descartes, if we are intellectually scrupulous and ask what
we really know, then we must pursue the regress of reasons for reasons
until we reach the claims “I exist” and “I think”, which are combined
in the famous cogito. Descartes presupposes that these claims cannot be
coherently doubted and that what cannot be coherently doubted must be
true. These foundations of knowledge are, as Wittgenstein correctly notes,
“ungrounded presuppositions”. Wittgenstein does not argue against this
picture—indeed arguing against pictures is futile. Instead, he replaces it
with a better, less misleading, one. On Wittgenstein’s alternative naturalist picture, giving reasons is something we do without any philosophical
theory of where reasons end or what justifies this terminus. Applying the
picture, we might say that we give reasons if and when necessary, depending on the practical circumstances of the case (e.g. its importance, who
is asking, how much time and effort we are willing to make, and so on),
but there is simply no such thing as giving reasons to remove all possible
doubts. To adapt one of Wittgenstein’s remarks, we might say, “[A word
or sentence] is in order—if, under normal circumstances, it fulfills its
purpose”62—where the question of purpose is an occasion-sensitive matter.
So, there are at least three ways in which Wittgenstein handles the
theme of human nature within what I am calling his liberal naturalism:
1. Reminding the reader of obvious facts about human nature which
are so familiar they escape notice, e.g. that we walk, and chat, and
eat food with our hands perhaps with the aid of utensils, and that we
tend not to eat our dead, and so on.
2. Imagining possible explanations of the way we learn words, say,
for sensations—which might involve an appeal to the sort of facts
described in 1). These are not empirical hypotheses and so do not
depend on empirical confirmation. They function simply to break the
60 PI, §1, §87, §326, §485, and Part ii, sect. v, §33.
61 OC, §110.
62 PI, §87.

Human Nature

51

grip of some picture that is holding us captive, countering the sense
that things must go a certain way that is apparently dictated by the
picture (say, of meaning as objects named).63
3. Providing counter-intellectualist imaginative pictures of the human:,
e.g. the picture of man as a “primitive being to which one grants
instinct but not ratiocination”64; or “The human body is the best
picture of the human soul.”65 These are used non-argumentatively
to remove a picture that we fail to notice but which, just because of
that, colors our view of the world unawares. A bad picture seems to
dictate a certain application which is often projected into our view of
the world without our even being conscious of it.
All of these aspects of the appeal to human nature are directed against intellectualist views of mind, meaning, and language that treat what are, in fact,
natural and familiar aspects of our life as though they were strange and
unheard of. But whilst, as Cavell saw, philosophy is naturalized, Wittgenstein’s vision of human nature is of us as, repeatedly, tempted into unnatural frames of mind for, without that, philosophy would lose its point.66

Naturalism as a World-Picture
In On Certainty, Wittgenstein offers an original vantage on the question
of naturalism from what we might call an imaginative point of view. The
suggestion is that metaphysical theories, for all of their apparent intellectual sophistication, ultimately rest on an imaginative core that is impervious to reason. Naturalism and religion (i.e. perhaps the most pervasive
and important form supernaturalism has taken in the history of human
thought) are, or rest on, two different imaginative structures or “worldpictures”.67 Elements of the naturalist world picture include:
1. The earth has existed for a long time before our births.68
2. The causal connections discovered by the sciences characterize things
in general not just for a certain time and place.69
63
64
65
66

Ibid., §115.
OC, §475.
PI, Part ii, sect. iv, §25.
At Ibid., §109 Wittgenstein remarks, “All explanation must disappear, and description
alone must take its place. And this description [of the grammar of particular concepts]
gets its light—that is to say, its purpose—from the philosophical [or metaphysical]
problems”.
67 OC, §93–95. Anthony Kenny argues that this conception “gives the most satisfactory picture of the relationship between Wittgenstein and naturalism,” at least, where
naturalism is understood in terms of human nature. “Whose Naturalism? Which Wittgenstein?”, American Philosophical Quarterly, vol. 48, no. 2 (2011): 113–118, 115.
68 Ibid., §203, §233–234, §261–262.
69 Ibid., §167.

52

David Macarthur

3. Part of this world-picture includes “the picture of the earth as a ball
floating free in space.”70
4. Another part is that material things are made of (invisible) atoms.71
And elements of the Judeo-Christian religious picture include:
1. God created the heavens and the earth.
2. There will be a Last Judgment and the good resurrected to live in
heaven with God.72
3. “God’s eye sees everything”.73
4. Human beings have a soul by the grace of God.
If naturalism and religion are identified with large-scale pictures of the
world as a whole—rather than the super-structure of philosophical theorizing built on or around them—then they provide the taken-for-granted
background against which we judge truth and falsity, as Wittgenstein
remarks:
But I did not get my picture of the world by satisfying myself of its
correctness: nor do I have it because I am satisfied of its correctness. No: it is the inherited background against which I distinguish
between true and false.74
That naturalism is best interpreted as a world-picture helps explain an
oft-noted feature of it, namely, that it is frequently left unarticulated;
and when it is given expression, it is often in the form of crude, highly
ambiguous slogans.75
Understood as a world-picture, naturalism is not subject to rational criticism being neither true nor false, justified nor unjustified.76 But, then, the
same can then be said for religion. Both the naturalist and religious worldpicture are typically acquired in childhood through a particular upbringing
and education, rather than through any deliberative process of reasoning.77

70 Ibid., §146.
71 Ludwig Wittgenstein, Lectures and Conversations on Aesthetics, Psychology and Religious Belief, ed. Cyril Barrett (Berkeley: University of California Press, 1967), 32.
72 Ibid., 57.
73 Ibid., 71.
74 OC, §94.
75 See Barry Stroud, “The Charm of Naturalism”, in Naturalism in Question.
76 Wittgenstein remarks that world-pictures “lie apart from the route travelled by inquiry”,
as is clear if someone challenges you to show that the world could not have come into
existence five minutes ago together with “old” rocks, fossil records, memories apparently extending back over the course of our lives, etc.
Ibid., §88.
77 So they are not to be identified with any fully worked out philosophical theory.

Human Nature

53

The question for Wittgenstein, then, becomes what the relation is between
these world-pictures, reflective thought and our forms of life.
What is of particular interest is Wittgenstein’s relatively sympathetic
treatment of the religious world-picture of Christianity in comparison to
the naturalistic world-picture. It is the latter, not the former, that Wittgenstein thinks tends to force a metaphysical construal upon us consisting
of, let us suppose, materialism (including a behaviorist view of mind)
and causal fundamentalism, i.e. the claim that there is a single fundamental causal order.78 Although Wittgenstein admits that the religious
world-picture leads some to a superstitious interpretation involving
belief in a supernatural God with super-human qualities, on the whole
Wittgenstein’s focus is on sharply distinguishing religious “belief” from
empirical or scientific belief—e.g., the latter, but not the former, relies on
evidence—as well as wavering over the question whether a non-believer
can be said to understand the person of faith.
Wittgenstein prefers not to see religion in terms of a theistic metaphysics. He is much more impressed by the way religious “beliefs” have
the power to change one’s entire life in a way that has no parallel in
philosophy:
Amongst other things Christianity says, I believe, that all sound doctrines are all useless. That you have to change your life. (Or the direction of your life) . . . Wisdom is passionless. By contrast Kierkegaard
calls faith a passion.79
It is of particular importance that when it comes to the human Wittgenstein borrows the language of the religious world-picture when he
speaks of the human as a “soul”: “My attitude towards him is an attitude
towards a soul. I am not of the opinion that he has a soul”.80 Wittgenstein is here pointing to the misleadingness of the grammar of having
in this context as opposed to that of being. Being a soul is being a person whose mindedness is expressed in gestures, facial expressions, and
bodily behaviors. Having a soul suggests something supernatural which,
somehow, the body “has” and which we have insufficient or no evidence
for—hence something that is a matter of mere opinion (in a pejorative
sense). The point is to avoid conceiving the body as a mindless lump of
matter which is, at best, caused to move in certain ways—a Cartesian picture shared by both the supernaturalist reading of the human as having
a “soul” in some superstitious spiritual sense and also by the naturalist
metaphysician who denies the existence of any such transcendent soul.
78 See, e.g. Donald Davidson, “Causal Relations”, Journal of Philosophy, vol. 64 (1967):
691–703.
79 CV, 61.
80 PI, Part ii, sect. iv.

54

David Macarthur

That a person, an embodied mindedness, is a soul is Wittgenstein’s way
of saying that a person is neither something supernatural nor something
wholly explicable in scientific terms.

References
Allen, B., Vanishing Into Things. Cambridge, MA: Harvard University Press,
2013.
Cavell, S., “Must We Mean What We Say.” In Must We Mean What We Say, Cambridge: Cambridge University Press, 1969, ch. 1.
———. “Postscript (2002) to The Investigations Everyday Aesthetics of Itself.”
In Naturalism in Question, edited by Mario De Caro and David Macarthur,
275–280. Cambridge, MA: Harvard University Press, 2004.
Davidson, D., “Causal Relations.” Journal of Philosophy, vol. 64 (1967): 691–703.
———. “Could There Be a Science of Rationality.” In Naturalism in Question,
edited by M. De Caro and D. Macarthur, 152–172. Cambridge, MA: Harvard
University Press, 2004.
Dewey, J., “Anti-Naturalism in Extremis.” In Naturalism and the Human Spirit,
edited by Y. Krikorian, 1–16. New York: Columbia University Press, 1944.
Dupré, J., The Disorder of Things: Metaphysical Foundations for the Disunity of
the Sciences. Cambridge, MA: Harvard University Press, 1995.
Hacker, P. M. S., “A Plague on Both Your ‘Isms’.” Amercian Philosophical Quarterly, vol. 48 no. 2 (2011): 97–111.
Jackson, F., From Metaphysics to Ethics. Oxford: Clarendon, 1998.
———, Mind, Method and Conditionals: Selected Essays. London: Routledge,
1998.
Kenny, A., “Whose Naturalism? Which Wittgenstein?” American Philosophical
Quarterly, vol. 48 no. 2 (2011): 113–118.
Macarthur, D. “Taking the Human Sciences Seriously.” In Naturalism and Normativity, edited by M. De Caro and D. Macarthur, 123–141. New York: Columbia
University Press, 2010.
McDowell, J., Mind and World, 2nd ed. Cambridge, MA: Harvard University
Press, 1996.
McGinn, M., “Wittgenstein and Naturalism.” In Naturalism and Normativity,
edited by M. De Caro and D. Macarthur, 322–351. New York: Columbia University Press.
Pears, D., “Wittgenstein’s Naturalism.” The Monist, vol. 78 no. 4 (1995): 411–424.
Putnam, H., “Wittgenstein on Religious Belief.” In Renewing Philosophy. Cambridge, MA: Harvard University Press, 1992, ch. 7.
———, The Collapse of the Fact/Value Dichotomy and Other Essays. Cambridge,
MA: Harvard University Press, 2002.
———, “The Content and Appeal of Naturalism.” In Naturalism in Question,
edited by M. De Caro and D. Macarthur, 59–70. Cambridge, MA: Harvard
University Press, 2004.
Quine, W. V., “Reply to Putnam.” In The Philosophy of W. V. Quine, edited by
L. E. Hahn and P. A. Schillp. La Salle: Open Court Press, 1986.
Rhees, R., Without Answers. London: Routledge, 1969.

Human Nature

55

Stroud, B., “The Charm of Naturalism.” In Naturalism in Question, edited by
M. De Caro and D. Macarthur, 21–35. Cambridge, MA: Harvard University
Press, 2004.
Wittgenstein, L., Tractatus Logico-Philosophicus, translated by C. K. Ogden.
London: Kegan-Paul, 1922.
———, Lectures and Conversations on Aesthetics, Psychology and Religious
Belief, edited by Cyril Barrett. Berkeley: University of California Press, 1967, 32.
———, Zettel, 2nd ed., translated by G. E. M. Anscombe. London: Blackwell,
1967.
———, The Blue and Brown Books. Oxford: Blackwell, 1969.
———, On Certainty, edited by G. E. M. Anscombe and G. H. von Wright, translated by Denis Paul and G. E. M. Anscombe. Oxford: Blackwell, 1969.
———, Philosophical Grammar, edited by Rush Rhees, translated by Anthony
Kenny. Oxford: Blackwell, 1974.
———, Philosophical Investigations, 4th ed., translated by G. E. M. Anscombe,
Peter Hacker and Joachim Schulte. Oxford: Blackwell, 2009.

3

Naturalism in the
Goldilocks Zone
Wittgenstein’s Delicate
Balancing Act
Daniel D. Hutto and Glenda Satne

1. Wittgenstein and the Three Naturalisms
Naturalism comes in many stripes and strengths. At one extreme, naturalists of a super strict sort advance a unification agenda along combined
methodological and ontological fronts. They hold that all legitimate
knowledge of the world must be the product of a single explanatory
enterprise and that what is discovered through such scientific endeavors
must conform to a single metaphysical scheme. Such a program seeks to
collapse philosophy into science and all sciences into hard science, and
presents an accompanying austere vision of nature. Arguably, this view
of nature is impoverished in several ways and, moreover, it is untenable
in that it lacks the resources even for making sense of the practice of hard
science.
At the other extreme, by way of response to the confining and austere
conception of nature proposed by strict naturalists, liberal naturalists are
pluralists who accept the existence of diverse ways of knowing and entities other than those recognized by the sciences. Such liberals putatively
retain their naturalistic credentials because they draw the line at admitting supernatural entities into the furniture of reality. Arguably, this view
of nature is overly permissive because it fails to make the connections
between different renderings and realms of reality intelligible, leaving us
with a fractured understanding of the world and a picture of it as fundamentally divided.
Neither of these naturalisms is satisfactory. Both generate intractable
problems because they mischaracterize the relevant natural facts. This
paper argues that in Wittgenstein’s philosophy we find a third version of
naturalism that proves more habitable—one that occupies a Goldilocks
Zone and thus provides the right conditions for understanding how philosophy and the sciences can be distinct and yet productively connected.
We argue that the kind of naturalism we find in Wittgenstein’s work is
the most balanced sort of naturalism, the best way to avoid the problematic features of the more extreme versions of strict or liberal naturalism.
After exposing the basic contours of Wittgenstein’s naturalism, we argue

Goldilocks Zone

57

that an enriched version of it can overcome the problems faced by its
competitors.

2. Strict Naturalism
The strictest of naturalists are motivated by an uncompromising unification agenda. They insist that any bona fide naturalist must subscribe to
such an agenda. By their lights, card-carrying naturalists have no choice
but to use the methods of the hard sciences—and those alone—as a
means of acquiring knowledge.
For example, Rosenberg holds that “science can’t accept interpretation
as providing knowledge of human affairs if it can’t at least in principle be
absorbed into, perhaps even reduced to, neuroscience”.1 And, ideally, the
fate of neuroscience—like all special sciences—will be to be, at least in
principle, absorbed into, perhaps even reduced to, a single hard science, a
yet-to-be articulated complete physics. The aim to reduce all sciences to
one hard science is driven by the demand that “natural science requires
unification”.2
Within philosophy, the situation is the same: it can only provide
knowledge insofar as it uses the methods of the hard sciences. Carruthers, for example, sees himself as conducting his philosophical work
in precisely this vein: he regards it as empirically sensitive theory building and testing. Focusing on the philosophy of mind, Carruthers holds
that it is not just continuous with but “an exercise in theoretical psychology (Compare theoretical physics, which uses other people’s data to
develop and test theories) . . . it is one that naturalistically-inclined philosophers of the present will recognize as a kind of philosophy. Indeed,
in my view it is a mistake to address questions in the philosophy of
mind in any other way”.3
Extreme naturalists of this kind connect their methodological demands
with particular ontological commitments. They hold that reductionist
explanations are required of any phenomenon if it is to qualify as natural, and thus only the hard sciences decide questions about what there is
in nature. They maintain that, “reality contains only the kinds of things
that the hard sciences recognize”.4 As explained, the unification agenda
results in zero tolerance for the recognition of any phenomenon that will
not reduce to the hard sciences.
1 Alexander Rosenberg, “Can Naturalism Save the Humanities,” in Philosophical Methodology: The Armchair of the Laboratory? edited by Matthew C. Haug (London: Routledge, 2014), 41.
2 Ibid., 41.
3 Peter Carruthers, The Opacity of Mind (Oxford: Oxford University Press, 2011), xiii.
4 Alexander Rosenberg, “Why I Am a Naturalist,” in Philosophical Methodology: The
Armchair or the Laboratory? edited by Matthew C. Haug (London: Routledge, 2014), 32.

58

Daniel D. Hutto et al.

A familiar criticism of extreme naturalism is that its unification maxim
is philosophically motivated and hence imposes extra-scientific ideologically driven constraints on scientific inquiry—constraints that we have
little or no reason to suppose will pay off in the end.5 Thus, De Caro
and Macarthur observe that “the ontological doctrine is a metaphysical
thesis not a scientific one”.6 These observations connect with the serious
concern that the austere program of strict naturalism is too restrictive
about what counts as genuine scientific inquiry. In using Occam’s razor
injudiciously and incautiously, it cuts away too much—indeed, so much
that it leaves itself bereft of the resources needed for making sense of the
practices of the hard sciences.
Williamson advances this damning worry in the form of a sharp incompleteness argument against strict naturalism’s idea that all truths are discoverable only by hard science. He formulates the argument as follows:
If it is true that all truths are discoverable by hard science, then it is discoverable by hard science that all truths are discoverable by hard science.
But it is not discoverable by hard science that all truths are discoverable
by hard science . . . ‘Are all truths discoverable by hard science?’ is not
itself a question of hard science. Truth is a logical or semantic property,
discoverability an epistemic one, and hard science a social process.7,8
On the assumption that these analyses hold good, strict naturalism is a
self-defeating philosophical program that faces insuperable internal difficulties. In short, Strict Naturalism is just too strict.

3. Liberal Naturalism
Acknowledging the fatal shortcomings of a scientific Strict Naturalism, a
number of philosophers have proposed an alternative “liberal or pluralistic form of naturalism”.9 Liberal Naturalism rejects Strict Naturalism’s
5 See Steven Horst, Beyond Reduction: Philosophy of Mind and Post-Reductionist Philosophy of Science (Oxford: Oxford University Press, 2007), 21.
6 Mario De Caro and David Macarthur, “Introduction: Science, Naturalism, and the Problem of Normativity,” in Naturalism and Normativity, edited by Mario De Caro and
David Macarthur (New York: Columbia University Press, 2010), 4.
7 Timothy Williamson, “The Unclarity of Naturalism,” in Philosophical Methodology:
The Armchair or the Laboratory? edited by Matthew C. Haug (London: Routledge,
2014), 37.
8 Macarthur (2015) makes a similar argument, arguing that strict naturalism in stating its
methodological and ontological maxims “is committed to a form rational normativity
[what is rationally normative for a philosophical practice] that it cannot explain with the
resources available to it” (p. 576).
9 Mario de Cario and David Macarthur, “Introduction: The Nature of Naturalism,” in
Naturalism in Question, edited by Mario De Caro and David Macarthur (Cambridge,
MA: Harvard University Press, 2004), 3.

Goldilocks Zone

59

central credo, namely the idea that “science is, or ought to be, our only
genuine or unproblematic guide in matters of method or knowledge or
ontology or semantics”.10 Consequently, they also reject the restrictive
vision of nature proffered by Strict Naturalists, asserting that “there are
whole domains of fact with respect to which present-day science tells us
nothing at all”.11
Liberal Naturalism—with a capital L and N—is a broad church movement that is still developing. In 2004, its progenitors spoke of it as a
position that was being actively articulated, describing their initial offerings as a kind of ‘roadmap’—one that originally aimed to provide the
“outlines of a new non-reductive form of naturalism and a more inclusive
conception of nature than any provided by the natural sciences”.12
Apparently, Liberal Naturalists abound—at least in philosophy. According to its chief spokespersons, De Caro and Macarthur, anyone who propounds some or other brand of non-scientific naturalism or non-standard
scientific naturalism makes the cut. They identify John Dupré (for his
‘pluralistic naturalism’), Jennifer Hornsby (for her ‘naive naturalism’),
John McDowell (for his ‘liberal naturalism’), and Barry Stroud (for his
‘more open minded or expansive naturalism’) among its ranks.13 Notably, Hilary Putnam, Huw Price, Akeel Bilgrami, Stanley Cavell, and Carol
Rovane, amongst others, also make the list.14
Liberal Naturalists are quite a motley crew. Some, like Dupré, allow
that science should be our only guide to what exists but claim that
“the limits of the ‘scientific’ are broader, and looser, than the orthodoxy
suggests”.15 Others, like McDowell, attempt to articulate a different conception of nature than that propounded by the kinds of naturalism that
takes the standard scientific image as their touchstone.
Despite these differences, De Caro and Macarthur claim that four features form the core of Liberal Naturalism.16 First, Liberal Naturalists tend
to focus on human nature, which they take to have unique characteristics, rather than concerning themselves with other aspects of the natural
world. Second, Liberal Naturalism acknowledges and seeks to incorporate non-reducible normative facts within the broader conception of
nature that it propounds. Third, Liberal Naturalists, although they reject
the idea of a first philosophy that grounds all domains of knowledge,

10 Ibid., 17.
11 Hilary Putnam, Realism With a Human Face (Cambridge, MA: Harvard University
Press, 1990), 143.
12 De Caro and Macarthur, “Nature of Naturalism,” 1.
13 Ibid., 14.
14 David Macarthur, “Liberal Naturalism and Second-Personal Space: A Neo-Pragmatist
Response to the ‘Natural Origins of Content’,” Philosophia, 43 (2015): 574, fn 19.
15 De Caro and Macarthur, “Nature of Naturalism,” 10.
16 Ibid.

60

Daniel D. Hutto et al.

claim that philosophy is autonomous and has a distinct methodology.
Finally, Liberal Naturalists are ardent pluralists: They advance pluralist
conceptions of both scientific and philosophical methods.
On the preferred version of Liberal Naturalism, advanced by its chief
spokespersons, it allows that there can be a multiplicity of kinds of things
in nature, and a multiplicity of legitimate ways of understanding and
investigating them. Thus some followers of Liberal Naturalism even go
as far as taking “seriously the plurality of explanations within the various
sciences and ordinary discourse, as well as the plurality of kinds of objects
to which they are committed”.17 Accordingly, one is not “beholden to fit
one’s ontology to the scientific image, Liberal Naturalism can admit . . .
non-scientific non-supernatural realities into the catalogue of the things it
acknowledges—or what we might (with some reservations) call its ‘ontology’ in a deflationary sense of that term”.18
There are two important general problems with Liberal Naturalism:
First, it has difficulty in making its metaphysical claims coherent and
keeping its metaphysics under control. Second, it does not provide an
adequate picture of how philosophy and various sciences connect.
With respect to the first concern, Macarthur’s talk of a deflationary
sense of ontology reveals his uneasiness with an out-and-out acknowledgment of a ‘plurality of kinds of objects’. However, not only does
such a gesture not address the underlying issue, it also does not explain
why or how we are justified in ruling out some but not other entities from our ontology. Without some restrictions on our methods, the
worry is that Liberal Naturalism will lead to a bloated and overpopulated ontology. On the face of it, Liberal Naturalism sports an insufficiently cautious and overly free attitude, one that risks metaphysical
extravagance.
Failure to deal with this problem makes it difficult to see why Liberal
Naturalism should be counted as a form of naturalism at all. Liberal
Naturalism suffers from the fact that without the backing of the sciences,
it appears to lack a principled criterion for demarcating what is natural
from what is not. To the extent that Liberal Naturalism allows ontological questions to be decided entirely by local standards, it appears to be
stuck when it comes to deciding, in a final tally, what is in and what is out
of nature in a principled way.
It should be noted that Liberal Naturalists cannot simply look to the
natural sciences to solve this problem, as De Caro and Voltolini do.19
Merely ruling out anything that contradicts the laws of nature potentially
17 Talia Morag, Emotion, Imagination, and the Limits of Reason (London: Routledge,
2016), 11, emphasis added.
18 See Macarthur, “Liberal Naturalism,” 574.
19 Mario De Caro and Alberto Voltolini, “Is Liberal Naturalism Possible?” in Naturalism
and Normativity, edited by Mario De Caro and David Macarthur (New York: Columbia University Press, 2010).

Goldilocks Zone

61

lets far too much into the ontology, for the natural will include any entities that make no attempt to intervene on the causal order or only do so
indirectly, without violating laws of nature.
A more promising strategy for dealing with the demarcation problem would be to reform the notion of nature offered by the sciences.
This is McDowell’s approach. In reacting against the excesses of an
imperialistic scientific naturalism, McDowell is steadfast in resisting
the demand that in order to qualify as bona fide natural phenomena,
normative phenomena, for example, must be explained in scientific
terms.
Focusing on the phenomena of meaning and understanding—like
other liberal naturalists—McDowell takes these to be wholly natural
phenomena despite the fact that they have an irreducibly normative
character—a character which, for McDowell, cannot be explicated in
non-normative terms.
By McDowell’s lights, it is simply a fact that certain creatures—those
that have benefited from being initiated into special kinds of social
practices—become capable of having meaningful or contentful states of
mind. For him, through such a process of enculturation such creatures
obtain a second nature and enter into the space of reasons.20
A key feature of McDowell’s view is that the space of reasons is autonomous and beyond the explanatory reach of the sciences. Thus McDowell
makes clear that “we must sharply distinguish natural-scientific intelligibility from the kind of intelligibility something acquires when we situate
it in the logical space of reasons”.21 If McDowell is right, then we are not
capable of explicating or explaining how it is that capacities for contentful thought come to obtain in nature. Foregoing explanation, we can only
simply affirm “our right to the notion of second nature”.22
This brings us face to face with the fundamental general problem for
Liberal Naturalism: It is that the picture of nature offered by the liberals
divides it into diverse domains of entirely disconnected fact.
Despite first appearances, McDowell’s liberal naturalism is not satisfying. It leaves us with two senses of nature—demarcated by the manifest
image, on the one hand, and the scientific image, on the other—which are
forever divided.
The price of buying into Liberal Naturalism is that it becomes unclear
how different domains of reality relate, or even why one should think
that they are connected at all. Liberal Naturalists acknowledge a plurality
of explanations about diverse domains of fact, but they apparently lack

20 See John McDowell, “Wittgensteinian Quietism,” Common Knowledge, 15/3 (2009):
287.
21 John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994),
xix.
22 Ibid., 94–95.

62

Daniel D. Hutto et al.

the resources for understanding how different explanations about diverse
realities are complementary rather than competing.23
Liberal Naturalists stress that pluralism does not equate to the view
that science cannot make valuable contributions to our understanding of
nature. For example, they do not reject the idea that philosophy might
be informed by science. As such, they “don’t want to deny that [scientific] findings may provide the impetus to philosophical reflection, or that
they may help to undermine one’s philosophical conclusions”.24 Still, it is
entirely unclear how such a dialogue is meant to work given the vision of
a divided nature that they sketch.
Anyone aiming to provide a satisfying Liberal Naturalism has to live
by the ‘hope’ that they can “bring the relations between these two senses
of nature into better focus”.25 How might this be done? De Caro and
Macarthur suggest that a way of bringing the two images into better
focus would be by revising McDowell’s thesis of autonomy of the space
of reasons “on the basis of more nuanced conceptions of the scientific
image and the manifest images”.26 Yet, as things stand, we have yet to
see a detailed account of what such nuanced revisions would look like,
and until such details are on the table, it would not be unreasonable to
suspect that they might not be forthcoming.
In the end, Liberal Naturalism offers an important corrective to Strict
Naturalism. It gets methodological pluralism right, but struggles when
it comes to ontological issues as well as in making the appropriate connections between philosophy and science. Still, all in all, in key respects,
Liberal Naturalism is just too Liberal.

4. From Liberal to Liberating Naturalism
Is there a version of naturalism that can avoid reductionism while embracing ontological monism? Could a naturalism do that while also respecting methodological pluralism in philosophy and the sciences? Moreover,
is there a kind of naturalism that could achieve all of the above while also
showing how philosophy and the sciences can connect productively and
enter in cooperative dealings? As the previous sections reveal, to achieve
all of this at once would be a delicate balancing act.
We think that there is a kind of naturalism that can pull off such a feat.
Such a naturalism has two important components. Firstly, it must liberate us from certain mystifying tendencies of thought: hence, it must be
23 Accordingly, “liberal naturalism does not privilege causation at the physical level, as if
physical causes are the only ‘real’ causes. The same event may be causally explained by
various explanatory practices, each of which would account for an aspect of the event
that is describable in terms of that specific practice” (Morag 2016, p. 11).
24 De Caro and Macarthur, “Nature of Naturalism,” 15.
25 De Caro and Macarthur, “Introduction,” 16.
26 Ibid., 17.

Goldilocks Zone

63

a liberating naturalism. Secondly, it will make clear how philosophy and
various sciences, using their own special methods, can take an interest in
the same subject matter in a relaxed, complementary manner: hence, it
must be a relaxed naturalism.
In this section, we take the first step towards articulating the kind of
liberating, relaxed naturalism that we think ought to be preferred over
its rivals. We postpone the question of how to best deal with questions
of ontological monism and methodological pluralism until the next section. Our first task is to introduce the idea of a liberating naturalism,
showing in what crucial respects it differs from liberal naturalism. In this
connection, we look to Wittgenstein for inspiration. As Pears observes,
Wittgenstein’s later philosophy promotes a naturalism that is inspired
not by maxims of unification or pluralism, but by the idea “that the right
method in philosophy is not to theorize about things but to describe them
as we find them in daily life”.27
Descriptions of the kind Wittgenstein provided serve as elucidations—
they help us to see what needs to be seen: facts that lie right before our
eyes but which can, due to our own tendencies, become blocked from our
view. Philosophers engaged in philosophy, as Wittgenstein conceived of it,
provide clarifications but do not add substantively to our knowledge of
the world; they do not discover new facts but help us to see aright facts
with which are already familiar.
Wittgenstein’s reminders—his perspicuous representations—are designed
to put us back into contact with relevant facts about our everyday practices and customary ways of going on. They remind us of:
the fact that we act in such-and-such ways, e.g. punish certain actions,
establish the state of affairs thus and so, give orders, render accounts,
describe colours, take an interest in others’ feelings.28
The facts in which Wittgenstein is interested are already quite familiar to
us. This is because they are part and parcel of what it is to be a competent
participant in the relevant practices. For this reason, Wittgenstein directs
us to attend to what “lies before our eyes” and to recognize that “What
has to be accepted, the given—it might be said—are facts of living [forms
of life]”.29,30
Supplying illuminating descriptions that clarify the facts of our situation is at the heart of Wittgenstein’s brand of naturalism. His kind of
27 David Pears, “Wittgenstein’s Naturalism,” The Monist, 78/4 (1995): 411.
28 Ludwig Wittgenstein, Remarks on the Philosophy of Psychology, Volume 1 (RPP1),
edited by G.E.M. Anscombe and G.H. Von Wright (Oxford: Basil Blackwell, 1980),
§630, emphases added; Philosophical Investigations (PI) (Oxford: Basil Blackwell,
1953), 226e.
29 Wittgenstein, PI, §415.
30 Wittgenstein, RPPI, §630, emphases added; PI, 226e.

64

Daniel D. Hutto et al.

naturalism is, in this crucial respect, unlike scientific forms of naturalism
that seek to provide penetrating explanations of various phenomena. It
is descriptive work that is needed since only such work can acquaint
us with the important facts needed to gain philosophical understanding.
Such activity is not speculative: it is not a matter of advancing hypotheses
or theories.
Wittgenstein always emphasizes the absolute gulf between the descriptive work of philosophy and the advancing of explanations and conjectures.31 In doing philosophy, Wittgenstein instructs, “we may not advance
any kind of theory. There must not be anything hypothetical in our considerations. We must do away with all explanation, and description
alone must take its place”.32 Famously, Wittgenstein tells us, “Philosophy simply puts everything before us, and neither explains nor deduces
anything.—Since everything lies open to view there is nothing to explain.
For what is hidden, for example, is of no interest to us”.33
Indeed, Wittgenstein highlights the ways in which coming to see things
aright can only be achieved by putting aside philosophical commitments
and explanatory ambitions. Thus, he tells us, “It often happens that we
only become aware of the important facts, if we suppress the question
‘why?’ ”.34 Certainly, for Wittgenstein, it is not the job of the philosopher
to attempt to provide explanations of facts of interest.
Though the relevant facts are often obscured, through philosophical
work they can become “surveyable by a rearrangement”.35 Just as the
weeds of an overgrown garden need to be removed if we are to see its
beauty, so we gain philosophical clarity about some subject matter by
“clearing misunderstandings away”.36 Importantly, though, the philosophical revelations achieved at the end of this process need no further
vindication; they are not established by appeal to additional evidence.
It is for this very reason that in conducting his philosophical investigations, Wittgenstein tells us that he is neither “doing natural science, nor
yet natural history”.37,38 This is a point to which we will return to discuss
at length in the next section.

31
32
33
34
35
36
37
38

Wittgenstein, PI, §§654–656.
Ibid., §109.
Ibid., §126.
Ibid., §471.
Ibid., §92.
Ibid., §90.
Ibid., 230e.
Equally Wittgenstein was not trying to divine a priori necessities—he wasn’t doing
traditional metaphysics. Indeed, by his lights, the essential thing about metaphysics
is that it, problematically, “obliterates the distinction between factual and conceptual
investigations” (Zettel §458). Wittgenstein’s naturalistic project turns its face against
metaphysical attempts to divine the essence of things through an analysis of what is
hidden and all attempts to discover a set of explanatory super-facts that fix all other

Goldilocks Zone

65

Despite being atheoretical in character, the philosophical work of getting clear about the facts of our situation is anything but easy. A major
reason for this is that in doing philosophy the reason we often lack a
clear view of the terrain of interest is because intellectual obstacles of
our own manufacture impair our vision. Assumptions about what must,
or has to be the case can prevent us from understanding a topic under
investigation.
Wittgenstein everywhere denounces the practice of making philosophical pronouncements about ‘what there must be’. Picking up again on the
difference between description and explanation in decrying this tendency,
he tells us, “Grammar does not tell us how language must be constructed
in order to fulfil its purpose, in order to have such-and-such an effect on
human beings. It only describes and in no way explains”.39 Hence, he
enjoins, “Don’t think, but look!”40
To see things aright requires us to break free, often again and again—
from certain irresistible but distorting pictures or ways of thinking about
various subject matters—ways of thinking that irresistibly attract us and
“bewitch our intelligence”.41
Gripping pictures can block our view. Describing his own struggles with
these, Wittgenstein tells us: “a picture held us captive. And we could not
get outside it, for it lay in our language and language seemed to repeat it
to us inexorably”.42
Recognizing the power of philosophical pictures and how hard it can
be to get free of their grip is pivotal when it comes to understanding
Wittgenstein’s distinctive brand of naturalism and how it differs fundamentally from Liberal Naturalism. Illustrating the difference, Marie
McGinn provides an apt, incisive criticism of the motivations behind
and central commitments of McDowell’s liberal naturalism, emphasizing how and where McDowell’s thinking deviates from Wittgenstein’s on
the issues of meaning and understanding.43 Despite acknowledging the
attractive aspects of McDowell’s philosophy, McGinn maintains that, on
close scrutiny, the kind of liberal naturalism he advocates breaks faith

39
40
41
42
43

facts. An animating insight that drives his philosophy—one that is shared with other
naturalisms—is his recognition that there is no necessary a priori order to things. By
his lights, there is no possibility of pursuing an analytic project of charting, surveying
or articulating metaphysics. See Ludwig Wittgenstein, Zettel, translated by G.E.M. Anscombe, edited by G.E.M Anscombe and G.H. Von Wright, 2nd edition (Oxford: Basil
Blackwell, 1967).
Wittgenstein, PI, §496.
Ibid., §66.
Ibid., §109.
Ibid., §115.
Marie McGinn, “Liberal Naturalism: Wittgenstein and McDowell,” in Philosophical
Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug (London:
Routledge, 2014).

66

Daniel D. Hutto et al.

completely, both in letter and spirit, from the sort of naturalism Wittgenstein promotes.
On McGinn’s analysis, what really divides these two thinkers boils
down to the fact that McDowell’s liberal naturalism, presumptuously,
continues to incorporate elements of a philosophical picture of what
having meaningful thoughts and understanding must involve, rather
than stopping with and being satisfied with a careful description of our
practices that involve meaning and understanding. Indeed, McGinn sees
liberal naturalism of a McDowellian stripe as embracing philosophical
commitments of exactly the sort that a liberating naturalism of the kind
inspired by Wittgenstein seeks to purge and expose as both mystifying
and unnecessary.
What are the offending commitments? A full, detailed analysis is beyond
the scope of this paper, but McGinn complains of McDowell’s insistence
that during acts of meaning and understanding we somehow come into
contact with mental contents—mental items with very special properties.
Moreover, he holds that anyone who fails to acknowledge this fact of our
situation will be bereft of the resources needed to properly characterize
our everyday acts of meaning and understanding.
McDowell holds that “someone’s understanding something is a fact
about her”.44 For him, philosophical work is needed to say just “what
sort of fact” this is—namely, to appropriately characterize it, casting
aside problematic candidates. In his final analysis, McDowell holds an
appropriate characterization must recognize the existence of mental contents that come before our minds and that failure to recognize this would
be to overlook an essential feature of acts of meaning and understanding.
In his own words, any account of meaning and understanding that failed
to acknowledge that meanings come before our minds would be to risk
‘under-mentalizing’ the relevant behavior.45
For McGinn, the problem with liberal naturalism lies, at root, precisely
in this conviction about what acts of meaning and understanding must
essentially involve. Summing up, as McGinn puts it, “in holding that we
need to preserve the idea that ‘meanings come to mind’, McDowell is
committed to the very idea of shadows that Wittgenstein’s naturalism is
intended to liberate us from”.46
For McGinn, to buy into the idea that we grasp mental contents and
bring meanings to mind, as McDowell does, is to resurrect the very picture about what following a rule actually comes to in practice that Wittgenstein sought to expose as a superfluous superstition.
44 John McDowell, “Meaning and Intentionality in Wittgenstein’s Later Philosophy,” in
Mind, Value and Reality (Cambridge, MA: Harvard University Press, 1998), 269.
45 Ibid., 276; see also John McDowell, “Intentionality and Interiority in Wittgenstein,” in
Mind, Value and Reality (Cambridge, MA: Harvard University Press, 1998).
46 McGinn, “Liberal Naturalism,” 73–74, emphasis added.

Goldilocks Zone

67

To be sure, if McDowell has fallen into error here, his mistake is a very
subtle one. When pressed to explain what meaning, understanding or following a rule involves, as Wittgenstein was at pains to expose, a natural
response is to try to articulate the rule to which we are attempting to
be faithful: the one that governs our particular attempts to apply it and
which—somehow—stands over and above and governs such attempts.
This taps into the platitudinous fact that getting it right depends on successfully aligning our behavior with standards which provide independent correctness conditions.
When we mean something by using a word or understand what a word
means we are accountable to something beyond us—something that
places a binding force on whether or not we have meant something by
a word or have understood it correctly. The relevant possibilities are all
somehow timelessly ‘laid out in advance’, predetermining all correct and
incorrect applications—past, future and present.
To explain how this can be so it seems necessary to postulate the existence of something independent from our practices—something above
and beyond our practices, something that can act as a standard that
determines whether any given application is correct or incorrect. Moreover, we will feel compelled to posit a means of access to these standards;
we need to have a grip on what the rule demands of us if it is to instruct
and guide our attempts to apply it.
Through these familiar steps of reasoning, it is easy to see how anyone might come to think that we must bring mental contents before our
minds if we are to mean or understanding anything by our words. Such
mental contents will need to have ‘queer’ properties if they are to simultaneously fix meanings and act as guides to our acts of meaning and
understanding, and our minds will need equally ‘queer’ properties if they
are to get a grip on such things. Trying to imagine how our minds operate
with such mental contents would be akin to trying to imagine operations
of some kind of ‘ethereal machine’, one that cannot err or malfunction.
In trying to articulate the philosophical picture behind such thinking
we seem to need to posit the existence of a device that can do what no
ordinary physical mechanism can, for only such a mysterious mechanism
could embody a rule.47
McDowell, as a sophisticated reader and exegete of Wittgenstein on
topics of meaning, understanding, and rule following, is well aware of
the problems attendant with the kind of Platonism associated with the
super-rigid machine notion. As he notes, such a picture is problematic in
that it fosters the idea that it is possible to grasp “patterns that extend
to new cases independently of our ratification”.48 But, at the same time,
47 PI, §193–194.
48 John McDowell, “Wittgenstein on Following a Rule,” in Mind, Value and Reality (Cambridge, MA: Harvard University Press, 1998a), 256, emphasis added.

68

Daniel D. Hutto et al.

he recognizes the problem of placing too much weight on the anti-realist
idea that the patterns in question depend entirely on our practices.
McDowell offers a transcendental argument which he takes to reveal
that old-school Platonism and anti-realism cannot be our only options:
“there must be a middle position”.49 On his analysis, he sees what he
has to offer on this score as wholly in tune with what Wittgenstein has
to say on the matter. According to McDowell, Wittgenstein concludes
“not that concepts have no normative status, but that the patterns they
dictate are not independent of our ratification”.50 To see things aright,
what has to be rejected is the erroneous idea that “a genuine fact must
be a matter of the way things are in themselves, utterly independently
of us”.51
Clarifying these possibilities, McDowell distinguishes two things that
can travel under the banner of Platonism, one is a philosophically problematic view connected with the “imagery of super-rigid machinery”,
whereas the other only propounds ideas that are “simply part of the conception of meaning as reaching normatively into the objective world”.52
As McDowell observes, it is possible to recoil from the first kind of
Platonic realism while embracing the second version, since it has “nothing to do with rejection of the truth-conditional conception of meaning,
properly understood. That conception has no need to camouflage the
fact that truth-conditions are necessarily given by us, in language that we
understand”.53
McDowell promotes the second brand of Platonism because he takes it
to be that “a condition for the possibility of finding real application for
the notion of meaning at all is that we reject anti-realism”.54 For him,
“there cannot be a position that is both anti-realist and genuinely hospital to meaning”.55
Thus, to avoid an unwelcome anti-realism McDowell recommends
embracing a kind of Platonism—albeit, one that under the auspices of
Liberal Naturalism, can be billed as a “naturalized Platonism”.56 Thus,
as explored already in the previous section, the long and short of it is that
this commitment is what inspires McDowell’s particular brand of liberal
naturalism—one that seeks to promote a:
form of realism which conceives of meaning and intending as inner
states which have an intrinsic determinate content. His conception of

49
50
51
52
53
54
55
56

Ibid., 256, emphasis original.
Ibid., 248.
Ibid., 254.
McDowell, “Meaning and Intentionality,” 273.
McDowell, “Wittgenstein on Following a Rule,” 255.
Ibid., 249.
Ibid., 262.
McDowell, “Wittgensteinian Quietism,” 88.

Goldilocks Zone

69

liberal naturalism is precisely that it can embrace these intrinsically
normative states as part of the furniture of the natural world.57
There is much to agree with and admire in McDowell’s analysis and
what it says about meaning and understanding. For example, we think
McDowell is right to think that the capacity to think contentful thoughts
is only gained if we have acquired a second nature. Only those that have
benefited from the right kind of training gain this second nature. Thus,
only those who have been trained and inducted into special sorts of
socio-cultural practices are able, for example, to mean things by their
words and to understand the meaning of words. It is through immersion
and participation in specific sorts of transformative practices that some
creatures become capable of such mental feats.
Indeed, as these kinds of naturalistic observations go, McDowell and
Wittgenstein are much aligned; for both meaning and understanding are
“natural responses of those who have undergone a certain training”.58
Thus, McDowell holds that being able to bring a content ‘before one’s
mind’ is not a basic capacity, as many representationalists suppose, but
an acquired one: it is the outcome of a special kind of enculturation. Thus
he agrees with Wittgenstein that grasp of meaning is not a basic ability
but instead that “a process of initiation into a custom is required in order
for anyone to perceive what the rule requires”.59 Thus, for McDowell,
to be able to think contentful thoughts depends on being appropriately
embedded in and having mastered a special set of practices.
We agree with McDowell in acknowledging the crucial enabling roles
that public practices play in making contentful thought possible. Despite
this, we hold that to adopt his new school naturalized Platonism goes
beyond Wittgenstein in attempting to add unnecessary philosophical elements in the characterization of our meaningful practices. In particular,
there is no need to recognize the existence of special kinds of irreducible
normative facts above and beyond the general facts of nature of which
Wittgenstein speaks in order to provide a satisfactory account of meaning and understanding.
It is precisely at this juncture that McDowell moves away from a liberating naturalism and, as a consequence, feels the need to expand our
conception of nature.60 If McDowell is right, we must introduce a special

57
58
59
60

McGinn, “Liberal Naturalism,” 67, emphasis added.
Ibid., 72.
Ibid., 70.
Thus, as McGinn (2014) observes: “It is no doubt a contingent matter that there are
natural beings that acquire the status of rational animals, but given this is the case, we
must, McDowell argues, have a conception of nature that can accommodate the existence of inner states with normative force, for without that the whole idea that there
are rational animals, whose action are describable in normative terms, is undermined”
(p. 68, emphasis added).

70

Daniel D. Hutto et al.

ingredient into the mix in order to do justice to the story—to tell it
wholly. By his lights, we have no choice but to acknowledge that, albeit in
a special sense, thinking and understanding meaningful thoughts involves
grasping contents that come “to mind”.61 By presenting the issues in this
way, it appears that further metaphysical tinkering and theoretical work
is required in order to accommodate the normative facts within our existing picture of nature.
This is the crux. A more careful examination of our actual practices
reveals that there is no need to introduce irreducible, inexplicable properties into the story in the first place: they play no part of meaning and
understanding, and to think otherwise is to be swayed by an illegitimate
philosophical demand. Here again, McGinn astutely sums up the situation: “There is no need to make this strange idea intelligible . . . Nothing
is lost by abandoning it”.62
In the end, we agree with McGinn’s assessment that McDowell’s
liberal naturalism is steeped in philosophical mythology—that what
“McDowell’s liberal naturalism sets out to preserve, amounts to a commitment to the ‘shadowy beings’ which Wittgenstein’s naturalism aims to
overcome”.63
The salient lesson to be taken away from the preceding analysis is that
even for the sharpest minds, misguided philosophical demands can generate unwelcome commitments that can be quite subtly tangled up with
more innocuous, indeed platitudinous, observations about the relevant
facts.
This can be so even for those who are actively on guard against such
unwanted influences.64 To avoid such a result, what is wanted is not a
Liberal Naturalism but a Liberating Naturalism—one that leads back to
a more open conception of nature that does not expand nature to include
a variety of special sorts of facts but rather places the relevant facts in
plain view, once obscuring obstacles to our sight have been removed.

61 McDowell, “Intentionality and Interiority,” 304.
62 McGinn, “Liberal Naturalism,” 84. Expanding on this thought she writes: “If we are
tempted to suppose that [rule-following, meaning and understanding] . . . depends on
the subject being in a mental state which mysteriously anticipates the future, then Wittgenstein’s descriptions of what goes on in particular cases—of the criteria by which we
judge these things, of the context in which we learn to use the relevant expressions, and
so on—can enable us to see that no such idea comes into it” (McGinn 2014, p. 84).
63 Ibid., 67.
64 It is perhaps surprising that we should accuse McDowell of falling into this trap since,
as a prominent quietist, he has utterly disavowed philosophical projects that attempt
to theorize. As Macarthur (2017) observes, “the most fully explicit example of quietist
therapy in contemporary philosophy is John McDowell’s Mind and World, 2nd edition (1996)” (p. 352). Nevertheless, as Macarthur also acknowledges, there are aspects
in presentation of McDowell’s work that suggest “an element of dogmatism that is
entirely lacking in Wittgenstein’s writings” (2017, p. 259).

Goldilocks Zone

71

5. From Restrictive to Relaxed Naturalism
A liberating naturalism—one that gets us to focus on and clarify our
understanding of certain general facts of nature—provides the resources
needed for understanding how philosophy and the various sciences can
be concerned with the same facts differently—bringing diverse interests
to the table and employing diverse methods to pursue those interests—
without conflict. Importantly, unlike Liberal Naturalism, Wittgenstein’s
Liberating Naturalism avoids dividing reality into distinct domains of
fact that cannot be intelligibly connected. Rather, Wittgenstein’s descriptive approach to philosophy operates in a framework that allows us to
think about nature in ways that enable us to see how philosophy and the
sciences can connect and cooperate in complementary endeavors.
In more programmatic terms, this is because Wittgenstein’s Liberating Naturalism can be partnered with what we have elsewhere dubbed
Relaxed Naturalism.65 Relaxed Naturalism, unlike Strict Naturalism does
not assume that all sciences must be unified through reduction but holds
that the diverse methods and findings of the natural and social sciences can
be relevant to investigating and understanding facts of interest. Relaxed
Naturalism draws upon and seeks to harmoniously integrate the discoveries from a wide range of empirical sciences as determined by what is
needed to make best sense of the topic under scrutiny. It allows that it may
be appropriate, and sometimes indeed necessary, to include results that
span not only the hard sciences, but also the human and social sciences.
To take a case in point, consider what is required when it comes to
accounting for the acquisition of our second nature—“the set of skills
and habits that are natural to properly enculturated humans . . . habits of
interpretation, judgment, reflective epistemic responsibility, and sensitivity to relations of entailment and support”.66
Relaxed Naturalists maintain that to fully understand the nature of
these capacities requires synthesizing findings from, inter alia, anthropology, developmental psychology, comparative psychology, cognitive
archaeology, and social neuroscience. Thus, like Godfrey-Smith, we too
“argue for the viability and importance of a form of investigation of these
features that McDowell seeks to resist.”67,68

65 Daniel D. Hutto and Glenda Satne, “The Natural Origins of Content,” Philosophia,
43/3 (2015); see also Daniel D. Hutto and Erik Myin, Evolving Enactivism: Basic
Minds Meet Content (Cambridge, MA: MIT Press, 2017).
66 Peter Godfrey-Smith, “Dewey, Continuity and McDowell,” in Naturalism and Normativity, edited by Mario De Caro and David Macarthur (New York: Columbia University
Press, 2010), 314.
67 Ibid., 314.
68 Resistance arises because “[f]or McDowell, the way in which second nature counts as
natural to us makes it unnecessary to engage in a philosophical project of locating these
features in the world as conceived by science” (Godfrey-Smith 2010, p. 314).

72

Daniel D. Hutto et al.

Crucially, taking Liberating Naturalism as a starting point provides a
philosophical framework for thinking about the relevant facts of nature
in a way that allows for the possibility of the sorts of investigations and
integrative work that Relaxed Naturalists conduct.
When it comes to making philosophical sense of our acquired capacities—
those that define our second nature—a Liberating Naturalism frees us
from tempting but ultimately mystery-generating tendencies of thought
and instead gets us to focus on the facts and features of the relevant
practices. Of course, as McGinn emphasizes, this is “not the end of the
story”.69 We will want answers to further questions about, for example,
how we know what to do in following rules or using our words and how
we decide what counts as doing so correctly and incorrectly, and so on.
At this juncture, we cannot do better, philosophically, than to, once again,
describe our practices with care.
Those who have taken this lesson on board will come to see that the
philosophical investigation of these issues comes to an end with an acceptance that it is a fact of our natural history that these practices have the
shape that they have.
It is vitally important to properly understand the status of these observations about our practices. Are these remarks theoretical in character?
Do we need to provide empirical evidence for these claims? Dromm
identifies a standard interpretation of Wittgenstein’s remarks on language learning—ubiquitous in the literature—that takes him to be doing
just that—namely, advancing empirical claims of a broadly theoretical
character.70
We agree with Dromm (2008) that we must resist understanding Wittgenstein’s remarks as armchair attempts at empirical theorizing.71 Yet
in doing so, we must also be careful to understand the sense in which
Wittgenstein was calling our attention to ‘important possibilities’. For,
although Wittgenstein is not engaged in speculative natural history, he
is seeking to bring very general facts into view: He is interested in “the
correspondence between concepts and very general facts of nature”.72 In
this regard, he makes it clear that “we are not doing natural science; nor
yet natural history—since we can also invent fictitious natural history for
our purposes”.73 It is precisely for this reason that Wittgenstein’s remarks
do require additional empirical evidence to establish their credentials.

69 Marie McGinn, “Wittgenstein and Naturalism,” in Naturalism and Normativity, edited
by Mario De Caro and David Macarthur (New York: Columbia University Press,
2010), 337.
70 Keith Dromm, “Imaginary Naturalism: The Natural and Primitive in Wittgenstein’s
Later Thought,” British Journal for the History of Philosophy, 11/4 (2003).
71 Keith Dromm, Wittgenstein on Rules and Nature (London: Bloomsbury, 2008), xvii.
72 PI, 230e.
73 Ibid., 230e.

Goldilocks Zone

73

Once their status is properly understood, it is apparent why it is a mistake
to suppose that, “the claims established by philosophers depend on the
same kind of empirical support as scientific theories”.74
Nonetheless, even though Wittgenstein is not doing natural science, he
is providing philosophical reminders about facts of our natural history.
Thus he affirms that:
What we are supplying are really remarks on the natural history
of human beings; we are not contributing curiosities however, but
observations which no one has doubted, but which have escaped
remark only because they are always before our eyes.75
The relevant facts are there to see for anyone with eyes to see—that is, for
anyone who is not in the grip of a picture.76
Liberating Naturalists are always in the business of philosophical clarification. Following Wittgenstein’s lead, they provide perspicuous representations of general facts of nature. Yet this work is wholly
compatible with other kinds of philosophical endeavors—in particular,
it can sit alongside attempts to enrich our understanding of these facts
of nature further by synthesizing the contributions of various sciences.
For example, philosophers can help to illuminate “how the ‘habits of
thought and action’ involved in our use of normative concepts relate to
other facts about us and how these habits function as human cognitive
tools”.77 Importantly, “when this philosopher says that such an investigation should mesh with what we learn from science, do not think ‘physics’
when he says ‘science’ ”.78
Under the auspices of Relaxed Naturalism, such philosophical efforts
would take the form of productively engaging in “work of synthesis and
argument integrating ideas and suggestions from many distinct research
traditions”.79 Operating in this synthetic mode, philosophers “exploit and
depend on data, but do not provide new data”.80 Crucially, in doing such
work, “philosophers can contribute a distinct kind of competence”.81
74 David Papineau, “The Poverty of Conceptual Analysis,” in Philosophical Methodology:
The Armchair or the Laboratory? edited by Matthew C. Haug (London: Routledge,
2014), 167.
75 PI, §415.
76 Dromm (2008) holds that what Wittgenstein has to say about “language-learning is not
among those facts that ‘lie before our eyes’ ” (p. 71). Yet arguably, this is a mistake. See
Hutto (2013) for a discussion.
77 Godfrey-Smith, “Dewey, Continuity and McDowell,” 316.
78 Ibid., 316.
79 Kim Sterelny, The Evolved Apprentice (Cambridge, MA: MIT Press, 2012), ix.
80 Ibid., ix.
81 Barry C. Smith, “Philosophical and Empirical Approaches to Language,” in Philosophical Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug
(London: Routledge, 2014), 296.

74

Daniel D. Hutto et al.

Likewise, “if philosophers need empirical input they do best to turn to
practicing scientists”82 (Smith 2014, p. 296). This way of understanding
the distinctive contributions of philosophers and scientists acknowledges
that “interdisciplinary work is hard and to do it one has to start from a
strong disciplinary base. Good fences make good neighbours”.83,84
There is no bar to combining the sort of philosophical work Relaxed
Naturalists seek to conduct with the philosophical work of Liberating Naturalism—namely, “with a deep critique of the standard ways in
which these problems have usually been generated and addressed in the
philosophical tradition”.85,86 Indeed, if the foregoing analyses of other
forms of naturalism are accurate then Liberating, Relaxed Naturalism
may be best placed to pursue what Sellars takes to be the central aim of
philosophy, namely “to understand how things in the broadest possible
sense of the term hang together in the broadest possible sense of the
term”.87
A Naturalism that is both Liberating and Relaxed constitutes a coherent naturalism—one that, when compared with the existing alternatives,
appears to get things just right.

6. Conclusion
Our review of the main variants of naturalism—Strict and Liberal—reveals
that by following Wittgenstein’s lead, we find our way to a third variant, a
Liberating, Relaxed naturalism. It is thus possible for naturalists to inhabit
a fertile, viable zone—a Goldilocks’ zone: an environment in which the
contributions of philosophers and diverse sciences can be brought together

82 Ibid., 296.
83 Ibid.
84 This echoes and respects a constant theme in Wittgenstein’s thinking, from early to late,
which is that “Philosophy is not one of the natural sciences. (The word ‘philosophy’
must mean something whose place is above or below the natural sciences, not beside
them’)” Ludwig Wittgenstein, Tractatus Logico-Philosophicus, translated by C. K.
Ogden (London: Routledge and Kegan Paul, 1922).
85 Godfrey-Smith, “Dewey, Continuity and McDowell,” 315.
86 This reveals why, contra Macarthur (2015), Relaxed Naturalism need not be thought
of as simply a looser kind of Scientific Naturalism. For when allied with Liberating
Naturalism, a Relaxed Naturalism is not just a broader and more pluralistic variety
of Scientific Naturalism that encompasses sciences beyond hard sciences (p. 569). For
a broad version of Scientific Naturalism would, as Macarthur characterizes it, simply
expand the range of sciences and thus methods that might determine which entities to
be included as bona fide parts of nature (see also De Caro and Macarthur 2010, pp.
5–6). And, of course, Relaxed Naturalists not only provide synthesis and argument,
they are in the business of clarification as well.
87 Wilfrid Sellars, “Philosophy and the Scientific Image of Man,” reprinted in Science,
Perception and Reality (London: Routledge and Kegan Paul, 1962/1963), 1.

Goldilocks Zone

75

to improve our understanding of the natural world. This is a healthy
domain in which philosophy’s distinctive work of elucidating, clarifying,
and synthesizing can help us to see how our various contributions to our
ways of understanding the natural world connect and cohere.

References
Carruthers, Peter. The Opacity of Mind. Oxford: Oxford University Press, 2011.
De Caro, Mario, and David Macarthur. “Introduction: The Nature of Naturalism.” In Naturalism in Question, edited by Mario De Caro and David Macarthur, 1–20. Cambridge, MA: Harvard University Press, 2004.
———. “Introduction: Science, Naturalism, and the Problem of Normativity.” In
Naturalism and Normativity, edited by Mario De Caro and David Macarthur,
1–19. New York: Columbia University Press, 2010.
De Caro, Mario, and Alberto Voltolini. “Is Liberal Naturalism Possible?” In
Naturalism and Normativity, edited by Mario De Caro and David Macarthur,
70–86. New York: Columbia University Press, 2010.
Dromm, Keith. “Imaginary Naturalism: The Natural and Primitive in Wittgenstein’s Later Thought.” British Journal for the History of Philosophy 11/4
(2003): 673–690.
———. Wittgenstein on Rules and Nature. London: Bloomsbury, 2008.
Godfrey-Smith, Peter. “Dewey, Continuity and McDowell.” In Naturalism and
Normativity, edited by Mario De Caro and David Macarthur, 304–321. New
York: Columbia University Press, 2010.
Horst, Steven. Beyond Reduction: Philosophy of Mind and Post-Reductionist
Philosophy of Science. Oxford: Oxford University Press, 2007.
Hutto, Daniel D. Wittgenstein and the End of Philosophy: Neither Theory Nor
Therapy. 2nd edition. Basingstoke: Palgrave, 2006.
———. “Enactivism: From a Wittgensteinian Point of View.” American Philosophical Quarterly 50/3 (2013): 281–302.
Hutto, Daniel D., and Erik Myin. Radicalizing Enactivism: Basic Minds Without
Content. Cambridge, MA: MIT Press, 2013.
———. Evolving Enactivism: Basic Minds Meet Content. Cambridge, MA: MIT
Press, 2017.
Hutto, Daniel D., and Glenda Satne. “The Natural Origins of Content.” Philosophia 43/3 (2015): 521–536.
Macarthur, David. “Liberal Naturalism and Second-Personal Space: A NeoPragmatist Response to the ‘Natural Origins of Content’.” Philosophia 43 (2015):
565–578.
———. “On Metaphysical Quietism and Everyday Life.” In The Cambridge
Companion to Philosophical Methodology, edited by Giuseppina D’Oro and
Soren Overgaard, 249–273. Cambridge: Cambridge University Press, 2017.
McDowell, John. Mind and World. Cambridge, MA: Harvard University Press, 1994.
———. “Intentionality and Interiority in Wittgenstein.” In Mind, Value and Reality, 297–324. Cambridge, MA: Harvard University Press, 1998.
———. “Meaning and Intentionality in Wittgenstein’s Later Philosophy.” In
Mind, Value and Reality, 263–278. Cambridge, MA: Harvard University Press,
1998.

76

Daniel D. Hutto et al.

———. “Wittgenstein on Following a Rule.” In Mind, Value and Reality, 221–
262. Cambridge, MA: Harvard University Press, 1998.
McDowell, John. “Wittgensteinian Quietism.” Common Knowledge 15/3 (2009):
365–372.
McGinn, Marie. “Wittgenstein and Naturalism.” In Naturalism and Normativity,
edited by Mario De Caro and David Macarthur, 322–351. New York: Columbia University Press, 2010.
———. “Liberal Naturalism: Wittgenstein and McDowell.” In Philosophical
Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug,
62–85. London: Routledge, 2014.
Morag, Talia. Emotion, Imagination, and the Limits of Reason. London: Routledge, 2016.
Papineau, David. “The Poverty of Conceptual Analysis.” In Philosophical Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug, 166–
194. London: Routledge, 2014.
Pears, David. “Wittgenstein’s Naturalism.” The Monist 78/4 (1995): 411–424.
Putnam, Hilary. Realism With a Human Face. Cambridge, MA: Harvard University Press, 1990.
Rosenberg, Alex. “Can Naturalism Save the Humanities?” In Philosophical
Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug,
39–42. London: Routledge, 2014.
———. “Why I Am a Naturalist.” In Philosophical Methodology: The Armchair
or the Laboratory? edited by Matthew C. Haug, 32–35. London: Routledge,
2014.
Sellars, Wilfrid. “Philosophy and the Scientific Image of Man.” Reprinted in Science, Perception and Reality. London: Routledge and Kegan Paul, 1962/1963.
Smith, Barry C. “Philosophical and Empirical Approaches to Language.” In Philosophical Methodology: The Armchair or the Laboratory? edited by Matthew
C. Haug, 294–317. London: Routledge, 2014.
Sterelny, Kim. The Evolved Apprentice. Cambridge, MA: MIT Press, 2012.
Williamson, Timothy. “The Unclarity of Naturalism.” In Philosophical Methodology: The Armchair or the Laboratory? edited by Matthew C. Haug, 36–38.
London: Routledge, 2014.
Wittgenstein, Ludwig. Tractatus Logico-Philosophicus. London: Routledge, 1922.
———. Philosophical Investigations. Oxford: Basil Blackwell, 1953.
———. Zettel. 2nd edition. Translated by G.E.M. Anscombe, edited by G.E.M
Anscombe and G.H. Von Wright. Oxford: Basil Blackwell, 1967.
———. Culture and Value. Chicago: University of Chicago Press, 1980.
———. Remarks on the Philosophy of Psychology, volume I, edited by G.E.M.
Anscombe and G.H. Von Wright. Oxford: Basil Blackwell, 1980.

Part II

Language
Self, Truth, and Mathematics

4

Sensations, Natural Properties,
and the Private Language
Argument
William Child

Introduction
I start with three questions about our concepts of conscious states.
(1) How do I understand what it is for me to be in pain? (2) How do I
understand what it is for someone else to be in pain? (3) Is the property
of pain that I ascribe to others on the basis of what they say and do the
same as the property I ascribe to myself without evidence?
Wittgenstein’s response to those questions involves a general antiplatonism about properties or standards of similarity. What it is for
one thing to be similar to another, or to have the same property as
another, he thinks, is not dictated by reality itself; it depends on our
classificatory practices and the standards of similarity they embody.
It comes more naturally to us to classify things in some ways than
others. But no property or standard of similarity is intrinsically or
objectively more natural than any other: more natural simplicter.
Wittgenstein’s view stands in sharp opposition to the contemporary
doctrine of natural properties.1 For the natural properties view holds
precisely that there is an objective hierarchy of naturalness amongst
properties, a hierarchy that does not depend in any way on our concepts or practices.
I want to explore the interaction between Wittgenstein’s position about
sensations and sensation language and his anti-platonism. Some antiWittgensteinian views in the philosophy of mind make essential appeal
to the idea of natural properties. What are the prospects for such views?
Do they pose a threat to a broadly Wittgensteinian position about sensations and sensation-language?

1 Locus classicus: David Lewis, “New Work for a Theory of Universals,” Australasian
Journal of Philosophy 61 (1983) and “Putnam’s Paradox,” Australasian Journal of Philosophy 62 (1984). For development and discussion see Theodore Sider, Writing the
Book of the World (Oxford: Oxford University Press, 2011).

80 William Child

1. Wittgenstein’s Position and the Rejection
of Natural Properties
One way of answering the questions with which we started involves what
we might call the model of introspection and identity. On this view the
answers to questions 1 and 2 are, in the broadest outline, as follows:
1. I understand what it is for me to be in pain from the first-person case.
(As Wittgenstein puts this view: ‘it is only from my own case that I
know what the word “pain” means’.)2 I single out a kind of sensation
on the basis of how it feels in the first-person case, and I associate the
word ‘pain’ with sensations of that introspectively-identified kind.
2. I grasp what it is for someone else to be in pain by extension from the
first-person case: what it is for someone else to be in pain is for her
to be in the same kind of conscious state that I am in when I am in
pain. (As Wittgenstein puts the view: ‘if I suppose that someone has a
pain, then I am simply supposing that he has just the same as I have
so often had’.)3
Wittgenstein famously argues against both parts of that account: the
idea that we can single out the property of pain by introspective ostension in the first-person case; and the idea that we can understand the
word in third-person ascriptions by appeal to the idea of another person’s being in the same kind of state as me. At both points, his argument essentially depends on his general opposition to the metaphysics of
natural properties.
Wittgenstein’s opposition to stage 1 of the introspection-and-identity
model is a central aspect of the private language sections of Philosophical
Investigations. A crucial theme of those sections is that one cannot give
meaning to a word for a kind of sensation by introspective ostension, in
a way that does not rely on any links with behavior or external circumstances. That is the burden of Wittgenstein’s discussion of the private
sensation diarist in Philosophical Investigations §258. The point of that
discussion, as I read it, is this. In order to give a meaning to a word, I must
establish a standard of correctness for uses of the word: a standard that
distinguishes between a correct and an incorrect application of the word.
The private linguist thinks she can establish such a standard of correctness by concentrating her attention on the sensation she is having at a
particular point and undertaking to use the word ‘S’ for all sensations of

2 Ludwig Wittgenstein, Philosophical Investigations, 4th edn., ed. P. M. S. Hacker and
Joachim Schulte, trans. G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte
(Oxford: Blackwell, 2009), §293.
3 PI §350. We shall see later how, on this approach, an answer to question 3 falls naturally
out of the answer to question 2.

Sensations

81

the same kind. But, Wittgenstein asks, what is it for something else to be
the same kind of sensation as this one? There are indefinitely many different possible standards of similarity. And no one way of classifying sensations as similar or different is any better, or more correct, than any other.
What it takes for one sensation to belong to the same kind as another is
not determined by the nature of things; it has to be understood by reference to a humanly-created standard of similarity. But in order for there to
be a humanly-created standard of similarity, there has to be a practice of
classification; a technique of applying a word.4 And the private linguist,
Wittgenstein argues, cannot establish a classificatory practice by a oneoff act of introspective association.
Wittgenstein’s argument depends on his general anti-platonist insistence that the world itself does not dictate what it takes for one thing to
belong to the same kind as another. For suppose there is simply a fact of
the matter, independent of our concepts and our practices of classification, about what it takes for one sensation to belong to the same kind as
another. In that case, there will similarly be a practice-independent fact
of the matter about whether the sensation the private linguist has today
is the same kind of sensation as the one she had yesterday when she
introduced her word ‘S’. So when the diarist attends introspectively to a
particular sensation and says, “I shall use the word ‘S’ to refer to sensations of the same kind as this one”, she will, pace Wittgenstein, thereby
establish a standard of correctness for future applications of the word ‘S’.
Or so, at least, the property-naturalist will argue.
We can trace a similar dialectic in connection with Wittgenstein’s opposition to stage 2 of the introspection-and-identity model. Suppose that,
contrary to the private language sections, I do understand the word ‘pain’,
in the first instance, solely on the basis of my introspective awareness of
my own pain. On the introspection-and-identity model, I then grasp what
it is for someone else to be in pain by appeal to the principle that for someone else to be in pain is for her to be in the same kind of state that I am in
when I am in pain. Wittgenstein famously rejects that idea:
That gets us no further. It is as if I were to say, ‘You surely know what
“It’s 5 o’clock here” means; so you also know what “It’s 5 o’clock on
the sun” means. It means simply that it is just the same time there as it
is here when it is 5 o’clock.’—The explanation by means of sameness
does not work here. For I know well enough that one can call 5 o’clock
here and 5 o’clock there ‘the same time’, but I do not know in what
cases one is to speak of its being the same time here and there.
In exactly the same way, it is no explanation to say: the supposition that he has a pain is simply the supposition that he has the same

4 For the point about a technique, in this context, see PI §262.

82 William Child
as I. For what’s surely clear to me is this part of grammar: that one
will say that the stove has the same experience as I if one says: it’s in
pain and I’m in pain.5
His point is this. The principle “for someone else to be in pain is for her to
be in the same state that I am in when I am in pain” is true. But I cannot
extract a conception of what it is for someone else to be in pain merely
from the principle about sameness of state, by itself. For there are indefinitely many different possible states of S that could count as S’s being in
the same state that I’m in when I feel pain. Appealing to the notion of
sameness of state, then, is not by itself enough to specify what it takes for
S to be in pain; I need some substantive understanding of what it takes
for S to be in the same state as me.
As before, Wittgenstein’s position depends on his anti-platonism. For
suppose there are natural properties. Then, it seems, we can appeal to
sameness of property in explaining what it is for someone else to be in
pain. For in that case the explanation, “for S to be in pain is for her to be
in the same state that I am in when I’m in pain”, does single out a definite
condition for the truth of ‘S is in pain’.

2. Peacocke on Concepts of Conscious States
Christopher Peacocke has offered an account of our concepts of conscious
states that directly challenges Wittgenstein’s view by endorsing a version
of the natural properties view and accepting the model of introspectionand-identity.6 Examining that account is a helpful way to explore the
issues about natural properties and sensation concepts.
According to Peacocke, our grasp of the concept pain is ‘anchored’
in the first-person case: in our awareness of our own feelings of pain.
At a first pass, Peacocke’s proposal is that possessing the concept pain
involves two things:
(i) knowing what it is like to feel pain;
(ii) having tacit knowledge that for someone else to be in pain is for them
to be in the same conscious state that one is in oneself when one is in
pain.
But that is only a first pass. To understand the idea of someone else being
in the same conscious state that I am in when I am in pain, I must grasp
the idea of someone other than myself being a subject of experience at all.

5 PI §350.
6 Christopher Peacocke, Truly Understood (Oxford: Oxford University Press, 2008),
Chapter 5.

Sensations

83

And in Peacocke’s view, just as my grasp of the concept pain is anchored
in my knowledge of what it is like to be in pain, so my grasp of the concept conscious subject is anchored in my knowledge of what it is like to
be a conscious subject. He holds, further, that the concepts conscious
state and conscious subject are interdependent: a conscious state is, in
its nature, a state of a conscious subject; and a conscious subject is, in
its nature, a thing that has conscious states. In line with that interdependence, Peacocke’s full view is that grasping the concept pain involves
satisfying this more complex condition:
(i) knowing what it is like to feel pain;
(ii*) having tacit knowledge that for someone else to be in pain is (a) for
them to be something of the same kind as me (a subject); and also
(b) for them to be in the same state I’m in when I’m in pain.7
In what follows, I shall ignore this complication, and focus just on the
simple account given by clauses (i) and (ii) above. But we should remember that Peacocke’s full account is more complex.
We can note three important features of this account. First, Peacocke’s idea
is that one’s grasp of what it is for someone else to be in pain is explained by
one’s grasp of what it is for one to be in pain oneself, together with knowledge of the identity set out in (ii). As I have said, Wittgenstein does not deny
that, when S is in pain, S is in the same state that I am in when I am in pain.
What he denies is that we can appeal to that identity in explaining what it is
for S to be in pain. But that is precisely what Peacocke thinks we can do: his
account deliberately and explicitly rejects this Wittgensteinian view.
Second, as Peacocke stresses, his account assumes a doctrine of natural properties.8 Mental properties, he thinks, are ‘properties in their own
right’; they do not need to be reduced to anything else in order to count as
genuine properties. By the same token, the notion of identity of conscious
state across different subjects is ‘a notion in good standing’ which does
not require ‘further elaboration, in terms of functional role’ or anything
else. So to grasp the idea of someone else being in the same conscious state
that I am in when I am in pain, I do not need any further specification of
what it takes for someone to be in the same conscious state as that. To see
why Peacocke’s account requires this sort of naturalism about properties,
suppose that the natural properties view was false; so, considered in itself,
any conscious state in someone else would have an equally good claim to

7 For this formulation, see Peacocke, Truly Understood, 175.
8 As Peacocke puts it, his account assumes ‘a certain irreducible realism’ about properties
(Truly Understood, 179; all quotations in the present paragraph of my main text are
taken from that page). The kind of realism about properties that Peacocke has in mind
is some version of the natural properties view; that is what is needed if the appeal to the
notion, ‘same conscious state as this’, is to do the work that he requires.

84 William Child
be counted the same kind of conscious state as my state of pain. Then the
clause, “for S to be in pain is for S to be in the same conscious state that I
am in when I am in pain”, would not specify a determinate condition for
the truth of ‘S is in pain’. On the assumption that there are natural properties, however, it is at least arguable that we can appeal to a basic notion
of sameness of conscious state to specify what it takes for S to be in pain.
Third, Peacocke claims two advantages for his account of concepts of
conscious states. In the first place, it explains the uniformity of meaning
between first-person and third-person ascriptions of pain and other conscious states. Since what it is for someone else to be in pain is explained
in terms of her being in the same conscious state that I am in when I’m in
pain, there is no danger that the property of pain that I ascribe to other
people might be different from the property of pain that I ascribe to
myself.9 By contrast:
other theories, and especially some forms of ‘criterial’ accounts favoured
by some neo-Wittgensteinians, have famously had difficulties in explaining how the same thing is meant in . . . first-person and third-person
psychological ascriptions.10
A second advantage Peacocke claims for his account is that it explains
how it is possible to understand the thought that someone else is in pain
without knowing what would count as evidence that she is in pain.11 He
takes it as obvious that understanding third-person ascriptions of pain is
independent of knowing what counts as evidence that someone else is in
pain—citing with approval Rogers Albritton’s observation that one might
have the concept toothache yet simply have “no notion how people with
toothache [are] likely to behave, these days”.12 Wittgensteinian views,
Peacocke thinks, again have difficulty in accommodating that insight.
Is Peacocke’s account successful? Does it offer a viable alternative to
Wittgenstein’s position?

3. Approaching the Middle Ground
We have seen that Wittgenstein’s rejection of natural properties plays
a crucial role both in his argument against the possibility of defining
9 This explains the point advertised in note 3 above: that, on an account that appeals to
identity in answering our initial question 2 (“How do I understand what it is for someone else to be in pain?”), the answer to question 2 brings with it an answer to question
3 (“Is the property of pain that I ascribe to others on the basis of what they say and do
the same as the property I ascribe to myself without evidence?”).
10 Peacocke, Truly Understood, 165.
11 Peacocke, Truly Understood, 177.
12 Rogers Albritton, “On Wittgenstein’s Use of the Term ‘Criterion’,” in Wittgenstein: The
Philosophical Investigations, ed. G. Pitcher (London: Macmillan, 1968), 248.

Sensations

85

sensation words by introspective ostension and in his argument against
the possibility of appealing to identity to explain what it is for someone
else to be in pain. So, on the face of it, someone who accepts the doctrine of natural properties has a ready way of bypassing Wittgenstein’s
concerns about both aspects of the model of introspection and identity.
Peacocke’s account of our concepts of conscious states takes exactly that
route.
I shall argue, however, that the opposition between Wittgenstein’s antiplatonist account of our concepts of sensations and the view of sensationconcepts that is suggested by the natural properties view is not as
straightforward as it may initially seem. On the one hand, someone who
takes the natural-properties view cannot avoid making some appeal to our
classificatory practices in explaining our concepts of conscious states. On
the other hand, the content of Wittgenstein’s anti-platonism needs careful
unpacking and, if it is developed in a way that preserves common-sense
intuitions, there is less distance than there may initially seem to be between
the Wittgensteinian account of sensation-concepts and the account we get
from a plausible application of the natural properties view.
My suggestion is this. If we start with the natural properties view, we
will find ourselves moving towards Wittgenstein’s practice-based account
of our concepts of conscious states. If we start with Wittgenstein’s antinaturalist account, we will find ourselves moving towards some form of
natural properties view. I do not say that we will reach the same position
whatever our starting point. But I do say that we should explore the
middle ground that lies between an extreme naturalism and an extreme
anti-platonism about sensation properties. The correct view surely lies
somewhere in this middle ground.
3.1 From Natural Properties to Classificatory Practices
Suppose we accept the metaphysics of natural properties. Any set of things
shares, or defines, a property; but some properties are objectively more
natural than others. What impact does that view about the metaphysical
status of properties have on the question, which properties our words
or concepts pick out? Lewis’s suggestion is this. It is possible for a word
or concept to pick out a highly unnatural property; ‘quus’ and ‘grue’
are examples of that. But relatively natural properties are more eligible
referents for our words than less natural properties; they are the default
referents. Unless we do something to upset that default (for example,
by explicitly defining a word so as to pick out a highly unnatural property), our words and concepts will pick out relatively natural properties
rather than relatively unnatural properties. On this view, the naturalness
of properties plays a role in determining the reference of our words. But
naturalness does not determine reference all by itself; our classificatory
practice plays some role, too.

86 William Child
One reason why practice plays an essential role in determining the
reference of our word ‘pain’, or our concept pain, even if we accept the
metaphysics of natural properties, is that someone who feels pain on
an occasion has an experience that instantiates a number of different
properties. Her sensation is a pain. But it is something more specific: an
intense, gnawing toothache. And it is something more general: a bodily
sensation. Consider an account of sensation-concepts like Peacocke’s,
on which my grasp of the concept pain is anchored in my singling out
sensation-properties in the first-person case. What determines which particular property I do single out; what determines that my word ‘pain’, say,
applies to pains in general, rather than being restricted to toothache? Peacocke’s answer is, crudely, that I am disposed to apply my word ‘pain’ to
any sort of pain I feel, and not just to toothache; my use of the word is as
he puts it, “keyed to instances in [me] of the property of being a pain”,13
rather than instances of some more restricted property. Even given the
existence of natural properties, then, the meaning of a sensation term is
not established by ostension alone; classificatory practice plays a crucial
role in determining which property I am picking out.14
Another reason for thinking that classificatory practice must play a
role in determining the referents of our sensation-terms starts with an
example from Wittgenstein:
A tribe has two concepts, akin to our ‘pain’. One is applied where
there is visible damage and is linked with tending, pity etc. The other
is used for stomach-ache, for example, and is tied up with mockery
of anyone who complains. ‘But then do they really not notice the
similarity?’—Do we have a single concept everywhere where there
is a similarity? The question is: Is the similarity important to them?
And need it be so?15
It seems clear that the people in Wittgenstein’s example have legitimate
concepts. Call the concept they apply to pain with visible damage pain1
and the concept they apply to pain without visible damage pain2. And
suppose that, in the objective hierarchy of naturalness, the property of
pain is the most natural property that is instantiated in the kinds of circumstance in which these people apply their concept pain1 to someone.

13 Peacocke, Truly Understood, 189.
14 For a similar stress on the role of recognitional abilities in determining which property I
pick out with a sensation-concept that I introduce by introspective ostensive definition,
see Brian Loar, “Phenomenal States (Second Version),” in The Nature of Consciousness,
ed. Ned Block et al. (Cambridge, MA: MIT Press, 1997).
15 Ludwig Wittgenstein, Remarks on the Philosophy of Psychology, Volume II, ed. G. H.
von Wright and H. Nyman, trans. C. Luckhardt and M. Aue (Oxford: Blackwell, 1980),
§638.

Sensations

87

Should we say that their concept pain1 therefore refers to pain, rather
than the more restricted property, pain1? If we interpret their concept that
way, we must say that, when they refuse to apply their concept pain1 in
cases of pain without visible damage, they are simply wrong about the
extension of the concept. But that seems highly implausible. The situation seems rather to be this. Our concept pain picks out the property
of pain. The other people’s concept pain1 picks out a different property:
the property of pain-with-visible-damage. And the difference between the
reference of our concept and the reference of theirs is not a matter of any
difference in the properties that are instantiated in people who belong
to the two different communities. It is, rather, a matter of the difference
between our practice of using our concept pain and their practice of using
their concept pain1. As before, even given the existence of natural properties, which property a word or concept picks out is in part a matter of
our classificatory practice.16 Once we see that point in the current case, it
seems clear that our practice of applying our words in the way we do will
have a crucial role in determining the reference of our concepts in very
many other cases, too.
For these reasons, then, even if we accept the metaphysics of natural
properties, and even if we accept that properties that are relatively natural are more eligible referents for our words than properties that are less
natural, we must also accept that our classificatory practices play a vital
role in determining which properties our words pick out. If we start from
the natural-properties view, we are driven towards the Wittgensteinian
view that what it takes for one thing to count as similar to another, or as
having the same property as another, is not dictated by reality itself but
depends in part on our classificatory practices.
3.2 From Classificatory Practices to Natural Properties
On Wittgenstein’s view, I said, what it takes for one thing to count as
similar to another is not determined by reality itself but depends in part
on our classificatory practices: on what we find it natural to count as
similar to what. But how exactly does similarity, or sameness of property, depend on our practices? We can distinguish two different ways of
developing the basic idea. One development leads to an extreme antirealism about properties and standards of similarity. The extreme view is
certainly distinct from any version of the natural-properties view. But it
is deeply counterintuitive and it is hard to think that this is the view that
Wittgenstein intends. A second way of developing the basic idea leads
16 Note the obvious similarity between this argument and Burge’s argument about the way
in which the reference of a term like ‘arthritis’ depends on the social practice of applying it in a particular way (see Tyler Burge, “Individualism and the Mental,” in Midwest
Studies in Philosophy IV (1979)).

