Vous êtes sur la page 1sur 187

Wittgenstein and Merleau-Ponty

Ludwig Wittgenstein and Maurice Merleau-Ponty are two of the most

­important philosophers of the twentieth century, yet they work within tra-
ditions that are generally regarded as standing in contrast to one another.
However, as this outstanding collection demonstrates, there are many simi-
larities between them. Both reject a Cartesian picture of the mind and offer
an alternative that does justice to the role played by bodily action, language,
and our membership within a community that shares a way of life.
This is the first collection to compare and contrast the work of these two
major philosophers. Fundamental topics and problems discussed include
their engagement with Gestalt psychology; the role of community in their
philosophies; the parallels between Merleau-Ponty on representation and
depiction and Wittgenstein on saying and doing; their treatment of expres-
sion; solipsism; and indeterminacy, knowledge and certainty.
It is essential reading for anyone studying the work of Wittgenstein and
Merleau-Ponty, as well as those interested in phenomenology, philosophy of
mind, and philosophy of language.

Komarine Romdenh-Romluc is a senior lecturer in philosophy at the Uni-

versity of Sheffield, UK. She is author of The Routledge Philosophy Guide-
book to Merleau-Ponty and Phenomenology of Perception (2011).
Routledge Research in Phenomenology
Edited by
Søren Overgaard,
University of Copenhagen, Denmark
Komarine Romdenh-Romluc,
University of Sheffield, UK
David Cerbone,
West Virginia University, USA

1  Phenomenology and the Transcendental

Edited by Sara Heinämaa, Mirja Hartimo and Timo Miettinen

2  Philosophy of Mind and Phenomenology

Conceptual and Empirical Approaches
Edited by Daniel O. Dahlstrom, Andreas Elpidorou, and Walter Hopp

3  Phenomenology of Sociality
Discovering the ‘We’
Edited by Thomas Szanto and Dermot Moran

4  Phenomenology of Thinking
Philosophical Investigations into the Character of
Cognitive Experiences
Edited by Thiemo Breyer and Christopher Gutland

5  Wittgenstein and Merleau-Ponty

Edited by Komarine Romdenh-Romluc
Wittgenstein and

Edited by
Komarine Romdenh-Romluc
First published 2017
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
© 2017 selection and editorial matter, Komarine Romdenh-Romluc;
individual chapters, the contributors
The right of the editor to be identified as the author 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
Names: Romdenh-Romluc, Komarine, editor.
Title: Wittgenstein and Merleau-Ponty / edited by Komarine
Description: 1 [edition]. | New York : Routledge, 2017. | Series:
Routledge research in phenomenlogy ; 5 | Includes bibliographical
references and index.
Identifiers: LCCN 2016053812 | ISBN 9780415625128 (hardback :
alk. paper)
Subjects: LCSH: Wittgenstein, Ludwig, 1889–1951. | Merleau-Ponty,
Maurice, 1908–1961. | Phenomenology.
Classification: LCC B3376.W564 W52135 2017 | DDC 192—dc23
LC record available at https://lccn.loc.gov/2016053812
ISBN: 978-0-415-62512-8 (hbk)
ISBN: 978-1-315-75484-0 (ebk)
Typeset in Sabon
by Apex CoVantage, LLC

Abbreviations of Cited Works by Wittgenstein

and Merleau-Ponty vii

Introduction 1

1 Wittgenstein and Merleau-Ponty on Gestalt Psychology 11


2 Expression 31

3 Merleau-Ponty and Wittgenstein on Mindreading:

Exposing the Myth of the Given Mind 49

4 Community Without Conservatism: Wittgenstein

and Merleau-Ponty on the Sociality of Subjectivity 66

5 The World and I 81


6 Painting and the Promiscuity of Vision 100


7 The Recovery of Indeterminacy in Merleau-Ponty

and Wittgenstein 114
vi Contents
8 Wittgenstein and Merleau-Ponty on Knowledge and Certainty 137

Contributors 175
Index 177
Abbreviations of Cited Works by
Wittgenstein and Merleau-Ponty

BB 1969. Blue and Brown Books. Oxford: Blackwell.
CV 1980. Culture and Value. Edited by G. H. Von Wright. Oxford:
LC 1966. Lectures and Conversations on Aesthetics, Psychology, and
Religious Belief. Edited by C. Barrett. Oxford: Blackwell.
LPP 1988. Lectures on Philosophical Psychology 1946–7, notes by P. T.
Geach, K. J. Shah and A. C. Jackson. Edited by P. Geach. Hemel
Hempstead: Harvester Wheatsheaf.
LW II 1992. Last Writings on the Philosophy of Psychology; Volume II.
Edited by G. H. von Wright and H. Nyman, translated by C. G.
Luckhardt and M.A.E. Aue. Oxford: Blackwell.
LW I 1982. Last Writings on the Philosophy of Psychology, Volume I.
Edited by G. H. von Wright and H. Nyman, translated by. C. G.
Luckhardt and M.A.E. Aue. Oxford: Blackwell.
OC 1972. On Certainty. Edited and translated by G.E.M. Anscombe.
New York: Harper and Row.
PI 1968. Philosophical Investigations. Translated by G.E.M. Ans-
combe. Oxford: Blackwell.
PO 1993. Philosophical Occasions 1912–1951. Edited by J. Klagge
and A. Nordman. Indianapolis: Hackett.
PR 1975. Philosophical Remarks. Edited by R. Rhees, translated by
R. Hargreaves and R. White. Oxford: Blackwell.
RC 1977. Remarks on Colour. Oxford: Blackwell.
RFM 1978. Remarks on the Foundations of Mathematics. Edited by
G. H. von Wright, R. Rhees and G.E.M. Anscombe, translated by
G.E.M. Anscombe. Oxford: Blackwell.
RPP I 1980. Remarks on the Philosophy of Psychology, Volume I. Edited
by G.E.M. Anscombe and G. H. von Wright, translated by G.E.M.
Anscombe. Oxford: Blackwell.
RPP II 1980. Remarks on the Philosophy of Psychology, Volume II. Edited
by G. H. von Wright and H. Nyman, translated by C. G. Luckhardt
and M.A.E. Aue. Oxford: Blackwell.
viii  Abbreviations of Cited Works by Wittgenstein and Merleau-Ponty
TLP 1974. Tractatus Logico-Philosophicus. Translated by D. F. Pears
and B. F. McGuiness. London: Routledge.
Z 1967. Zettel. Edited by G.E.M. Anscombe and G. H. von Wright,
translated by G.E.M. Anscombe. Oxford: Blackwell.

CPP 2010. Child Psychology and Pedagogy: The Sorbonne Lectures
1949–1952. Translated by T. Welsh. Evanston, IL: Northwestern
University Press.
HLP 2002. Husserl at the Limits of Phenomenology. Translated by
L. Lawlor. Evanston, IL: Northwestern University Press.
PhP 2011. Phenomenology of Perception. Translated by D. Landes.
London: Routledge.
PrP 1964. The Primacy of Perception. Edited by J. Edie, translated by
various. Evanston, IL: Northwestern University Press.
PW 1973. Prose of the World. Edited by C. Lefort, translated by
J. O’Neill. Evanston, IL: Northwestern University Press.
S 1964. Signs. Translated by R. C. McCleary. Evanston, IL: North-
western University Press.
SB 1963. The Structure of Behavior. Translated by A. L. Fisher. Bos-
ton, MA: Beacon Press.
SNS 1964. Sense and Non-Sense. Translated by H. L. Dreyfus and P. A.
Dreyfus. Evanston, IL: Northwestern University Press.
WP 2004. The World of Perception. Translated by O. Davis. London:
VI 1968. The Visible and the Invisible. Edited by C. Lefort, translated
by A. Lingis. Evanston, IL: Northwestern University Press.

Numbered remarks are marked §. Other references are to page numbers.

Komarine Romdenh-Romluc

Maurice Merleau-Ponty and Ludwig Wittgenstein are two of the most

important writers in twentieth century, Western philosophy. Merleau-Ponty
is a leading figure in the phenomenological movement. Whilst there has been
sustained interest in his work since his untimely death, it has undergone
something of a renaissance over the past few years, as theorists from diverse
philosophical backgrounds, and from a variety of different disciplines,
appeal to it as a rich source of ideas that can help illuminate many contem-
porary debates. Wittgenstein’s work is perhaps best known in the context of
British analytic philosophy, where his influence cannot be underestimated.
Whilst these philosophical movements are often held up as standing in some
kind of opposition, there are various overlaps and interconnections between
them. It is thus perhaps not entirely surprising that there are various striking
parallels between the ideas of Merleau-Ponty and Wittgenstein. The aim of
this volume is to explore some of these points of comparison.
One thread that runs through all the papers in this volume is an important
similarity between their respective philosophical projects. Both begin with
descriptions of everyday phenomena. For Merleau-Ponty, this marks his
explicit commitment to the phenomenology movement. The extent to which
Wittgenstein read phenomenological works, or was influenced by writers in
this tradition, is unclear. However, as Spiegelberg (1981) notes, Wittgenstein
uses the term ‘phänomenologie’ in relation to his own work in a few places,
and whilst it is not obvious that this signalled any explicit engagement with,
or commitment to, the phenomenological tradition, it does perhaps indicate
his interest in descriptions of ordinary phenomena as the starting-point for
philosophy.1 On the face of it, Merleau-Ponty and Wittgenstein describe dif-
ferent things. Merleau-Ponty’s central concern is experience. Wittgenstein,
on the other hand, focuses on linguistic behaviour, captured by his famous
notion of language games. However, as Katherine Morris notes in her paper
for the volume, it would be a mistake to think that this marks a significant
gulf between their approaches. As the idea of a language-game makes clear,
our language is embedded in our forms of life, and if we want to understand
the former, we need some grasp of the latter. One way to think of a form of
life is as some general way (or collection of loosely related ways) of engaging
2 Introduction
with the world. It includes a set of practices, taken-for-granted ideas, behav-
iour, experiences, and so on—all the miscellanea that make up the largely
implicit ‘background’ to our existence. Thus Wittgenstein’s interest in lan-
guage takes in a lot more than just a set of symbols and their syntax. It is
wider, even, than just a focus on explicit language use. Similarly, Merleau-
Ponty’s conception of experience is not of an inner datum, available only to
introspection. Instead, experience means, for him, the life-world—the world
as we experience it and interact with it—together with the details of our
own lives, our practices, and our behaviour, including our linguistic activi-
ties. It follows that whilst their investigations do not completely coincide,
there is significant overlap between them.
In addition, both Merleau-Ponty and Wittgenstein argue for the need to
overthrow a distorted picture of the world and our place in it, which they
seek to do by examining everyday phenomena. For Merleau-Ponty, this is
the framework of Objective Thought: a conception of the world as reducible
to its most basic atoms, which stand in causal connections to one another.
This basic understanding of the world gives rise to two apparently opposed
positions. These are Empiricism, which takes consciousness to be just one of
the things in the world, and so a physical entity that is reducible to its basic
atoms and whose behaviour can be fully explained in terms of causal laws;
and Intellectualism, which holds that consciousness constitutes the world
and as such, is unlike anything within it. Neither of these views correctly
captures the real nature of things, according to Merleau-Ponty. The picture
to be overthrown by Wittgenstein’s work is presented less as a global way
of thinking and more a ragtag bag of different, deeply entrenched ideas.
These include a Cartesian picture of the mind as a soul hidden in the body;
an Augustinian conception of language as a uniform, formal system; and the
idea of the self as essentially isolated, to name just three. It is hopefully clear
from just this brief summary that there are some parallels between at least
some of the ideas the two philosophers reject.
If one wishes to start from the everyday phenomena, an obvious question
to ask is: what are they, and how should we characterise them? In answering
this question, Merleau-Ponty draws heavily on Gestalt psychology—a field
he studied extensively. The notion of the Gestalt—a meaningful unity that
characterises perceptual experience—assumes a central importance in his
work, since he takes it to reveal a deep truth about the status of our world,
our relation to it, and our own nature. This interest in Gestalt psychology
is something he shares with Wittgenstein. The latter was much less familiar
with Gestalt psychology but studied a few works in much depth. He was
particularly interested in the ideas of Wolfgang Köhler, which he addressed
at length in his last lectures. In the first paper of this volume, Katherine
Morris examines Merleau-Ponty’s and Wittgenstein’s respective reactions to
Köhler’s work, using this as a springboard to reflect more widely on simi-
larities between their philosophical ideas.
Introduction  3
Wittgenstein and Merleau-Ponty both criticize Köhler’s descriptions of
Gestalt phenomena. On the face of it, this is for very different reasons. Witt-
genstein argues that ambiguous figures that can be seen in more than one
way, such as the duck-rabbit, constitute a paradox. In the shift from duck
to rabbit or vice versa—what is commonly known as a Gestalt-switch—­
everything remains the same, yet something also alters. This leads him
to distinguish two different senses of the word ‘see’. Morris tells us that
Wittgenstein draws this distinction with two examples: ‘I see this’, and, ‘I
see a likeness between these two faces’. The answer to the paradox is that
what remains the same throughout a Gestalt switch is seen in the first sense,
whilst what changes is seen in the second sense. He criticizes Köhler for not
recognising this difference between types of seeing. Wittgenstein connects
the second sense of seeing with the perception of other phenomena, which
are ubiquitous in our experience of the world, such as depth, the perception
of a figure against a background, and awareness of three-dimensionality. He
then emphasizes the idea that to understand the second sense of perception,
we have to acknowledge that we can perceive meanings. Merleau-Ponty, in
contrast, criticizes Köhler and the other Gestalt psychologists for failing to
see the implications of their discoveries. Morris presents two of Merleau-
Ponty’s lines of thought in this regard. First, he takes Gestalt phenomena to
put pressure on certain dichotomies that characterize our conceptual frame-
work. These include: cause versus reason, explicit awareness versus being
unaware, mind versus world, and consciousness versus body. Second, he
argues that to properly describe Gestalt phenomena, we must reconceive
the body and acknowledge the existence of distinctively bodily forms of
knowledge, intelligence, consciousness, and intentionality.
Despite these differences, Morris argues that there are a number of par-
allels between their approaches. There are some interesting resonances
between the particular ideas they each marshal to make their respective
cases against aspects of Köhler’s thought. Both Merleau-Ponty and Wittgen-
stein can also be viewed as preoccupied with a single task: to restore mean-
ing to human life in the face of an indefensible scientism.
Merleau-Ponty puts the idea of the Gestalt to use in understanding expres-
sion. The latter notion also plays an important role for Wittgenstein, and
although he does not make explicit use of Gestalt psychology in his discus-
sion of this topic, some of the claims he makes about expression coincide
with Merleau-Ponty’s. Kathleen Lennon’s essay considers some central ways
in which expression figures in their work, and the accounts they give of it.
Both thinkers employ the notion to provide an alternative to the picture of
the mind as an internal, Cartesian realm and as a means to do away with
dichotomies between the inner and the outer, self and others. The idea of
expression is perhaps most readily associated with their discussions of bodily
gestures as manifesting psychological life—both our own and those of oth-
ers we encounter. As Lennon shows, the philosophies of Merleau-Ponty and
4 Introduction
Wittgenstein are very close on this topic. The notion of the Gestalt is helpful
to capture how seeing the psychological written on the body is irreducible
to mere physiology, without thereby being something internal. Crucially, a
Gestalt is nothing over and above the elements in which it is realized, but
at the same time, cannot be reduced to those elements. This is aptly dem-
onstrated by a figure like the duck-rabbit. It can be seen as either a duck or
a rabbit, but this switch in perceptual meaning occurs without any change
in the lines and shapes that make up the diagram. Yet the perceptual mean-
ing is not something extra to those elements either; take away the lines and
shapes, and no perceptual meaning persists. Similarly, a body’s expressing
sadness is manifest in its gestures and interactions with the world, but it is
not reducible to a series of brute physiological facts. Both writers also con-
nect grasping the Gestalt of an expressive body with its soliciting a mean-
ingful response from the perceiver. We are not detached observers of others;
we interact with them, sharing a way of life against the background of a
common world. Both Merleau-Ponty and Wittgenstein are interested in the
expressive power of art, which both see as extending the expressive ges-
tures of the body. They emphasize the fact that—like bodily gestures more
­generally—artistic expression gains its significance from its place in a shared
way of life. Meaning is generated intersubjectively. Finally, Merleau-Ponty
employs the notion of expression to characterize the world—something that
goes beyond anything we find in Wittgenstein’s thinking. He holds that the
world itself is expressive. It is characterised by Gestalten, invites responses
from us, and is expressed by the gestures we perform in answer to these
invitations—gestures that help give shape to the world. Similarly, the idea of
expression is also key to understanding our ontological status for Merleau-
Ponty. We are beings that weave together the inner and outer, the physical
and psychological self and world.
Søren Overgaard takes up one of the themes touched upon by Lennon’s
piece. He considers how we come to know about other people’s psycho-
logical lives and relates ideas from Wittgenstein and Merleau-Ponty to
an important contemporary debate in this area. The ability to discern
how things are with another subject is often called ‘mindreading’. There
are three central theories that seek to explain how we mindread others:
‘theory-­theory’, which claims we do this by employing a theory that con-
nects observable behaviour with psychological states; ‘simulation theory’,
which states that we do this by putting ourselves in the other’s position to
see what we would think or feel in his or her circumstances and then ascribe
this to him or her (this can either be a conscious activity of imagining or a
subpersonal process); and ‘perception theory’, a relative newcomer to the
debate, which claims that we just perceive how the other is feeling or what
he or she is thinking. Proponents of perception theory often cite Wittgen-
stein and ­Merleau-Ponty as forerunners of this account. Overgaard particu-
larly discusses the work of Shaun Gallagher in this regard. As Lennon shows
in her paper for this volume, both Merleau-Ponty and Wittgenstein argue
Introduction  5
that we immediately perceive others’ bodies as expressing their psychologi-
cal lives. However, Overgaard argues that perception theorists move too
quickly from the recognition that minds are not hidden but perceptible fea-
tures of the world, to claim that theories that posit a role for inference
in mindreading are thereby redundant. Such a move depends on a thesis
explicitly rejected by both Merleau-Ponty and Wittgenstein, which Over-
gaard dubs ‘the myth of the given mind’. Whilst we perceive others’ psy-
chological states as expressed by their bodily gestures, we see them in a
different sense to the way in which we see something like colour. To perceive
another’s gesture as expressing her psychological life is an achievement, and
one that must be explained. Merleau-Ponty and Wittgenstein both take this
to be made possible by our insertion in a particular socio-cultural milieu—a
particular community that reciprocally recognizes each other’s expressive
gestures. Theorists like Gallagher are thus wrong to suppose that accounts
like theory theory and simulation theory are instantly made redundant by
the recognition that others’ psychological lives are perceptible.
Merleau-Ponty and Wittgenstein both seek to provide an alternative
notion of the self to the dominant Cartesian view of a disembodied individ-
ual, reaching universal truths through deployment of rational principles.
Merleau-­Ponty’s thoroughgoing reconceptualization of the body springs
to mind most readily in this regard. But, he also emphasised the irreduc-
ibly social nature of the self—something to which Wittgenstein also paid
a great deal of attention. However, it has been suggested that an account
that takes the self to be shaped by its socio-cultural environment cannot
accommodate the possibility of political change. This seems to be implied
by Wittgenstein’s view. Chantal Bax considers the ideas he puts forward in
On Certainty. There he argues that certain propositions, such as that the
earth exists, are not things we know because they are not things we can jus-
tify. Moreover, they seem more certain than other propositions, and they
are not usually things we call into question or debate. These certainties, or
hinge propositions, are instead things we have to take for granted in order
to gain any knowledge. They form the background to our engagement
with the world. Moreover, we gain them as children via training, and often
they are never explicitly entertained. Wittgenstein also suggests that we can
only interact with other people on the basis of a shared number of certain-
ties. It is clear how this implies that the self is reactionary. Since certain-
ties are not explicitly entertained, are difficult to question, and have to be
taken for granted to interact with the world and other people, the self does
not seem to be capable of calling into question the set of certainties that
ground her engagement with the world, and so these seem to be beyond
her powers to alter.
Merleau-Ponty also takes our engagement with the world to be founded
on a socio-cultural background that is taken for granted and something of
which we are only implicitly aware. But his account makes it clear how the
subject can actively change her socio-cultural context. Her situation neither
6 Introduction
causes her to behave in certain ways, nor is this a matter of explicit judge-
ments. Instead, her situation motivates her but in a way that is indeter-
minate. The facts of her situation solicit her to act in certain ways, but a
number of courses of action are consistent with what is suggested by her
socio-cultural context. Having acted in one way, this can bring other pos-
sibilities to light, which can motivate further actions. Chantal Bax describes
Merleau-Ponty’s example of how a member of the proletariat can come to
class-consciousness. He may first hear of workers in another factory gain-
ing a higher wage as a result of a strike. The possibility of striking himself
becomes salient for him. As a result, similarities between his life and those
of labourers elsewhere begin to show up for him, and he begins to see pos-
sibilities for solidarity. In this way, he can come to see himself as part of
an oppressed class. Finally, Chantal Bax shows how some of Wittgenstein’s
remarks resonate with this picture of the self and its relation to its socio-
cultural environment. He allows that certainties do not form an orderly set.
There are gaps and inconsistencies, which allow for the possibility of change.
Since both Wittgenstein and Merleau-Ponty emphasize the constitutive
role played by the community in our psychological capacities, it is perhaps
surprising that they take solipsism—the view that the self is, in some sense,
alone—seriously. I address this aspect of their writing in my contribution
to this volume. There are different forms of solipsism. One might hold
that the self is completely alone in a world of non-selves or that it is only
one of a special kind of self—a god amongst mortals. One might take such
claims to be about what really exists; or claims concerning what one can
know; or claims about the character of experience. I take both Merleau-
Ponty and Wittgenstein to be interested in the latter, phenomenological sort
of solipsism. My discussion begins with something of a puzzle raised by
­Merleau-Ponty’s discussion of solipsism in Phenomenology of Perception.
One might think that his conception of consciousness as embodied means
that we simply perceive other selves and so solipsism is just not an issue for
him. However, as Lennon and Overgaard note in their papers for this vol-
ume, it does not follow from the mere fact that minds are perceptible that
we actually perceive them as minds. Neither does it follow from the fact
that other human beings are selves just like me that I perceive them as the
same kind of subject as me. Thus he holds that, in spite of his account of
embodied consciousness, solipsism is still an issue for him. Both Merleau-
Ponty and Wittgenstein take there to be an important connection between
having a perspective on the world and solipsism. Merleau-Ponty says little
about this, but Wittgenstein offers a detailed description of it. He suggests
that the self is a bare point of view on the world and as such, cannot be
identified with any item in it. This includes one’s own body. A bare per-
spective is not an inner Cartesian realm. It is not the sort of thing that can
own mental states, at all. Thus one’s psychological life must be thought of
as happening in the world. It is something on which one has a perspective,
but it is no more ‘owned’ by me than other worldly events. This picture is
Introduction  7
solipsistic because the world is presented as revolving around me. As such,
it is my world—the disembodied, godlike perspective at its centre. Wittgen-
stein’s description of solipsistic experience has a striking resonance with
some of the experiences that characterise schizophrenia. However, it cannot
stand on its own, since it does not offer a description of non-schizophrenic
experience. Merleau-Ponty’s analysis of bodily experience provides what is
missing from Wittgenstein’s account. The non-schizophrenic subject does
not experience herself as a bare perspective but as a bodily point of view
on the world. This is possible because there is a form of bodily awareness
that is ‘adverbial’ in structure and does not present the body as an object on
which the subject has a perspective. Second, the momentary transfer of the
body schema allows the subject to experience the world as revolving around
other subjects, thus decentring her own perspective, and allowing her to
experience the world as one shared with other similar selves.
Parts of Merleau-Ponty’s and Wittgenstein’s respective projects of disman-
tling what they show to be distorted views of the world centre on the notion
of representation—something that is central to understanding the mind and
its relation to the world. Taylor Carman’s piece for this volume explores
the nature of representation through the idea that there is always a surplus
of meaning over and above the representational content. Both Merleau-
Ponty and Wittgenstein put forward this claim. Merleau-Ponty’s discussion
focuses particularly on pictures or images, whilst Wittgenstein’s distinction
between saying and showing indicates that even symbolic forms of represen-
tation such as language have a meaning—show something—beyond what
they say or depict. Whilst Wittgenstein and Merleau-Ponty do not give us
quite the same idea of this surplus, bringing them together like this reveals
something interesting. Merleau-Ponty’s focus on the residue of meaning
found in images directs our attention to the fact that whilst they may have
conventional aspects, they are not purely conventional in the same way as
language. It is arbitrary that English speakers use the noise ‘cat’ to represent
a certain sort of feline being, but whilst there are many different ways to
picture a cat, not just any arrangement of lines on a page will do. Yet at the
same time, images are not of their objects in virtue of resembling them, if
this is understood as sharing a sufficient number of properties. A portrait
of Toussaint Louverture is a flat arrangement of coloured paint on a can-
vas, but the man himself was a flesh-and-blood, living being. Images are
nevertheless likenesses of their objects in that perceivers see them as alike.
Thus a statement of what an image is of, or represents—for example, Tous-
saint Louverture riding into battle—can never capture what is essential to
images: the visual presence of their objects. Despite this difference between
images and language, Wittgenstein’s idea that language shows something
beyond what it says or represents points to an important way in which
pictures and language are similar. Carman suggests that this is captured by
Merleau-­Ponty’s thesis that the different types of representation lie on a con-
tinuum with perception itself. All are forms of expression. Even perception
8 Introduction
expresses the world because it is not the passive recording of data, but the
manifestation of the body’s ‘apprehension’ of its environment. Types of
representation differ in their degree of closeness to perceptual engagement
with the world. Images are very close to perception and so can conjure up
visual presence in a way that language cannot. But even more remote forms
of expression, such as non-poetic language, are nevertheless rooted in our
bodily engagement with, and perceptual grasp of, the world, and so retain a
surplus of meaning over and above what they represent.
The notion of representation is closely connected to another important
topic discussed by Merleau-Ponty and Wittgenstein: indeterminacy. David
Cerbone considers their treatment of this issue. Both writers draw attention
to indeterminacy at the heart of our meaningful engagement with the world.
Neither experience nor language admit of the sort of determinacy we usu-
ally expect to find there. The foil to Cerbone’s discussion of Merleau-Ponty
and Wittgenstein is Quine’s thesis of the indeterminacy of translation—one
of the most influential treatments of the indeterminacy of meaning in ana-
lytic philosophy. Quine argues that the linguistic behaviour of any commu-
nity is compatible with two or more incompatible manuals translating their
language. Moreover, there is no fact of the matter concerning which manual
is the correct one. Quine takes this indeterminacy to show that issues to
do with meaning are not about properly objective matters of fact. This has
wider repercussions for questions to do with the mind itself. Quine argues
that ‘intentional idioms’ such as ‘x believes that p’ fall short of the gold stan-
dard of scientific theorizing. Against the background of this thesis, Merleau-
Ponty and Wittgenstein’s ideas come sharply into view. Like Quine, both
thinkers point to what Cerbone calls ‘the fact of indeterminacy’ in our per-
ceptual experience, our language, our meaningful behaviour, and so on. But
unlike Quine, Merleau-Ponty and Wittgenstein do not take indeterminacy
to be a negative feature of our meaningful engagement with the world—
a falling-short of some more exacting ideal. Instead, both Merleau-Ponty
and Wittgenstein critique the pictures that hold such an ideal before us as
fundamentally flawed. Two central examples are Wittgenstein’s discussion
of rule-following, where he argues that a completely determinate language
would require an infinite set of rules specifying its use. To demand deter-
minacy is thus to require something impossible and something that we do
not need as we get along perfectly well without it. Merleau-Ponty connects
indeterminacy in perceptual experience with having a point of view. Perceiv-
ing the world from a location within it means that at any particular time,
parts of it will be hidden from view and so experienced as there but indeter-
minate. The demand that experience present a fully determinate universe is
destructive of perspective and so of experience itself. Both writers connect
indeterminacy with flexibility or variability. They point out that the inde-
terminacy displayed by meaningful behaviour gives it a flexibility that dis-
tinguishes it from activity that is purely mechanistic. Wittgenstein describes
the variability of the expressive body as its being ‘full of soul’ (i.e., its being
Introduction  9
shot through with subjectivity). Merleau-Ponty makes essentially the same
connection. Mechanism—the strict following of circumscribed rules with no
variation—is the antithesis of expression, which (Lennon discusses in this
volume) is essentially creative and so admitting of variation, even as it makes
use of already instituted meanings, which have somewhat determinate rules
for their usage. It follows that, for both Merleau-Ponty and Wittgenstein,
indeterminacy is a constitutive feature of our meaningful engagement with
the world. It is not the falling short of an ideal but that without which there
could be no meaning.
The final paper by Thomas Baldwin takes up some of the themes
touched on by other papers in this volume. His central concern is to trace
the points of contact and departure between Wittgenstein’s account of the
certainties—what are sometimes called, ‘hinge propositions’—that under-
pin our everyday engagement with the world, and Merleau-Ponty’s notion
of perceptual faith—his claim that the world in general is something of
which we can be absolutely certain. A striking point of comparison between
the two thinkers is the peculiar status they both accord to that which is
certain. For Wittgenstein, hinge propositions are not items of knowledge in
the usual sense, since they lack justification. But this does not mean they are
just dogmatically asserted either. They make possible our everyday engage-
ment with the world. Since they must be presupposed for even the sceptic’s
doubt to be cogent, they cannot themselves be doubted. Merleau-Ponty also
advances a position that lies between these two alternatives. He argues—
contra ­dogmatism—that experience of any particular thing is fallible. But the
sceptic’s doubts only make sense against a background of contact with the
world, which is likewise presupposed by any thought or action as the condi-
tion of its possibility. However, despite these and other similarities between
their accounts, there are, nevertheless, some important differences. Wittgen-
stein’s picture is ultimately holistic, despite the role played by hinge proposi-
tions in the system. These are not completely beyond r­ evision—something
Chantal Bax emphasizes in her piece for this collection. ­Merleau-Ponty, on
the other hand, sees the body’s non-thetic, pre-predicative understanding of
the world as the foundation for conceptual thought. This opposition is miti-
gated somewhat by Wittgenstein’s conception of non-linguistic behaviour
as playing a foundational role for the first languages and Merleau-Ponty’s
account of a dialectical interplay between conceptual thought and bodily
A more significant disagreement concerns the issue of idealism. Merleau-
Ponty’s position is ultimately a form of transcendental idealism. The lived
world that is presented in experience, the only one we can know about,
comes into being through perception. The perceiving subject thus plays a
constituting role with respect to the world. Like Kant, Merleau-Ponty’s
account allows for a notion of objective empirical reality, which he secures
on the basis of the claim that there is a commonality to adult human experi-
ence. Indeed, Baldwin reads him as asserting the superiority of the mature
10 Introduction
experience of civilised people, which functions as the standard from which
the experiences of children, the sick, animals, and so on, deviate—a move
that is problematic. Unlike Kant’s transcendental ego, however, Merleau-
Ponty’s constituting subject is bodily. (It should be noted that this is only
one aspect of Merleau-Ponty’s position. As a part of the world, the embod-
ied subject is also constituted in perception. His ultimate account of the
world and subject sees them as mutually constituting moments of a single
existence that bifurcates in perception.) Wittgenstein’s work, on the other
hand, contains nothing akin to transcendental idealism. Baldwin traces
ways in which Wittgenstein’s later thinking endorses a form of naturalism
and as such, stands in stark contrast to Merleau-Ponty’s account. In the end,
this difference in their ontological stances marks an important difference
between them.

1 The issue of Wittgenstein’s phenomenology is controversial. It has been discussed
by various authors in addition to Spiegelberg (1981), e.g., Gier (1981), Hintikka
(1996), Park (2012), Monk (2014), and Vrahimis (2014).