88 William Child
to a modest realism about properties. That is a plausible view in its own
right, and is plausibly ascribed to Wittgenstein. But it appeals to a notion
of sameness of property that cannot be spelled out in terms of our actual
human capacities to classify things as being similar to, or different from,
one another. So, I shall argue, even if we start from a Wittgensteinian,
practice-based conception of properties, we cannot avoid accepting some
element of the natural properties view.
Suppose we introduce some word by reference to a set of examples:
“The word ‘F’ applies to this, this, and this, and to anything else that is
relevantly similar to these things”. The general anti-platonist intuition
is that what it takes for something to be relevantly similar to the examples is a matter of our classificatory practice. But how exactly should we
understand the role of our classificatory practice in defining a standard
of similarity?
The extreme view is that what it takes for a thing to be similar to the
examples is determined by what we actually judge to be similar when we
consider the matter and classify the thing as being F or not-F (or what
we would judge to be similar if we were to consider it and classify it as F
or not-F). Whether or not the application of the word ‘F’ in a particular
case is correct, then, is directly determined by what we actually say (or
would say) when we consider that particular case and reach a verdict. But
that is a radically revisionist view. We ordinarily think that whether some
word ‘F’ applies to an object depends on two things: the meaning that
we have given to the word; and the way the object is. It does not depend
on what we would in fact say if we were to consider the question and
reach a verdict about whether or not the object is F. On the contrary, we
think, when we do consider an object and judge that it is F, our judgment
is true or false in virtue of whether or not the object meets the standard
of F-ness that we have already laid down. But the current view explicitly
rejects that common-sense thought. Similarly, we ordinarily think that
the standards of similarity and correctness associated with our words
and concepts extend to new cases in a way that is not restricted by our
actual capacity to apply them. Our actual human capacities are limited.
There are calculations that are too large or complex for us to be able to
complete; objects that are too small or inaccessible for us to be able to
observe and classify; environments that are too hostile for us to be able
to investigate. But, we think, the standards of similarity and correctness
that we establish in connection with cases that we are able to calculate
or classify apply determinately to these other cases. There is, we think, a
fact of the matter about the correct result of applying our mathematical
operations to a new case, even if it exceeds our calculative capacities;
there is a fact of the matter about whether some object that is too small
or inaccessible for us to classify is in fact similar to other things by our
standards of similarity; and so on. But, as before, the extreme view rejects
that ordinary thought. If what counts as similar to what is determined by

Sensations

89

what we actually classify (or would classify) as similar, our standards of
similarity simply do not extend to cases that transcend our classificatory
capacities. That is a possible view. And it is a view that some interpreters
have ascribed to Wittgenstein.17 But it involves the wholesale rejection
of basic elements of our ordinary thought. If that is the view that Wittgenstein intends, then his image of himself as a philosopher who leaves
everything as it is seems radically mistaken.
That is one way of understanding the basic anti-platonist intuition that
what it takes for one thing to be similar to another is a matter of our classificatory practices. The alternative way of understanding the basic intuition shares the idea that there are numerous different possible standards
of similarity that are consistent with an initial set of examples. It shares
the idea that no one of these possible standards of similarity is objectively
the correct one, or objectively ‘straighter’ or more natural than any other.
And it shares the idea that what ties a word or concept to a particular
standard of similarity is that that is the standard embodied in the way we
find it natural to classify things. But, on the modest view, once we have
done enough to single out a particular property by classifying things as
similar to, or different from, one another in the way we do, there is simply
a fact of the matter about whether or not something has that property:
a fact of the matter that is not determined by how we actually classify it
when we come to consider the matter (or how we would classify it if we
were to consider the matter). Our classificatory practices play a crucial
role in establishing standards of correctness for our words and concepts.
But the standards we establish are not limited by our own actual ability
to apply them; there are facts of the matter about how our concepts apply
to things in cases to which we ourselves are unable to apply them.
It seems fair to say that this modest realism has something in common with the natural properties view, insofar as it involves a view of the
naturalness or sameness of properties in which what it takes for one
thing to have the same property as another is not limited by our actual
human capacities to tell whether or not they have the same property. The
modest realist insists, for instance, that there is a fact of the matter about
whether some object that is too small or inaccessible for us to examine
has the same property that we attribute to objects that we can examine

17 See, e.g. Michael Dummett, “Wittgenstein on Necessity: Some Reflections,” in his
The Seas of Language, (Oxford: Oxford University Press, 1994). Dummett ascribes
to Wittgenstein the view that “it is only our doing [a] calculation and ‘putting it in the
archives’ that constitutes its result as being that obtained by doing it correctly; so, if
we never do that calculation, there is no one correct result” (Dummett, “Wittgenstein
on Necessity,” 459). The view that Dummett attributes to Wittgenstein is even more
radical than the position discussed in the text, since it rules out any appeal to subjunctives about the results we would get if we were to do a calculation that no-one ever
actually performs.

90 William Child
and categorize. But what does it take for this small or inaccessible object
to have the same property as the property that we identify in objects that
we can examine? It cannot be that we actually classify it (or would classify it) as having the same property; for, in the nature of the case, our classificatory capacities do not extend to the object in question. The modest
realist’s view must be that, once we have singled out a property through
our classificatory practice in cases we can consider, there is simply a fact
of the matter about what it is for a tiny or inaccessible object to have the
same property as that. And that use of the notion of sameness of property, as I have said, has something significant in common with the natural
properties view. The picture is one in which, in picking out the properties
we do, we latch onto properties that are there anyway: properties whose
status as genuine properties does not depend on us, our concepts, or our
practices of classification. That is what allows us to single out properties
in a way that extends beyond our own ability to identify instances.

4. Defending the Middle Ground
I have argued that an adequate account of our concepts of conscious
states must occupy a middle ground, which gives a role in determining
the content of those concepts both to our actual classificatory practices
and to a notion of naturalness or sameness of properties that cannot
be understood simply in terms of what we do in fact find it natural to
classify as similar to what. And I have suggested that Wittgenstein’s
view has just that form. But does this middle ground really exist? We
can approach that question by considering Peacocke’s critique of a ‘neoWittgensteinian’ view.
Recall that, on Peacocke’s view of the concept pain, what comes first
in the order of explanation is the property of pain. The account appeals
to that property at two stages: first, in explaining our grasp of the concept pain in the first-person case; and then in explaining how we understand ascriptions of pain to other people. Now, Peacocke anticipates a
neo-Wittgensteinian objecting to his account on the grounds that “there
simply is no property of being in pain that has all the characteristics
needed by [Peacocke’s] account”.18 “Under this response”, he says, “it
is legitimate to speak of the property of being a pain, but this property of a mental event must be regarded as a construct out of human
conceptual reactions and expressive capacities, in a way that is incompatible with its being causally and rationally explanatory of thinkers’
first-person judgments that they are in pain”.19 Peacocke argues that
18 Peacocke reports that this objection was suggested to him by Crispin Wright. The objection is closely related to the points I made in section 3.1 above.
19 Peacocke, Truly Understood, 190. Subsequent quotations in this paragraph of the text
are taken from pp. 190–1 of Peacocke’s book.

Sensations

91

this neo-Wittgensteinian view is inadequate. He faces it with a dilemma.
Suppose, first, that we treat the property of pain “as a construct out of
human conceptual reactions and expressive dispositions”. In that case,
our account will fail to make sense of the range of applications of the
concept pain that we actually take to be intelligible. We can, for instance,
understand the thought that octopuses, dolphins, and whales feel pain.
But such creatures have no “distinctively human reactions and expressive dispositions”. So if we understand the property of being in pain as a
construct out of such reactions and dispositions, we cannot understand
the thought that these creatures feel pain. In response to that horn of the
dilemma, the neo-Wittgensteinian may offer a different view, saying that
“a role in human conceptual and expressive life” plays an essential role in
singling out the property of pain—in fixing the reference of the concept
pain—but that “the very same conscious property could be instantiated
by creatures for which it does not have that role”. But, Peacocke objects,
if the neo-Wittgensteinian takes that view, the position collapses into a
version of Peacocke’s own account. For “if there is such a real conscious
property that has a nature independently of human conceptual reactions
and expressions, then that property can play a role in making first-person judgements rational”, and it can play a role in explaining how we
can understand third-person ascriptions of pain without knowing what
would count as evidence that they were true. And that is just the view
that Peacocke is promoting.
Care is needed in assessing Peacocke’s objection to the ‘neo-Wittgensteinian’
view, for two reasons. On the one hand, as we saw in 3.1, Peacocke’s own
account itself gives some role to our classificatory practices in specifying
exactly which property it is that a person picks out in the first-person
case, because it gives a role to her disposition to apply the word ‘pain’
to all pains she feels, and not simply to a wider, or a more restricted,
set of feelings. So Peacocke’s own view is not the most uncompromising
application of the natural-properties view. On the other hand, as we saw
in 3.2, the idea that the property of pain is an artefact or ‘construct’ of
our classificatory practices can be developed in different ways: developed
one way, it involves a radical anti-realism; developed in another way, it
leads to a modest form of realism. I will proceed by considering the effectiveness of Peacocke’s complaint considered as an argument against the
kind of modest realism about properties that I described in section 3.2. It
may be that that is not exactly the kind of neo-Wittgensteinian view that
Peacocke has in mind. Since, however, he suggests that every form of neoWittgensteinianism about properties will be vulnerable to versions of the
objections he raises, our discussion will certainly be relevant to assessing
his argument.
The first horn of Peacocke’s dilemma is that, if the neo-Wittgensteinian
presses the idea that pain is a “construct out of human conceptual reactions and expressive dispositions”, she will be unable to make sense of

92 William Child
ascriptions of pain to certain non-human animals, ascriptions that obviously do make sense. But that criticism seems ineffective against the position I have outlined. Nothing in Wittgenstein’s position requires that a
state of some other mammal only qualifies as a pain if the creature itself
has human-style conceptual reactions to pain or expresses pain in exactly
the ways that human beings express it. What matters for the intelligibility of ascribing pain and other experiences to creatures of some kind is
whether the states in question play a role in the creatures’ lives that is
sufficiently similar to the role that pain plays in our lives. And a state of a
whale or a dolphin can meet that standard without the creature needing
to have concepts or specifically human interests, and without the creature
behaving in exactly the ways that humans do. So Wittgenstein’s view does
make room for understanding ascriptions of pain and other sensations
to non-human animals whose lives and behavior are very different from
ours. It is plausible that, when we move beyond the simple case of pain,
we will quickly reach cases where, on Wittgenstein’s approach, the concepts we apply to human beings will not be straightforwardly applicable
to creatures that lack the dispositions and conceptual capacities that we
humans possess. It is hard to see how to make sense of ascriptions of
hope or grief, for instance, to creatures whose interests, capacities, and
reactions are radically different from the kinds of interests, capacities,
and reactions that we have.20 But that is as it should be: it really is unclear
that we can make sense of a whale, a cat, or a sheep experiencing hope,
envy, hatred, and so on.
So much for the first horn of the dilemma that Peacocke poses for a
neo-Wittgensteinian view. What of the second horn: the argument that, if
we regard human reactions to, and expressions of, pain simply as singling
out a property of pain that can play a role in causally explaining selfascriptions of pain, and in making sense of ascriptions of pain in cases
where we have no idea what would count as evidence for the presence of
pain, then our position collapses into Peacocke’s version of the natural
properties view?
In the first place, there is no reason why a Wittgenstein view of pain
cannot accept that the judgment ‘I am in pain’ is a causal response to
pain. It is an obvious, common-sense truth that we ascribe pain to infants
long before they exhibit any linguistic or conceptual reactions to pain.
Equally obviously, an infant’s being in pain stands in causal relations
to other events and states of affairs. It is caused by physical impacts
and injuries. It causally affects the child’s future behavior; the burned
child avoids the flame because getting burned was painful. And it causally affects other people’s behavior; we treat the child with sympathy
and concern because she is in pain. I see nothing in Wittgenstein that

20 For brief comments on the cases of hope and grief, see PI, part II §§1–2.

Sensations

93

conflicts with these common-sense truths. Now consider what happens
when the child learns to apply the word ‘pain’ to herself. We teach the
child to judge ‘I’m in pain’ in circumstances where we already know, on
independent grounds, that she is in pain; in doing so, we are teaching her
to respond to her pain by judging that she is in pain. Once the teaching
has succeeded, the response of judging ‘I’m in pain’ is, as Peacocke puts
it, ‘keyed to’ the occurrence of pain in the child. There is no reason for
a Wittgensteinian to deny that, when the child judges ‘I’m in pain’, she
makes that judgement because she is in pain. And there is no reason for
her to deny that that is a causal because; the child’s judging ‘I’m in pain’
is causally explained by her being in pain.
But recognizing that our self-ascriptions of sensations are causally
responsive to the sensations we self-ascribe does not undermine the Wittgensteinian idea that our classificatory practices play an essential role in
determining which properties we pick out. Recall Wittgenstein’s example of the people who have two concepts—pain1 and pain2—where we
have the single concept pain. The reference of these people’s terms ‘pain1’
and ‘pain2’ is not fixed merely by the nature of the properties instantiated in this community. What determines the reference of the terms is in
part their actual classificatory practice. But that is completely compatible with the point that self-ascriptions are causally responsive to occurrences of the properties that are thereby self-ascribed. In Wittgenstein’s
imagined community, self-ascriptions of the form ‘I’m in pain1’ are by
and large causally explained by the presence of instances of pain1; selfascriptions of the form ‘I’m in pain2’ are, by and large, causally explained
by the presence of pain2.
In the second place, I have already argued that the Wittgensteinian
view, as I have characterized it, can accommodate the fact that we can
understand third-person ascriptions of pain in cases where we have no
idea what would count as evidence that someone else is in pain. Our
classificatory practice has an essential role to play in determining which
property we are picking out; but once we have singled out that property,
we can go on to make sense of ascriptions of the same property in cases
that transcend our actual human capacity to tell whether or not something has the property. So I think Peacocke is wrong to suggest that if we
take that view of properties then we have given up a neo-Wittgensteinian
position.
It might be suggested that my disagreement with Peacocke about the
prospects for a Wittgensteinian view is a purely verbal one. On the one
hand, it might be claimed, my allegedly Wittgensteinian form of modest
realism is not a distinctively Wittgensteinian position at all but is, rather,
a version of the kind of realism about properties that Peacocke himself
recommends. On the other hand, it may be suggested, what Peacocke
means by a ‘neo-Wittgensteinian’ position is what I have described as
extreme anti-realism about properties: the kind of view on which what it

94 William Child
is for x to belong to the same kind as y can be understood only in terms
of how we find it natural to classify x when we actually consider the case
(or how we would find it natural to classify x if we were to consider the
case). So, the suggestion goes, I am not in the end opposing Peacocke’s
view that there are really only two kinds of position available: an unacceptable anti-realism or constructivism about properties; and a sensible
form of realism. The difference is only over where to locate Wittgensteinian views in this dichotomy: I think Wittgenstein’s views lead us to the
modest realist position; Peacocke thinks they lead us to the unacceptably
anti-realist position. Settling that disagreement is of course important
for the interpretation and assessment of Wittgenstein’s views. But, on
this line of thought, there is no substantive philosophical disagreement
between me and Peacocke.
This irenic conclusion seems premature, however. I agree that Peacocke and I are maneuvering in the same general area. We both reject
the extreme anti-realist or constructivist view of properties. And we both
reject the most radical alternative, on which the naturalness of properties
determines which properties our words and concepts pick out all by itself,
with no contribution at all from our classificatory practice. But, as I suggested in 3.1, Peacocke’s position seems to me to give insufficient recognition to the role of our actual classificatory practices in determining which
properties our concepts of pain and other conscious states pick out. Different linguistic communities may employ concepts that pick out different
properties, even if there are no differences in the properties that are actually instantiated in those communities’ environments. We cannot make
sense of these differences in properties picked out unless we recognize that
our classificatory practice plays a larger role than Peacocke seems to allow
in defining or specifying what counts as sameness of property.

5. Conclusion
I suggested at the start of 3.1 that, if our account of concepts of conscious
states starts from a commitment to natural properties, we are bound to
recognize on reflection that there is a role for our actual classificatory
practices in determining which properties our concepts pick out. On the
other hand, if we start from the anti-platonist idea that the notions of
similarity or sameness of property must be understood in terms of our
classificatory practice, we are bound to recognize that we need a notion
of sameness of property that extends beyond our limited capacity to
apply our own concepts and categories.
I have argued for three conclusions. First, that the suggestion just outlined is correct; a account of our concepts of conscious states must, as
I have put it, occupy the middle ground that gives a role both to our
classificatory practices and to a notion of sameness or naturalness that
extends beyond our limited capacities to recognize similarity or sameness

Sensations

95

of property. Second, that Wittgenstein’s own view is a position of that
general kind; in the terms of 3.2, he advances a version of modest realism. Third, that Peacocke’s objection to neo-Wittgensteinian views is
unsuccessful. No doubt there is room for refinement and expansion of
these conclusions. But the correctness of these general conclusions—both
about the location of the right position in this debate and about the form
of Wittgenstein’s view—seems clear.21

References
Albritton, Rogers. “On Wittgenstein’s Use of the Term ‘Criterion’.” In Wittgenstein: The Philosophical Investigations, edited by G. Pitcher, 231–250. London:
Macmillan, 1968.
Burge, Tyler. “Individualism and the Mental.” In Midwest Studies in Philosophy,
volume IV. Edited by P. A. French, T. E. Uehling and H. K. Wettstein, 73–122.
Minneapolis: University of Minnesota Press, 1979.
Dummett, Michael. “Wittgenstein on Necessity: Some Reflections.” In The Seas
of Language, 446–461. Oxford: Oxford University Press, 1994.
Lewis, David. “New Work for a Theory of Universals.” Australasian Journal of
Philosophy 61 (1983): 343–377.
———. “Putnam’s Paradox.” Australasian Journal of Philosophy 62 (1984):
221–236.
Loar, Brian. “Phenomenal States (Second Version).” In The Nature of Consciousness, edited by Ned Block et al., 597–616. Cambridge, MA: MIT Press, 1997.
Peacocke, Christopher. Truly Understood. Oxford: Oxford University Press, 2008.
Sider, Theodore. Writing the Book of the World. Oxford: Oxford University
Press, 2011.
Wittgenstein, Ludwig. Remarks on the Philosophy of Psychology, volume II.
Edited by G. H. von Wright and H. Nyman, translated by C. Luckhardt and
M. Aue. Oxford: Blackwell, 1980.
———. Philosophical Investigations, 4th edition. Edited by P. M. S. Hacker
and Joachim Schulte, translated by G. E. M. Anscombe, P. M. S. Hacker and
Joachim Schulte. Oxford: Blackwell, 2009.

21 An earlier version of this paper was presented at the conference, ‘Wittgenstein, Philosophy of Mind, and Naturalism’, at the University of Bergen in June 2015. I am grateful
to the participants in that conference, and especially to the organizers, Kevin Cahill and
Thomas Raleigh, for very helpful comments and discussion.

5

Wittgenstein, Self-Knowledge,
and Nature
Annalisa Coliva

Introduction
The aim of this chapter is to explore Wittgenstein’s suggestive remarks
about self-knowledge—that is, our knowledge of our own mental states—
and their connection with naturalism. As is well known, Wittgenstein
was critical of naturalizing the epistemology of mind in behaviorist or
in more physicalist-oriented terms. Yet, he placed great emphasis on the
role of instinctive and natural manifestations of our own mental states.
These natural manifestations were necessary, in his opinion, to acquire
the relevant psychological concepts and thereby get in a position to give
new expression to our mental states through language. Furthermore, he
thought that these verbal manifestations would become “second nature”
to us. So much so, that just as it would not make sense to regard a spontaneous cry of fear as based on having that emotion in view and then
giving expression to it through the cry, he thought that at least some of
our psychological self-ascriptions—those he termed “avowals”—should
not be considered as judgments based on and justified by having the relevant mental states in view. Rather, he thought they should be seen as
immediate and spontaneous, though culturally ingrained, expressions of
the mental states that elicited them. A fortiori, in his view, psychological
avowals should not be considered as the result of inference to the best
explanation starting from the observation of one’s overt behavior.
Wittgenstein put his own variety of naturalism at the service of dissolving—
rather than solving—the very problem of self-knowledge. To put it crudely,
in Wittgenstein’s perspective there is no real epistemology of the mind,
when it comes to (at least some of) our psychological self-ascriptions. For,
despite their surface grammar, psychological avowals are not linguistic
manifestations of true and justified, or reliably formed, beliefs about our
own mental states, reached through the observation of one’s own mental
states or by means of inference to the best explanation starting with the
observation of one’s overt behavior. Rather, their function is merely expressive of one’s ongoing mental states, and their occurrence is just as instinctive, and not observationally or inferentially mediated, as a cry or a sigh.

Self-Knowledge

97

For these reasons, Wittgenstein is rightly considered the father of both
contemporary expressivism and of constitutivism regarding psychological avowals. In particular, his remarks can be seen as the origin of the
idea—central to contemporary expressivism—that the main function of
psychological avowals is expressive rather than descriptive. Contemporary expressivists, however, have also tried to move away from Wittgenstein’s anti-epistemological outlook.1 Contemporary constitutivists, by
contrast, are united in rejecting the idea that, at least in some central
cases, our psychological self-ascriptions are underwritten by an appropriate epistemic relation linking the subject and her first-order mental
states, such that the former can actually be taken to manifest true and
appropriately justified beliefs about one’s own mental states. Constitutivists, however, are critical of the idea that the relevant self-ascriptions
serve merely an expressive function and that they replace forms of more
instinctive behaviour.2
In my own work on self-knowledge, I have tried to combine these two
perspectives, by endorsing the anti-epistemological outlook characteristic
of constitutivism whilst also retaining a role for the idea that psychological concepts are primarily acquired as ways of expressing, rather than
naming, an on-going mental state. Still, in my view, these concepts can
and indeed are used to make judgments, whose characteristic traits—
such as transparency, authority, and groundlessness3—are not guaranteed by any peculiar epistemic relation holding between a subject and her
own mental states. Rather, they are features that hold a priori, once these
judgments are seen as falling within the scope of appropriately qualified
C-conditions.4
1 See in particular Dorit Bar-On, Speaking My Mind: Expression and Self-Knowledge
(Oxford: Oxford University Press, 2004). I have discussed Bar-On’s position at length in
Annalisa Coliva, The Varieties of Self-Knowledge (London: Palgrave, 2016).
2 See in particular: Akeel Bilgrami, Self-Knowledge and Resentment (Cambridge, MA:
Harvard University Press, 2006); Sydney Shoemaker, The First Person Perspective and
Other Essays (Cambridge: Cambridge University Press, 1996); Crispin Wright, “Wittgenstein’s Rule-Following Considerations and The Central Project of Theoretical Linguistics,” in his Rails to Infinity, edited by Crispin Wright (Cambridge, MA: Harvard
University Press, 2001), 170–213. I have discussed their positions at length in Annalisa
Coliva, The Varieties of Self-Knowledge (London: Palgrave, 2016).
3 I have presented, explained, and defended these traits in chapter 3 of Coliva, Varieties.
Transparency amounts to the idea that, when one has the relevant mental states and
concepts, one is ipso facto in a position to self-ascribe that mental state. Groundlessness
amounts to the view that such a self-ascription is not underwritten either by observation of one’s own mental states, or by inference to the best explanation starting with
the observation of one’s own behavior. More strongly, it can also entail the idea that the
mental state does not serve as a justification for its self-ascription. Finally, authority
amounts to the idea that, when the relevant conditions obtain, a subject’s psychological
self-ascription cannot sensibly be challenged.
4 See ibid., chapters 7, 8.

98 Annalisa Coliva
These developments will not be the subject of the present chapter. What
is important to note, however, is that Wittgenstein was well aware of the
limitations of his own expressivist position. He was very careful—indeed
much more careful than several contemporary expressivist theorists—not
to over-generalize the expressivist treatment. For he was very mindful of
the fact that exactly the same turns of phrase could sometimes be used
to express judgments about our own first-order mental states and that,
in those cases, they would manifest a subject’s own beliefs regarding her
first-order mental states, reached through a variety of epistemic methods, all open to error, at least in principle. Hence, just as much as he can
rightly be seen as the ancestor of contemporary expressivism and constitutivism with respect to self-knowledge, he should actually be seen as
the father of contemporary pluralism regarding self-knowledge as well.5

I. Against Introspectionism and Behaviorism:
The Case of Sensations
In this section, I review Wittgenstein’s criticism of introspectionism and
behaviorism in order better to situate his positive proposal with respect
to self-knowledge. These remarks will also help us see what kinds of naturalism Wittgenstein’s positive proposal eventually stands opposed to.
Let us therefore start with his attack on introspectionsim. As understood here, introspectionism is the view according to which mental states
are luminously presented to us—that is, their nature and, when applicable, their content—would be something they would wear on their sleeves
and that we would know by perceiving them (or, more mildly, by attending to them), while paying attention to their distinctive elements. Just to
fix the general idea, consider pain. It seems extremely natural to think
that that sensation has its own distinctive phenomenology (or quale) and
that in order to self-ascribe it, we need only to pay attention to its occurrence, recognize it for what it is and self-ascribe it, through an exercise of
the relevant concepts. These, in turn, would have to be acquired, at least
in the case of the concept of pain, by creating a mental word, demonstrative concept, or file—depending on your favorite account of atomic
concepts—in response to the occurrence of that very sensation with its
distinctive phenomenology.
It is quite clear that although Wittgenstein criticized the idea of a private language in the Philosophical Investigations, his remarks could be
applied to the case of the concept of pain, understood along the lines just
sketched. The key analogy is given by the fact that unless one confined

5 This is the position I myself have developed in ibid, and in Annalisa Coliva, “How To
Be a Pluralist About Self-Knowledge,” in Epistemic Pluralism, edited by Annalisa Coliva
and Nikolaj Pedersen (London: Palgrave, 2017), 253–284.

Self-Knowledge

99

that concept to a single occurrence of a demonstrative—viz. that—used
in connection with a specific occurrence of a given mental state—Mi—,6
that concept would have to have either a predicative or a complex
demonstrative structure—it should be either pain or that pain—and
should be applicable in potentially many occasions. The first encounter
with its referent on one occasion should therefore be able to set up a rule
for the future employment of that concept. The very point Wittgenstein
makes against the idea of a private language could then be raised against
the notion of a private concept of pain. For a procedure of private ostensive definition of a word like ‘pain’ or of a concept like pain or that
pain cannot set up a standard against which future applications could be
deemed correct or incorrect: “Whatever is going to seem right to me is
right. And that only means that here we can’t talk about ‘right’ ”.7 That
is to say, if ‘pain’/pain/that pain is the name I give to a sensation of
mine, or else it is a demonstrative or predicative concept whose content
is given by that sensation, its use over time counts as correct whenever
I would think so, even when I may use it to refer or think to a different
kind of sensation or to nothing at all. This, however, simply means that
the distinction between correct and incorrect uses of that term/concept
has vanished. If that distinction can no longer be drawn, though, there is
no rule-governed use of the term. Thus, the word has no meaning, given
Wittgenstein’s normative conception of language. Mutatis mutandis, the
private ostensive definition would only seem to give rise to a rule for the
future employment of the concept. Hence, the putative concept pain (or
the concept that pain) would just be an impression of concept.8

6 Arguably, that is what is proposed in chapter 5 of Bertrand Russell, The Problems of
Philosophy (Oxford: Oxford University Press, 1912). See also David Chalmers, “The
Content and Epistemology of Phenomenal Belief,” in Consciousness: New Philosophical
Perspectives, edited by Quentin Smith and Aleksandar Jokic (Oxford: Oxford University
Press, 2003), 220–272.
7 Ludwig Wittgenstein, Philosophical Investigations (Oxford: Basil Blackwell, 1953), § 258.
8 Two brief remarks. First, if, like Chalmers, op cit., one contented oneself with short-lived
demonstrative concepts, the ostensive procedure could go through, but it would actually
require much more than what proponents of it are usually prepared to recognize. For
it would require attending to one’s sensation as such, as opposed to merely having it.
Furthermore, it would require the ability to demonstratively refer to it in thought and to
be disposed to use that demonstrative concept in a potentially infinite number of occasions. Given the short-lived nature of that concept, further occurrences in thought would
not be genuinely demonstrative ones. Rather, they would be like definite descriptions in
disguise. Alternatively, they would have to rely on memory. In this case, Wittgenstein’s
criticism that anything it would seem right to a subject would be right would apply.
These aspects make the demonstrative strategy much more complex and at the same time
less intuitive and appealing than it first seems. Furthermore, it then becomes quite clear
that it would depend on having quite complex abilities, even conceptual ones, which
could hardly be developed in a purely private setting. For these reasons, far from what
Chalmers claims in his paper, it is dubious that his strategy can really avoid the kind of

100 Annalisa Coliva
Furthermore, according to Wittgenstein, the idea that we are acquainted
with our sensations, which are the referents of our psychological terms
(or concepts) such as “pain”, would lead to solipsism. That is to say, to
the idea that one can know only one’s own mind and merely surmise what
goes on in other people’s minds, if indeed there are other people at all. For
one cannot know what another person is referring to when she uses those
very terms, not even by analogy with one’s own case, since their referents
are foreclosed to one. Indeed, one cannot even know whether another
person is referring to anything at all, by using those terms. Hence, one
cannot know if it is a person—that is, someone capable of having sensations (and other mental states)—rather than a creature who resembles us
in its looks and behavior but who is not really enjoying any mental state.
However, if Wittgenstein is clearly against introspectionism for the reasons just rehearsed, he is equally critical of behaviorism, despite the fact
that his expressivism might be taken—and indeed, it has been taken—to
be a covert form of it. Crude behaviorism has it that there are no mental
states, just outward behavior. Although, as we shall presently see, for Wittgenstein the language of sensations essentially depends on, and sometimes
replaces their natural and instinctive manifestations, he is careful to stress
that this is not tantamount to denying the existence of sensations and
further mental states. Moreover, he was surely against the idea that we
would know of our sensations through inference to the best explanation
starting from the observation of our own overt behaviour. What he did
object to is simply the way in which, following introspectionism, we are
led to think of the language of sensations after the name-object referential
model, which immediately leads to the endorsement—albeit implicitly—
of the idea of a private language. As he wrote in a series of telling passages:
Suppose everyone had a box with something in it: we call it a ‘beetle’.
No one can look into anyone else’s box, and everyone says he knows
criticisms Wittgenstein developed against the idea of a private language and of private
ostensive definitions.
Second, if one held that concepts are words of mentalese (along the lines of Jerry
Fodor, Concepts: Where Cognitive Science Went Wrong [Oxford: Oxford University
Press, 1998]) brought about by the causal interaction with the relevant properties, we
would have no clear explanation of the conceivability of “inverted-spectrum” scenarios.
For, if the concept elicited by the causal interaction with green objects looking red is red,
after all, inverted-spectrum subjects would have a different concept than normal ones,
who would conceptualize the very same items as green. Yet, the very point of invertedspectrum scenarios is that despite different perceptual appearances, subjects would still
be able to categorize objects in exactly the same way. If, in contrast, the concept elicited
by the causal interaction with green objects looking red is green, there is no causally
respectable story, which could explain the asymmetry. To account for it, one should be
prepared to say that inverted-spectrum subjects have an abnormal conceptual faculty,
rather than simply an abnormal perceptual mechanism, such that different phenomenal
properties would give rise to the same concepts.

Self-Knowledge

101

what a beetle is only by looking at his beetle.—Here it would be
quite possible for everyone to have something different in his box.
One might even imagine such a thing constantly changing.—But suppose the word ‘beetle’ had a use in these people’s language?—If so it
would not be used as a name of a thing. The thing in the box has no
place in the language game at all; not even as a something: for the
box might even be empty.—No, one can ‘divide through’ by the thing
in the box; it cancels out, whatever it is. That is to say: if we construe
the grammar of the expression of sensation on the model of ‘object
and designation’ the object drops out of consideration as irrelevant.9
‘Are you not really a behaviorist in disguise? Aren’t you at bottom
really saying that everything except human behavior is a fiction?’—If
I do speak of a fiction, then it is of a grammatical fiction.10
The origin of this “grammatical” fiction, for Wittgenstein, resides in the
fact that we talk of mental processes and states, leave their nature undecided, yet fail to understand them because the analogy with states and
processes “falls to pieces”.
So we have to deny the yet uncomprehended process in the yet unexplored medium. And now it looks as if we had denied mental processes. And naturally we don’t want to deny them.11
Again:
So you are saying that the word “pain” really means crying?’—On
the contrary: the verbal expression of pain replaces crying and does
not describe it.12
Now, if the language of sensations and mental concepts do not work—in
fact, cannot work—on the basis of a name-object referential model, how
do they work? Wittgenstein raises and answers this very question in PI
§ 244:
How do words refer to sensations?—There doesn’t seem to be any
problem here; don’t we talk about sensations every day, and give them
names? But how is the connection between the name and the thing
named set up? This question is the same as: how does a human being
learn the meaning of the names of sensations?—of the word ‘pain’
for example. Here is one possibility: words are connected with the
9
10
11
12

PI, §293.
Ibid., §307.
Ibid., §307.
Ibid., §244.

102 Annalisa Coliva
primitive, the natural, expressions of the sensation and used in their
place. A child has hurt himself and he cries; and then adults talk to
him and teach him exclamations and, later, sentences. They teach the
child new pain-behavior.13
Hence, according to Wittgenstein, subjects do have sensations, which have
primitive and natural expressions. For instance, in the case of pain it will be
crying and moaning. Adults teach children to accompany and sometimes
replace that primitive behavior with a linguistic one, which becomes
more and more articulated, up to making use of an entire sentence like “I
am in pain”, after having gone through other less complex linguistic formulations, like “Ouch!”14 and “Pain!” Yet, the function of that eventual
sentence is not to describe one’s inner state of mind, nor, for that matter,
to describe one’s primitive pain-behavior. Rather, it is to express and give
voice to one’s sensation, just as a cry is the immediate expression of one’s
on-going pain. “I am in pain”, therefore, is not the linguistic manifestation of one’s belief to be in pain, reached through an inner procedure in
which one’s present sensation is recognized as pain and named that way.
“I am in pain”, for Wittgenstein, is just the refined linguistic, yet by no
means less immediate, expression of pain.
Several consequences follow from this alternative picture of the workings our psychological language. First, it can be apprehended only via
one’s instinctive behavioral manifestations. Hence, for Wittgenstein:
‘But doesn’t what you say come to this: that there is no pain, for
example, without pain-behavior?’—It comes to this: only of a living human being and what resembles (behaves like) a living human
being can one say: it has sensations; it sees; is blind; hears; is deaf; is
conscious or unconscious.15
An ‘inner process’ stands in need of outward criteria.16
The human body is the best picture of the human soul.17
The idea is that only of a living human body (and more generally of a
living animal body that behaves similarly to us), contra a human corpse,
13 Ibid., §244.
14 Notice that “ouch!” is a thoroughly linguistic item. In Italian it translates with “ahi!”.
The usual sequence in Italian is “ahi!”, “Bua”, “Male” (pronounced /'male/, not “male”
like in “male/female”), “Ho/fa male”. “Bua” is considered child-speech, “ahi!” the universal exclamation of pain, “male” the ungrammatical but still acceptable expression
of pain, if it is used by a child. Finally, “Ho/fa male” is the fully grammatical sentence
to be used to express one’s pain. That very sentence, however, can be used to describe
one’s own standing situation. For instance, when one goes to see a doctor and is asked
to describe one’s symptoms.
15 Ibid., §281.
16 Ibid., §580.
17 Ibid., part II, iv, p. 178e.

Self-Knowledge

103

or a zombie, can we correctly say that it does have sensations and feelings
and can thus teach a subject to express them in linguistic terms.
Secondly, “I am in pain” has an expressive function like “Ouch!” or
“Damn it!” That is, even if its surface grammar is that of a descriptive sentence, which can be meaningfully denied and embedded in suppositions—
“Suppose I am in pain”—and therefore in conditional statements—“If I
am in pain, I go to see the doctor”—its primary function is not descriptive. Hence, either it is not up to semantic evaluation or else, if it is, it is
only in a minimal sense.
Wittgenstein is often considered one of the inspirational sources of
minimalism about truth. The idea, that is, that there is no more to truth
than its disquotational function—“Grass is green” is true iff grass is
green—and similar platitudes—if “Grass is green” is true, “Grass is not
green” is not true. Surely, Wittgenstein flirts with minimalism from time
to time, but it is not clear that he really endorsed it, at least in connection with psychological avowals.18 Rather, as we shall presently see, he is
more drawn towards a radical form of semantic contextualism.19
Thirdly, not all uses of the very same sentence are on a par, for him.
For, as we have just seen, “I am in pain” is very often an avowal, hence
an immediate linguistic expression of one’s ongoing pain. Yet, that very
sentence can be used differently when it is embedded in negation or in
wider, especially suppositional contexts. However, given Wittgenstein’s
equation of meaning and use (PI 43),20 that is tantamount to saying that,
despite the identical linguistic form, different occurrences of “I am in
pain” can have different meanings. Indeed, we explain their function, that
is, their use, differently. Since, for Wittgenstein, “the meaning of a word
is what is explained by the explanation of the meaning” (PI 560), that
means that the very same sentence can actually have an altogether different meaning depending on the context of its utterance, not just a different
function. In particular, Wittgenstein distinguishes between the avowing,
merely expressive function, and the descriptive one, which occurs when,
in the context of a conditional statement, for instance, we suppose being
in pain and describe what we would or would not do, just as we would
do in the case of another person.
As we shall see in the following, for Wittgenstein it is important to
realize the existence of these semantic asymmetries, which we are often
oblivious to, given our tendency to think of language as working in a
uniform manner. In particular, it is important to realize that even psychological self-ascriptions can depend on assuming a third-personal stance
with respect to oneself. Still, different occurrences of “I am in pain” in

18 Pace Rockney Jacobsen, “Wittgenstein on Self-Knowledge and Self-Expression,” Philosophical Quarterly 46 (1996): 12–30.
19 Contra, ibid.
20 At least in a large amount of cases, not always, though.

104 Annalisa Coliva
an avowing and in a descriptive mode are related to one another and are
not like “I went to the bank” when used to refer to one’s going to the
financial institute or to the bank of the river. In both cases, the criteria for
using that sentence depend on characteristic forms of behavior. In the one
case, the sentence replaces the instinctive behavior elicited by an ongoing painful sensation; in the other, it is used after a process of inference
to the best explanation, based on one’s reflection on one’s behavior over
time, which is not the one elicited by an ongoing painful sensation, but
still produced by distressing feelings or moods. Moreover, when it comes
to “I am in pain” and to “S is in pain”, it should be kept in mind that
one’s avowals of pain replace an instinctive kind of behavior, which is the
same kind of behavior that allows us to say of some other person that
she is in pain, even though we could never be in a position to avow her
pain. Nevertheless, despite the fact that the similarity in human reactions
could be used to build some bridge between the use in the first-person
and in the third-person present of our psychological vocabulary, Wittgenstein is very much attracted to a strong form of semantic contextualism,21
which tends to deny the uniformity of meaning of “pain” across changes
of person and tense. So there is a family resemblance between “pain”
when avowed and “pain” when used to describe one’s own dispositions
or someone else’s pain, but no identity in meaning.

II. Wittgenstein’s Anti-Epistemological Account
There is a fourth element to Wittgenstein’s overall account that needs to
be taken into consideration carefully. Namely, his idea that the use of “I
know” in connection with one’s own psychological avowals is problematic. Here are some famous quotes:
In what sense are my sensations private?—Well, only I can know
whether I am really in pain; another person can only surmise it.—In
one way this is wrong, and in another nonsense. If we are using the
word ‘to know’ as it is normally used (and how else are we to use
it?), then other people very often know when I am in pain.—Yes, but
all the same not with the certainty with which I know it myself!—It
can’t be said of me at all (except perhaps as a joke) that I know I
am in pain. What is it supposed to mean—except perhaps that I am
21 By strong contextualism, I have in mind what is nowadays called “semantic eliminativism”, a view maintained by Wittgenstein and later in Charles Travis, Saying and
Understanding (Oxford: Blackwell, 1975). I do not mean anything having to do with
forms of assessment sensitivity. Nor is the label “contextualism”, as used here, indebted
to Kaplan’s notion of context and to Grice’s distinction between semantics and pragmatics. Crudely, for Wittgenstein, the notion of context roughly coincides with that of
language-game and is certainly not exhausted by the triad subject-place-time or any
slight extension of it. Moreover, given his equation of meaning and use, there is no room
for a sharp division between semantics and pragmatics.

Self-Knowledge

105

in pain? Other people cannot be said to learn of my sensations only
from my behavior, for I cannot be said to learn them. I have them.
The truth is: it makes sense to say about other people that they doubt
whether I am in pain; but not to say it about myself.22
I can know what someone else is thinking, not what I am thinking. It is correct to say ‘I know what you are thinking’, and wrong
to say ‘I know what I am thinking’. (A whole cloud of philosophy
condensed in a drop of grammar).23
Hence, according to Wittgenstein, other people can know my own mental
states. Thus, the typically solipsist claim “Only I can know my own pain”
is false. Others do too, even though not by analogy with what happens in
their own case, but through the kind of pain behavior that I exhibit. In fact,
“I know I am in pain” is nonsensical, when “I am in pain” is an avowal,
and if “I know” is taken to express an epistemic relation between a subject
and a proposition, or even a fact. Why so? To become clear about that, it is
useful to recall the criteria for the correct and meaningful use of “I know”
Wittgenstein presents in On Certainty.24 Here is a quick summary of them:
1. One must have reasons for one’s knowledge attribution;
2. These reasons must be stronger (viz. different and more secure) than
what they are supposed to ground;
3. There must be a method to find out whether what one claims to
know is the case;
4. The knowledge claim must be relevant;
5. It must make sense to say “I do not know”.
All these criteria would be violated in the case of “I know I am in pain”.
For what could one offer as a reason for “I know I am in pain” apart
from just repeating that one is in pain? (1) would thus be flouted. One
might think that the reason is the sensation itself. Surely, Wittgenstein
would have objected to that, since, in his view, sensations as such cannot
be reasons25 for their ascriptions. As he famously wrote:
When I say ‘I am in pain’ I am at any rate justified before myself”.—
What does that mean? Does it mean: ‘If someone else could know

22 PI, §246. cf. PI part II, xi, p. 221e.
23 Ibid., part II, xi, p. 222e.
24 Ludwig Wittgenstein, On Certainty (Oxford: Blackwell, 1969). For an extended examination, see chapter 2 of Annalisa Coliva, Moore and Wittgenstein: Scepticism, Certainty
and Common Sense (London: Palgrave, 2010). There I also draw the connection with
Wittgenstein’s earlier treatment of the use of “I know” in connection with one’s own
mental states, to point out the continuity of his thought.
25 Reasons, not causes, and not anything that would be subserve an externalist conception
of justification, like an entitlement in Burge’s sense of that term.

106 Annalisa Coliva
what I am calling “pain”, he would admit that I was using the word
correctly?’26
Clearly, the question must be answered in the negative, for him, and in
fact, he goes on to say: “To use a word without a justification does not
mean to use it without right [zu Unrecht]”.27
Now, in the recent literature on Wittgenstein’s epistemology, this
remark is often at the basis of epistemic interpretations, which tend to
attribute to Wittgenstein an early endorsement of non-evidential warrants or, as they are usually called nowadays, “entitlements”.28 As a matter of fact, however, to have a right to use the expression “I am in pain”,
in this context, is not to be interpreted in an epistemic way. What Wittgenstein is saying, rather, is that even if “I am in pain” is not justified by
one’s ongoing pain, to use that sentence to avow and therefore to express
or exhibit one’s pain is perfectly correct. In other words, the use of that
sentence is legitimate even if it is not backed by any epistemic warrant or
guarantee.
Let us suppose, however, that one’s sensations were one’s reasons for
the relevant self-ascriptions. In that case, they would certainly not be any
stronger than the very knowledge claims they are supposed to ground
(contra (2)). For, as we have seen, for them to play a rationalizing role at
all they would have to be given to, and be conceptualized as such by the
subject. But then one’s psychological self-ascription would be justified on
the basis of itself! (Or of another psychological self-ascription relevantly
similar to the one we would be trying to provide a rational ground for, and
which would pose exactly the same kind of epistemological problems).29
As to condition (3), clearly one does not find out whether one is in pain
(at least when “pain” is meant as meaning an occurrent sensation and not
a dispositional state one enjoys and self-attributes through observation
and inference to the best explanation). One is or has pain and gives voice
to it. Furthermore, a claim like “I know I am in pain” would—in the normal run of cases—be totally irrelevant (thus flouting (4)) and indeed odd
in the course of a conversation, up to the point that, if no proper context
for its occurrence could be provided, one would start doubting whether

26 PI, §289.
27 The correct translation is “incorrectly” rather than “without right”.
28 See, in particular, Crispin Wright, “Warrant For Nothing (and Foundations For Free?),”
The Aristotelian Society Supplementary Volume, 78 (2004): 167–212, and Michael
Williams, “Wittgenstein, Truth and Certainty,” in Wittgenstein’s Lasting Significance,
edited by Max Kölbel and Bernhard Weiss (London and New York: Routledge, 2004),
249–284.
29 I develop this point, in connection with a discussion of Peacocke’s account of selfknowledge, in Annalisa Coliva, “Peacocke’s Self-Knowledge,” Ratio, 21 (2008): 13–27.

Self-Knowledge

107

a subject is in her right mind, or would start thinking that perhaps she is
trying to make a joke, etc. Finally, in the normal run of cases, it does not
make sense to suppose that one might not know whether one is in pain
(contra (5)). Of course, there can be cases of confused sensations—an itch
which borders pain—but the phenomenon of vagueness cannot support
the idea that when we experience sensations, in the normal run of cases,
we may not know that we do.30 As to the possibility of unconscious sensations, they would be a conceptual solecism in this connection. Thus,
for Wittgenstein, the use of “I know” in connection with one’s avowals
would not make sense.
Still, there is a grammatical sense of “I know” which would be
appropriate in connection with one’s psychological avowals. Again, the
existence of such a grammatical use of “I know”, for Wittgenstein, is
clear from his remarks in On Certainty. Two aspects of the grammatical use of “I know” are particularly relevant with respect to avowals.
Namely: (1’) the fact that “I know” would actually mean “A doubt
is excluded” or “I cannot be wrong”31 and (2’) that the “I” is unimportant (OC 58).32 Here are the relevant passages in the Philosophical
Investigations:
The truth is: it makes sense to say about other people that they doubt
whether I am in pain; but not to say it about myself.33
If anyone said ‘I do not know if what I have got is a pain or something else’, we should think something like this, he does not know
what the English word ‘pain’ means; and we should explain it to
him.—How? Perhaps by means of gestures, or by pricking him with
a pin and saying: ‘See, that’s what pain is!’ This explanation, like any

30 The argument would simply be a non sequitur. Consider the analogy with colors.
Clearly, there are cases of vague predication, where one subject would say “This is red”
and the other “This is orange”, or where the same subject might oscillate between the
two. That, however, would not mean that she could not know, for focal cases of redness
(or orange), that a given object is instantiating that property.
Notice, however, that the analogy with colors breaks down, for Wittgenstein. For
while in the case of colors we can bring our experience to bear onto our judgment “This
is red”, for instance, in order to justify it, we cannot legitimately bring our experienced
pain to bear on its very self-ascription. Once more, in order for the sensation to justify
its self-ascription, according to Wittgenstein, it should be given to, and conceptualized
by the subject as such. Hence, it could only provide a circular justification for the very
self-ascription we would be trying to justify thereby. I take it that circular justifications
would be no justifications, for Wittgenstein. Hence, by his lights, the whole enterprise
of justifying our self-ascriptions of pain and other sensations would be idle.
31 OC, §59.
32 Ibid., §58.
33 PI, §246.

108 Annalisa Coliva
other, he might understand right, wrong, or not at all. And he will
show which he does by his use of the word, in this as in other cases.
That expression of doubt has no place in the language-game; but
if we cut out human behavior, which is the expression of sensation, it
looks as if I might legitimately begin to doubt afresh. My temptation
to say that one might take a sensation for something other than what
it is arises from this: if I assume the abrogation of the normal languagegame with the expression of a sensation, I need a criterion of identity
for the sensation [itself]; and then the possibility of error also exists.34
And with respect to (2’), the idea that “I” is not a referential expression
when used in psychological avowals, which is present in Wittgenstein
since his remarks in the Blue Book (pp. 66–67), makes its appearance
again in PI §404–411. Here are some passages:
In saying this [‘I am in pain’] I don’t name any person. Just as I don’t
name any person when I groan with pain. Though someone else sees
who is in pain from the groaning.35
It would be possible to imagine someone groaning out: ‘Someone
is in pain—I don’t know who!’—and our then hurrying to help him,
the one who groaned.36
‘But you aren’t in doubt whether it is you or someone else who has
the pain!’—The proposition ‘I don’t know whether I or someone else
is in pain’ would be a logical product, and one of its factors would
be: ‘I don’t know whether I am in pain or not’—and that is not a
significant proposition.37
‘I’ is not the name of a person, nor ‘here’ of a place, and ‘this’ is not
a name. But they are connected with names.38
To say that “I know” in connection with one’s sensations could have a
grammatical use, at most, means to say that there is no inner epistemology. Furthermore, it means to say that it is a characteristic trait of our
linguistic practice that subjects’ avowals are not challenged, in the relevant circumstances, unless there are reasons to doubt of their sincerity.
As Wittgenstein writes, although in the context of explaining away, as we
may put it, the epistemology of self-ascriptions of intentions:
‘Only you can know if you had that intention’. One might tell someone this when one was explaining the meaning of the word ‘intention’
34
35
36
37
38

Ibid., §288.
Ibid., §404.
Ibid., §407.
Ibid., §408.
Ibid., §410.

Self-Knowledge

109

to him. For then it means: that is how we use it. (And here ‘know’
means that the expression of uncertainty is senseless).39
It is important to dwell a bit longer on what it means for a remark to be
grammatical, according to Wittgenstein. The notion of grammar, which
is subject to continuous development over the years, from the immediately post-Tractarian works up to Wittgenstein’s last collection of notes,
namely On Certainty, is connected to the notions of analyticity, aprioricity and of conceivability. Early examples of grammatical sentences, for
Wittgenstein, are (i) “An object cannot be of two different colours all
over its surface at the same time” or (ii) “Patience is played alone”. The
idea, on the one hand, is that these are meaning constitutive propositions. Thus, it is constitutive of what we mean by “patience” that it is
played alone. If we changed the rules of playing patience and allowed
for multiple players, the very meaning of that word would alter. Similar
for what we call “(physical) object”. Now grammar is “autonomous” for
Wittgenstein. This means that it is not answerable to reality. This is clear
from Wittgenstein’s treatment of grammatical propositions such as (i).
We accept (i), not because objects cannot physically be of two different
colors all over their surface at the same time. Nor is it because the structure of our sensory experience is of that kind and excludes the possibility
of perceiving an object as having two colours all over its surface. The
order of explanation goes the other way round, in his opinion: given our
concepts, reality and experience are constrained in the way (i) prescribes.
Similarly, it is not because imagination fails us that we cannot conceive
of an object as being of two colors all over its surface at once (cf. PI §
251). Rather, if by “object” we mean what we do, we would not know
where to start from in order to conceive of a physical object as having
two different colors all over its surface at the same time. Hence, to say
that “I know I am in pain” can at most be a grammatical proposition
means to say that it makes explicit a fact about how we use the vocabulary of sensations, or equivalently for Wittgenstein, a structural aspect
of our conceptual scheme. Hence, in the normal run of cases, subjects’
pronouncements about their ongoing sensations are taken at face value.
Moreover, if someone who sincerely expressed a doubt about whether
she is or is not having an ongoing sensation, we would not understand
what she might actually mean (bar vagueness). The impression of those
words as having nonetheless a meaning would be due to the projection
of meaning onto them from their ordinary contexts of use, in which they
do have meaning.
Still, in philosophy, according to Wittgenstein, we constantly run the
risk of mistaking grammatical propositions for empirical ones, yet as

39 Ibid., §247.

110 Annalisa Coliva
having the remarkable property of indubitability. That is how the ideas of
metaphysical necessity and of infallibility arise, in his view. When taken
that way, however, Wittgenstein’s reaction is to declare them nonsensical;
when understood as being grammatical, instead, his reaction is to find
them obvious. As he writes:
‘This body has an extension’. To this we might reply: ‘Nonsense!’—
but are inclined to reply ‘Of course!’—Why is this?40
The answer to this question is precisely that, underneath their superficial
identity, different tokens of the same sentence may actually be employed
in such different a way as to amount to nonsense in one case, or to grammatical claims in the other case. The mark of the latter is that we find the
statement obvious, just as is obvious to anyone who knows the meaning
of “patience” or of “physical object” that patience is played alone and
that physical objects don’t disappear of their own making if not perceived
and that we categorize plenty of things as falling under that category.
“This hand is a physical object”, then, is obvious to anyone who speaks
our language and who has been brought up within our community.

III. Propositional Attitudes: Expressivism and Genuine
Self-Knowledge
Let us now turn to Wittgenstein’s account of propositional attitudes and
of our knowledge of them. According to him, we should pay attention
to the variety of mental states we can enjoy. In PI §574 he writes: “The
concepts of believing, expecting, hoping are less distantly related to one
another than they are to the concept of thinking”. The difference depends
on the fact that thinking is occurrent and has salient phenomenological
aspects to it. Believing, expecting and hoping, in contrast, have a dispositional element to them, and there is an internal relation between these attitudes and their contents. That is to say, it is constitutive of these attitudes
that they have a certain content and the content they have individuates
each of their tokens as the particular belief, desire or intention it is. This
means, for instance, that if some external cause stops me from desiring
that P, it does not mean that my original desire was a desire that had that
cause as its object. For example, if someone punches me and causes me to
stop wanting to have an ice cream, it does not mean that I wanted to be
punched. This relates also to Wittgenstein’s discussion of actions.41 Intentional actions have reasons and reasons, for him, cannot be their causes.

40 Ibid., §257.
41 PI, §611–648. Cf. Ludwig Wittgenstein, The Blue and the Brown Books (Oxford:
Blackwell, 1958), 23–24.

Self-Knowledge

111

In general, causes may be “detached” from actions without turning them
into different ones, whereas reasons cannot. Intentions, in contrast, are
internally related to their contents. They are not the causes of the actions
that fulfil them.42 In his view, moreover, intentional actions and willing
do not stop short of the action willed.43 Although there may be cases in
which trying and making an effort to bring about the action are relevant.
Nor are they individuated by specific (bodily) feelings,44 although they
may be accompanied by them.
When it comes to our knowledge of intentions and actions (as well
as of beliefs and desires), it depends on the circumstances. In the usual
run of cases, we express, avow or voice our own intentions. In that case,
there is no real epistemology of the mental, for the reasons we rehearsed
in connection with (putative) knowledge of our own sensations. That is,
the criteria for the empirical and therefore meaningful use of “I know”
are flouted. To go over them once more: we do not have independent
reasons for “I know I ψ that P” (where ψ is a propositional attitude verb
like intending, desiring, believing, hoping, willing, etc.). That is, reasons
other than the very mental state itself (contra (1)). A fortiori, even if one
conceded that the mental state could be a reason for our knowledgeable
self-ascriptions, it would not be any stronger a reason than what it is
supposed to ground (contra (2)). For, to play a rationalizing role that
intentional mental state would have to be given to, and be conceptualized as such by the subject, thus making the rendering the justification
circular. We do not find out our intentions, beliefs, etc. We do have them
and give immediate expression to them by saying or thinking “I intend
to φ” or “I /believe/desire that P”. In the normal run of cases, we do not
discover them either through introspection or inference to the best explanation given our behavior or feelings, which may, at least on occasion,
accompany the occurrence of these mental states (contra (3)). To say “I
know I ψ that P” in ordinary circumstances would not count as making a relevant contribution to the conversation. In particular, asserting
it would not add anything to one’s simple self-ascription of the relevant
mental state (contra (4)). Finally, in the usual run of cases, it would not
make sense to say or judge “I do not know whether I ψ that P” (contra
(5)). For that kind of judgment could only meaningfully occur in a context in which I haven’t yet formed the (specific) intention (or the belief or
the desire). Or else, it would make sense in cases where one is confused
about one’s own intentional mental state and would have to apply various kinds of third-personal methodologies to find out what one intends
to do (or desires or believes).

42 PI, §632, cf. BB, 23–24.
43 PI, §622.
44 Ibid., §625.

112 Annalisa Coliva
Now, if the expressivist story has to get purchase in relation to propositional attitudes like intending, believing, and desiring, it will have to be
the case that the relevant self-ascriptions are learned as ways of accompanying and possibly replacing pieces of instinctive behavior. The question
arises, however, whether there is such a distinctive kind of instinctive
behavior, which would manifest our intending, desiring, or believing.
While with pain, it seems clear that there is a characteristic pre-linguistic
manifestation, with propositional attitudes, it is not that obvious. Here is
Wittgenstein’s reply:
What is the natural expression of an intention?—Look at a cat when
it stalks a bird; or a beast when it wants to escape.45
The idea then is that also when it comes to wanting, intending and desiring there are characteristic primitive reactions. For instance, a child wants
to have a glass of water, tries to reach for it, the adult sees her doing so
and says “You want a glass of water. I’ll give it to you”. Again, a child
puts her coat on and stands by the door. The adult says “You want to go
out. We’ll go in a moment”. Little by little, the child is taught to accompany her primitive behavior with pieces of linguistic vocabulary, up to the
point where she herself is in a position to say “I want a glass of water/
to go out”.
The case of belief, however, is certainly more complicated. For there
do not seem to be specific primitive reactions which would signal one’s
believing that today it is sunny, or that one’s favourite toy is soft and tender. In particular, even if the child reaches for her toy and holds it close
to her face, we do not normally teach her to replace that behavior with
a linguistic one that makes explicit the attitude of belief. That is to say,
we do not say “You believe that your teddy-bear (say) is tender”, even
if, to us, the child’s behavior manifests that attitude. We simply voice the
embedded content of the attitude. It is perhaps for this reason that when
Wittgenstein explicitly talks about self-ascriptions of belief in the present
tense, he takes them to be equivalent to the assertion of their embedded
content, just in a more tentative way.
In PI II, x, while addressing Moore’s paradox—“I believe it is raining, but it isn’t”—Wittgenstein observes, first, that “I believe that” is
“transparent”.46 For “The expression ‘I believe that this is the case’ is

45 Ibid., §657.
46 The notion of transparency involved here is not equivalent to the one described in
fn. 3. Rather, it is the one subsequently built upon by Gareth Evans, The Varieties of
Reference (Oxford: Oxford University Press, 1982), Richard Moran, Authority and
Estrangement: An Essay on Self-Knowledge (Princeton: Princeton University Press,
2001), and several other theorists. I discuss this issue in chapter 5 of Coliva, Varieties.

Self-Knowledge

113

used like the assertion ‘This is the case’ ”.47 Perhaps, sometimes, as a “hesitant assertion [of P]”.48 This leads him to hold that “to believe” has a different meaning when used in the first person present, as opposed to in the
first person past, or within a suppositional context, or in the third person.
Here are some relevant passages, whose significance is better appreciated
by keeping in mind Wittgenstein’s equation of meaning and use (cf. PI
§43, §560):
Moore’s paradox can be put like this: the expression ‘I believe that
this is the case’ is used like the assertion ‘This is the case’; and yet the
hypothesis that I believe this is the case is not used like the hypothesis
that this is the case.49
Similarly: the statement ‘I believe it is going to rain’ has a meaning
like, that is to say a use like, ‘It’s going to rain’, but the meaning of ‘I
believed then that it was going to rain’, is not like that of ‘It did rain
then’.50
Of course, we do not find this claim intuitive: Semantic continuity across
changes of tense and person seems obvious to us. Here is Wittgenstein’s
response:
‘But surely “I believed” must tell just the same thing in the past as “I
believe” in the present!’—Surely √–1 must mean just the same in relation to –1, as √1 means in relation to 1! This means nothing at all.51
Hence, the idea seems to be that since √1 admits of two possible solutions—1
and –1—both of which belong to real numbers, while √–1 does not admit
of a solution within the domain of reals, “√” means something different
in the two cases. Although the solutions of the application of the function
are different depending on whether it is applied to 1 or –1, and indeed in
order to provide a solution to √–1 we need to extend numbers to imaginary ones, while no such an extension is needed in order to provide a
solution to √1, that does not mean that the function itself is ambiguous.
Thus, the analogy provided by Wittgenstein to support a strong form of
contextualism about meaning is dubious. Indeed, while one may happily
agree that the function of a sentence may vary contextually—from playing an expressivist role, to playing a descriptive one, and that the criteria
of its legitimate employment may vary accordingly—there is no need to

47
48
49
50
51

PI, 190e.
Ibid., 192e.
Ibid., part II, x, p. 190e.
Ibid.
Ibid.

114 Annalisa Coliva
follow Wittgenstein so far as saying that the meaning of the words occurring in it would change as well.
Notwithstanding Wittgenstein’s strong form of contextualism about
linguistic meaning, there is something else worth noticing. Namely, the
relevance to an appropriate treatment of self-knowledge of his paying
close attention to the different functions the same words, conceding that
they may retain the same linguistic meaning, can have depending on context. He writes:
This is how I think of it: Believing is a state of mind. It has duration;
and that independently of the duration of its expression in a sentence,
for example. So it is a kind of disposition of the believing person.
This is shown me in the case of someone else by his behavior; and by
his words. And under this head, by the expression ‘I believe . . .’ as
well as by the simple assertion.—What about my own case: how do I
myself recognize my own disposition?—Here it will have been necessary for me to take notice of myself as others do, to listen to myself
talking, to be able to draw conclusions from what I say!52
Does it make sense to ask ‘How do you know that you believe?’—
and is the answer: ‘I know it by introspection’? In some cases it will
be possible to say some such thing, in most not.53
What Wittgenstein seems to suggest here is that “S believes that P” is used
to ascribe a dispositional state of mind and that “I believe that P” may,
on occasion, although quite rarely, be used to describe one’s own dispositions and be applied on the basis of the same criteria that govern thirdpersonal ascriptions of belief. In the vast majority of cases, however, “my
own relation to my words is wholly different from other people’s. That
different development of that verb would have been possible, if only I
could say ‘I seem to believe’ ”.54 So, in some unusual cases we would
know of our beliefs as dispositions in a third-personal way—that is, by
inferring from the observation of our behavior, as well as of our words, to
their likely mental cause. In those cases, Wittgenstein notices, “it would
also be possible for someone to say ‘It is raining and I don’t believe it’,
or ‘It seems to me that my ego believes this, but it isn’t true’ ”. Hence, in
these cases, Moore’s paradox would actually disappear.55 Still, in the vast

52 Ibid., part II, x, p. 191e–192e.
53 Ibid., §587.
54 Ibid., 192e. Cf. Ludwig Wittgenstein, Remarks on the Philosophy of Psychology, edited
by G. H. von Wright and Heikki Nyman, translated by C. G. Luckhardt and M. A. E.
Aue (Oxford: Basil Blackwell, 1980), 3.
55 This is a remarkable, even though often under-appreciated point, of Wittgenstein’s
treatment of Moore’s paradox. I have built on it in Annalisa Coliva, “How To Commit
Moore’s Paradox,” The Journal of Philosophy, 112/4 (2015): 169–192.

Self-Knowledge

115

majority of cases, “I believe that P” is used in an expressive way and,
for him, is tantamount to asserting “P”, albeit in a tentative voice, so to
speak. This, as I have argued elsewhere,56 is deeply wrong, even though it
is right to remark upon the variety of uses of the very same words. Other
telling passages are the following ones:
We say ‘I am expecting him’, when we believe that he will come,
though his coming does not occupy our thoughts [. . .] But we also
say ‘I am expecting him’ when it is supposed to mean: I am eagerly
waiting for him. We could imagine a language in which different
verbs were constantly used in these cases. And similarly more than
one verb where we speak of ‘believing’, ‘hoping’, and so on. Perhaps
the concepts of such a language would be more suitable for understanding psychology than the concepts of our language.57
When someone says ‘I hope he’ll come’—is this a report about his
state of mind, or a manifestation of his hope?—I can, for example,
say it to myself. And surely I am not giving myself a report. It may be
a sigh; but it need not. If I tell someone ‘I can’t keep my mind on my
work today; I keep on thinking of his coming’—this will be called a
description of my state of mind.58
In PI §586, Wittgenstein makes similar observations and then goes on
to draw a difference between the “exclamation” ‘I’m longing to see
him!’’ and a different use of the same words: “But I can utter the same
words as the result of self-observation, and then they might mean: ‘So,
after all that has happened, I am still longing to see him’ ”. In PI §587,
talking about love, he writes: “It makes sense to ask: ‘Do I really love
her, or am I only pretending to myself?’ and the process of introspection
is the calling up of memories; of imagined possible situations, and of
the feelings that one would have if . . .”. In PI 588, discussing the role
of self-ascriptions of intentions, Wittgenstein writes: “ ‘I am revolving
the decision to go away tomorrow’ (This may be called a description
of a state of mind.) [. . .] I say at the end of a quarrel ‘All right! Then I
leave tomorrow!’; I make a decision”. References could be multiplied
ad libitum. They all go in the direction of distinguishing between different uses or functions the same words can have and, at least in the
case of self-ascriptions of intentions, between the self-ascription of a
disposition and of an ongoing mental state, with the attendant diagnosis of their underlying epistemology. None in the latter case, and thirdpersonal in the former.

56 Cf. ibid.
57 PI, §577.
58 Ibid., §585.

116 Annalisa Coliva

IV. Conclusion
We have now examined at length Wittgenstein’s expressivism. We can
thus sum up how it fares with respect to an adequate theory of selfknowledge. In the case of first-personal self-knowledge,59 it respects
groundlessness—that is, the idea that one’s psychological avowals are not
based on the observation of one’s own mental states, or on inference to
the best explanation starting with the observation of one’s own behavior.
Nor are they justified, for Wittgenstein, by one’s ongoing mental states.
Indeed, avowals are not expressions of beliefs about one’s own mental
states, which may be more or less justified, or not at all. Rather, they are
immediate expressions of the mental states one is in. Hence, they are simply not up for being epistemically supported one way or the other.
It also respects transparency—that is, the fact that if one has (or is
in) the mental state M, one is immediately in a position to avow it, provided, of course, one has the relevant concepts. Indeed, the expressive role
of avowals has built into it the very idea that they are direct linguistic
expressions of one’s on-going mental states.
Finally, authority—that is, the fact that in the relevant circumstances a
subject’s avowals are not challenged or doubted, unless there is any reason to suspect of a subject’s sincerity—is respected as well. In particular,
it is seen as a product of our linguistic practice. That is to say, once the
relevant conditions obtain, it is part of our practice of making psychological avowals that there is no room for doubt regarding the fact that a
subject, who does avow her own mental state, actually has it.
At the same time, Wittgenstein was well aware of the fact that many
of our psychological self-ascriptions—that is, those having a descriptive
function—are epistemically supported. He noticed that they were arrived
at not just by inference to the best explanation starting with the observation of one’s overt behavior, but also by means of inference to the best
explanation starting with the observation of one’s inner promptings, as it
were. Similarly, he was responsive to the fact that these epistemic procedures might sometimes go wrong and that just as much as the resulting
self-ascriptions would often express genuine instances of knowledge, they
might fail to do so, at least occasionally.
Still, Wittgenstein’s expressivism is not without problems. For he ended
up endorsing a strong form of semantic contextualism, which would pair
a change of function with a change in meaning. In connection with avowals of belief and other propositional attitudes, moreover, it is not clear
what kind of pre-linguistic behavior avowals would substitute. Finally,

59 In Varieties, I distinguish between first- and third-personal self-knowledge to make it as
clear as possible that psychological self-ascriptions, though prima facie identical, may
serve different purposes and may actually be arrived at in totally different ways.

Self-Knowledge

117

their equation with the assertion of their embedded contents is highly
problematic, as his treatment of Moore’s paradox reveals.
These problems notwithstanding, Wittgenstein’s treatment of selfknowledge, while critical of what Strawson would have called “hard”
naturalism, is certainly hospitable to “soft” or “catholic” forms of it.60
The stress on the instinctive behavioral manifestations and the idea that
avowals are linguistic accompaniments or even replacements of those,
and are thus constitutive our “second nature”, clearly testifies to that.
Equally, his stress on the possibility for some of our psychological selfascriptions to express genuine self-knowledge, arrived at through inference to the best explanation starting from various kinds of promptings,
attests to his endorsement of the application of third-personal, naturalistically amenable, epistemic methods to acquire genuine knowledge of our
own mental states.

References
Bar-On, D. Speaking My Mind: Expression and Self-Knowledge. Oxford: Oxford
University Press, 2004.
Bilgrami, A. Self-Knowledge and Resentment. Cambridge, MA: Harvard University Press, 2006.
Chalmers, D. “The Content and Epistemology of Phenomenal Belief.” In Consciousness: New Philosophical Perspectives, edited by Quentin Smith and Aleksandar Jokic, 220–272. Oxford: Oxford University Press, 2003.
Coliva, A. “Peacocke’s Self-Knowledge.” Ratio, 21 (2008): 13–27.
———. Moore and Wittgenstein: Scepticism, Certainty and Common Sense. London: Palgrave, 2010.
———. “How to Commit Moore’s Paradox.” The Journal of Philosophy, 112/4
(2015): 169–192.
———. The Varieties of Self-Knowledge. London: Palgrave, 2016.
———. “How To Be a Pluralist About Self-Knowledge.” In Epistemic Pluralism,
edited by Annalisa Coliva and Nikolaj Pedersen, 253–284. London: Palgrave,
2017.
Evans, G. The Varieties of Reference. Oxford: Oxford University Press, 1982.
Fodor, J. Concepts: Where Cognitive Science Went Wrong. Oxford: Oxford University Press, 1998.
Jacobsen, R. “Wittgenstein on Self-Knowledge and Self-Expression.” Philosophical Quarterly, 46 (1996): 12–30.
Russell, B. The Problems of Philosophy. Oxford: Oxford University Press, 1912.
Shoemaker, S. The First Person Perspective and Other Essays. Cambridge: Cambridge University Press, 1996.
Strawson, P. Skepticism and Naturalism: Some Varieties. London: Methuen, 1985.
Travis, C. Saying and Understanding. Oxford: Blackwell, 1975.

60 See chapter 1 of Peter Strawson, Skepticism and Naturalism: Some Varieties (London:
Methuen, 1985).

118 Annalisa Coliva
Williams, M. “Wittgenstein, Truth and Certainty.” In Wittgenstein’s Lasting Significance, edited by Max Kölbel and Bernhard Weiss, 249–284. London and
New York: Routledge, 2004.
Wittgenstein, L. Philosophical Investigations. Oxford: Blackwell, 1953.
———. The Blue and the Brown Books. Oxford: Blackwell, 1958.
———. On Certainty. Oxford: Blackwell, 1969.
Wright, C. “Wittgenstein’s Rule-Following Considerations and The Central Project of Theoretical Linguistics.” In Rails to Infinity, edited by Crispin Wright,
170–213. Cambridge, MA: Harvard University Press, 2001.
———. “Warrant For Nothing (and Foundations For Free?).” The Aristotelian
Society Supplementary Volume, 78 (2004): 167–212.

6

The End of an Affair
Charles Travis

A third realm must be acknowledged. What belongs to it accords with
Vorstellungen in that it cannot be perceived by the senses, but with
things in that it needs no bearer to the contents of whose consciousness
it belongs.1

Introduction
In 1911, on Frege’s advice, Wittgenstein went to Cambridge to study philosophy with Russell. Perhaps he was ill-advised. For though, it seems,
he did his best to depart from Russell, still, this was within a common
framework. Which took as given what most fundamentally distinguished
Russell from Frege. To the detriment, and ultimate demise, of Wittgenstein’s early work (as he himself saw in 1929).
What separated Frege and Russell so fundamentally can be put in few
words. Where Frege spoke of Sinn, thus thoughts (Gedanken), Russell
spoke of propositions (sometimes ‘objective sentences’). Russell failed
to grasp such Fregean notions. He thus failed to understand, or grasp
the point of, Frege’s two most important distinctions for understanding
the phenomena of thought expression, and of being true (this last most
central to Frege). Those distinctions are: between the logical and the psychological (being true and holding (forth as) true; between the particular
and the general. Tractarian Wittgenstein followed Russell, if not entirely
in vocabulary, at least in insensitivity, or indifference, to those distinctions, most particularly the second. Compared to such agreement, what
remained contentious between the two pales into insignificance.
But these few words themselves call for explanation. Again in brief, the
Tractatus falls foul of Frege’s anti-reductionist point, in “Der Gedanke”,
against theories of truth; a point turning essentially on a categorial
distinction between what is innocent of generality, and that for which

1 Gottlob Frege, “Der Gedanke”, Beiträge zur Philosophie des deutschen Idealismus, v. 1,
n. 2, 1918, 69.

120

Charles Travis

generality is its life. For all of which, Frege and the Tractatus agree on
this much: truth first comes into question with representing something
as being something (henceforth “representing-as”). It is this (even if not
in all cases) which is liable to be representing things as they are, or as
they are not—as Aristotle pointed out, just what being true (or false)
consists in. But Frege and the Tractatus have very different ideas as to
how this core idea unfolds. The Tractatus unfolds it in terms of a certain
presumed relation holding between representer and represented, depiction and depicted. What Frege shows, in effect, is that there could be no
such relation: No relation of the needed sort could cross the categorial
gulf between the two relata, which come together in representing-as.

1. 1902–1904
Russell did not understand Frege’s logical-psychological distinction. Such
is one complication in their debate between 1902 and 1904 on (presently) crucial notions. Such will be our starting point.
The topic of the logical, for Frege, is the pure business of being true
(or false); of ways the world may be made to matter to whether there is
truth, to whether, that is, things are or not as represented. Frege makes
the central item for logic’s purpose what he calls a thought (Gedanke).
A thought is what makes truth turn in some given determinate way—the
thought’s proprietary way—on how things are. It represents things as
being some given way for things to be. Its truth is thereby made hostage to how things are. It represents in a particular aspect of that verb,
roughly: For a thought to represent things as it does is for such to be what
one would do in expressing it. One might usefully think of a thought as
the representing of things as thus and so; as something to be done (rather
than any particular doing of this). A thought thus identifies a particular
achievement, one to be attained in thought-expression; in authorship.
A thought is purely that by which, as Frege puts it, “truth can come
into question at all”; one of the two elements which make for truth or
falsehood. Two thoughts can thus differ only by a feature which distinguishes two different ways for truth to be hostage to the world (i.e. the
represented): for a representation to be hostage to how what it represents as something in fact is. So it is only by such features that a thought
is identified as the thought it is. A thought, thus, cannot be an object
of sensory awareness. Thoughts are invisible. So, too, a thought must
be separated out from anything which has merely to do with achieving
its expression, such as visible or audible forms of words thus used, or
accompanying demonstrations.
Such are logic’s needs. By which we arrive at Frege’s notion Sinn. A
thought is the fundamental case of a Sinn. It identifies, and is identified by, a way to represent things as being. It captures something an act
of representing may be understood to do. Other Sinne are carved out

Affair

121

of whole thoughts. They are partial doings of what that whole thought
does. A whole thought represents things as being a certain way. It is true
or false outright. So the question which things it so represents lapses.
It makes truth outright turn on how things (mass noun) are. A proper
thought element makes truth turn in part on how things are, e.g., on how
some object is, or on what things fall under some given concept. A proper
predicative element represents things as being some given way. For it,
though, unlike the thought, the question which things is very much à
propos. An answer mentions objects. The (Logikgemäss) notion object is
thus introduced. There is nothing even faintly psychological about any of
this, contrary to what Russell seemed to think.
If a thought element is carved out of a whole thought, so, too, in a different way, is a concept. But concepts are not thought-elements in present
sense. As I will use the term here (Frege’s use is unsteady), for any way
for things to be there is a concept which is the concept of things so being.
It is intrinsically so: to be that concept is eo ipso to be of being that way.
If an object is what a concept is a concept of, I will say that the object
falls under the concept. One might also have said: If for something to be
as that object is would be for it to be what that concept is a concept of
being, then that object’s being as it is falls under the concept. But to avoid
confusion I will say instead: then the object’s being as it is is a case of, or
instances the concept. A concept generates answers to questions whether
something falls under it; equally whether something instances it. Following Frege we can suppose that it is identified as the concept it is by the
answers it thus generates. So a concept does not engage in representingas. It is representationally inert. Sid may fall under the concept to be a
wheezer. But it is all one to the concept whether he does or not. Such is
part of what suits a concept to the logical role it is to play; a role distinguished by the above from naming. For present purpose, though, this
is an aside. A concept is thus not a thought-element, any more than an
object is. (Objects do not make truth turn on themselves.) But it may be
put into service, inter alia, by a predicative thought-element: that element
may make truth turn on what falls under that concept.
A thought’s work, making truth turn in some determinate way on how
things are, may be broken up into subtasks. To do so is to decompose the
thought into proper elements. For an element to be the element it is is for
it to perform the subtask that it does. The tasks of proper elements are
ones which could be performed only in the context of a whole thought.
An artist who paints the nose on his subject’s face and then goes out for
a six pack and a chilidog has at least painted a nose, even if fate decrees
that he shall never return to paint more. A thought-element which makes
truth turn on how Sid is would do nothing at all but for the other elements in the decomposition to which it belongs. There is no such thing
as making truth turn on how Sid is, never mind turn how. Such might
be seen as Frege’s context principle. The elements of a decomposition

122

Charles Travis

jointly do precisely what the whole thought does. Such it is for there to
be a decomposition. It follows that a given thought may be—Frege tells
us always is—decomposable in any of many ways, none with a claim to
‘objective priority’.
What sort of thing is as a whole thought represents things if that
thought is true or false outright? If an object, then the same object for
every thought, since otherwise the question ‘Which one?’ would not
lapse, as it does. Anyway, objects combine with concepts (or what they
are of) to form things to be so or not outright. If the thought is that Sid
wheezes, Sid is the one thus represented as a wheezer. But what is it represented as such that Sid is a wheezer? Truth or falsehood do not await
an answer to this question. There are different vocabularies for handling
this issue. I will speak here of things as what plays the role, where such
is to be read as things in general, and as rejecting questions which ones.
‘Things’ should be thought of as a mass out of which objects (inter alia)
may be carved, not as itself an object.
The most crucial thing about a thought for our present purpose is its
generality. This shows up with the fact that there might still have been
that thought, that way to think things were things not just as they are.
Thus that things may be represented as they are not. There would still be
such a thing as things being such that Sid wheezes even if he did not. He
might still have been a wheezer even were he, let alone things, other than
he is/they are. For there to be such a thing as wheezing, the existence of
chilidogs is optional. Nor is such just stipulation. It is part of what fits
concepts to their role: to stand far enough apart from that which, in fact
falls under them for there to be some truths which are contingent. (A
concept must not be a name of what so falls.)
If an object does fall under a concept, it need not have done, nor, for
that matter, have been, for that concept to exist. On the other hand, if it
does so fall it must be intrinsic to the concept that this object, in being
as it need not have been, so counts. Such belongs to the idea that truth is
the joint work of precisely two, distinct partners: how things were represented as being; how things are. Nothing else could make a difference. It
must thus be intrinsic to a thought’s proprietary way of generalizing over
cases that it captures just what it in fact does; that it be true given how
things are if it is true at all. Mutatis mutandis for a concept. So a concept’s
(or a thought’s) proprietary way of generalizing—of generating answers
to questions whether this is/would be how things were represented/
what falls/fell under the concept—must be such as to generate precisely
the answers there are to those questions history provides to be posed.
Grasp of the thought, or concept, would be a capacity to recognize what
answers are thus generated; how to answer such questions to the right
proprietary standard. Thus, we see how thoughts and concepts belong to
thought, not history.

Affair

123

2. Russell
With this background we turn to Russell. ‘Proposition’ is a standard
philosophical waffle word. It comes squarely down on neither side of
that distinction between a sentence (or Satz, or Aussage) and a thought
(thinkable) which Frege highlights. (‘Satz’, too, on the evidence is fit for
waffling.) Writing in German, Russell sometimes uses ‘Satz’, sometimes
‘Proposition’. But in 1902–04 his waffle appears of a different order. He
writes,
One does not assert the thought, which is a private psychological
matter. One asserts the object of the thought, and this is, in my view, a
certain complex (one could say an objective sentence (Satz)) in which
Mont Blanc itself is a constituent. If one does not allow this, one thus
gets the conclusion that we actually know nothing about Mt. Blanc
itself. Accordingly, for me the Bedeutung of a sentence is not the true,
but rather a certain complex which, in given cases) is true.2
Both conceiving/envisioning[/imagining?] [‘vorstellen’] and judging always have an object. What I call a ‘proposition’ can be the
object of a judgement, and can equally well be the object of a representation. There are thus two ways in which one can think of an
object in the case where this object is a complex: one can picture
it, or one can judge it; but the object is the same in both cases (for
example, if one says, ‘the cold wind’ and if one says, ‘the wind is
cold’.) So for me the judgement stroke signifies a different manner
of being directed at the object. Complexes are true or false: if one
judges, he means to be treating a true complex; though naturally he
may be mistaken. But truth is not an ingredient of being true as green
is an ingredient of a tree.3
In this we already see the germ of the collapse of the Tractatus. Though
there is a bit here of vocabularies passing each other in the dark, and a
bit of plain misunderstanding of Frege’s point. Part of what is missed
is Frege’s logical-psychological distinction. For, as we have seen, there
is nothing a bit psychological about a Sinn or a Gedanke. Such things
are tied to proprietary ways of generalizing over the historical, whereas
human psychology, or that of other beings, belongs to history, to what
there simply is to represent as being one thing or another. Another perspective: If A and B are thoughts, then for A and B to be one is, for Frege,

2 Bertrand Russell, “Letter to Frege, 12.12.1904”, Gottlob Freges Briefwechsel mit D.
Hilbert, E Husserl, B. Russel, sowie ausgwählte Einzelbriefe Freges, eds. G. Gabriel, F.
Kambartel, and C. Thiel (Hamburg: Felix Meiner, 1980), 99.
3 Russell, “Letter to Frege, 24.5.1903”, Gottlob Frege’s Briefwechsel, 91.

124

Charles Travis

for any proof of A to be eo ipso a proof of B. What is proof of what is
neither for psychology to decide nor for it to discover.
As for Mt. Blanc being a thought element, the misunderstanding here is
still more fundamental. A thought-element, at least for Frege, is a partial
doing of what a thought does (whole thoughts coming first in logical and
ontological priority). A whole thought makes truth turn in some given
determinate way on how things are. (The two-party enterprise again.) As
it may be, it makes truth turn on how high Mt. Blanc is (whether more
than 4000m). Making truth turn (somehow) on how Mt. Blanc is may
be, in the context of the whole, doing part of what that thought does. We
may thus speak of an element which so does. But Mt. Blanc makes truth
turn neither on itself nor, for that matter, on anything else. It is something
on which truth can turn, not something which makes truth depend on
the world in any such way. Thus it is that Mt. Blanc cannot be a thoughtelement on Frege’s scheme of things. And, like the scheme or not, there
must be some ingredient in the business of being true which does what a
Fregean thought-element does; which plays party of the first part in that
two-party enterprise.
If thoughts were both private (the only thing, it seems, Russell can
understand by ‘psychological’) and that by which truth can come into
question at all, then, indeed, we would be cut off from the world in
thought, and we should never know (or so much as think) anything about
Mt. Blanc. But perhaps there is another reason he thinks Frege is cutting
us off from our environment. In “Der Gedanke”, Frege speaks of three
realms. The first is a realm of private objects of consciousness. In this
realm, truth makes no appearance at all. Nor does perception. We do not
see or hear the denizens of this realm, Frege assures us, we have them.4 In
the second realm, we find denizens of an environment, objects of perceptual awareness, about which (non-accidentally) there may be truths, with
all the objectivity which forms part of that notion. In the third realm, we
find thoughts, concepts, thought-elements. These, like occupants of the
first realm, are not possible objects of sensory awareness. Like occupants
of the second, however, they are objective. There are matters of fact, e.g.
as to how a given concept generalizes. Russell seems simply blind to the
possibility of such a third realm. If not the first, he is also certainly not the
last otherwise first-rate philosopher with such a blindspot. To which one
can only say: Without such a realm, there are no thoughts (in the nonpsychological sense); thus, correlatively, there is no truth.
What of Russell’s ‘objective complexes’? The problems with these
will be spelled out in detail later. For the moment, a complex is a structuring of a particular set of elements, in terms of given relations. From
his examples (that cold wind, Mt. Blanc reaching upwards as it does, a

4 Vide Frege, “Der Gedanke”, 69.

Affair

125

certain green tree), these would seem to be denizens of an environment:
the sort of thing which can pierce through one’s fleece into his chest as
he pushes north; the sort of thing planes must avoid by remaining at a
certain height, and so on. In what sorts of relations do these things stand
towards one another? For example, temporal ones: The wind preceded
the avalanche. Or spatial ones: The mountain is to the port side of our
flight path. Such elements, so structured, do not add up to anything either
true or false. What is true may, metaphorically, touch one’s heart, but
not, literally, pierce one’s chest. But we need not dictate what Russell’s
elements must be. Anyway, a complex which was true or false would first
of al be one which posed, or presented, a particular question of truth. Its
elements would be, or what made, particular contributions to that question (or its posing). Such elements would not be the sorts of things which
could pierce chests or rise to immense, or even minute, heights. But to see
this one must grasp, as Russell did not, the point of Sinn; see why making
truth turn on itself is not something Mt. Blanc might do.

3. Relating Realms
The realm of Vorstellungen (what is proprietary to a particular consciousness) does not concern us here. Russell seems to picture the relation between Frege’s second and third realms as though that third one
is so independent of the second (of an environment) that if thinkables
belonged there we would be cut off from anything to think about. But it
is not like that. As Frege insists, a concept cannot have ‘an independent
existence’. Several ideas coincide under this rubric. Here is one:
What I see as essential to a concept is that the question whether
something falls under it has a sense . . . A concept is unsaturated in
that it requires something which might fall under it; thus it cannot
exist on its own.5
A concept is identified by its proprietary way of generalizing, of answering, across an indefinitely extendible range of cases, questions whether
something falls under it. But this generalizing cannot stop anywhere short
of capturing or rejecting these particular cases, anywhere short of objects
actually falling under it or failing to. A concept maintains its decorous
distance across a sort of buffer zone from that which does fall under it. Its
logical role (notably its role in making for contingency in truth) requires
this. Nonetheless, what there is to fall under it must, in fact, do so or not.
The concept, in being of what it is, must decide this where there is such to
be decided. If a concept named what fell under it, perhaps polysemously,

5 Gottlob Frege, “1882 letter to Anton Marty”, Gottlob Frege’s Briefwechsel, 118.

126

Charles Travis

then if Sid were pear-shaped he would be necessarily so; if he were not,
then again this would be necessarily so. Truth would be analytic through
and through. Thus the buffer just described. But for the concept of being
pear-shaped to generalize as it does, for it to answer questions who falls
under it in its own proprietary way is, inter alia, for the way Sid is to be
a case of someone being pear-shaped. Thinking Sid to be so shaped does
not cut one off from the circumstance of his so being. Quite the contrary.
Mutatis mutandis for thoughts. A thought element (on some decomposition) may make the whole thought’s truth turn per se on how Sid is (in this
thought he is the one who must be pear-shaped). But for Sid there would
have been no such element in any thought. By contrast, a thought-element
may make truth turn on who is pear-shaped. Were Sid not shaped as he is,
there would not be this to make for truth or falsehood. If, otherwise shaped,
he were not pear-shaped, that way for things to be would not have had the
instances it does. It would, for all that, still be the way it is. Inter alia, there
would still be thought-elements making truth turn on which things are it.
Such distance between a concept and what falls under it, or between the
way things are represented and what is so represented, does not distance
the truth of a thought from things being as they are represented.
Being pear-shaped is what it is no thanks to Sid, or his being as he is.
For all of which, among the questions to which the concept of so being
generates an answer is the question of whether Sid being shaped as he is
is being pear-shaped. Russell (of this period) seems to think as though if
a thought needed a thought-element to make its truth turn on how Sid is,
then the thought itself would not really be about Sid, so that to subscribe
to its truth would not really be to subscribe to a view about him. But such
merely misunderstands the role Sinn is meant to perform.
Frege also stresses a still more fundamental point. Holding true, and
holding forth as true, cannot be (or merely be) predicating anything of
anything—truth, included. For the result of doing that is just something
which one can hold/hold forth as true, or, equally well, as untrue. Rather,
to hold a representation true—to take it to represent things as they
are—is to subscribe to its truth; to be prepared to treat it, hence what it
represents as it does, accordingly. To hold something forth as true is to
underwrite subscribing to its truth, thus so treating things. It is to offer
authority as to what one purports to be authoritative. Frege writes,
[The word] ‘wahr’ really only makes a failed attempt to identify
logic, in that what it [logic] is really about does not lie at all in the
word ‘true’, but rather in the assertive force with which a sentence is
uttered . . . That which most clearly contains indication of the essence
of logic is the assertive force with which a thought is expressed.6
6 Gottlob Frege, “Meine Grundlegenden Logischen Einsichten”, Nachgelassene Schriften, 2nd
ed., eds. H. Hermes, F. Kambartel, and F. Kaulbach (Hamburg: Feliz Meiner, 1983), 271.

Affair

127

For a thought to be true is for what it represents as being such-and-such
to instance (be a case of) things being the way it represents them. Assertive force is offering authority for, underwriting, so treating what it so
represents, namely, things being as they are. A thought represents the way
things are as a certain way. Two thoughts are distinguished from each
other only by what way they represent what they do. But it is not as if
the truth of a thought concerns only that identifying feature and not both
those sides of the representing-as relation. On the contrary, there is an
intrinsic connection between truth and the thing to do; how a thinker’s
goals are, or may be pursuable. To subscribe to the truth of a thought is
to subscribe to acting accordingly, or if not that, to recognizing so acting as the thing to do, as how the world is treatable. There is a thought,
Frege tells us, just where it has applications; where its truth matters in
determinate ways to questions what to do.7 If thinkables, Sinne of a certain central kind, are what are true or false, this by no means cuts us off
from knowledge of the world we live in. On the contrary, it precisely is
such knowledge. Frege’s third realm is exactly what he says it is and no
more. What inhabits it belongs to thought, not to an environment. But its
existence is not independent of our environment.

4. Russell’s View
What is true, Russell tells us, is a complex of a certain kind. A complex
consists of given elements structured (in the complex) in some given way.
E.g., a certain wind and cold (a rather unpleasant physical phenomenon).
Russell is less clear as to the sort of structuring he has in mind. In any
case, a complex (of his sort) is to be one to which a thinker can stand in
either of two ways. He may present it to himself (e.g. imagine or remember it). Or, it may be what he thinks in thinking that such-and-such. In
this last guise, it may be true or false. Curiously, mention of it in the one
role would take a very different grammatical form than mention of it in
the other would. How could this be?
Compare the case of sentences and thoughts. A sentence, say, “Sid is
pear-shaped”, used to express a thought, might be thought of as presenting that thought as decomposed in a given way—in the example,
say, into an element which makes truth turn in part on how Sid is, and
another making truth turn in part on who is pear-shaped. On Frege’s
view, another sentence—say, “The concept being pear-shaped is satisfied
by Sid”, might be thought of as presenting that same thought; but then,
of course, decomposed in a different way. The difference: Decomposing a
thought in different ways yields different elements, standing towards one
another in different relations. For Frege, to repeat, whole thoughts come
7 Vide Gottlob Frege, Grundgesetze der Arithmetik (Jena: Verlag Hermann Pohle Band
I/II. 1893/1903), §91.

128

Charles Travis

first. They are there to be carved up in any of many ways. A thought is
not a given complex. It does not have a unique structuring. Which allows
Frege to recognize such a relation between sentences (or what he calls
‘Aussagen’) and thoughts. For Russell, though, that which can be mentioned in those different grammatical forms is a complex. It must thus
be supposed to be identified, inter alia, by its proprietary structure. This
analogy between Russell and Frege thus breaks down.
In 1879, Frege tried out the idea of a language with only one predicate.
So a sentence for expressing something true or false (a ‘judgeable content’)
might look like this: “Lager/lager’s containing yeast”, or “Odysseus/
Odysseus’ having sailed”, or “Sid/Sid’s being rotund”. The predicate might
then be something like ‘obtains’, or ‘is so’, or ‘is a fact’. He proposed this
for Begriffsschrift. Thus, where ‘A’ stood in for a judgeable content, the
horizontal stroke, ‘—A’, was to be read: ‘the circumstance of A (or A’s)
being so. Adding a vertical stroke (‘|—A’) would then turn this into the
form of an assertion: that A (that A is so/a fact).8 So here we have a complex which is not itself either true or false, but which is, so to speak, but
one step away. It calls only for a predicate. Could such a complex be what
Russell had in mind?
Prescinding from the fact that thoughts (coming first) are multiply
decomposable, the occurrence of the same content-indicating letter, ‘A’,
might be taken to indicate that the same complex of elements occurs
twice, in the structure with the horizontal stroke only, and again in that
with the vertical stroke added. The same elements, perhaps. But not the
same structuring of them. But this is not the point to push hardest here.
That point can start from the observation that if the proposed single
predicate (‘is so/is a fact/obtains’) genuinely predicates some way for
something to be, then that way-to-be is of a very peculiar sort. (Perhaps in something like the way in which being true is a peculiar way
for something to be, if one at all.) For though a way for things to be
(equally concept thereof, equally what predicates some such of something) is, in general, identified as the way it is by its proprietary way of
generalizing (of yielding answers to questions whether such-and-such
is a case of so being), the proposed single predicate here can have no
such proprietary way of generalizing. Rather, when combined with,
say, Sid’s being rotund, it acquires the generality of that way for things
to be (for them to be such that Sid is rotund), when combined with,
“Lager’s containing yeast”, it acquires the generality of that way for
things to be, when combined with “there being no largest prime”, it
acquires the generality of that way for things to be, similarly for “every

8 In fact, in the context of Begriffsschrift it would mark something being asserted—add
assertive force. Vide Gottlob Frege, Begriffsschrift, eine der arithmetischen nachgebildete, Formelsprache des reinen Denkens (Halle a. S.: Louis Nebert, 1879), 2.

Affair

129

prime being divisible by 2”, and so on. What it does, in effect, is to
unleash a generality already present in, e.g., (the circumstance of) Sid
being rotund; that is, put just that generality to predicative work; work
in the project of representing things as being some given way for them
to be outright. So, second, such generality is already present in what
follows the horizontal stroke (in Sid’s being rotund). What the one
predicate ‘predicates’ of is not some historical state of affairs, say, Sid’s
rotundity from the time his romance with lager began up to the present (something to be experienced palpably). It is rather a sort of thing
liable to occur (be so) or not. It is thus, unlike that cold nortada, not
something one may recall (with yet a shiver). Unlike Mt. Blanc, it is not
located in an environment. Exactly not. It belongs to thought in just
the same way that Frege’s Gedanken do. Russell gets to say whether
this is what he had in mind. But it is embracing, rather than rejecting,
Frege’s notion Sinn.

5. Two Distinctions
Representing-as is the prerogative of thinkers, or, secondarily, of their
means of achieving it (their words), or of that abstraction from their so
engaging, something there is for them to do (‘der Gedanke’). It is marked
by its contrast with factive meaning: if Sid’s unsteady gait is due, not to
imbibing, but to an inner ear problem, then it cannot mean falsely that
he is in his cups. It does not mean he is in his cups at all. It is in opening
up that first possibility that representing-as brings truth, or questions of
it, onto the scene. Frege made two fundamental contributions to clarity
as to what thus comes into view. They are his distinction between the
psychological and the logical, and a distinction between the particular
and the general (one with which he was not always surefooted himself).
Russell, in ’02–’04, was insensitive to both.
Truth is a cooperative enterprise between precisely two parties. The
work of party of the first part is to fix what truth (or a certain truth)
depends on; when there would be (the relevant) truth. The work of party
of the second is, in being as it is, to make good on the demand thus
imposed—or, again, to fail to. Party of the first part must, inter alia,
assure that party of the second will, in fact, settle that question of truth
it thus poses. It cannot be settled only relative to some third party. Such
is the objectivity of truth. In representing-as, party of the first part is the
representing-as, or what does it. So where there is truth, party of the first
part is that truth. Party of the second is what is so represented—where
there is truth outright, whatever is a case of the relevant way for things to
be outright. As discussed already, party of the first part is always distinguished by its proprietary generality. To be a party of the first part is to
have and deploy generality: As Frege put it, “something reaching beyond
the particular case, whereby this is presented to consciousness as falling

130

Charles Travis

under some given generality”.9 Party of the second (in re truth outright) is
always devoid of such generality. It just is a particular case. If to be as Sid
is is to be rotund, nothing in that decides what another case of rotundity
might be. If this is all there is to say as to what it is to be rotund, then as
for Pia, sylph-like as she remains, such might or might not be rotundity if
there could be any such thing as her being or not at all. And so it is with
particular cases in general, always in re truth outright, so, too, for truth
of anything environmental.
It follows that what is true or false cannot be a particular case; cannot
thus be a complex of elements all of which are innocent of generality
(of the sort just scouted), or any structuring of elements which does not
arrive at deployment of such generality. If Russell’s complexes are meant
to contrast as he thinks they do with Frege’s Sinne, then he falls afoul
of this fact. The generality of a way for things to be is as fundamental
to representing-as, and thus to truth, as is that objectivity crystallized in
the idea of an essentially two-party enterprise. It is just by this generality
that the predicative distinguishes itself from the designative, making truth
turn on how some given object (or collection) is. It is just this distinction
which allows for contingency in truth.
A concept, a way for things to be, a thought: one cannot watch these
“reaching over and beyond the particular case”, generalizing so as to
capture or reject further cases. Such things are necessarily not objects of
sensory awareness. Features of which to enjoy such awareness are not
ones by which one way of generalizing is distinguished from another.
By contrast, one can watch Sid getting them in him, one can witness Mt.
Blanc rising into the sky (from where he stands). Being Mt. Blanc has a
generality under which that particular mountain and it alone falls, which
it alone instances. Nothing falls under the mountain in that sense. It alone
can be Mt. Blanc. But not even it can Mt. Blanc.
Russell insists that one does not assert a thought (Gedanke), but rather
the object of the thought “das Objekt des Gedankens”, a complex formed
of, e.g., Mt. Blanc and its stature, or that cold wind and attendant mean
molecular motion (or lack thereof). But truth is a phenomenon arising
with representing-as. There is none of that to be found in such complexes. There is no way things are according to that cold wind; nor in
that mountain in towering as it does. So Russell’s complexes are just the
wrong sort of thing to be either true or false. To begin (but not end) with,
they lack the requisite generality. Sid, of course, plays some role in the
thought that he is in his cups. He is that object (in his case, that obstacle
to free passage) on which, in that thought, truth is made to turn. But it
is not so made by him. He is not a partial doing of that which the whole
thought does, namely, to bring truth into question in a particular way.

9 Frege, “1876/1882: ‘17 Kernsätze zur Logik’ ”, Nachgelassene Schriften, 189–190.

Affair

131

Nor is he what so does. To repeat, a person cannot make truth turn on
himself. (Nor can this be done at all except within some decomposition
of a whole thought.)
Russell simply misses the difference between particular and general,
or the point of that distinction. As he misses the difference between (a)
Gedanke (count) and Denken (mass), and the two readings of this last:
on one an historical phenomenon, the Denken that is done, or in currency among given thinkers; on another a body of thought there is to be
thought. Accordingly, he misses the possibility for Frege’s Sinn (among
which Gedanken) to lie on the logical side of the logical-psychological
distinction, to be in no sense psychological. He thus arrives at, perhaps
is driven to, complexes which are ineligible for truth. His story collapses
at that point.
The particular does not generalize. Its home is firmly in the represented. It cannot do double duty on both sides of the representing-as
relation. It cannot, thus, form complexes which are either true or false.
Russell rejects Frege’s third realm. But what one loses with that is truth
überhaupt. Objectivity may seem (as it seems to have seemed to Russell)
to require locating objects of thought in the historical, or the environmental. But, so doing, the very idea of objectivity vanishes. The Tractatus
may not follow Russell in all above details. But it follows him in erasing,
or ignoring, the distinction between particular and general. Therein lies
its downfall, as we shall see.
At first it may seem that Russell’s concern is just to place such things
as Mt. Blanc within truth-bearers themselves. But it is perhaps better seen
as placing truth-bearers themselves within the realm to which Mt. Blanc
belongs; having them be, or consist of, solely such things (portending,
perhaps, Quine). As though if there is really to be such a thing as thought,
then thought itself must be situated in the world just as Mt. Blanc is. As
indeed it is if we read thought as our thinking, rather than as the thinking
there is to do (i.e. what there is to think). Well, there may be plenty of
objective complexes in the world, one of which, perhaps, is to be found
there at all only if Mt. Blanc is more than 4000 meters high. But if I were
now to try to speak on that topic, why should anything I say be taken
to speak of any of these? Rather than rendering thought real, the move
Russell tries to make simply abolishes it altogether.

6. Correspondence Theories
The Tractatus, I will argue, founders on the means it allows itself for
unfolding the notion to be true. We can see this by working through
Frege’s critique, in “Der Gedanke”, of correspondence theories. But for
this we first need to consider what a correspondence theory might be.
Such a theory is not simply the remark that a true thought (or statement)
is one which corresponds with the facts. Not, at least, if all this means

132

Charles Travis

is that a representation represents things other than itself as being something or other, so that its truth depends on how those further things are.
And not, again, if all it means is that if it is true that P then it is a fact/so
that P. Here I will suppose that a correspondence theory aims at satisfying two demands. First, it aims to answer the question what being true is;
and to do so in a non-question-begging way. Second, it aims to define a
relation between representations and what they represent as being suchand-such (that representation’s represented, or, to use a Tractarian term,
Abgebildete), in terms of which one can define when any given representation would be true: either for a representation to be true would be for
it to stand in that relation to an Abgibildete, or it would be for the Abgebildete it anyway so relates to to have some further property it might or
might not have. (The Tractatus’ story is essentially of this second sort.)
There are thus two forms for a correspondence theory to take. For an
account of the first form we might start from the truism that it is true
that P just in case it is a fact that P. Now, to generate answers to questions when a representation would be true, when it would have that feature, we must start with a domain of representations. If what we aim to
explicate is the phenomenon, or feature, of being true as such, we should,
with Frege, put whole thoughts first. Such will be the representations in
question, which identify particular questions as to being true. As to the
domain of items to which these are to relate, we might develop the truism by nominalizing. We might suppose that wherever it is a fact that P,
there is a corresponding fact, namely, the fact that P. This would not be a
representation, but might rather be conceived as some circumstance, or
state of affairs obtaining in, or some feature of, the way things are, something for one to steer a course through, which, we might suppose, would
be there to be encountered only if things were now such as to make for
a case of its being so that P. So we now have a domain of thoughts and a
domain of facts. What we would like to complete the account is a function which would map each item in that first domain onto one of two
things. For any thought that P, the function should map that thought
onto the fact that P if there is such a fact. If not, then the function should
map that thought onto some designated item, say, Russell’s left big toe.
Then the theory can tell us that a thought is true just in case the designated function maps that thought onto a fact, false just in case it maps
that thought onto Russell’s left big toe.
The rub here is evident. Letting all the above as to what a fact would
be pass, there is no problem saying what fact would thus correspond to
a given thought, were there such a fact. But to fix when a thought would
be true it remains to say when there would be such a fact. For which
thoughts there are such facts depends, rather heavily, in general, on how
the world is. (Grasping the thought that Sid is in his cups as the thought
it is does not yet enlighten one as to whether it is true.) The end of time
would be a bit too soon to demand a theory which actually took account

Affair

133

of how all such dependencies panned out. So what we still need is an
adequate explication of just how a thought, never mind which, depends
on things being as it represents them (for its being a fact/so that things
are thus). We need, that is, an adequate general explication of what the
dependency in question is. But to say this is to say that what is so far
unexplained is simply everything as to what being true is.
With which we return to the second option. On this option, there are
two domains, the first made up of thoughts, the second made up of ‘complexes’ of some sort, which are to count as Abgebildete, each for some
thought or other. The idea would be that, for each thought, there is some
proprietary such complex which is that which it in particular represents
as thus and so. So we now want a function which maps each thought
onto its proprietary Abgebildete. For example, suppose we thought that a
thought (Gedanke) has a unique (fundamental) decomposition, and that
these could be specified in some systematic way. The thought that Sid is in
his cups, for example, might be per se to be decomposed into an element
which made it about Sid, and another which made it about one being in
his cups. Then we might suppose we could identify things to be found in
what any whole thought represents-as the way it does, which form an
Abgebildete peculiar to the thought in question. Such might look like a
Russellian complex of the Russellian sort exemplified by ‘the cold wind’—
say, Sid and the worldly presence of a person being in his cups (whatever
that is), and those two items relating to each other as they in fact do.
Such sketches an idea for a way of associating each thought with some
worldly complex which is peculiar to it. On this sketch of an idea, for
each thought there would be such a complex, independent of whether
the thought was true or false. So we would not have to wait until the
end of time, not nearly that long, to be able to say what the value of the
relevant function from domain to range would be. The thought that Sid is
in his cups maps onto the same Abgebildete in this sense whether he is in his
cups or not. So far, let us suppose, so good. (Though in fact an immense
amount must be assumed to get us this far.) Now, of course, the same
rub. If both true and false thoughts map onto proprietary Abgebildete of
this sort, then everything remains to be said as to when a thought which
so did would be true: when an ordered pair of thought and Abgebildete
so relates that that first element in the pair is true. What remains undone
is to specify what property an Abgebildete would have when it was the
second member of such a pair. And this must be done in a non-questionbegging way, as per the first demand above on what would count as a
correspondence theory. So, for example, one could not say that the relevant property, for the value of the relevant function for the argument,
the thought that Sid is in his cups must have the property of being as
represented in that thought. Such is a property any Abgebildete would
have if the thought which abgebildet it is true. But to say this is not to say
in non-circular terms what it would be for something to be true.

134

Charles Travis

For either of these two forms of correspondence theory there is, of
course, another fundamental, but so far unmentioned problem at the very
start. To identify a function from thoughts to something—whether to a
domain of facts plus a designated object, or to some domain of Abgebildete, we must make intelligible, or presuppose to be intelligible, what the
relevant domain of the function is to be. What, notably, are its inhabitants? When would we have a given one in hand? What would it be to
be a given one? The obvious problem: a thought is identified entirely
and exclusively, by its way of making truth turn on how things are; its
proprietary way of bringing truth into question. But if the account is to
be non-circular, its aim to answer the question what being true is without presupposing that notion itself, then the notion of being true is not
yet available to it. Inter alia, it cannot be presupposed in identifying the
domain of the relevant function.
We might, then, try departing from Frege, letting arguments for the
function we seek be, not thoughts, but items such as sentences, identifiable as the items they are by visible or audible forms, or by other features had or not independently of how they represent something as being.
Then, indeed, e.g., “The sentence ‘Sid is in his cups’ is true just in case
Sid is not sober” is informative, if true, in a way that a parallel remark
about the thought that Sid is in his cups would not be. But then, as Frege
remarked, “One cannot forbid anyone from adopting any arbitrary producible event or object as a sign for whatever he likes”.10 Suppose what
was wanted was a function which mapped sentences such as this into, say,
abgebildete. Specifying the right function now simply becomes still more
problematic. Not only must we say, without circularity, what property
an abgebildete should have for true to be the value for what represents
things as a given way as argument, but also what property a sentence
should have to have as its value an abgebildete with such a property.
A last remark. Suppose a would-be correspondence theory proceeded
piece-wise, for the thought that Sid is in his cups, saying what its being
true would be, for the thought that Pia just nutmegged Max, what it
would be for it to be true, for the thought that sea water is saline what it
would be for it to be true, and so on. Well, as they say, there’s one borne
every minute. Another human, or anteater, a new blade of grass, and so
on. With each such thing comes a mass of thoughts there need not have
been but in fact are. Frege’s birth, we may suppose, is a highly contingent
matter (that unusually cold February, etc.). But for it there would be no
thoughts of Frege that he was thus and so, nothing to think about him
at all. So take any piece-wise theory and wait for tomorrow. There will
then be countless unforeseeable cases of items being true or false, and no
account of what it would be for them to be this.
10 Gottlob Frege, “Über Sinn und Bedeutung”, Zeitschrift für Philosophie und philosophische Kritik, NF 100, 1892, 26.

Affair

135

There is, thus, a fundamental problem with the idea of a correspondence theory of truth. For any given version of the theory, there may well
be a way of pushing that problem under the rug (by moving to another
form of theory). But, it seems, such merely makes the problem push up
from under in a different place. Such sketches the form a correspondence
theory might take, and at least gestures at a fundamental problem it
would then face. With which, we are now prepared to look at Frege.

7. Defining Truth
Early on in “Der Gedanke”, Frege presents a case which concludes,
oddly, thus: “From this it is probable that the content of the word ‘true’
is entirely singular and undefinable”.11 To understand an argument, or
train of thought, one must first look for its beginning and then for where
it aims to end. Frege’s case starts from a question:
Is a picture, as mere visible, touchable thing, really true, and a stone,
a leaf not true?12
To which he answers,
Clearly one would not call a picture true if an intention did not attach
to it. The picture is to be taken to represent something.13
His case expands on this last remark. As for termini, there are two: one
immediate, the other, more general, later on. The immediate conclusion:
What emerges from this as the only thing by which truth can come
into question at all is the Sinn of a sentence.14
The only thing as opposed to what? There are two answers to that. First,
the case is inter alia against Vorstellungen as truth-bearers. These, in
Frege’s sense, would be objects of (e.g., sensory) experience and, if truthbearers, of thought; would require a bearer and brook no two—would
belong exclusively to a given consciousness. Inter alia, were they truthbearers, their objects would cross a certain categorial line separating off
the sensible from the thinkable. Already reason to object. Aside from
that, though, good riddance to them, and bracket them here.
A second more general concern is evident in the above immediate conclusion. One might think words as such were the fundamental truth-bearers
11
12
13
14

Frege, “Der Gedanke”, 60.
Ibid., 59.
Ibid.
Ibid., 60.

136

Charles Travis

(that by which ‘truth comes into question at all’). Or one might, as in
Frege’s initial question, think pictures, as, e.g., the products of a Sunday
painter, could, on their own, raise questions of truth. Or, more generally,
something or other with a visual, or audible form; still more generally,
something identifiable as the item it is independent of when it would
be true. Frege’s main aim in the ensuing discussion is to help that idea
out of the world. That by which truth can come into question at all is,
he insists, a thought (Gedanke), as per above. Such, in one form, is the
main point to be established. (A thought is a kind of Sinn in Frege’s
technical sense; since whole thoughts come first, the fundamental kind.
Thoughts, as we have seen, have a non-linguistic nature.) Such fits
Frege’s longer-term conclusion, that a ‘third realm’ must be recognized.
(See above.)
Frege’s answer to his initial question is: A picture can only be true if an
intention (‘Absicht’) attaches to it. Perhaps ‘Absicht’ is not quite le mot
juste. What matters is that a picture, to be in the business of being true,
must first represent things as some given way. For that it cannot just be,
e.g, a pattern of paints on canvas. Insofar as it has such a Gestalte, said
Gestalte must be to be taken, understood, in some given way. It must
bear an understanding as to how it is depicting; thereby how it is that
things are being depicted as. What it does in being so understood is not
reducible to how the picture is anyway (e.g., what paint in what places),
independent of the understanding it is thus to be given. Exactly not: just
so that there is room for an understanding (or bearing one). The canvas
cannot be to be understood to have paint on it, as if there might be some
other way of perceiving it correctly. It just is painted (modulo, perhaps,
what is to count as being painted). How it depicts is another matter; one
which contrasts with this.
I will divide Frege’s case in fourths. The first quarter runs:
Accordingly one might suspect that truth consists in a correspondence of a depiction with what it depicts. A correspondence is a relation. The use of the word ‘true’, which is not a relation word, and
contains no reference to some further thing with which something is
supposed to correspond, speaks against this.15
The thought that Sid is in his cups is decomposable into, inter alia, a
predicative element which predicates being in one’s cups of someone.
If we ask whether that element thus predicates truly, we may fairly be
asked in return, ‘Predicates of what?’. For such representing may be true
of one object, false of another. By contrast, where there is truth outright,
there is, eo ipso, no question ‘true of what?’. In that sense, truth outright

15 Ibid., 59.

Affair

137

is indeed not a relation between a representation/depiction and an object.
If we insist that a relation relates objects, then indeed correspondence of
the sort truth might consist in is not a relation. On the other hand, the
truth of a thought depends on how things are. For there to be truth is thus
for something, how things are, to oblige. Such just is a sense in which to
be true is to correspond with the facts—though not thus with an object.
Whether the word ‘relation’ fits here can remain moot. Frege is thus not
off to a good start.
On, then, to the second quarter:
If I do not know that a picture is meant to represent Cologne Cathedral, I do not know with what I must compare it in order to decide
its truth.16
Some lines on a napkin, or color patches on a canvas, might or might not
depict Cologne Cathedral. They might be just doodling or paint spills. Or
they might depict a cathedral, though no one in particular. There would
thus be no question of truth. (If your picture of a cathedral looks more
like a car park, such is not how a cathedral looks. Still, falsehood is not
quite yet in the cards.) There is thus a distinction between properties a
picture would have no matter how it was to be understood (so no matter
how it represented things), and properties it has only understood as it
is to be. No properties of the first sort can decide whether the picture is
true, much less how it would be. There are indefinitely many manners of
depicting something. Think of a recreation of ‘Las Meninas’ by Lichtenstein, or Braque. One source of need for an understanding; one source of
things for understanding to accomplish.
There is a converse point. Suppose a particular picture, or drawing,
represented Cologne Cathedral as with four flying buttresses on its river
side. What follows as to how the picture looked? Nothing. Frege’s point
again: anyone can use any means to represent anything. A representation need not resemble what it represents. Thus it is that a sentence can
represent—if, that is, it bears a suitable understanding. A point impressed
on Wittgenstein by the Tractatus’ collapse: a sentence is “a picture which
hasn’t the slightest similarity with what it represents”.17
Frege’s crucial move comes in the third quarter, partly disguised by his
treating two issues at once:
[A] correspondence can only be complete if the corresponding things
coincide, thus are not different things at all. One can test the genuineness of a banknote by trying to make it coincide stereoscopically

16 Ibid., 59–60.
17 Ludwig Wittgenstein, The Blue and Brown Books (Oxford: Blackwell, 1958), 37.

138

Charles Travis
with a real one. But the attempt to make a gold piece coincide stereoscopically with a twenty mark note would be ridiculous. To make
a Vorstellung coincide with a thing would only be possible if that
thing were also a Vorstellung. And then if the first corresponded
completely to the second, they would collapse into one.18

The point about Vorstellungen is crucial to a case for the publicity of
thoughts. But such is not our present business. Prescinding from it, what
is left? There is this idea: complete correspondence is identity. Thus, where
a Bild represented-as (thus might be true), the correspondence between
it and its Abgebildete (depicted), no matter how we conceived the latter,
could not be complete. (Of course not: It is intrinsic to a way to represent
things that, if this is a way they are, things need not be just as they are for
such to be so. (The intrinsic generality of thought.))
How, then, would a Bild and its Abgebildete correspond if the Bild
were true? Here there is a useful comparison with the case of stereoscopic coincidence, e.g., that between a putative 20-mark note and a certain one—a coincidence there could not be between a bank note and a
gold piece. Where the comparison is intelligible, there is something it
would be for there to be complete stereoscopic coincidence. Such coincidence would not be identity (or stereoscopes would lose their point. To
grasp what it would be, one thus needs to understand what the relevant
notion of stereoscopic identity is to be. There is something such identity
is to be understood to be. On such understanding, two banknotes are to
be counted as identical (the same) if, while distinct, they share relevant
properties. If one is torn or stained, such does not count. There are certain visible features by which a banknote (of relevant sort) ought to be
identifiable as genuine. Those features, whatever they are exactly, must
be shared by the exemplary exemplar (the real thing) and the merely
putative instance. What is fixed here, with the help of stereoscopy, is a
particular proprietary sense of (being) the same.
Here we see the burden a correspondence theory of truth must bear. If
truth is correspondence, such must be so on the right proprietary understanding of corresponding. In effect, the theory must fix a particular
understanding of the same on which a picture and its depicted are the
same just in case that depiction is true. For which it must first, of course,
fix what a ‘depicted’ is to be: not just that on which truth outright always
turns, but rather such that, for each depiction (each true or false way to
represent things) there is its proprietary depicted. The problem now is
to see in what relevant ways a truth-bearer and what it depicts as suchand-such could be alike. For the note in Frege’s pocket and the one fresh
from the mint to match stereoscopically—on the relevant understanding

18 Frege, “Der Gedanke”, Beiträge zur Philosophie des deutschen Idealismus, 59.

Affair

139

of matching—is for them to share certain specifiable visible patterns (e.g.,
lines per centimeter in a certain region). By what sorts of properties of
a Bild and its Abgebildete could the two match, or fit each other? And
what sort of such match would there be just in case there were truth?
This last question is so far neutral as to what properties a truth-bearer
might have. So far, a truth-bearer might be sentence-like, or portrait like,
admitting of such things as visual features. Or it might be thought-like. A
dilemma; on either branch of which a correspondence theory is doomed.
A correspondence theory must relate representer to represented. It must
do so in such a way that there is truth just where the right relation holds
between these two. Suppose (first horn) that we conceive the representer,
the depiction/Bild, on the model of a sunday-painter Bild. In the Bild, then
the depiction of a cathedral—let it be the Stephandom—has two spireshapes and six flying-buttress shapes suitable placed around a cathedral
roof shape, and so on. And there is the Stephansdom, with its two spires
and six flying buttresses. There is certainly some kind of correspondence
here. But now Frege’s point: A shape on a canvas can only represent if it is
to be understood/taken in a certain sufficiently determinate way. And its
converse: And then how it represents things as being depends on how it
is thus to be understood: Any shape can be used to represent anything. So
the correspondence we have so far identified cannot be one which would
hold just where there was truth (where things were as represented).
The relation we are looking for cannot thus be found in topology or
geometry, or so on. No common structure found there can yield the theory we seek. What other structure, then, might a representation have?
Well, a representation is structured, or anyway structurable, in a certain
way (or certain ways). So perhaps the sort of structure we are after is
to be found there. In brief, what we are looking for is the sort of structure which might structure a thought. Reasonably enough since, as Frege
points out, a thought is just that by which truth can come into question
at all. But the problem with this plan is that the sorts of structures to
which a thought is liable are those in terms of features which a thought
has, more specifically, by which one thought may be distinguished from
another. But since a thought is just that by which truth can come into
question, such features are only definable in terms of truth. If they were
not, they would not be ones such that two thoughts which differed, one
in having, one not, some such feature would thereby differ in how things
would be if they were true.
The point here can be filled in as follows. A thought just is a determinate way to make truth turn on how things are. Or, if you like, just that
which makes truth turn in this way. To structure a thought is to decompose it into elements, each of which is a partial doing of what the whole
thought does—in the singular case, say, some singly designative element
and some proper predicative one. But what it would be for a given such
element to be the one it is is to be defined—is only definable—in terms of

140

Charles Travis

the notion to be true. To be a given such element is to do just that part of
what the whole thought does that that element does; to make truth turn
in just that part on how things are.
Trivially, a thought is true just in case things are as per that thought.
Frege’s case against ‘defining’ truth does not exclude such truisms. But
then such truisms do not pretend to be an account of what being true is,
or anyway a non-circular one. They do not meet the first of those two
demands a correspondence theory aspires to satisfy. Drop this demand
and it is easy to say when a thought would be true. But the problem for
the sort of theory, or ‘definition’ of truth against which Frege’s case is
directed is that to be able to say this of any given thought one would first
need to identify it, pleonastically, by identifying when it would be true.
There is a crucial moral to this story which it is now time to state. The
representation of things as some given way—a representer, where this is
conceived as just that by which things are represented as they are (that
is, a Gedanke)—does not, and cannot, share any (identifying) features
with that which it depicts as being such-and-such. Or at least not in the
case of truth outright, and not in any case where the depicted objects
are environmental. So truth cannot consist in the sharing of any such
features. Notably, a thought and what it represents as being such-andsuch are not structurable in terms of any one set of features, or features
of any one sort. So a representer and what it represents as such-and-such,
a Bild and its Abgebildete, are not eligible to share a structure. So truth
cannot consist in any such sharing. Sid stands before the Stephansdom,
chilidog in hand. Here is a scene structurable in the way such scenes
are. Sid is 20 meters from the main entrance, at an angle of 67°, and so
on. The thought that Sid stands before the Stephansdom is structurable,
for example, into a two-place predicate predicating standing before of a
pair <A, B>, and an element making truth turn on how the pair <Sid, the
Stephansdom> is, for example, two very different kettles of fish.
In the fourth quarter, Frege generalizes the point from correspondence
theories of truth to, as he puts it, any attempt to define it. Any definition,
he thinks, will founder on an inevitable, and vicious regress:
Could one not maintain that there is truth just where there is correspondence in a certain respect? But in which? What would we then
need to do to decide whether something was true? We would need
to investigate whether it was true that—say, a Vorstellung and something actual—corresponded in the indicated respect. And with this
we would again face a question of the same sort, and the game could
begin anew. Any attempt to define truth as correspondence thus
comes to grief. But any other attempt to define truth would also thus
fail. For in a definition one would indicate a certain identifying mark.
And when it came to applying this to a particular case, it would then

Affair

141

always come to the question whether it we truth that this mark was
encountered. Thus one would go around in circles.19
In this generalizing of the result, we also find a new form of argument
to make the case. This argument turns on a certain principle. Frege states
the idea in a number of places, for example,
One can certainly say, ‘the thought that 5 is a prime number is true’.
But if one looks closer, he notes that really nothing more is said
thereby than in the simple sentence ’5 is a prime number’.20
[T]he word ‘true’ has a sense which contributes nothing to the
sense of whole senses in which it occurs as a predicate.21
This last states the idea in general. It can be put as follows: truth is an
identity under predication. (The thought that P and the thought that it is
true that P are one.) Call this principle IUP. Suppose now that being true
were equivalent to, or the very same thing as, being F, for some way, F,
for a thought to be. Then being F would also need to be an identity under
predication. Suppose that being true were analyzable non-circularly as
being F. Then what it would be for something to be F is independent of
what it would be for something to be true. Where there is, and where
there would be, a case of something being F is fixed, definable, without
appeal to the notion of being true. Then, the point is, being F cannot be
an identity under predication. For establishing that, say, the thought that
Sid is in is cups is F would now be a different matter from establishing
whether Sid is in his cups, whereas by IUP establishing that the thought
that Sid is in his cups is true is not a different matter than establishing
whether Sid is in his cups, since there is just one thing to be established.
So if we insisted that being true is being F (that being F is just what being
true is), we would violate IUP.
It is crucial here that being F must be such that an account of being
true on which being true was being F would, if correct, satisfy the analytical ambitions of a correspondence theory as per the previous section:
for being F to be what being true is would be for the notion being true
to reduce to something which does not itself presuppose this; to a way
there would anyway be for an item to be whether or not there was such
a phenomenon as representing-as. For suppose we do not insist on this.
Now consider the idea that a thought is true just in case it is such that
things are as it represents them. What, now, of the property a thing has

19 Ibid., 60.
20 Frege, “Über Sinn und Bedeutung”, 34.
21 Frege, “Meine Grundlegenden Logischen Einsichten”, 272.

142

Charles Travis

just when things are as it represents them? There is no reason not to
regard this as an identity under predication. It has as good a claim to this
as does to be true. But then taking this to be what being true is does not
land one in a regress, since IUP applies here as above.
The case here does, of course, rest on IUP (a sort of idea of finite
descent). I think both that, when properly formulated, it is correct, and
that Frege makes good and proper use of that idea (in insisting that the
word ‘true’ only gestures at what really belongs to assertive force, and
thus cannot correspond to a thought-element).22 But, as just shown, a
proper formulation does not rule out what Frege promises at the outset in “Der Gedanke” in the remark, “the meaning of the word ‘true’ is
unfolded in the laws of truth”.23 It is hardly the case that there are no
laws of truth. However, a further exploration of IUP and its credentials is
not part of present business. So I rest with the autobiographical.

8. The Tractatus
I turn now to the Tractatus account of representing-as. This begins with
an account of what there is to represent as something:
(1)
The world is everything that is the case.
(1.1) The world is the totality of facts, not of things.
(1.11) The world is determined by the facts, and by the fact that these are
all the facts.24
Facts rather than objects are put in lead position here. Such could be seen
as a material mode correlate of Frege’s insistence that whole thoughts
come first (or immediately after being true itself). Which could be the
beginning of a good point. But it is not.
The next step is to explicate representing-as. This begins as follows:
(2.1)

We make for ourselves pictures of facts.
...

(2.12) A picture is a model of reality (Wirklichkeit).
(2.13) The elements of the picture correspond in it to [given] objects.
(2.131) The elements of the picture stand in for (‘vertreten’) these objects.25
A picture depicts Wirklichkeit as being some given way. Such a picture
is made up of (given) elements structured in a given way. It is uniquely

22 Vide ibid., 272.
23 Frege, “Der Gedanke”, Beiträge zur Philosophie des deutschen Idealismus, 59.
24 Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. D. F. Pears and B. F.
McGuiness (London: Routledge and Kegan Paul, 1922), 5.
25 TLP, 9–10.

Affair

143

decomposable. The picture depicts these elements as structured in a
given way. It is a picture of them as so structured. Part of its depicting
as it does is fixing a relation, in fact a function, between these elements
and discernible elements in what it depicts. It maps each of its elements into some distinguishable element of the depicted. The picture
is a model of what it depicts. In the model, its elements stand in for
the elements of the depicted they map into. Let us call the function the
picture thus determines a V-function (‘V’ for Vertretung). For it to be
the picture it is, I will suppose, is (inter alia) for it to determine the
V-function it does.
The idea of a Vertretung-function gives us means for associating with a
given picture, as its depicted, not just Wirklichkeit—though every picture
(of present sort) does depict Wirklichkeit—but something specific to that
Bild. This could be the range of its V-function—some set of objects—
related to each other (structured, organised) as they are in fact (a condition of, and in, the world). The idea of a proprietary depicted—for each
depiction, its unique Abgebildete—promises a possible way of counting
Abgebildete. Suppose a depiction depicts Porto’s Sé. What are the elements of the Sé? What objects make it up? There is, as noted previously,
no answer to this question. It is ill-formed. For the notion object does not
itself give us any way of counting same. But suppose a depiction fixes a
given V-function. For a given function, a given range. So we can take it
as intrinsic to its depicting (to its being the depiction it is) that what it
depicts is essentially composed of the items in that range. That item (as
opposed to the Sé) could not be carved up differently. Different sets of
elements, on this way of thinking, would eo ipso identify different (if any)
depicteds (Abgebildete). Such makes an Abgebildete in our present sense
nothing like what is depicted by the Sunday painter, e.g, the Sé.
With this conception of an Abgebildete comes an obvious way for a
depiction to be correct or incorrect: the depiction depicts the elements
in its depicted as structured, related to each other, in such-and-such
way. They may or may not so be. The picture is correct (true) just where
they are, incorrect (false) just where they are not. Which brings us to
the Tractatus’ own account of true and false, or correct and incorrect,
representing:
(2.14) The picture consists in the elements being related to each other
in a determinate manner and way.
...
(2.15) That the elements of the picture relate to each other in a given
way presents it as so that things so relate to each other . . .
...
(2.21) The picture agrees with reality or not.; it is correct or incorrect,
true or false.
...

144

Charles Travis

(2.221) Its sense is what the picture represents.
(2.222) Its truth, or falsity, consists in the agreement or disagreement of
its sense with reality.
(2.223) In order to recognize whether the picture is true or false, we must
compare it with reality.26
The sense of a picture would then consist in the picture’s depicting the
image of its elements under its V-function as relating to one another as
its elements themselves relate in the picture to each other. Agreement
with reality would then consist in that image (the depicted) so relating to
one another. If a picture’s sense is what it represents (how it represents
things), there remains the question what a picture does represent on this
account. Is it to be compared with a sentence, say, which may represent,
speak of, (there being) pheasant on a (contextually definite) lawn? Or a
thought which is of, say, there now being pheasant on Pia’s lawn? This
last, one might hope, if such pictures are to be true or false. Whether it
could do either of these things remains to be seen.
As the Tractatus acknowledges, on this account of truth and falsehood,
a depiction can be so much as true or false only if there is something
(suitable) in common to it and its proprietary Abgebildete. The Tractatus
puts this thus:
(2.16) In order to be a picture, a fact must have something in common
with what is depicted.
(2.161) There must be something identical in both the picture and the
depicted in order for the one to be able to depict the other at all.27
What, then, must be in common to a depiction and what it depicts (as
thus)? The Tractatus answer:
(2.17) What a picture must have in common with reality in order, in its
own manner, to be able to depict either truly or falsely, is its form
of depicting.
(2.18) What every picture, of whatever form, must have in common
with reality, in order to be able to depict it at all—truly or
falsely—is the logical form, that is, the form of reality.
(2.181) Where the form of depiction is logical form, the picture is called
a logical picture.
(2.182) Every picture is also a logical one. (By contrast, not every picture
is spatial.) . . .
(2.2)
A picture has its logical form in common with the depicted.28
26 Ibid., 10–12.
27 Ibid., 10.
28 Ibid., 11–12.

Affair

145

So reality (the depicted) has a form (one intrinsic to it). And this is logical form. So it is a form of what engages in the business of being true:
a representation. Such a form would cross categories. It would be present in thought, in depicting, and also in what thought is true or false
of, Wirklichkeit. A false picture, of course, would have a form which its
depicted lacked. But this would still be a form of a genre available both
to representers (thoughts) and represented. If there are such forms, there
may indeed be something in common to depiction and depicted, representer and represented. If, but only if. But we have seen already, in the
collapse of correspondence theories, why there cannot be such.
Young Wittgenstein has now set himself up to be gored by the horns
of a bull we have already engaged. As things stand, he still gets to choose
his horn. For he has left it unclear which he has in mind: Whether his
depictions (representers) are to be conceived on the model of a painting
of the Sé, or on the model of a thought. Perhaps he intentionally omitted
distinguishing these two categorically different things. But such is at best
a way to make things seem alright. The damage is the same either way.
Suppose we think of representers, Gedanke-like, as distinguished from
one another solely by features which make for different ways for truth
to turn on how things are. Then they are structurable solely by relations
defined in terms of being true, as a Gedanke is structurable in terms of
elements each in that same business which is its. A representer so conceived is not the sort of thing which might, e.g., come before one’s eyes.
Nor is it some structuring of things which might. The point generalizes
from the visible to any items identifiable as the ones they are independent
of how they make truth depend on what they do. Relations of which such
things are relata are not ones in which thought-elements might stand,
hence not ones in terms of which a representer as now conceived might
be structured.
Conversely, what comes before the eyes is not the sort of thing which
can stand in those relations out of which a Gedanke-like representer
might be structured. What come before the eyes are bits of history, things
being as they are. One might generalize over and beyond them. But they
do not. Sid, Mt. Blanc, are not ways for truth to turn on things; nor what
does so. Particular cases are silent on what another case of any way they
are might be. The point starts with what any thought represents as something: things (as a whole) being as they are. This surely does not represent
anything as anything. Its business is simply to be as it is. A fortiori it is
innocent of any feature which might identify a given thinkable, or given
truth, as the one it is. So, too, for what there is to carve out of it, notably objects—items which may combine with partial ways to represent it
being to form a way for it to be or not outright. Sid and Mt. Blanc are
some simple examples. What a representer represents as such-and-such
cannot, in such representing, stand in those relations in terms of which
such representing is structurable.

146

Charles Travis

Pia drinks Billecart-Salmon (that magical potion concocted since 1818
in Mareuil-sur-Ay). There are thus relations she bears to that potion in
relating to it as she does, among which some Sid bears to a different
potion (lager) in relating to it as he does. Sid’s relating to lager as he does
is part of his way of relating to Billecart-Salmon, a way thus quite different from Pia’s. A thought (thinkable) might represent Pia, Sid, Billecart,
lager, as standing in such relations to one another. They might structure
what it thus depicts. But they do not sructure the representing as such.
Now the other horn. Suppose we conceive a representer as itself identifiable by features independent of what identifies the representing accomplished as the representing that it is. It might, for example, be a sentence,
with its syntactic identity. Or a visible image, with its geometric structure.
Such a representer does have features shareable by what it depicts as
such-and-such. Sub-images in such an image and values of the relevant
V-function may, e.g., share spatial structure. But now Frege’s point. You
can use any arbitrarily occurring object or event to represent anything.
So samenesses definable in terms of shared features of the just-mentioned
sort settle nothing in themselves as to whether there is truth (nor as to
how (if at all) things were represented as being). To settle this, the sentence, or image, must be assigned just those features which would identify some representer of our first kind: a Gedanke, something thinkable.
Nothing so far tells us what those features must identify. Whatever it is,
we are thus back on the first horn, saddled with what cannot share a
structure with what it depicts.
At 2.18, Wittgenstein speaks of what any picture and its depicted must
have in common as “the logical form, that is, the form of reality.” But
there is no such thing as ‘the conceptual structure of reality’—a structuring of generalities under which reality might fall or not. Logical form
cannot be the structure of reality because reality has no such form. It does
not generalise beyond a particular case (itself, perhaps?) to reach something general under which to present this. Conversely, reality, anyway an
environment, might be structured in some objectively fundamental way
for some given purposes, e.g., those of physics. It may not permit certain
would-be ways of generalizing across or beyond it. Such is for physics
to say. But just as the fact of Sid loving lager cannot of itself fix what
another case of someone loving lager might be, so things being as they are
cannot decide how one must generalize beyond things so being to other
ways things might have been.
Thus, the Tractatus picture of representing-as collapses. Not that this is
quite the way Wittgenstein saw its collapse in 1929. The trouble struck
him when he approached what the Tractatus left as an outstanding problem: what an atomic proposition might be. The immediate symptom
emerged in the realization that this ‘problem’ had no solution. The trouble lay in what made this appear to be a problem at all. The mistaken

Affair

147

assumption: that there is such a thing as a class of propositions, ‘the
atomic ones’29; an assumption at work here:
3.25 There is one and only one complete analysis of a proposition
(Satz).30
Such an idea could not have arisen had young Wittgenstein followed
Frege:
I do not believe that for each judgeable content there is just one way
in which it can be decomposed, or that one of the possible ways may
always claim an objective priority.31
For Frege whole thoughts (thinkables) come first; other notions involved
in representing-as are to be defined in terms of these; in terms, that is, of
truth outright. To be a proper predicative thought-element, e.g., is to play
a particular role in a decomposition of such a whole; thus to do what
could only be done at all in the context of that whole. From the thought
that Pia has a red dress on, one may carve a part which makes truth turn
on who has a red dress on. But there is no such thing as making truth
turn on this, fertig. There is so far nothing for truth to do. But decompose
a thought in some one such way, and one is immediately struck with the
possibility of others. Such is what Wittgenstein came to realize in the
wake of 1929 (by, say, 1931). It is what he is on about in the Investigations discussion of logical atomism, e.g., around §§47–48, in stressing
that the notion simple, like the notion same, needs filling out to have an
application. Simplicity an sich, like sameness an sich, is not a feature for
anything to have or lack. With which 3.25 is consigned to the dustbin.
But this merely points at issues to be treated elsewhere.
The deeper root cause of the fall of the Tractatus is inattention to the
second of Frege’s two fundamental distinctions: that between the particular and the general—a neglect inherited from Russell. For if it is the
general—the depiction, and not the depicted—which is to be analyzed, as
per 3.25 or otherwise, it is fairly obvious that there can be no one right
way of carving up a proprietary way of generalizing over particular cases
into parts. Suppose we carve up for things to be such that Sid is pear-shaped
into a part, for things to be such that relevant things are pear-shaped; for
things to be such that Sid is relevant ways. Each of these parts, like the
whole, denizens of Frege’s third realm, their homes representing-as, has
its own proprietary generality. There are those particular cases which are
29 Vide Ibid., §2.061–3, 9.
30 Ibid., 18–19.
31 Frege, 1882 letter to Anton Marty, 118.

148

Charles Travis

ones of an item being pear-shaped, and those which are ones of an item
being Sid, or of Sid being some way or other. The generality of our initial
whole is the resultant (as it were vector sum) of those of these parts. But
if so, it is also, inevitably, the resultant of the generalities of indefinitely
many sets of different parts. Thus simplicity an sich classifies with sameness an sich (a point stressed in recent times by Noam Chomsky among
others). Thus the collapse, not just of that wild goose chase on which the
Tractatus sent Wittgenstein in 1929, but the whole Tractarian picture of
representing-as. The Tractatus thus came to grief for much the same reason as correspondence theories of truth come to grief.

9. Thenceforth
In the years 1929–1931, Wittgenstein wrestled with the failure of his
early efforts—while also following assiduously developments in foundations of mathematics, with a special interest in Hilbert. By 1931, a
changed view had emerged. What happened? Here, in briefest and
sketchiest form, is an answer. In pursuit of that latter interest, Wittgenstein returned to reading Frege (at least Grundgestze vol. 2, specifically
Frege’s attacks on ‘formalist arithmetic’, chez Thomae and Heine). One
thing which emerged from this, I believe, is the Investigations idea of a
language game. Another theme, though, for expansion elsewhere. Russell and early Wittgenstein leaped with insouciance across that categorial
gulf Frege stressed, between the particular and the general, in particular,
between the second and third of Frege’s three realms. In these years of,
inter alia, renewed interest in Frege, Wittgenstein took the lesson not to
do so. He came to see the importance of this distinction.
But, as Wittgenstein saw, it is easy to mystify Frege’s third realm and
that distinction it represents. It is easy, e.g., to make it seem puzzling (as
it appears to have seemed to Russell) that a way for things to be—what
belongs in the third realm—can reach all the way to what belongs in the
second, things being as they in fact are. That independence of truth from
what we think true which Frege stresses can also be read as assigning such
things as thoughts, or concepts, mystical powers to reach all the way to
particular cases not yet, and perhaps never, part of history. How business
is conducted in the third realm can come to seem even more distanced
from how we think of things than it in fact is. Things are not helped here
by Frege’s own occasional blind spots, notably what shows up in the idea
(pretty clearly insisted on in Grundgestze vol. 2) that Begriffsschrift is a
language—an idea Frege felt pushed to by his own (mis)perception as to
how his dispute with Kant about arithmetic would have to play out.
So Wittgenstein’s turn towards Frege came with a rather heavily underlined asterisk or two. Frege, but with an improved view of the place of
the logical in the psychological, and of the relation of logic to thought,
and to language; thus, too, an improved view of logic’s province and its

Affair

149

inexorability and universality such as that may be. Frege at least showed
us where there are questions of the logical’s place in the psychological,
and something as to how these should be framed. So one could see later
Wittgenstein as, in large part, just a purer version of Frege.
The point which put paid to the Tractatus begins in Frege as targeting correspondence theories of truth, but becomes, in his hands, deeply,
and generally, anti-reductionist. It puts paid to any idea of reducing the
business of being true, thus equally that of reducing representing as—
representational or conceptual content—to anything else. It applies
equally to would-be reductions of any of this to what might be perceived,
more generally, to anything environmental, as to reductions to something
psychological, in Frege’s phrase, what “belongs to the contents of someone’s consciousness” (or which might be proprietary to any one). The idea
that being true, so the conceptual, does not reduce to the environmental
has proven, for many philosophers, the hardest to swallow. But that point
can also be put this way: Naturalism is not the price of objectivity. Such
is just what one recognizes in recognizing Frege’s ‘third realm’. That such
must be recognized is what is shown by his argument against the ‘definability’ of truth (that is, against any account of truth on which being true
fails to be an identity under predication).
All this, though, is but the briefest of sketches, likely too brief to be of
any use. It may also seem dogmatic, and to many, eccentric. As I stand
here I can no other. What I want to stress most, though, is that later,
Wittgenstein was concerned with the same philosophical problems as the
Tractatus was, as philosophers have long been, and as many of us are
today. These problems concern the relation of thought, or representation,
to what there is to be thought about; to the world. He was, throughout,
a philosopher, not a post-philosopher, or some other alternative to philosophy. Or so I mean to suggest.

References
Frege, Gottlob, Begriffsschrift, eine der arithmetischen nachgebildete, Formelsprache des reinen Denkens, Halle a. S.: Louis Nebert, 1879.
———, “Ueber Sinn und Bedeutung”, Zeitschrift für Philosophie und philosophische Kritik, NF v. 100, 1892, 25–50.
———, Grundgesetze der Arithmetik, Jena: Verlag Hermann Pohle, Band I/II,
1893/1903.
———, “Der Gedanke”, Beiträge zur Philosophie des deutschen Idealismus, v. 1,
n. 2, 1918, 58–77.
———, “1882: letter to Anton Marty”, Gottlob Frege’s Briefwechsel mit D. Hilbert, E Husserl, B. Russel, sowie ausgwählte Einzelbrefe Freges, G. Gabriel,
F. Kambartel, and C. Thiel, eds., Hamburg: Felix Meiner, 1980, pp. 117–119.
———, “1876/1882: ‘17 Kernsätze zur Logik’ ”, Nachgelassene Schriften, 2nd
ed., H. Hermes, F. Kambartel, and F. Kaulbach, eds., Hamburg: Feliz Meiner,
1983, 189–190.

150

Charles Travis

———, “1915: ‘Meine grundlegenden logischen Einsichten’ ”, Nachgelassene
Schriften, 271–272.
Russell, Bertrand, “1903: Letter to Frege, 24.5.1903”, Gottlob Freges Briefwechsel mit D. Hilbert, E Husserl, B. Russel, sowie ausgwählte Einzelbrefe Freges,
G. Gabriel, F. Kambartel, and C. Thiel, eds., Hamburg: Felix Meiner, 1980,
87–90.
———, “1904: Letter to Frege, 12.12.1904”, Gottlob Freges Briefwechsel, 96–99.
Wittgenstein, Ludwig, Tractatus Logico-Philosophicus. Translated by D. F. Pears
and B. F. McGuiness, London: Routledge & Kegan Paul, 1922/1967.
———, The Blue and Brown Books, Oxford: Blackwell, 1958.

7

Later Wittgenstein and the
Genealogy of Mathematical
Necessity
Sorin Bangu

It is like finding the best place to build a road across the moors. We may
first send people across, and see which is the most natural way for them
to go, and then build the road that way.
(LFM, 95)1

Introduction
According to P.M.S. Hacker, Wittgenstein’s central preoccupation was
not to elucidate why the propositions of mathematics are necessary, but
“the prior question”
[W]hat is it for a proposition to be a ‘necessary proposition’, i.e. to be
a proposition of mathematics?2
This point is of utmost importance for setting up the right direction in
understanding Wittgenstein’s approach to mathematical necessity. Indeed,
1 Throughout the paper, I reference Wittgenstein’s works in the usual way: LFM for Wittgenstein’s Lectures on the Foundations of Mathematics, Cambridge 1939, ed. C. Diamond (Chicago: Univ. of Chicago Press, 1976), PI for Philosophical Investigations, 3rd
ed., trans. G. E. M. Anscombe (Oxford: Blackwell, 1968), PG for Philosophical Grammar, ed. R. Rhees, trans. A. J. P. Kenny (Oxford: Blackwell, 1974), RFM for Remarks
on the Foundations of Mathematics, ed. G. H. von Wright, R. Rhees, and G. E. M.
Anscombe, rev. edn. (Oxford: Blackwell, 1978), and LWWC for Ludwig Wittgenstein
and the Vienna Circle, Notes recorded by F. Waismann, ed. B. F. McGuinness (Oxford:
Blackwell, 1967/1979).
2 P. M. S. Hacker and Gordon Baker, Wittgenstein: Rules, Grammar and Necessity (Oxford:
Blackwell, 2009), 243. Hacker insists that for Wittgenstein, necessary propositions form
a heterogeneous group. In addition to mathematical propositions, there are logical propositions and metaphysically necessary propositions (modus ponens and ‘every event has
a cause’, respectively). Although I see more similarities than differences here, in this
paper I focus only on mathematics. A recent insightful take on logic is Penelope Maddy,
The Logical Must: Wittgenstein on Logic (New York: Oxford Univ. Press, 2014). H. J.
Glock discusses both Wittgensteins’ (early and late) take on necessity, in general terms,
in Glock, “Necessity and Normativity,” in The Cambridge Companion to Wittgenstein,
eds. H. Sluga and D. Stern (Cambridge: Cambridge Univ. Press, 1996), 198–225.

152

Sorin Bangu

an interpretive reconstruction of Wittgenstein’s views should primarily
aim to answer this question—and yet, despite the attention Wittgenstein’s philosophy of mathematics received among commentators, such
an answer is still incompletely articulated.3 Once this is done, the other
task is to show how this answer enables us to deal with the traditional
skeptical questions raised in relation to mathematics (e.g., what if we are
wrong that 25 × 25 = 625? Is this revisable?)4 Here I hope to contribute
to accomplishing the first task; at the end, however, I will address in passing the second one too.
I proceed as follows. Section 1 summarizes Hacker’s discussion, which
covers enormous ground, and in considerable depth, but also leaves several issues unresolved. Section 2 proposes some solutions, and elaborates
in detail what I take (Wittgenstein’s) answer to the ‘prior’ question to be.
Section 3 concludes the paper, and suggests, very briefly, a way to deal
with skepticism.

1. Necessity via Normativity
For Hacker, in Wittgenstein’s view the equations of mathematics are to
be “compared with expressions of rules. They stand to their application
as a rule stands to its application. They are not descriptive but normative.
They do not state facts, but determine ways of describing how things, in
fact, are”.5
Such ‘necessary propositions’ are not descriptions—neither of the
world or its putative metaphysical structure, nor of a super-empirical
reality. They are norms of representation or, more prosaically, inference tickets.6
The license-to-infer idea is explicated further:
In particular, equations are substitution-rules, which are characteristically applied to the propositions of ordinary language (LWWC
156; PG 347). They license transformations of empirical statements, and they exclude certain inferences as invalid. For example,
‘12 × 12 = 144’ legitimates transforming the statement ‘There were

3 This, I am afraid, applies to Hacker’s own comprehensive presentation as well—
although, in all fairness, he describes it as only “a prolegomenon” to the proper study of
Wittgenstein’s philosophy of mathematics and logic. See Hacker and Baker, Wittgenstein,
243.
4 I will not take up examples more advanced than this, so perhaps the title of the paper
should mention arithmetical necessity.
5 Ibid., 262.
6 Ibid.

Necessity

153

12 rows each with 12 soldiers on parade’ into ‘There were 144 soldiers on parade’ without counting them up afresh.7
In a similar vein,
Geometry [. . .] consists neither of descriptions of ideal geometrical
objects (Platonism) nor of the description of the a priori structure
of our spatial intuition (Kant); nor does it consist of uninterpreted
calculi the primitives of which are implicitly defined by the axioms
(Hilbert). Rather, a geometry for space consists of complex interwoven norms of representation for the description of spatial objects
and their spatial properties and relations. Alternative geometries for
space are not alternative theories of space, but alternative grammars
of space.8
I quote Hacker at length because I believe that these points are in essence
correct.9 Instead of advancing a different and ‘better’ philosophical theory of necessity, Wittgenstein’s strategy was rather to suggest that we
see the mathematical equations in a different light (not as declarative/
descriptive, but as imperative/normative), a perspective shift which not
only weakens the temptation of Platonism, but also, as we’ll see, makes
the skeptical worry vanish.10 (Historically, Platonism is the most prominent philosophical theory of necessity; for Platonists, ‘12 × 12 = 144’
and ‘the angles of a triangle add up to 180°’ are descriptions of what
the mind’s eye ‘sees’ in the platonic heaven: atemporal, acausal abstract
objects, and their immutable properties.)11

7
8
9
10

Ibid., 269.
Ibid., 251.
But see footnote 56.
The shift is the same as the ‘possibility’ mentioned in PI, §244: regard sensation talk
not as descriptive but as expressive. If so, the temptation to posit an essentially private
reality also diminishes.
11 Wittgenstein’s dismissal of Platonism is obvious enough, and thoroughly discussed, so
I will not dwell on it here. See Steven Gerrard, “A Philosophy of Mathematics Between
Two Camps,” in Sluga and Stern’s Cambridge Companion, 171–197; David Finkelstein,
“Wittgenstein on Rules and Platonism,” in The New Wittgenstein, eds. Alice Crary and
Rupert Reed (New York: Routledge, 2000), 53–73; Robert Fogelin, Taking Wittgenstein
at His Word (Princeton: Princeton Univ. Press, 2009), and my “Later Wittgenstein’s Philosophy of Mathematics,” entry in the Internet Encyclopedia of Philosophy, at http://
www.iep.utm.edu/wittmath/ (especially section 3). Recently, Alvin Plantinga’s writings
attempted a revival of a variant of Platonism about necessity, which Michael Loux presents and criticizes his Metaphysics: A Contemporary Introduction (London: Routledge,
2006). There are non-Platonist theories too, such as David Lewis’s influential modal
(‘possible world’) realism, a conception riddled with such major and obvious difficulties
that it sounds almost like a non-starter. See Lewis, On the Plurality of Worlds (Oxford:
Blackwell, 1986). According to Stalnaker, the theory stumbles upon basic questions

154

Sorin Bangu

Below, I offer additional textual support for the ‘equations as rules’
idea, by citing one relevant paragraph from LFM (not mentioned by
Hacker); in it, the thought is introduced explicitly, albeit in an oblique
and cautious manner. Wittgenstein advances it against the background of
a serious worry,12 that this central idea may not come across the way it is
meant to—as a mere proposal (suggestion, possibility) with therapeutic
effect, but rather as a yet another philosophical theory.13 The passage reassures his audience that he is not betraying one of the key-principles of
his whole investigation, which was to avoid these ‘opinions’14 and to state
only “what everyone admits”:15
The only thing which I have a right to want to make you say is, ‘Let’s
investigate whether so-and-so is the case’. For instance, I have no
right to want you to say that mathematical propositions are rules
of grammar. I only have the right to say to you, ‘Investigate whether
mathematical propositions are not rules of expression, paradigms—
propositions dependent on experience but made independent of it.
Ask whether mathematical propositions are not made paradigms or
objects of comparison in this way’.16
There is another important insight in this paragraph (the reference to
‘experience’), and we will return to it soon; for the moment, let’s focus
on two immediate problems arising from understanding necessity as
normativity. The first one is addressed by Hacker but, I believe, not adequately solved; the other is, as far as I can see, not even considered. The
solutions to the two problems, sketched in sect. 2, are interconnected.
The first difficulty is that mathematical equations (e.g., 2 + 2 = 4),
when taken as rules (“Accept only 4 as a result of adding 2 and 2!”), just
are not the kind of things that can be true or false. Understood as orders
and imperatives, rules do not have truth-values. On the other hand, as
Hacker correctly remarks,17 Wittgenstein often talks about mathematical

12
13

14
15
16
17

such as “If necessity is true in all possible worlds, what explains why there are just the
possible worlds that there are?” See Robert Stalnaker, “Modality and Possible Worlds,”
in A Companion to Metaphysics, eds. J. Kim, E. Sosa, and G. Rosenkrantz (Oxford:
Blackwell, 2009), 54.
Again, not addressed in Hacker and Baker, Wittgenstein.
Moreover, if so, one may ask, what kind of evidence does he have for it? Is this his
opinion? Or his discovery of a hidden fact about mathematics, which we have so far
missed? Even if it is true, what is ‘hidden’ “is of no interest to us”, as he says in PI, §126.
LFM, 55; 103.
PI, §599.
Wittgenstein, LFM, 55.
Hacker and Baker, Wittgenstein, 271.

Necessity

155

equations as being true or false (see LFM, 107, among other places).
Hacker realizes that this is an issue,18 and tackles it as follows:
This objection can readily be countered. Although we do not ascribe
truth-values to a rule, we have no qualms at all about ascribing truth
or falsehood to the statement of a rule—as when we say that it is true
that the chess king moves one square at a time [. . .] (This is comparable to our not ascribing truth or falsity to a fact, but only to a
purported statement of a fact.) So there is nothing anomalous about
holding that arithmetical equations are rules—that ‘2 + 2 = 4’ is the
statement or expression of a rule—and that it is true that 2 + 2 = 4.19
I confess I do not find this convincing. Moreover, I shall argue, Wittgenstein
actually does not rely on the distinction rule v. statement-of-a-rule, as he
grounds his account in a different distinction, which does put him in the
position to be able to call arithmetical sentences rules, while also assigning them truth-values. I will sketch this account in the next section, after I
explain why the solution above does not work. (I will also sketch the second
difficulty). The account in section 2 should offer solutions to both problems.
The first problem first. We do say that the string of words “the king
moves one square at a time” has a truth-value: it is true—in chess. But,
when do we say this? Imagine a child watching a game of chess for the
first time. She is told which piece is the ‘king’, she follows it for a while on
the board, and she sees the two people sitting in front of each other moving it one square at a time. Then she reports this regularity (call it R) to a
friend next to her, also not knowledgeable about chess: “The king moves
one square at a time”. Importantly, it is precisely because she utters R
in these circumstances that it can be characterized as ‘true’. But, has she
stated a rule here? Of course not; she has no idea about that (yet). She
just accurately described what she saw happening regularly on the board;
so, R can be called ‘true’ precisely when it is not the statement of a rule.
On the other hand, once the child realizes what is actually going on
between the two people she has been watching (i.e., that they are players,
trying to win a game governed by strict rules), she may want to explain
all this to her friend. She thus may utter sentence R again. Now, by uttering R, she does state a rule. Were a player to move the king differently, she
would feel entitled to react—e.g., to signal this to the other player, who

18 As does Shanker, also without presenting a solution. See Stuart Shanker, Wittgenstein
and the Turning Point in the Philosophy of Mathematics (London: Routledge, 2006),
283. Surprisingly, there is no discussion of this in the otherwise enlightening paper by
Severin Schroeder, “Mathematical Propositions as Rules of Grammar,” Grazer Philosophische Studien 89 (2014): 21–36.
19 Hacker and Baker, Wittgenstein, 277–278.

156

Sorin Bangu

maybe has not noticed it. Now she changed her attitude toward this regularity (more about this later). In these new circumstances, R is revealed
to be an abbreviated form of the more explicitly formulated imperative
R*: “Move the king only one square at a time!” Now R* (and R) does
not have a truth-value, and the first problem persists: when one utters,
or uses R to state a rule, it does not (since it cannot) have a truth-value.
Before we discuss the second problem, let us note the dual function of
sentences like R, which leads to a distinction of major relevance later on:
the same string of words, declarative on the surface, can be used to make
both descriptions and prescriptions—to express both regularities and
rules.20 Now, when it comes to understanding mathematical equations
as rules, the other immediate question concerns the nature of these rules:
Why these rules, and not others? What makes them special (if anything)?
Can they be changed? How did we find them in the first place? And so
on. These questions mirror those asked in the investigation of the nature
and legitimacy of the judicial laws: Are the rules of conduct mere conventions among people, or are they somewhat ‘special’—e.g., discovered by
(revealed to) us, as residing in an unworldly, platonic-like, divine realm?
This series of inquires is particularly pressing since it indicates concerns about the viability of the whole necessity-as-normativity idea: is it
the case that Wittgenstein doubts the objectivity of mathematics?21 Conversely, if objectivity is upheld, aren’t we just back into platonism? Or, if
the rules are up to us, then it seems legitimate to be worried about Wittgenstein’s sympathy for conventionalism.22

2. Regularities and Rules: A Genealogical Account
The account I shall sketch below, in the hope of capturing some of
Wittgenstein’s central ideas, should solve the two problems above, by
answering the question posed at the outset: What is it for a proposition to be a proposition of mathematics? The account has a genealogical character: what elucidates the nature of such propositions is their
origin—and, because of this origin, they can be put to a certain use. If
convincing, the account should also be able to give more than a hint as
20 As we will see in a moment, this is the relevant distinction I hinted at above, not the
one between rules and statements of rules. I found Mark Steiner’s presentation of the
rule-regularity distinction most illuminating; moreover, he argues that Wittgenstein’s
discovery of it (at the end of 1930s) amounted to a ‘silent revolution’ in his thinking. See
Mark Steiner, “Empirical Regularities in Wittgenstein’s Philosophy of Mathematics,”
Philosophia Mathematica 17.1 (2009): 1–34. I discuss this paper, as well as Fogelin’s
similar ideas, in my “Later Wittgenstein’s Philosophy of Mathematics” and “Wynn’s
Experiments and Later Wittgenstein’s Philosophy of Mathematics.”
21 This is a worry seemingly supported by lines such as “The Mathematician Is an Inventor, Not a Discoverer,” in RFM, I-168.
22 In a famous 1959 article, Michael Dummett calls him a “full-blooded conventionalist”.

Necessity

157

to how Wittgenstein deflects the skeptical doubt about (the necessity of)
mathematical truths.
The account consists of two components, each of them deserving to
be called ‘naturalist’. (I do not have a definition of naturalism; my hope
is that in the end it will be unquestionable that the label fits.) The first
component has to do with certain methodological aspects Wittgenstein
deploys when reflecting on mathematics (unsurprisingly, it is the same
methodology he uses in Philosophical Investigations, and everywhere else
in his later philosophy.) The second element concentrates on the abovementioned relation between the mathematical statements and ‘experience’.
The term of art Wittgenstein introduces here is quite suggestive; he talks
about ‘hardening’ of empirical propositions into rules in RFM, VI-22. I
will examine this ‘hardening’ process, and of some help here will be a legal
analogy, suggested by Wittgenstein himself in RFM I-116, between (the
rules of) mathematics and (the laws of) jurisprudence. The key-idea is that
“custom is a source of law”, a locution common in legal theory.23 This
passage from Raz is illustrative for my train of thought in sect. 2.2 and 3:
Of course, the word ‘law’ designates, among other things, rules or
norms, and the concept of a rule probably emerged from a concept of
a law which did not separate natural law from customary practices,
nor either of them from a normative law.
(My emphasis)24
2.1 Methodology
A central (but neglected) aspect of Wittgenstein’s philosophy of mathematics is his strategy to treat the concepts and notions typically considered mathematical technicalities from the perspective of their role in
ordinary, everyday language:
An important problem arises from the subject itself: How can I—or
anyone who is not a mathematician—talk about this? What right
has a philosopher to talk about mathematics? [. . .] I can as a philosopher talk about mathematics because I will only deal with puzzles which arise from the words of our ordinary everyday language,
such as ‘proof’, ‘number’, ‘series’, ‘order’, etc. Knowing our everyday
language—this is one reason why I can talk about them.25

23 See David J. Bederman, Custom as a Source of Law (Cambridge: Cambridge Univ.
Press, 2010).
24 Joseph Raz, “Can There Be a Theory of Law?” in The Blackwell Guide to the Philosophy of Law and Legal Theory, eds. M. Golding and W. Edmundson (Oxford: Blackwell,
2005), 336.
25 Wittgenstein, LFM, 13–14.

158

Sorin Bangu

This perspective legitimizes the authority of his entire enterprise (once
again). But it also makes possible to apply the method he introduced in
the well-known family resemblance passages in the Investigations §65–
71. Thus, when analyzing ordinary language notions, a philosopher’s
concern is to “look and see”, not to ‘think’26—so it is similar to that of a
naturalist painter: not to distort.27 And what Wittgenstein notices when
he looks into the employment of such natural language concepts is a challenge to what has been called ‘intellectualism’: Language users are unable
to identify comprehensive, essentialist definitions of many concepts (and
‘game’ is the famous example here; not accidentally, ‘number’ is another).
This is usually taken as an attack on essentialism, but some commentators suggested that this is only a secondary goal for Wittgenstein here,28
if at all.29 In fact, what he wants is to ask “a question about a human
being”,30 lending itself to an unequivocal, and philosophically illuminating anti-intellectualist answer, that “everyone admits”: when we use the
word ‘game’, are we able to give a definition of it? The answer is ‘no’, the
definition—of ‘game’, ‘number’, etc.—eludes us; despite this, we are typically able to (learn to) apply these concepts correctly.
Then, it is not surprising that both RFM and LFM are pervaded with
the same kind of question, ‘about a human being’. The key-move is, once
again, to “look at what happens”31 when we deal with arithmetic, and to
ask questions about human abilities and practices, answerable by drawing on what is in front of our eyes—as opposed to ‘hard’ mathematical (calculational, technical) questions. During a discussion with Turing,
recounted in the LFM, Wittgenstein emphasized this methodological
point, about the kind of questions he poses:
I did not ask, ‘How many numerals are there?’ This is immensely
important. I asked a question about a human being, namely, ‘How
many numerals did you learn to write down?’32
So, what does Wittgenstein see when he looks at arithmetic from this perspective? What first stands out is how seriously we take it, how central and

26 PI, §66.
27 According to the art historian Harris, “naturalism is the name for a visual style and
active intention: that is, the desire to depict, as accurately as possible, the appearances
of things in the world.” J. Harris, Art History: The Key Concepts (London: Routledge,
2006), 211.
28 See Gordon Baker and P. M. S. Hacker, Wittgenstein: Understanding and Meaning
(Oxford: Blackwell, 2005), 214–215.
29 I argued that it is not, in my “Later Wittgenstein on Essentialism, Family Resemblance
and Philosophical Method”.
30 LFM, 32.
31 RFM, I-6.
32 LFM, 32.

Necessity

159

important it is for us. Arithmetical calculations are not esoteric matters,
technicalities debated among experts; nor are they mere hobbies, unserious pursuits. They are deeply embedded in our life, and ubiquitous.33
For what we call ‘counting’ is an important part of our life’s activities.
Counting and calculating are not—e.g. simply a pastime. Counting
(and that means: counting like this) is a technique that is employed
daily in the most various operations of our lives. And that is why we
learn to count as we do: with endless practice, with merciless exactitude; that is why it is inexorably insisted that we shall all say ‘two’
after ‘one’, ‘three’ after ‘two’ and so on.34
The other salient aspect is the silence, peace, the muted and deep
agreement among people when dealing with mathematical issues. In this
sense—to avoid contradicting the remark above—even disinterest and
neglect: “We acknowledge it [a mathematical proposition] by turning our
back on it”.35 We are also reminded that
We have all of us worked out certain multiplications. And actually
there are no disagreements about the result of a multiplication—so
that we don’t know what to believe because we always have a headache, or all the people get different results. This hasn’t happened; that
is immensely important.36
That there is hardly ever any discrepancy between ways of counting, and when there is we are able to clear it up usually, is of immense
importance.37
Both these evident aspects—seriousness (‘merciless exactitude’) and nonlitigiousness (‘no disagreements’)—will be relevant in the discussion
below.
2.2 Hardening
Going back to our ‘prior’ question, let us address it from the methodological perspective sketched above. To summarize what I will argue next:
According to Wittgenstein, if we look at the propositions of mathematics,
we see no intrinsic feature of them, something about their content, that

33 The world’s oldest writing system we know about, developed in Mesopotamia, derived
from methods of keeping accounts of goods and events; it is telling that the signs
imprinted into clay recorded both letters and numbers.
34 RFM I-4.
35 Ibid., IV-35.
36 LFM, 108. See also PI §240 and RFM III-15, VI-39.
37 LFM, 258.

160

Sorin Bangu

makes them necessary.38 Instead, what qualifies a proposition as a ‘mathematical’ (necessary) one is a certain extrinsic feature, an “attitude”39 we
have toward it (and toward the technique, or procedure, establishing it).
To flesh out a more complete story, we need first to recall the proposal
from LFM, 95 already cited above:
Mathematical propositions are [. . .] propositions dependent on
experience but made independent of it.
This echoes points made in several previous lectures. In Lecture IV, we are
told that a mathematical proposition is:
Independent of experience because nothing which happens will ever
make us call it [25 × 25 = 625] false or give it up.40
Dependent on experience because you wouldn’t use this calculation if things were different. The proof of it is only called a proof
because it gives results which are useful in experience.41
Moreover,
All the calculi in mathematics have been invented to suit experience
and then made independent of experience.42
[25 × 25 = 625] was first introduced because of experience. But
now we have made it independent of experience; it is a rule of expression for talking about our experiences.43
We also need to recall here RFM IV-21, where Wittgenstein explicitly
connects the descriptive (law-like regularities, holding in ‘experience’)
with the normative (rules):
The twofold character of the mathematical proposition—as law and
as rule.
(‘Law’ here has a descriptive sense, while ‘rule’ has a normative meaning.
This brings us back to the duality44 I signaled above, and this notion will
help solve the first problem, the truth-value assignment). On the descriptive

38 This is, I submit, part of what RFM VI-24 intends to convey: “You say you must; but
cannot say what compels you”. See also RFM VI-31: “the difficult thing here is not, to dig
down to the ground; no, it is to recognize the ground that lies before us as the ground”.
39 RFM VII-67.
40 LFM, 41.
41 Ibid., 42.
42 Ibid., 43.
43 Ibid., 44.
44 RFM IV-13 is also relevant: “I should like to be able to describe how it comes about
that mathematics appears to us now as the natural history of the domain of numbers,
now again as a collection of rules”.

Necessity

161

side, there are no special difficulties. We all know that some things happen
in a regular, law-like fashion while others just do not. It is an empirical regularity that pebbles always fall downwards when left unsupported in the
air—and also that they don’t suddenly disintegrate or multiply when laid
down on the ground to be counted. Similarly, it is a psychological regularity that a vast majority of people behave in (roughly) the same way when
experiencing similar circumstances (e.g., when hurt, when surprised), and
even most surely so when they receive the same kind of persistent, exacting, ‘merciless’ training—in counting pebbles, for instance. Here we should
note that ‘training’ is, in many of Wittgenstein’s employments e.g., in PI
§5, a translation of the German word ‘Abrichtung’, used only when one
talks about animals! Some of these regular reactions may have an innate
basis, and many are induced by training-abrichtung; either way, “everyone
admits” that such regularities exist. Now, what about the normative side?
A good way to explain the emergence of normativity is through a little story. Once upon a time there was a kingdom where people knew
how to count up to 1000, but they had not yet invented multiplication.45
(Moreover, these people never cared about concepts such as ‘necessity’
and ‘possibility’; we can even speculate that this describes a period of the
actual human history.) But one day, the Queen laid down 25 batches of
25 pebbles each on the shiny floor of her palace, and kept counting them
carefully every morning:

Figure 7.1 The Queen’s pebbles
45 In LFM, 95, Wittgenstein himself hints at such a thought-experiment: “Suppose we in
this room are inventing arithmetic”.

162

Sorin Bangu

She reached 625 almost always, 626 once, and 624 on a couple of
occasions. She then decided to ask the royal court to do the counting too.
All of them reached 625, even the King; one duke, however, got 619. (The
Queen noticed that he was a bit tipsy when he counted, so she decided
to dismiss this result.) Another (sober) subject got 626, and the Bishop
got 624. She then used her royal prerogatives and ordered everyone (aristocrats, merchants and peasants) to arrange pebbles on the ground and
start counting. An overwhelming majority, of the order of tens of thousands, reported 625, but three people, working independently, found the
result to be exactly 601. This coincidence caught the Queen’s attention—
how can this be?! It turned out that these people understood the order
‘put 25 batches of 25 pebbles each on the ground and count them’ in this
way46 (Figure 7.2):

Figure 7.2 An arrangement of 25 batches of 25 pebbles each, i.e., 24 horizontal
batches and one vertical.

However, when told that their procedure was not the same as the others, they did it again and two of them got 625. One of them still insisted
that what he did was right, and so the result is 601. Yet, in the end, he
gave up when the Queen, too tired to argue, threatened to exile him. After
taking all the reports into account, the Queen and virtually everyone else
came to believe that the result was the ‘peak’ value of 625 pebbles.
46 The setup is inspired by Wittgenstein, RFM I-38, and Fogelin, Taking Wittgenstein, 98.