Gier, N. 1981. Wittgenstein and Phenomenology: A Comparative Study of the Later
Wittgenstein, Husserl, Heidegger, and Merleau-Ponty. Albany: State University
of New York Press.
Hintikka, J. 1996. The idea of phenomenology in Wittgenstein and Husserl. In
J. Hintikka (ed.), Ludwig Wittgenstein: Half-Truths and One-and-a-Half-Truths.
Dordrecht: Kluwer Academic Publishers, pp. 55–77.
Monk, R. 2014. The temptations of phenomenology: Wittgenstein, the synthetic a
priori and the ‘analytic a posteriori’. International Journal of Philosophical Stud-
ies 22 (3): 312–340.
Park, B. C. 2012. Phenomenological Aspects of Wittgenstein’s Philosophy. Dor-
drecht: Springer.
Spiegelberg, H. 1981. The puzzle of Wittgenstein’s Phänomenologie (1929–?). In
H. Spiegelberg (ed.), The Context of the Phenomenological Movement. The
Hague: Nijhoff, pp. 202–28.
Vrahimis, A. 2014. Wittgenstein and the phenomenological movement: Reply to
Monk. International Journal of Philosophical Studies 22 (3): 341–348.
1 Wittgenstein and
Merleau-Ponty on
Gestalt Psychology1
Katherine J. Morris

Most famous for their researches on perception (and visual perception in

particular), the Gestalt psychologists developed their theories, mainly in
Germany, over roughly the same period as Husserl was developing phenom-
enology, and to some degree in dialogue with Husserl;2 this was also the
period of Wittgenstein’s ‘early’ philosophy. Both Wittgenstein (in his ‘later’
period)3 and Merleau-Ponty (throughout his philosophical career) engaged
directly with Gestalt psychology. Wittgenstein read rather little, although
things that interested him for whatever reason he read in great depth. This
category includes Köhler’s Gestalt Psychology (1947, hereafter GP),4 to
which Wittgenstein came late in his career;5 he discussed Köhler extensively
in his last lectures in 1947 (students’ notes on these lectures are published
as Lectures on Philosophical Psychology 1946–7).6 Schulte goes so far as to
say that GP ‘was the single most important influence on Wittgenstein during
those years’ (1993: 76). Wittgenstein’s Köhler period was roughly contem-
poraneous with Merleau-Ponty’s Phenomenology of Perception. Merleau-
Ponty was much more of a typical academic than Wittgenstein; he read
widely in the psychology of his day, and he regularly characterised Gestalt
psychology as ‘the new psychology’, as opposed to the empiricist and intel-
lectualist psychology that had dominated previously.
Because Köhler is the main common denominator in their reading about
Gestalt psychology, all of the references to Gestalt psychology in this
essay will be to Köhler and most to the book Gestalt Psychology. I will
focus on Wittgenstein’s and Merleau-Ponty’s critiques of Gestalt psychol-
ogy’s descriptions of the perceived world.7 At first sight, Wittgenstein’s and
Merleau-Ponty’s responses to these descriptions seem almost diametrically
opposed. Wittgenstein was clearly intrigued by some of the phenomena to
which the Gestalt psychologists drew attention but thought their descrip-
tions of these phenomena confused. Merleau-Ponty, by apparent contrast,
refers to the Gestalt psychologists as ‘the very psychologists who described
the world as I did’ (PrP 23); his primary objection is Gestalt psychology’s
failure to see the revolutionary consequences of its own discoveries. I will
ultimately suggest that this appearance of diametrical opposition is mislead-
ing. Section 1 outlines GP’s descriptions of the perceived world. Sections 2
and 3 bring out Wittgenstein’s and Merleau-Ponty’s principal objections.8
12  Katherine J. Morris
Section 4 tries to bring Wittgenstein and Merleau-Ponty into dialogue with
one another. The final section (5) reflects on this dialogue by way of drawing
a few conclusions.

1.  Gestalt Psychology’s Descriptions of the Perceived World

A great deal of GP is polemical, arguing against two strands of empiricist
psychology, which then dominated the field: behaviourism (about which
I can say no more in the present context) and what was known as intro-
spectionism, whose basic premise is ‘the all-important distinction between
sensations and perceptions, between the bare sensory material as such and
the host of other ingredients with which this material has become imbued
by processes of learning’ (GP 43). On this view, the sensation is ‘the genuine
sensory fact’, by contrast with those ‘mere products of learning’ (GP 44).
When we add to this move a bit of basic information about physics, optics,
and the physiology of seeing, it can seem irresistible to assert that the ‘genu-
ine sensory fact’ causally depends solely on the image on the retina, and that
everything else we say we see isn’t strictly speaking seen, since it brings in
further knowledge. Note, just for example, that this would imply that we
cannot strictly speaking see depth (distance, three-dimensionality), since the
image on the retina is two-dimensional.
Gestalt psychology’s conception of perception stands in sharp contrast to
this. The most basic notion in Gestalt psychology is that of organisation,
which, Köhler argues (GP 81), empiricism cannot accommodate.
This term ‘organisation’ points us in the direction of what the Gestalt
psychologists call ‘circumscribed units’: ‘In most visual fields the contents
of particular areas “belong together” as circumscribed units from which
their surroundings are excluded’ (GP 80–1), for example, ‘things: a piece
of paper, a pencil, an eraser, a cigarette, and so forth’ (GP 81).9 ‘Gestalten’
include both these segregated units and ‘groupings’ of such units: ‘a given
unit may be segregated and yet at the same time belong to a larger unit’
(GP 84; cf. GP 95, 120). For instance, in Figure 1.1, each patch is itself a
segregated unit, but the patches are also grouped into two groups of three
patches each, as ‘everyone beholds’ (GP 83).
The groups consist of the three on the left and the three on the right, even
though they could equally just be six patches, or the two groups consisting
of the three on top and the three below, etc.

Figure 1.1
Wittgenstein and Merleau-Ponty on Gestalt Psychology  13

Figure 1.2

Of particular interest to Wittgenstein are what the Gestalt psycholo-

gists call ‘reversals’ in ambiguous figures (i.e., figures which spontaneously
admit of different organisations). Consider Figure 1.2: whereas we might at
first see ‘an object formed by three narrow sectors’, we may ‘suddenly see
another pattern. Now the lines which belonged together as boundaries of a
narrow sector are separated; they have become boundaries of large sectors.
Clearly, the organisation of the pattern has changed’ (GP 100). Thus (pace
the introspectionists) ‘[w]ith a constant pattern of [retinal] stimuli, we may
see . . . two different shapes’ (GP 107).
Of especial importance to the phenomenologists are various charac-
teristics of gestalten, especially those characteristics that contribute to
the unity of the gestalt. According to the Gestalt psychologists, perceived
objects maintain their apparent size, shape, colour, etc.—and hence their
unity—through variations in distance, orientation, lighting, and so on.
Hence they will speak of various ‘constancies’: ‘size constancy’, ‘shape con-
stancy’, ‘colour constancy’, and so on. (Note that the introspectionists can-
not allow such constancies, since, for example, the projection of a circular
object such as a plate on the retina will vary in shape if the object is tilted.)
Also important to the phenomenologists are such essential characteristics
of gestalten as figure-ground structure (which, again, is impossible for the
introspectionists to accommodate). A ‘circumscribed unit’ (the figure) is
segregated from its surroundings (the ground), the figure having ‘the char-
acter of solidity or substantiality’, the ground being ‘loose or empty’ and
‘unshaped’ (GP 120). (Figure 1.2 above displays a figure-ground ambigu-
ity: as long as we are seeing the cross with the slender arms, the area of the
cross with the wider arms ‘is absorbed into the background, and its visual
shape is non-existent’, (GP 107); in addition, ‘the oblique lines are the
boundaries of the shapes which are seen at the time’, i.e., in this case they
‘belong to’ the slender cross, (GP 108, etc.) )
14  Katherine J. Morris
2.  Wittgenstein’s Criticisms of Köhler’s Descriptions
Wittgenstein’s best-known discussion of the Gestalt psychologists’ descrip-
tions occurs in his treatment of ‘aspect-seeing’ in PI II §xi.10 His initial focus
is on ‘reversals’ of ambiguous figures; Wittgenstein reproduces a number
of such figures, including most famously the ‘duck-rabbit’ (from Jastrow
1899), but also the ‘double cross’ (PI II 207), which is similar to Figure 1.2
above. Wittgenstein speaks of the different organisations of these figures
as ‘aspects’ and of reversals as ‘changes of aspect’. What appears to strike
Wittgenstein about reversals is a kind of paradox, which he expresses thus:
‘One would like to say: “Something has altered, and nothing has altered” ’
(RPP I §966).11 This ‘paradox’ is essential to a change of aspect.

W1.  Two (or Three?) Uses of the Word ‘See’

Wittgenstein is as resistant as Köhler to the empiricist temptation to say we
don’t (strictly speaking) see aspects: ‘if someone wanted to correct me and
say I don’t really see [these things, but only shapes and colours], I should
hold this to be a piece of stupidity’ (RPP I §1101).12 Wittgenstein’s central
objection to Köhler is that he supposes that we see organisation in the same
sense that we see colour and shape (cf., e.g., RPP I §1023). Not only does
this involve a conflation, it means that Köhler can neither express nor dis-
solve the ‘paradox’.
Köhler’s view apparently implies that what one sees changes in a change
of aspect in the same sense that what one sees would change were the
colours and shapes to alter. (A change in organisation ‘amounts to an actual
transformation of given sensory facts into others’ (GP 99).) Wittgenstein
clearly takes there to be a powerful temptation to say this (RPP I §534, cf.
RPP I §535, §536, §1107). However, to give in to this temptation is to leave
oneself unable to recognise the paradox which is of the essence of aspect
change: there is no sense, in this case, in which everything remains the same.
Wittgenstein claims that to ‘put the “organisation” of a visual impression
on a level with colours and shapes’ is to proceed ‘from the idea of the visual
impression as an inner object’ (PI II 196); indeed this idea is commonly
seen as the real target of these remarks.13 From this perspective, he imagines
Köhler saying that the ‘outer picture’ (e.g., the drawing of the duck-rabbit)
has remained the same while the ‘inner picture’ has changed. However, this
would get him nowhere: one ought to be able to represent the ‘inner picture’
with a drawing (an ‘outer picture’), since the notion of an inner picture is
modelled on that of an outer picture; but if one tried to represent what the
duck-rabbit was like before the change of aspect and what it is like now
simply by making an exact copy of what one sees in each case, ‘no change is
shewn’ (PI II 196; cf. RPP I §1041; PI I §196).
Wittgenstein asserts that the phenomena of aspect-seeing and change of
aspect point to ‘the difference of category between the two “objects” of
Wittgenstein and Merleau-Ponty on Gestalt Psychology  15
sight’ (PI II 193). Here, the two ‘ “objects” of sight’ are not the ‘before’ and
‘after’ of an aspect-change as they were for Köhler; rather, they correspond
to ‘[t]wo uses of the word “see” ’, which he exemplifies thus: ‘ “I see this”
(and then a description, a drawing, a copy)’ as opposed to ‘I see a likeness
between these two faces’ (PI II 193; cf. RPP I §964). How do these exem-
plars relate to changes of aspect? We have seen already that any attempt to
represent the ‘before’ and ‘after’ in a change of aspect with a drawing or
copy would show no difference; rather, one might (for example) convey the
change by first grouping the drawing with a number of other drawings of
ducks, and then with a number of drawings of rabbits (cf. PI II 196–7),14
and drawing attention to the likenesses in each case. Once we have distin-
guished these two uses of the word ‘see’, the apparent paradox vanishes:
what you see in the first use of the word ‘see’ does not change, whereas what
you see in the second use changes.
Importantly, the distinction, or at least some distinction, between differ-
ent uses of the word ‘see’ has application beyond the rather special case of
ambiguous pictures, even if other cases do not involve the ‘paradox’; since
all perceived objects (not just pictures, much less just ambiguous pictures)
are, according to the Gestalt psychologists, organised, many of the same
issues arise for figure-ground organisation, three-dimensionality and so on.
Wittgenstein clearly recognises this; thus RPP I §1023 reflects on Köhler’s
commitment to the idea that ‘object’ (figure) and ‘ground’ are ‘visual con-
cepts like red and round’ (cf. RPP I §1118: ‘Indeed, you may well say: what
belongs to the description of what you see, of your visual impression, is not
merely what the copy shews but also the claim, e.g., to see this “solid”, this
other “as intervening space” ’). Wittgenstein suggests that if one were to ask
what, in a drawing, ‘corresponded to the words “object-like” ’, the answer
would be ‘the sequence, the order, in which we made the drawing’, which is
surely not in the drawing in the sense that the colours and shapes are (RPP
I §1023). Again, RPP I §85 considers the question of whether depth can
‘really be seen’; he goes on to suggest that the sense in which we see colour
and shape and the sense in which we see depth are different senses, the one
perhaps to be represented ‘using a transparency’, the other ‘by means of a
gesture or profile’ (cf. also RPP I §86).
It is not entirely clear whether the sense in which we see depth, or see a
thing ‘as a thing’ or as ‘object-like’, is the same as the sense in which we see
aspects, thus it is not clear whether we should be speaking of a third use of
‘see’ here. For example, in the case of aspect-seeing, Wittgenstein’s distinc-
tion between two uses of ‘see’ aims to capture ‘what remains the same’ as
opposed to ‘what changes’. He uses the phrase ‘shapes and colours’ repeat-
edly to allude to the first sense, as a kind of shorthand for ‘what remains
the same’. When it comes to its application to seeing shapes and colours as
opposed to seeing a thing ‘as a thing’ or as three-dimensional, can this phrase
still be understood as a shorthand for ‘what remains the same’? And is seeing
a thing as a thing or as three-dimensional ‘subject to the will’, as Wittgenstein
16  Katherine J. Morris
asserts that aspect-seeing is? (Wittgenstein presents this as a further criticism
of Köhler: he allegedly ‘does not deal with’ (RPP I §971) the fact that ‘an
aspect is subject to the will’, i.e. that ‘it makes no sense to say “See it red”;
whereas it does make sense to say “See it as . . .” ’ (RPP I §899)—also to ask
someone to try to see it as . . .—which ‘touches the essence’ of aspect-seeing
(cf. RPP I §976), and thereby sheds light on the second sense of ‘see’.) In
fact, the ‘analytical attitude’ (GP 99; cf. RPP I §§1110–11), introduced in
the next section, may be understood as ‘subject to the will’; if it is what we
adopt in order to see ‘shapes and colours’ as opposed to ‘seeing a thing as a
thing’ or ‘seeing it as three-dimensional’ (as we do in the ‘normal attitude’),
then seeing in the first sense is ‘subject to the will’ (although we still cannot
say ‘See it red’, we can say ‘See it as shapes and colours as opposed to seeing
it as a thing’). As Wittgenstein rightly says, ‘[t]he concept of “seeing” makes
a tangled impression’ (PI II 200)! I will refer to the use in which we can say
that we see depth, or see a thing ‘as a thing’, as a third use, but the tangled-
ness of the concept must always be borne in mind.

W2.  Seeing Meanings

Consideration of a further related objection will enrich our understanding
of the second and third uses of ‘see’: ‘It is—contrary to Köhler—precisely
a meaning [Bedeutung] that I see’ (RPP I §869; the context suggests that
he is here talking about his second use of ‘see’, but I think that he would
extend this to the third use).15 What is Wittgenstein’s target here? As part
of his polemic against empiricist psychology, Köhler argues that ‘sensory
units’ (gestalten) ‘existed as units’ before the addition of ‘meaning’, i.e.,
before acquiring names, rich symbolism, or practical uses; indeed existence
as a ‘sensory unit’ is a condition for the possibility of ‘meaning’ (GP 82).
His point is that, pace the empiricist psychologists, gestalten cannot be
explained away as the product of learning such meanings.
I don’t think that Wittgenstein means this by ‘meaning’ in this context;
thus it may be said that this objection to Köhler misses its target. None-
theless I am inclined to think that he sees the word ‘meaning’ as captur-
ing something important: he wants Köhler to use the word ‘meaning’ more
widely than he does, and this is his objection. What does the word ‘meaning’
capture here for Wittgenstein? In part taking my cue from Wittgenstein’s
suggestion that there is a ‘connection between the concepts of seeing an
aspect and of experiencing the meaning [Bedeutung] of a word’ (PI II 214),
I submit that ‘meaning’ is what is missing from the lives (PI II 214; cf. RPP
I §202, LPP 181f.) of those (hypothetical, in Wittgenstein’s view) individuals
who are ‘meaning-blind’, ‘aspect-blind’,16 or ‘gestalt-blind’ [Gestaltblinde]
(RPP II §478, also at RPP I §170 where it is translated as ‘form-blind’).17
Wittgenstein notes that ‘the names of famous poets and composers seem
to have taken up a peculiar meaning into themselves. So that one can say:
the names “Beethoven” and “Mozart” don’t merely sound different; no,
Wittgenstein and Merleau-Ponty on Gestalt Psychology  17
they are also accompanied by a different character’ (RPP I §243). (‘It is as
if the name together with these works, formed a solid whole . . . The name
turns into a gesture; into an architectonic form’, (RPP I §341; cf. LPP
167).) The ‘meaning-blind man’ (Bedeutungsblinder), however, ‘would
not feel that the names . . . were distinguished by an imponderable Some-
thing’ (RPP I §243; cf. PI II 214; RPP II §571). He would be similar to
‘those for whom spelling is just a practical question’, who are insensitive
to changes in the orthography of a word (RPP II §572, cf. Z §184), or
who lacks the sensibility expressed here by Grillparzer: ‘I cannot describe
the dreadful impression which the h in the English word ghost makes on
me’ (quoted in Schulte 1993: 69). Again, ‘normal’ people, when listening
to someone reading a poem out loud, will attend to intonation, and will
take note if the reader pronounces a word with the wrong intonation so
that it ‘stands out too little or too much’ (PI II 214); if the person is read-
ing the poem himself and gets the intonation wrong, we may say to him:
‘You must hear this word as . . . then you will say the sentence properly’.
The meaning-blind man is unable to benefit from this instruction (RPP
I §247); Wittgenstein explicitly compares this to directions for playing a
passage of music (cf. RPP I §350). Overall, the impression made by the
meaning-blind man might be described as ‘prosaic’ (RPP I §342). Again,
the aspect-blind person is someone who ‘never sees anything as anything’
(cf. RPP II §478, cf. PI II 213), who can ‘gather various things about
the landscape’ from a photograph but cannot exclaim ‘What a glorious
view!’ (RPP I §168); we might ‘picture him as making a less lively impres-
sion than we do, behaving more “like an automaton” ’ (RPP I §198)—‘as
it were sleep-walking’ (RPP I §178). These points, once again, extend
well beyond aspect-seeing. Wittgenstein imagines a man with ‘the talent
to copy objects . . . very exactly, and yet he might keep on making small
mistakes against sense [Sinn]; so that one could say “He doesn’t grasp an
object as an object” ’ (RPP I §983; cf. RPP I §423, RPP I §978) (i.e., he
is ‘gestalt-blind’).

3.  Merleau-Ponty’s Criticisms of Köhler’s Descriptions

Despite his praise for Gestalt psychology’s descriptions of the perceived
world, Merleau-Ponty charges it with a failure to draw the correct conse-
quences from its own discoveries, and consequently with nothing less than
misconceiving the perceived world, the body, and the relations between
them.18 I will develop two strands of this critique.

MP1.  Persistent Dualisms

The issue here has to do with what to make of empiricist discoveries about
perceptual experience within what Gestalt psychologists call the ‘analyti-
cal’ (as opposed to the ‘normal’) attitude in perception (e.g., cf. GP 47).
18  Katherine J. Morris
To persuade us that we really do see what they say we see, the empiricist
psychologists ask us to attend to what we are looking at in particular ways,
often with the assistance of various laboratory techniques and apparatus.
So, for example, to ‘show’ that we really see a stack of paper as a multi-
plicity of shades of grey rather than the uniform white which we claim, or
that a circular plate seen at an angle really looks elliptical, they will ask us
to squint our eyes or to look at the paper or the plate through a cardboard
tube or a hole in a screen. Similar procedures are used to persuade us that
a man really does look smaller as he gets further away. This last case also
feeds into the perception of depth, distance or relief, which (according to
the empiricists) we don’t strictly speaking see; rather, we have learned that a
man who looks smaller, whose perception involves less ocular convergence,
and between whom and us there are many intervening objects, etc., is fur-
ther away. Köhler argues that the introspectionist unjustifiably reads the
experience we have while adopting ‘the analytical attitude’ into that which
we have when taking ‘the normal attitude’ in perception. In fact, ‘the experi-
ences revealed by introspection depend upon the attitude of introspecting.
One cannot show that they also exist in the absence of this attitude’ (GP 52,
cf. GP 102, PhP 8, PhP 505 n.26).
The Gestalt psychologists rightly insist that we do perceive depth and
distance and that perceived sizes, shapes and colours are constant. Yet these
factors (so-called visual cues) which are revealed in the analytic attitude
(e.g., ocular convergence, apparent size, intervening objects, etc., in the case
of distance, the lighting as well as the variety of colours which show up
on the parts of the object in shadow or bright sunlight) clearly enter into
perception somehow (PhP 269). The Gestalt psychologists don’t take the
intellectualist line that we infer depth or constant colour from visual cues;
what they say is that these factors ‘are not signs, but conditions or causes’
of the perception of depth or colour, of which we are unaware (PhP 269;
cf. PhP 48).
For Merleau-Ponty, the Gestalt psychologists’ consignment of these fac-
tors to the realm of ‘conditions and causes’ signals their inability to tran-
scend the ‘either cause or (inferential) reason’ dichotomy (which he sees as
part of a more general mind/body dualism). This is where Merleau-Ponty
puts ‘motives’: the ‘visual cues’—the ‘motives’ of perception—are meaning-
ful: they don’t function causally, but nor do they function as reasons from
which we infer a conclusion. Thus he speaks of ‘the silent language whereby
perception communicates with us’, in which ‘interposed objects, in the natu-
ral context, “mean” a greater distance’ (PhP 50). Similarly, the Gestalt psy-
chologists’ inference from ‘not expressly aware’ to ‘unaware’ signals their
inability to transcend a correlative dichotomy; here Merleau-Ponty puts the
notion of being non-expressly (‘non-thetically’ or ‘non-positionally’) aware
of the motives of perception (PhP 270, cf. PhP 50: these ‘are tacitly known
to perception in an obscure form, and they validate it by a wordless logic’).
Again, ‘the decisive factor in the phenomenon of colour constancy . . . is the
Wittgenstein and Merleau-Ponty on Gestalt Psychology  19
articulation of the totality of the field, the wealth and subtlety of its struc-
tures’ (PhP 321), including in particular the structure ‘lighting-thing lighted’
(PhP 320). The articulation of the field and the lighting are, not ‘causes’ or
‘conditions’ of colour constancy, but ‘moments’ of the phenomenon of colour
constancy and internally related to it via a relation of ‘motivation’ (PhP 322).
Thus, what the empiricist psychologists’ ‘analytical attitude’ does is to
make us thetically aware of the motives of perception. It is, Merleau-Ponty
argues, crucial to normal perception that we are not thetically aware of
these motives. We have all seen the play of shadows and the play of light,
but in ordinary life, ‘it hides itself in making the object visible. To see the
object, it is necessary not to see the play of shadows and light around it’
(SNS 16). Suppose that I look at a table spread with sheets of paper, some of
which are in shadow. ‘If I do not analyse my perception but content myself
with the spectacle as a whole, I shall say that all the sheets of paper look
equally white’. But if I fix my gaze or look at the sheets in shadow through a
matchbox lid, that is, disrupt the articulation of the field, they ‘change their
appearance: this is no longer white paper over which a shadow is cast, but a
grey or steely blue substance, thick and not definitely localized’ (PhP 234–5).
Again, a feebly lighted white wall which in some sense appears white ‘to the
unhampered vision’ ‘appears a bluish-grey if we look at it through the win-
dow of a screen which hides the source of light’ (PhP 320); hiding the source
of light disrupts the ‘lighting/thing-lighted’ structure, and looking through
the screen disrupts the entire articulation of the field, since when we look
through the screen we no longer perceive ‘subordinated wholes, each with
its own distinctness, standing out one against the background of another’
(PhP 321): in the analytic attitude, ‘we no longer see real bodies, such as the
wall or the paper’ (PhP 320).

MP2.  Reconceptualizing the Body

Second, Merleau-Ponty argues that the Gestalt psychologists’ descriptions
imply a reconceptualization of the body which they never fully undertake.19
To take a particularly blatant example, for them, ‘colour constancy’ comes
about because (inter alia) ‘the eye “takes the lighting into account” ’ (Guil-
laume quoted in PhP 548 n.23). What can that possibly mean without a
reconceptualization of the body?
Merleau-Ponty argues that to understand perception, we need to be pre-
pared to speak of bodily awareness, bodily knowledge, bodily intentional-
ity, and bodily purpose (even teleology), none of which could possibly make
sense within a Cartesian conception of the body. The non-thetic awareness
referred to earlier is the body’s awareness, neither explicit nor articulate.
The body is aware of itself; for example, it is aware of the convergence
or divergence of the two eyes as they rotate in their sockets: ‘perception
and experience of one’s own body are mutually implied’ (PhP 522 n.71; cf.
PhP 211); it is also aware of the lighting, the intervening objects, etc. These
20  Katherine J. Morris
seemingly very different ‘objects’ of bodily awareness are brought together
in the idea of a ‘comprehensive bodily purpose’ (PhP 101). The two eyes
converge or diverge because the body is focusing the eyes on the near or
distant object. And focusing is a ‘prospective activity’ (PhP 241), a ‘purpo-
sive’ activity which has an aim: one focuses in order to see the thing. By the
same token, instead of my perceiving intervening objects, apparent size, and
so on, I perceive a unified object within a total spectacle, and it is the body
which ‘brings about [this] synthesis’ (PhP 241). The body’s intentionality is
‘operative intentionality’, ‘that which produces the natural and antepredica-
tive unity of the world’ (PhP 1xxxii). Thus the thing—the unified thing with
constant size, shape and colour, in focus—is ‘the goal of a bodily teleology’
(PhP 337). ‘The thing’ is what gives me the ‘best hold’ (sometimes called
the ‘maximum grip’) I can have upon the world (thus one could equally say
that ‘maximum grip’ is ‘the goal of a bodily teleology’). The body’s ‘gaze’
is precisely ‘an apparatus capable of responding to the promptings of light
in accordance with their sense’ (PhP 323); it ‘ “knows” the significance of a
certain patch of light in a certain context; it understands the logic of light-
ing’ (PhP 341), and it uses that ‘knowledge’ to aim at ‘a certain “hold” . . .
on its surroundings’ (PhP 278). In order to ‘inhabit’ the world, in order
to have the best ‘hold’ I can upon it, that world must consist of unified,
constant ‘things’, and my body is precisely the power to ‘contrive’ this, to
‘contrive’ the organization of the field: it is ‘the general power of inhabiting
all the environments which the world contains, the key to all those transpo-
sitions and equivalences which keep it constant’ (PhP 325).
Again, it is vital for normal perception (because ‘the natural and antepre-
dicative unity of the world’ is vital for our life) that the body keep its own
operations hidden, so that in the normal attitude of perception we see not
lighting but things which have their ‘own’ colour (PhP 320). To fix the gaze,
as opposed to exploring with the gaze, is to step into the analytical attitude
and hence to ‘separat[e] the region under scrutiny from the rest of the field,
interrupting the total life of the spectacle’ (PhP 325). Lighting and reflection
‘play their part only if they remain in the background as discreet intermedi-
aries, and lead our gaze instead of arresting it’ (PhP 323).

4.  Further Reflection on These Criticisms

On the face of it, Wittgenstein and Merleau-Ponty are attacking Gestalt
psychology from very different angles; but are their views really as diametri-
cally opposed as they appeared at the beginning? Let’s look back at each of
the criticisms outlined in the previous two sections.

W1.  Two (or Three?) Uses of the Word ‘See’

Merleau-Ponty pays far less attention to ‘reversals’ than does Wittgenstein.20
Still, as we noted, Wittgenstein’s distinction between various uses of ‘see’
Wittgenstein and Merleau-Ponty on Gestalt Psychology  21
applies not just to ambiguous pictures but to all perceived objects. Wittgen-
stein’s distinctions—in particular the distinction between the first and the
third uses of ‘see’—invite comparison to the distinction which both Köhler
and Merleau-Ponty draw between the analytic and the normal attitude in
perception, in particular to the way it is spelled out by Merleau-Ponty.
A striking point of comparison is the fact that both Wittgenstein and
Merleau-Ponty use painting to get at, in Wittgenstein’s case, the first use
of ‘see’, in Merleau-Ponty’s case the analytic attitude. (Remember too that
one expresses what is seen in Wittgenstein’s first sense of ‘see’ by a painting
or drawing, as opposed to a comparison, or a gesture, etc. See also RPP
I §267, §1077.) For Merleau-Ponty, a painter learns to adopt the analytic
attitude: he may hold up his paintbrush at arm’s length alongside the distant
tree in order to focus on its ‘apparent size’ and to compare its apparent size
to that of the farmer in the field. He squints so that the colours he sees ‘are
determined by the quantity and quality of reflected light’ (PhP 320). By such
means, he makes himself thetically aware of the motives of the perception
of distance and of colour constancy; he ‘interrogates’ the mountain with his
gaze, asking it to ‘unveil the means’ ‘by which it makes itself a mountain
before our eyes’; he finds ‘[l]ight, lighting, shadows, reflections, color’ which
‘exist only at the threshold of profane vision’; and his gaze ‘asks them what
they do to suddenly cause something to be and to be this thing’ (PrP 166).
Wittgenstein makes a number of scattered remarks about painting. Many
focus on colour, and in particular on the distinction between what colour
something is and what colours a painter would use on the canvas to paint
that thing: ‘This paper is lighter in some places than in others; but can I say
that it is white only in certain places and gray in others??—Certainly, if
I painted it, I would mix a gray for the darker places’ (RC III §56, cf. RC III
§171, §244). Here he—like Merleau-Ponty and the Gestalt psychologists—
is evidently reluctant to say that parts of the paper are grey (cf. RC III §246,
RPP I §442); that one would have to mix a grey in order to paint those parts
does not demonstrate that they are, although it does provide a ­reason—one
which an empiricist psychologist might use—for saying so. Again, he repeat-
edly asks us to imagine a painting being cut up into small pieces, each of
which ‘should appear as a flat colour-patch. Only together with the other
pieces does it become a bit of sky, a shadow, a highlight, a concave or con-
vex surface, etc.’ (RC III §266; cf. RC I §60, RC III §53). We might say
(and some empiricist psychologists would say) ‘that this puzzle shows us
the actual colours of the various spots in the picture’ (RC III §267; RC III
§268 clearly identifies this as a temptation that ought to be resisted). It is
significant that it took Western painters centuries to discover the motives of
perception; as Merleau-Ponty notes, a reflection on the eye, although ‘not
seen as such’, makes the face a living face, as painters eventually discovered;
without it, ‘the eye remains dull and sightless as in the paintings of the early
masters’ (PhP 322). Cf. Wittgenstein’s remark: ‘High-light or reflection:
when a child paints, it will never paint these’ (RPP I §1105).
22  Katherine J. Morris
Thus it is possible to read both Wittgenstein and Merleau-Ponty not only
as drawing parallel distinctions between uses of the word ‘see’ but as saying
that a painter must set out to see in the first sense of the word ‘see’ (to adopt
the analytical attitude so as to become expressly aware of the motives of
perception) in order to get us, the viewers, to see in the third sense of ‘see’.

W2.  Seeing Meanings

Merleau-Ponty doesn’t explicitly accuse Köhler of failing to recognise that
seeing ‘organisation’ is seeing a meaning. Rather, he simply asserts that, as a
consequence of figure-ground structure, even a simple black patch on a white
background is ‘charged with a meaning’ (PhP 4, cf. PhP 13–14). He regu-
larly characterises perception as ‘physiognomic’ (PhP 134), so that ‘the object
“speaks” and is significant’ (PhP 133). It is noteworthy that Wittgenstein uses
this same word ‘physiognomy’ in a linked way: words, he says, have a ‘physi-
ognomy’, a ‘familiar face’ (RPP I §6, RPP I §322, RPP I §328, PI II 181).
Even more striking, however, is Wittgenstein’s consideration of the various
forms of ‘blindness’: meaning-blindness, aspect-blindness, gestalt-­blindness,
etc.21 Those familiar with Merleau-Ponty cannot read Wittgenstein’s descrip-
tions of what such a person’s life would be like without thinking of the actual
abnormal individuals, especially Schneider, whose experiences Merleau-
Ponty discusses in such detail, as well as the sorts of cases presented more
recently by neurologists such as Oliver Sacks and Jonathan Coles. (Recall
that Schneider is characterised, inter alia, by the fact that ‘the world no lon-
ger has any physiognomy for him’ (PhP 134), except when he is engaged in
a familiar task: e.g., he is no longer able to ‘see the meaning’ of a fountain
pen, but must laboriously work out what object he is faced with.) Such cases
might help us to imagine in more detail than Wittgenstein was able to how
aspect-blindness or form-blindness shows up in an individual’s life.22 Just for
example, think of Wittgenstein’s description of the man with ‘the talent to
copy objects . . . very exactly’, but who ‘might keep on making small mis-
takes against sense [Sinn]; so that one could say “He doesn’t grasp an object
as an object” ’ (RPP I §983; cf. RPP I §423, RPP I §978). Wittgenstein does
not tell us what ‘a mistake against sense’ might look like (nor does he here
mention ‘the sequence, the order, in which we made the drawing’, which
in RPP I §1023 was what in a drawing corresponded to the words ‘object-
like’); but readers are invited to have a look at the videos on YouTube of Ste-
phen Wiltshire (a hugely talented and successful young artist diagnosed with
autism) engaged in drawing complex cityscapes.23 We might be persuaded,
at least by the sequence in which he draws, that he seems to treat the scene
before him as a complex patterned mosaic, not as a scene.