Necessity

163

The point of this story is to illustrate how certain (arithmeticallysounding) beliefs—expressed as descriptive propositions, e.g., ’25 batches
of 25 pebbles each is 625 pebbles’—may emerge. They have the following
characteristics:
(i) are ‘dependent on experience’47 (verifiable empirically, quickly convergent, stable) and, in the end, peacefully accepted by (virtually)
everyone.
(ii) it is considered important to settle them, and thus the (extremely
few) dissenters are taken seriously and silenced if necessary.
On this royal road to normativity the next thing to note is that beliefs
like these form a set of propositions which serves, in the hands of a more
practically-inclined community (or Queen), as the pool of candidates to
fulfill the role of basic rules needed to implement social uniformity—
once the community finds desirable to ensure it (e.g., to avoid protracted
legal negotiations, to facilitate commerce, etc.). Propositions like the one
above—recall, originally dependent on experience, (almost) universally
accepted as true—are now made independent of experience. They are
made so by the community, most likely gradually and not by a completely transparent and explicit decision.
‘Independence’ means that no empirical facts, or procedures will be
recognized (explicitly, by the community) as capable of challenging these
propositions. Thus, one may say that such an empirical proposition has
become harder to overthrow. But this is not strong enough; in fact, it
has become meaningless to even try to do so—to make a proposition
independent of experience is to change its status, to look at it in a different light: not as a (descriptive) statement, but as a (normative) rule. A
monetary analogy may be useful to understand what happens here. If I
give you a 1$ bill (currently in circulation) and tomorrow you give me
another 1$ bill (in circulation), this is (i) a neutral, monetary transaction,
and (ii) a non-commercial one: we just exchange money, neutrally, and
without selling or buying anything. But, if in a second transaction, tomorrow you give me a 1$ bill issued in 1869 (i.e., not in circulation) for a
current 1$ bill I gave you today, this transaction is not neutral; it is not
even a monetary one. Moreover, there is no definite amount of current
dollars I should give you to make it so. This is not a monetary transaction because I give you money, but you give me an object (an old piece of
paper, that once had the role of money). So, we don’t exchange money for
money, exchange which, if it were one current dollar for another current
47 In the quote above (LFM, 42), Wittgenstein explains dependence on experience in counterfactual terms: “if things were different” then we would have considered a different
proposition. In our story, things would have been “different” if, for instance, we lived
in a world where pebbles would materialize (or disappear) instantly.

164

Sorin Bangu

dollar, it would have been ‘monetary’ (and ‘neutral’—because, if you give
me two current dollars tomorrow, neutrality is lost, since this amounts to
lending and borrowing, at an astonishing interest rate!) This leads to (ii).
Our second transaction above is, after all, a commercial one. In essence, I
buy from you (i.e., give you money for) an object, an old piece of paper;
and, as it happens, it is a pretty bad transaction for you, since a 1$ bill
issued in 1869 would cost a collector today significantly more than one
current dollar! Regardless, we have now entered the commercial realm.
To return to the analogy with the propositions of mathematics, they
are much like the dollar bills issued in 1869. They look like empirical/
descriptive, truth-valuable ones (dollars in use), but are not so. This
old bill per se will not buy you anything today (say, if you put it in a
vending machine). The lack of purchase power stands for the impossibility to refute a mathematical proposition by invoking empirical matters. The question “how many empirical facts do I need to show you in
order to renounce a mathematical proposition?” has no sensible answer,
just as “how many current dollars should I ask someone who wants to
exchange, monetarily and neutrally, a 1$ bill not in use?” does not have
one. (Again, the latter question is not: “How many current dollars should
I ask a collector to pay me for the very old 1$ bill I found in my greatgrand father’s attic?”)
Making rules involves instituting restrictions about what people are
allowed to do, i.e., the techniques and procedures they employ. If an
individual insists on getting recognition for the technique leading from
“25 groups of 25 objects each” to “601 objects”, the community has to
react by taking exclusionary measures against that individual. As Figure
7.2 shows, what he did “can be made out to accord with the rule”, as
PI §201 warns us. He can’t be proved wrong by pointing out that he
counted some pebbles twice, since there are situations when we do this:
we say that the equation x2 – 6x + 9 = 0 has two roots, so we count the
root ‘+3’ twice (LFM, 156); we also talk about the Christian Trinity, etc.
Thus, before instituting the rules the community was inclined to take
such punitive measures as a matter of course (the Queen thought that
some of the reported results should be dismissed.). But after this happened such measures are taken with merciless regularity; the community
has to take them, as it sees itself entitled to do so, since everyone was
warned in advance, and trained in arithmetic, abrichtung-style. While
social exclusion-type measures are taken both before and after setting the
rules, they have a different nature in the two cases: de facto and de jure.
RFM VI-22, perhaps the most important paragraph in all of Wittgenstein’s later philosophy of mathematics, gives us the ‘hardening’ notion:
The justification of the proposition 25 × 25 = 625 is, naturally, that if
anyone has been trained in such-and-such a way, then under normal
circumstances he gets 625 as the result of multiplying 25 by 25. But

Necessity

165

the arithmetical proposition does not assert that. It is so to speak an
empirical proposition hardened into a rule. It stipulates that the rule
has been followed only when that is the result of the multiplication.
It is thus withdrawn from being checked by experience, but now
serves as a paradigm for judging experience.
(Bold type added.)
Commenting on this passage, Fogelin writes:
It is not altogether clear what Wittgenstein is saying here [. . .] Here
we can imagine someone saying, ‘Wittgenstein has things backward.
It is the recognized necessity of 25 × 25 = 625 that leads us to adopt
it as a rule, and not its adoption as a rule that makes it necessary’.
Wittgenstein sees the force of this primitive complaint, and does not
think that it can easily be swept aside.48
The complaint imagined by Fogelin captures the strong temptation to
think of necessity in what I called intrinsic terms (as located in the proposition itself). Wittgenstein’s move here is exactly as Fogelin describes it
(his initial hesitation notwithstanding): Namely, to advance the opposite
thought, that it is extrinsic. Nevertheless, although he identifies the move
correctly, I am afraid he misses the very reason why Wittgenstein makes
it (as his hesitation, and the rest of the commentary, show): The rule’s
special origin, a key-idea to which I now turn.
We are now closer to seeing why the account of normativity is
‘genealogical’—and also ‘naturalist’. We are thus in the position to address
the second difficulty introduced above: Are the rules arbitrary, mere conventions, changeable at will? Is mathematical objectivity undermined?
It is the origin, or genealogy of an arithmetical proposition/rule that
qualifies it to be called ‘necessary’—it has to be the result of hardening.
Here, 25 × 25 = 625 originates in the one above, “(almost) everyone finds
that 25 batches of 25 pebbles each is 625 pebbles”. Before I elaborate, let
us note that, importantly, not all rules are like that. A comparison with
Hacker’s rule of chess above will be edifying here. The rule “the king
moves one square at a time” does not have this kind of origin—so, in
this sense, it is completely arbitrary, a mere convention we have to accept
if, and only if we want to play chess. This rule cannot be traced back
to any regularity of the kind presented in the story above. In short, and
crucially, this rule does not originate as a ‘peak’ result of an ‘experiment’,
and thus it is not generated by hardening.49 To challenge, or reject it is

48 Fogelin, Taking Wittgenstein, 102.
49 This is an important point missing in Hacker’s presentation: One can’t see how the rule
of chess is different from the mathematical rule.

166

Sorin Bangu

not a serious, important matter: to do so is just to refuse to play chess.
Challenging it may even be seen as constructive, and not deviant. One
may argue that a version of chess in which the king is allowed to move,
say, three squares at a time is more entertaining (and, after all, the rules
of chess have changed throughout history.)
So, what is it for a proposition to be a proposition of mathematics?
I have sketched here a two-stage genealogical account. What separates
them is the ‘hardening’ moment. At the first stage, the following regularity holds as a matter of fact: an overwhelming majority of people does
certain things in a certain way, i.e., uses the arrangement in Figure 7.1,
and reports the same result, 625. In a nutshell, the first stage embodies the observation that ‘peaks appear’. But note that these are contingent facts: namely, that such convergence of technique and results exist
and that they are pervasive (there is obviously nothing special about this
multiplication).50 Wittgenstein’s philosophy of necessity does not attempt
any ‘explanation’ of their occurrence.51 At this stage, there are no multiplication rules, no ‘mathematics’, only stable patterns of behavior:
‘(almost) everyone gets 625’. And, of course, “the arithmetical proposition does not assert that”—that, i.e., that ‘everyone gets 625’. This is so
indeed; this is a report of what happens, an empirical proposition, not an
arithmetical one.52
Thus, at the first stage we don’t have any arithmetic, only stable behavioral patterns (‘peaks’) which, importantly, include another social habit:
communities tend to exclude the deviants (when the matters are perceived
as serious, important). These kinds of stabilities encountered at this first,
pre-hardening stage, may be, as we saw, extremely useful socially, and the
community may be interested in re-enforcing them. Thus, we move on to
a second stage, when explicit rules are introduced.
But we now know where these needed rules come from. Crucially, they
are not mere inventions, figments of one’s (the Queen’s, or the community’s) imagination, i.e., arbitrary, random, whimsical. The temptation
(among some commentators) to say that for Wittgenstein the mathematical propositions are established ‘by convention’ is understandable. Yet,
when confronted with LFM, 107, there is little point in saying this:
50 As the numbers get bigger, it is very unlikely that such peaks appear (the infinite presents new challenges). Then we ‘harden’ by appealing to proofs; but an analysis of this
notion in Wittgenstein requires discussing his view of logical necessity, something I
cannot do here.
51 As far as I read him, he accepts that anthropology, evolutionary psychology, or some
other science may provide one.
52 Speaking of ‘almost everyone’, LFM, 95 even gives us a percentage: “Well, suppose that
90 per cent do it [the multiplication 123×489] all one way. I say, ‘This is now going to
be the right result.’ The experiment was to show what the most natural way is—which
way most of them go. Now everybody is taught to do it—and now there is a right and
wrong. Before there was not.”

Necessity

167

Mathematical truth isn’t established by their all agreeing that it’s
true—as if they were witnesses of it. Because they all agree in what
they do, we lay it down as a rule, and put it in the archives. Not until
we do that have we got to mathematics. One of the main reasons for
adopting this as a standard, is that it’s the natural way to do it, the
natural way to go—for all these people.
Neither the community, nor the Queen just decrees that the result is
625—this peak first appears in practice. So, we are dealing here with
very special conventions: what is stipulated (perhaps never explicitly!) is
that we change our attitude toward the empirical proposition ‘everyone
gets 625’. This is not just any proposition, but an already maximally
stable one (the result of applying an unanimously accepted technique).53
It is this proposition, which satisfies (i)-(ii) above, that is hardened into a
norm, or rule: ‘Get 625 when multiply 25 and 25!’
Thus, the rules at the second stage are natural, being generated from
(extremely stable) human habits, from the class of the accepted empirical
propositions identified at the first stage. Since the propositions-regularities
in this class have very large inertia, the rules they engender inherit this
inertia. The objectivity of mathematics is thus preserved, since the rules,
too, are stable, fixed; new rules cannot be introduced on a whim, and
existing ones cannot be changed as one wishes (unless, presumably, one
modifies their habitual basis, i.e., the whole form of life to begin with).54
Recall, moreover, that hardening is an extrinsic process; the result of the
community’s change in attitude toward a set of empirical propositions—
and not intrinsic, the result of recognizing something unassailable in
the content of these propositions themselves.55 Thus, what happens is

53 In LFM, 107, when speaking about “general consensus”, Wittgenstein says (cryptically)
that “There is something true in this”. We should now be able to see what this means,
what role consensus plays.
54 I am here in agreement with a point Barry Stroud made a long time ago—although he
writes about ‘logical’ necessity, the example and the context make it clear that he has
mathematical necessity in mind (the continuation he mentions refers to the famous +2
series in PI §185): “Logical necessity, he [Wittgenstein] says, is not like rails that stretch
to infinity and compel us always to go in one and only one way; but neither is it the
case that we are not compelled at all. Rather, there are the rails we have already traveled, and we can extend them beyond the present point only by depending on those that
already exist. In order for the rails to be navigable they must be extended in smooth and
natural ways; how they are to be continued is to that extent determined by the route
of those rails which are already there. I have been primarily concerned to explain the
sense in which we are ‘responsible’ for the ways in which the rails are extended, without
destroying anything that could properly be called their objectivity” (My emphases).
Barry Stroud, “Wittgenstein and Logical Necessity,” The Philosophical Review 74.4
(1965): 518.
55 Wittgenstein accepts, I think, the idea that made Quine famous: no content is immune
to revision—if one is willing to pay the price for challenging it.

168

Sorin Bangu

that certain empirical propositions (universally and peacefully accepted
declaratives, descriptions of ‘experimental’ peaks) become infinitely ‘hard’
to overthrow, i.e., become ‘grammatical’56 (paradigms, objects of comparison, prescriptions, imperatives to be enforced—firmly, through Abrichtung). RFM VII-67 reads as follows:
We say: ‘If you really follow the rule in multiplying, you must all
get the same result’ [. . .] This is [. . .] the expression of an attitude
towards the technique of calculation, which comes out everywhere
in our life. The emphasis of the must corresponds only to the inexorableness of this attitude both to the technique of calculating and to
a host of related techniques.
Given their origin, the rules themselves are touched by contingency in two
ways. Thus, one’s hopes for an absolutely secure account of necessity are
illusory. First, the existence of the pre-hardening, convergence stage (the
existence of peaks) is, as mentioned, a natural fact, which might not have
obtained. Moreover, the hardening itself, i.e., the transition from the pre- to
the post-hardening (convention-instituting) stage, might not have happened
either: there might not have been any felt need to institute rules—although
Wittgenstein once remarked (RFM I-74) that we have a “deep need” for
conventions. Yet, the disappointment should not become despair, since
contingency may not be such a major threat after all. We live in a world in
which peaks do emerge, and the transition does happen: were this not the
case—imagine there was no peak, but a whole range of results of 25 × 25
was reported (23, 19, 983, 77, etc.); or imagine that more than one peak
appeared (e.g., 625 and 601)—we would not have the ‘mathematics’ as we
know it today.
To make this thought more precise, let me make clear that the present
account allows the possibility of some radically different, alien community or species whose experimental peaks, when faced with the rows of
pebbles, are different from ours. They may converge and ‘peak’ on the
answer 1250, say, because, due to some bizarre physical law, a pebble
doubles itself instantly when one counts it. (There are similar stories in
Wittgenstein’s writings. Recall footnote 47.) This could then be ‘hardened’
into the rule 25 × 25 = 1250. This alien community would likewise not

56 The metaphor of hardening, as the mark of the ‘grammatical’, was perhaps inspired
by the idea that one needs rigid rods to make sense of measuring. Note also, as
Steiner astutely observes, that ‘grammar’ undergoes a major change, from Philosophical Grammar (“Grammar is not accountable to any reality”; X, 184) to LFM (where
arithmetical-qua-grammatical rules supervene on empirical regularities), change missed
by all other commentators, including Hacker—hence my agreement with him in section 1 is actually qualified. See Mark Steiner, “Empirical Regularities in Wittgenstein’s
Philosophy of Mathematics,” Philosophia Mathematica 17.1 (2009): 10.

Necessity

169

be able to alter this rule on a whim or at will—to do so, they would have
to fundamentally change their whole form of life. Just like ours, this rule
would not be arbitrary, but the natural result of the empirical regularities
holding in that region of the Universe. If one now asks what resources the
present account has for saying that this sort of alien mathematical community would be objectively wrong, then the answer is simply—none. But,
importantly, this should not worry us. Asking to establish this is to ask for
too much; moreover, the problem we seem to face may not occur in the
first place. For one thing, if the aliens are so different from ‘us’, in so many
fundamental respects, even if we may wish to present some arguments to
them to this effect, there would be no way to say whether they understand
them. (By arguments, I mean more substantial reasons than to say ‘look,
in the platonic heaven the result is 625, not 1250’.) For another, given the
deep chasm between the two forms of life, there would presumably be
no way to decide even more basic things, such as whether they actually
understand ‘counting’, ‘equal’, ‘different’, ‘row’, etc., the same way we do.
To push things even further, their form of life may not even contain the
language game of asking for, and giving, reasons.
Back to our world, once hardening is completed we can begin to talk
about necessity. These ‘hardened’ propositions constitute the framework
within which we now perform legitimate inferences (like the ones Hacker
mentioned above) between regular, empirical propositions: From “there
are 12 groups of 12 soldiers each in the parade”, we are entitled to move
to “there are 144 soldiers in the parade”. Crucially, these are moves (‘inferences’) which a very large majority of us makes anyway. What is left
unhardened is (called) ‘possible’, and is decidable by ‘experiment’. Thus,
we can talk about necessity only after the transition between the two poles
Wittgenstein speaks in RFM VII-30 has been completed: “It might be said:
experiment—calculation are poles between which human activities move”.
Returning to the fundamental problem posed at the very beginning, we
now see that the equation 25 × 25 = 625 lacks a truth-value in a strict
sense (is a rule). Yet it has one (is true), in a derivative sense. What is
true is the statement M: “Someone follows the order multiply 25 and 25
when they get 625”. Hence, if a child gets some other result, say 601, we
must say that they only thought they multiplied these numbers, or that
they thought they followed the rule, when in fact they did not. (Or, more
charitably, that they intended to follow the order to multiply, but failed—
regardless of how ‘certain’ they have felt, or what neural mechanisms
fired up in their brains.) Moreover, there is nothing wrong with saying
that we have mathematical beliefs. To believe that 25 times 25 is 625 is
just to believe proposition M above. And we don’t say that the multiplication order was properly followed as a result of some arbitrary decision,
but on the ground of the natural, empirical regularity (i.e., because of the
agreement in following the procedure in Figure 7.1, as opposed to some
other procedure).

170

Sorin Bangu

Finally, ‘must’ means something social too: that abrichtung-style measures are implemented to ensure that everyone follows the rule. They
make us ‘blind’ to all the virtually infinite possibilities to deviate, one
such possibility being exemplified by Figure 7.2. (Note also, in passing,
that this is what ‘blind’ means in “I follow the rule blindly” in PI §219).57
For the few ones who are not capable to become ‘blind’ in this way,
or refuse to, the consequences are serious, ranging from not graduating
elementary school to being locked up in the psychiatric ward.

3. The Skeptic Who Is Not: Concluding Remarks
To paraphrase, in the beginning was the Habit—although it is not only
the human psychological-behavioral learning and reaction patterns, but
also the stability of the physical world, that lead to the appearance of the
‘peaks’. Our interests come into play expressed as an attitude change—
we regard some of these regularities and techniques as worth “petrifying”
(as Wittgenstein puts it in LFM, 98); thus, the norms emerge.58 These
rules, once formalized, become what our textbooks now call ‘arithmetic’
(‘mathematics’). We perform ‘experiments’, like the Queen’s subjects, but
after ‘hardening’ on we can also make ‘calculations’. The question now is
what is the skeptic doing when challenging not the experiment, but the
calculation that 25 × 25 = 625.
Given the account sketched here, the challenger may object in two
ways. First, to the attitude change step—and say: “The empirical regularity should not have been adopted as a rule”. But this does not touch
the heart of the matter, since in saying this the skeptic does not reject
the equation above as a calculation, i.e., as an arithmetical proposition.
Rather, his objection is of a pragmatic nature; he only disagrees that
turning a particular hyper-stable regularity into a rule best serves our
interests.
The deepest objection is thus different. In essence, the skeptic claims
that the arithmetical norm lacks any binding power. He sincerely does not
believe that 25 × 25 must be 625, since, as far as he can see, alternatives

57 Here I should point out that the present account falls somewhere in between Wright’s
and McDowell’s takes on the rule-following problem. I acknowledge, with Wright, that
for Wittgenstein our having primitive dispositions (to react, and learn) is fundamental—
it is what leads to the peaks. But this is not enough, so on McDowell’s side, I also accept
that, through Abrichtung, the regularities become second nature, once regarded as normative. See the papers collected in Crispin Wright, Rails to Infinity (Cambridge, MA:
Harvard Univ. Press, 2001), and John McDowell, Mind Value and Reality (Cambridge,
MA: Harvard Univ. Press, 1998).
58 I am here in broad agreement with Penelope Maddy, whose analysis of logical necessity
stresses “the familiar Wittgensteinian trio of our interests, our nature, and the world’s
regularities.” Maddy, The Logical Must, 78.

Necessity

171

are available. He is of course aware that the community will ostracize
him for believing this; yet, to acknowledge 625 as the result because of
his fear of social exclusion is not the same as to recognize the rational
force of the calculation/rule itself. (We met an incarnation of such a skeptic before, as the subject who insisted that 25 × 25 = 601.) To deal with
this, we need to return to the previously introduced legal analogy, and
recall Aristotle’s famous pronouncement, in Book II.8 of his Politics, that
“the law has no power to secure obedience except habit”.59
Aristotle advances here a (the?) reason for the suppression of the
inclination to rebel against the laws of conduct in society: (many of) these
laws and norms are generated out of stable social customs.60 The analogue of the social rebellion is mathematical skepticism; and, following
the parallel further, we can suggest the ‘solution’ to it. If one already is
in the grip of a certain habit, one feels little (if any) temptation to rebel
when the habitual turns into the required—or, the virtually exceptionless is into ought. Thus, if the community is interested in implementing uniformity by introducing a norm (here: a calculation), then, if the
norm is of this kind (i.e., regularity-generated), there should be little
surprise that it does ‘secure obedience’. So, when it comes to accounting
for non-litigiousness, I submit that the arithmetical case is not essentially different from the legal case, only significantly more robust. In the
genealogy of arithmetic, the percentage of deviants we encounter is by
far the lowest; in fact, it is extremely low, negligible indeed. It is exactly
this maximal stability that partly makes possible the language game of
‘arithmetic’.
Thus, on the present account, the Wittgensteinean solution to the
skeptical worry is deflationary. Just like the percentage of those (inclined
to) voicing it, the worry is negligible—inaudible. Were this not so, there
would not be any regularities to harden, hence no calculations, and thus
no mathematics to be skeptic about.

Acknowledgements
I thank the editors, Kevin Cahill and Thomas Raleigh, for inviting me to
contribute and for detailed comments. I am indebted to Penelope Maddy,
Alois Pichler, Simo Säätelä, Kaveh Lajevardi, Sorin Costreie, Constantin Brincus, Mircea Flonta, Andrew Lugg, Constantine Sandis, Nuno
Venturinha and Alex Dragomir for reading and commenting on various
drafts of the paper.
59 Aristotle, Politics, trans., with Introduction and Notes, by C.D.C. Reeve (Indianapolis
and Cambridge: Hackett Publishing Company, 2010), 49.
60 See the discussion in Hermann Kantorowicz, The Definition of Law, ed. by A. H.
Campbell (Cambridge: Cambridge Univ. Press, 2014/1958), especially the section on
‘Law and Social Custom’.

172

Sorin Bangu

References
Aristotle, Politics. Translated with Introduction and Notes by C.D.C. Reeve, Indianapolis and Cambridge: Hackett Publishing Company, 2010.
Baker, G. and P. M. S. Hacker. Wittgenstein: Understanding and Meaning, 2nd
ed., Oxford: Blackwell, 2005.
Bangu, S. “Later Wittgenstein on Essentialism, Family Resemblance and Philosophical Method.” Metaphysica 6.2 (2005): 53–73.
———. “Later Wittgenstein’s Philosophy of Mathematics.” The Internet Encyclopedia of Philosophy, accessed May 6, 2012, www.iep.utm.edu/wittmath/
———. “Wynn’s Experiments and Later Wittgenstein’s Philosophy of Mathematics.” Iyyun: The Jerusalem Philosophical Quarterly 61 (2012): 219–241.
Bederman, D. J. Custom as a Source of Law, Cambridge: Cambridge University
Press, 2010.
Dummett, M. “Wittgenstein’s Philosophy of Mathematics.” Philosophical Review
LXVIII (1959): 324–348.
Finkelstein, D. “Wittgenstein on Rules and Platonism.” In The New Wittgenstein,
edited by A. Crary and R. Reed, 53–73, New York: Routledge, 2000.
Fogelin, R. Taking Wittgenstein at His Word, Princeton: Princeton University
Press, 2009.
Gerrard, S. “A Philosophy of Mathematics Between Two Camps.” In H. Sluga
and D. Stern, eds. The Cambridge Companion to Wittgenstein, 171–197, Cambridge: Cambridge University Press, 1996.
Glock, H.-J. “Necessity and Normativity.” In Sluga and Stern, eds., 198–225,
1996.
Hacker, P.M.S. and G. Baker. Wittgenstein: Rules, Grammar and Necessity, 2nd
ed., extensively revised by P.M.S. Hacker, Oxford: Blackwell, 2009.
Harris, J. Art History: The Key Concepts, London: Routledge, 2006.
Kantorowicz, H. The Definition of Law, edited by A. H. Campbell, 1st ed. 1958,
Cambridge: Cambridge University Press, 2014.
Lewis, D. On the Plurality of Worlds, Oxford: Blackwell, 1986.
Loux, M. J. Metaphysics: A Contemporary Introduction, 3rd ed., London: Routledge, 2006.
Maddy, P. The Logical Must: Wittgenstein on Logic, New York: Oxford University Press, 2014.
McDowell, J. Mind, Value and Reality, Cambridge, MA: Harvard University
Press, 1998.
Raz, J. “Can There Be a Theory of Law?” In The Blackwell Guide to the Philosophy of Law and Legal Theory, edited by M. P. Golding and W. A. Edmundson,
324–342, Oxford: Blackwell, 2005.
Schroeder, S. “Mathematical Propositions as Rules of Grammar.” Grazer Philosophische Studien 89 (2014): 21–36.
Shanker, S. G. Wittgenstein and the Turning Point in the Philosophy of Mathematics, London: Routledge, 2006.
Sluga, H. and D. Stern, eds. The Cambridge Companion to Wittgenstein, Cambridge: Cambridge University Press, 1996.
Stalnaker, R. “Modality and Possible Worlds”. In A Companion to Metaphysics,
edited by J. Kim, E. Sosa, and G. Rosenkrantz, 45–55, 2nd ed., Oxford: WileyBlackwell, 2009.

Necessity

173

Steiner, M. “Empirical Regularities in Wittgenstein’s Philosophy of Mathematics.”
Philosophia Mathematica 17.1 (2009): 1–34.
Stroud, B. “Wittgenstein and Logical Necessity.” The Philosophical Review 74.4
(1965): 504–518.
Wittgenstein, L. Lectures on the Foundations of Mathematics, Cambridge 1939,
edited by C. Diamond, Chicago: University of Chicago Press, 1976.
———. Ludwig Wittgenstein and the Vienna Circle, Notes recorded by F. Waismann, edited by B. F. McGuinness, Oxford: Blackwell, 1967/1979.
———. Philosophical Grammar, edited by R. Rhees, translated by A. J. P. Kenny,
Oxford: Blackwell, 1974.
———. Philosophical Investigations, 3rd ed., translated by G. E. M. Anscombe,
Oxford: Blackwell, 1968.
———. Remarks on the Foundations of Mathematics, edited by G. H. von Wright,
R. Rhees, and G. E. M. Anscombe. Revised ed., Oxford: Blackwell, 1978.
Wright, C. Rails to Infinity, Cambridge, MA: Harvard University Press, 2001.

Part III

Animal Minds, Human
Psychology

8

Minding the Gap
In Defense of Mind-Mind
Continuity1
Dorit Bar-On

If a lion could talk, we could not understand him.
Ludwig Wittgenstein2

Introduction
As commonsense would have it, although lions and other non-human animals surely cannot speak, they can communicate with—and understand—
each other. Common folks, as well as practicing ethologists, comparative
psychologists, and evolutionary anthropologists—not to mention zookeepers, park rangers, and pet owners—readily attribute to non-human
animals sensations and feelings (hunger, thirst, fatigue, agitation, pain, and
pleasure), intentional actions and intentional states or attitudes (trying to
do this or that, looking for or wanting to get something, being startled/
surprised/puzzled/scared by specific things), and so on. However, according to the view I call “continuity skepticism”, there is an unbridgeable
gap (or “deep chasm”, as some put it) separating human minds from the
minds of non-human animals.3 Animals’ mental lives, understanding, and
communication are said to be so profoundly different from ours that it
makes little sense to locate them along a natural continuum that culminates in our own mentality and language.4
1 This paper is a sequel to (and overlaps with parts of) Dorit Bar-On, “Expressive Communication and Continuity Skepticism,” Journal of Philosophy 110(6) (2013): 293–330,
and Dorit Bar-On and Matthew Priselac, “Triangulation and the Beasts,” in Triangulation From an Epistemological Point of View, eds. Maria Cristina Amoretti and Gerhard
Preyer (Berlin: Ontos Verlag, 2011), 121–152. Special thanks to Matthew Priselac and
Kevin Richardson, and to Carol Voeller, for help in preparing a talk on which this paper
is based. And thanks to Kevin Cahill and to audiences at conferences and colloquia held
at the universities of Alabama, Auburn, Bergen, Freiburg, and Hertfordshire, where earlier versions of this paper were presented in recent years.
2 Ludwig Wittgenstein, Philosophical Investigations, 3rd ed. Translated by G. E. M. Anscombe. (Oxford: Basil Blackwell, 1968), 223.
3 The sense in which the view is skeptical is explained in Section 1 below.
4 Historically, debates regarding animal mentality divided empiricists and rationalists. The
rationalist tradition has contemporary supporters even among evolutionary biologists,

178

Dorit Bar-On

My focus here will be on a particular version of the “gap” claim,
according to which not only are there significant ‘synchronic’ differences
between relevant capacities of extant non-human animals and ours, but
we should also doubt that any feasible precursors of our own mentality and behavior could be found in the capacities of some extinct species. So, diachronically speaking (on the skeptic’s view), there is—to
use a well-known metaphor—a “Rubicon” that no non-human animal
could have crossed. And this tells against the possibility of constructing a “natural history” of human minds.5 I begin by briefly outlining
a radical version of this skeptical view—defended by Davidson—which
denies that there could be any middle ground between animal mentality
and communication, on the one hand, and human reflective understanding and interpretation, on the other (Section 1). In Section 2, I present a
form of communication that we share with non-linguistic and prelinguistic creatures, namely, expressive communication—which, I will argue, is
apt to present a “synchronic middle ground”, poised between two poles
that the Davidsonian skeptic contrasts. In Section 3, I argue that proper
appreciation of the role expressive behavior plays in the lives of creatures capable of it, and the kind of communication it affords, point to a
“diachronic bridge” that could feature in a natural history connecting us
with our pre-human ancestors.6 Finally, in Section 4, I briefly consider
some possible implications of the existence of such a natural history for
a philosophical understanding of the relationship between human and
non-human mindedness.7
ethologists, and comparative psychologists. For two recent volumes, see Susan Hurley
and Matthew Nudds, eds., Rational Animals? (New York: Oxford University Press,
2006); and Robert Lurz, The Philosophy of Animal Minds (Cambridge: Cambridge University Press, 2009).
5 This “diachronic” claim, note, is much more radical than any claim about deep synchronic differences between human and non-human mentality. The presence of extant
gaps, however wide, is in principle consistent with the existence of extinct intermediaries.
See Bar-On, “Expressive Communication,” 295f.
6 I defend this position in several places—see e.g. Bar-On and Priselac, “Triangulation
and the Beasts,” Bar-On, “Expressive Communication”; and Dorit Bar-On, “Sociality,
Expression, and This Thing Called Language,” Inquiry 59(1) (2016): 56–79.
7 I take the view I outline in Sections 2–4 below to be broadly Wittgensteinian (both in
its search for a natural history of human mindedness, and in its appeal to expressive
behavior as relevant to the origins of meaning). Although it’s not entirely clear what
Wittgenstein’s view is on the question of the origins of human mind and language, what
I say below has several points of contact with a number of remarks he makes on animals, mentality, and expression, and on meaning, understanding, and interpretation, in
the Philosophical Investigations; Ludwig Wittgenstein, Remarks on the Philosophy of
Psychology, Volume I, eds. G. E. M. Anscombe and G. H. von Wright, trans. G. E. M.
Anscombe (Oxford: Basil Blackwell, 1998); Ludwig Wittgensteign, Remarks on the Philosophy of Psychology, Volume II, eds. G. H. von Wright and Heikki Nyman, trans. C.
G. Luckhardt and M. A. E. Aue (Chicago: University of Chicago Press, 1998); and elsewhere. Wittgenstein exegesis, however, goes beyond my scope in this paper.

Minding the Gap

179

1. Davidson’s Continuity Skepticism8
Contemporary continuity skeptics, unlike some of their historical predecessors, typically disavow Cartesian dualism and are self-proclaimed naturalists. Given that, we can, for present purposes, set aside metaphysical
arguments purporting to show that (human) minds are not part of the natural world, as well as arguments that trade on essentially epistemic, or else
methodological problems to do with our ability to know the minds of nonhuman animals. Continuity skepticism as I understand it here is motivated
by observations that purport to establish some fundamental differences
between human thought and language, on the one hand, an the conduct
and communication of all existing non-human animals, on the other
hand. The former is said to have essentially intentional, flexible, objective,
reflective, rule-governed, symbolic, world-directed, propositional, reasonbased, norm-governed character. The latter: merely responsive, stimulusbound, passion- or need-driven, pattern-governed, non-symbolic, merely
world-involving, and thoroughly manipulable character. Given these vast
“synchronic” differences, some philosophers have maintained that the
application of our concepts of intentional action, meaning, semantic content, reference, propositional attitudes, etc. to non-human creatures is at
best a matter of analogy or “metaphorical extension”. Importantly, the idea
that there are distinctive, unique, and essentially human capacities that do
not have even remote analogues or precedents among non-human animals
survives the recognition that various species of non-human animals are not
only sentient, but also possess considerable intelligence and are capable of
goal-directed behavior and self-initiated action, as well as complex cognition and affect, problem solving, memory, and group organization.9
We can discern three key moves made by continuity skeptics, in increasing order of strength—and decreasing order of plausibility10:
(i) Observation of significant, deep human/non-human differences
(“synchronic discontinuity”)
Many non-human animals enjoy sentience, as well as possessing “reliable
differential responsive dispositions”. But human beings also enjoy sapience, which “consists in knowing one’s way around the space of reasons”.11
(ii) Denial of mental commonalities between us and “dumb animals”
(“synchronic disconnect”)
8 For a fuller discussion, see Bar-On, “Expressive Communication,” Sections 1 & IID.
9 For relevant references, see Bar-On, “Expressive Communication.”
10 As my aim here is not exegetical, I provide only a highly selective sample of relevant citations below, which obviously cannot do justice to the richness and subtlety of the works
cited. For additional relevant quotations, see Bar-On, “Expressive Communication.”
11 Robert Brandom, Making It Explicit (Cambridge: Harvard University Press, 1994), 5.

180

Dorit Bar-On

Human mental capacities cannot be “factorized” into those we fully
“share with dumb animals” and those we do not.12 Even as regards sentience, we should recognize that “there are two species . . . one permeated
by spontaneity and another independent of it”.13 Even “our embodied
coping” amounts to “more than the embodied coping of nonrational
animals”.14
(iii) Rejection of diachronic continuity in “the order of explanation”
(“diachronic discontinuity”)
The undeniable claim that “there were non-linguistic animals before
there were linguistic ones, and the latter did not arise by magic” is only a
claim about the causal “order of being”.15 In the “order of explanation”,
the irreducibly normative “intentionality of non-linguistic creatures is
dependent on, and . . . derivative from, that of their linguistically qualified interpreters, who as a community exhibit a nonderivative, original
intentionality.”16,17 It is therefore philosophically futile to raise the question of “where we come from”.18
Continuity skeptics, like more familiar skeptics in other areas, are
prepared to distance themselves equally from the deliverances of both
commonsense and the relevant sciences.19 Our commonsense mentalistic attributions to animals are said by the skeptic to be shot through
with uncritical application of concepts that are inherently tailored to
understanding our minds. They are hopelessly anthropomorphic and
‘too thick’ to serve as a starting point for a philosophical explanation of
our relation to the beasts. On the other hand, scientific accounts of our
mental and linguistic capacities that portray them as continuous with
non-human capacities (insofar as they stay clear of uncritical anthropo-

12 John McDowell, Mind and World (Cambridge: Harvard University Press, 1996), 69.
(See also page 64).
13 Ibid., 69.
14 John McDowell, “What Myth?” Inquiry 50(4) (2007): 344.
15 Brandom, Making it Explicit, 155.
16 Ibid., 152.
17 For the different ‘orders’, see Wilfrid Sellars, “Empiricism and the Philosophy of Mind,”
in Minnesota Studies in the Philosophy of Science, vol. 1, eds. Herbert Feigl and Michael
Scriven (University of Minnesota Press, 1956), 253–329; and Wilfrid Sellars, “Mental
Events,” Philosophical Studies 39(4) (1981): 325–345. For a recent critique of Brandom’s view from an evolution-friendly philosophical perspective, see Daniel Dennett,
“The Evolution of ‘Why?’,” in Reading Brandom: On Making It Explicit, eds. Bernhard
Weiss and Jeremy Wanderer (New York: Routledge, 2010), 48–62.
18 Brandom, Making it Explicit, 4.
19 A good case in point is the Quinean or Kripkensteinian meaning skeptic, who dismisses
both the commonsense confidence that alternative interpretations of speakers’ utterances can be ruled out and scientific attempts to ground interpretation choices in, say,
facts about speakers’ behavioral dispositions or cognitive organization.

Minding the Gap

181

morphism) are regarded by the skeptic as overly reductionist and ‘too
thin’ to deliver a philosophically credible account of the emergence of
our minds in nature.20
1.1 Davidson’s Triangulation
The three claims introduced above are all explicitly endorsed by Davidson. A key feature of human thought and language, which all animal
thought and communication lack, is the objectivity of semantic content—
its being “true or false independent . . . of the existence of the thought
or the thinker”.21 Such thought requires possession of concepts, whose
employment involves rule-following (as opposed to merely behaving in
accordance with rules), which brings in its train the possibility of genuine
error.22 Davidson, like Brandom and McDowell (each of whom, in turn,
regards himself as following Wittgenstein on this point), thinks that the
possibility of genuine error requires the rule-follower’s awareness of that
possibility, and thus having objective thought requires the thinker to have
an awareness or grasp of objectivity, something that goes beyond the
capacities of non-human animals.23,24
In “The Emergence of Thought”, Davidson directly addresses the question of “the emergence of mental phenomena”—which, he says, is the
“conceptual problem . . . of describing the early stages in the maturing
of reason . . . that precede the situation in which [mentalistic] concepts
have clear application”.25 There (and elsewhere), he explicitly endorses
continuity skepticism:
There cannot be a sequence of emerging features of the mental, not
if those features are to be described in the usual mentalistic vocabulary. Of course . . . each stage in the emergence of thought can be
20 For more empirical versions of continuity skepticism, see Bar-On, “Expressive
Communication.”
21 Donald Davidson, Subjective, Intersubjective, Objective (New York: Oxford University
Press, 2001), 130.
22 For an earlier articulation of this claim, see Jonathan Bennett, Rationality: An Essay
Towards an Analysis (New York: Routledge & Kegan Paul, 1964), 87f.
23 Davidson, Subjective, Intersubjective, Objective, 130.
24 Brandom and McDowell agree with Davidson that even perceptual thought, if it is to
enjoy objectivity, requires reflective grasp of the contrast between subjective and objective). See Brandom, Making It Explicit, 48 and 63; and McDowell, Mind and World,
114ff. For a critique of the “Individualist Representationalism” betrayed by this view,
see Tyler Burge, Origins of Objectivity (Oxford: Oxford University Press, 2010), Ch. 2
and passim. Burge goes on to offer a continuist view regarding objective thought. In
Dorit Bar-On and Matthew Priselac, “Origins: Subjective, Objective, Intersubjective”
(in progress), we evaluate the extent to which Burge’s defense of continuity succeeds in
engaging Davidson’s skepticism.
25 Davidson, Subjective, Intersubjective, Objective, 127.

182

Dorit Bar-On
described in physical terms. But this will fail as an explanation of
the emergence of the mental since we . . . cannot expect to find, a
way of mapping events described in the physical vocabulary onto
events described in the mental vocabulary. . . . In both the evolution
of thought in the history of mankind, and the evolution of thought
in an individual, there is a stage at which there is no thought followed by a subsequent stage at which there is thought. . . . What
we lack is a satisfactory vocabulary for describing the intermediate
steps.26

To support the seemingly questionable move from the last claim, about
lack of vocabulary, to the claim that there can be no intermediate steps or
emergence, Davidson introduces the idea of triangulation. This is the idea
that contentful thought about an objective world, as well as meaningful
linguistic communication, require “the existence of a triangle” whose
base connects two subjects, S1 and S2, and whose apex is an object in
the world, O.27 In defense of his continuity skepticism, Davidson invites
us to contrast a “pure” triangular scenario involving non-human animals with the “reflective” triangular scenarios we are familiar with in our
own intersubjective experiences. Subjects in pure triangulation, Davidson allows, can “classify” and “generalize” and form “habitual inductions”, even learned ones, grouping various stimuli together “by virtue
of the similarity of the[ir] responses”.28 He even allows that they can
come to associate each other’s responses to O with O. For example, S1
could respond to S2’s O-reaction as S2 responds to O, and vice-versa.
This makes room for a simultaneous discrepancy that is at the heart of
objectivity (as Davidson understands it), whereupon “space is created”
for the concept of error to develop.29
But although Davidson thinks that this sort of scenario is necessary for
providing a conceptual foundation for objective thought, he insists that
it’s insufficient for its emergence. This is because nothing in the intersubjective interactions of pure triangulation supports the attribution of
reflective grasp of the concepts of error, belief, truth, etc. From each subject’s point of view, the other subject’s behavior is simply something that
can be correlated (or not) with items in the world—objects, events, state
of affairs—as smoke is correlated with fire, or deer tracks with the recent
presence of deer. Any disagreement between them would amount to no
more than behavioral discord. What’s missing is one subject’s treating

26 Ibid., emphasis added.
27 Ibid., 121.
28 Donald Davidson, “Externalisms,” in Interpreting Davidson, eds. Petr Kot̓átko et al.
(Stanford: CSLI Publications, 2001), 5.
29 Donald Davidson, “Comments on Karlovy Vary Papers,” in Interpreting Davidson, eds.
Petr Kot̓ átko et al. (Stanford: CSLI Publications, 2001), 285–308.

Minding the Gap

183

Intersubjective Object-Directed Relations in
“Pure” Triangulation
O

S1’s O-responses

!

!
S1

!

S2’s O-responses

Mediated O-responses
purely behavioral disagreement

S2

Figure 8.1 “Pure” Triangulation

another as a subject who has a take on the world, which take can fit or
fail to fit with the way things are.30
Indeed, Davidson thinks that nothing short of linguistic communication between the two subjects could move us significantly beyond pure
triangulation, for “[o]nly when language is in place can creatures appreciate the concept of objective truth . . . [and] make use of the triangular
situation to form judgments about the world”.31 In reflective triangulation, we have language speakers, capable of responding to objects with
meaningful, true or false utterances. On a given occasion, S1 may produce a sentence (say, “There’s a leopard nearby”) which S2 presumes S1
to hold-true, and yet which he (S2) takes to be false. Genuine objectivity
is provided for via the possibility of each subject recognizing a potential
gap between what is held to be true (and thus believed) and what is

30 See Davidson, Subjective, Intersubjective, Objective, 121; and Davidson, “Externalisms,” especially pp. 11–16. See also Naomi Eilan, “Joint Attention, Communication,
and Mind,” in Joint Attention: Communication and Other Minds, eds. Naomi Eilan
et al. (Oxford: Oxford University Press, 2005) for a helpful interpretation of this
point.
31 Davidson, Subjective, Intersubjective, Objective, 131. See also Donald Davidson,
Problems of Rationality (New York: Oxford University Press, 2004), 140–141; and
Davidson, “Externalisms,” 13. Also, compare McDowell, Mind and World, 114f., who
argues that a perceiver’s grasp of the difference between subjective and objective is
required to render her environment “more than a succession of problems and opportunities” and to allow both a [subjective] self and an [objective] world to “be in view.”

184

Dorit Bar-On

Intersubjective Object-Directed Relations in
“Reflective” Triangulation
O

S1’s O-sentences

S1

S2’s O-Sentences

Mutual interpretations
genuine intersubjective

S2"

disagreement becomes possible

Figure 8.2 “Reflective” Triangulation

the case.32 Reflective triangulation, Davidson maintains, is different in
kind from pure triangulation, and proper understanding of the reflective
capacities manifested in it would give the lie to any attempt to interpose
a significant intermediary between them.
By insisting that reflective triangulation is not only sufficient but also
necessary for the emergence of objective thought Davidson, in effect,
gives up on the possibility of explaining the natural emergence of objective thought and linguistic communication.33 Such an explanation would
require interposing a viable intermediary between the pure and reflective
triangulations that he contrasts. This, in turn, would require supposing
that we could “have an analysis of thought” or “a reduction of the intentional to the extensional”. Yet Davidson (like Brandom and McDowell)
thinks that such reduction “is not to be expected”, due to the irreducibly
normative character of objective thought.34
We can see Davidson’s continuity skepticism as resting on the following claims:

32 For more on the role language plays, see especially Davidson, Subjective, Intersubjective,
Objective, 111ff. and 119ff. Unlike Quine, Davidson is not a wholesale skeptic about
the legitimate application of the intentional idiom. For him, once mutual linguistic
interpretation is in place, there is room for genuine rule-following, conceptualization,
and the possibility of genuine error and disagreement.
33 Davidson, “Externalisms,” 13.
34 Ibid., emphasis added.

Minding the Gap

185

1. The objectivity of thought (as well as its intentionality, in Brentano’s
sense) has an irreducibly normative character.
“Irreducibility”
2. To have objective, intentional thought, a subject must have grasp of
the concept of objectivity.
“Grasp”
3. Grasp of the relevant notion of objectivity requires intersubjectivity—
it requires understanding of the idea of another subject’s take.
“Intersubjectivity”
4. Intersubjective interactions in all but reflective triangulation are unfit
to provide such understanding. (Alleged intermediate cases either
collapse into the pure case, or are assimilated into the reflective case.)
“Reflective Interpretation”
None of these Davidsonian claims is unproblematic. Quinean eliminativists,35 Dennettian instrumentalists,36 and proponents of “new-wave
metascience”,37 or of “embodied cognitive science”38 may accept the irreducibility of intentional idioms (1) but for that reason maintain that no
such notions could have room in a serious science. In a clear sense, these
philosophers are not engaging the Davidsonian skeptic, as they are in no
way concerned to identify a stable intermediary case between Davidson’s
pure and reflective cases (by tackling 4 head on). They can perhaps be
seen as providing a “skeptical response” to continuity skepticism.
In recent years, several philosophers of mind who subscribe to
representationalism—most notably, Tyler Burge—have taken issue with
what they regard as the hyperintellectualism manifest in Davidson’s “grasp”
requirement (Claim 2).39 In contrast to eliminativists, these philosophers

35 See Willard Van Orman Quine, Word & Object (Cambridge: MIT Press, 1960), Ch.
VI. See also Paul Churchland, “Eliminativism and Propositional Attitudes,” Journal of
Philosophy 78 (1981); and Patricia Churchland, Neurophilosophy (Cambridge: MIT
Press, 1986).
36 See Daniel Dennett, The Intentional Stance (Cambridge: MIT Press, 1987).
37 See John Bickle, Philosophy and Neuroscience: A Ruthlessly Reductive Account (Norwell, MA: Kluwer Academic Publishers, 2003).
38 See Anthony Chemero, Radical Embodied Cognitive Science (Cambridge: MIT Press,
2009).
39 Burge argues that most post-Kantian philosophers of mind have either fallen into
hyperintellectualism, clustering into two families of “Individual Representationalism”
(see Burge, Origins of Objectivity, Part II), or have been in the grip of reductionist
naturalism. Burge advocates what can be seen as a naturalist, nonreductionist, antiindividualist representationalism. (But NB: Burge’s view is only anti-individualist in the
sense that it posits constitutive connections to the world outside the individual, not in
the sense that it assigns any essential connection between mindedness and relations to
other individuals).

186

Dorit Bar-On

believe that intentional notions do have a substantial role to play in the
scientific explanation of perception, behavior, and action. They maintain
that, once it is denied that having objective thought requires individuals
already to have the “psychological resources to represent preconditions of
objectivity”,40 the road is clear to recognizing that current scientific theories of psychological phenomena (perception, for example) make warranted attributions of intentional states when characterizing the mental
life and explaining the behavior of non-human animals (including not
only mammals, but even invertebrates). On this representationalist view,
by the time we get to (what Davidson describes as) pure triangulation,
objective, genuinely representational thought is already in place. Thus, this
view is clearly apt to fund substantial continuities between the representational capacities of non-human animals and ours. However, proponents
of representationalism often also reject Davidson’s Claim 3 (which assigns
a special role for intersubjective interactions). So, like the eliminativists
(though for different reasons), they do not appear concerned to tackle
Claim 4 head on. In any event—and relatedly—from the perspective of
the skeptic, representationalists may seem guilty of an overly permissive
conception of what it takes to have objective thought; so the skeptic may
deny that the representationalist pays sufficient heed to Claim 1.
The representationalist framework has been under much attack in
recent years, under the heading of “extended mind”, “embodied cognition”, “enactivism”, etc.41 In an effort not to over-intellectualize psychological phenomena, anti-representationalists undertake to offer
‘minimalist’ accounts of all pre-verbal mentality and intelligent behavior,
as well as intersubjective interactions, in purely non-representationlist,
non-intentional terms. (Anti-representationalist variously appeal to
mere sensitivity to natural signs, goal-directedness, response-detection,
dynamic embodiment, coordinated action routines, anticipatory mechanisms, and so on.) Although non-human animals and pre-verbal children
possess various wholly nonconceptual modes of thought and attitudes,
propositional attitudes (which are conceptually structured) are the prerogative of language speakers.42 The anti-representationalist finds a deep
gulf between the merely enacted, non-conceptual states of non-linguistic
subjects and the fully propositional thought of language speakers. One
self-proclaimed “radical enactivist”, Dan Hutto, indeed explicitly refers
40 Burge, Origins of Objectivity, 139.
41 See Chemero, Radical Embodied Cognitive Science; Andy Clark, Supersizing the Mind:
Embodiment, Action, and Cognitive Extension (New York: Oxford University Press,
2008); and J. Kevin O’Regan and Alva Noë, “A Sensorimotor Account of Vision and
Visual Consciousness,” Behavioral and Brain Sciences 24 (2001): 939–1031.
42 In a similar spirit, Michael Dummett had earlier proposed that animals can only
engage in “essentially spatial,” dynamic, perception-based, and context-bound “protothoughts” that lack all conceptual structure. See his Origins of Analytical Philosophy
(London: Duckworth and Cambridge: Harvard University Press, 1993), 123f.

Minding the Gap

187

to “a major cognitive Rubicon” between the “basic minds” of non-verbal
creatures and our own “superminds”, the crossing of which is marked by
becoming language users.43 But, clearly, the bigger the Rubicon, and the
more committed one is to the role of language in getting creatures over
the representation “hump”, the more one plays into the hands of the
Davidsonian skeptic, since the more difficult it would seem to be to envisage an account of the emergence of propositional thought and language
in both ontogeny and phylogeny.
Thinking again in terms of Davidson’s triangles and his continuity
skepticism, my present complaint is that representationalists are committed to portraying even pure triangulation in terms that are “too rich”.
For on their view the languageless thoughts of subjects even in pure triangulation already exemplify key features of our own thought (they don’t
simply foreshadow them). On the other hand, anti-representationalists
seem committed to characterizing even intermediate triangulation in
terms that are “too poor.” On their view, the thought of “basic minds”
not only doesn’t exemplify but in no way foreshadows the thought of
“superminds” like ours. Thus, like Davidson, the anti-representationalist
would seem committed to denying the possibility of explaining the natural emergence of propositional thought and language (in both ontogeny
and phylogeny). I agree with representationalists that we should reject
Davidson’s “grasp” requirement. I also agree with enactivists who reject
the Davidsonian conception of basic mutual understanding as requiring
reflective-interpretive mindreading. However, my concern here is to see
whether we can provide a “straight” response to the continuity skeptic by
offering a viable candidate for a middle-ground that tries to accommodate the distinctive role Davidson assigns for intersubjectivity in grounding objectivity. To do this, I submit, one must identify a distinct type of
animal intersubjective interactions that can be seen to foreshadow key
features of human mentality as the skeptic understands it, without yet
fully exemplifying all of them.44

2. Expressive Communication
Thinking in terms of Davidson’s sharply contrasting triangles, the challenge is to try to bridge the diachronic gap without either underestimating the rich character of linguistic interactions in reflective triangulation
or overestimating the power of non-linguistic interactions in pure triangulation. This requires a characterization of a genuinely intermediate
43 Daniel Hutto, Folk Psychological Narratives: The Sociocultural Basis of Understanding
Reasons (Cambridge: MIT Press, 2008), 96.
44 For relevant discussion, see Dorit Bar-On, “Crude Meaning, Brute Thought (Or: What
Are They Thinking?!),” Journal for the History of Analytic Philosophy (special issue,
forthcoming).

188

Dorit Bar-On

triangulation that neither collapses into pure triangulation nor illicitly
presupposes capacities that are only possible in reflective triangulation.
Toward this end, I would like to propose for consideration expressive
behavior. In his seminal work The Expression of the Emotions in Man
and Animals,45 Darwin identifies expressive behavior as representing an
important common ground between ‘man and animals’. He had in mind a
wide range of vocal, facial, and postural displays, including (among others) yelps, growls, lip smacks, distress, food, and alarm calls, fear barks
and grimaces, “play faces” and play bows, cowering demeanor, and so
on.46 Darwin provides a rather nuanced characterization of these sorts
of behaviors—as being at once physiological and psychological, linking
bodily changes and movements intimately with complex, world-directed
emotional states.47 Yet many philosophers, as well as theorists of animal
communication and language evolution, often assimilate paradigmatic
natural expressions too closely to mere physiological symptoms, such
as red spots on the skin and sneezes, or to nonvoluntary displays that
merely convey information about biologically significant features of the
displayer (like a peacock’s tail), portraying them simply as reliable natural signs of the internal states that regularly cause them.48
With Darwin, I believe that a purely causal construal fails to do justice
to the richness and complexity of expressive behaviors.49 Elsewhere,50
45 Charles Darwin, The Expression of the Emotions in Man and Animals (London: John
Murray, 1872).
46 See Bar-On, “Origins of Meaning” for references to works that describe and analyze
facial, vocal, postural and gestural expressions. There is clearly a degree of heterogeneity
in the class of behaviors I am including here. I am relying on a measure of pre-theoretic
understanding, which can only provide an initial basis for future theorizing.
47 Darwin was arguing against an earlier work by Charles Bell—Essays on the Anatomy
and Philosophy of Expression (London: John Murray, 1824)—that the expressive
behaviors of human and non-human animals had shared ancestry (whereas Bell maintained that there were divinely created human muscles designed for the expression of
uniquely human feelings).
48 See, for example, Paul Grice, “Meaning,” The Philosophical Review 66 (1957): 377–88;
William Alston, “Expressing,” in Philosophy in America, ed. Max Black (Ithaca: Cornell
University Press, 1965), 15–34; and Jonathan Bennett, Linguistic Behavior (Cambridge:
Cambridge University Press, 1976). See also, in the animal communication literature,
e.g. John Maynard Smith and David Harper, Animal Signals (Oxford: Oxford University Press, 2003), especially chapter 7; Stephen R. Anderson, Dr. Dolittle’s Delusion:
Animals and the Uniqueness of Human Language (New Haven: Yale University Press,
2004), Ch. 2 and passim; and W. Tecumseh Fitch, The Evolution of Language (Cambridge: Cambridge University Press, 2010), Ch. 4 and passim.
49 For the account of expressive behavior that follows, see Bar-On, Speaking My Mind: Expression and Self-Knowledge (Oxford: Clarendon Press, 2004), Chapters 6–8. See also Mitchell
Green, Self-Expression (Oxford: Oxford University Press, 2007). For some reservations
concerning Green’s account, see Dorit Bar-On, “Expressing as Showing What’s Within: On
Mitchell Green’s Self-Expression,” Philosophical Books 51(4) (2010): 212–227.
50 Bar-On “Expressive Communication”; Bar-On and Priselac, “Triangulation and the
Beasts”; Dorit Bar-On “Origins of Meaning”; and Dorit Bar-On and Mitchell Green,
“Lionspeak: Communication, Expression, and Meaning,” in Self, Language, and World:

Minding the Gap

189

I have argued that animal expressive interactions exhibit features that
foreshadow significant aspects of human linguistic communication,
despite not being underwritten by complex communicative intentions or
ingenious insight. Here I can only rehearse some of the key ideas that
bear on the challenge posed by continuity skepticism.

2.1 Expressing Is a Form of Showing
Consider first expressive behavior in our own species. Upon seeing a
friendly dog, little Johnny’s face may light up; or he may let out an excited
gasp, pointing at the dog; or he may emit a distinctive sound (“Uh!”),
or call out: “doggy!” as he reaches to pet the dog; or he may exclaim:
“Wanna pet the doggy!” perhaps with no reaching. Jonny’s facial expression and his gasp are what are sometimes described as “purely natural”
expressions; whereas his eager reaching and subsequent utterances are
expressive behaviors he voluntarily or perhaps even intentionally engages
in to give vent to his desire to pet the dog. Among the utterances, note,
are English sentences, which have conventional linguistic meaning, and
express in (what Sellars calls) the semantic sense propositions. Still, these
all seem genuine instances of expressive behavior. What renders them so
has to do with similarities among the performances or acts, which equally
serve to give vent to Johnny’s state of mind. These similarities obtain
despite significant differences among the expressive vehicles used. One
can give expression to—express in the mental-state sense—one’s amusement at a joke by laughing (where we may assume that laughter does not
stand in a semantic representational relation to being amused), as well
as by uttering a sentence with a structured meaning, such as “This is so
funny!” We have here similar expressive performances that use different
vehicles of expression.51
An idea that takes its inspiration from earlier philosophical work on
expression, including remarks by Wittgenstein (as well as Ayer) is that
distinctively expressive performances are not merely symptoms, nor even
simply natural signs that indicate the states of mind that cause them.
Unlike verbal descriptions or reports, such behaviors show expressers’

Problems From Kant, Sellars, and Rosenberg, eds. James R. O’Shea and Eric M. Rubenstein (Atascadero, CA: Ridgeview Publishing Co., 2010), 89–106.
51 Wilfrid Sellars, “Language as Thought and as Communication,” Philosophy and
Phenomenological Research 29 (1969): 506–527, distinguishes expressing in the
semantic sense from expressing in the causal and the action senses. Bar-On, Speaking My Mind, distinguishes between an act of expressing and its product, on the
one hand, and between the process and vehicle of expressing, and defends a “neoexpressivist” construal, according to which an avowal such as “I’m so glad to see
you!” “a-expresses” the speaker’s joy at seeing the addressee, using a vehicle that
“s-expresses” the self-ascriptive proposition that the speaker is glad to see her
addressee (see especially Chs. 6–8).

190

Dorit Bar-On

states of mind to a suitably endowed audience, as opposed to hiding
them, on the one hand, and as opposed to intentionally and articulately
telling about the states, on the other.52 On the expresser’s side, the showing behavior relevant to expressing is behavior that springs immediately
from—and directly exhibits, displays, or betrays—the expressed state of
mind, as opposed to simply providing information or giving evidence
about it (the way, e.g., someone taking an aspirin shows that they are
in some kind of pain). On the audience’s side, the relevant contrast is
between behavior that allows some kind of direct recognition or uptake
of the expressed state, as opposed to requiring, say, inference (however
secure) based on various features of the behavior coupled with contextual information and background knowledge.53
2.2 What Non-Verbal Expressive Behavior Shows
Unlike English sentences used in expressing one’s states of mind, animals’
facial and bodily expressions, their calls and other affective displays, are
not expressive vehicles that bear a conventional, representational relation
to the states of mind the animals express.54 Nonetheless, even naturally
expressive behavior can directly manifest various aspects of expressed
states of mind—their quality, degree and intentional objects. A natural
expression can display the location of a pain in the chest, as well as its
severity, rage, as opposed to panic, at a specific attacker, extreme or mild
curiosity at a doll disappearing behind a screen, and so on.55 A dog’s cowering demeanor upon encountering another, or a vervet monkey’s alarm
call, will show to a suitably endowed recipient the kind of state the animal
52 For discussion of the relevant kind of showing (as well as references), see Bar-On,
Speaking My Mind, Ch. 7, 8; Bar-On, “Expressing as Showing”; and Dorit Bar-On,
“Expression: Acts, Products, and Meaning,” in Meaning Without Representation:
Essays on Truth, Expression, Normativity, and Naturalism, eds. Steven Gross et al.
(Oxford: Oxford University Press, 2015). And see James Sias and Dorit Bar-On, “Emotions and Their Expression”, in The Expression of Emotion, eds. Catherine Abell and
Joel Smith (Cambridge: Cambridge University Press, 2016).
53 A related distinction is drawn by ethologists and biologists when they describe animals’
“affective displays” as “merely expressive”, meaning that they are directly tied to, and
directly manifest animals’ affective states. Such displays are contrasted with intentionally produced behaviors that are designed to provide an audience with information
about the producer or her environment.
54 Contra Andrew McAninch, Grant Goodrich, and Colin Allen, “Animal Communication and Neo-Expressivism,” in The Philosophy of Animal Minds, ed. Robert W. Lurz
(Cambridge: Cambridge University Press, 2009), 128–155, who use the present framework to argue that animals’ alarm calls do express propositions.
55 Of course, not all perception-enabling showing is expression. For discussion, see BarOn, Speaking My Mind, 269ff and 298ff. Reflection on the role naturally expressive
behavior plays in the lives of creatures capable of it suggests that part of the effectiveness of such behavior is its capacity to meet with immediate and appropriate reactions
on the part the designated audience, which are likelier to ensue if the behavior enables
perception-like uptake, as opposed to, say, inference about hidden causes of behavior.

Minding the Gap

191

is in, the state’s quality or degree (e.g., how afraid it is), the state’s intentional object (of whom it’s afraid), and the state’s dispositional “profile”
(i.e. how the animal is disposed to act).56 These sorts of performances are
Janus-faced: they point inward—to the animal’s expressed state of agitation, fear, anger, etc.—at the same time as they point outward—toward
the object or event at which the state is directed. And they often reveal
the relevant behavior’s cause or motivation at the same time as they
foretell the expresser’s impending behavior and move others to respond
appropriately.57
Expressive communication is different from—and much less ubiquitous than—animal signals that are designed to convey information
about the producer’s identity, and various biologically significant attributes (such as readiness to mate, or fitness). Inasmuch as expressive
performances are keyed to objects and features of an animal’s environment as apprehended by the animal (and in that sense “psychologically
filtered”), they contrast with automatic physiological reactions and
hormonally triggered behavioral changes, and may be said to exhibit
a measure of intentionality or subjective directedness, even if not produced intentionally. And in contrast with perceptual and other, more
passive states, which are also often said to exhibit intentionality, expressive communication also has an active dimension. A creature giving
behavioral expression to a present state of mind—e.g., a dog bowing
playfully—shows designated receivers how he is disposed to act, as well
as how they should act or what to do. Moreover, unlike rote, automatic,
instinctive, or reflexive behaviors, expressive behaviors of a wide range
of animal species can be brought under voluntary control, intensified or
toned down.58 Such control prefigures the sorts of intentional production of which humans are capable.59
56 For discussion of play bows, see A. Miklosi, J. Topál, and V. Csánvi, “Comparative Social
Cognition: What Can Dogs Teach Us?” Animal Behavior 67(6) (2003): 995–1004.
57 For an early occurrence of the idea that expressive behavior shows what’s within while
pointing to what’s without, see Alan Tormey, The Concept of Expression (Princeton:
Princeton University Press, 1971), 27f. and passim. The Janus-face character discussed
here is different from the dual force ascribed by Millikan to “pushmi-pullyu” representations. See Ruth Millikan, “On Reading Signs: Some Differences Between Us and
the Others,” in Evolution of Communication Systems: A Comparative Approach, eds.
D. Kimbrough Oller et al. (Cambridge: MIT Press, 2004), 15–29; and Ruth Millikan,
Varieties of Meaning: The 2002 Jean-Nicod Lectures (Cambridge: MIT Press, 2004).
58 There is considerable experimental evidence that the production of alarm and other
calls, as well as other expressively communicative gestures, can be brought under control in all primates, many mammals, and even birds; there is also evidence of various
flexible “audience effects” in the production of calls in a number of species. For more on
this, see Fitch, Evolution of Language, section 4.9.3; Charles T. Snowdon, “Contextually Flexible Communication in Non-human Primates,” in Evolution of Communicative
Flexibility: Complexity, Creativity, and Adaptability in Human and Animal Communication, eds. D. Kimbrough Oller et al. (Cambridge: MIT Press, 2008), 71–93; and see
Bar-On “Origins of Meaning,” Section 4 for discussion (and additional references).
59 See Bar-On, “Origins of Meaning,” and “Expressive Communication”.

192

Dorit Bar-On

This should help give a sense of ways that naturally expressive behavior
can foreshadow linguistic behavior. But more can be said. Unlike linguistic utterances, episodes of naturally expressive behavior do not deploy
structured vehicles; the vehicles they deploy (facial contortions, vocalizations, bodily postures, etc.) do not seem to have recombinable parts or
components. Even so, natural expressive vehicles (as well as acquired
nonverbal vehicles) have various dimensions of complexity. They inherit
their complexity from the complexity of expressed psychological states.
As Sellars helpfully observes, a single state, which may not have any distinct parts or components corresponding to referential or predicative
parts of speech, may nevertheless have both a predicative and a characterizing function by virtue of its multiple aspects rather than its distinct
parts.60 The relevant psychological states could be understood as nonpropositional affective and action-guiding states that are directed at (or
are “about”) certain environmental objects: fear of x, anger/excitement
at y, attending to z. Moreover, these prepositional attitudes (as I refer
to them) can be usefully thought as prefiguring the propositional attitudes.61 And, importantly, to the extent that expressive signals transparently reveal aspects of the complex states they express, their use in
communication can be seen as foreshadowing the use of articulate, linguistic vehicles, despite the fact that they, like the states they are used to
express, lack composite structure.
2.3 Animals’ Expressive Behavior as Communicative,
Overt, and Social
Continuity skeptics often mention animals’ expressive behaviors only to
dismiss them as candidate forerunners of the symbolic utterances used
in linguistic communication.62 Yet I argue that, properly understood,
expressive communication possesses significant features that can be
seen to represent natural precursors of certain psychological, semantic

60 To illustrate, suppose “a” refers to a, “b” to b, italicization represents something as red,
bold font represents something as blue and one symbol being to the left of the other
represents its being larger than the other. On Sellars suggestion, the complex symbol
“ab” shares the propositional but not the logical (compositional) form of the sentence
“Red a is larger than blue b.” For more on this, see Jay Rosenberg, Wilfrid Sellars: Fusing the Images (Oxford: Oxford University Press, 2007), 105ff.
61 See Tormey, Concept of Expression, 10f.
62 See, e.g., John McDowell, “Meaning, Communication, and Knowledge,” in Philosophical Subjects: Essays on the Work of P. F. Strawson, ed. Zak van Straaten (Oxford:
Clarendon Press, 1983), 117–139; Bennett, Linguistic Behavior, §62; and Davidson,
Subjective, Intersubjective, Objective, 129; Anderson, Dolittle’s Delusion, Chapters 2,
7, and passim; James Hurford, The Origins of Meaning: Language in the Light of
Evolution (Oxford: Oxford University Press, 2007), Chapter 6 and passim; and Fitch,
Evolution of Language, Chapters 4 and 13.

Minding the Gap

193

and pragmatic aspects of linguistic communication. Elsewhere, I have
argued that animals’ expressive communication is best thought of as a
form of world-directed, overt, intersubjective, and social communicative behavior that is designed by nature (as opposed to being designed
through individual intention or culture) to show, or exhibit the presence
and character of expressers’ states of mind to suitably endowed observers, so as to move them to act in appropriate ways (toward the expresser
or the object of her expressed state), in part by foretelling the expresser’s
impending behavior.63 If I am right, expressive communicative interactions go beyond the merely discriminative, responsive behaviors of pure
triangulation, yet they do not require propositional thought (on the part
of expressers) or reflective interpretation (on the part of the audience).

3. Expressive Communication and Intermediate
Triangulation64
3.1 Intermediate Triangulation
Going back to Davidson’s triangles, suppose, briefly, that S1 produces an
alarm call, which is naturally designed to show conspecifics, S2 included,
his imminent flight from some specific type of nearby threat (some predator O), so as to move S2 to do the same.65 Having observed the behavior,
S2 is in a position to respond to it in some way that is not merely responsive to the presence of O (as indicated by the behavior) but is also anticipatory of S1’s subsequent behavior. Instead of also fleeing, for example,
S2 may, upon hearing the alarm call and spying no predator, respond
to S1’s alarm call by, say, moving toward S1 to consume S1’s soon-tobe-abandoned meal. When S1’s O-behavior betrays his impending flight,
the possibility opens for S2’s response to the behavior to match it or
not, depending on whether or not S2’s response is itself a bit of behavior
appropriate to the presence of O. S2’s behavior departs both from S1’s
(anticipated) behavior and from the appropriate responsive behavior;
and it can be said to embody O-related disagreement with S1’s behavior.
Here it looks as though S2 is treating S1’s O-related behavior as separable
from the (imminent) presence of O (as assessed by S2). S2 is keeping two

63 See Bar-On and Green, “Lionspeak”; Bar-On, “Origins of Meaning”; and Dorit Bar-On,
“Communicative Intentions, Expressive Communication, and Origins of Meaning,” in
Routledge Companion to the Philosophy of Animal Minds, eds. Kristin Andrews and
Jacob Beck (New York and Oxford: Routledge, 2018), 301–312.
64 For fuller discussion, see Bar-On and Priselac, “Triangulation and the Beasts.”
65 For present purposes, we need not attribute to S1 and S2 the concept PREDATOR, but
only whatever discriminatory dispositions vis-à-vis O that Davidson allows subjects
to have in pure triangulation. See again, e.g., Davidson, “Externalisms,” 12–13; and
Davidson, Subjective, Intersubjective, Objective, 117ff.

194

Dorit Bar-On

distinct but simultaneous tabs, as it were, on the world and on S1’s reaction to it. The right space seems to be open for crediting S2 with treating
S1 as having his own take on the situation. For O is no longer merely an
external cause serving as a point of intersection of S1 and S2’s discriminatory responses; nor is S1’s behavior treated merely as a natural O-indicator by S2. Instead, S2’s responsive behavior is one that takes account of
what amounts to S1’s getting things wrong (from S2’s perspective). S2, if
you will, minds a gap between S1’s take and how things are.
We have here what we may call “intermediate triangulation”, which a
proponent of continuity can interpose diachronically between Davidson’s
pure and reflective triangulations. Intermediate triangulation is poised
between the “thin”, pure case, in which S2 simply responds or fails to
respond to S1’s O-behavior with her own O-behavior, on the one hand,
and the “thick,” reflective case in which S2 judges that S1’s O-behavior
is incorrect (specifically, that S1 has uttered a false sentence, betraying
a false belief), on the other hand. Intermediate triangulation features
object-centered intersubjective interactions that allow for disagreement
but do not presuppose reflective grasp of objectivity or even possession of
propositional thought on the part of the relevant subjects. So it is clearly
different from the thick, reflective case. But it is also different from the
thin case. To see the crucial difference, recall that, on the expresser’s side,
expressive behavior shows the affective state the expresser is in partly
by revealing how he, S1, is prepared to react in light of O’s presence.
Intersubjective Object-Directed Relations in
“Intermediate” Triangulation

O

S1’s O-centered
expressions

S1

S2’s O-centered
expressions

Production/uptake of
O-centered expressive behavior
‘embodied’ affective disagreement

Figure 8.3 “Intermediate” Triangulation

S2

Minding the Gap

195

Faced with S1’s expressive performance, the observer, S2, can anticipate
not only O’s presence, but also the behavior on S1’s part that the performance foretells. In that sense, S2 is responding to S1’s expressive performance as O-centered behavior. Moreover, since expressive behavior
also has the function of moving suitably attuned observers to behave in
certain ways, proper uptake of S1’s expressive behavior requires a certain
O-related reaction on S2’s part. This interlocking of O-centered intersubjective interactions makes room for a broader range of intersubjective
mismatches/disagreements than Davidson allows in the pure case.
3.2 Some Synchronic Evidence?
The above scenario (like the Davidsonian ones after which it is modeled)
is an imaginary one. But recent observations of animals’ intersubjective
interactions in the wild, as well as in captivity, and controlled experiments comparing the behaviors of higher primates and prelinguistic children, suggest that intermediate triangulation is more than mere fiction.66
And, although there is much disagreement among researchers regarding the degree to which the relevant interactions resemble adult human
transactions, it is uniformly agreed that the interactions do not admit of
any simple explanation in terms of conditioned reflexes, innate or rigidly
determined species-wide signaling, or learning history. Just for one example, Crockford et al.67 report recent triangulation-relevant experiments
with chimpanzees in the wild, who emit snake calls highly selectively,
exhibiting fine-tuned sensitivity to whether or not the call receivers have
themselves seen the snake, whether they have previously been within earshot of the call, how far away they are relative to the caller, and how
affiliated they are with the caller. While it may be debatable whether the
callers are mindreaders who “assess the state of knowledge” of the receivers (as the authors suggest), it seems undeniable that the callers are highly
attuned, specifically, to other subjects’ attention to—as well as impending
behavior toward—a salient object of potential mutual interest or significance, as evidenced by the intricate pattern of their call production. And
the call receivers are moved to take specific actions to avoid a threat
that is perceived by the caller but invisible to them and of which the
call informs them, carefully skirting the path where the threat (and the
caller) is located. Also relevant are experiments showing that Capuchin
monkeys, for example, “when shown a grasping hand disappear behind a
screen without having been shown if there is or not anything behind the
screen” are “[surprised] if upon lifting the screen, no object was found

66 For relevant references, see Bar-On, “Origins of Meaning.”
67 Catherine Crockford et al., “Wild Chimpanzees Inform Ignorant Group Members of
Danger,” Current Biology 22 (2011): 142–146.

196

Dorit Bar-On

behind” (where surprise is measured using standard criteria applied to
pre-verbal infants).68
3.2 A “Diachronic Bridge”
Although I have accepted the continuity skeptic’s claim of synchronic
discontinuity, I have rejected the skeptic’s claim of diachronic discontinuity. In response to the skeptic’s challenge, I have explained how we could
begin to draw the outlines of a bridge across the diachronic gap between
us and our non-human ancestors.
The response I have offered is intended to represent a middle-ground
between representationalists and anti-representationalists responses to
continuity skeptics. In contrast to representationalists, I have not assumed
that creatures engaged in what Davidson describes as “pure triangulation” already have the wherewithal for the kind of objective thought that
intersubjectively engaged language speakers are capable of. Intermediate triangulation, as I have characterized it, is the prerogative of creatures capable of a special kind of minded, intersubjective, world-directed
engagements and interactions that, however, do not yet exemplify all the
key trappings of human objective, propositional thought. At the same
time, in contrast to anti-representationalists, I have argued that, properly
construed, expressive communication of the sort we share with existing
non-human animals does possess characteristics that foreshadow the sort
of linguistic communication of which creatures like us are capable. It is
thus apt to aid us in our search for a diachronic bridge of the sort the
continuity skeptic claims to be impossible.69

4. Continuity Skepticism and Mind-Mind Dualism
In outlining a “straight” response to the continuity skeptic, my strategy
has been to try to explain how the diachronic gap between us and our
non-human ancestors could have been bridged, even given the palpable
synchronic distance separating us from existing non-human animals.
But what would be the significance of such a diachronic bridge for our
philosophical understanding of the relationship between the minds of
reflective beings like us and the minds of non-human animals? By way of
conclusion, I would like to offer a few suggestive remarks.70

68 Juan Carlos Gómez, “Embodying Meaning: Insights From Primates, Autism, and Brentano,” Neural Networks 22(2) (2009): 195.
69 For relevant discussion of the need for an intermediary stage involving capacities that
foreshadow without exemplifying the capacities present in the reflective case, see BarOn, “Crude Meaning.”
70 In earlier work, I was concerned to argue that it is wrong to think that explaining
(as opposed to explaining away) certain puzzling phenomena—notably, so-called

Minding the Gap

197

Continuity skeptics often support their view by marshalling considerations concerning the distinctive nature of our own Mind—notably,
its essentially normative, reason-responsive, self-conscious, and “selfgoverned”, reflective character. Given that character, they invite us to
recognize that our Mental states (properly so described) have essential,
constitutive connections to the possession and exercise of certain distinctive capacities that non-human animals (and even preflective humans,
NB) clearly lack.71 Different proponents emphasize different constitutive
capacities; but they share the idea that the appearance of a genuinely
mental common ground between human and non-human minds is illusory. So the reality to which our mental attributions point is bifurcated:
our minds and the minds (such as they are) of non-human animals constitute substantially distinct kinds. At best, the so-called minds of nonhuman animals simply belong to a different species of mind, so that any
attributions of mental states to non-human animals must be made “in a
different register”. The result is (what I have elsewhere dubbed) Mindmind dualism.72
A familiar complaint against Mind-mind dualism is that it circumscribes the realm of the Mental too artificially narrowly. No (properly)
Mental states can be attributed to nonreflective non-humans, except

first-person authority—requires embracing constitutivism regarding mentality, with its
attendant Mind-mind dualism (see below). My interest here has been different, though I
come by it via the positive “neo-expressivist” view of first-person authority that I favor.
For relevant discussion, see Bar-On, Speaking My Mind, Ch. 7.
71 I will here set aside the case of prereflective humans, which may require a separate
treatment. Clearly, however, anyone who (like Davidson) takes language to be coeval
with mindedness owes us a story about what sets apart the prereflective from the unreflective, so that we could at least make sense of emergence within our own species
(i.e., in ontogeny), even if we become convinced by skepticism concerning evolutionary
emergence.
72 The general view is articulated in, inter alia, Brandom, Making It Explicit; Robert Brandom, Articulating Reasons (Cambridge: Harvard University Press, 2000); McDowell,
Mind and World; John McDowell, Having the World in View Having the World in
View: Essays on Kant, Hegel, and Sellars (Cambridge: Harvard University Press,
2013); Christine Korsgaard, Self-Constitution: Agency, Identity, and Integrity (Oxford:
Oxford University Press, 2009); Matthew Boyle, “Essentially Rational Animals,” in
Rethinking Epistemology, vol. 2, eds. Günter Abel and James Conant (Berlin: Walter
de Grutyer Verlag, 2012), 295–428; and Eric Marcus, Rational Causation (Cambridge,
MA: Harvard University Press, 2012). Relevant are also discussions in, e.g., Sydney
Shoemaker, The First-Person Perspective and Other Essays (Cambridge: Cambridge
University Press, 1996); Paul Boghossian, “What the Externalist Can Know A Priori,”
Proceedings of the Aristotelian Society 97 (1997): 161–175; Crispin Wright, “SelfKnowledge: The Wittgensteinian Legacy,” Royal Institute of Philosophy Supplement 43
(1998): 13–45; Akeel Bilgrami, Self-Knowledge and Resentment (Cambridge: Harvard
University Press, 2012); Richard A. Moran, Authority and Estrangement: An Essay on
Self-Knowledge (Princeton: Princeton University Press, 2001); and Jane Heal, “On First
Person Authority,” The Aristotelian Society 102(1) (2002): 1–19.

198

Dorit Bar-On

“in a different register”. (And it is far from clear how such states could
be attributed to prereflective humans in the same register as they are
attributed to us.) But even as regards fully formed reflective humans, we
must exclude unconscious emotions, wishes, and thoughts, feelings that
are unnoticed or not attended to, and so on. Worse, it would seem that
many of the conscious states of adult humans—brute sensations, irritations, passing thoughts and perceptions, non-specific anxieties, intrusive cravings and impulsive wants and whims, irrational judgments,
beliefs and desires, and so on—would all need to be relegated to a second class, “lower case” mental status. Indeed, some Mind-mind dualists
explicitly restrict their mind-related theses to what they term rational
beliefs and intentions, “judgment-sensitive” wishes, desires, and preferences, intentional states understood as commitments, and so on.73 But,
to the extent that they take only such states of mind to belong in the
category of the Mental properly so-called, they still seem committed
to a rather extreme bifurcation within the mental lives of even adult
human beings. For, they still require separating human psychology into
two distinct realms: the “merely psychological” realm, the denizens of
which are at best dispositions, or passive-responsive occurrences that
only have powers to move us causally (or perhaps teleologically), on
the one hand, and the genuinely Mental realm, the denizens of which
all lie within “the space of reasons”, which is normatively governed and
reflective, on the other.74

73 See, e.g., Moran, Authority and Estrangement, xxxiii, and Bilgrami, Self-Knowledge
and Resentment, p. 1. Alternatively, one can insist on judgment-sensitivity and reasonresponsiveness as necessary conditions on a state being one of belief, intention, desire, etc.
74 Kevin Cahill has suggested to me that since McDowell, in particular, is not committed
to an overly intellectualized view of what it is to be in the space of reasons, this may
exempt him (and other Mind-mind dualists, so called) from the charge of introducing
a bifurcation within our own mentality. However, whether or not this is so will obviously depend on McDowell’s account of what it takes to get into (or get “caught in”)
the space of reasons. This is a complicated issue that cannot be taken up here. Also,
several of the authors mentioned here are explicitly committed to a separation between
e.g. “judgment-sensitive” mental states and passive states of mind that we find ourselves
in, reserving the epithet “(genuinely) mental” for the former. (See above, footnotes 72
and 73.) Furthermore, even if being in the space of reasons does not require actively
reflecting on or explicitly considering relevant reasons, it remains the case that (on the
view in question) states that are within the space of reasons must be at least responsive
to reasons. And this would seem to exclude a wide range of psychological states (notably, ones we appear to share with prereflective and non-reflective creatures, as well as
reflectively challenged adult humans). Proponents of the view could, in turn, play down
what is involved in being responsive to reasons. (At one extreme would lie the idea that
being responsive to reasons is just a matter of being disposed to act—or being liable to
change—in ways that accord with such reasons.) This may help forestall the charge of
Mind-mind dualism; but by the same token, it risks undermining the claim that there
is a genuine gulf between our minds and those of (even existing) non-human animals.
Needless to say, the issue merits additional and more careful discussion.

Minding the Gap

199

If Mind-mind dualism is to fare better than its half-namesake—MindBody dualism—it seems that it would need to tell us what to make of the
following sort of seemingly intelligible questions: How could Minded
creatures come to be in nature? Or (as one might put it): How could
“rational souls” emerge in a natural world populated only with animal
souls? How, within our species, can/do incipient Minds emerge? And
how, within each of us, can the “merely psychological” part of our soul
co-exist and be integrated (causally, as well as rationally) with our reflective mind?75 My attempt to meet the continuity skeptic’s challenge was
offered in the spirit of trying to make philosophical space for raising, and
answering, these questions.
Thus, recall Davidson’s claim that the question of “the emergence of
mental phenomena” is the “conceptual problem . . . of describing the
early stages in the maturing of reason . . . that precede the situation in
which [mentalistic] concepts have clear application”.76 Davidson, we
saw, argues that “[t]here cannot be a sequence of emerging features of
the mental”, since “we lack . . . a satisfactory vocabulary for describing
the intermediate steps”.77 Unlike other “straight” responses to continuity
skepticism, my response did not depend on appealing to relevant sciences
that purportedly show creatures engaged in pure triangulation already
to possess central features of the kind of objective thought of which
reflective interpreters are capable.78 But I have also not simply taken our
uncritical commonsense attributions of mentality to non-human animals
at face value. Instead, I have accepted the skeptic’s understanding of the
crucial differences between pure and reflective triangulation, and tried
to engage him by offering a potential candidate for an intermediate case.
If I am right, the domain in which to look for Mind-mind continuity
is not that of animals’ intelligent problem solving, or self-initiated pursuit of individual goals, or strategic manipulation of the world around
them (other individuals included). I have instead proposed, on the one
hand, that we should focus on the specific character, distinctive texture,
and natural significance of expressive behavior, and the kind of communication it affords: its intersubjective world-directedness, overtness,
essential sociality, and connection to action. And, on the other hand, I
75 Where this question is to be read on analogy with the question: how can an animal’s
“nutritive soul” co-exist and be absorbed into its “animal soul?” It’s worth noting
that contemporary Mind-mind dualists have in fact acknowledged the need to articulate their view in ways that could potentially allow responses to these questions. See
especially Korsgaard, Self-Constitution; Boyle, “Essentially Rational Animals”; and
Marcus, Rational Causation.
76 Davidson, Subjective, Intersubjective, Objective, 127.
77 Ibid.
78 See, e.g., Burge, Origins of Objectivity; and Peter Carruthers, “Invertebrate Concepts
Confront the Generality Constraint (and Win),” in The Philosophy of Animal Minds,
ed. Robert W. Lurz (Cambridge: Cambridge University Press, 2009).

200

Dorit Bar-On

have emphasized the significance of the fact that such behavior (unlike
informative signaling) is not ubiquitous in the animal kingdom, but is
characteristic of only some non-human animal species. And, moreover, it
provides a common ground between non-human animals and the young
members of our species, as well as our own fully formed selves.
It is the overt showing and direct uptake of states of mind distinctive
of social-expressive animals that, I have argued, underwrites a capacity
for non-reflective grasp of others’ “subjective take,” and thus, despite
being non-reflective, marks the dawn of a different kind of orientation. It
is an orientation that involves a more active, correctible, even incipiently
critical, stance toward the world, others, and oneself. No doubt, socialexpressive souls still lack features thought to be the hallmarks of rational
souls. But, for all that, they may possess what it would have taken to
cross one significant threshold that separates parts of the non-reflective
animal world from the reflective parts that we adult humans occupy.

References
Alston, William. “Expressing.” In Philosophy in America, edited by Max Black,
15–34. Ithaca: Cornell University Press, 1965.
Anderson, Stephen R. Dr. Dolittle’s Delusion: Animals and the Uniqueness of
Human Language. New Haven: Yale University Press, 2004.
Bar-On, Dorit. Speaking My Mind: Expression and Self-Knowledge. Oxford:
Clarendon Press, 2004.
———. “Expressing as Showing What’s Within: On Mitchell Green’s SelfExpression.” Philosophical Books 51(4) (2010): 212–27.
———. “Expressive Communication and Continuity Skepticism.” Journal of Philosophy 110(6) (2013): 293–330.
———. “Origins of Meaning: Must We ‘Go Gricean’?” Mind and Language
28(3) (2013): 342–75.
———. “Expression: Acts, Products, and Meaning.” In Meaning Without Representation: Essays on Truth, Expression, Normativity, and Naturalism, edited
by Steven Gross, Nicholas Tebben, and Michael Williams, 180–209. Oxford:
Oxford University Press, 2015.
———. “Sociality, Expression, and This Thing Called Language.” Inquiry 59(1)
(2016): 56–79.
———. “Communicative Intentions, Expressive Communication, and Origins
of Meaning.” In Routledge Companion to the Philosophy of Animal Minds,
edited by Kristin Andrews and Jacob Beck, 301–312. New York and Oxford:
Routledge, 2018.
———. “Crude Meaning, Brute Thought (Or: What Are They Thinking?!).” Journal for the History of Analytic Philosophy (forthcoming).
Bar-On, Dorit and Mitch Green. “Lionspeak: Communication, Expression, and
Meaning.” In Self, Language, and World: Problems From Kant, Sellars, and
Rosenberg, edited by James R. O’Shea and Eric M. Rubenstein, 89–106. Atascadero, CA: Ridgeview Publishing Co., 2010.
Bar-On, Dorit and Matthew Priselac. “Triangulation and the Beasts.” In Triangulation From an Epistemological Point of View, edited by Maria Cristina
Amoretti and Gerhard Preyer, 121–152. Berlin: Ontos Verlag, 2011.

Minding the Gap

201

Bell, Charles. Essays on the Anatomy and Philosophy of Expression. London:
John Murray, 1824.
Bennett, Jonathan. Rationality: An Essay Towards an Analysis. New York: Routledge & Kegan Paul, 1964.
———. Linguistic Behavior. Cambridge: Cambridge University Press, 1976.
Bickle, John. Philosophy and Neuroscience: A Ruthlessly Reductive Account.
Norwell, MA: Kluwer Academic Publishers, 2003.
Bilgrami, Akeel. Self-Knowledge and Resentment. Cambridge: Harvard University Press, 2012.
Boghossian, Paul. “What the Externalist Can Know A Priori.” Proceedings of the
Aristotelian Society 97 (1997): 161–75.
Boyle, Matthew. “Essentially Rational Animals.” In Rethinking Epistemology,
vol. 2, edited by Günter Abel and James Conant, 295–428. Berlin: Walter de
Grutyer Verlag, 2012.
Brandom, Robert. Making It Explicit. Cambridge: Harvard University Press,
1994.
———. Articulating Reasons: An Introduction to Inferentialism. Cambridge:
Harvard University Press, 2000.
Burge, Tyler. Origins of Objectivity. Oxford: Oxford University Press, 2010.
Carruthers, Peter. “Invertebrate Concepts Confront the Generality Constraint
(and Win).” In The Philosophy of Animal Minds, edited by Robert W. Lurz,
89–107. Cambridge: Cambridge University Press, 2009.
Chemero, Anthony. Radical Embodied Cognitive Science. Cambridge: MIT Press,
2009.
Churchland, Patricia. Neurophilosophy. Cambridge: MIT Press, 1986.
Churchland, Paul. “Eliminativism and Propositional Attitudes.” Journal of Philosophy 78 (1981): 67–90.
Clark, Andy. Supersizing the Mind: Embodiment, Action, and Cognitive Extension. New York: Oxford University Press, 2008.
Crockford, Catherine, Roman Wittig, Roger Mundry, and Klaus Zuberbühler.
“Wild Chimpanees Inform Ignorant Group Members of Danger.” Current Biology 22 (2011): 142–46.
Darwin, Charles. The Expression of the Emotions in Man and Animals. London:
John Murray, 1872.
Davidson, Donald. “Comments on Karlovy Vary Papers.” In Interpreting Davidson, edited by Petr Kot̓átko, Peter Pagin, and Gabriel Segal, 285–308. Stanford:
CSLI Publications, 2001.
———. “Externalisms.” In Interpreting Davidson, edited by Petr Kot̓átko, Peter
Pagin, and Gabriel Segal, 1–16. Stanford: CSLI Publications, 2001.
———. Subjective, Intersubjective, Objective. New York: Oxford University
Press, 2001.
———. Problems of Rationality. New York: Oxford University Press, 2004.
Dennett, Daniel. The Intentional Stance. Cambridge: MIT Press, 1987.
———. “The Evolution of ‘Why’?” In Reading Brandom: On Making It Explicit,
edited by Bernhard Weiss and Jeremy Wanderer, 48–62. New York: Routledge,
2010.
Dummett, Michael. Origins of Analytical Philosophy. London: Duckworth and
Cambridge: Harvard University Press, 1993.
Eilan, Naimi. “Joint Attention, Communication, and Mind.” In Joint Attention:
Communication and Other Minds, edited by Naomi Eilan, Christoph Hoerl,

202

Dorit Bar-On

Teresa McCormack, and Johannes Roessler, 1–34. Oxford: Oxford University
Press, 2005.
Fitch, W. Tecumseh. The Evolution of Language. Cambridge: Cambridge University Press, 2010.
Gómez, Juan Carlos. “Embodying Meaning: Insights From Primates, Autism, and
Brentano.” Neural Networks 22(2) (2009): 190–96.
Green, Mitchell. Self-Expression. Oxford: Oxford University Press, 2007.
Grice, Paul. “Meaning.” The Philosophical Review 66 (1957): 377–88.
Heal, Jane. “On First Person Authority.” The Aristotelian Society 102(1) (2002):
1–19.
Hurford, James. The Origins of Meaning: Language in the Light of Evolution.
Oxford: Oxford University Press, 2007.
Hurley, Susan and Matthew Nudds, ed. Rational Animals? New York: Oxford
University Press, 2006.
Hutto, Daniel. Folk Psychological Narratives: The Sociocultural Basis of Understanding Reasons. Cambridge: MIT Press, 2008.
Korsgaard, Christine. Self-Constitution: Agency Identity, and Integrity. Oxford:
Oxford University Press, 2009.
Lurz, Robert. The Philosophy of Animal Minds. Cambridge: Cambridge University Press, 2009.
Marcus, Eric. Rational Causation. Cambridge: Harvard University Press, 2012.
Maynard Smith, Jonathan, and David Harper. Animal Signals. Oxford: Oxford
University Press, 2003.
McAninch, Andrew, Grant Goodrich, and Colin Allen. “Animal Communication
and Neo-Expressivism.” In The Philosophy of Animal Minds, edited by Robert
W. Lurz, 128–55. Cambridge University Press, 2009.
McDowell, John. “Meaning, Communication, and Knowledge.” In Philosophical
Subjects: Essays on the Work of P. F. Strawson, edited by Zak van Straaten,
117–39. Oxford: Clarendon Press, 1983.
———. Mind and World. Cambridge: Harvard University Press, 1996.
———. “What Myth?” Inquiry 50(4) (2007): 338–51.
———. Having the World in View: Essays on Kant, Hegel, and Sellars. Cambridge: Harvard University Press, 2013.
Miklosi, Adam, József Topál, and Vilmos Csánvi. “Comparative Social Cognition: What Can Dogs Teach Us?” Animal Behavior 67(6) (2003): 995–1004.
Millikan, Ruth. “On Reading Signs: Some Differences Between Us and the Others.” In Evolution of Communication Systems: A Comparative Approach,
edited by Kim Oller, Kim Plunkett, and Ulrike Griebel, 15–29. Cambridge:
MIT Press, 2004.
———. Varieties of Meaning: The 2002 Jean-Nicod Lectures. Cambridge: MIT
Press, 2004.
Moran, Richard A. Authority and Estrangement: An Essay on Self-Knowledge.
Princeton: Princeton University Press, 2001.
O’Regan, J. Kevin, and Alva Noë. “A Sensorimotor Account of Vision and Visual
Consciousness.” Behavioral and Brain Sciences 24 (2001): 939–1031.
Quine, Willard Van Orman. Word & Object. Cambridge: MIT Press, 1960.
Rosenberg, Jay. Wilfrid Sellars: Fusing the Images. Oxford: Oxford University
Press, 2007.

Minding the Gap

203

Sellars, Wilfrid. “Empiricism and the Philosophy of Mind.” In Minnesota Studies in the Philosophy of Science, vol. 1, edited by Herbert Feigl and Michael
Scriven, 253–329. Minneapolis: University of Minnesota Press, 1956.
———. “Language as Thought and as Communication.” Philosophy and Phenomenological Research 29 (1969): 506–27.
———. “Mental Events.” Philosophical Studies 39(4) (1981): 325–45.
Shoemaker, Sydney. The First-Person Perspective and Other Essays. Cambridge:
Cambridge University Press, 1996.
Snowdon, Charles T. “Contextually Flexible Communication in Non-human Primates.” In Evolution of Communicative Flexibility: Complexity, Creativity, and
Adaptability in Human and Animal Communication, edited by D. Kimbrough
Oller and Ulrike Griebel, 71–93. Cambridge: MIT Press, 2008.
Tormey, Alan. The Concept of Expression. Princeton: Princeton University Press,
1971.
Wittgenstein, Ludwig. Philosophical Investigations, 3rd ed., translated by G. E.
M. Anscombe. Oxford: Basil Blackwell, 1968.
———. Remarks on the Philosophy of Psychology, vol. I, edited by G. E. M.
Anscombe and G. H. von Wright, translated by G. E. M. Anscombe. Oxford:
Basil Blackwell, 1998.
———. Remarks on the Philosophy of Psychology, vol. II, edited by G. H. von
Wright and Heikki Nyman, translated by C. G. Luckhardt and M. A. E. Aue.
Chicago: University of Chicago Press, 1998.
Wright, Cripsin. “Self-Knowledge: The Wittgensteinian Legacy.” Royal Institute
of Philosophy Supplement 43 (1998): 13–45.

9

Rational Animals
Julia Tanney

1. Our Boys
Our day starts at just before 7:00 am in time to get the 6:00 start of the
Today Show on BBC’s Radio 4. My husband gets up to make tea, as I
doze in and out of sleep. He arrives with two mugs, sets his own on the
dresser a few steps away from the bed, and walks around to put mine at
my side. As he walks to the dresser to fetch his own, the dog, who had
been fast asleep at the foot of the bed, gets up and plops himself down
on my husband’s pillow. When my husband turns around with his mug
of tea, he gives the dog a look. Then he sits at the side of the bed. I turn
around and watch the unfolding scene, which I know so well.
“What do you think you are doing?”
No reply.
“Um, do you think you might move over so that I can get back in
bed?”
No reply.
“Solutré”.
The dog looks up.
Then, in choreographed maneuver that has developed over the years,
as my husband slides carefully into bed, I roll the dog over toward the
middle, where he now lies on his back, fully ensconced between us. His
four paws are up in the air as he relishes a long massage à deux on his
tummy and his paws (from both of us) and lots of kisses on his muzzle
(from me). Then, he scrambles under the blankets for a few minutes, finds
a place that maximizes body contact, and finally, when it gets too hot,
goes back to his place at the foot of the bed.
The cat, meanwhile, had gotten up as soon as my husband had set
off for the kitchen, following him to request his breakfast. I can just
make out their conversation from the bedroom. Another daily ritual
occurs there: Vinzelles will not, on these occasions, jump onto the
counter to get his food as he does on others. My husband is to pick

Rational Animals

205

him up, give him a cuddle, and then plop him in front of his plate.
If we stay in bed long enough, the cat wanders back to the bed and
creeps slowly up my body until he greets me face to face. We then
engage in a long amorous nose-butting routine before he lies down
on my chest with our heads turned in the same direction as we doze
together cheek to cheek.
“Vous avez une relation très fusionnelle avec votre chat, Madame”, a
veterinarian told me one day in her office, as I stood there calmly with
the cat wrapped around my shoulders as the dog was being examined.
She is right: Vinzelles—an Egyptian Mau—is my teddy bear, alive, warm
and purring. We go to sleep using each other as pillows, and we wake up
in the same position.
Just as well, because my husband has une relation très fusionnelle with
the dog. Solutré—a Schipperke—knows how to get an emotional reaction out of my husband: he gets under his skin, as it were, in a way which
leaves me amazed, amused, and admiring.
We talk to our boys and they talk back. They make requests for cuddles, they demand food, and they play games. They both go to the same
spot and demand, in unison, their daily imported freeze-dried venison
tongue. One of Solutré’s favourite games is to take a piece from my hand,
and then to toss it into the air and let it fall to the ground. He sits, he
watches, and he waits.
“Doesn’t he want it?” my husband asks.
“Yes, of course he does”.
(In fact, he would have sniffed it and walked away with his head in the
air if he had not. He will ask for it, but will refuse to take it from me in
the morning, for example.)
“He is waiting from someone to play his game with him”, I continue.
“Well, you play it, then. It’s your game”.
“Solutré”, I say, “is that for me? Are you going to give that to me?”
I reach down to take the piece of tongue, and the dog growls. I put my
face next to his, and I growl back. He growls louder and bares his teeth.
I do the same. I reach out again and the dog snatches the morsel in his
mouth and begins to chomp. We play this game, with variations on the
theme, almost every single day. Sometimes I can grab it away, sometimes
he snatches it first.
We are constantly asking the dog (and each other) what he wants when
he paws our knees, jumps on our laps, and starts barking. He tells us
what he wants when he finds his lead, or fetches our shoes. We apologize
to him after having insisted that he need not go out but he has proved
otherwise. After all, he is the authority on this particular matter.

206

Julia Tanney

Solutré knows with (what seems at times to be) an uncanny ability that
we are going out. At first, this ability seemed to be partly grounded in
word-associations: he picked up ‘walk’ very young and eventually mastered ‘perambulation’ which—even if you say the word alone—still elicits
a response. Recently, the amassing of empty wine bottles has started to
provoke a reaction and my husband has been overheard to say “I know
you know that I am going to the tip. And yes, you’re coming too”. I
learned belatedly that the dog understood my putting on make-up to
mean that I was about to go out. When he was a puppy, he would stick
his nose up to my eyes and mouth and sniff my mascara and my lipstick.
Then he would start prancing around, tail wagging, in a way that suggested that something interesting was about to happen. I only fully took
this on board when I realized that he seemed to think I was about to go
out when in fact I was not. My putting on a coat excites a similar reaction. Reflecting on what made him mistakenly believe I was about to
leave, I remembered that he had not only smelled my make-up, but he
had also been watching me. Perhaps I spent more time in front of the
mirror than normal? Was it the high-heels? So does the cat seem to know
when I am about to take the dog out for his last walk of the day, for he
appears at the front door at just the right time. Perhaps he will, perhaps
he will not let me to pick him up and wrap him around my shoulders as
we go for our perambulation around the village.
This is but a glimpse into our lives with our boys. There are many other
practices in which we engage: normative practices, albeit primitive ones.
There are ways of playing them as history dictates they should be played
and there are ways of going against the rules that have taken hold as a
part of the habitual rituals that have developed over our decade together.

2. Dog Science
Alexandra Horowitz’s 2009 book The Inside of a Dog: What Dogs See,
Smell, and Know is a fascinating attempt to navigate the murky waters of
dog ethnography with her scientist’s creed not to anthropomorphize on the
one hand—by this she means “attributing to [dogs] the feelings, thoughts,
and desires that we use to describe ourselves”—and her dog lover’s desire
to tell us about a beloved member of her close family and its species on
the other. Although professionally submerged in “masterful texts” that
“describe everything from hormonal and genetic explanations for the social
behaviour of animals, to conditioned responses, fixed action patterns, and
optimal foraging rates, in the same steady, objective tone”, she recognizes
that most of her and other dog owners’ questions go unanswered. “Science,
as practiced and reified in texts”, she says, “rarely addresses our experiences
of living with and attempting to understand the minds of our animals”.1
1 Alexandra Horowitz, The Inside of a Dog: What Dogs See, Smell, and Know (New York:
Scribner, 2009), 3 (all quotations this paragraph).

Rational Animals

207

Her solution to steer this course is as follows. She describes her relationship with her own dog Pumpernickel—who was never, she tells us, a
subject in her professional studies—in almost poetical terms, as an epigram to each section. With her dog-subjects, she makes free use, as her
subtitle suggests, of descriptions decidedly ‘thick’ with mentalistic vocabulary. She augments and adjusts these with suggestions from her studies
of watching dogs play, together with recent studies in animal evolution,
biology, physiology, and ethnography to help us to understand the species based on “what dogs actually have the capacity to feel, know, and
understand”.2 Very occasionally theoretical quibbles creep in.
In learning about the dog’s nose, eyes, and ears, we get a glimpse of
the physiological grounds of its sensory capacities. This, in turn, gives us
a different perspective on his umwelt or the features in his environment
that will be salient to a creature with such capacities. In summarizing
years of research in dog parks, in which dog interactions and play were
filmed and then replayed in slow motion, we are treated to descriptions
of a complex dance of play the subtle details of which we might not have
noticed otherwise.
In thinking about the break in their evolutionary history when domestication and selective breeding occurred, we are invited to reject as a myth
the idea that dogs are best understood (and trained) as pack animals.
They do not generally hunt; they do not form social packs with other
animals; they tend to live a stable life with their human family and not be
pushed out of the pack depending on the availability of prey, and so on.
As for perception, there are important differences. “As we see the
world”, says Horowitz, “the dog smells it”,3 and his world is at least as
rich, if not more, as a result. We learn that dogs are built for sniffing: A
special photographic method has revealed that a slight wind generated
by an exhale, displacing air already in the nose, helps to pull more scent
in, creating a current of air over it. Whereas human noses have only six
million receptor sites with attendant hairs to catch molecules perceived
as scents, sheepdogs have over 200 million and beagles over 300 million.
Together with their genetic makeup coded for detecting smells, a dog’s
“smell experience is exponential” as “combinations of sites fire together
to send information to the brain”.4 A beagle’s sense of smell may be millions of times more sensitive than our own. “The olfactory bulbs of the
dog brain make up about an eighth of its mass: proportionally greater
than the size of our central visual processing center, the occipital lobes,
in our brains”.5 Not only that: the ‘vomeronasal’ organ—a special sac
above the roof of the mouth to detect pheromones—as well as a moist
nose exterior which allows molecules to stick so that they can dissolve
2
3
4
5

Ibid., 14.
Ibid., 68.
Ibid., 71–72.
Ibid., 72.

208

Julia Tanney

and travel through interior ducts to this organ—enhance the dog’s ability
to smell the world. Like sharks, dogs “see in scent”. With this amazing
capacity Horowitz says,
A dog knows if you’ve had sex, smoked a cigarette (done both these
things in succession), just had a snack, or just run a mile [. . .] they
can also smell your emotions.6
And, as we dog lovers already know:
Dogs talk. They communicate; they declare; they express themselves.
This comes as no surprise; what is surprising is how often they are
communicating, and in how many ways. They talk to each other, they
talk to you, and they talk to noises on the other side of closed doors
or hidden in high grasses.7
Over the course of a walk with several dogs, she tells us, the dogs may
scold one another, confirm friendships, court each other, declare dominance, rebuff advances, claim ownership of a stick, or assert allegiance to
their person. Dogs, like so many non-human animals, have evolved innumerable, non-language-driven methods to communicate with one another.
As this short sample shows, Horowitz’s descriptions are peppered with
an unabashed use of mental predicates. Careful, however, not to risk misunderstanding by attributing to dogs capacities that outrun the evidence,
she sometimes retreats. About Rico, the border collie who learned the
names of over 200 toys (and was able to pick out, upon hearing an unfamiliar name, the new toy in the bunch), she says:
Rico was not using language, of course, in the way we, or even young
children, do. One can debate how much he was understanding, or if
he was even doing anything other than showing a preference for the
new object. On the other hand, he was showing an astute ability to
satisfy the humans making various sounds by picking up the referents of those sounds.8
Occasionally, she says things in which adherence to a controversial picture of the mind can be sensed:
. . . the subjective experience of animals is notoriously difficult to get
at scientifically. No animal can be asked to relate its experience in
voice or on paper, so behavior must be our guide.9
6
7
8
9

Ibid., 79.
Ibid., 90.
Ibid., 97.
Ibid., 145.

Rational Animals

209

And:
Behavior has its pitfalls, too, since we cannot be positive that any two
individuals’ similar behavior indicates similar psychological states.
For instance, I smile when I am happy [. . .] but I may also smile
out of concern, uncertainty, or surprise. You smile back at me: it too
might be happiness—or ironic detachment. To say nothing of the
near impossibility of determining whether your ‘happiness’ feels like
mine does.10
More problematically, though thoroughly consistent with present-day
research, Horowitz even considers whether dogs have, understand, or
grasp concepts (such as that of imitation, of the self, of time, or of death)
or whether they have theories of mind.11
Horowitz’s attempt to describe her life with her own dog on the one
hand, and to summarise years of research with her dog-subjects on the
other, is fascinating. Helping herself to a generous assortment of ‘commonsense’ mental predicates, she is nonetheless concerned to adhere to
‘correct’ scientific methodology, especially when it comes to her research
subjects. While admiring her general approach and wishing to defend it
on philosophical grounds, I am nonetheless concerned to tease out what
seem to be residual allegiances to a Cartesian and Intellectualist picture
of the mind, which, albeit rife in cognitive-scientific research, are worth
interrogating.

3. Philosophy
Wittgenstein reminds us that there is a family of structures, more or less
related, that can be discerned in the use of our natural language expressions as opposed to the existence of “one thing in common which makes
us use the same word for all”.12 This includes, of course, the kinds of
mental predicates we are considering here. There are differences, in other
words, in what we count as ‘acting for a (particular) reason’, ‘intending’,
‘wanting’, ‘knowing’, ‘seeing’, ‘believing’, and so on. Dogs and cats, at
least, are unquestionably minded in respect to the most important of the
considerations that would justify these testimonials. I would not be able
to describe our lives with our animal-boys, nor would Horowitz be able
to relate the conclusions of her research, unless we were both allowed to
make use of such mental predicates in our narratives.
Of course, we may be convinced to revise our descriptions—in many
cases Horowitz adduces good reasons for our doing so for dogs—but
10 Ibid., 145.
11 Ibid., 191ff.
12 Ludwig Wittgenstein, Philosophical Investigations, 3rd ed. Translated by G. E. M. Anscombe (Oxford: Basil Blackwell, 1968), §65.

210

Julia Tanney

there are limits here and in any case, as she herself demonstrates, this
would not involve depriving ourselves of mental verbs such as ‘see’,
‘understand’, ‘need’, ‘believe’, ‘want’, ‘hope’, ‘think’, and ‘know’. (A careful reader will have noticed that this includes the expressions ‘know how’
and ‘know that’.)
Before going into more detail about how I would defend my descriptions, if challenged, notice that the action-descriptions we use are themselves fully charged. Consider first, and by way of contrast, typically
human activities such as calculating, debating, washing, shopping, giving
a lecture, going to the doctor, attending a concert, or playing the piano.
That intention is (logically) internal to these descriptions shows up most
clearly when we consider how difficult it is to figure out what one could
be asking in enquiring whether someone is, say, shopping intentionally.
Though we are tempted to say, “yes, of course”, we should hesitate to
accept the question until it is clear what is being ruled out. That the shopper has been hypnotized? That he is addicted? Only when it is clear what
would count as shopping unintentionally and that this is a live possibility
in the circumstances at hand would it be appropriate to respond affirmatively or negatively. It is part of the background—a presupposition of
this particular linguistic activity—that, unless special circumstances are
introduced, to describe a human being in one of these ways is all it takes
to understand his or her actions as intentional. The problem with the
question is the implicit suggestion that something else may be required
in addition to what is already implied in the description—perhaps the
occurrence of a bit of prior cogitation—for a positive answer.
As this is true for human beings, so is it true for animals: ‘stalking’,
‘playing’, ‘hunting’, ‘searching’, ‘cuddling’, ‘asking for/demanding attention’, ‘fetching’, ‘playing the “venison tongue” game’, are, like the ones
above, descriptions that are logically connected to intention and desire in
the sense just described. To spell it out again: We would not understand
what it could mean, unless special considerations were adduced, to suggest that the animals were stalking or playing but not intentionally so; or
asking for attention but not wanting it, or chasing a lizard but not seeing,
thinking, or believing one was there, and so on. This contrasts with the
view that we find in many philosophical discussions as well as in animal
science (vestiges of which are detectable in Horowitz’s hesitations) that
thick action-descriptions depend for their correctness on the existence of
a hidden, antecedent mental event that is both contingently connected to
‘mere behavior’ and provides the rational content to elevate this behavior to an intentional action. The occurrence of this alleged mental event
would be, so the idea goes, both the causal precursor of, and the metaphysical thickening agent for, ‘thinner’ or unleavened bodily movements.13
13 For more discussion, and a thorough criticism of this view, see my “Remarks on the ‘Thickness’ of Action-Description: with Wittgenstein, Ryle, and Anscombe” in Philosophy of

Rational Animals

211

Instead of construing mental predicates as picking out hidden mental
states that we infer from behavior and other ‘outward’ signs, it might be
less harmful to construe them as characterizing what we perceive in the
circumstances. Consider, for example, Gary L. Hagberg’s suggestion that
We see human mental states, emotions, nuanced expression (of a kind
and with variations far more subtle than philosophy often acknowledges) in persons, not in bodies that provide evidence for hidden
ghostly mental entities.14
But so too do those whose families, like mine, include cats and dogs see
desires, intentions, thoughts, fears, hopes, jealousy, grief, anxiety, and so
forth in these animals in the particular circumstances and environment,
and not in bodies that ostensibly provide evidence for hidden ghostly
mental entities. As Hagberg explains,
Physiognomic perception, rightly understood, is in truth not in the
first instance dualistic; unlike the conceptual picture of dualistic selfhood, it does not place us at an inferential distance from the otherwise
hidden expressive content. And in not introducing an inferential gap
between the inner content of the expression and its (we think merely
contingently) attached outward manifestation, it does not establish
from the outset a philosophical problem asking for an explanation, a
theory, of how that gap is crossed.15
Both the Cartesian—who points to the inner determinants of content—
and the behaviorist—who focuses on the outer, “have buried in their
conceptual substrates a picture that they share in common beneath their
more visible differences”.16
Wittgenstein suggests that the very first step of the philosophical problem we are confronted with here is one that altogether escapes notice:
We talk of processes and states and leave their nature undecided.
Sometime perhaps we shall know more about them—we think. But
that is just what commits us to a particular way of looking at the
Action From Suarez to Anscombe” (working title), edited by Constantine Sandis. Forthcoming in Philosophical Explorations, Special Issue (21:1) 2018 and Routledge, 2019.
14 Garry L. Hagberg, “Wittgenstein, the Human Face, and the Expressive Content of
Poetry: On Bernard Rhie and Magdalena Ostas,” (Nonsite.Org- Issue #4: No Quarrel
(Part 2), Responses, 2011/2012), 131. http://nonsite.org/wp-content/uploads/2011/06/
nonsite_issue_4.pdf. He goes on to suggest (on the same page) that this is key to clarifying “in a way true to the wondrously complex and intricate phenomenology of facial
recognition” what we see in works of art and how we see it.
15 Ibid., 125.
16 Ibid., 127.

212

Julia Tanney
matter. For we have a definite concept of what it means to learn
to know a process better. (The decisive movement in the conjuring
trick has been made, and it was the very one that we thought quite
innocent.)17

We talk of mental processes and mental states and wonder about their
nature: Are they nothing over and above behavior or are they something
else as well? Are they identical to or realized in physical states? Do they
supervene upon them? If so, which physical states? Those of the brain?
Or do they somehow reach beyond the brain and the body? Wittgenstein
continues:
—And now the analogy which was to make us understand our
thoughts falls to pieces. So we have to deny the yet uncomprehended
process in the yet unexplored medium. And now it looks as if we
had denied mental processes. And naturally we don’t want to deny
them.18
Consider how Ryle depicts this same dilemma—one which, as he
describes it, pits the behaviorist (and reductionist) against the Cartesian
(and Platonist).19 The attraction of behaviorism is that it does not insist
on occult happenings as the basis upon which mental terms are given
meaning, and points to the perfectly observable considerations that are
by and large employed when we are called upon to defend or correct our
employment of these mental terms. The problem with behaviorism (as
Hagberg’s comments anticipate) is that it has a too-narrow view both of
what counts as behavior and of what counts as observable. The attraction of Cartesianism is that it recognizes in a way that behaviorism does
not that there may be crucial differences between creatures who—on a
certain restrictive notion of behavior—do indeed behave identically. The
problem with Cartesianism is that it attempts to account for these differences by hypothesizing the existence of occult or hidden causes.
If these were the only options, then we can understand the temptation
to swing first one way and then the other. To deny mental processes—or,
as I would gloss this—to revoke the conceptual apparatus that mental
predicates make available20—would be to deprive ourselves of a key way
in which we render intelligible ourselves and others.

17 PI, §308.
18 Ibid.
19 Gilbert Ryle, “Adverbial Verbs and Verbs of Thinking,” in On Thinking, ed. K. Kolenda
(Oxford: Basil Blackwell, 1979), 17–32.
20 Why do I prefer to gloss it thus? There are, of course, mental states and mental processes.
Anxiety and depression are examples of the first and calculating and remembering are
examples of the second. It is not only unhelpful, however, but also dangerous simply

Rational Animals

213

We do not want to deny mentality. But we do want to rid ourselves of
the Cartesian impulse to suppose that mental adjectives such as ‘jealous’,
‘afraid’, ‘lonely’, ‘sad’, or adverbs such as ‘intentionally’ or ‘deliberately’,
or verbs such ‘believes’, ‘thinks’, ‘knows’, ‘wants’, or combinations such
as ‘acting thus because . . .’, and so on, refer to hidden events, states or
processes that are unknowable in all but our own case. Or worse: that
it is only beings like us—those, for example, with self-reflective awareness (which in turn requires fully mastered language-use)—who are the
proper subjects of this conceptual apparatus. But we can avoid the horns
of the dilemma by both insisting on the appropriateness of employing these mental expressions in describing what we see (in contrast to
what we infer) and by pointing to familiar and observable considerations to defend or correct our employment of these terms. For it is these
considerations—and not their ostensible hidden referents—that disclose
the proper uses and thus the meaning of these expressions.
In order to bring this point home, let us look in more detail how I
am prepared to justify my descriptions of what my animals see, believe,
understand, and know. I shall defend my claim that both my dog and my
cat understand “Up to the terrace!” and “Nap time! Au lit!” by pointing
to the fact that they almost always come running and rush ahead of me
to the terrace door and then outside; or they both hop onto the bed and
plop themselves down, even before I arrive, in readiness for our afternoon ‘pack-naps’.
In describing the considerations by which I would defend my claim,
I am at the same time suggesting that this is what we ought to count as
understanding “Up to the terrace!” and “Nap time! Au lit!” in the circumstances just described: as an invitation to our boys to enjoy the fresh air
or to settle in bed for our afternoon nap. Such circumstances, of course,
essentially involve the capacities of highly intelligent domestic animals.
If the suggestion as to what we should count as understanding in these
particular circumstances seems foreign, it may be worth reflecting on the
methods we adopt when considering human beings. It is part of our job,
for example, as teachers not simply to examine our students to test their
understanding, but to make decisions about what we will count as such.
Indeed, in the UK, for example, one is required to make this decision
explicit by specifying in advance the learning objectives of one’s ‘modules’ and by spelling out what will count as evidence that these objectives

to assume that where there is a mental predicate there is a corresponding state or
process, since these categories themselves demand features (such as genuine duration)
which may or may not be applicable in any particular ascription. Belief, knowledge,
understanding, and intention, for example, are typically ascribed in situations in which
questions about the existence of a ‘corresponding’ state or process is irrelevant or
unintelligible. The locus classicus of this view is Gilbert Ryle, The Concept of Mind
(London: Routledge, 2009), first published (London: Hutchinson, 1949).

214

Julia Tanney

have been met. Even thus specified, the standard achieved for the various types of considerations in play in any given assessment—factual
knowledge, research, argumentation, individuality, creativity, prose style,
and spelling—will often differ and will thus need to be balanced. There
are no higher-order rules that determine this: It is, famously, a matter
of ‘academic judgment’. That ad hoc decisions are required to adjudicate between these potentially conflicting criteria is all too evident when
double-marking and external examining is required. What the students
receive credit for is, for better or for worse, what the examiner decides
to count as understanding in those circumstances. Clear cases of cheating
will force her to revise these decisions. But the fact is that in doing so she
will search for different, observable considerations—perhaps, if it concerns their final mark, she will take into account their comments in class,
or examine them viva voce—taking us no closer to the Cartesian view
that in crediting a student with understanding, we are making inferences
to special faculties or other occult determinants of action.
This helps to put a finger on what I found curious about Horowitz’s
reticence about Rico. If a dog is demonstrably able to pick out the toy his
owner calls out among 200 other ones, it is not clear what anyone could
mean when they ask, “Yes, but does he really understand the words for
those toys?” The fact that the dog underwent training, that he learned
the names of such a large number of toys, and that the experiment is
repeated again and again is enough to rule out a happy accident. This
ability is impressive indeed: Rico and his trainer deserve credit. When the
dog is deemed to have understood the words for these toys, this is due in
part to our (implicit or acknowledged) agreement that fetching the toys
under the specified conditions is what we will count as ‘understanding
the words for toys’ when considering Rico. Fetching the toy named, after
all, was the point of the trainer’s call.21
This concern is related to a broader issue about understanding that is
sometimes put by asking, not merely whether an individual can perform
to the standard required, but whether her doing so is evidence that she has
grasped the rule which determines the correct way to go on. The alleged
rule that constitutes word-meanings is traditionally known as a ‘concept’.
Although this philosophical notion of grasping, having, or understanding
a concept is fashionable, it is not at all clear what is implied by those who
employ it. Its use may be innocent enough: Perhaps it is merely a way

21 Recall Wittgenstein’s builder’s assistant, who understands “Slab!” just in case he takes
to the builder the slab that he had been trained to bring on command (PI, §2). In that
primitive language game as it as been set up, bringing the slab is what counts as the
assistant’s understanding “Slab!” and no question of mental images, for example, or
other hidden processes are in play. Wittgenstein insists on this point, and it is the very
starting point of his demolition of the view that occult happenings are the basis upon
which mental terms are given meaning.

Rational Animals

215

of focusing attention on misunderstandings that might arise by employing certain terms in our descriptions of animals (if, for example, these
characterizations suggest capacities for which there is no evidence). But
there is a danger in talking this way. The notion of ‘grasping’ a concept
harks back to the view that when an individual—to wit, the human being
or ‘rational animal’ par excellence—is praised for having acted intelligently or with her wits, it is not her performance per se that deserves
praise. Nor is it her performances in the light of the fact that she has been
trained, that she can repeat them, or indeed, even (for a fully acculturated language-speaker) that she can comment upon her performances
if required. What would decide the matter (if only it were decidable!) is
whether the successful move had been piloted by the intellectual grasp of
truths, or rules, and connections between them.
Even when it concerns language-speaking human beings this picture
is fraught with difficulties. One, which concerns concepts construed as
word-meanings, is the presupposition that there are context-free and
therefore purpose-independent considerations that would entail or logically determine the correct application of the relevant words or expressions. This picture has already been cast into doubt with the rejection of
the idea that there is “some one thing in common which makes us use the
same word for all”. As a moment’s reflection on our linguistic practices
will make clear (and this is a point which I am independently driving
home in considering our descriptions of animals), there may be different
considerations that warrant our use of expressions such as ‘understands’,
‘intends’, ‘believes’, ‘knows’, and so forth. If the situation or circumstances surrounding any particular employment of these expressions do
not make them manifest, we can, in normal discourse, ask the one who
employs one of these expressions to spell out what she would count as
satisfying it and thus make explicit which considerations are in play: in
other words, what she means by employing it. This opens up the possibility of agreement or disagreement about whether the conditions have
been satisfied on the one hand, and whether the particular conditions
that have been proposed are appropriate for setting the standard in the
circumstances, on the other. This is a vast improvement on the dialectical
situation we would find ourselves in if the ‘fact’ that answers to the question “Does she understand?” were to be forever inaccessible to us.
A second difficulty with this picture is that although it purports to
provide an explanation of a rational ability, there is nothing that could
possibly count as evidence—save the manifestation of the ability itself—
that the alleged ‘rule-book’ has indeed been correctly grasped and implemented in the appropriate circumstances. Further, if the cognitive grasp of
rules is required for intelligent performance, then surely this would have
to be the case for an intelligent application of the rules themselves. But
this would lead to a vicious regress of the sort one finds in the remarks on
rule-following in the Investigations and in Ryle’s arguments against the

216

Julia Tanney

intellectualist in The Concept of Mind. This exaggerated rationalism—a
spruced-up version of the Cartesian view—is bad enough when applied
to adult, competent, language-using human beings. But such individuals,
at least, unlike infants, animals, and the linguistically disabled, have the
ability to explain what they mean in using an expression, or what they
intend by acting in a certain way. So it is, at a stretch, understandable
how the confusion arises: One has mistaken what is in effect a secondorder ability to explain, correct, or justify—the manifestation of which is
required in certain conceivable circumstances—as a necessary condition
of intelligent action required in all. Once this mistaken move has been
made, and we are confronted with an unquestionably intelligent action
upon which there has been no overt second-order mental operation, we
(erroneously) assume there must have been a covert one. Thus, we arrive
at the hidden mechanism posited by the intellectualist. But what could it
possibly mean to suggest this mechanism—a non-conscious derivative of
high-order theorizing—may or may not be at play for dogs?
Similar problems arise about the question whether an animal could have
a ‘theory of mind’. The notion of a theory of mind was originally introduced as a hypothesis about what underlies and explains the languagespeaking human being’s ability to ascribe ‘mental states’ to others.22 Since
animals do not have the capacity to engage in the linguistic practice we
call ‘ascribing mental states to others’ or to deploy mental predicates, the
question is usually understood as asking whether there are grounds for
supposing animals can, for example, detect intentions or engage in pretence. Surely, this question can be asked without the intellectualist trappings. Aside from the explanatory regress and lack of evidence that such
a theory has been deployed—even in the case of adult, language-speaking
human beings—what are we to make of the sheer absurdity of supposing
that the animal possesses the ability to deploy commonsense generalizations about the mind and rational action in the first place? Surely, this is
cognitivism gone mad.
We should agree with Wittgenstein and Ryle that it would be a mistake to think that there is some Faculty, a causally efficacious disposition, the possession of concepts or theories, or any other hidden mental
processes that is supposed to both explain in a causal and logical/rational
sense our capacity to ‘understand the mental states’ of others. To think
so in the case of human beings is already a muddle—but at least there is
something that counts as a human being’s applying a theory (which of
course is a second-order practice that comes with its own standards of
correctness). The trouble comes when this practice is converted into to a
‘well-oiled mechanism’ that seemingly operates behind the scenes. Once

22 For an early review of the literature, see Martin Davies and Tony Stone, eds., Folk
Psychology—The Theory of Mind Debate (Oxford: Blackwell, 1995).

Rational Animals

217

that move is made, the temptation to extend the theoretical capacities to
non-linguistic creatures beckons.
Instead of asking whether Rico really understands the names for the
toys, we would be on more solid ground by asking whether we are willing to count the ability to fetch the objects named as grounds for crediting him with understanding their names. In this respect, it may be worth
remembering that parents seem to have no hesitation giving young children credit for understanding the words for general terms when they
are able to point to samples or when they respond with words to another’s pointing. When we celebrate a young child’s success by exclaiming
“Look! She understands ‘dog’!” when she is able to point to a dog or
several dogs or real dogs and picture dogs, then we are, at the same time,
suggesting that these abilities are what we will count, on such occasions,
as “understanding ‘dog’ ”. Clearly, the considerations for what we will
count as understanding will change as the child matures—and they will
go on changing as she establishes various areas of expertise. But that
does not mean that the young child or even the dog did not really understand when they pointed to, or fetched, the correct objects. Whether
the child or the dog really understands, again, depends upon what we
are counting as understanding for the child and the dog in the circumstances and whether or not those considerations are satisfied. It would
be altogether inappropriate to suggest that the young child, just learning to speak, does not understand the word for ‘dog’ since she does not
yet ‘possess the concept’. Since, again, there is nothing that could count
as evidence for the existence of an occult mechanism that guarantees
success upon the cognizing of rules for the application of these terms,
all we could mean by this suggestion is that we will await evidence of
additional abilities before attributing understanding.
Of course, we may have to revise our standards, in any given case, for
what is to merit the testimonial ‘understands’. Just as there may be cheating on an exam, so there might be a trick of sorts at play for any ostensible achievement. If it turns out the child just learning the word points to
cats and horses and well as dogs, we might demur from crediting her with
understanding ‘dog’. Or, perhaps we find that Rico is like Hans the Horse
who appeared to be counting, but in the end was found to be acutely
sensitive to his owner’s actions, reactions, and expressions.23 If this were
to be so, we would perhaps be well advised to revise our descriptions. But
to do so would be in effect to revise what we are prepared to count as
setting the standard for understanding in the circumstances.
Cartesian leanings seem to creep in again with the suggestion that since
no animal can be asked to relate its experience in voice or on paper,
behavior must be our guide. What ‘experiences’ are being imagined here?

23 See Horowitz, The Inside of a Dog, 163–164.

218

Julia Tanney

Patrick Ness, in The Knife of Never Letting Go, captures the irony here
in the first lines of his novel, spoken by the main character, Todd.
The first thing you find out when yer dog learns to talk is that dogs
don’t got nothing much to say. About anything.
Indeed, his dog Manchee’s first utterance in the book is: “Need a poo,
Todd”. His second is “Squirrel!” as he runs off in hot pursuit.24
Moreover, third-person attributions of mental predicates are always
answerable to (may be challenged or justified by appeal to) the natural
expressions, reactions, and actions of the one to whom they are attributed. In this sense, ‘behavior’ is always our guide, since it is by these
considerations, together with the circumstances (e.g. absence of food,
injury to the body) that we teach children, for example, when and how to
articulate their needs and wants. Without primitive natural expressions,
reactions, and actions (by which I mean to include, of course, stillness or
lethargy), we are deprived, at least in the first instance, of the grounds for
describing an animal or a child as in pain, hungry, wanting or acting with
intention. True enough, as the child matures, she can learn to suppress
some of her natural reactions and even to pretend. But as Wittgenstein
reminds us, “A child has much to learn before it can pretend. (A dog cannot be a hypocrite, but neither can he be sincere)”.25 Given enough time,
even pretense, feigning, and acting are in principle detectable: And the
evidence comes, not from the individual’s ‘subjective experiences’ or what
she describes in her stream of conscious thoughts, but from inconsistencies in other things she says and does.
In any case, finally, Solutré’s insistent barks are as expressive of his
wanting something—and thus what is on his mind—as are his yelps of
alarm when the cat makes a grab for his couilles. So too is the cat’s climbing onto my chest, poking his nose into my eye, and eventually resting his
muzzle on my cheek and purring, as expressive as what is going on in his
heart as is my husband’s hand touching mine. But ‘going on in his mind’
is as metaphorical as ‘going on in his heart’. Why should the former be
construed literally? Again, these descriptions of my dog, cat, and husband, come with their logically internal connections to the concepts of
wanting, hurting, needing, intending, and feeling affection. What inclines
us to construe them as picking out ‘inner experiences’ to be communicated at will, if only the means were available? Why not view our use of
such notions instead as deriving from primitive, natural manifestations
which are both psychologically and logically prior to the more complex,
learned ability to articulate or suppress what and how one thinks and
feels? And, finally, why project this picture of a fully-fledged language
24 Patrick Ness, The Knife of Never Letting Go: Chaos Walking: Book 1 (Somerville, MA:
Candlewick Press, 2008).
25 Wittgenstein, PI, IIxi.

Rational Animals

219

speaker’s abilities onto the domestic animal who has no such capacity
to hide its thoughts, or even to think about them, and thus for whom
it would be nonsense to imagine it might, if only it could speak, freely
express what it has chosen not to divulge?
The correct employment of mental predicates—as with all natural
language predicates—involves being able to explain and defend what
we are counting as satisfying them on the occasion. The considerations
by which any two people might justify their employment of the same
mental expression may not only differ, but also be incompatible, thus
necessitating either explicit decisions or implicit agreement as to which
will be appropriate in any particular case. When the subjects of our psychological ascriptions are fully acculturated language speakers there are
a variety of considerations that come into play. With understanding,
as with belief, desire, hope, or love, there are ‘static’ tests which may
have to do with the way an individual is feeling, thinking, or acting at
some particular moment. There will also be those tests—arguably more
important—that extend over time. A third, parasitic on the first two,
considers what the individual would say in spontaneous avowals or in
self-reflective interpretation. These different ‘tests’, considerations, or
criteria may overlap: A person, for example, who is in love may satisfy
all three. That is, her sincere avowals, her present feelings and her future
actions may all support this description. But sometimes these considerations may diverge and we will be forced to choose which of them will
guide our decision to apply or to refuse the application of any particular
term in any particular case.26
The kinds of considerations we may appeal to when we are considering
domestic animals is reduced as a result of their incapacity for speech and
the kind of thinking this brings in train. But this does not mean that they
are not appropriate subjects of mental descriptions. Nor does it mean
that we can only have an impoverished view of what they really think,
want, and feel, because the evidence is out on what is really going on
inside their heads.
There is a lot more to be said about how mental terms function and
how this plays out in my characterization of my boys. When I claim that
Solutré knows I am about to go out, I am not only telling you something
about him. I am also implying something about myself: To wit, that I am,
indeed, about to leave. If I were to describe him as believing this to be so

26 It would be tempting, but a mistake, to suppose that the satisfaction conditions for
these so-called ‘dispositional’ concepts (so called when it was recognized that the tests
extending over time often trump the test for the present condition of the individual or
what she would say about her condition) are satisfied by the presence of a complex,
“bobbin-like” state: “as something always there from which behaviour follows [. . .]
analogous to the structure of a machine and its behaviour”. Ludwig Wittgenstein, Wittgenstein’s Lectures, Cambridge, 1932–1935, ed. Alice Ambrose (Oxford: Blackwell,
1982), 90–91.

220

Julia Tanney

when I put on my make-up and high-heels, I leave it open whether I was
in fact going out. When I say that Vinzelles seems to know when I am
about to take the dog out for a walk, here I am showing a reluctance, to
use Austin’s expression, to stick my neck out about how to explain his
uncanny appearance at the front door. In employing these predicates, I
am signaling the evidential status of my claim, or perhaps even my own
commitment to or willingness to endorse what I put forward. The use of
such predicates, in other words, extends beyond what might be called
‘explaining one’s behavior by describing one’s mental states’.27

4. Conclusion
I have argued that the ‘observable’ conditions that are satisfied when we
count one another as understanding, knowing, seeing, believing, wanting, thinking, and so on can also be used without apology for animals:
at least when their actions, reactions, and ‘physiognomy’ contribute to
the defence of these characterizations in ways and in circumstances that
show relevant similarities to our own. For this reason, to paraphrase
Wittgenstein, it would make sense to attribute to Solutré the expectation
that his ‘mummy’ was coming home.28 For he was, I am told, amazingly
accurate at predicting correctly that the time to fetch me—at the RER
station every week during term time—was nigh. By contrast, it does not
make sense to attribute to him the expectation that I would be coming
home the day after tomorrow. For nothing in his repertoire would provide any grounds for saying so.
Nor can we credit the dog or cat with participation in the multifarious
more complex normative practices (even putting aside the obvious intellectual and moral ones) with which we engage. As Ryle says,
A lion, unlike a man, neither respects nor flouts technical standards
of craftsmanship; it is neither an artist nor a philistine; it is neither
tactless nor tactful, courteous nor brusque; it has neither a good nor
a bad head for business; it is neither a good nor a bad player of tennis
or chess or hide-and-seek; it is neither sentimental nor unsentimental
nor cynical.29
As this quotation suggests, however, it is not that the animal is not a
good craftsman or artist, or that it is not tactful or courteous, or that it is
27 For more discussion, see my “What Knowledge is Not: Reflections on Some Uses of the
Verb ‘To Know’,” in Knowledge in Contemporary Philosophy, eds. Markos Valaris and
Stephen Hetherington (New York and London: Bloomsbury Academic, forthcoming,
2018).
28 Wittgenstein, PI, Iii.
29 Gilbert Ryle, “A Rational Animal,” in Collected Papers, Vol. 1: Collected Essays (London: Routledge, 2009), 431.

Rational Animals

221

not a good businessman. It is rather that there is nothing, given its capacities, that would count as its being either good or bad. Further, insofar as
we enjoy what Ryle calls “the egotistical pastime of giving to mankind
testimonials which we withhold from other creatures”:
we shall certainly lay great, though not exclusive emphasis on his
past performances and his future promises as a theorist, that is, an
advancer of knowledge, no matter whether this be knowledge of
nature, mathematical knowledge or knowledge of human ways and
human callings.30
Nor would anything within the observable repertoire of animals count
as grounds for supposing them to have the capacity to reflect on their reasons for acting. So one consideration I might count in favor of attributing
‘a reason for doing such and such’ to you or to me—what we would give
as our reason—is ruled out here. This has some interesting consequences:
nothing would count as and thus it makes no sense to speak of Solutré’s
or Vinzelles’ being self-deceived or weak-willed.
Are domestic animals rational? Perhaps, as Ryle suggests, since they are
neither rational nor irrational, the question should be rejected. But they
count as acting for reasons in the senses spelled out in this paper and they
fail to count as acting for reasons in some of the other possible senses also
described herein. To credit animals with intelligence or mental capacities
in general will require deciding what we are to count as deserving the epithet and ruling out a myriad of possible misunderstandings. But, as Ryle
insists, “[w]hat we must not do”—and here I would add for either human
beings or for animals—“is to confuse testimonials with explanations”.31
Whatever we decide depends neither upon the presence or the absence of
‘elemental agencies of forces’ such as Theoretical or Practical Reason, nor
upon hidden, because ‘inner’, mental causes of overt behavior.

References
Davies, Martin and Stone, Tony, eds. Folk Psychology—The Theory of Mind
Debate. Oxford: Blackwell, 1995.
Hagberg, Garry L. “Wittgenstein, the Human Face, and the Expressive Content
of Poetry: On Bernard Rhie and Magdalena Ostas”. Nonsite.Org- Issue #4:
No Quarrel (Part 2), Responses, 2011/2012. http://nonsite.org/wp-content/
uploads/2011/06/nonsite_issue_4.pdf
Horowitz, Alexandra. The Inside of a Dog: What Dogs See, Smell, and Know.
New York: Scribner, 2009.
Ness, Patrick. The Knife of Never Letting Go (Chaos Walking, Book 1). Somerville MA: Candlewick Press, 2008.
30 Ibid., 447.
31 Ibid., 447.

222

Julia Tanney

Ryle, Gilbert. The Concept of Mind. London: Routledge, 2009. First published
London: Hutchinson, 1949.
———. “Adverbial Verbs and Verbs of Thinking”. In On Thinking, edited by K.
Kolenda, 17–32. Oxford: Basil Blackwell, 1979.
———. “A Rational Animal”. In Collected Papers, Vol. 1: Collected Essays. London: Routledge, 2009, 428–447. Originally delivered as the Auguste Comte
Memorial Lecture (London: Athlone Press, 1962).
Tanney, Julia. “Remarks on the ‘Thickness’ of Action-Description: with Wittgenstein, Ryle, and Anscombe” in Philosophy of Action From Suarez to Anscombe”
(working title), edited by Constantine Sandis. Forthcoming in Philosophical
Explorations, Special Issue (21:1) 2018 and Routledge, 2019.
———. “What Knowledge Is Not: Reflections on Some Uses of the Verb ‘To
Know’ ”. In Knowledge in Contemporary Philosophy, edited by Markos Valaris
and Stephen Hetherington. New York and London: Bloomsbury Academic,
forthcoming, 2018.
Wittgenstein, Ludwig. Philosophical Investigations, 3rd ed. Translated by G. E.
M. Anscombe. Oxford: Basil Blackwell, 1968.
———. Wittgenstein’s Lectures, Cambridge, 1932–1935, edited by Alice
Ambrose. Oxford: Blackwell, 1982.

10 Modes of a “Complicated Form
of Life”
Expression and Human-Animal
Continuity
Stina Bäckström
Introduction
The later Wittgenstein was obviously not a naturalist about psychological phenomena in many of the senses philosophers attach to this term. He
was no friend of the reductionist positions of his time, be it behaviorism
or identity-theory materialism. Neither did he think that philosophical
psychology is continuous with scientific psychology, but defends philosophy as a distinctive discipline. On top of this, there is a quite general
difficulty of aligning Wittgenstein with any “—ism”, given his view of
philosophy as aiming not at doctrines, but at clarity and the dissolution
of problems.
In this essay, I will not directly address the question where to place
Wittgenstein with respect to the various naturalist positions of today’s
philosophical currency. However, I will discuss one question of importance to naturalism, namely how to conceive of the relation between the
minds of human animals and the minds of other animals.1 I will approach
this question via considerations about expression in animals and humans.
Such an approach to questions about the psychological is distinctively
Wittgensteinian. Wittgenstein showed how we can make progress with
philosophical problems regarding the mind by turning to how the mind
is expressed (and, more broadly, to the surroundings of such expressions).
I want to argue that there is material in Wittgenstein to oppose the view
that human and animal expression is continuous—at least on one understanding of what continuity means. So first we need to know what view,
more precisely, I want to oppose.

I. The Continuity-View
The continuity-view is constituted by two interlocking claims: 1) There is
a significant overlap in expressive behavior between humans and animals
1 I will in what follows refer to human animals as “humans” and other animals as “animals”. This colloquial way of marking the distinction serves as well as other constructions and it is the least cumbersome. By using it, I do not mean to deny or downplay the
fact that a human being is an animal.

224

Stina Bäckström

2) Adding new expressive repertoire to include linguistic expression does
not fundamentally alter the entire spectrum of expressive behavior. What
these commitments articulate, in different ways, is the idea that there are
at least some forms of expression in human beings that are not structurally or qualitatively different from expression in animals.2
Consider the following passage from Alice Crary’s recent book Inside
Ethics:
Expressive behavior is helpfully conceived as forming a continuum
that runs from, on the one hand, behavior that is wholly unlearned
or part of creatures’ original natures to, on the other, behavior that
is fully linguistic (i.e., behavior of the sort exhibited by a speaker
when she expresses herself by saying something.) Thus conceived,
expressive behavior is the domain of animals as well as human
beings. Whereas only mature human beings exhibit behavior that
falls all along this continuum, and whereas animals of some kinds
exhibit only unlearned behavior, many animals exhibit behavior that
ranges from the fully unlearned up to quite sophisticated intermediate points.3
The picture Crary sketches here might seem like nothing but sheer common sense. (And indeed, Crary presents her philosophy of mind as a
“common sense realism”.)4 For, clearly, many animals have a rich expressive repertoire, just like we do. Here is one example: During a recent
stroll, I saw a swan defending its territory against an intruding duck by
assuming a hostile posture, chest up and wings raised, hissing threateningly. For Crary, continuity of expressive behavior is the best way
of respecting my perception of the swan as expressing hostility. When
Crary wants to “tell a story about conditions of understanding expressive
behavior—and of thereby understanding aspects of mind—that is pertinent to human beings and animals alike”, she shows her adherence to the
idea that the differences between human and animal expression are not
structural or qualitative.5 On her view, one and the same philosophical
account embraces both animal and human expression, at least in their
central aspects. If we take Crary to merely be endorsing common sense,
2 Matthew Boyle’s recent work on rationality as a transformative capacity provides an
excellent way into thinking about structural or qualitative differences in human and
animal mentality. See Matthew Boyle, “Additive Theories of Rationality: A Critique,”
European Journal of Philosophy, 24/3 (2016), 527–555, and “Essentially Rational Animals,” in Rethinking Epistemology, eds. Günther Abel and James Conant. Vol. 2 of Berlin Studies in Knowledge Research (Berlin/Boston: De Gruyter, 2012), 395–428.
3 Alice Crary, Inside Ethics: On the Demands of Moral Thought (Cambridge, MA and
London: Harvard University Press, 2016), 68.
4 Ibid., 40–47.
5 Ibid., 68.

Modes

225

opposing her view will appear to constitute skepticism about animal
expression, i.e. the position that animals do not express states of mind
in any significant sense at all. The fact that we perceive them as doing
so should, on the skeptic’s view, be regarded as some kind of systematic
misconception—as “intellectually irresponsible anthropomorphism”, to
use a phrase from Dorit Bar-On.6
Bar-On is another advocate of the continuity-view. She argues that
although there might be parts of human mentality—such as reasoning
and speaking—that are qualitatively different from anything we find in
animals, there is still a common ground of expressive behavior. Bar-On
says,
[W]hat facilitates our understanding of animal minds is the fact that,
like us, our fellow beasts engage in a distinctive kind of behavior:
expressive behavior. Even those skeptical about the cognitive powers
and language-related capacities of animals [. . .] would likely agree
that animals as biologically diverse as apes, canines, felines, dolphins,
and birds are able to engage in behaviors expressive of pain, fear,
agitation, aggressiveness, excitement, contentment, playfulness, even
puzzlement, and so on.7
Here, Bar-On is claiming both that expressive behavior is one distinctive
kind of behavior exhibited by us and other animals and that we can separate that form of behavior out from our “language related capacities”.
Thus, on her view, expression in humans and animals is continuous, in
the sense expressed by the two commitments.
Both Bar-On and Crary claim their continuity-views to be inspired by
Wittgenstein. I think, however, that there are passages in Wittgenstein
where he opposes the continuity-view of human and animal expression.8
Contrary to the continuity-view, we can use Wittgenstein to argue that
in (mature) human beings, expressions, even non-verbal expressions of
pain, fear, agitation, and so on, are modes, as he himself puts it, of a
“complicated form of life”, a life led in language.9
Bar-On and Crary are right, however, that Wittgenstein is not a skeptic
about animal expression. In the last section of the paper, I will discuss
6 Dorit Bar-On, “Expressive Communication and Continuity Skepticism,” The Journal of
Philosophy, 110/6 (2013), 293–294.
7 Ibid., 296.
8 In developing these passages, I am aware that my reading is underdetermined by the,
often elusive, textual evidence. What is important is that this is a powerful view that also
makes sense of difficult material in Wittgenstein.
9 Ludwig Wittgenstein, Philosophical Investigations, 3rd ed. Tr. G. E. M. Anscombe
(Oxford: Blackwell, 2001), 148. The main contrast in this paper is that between mature
human beings and animals. There are, of course, other contrasts deserving of consideration. Children/mature human beings is one, domestic animals/wild animals is another.

226

Stina Bäckström

two different ways in which we might avoid both skepticism and the
continuity-view. First, I want to turn to the passages in Wittgenstein
where we can find arguments against the continuity-view.

II. Expression, Articulability, and Time
I begin with two successive remarks from Wittgenstein’s Philosophical
Investigations:
647. What is the natural expression of intention?—Look at a cat
when it stalks a bird; or a beast when it wants to escape.
((Connexion with propositions about sensations.))
648. ‘I no longer remember the words I used, but I remember my
intention precisely; I meant my words to quiet him’. What does my
memory shew me; what does it bring before my mind? Suppose it did
nothing but suggest those words to me!—and perhaps others which
fill out the picture still more exactly.—(‘I don’t remember my words
any more, but I certainly remember their spirit’.)
In these remarks, Wittgenstein is juxtaposing two cases. First, he asks
us to consider a cat stalking a bird as a natural expression of intention.
Second, he describes someone who is recalling a past intention, remembering what she meant with a particular remark in the past. Her saying,
“I meant my words to quiet him” is an expression of intention, albeit a
past intention.
In considering these cases in succession, is Wittgenstein asking us to see
them as continuous? Is the person who is expressing her past intention
doing something more sophisticated but qualitatively similar to what
the cat is doing? There is something puzzling about this question. When
we try to answer it, one thing that emerges is that nothing in the cat’s
expressive repertoire counts as expressing a past intention. Its expressions appear confined to its present.
In § 650, Wittgenstein offers us a new case, one that also invites reflection on human/animal expression and time. But in this paragraph the
psychological phenomenon at issue is a different one, namely fear: “We
say a dog is afraid his master will beat him; but not, he is afraid his
master will beat him tomorrow. Why not?” We can put this remark in
terms of expression: We say the dog is expressing his fear that his master
will beat him, but not that he is expressing his fear that his master will
beat him tomorrow. This latter sort of expression is, as far as we know, a
human phenomenon.
Time enters into the comparison between animal and human expression in two different ways in these passages. First, we have the recognition that there is nothing in the stalking cat’s repertoire that counts as

Modes

227

expressing a past intention. Second, we note that the frightened dog’s
expression of fear can only take an object that is, in some way, in its
presence, whereas a human expression of fear can take an object that is
in the (indefinite) future. The life of a dog or a cat, say, certainly sustains
the idea that things, sounds, or smells can evoke fear, even if there is
nothing actually dangerous on the scene in the moment. So, if the dog’s
cruel master has previously beaten the dog with a stick, merely seeing the
stick or a similar one probably will fill the dog with fear of a beating. The
point is that there is no room for clarifying what the dog is afraid of by
situating the object of its fear at a point in time. It is hard to find a clear
application for the question: Is the dog afraid it will get beaten now or in
two days when its master comes home? It is, as it were, always afraid of
being beaten now.10
Going beyond our present in these two ways—expressing our past
mental states as well as our present and expressing mental states whose
objects are in the indefinite future (or past)—is something we do by
using language. It is by deploying the tensed expression “I meant . . .”
that we express a past intention. And it is by saying such things as “I’m
afraid he’ll beat me tomorrow” that we express a fear of an object in
some future point in time. No advocate of continuity would deny that
linguistic forms of expression are quite special, and part of why they are
special is precisely because they embody a relation to time in the way
just outlined. Why should we not think, as Crary and Bar-On encourage
us to, that the relation to time made possible by language merely adds a
layer of complexity to our expressive lives, on top of the natural expressive repertoire we share with other animals? On the view I am now
imagining, some human expressions are precisely examples of natural
expressions, i.e. of a piece with the cat’s stalking of the bird. One might
think that a person’s reaching for an apple or walking over to the coffeemaker are precisely natural expressions in this sense. Why not think that
Wittgenstein is inviting us to see (at least some of) our own intentional
actions in the same light, i.e. as “natural expressions” of our intentions?
I do not think this is right. As Crary rightly points out, both human and
animal expression are understood contextually, in the sense that what
happens in the moment comes into view given a conception of the place
of that sort of behavior in the situation (and, even more broadly, in the

10 My interpretation of these passages in Wittgenstein is inspired by a comment on PI
§650 by Sebastian Rödl: “Wittgenstein does not want to say that the fear of the dog
does not reach as far into the future as that of man. Perhaps he wants to say that expecting is something different in a dog from what it is in a man: the form of expectation of
the dog is not time, its expecting, no time-consciousness [. . .] Perceiving, remembering,
and expecting do not expand their scope in man. They are tuned to a new key. This new
key is time”. Sebastian Rödl, Categories of the Temporal (Cambridge, MA: Harvard
University Press, 2012), 78–79.

228

Stina Bäckström

place of the life of the animal or human being).11 A cat’s stalking motions
might be a part of playing where no real prey or food is on the scene.
Or, they might be a part of a hunt of prey. There can be, for this reason,
a form of openness or indeterminacy to an animal’s expression, an indeterminacy that can be settled by continuing to observe what the animal
is doing. In the human case, however, the temporal dimension tied to
language-use gives a different structure to the indeterminacy of what is
immediately given.
The point is not that psychological phenomena are elusive or even unobservable. In neither the human nor the animal case, there is always—or
even most often—any such indeterminacy regarding what is going on. It
is often perfectly open to the naked eye what the animal or human being
is doing.12 That I am sitting in front of my computer writing is something,
as Anscombe puts it, “anyone grown to the age of reason in the same
world” would know as soon as he saw me.13 However, there are situations where we need to draw on a wider context of intelligibility in order
to understand what a person or animal is doing or otherwise expressing.
And such situations can tell us something important. In the human case,
we can ask the person what they are doing and get their own account or
description of the situation. This possibility is intimately connected to the
relation to time enabled by language. In heading to the coffee-maker, am
I intending to make myself a coffee (now) or prepare the machine so I can
put it on when I come back from the meeting (later)? In the case of an
animal, the material we need to settle expressive indeterminacy is within
the scope of what we, broadly speaking, can see. (This is not the point
that we always can settle all questions with respect to animals—animals
can be difficult to interpret—but rather how we do settle the questions
we can in fact settle.) What the cat is doing in the next moment will cast
immediate light on what it is expressing; if it is, say, playfully chasing or
hunting prey.14
Now, someone might object that the difference here is precisely one
of degree of complexity: if you follow the person around and observe
him for long enough, you might eventually be able to settle the question
what he is up to. And this is of course true. My point is that with human
beings what we see is indicative of what the person would say, if asked,
and what the person says clarifies what we see. There is thus a sense in

11 Crary, Inside Ethics, 68.
12 In chapter 2 of Inside Ethics, Crary has an excellent discussion of observability and
objectivity with regard to psychological properties.
13 G. E. M. Anscombe, Intention, Paperback ed. (Cambridge, MA: Harvard University
Press, 2000), 8.
14 In my “Seeing People and Knowing You: Perception, Shared Knowledge and Acknowledgment,” European Journal for Philosophy of Religion, 5/4 (2013), 55–73, I develop
this point in relation to the epistemological problem of other minds.

Modes

229

which the absence of an account can make our understanding of a human
action or expression poor, less than ideal. Say the person is heading to the
coffee-maker to note its brand in order to make sure that when he buys
one it will not be that particular brand. What we can see is him walking
up to the machine, taking a quick look at it, and then moving on to do
something else. Even if he, sometime later, acts on the plan to buy a new
coffee-maker, we will not be able to make the connection, unless he says
something about it.15
Of course, our intentions need not always actually go beyond the present moment, we might be, say, out for a walk but going nowhere for no
reason. Or they might go beyond the present moment but in a way where
the progression of the situation perfectly well settles the question what
we are doing. I might go over to the coffee-maker, make myself a coffee,
and sit down to drink it. The important point is that the possibility of verbal articulation, bound up with the relation to time, changes what we can
say about, and thus how we understand, even such cases. Such actions
are in the present—we are saying something informative when we thus
qualify them. “Nowhere for no reason” answers a question about the aim
of the action, a question whose other possible answers include things in
the indefinite future and which thus cannot be posed even about (not to
mention to) the cat stalking the bird.16
Now, is not this interplay between the verbal and the non-verbal particular only to some psychological phenomena and their expressions?
My examples, intentional action and fear, might appear biased towards
discontinuity since they clearly have what philosophers sometimes call
“intentional content”. How about phenomena whose intentionality is,
at least, a matter of dispute? Expression of physical pain might seem like
a better candidate for a shared form of expression. There is reason to
doubt, however, that non-verbal expressions of pain are any more separable from language-use than expressions of intention or fear.
Say I scream in pain and withdraw my hand upon burning it on a
hot oven. Since this expression does not seem to be about some present
or future object, it might seem as if we do not have the same kind of
temporal indeterminacy in this case and hence not the same dependence
15 For a discussion of this point in relation to practical knowledge, see Michael Thompson’s “Anscombe’s Intention and Practical Knowledge,” in Essays on Anscombe’s
Intention, eds. Anton Ford, Jennifer Hornsby, and Frederick Stoutland (Cambridge,
MA: Harvard University Press, 2011), 198–211.
16 One way of characterizing the move I make here is to say that I am isolating formal
features of expression. Here I am inspired by Anscombe’s remark in Intention, 84, that
she is isolating a “form of description of events.” Martin Gustafsson and I develop the
idea of formal features in our “Skill, Drill, and Intelligent Performance: Ryle and Intellectualism,” Journal for the History of Analytic Philosophy, 5/5 (2017), 41–55. See also
Anton Ford, “The Arithmetic of Intention,” American Philosophical Quarterly, 52/2
(2015), 129–143.

230

Stina Bäckström

on verbal articulation. For a contrary picture, consider another remark
in Philosophical Investigations, §666, where Wittgenstein imagines
someone in pain who is simultaneously hearing someone tuning a piano
nearby, uttering, “It’ll stop soon”. Wittgenstein says,
It certainly makes quite a difference whether you mean the pain or
the piano-tuning!—Of course; but what does this difference consist
in? I admit, in many cases some direction of the attention will correspond to your meaning one thing or another, just as a look often
does, or a gesture, or a way of shutting one’s eyes which might be
called ‘looking into oneself’.
Wittgenstein is here considering a case where there is indeterminacy in an
utterance. Interestingly, Wittgenstein claims that, “in many cases”, the criteria we use to settle the issue what the person meant are non-verbal expressions: a look, a gesture, and a way of shutting one’s eyes. Let us imagine
that it is a non-verbal expression (such as a groan or a grimace) and not
an utterance that is at issue in the passage. With such an expression, too,
we might wonder whether it is an expression of pain or an expression of
discomfort over the piano-tuning. This question might either be settled by
asking the person, or by observing other non-verbal expressions. I take
Wittgenstein here to be pointing to precisely the sort of interplay between
what we see and what the person says that I argued for in the case of intention. One of Wittgenstein’s points in the above passage is that both verbal
and non-verbal expressions—even of pain—are parts of such an interplay.
They can thus be seen as belonging to the same space of intelligibility.
Such an interplay also suggests that it is problematic to think that nonverbal expressions of pain are void of a temporal dimension. Say you
hear me scream from the other room. You run to see what is the matter,
find me with my hand under running cold water and ask: “What happened?” I respond: “I burned my hand. It really hurt for a second, but now
I’m sort of numb”. It is natural to think that I do not merely explain or
describe a noise I happened to emit (as I might explain a squeaky sound
my shoes make), but rather put words in place of the scream, articulate it
verbally. I use the past tense to say what my previous non-verbal expression meant. At the end of §249, where Wittgenstein discusses the meaning
of sensation-terms, he says: “the verbal expression of pain replaces crying
and does not describe it”. In growing up, we become able to use “Ouch!”
or “I am in pain” instead of merely crying out. But of course, we keep crying and screaming, even when we have learned to express ourselves verbally. My point is that once the replacement is in effect (something that is
a gradual transition, of course), even the cry can be seen to be different—it
is now of a piece with its verbal replacements—it can be elaborated verbally. In this sense, even though a cry, or the utterance “Ouch” is not itself
tensed, non-verbal expressions have a temporal dimension.

Modes

231

I thus disagree with the following point, expressed by Crary: “within a
given species, individual animals exhibit the same sort of natural expressive variability that human beings do”.17 Animal expression is variable,
but the variability is not of the same sort as the one we find in humans. In
humans, language-use provides for new possibilities—possibilities I have
described in terms of an interplay between verbal and non-verbal expressions. Since verbal expressions can clarify the non-verbal and vice versa,
there is a sense in which the possibility of linguistic articulation alters the
form expressive variability takes.
One objection at this point could be that we need the continuity-view
in order to account for both the fact that the human species is an evolved
species and the fact that individual human beings grow into language and
reason. These facts are central to both Bar-On and Crary. However, these
facts, no more than the common-sense view that animals are expressive,
force us to accept the continuity-view. What the denial of the continuityview does imply, however, is that the developmental paths in question
cannot be conceived of as simply cumulative—that is, as a process where
children or pre-human species add capacities on top of each other until
they reach the stage of human maturity or the human species. Rather,
such a process needs to be understood as transformative, as yielding new
structures of interlocking capacities.
I thus think that Wittgenstein’s reflections on expression and, more
specifically, his comparisons between animal and human expression, can
point us to how language permeates our expressive repertoire. But other
passages perhaps suggest a different view on his part. Let us consider a
variation on §650 from Philosophical Investigations, part II:
One can imagine an animal angry, frightened, unhappy, happy, startled. But hopeful? And why not?
A dog believes its master is at the door. But can he also believe his
master will come the day after tomorrow? And why not?
Can only those hope who can talk? Only those who have mastered
the use of a language. That is, the phenomena of hope are modes
of this complicated form of life. (If a concept refers to a character
of human handwriting, it has no application to beings that do not
write.)18
Read one way, this passage may express a dissatisfying position. Wittgenstein might then be seen as embracing continuity about certain psychological phenomena and then singling out hope as a distinctively human
phenomenon, thus reverting to skepticism about dog-hope. That is, he

17 Crary, Inside Ethics, 78.
18 PI, Part II, 148.

232

Stina Bäckström

would then be saying that only those who can talk can hope. This view
is dissatisfying precisely because there are situations where it is plainly
open to view that a dog is being hopeful and one such situation is precisely that which Wittgenstein describes: a dog sitting at the door, waiting
for its master to come home. Why does he draw the line with hope, one
wonders, when he thinks that animals can be angry and frightened?
I think we have the resources to articulate what Wittgenstein should
have said here, so that we do not saddle ourselves with the implausible
thought that a dog can not be hopeful, or express hope. I want to put the
thought this way instead: dogs can express hope, just as they can express
anger, fear, happiness, and a startled reaction. But when we talk about all
those things: hope, anger, fear, etc, in the context of a human life, we are
talking about phenomena that are modes of a language using form of life,
where their interplay with the verbal, their articulability, is a part of their
context of intelligibility, i.e. that which makes them what they are. If this
is right, it brings us back to the question brought up at the very beginning
of this paper. If we deny the continuity-view, can we respect the thought
that animals have a rich expressive repertoire? How are we to avoid saddling ourselves with a view according to which the concept of expression
simply has no application in the life of animals?

III. Forms of Conceptual Unity: Categorial Generality
and Family Resemblance
The way the debate about continuity is framed, we can get the impression
that the only available alternative to the continuity-view is skepticism
about animal expression. Bar-On at one point puts such skepticism, helpfully, in terms of the application of concepts:
[S]ome philosophers have maintained that the application of our
concepts of intentional action, meaning, semantic content, reference,
propositional attitudes, and so on to non-human creatures is at best
a matter of analogy or ‘metaphorical extension’.19
Bar-On’s thought might be put in terms of a challenge to the person who
wants to deny the continuity-view: to account for the conceptual unity
between human and animal expression, in a way which doesn’t merely
make reference to analogy or metaphorical extension.
Wittgenstein’s way of approaching such a challenge would be to not
take for granted that we must choose between saying that “expression”
as applied to humans and as applied to animals has univocal meaning,
saying that it is an ambiguous word, and saying that its application to
19 Bar-On, “Expressive Communication and Continuity Skepticism,” 297.

Modes

233

animals is a matter of analogy or metaphor. Before I develop this thought,
I would like to review another attempt to meet the challenge, namely John
McDowell’s. He, quite famously, argues against continuity in the case of
perception and agency.20 And he deals with the challenge by claiming that
the similarities between human and animal perception and agency are a
matter of their being species of a common genus. Elaborating and endorsing what he takes to be Wilfrid Sellars’s view on perception McDowell
writes,
I do not believe we miss anything Sellars has reason to insist on if we
take him to conceive of [human perception] as a species of a genus,
which, for all he cares, can be recognized as being instantiated also in
the lives of at least some non-human animals, and, we might add, in
the lives of human children.21
Saying that McDowell thinks of perception in (mature) human beings as
a species of a genus is not enough, however, to differentiate his view from
the continuity-view. Crary and Bar-On might be happy to reformulate
their views in the language of genus-species. After all, human expression is in their views a kind of expression, with the distinctive feature of
including a linguistic repertoire. We can keep the central commitments of
the continuity-view while reformulating them in terms of a genus-species
relation, where human expression is one species, differentiated from animal expression by including a linguistic repertoire.
McDowell is, however, appealing to a rather different genus-species
relation—one that is not available to the advocate of the continuity-view.
The genus-species relation McDowell is making use of should be understood in terms of what Anton Ford has called “categorial generality”.22
Where we are dealing with categorial generality, there is no feature that
differentiates the species from the genus—at least no feature that we
can understand apart from understanding the species. Consequently, we
do not arrive at an adequate account of the species by starting from an
20 Crary elaborates her philosophy of mind using resources from John McDowell. This
is surprising, given her commitment to the continuity-view. Crary does not distinguish
sufficiently clearly between endorsing the continuity-view and claiming that “some animals have qualities of mind very much like our own”. Crary, Inside Ethics, 105. I agree
with Crary that McDowell’s position is consistent with there being “significant similarities between the characteristic mental capacities of non-rational animals and those of
rational human beings” and that it leaves perfectly open the possibility of investigating
how the “different stages on the way to rationality resemble each other”. Ibid. But there
is more than one way of understanding difference and similarity. Significant similarities
do not imply the continuity-view.
21 John McDowell, Perception as a Capacity for Knowledge (Milwaukee: Marquette University Press, 2011), 14–15.
22 Anton Ford, “Action and Generality,” in Essays on Anscombe’s Intention, 76–104.

234

Stina Bäckström

account of the genus and then adding a differentia to yield an account of
the species. Rather, an account of the genus is dependent on an account
of the species. The characterization of the genus is more abstract, but the
account of the species is the source of our understanding of the genus.
Ford says, “The categorial species, though less abstract, is more fundamental than its genus”.23
Consider two examples Ford uses to illustrate categorial generality.
Horse is a species of animal, but is there some differentia that accounts
for Bukefalos’s being a horse, over and above being an animal? No, argues
Ford, Bukefalos’s being a horse is “nothing but the determinate form that
his animality takes”.24 The specific features that belong to being a horse
cannot be understood as independently available properties, properties
we can use to get from “sheer animality” to “horseness”. Another, more
well-worn, example is red as a species of the genus color. A red thing is
colored, but its being red does not consist in some feature over and above
its color, its being red just is its being colored.25
An important aspect of categorial generality is that an account of the
species cannot proceed in two steps, by describing the genus first and
then describing the features that differentiate the species. This is precisely
what McDowell has argued in the case of perception. We cannot arrive
at an appropriate account of human perception in a two-step process:
First, we consider what it is to merely take in the world in sensory perception (in a way neutral between animal and human perception); then,
we add rational or linguistic capacities on top of sensory perception to
yield specifically human perception.26 McDowell’s criticism of this sort of
picture—often framed in terms of an attack on the Myth of the Given—
should lead us to conclude that he envisages a categorial genus-species
relation. Whereas a commitment to the continuity-view suggests precisely
that a two-step process of understanding specifically human capacities is
possible, where the first step is an account common to both humans and
animals.
If we conceive human and animal perception in terms of a categorial
genus-species relation, we can respect our common sense view that there
are close similarities between human and animal perception. They are
both, for instance, ways to acquire knowledge. What is blocked is the
idea that we can understand this abstract characterization as a neutral

23 Ibid., 83.
24 Ibid., 83–84.
25 As Ford recognizes, color is less controversial as a case of categorial generality, than
the animal species case. Hence, Ford presents an argument for why we should indeed
accept categorial generality in this latter case, too. His point is, however, not dependent
on our verdict in particular cases, but rather on our accepting that categorial generality
is a genuine philosophical possibility.
26 John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994).

Modes

235

“common factor”, or independently understood feature, in both human
and animal perception. It is rather something we understand by providing an account of its determinate forms, one of which is human—rational
and linguistic—perception.27
McDowell thus, at least prima facie, meets the challenge of accounting
for conceptual unity while denying the continuity-view. Now, what about
Wittgenstein? I said in the beginning of this section that Wittgenstein’s
approach to the challenge would be to resist settling on either a yes or no
answer to the question whether “expression” has the same meaning when
applied to human beings and to animals. Concepts are variable and open,
on his view, and there might be considerations regarding their criteria
that pull both in the direction of unity and disunity. I suspect he would
argue that this is precisely the case for “expression” as applied to humans
and to animals. As long as we manage to keep the criteria straight, we
will stop demanding that there must either be complete unity or complete
disunity.28
Wittgenstein’s term for the flexible and open unity of concepts is family resemblances. Wittgenstein wonders, in Philosophical Investigations
§66, whether, when we consider the concept “game”, we find any one
feature in common to all the things we call games. Reflection on this
question leads him to conclude that the requirement that there needs
to be any such feature for the concept to be applicable is misguided.
According to Wittgenstein, when we discover that our use of a concept
exhibits considerable variation and we cannot find any feature common
to all contexts, we can think of the unity of the concept in terms of “a
complicated network of similarities overlapping and criss-crossing”. The
resemblances of members of the same family is one metaphor Wittgenstein uses to describe such a network. Spinning a thread with fibers that
overlap but no one fiber running through the entire thread is another.29
27 Perception in human beings, according to McDowell, is a capacity for self-conscious
knowledge, knowledge of the kind that gets articulated in judgments and expressed in
assertions. What perception is in animal life, is not a big concern for McDowell, the
important thing is that in so far as an animal does not issue judgments on the basis of
what it sees, its form of perception needs to receive a different philosophical account
than the one he provides for human beings. In Mind and World, McDowell provided
some (quite problematic) positive suggestions about how to characterize animal perception and agency, but in later work he restricts himself to the merely negative point. The
negative point leaves open the thought that there may be many species of non-rational
forms of perception, each depending on how perception is caught up with other capacities central to the animal species in question.
28 Here I am inspired by Roger Teichmann’s paper “The Identity of a Word,” American
Catholic Philosophical Quarterly, 90/2 (2016), 317–335. On his view, developed from
Anscombe and Wittgenstein, the question whether a certain word is used with the same
meaning in different contexts can in philosophy receive both a yes and a no answer,
“relative to different criteria of synonymy and (thus) different purposes” (334).
29 PI, §§ 66–67.