MP1.  Persistent Dualisms

Wittgenstein, like Merleau-Ponty, tackles many dualisms. One imagines
that he would be in broad sympathy with Merleau-Ponty’s challenges to
Wittgenstein and Merleau-Ponty on Gestalt Psychology  23
the ‘reason/cause’ and ‘expressly aware/unaware’ dichotomies; at the same
time, there is little evidence that he was aware of these as dualisms in need
of challenge. Consider this passage: ‘we often see a distant object merely as
distant and not as smaller . . . Thus we cannot say “I notice that he looks
smaller, and conclude from that that he is further away”, but rather I notice
that he is further away, without being able to say how I notice it’ (RC III
§171). Here he rejects the intellectualist, inferential-reason-giving interpre-
tation of ‘visual cues’, and it is even more evident that he would reject a
causal account as well. At the least this leaves space for Merleau-Ponty’s
‘motives’ of perception, yet there is no sign of an attempt on Wittgenstein’s
part to carve out that space himself. One might argue that his suggestion
that we are ‘unable to say’ how we ‘notice’ distance indicates an incipient
conception of our being neither ‘expressly aware’ nor ‘unaware’ of those
visual cues, but this is never spelled out. (See also RPP I §443: ‘we are
hardly ever conscious of the unclarity of the periphery of the visual field’,
cf. PhP 6.)

MP2.  Reconceptualising the Body

There is nothing in Wittgenstein that remotely resembles Merleau-Ponty’s
radical and thoroughgoing reconceptualization of the body. For all that,
there are indications throughout of his recognition of a widespread picture
of the human body as a mere object or a machine and his eagerness to chal-
lenge that picture. To take just one pertinent example, in the context of how
it comes about ‘that I see a tree standing up straight even if I incline my head
to one side, and so the retinal image is that of an obliquely standing tree’
and the (empiricist) answer: ‘Well, I am conscious of the inclination of my
head, and so I supply the requisite correction in the way I take my visual
impression’, Wittgenstein remarks: ‘. . . Well, who says that a living crea-
ture, an animal body, is a machine in this sense?’ (RPP I §918). And anyway
one could hardly say that ‘[t]he human body is the best picture of the human
soul’ (PI II 178) if one were thinking of the human body as a machine.

5.  Further Reflections and Concluding Remarks

The previous section will, I hope, have demonstrated that Wittgenstein’s
and Merleau-Ponty’s responses to GP’s descriptions of the perceived world
are by no means as radically at odds as they may have seemed at first sight.
But how close are they really?

Reflections on W1
It is easy to argue ‘a priori’ that Wittgenstein and Merleau-Ponty were ‘doing
the same thing’ by juxtaposing such Wittgensteinian methodological pas-
sages as ‘We must do away with all explanation, and description alone must
take its place’ (PI I §109) with Merleau-Ponty’s claim that phenomenology
24  Katherine J. Morris
‘is a matter of describing, not explaining or analysing’ (PhP xxi); but it is
also easy to argue ‘a priori’ that Wittgenstein and Merleau-Ponty were not
‘doing the same thing’ because what Wittgenstein describes is language, and
what Merleau-Ponty describes is experience. We suggested in the previous
section that it is possible to read both Wittgenstein and Merleau-Ponty as
drawing parallel distinctions between uses of the word ‘see’. And here, the
form which the a priori objection will take is that Merleau-Ponty isn’t really
interested in uses of the word ‘see’; he is interested in seeing, and conversely
that Wittgenstein isn’t really interested in seeing, he is only interested in how
we talk about seeing. It seems to me, first, that no such tension made itself
felt in our (very non-a priori) discussion; and it seems to me, second, that the
idea of such a tension is at least in part a product both of a narrow concep-
tion of language and of a narrow conception of experience—a conception
which would require the philosopher to focus inwards when attempting to
describe experience, and which would make language and experience exter-
nally related to one another.
Experience: There is a great deal to say here, but I will confine myself to
saying this much: that phenomenology is not introspection, and ‘describ-
ing experience’ for the phenomenologists requires focusing not inwards but
‘outwards’—that is to say, on the world (the lifeworld, and our being-in-the-
lifeworld). Wittgenstein, it may be argued, was generally suspicious of the
term ‘experience’ because he associated it with a conception of experience,
something like that held by empiricists, which saw it as something ‘inner’
and ‘private’. Merleau-Ponty was as critical as Wittgenstein of such a con-
ception (PhP 57–8).24
Language: Again there is a great deal to say here at which I can only
gesture, but it is clear that neither author holds such a narrow conception
of language, once we reinstate a proper phenomenological conception of
experience. Both Wittgenstein’s challenges to the ‘Augustinian’ conception
of language (PI 1ff.) and Merleau-Ponty’s challenges to the logical positivist
conception of language (PhP xxix) (as well as the empiricist and intellectu-
alist conceptions of word-meaning, PhP 182) are aimed at reinstating an
internal relation between language and the lifeworld (‘to imagine a language
means to imagine a form of life’, PI I §19). Merleau-Ponty’s description
of ‘the meaning of speech’ as ‘the way in which it handles this linguistic
world or in which it plays modulations on the keyboard of acquired mean-
ings’ (PhP 192)—here we have ‘a kind of habituation, a use of language as
a tool or instrument’ (PrP 99)—invites comparison to Wittgenstein’s sug-
gestion that we view ‘the meaning of a word [as] its use in the language’
(PI I §43).25 To be sure, Merleau-Ponty is here describing what he calls
‘second-hand’ as opposed to ‘authentic’ speech (cf. PhP 530 n.6), or the
spoken word as opposed to the speaking word (PhP 202). (This distinc-
tion is not exactly well-drawn in PhP. Paradigms of ‘speaking speech’ or
‘authentic expression’ include the speech of the writer, artist or philosopher
(PhP 203), or ‘that of the child uttering its first words, of the lover revealing
Wittgenstein and Merleau-Ponty on Gestalt Psychology  25
his feelings, of the “first man who spoke” ’ (PhP 530 n.7). These are very
different examples.) That is, whereas Wittgenstein sees ‘use’ (normally for
him ‘customary’ or ‘habitual’ use) as ‘life-giving’ (‘Every sign by itself seems
dead. What gives it life?—In use it is alive’, PI I §432), Merleau-Ponty may
seem to be saying the opposite. Yet he is not: after all, ‘speaking speech’ is
used—it is speech—and in any case, he insists on the dialectical interplay
between ‘spoken speech’ and ‘speaking speech’. The ‘acquired fortune’ of
‘spoken words’ required ‘speaking words’ in order to be acquired in the first
place, but is also what makes new ‘speaking words’ possible (PhP 203). And
although Wittgenstein does not explicitly make Merleau-Ponty’s distinction
between the spoken word and the speaking word, his discussion of the expe-
rience of meaning (brought out as W2 above) may be argued to point in the
same direction as Merleau-Ponty’s ‘speaking word’. (Remember that the
meaning-blind man, he who lacks the experience of meaning, is described
as making a ‘prosaic’ impression, RPP I §342; his words, we may say, are
spoken rather than speaking.)
It does not follow from these remarks that describing language and
describing experience are the same enterprise; but it does follow that they
are internally related enterprises.

Reflections on MP1 and MP2

We might, however, argue that descriptions (be they of language or of expe-
rience) have different purposes for Wittgenstein and Merleau-Ponty. (Witt-
genstein: ‘this description gets its light, that is to say its purpose, from the
philosophical problems’, PI I §109; cf. ‘The work of the philosopher con-
sists in assembling reminders for a particular purpose’, PI I §128.) The later
Wittgenstein has been read in multiple ways; even those who take seriously
the analogies which Wittgenstein drew between his method and psycho-
therapy differ in important respects. Some see Wittgenstein’s descriptions
of the uses of words as his primary task—he aims to construct a complete
‘logical geography’ of language—and his therapeutic moves against various
philosophical dogmas and prejudices as subservient to that aim, inasmuch
as those in the grip of dogmas and prejudices are apt to misdescribe the
uses of words. Others see his primary task as therapeutic, as freeing people
of philosophical dogmas and prejudices, and his descriptions of the uses of
words as subservient to that purpose, inasmuch as reminding people how
they ordinarily use words may make them reflect on what has compelled
them to misdescribe that use.26 Now, Merleau-Ponty’s philosophy clearly
has a therapeutic purpose too: to liberate his interlocutors from the ‘preju-
dice of objective thought’; yet arguably his approach is more easily assimi-
lable to the first way of reading Wittgenstein than the second: his ‘therapy’ is
instrumental to his primary end rather than being that end. If we (as I would
urge) read Wittgenstein in the second way, what becomes of our reflections
on MP1 and MP2?
26  Katherine J. Morris
We noted a general tendency on Wittgenstein’s part to resist dualisms,
although he didn’t appear to be particularly sensitive to the ‘reason/cause’
and ‘expressly aware/unaware’ dualisms which Merleau-Ponty picked up
on here. Had he been made so, there seems to be little doubt that he would
be as resistant to these as to others. What though would be his response to
Merleau-Ponty’s strategy? On the one hand, proposing a third possibility
that cannot be fitted to either one of the two terms of the dichotomies is
an important way of challenging dichotomies; on the other, I suspect that
he would see a danger here: that Merleau-Ponty has simply replaced these
dichotomies with trichotomies, and that these could become as dogmati-
cally entrenched in our thinking as the original dichotomies.

Here we noted that Wittgenstein offers, as an alternative possibility, a non-
mechanical conception of the body. He thinks of philosophical prejudices
as ‘musts’, as blindnesses to alternative possibilities; and a vital therapeutic
task is, precisely, describing such alternatives.27 One may argue that in order
for the therapeutic task to be successfully achieved, it is insufficient simply
to moot such an alternative (‘Well, who says that a living creature, an ani-
mal body, is a machine in this sense?’); one’s interlocutor needs to be put
in a position of seeing the possibility as a real possibility, and that requires
just the sort of detailed elaboration which Merleau-Ponty so elegantly pro-
vides. Yet it is far from clear that Merleau-Ponty would be content with
our seeing his reconceptualization of the body as ‘elaborating an alternative

Reflections on W2
At the same time, it hardly seems like a coincidence that the intellectual prej-
udices against which Wittgenstein did battle were very much those against
which Merleau-Ponty fought, and (by the same token) that the conception
of the body put forward by Merleau-Ponty was, from Wittgenstein’s per-
spective, an ‘alternative possibility’ of which Western philosophy and the
culture in which it is embedded has largely lost sight.
It is commonly said, even in prefacing compare-and-contrast enterprises,
that Wittgenstein and Merleau-Ponty are ‘products of the very different tra-
ditions of analytic philosophy and phenomenology’ (Marsh 1975: 244; cf.
Epstein 1975: 221). However, it may be pointed out that these two tradi-
tions have many historical interconnections, and that in any case the very
fact that Wittgenstein and Merleau-Ponty were reading at least some of the
same works—GP, to mention a pertinent example, as well as some of GP’s
targets such as James—implies that their ‘traditions’ overlapped. It may also
Wittgenstein and Merleau-Ponty on Gestalt Psychology  27
be pointed out on the one hand that Wittgenstein has never sat comfortably
in ‘the analytic tradition’ (in part because he resisted all ‘ologies’ and ‘isms’),
and on the other that phenomenology for Merleau-Ponty is less an ‘ology’
or an ‘ism’ than ‘a manner or style of thinking’ (PhP xxi), and one that
Merleau-Ponty very much made his own.
It also seems to me that ‘history of philosophy’ is not the only kind of his-
tory that is relevant to understanding a thinker; it is noteworthy that both
PhP and PI were ‘war books’ (cf. Read 2010). GP (like phenomenology)
self-consciously grew out of a ‘crisis’: ‘science seemed incapable of dealing
with the most significant human problems. Rather than abandoning nat-
ural science,’ the Gestalt psychologists ‘ “proposed that the difficulty was
not with science itself, but with the current conception of natural science
among psychologists”.’ Their reformed conception, they hoped, ‘would do
justice to the intrinsic meaning and value in human experience’ (Ash 1998:
2, embedded quotation from M. Henle).28 We may see both Wittgenstein
and Merleau-Ponty as responding to GP out of a sense of ‘the wretched
effect that the worship of science and the scientific method has had upon
our whole culture’ (Monk 1990: 404); Gestalt psychology had seen the
problem—that is what is so attractive about it for them—but its solution
failed to ‘do justice to the intrinsic meaning and value in human experience’
because it remained wedded to being a science.
Could this not be precisely the explanation of Wittgenstein’s otherwise
rather peculiar insistence that Köhler’s use of the word ‘meaning’ is too
narrow—that to see the world from a scientific point of view endangers
the meaningfulness of the world, that our culture is in danger of becoming
meaning-blind? And does this not make sense of the fact that Wittgenstein’s
and Merleau-Ponty’s therapeutic targets were so similar?

1 A much earlier version of this paper was presented at a workshop on Wittgen-
stein’s Philosophy of Psychology (organised by the Nordic Network for Witt-
genstein Research) at the Norwegian Institute in Rome, December 2009; a more
recent version at the second Wittgenstein Colloquium at Porto Alegre in Brazil
(May 2013). I am extremely grateful to the participants for their helpful com-
ments on both occasions.
2 See, e.g., Ash (1998), especially chapters 3 and 5.
3 Wittgenstein and Merleau-Ponty scholars divide up their corpuses in various
ways. All such divisions are, needless to say, contestable, and I will not enter into
the issues here.
4 The best-known names among the Gestalt psychologists are Max Wertheimer,
Kurt Koffka, and Wolfgang Köhler.
5 There is some indirect evidence that Wittgenstein came across GP rather earlier
(a lecture by Waismann, with whom Wittgenstein worked closely, which men-
tions Köhler by name; I am grateful to Matthieu Marion for calling this to my
attention), but he certainly did not engage with it in detail until much later.
6 Wittgenstein’s written remarks during this period are gathered together as
Remarks on the Philosophy of Psychology; some of this material was culled
28  Katherine J. Morris
in Zettel, and Wittgenstein himself gathered some of them into a short manu-
script which was originally published as Part II of the Philosophical Investiga-
tions. Also of interest are the Remarks on Colour, based on notes written around
7 It would be of equal interest to explore their respective responses to the Gestalt
psychologists’ conception of behaviour and critique of behaviourism, but that
must await another occasion. Also of great interest would be an exploration of
Wittgenstein’s and Merleau-Ponty’s criticisms of Köhler’s doctrine of ‘isomor-
phism’, according to which ‘in a given case the organisation of experience and
the underlying physiological facts have the same structure’ (GP 177, cf. GP 201),
or of how these two thinkers respectively viewed the relationship between phi-
losophy and psychology, but these explorations too must await another occasion.
8 In each case I pick out two linked objections. Wittgenstein’s and Merleau-Ponty’s
respective criticisms of Gestalt psychology are both scattered, and not everyone
will select exactly the same strands to focus on.
9 He notes that behaviourists often treat such circumscribed units as ‘stimuli’—
contrary to their own empiristic assumptions (GP 97).
10 There is a considerable literature on Wittgenstein’s discussion of aspect-seeing
with which I cannot engage here, since my focus is not on aspect-seeing per se
but on his criticisms of Köhler; Day and Krebs (eds.) 2010 is an excellent recent
collection that explores many aspects of aspect-seeing.
11 The more famous ‘I see that it has not changed and yet I see it differently’
(PI II 193) may mislead us (and did momentarily mislead Mulhall 2010: 255, and
possibly Baz 2010: 232) into thinking that the two uses of ‘see’ in this sentence
are the two uses which begin his discussion. In fact, the italicised ‘see’ is properly
understood as ‘I see that what I see (in one use of ‘see’) has not changed’.
12 In this passage Wittgenstein is actually referring to seeing facial expressions, but
it is clear that he would say the same thing about standard cases of ‘aspects’.
Wittgenstein’s contrast is regularly between seeing aspects and seeing colours
and shapes (pace Schulte who sees the latter half of the contrast as seeing objects
such as trees, 1993: 82.).
13 This argument is treated in more detail in Mulhall (1990: 9ff).
14 Cf. RPP I §868: ‘It is as if one saw a picture: one time together with one group,
and then another time with another one.’ The ‘as if’ means: ‘that process might
be a representative of the actual one, it would have the right “multiplicity”.’
15 According to Schulte, ‘the point is that the hypothesis that we can perceive
gestalts without giving them a meaning is vague and useless if we are trying to
clarify basic concepts like “perceive”, “see”, “feel”, and so forth’ (1993: 83).
This seems to me singularly unhelpful.
16 Cf. RPP I §189; cf. also this: ‘the interest here [in the ‘experience of mean-
ing’] does not depend on the concept of the “meaning” of a word, but on the
range of similar psychological phenomena which in general have nothing to do
with word-meaning’ (RPP I §358), by which ‘psychological phenomena’ I take
it he means aspect-seeing and its kin. That aspect-blindness is linked to meaning-
blindness is clear from this: ‘Anyone who cannot understand and learn to use the
words “to see the sign as an arrow”—that’s whom I call “meaning-blind” ’ (RPP
I §344).
17 He equates ‘gestalt-blind’ with ‘aspect-blind’ (RPP II §478).
18 Merleau-Ponty’s criticisms of Gestalt psychology are treated in much more detail
in Morris (2012: Chapter 4).
19 They define the body schema as ‘a total awareness of my posture in the intersen-
sory world, a “form” [gestalt]’; adding that this body schema is ‘dynamic’ (PhP
102) does not correct for the implicit assumption that the body is an object (i.e.,
a gestalt).
Wittgenstein and Merleau-Ponty on Gestalt Psychology  29
20 A rare exception is the discussion of the Necker cube; here Merleau-Ponty
argues—in an interesting parallel and contrast to Wittgenstein’s suggestion that
aspects are subject to the will—that ‘when I focus upon the face ABCD . . . that
does not only mean that I bring it to the state of being clearly visible, but also
that I make it count as a figure and as nearer to me than the other face; in a word
I organise the cube’ (PhP 275), adding that the gaze is the ‘perceptual genius’
here. Wittgenstein too might want to say: not so much that the cube ‘is organ-
ised’ but that we ‘organise’ it: see RPP I §1121.
21 Also blindness to the expression of a face (PI II 210).
22 There are issues here which go well beyond the scope of the present essay: what
are the advantages and disadvantages of using actual as opposed to imagined
abnormal individuals? (Actual cases may be aids to the imagination, as Lévy-
Bruhl’s Primitive Mythology was to Husserl’s ‘philosophical imagination’, PrP
90; real engagement with ‘abnormal’ individuals might allow us to treat them
with more sympathy than Wittgenstein does [‘prosaic impression’, ‘like an
automaton’]; on the other hand, with actual individuals one has to begin with
the person and attempt to describe their problem, whereas with imagined indi-
viduals the direction of fit goes the other way.)
23 www.youtube.com/user/stephenwiltshire
24 Some of these points are made by Overgaard and Zahavi (2009).
25 See, e.g., Carman (2008: 24); although Romdenh-Romluc does not explicitly men-
tion Wittgenstein, her elaboration of Merleau-Ponty’s conception of ‘thoughts
and their expression’ (2011: 186ff.) will resonate with Wittgensteinians.
26 See Morris (2007).
27 See Morris (2007).
28 Much has been written about the so-called ‘crisis’; an apposite source might be
Köhler (1976 [1938]).

Ash, M. G. 1998. Gestalt Psychology in German Culture 1890–1967: Holism and
the Quest for Objectivity. Cambridge: Cambridge University Press.
Baz, A. 2010. On learning from Wittgenstein, or what does it take to see the gram-
mar of seeing aspects? In W. Day and V. J. Krebs (eds.), Seeing Wittgenstein
Anew. Cambridge: Cambridge University Press, pp. 227–248.
Carman, T. 2008. Merleau-Ponty. London and New York: Routledge.
Day, W. and Krebs, V. J. 2010. Seeing Wittgenstein Anew. Cambridge: Cambridge
University Press.
Epstein, M. F. 1975. The common ground of Merleau-Ponty’s and Wittgenstein’s
philosophy of man. Journal of the History of Philosophy 13 (2): 221–234.
Jastrow, J. 1899. The mind’s eye. Popular Science Monthly 54: 299–312.
Köhler, W. 1947. Gestalt Psychology: An Introduction to New Concepts in Modern
Psychology. New York: Liveright.
Köhler, W. 1976. The Place of Value in a World of Facts. New York: Liveright
(Original publication 1938.)
Marsh, J. L. 1975. The triumph of ambiguity: Merleau-Ponty and Wittgenstein.
Philosophy Today 19 (3): 243–255.
Monk, R. 1990. Ludwig Wittgenstein: The Duty of Genius. London: Vintage.
Morris, K. J. 2007. Wittgenstein’s method: Ridding people of philosophical preju-
dices. In G. Kahane, E. Kanterian, and O. Kuusela (eds.), Reading Wittgenstein.
Malden, Oxford and Carlton: Wiley-Blackwell, pp. 66–87.
30  Katherine J. Morris
Morris, K. J. 2012. Starting with Merleau-Ponty. London and New York: Continuum.
Mulhall, S. 1990. On Being in the World: Wittgenstein and Heidegger on Seeing
Aspects. London and New York: Routledge.
Mulhall, S. 2010. The work of Wittgenstein’s words: A reply to Baz. In W. Day and
V. J. Krebs (eds.), Seeing Wittgenstein Anew. Cambridge: Cambridge University
Press, pp. 249–267.
Overgaard, S. and Zahavi, D. 2009. Understanding (other) minds: Wittgenstein’s
phenomenological contribution. In E. Zamunev and D. Levy (eds.), Wittgen-
stein’s Enduring Arguments. London: Routledge, pp. 60–86.
Read, R. 2010. Wittgenstein’s Philosophical Investigations as a war book. New Lit-
erary History 41 (3): 593–612.
Romdenh-Romluc, K. 2011. Merleau-Ponty and Phenomenology of Perception.
London and New York: Routledge.
Schulte, J. 1993. Experience and Expression: Wittgenstein’s Philosophy of Psychol-
ogy. Oxford: Clarendon Press.
2 Expression
Kathleen Lennon

The parallels between Wittgenstein and Merleau-Ponty on expression are

most often invoked in the context of both resisting a picture of the mind as
an inner private realm and interconnectedly, providing an account of our
understanding of, and relationships with, others. Expression, then, is the
thread which unravels the distinction between the inner and the outer, self
and others. In this context their views are so close that some quotations
could come from either writer. These parallels are closest between, for Witt-
genstein, Part II of the Philosophical Investigations, Zettel and the Remarks
on the Philosophy of Psychology; and for Merleau-Ponty, the Phenomenol-
ogy of Perception. It is therefore with the expressiveness of bodily gestures
that this paper will start. Despite similarities, even in this area, there are
some differences of concern between the two writers. For both, expressions
of emotions and thoughts are inter-subjectively perceptible and as expres-
sions, form a potential communicative link with others, which the gesture
embodies and initiates. Wittgenstein is particularly preoccupied with the
way in which the expressiveness of a face can be seen, with the distinc-
tion between physiology and physiognomy, with what anchors our judg-
ments of similarity and differences between faces. Merleau-Ponty, though
addressing these issues, is also concerned with the coming into being of
expressive sense, the distinction between instituting and instituted meaning,
and the way in which gestures manifest the subjectivity of ourselves and
others by making manifest the world which we share. His insistence that
gestures capture a world, evident in the Phenomenology, is developed and
explored throughout his later work. These differences become more marked
when each extends the account of bodily gestures to provide accounts of
the expressiveness of works of art. For both, the expressive character of art
works is neither the projection of the pregiven subjectivity of the artist, nor
the imitative reproduction of something external to the work; but instead the
creation of a content, a gestalt which can be grasped inter-subjectively and
integrated into the practices of a culture. Wittgenstein’s account of aesthetic
practices sees them as gaining their sense from the cultural forms of life in
which they are embedded. But there is nothing here of the artist ‘lending his
32  Kathleen Lennon
body to the world’ (PrP 162) so that it can be expressively accomplished
through him, which is at the heart of Merleau-Ponty’s account. Particularly
in his later philosophy (and in passages in the Phenomenology), Merleau-
Ponty develops a philosophy of expression that is without obvious parallels
in Wittgenstein, but invites comparison with later writings of Heidegger.
Our bodily, linguistic, and painterly practices are viewed as ‘singing the
world’ (PhP 193), a shared perceived world that is ‘accomplished’ in our
expression of it.
I will start, however, with the shared ground.

1.  Bodily Expressions

For both Wittgenstein and Merleau-Ponty, being enminded or having a soul
is captured with reference to the expressive body. ‘Expression is not one of
the curiosities that the mind may purpose to examine’, says Merleau-Ponty
‘but is its existence in act’ (S 79). ‘The face is the soul of the body’ (CV,
23e) says Wittgenstein, for whom the face is always the expressive face,
and not simply a part of our physiology. Both writers therefore reject an
account of the mind as a private inner realm, for which public expressions
are simply signs or reminders from which observers make inferences to the
inner realm of others. In opposition to accounts of expression which retain
a dualism of the expression and what is expressed, as if what is expressed
somehow lies behind its expression, both writers emphasise the direct avail-
ability of expressive content. Experiences, sensations, emotions are visible
on the body for us all to see. Thus Wittgenstein: ‘consciousness in the face
of another. Look into someone else’s face and see the consciousness in it,
and also a particular shade of consciousness. You see on it, in it, joy, indif-
ference, interest, excitement, dullness etc.; the light in the face of another’
(Z §225). ‘We see emotion . . . we do not see facial contortions and make
inferences from them . . . to joy, grief, boredom. We describe a face imme-
diately as sad, radiant, bored, even when we are unable to give any other
description of the features. Grief one would like to say, is personified in the
face’ (RPP II §570). ‘It is possible to say “I read timidity in this face” . . .
the timidity does not seem to be merely associated . . . with the face; but fear
is there, alive, in the features. If the features change slightly, we can speak
of a corresponding change in the fear’ (Z §376). Merleau-Ponty writes ‘The
operation of expression . . . does not simply leave . . . a reminder; it makes
the signification exist’ (PhP 188). ‘Faced with an angry or threatening ges-
ture, I have no need, in order to understand it, to recall the feelings which
I myself experienced when I used those gestures on my own account . . . I do
not see anger or a threatening attitude as a psychic fact hidden behind the
gesture, I read anger in it. The gesture does not make me think of anger, it
is anger itself’ (PhP 190). ‘The smile, the relaxed face, and the cheerfulness
of the gestures actually contain the rhythm of the action or of this joy as a
particular mode of being in the world’ (PhP 192). Joy, Wittgenstein says, is
Expression  33
‘neither any inward nor any outward thing’ (Z §487), an item which can be
inspected (first person) or hypothesised about (third person). Joy is a mode
of bodily being, and the joyful person is the one whose body manifests joy, a
manifestation invoking a response by others (more on this below).1
Bodily expressions are therefore public and observable, not hidden,
though (see below) there are circumstances in which we might find them dif-
ficult to read, or we might be mistaken about them. What is involved in the
process of perceiving gesture is grasp of a certain kind of gestalt,2 recogni-
tion of a certain kind of patterning of the body as that of fear, or joy or grief.
Such a patterning is one we can recognise without paying attention to the
material features of the bodies (in Wittgenstein’s case particularly the face)
that are involved. As Wittgenstein points out ‘descriptions of facial expres-
sions’ do not consist ‘in giving the measurements of the face . . . One may
note an alteration in a face and describe it by saying that the face assumed a
harder expression—and yet not be able to describe the alteration in spatial
terms’ (RRP I §919). And Merleau-Ponty remarks, ‘The movement . . . is
irreducible’ (PhP 191). Moreover the gestalt which is distinctive of particu-
lar expressions has woven into it a pattern of movements over time, the
broader context and background, and crucially the responses of others.
Although Wittgenstein would not use the term, both writers offer a phe-
nomenology of expression; the way bodies are experienced, when taken as
expressive. Wittgenstein stresses that the criteria of similarity and difference
for expressive bodies are quite distinct from physiological similarity and
difference. We are aware of the body in terms of what it expresses. The
material shape may only be available to us as, for example, the shape of
pain. Different physiologies are grouped together in so far as they are all
recognisable as ‘expressing pain’. There may be no way of grouping all the
bodily physiologies which express pain without this description. Without
the concept of ‘expression of pain’ the group may be shapeless. What we
perceive, when recognising expressive content, is not physiology, physical
features as captured by scientific categorisation, but physiognomy, ‘face or
form as an index of character’ (Shorter Oxford English Dictionary). This
allows us to perceive resemblances across faces which are physiologically
very different, and differences in faces, in different contexts, which may be
physiologically similar. David Cockburn remarks ‘two faces that, in most
contexts, would strike me utterly differently suddenly come together in a
way such that I want to say that I saw in each of them just the same reaction
to a humorous remark or to a dreadful piece of news . . . the same emotion
can find a grip on radically different facial features’ (Cockburn 2009: 132).
‘ “Similar expression” takes faces together in a quite different way from
“similar anatomy” ’ (RPP I §1068). In relation to the question of in what
way bodies need to resemble each other, in order to be detectable as express-
ing the same or similar emotions, the answer might just be, that they need
to resemble each other, simply, in both being e.g., recognisably sad. There
may be no other informative answer. The emotional characterisations here,
34  Kathleen Lennon
which often find expression in patterns of behaviour over time, have auton-
omy from the physiological ones, though requiring that physiology to have
an anchorage.3 Nonetheless these characterisations are teachable, learnable,
and projectable to new cases. Judgements of similarity and difference are
made possible by inter-subjective agreement over a range of cases. For these
writers understanding expressions is something which we are initiated into
within culture. What we are being given with such training is a certain kind
of perceptual sensitivity to patterns manifest on bodies over time.
Wittgenstein discusses perceiving expressions in the context of a more
general discussion of seeing as. ‘Seeing the resemblance of one face to
another, the analogy of one mathematical form to another, a human form in
the lines of a puzzle picture . . . all these phenomena are somehow similar
and yet again very different’ (RPP I §316). He is insistent that ‘seeing’ is the
correct word here. ‘ “I see that the child wants to touch the dog, but doesn’t
dare”. How can I see that?’ (RPP I §1066). ‘Do I learn the meaning of the
word “sad”- as applied to a face—in just the same way as the meaning
of . . . “red” ? No, not in quite the same way, but still in a similar way.’ (RPP
I §1071). ‘Is it here as if I were perceiving a “fourth dimension”? Well yes
and no. Queer, however, it is not’ (RPP I §1074). ‘ “If you will only shake
free of your physiological prejudices, you will find nothing queer about the
fact that the glance of the eye can be seen too” . . . On the other hand . . .
I should contradict anyone who told me I saw the glance “just the way”
I see the shape and colour of the eye’ (Z §223). Crucial here is both back-
ground and context. Behaviour is expressive only as contextualised, only
against particular backgrounds, backgrounds which, Wittgenstein points
out, are not themselves articulable. It is this context which will distinguish
expressions of fear, for example, from mere pretence, engaged in, maybe,
for the purposes of explaining what fear is. ‘What determines our judgment,
our concepts and reactions, is not what one man is doing now, an individual
action, but the whole hurly-burly of human actions, the background against
which we see any action . . . Seeing life as a weave, this pattern, (pretence
say), is not always complete and is varied in a multiplicity of ways. But we,
in our conceptual world, keep on seeing the same, recurring with varia-
tions . . . For concepts are not for use on a single occasion . . . And one pat-
tern in the weave is interwoven with many others’ (Z §§567–9). ‘ “Grief”
describes a pattern which recurs, with different variations in the weave of
our life. If a man’s bodily expression of joy alternated, say, with the ticking
of a clock, here we should not have the characteristic formation of the pat-
tern of sorrow, or of the pattern of joy’ (PI II 174). ‘Pain has this position in
our life; it has these connexions; (That is to say: we only call “pain” what
has this position, these connexions) . . . Only surrounded by certain normal
manifestations of life is there such a thing as an expression of pain; only sur-
rounded by an even more far reaching manifestation of life, such a thing as
the expression of sorrow or affection. And so on’ (Z §§533–4).
Expression  35
The kind of seeing which is involved here, in which the expressive con-
tent requires the colours and shapes of the face, and yet transcends them,
preoccupied much of Wittgenstein’s writing on expression. It was also a key
feature of phenomenological writing. So Merleau-Ponty: ‘Behaviours create
significations that are transcendent in relation to the anatomical structure,
and yet immanent to the behaviour as such, since behaviour can be taught
and can be understood’ (PhP 195). The expressive significance, which is seen
in the behaviour and yet takes us beyond it, is here reflecting a characteristic
which for Merleau-Ponty marked all our perceptual encounters. The world
we perceive has a meaning for us, a salience and significance which takes
us beyond brute sensory givens, weaving into the gestalt the absent and
the elsewhere, laden with ‘immense latent content, of the past, the future
and the elsewhere, which it announces and which it conceals’ (VI 114).
For him even to recognise an object as red requires an awareness of other
actual and possible reds, and to take an impression as that of a continu-
ous thing requires holding onto its past and potential future as well as to
the sides which are not immediately visible. Other possible reds are alive
in the red which we see.4 ‘This red is what it is only by connecting up . . .
with other reds about it, with which it forms a constellation . . . a certain
node in the woof of the simultaneous and the successive . . . A punctuation
in the field of red things, which includes the tiles of roof tops, the flags of
gatekeepers and of the Revolution . . . ; a naked colour . . . is not a chunk of
hard indivisible being’ (VI 132). These connections are part of the texture
of the perceptual experience. In a similar way, the gestalt of a gesture is a
movement across time in which a visible or positive presence carries with it
expressive depth. In his later work he articulates these features of perception
utilising the terms the visible and invisible. The visible, what we might ini-
tially characterise as present sensory data, is woven through with the invis-
ible, a sense or salience, invoking the present and the elsewhere, available to
us in the visible.5 A recurrent metaphor here is that of pregnancy. The visible
is pregnant with the invisible. The invisible is not the non visible. It is made
manifest through the visible, so that ‘the animals painted on the walls of
Lascaux are not there in the same way as the fissures and limestone forma-
tions. But they are not elsewhere. They inhabit those rocky surfaces’ (PrP
164). Here we are reminded of Wittgenstein’s recognition that the joy is not
on the face in the same way as its lines and crevices, but it is not elsewhere
either. It is made manifest in those lines.
Although all perception involves both the visible and the invisible, for
Merleau-Ponty, as for Wittgenstein, ‘the sense of the gesture is . . . not pre-
sented to me as a thing’. If it were then ‘it would not be clear why my
understanding of gestures should be restricted, for the most part to human
gestures’ (PhP 190). At first this remark seems puzzling. He contrasts the
understanding of human gestures with understanding the sexual gestures of
the dog (which actually don’t seem difficult to grasp) or ‘the beetle or the
36  Kathleen Lennon
praying mantis’; and also points out the difficulty of understanding emo-
tions of people ‘in milieus too different from my own’ (PhP 190). These
remarks become clearer, however, when we realise that for both writers the
expressive content of a gesture is internally related to the response which it
invites/requires from others. What response is invited is part of the invisible
which we perceive in the visible behaviour. So to grasp expressive content
as expressive, we must be engaged. We cannot be detached, for our own
responsiveness is required. For both authors, bodily expressions are com-
municative. Expressions, as opposed to mere bodily reactions, must be able
to be taken up, responded to by others. If to see bodily gestures as expres-
sions is to grasp them as part of the weave of life of those whose expressions
they are, it is also to grasp them as invitations for a response. To recognise
expressive content is to grasp how that expression is woven into a pattern of
inter-subjective life. Expressions of pain or grief, in most contexts,6 prompt
responses of comfort and solicitude from others. Overgaard remarks ‘seeing
another suffers . . . is recognising . . . something ought to be done’ (2007:
146). Exactly how is the response of the observer tied to the perceptual
recognition of expressive content? Wittgenstein sometimes says that the
response is one which comes spontaneously or naturally to us. And this can
suggest that the link between the pattern detected and our response to it is
a causal, and thereby contingent, one. However, what is at issue here is a
normative, rather than a causal relation. Whether or not we respond to the
pain of another, in grasping that pain is what is being expressed, we grasp
that certain kinds of responses are appropriate The expressive shape here
manifest in the responses it invites.7
It would be a mistake to regard such recognition as necessarily involving
cognitive judgments. For Merleau-Ponty the grasping of such bodily gestalts
depends on a reciprocity between other bodies and our own. My body takes
a responsive shape during my interactions with others, and the shape it
takes reveals the expressive content the body I am encountering has for me.
As he points out ‘I do not understand the gestures of others through an act
of intellectual interpretation . . . I join with it in a sort of blind recognition’
(PhP 191). For both authors, it is important to understand this interaction
properly and not confuse it with an epistemic operation. ‘Communication
or the understanding of gestures is achieved through the reciprocity between
my intentions and the other person’s gestures and between my gestures and
the intentions which can be read in the other person’s behaviour.’ ‘The ges-
ture is in front of me like a question . . . and invites me to join it’ (PhP 190–
1). To make sense of Merleau-Ponty’s remarks about the sexual behaviour
of the dog, it seems that the difference is that, for him, the sexual behaviour
of the dog and certainly the beetle is not experienced in this way by us. It
is not addressed to us, and it is not grasped by our own bodily responsive-
ness. (Although it certainly seems to be the case that we experience many
animal expressions as inviting our responses. Think of a dog’s whimpering
or adopting either a threatening or submissive posture towards us. I may
Expression  37
respond by shrinking back or reaching out to stroke and console the dog,
especially if it is my dog. Moreover, animals and young children reveal by
their responses their grasp of the expressions of others).8 For both authors
this way of experiencing others is distinguished from ‘an epistemic opera-
tion’. Wittgenstein famously writes ‘My attitude towards him is an attitude
towards a soul. I am not of the opinion that he has a soul’ (PI II §iv).
Grasping a gesture as expressive is therefore to be distinguished from a mere
description of it in which we are not implicated. It is to immediately rec-
ognise the possibilities for us, for our life in relation to the other whose
expression we grasp.
The reciprocal nature of our expressive perception allows, nonetheless,
for the possibility that we might recognise that someone is expressing some-
thing we cannot grasp. A central fact in our relations with others is the
extent to which they can elude our comprehension, confront us with a sub-
jectivity which is always other to our own.9 There are times when others
remain quite opaque to us. We are not able to read their expressions and
consequently find our feet with them. Expressions can be ambiguous, and
our perceptions of them open to revision and re-evaluation. Both writers
draw attention to the difficulties we may have in making sense of people
from different cultural or historical contexts. Wittgenstein imagines scenar-
ios in which our inter-subjective practices simply cannot get off the ground.
For these writers, however, inference to something supposedly lying behind
the expressions will not help. What is necessary is to be able to see the
gestures in a certain way—a way that makes their position in the life of the
subject, and potentially ourselves, clear.
If it is the case that bodily forms are perceived in terms of, for example,
the emotion they express, it is also the case that we need such bodily expres-
sions to grasp the emotion in question—just as we frequently need linguistic
expression to grasp a thought. Wittgenstein illustrates this with the case of
fear: ‘what is fear? What does “being afraid” mean? If I wanted to define
it at a single showing—I should play act fear’ (PI II 188e). What we learn,
in being taught affective terms, is what counts as a bodily expression of
particular emotions. The emotion is characterised as what gets expressed in
certain ways: ‘there is no step, after I have learned what anger is, of notic-
ing that anger characteristically finds expression in these ways’ (Cockburn
2009: 127). So we cannot characterise what counts as the relevant bodily
features for expressing a given emotion without utilising their emotional
characterisation and we cannot make evident what is involved in having
a certain emotion without utilising its bodily vehicle. In a parallel way the
content of a thought requires its linguistic or bodily expression and does not
lie behind it as a package to be simply conveyed to another. Merleau-Ponty
writes: ‘in the sense that speech expresses thought . . . it is necessary to rec-
ognise a primordial operation of signification in which the expressed does
not exist apart from the expression’ (PhP 169). ‘Thought tends towards
expression as if towards its completion . . . it is through expression that
38  Kathleen Lennon
thought becomes our own’ (PhP 182–3). ‘Speech does not translate a ready-
made thought . . . speech accomplishes thought’ (PhP 183). Nonetheless
for both writers, it makes no sense to consider a complete expression of a
thought or emotion. Our states become more determinate via their expres-
sions. In Merleau-Ponty’s terms the expression accomplishes the thought or
feeling but does not exhaust it. New expressions, sometimes in new medi-
ums, can expand our understanding of what grief or fear or sadness is, or
may render more perspicuous to us the contents of our thoughts, or more
generally, and in Merleau-Ponty’s terms, make clear to us aspects of our
existence. ‘The body expresses . . . existence not [as] . . . an external accom-
paniment of it, but because existence accomplishes itself in the body’ (PhP
169) in ways which we may not anticipate, but can recognise on encounter.
(More on this below.)