236

Stina Bäckström

One thing that the notion of family resemblance accomplishes is a way
of resisting the idea that univocality in a concept needs to be accounted
for by providing one or more features that invariably characterize all its
applications. Hence this notion also equips us to think that even though
expression in human life is qualitatively different from animal expression, we are not forced to conclude that we have entirely different senses
of the word “expression” on our hands, or that we are dealing with analogies or metaphorical uses. For what this line of reasoning allows us to
say is that there need not be any one feature, or collection of features,
that account for the fact that we use “expression” and cognate concepts
when we talk about human beings and other animals. What we need are
ways of relating the different uses, ways we might very well not be able
to spell out in non-circular terms.30
Consider fear and its expressions in dogs and human beings. Expressions of fear are in dogs and human beings related to the prospect of
danger; they expose a felt vulnerability. The dog and the person alike
might try to flee from the danger, or get paralyzed. However, what danger is in human life, what it is to have a prospect of danger, what counts
as fleeing—all these things are different in dog-life. A person, I noted
earlier, might fear that something will happen at a future time. To take
an example fresh on my mind, I might fear that my children will experience war due to the rise of nationalism in the Western world. This
prospect might strike me as I hear the outcome of the American presidential election. As it does not make sense to flee or take some other
immediate action, I might sit down and cry, overwhelmed by my own
powerlessness.
There are situations, of course, where human beings fear an immediate
danger to their physical integrity and where they flee, scream, or fight, as
we say, like a hunted, caged, or wounded animal. But it would be wrong
to conclude from this possibility that “behaving like an animal” picks
out a kind of behavior or expression that is open to animals and human
beings alike. The purpose and meaning of such formulations as “behaved
like an animal” is itself an interesting topic. All I am contending here is
that such formulations do not pick out ways of behaving that can be
exemplified by human beings and animals alike.
That the notion of family resemblances can be used to articulate a position between the continuity-view and skepticism about animal expression is likely to be met with resistance. It is true that it will only be of
use to me if we understand Wittgenstein to be describing ways of transitioning between uses of concepts that are not simply a matter of adding or subtracting features. I think that the family-resemblance metaphor
is helpful in capturing such transitions (the rope-metaphor perhaps less

30 This is a point Teichmann makes nicely in his “The Identity of a Word”.

Modes

237

so). In comparing and contrasting members of the same family, we are
emphasizing aspects of given unities: faces and physiognomies. When we
note that two people have the same nose, we are capturing something
about how their noses figure in their faces. We are not saying that their
noses are both, say, exactly two inches long. Crucially, we need the unities
to be in place before any meaningful similarities might emerge.
Is Wittgenstein’s idea of open-ended unity, as captured by the notion
of family resemblances, a distinctive alternative to McDowell’s view?31
Wittgenstein’s idea of a criss-crossing pattern of similarities suggests,
prima facie, a less principled order than the notion of categorial generality. How deep do the differences between these two different ways of
accounting for conceptual unity go? The initial appearance that we are
dealing with two different possibilities here might start to look dubious
if we focus on the primacy of the species over the genus characteristic
of categorial generality. On Ford’s account of categorial generality our
understanding of the genus derived from our understanding of its species
forms. The abstract characterizations of the genus appear, then, to have
no content apart from what is given by its more determinate appearances.
Thus spelled out, this picture seems quite close to what Wittgenstein says
about the concept “game” in Philosophical Investigations §69, “How
should we explain to someone what a game is? I imagine that we should
describe games to him, and we might add: ‘This and similar things are
called “games” ’.”
Someone might still insist that there is a difference, one revealed by the
central examples of each view: games for Wittgenstein and colors and
animal species for Ford. Color and animality are prima facie much more
suited to be thought of as categories than games. For it does not seem
appealing at first blush to think of the unity the different colors exhibit
as forming a pattern of overlapping and criss-crossing similarities. The
different colors are simply, and in the very same respect, colors, we might
want to say. Similarly, different animal species form, it might seem, a
much tighter unity than all the different activities we call games. Games
are, after all, human inventions and as such directly dependent on us and
our interests. (It is no accident that animal species are often put forward
as candidates for being so called natural kinds.)

31 The following reflections I take to be preliminary, opening up a field of inquiry rather
than providing definite answers. One path of such an inquiry would lead to considering Wittgenstein’s place with respect to the problem of naturalism as it figures in the
tradition stemming from Aristotle and continuing via Kant and Hegel. For a criticism
of McDowell on the grounds that he (in the purported spirit of Wittgenstein) fails to
answer questions regarding the unity of the concept of nature, see Christoph Halbig,
“Varieties of Nature in Hegel and McDowell”, in John McDowell: Experience, Norm,
and Nature, ed. Jakob Lindgaard (Oxford: Blackwell, 2008), 72–91. See also McDowell’s response in the same volume.

238

Stina Bäckström

Wittgenstein, however, would insist on blurring the line between categorial unity and mere family resemblance. The appearance of a sharp
distinction is based, I imagine he would say, on considering a too narrow
set of examples. He would urge us, as he does with the concept “game”,
not to think but look and see.32 In Remarks on Colour, Wittgenstein asks
us to consider questions such as, “Is there a transparent white?” “When
we say gold-coloured, or silver-coloured, what do we mean?” “Why can
there be flaming red, but not flaming grey?” “Are there three (red, yellow, and blue) or four (red, yellow, blue, and green) primary colors?”
What reflection on such questions yields is, on Wittgenstein’s view, a recognition that what first appeared as a tight unity, color, in fact harbors
significant internal variation. When we think that color is a tight unity
of properties that are colors in the very same respect, are we thinking
of the unity of red, blue, and yellow? Or of all the different shades on a
spectrum? Or one that also comprises white, black, and brown, and the
different combinations that they play a part in? One might speculate that
Wittgenstein would urge us to make a similar move with respect to animal species. Hence there is reason to think that Wittgenstein would not
place much confidence in the idea that categorial generality is something
fundamentally different from family-resemblance.
Let us now return to the quote from Wittgenstein with which I ended
the previous section:
Can only those hope who can talk? Only those who have mastered the
use of a language. That is, the phenomena of hope are modes of this
complicated form of life. (If a concept refers to a character of human
handwriting, it has no application to beings that do not write.)33
Wittgenstein is in this passage emphasizing the distance between doghope and man-hope, to the point where it seems as if he is altogether
denying that we are right to talk about an animal as hopeful. But given
his own conception of how fluid and variable our concepts are, we might
instead reformulate his point this way: It is perfectly right to talk about a
dog sitting and waiting for its master as hopeful, and this is not a case of
analogy or metaphor. But it might be useful, especially when philosophizing about the animal-human relation, to keep in view how different the
surroundings of dog-hope are from those of man-hope—and not insist
on hinging our right to use the same concept on some, independently
understood, shared feature between the cases.
The main aim of this paper has been to argue that there is material in
Wittgenstein’s writing on animal and human expression to construe a
position that is neither the continuity-view, nor skepticism about animal
32 PI, §66.
33 PI, Part II, 148.

Modes

239

expression. On this position our entire expressive repertoire is caught
up in an inter-play between the verbal and the non-verbal. “Expression”
should be seen as a family-resemblance concept, where we can find similarities in criteria for application not between independently understood
common features but in aspects of given unities.

References
Anscombe, G. E. M. Intention. Paperback ed. Cambridge, MA: Harvard University Press, 2000.
Bäckström, S. “Seeing People and Knowing You: Perception, Shared Knowledge
and Acknowledgment.” European Journal for Philosophy of Religion, 5/4
(2013): 55–73.
Bäckström, S. and M. Gustafsson. “Skill, Drill, and Intelligent Performance:
Ryle and Intellectualism.” Journal for the History of Analytical Philosophy, 5/5
(2017): 41–55.
Bar-On, D. “Expressive Communication and Continuity Skepticism.” The Journal
of Philosophy, 110/6 (2013): 293–330.
Boyle, M. “Additive Theories of Rationality: A Critique.” European Journal of
Philosophy, 24/3 (2016): 527–555.
———. “Essentially Rational Animals.” In Rethinking Epistemology, eds. G. Abel
and J. Conant. Vol. 2 of Berlin Studies in Knowledge Research. Berlin and Boston: De Gruyter, 2012: 395–428.
Crary, A. Inside Ethics: On the Demands of Moral Thought. Cambridge, MA and
London: Harvard University Press, 2016.
Ford, A. “Action and Generality.” In Essays on Anscombe’s Intention, eds. A.
Ford, J. Hornsby, and F. Stoutland. Cambridge, MA: Harvard University Press,
2011: 76–104.
———. “The Arithmetic of Intention.” American Philosophical Quarterly, 52/2
(2015): 129–143.
Halbig, C. “Varieties of Nature in Hegel and McDowell.” In John McDowell:
Experience, Norm, and Nature, ed. J. Lindgaard. Oxford: Blackwell, 2008:
72–91.
McDowell, J. Mind and World. Cambridge, MA: Harvard University Press, 1994.
———. Perception as a Capacity for Knowledge. Milwaukee: Marquette University Press, 2011.
———. “Responses.” In John McDowell: Experience, Norm, and Nature, ed. J.
Lindgaard. Oxford: Blackwell, 2008: 200–267.
Rödl, S. Categories of the Temporal. Cambridge, MA: Harvard University Press,
2012.
Teichmann, R. “The Identity of a Word.” American Catholic Philosophical Quarterly, 90/2 (2016): 317–335.
Thompson, M. “Anscombe’s Intention and Practical Knowledge.” In Essays on
Anscombe’s Intention, eds. A. Ford, J. Hornsby, and F. Stoutland. Cambridge,
MA: Harvard University Press, 2011: 198–211.
Wittgenstein, L. Philosophical Investigations, 3rd ed., trans. G. E. M. Anscombe.
Oxford: Blackwell, 2001.
———. Remarks on Colour, ed. G. E. M. Anscombe, trans. L. L. McAlister and
M. Schättle. Oxford: Basil Blackwell, 1977.

Part IV

Naturalism and
Meta-Philosophy

11 Wittgenstein, Hume,
and Naturalism
Benedict Smith

Introduction
One of the many things that makes Wittgenstein’s philosophical work
distinctive is the sustained interest in the nature of philosophy itself. This
interest is expressed throughout his writings from the earliest notes and
drafts of the Tractatus to his remarks up to the late 1940s and including the months before he died. It is unusual to find a person for whom
the character of philosophy received so much attention and over such
a sustained period of time. Throughout his writings, albeit in different
ways, Wittgenstein insisted on the independence of philosophy from science. And that insistence has been taken as evidence that he was an antinaturalist. That would follow, however, only if naturalism is a view about
how philosophy should privilege science. But that understanding of naturalism is contestable. Clarifying Wittgenstein’s relation to naturalism is
not easy but one way is to consider how his work aligns with the naturalism of Hume. Initially, this may seem a questionable strategy, since the
two appear to differ substantially over basic philosophical commitments
and are seemingly quite opposed to one another on questions of methodology. But a comparison with Hume is illuminating in a number of ways.
For some, it is not obvious that Wittgenstein is any kind of naturalist at
all, let alone one who shares that view with a figure often interpreted to
have championed what is now known as scientism. And there are broader
issues too. If Hacker is right Hume “made almost every epistemological and metaphysical mistake Wittgenstein could think of”.1 There are
indeed many ways that Hume’s work contrasts with Wittgenstein’s but
there are deep affinities too. These affinities can be overlooked because
of a tendency to misread Hume’s naturalism as a forerunner to what now
counts as scientific naturalism and, relatedly, to construe Wittgenstein’s
suspicions with regard to the use of scientific method in philosophy to

1 P.M.S. Hacker, Insight and Illusion: Themes in the Philosophy of Wittgenstein (Oxford:
Clarendon Press, 2009), 218. See also O. Hanfling, “Hume and Wittgenstein,” Royal
Institute of Philosophy Lectures 9 (1975): 47.

244

Benedict Smith

express an anti-naturalism. Against this last idea, Strawson writes of how
the resemblances between them and the ‘echoes’ of Hume in Wittgenstein’s work are more striking than the differences.2 And I think in the
context of clarifying Wittgenstein’s naturalism the affinities are particularly instructive as are the lessons for how we might characterize contemporary philosophical naturalism more broadly.
Pears describes Wittgenstein’s naturalism as inspired by a conception
of philosophical method which consists in portraying what we find in
daily life;3 an echo, I take it, of Hume’s conception of naturalistic method
as grounded in the “cautious observation of human life” and of characterizing the features of our lives “as they appear in the common course
of the world”.4 There is (almost) no opposition to the idea that Hume
is a naturalist although there are significant differences over which kind
of naturalism is most relevant. As indicated, one approach is to portray
him as paving the way for scientism and thus as quite inconsistent with
Wittgenstein’s anti-scientism. But there are other kinds of naturalism in
Hume, those that have been interpreted as more ‘liberal’ and at least nonscientistic in the sense developed by a number of contemporary authors.5
Despite other changes during the development of Wittgenstein’s thought
a theme that persists is the insistence that philosophy and science are
distinct. That view is taken to expresses a clear anti-naturalism. According to Flanagan, for example, Wittgenstein is anti-naturalist because he
denies that philosophy is more or less related to any of the natural sciences: psychology, and Darwinism, for example, are taken by Wittgenstein to be both equally independent of philosophy and there is no sense
that philosophy can be informed at all by any of the natural sciences.6
Flanagan chooses to use Quine’s view as the arbiter of what counts as
naturalism, at least for the purposes of characterizing Wittgenstein as an

2 P.F. Strawson, Skepticism and Naturalism: Some Varieties (London: Methuen & Co.,
1985), 14. For others that emphasize the affinities between Hume and Wittgenstein see,
for example, P. Jones, “Strains in Hume and Wittgenstein,” in Hume: A Re-evaluation,
eds. Donald W. Livingstone and James T. King (New York: Fordham University Press
1976), 191–209; B. Williams, Truth and Truthfulness: An Essay in Genealogy (Princeton,
NJ: Princeton University Press, 2002).
3 D. Pears, “Wittgenstein’s Naturalism,” The Monist 78, no. 3 (1995): 411.
4 D. Hume, A Treatise of Human Nature (Oxford: Oxford University Press, 1978), xix.
Doubtless there are other ‘naturalisms’ in Hume that would fit uneasily with this kind,
those that motivate Hacker’s remark above, for example. There has been a tendency to
underplay the diversity of Hume’s view in this regard.
5 See for example, the essays in M. de Caro, and D. Macarthur, eds., Naturalism in Question (Cambridge, MA: Harvard University Press, 2004) and M. de Caro, and D. Macarthur, eds., Naturalism and Normativity (New York: Columbia University Press, 2010).
6 O. Flanagan, “Varieties of Naturalism,” in The Oxford Companion to Religion and Science eds. Philip Clayton and Zachary Simpson (Oxford: Oxford University Press, 2006),
432.

Hume

245

anti-naturalist.7 But Quine’s view need not be taken as the arbiter and
there are a number of reasons—some of them Humean—to resist doing
so. Despite Quine’s appeal to Hume as a forerunner of the “epistemological enterprise in [its] new psychological setting”,8 the latter would not,
I think, easily recognize the idea that epistemology studies “a physical
human subject”.9 The principal subject matter of what Hume calls “the
science of human nature”10 is not a physical human subject but persons;
not transitions between experiential input in the form of “patterns of
irradiation in assorted frequencies” and conceptual output in the form
of descriptions of the world and its history as, Quine puts it.11 Hume’s
naturalism encompasses more than a psychological characterization of
belief-formation.
In the next section, I consider Bernard Williams’s remarks about how
Wittgenstein rejected explanation and conceived the subject matter of
philosophy as exclusively a priori. I suggest that Wittgenstein had a
nuanced view about what explanation can provide and, in the following section, develop how that is related to experience and belief. I then
sketch the implications for ways that we can understand philosophical
naturalism.

1. The Subject Matter of Philosophy and the a Priori
The scientistic naturalism to which Wittgenstein is opposed involves,
amongst other things, a “smug and unexamined assurance that what
wants explanation is obvious, and that scientific tools are immediately
applicable”, as Goldfarb puts it.12 But in the course of exposing that assurance, Wittgenstein’s view allegedly incorporates a problematic opposition
to explanation as such; problematic since it seems to embody a dogmatic
skepticism about the value of any empirical investigation in the context
of clarifying and responding to philosophical problems. And this assumes
a distinctive view about the subject matter of philosophy. A related interpretation is proposed by Bernard Williams who suggests that according

7 Flanagan also suggests that Wittgenstein’s ‘anti-naturalism’ is allegedly inferior to other
versions, O.K. Bouwsma’s for example, because Wittgenstein merely stipulates antinaturalism and does not argue for it. I have discussed Wittgenstein’s relation to scientism
in B. Smith, “Wittgenstein, Naturalism and Scientism,” in Wittgenstein and Scientism,
eds. Jonathan Beale and Ian James Kidd (London: Routledge, 2017), 209–224.
8 W.V.O. Quine, “Epistemology Naturalized,” in Knowledge: Readings in Contemporary
Epistemology, eds. Sven Bernecker and Fred Dretske (Oxford: Oxford University Press,
2000), 274.
9 Quine, “Epistemology Naturalized,” 273.
10 Hume, Treatise, xvii.
11 Quine, “Epistemology Naturalized,” 274.
12 W. Goldfarb, “Wittgenstein, Mind, and Scientism,” The Journal of Philosophy 86, no.
11 (1989): 367.

246

Benedict Smith

to Wittgenstein “philosophy had nothing to do with explanations—
not merely scientific explanations . . . but any explanations at all”. As a
result, Wittgenstein regarded the subject matter of philosophy as being
“exclusively a priori”.13 But the view that philosophy’s subject matter
is a priori in the sense Williams intends cannot, it seems to me, be easily
accommodated by Wittgenstein’s naturalism. To show why involves considering what it is about explanation, particularly empirical explanation,
that Wittgenstein opposed.
Wittgenstein famously suggests that what it takes to frame our subject
matter in the right way is ‘description’ since explanation, a core component
in the scientific attitude, is to be done away with.14 Description is a process through which we remind ourselves of the familiar ordinary practical
contexts within which meaning, understanding, word-use, concept application, and so on, are embedded and inextricably entwined with our agency.
According to Wittgenstein it is this familiarity that makes our subject matter
potentially hard to get into focus: “One cannot notice something because it
is in front of one. Our subject matter is there if only we remove the prejudice that obscures it”.15 Our ‘disquietudes’ are not properly understood let
alone addressed by providing explanations that attempt to reach behind
the phenomena, as it were. Rather, they are addressed through achieving
the right kind of perspective on what is already there as it is constituted
in the midst of our practical existence. So in that sense, we are reminded of
what we are already familiar with and understand, since our practices—
our lives—are partly made by such understanding. This is unlike a different, more ‘scientific’, subject matter which is not already partly constituted
by our practices, and thus illuminating it cannot be achieved by “description alone”.16 It is the failure to differentiate the subject matter in each
case that results in a tendency to treat our philosophical questions as if
they were scientific in structure and thus admit of explanations that, when
successful, identify new and perhaps quite unfamiliar facts. The scientific
attitude, one that embodies a certain kind of explanatory urge, assumes
that we need to look past the manifest subject matter to something more
basic “as if we had to penetrate phenomena”.17
The a priori in Williams’s sense concerns a subject matter that contains
potential objects of understanding that are fully intelligible independently
of all experience. But adverting to what is intelligible in this sense is a
priori in a problematic way. It is possible to employ this use of a priori, as

13 B. Williams, “Philosophy as a Humanistic Discipline,” Philosophy 75 (2000): 493. See
also Williams, Truth and Truthfulness, 283 n.23.
14 Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe, 2nd ed.
(Oxford: Blackwell, 1997), §109.
15 PI §340.
16 PI §109.
17 PI §90.

Hume

247

Williams does, but it arguably encompasses too much. For example, this
a priori would also capture a significant part of how Hume characterizes
his basic methodological approach and thus it brings into question its
suggested relevance in the specific case of Wittgenstein. I will come back
to this in the next section. But the opposition between explanation and
description is not best read, I think, as expressing skepticism about explanation as such, as if our subject matter is strangely inexplicable whilst
being at the same time describable. The skepticism about explanation is
directed to a particular kind of attitude toward inquiry not a rejection of
the idea that the features under question can be explained at all.
Wittgenstein does not provide much explicit detail as to which kinds of
explanation his critical remarks are directed against and to which they are
not. At any rate, the idea that he was somehow simply anti-explanation
is not obviously right. It would be peculiarly un-Wittgensteinian to suppose that there is something illegitimate in the very idea of explanation;
un-Wittgensteinian partly because that would itself rely on a form of
essentialism about concepts and our forms of understanding that he
clearly rejected. Just as there are many things we call ‘games’, “there are
all sorts of things we call “explanation of meaning”.18 So we ought to
be mindful of the complex status of explanation, a complexity that might
be overlooked if one considers only those remarks that appear to dismiss
the value of explanation outright.
In The Big Typescript, for example, and in subsequent work, Wittgenstein often explores the relation between explanation and understanding, particularly the way that explanation ‘correlates’ with understanding
as he puts it. In some cases explanation can potentially exorcize misunderstanding and hence is used in a different way to how explanation is
often referred to in the Philosophical Investigations.19 In the latter work,
explanation takes a more specific and a particularly empirical shape and,
to that extent, it was deemed misleading in the context of addressing
philosophical problems. Despite some inevitable vagueness, Wittgenstein
had a specific form of explanation as the principal target of his criticism
and in a way that was connected with his proposals that an alternative attitude, description, would faithfully preserve our subject matter.
But preservation is, I take it, an activity that can often demand a good
deal of work, one that involves “clearing misunderstandings away”,20
and so the preservation in question is more like the conservation of a
living dynamic natural environment, not like preserving an object in a
form of suspended animation such as a cryogenic stasis. That can seem a
peripheral point, but it highlights how the wider context of Wittgenstein’s
18 Ludwig Wittgenstein, The Big Typescript TS 213, eds. and trans. C.G. Luckardt and
M.A.E. Aue (Oxford: Blackwell, 2005), 59.
19 BT 17.
20 PI §90.

248

Benedict Smith

skepticism about explanation does not imply a dogmatic refusal to seek
ways to change and advance our understanding.
A key theme in the Philosophical Investigations is resistance to the idea
that grasp of the meaning of a concept amounts to possessing the right
kind of mental state, a state that constitutes knowing how to extend a
series of numbers according to a learned formulation, say. Wittgenstein’s
suggestive remark in this context that the grammar (in his sense) of
‘know’ is related to that of ‘can’ and ‘is able to’21 points to the essentially
practical character of what understanding amounts to, an understanding that is not locatable inside a person’s head or across instances of
past behavior. The investigations of rule following involve a dialectic centered on the differing perspectives of instructor and learner in order to
reveal the character of understanding, in particular the kind of practical
context-bound mastery a person comes to take on, an ability that exhibits
a change in the “way of looking at things”.22 This change is grounded in a
capacity to compare cases and examples and is reflected in what we come
to know through a grammatical investigation: “the kind of understanding that consists in ‘seeing connexions’ ”.23
The connections here are not the ones posited by rival empirical
explanations, connections that purportedly underpin the phenomena in
question: The connections and our sensitivity to them constitute the phenomena when regarded by a clear view enabled by a grammatical investigation. In the Remarks on the Foundations of Mathematics Wittgenstein
writes that in the face of disquietudes “[p]hilosophical dissatisfaction
disappears by our seeing more”.24 So the question of development or
in some sense a change of view on our subject matter is not in question.
What is problematic is to attempt a change of view by using explanations
of the sort typically employed in scientific practice and illicitly borrowed
for philosophical purposes. So there are forms of activity, particularly
being able to compare cases, to catch on to patterns and form awareness
of how examples are related, that Wittgenstein thinks are crucial and
irreducible components in what constitutes understanding but resist, at
least in the relevant sense, explanation.
For Pears, there is no reason for Wittgenstein, or anyone else, to rule
out a priori from our investigations the value of empirical findings, at
least from our responses to the problems that we have discovered even
if they arise from the misunderstanding of language.25 This is a similar

21
22
23
24

PI §150.
PI §144.
PI §122.
Ludwig Wittgenstein, Remarks on the Foundations of Mathematics, eds. G.H. von
Wright, R. Rhees, and G.E.M. Anscombe, revised ed. (Oxford: Blackwell, 1978), II 85.
Wittgenstein’s emphasis.
25 Pears, “Wittgenstein’s Naturalism,” 423.

Hume

249

idea to Williams’s. Both Pears and Williams regard Wittgenstein’s suspicions about the ‘scientific way of thinking’ to dogmatically rule out
the very idea that the accumulation of facts and empirical explanations
could ever contribute to our understanding. But Williams also notes that
there are kinds of explanation that Wittgenstein embraced, “philosophical explanations”, which are distinctive since these are “like elucidations
or reminders”.26
I am suggesting that there is more to Wittgenstein’s apparent animus
towards explanation than unconditional rejection and also that there is
a dynamic role for the kinds of explanation that Wittgenstein acknowledged. A ‘philosophical explanation’ or an elucidation can, I think, be
more than a simple reminder but involve a process of coming to a new
kind of understanding even if that is not one grounded in awareness of
new facts lying behind the phenomena in question. In my view, elucidatory or philosophical explanations in Williams’s sense are also forms of
naturalistic explanation but in which there is no place for an alleged
explanans that underpins what otherwise “lies open to view”.27 Elucidatory explanations are normative to the extent that they help to make
phenomena intelligible by contextualizing them in amongst our lived
experiences which constitute their “original home”.28
At one point in the Philosophical Investigations, Wittgenstein characterizes the way in which potential explanations of meaning must make
use of language “full-blown”.29 That is, language, as the etymology suggests, as fully developed, as in blossom as it were. Competence with
providing and receiving normative explanations requires immersion in
and familiarity with a ‘language-game’. Wittgenstein emphasizes here, as
elsewhere, that what comes to be embodied as a form of agency rooted
in enculturation reflects how (some) “explanation has its foundation in
training”.30 Again, a contrast would be with the scientistic “causal point
of view”31 that cannot make intelligible the normative possibilities of
meaning that run through our understanding that is exhibited in ordinary
life.32 Wittgenstein’s non-reductive naturalism, just as Hume’s, conceives
our subject matter as irreducibly situated in socially articulated practical
contexts and risks irreparable distortion if detached from them. That the
26 Williams, “Philosophy as a Humanistic Discipline,” 493.
27 PI §126.
28 PI §116. For Hanfling, these are ‘manifest explanations’ in contrast to the explanations
of science that have ‘hidden’ facts as their target. See O. Hanfling, Wittgenstein and the
Human Form of Life (London: Routledge, 2002), 93.
29 PI §120.
30 Ludwig Wittgenstein, Zettel, eds. G.E.M. Anscombe and G.H. von Wright (Berkeley:
University of California Press, 1967), §419.
31 Ludwig Wittgenstein, Culture and Value, rev. 2nd ed. ed. G.H. von Wright, trans. Peter
Winch (Oxford: Blackwell, 1998), 45.
32 PI §90.

250

Benedict Smith

subject matter of philosophy includes the taken for granted conditions
of human life presupposed by the explicit activities of reflecting on what
we ordinarily do is, in some sense, to characterize that subject matter as a
priori, as available for thought independently from experience. But until
that sense is clarified we risk missing the other ways that Wittgenstein’s
thought regarded experience as inalienable to the project of clarifying
the subject matter of philosophy and of illuminating how meaning and
understanding permeate our lives.

2. Hume and Wittgenstein on Experience and Belief
In §97 of the Philosophical Investigations Wittgenstein writes that, like
other words such as ‘language’ and ‘world’, if the term ‘experience’ is to
have a use “it must be as humble a one as that of the words ‘table’, ‘lamp’,
‘door’ ”. This suggests that we understand ‘experience’ to be ‘full-blown’
in the sense suggested in the previous section. That is, that in referring to
experience and its role in meaning and understanding we need not use it
in a narrow, peculiarly epistemological sense, even if philosophers have
characteristically done so in a way that Wittgenstein found objectionable.
By the epistemological sense of experience, I mean, roughly, the idea of
experience as playing a grounding or foundational role in our understanding. The discussion of what it is to grasp and follow a rule is again
helpful. This epistemic sense of experience mistakenly portrays what it
is to grasp a rule as an experience—a “special experience”—internal to
the mind of a person in the midst of training, an experience that supposedly marks the transition between unsuccessful and successful grasp of
a rule.33 But experience understood in an ordinary and ‘humble’ sense
refers to a more general, socially articulated meaning that reconnects
with the dimension of our practical embodied agency. So when he writes
at §655 that “[t]he question is not one of explaining a language-game by
means of our experiences, but of noting a language-game”, what is problematic is the aspiration to use experience as somehow an explanation or
justification for outward behavior. Noting a language-game is an activity that takes in our experiences in a wider sense, that captures what we
say, think or do in the course of ordinary life; a more phenomenological
characterization of experience.34

33 PI §155.
34 In 1974, M. O’C. Drury wrote of an exchange with Wittgenstein in which the latter
described his work as “phenomenology”. See M.O’C. Drury, “Conversations with Wittgenstein,” in Recollections of Wittgenstein, ed. R. Rhees (Oxford: Oxford University
Press, 1984), 116. See also Rhees’s comment at x–xi. At the time in question (c.1930),
Wittgenstein was working on notes that were later published as Philosophical Remarks
in which he makes the connection between his work and phenomenology. In other
places, The Big Typescript, for example, Wittgenstein also suggests that a grammatical

Hume

251

Of course, the aim here is not to assess the complex connections
between Wittgenstein and the phenomenological tradition but to gain
a clearer view of the role of experience, in the way I am suggesting
that it might understood, in relation to Hume’s naturalism. Roughly, a
widespread view is that Hume’s ambition was to provide a scientifically
respectable account of belief acquisition in the form of explaining the
experiential foundation of our ideas in ways that avoid abstruse metaphysical assumptions. That is Hume’s so-called explanatory aim to be
realized through the ‘experimental method’ that he advocated. On the
other hand, Hume aimed to clarify the normative or rational credentials
of belief by assessing whether belief can be justified, principally by one of
two routes: by sense experience (‘the present testimony of sense’) or by
a form of conceptual analysis (‘relations of ideas’). Notoriously many of
our basic beliefs such as that the external world exists, that causal relations hold between objects, that objects exist when unperceived, cannot
be justified and thus a form of skepticism follows. An influential interpretation holds that Hume’s explanatory aim contains the resources to
respond to, if not refute, skepticism: Unavoidable psychological mechanisms inevitably lead to the formation of belief such that the search for
independent rational justification is shown to be incoherent.35 Nevertheless, Hume’s view is that the process of belief acquisition leads to our
beliefs becoming ‘stable’ in a way that reliably connects them to truth.
Hume has a naturalistic theory of justification, then, since truth-oriented
stability is explained as grounded in psychological dispositions.36 And,
according to Pears, had Hume known how the brain works his account
would have moved “beyond psychology into neurology”.37 But it is
questionable whether Hume’s commitment to the ‘experimental method’
implies a naturalistic reduction in this sense.
Hume appealed to features of our experience which presupposed an
interpersonal environment, claims to which he assumed the community

investigation is a form of phenomenological investigation. There have been a number
of attempts to unravel the intriguing connection that Wittgenstein made with phenomenology. See, for example: T.N. Munson, “Wittgenstein’s Phenomenology,” Philosophy
and Phenomenological Research 23, no. 1 (1962): 37–50; H. Spiegelberg, “The Puzzle
of Wittgenstein’s Phänomenologie,” American Philosophical Quarterly 5, no. 4 (1968):
244–256; N.F. Gier, Wittgenstein and Phenomenology: A Comparative Study of the
Later Wittgenstein, Husserl, Heidegger and Merleau-Ponty (Albany: SUNY Press,
1981) and, more recently, R. Monk, “The Temptations of Phenomenology: Wittgenstein, the Synthetic a Priori and the ‘Analytic a Posteriori’,” International Journal of
Philosophical Studies 22, no. 3 (2014): 312–340.
35 For example, N. Kemp Smith, The Philosophy of David Hume (London: MacMillan,
1941).
36 For example, L. Loeb, Stability and Justification in Hume’s Treatise (Oxford: Oxford
University Press, 2002).
37 Pears, “Wittgenstein’s Naturalism,” 412.

252

Benedict Smith

would readily assent on the basis of reflection. Yet what ‘experience’
means for Hume is not straightforward, taking in experience that is introspectively accessible by an individual but also experience that cannot be
made intelligible other than in social contexts. The latter assumes a prior
intersubjective understanding that could be made explicit by reflection.
It is on the basis of prior understanding shared across communities, even
humanity perhaps, that one might be persuaded of Hume’s views about,
say, the role of sentiment in ethics and of the role of perceptual anticipation
in the context of causal inference. According to Hanfling, the psychology
invoked by Hume is not that of empirical science but “introspective and
in that sense a priori. He wants us to notice that certain things are as he
says, not to learn it from him”.38 Under one light, Hume’s naturalism is
a commitment to—even laying the foundations for—what is now known
as scientism. Under another, it is a naturalism that consists in attending
to our lived experiences and practice.39 If the latter is right, then Hume’s
method, whilst naturalistic, is non-reductionist and more descriptive in a
Wittgensteinian sense. What one comes to notice from such descriptions
are forms of experience that constitute ordinary life and thus the subject
matter of our investigation and the experimental method used to investigate it can be viewed as naturalistic but non-scientistic.
What Hume considered as ‘experiments’ and ‘experimental reasoning’
included diverse reflections on human life that often included considering
historical events and practices:
[R]ecords of wars, intrigues, factions, and revolutions are so many
collections of experiments, by which the politician or moral philosopher fixes the principles of his science, in the same manner as
the physician or natural philosopher becomes acquainted with the
nature of plants, minerals, and other external objects, by the experiments which he forms concerning them.40
There is then a connection between philosophy and natural science in so
far as we need to form understanding in light of experiments. But there
is no suggestion of reducing what Hume’s calls “the science of man”41
or sometimes “the science of human nature”42 to natural philosophy. In

38 O. Hanfling, “Hume and Wittgenstein,” Royal Institute of Philosophy Lectures 9
(1975): 51.
39 For further discussion see B. Smith, “Naturalism, Experience, and Hume’s ‘Science
of Human Nature’,” International Journal of Philosophical Studies 24, no. 3 (2016):
310–323.
40 D. Hume, Enquiries Concerning Human Understanding and Concerning the Principles
of Morals (Oxford: Oxford University Press, 1975), 83–84.
41 Hume, Treatise, xv.
42 Hume, Treatise, xvii.

Hume

253

fact, quite the contrary. In the very last section of the Philosophical Investigations Wittgenstein warns against the assumption that the existence of
“experimental methods” in psychology contain the resources for solving
the problems “which trouble us”; “problem and method pass each other
by”, he writes.43 One might see a Humean view as the target, yet doing
so would assume a narrow conception of what Hume meant by experiment and experimental method. The science of man qua observation of
ordinary life is basic and in fact presupposed by natural philosophy and
other forms of enquiry: “[T]he science of man is the only solid foundation for the other sciences”.44
Since the science of man in Hume’s sense is not dealing with inanimate
objects in the world that can be manipulated and placed in experimental
settings at will in order to “observe the results”, a different approach is
needed. But in such an approach a form of observation is nevertheless
still central. Hume writes:
We must therefore glean up our experiments in this science from a
cautious observation of human life, and take them as they appear in
the common course of the world . . . Where experiments of this kind
are judiciously collected and compared, we may hope to establish
on them a science, which will not be inferior in certainty, and will
be much superior in utility to any other of human comprehension.45
So, as we should expect, the philosophical study of human life and our
practices is a different activity from observing how inanimate phenomena
interact under the laws of physics. There is something distinctive about
the source and object of the observations in the context of human life
since what we are aiming at is an understanding of ourselves.
In Hume’s treatment of causal belief one angle is a skeptical one: An
impression of causation cannot be located as the experiential source of
the relevant idea and, since denying a casual relation does not lead to
a contradiction, the justificatory possibilities provided by Hume’s twofold distinction alluded to above are exhausted. But there is also a different angle, one that seeks to uncover a non-rational “determination to
carry our thought from one object to another” embodying our practice
of causal inference.46 So one epistemic method of assessing causal inference is to “trace up” the origin of belief in the attempt to identify the
relevant impression, a form of genetic analysis of ideas. An alternative
approach illuminates our “natural instinct” for belief, as Hume put it;47

43
44
45
46
47

PI II, xiv.
Hume, Treatise, xvi.
Hume, Treatise, xiii–xix.
Hume, Treatise, 165. See also Hanfling, “Hume and Wittgenstein,” 59.
Hume, Enquiries, 151.

254

Benedict Smith

an example of how “Nature, by an absolute and uncontroulable necessity has determin’d us to judge, as well as to breathe and feel”.48 That
insight is not part of a refutation of skepticism but rather describes what
is presupposed by it.
Hume writes that believing in the deliverances of sense-perception and
our believing in an independent external world is a pre-rational natural
instinct:
It seems evident, that men are carried, by a natural instinct or prepossession, to repose faith in their senses; and that, without any reasoning, or even almost before the use of reason, we always suppose an
external universe.49
This capacity for what commentators have called “natural belief”50 is
“antecedently implanted in the mind, and render’d unavoidable”.51
However, the “slightest philosophy” can destroy the “primary instincts
of nature” and lead us to “embrace a new system” according to which
mind and world are, from this new, detached perspective, irrecoverably
alienated from one another.52
In Hume’s presentation, the term belief is consistently and deliberately
employed to characterize this pre-rational natural instinct and that a
principal aim of Book 1 of the Treatise is to demonstrate how “belief is
more properly an act of the sensitive, than of the cogitative part of our
natures”.53 This at least indicates, I think, that Hume’s naturalism is not
the imposition of a scientific way of thinking in an attempt to understand
our practices. Rather it involves showing how the “maxims of common
life” are apparently subverted by philosophical commitments driven by
a particular and peculiarly philosophical style of questioning. Put in a
Wittgensteinian light, this subversion affects what is then regarded as
acceptable responses to the questions that arise. These responses must
take the form of explanations of the kind Wittgenstein is suspicious of.54
Under a restrictive conception of what naturalism amounts to one might
assume that Hume is imagining that the unavoidable character of belief
in the external world, say, is a fact that can be respected only by reading into the view a kind of proto-Quineanism, an account of belief formation immunized from the fruitless endeavor of rational or normative

48
49
50
51
52
53
54

Hume, Treatise, 183.
Hume, Enquiries, 151.
Kemp Smith, David Hume, 114.
Hume, Treatise, 183.
Hume, Enquiries, 152.
Hume, Treatise, 183.
The ‘must’ here is a feeling symptomatic of the “causal point of view”. See CV, 45.

Hume

255

assessment and which instead invokes sub-personal mechanical processes.55 Yet what Hume is emphasizing, on the interpretation suggested here,
is that belief-formation is a practical achievement that requires immersion in a social world and in common life. “[P]hilosophy.”, writes Hume,
“finds herself extremely embarrassed” by how toothless our reasoning is
in the context of responding to skepticism, a skepticism that philosophy
animated in the first place.56 But it is a matter of contention, or at least
it should be, what form Hume’s naturalism adopts here as a response.
The detached perspective of philosophical reflection and the skepticism
it necessarily induces need not be countered by appealing to brute causal
processes, as if that is the aspiration of Hume’s science of man. Rather,
the naturalism prioritizes the origin and role of belief in the context of
ordinary life which, for Hume, is irreducibly interpersonal.
What we are given is not some reductive account of belief but a description and characterization of a relevant phenomenon, in this context the
phenomenology of belief. According to Hanfling, part of what Hume is
doing here is to attend to the circumstances in which we acquire beliefs
and call something, for example, a cause.57 These are circumstances in
which our natural belief is formed and need not require the operation of
any explicit intellectual endorsement. These are also circumstances which
are world-involving and interpersonal, not just the circumstances in a
person’s brain. To say that a caused b is an accomplishment not of reason, nor of any direct experience of a worldly cause but of custom which
operates, unlike a brute causal process, against a background of training
provided and regulated by other persons in the context of inhabiting a
shared world. There are different kinds of belief that Wittgenstein discusses and one of them, famously, lies at the “foundation of well-founded
belief” but is itself “not founded” as he describes it in On Certainty.58
Such belief, it seems to me, is an ‘echo’ in Strawson’s sense of Hume’s
description of the role of belief construed as natural instinct. According to Wittgenstein “believing is a natural act for humans”,59 and one
that requires a taken for granted background of human life. Despite
important differences between Hume and Wittgenstein, there is a sense in
which they are both concerned with what is presupposed by the activity
of explicit reflection, what is needed in order for our reflection to have
a subject matter in the first place. This, I have suggested, is constituted
by our ordinary experience. Wittgenstein regarded the subject matter of
philosophy as a priori in a sense that marks out empirical investigation

55
56
57
58

See Quine, “Epistemology Naturalized”.
Hume, Enquiries, 152.
Hanfling, “Hume and Wittgenstein,” 58.
Ludwig Wittgenstein, On Certainty, eds. G.E.M. Anscombe and G.H. von Wright,
trans. Denis Paul and G.E.M. Anscombe (Oxford: Blackwell, 1969), §253.
59 BT 575.

256

Benedict Smith

as inappropriate for informing philosophical thought. But this idea does
not encompass ordinary experience to which Wittgenstein assigned a fundamental role.

3. Concluding Remarks: Naturalism and Philosophy
Wittgenstein’s naturalism can be articulated, as can Hume’s, as a commitment to the philosophical importance of ‘the everyday’.60 For both Hume
and Wittgenstein the everyday is the irreducible context which is presupposed by our questions, doubts, reasons and responses and inattentiveness to the taken for granted background of ordinary life can lead to a
gross distortion of our subject matter. This background to our thought
and action resists being accommodated from within a naturalistic outlook
if that outlook is restricted to scientific naturalism. Hume wrote of the
philosophical importance of “the gross earthy mixture” of common life,
Wittgenstein the “rough ground” of day to day practice and the “phenomena of every-day”.61 And both insisted that we return to this subject
matter, to return from the “fairy land”62 of metaphysics or to recall language from the “holiday” on which it goes thereby generating philosophical problems.63
Seen from the point of view of common life, from the perspective of
embodied, interpersonal experience, philosophical problems (particularly
skepticism) seem “so cold, strain’d, and ridiculous” as Hume put it.64
But by this Hume was not suggesting that philosophy as such is thereby
shown to be suspect. Rather, philosophy as well as many other aspects of
our practices together such as having friends and family, working, relaxing, learning, creating, investigating, and so on, are part of what a flourishing human life consists in. This reflects how
[N]ature has pointed out a mixed kind of life as most suitable to
human race . . . Indulge your passion for science, says she, but let
your science be human, and such as may have a direct reference to
action and society.65
After warning of the “pensive melancholy [and] endless uncertainty” that
“abstruse thought and profound researches” will bring on, Hume ends
here with the famous line: “Be a philosopher; but, amidst all your philosophy, be still a man”. Hume seemed to think that the mixed kind of life

60
61
62
63
64
65

See Williams, Truth and Truthfulness, 10.
Hume, Treatise, 272. PI §107, §436.
Hume, Enquiries, 72.
PI §38.
Hume, Treatise, 214.
Hume, Enquiries, 9.

Hume

257

was one that somehow struck a balance between the different elements
that constitute it and, whilst he did not think that each of us is equally
suited to or even capable of pursuing philosophical thought, to be drawn
to philosophical questions was a natural aspect of the human condition.
According to Hume, reason, which inevitably leads to skeptical questions, is nevertheless employed “only because we cannot help it”;66 but
even if this is right, it leaves open the possibility that a liveable balance
might be struck between philosophical inquiry and the fruitful participation in other activities. The mixed life that Hume imagined involved
a domestication of reason and so of philosophy, a view not so easily
accommodated by Wittgenstein.
According to Williams,
Wittgenstein inherited from Kant a concern with the limits of understanding, from Frege and Russell an interest in the conditions of linguistic meaning, and from himself a sense of philosophy as a quite
peculiar and possibly pathological enterprise.67
Philosophy may be possibly pathological but not, I take it, because it is
philosophy but because of a particular conception of its aims and methods or of a particular way of inhabiting a philosophical perspective. As
discussed Wittgenstein was critical of the idea that scientific method is
in any way appropriate for identifying and responding to philosophical
problems. The way that naturalism is now often understood assumes the
opposite view: that philosophy is or should be part of science in some
sense. As I have emphasized, what motivates the view that Wittgenstein
was anti-naturalist is the assimilation of naturalism to a scientific perspective. Whilst we need not countenance any such assimilation there is a
danger here from the other direction as it were.
Wittgenstein regarded the scientific attitude as pervasive, as shaping
our very conception of what explanation and understanding any phenomena consists in. But resisting that can also encourage the idea that
the existence of a philosophical problem as such is already the product
of a mistaken departure from ordinary life. To be sure, Wittgenstein saw
the difficulties—logical, conceptual, psychological, and moral—that arose
from the misguided ways that explanation and understanding are sought.
But that reflects how our understanding of thought and practice is under
the sway of the scientific attitude. Dissolving the general authority that scientific naturalism has been granted should not threaten philosophy itself.
Williams portrays Wittgenstein’s view of philosophy as a possibly
pathological enterprise as connected to the way the latter is said to have

66 Hume, Treatise, 657.
67 Williams, “Philosophy as a Humanistic Discipline,” 493.

258

Benedict Smith

conceived the subject matter of philosophy as exclusively a priori. I suggest that that is more or less right if understood as a point against assuming the supposed authority of empirical explanation. But it cannot easily
accommodate the fact that a fundamental subject matter of philosophy
is an impulse to understand ourselves as embodied practical agents. Wittgenstein’s way of contributing to this ambition was to not to regard our
subject matter as available empirically, yet neither did it assume that it
was exclusively a priori. For Wittgenstein, as for Hume, the careful observation of human life is central to a compelling version of philosophical
naturalism.68

References
Caro, M. de and Macarthur, D. ed. Naturalism and Normativity. New York:
Columbia University Press, 2010.
———. Naturalism in Question. Cambridge, MA: Harvard University Press,
2004.
Drury, M.O.’C. “Conversations With Wittgenstein.” In Recollections of Wittgenstein, edited by R. Rhees, 97–171. Oxford: Oxford University Press, 1984.
Flanagan, O. “Varieties of Naturalism.” In The Oxford Companion to Religion
and Science, edited by Philip Clayton and Zachary Simpson, 430–452. Oxford:
Oxford University Press, 2006.
Gier, N.F. Wittgenstein and Phenomenology: A Comparative Study of the Later
Wittgenstein, Husserl, Heidegger and Merleau-Ponty. Albany: SUNY Press,
1981.
Goldfarb, W. “Wittgenstein, Mind, and Scientism.” The Journal of Philosophy 86,
no. 11 (1989): 635–642.
Hacker, P.M.S. Insight and Illusion: Themes in the Philosophy of Wittgenstein.
Oxford: Clarendon Press, 2009.
Hanfling, O. “Hume and Wittgenstein.” Royal Institute of Philosophy Lectures 9
(1975): 47–65.
———. Wittgenstein and the Human Form of Life. London: Routledge, 2002.
Hume, D. Enquiries Concerning Human Understanding and Concerning the
Principles of Morals. Oxford: Oxford University Press, 1975.
———. A Treatise of Human Nature. Oxford: Oxford University Press, 1978.
Jones, P. “Strains in Hume and Wittgenstein.” In Hume: A Re-evaluation, edited
by Donald W. Livingstone and James T. King, 191–209. New York: Fordham
University Press, 1976.
Kemp Smith, N. The Philosophy of David Hume. London: MacMillan, 1941.
Loeb, L. Stability and Justification in Hume’s Treatise. Oxford: Oxford University
Press, 2002.

68 A version of this paper was presented in Bergen at the ‘Wittgenstein, Philosophy of
Mind, and Naturalism’ conference in 2015 organized by Kevin M. Cahill and Thomas
Raleigh. I am very grateful for the stimulating conversations that took place. In particular, my thanks to Stephen Burwood, Eugen Fischer, Thomas Raleigh, Bjørn Ramberg,
and especially to Kevin M. Cahill, who provided very helpful comments on a previous
draft.

Hume

259

Monk, R. “The Temptations of Phenomenology: Wittgenstein, the Synthetic a Priori and the ‘Analytic a Posteriori’.” International Journal of Philosophical Studies
22, no. 3 (2014): 312–340.
Munson, T.N. “Wittgenstein’s Phenomenology.” Philosophy and Phenomenological Research 23, no. 1 (1962): 37–50.
Pears, D. “Wittgenstein’s Naturalism.” The Monist, 78, no. 3 (1995): 411–424.
Quine, W.V.O. “Epistemology Naturalized.” In Knowledge: Readings in Contemporary Epistemology, edited by Sven Bernecker and Fred Dretske, 266–278.
Oxford: Oxford University Press, 2000.
Rhees, R. ed. Recollections of Wittgenstein, Oxford: Oxford University Press, 1984.
Smith, B. “Naturalism, Experience, and Hume’s ‘Science of Human Nature’.” International Journal of Philosophical Studies 24, no. 3 (2016): 310–323.
———. “Wittgenstein, Naturalism and Scientism.” In Wittgenstein and Scientism,
edited by Jonathan Beale and Ian James Kidd, 209–224. London: Routledge,
2017.
Spiegelberg, H. “The Puzzle of Wittgenstein’s Phänomenologie.” American Philosophical Quarterly 5, no. 4 (1968): 244–256.
Strawson, P.F. Skepticism and Naturalism: Some Varieties. London: Methuen &
Co., 1985.
Stroud, B. Hume. London: Routledge, 1977.
Van Peursen, C.H. “Edmund Husserl and Ludwig Wittgenstein.” Philosophy and
Phenomenological Research 20, no. 2 (1959): 181–197.
Williams, B. “Philosophy as a Humanistic Discipline.” Philosophy 75 (2000):
477–496.
———. Truth and Truthfulness: An Essay in Genealogy. Princeton, NJ: Princeton
University Press, 2002.
Wittgenstein, L. The Big Typescript TS 213, edited and translated by C.G. Luckardt and M.A.E. Aue. Oxford: Blackwell, 2005.
———. On Certainty, edited by G.E.M. Anscombe and G.H. von Wright translated by Denis Paul and G.E.M. Anscombe. Oxford: Blackwell, 1969.
——— Philosophical Investigations, translated by G.E.M. Anscombe, 2nd ed.
Oxford: Blackwell, 1997.
———. Zettel, edited by G.E.M. Anscombe and G.H. von Wright. Berkeley: University of California Press, 1967.

12 Wittgensteinian ‘Therapy’,
Experimental Philosophy, and
Metaphilosophical Naturalism
Eugen Fischer

Introduction
There is a striking difference between scientists and scientifically-minded
philosophers: Scientists have not been afraid of engaging with Wittgenstein. In cognitive psychology, for example, his ideas about familyresemblance concepts inspired theoretical and experimental work on
prototypes in categorisation.1 In computational linguistics, his idea that
questions about words’ meaning can be addressed by examining their use
inspired the research programme of distributional semantics,2 with seminal contributions by the group of his student Margaret Masterman.3 In
contrast, Wittgenstein’s apparently anti-naturalistic views on philosophical method have generally prevented meaningful engagement with Wittgensteinian ideas by philosophers who wish to deploy scientific methods
or findings. Conversely, many Wittgensteinian philosophers equated such
methodological naturalism with ‘scientism’ and largely ignored developments in the cognitive or other sciences.
This chapter will argue that the advent of experimental philosophy, the
most ‘hands-on’ form of methodological naturalism to date, facilitates a
fresh, mutually beneficial, dialogue between Wittgensteinian philosophers
and methodological naturalists:4 I will argue that a prominent strand of
experimental philosophy promotes a new kind of methodological natu-

1 See, for example, Eleanor Rosch, “Principles of Categorisation,” in Cognition and Categorization, eds. E. Rosch and B. Lloyd (Hillsdale, NJ: Erlbaum, 1978), 27–48, and J.A.
Hampton, “Testing the Prototype Theory of Concepts,” Journal of Memory and Language 34 (1995): 686–708.
2 See K. Erk, “Vector Space Models of Word Meaning and Phrase Meaning: A Survey,”
Language and Linguistics Compass 6 (2012): 635–653 and P.D. Turney and P. Pantel,
“From Frequency to Meaning: Vector Space Models of Semantics,” Journal of Artificial
Intelligence Research 37 (2010): 141–188.
3 See K. Spärck-Jones, Synonymy and Semantic Classification (PhD. diss., University of
Cambridge. Cambridge Language Research Unit, 1964).
4 This methodological stance is independent from naturalism as a metaphysical position. See J. Collins, “Naturalism Without Metaphysics,” in Experimental Philosophy,
Rationalism, and Naturalism, eds. E. Fischer and J. Collins (London: Routledge, 2015),
85–109. This paper is exclusively concerned with the methodological stance.

Experimental Philosophy

261

ralism, ‘meta-philosophical naturalism’, which is consistent with broadly
Wittgensteinian aims and strictures (Section 1). This new naturalism
facilitates research that can provide empirical foundations precisely for
the ‘therapeutic’ aspects of Wittgenstein’s conception of philosophy that
prima facie seem most antithetical to methodological naturalism (Section
2). Recent psycholinguistic research is a case in point: It can vindicate Wittgensteinian ideas about how an ‘urge to misunderstand’ engenders merely
apparent problems—like the classic ‘problem of perception’5—which call
for a therapeutic approach (Section 3). Using cognitive therapy6 as a model,
we will see that these insights into the cognitive sources of the problem
already provide the best part of what is needed for therapy (Section 4).
This argument will not address exegetical concerns. (There is no question of attributing to Wittgenstein concepts developed only decades after
his death, and little point in speculating how his views would have evolved
in response to them.) Rather, this programmatic paper seeks to bring into
view fresh avenues for philosophical research: Research in experimental
philosophy can provide new, empirical, foundations for some key aspects
of Wittgenstein’s conception of philosophy (while remaining orthogonal
to some others). Conversely, taking into account Wittgensteinian ideas
about the nature and genesis of some philosophical problems can open
up fresh applications for the tools of experimental philosophy—in a
natural extension of one of its prominent strands.

1. A New Naturalism
1.1 Metaphilosophical Naturalism
Experimental philosophy imports empirical methods and findings from
psychology into philosophy. Different projects in experimental philosophy deploy the new means towards different ends. The ‘Warrant Project’
responds to the practice, common in mainstream analytic philosophy,
to use intuitions about hypothetical cases, considered in thought experiments, as evidence for or against philosophical theories.7 It is interested
in the evidentiary value of specific intuitions: in whether the mere fact
5 See T. Crane and C. French, “The Problem of Perception,” in The Stanford Encyclopedia
of Philosophy, ed. N. Zalta, Summer 2015. http://plato.stanford.edu/entries/perceptionproblem/. See also A.D. Smith, The Problem of Perception (Cambridge, MA: Harvard
University Press, 2002).
6 See J.S. Beck, Cognitive Therapy: Basics and Beyond, 2nd ed. (New York: Guilford,
2011) and A.T. Beck, N.A. Rector, N. Stolar, and P. Grant, Schizophrenia: Cognitive
Theory, Research, and Therapy (New York: Guildford, 2008).
7 See Y. Cath, “Reflective Equilibrium,” in Oxford Handbook of Philosophical Methodology, eds. H. Cappelen, T. Szabo Gendler, and J. Hawthorne (Oxford: Oxford University
Press, 2016), 213–230, J. Nagel, “Intuitions and Experiments: A Defence of the Case
Method in Epistemology,” Philosophy and Phenomenological Research 85 (2012): 495–
527, and J.M. Weinberg, “Intuitions,” in Oxford Handbook of Philosophical Methodology, 287–308.

262

Eugen Fischer

that thinkers have them, as and when they do, speaks for their truth.8 It
seeks to develop a naturalised epistemology of intuitions which determines intuitions’ evidentiary value on an empirical basis.
The Warrant Project’s most ambitious strand, variously characterised
as ‘cognitive epistemology’,9 the ‘Sources Project’,10 or the ‘underwater
part’ of ‘iceberg epistemology’,11 seeks to develop and experimentally test
explanations of intuitive judgments which trace their source to automatic
cognitive processes that go on below the waterline of conscious awareness. This research seeks an understanding of these underlying processes
that allows us to determine under what conditions we may (not) trust the
intuitions they generate.12 One approach seeks to trace intuitions back
to processes that are generally reliable but predictably lead to cognitive
illusions,13 under specific circumstances.14 Such explanations can vindicate intuitions generated under normal circumstances15 and debunk intuitions formed under specific circumstances identified as vitiating.16
Experimental philosophy is often regarded as the epitome of methodological naturalism. But at any rate, the Sources Project promotes a new
form of naturalism. Traditional or first-order methodological naturalism
seeks to address philosophical problems about a topic X (say, the mind or
perception) by building on scientific findings about X. The present projects, by contrast, wish to contribute to the resolution of philosophical

8 See J. Alexander, Experimental Philosophy (Cambridge: Polity 2012), R. Mallon,
“Experimental Philosophy,” in Oxford Handbook of Philosophical Methodology,
410–433, and S. Stich and K. Tobia, “Experimental Philosophy and the Philosophical
Tradition,” in Blackwell Companion to Experimental Philosophy, eds. J. Sytsma and W.
Buckwalter (Malden, MA: Wiley Blackwell, 2016), 5–21.
9 E. Fischer, “Philosophical Intuitions, Heuristics, and Metaphors,” Synthese 191 (2014):
569–606.
10 J. Pust, “Intuition,” in The Stanford Encyclopedia of Philosophy, ed. N. Zalta, Winter
2012. http://plato.stanford.edu/ archives/win2012/entries/intuition/.
11 D.K. Henderson and T. Horgan, The Epistemological Spectrum: At the Interface of
Cognitive Science and Conceptual Analysis (Oxford: Oxford University Press, 2011).
12 J.M. Weinberg, “Humans as Instruments: Or, the Inevitability of Experimental Philosophy,” in Experimental Philosophy, Rationalism, and Naturalism, eds. E. Fischer and J.
Collins (London: Routledge, 2015), 171–187.
13 R. Pohl, ed., Cognitive Illusions (New York: Psychology Press, 2004).
14 E.g., one line of research traces intuitive knowledge-attributions to a ‘mind-reading’
competency, which is argued to be generally reliable (K. Boyd and J. Nagel, “The Reliability of Epistemic Intuitions,” in Current Controversies in Experimental Philosophy,
eds. E. Machery and E. O’Neill (London: Routledge, 2014), 109–127) but subject
to biases, including an egocentrism (J. Alexander, C. Gonnerman, and J. Waterman,
“Salience, and Epistemic Egocentrism,” in Advances in Experimental Epistemology, ed.
J. Beebe (London: Bloomsbury, 2015), 97–118) and focal bias (M. Gerken and J. Beebe,
“Knowledge in and Out of Contrast,” Nous 50 (2016): 133–164).
15 E.g., Nagel, “Intuitions and Experiments”.
16 E.g., E. Fischer and P.E. Engelhardt, “Intuitions’ Linguistic Sources: Stereotypes, Intuitions, and Illusions,” Mind & Language 31 (2016): 65–101.

Experimental Philosophy

263

problems or debates about X by turning to scientific—psychological—
findings about the way people think about X, in general or when
doing philosophy. Fischer and Collins called this ‘meta-philosophical
naturalism’.17 This new ‘higher-order’ naturalism is distinct from, and
implies no commitment to, first-order methodological naturalism.
This independence is illustrated by recent work on the ‘problem of
perception’. This problem arises from paradoxical arguments that build
on intuitions about cases of illusion and hallucination.18 Their proponents tend to consider verbal descriptions of hypothetical cases, rather
than any actual cases; indeed, they stress they merely assume that the
sort of hallucinations they are describing are possible. Recent contributions to cognitive epistemology19 explore the hypothesis that their intuitions about what else is true in the cases described are due to the routine
language comprehension process of stereotypical enrichment, which has
us automatically extract a maximum of information from verbal and
written utterances.20 This body of research develops and experimentally
tests psycholinguistic explanations of these intuitions, with a view to
assessing whether philosophers have warrant to accept their intuitions
in the absence of further argument. To help assess philosophical claims
about sense-perception, this research thus draws on, and generates,
empirical findings about how people talk, think, and reason about senseperception. Where first-order naturalism would have philosophers turn
to the psychology or neuroscience of perception, this work turns to the
psychologies of language and judgment. This contribution to the Warrant Project exemplifies meta-philosophical naturalism but violates the
strictures of first-order methodological naturalism, demonstrating their
independence.
1.2 A Wittgensteinian Perspective
As we shall see, this new naturalism—unlike its first-order cousin—is open
to the pursuit of broadly Wittgensteinian aims, within broadly Wittgensteinian strictures. Much work in the Warrant Project is directed at discrediting

17 E. Fischer and J. Collins, “Rationalism and Naturalism in the Age of Experimental
Philosophy,” in Experimental Philosophy, Rationalism and Naturalism, eds. E. Fischer
and J. Collins (London: Routledge, 2015), 3–33.
18 See T. Crane and C. French, “The Problem of Perception,” in The Stanford Encyclopedia of Philosophy; H. Robinson, Perception (London: Routledge, 2001); A.D. Smith,
The Problem of Perception (Cambridge, MA: Harvard University Press, 2002).
19 E. Fischer and P.E. Engelhardt, “Intuitions’ Linguistic Sources”; “Diagnostic Experimental Philosophy”, Teorema 36 (2017): 117–137; “Stereotypical Inferences: Philosophical
Relevance and Psycholinguistic Toolkit,” Ratio (2017). doi 10.1111/rati.12174.
20 S.C. Levinson, Presumptive Meanings: The Theory of Generalized Conversational
Implicature (Cambridge, MA: MIT Press, 2000).

264

Eugen Fischer

intuition-driven philosophical theorizing.21 This critical impetus is consistent with different metaphilosophical outlooks: It is consistent with the
view that philosophers should construct theories that rely not on intuitions,
but on scientific evidence, to support their claims and solve their problems
(first-order naturalism). But it is also consistent with the more Wittgensteinian view that—some—philosophical problems should not be addressed by
constructing philosophical theories but by ‘diagnostic’ approaches that ‘dissolve’ them, on the basis of (e.g., psycholinguistic) insights into ‘the workings of language’ or other processes that shape philosophical thought.
Let’s develop the commonly neglected second possibility from a currently prominent perspective. Ongoing metaphilosophical debates concern the role of intuitions in philosophy.22 What role intuitions actually
play in philosophical work is an empirical question. We can address it
through case-studies on philosophical texts: by identifying expressions of
intuitive judgments in these texts, and examining the dialectical, heuristic,
argumentative, and justificatory roles intuitions play in them.23 To what
extent philosophical authors attribute evidentiary value to intuitions
and adduce them as evidence for theories is a focus of ongoing debate.24
However, a series of case-studies on early modern and 20th-century analytic texts25 revealed that intuitive judgments also play a role at a prior
and perhaps more fundamental stage of philosophical thought, namely,
in generating distinctive, and distinctively philosophical, problems.
A good example is the aforementioned ‘problem of perception’26: “How
is it possible that we perceive physical objects around us, when we use our
five senses?” As standardly conceived (ibid.), this question is motivated by
at least two related paradoxes, known as ‘arguments from illusion’ and
‘from hallucination’. Both swiftly lead from brief descriptions of cases of
illusion or hallucination, via intuitions about what else is also true of these
cases,27 to the conclusion [q] that when we use our senses, we are aware
21 See Alexander, Experimental Philosophy, 70–88; cf. J.M. Weinberg, “How to Challenge
Intuitions Empirically Without Risking Scepticism,” Midwest Studies in Philosophy 31
(2007): 318–343.
22 See H. Cappelen, Philosophy Without Intuitions (Oxford: Oxford University Press,
2012); Fischer and Collins, “Rationalism and Naturalism,”; J.M. Weinberg, “Intuitions,” in Oxford Handbook of Philosophical Methodology, eds. H. Cappelen, T.
Szabo Gendler, and J. Hawthorne (Oxford: Oxford University Press, 2016), 287–308.
23 See Cappelen, Philosophy Without Intuitions and E. Fischer, Philosophical Delusion
and Its Therapy (New York: Routledge, 2011).
24 See Cappelen, Philosophy, M. Deutsch, The Myth of the Intuitive (Cambridge, MA:
MIT Press, 2015), and J.M. Weinberg, “Humans as Instruments: Or, the Inevitability of
Experimental Philosophy” in Experimental Philosophy, Rationalism, and Naturalism,
eds. E. Fischer and J. Collins (London: Routledge, 2015), 171–187.
25 See Fischer, Philosophical Delusion.
26 See Crane and French, “The Problem of Perception,”; A.D. Smith, The Problem of
Perception (Cambridge, MA: Harvard University Press, 2002).
27 See Robinson, Perception, 54.

Experimental Philosophy

265

only of subjective perceptions or sense-data. The apparent clash with the
common-sense conviction that [p] we see the tables and chairs around
us, then provides a distinctive pre-scientific motivation for the question
of how this is possible, namely, in view of the apparent conflict28: “How
is it possible that p, given that q?” This apparent conflict may engender
a sense of intellectual disquiet, make the very possibility of the otherwise
humdrum fact (that p) seem puzzling and engender the kind of sense of
wonder in the light of familiar facts or phenomena that Plato regarded as
the starting-point of philosophising (Theaetetus 155b-d).29
Classical theoretical responses (indirect realism, phenomenalism, etc.)
seek to solve the problem and answer the question, mainly by showing
that the parties to the apparent conflict are compatible, when properly
understood. Such theories typically seek to honor the underlying intuitions and reconcile them with the background beliefs with which they
appear to clash. Diagnostic responses, by contrast, seek to reconstruct
the underlying reasoning that motivates the question and expose defects
in that reasoning, with a view to showing the question ill-motivated or
meaningless (e.g., Austin).30 Where successful, such responses can be said
to ‘dissolve’ the initial problem by showing its proponents that they have
no right to believe there is any such difficulty [q] as they took to stand
in the way of the familiar fact [p]: that they imagined a difficulty where
they have no warrant to believe there is one [q is unwarranted]. In such a
case, their question (‘How is it possible that p, given that q?’) may well be
intelligible. But it will be ill-motivated and articulate a ‘pseudo-problem’
engendered by an imaginary difficulty.31
28 See Fischer, Philosophical Delusion, 206–11; cf. Paul Horwich, Wittgenstein’s Metaphilosophy (Oxford: Oxford University Press, 2012), 50–60.
29 The paradoxes underlying such problems can have roots, identified by Wittgenstein,
ranging from adherence to unrealistic ideals of precision and rigour in natural language
(PI §§ 98–103) to systematic reliance on “false analogies accepted into language” (BT
409, cp. BT 408, 427, PI §§ 90, 94, 11, 112, 115). In the light of recent metaphor
research in psychology and artificial intelligence, the latter may be conceived as systematic but task-specific overreliance on linguistically realised conceptual metaphors (see
Fischer, “Intuitions, Heuristics, Metaphors” and “Two Analogy-strategies: The Cases
of Mind Metaphors and Introspection.” Connection Science (2017).) For a helpful
general, if idealised, model of how distinctively philosophical problems arise through
paradox, see Horwich, Wittgenstein’s Metaphilosophy, 50–60, cp. 25–9. See fn. 33 and
36 for further Wittgenstein references.
30 J.L. Austin, Sense and Sensibilia (Oxford: Oxford University Press, 1962).
31 More traditional diagnostic approaches (e.g., ‘therapeutic diagnosis’, as defined, for
example, by Michael Williams) seek to show the targeted questions meaningless, rather
than ill-motivated, but convinced few philosophers without prior commitment to them.
See Michael Williams, Unnatural Doubts: Epistemological Realism and the Basis of
Scepticism (Princeton, NJ: Princeton University Press, 1996), xvi. For the development
of the present notion of epistemic (rather than semantic) problem-dissolution, see Eugen
Fischer, “Wittgenstein’s Non-cognitivism—Explained and Vindicated,” Synthese 162
(2008): 53–84 and his Philosophical Delusion, 67–72. Severin Schroeder magisterially

266

Eugen Fischer

Where familiar first-order naturalism has philosophers draw on scientific findings to support a theoretical response to problems that may be
imaginary, the metaphilosophical naturalism associated with experimental philosophy’s Sources Project remains open to either kind of response:
Whether theoretical or diagnostic approaches are appropriate depends, for
any particular problem of this form, on whether the motivating reasoning
actually is sound or defective. Typically, this reasoning is largely intuitive
and proceeds from, or crucially involves, intuitive judgments. Where this is
the case, experimental philosophy’s Sources Project can determine which
response is appropriate. Its ‘restrictionist’ proponents have sought debunking explanations of intuitions.32 If successful, such explanations of intuitions
which underlie philosophical problems provide empirical foundations for
diagnostic approaches and vindicate their broadly Wittgensteinian aim of
‘dissolving’ particular philosophical problems without residue.33
At least where they expose misjudgments or fallacies, competent thinkers would not be expected to commit, diagnostic analyses need empirical
foundations: On their own, logical reconstructions of arguments cannot
establish that glaring misjudgments or fallacies have been made: Hermeneutic principles of charity impose constraints on attributions of ‘irrationality’ and make it difficult to attribute such cognitive misdeeds to
competent thinkers.34 Medium-strength principles of charity demand that
we justify such attributions by providing empirically supported explanations of why competent thinkers make the mistakes attributed to them.35
Such an empirically grounded diagnostic approach keeps in line with
Wittgenstein’s provocative suggestion that “taking care of a philosophical problem is not a matter of pronouncing new truths about the subject of the investigation” (BT 416). Debunking explanations of relevant
intuitions will establish new truths about the underlying cognitive processes and the way people think about the subject of the philosophical
investigation (e.g., sense-perception). But, unlike theoretical responses

32

33

34

35

develops Wittgensteinian ideas about genesis and treatment of ‘pseudo-problems’. See
Severin Schroeder, Wittgenstein: The Way Out of the Fly-bottle (Malden, MA: Polity,
2006), 151–168.
See Weinberg, “Challenge”. Cf. J. Knobe, and S. Nichols, “An Experimental Philosophy
Manifesto,” in Experimental Philosophy, eds. J. Knobe and S. Nichols (Oxford: Oxford
University Press, 2008), 3–14.
Ludwig Wittgenstein, The Big Typescript TS 213, eds. and trans. by C.G. Luckardt
and M.A.E. Aue (Oxford: Blackwell, 2005), 421. For brevity’s sake, we here focus on
intuitions. The Sources Project also targets—e.g., religious or moral—beliefs (Knobe &
Nichols “Manifesto”) whose debunking explanation helps dissolve problems motivated
by reasoning presupposing them.
See J.E. Adler, “Fallacies and Alternative Interpretations,” Australasian Journal of Philosophy 72 (1994): 271–282 and M. Lewinski, “The Paradox of Charity,” Informal
Logic 32 (2012): 403–439.
See P. Thagard and R.E. Nisbett, “Rationality and Charity,” Philosophy of Science 50
(1983): 250–267.

Experimental Philosophy

267

to philosophical problems, they will not even purport to establish new
truths about that subject.
Accordingly, an empirically grounded diagnostic approach is also consistent with Wittgenstein’s striking repudiation of philosophical theorizing:
We may not advance any kind of theory. There must not be anything
hypothetical in our considerations. We must do away with all explanation, and description alone must take its place. And this description gets . . . its purpose from the philosophical problems. These are,
of course, not empirical problems; they are to be solved by looking
into the workings of our language . . .36
Of course, cognitive epistemology (the Sources Project) seeks to explain
why thinkers have certain intuitions (e.g., about certain linguistically
described cases), and these explanations are informed by psychological
theories (e.g., about language comprehension) and rely on hypotheses
which are experimentally testable. But these explanations, theories, and
hypotheses are psychological, not philosophical. They are used to empirically answer metaphilosophical questions about how thinkers speak,
think, and reason about X, when doing philosophy, not philosophical
questions about X. In the context of a diagnostic approach, these questions are not treated as empirical, but as expressions of confusion, resulting from defective reasoning—driven also by routine language processes
(like the above-mentioned process of stereotypical enrichment).
At any rate, one strand of experimental philosophy can thus contribute to
giving fresh content to Wittgensteinian ideas that prima facie seem in direct
conflict with methodological naturalism. Next, we shall consider how it
cannot merely lend empirical substance to diagnostic approaches but provide a fresh empirical rationale for therapeutic conceptions of philosophy,
which may appear to be even more opposed to a naturalist outlook.

2. Empirical Foundations for Therapeutic Philosophy
2.1 Philosophy as Therapy
More or less substantive therapeutic conceptions of philosophy can be
obtained by identifying similarities between kinds of therapy and ways
of doing philosophy, at the level of individual features (surface-similarities) or of relations between several elements in each domain (analogies). Surface similarities motivate the metaphorical extension of single
words (e.g., ‘elephant’), namely, to attribute one or more stereotypically

36 Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe, 3rd ed.
(Oxford: Blackwell, 2001), §109.

268

Eugen Fischer

associated features (clumsy, huge, has excellent memory), to elements of
a new domain (‘John is an elephant’).37 This motivates superficial talk
of ‘retail therapy’ (like literal therapy, it—supposedly—makes someone
feel better). By contrast, comprehensive analogies license the extension of
several related words (“you don’t have to be bright to see her point; it’s
clear”). These extensions are motivated by conceptual metaphors,38 i.e.,
comprehensive mappings from one domain (e.g., vision) to another (cognition), which preserve inferential relations between the terms extended.39 Where comprehensive analogies between domains40 license such
extension, a new, wider category (e.g., of ‘vision’, or ‘therapy’, or ‘health’)
may be formed.41
A substantive therapeutic conception of philosophy can be obtained
by identifying comprehensive analogies between particular areas or practices of philosophy, and either specific forms or the generic domain of
psychotherapy.42 On the ‘generic approach’, the metaphorical extension
of ‘therapy’ is part-and-parcel of the wholesale extension of a family of
related generic terms (including ‘health’, ‘illness’, ‘disease’, ‘symptom’,
‘diagnosis’, ‘therapy’), first (and not without problems) from somatic to
psychological phenomena,43 then, second (and again not without problems), from these psychological phenomena to philosophy.44 Accordingly,
37 See H. Bortfeld and M.S. McGlone, “The Continuum of Metaphor Processing,” Metaphor and Symbol 16 (2001): 75–86 and John Searle, “Metaphor,” in Metaphor and
Thought, ed. A. Ortony, 2nd ed. (Cambridge: Cambridge University Press, 1993),
83–111.
38 See G. Lakoff and M. Johnson, Metaphors We Live By (Chicago: Chicago University
Press, 1980) and G. Lakoff and M. Johnson, Philosophy in the Flesh (New York: Basic
Books, 1999).
39 See Fischer, “Intuitions, Heuristics, Metaphors”.
40 See D. Gentner, “Structure Mapping: A Theoretical Framework for Analogy,” Cognitive
Science 7 (1983): 155–170.
41 See B.F. Bowdle and D. Gentner, “The Career of Metaphor,” Psychological Review 112
(2005): 193–216.
42 The ‘specific approach’ led some followers and scholars of Wittgenstein to construct
more (e.g., M. Lazerowitz, “The Passing of an Illusion,” in Necessity and Language, eds.
M. Lazerowitz and A. Ambrose (Beckenham: Croom Helm, 1985), 200–240) or identify less (e.g., G.P. Baker and P.M.S. Hacker, Wittgenstein: Meaning and Understanding:
Essays on the Philosophical Investigations (Oxford: Blackwell, 1983), 288; E. Harcourt,
“Wittgenstein and Psychoanalysis,” in The Blackwell Companion to Wittgenstein, eds.
H.-J. Glock and J. Hyman (Oxford: Blackwell, 2017), 651–665) comprehensive analogies between Wittgenstein’s approach and the specific form of psychotherapy dominant
in his day, viz., Freudian psychoanalysis (cf. G.P. Baker, “Wittgenstein’s Method and
Psychoanalysis,” in his, Wittgenstein’s Method: Neglected Aspects (Oxford: Blackwell, 2004), 205–222, B.A. Farrell, “An Appraisal of Therapeutic Positivism,” Mind 55
(1946): 25–48 and 133–150, and J. Wisdom, “Philosophy and Psychoanalysis,” in his
Philosophy and Psychoanalysis (Oxford: Blackwell, 1964), 169–181.
43 See M. Boyle, Schizophrenia: A Scientific Delusion, 2nd ed. (London: Routledge, 2000).
44 See E. Fischer, “Diseases of the Understanding and the Need for Philosophical Therapy,” Philosophical Investigations 34 (2011): 22–54 and P. Tarras, “ ‘Philosophie’

Experimental Philosophy

269

recent proponents of therapeutic conceptions of philosophy have begun
to spell out the underlying analogies between the domain of psychotherapy and intended applications in philosophy45 and to examine how their
careless handling may mislead.46
In a nutshell:47 A common notion of mental health is shaped by Plato’s
ideal of a balanced, rational agent who can master his feelings and impulses
sufficiently well to be truly autonomous (Republic 443c—444e). People
count as mentally ill when they fall significantly short of this ideal and
have emotional problems (unwarranted and distressing or disabling emotions they cannot control) or behavioral problems (unwarranted behavior
they cannot control), due to a disease, i.e., a literally or metaphorically
‘inner’ process that is not under people’s direct control, which brings
about symptoms (emotions, thoughts, mental states) that are constitutive
of or engender those problems. A process may qualify as pathological, i.e.,
as a ‘disease’, in virtue of these consequences, rather than any abnormality
of or in its course.48 A therapy is a procedure that seeks to cure the illness, i.e., to put an end to the emotional or behavioural problems (symptom relief) and prevent their recurrence (relapse prevention). It should be
based on a diagnosis that identifies the disease or process causing those
problems, in a given case. To further extend these notions to philosophy,
we need to identify an analogous structure in that domain. This has been
attempted for philosophical efforts addressing emotional and behavioral
problems which arise in ordinary life (‘philosophical therapy’, e.g., Stoics)
and in philosophical reflection (‘therapeutic philosophy’), respectively.
Wittgenstein can be regarded as engaged in efforts of the second kind.
He often refers to emotional responses to philosophical questions or
ruminations and explicitly seeks to liberate himself and others from these
intellectual ‘disquietudes’.49 Since he takes them to be engendered by
wrong analogies,50 ‘disorder in our concepts’,51 misleading expressions,52
and ‘misinterpretation’,53 he clearly regards them as unwarranted, if

45

46
47
48
49
50
51
52
53

Grammatisch Betrachtet: Wittgensteins Begriff der Therapie.” Kriterion—Journal of
Philosophy 28 (2014): 75–97.
See K. Banicki, “Philosophy as Therapy: Towards a Conceptual Model,” Philosophical
Papers 43 (2014): 7–31. See also Fischer, Philosophical Delusion and Fischer, “Diseases
of Understanding”.
Cf. B. De Mesel, “On Wittgenstein’s Comparison of Philosophical Methods to Therapies,” International Journal of Philosophical Studies 23 (2015): 566–583.
For more detail, see Fischer, “Diseases of Understanding,” 43–49.
D. Clouser, C. Culver, and B. Gert, “Malady,” in What Is Disease? eds. J. Humber and
R. Almeder (Totowa, NJ: Humana Press, 1997), 173–219.
BT 409, 415, 416, 421, 431, PI §111. See also Ludwig Wittgenstein, Culture and Value,
rev. ed. (Oxford: Blackwell, 1998), 50.
BT 409.
BT 421.
BT 416.
PI §111.

270

Eugen Fischer

natural.54 To such unwarranted emotional responses to philosophical
questions, we can add unwarranted behavioral responses: Wherever
such a question is ill-motivated, efforts to answer them will be equally
unwarranted—at any rate, in the absence of an independent rationale.
However, to amount to emotional or behavioural problems in need
of therapy, such unwarranted disquietudes and efforts need to be due
to some sort of disease, i.e., to (literally or figuratively) inner processes
that are not under the thinkers’ direct control. Many metaphilosophical remarks of Wittgenstein’s are consistent with such an aetiological
view: Wittgenstein attributes the ‘disquietudes’ he targets to an ‘urge to
misunderstand’,55 i.e., a propensity that is not under subjects’ control. He
identifies a propensity to spontaneously make unwarranted inferences, in
line with certain invalid ‘thought-schemas’,56 from common-sense convictions or observations to puzzling conclusions.57 Through a series of
simple ‘experiments’,58 he elicits such inferences in his own thought and
thus exposes certain urges he may share with others.59 When such uncontrollable urges to leap to conclusions engender distressing disquietudes
(or simply have thinkers embark on unwarranted endeavours) the latter
can qualify as emotional (or behavioral) problems, and the urges as diseases which he diagnoses, for a start, in himself,60 by eliciting the spontaneous conclusions that are their symptoms. Accordingly, Wittgenstein
maintains “the philosopher is someone who has to cure in himself many
diseases of the understanding, before he can arrive at the notions of common sense”,61 namely, before he can return to the common-sense notions
from which unwarranted inferences he could not help making drove him
to bogus puzzles. Where these problems are entirely bogus, the only genuine problems facing the philosopher are the emotional and behavioral
problems they engender. In this case, one may reasonably regard the resolution of such problems as one’s ‘entire [!] task’, as Wittgenstein strikingly
does.62
54
55
56
57
58
59

CV 22.
PI §109.
PI §597.
See Fischer, Philosophical Delusion, 257–261.
PI §167.
E.g. PI §§ 166, 169, 173, 174, 176; cf. E. Fischer, “A Cognitive Self-therapy—PI 138–
97,” in Wittgenstein at Work: Method in the “Philosophical Investigations”, eds. E.
Ammereller and E. Fischer (London: Routledge, 2004), 86–126.
60 For exegetical argument that Wittgenstein often is his own main ‘patient’, see Fischer,
“Cognitive Self-therapy” and Philosophical Delusion, 261–264.
61 CV 50.
62 BT 421, CV 50. In these passages, Wittgenstein asserts dispelling certain disquietudes
is the sole purpose of his philosophizing. This is reasonable for him: Since he is generally careful not to accept the puzzling conclusions to which his urges have him leap, he
does not engage in any theoretical efforts to reconcile them with common sense, and
his urges fail to engender anything describable as behavioral problems.

Experimental Philosophy

271

This conception of philosophy posits the sort of analogy to psychotherapy that would motivate the extension of the concept-cluster
reviewed from the domain of psychotherapy to that of philosophy: If
cognitive processes we cannot control drive us, in philosophical thought,
to agonize over ill-motivated questions and make unwarranted efforts to
answer them, i.e., if such processes make our lives miserable with gratuitous worries and have us misspend them on pointless endeavors we
would desist from if we were more fully masters of ourselves, then we
philosophers can be said to have ‘emotional and behavioral problems’
that are caused by ‘diseases’ and call for ‘therapy’.
2.2 Empirical Foundations
This argument for the need for therapy in philosophy relies on an empirical assumption:
(EA) There are ill-motivated philosophical questions which are formulated due to cognitive (or other psychological) processes thinkers cannot directly control.
These processes may be normal in their nature and course and qualify
as ‘diseases’ in virtue of causing emotional and behavioral problems,63
under specific circumstances. But without (literally or metaphorically)
‘inner’ sources beyond thinkers’ direct control, unwarranted worries
about ill-motivated questions, and pointless efforts to solve them, will
not qualify as emotional and behavioral problems, respectively, and there
will be no need for anything analogous to psychotherapy.
So, is that crucial assumption true? This metaphilosophical question
brings in experimental philosophy: The strand we discussed above is all
about empirically developing and evaluating assumptions like this. The
Sources Project considers assumptions about the cognitive sources of
philosophically relevant intuitions and beliefs,64 including intuitions and
beliefs that motivate philosophical questions and problems; it turns such
assumptions into experimentally testable hypotheses, and experimentally
tests them; and it focuses precisely on assumptions about the role of automatic cognitive processes which we cannot directly control.65
One of the major findings of cognitive and social psychology66 over
the last decades has been the extent to which our judgments, decisions,

63 Cf. Clouser, et al., “Malady”, 190.
64 See Knobe and Nichols, “Manifesto”, 7–8.
65 See Fischer, “Intuitions, Heuristics, Metaphors” and J. Nagel, “The Psychological Basis
of the Harman-Vogel Paradox,” Philosophers’ Imprint 11(5) (2011): 1–28.
66 See J.St. B.T. Evans, Thinking Twice: Two Minds in One Brain (Oxford: Oxford University Press, 2010), D. Kahneman, Thinking Fast and Slow (London: Allen Lane, 2011),

272

Eugen Fischer

and actions are shaped by cognitive processes which are automatic: effortless, unconscious, initiated regardless of the subject’s goals, and not alterable in their course, once initiated.67 All these properties are operationally
defined (e.g., ‘effortless’ means performance requires little attention, i.e., is
little affected by multitasking); they are thus measureable and gradable.68
Intuitions, as conceived by cognitive psychologists—and experimental
philosophy’s Sources Project—are judgments that are generated by automatic inferences, i.e., by highly automatic processes that duplicate rulegoverned inferences, in particular inferences governed by heuristic rules.69
The speed and amount of effort required are used as a metacognitive cue to
assess plausibility.70 Intuitions result from highly automatic processes and
immediately strike the thinker as plausible, regardless of whether she accepts
those judgements upon further reflection. Some intuitions are cognitive illusions: predictable misjudgments which are automatic in origin, modifiable
by conscious reflection, but typically strike thinkers as intuitively compelling even once they have realized they cannot be right (if they realize).71
Above (Section 1.2), we noted that some philosophical questions are
motivated by perceived conflicts between intuitions, or consequences
derived from them, and background beliefs or common-sense convictions. Where such intuitions, first, qualify in the aetiological psychological sense explained and, second, are cognitive illusions unsupported by
further argument, we are dealing with an ill-motivated pseudo-problem
to whose formulation cognitive processes beyond our direct control make
an essential contribution. We highlighted an approach from the Sources
Project that seeks to construct psychological explanations of intuitions
that trace them back to automatic cognitive processes that are generally
reliable but predictably lead to cognitive illusions under specific circumstances (Section 1.1; cf. fn.14). By explaining the intuitions at the root of
such a philosophical problem, and helping to expose them as cognitive
illusions, such ‘GRECI explanations’72 can support the empirical assumption (EA) on which the present therapeutic conception of philosophy
relies, and show that, sometimes, philosophy needs therapy.