2.  Singing the World

For both writers, what is revealed by bodily expressions is our subjectivity;
not as an interiority hidden behind the body, but as a mode of being embod-
ied. So Merleau-Ponty says: ‘A woman passing by is not first and foremost
a corporeal contour for me . . . She is a certain manner of being flesh which
is entirely given in her walk or even in the simple click of her heel . . . a way
of inhabiting the world . . . -in short . . . a certain relationship with being’
(S 54). What marks out his characterisation, however, is that the subjectiv-
ity made manifest is a way of inhabiting a world, a world that remains
largely implicit in Wittgenstein’s account. Consequently, for ­Merleau-Ponty,
the inter-subjectivity, which expression requires, is an inter-subjectivity
anchored in perception of a shared world and expressive gestures reveal
characteristics of that world. Grasp of the subjectivity of others and expres-
sive grasp of the world are interdependent: ‘I communicate . . . with a speak-
ing subject, with a certain style of being, and with the world that he aims at’
(PhP 189). Expression is ‘the simultaneous articulation of . . . body and . . .
world’. Gestures ‘presuppose a perceived world shared by everyone in which
the sense of the gesture unfolds and is displayed’ (PhP 195).10
In the Phenomenology of Perception, after discussing bodily expressions,
Merleau-Ponty says ‘this revelation of an immanent or nascent meaning in
the living body extends, as we shall see, to the entire sensible world, and
our gaze . . . will discover the miracle of expression in all other “objects” ’
(PhP 203–4). Here he is drawing attention to the fact that we experience the
world itself as expressive. The world of everyday experience has a salience
and significance for us, a physiognomy or character, which anchors and
motivates our responses to it: ‘the object ‘speaks’ and is significant . . . the
perceptual field . . . a wave of significance’ (PhP 131). This expressive world
is a world for engagement, offering possibilities for our own becoming in
relation to it, the salience and significance we see linked, not causally, but
Expression  39
constitutively, to the recognition of certain responses as appropriate to it, to
the holding of our bodies in readiness to respond.
To experience the world as expressive is not to go back to a view in which
the world is regarded as itself a manifestation of consciousness or spirit. The
expressiveness of the world is not a manifestation of a consciousness, such
as bodily expressions capture. Nonetheless, neither is its expressiveness the
result of a projection of significance and salience, derived from conscious
subjects and projected onto a neutral reality. As with bodily expressions, to
experience the world as expressive is to experience it in terms of a gestalt
which weaves together, in Merleau-Ponty’s terms, both visible and invisible
characteristics, including possibilities for responses of ourselves and others
towards it. We experience the world in this way, however, because we are
able to express it, finding the world expressive and being able to express
it, for Merleau-Ponty equivalent claims, made possible by our embodied
anchorage in it. We are able to accomplish, or make manifest, features of
our world by gestures, including linguistic gestures of our own; such ges-
tures offering to ourselves and others a physiognomy of things. This, how-
ever, is not an act of a constituting subject, but of an expressive one: ‘I am
able . . . being connatural with the world, to discover a sense in certain
aspects of being, without having myself endowed them with it through any
constituting operation’ (PhP 217). The capacity for expression, that is for
bodily gestures which make manifest, to ourselves and others, aspects of
our shared world, is one of our fundamental carnal possibilities. ‘Quality,
light, colour, depth, which are there before us, are there only because they
awaken an echo in our bodies and because the body welcomes them’ (PrP
164). ‘Things . . . arouse in me a carnal formula of their presence’ (PrP 164).
What is offered is a world which comes into focus alongside us, by means
of creative expressive acts, whose disclosiveness rests on their being able to
be recognised and taken up by others.11 Our expressions, linguistic or oth-
erwise, are not operating as signs, to draw attention to a world whose char-
acteristics are determinate independently of our expressive relations to it.
Rather our gestures draw our attention to a shared world which is such that
it can be expressed in this way, accomplished, as he says, through us: ‘words,
vowels, and phonemes . . . so many ways of singing the world’ (PhP 193).
For Merleau-Ponty the initiation of expressive content is a central focus
of concern. An ‘initiating gesture- gives a human sense to the object for
the first time’ (PhP 200). It is such initiation that offers us insight into
the nature of expression itself, and into the way in which a capacity for
expression distinguishes our mode of being in the world. ‘All perception,
all action which presupposes it . . . every human use of the body is already
primordial expression, the primary operation which first constitutes signs as
signs . . . -inaugurates an order and founds an institution or tradition’ (S 67).
In the Phenomenology of Perception, and more explicitly in later works, he
returns time and again to the opening of signification, the creative origin of
40  Kathleen Lennon
expressive communication, which makes expression possible. Expressions
require a background of already established (instituted) gestures or forms
of speech, which lie behind them as a condition of their possibility, but they
also require an originating creative (instituting) moment, in which meaning
takes hold, signification becomes actualised. He says in the Phenomenology,
talking about speech but making it clear that this applies to expressions of
all kinds, ‘constituted speech, such as it plays out in everyday life, assumes
that the decisive step of expression has been accomplished’ (PhP 189), and
this requires that we ‘rediscover the primordial silence beneath the noise of
words . . . and describe the gesture which breaks this silence’ (PhP 190).
An initiating gesture then brings into view an aspect of the world, makes it
accessible to ourselves and others. This gesture is a creative one; its expres-
sive possibility resting on it finding a public, on it succeeding as an act
of communication. The necessity of creative initiating acts applies to all
kinds of expression for Merleau-Ponty, including bodily expressions, where
his account shares many features with Wittgenstein’s. For Wittgenstein, the
basis for our expressive encounters is anchored in our primitive reactions:
crying out in pain, responding to a smile with a smile, reaching out to com-
fort. On these our more complex inter-subjective responses are built. But
there is much that is common between the two writers. As noted above,
Merleau-Ponty stresses the ‘blind recognition’ with which we respond to
others, (and to the world). But such a ‘primitive’ response can nonetheless
form an instituting act on which later responses build. For both writers, this
depends on it being recognised and taken up by others.12 For Merleau-Ponty
the way in which an expression is taken on by others is a creative taking up.
Not even when we are re-employing instituted expressions is there a simple
repetition of gestures. Each expression manifests both creation and repeti-
tion and is both indebted to an instituted practice and an opening towards
an indeterminate future.13
For Merleau-Ponty this initiating expressive process is illustrated particu-
larly with reference to language (the subject of a separate paper in this vol-
ume),14 works of art (see next section), and perhaps surprisingly, geometry.
In the Phenomenology, he discusses a demonstration by a geometer of the
characteristics of a triangle (PhP 404–7). This demonstration he suggests
is not reached via a logical definition of ‘triangle’, but by a constructive
engagement with a triangle, an extension of its lines (even if only in imagi-
nation), which makes manifest its possibilities, ‘a configuration . . . towards
which my movements are directed’ (PhP 405). This is possible ‘because my
perception of the triangle [is] not so to speak congealed and dead; the draw-
ing of the triangle on the paper [is] . . . shot through by lines of force,
untraced yet possible directions . . . born everywhere in it . . . bursting with
indefinite possibilities, of which the construction actually drawn is merely
one particular case’ (PhP 406).15 The proof of the geometer is therefore a
creation in the face of an initial triangle encountered as open with possibili-
ties. He reiterates this point in a discussion in The Prose of the World, ‘when
Expression  41
I introduce a new line into a drawing that changes its signification’ (PW
119). These creative acts do not arise in a void but against the background
of previous significations which leave open possibilities. The creativity of
the geometric proof does not, however, detract from its necessity, ‘relations
seem evident to us once the expression is accomplished’ (S 76). The geom-
eter in providing us with insights about the figure is making evident some-
thing new about it, which others, if they follow the demonstration, must
also find there. Here, of course, we are reminded of Wittgenstein’s discus-
sion of following a rule in mathematics, where the proof of validity has a
similar structure (PI I §198ff). Wittgenstein, however, does not characterise
such rule following as resting on expressive acts or suggest, in the way of
Merleau-Ponty, that these acts ‘give us a world to express and think about
that envelops and exceeds [our] perspectives, a world that announces itself
in lightening signs as a spoken world or an arabesque’ (S 52).
The claim that our expressive acts yield a world announcing itself has a
metaphysical ring which Wittgenstein would not endorse. Consequently,
for him, although our expressions can be more or less successful, in respect
of their succeeding as acts of communication which can be taken up by
others; and although they can be sincere or insincere, they are not com-
municative acts which can be assessed as true or false, disclosive or non-
disclosive. In contrast, for Merleau-Ponty, ‘our expressive significations of
the world . . . have a truth: ‘truth which does not resemble things . . . and
without any predestined instruments of expression . . . which is nonethe-
less truth’ (S 57). It is truth if, in a move which here brings him close to
Heidegger (1971), the world announces itself in the expressive act. Aspects
of it become available to us.

3.  The Work of Art

The similarities and differences which we find between the two authors
become crystallised in their account of the expressiveness of works of art.
Wittgenstein discusses the way different parts of our body could become
expressive, ‘a friendly mouth, friendly eyes. How would one think of a
friendly hand?—probably open and not as a fist- And could one think of the
colour of a man’s hair as an expression of friendliness or the opposite . . .
We say; “He has a black look,” perhaps because the eyes are more strongly
shadowed by the eyebrows; and now we transfer the idea of darkness to
the colour of the hair’ (Z §506). ‘Friendly hair’: someone makes use of the
expression and others respond to it and take it up and use it. It becomes
woven into our inter-subjective practices. The possibility of extending our
emotional vocabulary to new bodily expressions also leaves space for the
extension of such vocabularies into other areas. We can see the willow tree
as sad, the blood hound’s face as sad; and consequently come to see emotion
in works of art with differing mediums. ‘The expressive operation of the
body’ says Merleau-Ponty, ‘develops into painting and art’ (S 70).
42  Kathleen Lennon
Both writers firmly reject a subjectivist account of expressive content in
art, such that what is expressed depends on the feeling of the artist or, in
some accounts, the feelings of the audience confronted with the work.16 It
is quite absurd, says Wittgenstein ‘if you say that an artist wants the feel-
ings he had when writing to be experienced by someone else who reads his
work . . . I can . . . understand a poem, [e.g.], . . . -but what he may have
felt in writing it doesn’t concern me at all’ (CV 58e). Nonetheless we can
call a work of art ‘an expression of feeling’, but not because it originates in
the subjective feeling of its maker. It is an expression of feeling, in itself, as
the thing it is (of course a thing in the phenomenal world), anchored in the
cultural formation in which it is encountered. Merleau-Ponty also rejects the
individualism and subjectivism which leads to an account of style in paint-
ing (put forward, for example, by Malraux) as the projection of the personal
meanings and values of the painter (S passim). Style is an inter-subjective
quality of the work; its expressive quality, a gestalt, a physiognomy, which
reaches out to the viewer and invites a response. It is not something ‘shut
up in the depths of the . . . individual’ (S 53). As such it is not something
that a painter could grasp about himself until he sees the pattern manifest
in the work he produces. In a way that parallels bodily expressions, expres-
sive quality is a character seen in the materiality of the work of art, but
transcending it. So Wittgenstein: ‘Tender expression in music. It isn’t to be
characterised in terms of degrees of loudness or tempo. Any more than a
tender facial expression can be described in terms of the distribution of
matter in space’ (CV 82e). ‘If I say of a piece of Schubert’s music that it
is melancholy, that is like giving it a face’ (LC 4). For Merleau-Ponty the
expressive salience is the invisible, traceable in the visible materiality, reso-
nating with a background of instituted meaning and opening a conversation
with the future.
As with bodily expressions the expressive content of works of art invites/
requires/gives reasons for/responses from those who encounter it. ‘You
could say’ says Wittgenstein, ‘that in so far as people understand it, they
“resonate” in harmony with it, respond to it’ (CV 58e). ‘The accomplished
work’ says Merleau-Ponty, ‘is . . . not the work which exists in itself, like a
thing, but the work which reaches the viewer and invites him to take up the
gesture which created it’ (S 51). The responses here are not simply causally
prompted. As with responses to bodily gestures, they are internally related
to the character of the work, that character distinguished by the ranges of
responses it invites or motivates. Wittgenstein, in particular, points out the
variety and complexity of our responses to aesthetic objects and the way
in which they fit into our forms of life. What he is at most pains to resist is
that our understanding requires some kind of inner experience in us. Under-
standing music can be shown by the way I play it; by the way I move my
body in response, dancing, tapping my fingers; by the way I compare pieces
together. But it is also shown by the role a piece of music plays within my
life, when and where it is played, the kind of occasions it is considered
Expression  43
suitable for. ‘This musical phrase is a gesture. It insinuates itself into my
life’ (CV 73). ‘Appreciating music is a manifestation of the life of mankind’
(CV 70). A poem may be read a certain way, with certain kinds of rhythm or
copied into notebooks, passed to others, read out at weddings, or just read
again and again (LC 4). A piece of china displayed or carefully wrapped and
put in a drawer.17 To respond with understanding to some works of art is
something for which we may need training. We may need to be persuaded
or nudged to see or hear it in a certain way, and thereby to respond. John
McDowell describes how such perception can be facilitated. ‘One exploits
contrivances similar to those one exploits in other areas where the task is
to back up the injunction “See it like this”: helpful juxtaposition of cases,
descriptions with carefully chosen terms and carefully placed emphasis, and
the like. (Compare, for instance what one might do and say to someone
who says “Jazz sounds to me like a mess, a welter of uncoordinated noise”)’
(McDowell 1998: 85).
There are also times, with works of art, as with people, when we cannot
grasp expressive content as expressive. We do not know how the pieces
were woven into the lives of people at the time they were created or how to
weave them into our lives now. We may turn to a glossary to inform us of
the significance of the items. In Van Eyck’s Virgin and Child,18 for example,
we seek the significance of the parrot in the child’s hand, or the identity of
the characters apparently carved into the pillars on either side. Yet a glos-
sary cannot do the job. It assumes that what is expressed by these symbols
can be expressed in some other way. Yet there will have been a myriad of
ways in which these items were woven into the (religious) lives of audiences
at that time. Consequently, we cannot experience their meaning directly. In
Wittgenstein’s terms, we are not ‘at home’ with them. We cannot respond to
them spontaneously, or with ease. For a fifteenth-century audience, no glos-
sary would be required. They would respond to these items with immediate
recognition (rather as those of us raised with Christian iconography might
to the presence of a lamb beside a depiction of the Christ child).
For both Wittgenstein and Merleau-Ponty, it is not possible to separate
expressive content from its mode of expression. Such content is fixed neither
by its relation to a prior feeling of the artist nor by an imitative relation to
a world it is attempting to copy. Wittgenstein discusses the possibility of
describing in words ‘the expression of God in Michelangelo’s “Adam” ’ and
concludes ‘you can’t at all transmit the impression by words . . . you’d have
again to paint’ (LC 38–9).19 Consequently what is expressed is not deter-
minate in advance of the creation of the work of art, but emerges with it.
Merleau-Ponty discusses a slow-motion film of Matisse in front of his can-
vas ‘Matisse . . . looked at the still open whole of his work in progress and
brought his brush towards the line which called for it in order that the paint-
ing might finally be that which it was in the process of becoming’ (S 45–6).
(Such an account can explain then how a painter like Jackson Pollock can
drip paint on to the canvas and yet create work which is found expressive
44  Kathleen Lennon
and able to be taken up within a culture). Merleau-Ponty stresses the anal-
ogy between painting and speaking: ‘The case is no different for the truly
expressive word . . . The expressive word does not simply choose a sign for
an already defined signification, as one goes to look for a hammer in order
to drive a nail . . . It gropes around a significative intention . . . [for] . . . a
particular word . . . the only possible one if that signification was to come
into the world’ (S 46). For Wittgenstein expression is accomplished if a
work can be taken up and woven into a form of life. For Merleau-Ponty
the work initiates a conversation and opens a field for further creative acts
in response. He is particularly concerned with works of art as both advents
and continuations. They are advents in that they introduce something and
bring a fecundity which suggests many responses, including creative ones.
(‘Advent is a promise of events’ (S 70).) But they also make sense as a
response to particular cultural moments, in which other creative acts lie in
the background. So when ‘Brunelleschi built the copula of the cathedral in
Florence’, he both ‘broke with the closed space of the Middle Ages’ (S 41)
and wrought a transformation which those spaces had left open; and which,
in its turn, insisted on a continuation with further expressions (the expres-
sion thus both instituting and instituted).
Both writers were concerned with judgements in aesthetics, with the art-
work as getting something right, but they give different accounts of this.
‘Architecture is gesture’ (CV 42e) says Wittgenstein, and part of what he
means by this has been captured above. It is an intervention, something with
a communicable gestalt that can be put to use in our lives. But he also says
something more puzzling. ‘Work in philosophy—like work in architecture in
many respects—is really more work on oneself . . . On how one sees things.
(And what one expects of them.)’ (CV 24). Here we have a sense of archi-
tecture as setting one questions or puzzles; that with the design we are try-
ing to make something clear. This fits with another remark he makes: ‘The
queer resemblance between a philosophical investigation (perhaps especially
in mathematics) and one in aesthetics. (E.g. what is bad about a garment,
how it should be etc.)’ (CV 25).20 Leaving on one side for this paper what
this tells us about his conception of philosophy, what is interesting for our
purposes is the comparison between architecture and mathematics. What
we are trying to make clear, or get right is, however, something internal to
the realm of aesthetic practice. And our judgements may only be articulated
in a gesture: ‘You design a door and look at it and say: “Higher, higher,
higher . . . oh, all right” (Gesture)’ (LC 13). ‘What does a person who knows
a good suit say when trying on a suit at the tailors? “That’s the right length”,
“too short” . . . a good cutter may not use any words at all, but just make a
chalk mark and later alter it’ (LC 5). The link with mathematics here echoes
his discussion of rule following in other places, with the correctness of our
moves, in both, resting on gaining the recognition of an informed audience;
an audience who may well have needed training. What is not needed to
legitimate the moves are references to something outside of the practice,
Expression  45
something which our buildings or formulae either do or do not manage to
capture. The work of art is its own kind of thing. It is not significant because
it reveals something about a different reality. But of course, in so far as our
aesthetic practices are woven into our wider forms of life, they point to and
get anchorage in this wider environment, which they require to make sense.
For Merleau-Ponty the work of art continues the expressive operations of
the body in breaking the silence of the world: ‘the quasi eternity of art is of
a piece with the quasi eternity of incarnate existence; . . . the use of our bod-
ies and our senses, in so far as they involve us in the world’ (S 70). ‘There is
no choice to be made between the world and art . . . for they blend into one
another’ (S 48). He pays most attention to the practices of painting, in so far
as they concern and make manifest the visibility and invisibility (in the sense
of what is in visibility) of the world. ‘It is by lending his body to the world
that the artist changes the world into painting’ (PrP 162). Merleau-Ponty is
not here concerning himself with a purely figurative art, sometimes viewed
as trying to reproduce on canvas something bearing a relation of similarity
to the observed scene. Art for him, figurative or not, is not imitative, the
artist does not ‘appropriate what he sees; he merely approaches it . . . he
opens on to the world’ (PrP 162), and by lending his body makes evident a
physiognomy of existence. What art is bringing to expression is the salience
and significance of the world we encounter, of colour and of space, draw-
ing our attention to the possibilities of a line. Although he returns time and
again to Cezanne, and his workings and reworking of the way a perceived
scene comes into focus for us as a world of things; he also engages with
colour in the work of Matisse and the line in Klee: ‘as Klee said the line
no longer imitates the visible; it “renders visible,” it is the blueprint of the
genesis of things. Perhaps no one before Klee had let a line muse’ (PrP 183).
Art does not describe or depict an already determinate reality, but as with
other expressive gestures, brings a world into focus, opening, a ‘hollow’ or
‘fold’ in existence. Again Merleau-Ponty’s thinking echoes that of Heidegger
(1971), and although he does not talk of art as disclosive or unconcealing,
art, along with other expressive gestures, is capable of a truth in a way
that is similar to this Heideggerian picture. It is accomplishing the world
alongside other expressive gestures. This is what distinguishes his account
from that of Wittgenstein. That art offers us a physiognomy with which to
approach the perceived world, something in terms of which we can attend
to its ‘allusive logic’ (S 57) is not a thought that we find in Wittgenstein. Yet
it is striking that in both writers we find a comparison between the work of
the artist and that of the geometer or the mathematician; a creativity which
is yet accountable. In both we find that the anchorage of the correctness of
the judgements made, or of the gestures offered, requires what is essentially
the model Kant offers in the Third Critique, in his discussion of beauty:
recognition by appropriately trained others.21 For at the centre of both their
accounts of expressions of all kinds is a view of expressions as communica-
tions, which are unsuccessful if they cannot evoke recognition.
46  Kathleen Lennon
For Merleau-Ponty, however, particularly in his later work, expression
becomes the key to our ontological status, the intertwining of our bodies
with and within a world which we can bring to expression, providing a pic-
ture in which the binaries of subject and object are replaced. Such metaphys-
ical claims are not to be found in Wittgenstein, who, nonetheless, carefully
delineates our expressive practices to destabilise the grip of the misleading
metaphysical pictures, such as the gap between the inner and the outer, sub-
ject and world, which Merleau-Ponty is replacing.

1 There is an important asymmetry here, and for Wittgenstein without this asym-
metry there would be no such thing as expression: ‘the subject of pain is the
person who gives it expression’ (PI I §302). The subject of experience is the
one whose body is manifesting expressive content. The recognition of, and
response to, this expression by another puts them in a different relation to the
expressive content. (This provides insight not only into the way in which the per-
son who has the feeling, for example, is to be identified; but also into what it is to
be a subject of experience, namely to have an, actually or potentially, expressive
body. Further discussion is beyond the scope of this paper).
2 Both writers were reading and responding to Gestalt psychology. See paper by
Katherine Morris in this volume.
3 The character is experienced through the physiology; in Merleau-Ponty’s terms
the invisible that is seen in the visible, as discussed below.
4 Strawson discusses what is involved in perceiving a dog. ‘To perceive something
as a dog, when silent and stationary, is to see it as a possible mover and barker’
(Strawson 1974: 95).
5 This is not to suggest a picture in which we invest brute sensory data with salience
or meaning. We do not experience brute data but always salient perceptions; in
Merleau-Ponty’s terms, our experience is of the invisible in the visible.
6 David Cerbone has questioned (in correspondence) what direction that engage-
ment must take. If I am, for example, sadistic, then seeing the other as suffering
will involve my recognizing that it ought to continue or be increased rather than
assuaged or mitigated. That is a kind of engagement, although not a compassion-
ate form. Is the claim here that it is only with the compassionate response that
one really sees the other’s suffering? My own thought is that grasping the other
as suffering involves grasping the appropriateness of the need for alleviation,
whether or not this prompts any desire to alleviate. But this point needs further
7 There are some parallels here with the way John McDowell sees the content
of perceptual experience as providing reasons for beliefs about the world. See
McDowell (1996), and also the discussion in McDowell (1998), chapter 3.
8 Thank you to David Cerbone for this point.
9 Cf. Overgaard (2007).
10 This is not simply the point that emotions, for example, are about things, are
intentionally directed at a world, it is that the quality of the emotional state is
provided by the way in which the world manifests itself.
11 Although, of course, we must also leave room for the idea that there are things
in front of me that I do not now see—that do not now awaken anything in me.
Thanks to David Cockburn for this thought.
12 A bodily response which does not get taken up remains as that: a bodily response
to a situation. Not all of our responses get taken up into expressive practices.
Expression  47
13 For an extended discussion of this point, see my discussion in Chapter 5 of Len-
non (2015). Also very useful here is Landes (2013).
14 There is a question as to whether understanding language as anchored in, and
itself, an expressive gesture, undermines the distinction between expressive and
descriptive uses of language and expressive and descriptive content. Merleau-
Ponty seems to replace this with a distinction between the saying and the said.
The saying is the instituting moment in linguistic expression, which can become
sedimented into instituted language, the said, which is no longer creative, and
in which our own implicatedness in the original gesture is no longer evident.
Nonetheless such language in expressing the world rests on a prior instituting
15 See also the discussion in Hass (2008).
16 See Gombrich (1963). He discusses what he calls the ‘parcel post’ model of artis-
tic expression; ‘the artist . . . knows the meaning the red patch has for him
within. . . his private world. The idea [is] that all he need do is transform his
emotion into an “expressive” configuration and send it across to the sensitive
beholder, who will unwrap the parcel and take out the emotion’ (Gombrich
1963: 54).
17 I am grateful here for discussions with Dawn Wilson.
18 Jan Van Eyck, The Virgin and Child with Canon Van der Paele, 1436, Groeninge-
museum, Bruges. See discussion Lennon (2011).
19 Although we cannot replace visual expressive content with some other form, we
can manifest our grasp on such content, or encourage others to its recognition by
verbal or other expressions of our own.
20 For a discussion of the implications for his view of philosophy see Wilson
21 In Kant’s account of beauty in The Critique of Judgement (Kant 1987 first part,
first book, ‘Analytic of the Beautiful’), we have an account of the imagination in
which creativity comes into play. Rather than bringing a manifold under univer-
sal forms, aesthetic comprehension allows a creative apprehension of the form of
beauty, recognising a unity within a multiplicity. He argues here that we employ
our creativity in seeking the form of beauty, but that its validity depends on it
being recognisable by others.

Cockburn, D. 2009. Emotion, expression and conversation. In Y. Gustafsson,
C. Kronqvist, and M. McEachrane (eds.), Emotions and Understanding: Witt-
gensteinian Perspectives. Basingstoke: Palgrave Macmillan, pp. 126–144.
Gombrich, E. H. 1963. Meditations on a Hobby Horse. London: Phaidon.
Hass, L. 2008. Merleau-Ponty’s Philosophy. Bloomington and Indianapolis: Indiana
University Press.
Heidegger, M. 1971. The origin of the work of art. In his Poetry Language Thought,
translated by Albert Hofstadter. New York: Harper Collins, pp. 15–86.
Kant, I. 1987. Critique of Judgment. Translated by W. Pluhar. Indianapolis: Hackett.
Landes, D. 2013. Merleau-Ponty and the Paradoxes of Expression. London and
New York: Bloomsbury.
Lennon, K. 2011. Imagination and the expression of emotion. Ratio 24 (3): 282–298.
Lennon, K. 2015. Imagination and the Imaginary. London and New York: Routledge.
McDowell, J. 1996. Mind and World. Cambridge, MA: Harvard University Press.
McDowell, J. 1998. Mind, Value and Reality. Cambridge, MA: Harvard University
48  Kathleen Lennon
Overgaard, S. 2007. Wittgenstein and Other Minds: Rethinking Subjectivity and
Intersubjectivity with Wittgenstein, Levinas, and Husserl. London: Routledge.
Strawson, P. 1974. Imagination and perception. In his Freedom and Resentment.
London: Methuen, pp. 82–99.
Wilson, D. forthcoming. Giving a good simile: How Wittgenstein compares aesthetic
and philosophical enquiry. In G. Tomasi (ed.), Wittgenstein, Aesthetics and Art.
3 Merleau-Ponty and Wittgenstein
on Mindreading
Exposing the Myth of the
Given Mind1
Søren Overgaard

According to a widely held view, one of the primary ways in which we

make sense of other people is by ‘mindreading’ them. Mindreading, in this
context, refers to our perfectly ordinary capacity ‘to identify the mental
states of others, for example, their beliefs, desires, intentions, goals, experi-
ences, sensations and also emotion states’ (Goldman & Sripada 2005: 193).
Those who believe that we have such a capacity, and that it constitutes
one of our most important means of understanding others, disagree about
how we mindread: what the most fundamental or common strategies or
procedures for reading others’ minds might be. Until fairly recently, two
broad proposals were widely believed to exhaust the alternatives. Accord-
ing to the so-called ‘theory theory’ (or TT), mindreaders utilize a rich body
of information about mental states and how they are connected with other
mental states, with observable behaviour, and with events in the environ-
ment.2 Feeding current perceptual or other information (e.g., about another
person’s brow-knitting) into the stored information (which, e.g., connects
brow-knitting with anger) allows mindreaders to infer what mental state
the other is in (e.g., anger). The rival ‘simulation theory’ (or ST) maintains
that mindreaders put themselves in the other person’s ‘shoes’ and use their
own minds to work out what they would do, think or feel in the other’s
situation—and then attribute those intentions, thoughts, or emotions to the
other person (e.g., Gordon 1986; Goldman 1989; Heal 1995). Simulation-
ists need not maintain that we actively and consciously imagine ourselves
in other people’s shoes for the relevant simulation to occur; sometimes this
may happen subpersonally and unconsciously, e.g., by activation of so-
called ‘mirror neurons’ in our brains (see Gallese & Goldman 1998).3
Recently, however, both TT and ST have come under fire from theorists
who maintain that mindreading, at least in a vast range of cases, involves
neither theorizing nor simulating. Instead, they maintain, we may immedi-
ately and directly see or hear—that is, perceive—that others are in whatever
mental states they are in, and so have no need for extra-perceptual cogni-
tive operations of either the TT or the ST variety. Call the view in ques-
tion ‘perception theory’ (or PT). Something like PT has been defended by,
among others, Dan Zahavi (2011), Vasu Reddy (2008), Matthew Ratcliffe
(2007), and above all, Shaun Gallagher, whose views I’ll be discussing in this
50  Søren Overgaard
chapter.4 Advocates of PT often emphasize that their views were anticipated
by philosophers such as Wittgenstein and Merleau-Ponty (Gallagher 2008a:
538, 2008b: 167; Gallagher & Zahavi 2008: 184–5). While there is a lot of
truth in such claims, as we shall see, my main aim in this paper is to show
that Merleau-Ponty and (at least indirectly) Wittgenstein also expose a fal-
lacy that seems to underlie at least some of the standard PT criticisms of
ST and TT. The fallacy in question involves subscribing to what I shall call
‘the Myth of the Given Mind’. I argue, in other words, that a survey of some
of the things Merleau-Ponty and Wittgenstein have to say about mindread-
ing reveals that their PT heirs, in arguing against TT and ST, often embrace
a deeply problematic line of thought.
The paper is structured as follows. In the next section, I briefly present some
of Gallagher’s criticisms of (personal-level, conscious versions of) TT and ST.
In section 2, I show that there is good reason to think that M ­ erleau-Ponty
and Wittgenstein would be sympathetic to PT, and that they would share
most of Gallagher’s reservations vis-à-vis TT and ST. In section 3, I argue that
Merleau-Ponty and Wittgenstein at the same time help to expose a fallacious
line of thought implicit in Gallagher’s response to TT and ST. Finally, in the
concluding section, I reply to two objections to my critique of Gallagher.

1.  Gallagher Versus Hidden Minds

TT and ST are rival theories about how we mindread. While TT attri-
butes theoretical inferences to mindreaders, ST maintains that we use our
own minds to model or simulate what others might be thinking, feeling
or intending. Advocates of PT, however, maintain that both accounts are
problematic, and for at least partly overlapping reasons. TT, it is some-
times maintained, seems phenomenologically implausible: when confront-
ing someone exploding in a fit of rage, does it really seem to you that you
need to connect some emotionally neutral perceptual input involving knit-
ted brows, flushed cheeks, and shaking fists with stored general information
about the emotion such behaviour is usually associated with, before you are
able to detect the person’s anger? Hardly, the defender of PT maintains. Nor
does ST fare any better. It does not seem to you that you need to imagine
yourself in the angry person’s shoes, or simulate his or her behaviour, in
order to figure out what such behaviour could mean. Rather, in at least a
range of everyday situations it seems immediately evident to you that the
person is angry. Enter Perception Theory: In the sort of case under consid-
eration, you neither simulate nor theorize because there’s no need for such
things. You see (and hear) that the person is angry. This is information your
perception already makes available to you; no extra-perceptual cognitive
operation is called for. As we will see later, defenders of PT are right to view
Merleau-Ponty and Wittgenstein as important forerunners, both when it
comes to the sketched criticisms of ST and TT, and when it comes to the
general outline of a positive alternative.
Merleau-Ponty and Wittgenstein on Mindreading  51
Why do advocates of TT and ST overlook this alternative option? Accord-
ing to PT, they do so because they subscribe to a set of assumptions, the
upshot of which is that the PT alternative isn’t a real option. In Shaun Gal-
lagher’s recent formulation, the assumptions in question are these:

Hidden minds: The problem of social cognition is due to the lack of

access that we have to the other person’s mental states. Since we cannot
directly perceive the other’s beliefs, desires, feelings, or intentions, we
need some extra-perceptual cognitive process (mindreading or mental-
izing by way of theoretical inferences or simulation routines) that will
allow us to infer what they are.
(Gallagher 2012: 188)

Hidden Minds seems to be an argument.5 Let us try to identify the a­ rgument’s

premises and conclusion. First, there is the assumption that the mental states
of others are inaccessible to direct perception. Call this the ‘imperceptibil-
ity thesis’. The imperceptibility thesis is said to ground the conclusion that
mindreading must crucially rely on extra-perceptual cognitive resources of
an inferential (TT) or simulative (ST) sort. Call this conclusion the ‘mental-
izing thesis’. Before taking a closer look at the structure of the argument,
we need to get clear on what the imperceptibility and mentalizing theses
actually state.
Consider the imperceptibility thesis first. What is the claim that we can-
not directly perceive another’s mental states a claim about? Is it a claim
about perception (e.g., about the sort of objects or ‘contents’ perceptual
experiences can and cannot have)? Or is it a claim about the sorts of things
mental states are (e.g., that they are ‘inner’ and unobservable, as opposed to
the observable, outward behaviour)? Many things Gallagher writes suggest
that he thinks the latter. The imperceptibility thesis, he writes, is the view
‘that intentions and feelings are not things that can be seen. They are mental
states . . . that are hidden away (imperceptible) in the other person’s mind’
(Gallagher 2008a: 539). Its advocates defend ‘the Cartesian idea that other
minds are hidden away and inaccessible’ (Gallagher 2008b: 164). In fact, if
this interpretation of the imperceptibility thesis is correct, Gallagher is right
to think the thesis is widely subscribed to in the mindreading literature.6
Here is a representative sample of quotes (many more could be adduced):

[Children’s theories of mind] involve appeal to abstract unobservable

(Gopnik & Wellman 1995: 234)

Unlike behaviorists, normal adults attribute to one another . . . unob-

servable internal mental states, such as goals, thoughts, and feelings,
and use these to explain and predict behavior.
(Saxe et al. 2004: 88)
52  Søren Overgaard
How do [people], or their cognitive systems, go about the task of form-
ing beliefs or judgments about others’ mental states, states that are not
directly observable?
(Goldman 2012: 402)

Defenders of TT and ST may not agree on what it is about the nature of

mental states that makes the mental states of others hidden or unobserv-
able. Many advocates of TT seem to associate mental states with abstract,
theoretical entities or ‘constructs’ of the sort postulated in scientific theories;
other theorists seem to regard them (in this respect continuing a venerable
philosophical tradition) as ‘internal’ and ‘private’ states, directly accessible
only to their owner. But the imperceptibility thesis is a negative thesis: it
simply states that whatever mental states are, they are not the sort of thing
that can be perceived (at least by anyone but their owner). And Gallagher
seems right in suggesting that many advocates of TT and ST agree on at
least this much.
If the imperceptibility thesis is about mental states not being the sort
of thing that is observable, then the Hidden Minds argument should be
interpreted as maintaining that because mental states are like that, we need
extra-perceptual processes along the lines suggested by TT or ST in order
to mindread. That this is how Gallagher himself understands the argument
seems to be confirmed by an earlier formulation of Hidden Minds. Galla-
gher writes:

Note first that for TT and ST, some extra-perceptual cognitive elements
seem to be required because of the way the problem is framed. In the
standard versions of TT and ST . . . [t]he supposition is precisely that
the other person’s mental states are hidden away and are therefore not
accessible to perception. I cannot see into your mind; hence I have to
devise some way of inferring what must be there, based on evidence that
is provided by perception.
(Gallagher 2008a: 536)

It is, then, because TT and ST conceive of the mental states of others as ‘hid-
den away’ that those states are regarded as ‘not accessible to perception’,
and it is for this reason inference or simulation is needed in order to read
another person’s mind.
Again, if this is indeed the argument Gallagher means to be attributing to
advocates of TT and ST, he seems justified in doing so (again, many similar
quotes could be adduced):

One of the most important powers of the human mind is to conceive

of and think about itself and other minds. Because the mental states of
others . . . are completely hidden from the senses, they can only ever be
(Leslie 1987: 139)
Merleau-Ponty and Wittgenstein on Mindreading  53
. . . mental states, such as beliefs and desires, are private, internal,
and not observable in others. However, such states are theoretically
anchored in other relevant constructs. Thus, we infer others’ beliefs and
desires . . . from, among other things, perceptual experience . . . physi-
ological history . . . and emotional expressions and reactions.
(Wellman 1990: 107)

Mental states, and the minds that possess them, are necessarily unob-
servable constructs that must be inferred by observers rather than per-
ceived directly.
(Johnson 2000: 22)

Hidden Minds, then, on the interpretation we have given of it, seems to be

a line of thought that at least a fair number of advocates of TT and ST sub-
scribe to.7 Its conclusion—the mentalizing thesis—seems to exclude PT as
an account of mindreading. How, then, should a proponent of PT respond
to the argument? One possible response is to question the structure of the
argument: is there any good reason to think the mentalizing thesis follows
from the imperceptibility thesis? The argument looks like this:

P. (IT) mental states are imperceptible objects.

C. Hence (MT) we can only ever infer (or simulate) their presence in
other people.

It is not obvious that C follows from P. For one thing, it doesn’t follow from
the fact that X is strictly speaking invisible that one cannot see (as opposed
to infer) that X is present. Dretske (1973) imagines a breed of aliens that
can make themselves invisible only at the cost of creating a strong magnetic
field around their bodies. In this sort of scenario, one might visually detect
the presence of an invisible alien by the paperclips and butter knives that fly
across the room to outline a moving space-alien gestalt. Less speculatively,
it could be the case that, although mental states are strictly speaking not
perceptible, there could be something like what Reid calls ‘acquired per-
ception’ of their presence.8 Perhaps, over time, one acquires enough famil-
iarity with the patterns of outward behaviour to which, say, anger gives
rise to no longer need to infer that another person is angry. Perhaps one
then simply ‘sees’—in a perhaps somewhat attenuated or metaphorical
sense—that another person is angry; no inferences or simulation exercises
need be involved. It has been argued that something like this may happen
with respect to theoretical entities in natural science (Lavelle 2012); so why
couldn’t it happen in the case of another’s mental states?
Gallagher, however, does not question the structure of the argument as
such. His response, rather, is to reject the imperceptibility thesis. Gallagher’s
PT ‘rejects . . . the Cartesian idea that other minds are hidden away’ (Gallagher
2008b: 164). We can regard the mental states of others as ‘normally and fre-
quently apparent in their embodied and contextualized behaviors, including
54  Søren Overgaard
their vocalizations, gestures, facial expressions, eye gaze, and situated pos-
tures’ (Gallagher 2012: 188), rather than hidden ‘behind’ such behaviours.
Importantly, Gallagher suggests that doing so removes the explanandum of
mainstream mindreading research. It was supposed that we needed some
extra-perceptual cognitive element because the mental states of others were
hidden and unobservable, and TT and ST were rival accounts of what that
element consisted in. But if the imperceptibility thesis is rejected, then there
is no need for an extra element, and so there is no need for anything like TT
and ST. As Zahavi and Gallagher put it (with respect to TT in particular),

Theory theory argues that the mental life of others is invisible . . .

Supposedly the theory theory account has been developed in order to
explain a certain cognitive achievement, namely the move from the per-
ception of observable behavior to the attribution of unobservable men-
tal states . . . If in contrast, we concede that there is no move from the
perception of behavior to the attribution of mental states, but that we
rather perceive the mental life of others directly, we have . . . changed
the explanandum radically.
(Zahavi & Gallagher 2008: 239)

And by changing the explanandum in this way, we have rendered TT and

ST irrelevant. They were, after all, rival accounts of the extra-perceptual
element in mindreading, and such an element was thought to be needed
because of the imperceptibility thesis. PT, however, rejects precisely this the-
sis. To be precise, Gallagher’s claim is not that TT and ST accounts are now
without any relevance whatsoever. Rather, these types of accounts may still
be needed to account for mindreading in relatively rare cases where, for
whatever reason, someone’s mental states are not apparent in his or her
behaviour. Normally,

there is no puzzle to solve, no inference to make, since everything is just

out there and obvious. Perhaps, if . . . [e.g.] people were acting in an
odd way that didn’t match their words, or were saying odd things that
did not match their actions, we might have to resort to folk psychol-
ogy or to simulation in order to formulate some inference and to piece
it all together. These, however, are relatively rare or specialized cases
where some version of TT or ST or some hybrid theory might have
some relevance.
(Gallagher 2008b: 165)

To sum up Gallagher’s line of thought, TT and ST are rival accounts of the

extra-perceptual cognitive element in mindreading. There must be such an
element, many proponents of TT and ST seem to think, because mental
states just aren’t the sort of thing that can be seen or heard. If, however,
we reject that assumption—the assumption I’ve called the imperceptibility
Merleau-Ponty and Wittgenstein on Mindreading  55
thesis—then we render TT and ST irrelevant as theories about how we
mindread (at least in ordinary cases). In the next section, I show that
­Merleau-Ponty and Wittgenstein would be sympathetic to PT, and to much
of Gallagher’s criticism of TT and ST.

2.  Merleau-Ponty and Wittgenstein Versus Hidden Minds

It seems clear that Merleau-Ponty and Wittgenstein would agree with Gal-
lagher that neither TT nor ST offers an adequate account of paradigmatic
ordinary cases of mindreading. At least, they would question TT and
ST accounts of ordinary emotion detection. To take Wittgenstein first, he
seems to distance himself from a TT type of account in the following well-
known passages:

In general I do not surmise fear in him—I see it. I do not feel that I am
deducing the probable existence of something inside from something
outside; rather it is as if the human face were in a way translucent and
that I were seeing it not in reflected light but rather in its own.
(RPP II §170)

“We see emotion.”—As opposed to what?—We do not see facial con-

tortions and make the inference that he is feeling joy, grief, boredom.
We describe a face immediately as sad, radiant, bored, even when we
are unable to give any other description of the features.—Grief, one
would like to say, is personified in the face.
(RPP II § 570; cf. Z §225)

Wittgenstein’s worries about inferential models seem to echo PT’s insistence

on the phenomenological implausibility of TT: it simply does not seem to
us that we need to make an inference from seen facial contortions to the
emotion someone else is feeling. Rather, we immediately see the person as
angry or sad. Thus, ‘it is only when we cannot read the outer that the inner
seems to be hidden behind it’ (LW II 63), and inferences consequently seem
to be called for.
For the same reason, Wittgenstein would not be prepared to accept a
simulationist account either, as the following famous quote makes clear:

Consciousness in the face of another. Look into someone else’s face and
see the consciousness in it, and also a particular shade of consciousness.
You see on it, in it, joy, indifference, interest, excitement, dullness etc.
The light in the face of another.
Do you look within yourself, in order to recognize the fury in his
face? It is there as clearly as in your own breast.
(And what does one want to say? That someone else’s face stimulates
me to imitate it, and so that I feel small movements and muscular tensions
56  Søren Overgaard
on my own part, and mean the sum of these? Nonsense! ­Nonsense,—
for you are making suppositions instead of just describing. If your head
is haunted by explanations here, you will neglect to bear in mind the
facts which are most important.)
(RPP I §927; cf. Z §220)

According to Wittgenstein, then, fear, joy, grief, and so on, can be imme-
diately perceptible in human facial expressions (cf. RPP I §1070). There is
no gap between the ‘external’ face and the ‘internal’ mental phenomenon,
which must be bridged by means of inferences or simulation exercises.
Merleau-Ponty seems to press similar points. Contra the thesis that we
use our emotional resources to model or simulate another’s emotion—as
ST would maintain—Merleau-Ponty writes:

Modern psychology has . . . shown that the spectator does not look
within himself or within his inner experience for the sense of gestures he
witnesses. Consider an angry or threatening gesture. In order to under-
stand these gestures, I have no need of recalling the feelings I experi-
enced while I myself performed these same gestures.
(PhP 190)

Rather, I simply ‘perceive the other’s grief or anger’, and I do so ‘with-
out any borrowing from an “inner” experience of suffering or of anger’
(PhP 372). Nor would Merleau-Ponty be prepared to endorse a TT type of
account, and for similar reasons: ‘Similarly, I do not understand the other
person’s gestures through an act of intellectual interpretation’ (PhP 191).
Indeed, in quite the same way as Gallagher, Merleau-Ponty seems to suggest
that simulation or theoretical inference ‘can certainly provide a guide in the
methodical knowledge of others and when direct perception fails, but they
do not teach me about the existence of others’ (PhP 368); nor, we might
add, would they be my default ways of detecting others’ emotions.
So Merleau-Ponty and Wittgenstein are unified in denying that simula-
tionist and inferential models are adequate for at least a range of paradig-
matic cases of everyday mindreading. They also seem to agree that what
I have termed the ‘imperceptibility thesis’ is false. This is very clear in
­Merleau-Ponty’s writings. In Sense and Non-Sense, for example, he writes:

We must reject that prejudice which makes ‘inner realities’ out of love,
hate or anger, leaving them accessible to one single witness: the person
who feels them. Anger, shame, hate, and love are not psychic facts hid-
den at the bottom of another’s consciousness: they are types of behavior
or styles of conduct which are visible from the outside. They exist on
this face or in those gestures, not hidden behind them.
(SNS 52–3)
Merleau-Ponty and Wittgenstein on Mindreading  57
In line with Gallagher, what Merleau-Ponty recommends is that we do not
conceive of mental phenomena such as anger as ‘hidden’, ‘inner’ phenom-
ena, but rather view them as extending all the way to the perceivable bodily
behaviour.9 In the case of anger, for example, the angry gesture ‘does not
[merely] make me think of anger, it is the anger itself’ (PhP 190). The per-
son’s anger is not something that ‘takes place in some otherworldly realm,
in some shrine located beyond the body of the angry man. It really is here, in
this room and in this part of the room, that the anger breaks forth’ (WP 63).
A similar message emerges, though somewhat more ambiguously, in the
writings of Wittgenstein. In the Philosophical Investigations, he famously
claims that ‘If one sees the behaviour of a living thing, one sees its soul’ (PI
I §357). This is not just a point about the mind or ‘soul’ as such, but applies
to specific psychological states as well. For example, ‘One sees sadness inso-
far as one sees a person’s sad facial expression’ (LW I §769). In saying such
things, Wittgenstein’s point is ‘not that the inner is something outer’ (LW II
61), that is, that the mental simply is outward bodily behaviour.10 Rather,
his point is merely to deny that the mind and mental phenomena must be
thought of as essentially ‘hidden’ behind bodily behaviour (see LW II 35),
to which it is merely externally connected. Rather, the mental is paradig-
matically expressed in outward behaviour (LW I §947; LW II 68), where
the notion of expression must be understood as involving a ‘logical’ and
not merely empirical connection between the ‘inner’ and the ‘outer’ (LW II
63–4).11 People may of course conceal their thoughts, feelings, and emotions
(PI I §391). But this is precisely something that we do, and ‘that means that
it is not a priori they are always hidden’ (LW II 35).
This is not the place to go into the details of Merleau-Ponty’s and Witt-
genstein’s views. What matters in the present context is merely that they
would seem to share Gallagher’s reservations vis-à-vis the types of accounts
of mindreading offered by advocates of TT and ST, and that they agree
with Gallagher in rejecting what I’ve termed the imperceptibility thesis.
Now Gallagher, as we saw, seems to think that rejecting that thesis is suf-
ficient to render TT and ST accounts of mindreading—appealing, as they
do, to extra-perceptual cognitive operations—irrelevant save for relatively
rare cases where people effectively manage to completely conceal their emo-
tions (etc.), or simply behave in odd and indecipherable ways. Is this also a
thought we can find in Merleau-Ponty and Wittgenstein? The answer, quite
explicitly as far as the former is concerned, and implicitly in the case of
Wittgenstein, is negative, as we shall see in the next section.

3.  The Myth of the Given Mind

According to Merleau-Ponty, rejecting the imperceptibility thesis does not
get us very far. A fortiori, it does nothing to undermine an inferential account
of mindreading along the lines suggested by TT. Immediately after stating
58  Søren Overgaard
that the angry gesture ‘is the anger itself’ (i.e., immediately after rejecting
the imperceptibility thesis) Merleau-Ponty writes:

And yet, the sense of the gesture is not perceived like, for example, the
color of the rug. If it were presented to me as a thing, then it would not
be clear why my understanding of gestures should be restricted, for the
most part, to human gestures. . . . I do not even understand emotions in
primitive people, or in milieus too different from my own. . . . The sense
of the gestures is not given but rather understood, which is to say taken
up by an act of the spectator. The entire difficulty is to conceive of this
act properly and not to confuse it with an epistemic operation.
(PhP 190)

This passage contains a number of important points. First, even if the ges-
ture is anger made visible, and so even if anger is something that is straight-
forwardly perceivable and observable, it does not follow that the emotional
meaning of the gesture is ‘given’ to one when one sees the gesture (and
thus sees the other’s anger). If this is not immediately obvious, consider a
complex physical object such as a smartphone. Presumably, smartphones
have yet to reach all corners of the world, so there ought to be people some-
where who as yet have no idea what a smartphone is. Smartphones are
obviously visible, and a person unfamiliar with them can still see them. (To
think otherwise, as Dretske writes somewhere, is to confuse ignorance and
blindness.) Yet, whereas it would be immediately obvious to the average
youngish Western European that the object seen was a smartphone, some-
one unfamiliar with smartphones would presumably wonder what sort of
object they are looking at. Even though something is perfectly visible, then,
it clearly doesn’t follow that its meaning or significance (or the kind of thing
it is) is visible in the same way.12
So, even if the imperceptibility thesis is false, and some mental states are
not hidden, but somehow ‘embodied’ in observable behaviour, it only fol-
lows that those mental states can be seen (or heard). That is to say, it only
follows that we can see (or hear) what is in fact someone else’s mental state.
But that leaves the question concerning mindreading unanswered: how do
we recognize what we see as another’s mental state? To think that, once the
imperceptibility thesis has been rejected, there is no such further question, to
which TT or ST could plausibly provide answers, is to subscribe to what I’ll
call ‘the Myth of the Given Mind’. According to this myth, once the mental
states of others are conceived of as perceptible, it follows that they are given
to a perceiver as the mental states they are. This is a myth, as it is clear that
nothing of the sort follows.
Nor does Merleau-Ponty suppose otherwise. Remember, he states that the
angry gesture is anger, and so he clearly rejects the imperceptibility thesis.
And yet the meaning of the gesture is, as Merleau-Ponty says, ‘not given
but rather understood’, and that means we still need an account of how
Merleau-Ponty and Wittgenstein on Mindreading  59
we understand it. Have TT and ST been ruled out as potentially correct
accounts of how we grasp the meaning of the seen gesture? Not according
to Merleau-Ponty, for as the quote continues, ‘the entire difficulty’ is to give
an account of our act of grasping that does not conceive of the latter as ‘an
epistemic operation’. Nor is it clear how TT and ST could be ruled out, just
by rejecting the imperceptibility thesis. For example, it is hard to see why it
couldn’t be general information about the characteristic behaviour of angry
people that enabled us to grasp the meaning of the gesture, as a defender of
TT might suggest. If so, rejecting the imperceptibility thesis gets us precisely
nowhere in terms of rendering TT or ST irrelevant.
Wittgenstein makes at least partly identical points. He, too, observes that
our ability to understand another’s expressions of emotions depends on a
number of things, such as belonging to the same culture, or knowing the
person well. Chinese gestures, he states, may be as difficult for Europeans to
understand as Chinese sentences (Z §219). ‘I wouldn’t know, for instance,
what genuine gladness looks like with Chinese’ (LW II 89). And while Witt-
genstein is clear that I may see the glance of another’s eye, he is equally
emphatic that I don’t see it ‘ “just the way” I see the shape and colour of
the eye’ (Z §223). The meaning of another’s gestures and expressions is
not strictly given, then, but depends on our ‘power of interpreting’ them in
mental terms (RPP I §1106).
This need for interpretation has nothing whatever to do with the imper-
ceptibility thesis. We could imagine that other people’s thoughts and feelings
were always expressed ‘in Morse code or some such thing’ (RPP II §563). If
so, until I learn the code, those thoughts and feelings would be completely
inaccessible to me, despite the fact that, to paraphrase Gallagher, there is
a sense in which everything is just out there in the open. We don’t need a
‘mental thing which is hidden’ (RPP II §564) for other minds to be inacces-
sible; all we need are expressions, linguistic or otherwise, which we don’t
A series of remarks in Last Writings on the Philosophy of Psychology,
volume II, supports this interpretation of Wittgenstein.13 In order to illus-
trate the kind of thinking that underlies the idea that the mental states of
others are hidden and inaccessible, Wittgenstein imagines ‘that the soul is a
face, and when someone is glad this hidden face smiles’ (LW II 83). ‘Let it
be this way’, he continues, ‘but now we still want to know what importance
this smile . . . has’ (LW II 83). Wittgenstein’s point is precisely that reject-
ing the imperceptibility thesis does not remove the need for interpreting or
understanding what we see. This emerges when he states that ‘the inner
smile could replace the outer one [i.e., become visible in the same way as
the outer smile is], and the question about the meaning would (still) remain
unanswered’ (LW II 83).
Where Wittgenstein does not follow Merleau-Ponty is in explicitly acknowl-
edging that rejecting the imperceptibility thesis leaves the attractions of infer-
ential (or simulationist) accounts of mindreading intact. But that is probably
60  Søren Overgaard
to do with the fact that he considers such accounts implausible on other,
more phenomenological grounds, as we saw in section 3. At best, ST and
TT might be plausible as rival ‘historical’ hypotheses about how we might
have acquired the ability to perceive the meaning of another’s expressions and
gestures.14As theories about how we go about mindreading, however, they are
entirely implausible; yet, crucially, it is not the falsity of the imperceptibility
thesis that makes them so.
Now, given Merleau-Ponty’s and Wittgenstein’s emphasis on the fact that
rejecting the imperceptibility thesis does nothing to establish that we can
perceive the emotional meaning of an angry person’s gestures and facial
expressions; given that they maintain that understanding and interpreta-
tion must be involved; and given their insistence that nevertheless some-
thing like PT is the correct account of at least a range of everyday cases of
­mindreading—given all these things, what we urgently need from Merleau-
Ponty and Wittgenstein is an account of how perception can present us with
the psychological meaning of the gestures. Although properly addressing
this question is beyond the scope of this paper, let me give a few indications
of how I think such an account would go, at least in the case of Merleau-
Ponty. (Wittgenstein is trickier in this respect, because of his ‘quietist’ refusal
to offer positive philosophical theories.)
It is not insignificant that both Wittgenstein and Merleau-Ponty seem
to actively employ the metaphor—much criticized by advocates of PT—of
‘mind-reading’. It is, Wittgenstein says, only when I ‘cannot read’ (nicht
lesen können) the other’s behaviour that his mind seems to be ‘hidden
behind it’ (LW II 63). And for Merleau-Ponty, ‘I do not perceive the anger
or the threat as a psychological fact hidden behind the gesture, I read the
anger in the gesture [je lis la colère dans le geste]’ (PhP 190). Anger and
pain are things I may ‘read on someone’s face’ (PhP 25). Indeed, as he says
elsewhere, ‘Gestures are expressive in the manner of a language’ (CPP 446),
and thus to grasp their meaning ‘is to decipher a language’ (CPP 445).
Merleau-Ponty, for one, believes this does not undermine the perceptual
character of (much of our) mindreading. For even outside the context of
mindreading, perception may involve an element of ‘reading’, according to
Merleau-Ponty. He thus speaks of ‘the spontaneous method of normal per-
ception’ making the ‘concrete essence of the object immediately readable’
(lisible) (PhP 133).15 The crucial point is that when we’re accomplished at
reading—whether reading a text, ‘reading’ the ‘concrete essence’ of some
complex physical object, or ‘reading’ another’s emotion in their b ­ ehaviour—
the ‘vehicles’ of meaning (the printed signs, the knitted eyebrows) in a sense
withdraw or become transparent. As Wittgenstein observes, we are some-
times able right away to describe a face as happy or sad, whereas we would
be hard pressed to describe its features in purely ‘external’ physical or geo-
metrical terms (cf. LW II 62). The difficulty here is not to get from clearly
discernible physical features to hidden mental phenomena, but rather to
get from the clearly discernible mental phenomena to the purely physical
Merleau-Ponty and Wittgenstein on Mindreading  61
features, which can only be reached by means of a process of abstraction.
We need no inference or simulation exercise, then, to get from the vehi-
cles to the meaning, for we are already at the meaning. As Merleau-Ponty
writes, presumably intending to capture something that holds also for the
‘language’ of another’s gestures:

The wonder of language is that it makes itself be forgotten: my gaze is

drawn along the lines on the paper, from the moment that I am struck
by what they signify, I no longer see them. . . . The expression fades
away in the face of the expressed.
(PhP 422)

Much more would obviously need to be said about the Merleau-Pontian

view to get a clear picture of its commitments, and thus to gauge its plausi-
bility.16 I cannot undertake this task here. Instead, I will conclude by briefly
considering two rejoinders to the critique of Gallagher that I have developed
with help from Merleau-Ponty and Wittgenstein.