67
68
69

70
71
72

and T.D. Wilson, Strangers to Ourselves: Discovering the Adaptive Unconscious (Cambridge, MA: Harvard University Press, 2002).
J.A. Bargh, “The Four Horsemen of Automaticity,” in Handbook of Social Cognition,
eds. R. Wyer and T. Srull, (Hillsdale: Earlbaum, 1994), 1–40.
A. Moors and J. De Houwer, “Automaticity: A Theoretical and Conceptual Analysis,”
Psychological Bulletin 132 (2006): 297–326.
D. Kahneman and S. Frederick, “A Model of Heuristic Judgment,” in The Cambridge
Handbook of Thinking and Reasoning, eds. K.J. Holyoak and R. Morrison (Cambridge: Cambridge University Press, 2005), 267–293.
V.A. Thompson, J.A. Prowse Turner, and G. Pennycook, “Intuition, Reason, and Metacognition,” Cognitive Psychology 63 (2011): 107–140.
Pohl, Cognitive Illusions, 2–3.
As we have called them: Fischer and Engelhardt, “Intuitions’ Linguistic Sources,” 67.

Experimental Philosophy

273

If successful, this approach can simultaneously show that nothing need
be constitutionally wrong with the ‘patients’ targeted by therapeutic philosophy: Suppose that a successful GRECI explanation identifies vitiating
circumstances which obtain in the formulation of the given philosophical
problem and interfere with the generally reliable automatic process. Proponents of the problem are then in a similar position as mountaineers with
altitude sickness, whose symptoms (headache, weakness, etc.) can befall even
the physically fittest, due to the interaction of normal physiological processes with environmental conditions (oxygen shortage at higher altitudes
with lower atmospheric pressure): Their ‘symptoms’ (disquieting puzzlement
in the light of paradoxical intuitions) are ultimately due to the interaction
between a normal (and generally reliable) cognitive process and specific vitiating circumstances, which obtain in the formulation of the specific problem
and may recur systematically in certain kinds of philosophical thought.

3. A Case in Point
To lend substance to these general ideas and provide evidence for the
empirical assumption EA, this therapeutic conception of philosophy relies
on, let’s return to our previous example: The ‘problem of perception’73 is
generated by two paradoxes, the ‘arguments from illusion’ and ‘from hallucination’ (Section 1.2). These rely on particular intuitions about cases
of ‘illusion’ and hallucination74: For example, when a round coin, viewed
sideways, looks elliptical, the viewer is not aware of the round coin, but
of something elliptical; or: when Shakespeare’s Macbeth sees a dagger,
there is something the hallucinating man sees.75 Typically, proponents
of those arguments do not provide any independent arguments to support these intuitive judgments.76 Their warrant for accepting these intuitions in the face of tensions with background beliefs therefore depends
upon the intuitions’ evidentiary value: on whether the mere fact that the
arguments’ proponents have them, as and when they do, speaks for their
truth.77
Suppose a GRECI explanation traces these intuitions to a particular automatic cognitive process and shows that, in the formulation of the paradoxes,

73 See Crane and French, “The Problem of Perception” and Smith, The Problem of Perception.
74 See Robinson, Perception, 54.
75 These arguments first conclude that in these particular cases, subjects are aware of
sense-data, rather than physical objects; their second half then transfers this conclusion
to all cases of visual perception (Crane and French 2015; Smith 2002). We focus here
on the key intuitions driving the first half.
76 See Fischer and Engelhardt, “Intuitions’ Linguistic Sources”, 78, and Fischer, Philosophical Delusion, 167–200.
77 See E. Fischer, P.E. Engelhardt, and A. Herbelot, “Intuitions and Illusions: From Explanation and Experiment to Assessment”, 260, in Experimental Philosophy, Rationalism
and Naturalism, eds. E. Fischer and J. Collins (London: Routledge, 2015), 259–292.

274

Eugen Fischer

this process operates under vitiating conditions and engenders cognitive
illusions. Such an account will provide an undermining defeater78 and
show that the crucial intuitions lack evidentiary value—so that proponents of the argument typically lack warrant for accepting them and
the further conclusions that rely on them that generate the problem of
perception. This will show the problem ill-motivated. Simultaneously, it
will reveal that the problem was formulated due to automatic processes
thinkers cannot directly control. That is, such an account will support
the existence claim (EA) by showing that the problem of perception is a
case in point.
Such an explanation has been developed and experimentally tested
by Eugen Fischer and Paul Engelhardt.79 It traces the targeted intuitions
to the routine utterance comprehension process of stereotypical enrichment.80 This process automatically fills in detail: In the absence of explicit
indications to the contrary, competent language users spontaneously
infer that the situation talked about conforms to the stereotypes associated with the words used (e.g., that the ‘secretary’ is female). Verbs can
be associated with typical features of actions, agents, and objects (e.g.,
the object ‘seen’ is located in front of the viewer, the person to whom
something ‘seems F’ is inclined to think it is F, etc.).81 These features can
make up complex, internally structured situation-stereotypes, known as
‘generalized situation schemas’.82 Embedded in a communicative practice that requires speakers to make stereotype-deviations explicit,83 this
process is generally reliable but leads to cognitive illusions under specific
circumstances.84
The proposed account of arguments from illusion and hallucination
builds on the observation that, in formulating these arguments, philosophers use appearance- and perception-verbs in a rarefied phenomenal
sense which serves to describe subjects’ experience only85 and is devoid
of stereotypical (doxastic, spatial or existential) implications these verbs
78 See J. Pollock, “Reliability and Justified Belief,” Canadian Journal of Philosophy 14
(1984): 103–114.
79 See and E. Fischer and P.E. Engelhardt, “Diagnostic Experimental Philosophy,” Teorema 36(3) (2017): 117–137. See also Fischer and Engelhardt, “Intuitions’ Linguistic
Sources” and “Stereotypical Inferences”, and Fischer et al., “Intuitions and Illusions.”
80 See Levinson, Presumptive Meanings, 112–164.
81 See K. McRae, T.R. Ferretti, and I. Amyote, “Thematic Roles as Verb-specific Concepts,” Language and Cognitive Processes 12 (1997): 137–176.
82 D.E. Rumelhart, “Schemata: The Building Blocks of Cognition,” in Theoretical Issues
in Reading Comprehension, eds. R. Spiro, B. Bruce, and W. Brewer (Hillsdale, NJ: Erlbaum, 1980), 33–58.
83 See Levinson, Presumptive Meanings, 27–42 and 113–114.
84 See Fischer and Engelhardt, “Intuitions’ Linguistic Sources”.
85 See A.J. Ayer, The Problem of Knowledge (London: Penguin, 1956/1990), 90, and F.
Macpherson, “The Philosophy and Psychology of Hallucination,” in Hallucination:
Philosophy and Psychology, eds. F. Macpherson and D. Platchias (Cambridge, MA:
MIT Press, 2013), 1–38.

Experimental Philosophy

275

carry in their dominant use in ordinary discourse. Arguably, this kind
of situation recurs in philosophy, as philosophers repeatedly take words
that have well-established uses in ordinary discourse and adapt them for
special philosophical purposes.
While perfectly legitimate, such ‘tampering with words’ can have
‘unforeseen repercussions’:86 Dominant uses of words shape our habits
of inference. When we encounter a polysemous word far more frequently
in a dominant sense or use than in any of its other uses, we cannot help
making the kind of inferences we habitually, and typically rightly, make
when we encounter the word in that dominant sense.87 This leads to predictable fallacies: Even competent language users may leap to conclusions
licensed—only—by the dominant (typically literal) sense of a word, when
encountering the word in a more rarefied (technical or metaphorical) use.
This happens in particular
(i) when a rarefied use is not explicitly marked by riders like ‘in a special
sense’88 and
(ii) when the situation-stereotype associated with the dominant sense
can contribute to interpreting utterances using the rarefied sense.89
Conditions (i) and (ii) typically apply, respectively, to phenomenal uses of
appearance-verbs in arguments from illusion90 and to phenomenal uses
of perception-verbs in arguments from hallucination.91
On this account, the intuitions at the root of the latter arguments are
generated by contextually inappropriate stereotypical inferences, namely,
spatial and existential inferences, e.g., from “Macbeth sees a dagger” to
“There is something before Macbeth’s eyes”—if no physical object (as
per the assumption that he is hallucinating), then a sense-datum, which
must be before his mind’s eye. The crucial intuition about hallucination
is thus due to the interplay of two factors: Contextually inappropriate

86 See Austin, Sense and Sensibilia, 63.
87 See R. Giora, On Our Mind: Salience, Context, and Figurative Language (Oxford:
Oxford University Press, 2003), G.B. Simpson and C. Burgess, “Activation and Selection
Processes in the Recognition of Ambiguous Words,” Journal of Experimental Psychology: Human Perception and Performance 11 (1985): 28–39, and R.E. Till, E.F. Mross,
and W. Kintsch, “Time Course of Priming for Associate and Inference Words in a Discourse Context,” Journal of Verbal Learning and Verbal Behaviour 16 (1988): 283–298.
88 See S. Givoni, R. Giora, and D. Bergerbest, “How Speakers Alert Addressees to Multiple
Meanings,” Journal of Pragmatics 48 (2013): 29–40.
89 See Giora, On Our Mind and R. Giora, M. Raphaely, O. Fein, and E. Livnat, “Resonating With Contextually Inappropriate Interpretations: The Case of Irony,” Cognitive
Linguistics 25 (2014): 443–455.
90 See Fischer and Engelhardt, “Intuitions’ Linguistic Sources”.
91 See E. Fischer, P. Engelhardt, and A. Herbelot, “Salience Effects Drive Inappropriate
Inferences in Philosophical Arguments: An Interdisciplinary Investigation of Philosophical Paradox,” (under review).

276

Eugen Fischer

inferences are automatically made from a rarefied metaphorical use of
‘see’ (the phenomenal use); and the conclusion is swiftly integrated with
implicit background theories, namely, either introspective conceptions of
the mind92 or conceptions of ‘phenomenal space’,93 rather than being discarded in the light of a conflict with the contextual information that no
suitable physical object is around.
This account gives empirical content to Wittgenstein’s idea that some
philosophical problems are generated by “an urge to misunderstand the
workings of our language”,94 which has us think “as if our thinking was
based on a thought schema” and we were “ ‘unconsciously’ translating
from a more primitive mode of thought into ours”:95 The present inferences are based on a generalised situation schema which we know not
to (fully) apply to the given case. In making contextually inappropriate
stereotypical inferences that project it onto stereotype-deviant situations,
we are leaping to conclusions as if we had the urge to treat polysemous
words as having only one (their dominant) sense and were ‘unconsciously’
translating from a primitive ‘literalist’ mode of thought into ours. The
new account explains what Wittgenstein describes.
Thinkers lack conscious insight into automatic inferences.96 Nor can
logical reconstructions of arguments establish on their own that such inferences are made in clearly inappropriate contexts, as in the present example:
Medium-strength principles of charity97 demand that we justify attributions of clearly inappropriate inferences by providing empirically supported
explanations of why competent thinkers make them (cf. Section 1.2).
Accordingly, Fischer and Engelhardt98 justified crediting competent speakers with inappropriate spatial and existential inferences from phenomenal
uses of ‘see’, by invoking psycholinguistic theories that would account for
them (see above)99 and experimentally testing for the posited inferences.

92 See Fischer, “Intuitions, Heuristics, Metaphors” and E. Fischer, “Mind the Metaphor! A
Systematic Fallacy in Analogical Reasoning,” Analysis 75 (2015): 67–77.
93 See M. Hymers, Wittgenstein on Sensation and Perception (New York: Routledge,
2017), 6–13.
94 See PI §109.
95 See PI §597.
96 See Bargh, “Four Horsemen” and Moors and De Houwer, “Automaticity”.
97 See Thagard and Nisbett, “Rationality and Charity”.
98 See Fischer and Engelhardt, “Stereotypical Inferences”, “Diagnostic Experimental Philosophy”, and Fischer et al., “Salience Effects”.
99 The account outlined relies on the graded salience hypothesis (See Fein et al., “On the
Priority of Salience-Based Interpretations: The Case of Sarcastic Irony” Intercultural
Pragmatics 12 (2015): 1–32, Giora, On Our Mind, and Giora et al., “Resonating”), the
cued-schemas account of language comprehension/production and Levinson’s I-heuristic. See J.L. Elman, “On the Meaning of Words and Dinosaur Bones: Lexical Knowledge
Without a Lexicon,” Cognition 33 (2009): 547–582, Levinson, Presumptive Meanings,
and Fischer et al., “Salience Effects”.

Experimental Philosophy

277

If subjects really make these inferences (e.g., from ‘S sees X’ to X is in
front of S), they will be surprised when the verb is followed by a sequel
inconsistent with the inferred conclusion, as in (1), but not (2):
(1) Jeb sees the spot on the wall behind him.
(2) Matt sees the spot on the wall facing him.
When we are surprised, our eyes’ pupils expand; pupil dilation is an
index of surprisal.100 Accordingly, one can examine whether automatic
inferences are made by measuring pupil size during and after participants hear sentences like (1) and (2): When significant dilations occur
after sentences like (1) but not (2), this is evidence that the hypothesized
(spatial-directional) inferences are automatically made from ‘see’, in language comprehension.
To examine the hypothesis that participants make spatial inferences
(from ‘S sees X’ to X is in front of S) not only where ‘see’ is used in its
dominant perceptual sense (as in 1 and 2) but also—inappropriately—
where it is used in much less frequent (e.g., epistemic or phenomenal)
senses,101 experimentalists can use sentences like:
(3) Jack sees the problems he left behind.
(4) Joe sees the problems that lie ahead.
Here, a purely epistemic reading of ‘see’ (Macmillan Advanced Learners’
Dictionary [MEDAL], sense 4) and a purely metaphorical reading of the
spatial sequel (‘left behind’ = in the past, ‘lie ahead’ = in the future) are
readily available. On this interpretation, both (4) and (3) are perfectly
consistent, and (3) is utterly unsurprising—of course people can know
what problems they had in the past. But if the hypothesized inappropriate inferences occur, pupils will dilate when participants hear the likes of
(3)—but not of (4).
To follow up the hypothesis that relevant inferences are supported by
features (stereotypical associations) of the verb, we add otherwise identical sentences that replace ‘see’ by an otherwise similar verb without spatial associations. E.g., ‘is aware of’ is ordinarily used in an epistemic sense,
to attribute knowledge that may, but need not, be acquired through the

100 See D. Kahneman, Attention and Effort (Englewood Cliffs, NJ: Prentice-Hall, 1973)
and B. Laeng, S. Sirois, and G. Gredeback, “Pupillometry: A Window to the Preconscious?” Perspectives on Psychological Science 7 (2012): 18–27.
101 A corpus analysis supports the hypothesis about relative frequencies. See Fischer et al.,
“Salience Effects”.

278

Eugen Fischer

five senses (MEDAL, WordNet), and should have the same existential/
factive implications as ‘see’102 but lack spatial associations:103
(3*) Jack is aware of the problems he left behind.
(4*) Joe is aware of the problems that lie ahead.
The present hypothesis predicts pupils will dilate after the likes of (3) but
not of (3*).
Conclusions of automatic inferences can be suppressed,104 before they
can influence people’s judgments or further reasoning. To determine
whether the posited inferences were automatically made and went on to
influence judgments, Fischer and Engelhardt105 combined pupil measurements with a plausibility rating task: If conclusions are not suppressed,
their clash with the sequel will make participants judge sentences like
(1) and (3) significantly less plausible than sentences like (2) and (4), respectively, even more than a second after sentence offset—and even though it
is at least as plausible that people should know what problems they had
in the past (as per 3) than what awaits them in the hard-to-predict future
(as per 4). Similarly, ‘see’-sentences like (3) will strike participants as less
plausible than analogous ‘aware’-sentences (like 3*)—even though they
mean the same, on the readily available epistemic reading of ‘see’.
In both studies, pupil measurements suggested that participants indeed
made spatial inferences both—appropriately—from perceptual uses of
‘see’ (as in 1 and 2), and—inappropriately—from less common nonperceptual uses (as in 3 and 4), though not from analogous uses of ‘aware’
(as in 3* and 4*). Plausibility ratings suggested that conclusions of both
appropriate and inappropriate inferences influenced subsequent judgments: ‘See’-sentences like (3) were judged less, not more plausible than
sentences like (4), and less plausible than semantically apparently identical
‘aware’-sentences, like (3*), which were deemed exactly as plausible as the
likes of (4*): The manipulation of the spatial sequel affected the plausibility of ‘see’-, but not ‘aware’-sentences.
In these ways, experiments can examine hypotheses about what automatic inferences are made, appropriately or otherwise, from linguistic case descriptions and premises of arguments. The studies outlined
102 Fischer et al., “Salience Effects”, submit that versions of the argument from hallucination which use ‘aware’ involve inappropriate existential inferences from this verb’s
phenomenal use.
103 This hypothesis is supported by production experiments (See Fischer and Engelhardt,
“Diagnostic Experimental Philosophy” and “Salience Effects”).
104 See M. Faust and M.A. Gernsbacher, “Cerebral Mechanisms for Suppression of Inappropriate Information During Sentence Comprehension,” Brain and Language 53
(1996): 234–259.
105 See Fischer and Engelhardt, “Stereotypical Inferences” and Fischer et al., “Salience
Effects”.

Experimental Philosophy

279

support a debunking explanation of intuitions at the root of arguments
from hallucination. This explanation traces them back to contextually
inappropriate stereotypical inferences from rare uses of perception-verbs,
and thus exposes them as cognitive illusions.
Together with analogous work on the intuitions grounding arguments
from illusion,106 these studies contribute to showing that the problem of
perception generated by these two related paradoxes is, first, ill-motivated
and, second, formulated due to an automatic cognitive process thinkers
cannot directly control, namely, stereotypical enrichment. This provides
initial support for the empirical existence assumption (EA) on which
the outlined therapeutic conception of philosophy relies: There are illmotivated philosophical questions which are formulated due to cognitive
(or other psychological) processes thinkers cannot directly control. This
example also lets us understand how competent thinkers can come to
raise such questions: The relevant cognitive process is generally reliable;
but in the formulation of the paradoxical arguments that motivate the
question, it operates under vitiating conditions—(i) and (ii) above. To
what extent further philosophical problems fit this bill can be explored
by combining analytic case-studies on philosophical texts (to identify
relevant intuitions and their role in problem-genesis) with experimental
work (testing explanations of those intuitions), in future research.

4. Cognitive Philosophical Therapy
In this chapter, we have seen how a key strand of experimental philosophy
(the ‘Sources Project’) (Section 1.1), can facilitate diagnostic approaches to
characteristically philosophical problems (Section 1.2) and provide a therapeutic conception of philosophy in Wittgenstein’s wake (Section 2.1) with
empirical foundations (Section 2.2). A case-study provided a first ‘pudding
proof’ (Section 3). In closing, we now consider how the empirical findings
reviewed can contribute not merely to establishing the need for therapeutic
philosophy (Section 2) but also provide means for putting it into practice.
The aim of the diagnostic approach is to ‘dissolve’ particular philosophical problems by showing them ill-motivated.107 The further aims of therapy
are (a) to enable proponents of the ‘problem’ to rationally and effectively
give up pursuing and worrying about it, and (b) to prevent them from falling for similar pseudo-problems in the future (Section 2.1).108 A diagnostic
analysis that succeeds in identifying fallacies in the motivating reasoning
may fail to secure repudiation of the resulting problem, where automatic
cognitive processes drive effortless inferences to intuitions that continue to
106 See Fischer et al., “Intuitive Inferences” and Fischer and Engelhardt, “Intuitions’ Linguistic Sources”.
107 Cf. BT 421.
108 Cf. BT 421 and CV 50.

280

Eugen Fischer

strike thinkers as plausible109—and their clash with background beliefs as
worrying. In addition, such diagnostic analysis of one problem will not per
se prevent a thinker from falling for other ill-motivated problems driven
by the same cognitive processes. To achieve the goals of therapy, more
than reconstruction of motivating reasoning and identification of fallacies
appears to be required. But what more?
Cognitive behavioral therapy110 works with a variety of therapeutic
methods which can serve as useful models.111 This frequently used ‘talking cure’ seeks to lastingly increase patients’ rational autonomy to an
extent sufficient to end their emotional and behavioral problems and to
prevent relapse. The key idea is that rational autonomy can be enhanced
through metacognitive insight: by coming to understand how automatic
processes of which we are not aware shape our thoughts, feelings, and
behaviors, including the behaviors and emotions constitutive of symptoms. This new self-understanding allows patients to grow aware of,
and critically examine, dysfunctional judgments and beliefs they had
previously just presupposed; it facilitates a new, more detached attitude
towards them, which mitigates what distress and behavioral responses
they otherwise occasion.
Accordingly, cognitive therapy systematically builds on findings from
cognitive and social psychology, and cognitive neuroscience, to develop
empirically supported models of how symptoms arise in different conditions (depression, anxiety, psychosis, etc.).112 A key part of an individual therapy then consists in developing an understanding of how the
model applies to the individual patient, and helping her to acquire this
understanding.113 This often involves identifying and mitigating psychological motivations like the need to defend against low self-esteem, which
may help maintain problematic beliefs and inference-styles, and need
to be mitigated, e.g., through behavioral confidence-building exercises,
before the patient is willing and able to acquire a new self-understanding.
However, most behavioral accretions that turn cognitive into cognitive

109 Cf. PI§§ 109, 597. This situation is liable to recur, since automatic processes are not
under our direct control, but highly fluent (i.e. effortless), and fluency acts as a metacognitive cue for assessing plausibility (See Thompson et al., “Metacognition”.). Also,
subsequent reflection may seek to justify rather than correct plausible (but wrong)
intuitions (See Evans, Thinking Twice).
110 See J.S. Beck, Cognitive Therapy: Basics and Beyond. 2nd ed. (New York: Guilford,
2011).
111 Cp. PI §133.
112 See R.P. Bentall, Madness Explained (London: Penguin, 2003) and A.T. Beck, N.A.
Rector, N. Stolar, and P. Grant, Schizophrenia: Cognitive Theory, Research, and Therapy (New York: Guildford, 2008).
113 See Beck et al., Schizophrenia and H.E. Nelson, Cognitive-Behavioural Therapy with
Delusions and Hallucinations, 2nd ed. (Cheltenham: Nelson Thornes, 2005).

Experimental Philosophy

281

behavioral therapy are ultimately geared towards the achievement of
metacognitive insight.
For some health conditions, the therapeutic aims can be attained through
insight into the underlying automatic processes, without changing their
course or outputs. For example, cognitive therapists accept that they cannot put an end to auditory hallucinations (voices) in psychosis and rather
seek to change patients’ attitude towards their voices, so that they are
no longer distressed by them or inclined to obey their commands. Key
to this change of attitude is coming to understand how voices arise from
inner speech, through failure of source monitoring, in response, e.g., to
trauma, and that voices articulate own thoughts, not others’ opinions or
commands.114
I suggest that analogous metacognitive insight into how automatic
processes shape intuitions and philosophical thought are what more is
apparently required for therapeutic philosophy, in addition to the exposure of mistakes through the diagnostic approach.115 The philosophical
problem we considered was generated by unsound paradoxes driven by
intuitions that are cognitive illusions. The analogy to cognitive therapy
suggests that thinkers will come to effectively give up pursuing such a
problem the moment they are not only aware of the fallacies involved
in its motivating reasoning, but understand the automatic cognitive processes that generate the intuitions at its root, and how and why these
processes had them commit those fallacies and find the intuitive conclusions plausible. Again, therapy need not change the course and outputs
of relevant automatic processes which include fundamental comprehension processes. Rather, it needs to change thinkers’ attitude towards the
intuitions these processes deliver: replacing acceptance by detachment,
as default, and by rejection, wherever vitiating circumstances obtain. Key
to this is the acquisition of relevant metacognitive insight: of a shared
understanding of what automatic processes drive our intuitive inferences
and judgments and of precisely when and why they go awry.
Such understanding has benefits beyond the individual problem considered: It helps prevent falling for similar problems and devises measures
to guard against them. For example, insight into the conditions under
which stereotypical enrichment leads to contextually inappropriate inferences helps us identify such inferences in lines of thought that set up other
problems and to formulate case-descriptions in ways that do not invite
such inferences (e.g., explicitly marking rarefied uses and refraining from

114 See Bentall, Madness Explained, 347–377 and Nelson, Cognitive Behavioural Therapy, 228–247.
115 In fact, such metacognitive insight is required already to defend diagnostic accounts in
the light of plausible principles of charity (e.g., Thagard and Nisbett 1983) (see Section
1.2).

282

Eugen Fischer

giving special uses to words that already have clearly dominant uses in
ordinary discourse).
The relevant metacognitive insight or enhanced self-understanding begins
to be provided by experimental philosophy’s Sources Project. Just as clinical psychologists have systematically built on findings from cognitive and
social psychology and cognitive neuroscience to develop empirically supported models of how symptoms arise in different mental health conditions,
so experimental philosophers are systematically building on findings from
those very disciplines, to develop empirically supported models of how
philosophically relevant intuitions arise—and, more generally, of how automatic cognition shapes philosophically relevant thought. Just as the former
models provide the key input for cognitive behavioral therapy, so the latter models provide the key input for a therapeutic philosophy. In summary,
the findings from experimental philosophy’s Sources Project contribute to
providing a fresh empirical rationale for therapeutic conceptions of philosophy in the wake of Wittgenstein and can give us precisely the kind of metacognitive insight that is necessary to put such conceptions into practice.116

References
Adler, J.E. “Fallacies and Alternative Interpretations.” Australasian Journal of
Philosophy 72 (1994): 271–282.
Alexander, J. Experimental Philosophy. Cambridge: Polity, 2012.
Alexander, J., Gonnerman, C., and Waterman, J. “Salience, and Epistemic Egocentrism.” In Advances in Experimental Epistemology, edited by J. Beebe, 97–118.
London: Bloomsbury, 2015.
Austin, J.L. Sense and Sensibilia. Oxford: Oxford University Press, 1962.
Ayer, A.J. The Problem of Knowledge. London: Penguin, 1956/1990.
Baker, G.P. “Wittgenstein’s Method and Psychoanalysis.” In his, Wittgenstein’s
Method: Neglected Aspects, 205–222. Oxford: Blackwell, 2004.
Baker, G.P., and Hacker, P.M.S. Wittgenstein: Meaning and Understanding. Essays
on the Philosophical Investigations. Repr. with corrections 1992. Oxford:
Blackwell, 1983.
Banicki, K. “Philosophy as Therapy: Towards a Conceptual Model.” Philosophical Papers 43 (2014): 7–31.
Bargh, J.A. “The Four Horsemen of Automaticity.” In Handbook of Social Cognition, edited by R. Wyer and T. Srull, 1–40. Hillsdale: Earlbaum, 1994.
Beck, A.T., Rector, N.A., Stolar, N., and Grant, P. Schizophrenia: Cognitive Theory,
Research, and Therapy. New York: Guildford, 2008.
Beck, J.S. Cognitive Therapy: Basics and Beyond. 2nd ed. New York: Guilford, 2011.
Bentall, R.P. Madness Explained. London: Penguin, 2003.
Bortfeld, H., and McGlone, M.S. “The Continuum of Metaphor Processing.”
Metaphor and Symbol 16 (2001): 75–86.
Bowdle, B.F., and Gentner, D. “The Career of Metaphor.” Psychological Review
112 (2005): 193–216.

116 I thank Kevin Cahill, Wolfgang Kienzler, and Severin Schroeder for helpful comments.

Experimental Philosophy

283

Boyd, K., and Nagel, J. “The Reliability of Epistemic Intuitions.” In Current Controversies in Experimental Philosophy, edited by E. Machery and E. O’Neill,
109–127. London: Routledge, 2014.
Boyle, M. Schizophrenia: A Scientific Delusion. 2nd ed. London: Routledge, 2000.
Cappelen, H. Philosophy Without Intuitions. Oxford: Oxford University Press,
2012.
Cath, Y. “Reflective Equilibrium.” In Oxford Handbook of Philosophical Methodology, edited by H. Cappelen, T. Szabo Gendler, and J. Hawthorne, 213–230.
Oxford: Oxford University Press, 2016.
Clouser, D., Culver, C., and Gert, B. “Malady.” In What Is Disease?, edited by J.
Humber and R. Almeder, 173–219. Totowa, NJ: Humana Press, 1997.
Collins, J. “Naturalism Without Metaphysics.” In Experimental Philosophy, Rationalism, and Naturalism, edited by E. Fischer and J. Collins, 85–109. London:
Routledge, 2015.
Crane, T., and French, C. “The Problem of Perception.” In The Stanford Encyclopedia of Philosophy, edited by N. Zalta, Summer 2015. http://plato.stanford.
edu/entries/perception-problem/
De Mesel, B. “On Wittgenstein’s Comparison of Philosophical Methods to Therapies.” International Journal of Philosophical Studies 23 (2015): 566–583.
Deutsch, M. The Myth of the Intuitive. Cambridge, MA: MIT Press, 2015.
Elman, J.L. “On the Meaning of Words and Dinosaur Bones: Lexical Knowledge
Without a Lexicon.” Cognition 33 (2009): 547–582.
Erk, K. “Vector Space Models of Word Meaning and Phrase Meaning: A Survey.”
Language and Linguistics Compass 6 (2012): 635–653.
Evans, J.St.B.T. Thinking Twice: Two Minds in One Brain. Oxford: Oxford University Press, 2010.
Farrell, B.A. “An Appraisal of Therapeutic Positivism.” Mind 55 (1946): 25–48
and 133–150.
Faust, M., and Gernsbacher, M.A. “Cerebral Mechanisms for Suppression of
Inappropriate Information During Sentence Comprehension.” Brain and Language 53 (1996): 234–259.
Fein, O., Yeari, M., and Giora, R. “On the Priority of Salience-based Interpretations: The Case of Sarcastic Irony.” Intercultural Pragmatics 12 (2015): 1–32.
Fischer, E. “A Cognitive Self-Therapy—PI 138–97.” In Wittgenstein at Work:
Method in the “Philosophical Investigations”, edited by E. Ammereller and E.
Fischer, 86–126. London: Routledge, 2004.
———. “Wittgenstein’s Non-Cognitivism—Explained and Vindicated.” Synthese
162 (2008): 53–84.
———.“Diseases of the Understanding and the Need for Philosophical Therapy.”
Philosophical Investigations 34 (2011): 22–54.
———. Philosophical Delusion and Its Therapy. New York: Routledge, 2011.
———. “Philosophical Intuitions, Heuristics, and Metaphors.” Synthese 191
(2014): 569–606.
———. “Verbal Fallacies and Philosophical Intuitions: The Continuing Relevance
of Ordinary Language Analysis.” In J.L. Austin on Language, edited by Brian
Garvey, 124–140. Basingstoke: Palgrave Macmillan, 2014.
———. “Mind the Metaphor! A Systematic Fallacy in Analogical Reasoning.”
Analysis 75 (2015): 67–77.
———. “Two Analogy-strategies: The Cases of Mind Metaphors and Introspection.” Connection Science (2017): doi:10.1080/09540091.2017.1350937

284

Eugen Fischer

Fischer, E., and Collins, J. “Rationalism and Naturalism in the Age of Experimental Philosophy.” In Experimental Philosophy, Rationalism and Naturalism,
edited by E. Fischer and J. Collins, 3–33. London: Routledge, 2015.
Fischer, E., and Engelhardt, P.E. “Intuitions’ Linguistic Sources: Stereotypes, Intuitions, and Illusions.” Mind & Language 31 (2016): 65–101.
———. “Stereotypical Inferences: Philosophical Relevance and Psycholinguistic
Toolkit.” Ratio (2017): doi:10.1111/rati.12174
———. “Diagnostic Experimental Philosophy.” Teorema 36(3) (2017): 117–137.
Fischer, E., Engelhardt, P.E., and Herbelot, A. “Intuitions and Illusions: From Explanation and Experiment to Assessment.” In Experimental Philosophy, Rationalism and Naturalism, edited by E. Fischer and J. Collins, 259–292. London:
Routledge, 2015.
———. (under review). “Salience Effects Drive Inappropriate Inferences in Philosophical Arguments: An Interdisciplinary Investigation of Philosophical Paradox.”
Gentner, D. “Structure Mapping: A Theoretical Framework for Analogy.” Cognitive Science 7 (1983): 155–170.
Gerken, M., and Beebe, J. “Knowledge in and Out of Contrast.” Nous 50 (2016):
133–164.
Giora, R. On Our Mind: Salience, Context, and Figurative Language. Oxford:
Oxford University Press, 2003.
Giora, R., Raphaely, M., Fein, O., and Livnat, E. “Resonating With Contextually Inappropriate Interpretations: The Case of Irony.” Cognitive Linguistics
25 (2014): 443–455.
Givoni, S., Giora, R., and Bergerbest, D. “How Speakers Alert Addressees to Multiple Meanings.” Journal of Pragmatics 48 (2013): 29–40.
Hampton, J.A. 1995. “Testing the Prototype Theory of Concepts.” Journal of
Memory and Language 34 (1995): 686–708.
Harcourt, E. “Wittgenstein and Psychoanalysis.” In The Blackwell Companion to
Wittgenstein, edited by H.-J. Glock and J. Hyman, 651–665. Oxford: Blackwell, 2017.
Henderson, D.K., and Horgan, T. The Epistemological Spectrum: At the Interface of Cognitive Science and Conceptual Analysis. Oxford: Oxford University
Press, 2011.
Horwich, P. Wittgenstein’s Metaphilosophy. Oxford: Oxford University Press,
2012.
Hymers, M. Wittgenstein on Sensation and Perception. New York: Routledge, 2017.
Kahneman, D. Attention and Effort. Englewood Cliffs, NJ: Prentice-Hall, 1973.
———. Thinking Fast and Slow. London: Allen Lane, 2011.
Kahneman, D., and Frederick, S. “A Model of Heuristic Judgment.” In The Cambridge Handbook of Thinking and Reasoning, edited by K.J. Holyoak and R.
Morrison, 267–293. Cambridge: Cambridge University Press, 2005.
Knobe, J., and Nichols, S. “An Experimental Philosophy Manifesto.” In Experimental Philosophy, edited by J. Knobe and S. Nichols, 3–14. Oxford: Oxford
University Press, 2008.
Laeng, B., Sirois, S., and Gredeback, G. “Pupillometry: A Window to the Preconscious?” Perspectives on Psychological Science 7 (2012): 18–27.
Lakoff, G., and Johnson, M. Metaphors We Live By. Chicago: Chicago University
Press, 1980.
———. Philosophy in the Flesh. New York: Basic Books, 1999.

Experimental Philosophy

285

Lazerowitz, M. “The Passing of an Illusion.” In Necessity and Language, edited
by M. Lazerowitz and A. Ambrose, 200–240. Beckenham: Croom Helm, 1985.
Levinson, S.C. Presumptive Meanings: The Theory of Generalized Conversational
Implicature, Cambridge, MA: MIT Press, 2000.
Lewinski, M. “The Paradox of Charity.” Informal Logic 32 (2012): 403–439.
Macpherson, F. “The Philosophy and Psychology of Hallucination.” In Hallucination: Philosophy and Psychology, edited by F. Macpherson and D. Platchias,
1–38. Cambridge, MA: MIT Press, 2013.
Mallon, R. “Experimental Philosophy.” In Oxford Handbook of Philosophical
Methodology, edited by H. Cappelen, T. Szabo Gendler, and J. Hawthorne ,
410–433. Oxford: Oxford University Press, 2016.
McRae, K., Ferretti, T.R., and Amyote, I. “Thematic Roles as Verb-specific Concepts.” Language and Cognitive Processes 12 (1997): 137–176.
Moors, A., and De Houwer, J. “Automaticity: A Theoretical and Conceptual
Analysis.” Psychological Bulletin 132 (2006): 297–326.
Nagel, J. “The Psychological Basis of the Harman-Vogel Paradox.” Philosophers’
Imprint 11(5) (2011): 1–28.
———. “Intuitions and Experiments: A Defence of the Case Method in Epistemology.” Philosophy and Phenomenological Research 85 (2012): 495–527.
Nelson, H.E. Cognitive-Behavioural Therapy With Delusions and Hallucinations.
2nd ed. Cheltenham: Nelson Thornes, 2005.
Pohl, R. (ed.) Cognitive Illusions. New York: Psychology Press, 2004.
Pollock, J. “Reliability and Justified Belief.” Canadian Journal of Philosophy 14
(1984): 103–114.
Pust, J. “Intuition.” In The Stanford Encyclopedia of Philosophy, edited by E.N.
Zalta, 2012. http://plato.stanford.edu/ archives/win2012/entries/intuition/.
Robinson, H. Perception. London: Routledge, 2001.
Rosch, E. “Principles of Categorisation.” In Cognition and Categorization, edited
by E. Rosch and B. Lloyd, 27–48. Hillsdale, NJ: Erlbaum, 1978.
Rumelhart, D. E. “Schemata: The Building Blocks of Cognition.” In Theoretical
Issues in Reading Comprehension, edited by R. Spiro, B. Bruce, and W. Brewer,
33–58. Hillsdale, NJ: Erlbaum, 1980.
Schroeder, S. Wittgenstein: The Way Out of the Fly-Bottle. Malden, MA: Polity, 2006.
Searle, J. “Metaphor.” In Metaphor and Thought, edited by A. Ortony, 2nd ed.,
83–111. Cambridge: Cambridge University Press, 1993.
Simpson, G.B., and Burgess, C. “Activation and Selection Processes in the Recognition of Ambiguous Words.” Journal of Experimental Psychology: Human
Perception and Performance 11 (1985): 28–39.
Smith, A.D. The Problem of Perception. Cambridge, MA: Harvard University
Press, 2002.
Spärck-Jones, K. Synonymy and Semantic Classification. PhD thesis, University of
Cambridge. Cambridge Language Research Unit, 1964.
Stich, S., and Tobia, K. “Experimental Philosophy and the Philosophical Tradition.”
In Blackwell Companion to Experimental Philosophy, edited by J. Sytsma and
W. Buckwalter, 5–21. Malden, MA: Wiley Blackwell, 2016.
Tarras, P. “ ‘Philosophie’ Grammatisch Betrachtet: Wittgensteins Begriff der Therapie.” Kriterion—Journal of Philosophy 28 (2014): 75–97.
Thagard, P., and Nisbett, R.E. “Rationality and Charity.” Philosophy of Science
50 (1983): 250–267.

286

Eugen Fischer

Thompson, V.A., Prowse Turner, J.A., and Pennycook, G. “Intuition, Reason, and
Metacognition.” Cognitive Psychology 63 (2011): 107–140.
Till, R.E., Mross, E.F., and Kintsch, W. “Time Course of Priming for Associate
and Inference Words in a Discourse Context.” Journal of Verbal Learning and
Verbal Behaviour 16 (1988): 283–298.
Turney, P.D., and Pantel, P. “From Frequency to Meaning: Vector Space Models
of Semantics.” Journal of Artificial Intelligence Research 37 (2010): 141–188.
Weinberg, J.M. “How to Challenge Intuitions Empirically Without Risking Scepticism.” Midwest Studies in Philosophy 31 (2007): 318–343.
———. “Humans as Instruments: Or, the Inevitability of Experimental Philosophy.” In Experimental Philosophy, Rationalism, and Naturalism, edited by E.
Fischer and J. Collins, 171–187. London: Routledge, 2015.
———. “Intuitions.” In Oxford Handbook of Philosophical Methodology, edited
by H. Cappelen, T. Szabo Gendler, and J. Hawthorne, 287–308. Oxford:
Oxford University Press, 2016.
Williams, M. Unnatural Doubts: Epistemological Realism and the Basis of Scepticism. Princeton, NJ: Princeton University Press, 1996.
Wilson, T.D. Strangers to Ourselves: Discovering the Adaptive Unconscious.
Cambridge, MA: Harvard University Press, 2002.
Wisdom, J. “Philosophy and Psychoanalysis.” In his, Philosophy and Psychoanalysis, 169–181. Oxford: Blackwell, 1964.
Wittgenstein, L. Culture and Value. Rev. ed. Oxford: Blackwell, 1998.
———. Philosophical Investigations, translated by G.E.M. Anscombe. 3rd ed.
Oxford: Blackwell, 2001.
———. The Big Typescript TS 213, edited and translated by C.G. Luckardt and
M.A.E. Aue. Oxford: Blackwell, 2005.

13 Representationalism,
Metaphysics, Naturalism
Price, Horwich, and Beyond
Jonathan Knowles

Introduction
Until quite recently, a view that combined the idea that believing and
asserting are fundamentally a matter of representing a mostly mindindependent world with the idea that that world, including the mental
and representing parts, could be given a metaphysical rendering through
the concepts of natural science was standard within analytic philosophy.
With the recent (re-)emergence of deflationism about truth and pragmatist approaches to mind and language, this consensus is today at least
seriously challenged. Two central thinkers in this wave are Paul Horwich
and Huw Price. Though it is only Price that explicitly subscribes to ‘pragmatism’, both see themselves as what I call ‘anti-representationalists’,
and also—at least in a significant sense—anti-metaphysicians. Their antirepresentationalism consists both in a disavowal of truth and reference
as substantial relations explanatory of meaning, and in seeing different
domains of talk or ‘discourses’ as playing different kinds of roles in our
lives. Their anti-metaphysicalism, though more difficult to characterize
in a neutral way, shares at least the idea that a reductive, metaphysical
form of naturalism is misguided. They both nevertheless uphold a kind of
naturalistic commitment in aiming to provide a theory of language and
language-use that does not regard a normatively laden notion of content
as primitive.1 Beyond this however there at least appear to be some significant divergences: Price expresses allegiance to a more general naturalist position he calls subject naturalism, whilst Horwich explicitly declares
himself a non-naturalist when it comes to things like ethical values. We
shall be investigating this and other apparent differences between the
thinkers in the sequel.
1 Paul Horwich, Meaning (Oxford: Oxford University Press, 1998), ch. 8; Huw Price,
“What Should a Deflationist About Truth Say About Meaning?” in Truth, Philosophical
Issues, Vol. 8, ed. Enrique Villanueva (California: Ridgeview, 1997).

288

Jonathan Knowles

A further point of convergence consists in the fact they both see their
views as having important resonance (at least) with central themes of the
later Wittgenstein.2 Price writes:
For my part, I have considerable sympathy with what I take to be
two important ingredients of Wittgenstein’s view, on any adequate
interpretation. One of these ingredients plays down the theoretical
significance of the idea that a function of (a large part of) language
is to ‘describe’ or ‘represent’ reality. The other plays up the idea
that the language concerned has many different functions, in a way
that is not evident ‘on the surface’. I’ll call these ingredients nonrepresentationalism and functional pluralism, respectively.3
For Price, this picture of language has implications for a certain metaphysical project, as we shall see. Horwich meanwhile stresses Wittgenstein’s
skepticism to substantive theorizing in philosophy—to metaphysics more
generally—in terms of our bewitchment by similarity of linguistic form.
We assume “that to take [say] our mathematical discourse at face value
requires the adoption of a metaphysical picture with many attendant difficulties. [. . . But] arithmetical discourse [. . .] need [not] necessarily carry
with [it] the proprietary of puzzlement about the nature, location and
accessibility of [arithmetical] reality [. . .] Wittgenstein’s therapy [. . .]
relieves the grip of these tempting oversimplifications”.4
Though Wittgenstein is an important figure of the broader background,
the details of Wittgenstein’s views will not be central stage here (beyond the
extent to which Horwich’s own metaphilosophical position faithfully capture these, as he claims they do). Rather, my aim is systematically to understand how anti-representationalism, anti-metaphysicalism, and naturalism
might interrelate. Price’s work, perhaps uniquely in the contemporary
debate, has been devoted explicitly to these issues, leading to the development of a position he terms global expressivism.5 Horwich, however, also
has views on them, as witnessed not least in a recent exchange with Price.6
Insofar as he is a central thinker in the anti-representationalist camp, has
much in common with Price, and yet professes to differ from Price concerning the implications of anti-representationalism for naturalism and
2 See especially Ludwig Wittgenstein, Philosophical Investivations, 3rd ed. Translated by
G. E. M. Anscombe (Oxford: Basil Blackwell, 1968).
3 Huw Price, “Immodesty Without Mirrors—Making Sense of Wittgenstein’s Linguistic
Pluralism,” reprinted in Huw Price, Naturalism Without Mirrors (Oxford: Oxford University Press, 2011), 201.
4 Paul Horwich, Wittgenstein’s Metaphilosophy (Oxford: Oxford University Press, 2011),
15–16.
5 Price, Naturalism; Huw Price et al., Expressivism, Pragmatism and Representationalism
(Cambridge: Cambridge University Press, 2013).
6 Paul Horwich, “Naturalism, Deflationism and the Relative Priority of Language and
Metaphysics,” in Price et al., Expressivism, 178–186.

Metaphysics

289

metaphysics, it would appear fruitful to consider his take on these issues. In
this paper, I want to examine what the commitments amount to and how
they fit together for these two philosophers and to compare and critically
discuss the resultant overall views. I then want to suggest my own somewhat different understanding of how the ideas might fit together.7
The paper starts in section 1 with Price’s argument against metaphysics from anti-representationalism, leading to a sketch of his global expressivism. In section 2 I present Horwich’s critique of Price’s argument and
Price’s attempt to defuse this, before offering my own appraisal of Horwich’s position. I seek both to show that he is not as distant from Price as
he presents himself, and also to provide my own (I hope clearer) understanding of what anti-representationalism involves and of its relation to
metaphysics at a general level. This paves the way for a new critique of
Price’s anti-metaphysicalism in section 3, where I argue that Horwich—
though not for exactly the reasons he gives—is basically right that antirepresentationalism does not definitively rule out metaphysics. In section 4
I consider Horwich’s Wittgensteinean critique of metaphysics and philosophy more generally; a full treatment of this is not feasible here but I will
suggest it is at least unlikely to do the service of filling any gap left by Price’s
argument against metaphysics. Finally, in section 5, I sketch what I think
is a more promising anti-representationalist and anti-metaphysicalist, yet
also distinctively naturalistic, picture, one which draws on the Wittgensteinean (and, as we shall, Carnapian) ideas that Price and Horwich build
on, but also views about science and the nature of scientific enquiry that
can be identified perhaps most clearly in the work of Noam Chomsky.

1. ‘Naturalism Without Representationalism’
The above is the title of the paper in which Price’s argument against metaphysics from anti-representationalism is most succinctly formulated;8
it is also this argument that is the focus of Horwich’s attack on Price.
This section gives a presentation of it and of Price’s global expressivism
(henceforth GE).
The argument focuses on a metaphysical thesis Price calls object naturalism (henceforth ON): The reductive view that all truth or knowledge
7 I should stress that I will be concerned for the most part here, following Price, with
so-called ‘placement’ metaphysics, not metaphysics in its entirety (assuming this goes
beyond the former). I should also mention at the outset another thinker who recently
has had much to say about these issues, in a similar vein: Amie Thomasson. Though not
central stage here, her views will play an important role in section 2.
8 Huw Price, “Naturalism Without Representationalism,” reprinted in Price, Naturalism.
Other papers that cover similar ground but with somewhat different foci are Huw Price,
“The Semantic Foundations of Metaphysics,” reprinted in Price, Naturalism, and Huw
Price and Peter Menzies, “Is Semantics in the Plan?” in Conceptual Analysis and Philosophical Naturalism, eds. David Braddon-Mitchell and Roberta Nola (Cambridge, MA:
MIT Press, 2009).

290

Jonathan Knowles

is fundamentally of natural scientific character; that in a sense “all there
is is the world studied by science”.9 The problem for ON, according to
Price, is not so much that it is false, but that it is, or at least is in danger
of being, incoherent or at least somehow irrational. In this way, Price’s
attack on ON turns into a more general attack on the kind of metaphysical theorizing, naturalistic or non-naturalistic, that lies behind it.
The argument proceeds via a dilemma: One can either understand ON
as having a linguistic or conceptual starting point,10 or in a material manner; but on neither understanding is ON coherently motivated (at least
obviously). Though the latter understanding might appear to be more
straightforward and indeed to be more what is intended by the content
of ON, it is precisely one of Price’s aims to show that the motivation for
ON is not as clear as many seem to assume. Price considers the linguistic
understanding first, arguing that while metaphysical issues may start in
the linguistic realm—as questions about what we mean by this or that
term or locution, and how we might analyse it—a supporter of ON will
have at some point to assume that linguistic items represent bits of extralinguistic reality to take us from language to the world. ON will then
be understood as claiming that all the truth makers of our true claims
can, at least after these are subjected to a suitable process of analysis, be
seen as natural or physical truth-makers (in some appropriately demarcated sense of ‘natural’ or ‘physical’). This understanding of ON corresponds to the assumptions of what Price has called the ‘Canberra plan’,
as pursued by people like David Lewis and Frank Jackson. It depends on
there being substantial reference and truth relations, that is, ones which
explain meaning by relating bits of language to bits of the world—that
is, on representationalism.11 Thus, when we say that, e.g., experiences or
values are ultimately physical, we are saying that the referents of terms
like ‘pain’, ‘good’, and so on are in fact physical in nature.
ON has then a fairly clear content when construed on the linguistic
model. However, there is just one hitch: Representationalism is not something you get for free; rather, it is a substantial theoretical claim about
how language functions.12 Thus, one needs at least what Price calls a
subject naturalistic enquiry—into the nature of human language and

9 Price, “Naturalism,” 185. The view is more standardly referred to as metaphysical naturalism; sometimes physicalism. The epistemological and ontological versions of the
view are treated as equivalent by Price (given his purposes), and likewise here.
10 I follow Price in glossing the linguistic versus conceptual distinction for present purposes (and like him will talk mostly in terms of language).
11 Representationalism is not just the idea of there being truth-makers and truth-making
relations. Semantic deflationary views can for example accept that moral states of
affairs make moral statements true or false. Rather representationalism assumes that
there is some non-trivial specification of reference that also allows for non-trivial truthmakers in a way that in turn makes space for ON.
12 Indeed, it is not the only hitch for ON according to Price, for even if one accepts it there
may be problems in understanding referential relations determinately enough to allow

Metaphysics

291

thought, construed itself as a scientific project—prior to declaring ON
viable. Moreover, representationalism is something Price thinks we have
good reasons to be suspicious of as such a naturalistic theory of language; and that there is a promising alternative to it, semantic deflationism, which avoids the problems that infect representationalism. Though
I think Price’s arguments here are in fact quite convincing,13 the general tenability of semantic deflationism and the anti-represetationalism
which I see as attending it are things I am simply taking for granted here
(though what they amount to will be important to get clear on in the
following section). Our dialectical focus is thus the alleged link between
representationalism and metaphysics, in particular Price’s claim that if
you start your metaphysics in the linguistic realm and yet deny representationalism, there is going to be a problem in so much as making sense of
what a doctrine like ON could amount to.14
Let us now turn to the material understanding of ON, according to
which “we do metaphysics without semantic crutches”.15 Here Price
argues there are two fundamental problems. One rests on the idea that a
supporter of ON should be able to frame an argument for her position.
Such an argument can be mounted through the Canberra planners’ idea
that all semantic roles have to be filled by naturalistic occupiers; but this
takes us back to the linguistic conception, and the problems with representationalism. Yet without this starting point, it is just not clear how
one would frame a general argument for ON, and thus it wouldn’t get off
the ground.16 The second problem is that the possibility of a deflationary treatment of discourse shows that ‘the cat is out of the bag’, as Price
puts it: If we accept deflationism, we do not need to think in terms of any
worldly items to which our words may correspond to understand their
role in our lives, but rather can focus on explaining the talk itself.17
It is the latter idea that forms the project of GE, which Price describes
more fully in subsequent work.18 GE seeks to explain a discourse by
looking at its underlying function, taking as its starting point a view of
humans as natural, evolved beings in a natural environment, in accord

13

14

15
16
17
18

useful resolution of metaphysical questions; cf. Stephen Stich, Deconstructing the Mind
(Oxford: Oxford University Press, 1996).
For discussion and defence, see Jonathan Knowles, “Naturalism Without Metaphysics,” in Realism, Science and Pragmatism, ed. Kenneth Westphal (London: Routledge,
2014), §2.
Barring at least the eventuality that we can exhaustively analyze problematic claims,
such as those of ethics, in terms of physical/naturalistic ones—a vindication of ON
which Price it seems is in principle open to (pers. comm.), but presumably thinks so
unlikely as to be not worth mentioning. The idea of purely ‘conceptual’ metaphysics is
also discussed in the following section.
Price, “Naturalism,” 196.
Ibid., 197.
Ibid., 195 f.
Price, Naturalism, ch. 1; Price, Expressivism.

292

Jonathan Knowles

with subject naturalism. It is a global project because, though different discourses serve different functions, an anti-representational, pragmatic explanation applies to each—including our talk about the natural
world itself. Price allows that to explain the function of some vocabularies we will need to use their referring terms, while with others this
will not be necessary (that is, we will only need to use the referring
terms of other discourses).19 This distinction is coordinate with another
he makes between vocabularies that do and those that do not have
an e-representational character.20 E-representation for Price is a nonsemantic relation of something like covariation or causation between
the terms of a vocabulary and the items these terms genuinely semantically, albeit deflationistically, refer to in the natural world.21 The bet
is that in domains like ethics terms will not e-represent, whereas in, say,
that of middle-sized dry goods or science—the domain of the natural,
in a certain recognizable sense—they will. The distinction is not meant
to be absolute, thus allowing gradations of e-representationality; nor
is it given a priori which domains will and which will not turn out to
be e-representational. But however things exactly pan out, Price’s overall point is that his non-representational enquiry into our different discourses will not leave any significant question unanswered and hence no
room for metaphysical quandaries.
How successful is Price’s argument against ON or placement metaphysics more generally? The first stage of it seems sound, but I think one
might be forgiven for finding the second set of arguments against ON,
when understood as having a material starting point—even when embellished in the way I have just indicated—somewhat less than totally transparent. This it seems applies at least to Paul Horwich, to whose critique
of Price I now turn.

2. Horwich Versus Price
The obvious commonality between Price and Horwich consists in what
I am calling their anti-representationalism (henceforth AR): a commitment to semantic deflationism, and an embracement of something like
functional pluralism about language. As noted, they also both deny that
any theory of language or meaning must acknowledge as primitive some

19 Given this, ‘global expressivism’ might appear somewhat misleading as a label for Price’s
position. He is aware of this, however—sometimes instead using ‘global pragmatism’—
the important point for him being that the account is not semantically representationalist
(see, e.g., Price, Expressivism, 176). See also below.
20 Ibid., ch. 2.
21 Price uses the term ‘i-representation’ to characterize this latter, ‘language-internal’ relation. Neither i-representation nor e-representation correspond to representation in the
sense assumed by representationalism; in Price’s view, the latter makes the mistake of
merging what are in fact two separate notions. See ibid. for further discussion.

Metaphysics

293

normatively laden notion of content.22 However, it would appear their
views part ways with respect to further issues concerning metaphysics
and naturalism. In particular, Horwich rejects Price’s argument from
AR against ON and metaphysics more generally.23 Horwich rejects ON
alright, and indeed the viability of the kind of substantive philosophical
theorizing that metaphysics generally involves, but his reasons for doing
this are different and based on what he sees as Wittgenstein’s critique of
‘traditional philosophy’ (see section 4, below). He also states explicitly
that he considers himself a non-naturalist about things like ethics,24 a
commitment that might seem to place him amongst the ranks of the incoherent metaphysicians by Price’s standards.
It can seem odd that the two thinkers should really be so strongly at
odds as Horwich suggests, given their profound agreement on AR. In
this section, I will try to shed some light on this conundrum and identify
what is really at stake between the two, arguing that the gulf between
them is less deep than Horwich suggests—even though he is also right, as
we shall go on to see, to reject Price’s argument. I will do this partly with
reference to Price’s commentary on Horwich’s piece, though also by relying on my own critical evaluation of Horwich’s position, which I hope
will also lead to a better understanding of what AR involves and how it
relates to metaphysics at a general level.
First, however, we need to look at Horwich’s rendering of Price’s argument against ON and why he says he rejects it. According to Horwich,
Price sees metaphysics as having in the first instance a linguistic subject
matter; that is, Price thinks (according to Horwich) that “[m]etaphysical questions can be answered only insofar as they are transformations
of more basic linguo-conceptual questions”.25 To be then genuinely
metaphysical—to be about the world and not just our concepts—a view
like naturalism (by which Horwich means ON) requires us to transpose

22 There are some subtle differences between the two concerning their anti-representationalism.
To start with, Price sees our concept of truth as having a primitive and sui generis
normative function (Huw Price, “Truth as Convenient Friction,” reprinted in Price,
Naturalism), while Horwich sees the normativity attaching to truth-seeking as derivate
(Paul Horwich, “The Value of Truth,” Noûs 40 (2006)). Secondly, ’functional plurality’
is Price’s term (thus I say ‘something like), and in his hands involves a stronger leaning towards a form of expressivism, something Horwich, at least insofar as he follows
Wittgenstein, disavows (Wittgenstein’s Metaphilosophy, 62, footnote). Finally, there are
differences with respect to how ‘immodest’ they view their respective theories of meaning
as being (cf. Price, “What Should a Deflationist”). That said I think their commonalities
in respect of anti-representationalism are more significant than their differences (and, for
the record, I will not be questioning their ‘immodesty’ here, though for critical discussion of this, see David Macarthur, “Subject Naturalism, Scientism and the Problem of
Linguistic Meaning: Critical Remarks on Price’s ‘Naturalism Without Representationalism’,” Analasis: Revista de invetagacion filosofica 1 (2014)).
23 Horwich, “Naturalism”.
24 Ibid., §5.
25 Ibid., 115.

294

Jonathan Knowles

these questions from a linguistic to a material key; and this, in turn,
requires substantive semantic relations (ibid.). AR thereby shows naturalism is ‘impossible to establish’.26
Horwich has several critical remarks about this argument, but his central objection is that though linguistic analysis may be a necessary preliminary to metaphysical concerns—in the sense that one needs to ask
whether, say, ethical claims are so much as in the business of ‘saying how
things are’—it remains open once one has decided, as the case might be,
that they are, that they might nevertheless be seen as uniformly false, a
useful fiction, reducible to naturalistic truths of some kind, or whatever
more subtle account the placement metaphysician might come up with.27
Thus, the linguistic starting point does not after all preclude the typical
metaphysical gambits of ON, even if one rejects representationalism.
One can easily feel Horwich has misunderstood Price here—as Price
himself suggests in his reply. As Price puts it, GE isn’t meant to be “a way
of doing metaphysics in a pragmatist key [but] a way of doing something
like anthropology [. . .] not a matter of recasting issues of metaphysics
as issues about language, but of abandoning the metaphysical questions
altogether”.28 On the other hand, Price accepts that metaphysics understood as conceptual analysis is legitimate;29 a similar view is defended
by Amie Thomasson in her Price-friendly defence of a neo-Carnapian
approach to ontology.30 An important issue thus does seem to be raised
by Horwich—as indeed Price accepts31—which is to understand exactly
why, if one rejects representationalism, one can only engage in this and
not in material metaphysics—or metaphysics ‘proper’, as one might say.
Price, recall, had two arguments for this claim. Firstly, representationalist semantics is part of the Canberra plan’s standard ‘toolkit’ and gives a
framework for mounting an argument for ON. To this, Horwich replies
that there might be other motivations for ON and, moreover, that the point
doesn’t affect non-naturalistic metaphysical positions.32 Price largely concedes the force of this, stressing instead his ‘the-cat-is-out-of-the-bag’ argument. As Price explains, this builds on the idea that as far as ‘saying how
things are’ is concerned, there is no deep distinction between assertoric
vocabularies that do and don’t do this to be revealed by linguistic analysis;
that is, the fact-stating/non-fact-stating distinction is deflated along with
truth. Metaphysics is then bypassed by focusing on explaining how we
talk: to explain some vocabularies we will need to make use of the refer-

26
27
28
29
30
31
32

Ibid., 116.
Ibid., 123–124.
Price, Expressivism, 181.
Pers. comm.
Amie Thomasson, Ontology Made Easy (Oxford: Oxford University Press, 2015).
Price, Expressivism, 182.
Horwich, “Naturalism,” 123 fn.

Metaphysics

295

ring terms—those which have a more e-representational character—but
to explain others not. The bet is that with domains like ethics this will not
be the case. But the metaphysical—or, rather, meta-metaphysical—point is
that there is no further issue to be addressed or resolved.33
Is Horwich thus rebutted? Insofar as his objection rests on not seeing
how deflationary Price intends to be—indeed, as far as I can see, as he
himself should be as a good anti-representationalist—I think the answer
is yes; moreover, it does seem that Horwich’s main point against Price is
that the possibility of metaphysical enquiry is beholden to this question
of factuality that it is one of the points of prior linguistic analysis to sort
out. Price thus seems to win this particular skirmish. However, I do not
in the end think Price’s argument in fact rules out material metaphysics, as we shall in section 3. What I therefore want to do in the rest of
this section is offer a somewhat more general critique of Horwich’s position, viewed as the idea that there is in principle no connection between
AR and the possibility of metaphysics. This will then provide us with a
cleaner backdrop for understanding where the limitations in Price’s argument really lie, as well as how to understand the relationship between our
three central concepts more appropriately.
When all’s said and done, the intuitive heart of Horwich’s response to
Price seems to be the following: The fact that ‘X’ refers to X (or Xs) does
not express a substantive semantic relation does not logically entail that
there is nothing substantive to say about the existence or nature of X (or
Xs). Now I think we can all agree with that. However, it would be a mistake to think doing so amounts to accepting that AR has no implications
for one’s take on metaphysics whatsoever. After all, semantic deflationism
has to be couched within a wider theory of meaning. To begin with, as Horwich himself has pointed out,34 deflationism about truth brings deflationism about reference and satisfaction in its wake, and vice versa, in virtue
of their interdefinability. Furthermore, once reference is deflated, then so
plausibly is ontological commitment, something Thomasson, inspired by
Horwich, has argued explicitly.35 Moreover, since all facts about reference
and truth fall out of facts about use, of some kind—again, according to
Horwich not least, but presumably in fact any anti-representationalist—
and since ontological commitment falls out of facts about reference, it
seems we can conclude that we don’t need, as Thomasson has put it, ontological premises to introduce an ontologically committing vocabulary.

33 Price, Expressivism, 182–184.
34 Paul Horwich, From a Deflationary Point of View (Oxford: Oxford University Press,
2004), 79.
35 Amie Thomasson, “Deflationism in Semantics and Metaphysics,” in Metasemantics:
New Essays on the Foundations of Meaning, eds. Alexis Burgess and Brett Sherman
(Oxford: Oxford University Press, 2014). For example, for a singular term, ‘n’, to refer
(at all) is just for there to be an x such that x = n (ibid., 197–198).

296

Jonathan Knowles

The above is in line with a (certain) neo-Carnapian approach to ontology.36 On this view, our linguistic activity is conceived as divided into
different discourses or (to use Carnap’s phrase) frameworks that are individuated by proprietary sets of rules—discourses we to an extent naturally find ourselves with, but may also modify or even introduce ab initio
for various different practical or theoretical purposes. We use expressions
in accord with the relevant rules, and on that basis, relying on simple
logical inference and/or sense experience, arrive at existential claims. For
example, given the rules for arithmetic, we justifiably say that there is a
prime number between two and five and can therefore conclude, trivially,
that there are numbers. Given the rules for the discourse of medium-sized
dry goods, I can establish by sight that there is a chair in my office. And
so on. These are what Carnap would call answers to internal ontological
questions, and as such are perfectly kosher. External ontological claims
or questions, by contrast, are not, precisely because they are posed from
outside all linguistic frameworks and thereby lack the necessary setting
which gives them the only sense such questions can have.
Now one might think that a sense to ‘external’ ontological questions
could be given by focusing on what kinds of terms are in contact with
a fully mind-independent reality, as opposed to those that merely arise
from our deployment of terms in relation to appropriate rules of use.
However, the notion of ‘mind-independent reality’ is surely one that an
anti-representationalist should reject: for the idea of such a reality seems
impossible to extricate from the idea that reference is a matter of us or
our concepts standing in some genuine, substantive relation to something
non-linguistic or non-mental.37 Thus, external questions can only ever

36 Rudolf Carnap, “Semantics, Empiricism and Ontology,” Revue Internationale de Philosophie 4 (1950); Thomasson, Ontology; Huw Price, “Metaphysics After Carnap: The
Ghost Who Walks?” reprinted in Price, Naturalism. I say ‘certain’ because there are
other neo-Carnapian positions in the contemporary literature that adopt a different
take on the significance of Carnap’s views on ontology (e.g., Eli Hirsch, Quantifier Variance and Realism (Oxford: Oxford University Press, 2011)). I am not concerned here
with whose reading is most correct, but I should stress that the interpretation adopted
here is one with intimate connections to AR—something at least Hirsch does not
stress—and moreover whose pluralism is meant to be of a vertical rather than horizontal character (cf. Huw Price, “Metaphysical Pluralism,” reprinted in Price, Naturalism).
37 Though this point strikes me as deeply intuitive, it deserves a fuller treatment; space
considerations preclude this here. It is clearly related to Putnam’s model-theoretic
attack on metaphysical realism (cf. Hilary Putnam, Realism and Reason (Cambridge:
Cambridge University Press, 1983), for a recent defence of which see Tim Button, The
Limits of Realism (Oxford: Oxford University Press, 2013)). For an attempted defense
of the point on my own part, which also considers Michael Devitt’s view that realism
is independent of semantic issues, see Knowles, “Naturalism,” 205–207. Thomasson
provides an independent defence against the “we are talking about mind-independent
reality” riposte from the metaphysician, cf. Thomasson, Ontology, ch. 10.

Metaphysics

297

be pragmatic questions that ask why we talk the way we do, whether we
do well in adopting and employing these modes of discourse, and so on.
What all this shows then is that a kind of deflationary ontological pluralism emerges as a default position for an anti-representationalist. Given
the various different ways we talk, and given that relevant empirical and
other conditions are fulfilled, it follows trivially that we are ontologically
committed to at least most of what many traditional or so-called ‘serious’
metaphysicians have devoted large amounts of energy to questioning and/
or seeking to vindicate in other terms: mental states, ethical and aesthetic
values, medium-sized dry goods, numbers, and so on, as well as entities
of non-fundamental science. If this is the right way to understand AR—
and I submit it is hard to see any viable alternative—then on the face of
it, there are no deep questions of the kind a view like ON seeks to ask
and which Horwich thinks it leaves on the table. Moreover, being committed to, say, values, as something non-natural (i.e. non-scientifically
natural), as Horwich is, should not be understood as a substantial commitment that needs some special argument to support, but viewed simply
as an upshot out of the neo-Carnapian framework and the fact that we
engage rationally in ethical discourse. Pace his own apparent understanding of the situation, then, Horwich need not be seen as infringing Price’s
anti-metaphysicalist conclusion insofar as he describes himself as a nonnaturalist about ethics (and certain other domains).
I think this broader understanding of AR gives a better sense of how
the position relates to metaphysics at a more general level than Horwich’s
commentary on Price’s argument. Nevertheless, I think Horwich is in fact
also right that the argument does not ultimately succeed in showing that
AR rules out metaphysics. The next section explains why.

3. The Real Problems for Price’s Argument
I have just argued that AR has implications for metaphysics, in the sense
that the kind of overall view that naturally accompanies semantic deflationism will lead by default to a deflationary ontological pluralism. However, it is important to stress that this is a default. The question thus needs
to be addressed whether such a default could be coherently or rationally
abrogated consistently with upholding AR. As things stand, given our
natural and extant ways of talking, we are committed to the existence
of numbers, values and so on, along with chairs, protons and electrical fields. But is this in principle compatible with someone coherently
mounting a case against the fundamentality of the former quantities, or
for the idea that they should somehow be reduced to the latter? I believe
it is: There could, pace Price, be a project coherently described as (materialistically) metaphysical, even once one gives up representationalism.
To see this, it is important firstly to emphasize that AR’s rejection of the
idea of reference and truth playing substantive, meaning-constituting roles

298

Jonathan Knowles

in no way implies a rejection of the idea that our talk has semantic content: that it involves reference to entities and saying how things are—in
our ethical, mental and modal discourses as much as our naturalistic ones.
Now for Price of course all the theoretical and explanatory work will be
taking place at a different level from this—at the level where we focus
on use and function, which is moreover pluralistic. Nevertheless, he cannot and would not deny that all discourses do purport to say how things
are, and moreover in a univocal sense—at the same time of course as
different discourses say different things with respect to how things are.38
Now let us imagine an anti-representationalist who, for whatever reason,
wants to prioritize or valorize our scientific talk over all other talk—our
scientific or physical account of how things are over all other accounts—
and to seek to give an account of the latter exclusively in terms of, or that
is uniquely responsible to, the former. This needn’t proceed through conceptual analysis; rather, it could be understood in terms of what Carnap
and later Quine called explication,39 whereby a concept’s role in a theory
is taken over by some other in virtue of the latter being an in some way
more adequate concept that fulfils the same purpose (e.g., heat’s role was
taken over by that of temperature). Now, Price wants to say that this antirepresentationalist’s commitment to ON must be incoherent or at least
irrational—but can he be right?
Considering first the charge of incoherence, it seems this might be
made to stick if one also upheld an absolute division between analytic
and synthetic truth, or between rules of language and the substantive
truths expressed by these. For then the idea that say ethical values might
literally be reducible to physical states of affairs could be passed off as a
kind of ‘category mistake’. Now Carnap certainly sought to uphold such
a distinction; but Quine famously attacked and rejected it,40 and in this
he has been followed by many since. Moreover, there would seem noth38 Price commits himself to truth-conditional content both explicitly (Expressivism, 40)
and implicitly in view of his use of Brandom’s inferential semantics (ibid., 34), which
precisely seeks to understand the notion of content in relation to normative practices.
For a study of the various ways content inevitably figures in Price’s system, see also
Lionel Shapiro, “Linguistic Function and Content: Reflections on Price’s Pragmatism,”
Philosophical Quarterly 64 (2009).
Some might think this commitment to content can play no substantive role within
AR. Thus, in his defense of semantic deflationism, Price argues for a quietist position
about reference and truth, claiming that one needn’t deny in one’s theoretical voice
semantic properties, only not assert their existence and not make use of them (“Naturarlism,” 191–2). However, from the position of one involved in the relevant discourses,
including semantic discourses, it stands fast that there are ethical truths, say, as well as
physical ones, and in the same sense—in whatever deflationary sense of ‘truths’ this is,
from a theoretical perspective—and that is all my argument requires.
39 See, e.g., W.V.O. Quine, Word and Object (Cambridge, MA: MIT Press, 1960), 258 ff.
40 W.V.O. Quine, “Two Dogmas of Empiricism,” in From a Logical Point of View
(Cambridge, MA: Harvard University Press, 1953).

Metaphysics

299

ing about the analytic-synthetic distinction that would make it obviously
endemic to AR: indeed, Price presents his own GE as a natural successor
to Carnap’s own pluralism that precisely does without it, basing itself
instead on the idea of functional rather than semantic plurality.41 Further,
Quine himself seems reasonable to class as an anti-representationalist.
To start with, he subscribes to disquotationalism about truth, a form
of semantic deflationism;42 whilst the idea of there being a plurality of
linguistic frameworks is not one, arguably, he rejects, but rather sees as
lacking fundamental significance when it comes to “limning the true and
ultimate structure of reality”.43 I therefore see it as at least dialectically
germane to appeal to the idea of an anti-representationalist position that
rejects the analytic-synthetic distinction.44
The upshot of this is that a kind of ON which simply puts the scientific
discourse forward as the touchstone of what is true and real (or ‘true’
and ‘real’ if one insists) is not ruled out once one rejects AR. This might
require seeing this metaphysical program as somehow revisionary, but it
is not at all clear, independently of what has already been said, why this
should be problematic or make the position not qualify as a form of ON.
Coherence is one thing; rationality, however, another. Could Price
object that the kind of Quinean, anti-representationalist version of ON
just sketched is nevertheless deeply irrational—unmotivated, arbitrary,
or perverse, given its commitment to AR (as his argument sometimes
suggests is the idea)? I would not deny that it in some sense runs counter
to the pluralist spirit of AR; moreover, GE might on the face of it seem
more in keeping with the latter. However, my Quinean need not deny that
41 Price, “Metaphysics After Carnap,” 287.
42 W.V.O. Quine, Philosophy of Logic (Upper Saddle River, NJ: Prentice Hall, 1970).
43 Quine, Word and Object, 221. I note that this quote should not be seen as indicating a
commitment to some kind of metaphysical realism, that upholds the idea of reference
being a substantial relation to ‘reality-as-it-in-itself’. Quine is quite consistent in maintaining that what is real must be understood as a mere reflection of the referential terms
of the relevant discourses (see, e.g., W.V.O. Quine, “Reply to Dagfinn Føllesdal,” in The
Philosophy of W.V. Quine, eds. Lewis Hahn and Paul Schlipp (La Salle, IL: Open Court,
1986), 115. For further discussion, see Antti Keskinen, “Quine’s Conception of Objects
as Theoretical: Beyond Realism and Anti-realism,” in Realism, Science and Pragmatism,
ed. Kenneth Westphal (London: Routledge, 2014).
44 Of course, one might argue that such a construal of anti-representationalism is in fact
untenable. Thomasson would appear to come close to doing this insofar as she sees the
analytic-synthetic distinction as integral to her own neo-Carnapian program (Thomasson, Ontology, ch. 7; Amie Thomasson, Ordinary Objects (Oxford: Oxford University
Press, 2007), ch. 2). Horwich also upholds the analytic-synthetic divide as part of his
use-based theory of meaning (see e.g. Meaning, 60). To this I would respond, firstly,
that it is not clear that the distinction is really so integral to Thomasson’s project as she
seems to think, insofar as, by her owns lights, application conditions for terms needn’t
involve statable rules of inference (cf. Ontology, 92); and secondly, with respect to
Horwich, that I am not as convinced as he is that AR is well served by a positive theory
of meaning, even one based on use (see also Price, “What Should a Deflationist”).

300

Jonathan Knowles

one can do the kind of work that Price thinks GE does, reflecting on the
different natural functions our discourses have. But she will also point
out that roughly the same kind of naturalistic materials that GE helps
itself to—remember its commitment to subject naturalism!—can also
serve as the basis for a metaphysical placement project. In other words:
Since both my Quinean and Price commit to naturalism of at least a
somewhat substantive variety, and since both reject an absolute analyticsynthetic divide, it is just unclear what might be irrational about an antirepresentationalist form of ON that would not be irrational about GE.
There is more to say about this last point, especially in relation to
Price’s conception of what naturalism means for him and what GE aspires
to, but for reasons of space we must leave the issue there for now.45 Our
conclusion is that there is (modulo objections rebutted elsewhere) no
reason to think that a form of ON based on an anti-representationalist
framework is either an incoherent or irrational project, given the kind of
naturalism Price himself assumes.

4. Horwich’s Wittgensteinean Attack on T-Philosophy
I have argued, on the one hand, contra Horwich, that AR naturally
coheres with a neo-Carnapian picture of our linguistic practices that supports a default ontological pluralism, but on the other, contra Price, that
one could on the basis of such a picture both coherently and rationally
seek to metaphysically ‘critique’ these different ontological commitments
by valorizing a particular vocabulary, such as science (at least given that
vocabulary is valorized anyway). Insofar, AR-plus-metaphysics remains a
coherent and indeed live combination, and Horwich’s rejection of Price’s
argument is justified, though not exactly for the reasons he gives.
Horwich nevertheless thinks there are good arguments against ON and
that it is somehow irrational, and similarly for other reductive though
non-naturalistic programs. Here he draws on his Wittgenstein-inspired
critique of what he calls T-—for traditional—philosophy.46 Applied to
ON, this critique amounts to claiming that it illicitly treats areas like
the ethical and other common sense discourses as fit targets for a kind
of scientific deep explanation that we find in the sciences, including
45 For a fuller presentation of the critique of Price proffered in this section, see Jonathan
Knowles, “Global Expressivism and the Flight From Metaphysics,” Synthese (2016),
doi:10.1007/s11229-016-1166-1, accessed February 26, 2017. The details are of
interest though their role is essentially to rebut possible objections to my arguments,
especially the last. One thing I should emphasize here is that this does not depend on
whether one’s naturalism is precisely a form of physicalism, like Quine’s, or a somewhat
more liberal form that allows other natural scientific though possibly non-physical elements into its explanatory base, as Price’s plausibly is. The position in question does not
have to be Quine, but merely, in the senses discussed above, Quinean.
46 Horwich, Wittgenstein’s Metaphilosophy.

Metaphysics

301

biology and psychology. In particular, it involves—by scientific standards themselves—an illegitimate abstraction from central features of
the data, such as the oft-noted ‘weirdness’ of ethical properties, as well
as an equally illegitimate smoothing out of apparent counterexamples in
the dataset by appeal to an ideal of simplicity.47 For Horwich, the only
justification of this kind of treatment would be an antecedent faith in
an underlying ethical ‘deep reality’, corresponding to the kind of deeper
realities uncovered by science; but he thinks it is evident, on reflection,
that our ethical concepts are simply too unruly and too idiosyncratic to
expect that kind of treatment to apply to them.48
As suggested, Horwich’s Wittgensteinean critique of T-philosophy
extends beyond ON, for he also sees it as applicable to theorizing about
topics like metaphysical necessity or aesthetic value that it is not clear
could be seen as falling under the kind of ‘placement’ metaphysics ON is
concerned with. For present purposes I want to put these latter kinds of
cases aside.49 For now, the question is how it relates to the issues we have
been looking at here in relation to ON and AR. Does it, in particular, plug
a hole in Price’s argument against ON?
It is very unclear to me that the critique of ON Horwich gives advances
us very far beyond issues we have already canvassed. A natural reaction
to Horwich’s argument against ethical naturalism would be to see the sui
generis nature of our ethical concepts it points up as reflecting their place
in something like a Carnapian framework, ipso facto ensuring a reference
for them independently of any consideration of their ‘place in the natural world’. However, this would only hold by default, and thereby the
argument would again be vulnerable to the kind of Quinean revisionist
metaphysical gambit outlined in the previous section (modulo the provisos concerning the analytic-synthetic distinction; see note 44). I should
emphasize that I am not decrying the kind of Wittgensteinean therapeutic
reflection on the diversity our linguistic practices that Horwich points
up; indeed, it would seem to cohere with and support the neo-Carnapian
pluralistic picture. However, as far as at least Horwich’s Wittgenstein is
concerned, it is unclear it has any further significance; moreover, as we
have seen, this is not sufficient in itself to render placement metaphysics
incoherent.
We saw in the previous section that a possible response to this would
be to accuse ON of irrationality. There I argued that this was at least
not available to Price, who himself is a naturalist, albeit of a ‘subject’
(not ‘object’) variety. But could not Horwich disavow any antecedent
47 Horwich, “Naturalism,” 126.
48 Ibid.
49 For a more general critique of Horwich’s book on Wittgenstein’s metaphilosophy, see
Timothy Williamson, Review of Wittgenstein’s Metaphilosophy by Paul Horwich,
European Journal of Philosophy, 21 (2013).

302

Jonathan Knowles

commitment to naturalism (of a substantive variety at least), insisting indeed that naturalism itself represents an irrational philosophical
theory?50 If that is the case then both Price’s and Quine’s positive projects
would have to be rejected—but at least, a fortiori, we would also have
ruled out ON.
Exactly where Horwich stands on the issue of naturalism is a difficult
issue to divine. On the one hand, he explicitly disavows naturalism in
connection with domains like ethics and arithmetic. On the other hand,
he defends a non-normativist form of the use theory of meaning, motivated in part at least by a wish to avoid an unreduced appeal to normativity (see note 1)—something that would seem to presuppose some kind
of naturalistic commitment. So it is not clear his disavowal of naturalism
is very self-consistent.
Even putting this (admittedly somewhat ad hominen) point to one side,
however, I think it can be reasonably questioned whether some form of
naturalism, of a substantive kind, can really be seen as arising primarily
or at least exclusively from irrational ‘T-theoretic’ philosophical thinking.
As Price has recently argued,51 whilst renouncing the stringent ontological ideals of ON arguably has a lot going for it, we surely don’t want on
the other hand to land up in a kind of blanket quietism (or ‘soggy pluralism’, as Blackburn puts it52) which discerns no significant differences
between our practices and disallows all external perspectives on them.
Yet surely this is in a broad sense just what a naturalist philosopher or
scientist of the human seeks to provide us with (that is, in a way, one of
science’s roles). This desire to avoid a soggy pluralism and to admit the
perspective of science into our understanding of our various different
practices strikes me as laudable; moreover, it can, as I hope to show in
the final section, also help us finally to understand why placement-style
metaphysical questions might legitimately be demurred at. For now, however, the point is simply that naturalism is not, at least at all obviously,
the product of dubious philosophical theorizing, but rather represents
the legitimate aim of getting some kind of third-person leverage on ourselves and our activities. And, if that is so, I cannot see that Horwich’s
Wittgenstein-inspired points about the sui generis nature of (say) ethical
and arithmetical discourse suggest a reassessment of our conclusion from
the previous section: that ON is still on the table, even once we reject
representationalism.

50 He has indeed done this in discussion when I put the objection of the previous paragraph to him.
51 Huw Price, “Idling and Sidling Toward Philosophical Peace,” in Meaning Without Representation: Essays on Truth, Expression, Normativity, and Naturalism, eds. Steven
Gross et al. (Oxford: Oxford University Press, 2015).
52 Simon Blackburn, “Pragmatism: All or Some?” in Price, Expressivism, 113.