4. Objections
I have accused Gallagher of being implicitly committed to what I’ve called
the ‘Myth of the Given Mind’. One might object to this, however, that Gal-
lagher is clearly not committed to any such thing. The burden of Gallagher’s
paper ‘Direct perception in the intersubjective context’ (2008a) is precisely
to show that perception can be ‘smart’ enough to deliver information about
other people’s emotions and intentions. This suggests that Gallagher realizes
that rejecting the imperceptibility thesis is insufficient to establish something
like DP: what is needed is reflection, not on the nature of the mental, but on
the deliverances or accomplishments of perception. And as the title of that
paper indicates, and as Gallagher explicitly states in a number of places, the
smartness of perception is highly dependent on context. There is, then, no
suggestion that the psychological meaning of another’s behaviour is simply
‘given’ to anyone who cares to look, in every type of situation, and so on.
I concede all of this. Gallagher’s explicit views have nothing to do with
the Myth of the Given Mind. My claim in this chapter has merely been that
the way Gallagher construes the consequences of rejecting the impercepti-
bility thesis—in particular, his idea that such rejection renders inferential
and simulational accounts of mindreading irrelevant, save for a few odd
cases—implicitly commits him to the Myth. Avoiding such commitment
requires realizing that rejecting the imperceptibility thesis gets us precisely
nowhere when it comes to marginalizing (or rendering irrelevant) TT and
ST accounts of mindreading. Unlike Merleau-Ponty, Gallagher seems not to
realize this.
That leads me to a second objection. I have construed Hidden Minds and
the imperceptibility thesis as having to do with the nature of the mental,
62  Søren Overgaard
rather than the accomplishments or deliverances of perception. But perhaps
Gallagher really intends them to be about the latter, rather than the former.
If so, the charge of being committed to the Myth of the Given Mind must
be dropped. For if the imperceptibility thesis states that perception can-
not deliver the psychological meaning of another’s seen behaviour, and it
is the thesis understood in this way that is supposed to necessitate extra-­
perceptual cognitive operations along TT or ST lines, then rejecting that
thesis does pull the rug from under the feet of TT and ST.
My reply is as follows. If Gallagher meant to say what the present objec-
tion has him say, then two remarks are in order. First, some of his formula-
tions of the imperceptibility thesis and the Hidden Minds argument are, at
the very least, highly misleading. To speak, for example, of ‘the Cartesian
idea that other minds are hidden away’ (Gallagher 2008b: 164) seems con-
fusing, at best, if the ‘Cartesian’ idea in question is supposed to be an idea
about perception, rather than the nature of the mental. The same point
applies to Gallagher’s talk of the thesis ‘that intentions and feelings are not
things that can be seen. They are mental states . . . that are hidden away
(imperceptible) in the other person’s mind’ (2008a: 539). Second, on the
current reading of Gallagher’s claims, it seems questionable whether they
home in on assumptions that are widely accepted among defenders of TT
and ST. All but one of the quotes from TT and ST advocates that were
adduced in section 2 in support of Gallagher’s claim that TT and ST are
wedded to Hidden Minds were about the supposed nature of mental states,
not about perception. It was the supposed unobservability of such states
that was (unconvincingly) believed to motivate the mentalizing thesis, not
specific commitments about the accomplishments of perceptual experience.

5. Conclusion
To sum up, Merleau-Ponty and Wittgenstein help us to see that, pace Gal-
lagher, rejecting the imperceptibility thesis does not affect the raison d’être
of theory theory and simulation theory. In fact, the task of showing that
perception theory is preferable to the latter remains to be carried out. Gal-
lagher has made some useful pointers in this regard, as have Wittgenstein
and Merleau-Ponty; but I think it is safe to say that the matter has not been
settled. If this chapter has brought us a step closer to settling it, it is merely
by exposing as illusory one supposed reason for believing the task to be an
easy one.

1 I am grateful to audiences at Wuppertal and Nijmegen Universities for helpful
discussion. I thank Rasmus Thybo Jensen, Smail Rapic, Marc Slors, and László
Tengelyi, for comments on earlier versions of some of the material included here.
Special thanks to my commentators at the Nijmegen conference, Ken Aizawa
and Julian Kiverstein, and to Komarine Romdenh-Romluc for helpful comments
on the penultimate version of the paper.
Merleau-Ponty and Wittgenstein on Mindreading  63
2 Defenders of TT are themselves divided over whether the relevant body of infor-
mation is the result of theory formation on the basis of observation, testing, and
learning, as so-called ‘child-scientist theory’ maintains (e.g., Gopnik & Wellman
1995), or whether it is contained in a ‘module’ that is activated at some point in
development, as so-called ‘modular theory’ has it (e.g., Leslie 1994; cf. Baron-
Cohen 1995).
3 Neurons that fire both when we ourselves perform a particular action, say, and
when we observe another perform the same action.
4 It is not obvious that PT is incompatible with subpersonal versions of ST, or for
that matter subpersonal (e.g., modular) TT, although Gallagher tends to describe
the view as if it were. For the purposes of this chapter, I shall assume that Gal-
lagher is right to ascribe to TT and ST quite generally the claim that mindreading
involves extra-perceptual cognitive resources. But see Overgaard and Michael
(2015) for critical discussion.
5 I suppose one could also view it as an explanation of the supposed fact that we
need extra-perceptual cognitive machinery to mindread, though such a reading
strikes me as less natural. Note, in this connection, that in the paper that set the
‘mindreading’ debate going, Premack and Woodruff argued that the fact that men-
tal states ‘are not directly observable’ is a reason for viewing ‘a system of infer-
ences’ to such states as constituting a theory (Premack & Woodruff 1978: 515).
6 I know of only one example where the imperceptibility thesis is construed as a
thesis about perception (as opposed to the nature of the mental): ‘Our sensory
experience of other people tells us about their movements in space but does not
tell us directly about their mental states’ (Meltzoff et al. 1999: 17). I suspect,
though, that Meltzoff et al. may be speaking loosely here, rather than commit-
ting themselves to strong claims about perceptual experience.
7 It does not follow that Gallagher is justified in regarding TT and ST, quite gener-
ally, as being premised upon the Hidden Minds assumption, though I shall not
pursue this point here (see Overgaard & Michael 2015).
8 See Reid (1785/2000), chapter 6, section 20.
9 Some of Merleau-Ponty’s formulations may seem to commit him to behaviourism—
an offence of which Wittgenstein, too, has been accused. For some reflections on
ways to develop a perceptual account of mindreading without embracing behav-
iourism, see Overgaard (2012).
10 ‘It is, I think, misleading to describe the genuine expression as a sum of the expres-
sion and something else, though it is just as misleading . . . if we say that the genu-
ine expression is a particular behavior and nothing besides’ (PO §§302–3).
11 I cannot go into Wittgenstein’s (or for that matter Merleau-Ponty’s) account of
expression here. See Kathleen Lennon’s contribution to the present volume.
12 People, tomatoes, and oak trees are perfectly visible things. It does not follow
that—in fact, it is hotly debated whether—one sees the ‘kind properties’ of such
things. For example, do you, in addition to seeing the tomato’s redness and
roundness, also see its being a tomato? Not everyone thinks so. For some reflec-
tions on the significance of the question of whether perceptual experience can
have so-called high-level content, see Siegel (2010: 3–15), and Silins (2013).
13 See also Bax (2011), chapter 3, section 3.6.
14 ‘ “What I see can’t be the expression, because the recognition of the expression
depends on my knowledge, on my general acquaintance with human behaviour.”
But isn’t this merely an historical observation?’ (RPP I §1073).
15 Note that this point is obscured in Colin Smith’s translation, which has ‘recog-
nizable’ instead of ‘readable’. I am indebted to Rasmus Thybo Jensen for calling
my attention to this passage and its significance.
16 If the view sketched is indeed Merleau-Ponty’s, then he seems to understand at
least some basic cases of mindreading along the lines of what Reid calls ‘acquired
perception’ (see note 8 above).
64  Søren Overgaard
Baron-Cohen, S. 1995. Mind-Blindness: An Essay on Autism and Theory of Mind.
Cambridge, MA: MIT Press.
Bax, C. 2011. Subjectivity After Wittgenstein. London: Continuum.
Dretske, F. 1973. Perception and other minds. Noûs 7 (1): 34–44.
Gallagher, S. 2008a. Direct perception in the intersubjective context. Consciousness
and Cognition 17: 535–543.
Gallagher, S. 2008b. Inference or interaction: Social cognition without precursors.
Philosophical Explorations 11: 163–174.
Gallagher, S. 2012. In defense of phenomenological approaches to social cognition:
Interacting with the critics. Review of Philosophy and Psychology 3 (2): 187–212.
Gallagher, S. and Zahavi, D. 2008. The Phenomenological Mind: An Introduction
to Philosophy of Mind and Cognitive Science. London: Routledge.
Gallese, V. and Goldman, A. 1998. Mirror neurons and the simulation theory of
mind-reading. Trends in Cognitive Sciences 2: 493–501.
Goldman, A. I. 1989. Interpretation psychologized. Mind and Language 4: 161–185.
Goldman, A. I. 2012. Theory of mind. In E. Margolis, R. Samuels, and S. P. Stich
(eds.), The Oxford Handbook of Philosophy of Cognitive Science. Oxford:
­Oxford University Press, pp. 402–424.
Goldman, A. I. and Sripada, C. S. 2005. Simulationist models of face-based emotion
recognition. Cognition 94: 193–213.
Gopnik, A. and Wellman, H. 1995. Why the child’s theory of mind really is a theory.
In M. Davies and T. Stone (eds.), Folk psychology: The Theory of Mind Debate.
Oxford: Wiley-Blackwell, pp. 232–258.
Gordon, R. 1986. Folk psychology as simulation. Mind and Language 1 (2):
Heal, J. 1995. How to think about thinking. In M. Davies and T. Stone (eds.), Men-
tal Simulation. Oxford: Wiley-Blackwell, pp. 33–52.
Johnson, S. C. 2000. The recognition of mentalistic agents in infancy. Trends in
Cognitive Science 4: 22–28.
Lavelle, J. S. 2012. Theory-theory and the direct perception of mental states. Review
of Philosophy and Psychology 3: 213–230.
Leslie, A. M. 1987. Children’s understanding of the mental world. In R. L. Gregory
(ed.), The Oxford Companion to the Mind. Oxford: Oxford University Press,
pp. 139–142.
Leslie, A. M. 1994. ToMM, ToBy, and agency: Core architecture and domain speci-
ficity. In L. A. Hirschfeld and S. A. Gelman (eds.), Mapping the Mind: Domain
Specificity in Cognition and Culture. Cambridge: Cambridge University Press,
pp. 119–148.
Meltzoff, A. N., Gopnik, A. and Repacholi, B. 1999. Toddlers’ understanding of
intentions, desires, and emotions: explorations of the dark ages. In P. D. Zelazo,
J. W. Astington, and D. R. Olson (eds.), Developing Theories of Intention. Mah-
wah: Lawrence Erlbaum, pp. 17–41.
Overgaard, S. 2012. Other people. In D. Zahavi (ed.), The Oxford Handbook of
Contemporary Phenomenology. Oxford: Oxford University Press, pp. 460–479.
Overgaard, S. and Michael, J. 2015. The interactive turn in social cognition ­research:
A critique. Philosophical Psychology 28 (2): 160–183.
Merleau-Ponty and Wittgenstein on Mindreading  65
Premack, D. and Woodruff, G. 1978. Does the chimpanzee have a theory of mind?
Behavioral and Brain Sciences 4: 515–526.
Ratcliffe, M. 2007. Rethinking Common Sense Psychology: A Critique of Folk Psy-
chology, Theory of Mind and Simulation. Basingstoke: Palgrave Macmillan.
Reddy, V. 2008. How Infants Know Minds. Cambridge, MA: Harvard University
Reid, T. 1785/2000. An Inquiry into the Human Mind on the Principles of Common
Sense. Edited by D. R. Brookes. Edinburgh: Edinburgh University Press.
Saxe, R., Carey, S. and Kanwisher, N. 2004. Understanding other minds: Linking
developmental psychology and functional neuroimaging. Annual Review of Psy-
chology 55: 87–124.
Siegel, S. 2010. The Contents of Visual Experience. Oxford: Oxford University
Silins, N. 2013. The significance of high-level content. Philosophical Studies 162:
Wellman, H. 1990. The Child’s Theory of Mind. Cambridge, MA: MIT Press.
Zahavi, D. 2011. Empathy and direct social perception: A phenomenological pro-
posal. Review of Philosophy and Psychology 2 (3): 541–558.
Zahavi, D. and Gallagher, S. 2008. The (in)visibility of others: A reply to Hersch-
bach. Philosophical Explorations 11: 237–244.
4 Community Without Conservatism
Wittgenstein and Merleau-Ponty
on the Sociality of Subjectivity
Chantal Bax

1.  Social Subjectivity and the Question of Conservatism

According to both Wittgenstein and Merleau-Ponty, the human subject is
fundamentally social in nature. Although this insight is perhaps more read-
ily associated with Wittgenstein—famous for his remarks on practices and
forms of life—Merleau-Ponty also holds that the self cannot be understood
in isolation from its socio-cultural context, as can for instance be seen from
the chapter on ‘Others and the Human World’ in the Phenomenology of
Perception. Like Wittgenstein, Merleau-Ponty repeatedly underscores that
subjectivity is embedded as well as embodied. Indeed, in this contribution,
I will use the latter’s arguments to clarify something that has remained
underdeveloped in the former’s writings. For even if Wittgenstein is better
known for his social view on the self, it is Merleau-Ponty who has explored
some important implications of the claim that there is no subject without
While this claim has been defended by other influential critics of
Cartesianism—Heidegger and Foucault, to name just two—it has not
replaced the traditional individualistic account of subjectivity without
objections raised or questions asked. Commentators have pointed out that
an unspecified or unqualified social view on the self may have problematic
consequences, especially from an ethico-political perspective.1 For does the
claim that the subject is formed and shaped by its socio-cultural context not
effectively imply that subjects are predestined to carry on inherited ways of
living and thinking and thus lack the freedom to make choices for them-
selves? Does it not mean that the communities in which they are embedded
form monolithic wholes that allow no change in composition or organiza-
tion? Is the essentially social self, in other words, not inherently reactionary?
In what follows, I will address questions such as these, not because I think
that they inevitably follow from the claim that there is no subject without
community, but because it is all the same important to understand more
fully why an embedded account of subjectivity does not automatically have
conservative consequences. Even though some interpreters have described
Wittgenstein’s social view precisely as reactionary,2 I will investigate for
Community Without Conservatism  67
what reasons a socially situated subject is not necessarily preprogrammed to
reject what diverges from the norm. By reading Wittgenstein together with
Merleau-Ponty—who explicitly discussed the possibility of socio-political
change—I will explain in more detail why the embedded self is by no means
precluded from being open to what is new or different.
To this aim, I will first of all look into Wittgenstein’s social account of
subjectivity. I will focus primarily on On Certainty: remarks in which Witt-
genstein not only examines the conditions for the possibility of human
knowledge, but also most consistently investigates the social formation of
subjectivity of all of his writings. What is more, On Certainty makes all
too clear why critics have objected to the idea that the self is formed by
its socio-cultural context. Wittgenstein for instance describes children as
absorbing pre-given ways of thinking without further ado, and matter-of-
factly observes that, were we to meet someone with a different outlook on
things, ‘we should not just not share his opinion: we should regard him as
demented’ (OC §155). While this could be taken to mean that an essentially
social subject is indeed reactionary to the core, I will argue that this is not
necessarily the conclusion that Wittgenstein himself has drawn; he simply
did not work out the ethico-political implications of his view in much detail.
It is for this reason that I will then turn to Merleau-Ponty, for in addition to
contributing to the philosophy of subjectivity, the latter wrote extensively
on matters of politics.
Politics is even among the topics discussed in the Phenomenology of Per-
ception, the book in which Merleau-Ponty famously offers an embodied
account of the self. Focusing on this work, I will underscore that Merleau-
Ponty not only explains the subject as inherently corporeal, but also identi-
fies the social world as a permanent field of human existence. He, however,
does not take this to mean that the self is imprisoned in its situation, unable
to change or accept changes in social arrangements of its own accord. After
repeatedly characterizing the (non-pathological) subject as free, Merleau-
Ponty uses the final chapter of the Phenomenology to explain more fully
how a situated self can be still said to have freedom, and can more specifi-
cally still be said to be an actor in socio-political change. Merleau-Ponty’s
reflections can accordingly be used to answer the questions that Wittgen-
stein left unaddressed. After discussing the Phenomenology’s final chapter,
I will apply its insights to On Certainty and conclude that neither Merleau-
Ponty nor Wittgenstein takes the social self to be intrinsically conserva-
tive, because neither holds that its socio-cultural context is something rigid
and wholly determinate. It therefore does not determine the subject’s every
action and reaction in advance.

2.  Wittgenstein on Community in On Certainty

The insight that human life is an inherently social affair informs much of
Wittgenstein’s later writings, from the Investigations’ argument against
68  Chantal Bax
private language to the reflections on other minds in the Remarks on the
Philosophy of Psychology. It can, however, be said to be On Certainty in
which Wittgenstein most unmistakably shows that there is no subject with-
out community. In this discussion of human epistemology, Wittgenstein
explains that everything we say and do is made possible by our already tak-
ing numerous things to stand fast; assumptions we moreover make, not after
deliberate investigation and experimentation, but after having undergone
a process of socialization. Wittgenstein calls these basic presuppositions
‘certainties’, and they undermine the notion of a solitary and self-sufficient
subject by being neither individual accomplishments nor purely individual
Wittgenstein introduces the concept of certainty in response to Moore’s
attempt to refute scepticism. In papers such as Proof of an External World
(Moore 1993), Moore had tried to disprove the sceptic by means of state-
ments like ‘The earth already exists for many years’ and ‘I know this is a
hand’. In On Certainty, Wittgenstein tries to come to terms with the fact that
he both agrees and disagrees with Moore’s endeavour. Like Moore, Wittgen-
stein holds that sceptical questions do not have the devastating consequences
that the sceptic takes them to have, but at the same time, he feels that some-
thing is amiss with Moore’s insistence that he indisputably knows things like
his having a hand. That is to say, Wittgenstein does not contest the indisputa-
bility of what Moore claims to know, but this very indisputability makes him
wonder whether Moore can really be said to know these things.
In the course of On Certainty, Wittgenstein points to several differences
between Moore-style statements3 and normal or actual knowledge claims,
prompting him to coin a different term for the former. While he tries out dif-
ferent descriptions and is not always consistent in his terminology, he seems
to settle for the label ‘certainty’ to characterize convictions like Moore’s.
Certainties are what we unhesitatingly take for granted in all our doings
and sayings; unspoken assumptions about life and the world we simply and
confidently take for a fact—like there being external objects, our having
a body, and every human being’s having a mind. The difference between
knowledge claims and such certainties is profound, Wittgenstein maintains:
they ‘belong to different categories’ (OC §308).
Certainties for instance differ from knowledge claims in lacking clear pro-
cedures for justification, or in lacking justification procedures überhaupt.
While a statement like ‘The earth exists’ may seem to belong to a special
class of extremely well-founded knowledge claims, we would in fact have
a hard time explaining how this can be proven to be true. ‘Moore chooses
precisely a case,’ Wittgenstein points out, ‘in which we all seem to know
the same as he, and without being able to say how’ (OC §84). Simply stat-
ing that one knows the world exists, no matter how emphatically, does not
yet amount to a proof thereof, as Wittgenstein explains in another remark:
‘when writers enumerate all the things they know, that proves nothing
whatever. So the possibility of knowledge about physical objects cannot be
Community Without Conservatism  69
proved by the protestations of those who believe that they have such knowl-
edge’ (OC §488). It is, however, unclear what further means there are for
demonstrating such things.
Yet even if we do not know how to justify our certainties, they are exempt
from doubt and disagreement in a way that knowledge claims are not, Witt-
genstein observes. Statements like ‘Cell phone radiation causes cancer’ and
‘This spearhead dates from 700 BC’ are things we debate and discuss in an
attempt to come to an agreement, though we may not always succeed in
reaching one. Moore-style convictions, by contrast, usually do not come
up for discussion at all. We would normally not argue with someone saying
‘This is my hand’ (though we might wonder why she wants to make such
a statement to begin with).4 The things Moore claims to know, Wittgen-
stein explains, ‘are all of such a kind that it is difficult to imagine why any-
one should believe the contrary’ (OC §93). Indeed, Moore does not merely
mean to say something about his own convictions—when he declares to
have knowledge of his hand, he means ‘that any reasonable person in [his]
position would also know it, that it would be a piece of unreason to doubt
it’ (OC §325).
According to the analysis offered in On Certainty, this lack of debate
and discussion is not a matter of ‘hastiness or superficiality’ (OC §352). At
some point, namely, questions and investigations have to come to an end.
We would not be able to make any claim or perform any action if we would
have to answer all possible questions about them beforehand: ‘We just can’t
investigate everything, and for that reason we are forced to rest content with
assumption. If I want the door to turn, the hinges must stay put’ (OC §343).
Wittgenstein argues that it is precisely this hinge-like function that Moore-
style statements fulfil. Rather than amounting to knowledge claims, convic-
tions of the kind Moore lists make knowledge possible in the first place.
In line with Wittgenstein’s remarks about the lack of justification for cer-
tainties, however, they do not fulfil this function because they have been—
or can be—proven to be true beyond the shadow of a doubt. Certainties
allow us to look for a justification of specific knowledge claims, but they
are themselves unjustified, or do at any rate not originate in investigation
and validation. As Wittgenstein observes: ‘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’ (OC §94).
The things we take for granted in everything we do and say, in other
words, have simply been handed on to us in the course of our upbringing.
We need not even have learned certainties explicitly, Wittgenstein maintains.
While things like ‘This is a hand’ and ‘This is your foot’ are repeatedly
spelled out for children learning to speak their native tongue, most of what
we take to stand fast is transmitted in an implicit manner. Children for
instance ‘do not learn [. . .] that armchairs exist [. . .] —they learn to [. . .] sit
in armchairs’ (OC §476) and thereby come to take the former for granted.
70  Chantal Bax
Or as Wittgenstein also puts it: ‘[The child] swallows this consequence
down, so to speak, together with what it learns’ (OC §143). Even though
adults do not always explicitly convey their particular system of certainties,
children almost automatically make the world picture of their elders their
own: ‘After [the child] has seen this and this and heard that and that, he is
not in a position to doubt whether’ (OC §280) there are external objects,
say, or whether every human being has a mind.
This further confirms Wittgenstein’s placing knowledge and certainty into
different categories. For knowledge claims, as stated, are things we debate
and discuss with the aim of coming to an agreement, though we have no
guarantee that we will ever reach one. In the case of certainties, however,
those who have received the same upbringing do not have to come to an
agreement: they already are in agreement, attuned in a more fundamental
way than a deliberate quest for concord could ever achieve. Wittgenstein
takes certainties to stand fast, not just for the individual, but for an entire
socio-cultural group: ‘ “We are quite sure of it” does not mean just that
every single person is certain of it, but that we belong to a community which
is bound together by science and education’ (OC §298).
According to Wittgenstein, then, certainties are not instances of knowl-
edge but are what makes knowledge—as well as doubt—possible to begin
with. These basic presuppositions are moreover social in two important
respects: a person both owes her certainties to and shares them with the
other members of her community. It is this shared belief system that under-
lies everyday practices and that enables those who subscribe to the same
set of certainties to go about their lives together, discussing a host of things
except what they take to stand fast. Certainties can accordingly be said to
form the lubricant for smooth (or at least relatively smooth) interaction
between the members of a particular socio-cultural group.
This, however, also means that they can hamper communication with per-
sons from a different background. If people can only engage in conversation
and discussion on the basis of a shared set of certainties, then the absence of
a common belief system interferes with the possibility of such interaction, as
Wittgenstein points out in several remarks. Giving reasons comes to an end,
he explains, and ‘At the end of reasons’ (OC §612), or when ‘two principles
really do meet which cannot be reconciled with one another’ (OC §611),
normal ways of arguing break down. As a result, Wittgenstein observes,
‘One might simply say “O, rubbish!” ’ (OC §495) to someone with different
presuppositions; respond with swearwords and insults instead of explana-
tions and arguments. When a basis for the latter is lacking, after all, com-
munication with the other can no longer be uncomplicated and smooth.
Observations such as these may suggest that, on a Wittgensteinian account,
an essentially social self is intrinsically reactionary and always already averse
to what diverges from the norm. It is therefore not surprising Wittgenstein
has been interpreted as a thoroughly conservative philosopher. While these
interpretations are often based primarily on the Philosophical Investigations,
Community Without Conservatism  71
Wittgenstein’s best-known work, and a commentator like Nyíri also makes
use of the biographical fact that Wittgenstein hails from neo-conservative, fin
de siècle Austria,5 the gist of these interpretations similarly goes that the Witt-
gensteinian self is unable to look beyond the conventions of its community
and can only ‘deplore a movement away from the order and organic unity’
(Bloor 1983: 161) of the practices it knows. Read along reactionary lines,
Wittgenstein presents us with a subject that is immersed in its culture to such
an extent that it can only respond with intolerance to what is new or different.
Yet while there are certainly moments in both Wittgenstein’s life and work
that can be understood conservatively, both also contain material for chal-
lenging a reactionary interpretation. Wittgenstein’s interest in the books of
the cultural pessimist Spengler and the anti-Semite Weininger is for instance
well known, yet in the same sentence in which he mentions these authors as
influences on his thought, Wittgenstein also lists the left-wing satirist Kraus
and the progressive economist Sraffa.6 Wittgenstein’s own writings further-
more not only contain remarks about ‘obey[ing] the rule blindly’ (PI I §219)
and having to accept pre-given ‘forms of life’ (PI II 226), but also include
entries noting that linguistic practices are ‘not something [. . .] given once
for all’ (PI I §23) and that ‘a language game does change with time’ (OC
§256). Prior to the passages in which Wittgenstein allows for the possibility
of insulting a person with different certainties, moreover—to return com-
pletely to On Certainty—he observes that when someone is ‘contradicting
my fundamental attitudes’, I will simply ‘have to put up with it’ (OC §238)
rather than bully and offend her.
I take this to mean that the exact implications of Wittgenstein’s social
account of subjectivity are far from self-evident, and that he himself did not
draw any definitive conclusions as to the reactionary or progressive nature
of human beings. If Wittgenstein can be criticized from an ethico-political
point of view, it is perhaps because he did not explore these implications
in sufficient detail. This brings me to the writings of Merleau-Ponty, who
also took the human subject to be socially situated, but who—in contrast to
Wittgenstein—did not fail to investigate the political consequences of this
view. In the next section, I will therefore turn to Merleau-Ponty; not only
to show that he, too, offers a social account of human existence, but most
importantly to examine the reasons he gives as to why an embedded view on
the self does not necessarily have conservative consequences.7

3.  Merleau-Ponty on Freedom in the Phenomenology

Even though Merleau-Ponty is primarily known as a philosopher of human
corporeality, he in fact takes the bodily and the social nature of subjectiv-
ity to go hand in hand. His interest in humans as social beings does not
only come to the fore in the political essays he published from the 1940s
onwards; it also informs a book like the Phenomenology of Perception, and
not just its extensive footnote on historical materialism.8
72  Chantal Bax
What Merleau-Ponty for instance explains to be pathological about the
Schneider case—which he extensively discusses to bring out the usual nature
of our bodily being-in-the-world—is just as much the absence of true inter-
subjectivity from his life as his inability to perform abstract movements.
Schneider has a hard time making new friends, Merleau-Ponty recounts,
and while ‘He would like to be able to think about politics or religion, [. . .]
he never even tries. He knows that these regions are no longer accessible to
him’ (PhP 160). In the non-pathological case, by contrast, the social is part
and parcel of existence: a person ‘can certainly turn away from the social
world, but [. . .] cannot cease to be situated in relation to it’ (PhP 379).
Merleau-Ponty’s critique of the Cartesian Ego thus not only concerns its
lack of embodiment. In order to fully rethink subjectivity, he maintains, the
self should be explained as both embodied and embedded: ‘we must redis-
cover the social world, after the natural world, not as an object or a sum
of objects, but as the permanent field or dimension of existence’ (PhP 379).
One way in which Merleau-Ponty tries to capture the social dimension
of human life is in terms of anonymity and generality. To be sure, Merleau-
Ponty also uses these terms to characterize certain aspects of our bodily
being-in-the-world, for instance when he states that ‘Every perception takes
place within an atmosphere of generality and is presented to us as anony-
mous. [. . .] if I wanted to express perceptual experience with precision,
I would have to say that one perceives in me, and not that I perceive’ (PhP
223). In line with what I have just emphasized, however, Merleau-Ponty’s
notions of anonymity and generality do not only refer to certain shared
biological facts; they have a social meaning for him as well. Merleau-Ponty
accordingly mentions the same anonymous ‘One’, not just when he dis-
cusses perception, but also when he discusses socio-culturally informed (or
more clearly socio-culturally informed) behaviour: ‘One uses the pipe for
smoking, the spoon for eating or the bell for summoning’ (PhP 363).
Like Wittgenstein, then, Merleau-Ponty maintains that what a person says
and does, or how a person says and does things, should always already be
seen against the background of the community of which this person is part;
specific ideas about what one is supposed to do in which situations underlie
her doings and sayings. Merleau-Ponty’s notion of the ‘One’ is however no
more clearly non-reactionary than Wittgenstein’s idea of a shared system of
certainties. Indeed, it can be said to raise questions similar to those I men-
tioned in my introduction: do preconceptions as to what one is supposed
to do not mean that a person can only respond with intolerance to people
who happen to do things differently? Do they not mean that the conventions
of a particular society are impervious to change, no matter how much they
might call for reform—what is more, do they not mean that a person would
never even think of reform in the first place?
That Merleau-Ponty does not think these questions should be answered
in the affirmative can be seen from the Phenomenology’s final chap-
ter on freedom, a large part of which is devoted to the possibility of
Community Without Conservatism  73
communist revolution. To the extent that this strikes one as an odd exam-
ple for the phenomenologist to focus on, it should be noted that at the
time of writing the Phenomenology, Merleau-Ponty was—like many French
­intellectuals—­committed to the communist cause and accordingly thought
it of vital importance to understand the possibility of a proletarian over-
throw of the capitalist status quo.9 It should be possible to transform situ-
ations of exploitation from within, as Merleau-Ponty’s main problem here
can be paraphrased, and workers should therefore be able to acquire class-­
consciousness regardless of the circumstances in which they find them-
selves, but how exactly can we explain this potential to see through and
break with existing socio-­economic arrangements? In the freedom chapter,
­Merleau-Ponty accordingly makes unmistakably clear that he does not take
the socially situated self to be inherently conservative, always already averse
to new and unfamiliar things.10
After already addressing the topic of freedom a number of times in the pre-
ceding pages of the Phenomenology—Merleau-Ponty for instance describes
Schneider as being ‘ “bound” to the actual, and [to] “[lack] freedom” ’ (PhP
137). He starts the final chapter with a critical discussion of the account
of freedom offered by his close collaborator Sartre. Without going into the
details of their debate,11 suffice it to say that on Merleau-Ponty’s view, the
Sartrean emphasis on radical choice in effect undermines the very idea of
human freedom. For in order for an act to count as free, Merleau-Ponty
explains, it has to stand out against a background of unfreedom or of lesser
freedom. It only makes sense to call a particular action free, after all, if we
can contrast it with other actions or other aspects of human life that are not
of an unrestricted nature. So if we would truly be able to choose our entire
being-in-the-world at every single instant, none of our actions could actually
be called free, because no background of unfreedom would remain. What
is more, in a Sartrean world, there would effectively not be any use for free
actions anymore. For when all situations are subject created, there are no
longer subject preceding situations that in fact call ‘either for a decision that
confirms them or for one that transforms them’ (PhP 462) and that thus
invite us to exercise our freedom in the first place. ‘[F]reedom must have a
field’ (PhP 462), Merleau-Ponty sums up, and Sartre ends up doing away
with this necessary precondition, hence with freedom itself.
This means that Sartre’s intellectualist or idealist account ultimately fares
no better than the idea of objective thought12 that all human action is caus-
ally determined, Merleau-Ponty continues. Both positions fall prey to the
same false dichotomy: the subject is either radically different from the world
and able to impose its will on its surroundings without further ado, or is just
another thing in the world and completely at the mercy of outside laws and
processes. In Merleau-Ponty’s view, however, our options are not exhausted
by these two extremes, and as he accordingly declares: ‘We must again take
up the analysis of Sinngebung and show how it can be at once centrifugal
and centripetal’ (PhP 464). We must, in other words, show how there is a
74  Chantal Bax
dynamic interplay between the subject and its situation that neither makes
the self into the sole situation-creating and meaning-giving authority, nor
presents it as entirely determined by its situation or field.
This goes for both our being situated in a physical world and our being
situated in a social world, Merleau-Ponty points out, and while he starts
the chapter with examples of our bodily situatedness, he mainly works out
his alternative to idealism and objectivism by discussing a specific socio-­
historical issue: the aforementioned possibility of communist revolution. He
starts this part of the freedom chapter, too, by arguing against an overly
intellectualistic account, according to which ‘the revolutionary project is
[. . .] the result of a deliberate judgment’ (PhP 471) on the part of the worker.
This account is supposed to be an alternative to the objectivist explanation
that takes proletarian revolt to automatically follow from the dire circum-
stances of exploited workers. Merleau-Ponty agrees with idealism that such
a deterministic explanation won’t do: in the long footnote on the proper
interpretation of historical materialism earlier in the Phenomenology, he
already distanced himself from a causal, reductive account of revolution,
stating that ‘there is never a purely economic causality because the economy
is not a closed system’ (PhP 176).
Yet while idealism is thus correct in so far as it sees that economic situ-
ations by themselves do not guarantee the emergence of a workers’ move-
ment, Merleau-Ponty argues that it overreacts in its narrow focus on the
explicit deliberations of the worker; an overreaction, moreover, that no
more helps us understand the possibility of revolution than the causal
account. ‘I certainly do not become a worker [. . .] the day that I commit to
seeing history through the lens of class warfare’ (PhP 468–9), after all. Like
the idealist account of freedom more generally, the idealist account of revo-
lution merely forms the reverse side of the objectivist view. As an alternative
to both these positions, Merleau-Ponty then explains revolution as what he
calls an existential project.
According to Merleau-Ponty, that is, we should understand the rise of
class consciousness from the perspective of the situated lives of labor-
ers themselves. It is neither purely intellectual reflections nor economic
relations per se but rather ‘society or the economy such as I bear them
within myself and such as I live them’ (PhP 469) that account for the
possibility of communist revolution. This lived experience of economic
relations is emphatically not a matter of Marxist theory and reflection.
Indeed, as ­Merleau-Ponty explains in a detailed description of the way
class ­consciousness can arise, ‘there is no need for [a worker] to conceive
of himself as a proletarian in the sense a Marxist theoretician gives the
word’ (PhP 470). A factory worker may for instance simply hear about
the strike at another plant that resulted in a higher wage. To be sure, this
need not automatically make the recipient of the news class conscious,
Merleau-Ponty hastens to add; a person may have any number of reactions
upon hearing about the strike. But even so, events such as these change the
laborer’s horizon or field. Similarities between his life and that of other
Community Without Conservatism  75
workers become increasingly prominent, for instance, and more and more
occasions for the expression of solidarity present themselves. It is at this
point that the worker can begin to see himself as part of an entire class of
the oppressed and can start acting in the name of the proletariat, even if he
refrains from using that specific term.
The revolution, then, does not arise from a spontaneous decision, as ide-
alism would have. On Merleau-Ponty’s view, it is rather ‘prepared for by
a molecular process, it ripens in coexistence prior to bursting forth’ (PhP
471). Like freedom, in other words, revolution must have a field. This how-
ever does not mean that workers ‘bring about the revolution unwittingly’
(PhP 471), as objectivism would have. Laborers themselves still play an
active role in the revolutionary movement in the sense that it is up to them
to ‘crystallize what is latent in the life of all producers’ (PhP 471).
Upon offering this existential account of communism, Merleau-Ponty
formulates his findings in more general terms and returns to the ­problem
of Sinngebung, here applied to historical events. In line with his ­preceding
arguments, Merleau-Ponty points out that historical meaning can—­
contra idealism—not be a matter of free-floating subjects making spon-
taneous decisions. Contrary to objectivism, however, there is no such
thing as meaning in events by themselves either, for ‘history is powerless
to complete anything without the consciousnesses that take it up and
that thereby decide its course’ (PhP 475). According to Merleau-Ponty,
put differently, historical meaning is both centrifugal and centripetal; it
comes about in the exchange between the subject and its situation: ‘we
give history its sense, but not without history offering us that sense’ (PhP
475). Or as Merleau-Ponty also puts it, referring once again to the social
anonymity or generality I mentioned earlier: ‘What we call the sense of
events is not an idea that produces them, nor the fortuitous outcome of
their assemblage. It is the concrete project of a future that is elaborated
in social coexistence and in the One prior to every personal decision’
(PhP 475).
Hence, while Merleau-Ponty’s notion of the One may at first sight raise
questions as to the conservative nature of his socially situated subject, in the
freedom chapter he uses this exact same notion to explain the possibility of
communist revolution and of socio-historical change in general. In the final
paragraphs of the chapter, Merleau-Ponty draws his arguments to a close
and discusses in more detail why preconceptions as to what one ought to do
in which situations do not predestine the subject to passively accept existing
social arrangements.
On Merleau-Ponty’s account, its socio-historical context does not prede-
termine the self to do anything, whether of a reactionary or of a revolution-
ary nature—but this is precisely what makes social transformation possible.
Just as class consciousness does not causally result from exploitative eco-
nomic relations, it does not automatically arise from the ‘molecular pro-
cess’ of workers starting to notice similarities between their lives and that
of other laborers. Far from preventing revolution, however, this allows for
76  Chantal Bax
the revolutionary moment to occur ‘when the sense that was taking shape
in the One and that was merely an indeterminate possibility [. . .] is taken
up by an individual’ (PhP 476). What Merleau-Ponty calls ‘the One’, in
other words, is a matter of relative indetermination rather than of absolute
determination. It does not lay down the exact course that history should
take because it does not spell out in full detail what should be done in
which situations. The One should rather be understood in terms of general
suggestions or unspecified directions that still have to be implemented, and
that can be implemented in a number of different ways. Merleau-Ponty’s
‘molecular process’, for instance, could lead to both violent upheaval and
peaceful resistance.
What one is supposed to do, therefore, is not always already given in
advance. The socio-historical situation in which the subject finds itself is
not something wholly determining because it is not something wholly deter-
mined: ‘The world is always already constituted, but also never completely
constituted’ (PhP 480). Our situations may accordingly solicit us, or moti-
vate us, as Merleau-Ponty carefully puts it, but they do not force or dictate
us. In line with the idea that history is powerless in and of itself, it is always
still up to subjects themselves to take up the solicitations of their situation
and to make actual sense of what is only ambiguously present in history as
such. This means that a person can also ignore the solicitations of the One,
or respond to them in as of yet unheard of ways. An individual can steer
history ‘well beyond what seemed to be its sense’ (PhP 476), Merleau-Ponty
explains, because one’s past and milieu are only ‘moments of my total being
whose sense I could make explicit in different directions’ (PhP 482).
Hence, as Merleau-Ponty himself already pointed out in the footnote on
historical materialism, ‘what we said [. . .] about the existential conception
of “expression” and “signification” must again be applied here’ (PhP 176).
Similar to the accounts of meaning and perception offered in the earlier
chapters of the Phenomenology, the account of history offered in the free-
dom chapter presents it as a two-way affair, as an interchange between the
subject and its situation. This interchange is made both possible and nec-
essary by the fact that the situation in which the subject finds itself, is not
always already clearly determined and still waits to be made concrete. ‘An
existential theory of history is’ therefore ‘ambiguous,’ the footnote contin-
ues, ‘but it cannot be reproached for this ambiguity, for the ambiguity is in
the things’ (PhP 176).
It is because of this equivocality, moreover, that the subject is not predes-
tined to carry on existing situations or conventions until the end of time.
The ambiguity implies that the future or the future form of a particular con-
vention is never given. Husserl was right to maintain that ‘there is a “field
of freedom” and a “conditioned freedom”,’ Merleau-Ponty concludes, ‘not
because freedom is absolute within the limits of this field and nothing out-
side of it (for just like the perceptual field, this one too has no linear limits),
but because I have immediate possibilities and more distant possibilities’
Community Without Conservatism  77
(PhP 481), or because my non-linear field, precisely because of its being
non-linear, always offers me a number of more or less articulated options
to take up. This even prompts Merleau-Ponty to state, after repeatedly
emphasizing that we can never completely leave our situation behind, that
‘we carry with us—from the mere fact that we are in and toward the world
[. . .]—all that is necessary for transcending ourselves’ (PhP 483).

4.  Community Without Conservatism

As I explained in my introduction, the claim that there is no subject with-
out community raises questions as to the conservative nature of human
existence. Insofar as On Certainty suggests that the social self is indeed
reactionary to the core, the Phenomenology makes clear that the situated
subject is not inherently conservative, unable to change or accept changes
in social arrangements of its own accord. For on Merleau-Ponty’s account,
situations lack a clear delineation and are therefore not the kind of things
one can be imprisoned in or determined by. It is rather up to subjects them-
selves to give direction to the multi-interpretable conventions that happen
to be their own. This by no means guarantees that social arrangements are
never given a static and rigid interpretation; the history of communism itself
testifies to this fact. According to Merleau-Ponty, however—who unsurpris-
ingly came to criticize the Soviet interpretation of Marxism13—making one’s
social ­situation into a static and rigid thing is not the only possible or even
necessary response to its solicitations, and is moreover a response that can
always be undone.
Returning to Wittgenstein, something similar can be said about his social
account of subjectivity. To be sure, one would not automatically associate
On Certainty with indeterminacy and ambiguity, but reading these remarks
next to the Phenomenology allows us bring out things that may otherwise
have gone unnoticed. For while Wittgenstein explains that a particular cer-
tainty only has its foundational function because it is part of a ‘whole sys-
tem’ (OC §141) of presuppositions, the examples he gives of certainties are
numerous and diverse: he for instance mentions mathematical, historical as
well as autobiographical convictions. This means that the certainties that
underlie a person’s doings and sayings do not necessarily make for a neatly
ordered whole without overlaps, gaps, or inconsistencies. A system need not
be entirely seamless to function, and the idea of a well-defined set of presup-
positions could be said to be at odds with Wittgenstein’s emphasis on the
non- or pre-intellectual nature of certainties. As he himself observes: ‘Our
“empirical propositions” do not form a homogeneous mass’ (OC §213).
While the ‘our’ here is equivocal, it can be said to refer as much to every
individual’s as to a specific community’s set of certainties. It is after all in the
course of one’s upbringing that one comes to take these things to stand fast,
and even within a given society, not everyone’s initiation into the practices
of the community will be entirely the same. Those who never had a physics
78  Chantal Bax
class, to give one very simple example, will probably not take the exact
same things for granted as those who did receive such education.14 Like
a particular person’s set of certainties, therefore, a community’s system of
basic presuppositions need not make for a carefully ordered unity without
internal difference.
Even if some of Wittgenstein’s remarks suggest that certainties make for
a strict set of norms that allow no deviation, this means that there is not
necessarily a rigid and uniform whole to conform to in the first place. As a
result—and regardless of the fact that Wittgenstein focuses on examples in
which the contrast between certainties is big and straightforward—it is not
always already clear which new or unconventional doings and sayings can
still be said to be part of the community’s practices, and which ones cannot.
If the community’s certainties do not form a homogeneous mass, it should
be said to be a practical question which perspectives can be included or
excluded, rather than something that can be stated beforehand.
As is the case with Merleau-Ponty’s One, this does not guarantee that
communal practices will never be given a reactionary interpretation. It,
however, does mean that a conservative response to the new or different
is only one among a number of possible reactions, as already became clear
from the final quote I gave in the second section. A person could also decide
to just put up with the difference, for instance, or perhaps be ‘torn away
from the sureness of the game’ (OC §617) and realize that other perspec-
tives on life and the world are possible too. For as became clear from other
remarks I quoted at the end of the second section, Wittgenstein does not
maintain that practices are given once and for all. Contrary to his seeming
conservatism, he points out that certainties can change and become a topic
for discussion rather than unspoken agreement, or as he metaphorically
observes: ‘It might be imagined that some propositions [. . .] were hardened
and functioned as channels for such empirical propositions as were [. . .]
fluid; and that this relation altered with time, in that fluid propositions hard-
ened, and hard ones became fluid’ (OC §96).15 According to Wittgenstein,
in other words, it not a given that what we take for granted now will also
shape our practices in the future. Certainties can lose their indisputability,
and heterogeneity within a community’s world picture provides an explana-
tion for how such breaks or shifts arise.
Both in the case of Wittgenstein and in the case of Merleau-Ponty, then,
the claim that there is no subjectivity without community does not imply
that the self is intrinsically reactionary. Neither maintains that the subject
is predetermined to carry on inherited conventions until the end of time,
because both hold that it is a matter of our actual reactions and responses
how static or dynamic our socio-cultural situation will be. Even if this does
not make humans into inherently progressive beings, it does not condemn us
to conservatism either. Indeed, it does not condemn us to a pre-given view
on all social change at all.
Community Without Conservatism  79
1 See, e.g., Benhabib (1992), who argues that when one takes the subject to be
completely constituted by its community—rather than only being situated in it—
one does away with a locus for crucial ethico-political notions such as agency
and autonomy.
2 See, e.g., Nyiri (1982), Bloor (1983). Critics have also argued that Merleau-
Ponty is insufficiently progressive (see e.g. Butler (1989), Grimshaw (1999)), but
these criticisms mainly concern his notion of the body and amount to the claim
that the latter is, rather than being overly social, by no means social enough.
3 Wittgenstein also points out that the things Moore claims to know normally go
unexpressed, and function as certainties precisely by being of a non-propositional
nature; see e.g. OC 87, 159, 204, 402, 466, 467.
4 In some cases a statement like ‘This is my hand’ does not necessarily go without
saying (e.g., when it concerns amputation or a prosthetic device). According to
Wittgenstein, however, in such cases ‘This is my hand’ no longer (or not yet)
functions as a certainty.
5 See Nyíri (1982).
6 See Monk (1990: 316–17).
7 My understanding of Merleau-Ponty thus differs from that of the critics I men-
tioned in note 2, who argue that Merleau-Ponty’s notion of bodily subjectivity
is reactionary for overlooking the socio-cultural context of human life; I neither
take Merleau-Ponty to ignore the social nature of subjectivity nor to be a conser-
vative philosopher. See Oksala (2006) for a more detailed response to Butler and
8 See PhP 174–8. For more elaborate discussions of the relation between Merleau-
Ponty’s philosophy and his politics, arguing that the latter in fact forms the basis
for the former, see Whiteside (1988) and Coole (2007).
9 Wittgenstein also seems to have had an interest in Marxism; he for instance had
a number of communist friends (including the aforementioned Sraffa) and made
plans to visit the Soviet Union (see Monk 1990: 248, 343). I will not pursue this
particular similarity between Wittgenstein and Merleau-Ponty any further, also
because I am not interested in Merleau-Ponty’s view on communism per se but
only as a case study for his general view on socio-historical change.
10 If the contemporary reader precisely associates communism with conservatism
and repression, it should be noted that Marxism—as will also become clear
below—never was a matter of force or totalitarianism for Merleau-Ponty but
precisely a way to (eventually) overcome struggle and conflict. He acknowledged
the violence committed by the Stalinist regime at an early stage, and while this
did not bring him to reject Marxism immediately (i.e., in Humanism and Terror),
he began to publish criticisms of a dogmatic, Soviet-style interpretation of com-
munism (e.g. Adventures of the Dialectic) in the 1950s; see Flynn (2007).
11 See Compton (1982) for a more thorough discussion of Merleau-Ponty’s debate
with Sartre.
12 It should be noted that idealism can also be understood as a particular manifesta-
tion of objective thought rather than as its opposite, but in the freedom chapter
Merleau-Ponty describes them as being counterparts and does not use ‘objective
thought’ as a label under which idealism can (together with empiricism) be said
to fall.
13 See note 10.
14 Other examples can be given along the lines of Young, who points out that boys
and girls are not taught the same things about themselves and their place in the
world (see Young 1980; while she bases herself on Merleau-Ponty, her point can
80  Chantal Bax
also be made in Wittgensteinian terms, for certainties precisely show themselves
in what people do and say, or in how they do and say things). And as Cavell
explains, when children do not receive the same level or kind of affection, that
will also have an impact on what they come to take for granted, in this case
about inter-human relations (see Cavell 1979: 177).
15 Compare this to Nyiri’s interpretation, according to whom Wittgenstein does
allow for the possibility of change, but only when it emerges organically (Nyíri
1982: 61). That Wittgenstein employs a natural metaphor here however does not
mean that he rules out the possibility of change as a result of our own actions
and reactions. According to the river metaphor, after all, it is precisely because
of movements in the water, or changes at the level of the knowledge claims we
make, that the riverbed of certainties might shift.

Benhabib, S. 1992. Situating the Self: Gender, Community and Postmodernism in
Contemporary Ethics. Cambridge: Polity Press.
Bloor, D. 1983. Wittgenstein: A Social Theory of Knowledge. London: Macmillan
Butler, J. 1989. Sexual ideology and phenomenological description. In J. Allen and
I. M. Young (eds.), The Thinking Muse: Feminism and Modern French Philoso-
phy. Bloomington: Indiana University Press, pp. 85–100.
Cavell, S. 1979. The Claim of Reason: Wittgenstein, Skepticism, Morality, and Trag-
edy. Oxford: Oxford University Press.
Compton, J. J. 1982. Sartre, Merleau-Ponty, and human freedom. The Journal of
Philosophy 79 (10): 577–588.
Coole, D. 2007. Merleau-Ponty and Modern Politics after Anti-Humanism. Lan-
ham: Rowman & Littlefield.
Flynn, B. 2007. The development of the political philosophy of Merleau-Ponty. Con-
tinental Philosophy Review 40: 125–138.
Grimshaw, J. 1999. Working out with Merleau-Ponty. In J. Arthurs and J. Grim-
shaw (eds.), Women’s Bodies: Discipline and Transgression. London: Cassell,
pp. 91–116.
Monk, R. 1990. Ludwig Wittgenstein: The Duty of Genius. London: Jonathan Cape.
Moore, G. E. 1993. Selected Writings. Edited by T. Baldwin. London: Routledge,
pp. 147–170.
Nyíri, J. C. 1982. Wittgenstein’s later work in relation to conservatism. In B. Mc-
Guinness (ed.), Wittgenstein and His Times. Chicago: University of Chicago
Press, pp. 44–68.
Oksala, J. 2006. Female freedom: Can the lived body be emancipated? In D. Olkowski
and G. Weiss (eds.), Feminist Interpretations of Maurice Merleau-Ponty. Univer-
sity Park: Pennsylvania State University Press, pp. 209–228.
Whiteside, K. 1988. Merleau-Ponty and the Foundation of an Existential Politics.
Princeton: Princeton University Press.
Young, I. M. 1980. Throwing like a girl: A phenomenology of feminine body com-
portment motility and spatiality. Human Studies 3 (1): 137–156.
5 The World and I1
Komarine Romdenh-Romluc

Solipsism is the view that the I—my self—is, in some sense, alone. There are
different forms of solipsism, which vary along two dimensions. First, solip-
sistic views can differ in how alone they take the I to be: one might claim
that the I is the only self in a world of objects or non-selves. Or one could
take the I to be all that there is. Second, there are different views one might
take up concerning the nature of these claims. They can be understood as
metaphysical claims about what exists; epistemological claims about what
can be known; or as phenomenological claims about the character of expe-
rience. (These options are not mutually exclusive.) The standard view is
that solipsism in all its varieties is at best, a deeply unattractive position,
and at worst, absurd. It is surely undeniable that I share the world with
other people. Moreover, this fact features in my experience and is know-
able by me. Nevertheless, both Merleau-Ponty and Wittgenstein—two of
the twentieth century’s most profound and interesting thinkers—hold that
solipsism expresses something important about the human condition. My
aim in this paper is to articulate what they take solipsism to express. Much
has been written about Wittgenstein’s views on solipsism (see, e.g., Ans-
combe 1959; Hacker 1986; Diamond 1991; Pears 1996). Merleau-Ponty’s
ideas about solipsism and our relations with others have also received a
fair amount of attention in the literature (see, e.g., Madison 1981; Carman
2008; Romdenh-Romluc 2011; Morris 2012). Thus, one might wonder
what more there is to say on the topic. I hope to show here that revisiting
these ideas is a fruitful enterprise. I will argue that the form of solipsism that
most concerns Merleau-Ponty and Wittgenstein is phenomenological, and
their theories can be used to illuminate each other. Reading them together
yields a single account of human experience that reveals how its structure
makes phenomenological solipsism an ever-present possibility for us.

1. Merleau-Ponty, Embodied Consciousness,

and the Puzzle of Solipsism
Phenomenological solipsism holds that experience is solipsistic: I experience
myself as alone in some significant sense—my experience presents me as
82  Komarine Romdenh-Romluc
being the only self, or as the only one of a special kind of self. One may be
puzzled by my claim that Merleau-Ponty endorses anything like this posi-
tion, since he seeks to accurately describe the phenomenology of our lives,
and it seems that each of us ordinarily experiences herself as just one of
many similar selves who share a common world. Moreover, phenomeno-
logical solipsism seems to be ruled out by Merleau-Ponty’s conception of
consciousness as embodied. Solipsistic experience is instead associated with
the view of the mind and body as distinct. On this conception, I am directly
aware of my own mind through introspection, but I can never directly expe-
rience another. All I ever perceive are other bodies, but these are distinct
from other minds. My experience thus presents me as being the only self in
the world.2 In contrast, Merleau-Ponty conceives of mind and body as inter-
twined aspects of a single embodied consciousness. Since I perceive many
human bodies, my perceptual experience thus presents me as one of many
embodied selves who inhabit the world.
However, there is a difference between perceiving what is in fact another
self, and perceiving it as a self. Merleau-Ponty’s conception of conscious-
ness says nothing about the latter. If I never perceive another as a self, then
my experience has solipsistic tendencies, even if I perceive what are in fact
other selves, and I know this to be the case. Moreover, Merleau-Ponty sug-
gests that there is a solipsistic element to experience, which a thoroughgoing
phenomenology must capture. He writes, ‘we see a lived solipsism that can-
not be transcended’ (PhP 374). He continues, ‘[c]onsciousnesses present the
absurdity of a solipsism-shared-by-many, and such is the situation that must
be understood’ (PhP 376). It is clear that Merleau-Ponty takes the solipsism
in experience to be connected with having a perspective on the world—
something that he takes to be essential to being a subject. In his discussion of
intersubjectivity, he first searches for ways in which my experience presents
me as a subject, and presents others as similar to me. After laying out some
initial suggestions, he then objects to his analysis. He writes,

But is it really the other that we reach in this way? We, in effect, level
out the I and the You in an experience-shared-by-many, we introduce
the impersonal into the center of subjectivity, and we erase the individu-
ality of perspectives—but, in this general conflation, have we not caused
the alter Ego to disappear along with the Ego?
(PhP 372, my italics)

The worry that Merleau-Ponty raises here is that his analysis does not accom-
modate the fact that each self has a particular perspective on the world, and
thus it subsequently fails to be an account of the relations between selves
at all. It is after noting that each self necessarily has a perspective on the
world that he comes to the conclusion that living experience contains a
solipsism. Merleau-Ponty never explains in detail what he takes the connec-
tion between having a perspective and solipsism to be.
The World and I  83
2.  Wittgenstein and Solipsism
Wittgenstein offers an account of human experience as solipsistic, which
explains the connection Merleau-Ponty identifies between solipsistic experi-
ence and perspective.
Initially, one may be surprised by the claim that Wittgenstein endorses
phenomenological solipsism. In the Tractatus, Wittgenstein tells us that
‘the world is my world’ (TLP §5.62), ‘I am my world’ (TLP §5.63), and
that ‘what solipsism means is quite correct’ (TLP §5.62). These statements
amount to an endorsement of the claim that I am my world, which looks
rather like the view that the I is all there is (Bell 1996). Wittgenstein thus
appears to be concerned primarily with a metaphysical sort of solipsism.
Whilst metaphysics and phenomenology are not mutually exclusive, it is not
obvious that Wittgensteinian solipsism is, or is intended to be, an account
of lived experience. Why think, then, that Wittgensteinian solipsism has any
connection with phenomenology?
A reading given by Bell (1996) provides the resources to answer this ques-
tion. He takes Wittgenstein to tacitly accept what he calls the ‘Franklin
Requirement’. This is a condition alluded to by Christine Ladd-Franklin in
a letter she sent to Bertrand Russell. It states that for solipsism to be accept-
able, it must be fact-preserving (Bell 1996: 159). Bell connects this condi-
tion with the Wittgensteinian thesis that the language we use to describe
the world is, to some degree, conventional—i.e., the nature of the world
does not entirely dictate how we must describe it. I can, e.g., describe my
dog as ‘my dog’, or as ‘instantiated doghood’, or as ‘the furry Overlord’.
The nature of the world does not determine which of these descriptions
I should use. It is, of course, more usual to employ the phrase ‘my dog’, but
this is merely a matter of convention. It follows that there can be different
ways of describing exactly the same set of facts. We can thus distinguish
between descriptions of the world that posit a different set of facts to those
we usually take to obtain, and descriptions that acknowledge exactly the
same set of facts, but describe them differently. Bell calls the latter ‘alterna-
tive notations’ (Bell 1996: 163). The Franklin Requirement holds that for
some form of solipsism to be an acceptable philosophical position, it must
be an alternative notation; it must not posit a different set of facts to the
ones we usually take to obtain.
It is important to note that alternative notations are not competing
hypotheses that explain the same set of data. For example, there are two
main theories concerning the extinction of the dinosaurs. One theory states
that this happened quickly as a result of a cataclysmic event, such as an
asteroid colliding with the earth. The second theory states that the dinosaurs
became extinct gradually, over a period of millions of years. Both theories
are consistent with the available evidence. But they are not alternative nota-
tions because they posit different sets of facts. Contrast this example with the
following: ‘Imagine a language in which, instead, of saying “I found nobody
84  Komarine Romdenh-Romluc
in the room”, one said, “I found Mr. Nobody in the room” ’ (BB 69). These
two sentences are alternative notations. They describe exactly the same
fact—that the room is empty of people. In particular, the surface grammar
of the latter sentence should not lead us to think that it posits the existence
of an extra person!
One might think that the difference between competing hypotheses and
alternative notations is clear. However, a little more must be said on how
they are to be distinguished. Contrast the claim that it was mere chance that
led me to choose my dog (rather than any other) with the claim that Fate
brought us together. Are these competing hypotheses or alternative nota-
tions? One might opt for the former on the grounds that the latter posits the
existence of a mysterious force—Fate—which is missing from the former
explanation. Yet one might also argue that they are alternative notations on
the grounds that the positing of Fate makes no practical difference to the
situation. It does not change what can be observed, nor does it make any
difference to how one should behave. The only change that might flow from
a belief in Fate is a difference in one’s attitude towards events. Perhaps one
will be more inclined to accept what comes one’s way with grace, rather
than rail against one’s misfortunes. But even this is by no means certain.
The example raises the question of what it is for a description to be fact-­
preserving in the way required for it to be an alternative notation. Bell takes
Wittgenstein to hold that it must make no difference to our ordinary expe-
rience or practices. According to this criterion, the claim that it was mere
chance I chose Billy from the dog pound, and the alternative view that Fate
brought us together are alternative notations—my ordinary experience and
practices are left untouched, no matter which view I adopt. Thus, on this
reading, the Franklin Requirement holds that for some form of solipsism
to be acceptable, it must be an alternative notation that preserves the facts,
where this means that it must be a description that coheres with our lived
One might then wonder what might be gained from an alternative nota-
tion, given that—by definition—it makes no difference to everyday experi-
ence or practice. Wittgenstein’s answer is that it might reveal something of
philosophical significance. His view is that the essence of our grammar is
the essence of our world, i.e., the world as we find or live it—‘Essence is
expressed by grammar’ (PI I §371). The essence of our grammar is what
alternative notations have in common:

A proposition possesses essential and accidental features.

Accidental features are those that result from the particular way in
which the propositional sign is produced. Essential features are those
without which the expression could not express its sense.
(TLP §3.34)

The method of varying notations allows us to discover what is essential to

our grammar, thus revealing the essence of our world. It is interesting to note
The World and I  85
here that Wittgenstein is proposing something akin to the eidetic reduction.
The eidetic reduction is part of Husserl’s phenomenological method. Its aim
is to reveal the essential structures of consciousness. One does this by imagi-
natively varying the features of the phenomena with which one is presented.
This allows one to distinguish between those features that are essential to
phenomena of the type in question, and those that are merely accidental.
The upshot is that the Franklin Requirement entails that an acceptable
solipsism must be fact-preserving, which means that it must be a descrip-
tion of things that coheres with our everyday experience and practices. If an
alternative, solipsistic notation is available, it will help reveal the essential
structures of human experience. Wittgenstein’s commitment to the Franklin
Requirement, and his understanding of what this entails, means that we can
take his solipsism to be an account of lived experience; we can understand
Wittgenstein as doing phenomenology.
There is a further reason why one might be suspicious of my claim that
Wittgenstein endorsed phenomenological solipsism. The early Wittgenstein
certainly appears to endorse solipsism of some sort, as the remarks cited
above illustrate, but there is broad agreement amongst philosophers that
solipsism is absurd, or at the very least, highly implausible. Many com-
mentators thus deny that Wittgenstein’s statements on this subject should
be taken at face-value.3 Moreover, in the Philosophical Investigations, Witt-
genstein advances a number of considerations, collectively known as ‘the
private language argument’, which are almost universally read as a rejec-
tion of solipsism. Thus it is generally agreed that even if Wittgenstein was
once foolishly tempted into solipsism (and it is not certain that he was),
his mature view was that solipsism is untenable and should be rejected.
However, there is an alternative reading of Wittgenstein available, accord-
ing to which the early Wittgenstein did endorse solipsism, and the private
language argument is a development of the very same position (Bell 1996).
On this reading, Wittgenstein understands consciousness, or the subject,
in line with a well-established philosophical tradition that takes a self to
be essentially bound up with having a perspective or point of view on the
world. The idea has its roots in perceptual experience. Perception is egocen-
tric. One perceives the world as located in space around one. I see my dog
stirring in his bed a short distance to my right; I see a collection of house
plants a metre in front of me; my tactile experience presents the computer
keyboard under my hands, just in front of my body; I perceive the sofa
underneath me; and so on. It is part of this experience that I am aware of
myself as located relative to the things I perceive. I am aware of myself as
located at the centre of egocentric space—at the point from which things are
seen. Thus to be the subject of perceptual experience is to have a point of
view on the perceived world.
Wittgenstein then takes this notion of the subject to lead to the conclusion
that the self is, as it is often put, ‘systematically elusive’—one cannot come
across the self (qua self) in either inner or outer space. A classic presentation
of the Elusiveness Thesis is found in this passage from Hume.
86  Komarine Romdenh-Romluc
For my part, when I enter most intimately into what I call myself, I always
stumble on some perception or other, of heat or cold, light or shade, love
or hatred, pain or pleasure. I never catch myself at any time without a
perception, and never can observe anything but the perception.
(Hume 1978: 252)

Some version of this thesis has been endorsed by various philosophers,

including Kant (1929), Sartre (1989), and Ryle (1949). Various arguments
are offered in support of the Elusiveness Thesis. But the one on which Witt-
genstein relies turns on the thought that a perspective is not an item that one
can come across in experience. One perceives from a perspective, but this
perspective cannot itself be perceived. Anything one may come across in the
world is something on which one has a perspective; it is not identical with
that perspective. Wittgenstein writes,

Where in the world is a metaphysical subject to be found?