Metaphysics

303

5. How to Be an Anti-Representationalist
Anti-Metaphysicalist Naturalist
For someone suspicious of the traditional placement problems, rejecting
representationalism is certainly a step in the right direction. However, I
have been arguing that the form this insight takes in the overall views
of Price and Horwich still fails to give us the desired insulation from
metaphysics.
In my view, the way forward here is to look again at what a science of
humans—a subject naturalistic enquiry—might tell us about ourselves
and the capacities underlying our different vocabularies. On something
like the Canberra planners’ view of things, such enquiry would ideally
vindicate representationalism; while on Price’s ‘Sydney plan’,53 i.e. GE,
we would theorise different vocabularies in a ‘non-representationalist
key’ though employing the notion of ‘e-representation’ to mark a divide
between those vocabularies that do and those that don’t ‘track’ some
environmental feature. If we reject representationalism but also GE, what
alternatives for subject naturalism remain?
My suggestion is based on ideas I find in the work of Noam Chomsky,
though I develop them in ways that go well beyond anything Chomsky has
said (and might diverge from some of his other claims). There is some
inherent interest in developing an anti-representationalist philosophy on
the basis of Chomsky’s work in view of the in some sense clearly antirepresentationalist approach he adopts to language, i.e. one that rejects
the assumption of ‘[quoting Scott Soames] “the central semantic fact about
language [. . .] that it is used to represent the world”.54 Here I cannot go
into Chomsky’s ‘internalist’ views of meaning, but will develop instead an
anti-metaphysical form of AR that assumes the neo-Carnapian picture of
AR we have already sketched and builds on to these ideas of Chomsky’s
about the nature of science and naturalistic enquiry into the mind.55
The first of these is what I call radical anti-reductionism about science
(or RAR for short), a view which is meant to be more faithful to what
scientists actually do and say than the picture that philosophers typically
operate with.56 According to RAR, reduction is not simply a mere ideal
that much science in fact fails to live up to; rather, it is incoherent, for

53 The nomenclature is due to Jenann Ismael, “Naturalism on the Sydney Plan,” in Philosophical Methodology: The Armchair or the Laboratory? ed. Matthew Haug (London:
Routledge, 2015).
54 Noam Chomsky, New Horizons in the Study of Language (Cambridge: Cambridge University Press, 2000), 132. This quote is also used approvingly by Price; see Huw Price,
“Expressivism for two voices,” in Pragmatism, Science and Naturalism, eds. Jonathan
Knowles and Henrik Rydenfelt (Frankfurt: Peter Lang), 103, fn.
55 See especially Chomsky, New Horizons, essays 4 and 5.
56 See also John Dupré, The Disorder of Things: Metaphysical Foundations of the Disunity
of Science (Cambridge, MA: Harvard University Press, 1993), and James Ladyman and

304

Jonathan Knowles

there is no basic scientific level to which all others might be reduced.
Though unification is a genuine goal of science, it is only one of many,
and, in any case, standardly involves adjustments at both levels to be
unified.57 Furthermore, and relatedly, common-sense categories are not
eliminated or vindicated through scientific discoveries, but are simply left
behind as part of science—a view John Collins, again drawing on Chomsky’s views, has called metascientific eliminativism.58 Clearly such a view
presupposes that the representationalist idea on which only some of our
talk latches onto the ‘real’ is defective—so the neo-Carnapian framework
I have been developing remains central to our overall picture. What RAR
adds is that science itself does not yield a picture of a Quinean ‘desert
landscape’ or even something similar that one might use for either a metaphysical placement project or the kind of expressivist explanations that
Price proposes. Scientific theories concern themselves as such only with
the proprietary ‘level’ they operate at; they are not input to a metaphysical or indeed any other distinctively philosophical program of any kind.
Of course, one might as a philosopher seek to argue that these theories
fruitfully can be used in this way. But it should give philosophers, at least
those who call themselves naturalists, pause for thought that science itself
is not concerned with doing this. (We will find further reason to reject the
‘desert landscape’ view of what science provides below.)
Leaving things there, however, would precisely not do justice to what
scientific approaches presumably can tell us about our various different
discourses—it would in effect land us in the soggy pluralism a reasonable resistance to which we saw above was sufficient to motivate some
kind of naturalism and hence gave the lie to Horwich’s argument against
metaphysics.59 But how now is naturalism to be conceived, if not as
thinking launched from the perspective of desert landscapes or something similar? The idea here would be that there is nothing in RAR that
precludes the idea of subject naturalism, or, to use a Quinean term, a
naturalized epistemology, that is, a project that considers our “scientific
imaginings as activities within the world we imagine”60—though for us,
seen as a psychological enquiry into the mental and/or neural substrates

57
58
59

60

Don Ross et al., Everything Must Go: Metaphysics Naturalised (Oxford: Oxford University Press, 2007) for similar expressions of this kind of view.
Chomsky, New Horizons, 82.
John Collins, “Metascientific Eliminativism: A Reconsideration of Chomsky’s Review
of Skinner’s Verbal Behavoiur,” British Journal for the Philosophy of Science 58 (2007).
As with the argument against Price, I stress that what I reject in Horwich is not so
much his end-point but how he reaches it. I agree with him that metaphysical placement projects are irrational, but we need to find out what naturalism can be such that
metaphysics does not threaten as a consequence of it of before we can help ourselves to
that conclusion.
Quine, Word and Object, 5.

Metaphysics

305

of, not just science, but our various different domains of discourse.61
Nothing in RAR, in other words, precludes scientifically turning our
gaze upon ourselves and exploring the underlying bases of our various
discourses—both genealogically, and from a more cognitive science perspective, examining the sub-personal underpinnings of our thought and
experience. This is, of course, very similar to what Price’s GE seeks to do.
But on the kind of picture I am proposing, these studies have a different significance. On my view, what anthropological and/or psychological
studies can suggest is how, say, moral discourse in being underpinned by
motivational states of particular kinds is unlikely to become a systematic, scientific discipline—by contrast with other vocabularies which have
no such motivational underpinnings but are, as far as we can see, more
purely ‘conceptual’. Although such study is not ontologically significant,
and also has no relation to any notion like ‘e-representation’, it stands
to enrich, explain and possibly partially correct our common sense conceptions of what possibilities for systematic theorizing exist. Chomsky
has himself suggested a ‘science-forming faculty’62 as one of the mind’s
modules, and, in principle, this might be part of what we uncover as the
explanation for our ability to do physics and not, say, systematic ethics.
But my claims here are independent of there being such a faculty in any
substantive sense;63 I use it here just to illustrate the kind of idea that a
subject/epistemologically naturalistic science might make without adopting a ‘desert landscape’ conception of the tools available to it.
I should also underline that this view is in no way anti-realistic about
things like ethical values; rather, it suggests they figure in something like
(what others have called) the lifeworld: our everyday world of ordinary
objects and their properties that we interact with, but which also depend
constitutively for what they are on the possibilities we humans have for
action. In invoking this notion of a lifeworld, we also commit to the
idea that certain other properties, centrally those of the most theoretical reaches of science, do not belong in it, and would thus perforce be
referred to some other ‘world in itself’ (without any implication that this
is an ultimate ‘reality’). The legitimacy of a lifeworld is in my view in
turn mandated by a proper scientific understanding of the nature of conscious experience—as something inherently world-involving and not ‘in
the head’.

61 James Higginbotham sees Chomsky’s mentalistic psychology as a natural heir to
Quine’s program of naturalized epistemology; see his “The Place of Natural Language,”
in On Quine: New Essays, eds. Paolo Leonardi and Marco Santambrogio (Cambridge:
Cambridge University Press, 1993). It is in this piece I found the Quine quote.
62 Chomsky, New Horizons, 82–83.
63 For discussion, see John Collins, “On the Very Idea of a Science Forming Faculty,”
Dialectica 56 (2002).

306

Jonathan Knowles

These reflections clearly only point up ideas that would have to be
explored in greater detail to be vindicated. Nevertheless, if we accept the
outline they provide together with AR and RAR, we can, I think, more
clearly see the form of a thoroughly non-metaphysical picture of the relation between things like morality and science that nevertheless allows
science to have something to say about the respective forms of natural
understanding these categories involve—both through genealogy and
sub-personal psychology, and also through the systematic description of
the conscious, lived world.64

References
Blackburn, Simon. “Pragmatism: All or Some?” In Expressivism, Pragmatism
and Representationalism, edited by Huw Price, 67–84. Cambridge: Cambridge
University Press, 2013.
Button, Tim. The Limits of Realism. Oxford: Oxford University Press, 2013.
Carnap, Rudolf. “Semantics, Empiricism and Ontology.” Revue Internationale de
Philosophie 4 (1950): 20–40.
Chomsky, Noam. New Horizons in the Study of Language and Mind. Cambridge:
Cambridge University Press, 2000.
Collins, John. “Metascientific Eliminativism: A Reconsideration of Chomsky’s
Review of Skinner’s Verbal Behavoiur.” British Journal for the Philosophy of
Science 58 (2007): 625–58.
———. “On the Very Idea of a Science Forming Faculty.” Dialectica 56 (2002):
125–51.
Dupré, John. The Disorder of Things: Metaphysical Foundations of the Disunity
of Science. Cambridge, MA: Harvard University Press, 1993.
Higginbotham, James. “The Place of Natural Language.” In On Quine: New
Essays, edited by Paolo Leonardi and Marco Santambrogio, 113–139. Cambridge: Cambridge University Press, 1993.
Hirsch, Eli. Quantifier Variance and Realism. Oxford: Oxford University Press,
2011.
Horwich, Paul. From a Deflationary Point of View. Oxford: Oxford University
Press, 2004.
———. Meaning. Oxford: Oxford University Press, 1998.
———. “Naturalism, Deflationism and the Relative Priority of Language and
Metaphysics.” In Expressivism, Pragmatism and Representationalism, edited
by Huw Price, 112–27. Cambridge: Cambridge University Press, 2013.
———. “The Value of Truth.” Noûs 40 (2006): 347–60.
———. Wittgenstein’s Metaphilosophy. Oxford: Oxford University Press, 2012.

64 I would like to thank the audience at the Bergen conference where this material was
originally presented for their questions and feedback. Thanks also to Huw Price and
Paul Horwich for discussion of these issues, and to Kevin Cahill for his useful comments
and suggestions on a draft version of the paper.

Metaphysics

307

Ismael, Jenann. “Naturalism on the Sydney Plan.” In Philosophical Methodology:
The Armchair or the Laboratory? edited by Matthew Haug, 87–104. London:
Routledge, 2015.
Keskinen, Antti. “Quine’s Conception of Objects as Theoretical: Beyond Realism and Anti-realism.” In Realism, Science and Pragmatism, edited by Kenneth
Westphal, 93–114. London: Routledge, 2014.
Knowles, Jonathan. “Global Expressivism and the Flight From Metaphysics.”
Synthese. Accessed February 26, 2017. doi:10.1007/s11229-016-1166-1.
———. “Naturalism Without Metaphysics.” In Realism, Science and Pragmatism,
edited by Kenneth Westphal, 200–18. London: Routledge, 2014.
Ladyman, James, and Don Ross et al. Everything Must Go: Metaphysics Naturalised. Oxford: Oxford University Press, 2007.
Macarthur, David. “Subject Naturalism, Scientism and the Problem of Linguistic
Meaning: Critical Remarks on Price’s ‘Naturalism Without Representationalism’.” Analasis: Revista de invetagacion filosofica 1 (2014): 69–85.
Price, Huw. “Expressivism for Two Voices.” In Pragmatism, Science and Naturalism, edited by Jonathan Knowles and Henrik Rydenfelt, 87–113. Frankfurt:
Peter Lang, 2011.
———. Expressivism, Pragmatism and Representationalism. Cambridge: Cambridge University Press, 2013.
———. “Idling and Sidling Toward Philosophical Peace.” In Meaning Without
Representation: Essays on Truth, Expression, Normativity, and Naturalism,
edited by Steven Gross, Nicholas Tebben, and Michael Williams, 307–30.
Oxford: Oxford University Press, 2015.
———. “Immodesty Without Mirrors—Making Sense of Wittgenstein’s Linguistic Pluralism.” Reprinted in his Naturalism Without Mirrors, 200–27. Oxford:
Oxford University Press, 2011.
———. “Metaphysical Pluralism.” Reprinted in his Naturalism Without Mirrors,
34–53. Oxford: Oxford University Press, 2011.
———. “Metaphysics After Carnap: The Ghost Who Walks?” Reprinted in his
Naturalism Without Mirrors, 280–303. Oxford: Oxford University Press, 2011.
———. Naturalism Without Mirrors. Oxford: Oxford University Press, 2011.
———. “Naturalism Without Representationalism.” Reprinted in his Naturalism
Without Mirrors, 184–99. Oxford: Oxford University Press, 2011.
———. “The Semantic Foundations of Metaphysics.” Reprinted in his Naturalism Without Mirrors, 253–79. Oxford: Oxford University Press, 2011.
———. “Truth as Convenient Friction.” Reprinted in his Naturalism Without
Mirrors, 163–83. Oxford: Oxford University Press, 2011.
———. “What Should a Deflationist About Truth Say About Meaning?” In
Truth (Philosophical Issues, Vol. 8), edited by Enrique Villanueva, 107–15.
Ascatadero, CA: Ridgeview, 1997.
Price, Huw, and Peter Menzies. “Is Semantics in the Plan?” In Conceptual Analysis
and Philosophical Naturalism, edited by David Braddon-Mitchell and Roberta
Nola, 183–200. Cambridge, MA: MIT Press, 2009.
Putnam, Hilary. Realism and Reason. Cambridge: Cambridge University Press,
1983.
Quine, W.V.O. Philosophy of Logic. Upper Saddle River, NJ: Prentice Hall.
———. “Reply to Dagfinn Føllesdal.” In The Philosophy of W.V. Quine, edited
by Lewis Hahn and Paul Schlipp, 114–115. La Salle, IL: Open Court, 1986.

308

Jonathan Knowles

———. “Two Dogmas of Empiricism.” In his From a Logical Point of View,
20–46. Cambridge, MA: Harvard University Press, 1953.
———. Word and Object. Cambridge, MA: MIT Press, 1960.
Shapiro, Lionel. “Linguistic Function and Content: Reflections on Price’s Pragmatism.” Philosophical Quarterly 64 (2014): 497–506.
Thomasson, Amie. “Deflationism in Semantics and Metaphysics.” In Metasemantics: New Essays on the Foundations of Meaning, edited by Alexis Burgess and
Brett Sherman, 185–213. Oxford: Oxford University Press, 2014.
———. Ontology Made Easy. Oxford: Oxford University Press, 2015.
———. Ordinary Objects. Oxford: Oxford University Press, 2007.
Williamson, Timothy. Review of Wittgenstein’s Metaphilosophy by Paul Horwich. European Journal of Philosophy 21 (2013): e7–10.
Wittgenstein, Ludwig. Philosophical Investigations, 3rd ed. Translated by G. E.
M. Anscombe. Oxford: Basil Blackwell, 1968.

14 Do Pragmatic Naturalists Have
Souls? Should Anyone Be Paid
to Worry About It?
Bjørn Torgrim Ramberg

1
Let us say a soul is the psychic identity that something has. So worrying
about souls—whether there are any, what their distinguishing properties
are, how they come to be and cease to be, etc.—that is something that the
philosophy of mind does. However, philosophy of mind also indisputably
has a soul; the question of the place of mind (any mind, not just “ours”) in
the natural world. Its soul is a conflict, a problem of fit; the mental must
be, yet cannot be, a part of the material world: Subjectivity, intentionality,
phenomenal experience, agency, these are patently real, yet they appear
to be crowded out by the kinds of objects and properties that natural
science recognizes and by the explanatory scope that natural scientific
understanding has. So there is work be done to mark out or make up
appropriate space. Complicated families of positions shape the intellectual topography, but it has an organized structure, a unity. It is, at least in
rough outline, a familiar enough landscape to most philosophers, drawn
up in standard courses and introductory texts. This landscape is not static,
though, and such texts need to be updated every few years, if they are to
maintain their status as—predominantly—reporting textbooks. In time,
presentations of the landscape slide into the category of historical artifact,
partisan expressions of a historically conditioned and now transcended
perspective. So the enterprise goes on, and it is the problem of the soul that
shapes and directs it, and thus constitutes its evolving soul.
Pragmatic naturalism may be taken as a certain kind of response to this
enterprise. The first question of the title, however, suggests that there is
something peculiar about that response—that a pragmatic expression of
naturalism with regard to the problem of the soul lacks the kind of unity
and drive that this problem gives to—or at least until relatively recently
has provided for—the discipline we think of as the philosophy of mind.
The second of the two title questions is meant to indicate a set of issues
that are important but hard to make tractable. What sustains the idea
of philosophy as a subject and a discipline? What supports a claim that
philosophy be recognized and funded as a worthy intellectual project,
to be commended to bildung-aspiring young adults? What, in short, is

310

Bjørn Torgrim Ramberg

philosophy good for as a public endeavor? I take it that simply to pose
(and pursue) this question is not to answer it. There must be more.
What more? One might say that philosophy has been important; you
can’t understand much of the emergence of either modern politics or
modern science without some understanding of the history of philosophy. However, while it is true that a mandate of philosophy departments
is to interpret and convey past philosophy, if this exhausted the remit
of the discipline, then something clearly must, at some point or other,
have changed dramatically. Whatever role philosophy had in bringing
about modern scientific and political thinking, it did not serve that role
solely by reflecting on its own history. So while critical meta-reflection
and thoughtful self-narrative arguably are intrinsic to philosophy, philosophers must also be doing something else. What else?
The possibility to be considered is that pragmatic naturalists get themselves into a special kind of hole with regard to this second question. What
pragmatic naturalists share with most famous, canon-forming philosophers over the centuries, is a sense that there is something wrong with the
approaches they find on offer wherever and whenever it is that they grow
up to be philosophers. However, where other discontents have made their
mark by showing us how to get it right, and, thus constituting the history
of philosophy, pragmatic naturalists encounter challenges at just this point.
It is these challenges and the responses they draw that I want to consider.
One way to engage in meta-philosophical critique is to set out to get
clearer on what philosophy really is, what its proper ends are. Such selfdescription typically leads to explicit critique—even radical critique, which
contests fundamental aspects of the self-understanding with which its target
operates. Still, it may be internal critique in so far as it aims for improvement of philosophy in light of its own—now properly understood—ends;
it is a correction, rather than a displacement. However, for someone suspicious of the underlying idea—that philosophy has proper ends—this avenue
appears blocked. Internal critique is not going to be radical, because from
the internal standpoint no foil is available against which the practice as it is
found can be chastisingly held. The clarified real or proper ends of philosophy were supposed to play that part. Without this idea in play, the radical
impulse, the urge to be sweeping in one’s critical diagnosis, forces one into
an external stance—in effect, a stance against philosophy—in which one
makes claims about, for instance, the uselessness or worthlessness or selfdeluded nature of the enterprise that one is describing. The difficulty now,
though, is in convincing those engaged in the practice that they should listen
and take you seriously. On what grounds can you now claim their attention
as philosophers?
This problem—something like a dilemma—is faced by what I have
been calling pragmatic naturalism, at least in so far as its proponents
engage in meta-philosophical critique. Let us call the stance in question
PCPN: Philosophy Critical Pragmatic Naturalism. Not all pragmatic

Souls

311

naturalists fall under PCPN.1 But self-styled pragmatists, in the tradition
of classical pragmatism, tend to do so. This means, I will suggest, that
they must grapple with a tension between the urge to be sweeping and
the wish to be relevant. The meta-philosophically most interesting exponents of PCPN are those alert to the challenges of offering radical diagnoses without sacrificing what we may call practice-directed authority.
Guided by the challenges posed by this particular tension, I will proceed
by comparing elements of the meta-philosophical critiques offered by
three self-declared pragmatic naturalists; Richard Rorty, Huw Price, and
Philip Kitcher, all of whom have made it their business to offer sweeping,
reform-oriented critiques of philosophy.

2
In his recent Preludes to Pragmatism, Kitcher aligns explicitly with PCPN.
With approval, he says of James and Dewey:
they are out to focus philosophy on issues that matter to people.
Both are suspicious of the idea of timeless philosophical problems,
demanding to be tackled in each generation; both suppose that the
deepest philosophical challenges of an age depend on the previous
evolution of human life and culture.2
Kitcher has been “increasingly moved by this reformist approach to philosophy” and presents his essays in the volume as “investigations in the
spirit of the would-be pragmatist revolution”. Kitcher hopes “to renew
the James—Dewey project for our own times”.3
Kitcher’s stance is undeniably radical:
Pragmatism should not be domesticated and brought into the precincts of ‘normal philosophy’, so that James and Dewey can join the
pantheon of respectable philosophers. To paraphrase Marx, the point
is no to continue philosophy-as-usual, but to change it.4
1 Daniel Dennett is a case in point, as fellow pragmatist Richard Rorty notes:
“I enjoy metaphilosophy in a way that Dennett seems not to enjoy it”. Richard Rorty,
“Dennett on Intrinsicality,” in Truth and Progress, Philosophical Papers, vol. 3, 98–121
(Cambridge: Cambridge University Press, 1998), 119. Indeed, for Rorty, critical metaphilosophy, reform of the practice, is where the game is at:
I think, in short, that Dennett’s ‘urbane verificationism’ is a bit too urbane. It stops
short of the goal out of what seems to me misplaced courtesy to a half-defeated enemy.
(Rorty, “Dennett on Intrinsicality,” 119)
2 Philip Kitcher, Preludes to Pragmatism: Toward a Reconstruction of Philosophy (New
York: Oxford University Press, 2012), xiii.
3 Kitcher, Preludes, xiii.
4 Kitcher, Preludes, xiv.

312

Bjørn Torgrim Ramberg

Pragmatists like Robert Brandom and Hilary Putnam are too conservative for Kitcher, in so far as they “are inclined to find a closer connection
between pragmatism and central themes in “analytic” philosophy than
[Kitcher] would favor”.5 Kitcher wants to end the separateness of pragmatism not by modifying its vocabulary, but by changing philosophy in
line with the vision of Dewey. As he says, this is, “the revisionary hope
I take to lie at the heart of pragmatic naturalism”.6 What’s wrong with
philosophy? It is spending too much of itself on problems that are of
secondary, or of little, or of no importance, of no real value to the larger
community that supports the activity. Too much is scholastic, in the pejorative sense, different only in degree, Kitcher says, from some hypothetical group of lucky researchers “who decide, solely for reasons of personal
satisfaction, to spend their days counting the dust motes or musing on the
shifting patterns of the shadows on the floor”.7
So, Kitcher urges: “To revive pragmatism today is, I suggest, not to
invoke James and Dewey as allies in current debates, but to recognize
that our own scholastic conception of philosophy cries out for just the
reform they wanted”.8
In his radical stance, Kitcher recognizes Richard Rorty as close kin, with
one major proviso; where Rorty, Kitcher thinks, concludes that philosophy
has run its course, Kitcher, with Dewey, seeks to liberate philosophy, regarding it “as growing out of an impulse that is central to human nature”.9 Where
Rorty draws his get-over-philosophy consequences of pragmatism, Kitcher
sounds his optimistic preludes to a renewed pragmatist philosophy.10
Kitcher is right to see his project as more closely aligned with Rorty’s
than with the other new (or neo-) pragmatists he addresses. Rorty, too,
is concerned to hold on to the radical nature of the critique he traces to
James and Dewey. Here is Rorty, speaking, as its president, to the APA
Eastern Division, just as his magnum opus, Philosophy and the Mirror of
Nature (henceforth PMN), was coming off the press:
Peirce himself remained the most Kantian of thinkers—the most convinced other species of discourse could be assigned its proper place
and rank. It was just this Kantian assumption that there was such a
context, and that epistemology or semantics could discover it, against
which James and Dewey reacted. We need to focus on this reaction if
we are to recapture a proper sense of their importance.11
5
6
7
8
9
10
11

Kitcher, Preludes, xiii.
Kitcher, Preludes, xv.
Kitcher, Preludes, xiii.
Kitcher, Preludes, 192.
Kitcher, Preludes, 192.
Kitcher, Preludes, xvii.
Richard Rorty, “Pragmatism, Relativism, and Irrationalism,” in Consequences of Pragmatism, edited by Richard Rorty (Minneapolis: University of Minnesota Press, 1982), 162.

Souls

313

There is no super-context. That is the heart of Rorty’s appropriation
of his pragmatist predecessors. To show this, and show what it meant
for philosophy, was Rorty’s great project in PMN. And what it means,
Rorty there argues, is the end of epistemology and metaphysics as it has
been pursued since the early modern age—the end of representationalist
philosophy. We must stop this fruitless scholastic endeavor and turn to
something different.
While Rorty here may sound quite like Kitcher, the latter is concerned
to keep an important space between them. As Kitcher emphasizes, Rorty’s
radicalism terminates in an external stance. Kitcher, by contrast, thinks
that philosophy should not be given up, but radically changed.
A tempting reaction to this, however, is to say that this is merely a
difference of definition of terms. One suspects that is exactly how Rorty
would have responded; “look, Kitcher, we share the same heroes, the
same revisionary hope, the same impulse toward a liberating, humanizing change in the practice we call philosophy. We agree with Wittgenstein
and Dewey that philosophical problems are not to be solved, but gotten
over. We both think this could actually happen. Let us not quibble over
labels!”
We should grant this reaction the point that the contrast between Deweyan optimism and Rortyan pessimism about philosophy, a contrast that
Kitcher makes much of, is less straightforward than Kitcher suggests.
In fact, the distance between both the ends and the means of these two
champions of PCPN really is smaller than Kitcher’s juxtaposition suggests. Still, it is clear that there are revealing differences between the way
Rorty and Kitcher express the Deweyan revisionary hope for a humanized intellectual culture. These differences matter to the problem with
which we began; how to be a sweeping reformist critic while retaining
practice-directed authority. To see these differences play out, though, we
must first look a little more closely at the kind of view I have labeled
pragmatic naturalism.

3
There are two quite different things one might be pushing for in articulating pragmatism. There is a strong swell of contemporary pragmatists
(Robert Brandom, Cheryl Misak, and Robert Talisse are prominent representatives) who work to end the separateness of pragmatism by tailoring its vocabulary to fit the themes of mainstream epistemology and
philosophy of language.12 This mode, let us call it philosophical pragmatism, makes the notion of agency and intervention central, takes practice

12 For a recent work of historical narrative explicitly engaged in this project, see Cheryl
Misak, The American Pragmatists (New York: Oxford University Press, 2013).

314

Bjørn Torgrim Ramberg

to be the source of normativity, and expands the scope of means-ends
reasoning as far it will possibly go. Philosophical pragmatists develop
distinctive views on a wide range of recognizable themes.13 Pragmatism
in this key is a set of basic, integrated commitments, along the lines just
indicated, that undergird a series of theories on a series of topics, theories
that, one supposes as a pragmatist, will stand up to criticism better than
rival views do. Support will come at either end—from the plausibility of
the account of the phenomena, and from the force and plausibility of the
core commitments. Representatives of philosophical pragmatism of this
form have no trouble with their self-understanding as philosophers, or
with pragmatism as philosophical and as requiring philosophical work,
nor with the idea that their engagement in mainstream debate produces
thoroughly philosophy-internal criticism of rival positions. Pragmatism is
a distinctive family of positions, perhaps, but in a conventional, familiar
matrix. It isn’t, and ought not to be, in any way separate.
The other kind of pragmatism, by contrast, the kind to which Kitcher
and Rorty both belong, along with Price, is not a foundation for rival
theories of this or that. It is not a set of basic commitments of the sort
that gives rise to an integrated set of theoretical answers to familiar
philosophical questions in various domains, answers comprising distinctive theories that may compete with other theories in providing answers
to—more or less—the same questions.14 It is fundamentally and broadly
reformist. It is unhappy about philosophy—it thinks that philosophy is,
in some sense, in some way, barking up the wrong tree. Typically and
unsurprisingly, such pragmatists also take Wittgenstein’s assault on constructive philosophy as a central source of inspiration, and it would not be
misleading to emphasize the contrast with philosophical pragmatism by
calling this mode of pragmatist thinking Wittgensteinian pragmatism.15

13 A list of core topics might be: The nature of truth, knowledge and justification; the
nature, source, and possibility of objectivity; the nature of meaning or content; the
nature of value, particularly the nature and source of the force of moral values; the relation between facts and values, between the descriptive and the normative; the conditions
of the good life or human flourishing—or the flourishing of any creature capable of
some form of it; the nature of justice, of autonomy, of democracy; the nature of personal
identity and the relation between the self and its social and natural contexts of existence
and persistence.
14 I say “more or less” here, because almost all interesting theoretical rivalries in philosophy also concern—sometimes mostly concern—the nature of the question to be
answered.
15 In the case of Rorty and of Price, this assimilation needs no qualification, since both are
explicit about their debts to Wittgenstein and the Wittgensteinian elements of their attitude to philosophy. With Kitcher, the matter is more complicated. However, precisely
by virtue of what I will argue is distinctive about Kitcher’s meta-philosophical stance—
its predominantly ethical character—I believe there is a strong argument to be made
for the label also in his case. Of contemporary pragmatists, the one whose thought is
most explicitly a development of Wittgensteinian themes is David MacArthur (see for

Souls

315

One immediate challenge that Wittgensteinian pragmatism faces is to
delineate its object of dissatisfaction; that toward which a negative attitude is to be struck. A very common name for it among champions of
PCPN is ‘metaphysics’. Metaphysics, however, is hard to define. It may
appear as if the very act of delineating the area one wants to leave alone,
or get over, or be quiet about, is already to make too much noise. Certainly, this is a theme that has figured centrally in discussions about how
Wittgenstein ought to be taken, particularly with regard to sense and
nonsense.16 Here, though, we are at a point where champions of PCPN
show a distinctively pragmatic attitude; the vices they aim to expose are
not sins against sense or meaning as such, but projects built on more specific substantive assumptions. In Rorty’s case, these are summed up under
the label of representationalism, which amounts to the idea that systems
of linguistic or mental representation may be assessed for adequacy in
terms of their ability to capture the way the world really is.
Much of Rorty’s PMN is devoted to arguing that this idea leads us
nowhere interesting or useful, and certainly not to knowledge: Representationalist metaphysics, Rorty thinks, even in its Kantian form—where
the nature of the knowing subject, not the world in itself, is the target—is
delusional, in treating its own constitutive metaphors as magically obligatory. Certainly there is much to be said about knowing subjects, but
without the representationalist framework in place, the universality and
the necessity that such investigations may aspire to simply falls away.
Instead, we elaborate perspectives, and we evaluate concepts and vocabularies in a context-dependent historically shifting means-ends scheme,
where no pretense of transcendence or finality is present. We reflect on
science and on how to incorporate the insights of science in our lives. But
science, for Rorty, has no ontological priority—since nothing has. Science, morality, poetry, these are all evolved ways of coping. There is no
way the world is in itself. The only scale of measurement for vocabularies
of any sort is usefulness to agents’ purposes.
As is well known, Rorty’s case draws on arguments developed by central
figures in 20th-century analytic philosophy. He presents the case he makes
in PMN as Wittgensteinian in spirit, but in its detail, it is built largely from
elements provided by Quine and Sellars, filtered through some Davidsonian
arguments. By this route, the story developed in PMN arrives at a point
where Dewey, so Rorty claims, is already waiting. By this last claim Rorty

instance David Macarthur, “Wittgenstein and Expressivism,” in The Later Wittgenstein
on Language, edited by Daniel Whiting (London: Palgrave, 2010), 81–95).
16 Rorty’s summary of the issues that divide “pragmatic Wittgensteinians” from “resolute
Wittgensteinians” is offered in his “Wittgenstein and the Linguistic Turn.” (Richard
Rorty, “Wittgenstein and the Linguistic Turn,” in Philosophy as Cultural Politics, Philosophical Papers, Volume 4, edited by Richard Rorty (Cambridge: Cambridge University
Press, 2007), 160–175).

316

Bjørn Torgrim Ramberg

means, I take it, that if we accept his line of argument, we will see that the
sort of issues that representationalist assumptions lead us to confront will
dissipate once we follow those assumptions through to their conclusions.
So we are driven by philosophical argument to the conclusion that there
is no other source of philosophical problems than those we make up for
ourselves as we go—important enough, perhaps, many of them—for a time
and a place. And Dewey, Rorty suggests, offers just the right response to
that predicament, even if he didn’t have at his disposal the dialectical arsenal of PMN, and so could only adumbrate the arguments that get us there.
For Rorty, then, pragmatism is both a stance of philosophy and a
stance toward philosophy. Philosophical argument and reasoning lead to
the conclusion that epistemology and metaphysics—construed in representationalist terms—are pointless activities. We can still do philosophy
in a meaningful way, Rorty insists, but it should be edifying rather than
constructive, redescriptive rather than argumentative, and hermeneutic
rather than epistemological.
It is worth pausing briefly at these three imperatives of PCPN (Rortystyle). The edifying ambition in PMN is elaborated through a contrast
between two essentially different kinds of revolutionary philosophers;
there are those who offer new, better systems, and there are those whose
work reacts against systematic philosophy, aiming to undermine our faith
in constructive efforts. Rorty explicitly holds up Wittgenstein’s Philosophical Investigations as a work of the latter kind.17 Speaking of Wittgenstein, Heidegger, and Dewey, Rorty remarks:
Thus, their later work is therapeutic rather than constructive, edifying rather than systematic, designed to make the reader question his
own motives for philosophizing, rather than to supply him with a
new philosophical program.18
The point, for Rorty, in invoking Wittgenstein here is to emphasize that
philosophical activity may be valuable even when nothing is thereby constructed, in so far as reactive, edifying writers may aid us in breaking the
grip that certain vocabularies have on our thinking, by helping us see that
these vocabularies—representationalism in particular—are contingent elaborations of optional metaphors. The main strategy for such therapeutic philosophy is redescription. Insofar as it contrasts with argument, redescriptive

17 See Rorty, PMN, 368–372.
18 Rorty, PMN, 5–6. In PMN, Rorty more or less equates edifying philosophy with therapeutic philosophy. Twenty years later, however, responding to readers of Wittgenstein
who link the idea of philosophy as therapy with a notion of nonsense and of the deceptiveness of certain forms of language, it becomes important for Rorty to distinguish his
form of pragmatic naturalism from the therapeutic readers of Wittgenstein. In Rorty’s
terms, the problem with this approach is that it recommends a quietist stance based on
an essentialist picture of what philosophy is—albeit a negative picture.

Souls

317

activity is not aimed at direct rebuttal of some target view, but offers a presentation of the matter under discussion under some other aspect. Argument, in
this scheme, is about settling truth-values, while redescriptive innovation is
about proposing different truth-value candidates. Successful argument settles our beliefs, while successful redescription makes us care about different
things, or the same things in new ways. Hermeneutic philosophy, finally, is
what we have when we replace, as Rorty proposes, objectivity with solidarity as a guiding norm for intellectual conversation. Hermeneutic philosophy
is not informed by the goal of developing an overarching vocabulary in
which all rival claims or perspectives may be put, but rather with exploiting
incommensurabilities between perspectives or vocabularies or forms of life
so as to generate further and richer descriptive options and so to expand
the dialectical space in which we operate. We might say that hermeneutic
philosophy aims to expand discursive (pragmatic) reach, whereas epistemological philosophy enforces discursive (semantic) discipline.
These rough contrasts capture, I think, important concerns in Rorty’s
critique of philosophy. However, whatever one thinks of them, it is
pretty clear that these are not prescriptions that point in any particular
substantive direction. They generate no particular philosophical impetus; they indicate no road of inquiry. The deconstructive strategy that
Rorty deploys brings him to a point where all the steam seems to be let
out of the philosophical boiler. So while the process has been an internal
endeavor, the terminus of Rorty’s critique, as Kitcher notes, is external.
And there is no getting back inside—there seems to be nothing to get
back inside into. The ‘what more’ question posed initially—asking for a
working order of some sort—seems to have no answer in Rortyan terms.
Kitcher’s argumentative strategy, by contrast, is not deconstructive in
this way. He is closer to Dewey (and, perhaps, to be fair, to Rorty in later
years) in that he takes what is fundamentally a moral stance toward the
practice of philosophy—he does not deconstruct, but confront. Before
returning to Kitcher, though, it is necessary to say a little more about the
kind of naturalism that is operative in PCPN. And in this context, the variety of PCPN that Price has articulated will be particularly illuminating.
A consideration of naturalism will show both points of convergence and
points of contrast between Price and Rorty, which in turn will be of use
in our attempt to understand where PCPN leaves philosophy.
Both Rorty and Price habitually invoke Wittgenstein as they situate themselves meta-philosophically. A good place to start, then, might
be with the kind of naturalism that we might plausibly attribute to
Wittgenstein.

4
An excellent point of departure is Marie McGinn’s perspicuous summary
of features of Wittgenstein’s naturalism, which she takes to be “a fundamental and all-pervasive approach to philosophical perplexity”:

318

Bjørn Torgrim Ramberg
the importance of seeing things in context, of looking at particular
cases, of seeing connections, of looking at how something develops or
unfolds in time and of recognizing patterns; the rejection of explanation in favour of description; the use of analogies and comparisons;
the suspicion of abstractions, hypostatizations, and idealizations; the
avoidance of dogma; the appeal to the reader’s full sensuous awareness of phenomena and the attempt to make phenomena present to
the imagination; and finally, the consistent emphasis of doing over
knowing, on the application or employment of linguistic techniques
in everyday human activities and on the roots of our language-games
in primitive responses and reactions.19

As McGinn expounds Wittgenstein’s naturalism, it is easy to appreciate its
attraction for the kind of pragmatism I am pursuing here. Both the positive means and the objects of suspicion chime with the Rortyan emphasis on the therapeutic, the redescriptive, and the hermeneutic aspects of
philosophical practice. Moreover, McGinn’s list of features brings out the
integral connection in Wittgenstein’s philosophical practice between a
conception of how to do philosophy and a view of what its aims and commitments may be. And this is a useful way to frame the issue we are pursuing with regard to PCPN. Just at this point, however, complications arise.
In Wittgenstein’s case, the mandate of philosophical activity is, as McGinn
puts it, “to overcome the intellectual temptation [. . .] to idealize, create
abstractions, and hypostatize objects” in the reader that he engages and
thus to liberate the reader from philosophical perplexities.20 PCPN, however, differs from Wittgenstein in just this regard. PCPN does not want to
leave everything as it is, but, in the spirit of pragmatism, seeks to articulate a mandate for philosophy in the general project of promoting human
well-being. Philosophy, for PCPN, ought to be a positive force for change,
change to the better. Naturalistic therapy aimed at releasing us from philosophical perplexity cannot, from the point of view of PCPN, be the whole
story. It is not surprising to find, therefore, that there are other elements of
naturalism in PCPN than the Wittgensteinian therapeutic features.
Wittgensteinian naturalism contrasts with what we might call orthodox naturalism. Orthodox naturalists are the folks whose gut tells them
that science—the practice of systematic empirical inquiry—tells us what
the world, including its thinking things, is really like. This gloss on the
label is of course as far as we can get from a technical and precise definition. It is also of little use before the advent of modern physical science
makes the mind-body problem available. So while most lines of thought

19 Marie McGinn, “Wittgenstein’s Naturalism,” in Naturalism and Normativity, edited by
Mario De Caro and David Macarthur (New York: Columbia University Press, 2010), 347.
20 McGinn, “Wittgenstein’s Naturalism,” 347.

Souls

319

in this area have ancestry that goes back to antiquity, there is little point
in applying the label ‘naturalism’ before, say, the exchange between Descartes and Princess Elisabeth of Bohemia.21 Since then, though, orthodox
naturalism offers one reasonably tractable historiographical distinction
in philosophy concerned with the problem of the soul. And the formulation, casual though it is, does pay heed to both the epistemic and the
ontological dimensions of the commitment at issue, and also to its selfreflexive nature.
For the orthodox naturalist, the problem of the soul is just an instance—
a tricky and interesting instance, but just an instance, nonetheless—of a
general issue: How is it that any of the kinds of things not in the ontology
of science can be accommodated in our view of the world? Orthodox
naturalism faces a puzzle concerning the relation between physics and
everything that isn’t part of basic natural science. This is the purest version of what Price and other have called “the placement issue:”
If all reality is ultimately natural reality, how are we to ‘place’ moral
facts, mathematical facts, meaning facts and so on? [. . .] In cases
of this kind, we seem to be faced with a choice between forcing the
topic concerned into a category that for one reason or another seems
ill-shaped to contain it, or regarding it as at best second-rate—not a
genuine area of fact or knowledge.22
Orthodox naturalists typically approach their task as one of locating truth-makers of a suitably worldly and objective kind for the discourse under pressure. They want to show us what we are really talking
about when we use moral language, or wonder about the properties of
numbers—or, as in our case, wonder about the nature of our souls. The
stance is expressed in exemplary fashion by Jerry Fodor, when he says
that “if aboutness is real, it must be really something else”.23
Now we can state the tension that arises for champions of PCPN with
regard to naturalism in the following way. While their philosophy-critical

21 “And I admit that it would be easier for me to concede matter and extension to the mind
than it would be for me to concede the capacity to move a body and be moved by one to
an immaterial thing”. Princess Elizabeth to Descartes, May 1643. Quoted by Jaegwon
Kim (Jaegwon Kim, Philosophy of Mind, 3rd ed. (Boulder: Westview Press, 2010)),
from Daniel Garber (Daniel Garber, Descartes Embodied: Reading Cartesian Philosophy Through Cartesian Science (Cambridge: Cambridge University Press, 2000), 173).
22 Huw Price, “Naturalism Without Representationalism,” in Naturalism Without Mirrors, edited by Huw Price (New York: Oxford University Press, 2011), 74.
23 Jerry Fodor, Psychosemantics (Cambridge, MA: MIT Press, 1987), 97.
Other approaches, such as Davidson’s original anomalous monism, may be weaker
with respect to ontological commitments and softer on reduction, but nevertheless
remain ways respecting the orthodox form of the naturalist impulse, taken as a restriction on our catalogue of ontological items.

320

Bjørn Torgrim Ramberg

inclination toward Wittgensteinian naturalism suggests that both the
ontological and the epistemic dimensions of orthodox naturalism are
symptoms of uncured philosophical perplexity, their pragmatism nevertheless inclines them to feel, on the other hand, that there is something
right about what orthodox naturalism is trying to tell us about the scientific view of the world. Moreover, taking on board that lesson, in the
right form, and in the right way, would be an advance for human culture.
Bringing this cultural change about is exactly the sort of task that pragmatist philosophers should be engaging in.
In the paper that I have already quoted from, “Naturalism Without
Representationalism”, Price goes on to propose an alternative way of
respecting the naturalist gut instinct. Price writes:
Concerning naturalism itself, then, my argument is something like
this. To assess the prospects for philosophical naturalism, we need
a clear sense of the task of philosophy, in the areas in which science
might conceivably be relevant. Clarity about this matter reveals not
only that the approach commonly called naturalism is not the only
science-sensitive option for philosophy in these areas, but also that a
different approach is the preeminent approach [. . .]24
This different approach Price designates subject naturalism. In the spirit
of Hume, the subject naturalist takes as her starting point the notion
that we, thinking things, are natural creatures doing natural things, also
when we exercise the capacities of our souls—for instance, when we
pose and solve placement problems. We need not reconstruct the argument of Price’s paper here. The important point, for present purposes, is
that it leads Price, or rather allows Price, to treat orthodox naturalism
as a form of metaphysics, and metaphysics as something naturalistically
dubious. Price’s form of naturalism differs from orthodox naturalism
along just the demarcation line I aim to trace by speaking of pragmatic
naturalism, in so far as it self-consciously aligns itself against metaphysics. Price, then, is a pragmatic naturalist precisely because he does not
take the classical, ontological problem of the soul (material, yet not material) at face value. Indeed, he thinks the problem should be abandoned,
not solved. Hence Price, too, is a reformer. His philosophical arguments
diagnose struggles between realists and anti-realists as based on shared,
erroneous premises. In this respect, Price’s attack on metaphysics is
akin to Rorty’s. But where Rorty’s dialectical purpose is a transformative genealogical story meant to show that the foundation of the core
problems of philosophy is a set of optional, contingent metaphors, a
line of thought which pushes him exceedingly close to Wittgenstein’s

24 Price, “Naturalism Without Representationalism,” 185.

Souls

321

therapeutic conception of philosophical practice, Price takes himself to
be clarifying the constructive tasks of philosophy. Object naturalists—
the ontologists who worry about placement, have made a mistake on
their own terms (Price argues) by failing to see the priority of subject
naturalism. And once that is conceded, there is no obstacle other than
prejudice to a generalization of the expressivist (i.e., non-metaphysical
because non-representational) approach to all our assertoric discourses.
That means that the central task of solving the ontological problems that
representationalist semantics brings on in its obsession with the question
of what our various discourses are really about, is in truth a misguided
effort—it is based on an inadequately understood naturalism. Price sees
global expressivism as the dialectical outcome of a philosophical argument that in effect pragmatizes orthodox naturalism. He begins on what
appears to be, and is meant to be, common philosophical ground with
orthodox naturalists. And the upshot of the argument is that where philosophers in the past have taken themselves to be concerned with the
nature of things and our representations of them, what we should be
trying to illuminate is the purpose and function served by various different vocabularies for natural creatures like us. Thereby, “the expressivist
simply sidesteps the metaphysical conundrums that trouble her representationalist opponents, realists and anti-realists alike. (‘Those are not
my issues’, she tells them.)”.25

5
Price, then, unlike Rorty, offers a substantive, post-metaphysical replacement project. The replacement project that arises from Price’s deconstruction of orthodox naturalism requires some conceptual innovation, such
as Price’s own development of Sellarsian distinctions into what he calls
e-representations and i-representations.26 Here something is clearly left
for philosophers to do, at least in characterizing the program. But the
task is, as Price stresses, at heart an empirical project. Its goal is to understand naturalistically the use we engage in; the various things we do with
our different vocabularies, what functions they have and purposes they
serve, and also the various things that language does with us. Here, too,
the contrast with Wittgenstein is apposite. Price articulates a position that
appears Wittgensteinian in its rejection of representationalist metaphysics and its emphasis on the elucidation of practice, of what we do. However, in its systematic aim and its quasi-scientific ambition, it launches an
explanatory project—a project of knowledge—that implies a break with

25 Huw Price, Expressivism, Pragmatism and Representationalism (Cambridge: Cambridge University Press, 2013), 157.
26 For an elaboration of these notions, see Price, Expressivism, chapters 2 and 3.

322

Bjørn Torgrim Ramberg

Wittgenstein’s conception of philosophy as therapy, and his embodiment
of philosophy as a dialogical naturalizing practice.
Like Rorty’s philosophical arguments, Price’s lead to a position that
makes common projects of philosophical theory look very questionable
on their own terms. But Price’s development of subject naturalism (incidentally, heartily endorsed by Rorty) comes in the guise of an improvement of the understanding of the project to which orthodox naturalists
are already committed. Certainly, Price’s version of PCPN, if one accepts
it, makes a certain kind of metaphysical worry go away. And if one goes
with Price’s program, there is no doubt that reform is quite extensive—
ontology as the search for truth-makers may be found in many domains,
and if Price has his way its days would be over. But Price’s objection
to the metaphysics that he “sidesteps” is in one sense quite narrow. He
thinks it is based on a failure to see a philosophical point, in a sense of
“philosophical” that Price would expect to be entirely uncontroversial,
but that both Rorty and Wittgenstein might view with suspicion. The
sidestepping of metaphysics is argumentative and completely internal to
the basic project of naturalistic philosophy. The project goes on, though
now the inquiry to be undertaken is fundamentally empirical, and philosophy of language, Price-style, is continuous with the various sciences
of language, of communication, and of behavior generally. Still, though, it
seems that the project is not confined to these. For science to get its grip
on our linguistic practices, someone must be paying attention to what we
are doing and begin to describe it in a way that renders our discursive
practices available as objects of scientific investigation. But here again the
contrast with Wittgenstein is clear; Wittgenstein’s descriptive activity and
its appeal to forms of life are designed to free us from the temptations
to generate explanatory theory. Price, by contrast, expects from the right
kind of philosophical redescription of these practices that they emerge as
tractable objects of empirical science.
While some philosophers may think this Quinean transformation of
philosophy troubling, the grounds for the move are hardly external to the
project under scrutiny. In contrast to Rorty, Price provides a work order
for post-representationalist philosophers. And in contrast to Kitcher,
Price does not confront the practice of philosophy on ethical grounds.
Perhaps, then, it is fair to conclude that in the incarnation of PCPN
that Price articulates, the tension between radical critique and practicedirected authority is resolved in favor of the latter. In Price’s program, it
is the ambition of radicalness that gives way. One could imagine a fresh
young object-naturalism enthusiast one day coming up with an argument
that Price couldn’t reply to, thus bringing metaphysics back in more or
less traditional, ontological form.
Rorty’s genealogical arguments are of a different sort, and the persuasion they effect (if they do) is of a different order. Here we are at a point
where the motivation for Rorty’s invocation of Wittgenstein and the idea

Souls

323

of philosophy as a kind of edifying activity (against philosophical perplexity) is most apparent. For readers once convinced by PMN cannot
be brought back inside metaphysics by some particular and particularly
clever argument. The reason is that such readers have been given a template for reading that genre, a template that robs its instances of moving
power.27 The cost—if it is a cost—is that Rorty institutes no new soul for
naturalism. His polemics against and redescriptions of representationalist activity play a dialectical role, as he put it, serving, when they work, a
liberating capacity—at their best, they free philosophers’ soul from what
have become stifling self-conceptions. Rorty’s critique is radical, in that it
dislodges practitioners of philosophy from the metaphors that structure
the vocabularies in which problems of epistemology and ontology are
phrased. But then—what? What about the “What more?” question? Wittgenstein offers at least a kind of challenge—perpetual vigilance against
the temptations of language to lead us into philosophical perplexity. But
Rorty, refusing quietism and dismissing the idea of a general diagnosis of
philosophical perplexity, seems to leave us with no particular thing to do,
nothing to get on with. Rorty’s dialectical fate suggests that PCPN, in its
radical version, may not have an answer to the “What more?” question.
And if it does not, why should philosophers take it seriously?
Perhaps, though, the problem is not that PCPN, in its radical, Rortyan
form, provides no answer to what philosophy should be or do. Perhaps
the problem is the level of abstraction at which the question is posed. In
effect, this is the hypothesis that a core message of PCPN is that philosophers should stop expecting that a proper understanding of philosophy
will provide them with a work-order. This is the thought I should like to
pursue in the final section.

6
Let us briefly consider the philosophy of mind again. The field has altered
in recent decades.28 It has as much steam as ever, but its soul clearly
has changed. A great many current practitioners are aiming for a kind
of understanding that science gives us—to understand various forms of

27 This may be what Robert Brandom has in mind when he argues for the point that
as they are absorbed in pragmatist thought, historicism and naturalism are mutually
reinforcing positions. (Robert B. Brandom, “Vocabularies of Pragmatism: Synthesizing
Naturalism and Historicism,” in Rorty and His Critics, edited by Robert B. Brandom
(Oxford: Blackwell Publishers, 2000), 156–183).
28 Compare, for instance, the aggressively essentialist, armchair-clinging attitude of Colin
McGinn’s 1982 textbook (Colin McGinn, The Character of Mind: An Introduction
to the Philosophy of Mind, 2nd ed. (Oxford: Oxford University Press, 1997) (original
edition published 1982)) with the various approaches represented in John Hawthorne’s
state-of-the art 2007 collection (John Hawthorne, Philosophy of Mind: Philosophical
Perspectives, Volume 21 (Oxford: Blackwell Publishers, 2007)).

324

Bjørn Torgrim Ramberg

consciousness, perception, conceptual, and cognitive capacities as empirical phenomena. Such theorists could not care less about a priori necessary
truths about mental phenomena, but they do not feel their self-labeling as
philosophers the least bit threatened by that. Nor is their work confined
to conceptual clarification or analysis or theory. They try to understand
the mind, and science is one very important way into the phenomena they
are interested in. Their work also deals with the nature of causation, the
nature of explanation, the nature of properties, because in our efforts to
get a scientific grip on the soul, the phenomena are such that assumptions
about the scaffolding concepts of science become salient. But these problems are frequently treated instrumentally—in a pragmatic spirit, worth
dealing with only in so far as they help improve naturalistic understanding. In such approaches, many of the themes of metaphysical philosophy
of mind live on, but thoroughly instrumentalized, co-opted by a pragmatic attitude.
So have these workers heeded a call to reform? I cannot offer much
of an argument, but I suspect that would be an entirely unhelpful way
to explain these developments. Certainly, this change is not a result of
some general critical argument about the dubiousness of metaphysics. But
the change does, perhaps, attest to some difference in attitude toward the
discipline.
And here, maybe, radical critique á la PCPN can play a part—by opening up, by liberating, rather than by directing. Rorty’s radicalness—clearly
maintained, in his case, at the expense of practice-directed authority—
suggests that if PCPN has anything like a soul, it must be an ironic soul. Its
discipline lies in its abstentions. If it does any useful work at all in its metaphilosophical mode, then I suspect that this is in so far as it is absorbed
as an attitude, one that is affecting but not directing what people—
philosophers—care about, and what they stop caring about. Maybe this
is all that radical meta-philosophical critique in a naturalistic key can ever
hope to achieve. If radical critique were to be directive, it will undermine
itself, or merely look ridiculous. When Price’s critique does neither of these
things, it is perhaps because it is not so radical after all.
But perhaps Price and Rorty do not exhaust the options. I should like
to close with a final glance at Kitcher. I said above that Kitcher does not
deconstruct, he confronts: Too many philosophers are spending too much
time on problems that are not relevant to human well-being. Kitcher’s
response to this is not to construct an argument that will lead step-by
step from wasteful metaphysics to meaningful project. Nor is it to deconstruct the offending activities by undercutting their self-understanding.
His response is rather to remind us of our civic obligations; to justify
what we do as being of some significance, in a wide, idealized community
of deliberation. This move derives its force from the fact that its ground
is not one over which philosophers have any special entitlement, yet it is
not possible to simply dismiss it as external, as not pertaining to what we

Souls

325

do. The ethical is general. So Kitcher can be both radical, like Dewey, and
Rorty, and yet retain authority in his critique.
However, the difficulty comes in saying what specifically is and isn’t
useful, justifiable, of promise, worth doing. Neither Rorty nor Price
have that particular problem, because they have arguments, which pick
out just those things that fall under the scope of their dialectic; object
naturalism with its attendant metaphysics, for Price, and the premises
of representationalism in philosophy, for Rorty. Kitcher goes an entirely
different way. Because there is no super-context, there is no fixed list of
useful things—we can, however, settle on what is useful deliberatively,
Kitcher suggests, through an idealization that looks like an adaptation
from Rawls.
My hunch, though, is that incommensurable views and interests will
ensure that such judgments of a deliberative community would be essentially contested. I doubt that the kind of deliberatively grounded assessment of importance or use that Kitcher proposes could have much power
to influence the choices of individuals already engaged in the target inquiry.
Moreover, the element of nonstrategic, non-instrumental curiosity-driven—
perhaps useless-seeming—investigation is arguably important for the enterprise. Seemingly pointless inquiries and constructions come to touch on
issues that matter. These, of course, are points the Kitcher himself would
insist on. But if that is the case, then where does the ethical critique of
philosophy lead us?
Perhaps Kitcher’s recipe can be transposed from the level of disciplinary content—problems and methods—to the level of disciplinary structure or practice. What if, instead of asking ourselves what philosophy
should be, what we philosophers should be working on, we were to ask
ourselves; how do we maximize the chances that philosophers spend
their time on useful things? Would we not then be pointing at practical,
tractable issues? One lesson that pragmatic naturalism teaches, in spite
of the rhetoric of some of its practitioners, is that when we talk about the
ends of philosophy, only the thinnest of abstractions will be uncontested.
Thicker, direction-giving project proposals such as Price’s, may, if they
have merit, gain momentum and entrench a particular understanding of
a way of doing what philosophy should be doing, for a time. But the
idea, invoked by Price, that the force of such proposals derives from a
proper understanding of the aims of philosophy—that is surely a hopeless one. This insight, if anything, PCPN should take from Wittgenstein.
As Rorty emphasizes, on the pragmatist’s view of Wittgenstein’s achievement, “he did not show metaphysics to be nonsense. He simply showed it
to be a waste of time”.29 The point is not to replace a misguided notion
of philosophy with a proper one, then, but rather to see that there is no

29 Rorty, “Wittgenstein and the Linguistic Turn,” 163.

326

Bjørn Torgrim Ramberg

particular, distinctive intellectual task that philosophy, as such, requires
of us. By contrast, Price’s suggestion that we first get clear on the proper
aims of philosophy and then figure out how to pursue them invites back
in an essentialism about philosophy that invariably encourages ascent to
free-floating abstractions and ensuing debates of a spurious, vicarious,
and ultimately counterproductive sort.
As Kitcher poses the question that pragmatic naturalism forces on us,
however, the call for reform is not based on some understanding of what
philosophy really is about. Rather, it is based on an appeal to significance
over which philosophy has no particular authority, but which philosophers, as part of a community, are under a general obligation to recognize. Kitcher, as I propose we take his reformist call, in effect encourages
us to lift our gaze from the argumentative texts, and to consider our
practice from the point of view of organization and structure. Might
our current forms of practice of philosophy be modified in ways that
would make us more likely to move the discipline in innovative, useful
directions—as, arguably, philosophy of mind has done in recent decades?
Are there aspects of present structure and organization and norms of
conduct that contribute to the production of philosophy at its worst,
that is, as insular, self-sufficient, self-righteous, privilege-protecting, dustmote-counting exercises? These are, obviously, multi-faceted issues, and
hardly ones with clear answers. We are already deeply enmeshed in them
when we ask about the significance for philosophy of organization and
of human composition—that is, of representation—not semantic but cultural, economic and social. Looked at this way, the question about the
ends of philosophy or the reform of philosophy isn’t really a theoretical
issue at all, not a question of the success of this or that line of deconstructive argument or program-pushing counter-argument. It is now a question of how we politically, practically, and institutionally facilitate the
sort of intellectual activity that Kitcher characterizes as an ideal:
Philosophy, so understood, is a synthetic discipline, one that reflects
on and responds to the state of inquiry, to the state of a variety of
human social practice, and to the felt need of individual people to
make sense of the world and their place in it.30
One must recognize that there is no implicit direction, no program, no
method on offer in this characterization. Transposed along the lines I am
now suggesting, the point of PCPN is to trigger a kind of conceptual unclogging; to foster a commitment to raise destructive argument whenever there
is a proposal or a tendency to close philosophy by tying it to a particular
substantive vision or a particular methodological orthodoxy; a commitment

30 Kitcher, Preludes, 216.

Souls

327

to issue perforating reminders that philosophy, both materially and intellectually, runs on steam generated from engines beyond the discipline; a
commitment to generate proposals for increased openness, flexibility, and
inventiveness in the effort to enrich the discursive spaces in which human
beings struggle to make sense, to be the best humans we can be. The lesson PCPN teaches is not that there is some flaw in this or that conception
of what philosophy is, but that there is no intrinsic end of philosophy at
all to which one might appeal in justification of means that do not stand
up to ethical scrutiny. PCPN, then, is not a source of intellectual fuel for
philosophical innovation; we should stop thinking that such fuel could be
meta-philosophically generated. Rather, the pragmatist anti-foundationalist
message of PCPN is a work-order of a different sort; it reminds us that it is
an integral part of the business of philosophy to improve the justice and the
openness and the representativeness of the practice as we find it, not just as
matter of internal house-cleaning but also in its relations to the communities and societies in which some of us have had the incredible luck to be paid
to do this kind of work.31

References
Brandom, Robert B. “Vocabularies of Pragmatism: Synthesizing Naturalism and Historicism.” In Rorty and His Critics, edited by Robert B. Brandom, 156–183. Oxford:
Blackwell Publishers, 2000.
Davidson, Donald. Essays on Actions and Events. Oxford: Oxford University
Press, 1982.
De Caro, Mario and David Macarthur (editors). Naturalism in Question. Cambridge, MA: Harvard University Press, 2004.
De Caro, Mario and David Macarthur (editors). Naturalism and Normativity.
New York: Columbia University Press, 2010.
Dennett, Daniel C. Consciousness Explained. Boston: Little, Brown and company, 1991.
Fodor, Jerry. Psychosemantics. Cambridge, MA: MIT Press, 1987.
Garber, Daniel. Descartes Embodied: Reading Cartesian Philosophy Through
Cartesian Science. Cambridge: Cambridge University Press, 2000.
Hawthorne, John. Philosophy of Mind: Philosophical Perspectives, volume 21.
Oxford: Blackwell Publishers, 2007.
Kim, Jaegwon. Philosophy of Mind, 3rd edition. Boulder: Westview Press, 2010.
Kitcher, Philip. Preludes to Pragmatism: Toward a Reconstruction of Philosophy.
New York: Oxford University Press, 2012.
Macarthur, David. “Wittgenstein and Expressivism.” In The Later Wittgenstein
on Language, edited by Daniel Whiting, 81–95. London: Palgrave, 2010.
McGinn, Colin. The Character of Mind: An Introduction to the Philosophy of Mind,
2nd edition. Oxford: Oxford University Press, 1997 (original edition published
1982).

31 I am happy to acknowledge my gratitude to the editors of this volume for careful reading and constructive suggestions that significantly improved this paper.

328

Bjørn Torgrim Ramberg

McGinn, Marie. “Wittgenstein’s Naturalism.” In Naturalism and Normativity, edited
by Mario De Caro and David Macarthur, 322–351. New York: Columbia University Press, 2010.
Misak, Cheryl. The American Pragmatists. New York: Oxford University Press,
2013.
Price, Huw. Naturalism Without Mirrors. New York: Oxford University Press,
2011.
Price, Huw. “Naturalism Without Representationalism.” In Naturalism Without Mirrors, edited by Huw Price, 184–199. New York: Oxford University Press, 2011.
Price, Huw. Expressivism, Pragmatism and Representationalism. Cambridge:
Cambridge University Press, 2013.
Rorty, Richard. Philosophy and the Mirror of Nature. Princeton: Princeton University Press, 1979.
Rorty, Richard. “Pragmatism, Relativism, and Irrationalism.” In Consequences of
Pragmatism, edited by Richard Rorty, 160–175. Minneapolis: University of Minnesota Press, 1982.
Rorty, Richard. “Introduction: Metaphilosophical Difficulties of Linguistic Philosophy.” In The Linguistic Turn, edited by Richard Rorty, 1–40. Chicago: The
University of Chicago Press, 1992 (first published 1967).
Rorty, Richard. “Dennett on Intrinsicality.” In Truth and Progress, Philosophical
Papers, volume 3, edited by Richard Rorty, 98–121. Cambridge: Cambridge
University Press, 1998.
Rorty, Richard. “Wittgenstein and the Linguistic Turn.” In Philosophy as Cultural
Politics, Philosophical Papers, volume 4, edited by Richard Rorty, 160–175.
Cambridge: Cambridge University Press, 2007.
Wittgenstein, Ludwig. Philosophical Investigations, 3rd edition. Translated by G.
E. M. Anscombe. Oxford: Basil Blackwell, 1968.

Contributors

Dorit Bar-On is a Professor of Philosophy at the University of Connecticut. She has published extensively in the philosophy of mind and language, as well as epistemology and metaethics. In her book Speaking
My Mind: Expression and Self-Knowledge (Oxford, Clarendon Press
2004), she developed a neo-expressivist framework for understanding first-person authority, later applied to ethical and other evaluative
discourses. In recent years, she has become interested in continuities
and discontinuities between animal expressive communication and
language. She is currently working on a book manuscript tentatively
titled Expression, Communication, and the Origins of Meaning. She is
the director of an eponymous interdisciplinary research group at the
University of Connecticut.
Sorin Bangu is a Professor of Philosophy at the University of Bergen, Norway. He has previously held positions at the University of Cambridge
and the University of Illinois at Urbana-Champaign. He is the author
of The Applicability of Mathematics in Science: Indispensability and
Ontology (Palgrave Macmillan, 2012) and the editor of Naturalizing
Logico-Mathematical Knowledge: Approaches from Philosophy, Psychology and Cognitive Science (Routledge, 2017). He has published
widely in Philosophy of Science and Philosophy of Mathematics, with
a particular interest in the work of Wittgenstein and Quine.
Stina Bäckström is a Senior Lecturer at Södertörn University and a
researcher at Åbo Akademi University. She received her PhD in philosophy from the University of Chicago in 2013. She works on issues in the
philosophy of mind and action, inspired by figures such as Anscombe,
Wittgenstein, Cavell, and McDowell. Her dissertation, “The Mind’s
Movement: An Essay on Expression”, treats acts of expression and
intentional action. Since then she has written papers on, for example,
what it is to depsychologize psychology, Ryle’s conception of skilled
action, knowing other minds, and expression and self-consciousness.
Kevin Cahill is a Professor of Philosophy at the University of Bergen. He
works mainly on Wittgenstein’s Philosophy and the Philosophy of the

330

Contributors

Social Sciences. His publications include The Fate of Wonder: Wittgenstein’s Critique of Metaphysics and Modernity (Columbia, 2011).
William Child is a Professor of Philosophy at the University of Oxford
and a Fellow in Philosophy at University College, Oxford. He is the
author of Causality, Interpretation, and the Mind (Oxford University Press, 1994) and Wittgenstein (Routledge, 2011), and co-editor,
with David Charles, of Wittgensteinian Themes: Essays in Honour of
David Pears (Oxford University Press, 2001). He has published widely
on Wittgenstein and on issues in the Philosophy of Mind.
Annalisa Coliva is a Professor of Philosophy at the University of California, Irvine. She is the author of ten books, in both English and Italian,
including The Varieties of Self-Knowledge. London (Palgrave, 2016),
Extended Rationality. A Hinge Epistemology (Palgrave, 2015) and
Moore and Wittgenstein. Scepticism, Certainty and Common Sense
(Palgrave, 2010). She has published numerous articles in Epistemology
and Philosophy of the Mind, with a particular focus on the work of
Moore and Wittgenstein.
Eugen Fischer (BPhil, DPhil, Oxford; Habilitation, LMU Munich) is a
Reader in Philosophy at the University of East Anglia. He has been a
Heisenberg Research Reader (DFG), Golestan Fellow at the Netherlands Institute for Advanced Study (NIAS), and Senior Research Fellow
at Collegium Budapest. He currently is perhaps the only Wittgenstein
scholar who is also an experimental philosopher. His main interests are
in the philosophy of philosophy and the philosophy of perception. His
main projects are the exploration of therapeutic conceptions of philosophy and the study of how automatic cognition (intuition) shapes philosophical thought. He has examined automatic stereotypical inferences
and analogical reasoning with metaphors (‘philosophical pictures’), in
the development of philosophical problems, arguments, and theories.
His empirical work on stereotypical inferences (with psychologist P.E.
Engelhardt) has pioneered the use of psycholinguistic methods in experimental philosophy. He is the author of Linguistic Creativity: Exercises
in ‘Philosophical Therapy’ (Springer 2000) and Philosophical Delusion
and its Therapy (Routledge 2011, p/b 2013). He has co-edited (with
Erich Ammereller) Wittgenstein at Work: Method in the “Philosophical
Investigations” (Routledge 2004) and (with John Collins) Experimental Philosophy, Rationalism, and Naturalism: Rethinking Philosophical
Method (Routledge 2015).
Daniel D. Hutto is a Professor of Philosophical Psychology at the University of Wollongong, Australia. He is the author of Folk Psychological Narratives: The Socio-Cultural Basis of Understanding Reasons
(MIT Press, 2008) and Wittgenstein and the End of Philosophy: Neither Theory Nor Therapy (Palgrave, 2003), and he is the co-author,

Contributors

331

together with Erik Myin, of Evolving Enactivism: Basic Minds Meet
Content (MIT Press, 2017) and Radicalizing Enactivism: Basic Minds
without Content (MIT Press, 2013). He has published very extensively
in Philosophy of the Mind and Cognitive Science, as well as frequently
writing about Wittgenstein and on Meta-Philosophy.
Jonathan Knowles is a Professor of Philosophy at the Norwegian University of Science and Technology in Trondheim. He took his doctorate
in 1995 at Birkbeck College, London, with a thesis on philosophy of
cognitive science and linguistics. Specializing initially in philosophy of
the mind, language, and psychology, he has since become interested in
metaphilosophical issues concerning the interrelations between naturalism, realism and representationalism and the possibility of metaphysical inquiry. He has published a number of articles in journals
such as Analysis, Philosophical Quarterly, Erkenntnis and Synthese
and is author of the monograph Norms, Naturalism and Epistemology: The Case for Science Without Norms (Palgrave 2003).
David Macarthur is an Associate Professor in the Philosophy Department at the University of Sydney. He works at the interface of contemporary pragmatism, Wittgenstein’s philosophy of psychology and
language, and philosophy of art. In addition to these topics, he has
published articles in leading philosophy journals and books on liberal
naturalism, skepticism, common sense, Stanley Cavell, perception, language, philosophy of architecture, and philosophy of photography and
film. He has co-edited three collections of papers with Mario De Caro
(Roma Tré): Naturalism in Question (Harvard, 2004); Naturalism and
Normativity (Columbia, 2010); and Philosophy in an Age of Science:
Physics, Mathematics and Skepticism (Harvard, 2012). He is currently
editing Pragmatism as a Way of Life: Hilary and Ruth-Anna Putnam
on the Lasting Legacy of James and Dewey for Harvard University
Press.
Thomas Raleigh is currently a Post-Doctoral Fellow at the Ruhr-University,
Bochum. He is also a Visiting Research Fellow at King’s College
London. He has previously held positions at the University of Vienna,
the Norwegian University of Science & Technology, Concordia University (Montreal), and U.N.A.M in Mexico City. His research is primarily in Philosophy of the Mind and Epistemology, with a particular
interest in the work of Wittgenstein. As well as the present volume,
he is also the co-editor, together with Jonathan Knowles, of Acquaintance: New Essays (forthcoming, Oxford University Press).
Bjørn Torgrim Ramberg is from Oslo, Norway, where he also began his
studies in philosophy in 1979. Going to Canada in 1982, he completed
his MA and PhD in philosophy at Queen’s University, submitting a
dissertation on Donald Davidson’s philosophy of language (1987).

332

Contributors

His first academic appointments were at Queen’s University, Harvard
University, and Simon Fraser University. Since 1997, he has been a
Professor of Philosophy at the University of Oslo, where he has also
been a Core Group member of the Centre for the Study of Mind in
Nature (CSMN). Ramberg has also been a visiting professor at the
University of Chicago (2012) and Nan Jing University (2012). He has
published on various aspects of pragmatism, interpretation and meaning, focusing mainly on work by Richard Rorty and Donald Davidson
and linking their work with the hermeneutical tradition in philosophy.
Ramberg has an abiding interest in the relation between philosophy
and other disciplines and between theoretical philosophy and life.
Glenda Satne is a Vice-Chancellor’s Fellow at the University of Wollongong. She was previously a Marie Curie Experienced Researcher at the
Center for Subjectivity Research at the University of Copenhagen. Her
research focuses primarily on Philosophy of the Mind, Social Ontology, and Philosophy of Language. She is currently engaged in several
collaborative international projects on Collective Intentionality, the
Second-Person, and the role played by culture in human evolution.
Benedict Smith was an undergraduate at the University of Glasgow and
a graduate student at the University of Warwick. He is currently a
Lecturer in Philosophy and Director of the MA at Durham University,
having previously been a Research Fellow at Durham. His research
interests include ethics, philosophy of the mind, history of philosophy,
and philosophy of psychiatry, and he teaches a variety of undergraduate and postgraduate modules, including metaethics, phenomenology,
and history of philosophy. His publications include Particularism and
the Space of Moral Reasons (Palgrave Macmillan 2011) and articles
on topics including motivation, trust, and the role of concepts in our
thought and practice. His current research projects include the nature
of belief, realism and moral perception which continue his long-standing
interest in the relation between naturalism and normativity.
Paul Snowdon is an Emeritus Professor of Philosophy at University College London, where he was previously, for many years, the Grote Professor of Philosophy of Mind and Logic. He is the author of Persons,
Animals, Ourselves (Oxford University Press, 2014) and the co-editor,
along with Stephen Blatti, of Animalism: New Essays on Persons, Animals, and Identity (Oxford University Press, 2016). He has published
extensively in Philosophy of the Mind, with a particular focus on perceptual experience and on personal identity.
Julia Tanney has written numerous articles on philosophy of the mind
and language, focusing especially on reason explanation, rule-following,
self-knowledge, and the nature of philosophical investigation. Her
book, Rules, Reason, and Self-Knowledge, was published by Harvard

Contributors

333

University Press in 2013. She is an international expert on the philosophy of Gilbert Ryle and the later Wittgenstein. Having spent most of
her career in the UK, with visiting posts in France, she now works independently, dividing her time between Paris and the South of France.
Charles Travis is an Emeritus Professor of Philosophy at King’s College
London. He is also currently a Professor at the Institute of Philosophy
of Porto, Portugal. Before coming to King’s, he was a Professor at
Northwestern University, and before that at the University of Stirling.
He has also held visiting professorships at the University of Michigan,
and at Harvard. He is the author of many books—including Thoughts
Footing (OUP, 2009), Unshadowed Thought (Harvard University
Press, 2000) and The Uses of Sense (OUP, 1989). Oxford University
Press has also published three volumes of his selected essays. His work
focuses primarily on the philosophy of language, thought, and perception, with an especial focus on the work of Frege and Wittgenstein.

Index

Albritton, Rogers 84
analogies 100–101, 265, 267, 268–269;
language 265, 268–269; and
metaphors 268; between philosophy
and psychotherapy 269–271,
280–282
animal minds 7–8, 48, 92, 102, 177–180,
191–192, 196–197
anti-representationalism 288, 289,
292–301, 303, 306
Aristotle 37n16, 40, 120, 171
authority 6, 97, 116, 126–127, 311, 313,
322, 324–326
automatic cognition 271–274, 276–282;
see also intuitions
Bar-On, Dorit 97n1, 225, 227, 231, 232,
233
behaviorism 8, 53, 96, 98, 100–101, 212
Blackburn, Simon 1, 2n4
Bouwsma, O.K. 245n7
Brandom, Robert 1, 2n4, 181, 181n24,
184, 312, 313, 323n27
Burge, Tyler 87n16, 105n25, 181n24,
185, 185n39
Canberra plan 5, 290, 291, 294, 303
Carnap, Rudolf 289, 296, 298, 299
Cartesianism 8, 53, 209, 211–214, 216,
217; see also Descartes, Rene
causation 35, 37, 39–41, 62n23, 92–93,
100n8
Cavell, Stanley 1, 1n2, 35n7, 42–43, 59
Chomsky, Noam 148, 289, 303–305
cognitive science 209
Collins, John 304
common sense 8, 44, 85, 88, 92–93,
209, 216

consciousness 2–3, 26–29, 79–80,
82–85, 91, 94, 102, 107, 124–125,
135; see also sensations
continuity: human-animal 223–227,
231–235
Crary, Alice 1, 8, 224, 225, 227, 228n12,
231, 233, 233n20
Davidson, Donald 7, 41n26, 53n78,
178, 179–181, 181n24, 182–187,
193–197, 199, 319n23
deflationism (semantic) 60, 103, 287,
291, 292, 295, 297–299
Dennett, Daniel 1, 311n1
Descartes, Rene 37n16, 50
Dewey, John 39, 311, 312, 313, 315,
316, 317, 325
Dromm, Keith 72, 73n76
Drury, Maurice O’Connor 250n34
Dupre, John 46n44, 59
epistemology 31, 34, 37n16, 38,
96–98, 104–106, 108, 116–117; selfknowledge 96–98, 114, 116; see also
inferences: inferential model of selfknowledge; skepticism: mathematical
experimental philosophy 9, 260, 261;
sources Project in 262, 266–267,
271, 279, 282; and warrant project
in 261–264
experiments 35, 261, 270, 276–278
expression 223–239
expressive behavior 48–51, 90–92, 96,
103, 120, 178, 187–197, 199–200
expressivism 2n4, 97–98, 100, 112–113,
116; global 288, 289, 292, 294, 299,
300, 303, 305, 315n15, 321
externalism 21–22, 87n16, 105n25

336

Index

Flanagan, Owen 244, 245n7
Fodor, Jerry 100n8, 319
Ford, Anton 233, 234, 234n25, 237
Frege, Gottlob 4, 4n8, 6, 26–28,
119–132, 135–137, 139–142,
146–149; Begriffsschrift 128, 148;
Der Gedanke 4n8, 119, 124, 131,
135, 138n18, 142; Sinn 119–120,
123, 125–127, 129–131, 136; third
realm 4, 119, 124–125, 127, 131,
136, 147–148; Vorstellungen 119,
125, 135, 138
genealogy 151, 165, 171
generality: 128–130, 138; categorical
232–234, 234n25, 237, 238
Gier, Nicholas F. 251n34
God 42, 46–47, 52
Goldfarb, Warren 245
grammar 48, 65, 96, 101, 102n14,
107–110, 127–128, 154, 168
Hacker, Peter M.S. 151–155, 165, 169,
243, 244n4
Hagberg, Garry L. 211, 212
hallucination 263; argument from 264,
273–276
Hanfling, Oswald 243n1, 249n28, 252,
253n46, 255
hardening 157, 159, 164–171
Hawthorne, John 323n28
Hilbert, David 148, 153
Hornsby, Jennifer 59
Horowitz, Alexandra 206–210, 214
Horwich, Paul 1, 9, 287–289, 292–295,
297, 299, 301–304
Hume, David 9, 243–245, 247, 249,
251–258
illusion 263; argument from 264,
273–276
inferences: contextually inappropriate
275–279, 281; inferential model of
self-knowledge 96, 10, 104, 111,
116; stereotypical 274–279
introspection 80–82, 85, 98, 100,
114–115
intuitions 88–89, 261–264, 272–274,
279, 281–282; evidentiary value of
261–262, 264, 273–274
Jackson, Frank 2, 36, 38–41, 290
Jones, Peter 244n2

Kierkegaard, Soren 53
Kitcher, Philip 10, 311–314, 317, 322,
324–327
Lewis, David K. 5–6, 79n1, 85, 153n11,
290
Loeb, Louis 251n36
logic 49, 119–121, 126, 145–146, 148,
151n2
Macarthur, David 58–60, 62, 70n64,
74n86, 315n15
Maddy, Penelope 1, 3, 151n2, 170n58,
171
mathematics 7, 19, 37, 88, 113, 148,
151–160, 164–171; see also
skepticism: mathematical
McDowell, John McDowell 1, 5, 8,
23n22, 41n26, 49, 59, 61–62, 65–71,
170n57, 181, 181n24, 183n31,
184n74, 233–235, 235n27, 237,
237n31
McGinn, Colin 323n28
McGinn, Marie 1, 33n2, 65–66, 69–70,
72, 317, 318
meaning: as use 20, 29–31, 45, 80, 103,
104n21, 113
mental: capacities 221; contents 66–67;
descriptions 219; expression 219;
predicates 218, 219; properties 83,
90; states 96–101, 105, 111, 116, 216
metacognitive insight 261, 264, 276,
280–288
metaphor 9, 22–23, 265, 267–268,
275–277
metaphysics 10, 36–39, 41–42, 60,
64n38, 85–87, 313, 315, 316,
320–325
Misak, Cheryl 313
Monk, Ray 251n34
Moore, G.E. (Moore’s Paradox)
112–114, 117
Moran, Richard 1, 112n46
Munson, Thomas N. 251n34
naturalism 2–3, 16, 21, 29, 33–36, 96,
98, 149; compatible with therapeutic
philosophy 263–267, 273–276;
metaphilosophical 261–263, 266;
methodological 260, 262–263, 266;
methodological vs. metaphysical 38,
57–58, 260; natural properties 5–6,
79–86; object 289–294, 297–302;

Index
orthodox 318–321; pragmatic 309,
310, 312, 313, 320, 325, 326;
strict vs. liberal 4–5, 45–46, 56–62,
65–66, 70, 117; subject 287, 292,
300, 303, 304, 320–322
necessity 110, 126, 151–157, 159, 161,
163, 165–171
norm 2, 152–154, 156–157, 160–161,
163–165, 167, 170–171
normativity 40–41, 61, 68–70, 314,
318n19
pain 44, 48–49, 79–87, 90–94,
98–109; see also sensations
paradox 43, 263–265, 273, 279, 281
Peacocke, Christopher 82–86, 90–94,
106n29
Pears, D.F. 1, 17–19, 244, 248–249, 251
perception 124; problem of 263–264,
273–276, 279
perplexity 317–318, 320, 323
philosophy: as continuous with science
56–57; diagnostic approaches in
264–265, 279, 281; therapeutic
conception of 261, 268–273,
279–282; see also Wittgenstein,
Ludwig, conception of philosophy
physicalism 35–36, 39–41
Plato 37n16
platonism/anti-platonism 5–7, 67–69,
79, 81–82, 85, 88–89, 153, 156, 169
pragmatism 10, 313, 314
Price, Huw 1, 1n4, 9–10, 59, 287–305,
311, 314, 317, 319–322, 324–326
principle of charity 266, 276
private language 6, 22, 79–81, 98–100,
104, 123–124, 153n10; see also
sensations: privacy of
pseudo-problems 265–266, 272;
dissolving of 96, 265–266, 273–274,
279
psycholinguistics 261, 263–264,
274–279, 282
Putnam, Hilary 22, 41n26, 59, 79n1
quietism 45–46, 70n64, 323
Quine, Willard V.O. 4, 36, 38n19, 131,
244–245, 298–300, 304, 305
realism/anti-realism 17–18, 68, 83n8,
87–89, 91, 93–94
regularity: empirical 155–156, 161,
164–166, 169–71

337

representation 10, 120–123, 126–127,
152–153; representing-as 123, 132,
142
representationalism 69, 315, 316,
319n22, 320, 321n25, 325
Rhees, Rush 33, 250n34
Rorty, Richard 10, 311–318, 320–325
rule 7, 99, 109, 152, 154–157, 160,
163–170
rule-following 20, 25–28, 31–32, 66–67,
72, 97n2, 99, 170
Russell, Bertrand 38, 99n6, 119–121,
123–133, 147–148; Russellian
propositions 119, 123
Ryle, Gilbert 8, 216, 220, 221
science 2–5, 17, 33–41, 56–62, 71–74,
309, 310, 315, 318–320, 322, 323,
324; physics 33–35, 46n46, 57, 73,
146; unity of 39
Sellars, Wilfrid 74
sensations 6, 17, 19, 22–23, 44, 48,
50, 79–83, 85–87, 92–93, 98–109;
first-person (self-ascription) vs. thirdperson ascriptions 81–82, 84, 92–93,
96–98, 103–104, 106–108, 111–112,
115–116; privacy of 22, 79–81,
98–99, 124; see also pain
skepticism: continuity 7–8, 177,
179–182, 184–185, 187, 189,
192, 196–197, 199; mathematical
152–153, 157, 170–171
Smith, Norman Kemp 251n35,
254n50
soul 51–54, 102, 309, 311, 313, 315,
317, 319, 321, 323, 325
Spiegelberg, Herbert 251n34
Strawson, Peter F. 117, 244, 255
Stroud, Barry 1, 59, 167n54
supernatural 5, 39, 42, 46–49, 52–53,
56, 60
theory of mind 216
therapy: cognitive 261, 279–282;
concept of 268–269; and philosophy
9, 45, 261, 268–273, 279–282
Thomasson, Amie 289, 294–296,
299
truth 58, 82, 84, 103, 119–122, 124–127,
129–132; correspondence theory of
131, 135–141, 145, 148–149; truthbearer 131, 135, 138–139; truthmaker 319, 322

338

Index

unity: conceptual 232, 235, 237, 237n31,
238
urge to misunderstand 261, 270, 276
Williams, Bernard 8, 244n2, 245–247,
249, 256n60, 257
Williamson, Timothy 58
Wittgenstein, Ludwig 1–3, 223,
225, 226, 227, 227n10, 230–232,
235–238; On Certainty 105, 107,
109; his conception of philosophy
22–25, 43–48, 56, 63–68, 149,

261, 263–267, 269–271, 313–318,
320–323, 325; his conception of
religion 52–54; and forms of life
6, 167–169; and metaphilosophical
naturalism 33–34, 263–267; his own
experiments 270; and pragmatism
314, 315; Tractatus LogicoPhilosophicus 34n5, 36, 38n18,
43, 119–120, 123, 131–132, 137,
142–149
Wright, Crispin 1, 23n22, 90n18, 97n2,
106n28, 170n57