You say that this is exactly the case of the eye and the visual field.
But really you do not see the eye.
And nothing in the visual field allows you to infer that it is seen by the
(TLP §§5.633–5.6331)

We find the same thought expressed elsewhere too,

Visual space has essentially no owner . . . The essential thing is that

the representation of visual space is the representation of an object and
contains no suggestion of a subject.
(PR 100)

Wittgenstein assumes these points are true, in particular, of one’s body. One
can perceive one’s own body, but this means that one has a perspective on it.
One does not experience it as the perspective from which things are perceived,

If I wrote a book called ‘The world as I found it’, I should also have
therein to report on my body and say which members obey my will and
which do not, etc. This then would be a method of isolating the subject
or rather showing that in an important sense there is no subject: that is
to say, of it alone in this book mention could not be made. If I wrote a
book called The world as I Found It.
(TLP §5.631)

It follows that since the self is the point of view on the world, the self cannot
be experienced as part of the world. In other words, the perspective on the
world that is the I has no extension—it is the geometric point from which
things are seen. ‘The I . . . shrinks to an extensionless point’ (TLP §5.64).
The World and I  87
The idea of the self as a geometric point of view grounds Wittgenstein’s
claim that I am the world. Bell points out that this claim can be read in two
ways. It is natural to read it as an identification of the world with the self, so
that the world ‘disappears’, leaving only the I. This is, of course, the usual
way to understand the strongest form of metaphysical solipsism: as the view
that the self is all there is. But Wittgenstein’s claim can also be read in the
other ‘direction’ so that the I disappears, leaving only the world. A point of
view is essentially a point of view on something. There can be no perspec-
tive without the existence of that upon which it is a point of view. In this
sense, the perspective can be identified with that on which it is a perspective:
the world. The latter is the reading intended by Wittgenstein—‘The self of
solipsism shrinks to a point without extension, and there remains the reality
co-ordinated with it’ (TLP §5.64). Read in this way, the identification of the
I with the world leaves the world exactly as it is, and so, ‘solipsism strictly
carried out coincides with pure realism’ (TLP §5.64).
One might suppose that Wittgenstein’s claim that only the world is
left means that my psychological life—the experiences, thoughts, desires,
wishes, emotions, and so on that I undergo every day—has been erased
from the picture. Since my psychological life is an important feature of my
ordinary experience, a description of things that left no place for it would
violate the Franklin Requirement. However, Bell argues that Wittgenstein’s
solipsism does not deny the existence of my psychological life. Instead, it
offers a different description of it to the one offered by the traditional pic-
ture. Since Descartes, it is usual to think of subjectivity as an inner realm,
distinct from the external world, and accessible only to the subject. My
psychological states populate my inner realm. Just as external space con-
tains dogs, cars, camels, trees, and all the other entities one encounters as
one makes one’s way about the world, so too inner space contains one’s
thoughts, feelings, desires, and so on. Whilst this picture is compelling,
many writers—­including Wittgenstein—have identified numerous difficul-
ties with it. Wittgenstein’s solipsism implies a rejection of this picture. The
shrinking of the self to an extensionless point is the disappearance of inner
space. The subject as a point of view on the world has no inner volume.
But this does not mean that one’s ‘inner’ states vanish. Instead, one should
understand one’s psychological life as happening in the world.
Bell (1996) connects this point with Wittgenstein’s no-ownership view of
the self, which denies that my experiences, thoughts, and other psychologi-
cal phenomena are owned.4 Once the self is conceived as a point of view
on the world, this denial makes sense. (Recall Wittgenstein’s remark that
‘visual space has essentially no owner’ (PR 100).) Experiences are under-
gone from a perspective. For example, you and I both experience my love
for my dog. But we each have a different perspective on it. You see it in my
gestures when I engage with him; the loving look on my face when I talk
about him; the concern in my voice as I speak to the vet when he is ill. I, on
the other hand, am the one who loves him, and I experience my love from
88  Komarine Romdenh-Romluc
this point of view. It manifests in such things as the lovable, familiar look
of my dog; the sense of urgency that pervades the world when he is ill and
needs medical attention. But a perspective is not the sort of thing that can
own experience. Wittgenstein writes,

One of the misleading representational techniques in our language is the

use of the word ‘I’, particularly when it is used in representing immedi-
ate experience . . . It would be instructive to replace this way of speak-
ing with another in which immediate experience would be representing
without using the personal pronoun, for then we’d be able to see that
the previous representation wasn’t essential to the facts.
(PR 88)

These ideas point to the following picture of how one’s psychological life
manifests to one. It is experienced as all out in the world, and on a par with
other worldly events:

Amongst other things, the world as I find it contains thoughts, feelings

of misery and happiness, sensations, perceptions, desires, and the like.
Some of these will indeed be contingently related to a particular body
[found at the Centre]—but then again, some of them will be contin-
gently related to ambient temperature, say, or the presence of food, or
the fortunes of Sheffield Wednesday. In addition to thoughts, feelings,
sensations and the like, which are events in the world, there will also
be such things as football matches, hurricanes, and eclipses of the sun.
And these too are events in the world. For me, as the Centre, none of
these occurrences involves any identification of an owner, or bearer, or
subject who has them. From this new perspective [that of Wittgenstein’s
alternative solipsistic notation] a hurricane is just as little, or as much,
“mine” as is a headache.
(Bell 1996: 65)

Finally, one may wonder how this position has any call to be named
‘solipsism’ since the I has disappeared, leaving only the world. The answer
to this is that the world that is left behind when the I disappears is ‘the real-
ity co-ordinated with’ the point of view that is me (TLP §5.64). In other
words, it is an essentially perspectival world. In this sense, it is my world.
I am the zero point around which it revolves. Bell tells us that the world is
one ‘pervaded’ by my subjectivity (Bell 1996: 168). This is what Wittgen-
stein means when he says, ‘The I occurs in philosophy through the fact that
“the world is my world” ’ (TLP §5.641).
I claimed earlier that Wittgenstein’s endorsement of the Franklin Require-
ment means that he intends his solipsism to be a phenomenological account
of experience, so to defend it, we must show that it does indeed capture
some aspect of human life. Bell (1996) suggests that it is an alternative
The World and I  89
notation describing the experience of any human subject, because adopting
this description leaves our ordinary experience and practices untouched.
But this does not seem credible. However, phenomenology—at least since
Merleau-Ponty—does not discriminate against any type of experiences. The
so-called pathological, the extraordinary, the rare, are all grist to the phe-
nomenologist’s mill in trying to understand the essential structure of human
subjectivity. It follows that a phenomenological defence of Wittgenstein’s
solipsism need not show that it captures the experience of any human sub-
ject. It will be sufficient to show that it correctly describes some form of
human experience. I will show that Wittgenstein’s solipsism accurately cap-
tures certain experiences that are characteristic of schizophrenia.5
On Wittgenstein’s picture, the I cannot be identified with the bodily self.
I am a perspective on the world, whilst the body is just another thing in the
world on which I have a perspective. This claim implies alienation from my
own body—it is not experienced as me. It implies further that my body is
experienced as lacking in subjectivity. In other words, it is presented to me
as a mere object, such that there is no essential difference between it and
inanimate things like rocks and chairs. Wittgenstein makes this alienation
vivid in a remark quoted above.

If I wrote a book called ‘The world as I found it’, I should also have
therein to report on my body and say which members obey my will and
which do not, etc. This then would be a method of isolating the subject
or rather showing that in an important sense there is no subject: that is
to say, of it alone in this book mention could not be made.
(TLP §5.631)

In this remark, Wittgenstein distinguishes the subject from the body and
conjures up an image of the latter as a puppet whose strings are pulled
by one’s will. Alienation from the body is typical of certain characteristic
forms of schizophrenic experience. Stanghellini states that ‘in schizophrenic
states, the body . . . is experienced as a sort of object that is detached from
the prime initiator of the movement, their actions detached from the energy
that should spontaneously feed it’ (Stanghellini 2004: 157). Schizophrenic
alienation from one’s own body, which is experienced as a mere object, is
expressed in the following first-person reports, ‘[There’s] no inside of the
body, but only a frame . . . food is falling into a vacuum . . . behind the
chest is nothing, only a big hole’ (Angyal 1936: 1042). ‘I’m blessed with a
bladder-emptyer that I can turn on and off, and an anal expeller’ (Stang-
hellini 2004: 155).
A second component to Wittgenstein’s solipsism is the notion of my own
psychological goings-on as events I encounter in the world, which are just as
much or as little ‘mine’ as hurricanes, and other impersonal worldly events.
My psychological life is not experienced as owned by me. This claim again
implies alienation—this time separation from one’s own psychological life.
90  Komarine Romdenh-Romluc
Once more, we find that this is characteristic of schizophrenic experience,
as the following remarks attest. ‘My first personal life has been lost and
replaced by a third person perspective’ (Parnas 2000: 124). ‘It is not me
the one who feels’, and ‘It is not me who feels—It feels’ (Stanghellini 2004:
157). ‘Feelings are not felt by me, things are not seen by me, only by my
eyes’ (Spitzer 1990: 393–4). ‘[I doubt] that it is me thinking, myself’ (Rosser
1979: 182).
The third component is the identification of the self with a bare point of
view on the surrounding space—a perspective on the world that is not part
of it—and the correlative claim that since the world is experienced from
my own perspective, it is my world. The alienation from one’s own body
in schizophrenia is an experienced separation from one’s bodily insertion
in the world. Unsurprisingly, this often goes hand in hand with experience
of oneself as a mere point of view on the world—a distant spectator who
does not interact with the spectacle. Moreover, just as Wittgenstein con-
nects the notion of the self as point of view with the solipsistic conception
of the world as my world, so too, we find that the schizophrenic experience
of oneself as a mere perspective is often accompanied by the sense that the
world is mine. Sass writes, ‘perhaps the most emblematic delusion of this
enigmatic illness [schizophrenia] is of being a sort of God-machine, a kind
of all-seeing, all-constituting camera eye’ (1999: 320). The experience is
expressed in statements such as the following. ‘I feel like an emperor inside
a pyramid’ (Stanghellini 2004: 155). ‘My thoughts can influence things’,
‘This event happens because I think it’, ‘To keep the world going, I must not
stop thinking’ (Spitz 1990: 393–4). ‘I could create the events of my universe
by just thinking them, believing them to be true . . . What really terrified me
was when I realized that I could conceive of wrenching the world from its
axis’ (Sass 1999: 330).
In summary, Wittgenstein’s solipsism takes as its starting point the idea
that the self is a point of view. Once this notion is in place, his solipsistic
picture flows from it. Any item one experiences is something on which one
has a perspective; it is not identical with that perspective. The self is thus
divorced from the body and slips out of the world. One’s psychological
life has to be part of the world on which one has a perspective—there is
no ‘where’ else to situate it. To experience the world from a perspective
is to experience oneself as the origin of egocentric space, with the world
laid out around one. In other words, it is to experience oneself as the zero
point, around which the world revolves. The I is thus a Godlike view on
its own world. Wittgenstein’s solipsistic notation is confirmed as an accu-
rate description of lived experience by the characteristic forms of schizo-
phrenic experience identified above, which are often found together in the
experience of a single subject (Sass 1999; Stanghellini 2004). However,
Wittgenstein’s solipsism cannot be a complete phenomenological account
because the schizophrenic experiences that confirm it stand in stark contrast
The World and I  91
to non-schizophrenic experience, which presents the subject as just one of
many similar selves who share a common world. Something is missing from
Wittgenstein’s account. To see what this is, we must return to Merleau-
Ponty’s philosophy.

3.  Merleau-Ponty and the Body

Merleau-Ponty, like Wittgenstein, identifies the subject with a point of view
on the world. But unlike Wittgenstein, he pays close attention to the experi-
ence of one’s own body, and it is this which—for most of us—mitigates the
solipsism that flows from being a perspective on the world. There are two
features of bodily experience that are particularly relevant in the present
context: the awareness of one’s perspective as embodied, and the transfer of
the body schema to other people. I will consider both of these in turn.
On Wittgenstein’s solipsistic picture, one’s body is experienced as just
one of the items in the world on which one has a perspective. His descrip-
tion implies alienation from one’s own body, which is experienced as apart
from, rather than identical with, oneself. The schizophrenic experiences that
fit this description highlight the fact that such alienation is absent from at
least some forms of non-schizophrenic awareness. Merleau-Ponty accounts
for this by describing one’s awareness of one’s perspective not as an exten-
sionless point of view, but as embodied. One might initially take Merleau-
Ponty’s claim to be the simple observation that one experiences one’s own
body as located in the same place as the point from which one’s perspective
originates. However, his claim runs deeper than this. To see the world from
a perspective is to experience items in the world as located relative to where
one appears to be. If one merely experienced one’s body as being in the same
place as one’s point of view, one would still experience it as located relative
to one’s apparent location. It would still be an object on which one had
a perspective—rather than being experienced as the embodiment of one’s
point of view. Merleau-Ponty holds instead that one experiences one’s body
as being the origin of egocentric space. It is presented as the zero point that
establishes its co-ordinates; it is not presented relative to those co-ordinates.
To put matters more simply, my body is not presented as just another item
that is located relative to me; I perceive worldly items as located relative to
my body.
The thesis that I experience my body as being the origin or zero point of
egocentric space means that it must feature in my perceptual field without
being something on which I simply have a perspective. Initially, this claim
may strike one as implausible. I experience it using the same senses that
I use to experience other worldly items—I can see, touch, hear, etc., my own
body. My senses give me an awareness of worldly things as located around
me in space—they provide me with a perspective on those things. Thus it
seems my sensory awareness of my body must provide me with a perspective
92  Komarine Romdenh-Romluc
on it. However, whilst Merleau-Ponty does not deny that I can experience
my body using my senses, he rejects this simple account of the situation. He
begins by considering what it is like for one’s two hands to touch each other,
and makes the following observations:

[T]he two hands are never simultaneously both touched and touching.
So when I press my two hands together, it is not a question of two
sensations that I could feel together, as when we perceive two objects
juxtaposed, but rather of an ambiguous organization where the two
hands can alternate between the functions of ‘touching’ and ‘touched’.
In speaking of ‘double sensations’, psychologists mean that, in the pas-
sage from one function to the other, I can recognize the touched hand as
the same hand that is about to be touching.
(PhP 95)

In this passage, Merleau-Ponty claims that when my two hands touch one
another, one is the object of tactile experience (the touched), whilst the other
is the subject (the touching). Moreover, there is a phenomenological differ-
ence between the two roles. I am aware of the touched hand in the same sort
of way that I am aware of other worldly objects when I touch them. But
my awareness of the touching hand is different. I am aware of the touching
hand as the subject of experience. I am aware of it, in other words, as part
of me. This means that in my experience, there can be no ‘distance’ between
me and the touching hand. There is always ‘distance’ between me and an
object on which I have a perspective, since experiencing oneself as having a
perspective on something is to experience it as located relative to me, and to
experience it as located relative to myself is to experience myself as separate
from it. It follows that when I experience my hand touching, my awareness
cannot present it as something on which I have a point of view.
How should we understand this experience? The difference lies in the
structure of awareness. It is usual to think of awareness as having what is
known as an ‘act-object’ structure. On this model, to be aware of something
involves an act of awareness that is directed at that thing. Perceptual aware-
ness of things in the world has this structure. Seeing a table, e.g., involves an
act of awareness that is directed at the table. Seeing Shmoo the cat involves
an act of awareness that is directed at Shmoo. But there is an alternative way
that awareness may be structured. On this account—what Moran (2001)
calls the ‘adverbial model’—to be conscious of x does not involve an act of
awareness directed at x. Instead, it describes the kind of thing or activity that
x is—a conscious thing or activity. We can get some purchase on this idea if
we consider what it is to dance joyfully. It is implausible to suppose that this
should be analysed as involving an act of joy that is directed at one’s danc-
ing. This implies that the dancing is independent from the joy, so that one
could engage in the very same dancing, without its being joyful. One could
dance sadly, e.g., by directing an act of sadness at the very same dancing.
The World and I  93
But this is surely wrong. To dance joyfully is for one’s dancing to have par-
ticular joyful qualities, which are different from those it has if one dances
sadly. Thus a better analysis takes ‘joyfully’ to describe the way in which
one dances, i.e., the sort of dancing it is. On the adverbial model of aware-
ness, certain cases of what it is to be conscious of something are treated in
the same way as joyful dancing. They do not involve acts of awareness that
are directed at some thing or activity. Instead, ‘conscious’ describes the sort
of thing or activity it is. The difference between the experience of being
touched, and the experience of touching can be analysed using this distinc-
tion. My awareness of my touching hand has an adverbial structure. It does
not consist in an act of awareness that is directed at my hand as its object.
I am not, in other words, aware of my hand. Instead, I am conscious in, or
better, my hand is sensitive. It follows that there is a form of bodily aware-
ness that does not present me with my body as an object on which I have a
perspective, because it is not, properly speaking, awareness of my body at
all. Instead, it is awareness in my body, or my body’s sensitivity.
My awareness in my body, or my body’s sensitivity, is the experience of
the subject of perception (myself) as a bodily being. As such, it is awareness
of myself as an embodied subject who is part of the world. This sense of
myself as a part of the world is reinforced by what Merleau-Ponty refers to
as the ‘double nature of sensations’. My experience of my body can switch
between the two forms of awareness, so that the touched can become the
touching, and vice versa. In this way, I experience myself, the touching sub-
ject, as something that—like all the worldly items I encounter—can be an
object for touching, and as such, something that is located in, and part of a
world of touchable objects.
Merleau-Ponty’s account of how one experiences one’s point of view as
embodied inserts the I back into the world. But it is not sufficient to com-
pletely dispel solipsism and capture the nature of non-schizophrenic experi-
ence. On his account so far, the subject is no longer a godlike perspective
on the world but a flesh-and-blood part of it. However, recall that on Witt-
genstein’s account, it is not just the elusiveness of the subject that leads to
solipsism. Experience is also solipsistic because it is perspectival and to be
aware of the world as laid out in space around one is to experience oneself
as the centre of that world. The fact that one experiences one’s perspective
as bodily does not alter the privileging of one’s own perspective as the zero
point around which the world revolves. The solipsism that flows from hav-
ing a point of view must be mitigated in non-schizophrenic experience by an
awareness that somehow de-centres one’s own perspective. Merleau-Ponty
holds that to de-centre the point from which I view the world, I must be able
to occupy the perspective of others. As we saw earlier, I experience my point
of view in perception as being the perspective of my body—I experience
my point of view as embodied. Correlatively, to occupy the perspective of
another must be to experience the other’s bodily perspective; it must be to
inhabit the other’s body. Merleau-Ponty explains how this is possible with
94  Komarine Romdenh-Romluc
his account of the momentary transfer of one’s body schema to the bodies
of others.
Merleau-Ponty draws on Schilder’s (1950) work on the body image to
make his case. Merleau-Ponty prefers the term ‘body schema’—schéma
­corporel—which emphasises its role in providing a framework for the sub-
ject’s experience of the world. In what follows, I will use his preferred ter-
minology. For Merleau-Ponty, the body schema is one’s embodied sense of
self; it is one’s bodily ‘grasp’ of the sort of creature one is. It comprises one’s
motor abilities; one’s sense of how one appears from the outside; cultural
beliefs the subject holds about her body; her emotional attitudes towards it;
etc. The body schema can be usefully thought of as comprising two com-
ponents: (i) a reasonably settled sense of oneself as an embodied creature
of a certain sort, with particular properties; (ii) a continuous awareness of
one’s changing posture as one moves through the world. The body schema
is only partly available to consciousness. Whilst I have used the term ‘grasp’
to describe it, the above definition should make it clear that it is not wholly
cognitive, although the subject’s thoughts about her body can enter into it.
Similarly, whilst I have described it as the ‘sense’ one has of one’s body, it is
not a wholly perceptual phenomenon either.
There is clearly an essential connection between one’s body schema and
the physical object that is one’s body. However, if the body schema simply
tracks and reflects one’s physical body, then it is unclear how it could be
transferred to another person. To understand how such a transfer is possi-
ble, we need to know more about the relation—as Merleau-Ponty conceives
it—between one’s physical body and one’s body schema.
His understanding of their relation is grounded in an important distinc-
tion: that between the living body (Leib) and the objective body (Körper).
The latter is the body conceived as a mere physical object, whilst the former
is the body as the subject experiences and lives through it. The subject’s
body is her means of being in the world. Her experience of it, therefore, is
not awareness of an object separate from herself; it is her conscious inhabit-
ing of the world. The fact that her body is her means of living in the world
means that her body as she experiences it (her living body) does not simply
follow the objective body’s contours. Moreover, this is not some defect in
her bodily awareness (i.e., we should not class all such experiences as illu-
sory) rather, this is what it is to live through the body. We can get some
initial purchase on this idea by considering a different case: the experience
of time passing. The units of time—seconds, minutes, hours—are uniform;
each second (minute, hour) is the same length as any other. But one does not
always experience them as uniform. It is a commonplace truism that time
passes quickly when one is doing things one loves and drags when one is
engaged in boring activity. More pronounced experiences of time stretch-
ing or speeding up often accompany the use of certain drugs. Such tempo-
ral experiences do not ‘match’ the uniformity of time’s units. But it would
be odd to thereby class them all as illusory. We might class very extreme
The World and I  95
experiences of time’s stretching or compressing as illusory, but in the major-
ity of cases we do not, because the stretching or compressing of time in
different contexts just is the way that we live in time. Merleau-Ponty makes
a roughly analogous claim about the subject’s experience of her own body.
There may indeed be cases where the subject’s experience of her body is so
different from its objective properties that it should be classed as illusory
(an example might be cases of lycanthropy where the subject experiences
herself as being a wolf). But in less extreme cases, we should not class bodily
experience that does not exactly map on to the objective body’s properties
as illusory because these differences are simply to do with the way that the
body is lived.
There are many examples to illustrate the way in which the subject’s expe-
rience of her own body does not exactly match its objective properties. I will
provide just two here. The first is given by Schilder (1950). After injuring his
hand in a car accident, he came to experience the space around it as having
a special significance, so that his hand hurt if people came too close to it.
Schilder states that after the accident, his sense of himself as an embodied
subject changed so that it included a sense of his hand as vulnerable. He
suggests that this illustrates a more general fact. The space immediately sur-
rounding one’s physical body is experienced as part of the living body. Sec-
ond, the subject lives her body partly as a collection of capacities for action.
Her capacities can be extended through the use of tools. When the subject
habitually uses a tool, her sense of herself expands to take in the tool, which
is subsequently experienced as part of her embodied self, albeit temporar-
ily. For example, just as one does not need to consciously calculate whether
a gap is wide enough to walk through—one can immediately see whether
or not one’s body will fit through it—so too, when one is practised at rid-
ing a bike, one can immediately see whether a gap between two bollards
is wide enough to cycle through. Moreover, one can have proprioceptive
(or perhaps quasi-proprioceptive) sensations in a tool. Schilder (1950) notes
that the blind person senses the ground with the end of their stick in much
the same way that one senses objects by touching them with one’s hands
(or other parts of one’s body). He no longer feels the stick in his hands as
an intermediary between himself and the world. Instead, sensations in his
hands are immediately transformed to yield experience of the world at the
surface of the stick.
The body schema is bound up with the subject’s experience of her body.
Component (ii) of the body schema is the continuous awareness of one’s
changing posture as one moves through the world (i.e., it just is the ongoing
experience of one’s body). Component (i) is the reasonably settled sense of
oneself as an embodied creature of a certain sort. We can think of compo-
nent (i) as affecting (ii) (and so affecting the experience of one’s body). In the
case of Schilder’s injured hand, his (i) sense of himself as an embodied sub-
ject changed to include a sense of his hand as vulnerable. This then affected
(ii) how he experienced his bodily self, so that the space around his hand
96  Komarine Romdenh-Romluc
took on a special significance for him. Similarly, we can think of a tool user’s
(i) sense of herself altering as she becomes habituated to using the tool so
that this then affects (ii) and she experiences the tool as a temporary part of
her own body. The body schema’s role in bodily experience means that it is
what underlies the living body. Since the living body does not exactly map
on to the contours of the objective body, the body schema’s function cannot
be to simply track and reflect the latter’s properties.
The distinction between the living body and the objective body, and the
body schema’s role in the former, opens up the possibility of the body sche-
ma’s expansion beyond the boundaries of the objective body. Indeed, this is
what happens in the two examples given above. The body schema expands
to encompass the space surrounding the objective body, and the tools habit-
ually used by the subject. Merleau-Ponty continues by claiming that not
only can the body schema be extended, it can also be momentarily trans-
ferred to the body of another subject. This allows the subject to perceive the
world as for-the-other. To understand this claim, we need to know a little
more about Merleau-Ponty’s account of perception and the body schema’s
role in it. Perception, for Merleau-Ponty, never presents the perceiver with
a ‘neutral’ world of things. Instead, she is confronted with an environment
that solicits her to act in various ways. She sees food as inviting her to eat
it; footballs appear to solicit kicking; and she perceives her dog as drawing
her to stroke and feed him. Perception has this character because it involves
the exercise of skills that have both a perceptual and motor component.
Merleau-Ponty holds that when the perceiver acquires a motor skill—such
as the ability to ride a bicycle—she also acquires the ability to perceive the
world in certain ways. She can perceive certain environments as appropri-
ate for exercising that skill. In the example given above, she perceives bikes
as for riding, and different sorts of terrain as more or less appropriate for
riding along. On this view, each perceptible object corresponds to a set of
the perceiver’s motor capacities—those that can be used to interact with it.
Actually perceiving the object involves using these motor capacities. We saw
above that Merleau-Ponty takes the body schema to incorporate the body’s
‘sense’ of what it can do, so the body schema is partly constituted by the
body’s motor abilities. Thus on his account, perceiving an object involves
using one’s body schema.
Merleau-Ponty then argues that when sees another subject, one’s body
schema is transferred momentarily to that person. This allows one to see the
surrounding world as soliciting the other to act:

My gaze falls upon a living body performing an action and the objects
that surround it immediately receive a new layer of signification: they
are no longer merely what I could do with them, they are also what this
behaviour is about to do with them.
(PhP 369)
The World and I  97
Merleau-Ponty holds that to see the world as for-the-other in this way is to
see the other as a centre of action, and so to see her as another ­subjectivity—
another view of the world. Correspondingly, to see the world as for-the-
other is to see it as laid out around this other perspective. This decentres my
own point of view, so that I no longer experience it as the zero point around
which the world revolves.

A vortex forms around the perceived body into which my world is

drawn and, so to speak, sucked in: to this extent, my world is no longer
merely mine, it is no longer present only to me, it is present to X, to
this other behaviour that begins to take shape in it. The other body is
already no longer a simple fragment of the world, but rather . . . a cer-
tain ‘view’ of the world. A certain handling of things—which were until
now mine alone—is taking place over there. Someone is using my famil-
iar objects. But who? I say that it is another person, a second myself.
(PhP 369–70)

4. Conclusion
Both Merleau-Ponty and Wittgenstein take there to be an important con-
nection between having a perspective on the world and solipsism. Merleau-
Ponty never makes this connection clear, but Wittgenstein offers an account
of it. On his solipsistic picture, the I is a bare perspective on the world and
cannot be identical with anything in it, including one’s body. One’s psycho-
logical life is part of the world on which one has a perspective, and no more
or less ‘owned’ by me than other worldly events, such as cyclones and foot-
ball matches. The world is my world because it is one that revolves around
me—the insubstantial, godlike point of view at its centre. Wittgenstein’s
solipsism beautifully captures certain experiences that are characteristic of
schizophrenia, but it is incomplete on its own, since it does not provide an
analysis of non-schizophrenic experience. Merleau-Ponty’s account of bodily
experience explains how the solipsism that flows from being a perspective on
the world is mitigated in non-schizophrenic experience. In the latter, the sub-
ject does not experience herself as a bare point of view on the world. Instead,
she experiences her point of view as embodied. This is made possible by
adverbial awareness of her own body. Second, the subject is able to experi-
ence the world as revolving around other subjects through the momentary
transfer of her body schema. This decentres her own perspective, so that she
experiences the world as one that is equally shared by others.

1 I would like to thank audiences at the University of Amsterdam, the University of
Sheffield, and Sussex University for their insightful comments on earlier drafts of
this paper, which is improved as a result.
98  Komarine Romdenh-Romluc
2 The conception of the mind and body as distinct is usually associated with episte-
mological solipsism: the view that I cannot know that other minds exist. However,
this lack of knowledge flows from the experiential situation. I cannot know that
other minds exist because the only mind I can experience is my own.
3  See, e.g., Pears (1987, 1996), Diamond (1991, 2000).
4 This thesis is associated primarily with Wittgenstein’s later writings such as Philo-
sophical Remarks, Blue and Brown Books, and Philosophical Investigations.
5 Sass (1993) also suggests that Wittgenstein’s solipsism can be employed to make
sense of certain important and puzzling aspects of schizophrenia. Unfortunately,
space prevents me from engaging with his ideas here. However, it is worth noting
an important difference between our accounts. Sass’s proposal is that the logic of
schizophrenic thinking is analogous to the logic of Wittgenstein’s solipsism. My
proposal, in contrast, is that Wittgenstein’s solipsism is a (partial) account of the
structure of experience, which (partially) explains how certain kinds of schizo-
phrenic experience have the character they do. I take it that this difference in our
approaches means that mine is able to escape some of the criticisms leveled at Sass
by Read (2001), although it is clear the latter would disagree with the interpreta-
tion of Wittgenstein offered here.

Angyal, A. 1936. The experience of the body-self in schizophrenia. Archives of Neu-
rology and Psychiatry 35 (5): 1029–1059.
Anscombe, G. E. 1959. An Introduction to Wittgenstein’s Tractatus. London:
Hutchinson University Library.
Bell, D. 1996. Solipsism and subjectivity. European Journal of Philosophy 4 (2):
Carman, T. 2008. Merleau-Ponty. London: Routledge.
Diamond, C. 1991. The Realistic Spirit. Cambridge, MA: MIT Press.
Diamond, C. 2000. Ethics, imagination and the method of the Tractatus. In A. Crary
and R. Read (eds.), The New Wittgenstein. London: Routledge, pp. 149–173.
Hacker, P. 1986. Insight and Illusion. Oxford: Clarendon Press.
Hume, D. 1978. A Treatise of Human Nature. Edited by L. A. Selby-Bigge, Revised
by P. H. Nidditch. Oxford: Oxford University Press.
Kant, I. 1929. Critique of Pure Reason. Translated by N. Kemp Smith. London:
Palgrave Macmillan.
Madison, G. B. 1981. The Phenomenology of Merleau-Ponty. Athens, OH: Ohio
University Press.
Merleau-Ponty, M. 1962. Phenomenology of Perception. Translated by Colin Smith.
London: Routledge.
Moran, R. 2001. Authority and Estrangement. Princeton: Princeton University Press.
Morris, K. 2012. Starting with Merleau-Ponty. London: Continuum.
Parnas, J. 2000. The self and intentionality in the pre-psychotic stages of schizophre-
nia. In D. Zahavi (ed.), Exploring the Self: Philosophical and Psychopathological
Perspectives on Self-Experience. Amsterdam: John Benjamins, pp. 115–147.
Pears, D. 1987. The False Prison: A Study of the Development of Wittgenstein’s
Philosophy, Volume I. Oxford: Clarendon Press.
Pears, D. 1996. The originality of Wittgenstein’s investigation of solipsism. Euro-
pean Journal of Philosophy 4 (2): 124–136.
Read, R. 2001. Approaching schizophrenia through Wittgenstein. Philosophical
Psychology 14 (4): 449–475.
The World and I  99
Romdenh-Romluc, K. 2011. Routledge GuideBook to Merleau-Ponty and Phenom-
enology of Perception. London: Routledge.
Rosser, R. 1979. The psychopathology of feeling and thinking in a schizophrenic.
International Journal of Psychoanalysis 60: 177–188.
Ryle, G. 1949. The Concept of Mind. Chicago: University of Chicago Press.
Sartre, J. P. 1989. Being and Nothingness: An Essay on Phenomenological Ontol-
ogy. Translated by H. Barnes. London: Routledge.
Sass, L. 1993. The Paradoxes of Delusion: Wittgenstein, Schreber, and the Schizo-
phrenic. Ithaca: Cornell University Press.
Sass, L. 1999. Schizophrenia, self-consciousness, and the modern mind. In S. Gal-
lagher and J. Shear (eds.), Models of the Self. Thorverton: Imprint Academic,
pp. 319–341.
Schilder, P. 1950. The Image and Appearance of the Human Body. London:
Spitzer, M. 1990. On defining delusions. Comprehensive Psychiatry 31: 377–397.
Stanghellini, G. 2004. Disembodied Spirits and Deanimated Bodies. Oxford: O
­ xford
University Press.
6 Painting and the Promiscuity
of Vision
Taylor Carman

Henry Moore did not start by looking at his model. He started by looking
at his stone.
—E. H. Gombrich

‘Every theory of painting is a metaphysics,’ writes Merleau-Ponty in his last

published work, Eye and Mind.1 His particular case in point is a passage
in the Optics (Descartes 1985) in which Descartes seeks to undermine the
Aristotelian-scholastic theory of sensible forms by severing (almost) all con-
nection between pictorial representation and visible resemblance. Descartes
observes (correctly) that pictures not only need not resemble their objects,
but often represent them precisely by not resembling them: ‘You can see this
in the case of engravings: consisting simply of a little ink placed here and
there on a piece of paper, they represent to us forests, towns, people, and
even battles and storms.’ In fact, an image cannot in principle resemble its
object too closely, for if it did, ‘there would be no distinction between the
object and its image’ (Descartes 1985: 113).
Descartes’s larger argument in the context is that seeing an object cannot
be a function of having in your mind a picture that resembles the object,
since not even physical pictures represent things simply by resembling
them. For a rationalist like Descartes, perceiving is thinking, and the unique
force and directness with which images seem to impress us is an illusion
of common sense, for ‘our mind can be stimulated by many things other
than images—by signs and words, for example, which in no way resemble
the things they signify’ (Descartes 1985: 112). Descartes’s critique of the
scholastic image theory of perception is devastating, but his argument goes
beyond that critique to the point of assimilating perception to cognition
and blurring the distinction between pictures and symbols: ‘The magic of
intentional species,’ Merleau-Ponty remarks, ‘loses its final argument if the
entire potential of a painting is that of a text to be read, a text totally free of
promiscuity between the seeing and the seen’ (PrP 171).
What is the ‘promiscuity’ of seeing? Merleau-Ponty expresses the same
idea, though more chastely, when he writes in Phenomenology of Perception
Painting and the Promiscuity of Vision  101
that ‘sensation is, literally, a communion’ of my body with my ‘existential
environment’ (PhP 219). Unlike texts, pictures materially embody and enact
vision’s free mingling with the visible, putting us in concrete contact with the
world. Perception is our immersion in the inexhaustible expanse of visible
reality that—far from being somehow ‘conceptual all the way out,’ as John
McDowell puts it—engulfs us, beneath and beyond the reach of thought
and representation (McDowell 1992: 8 n7, 11, 13, 42, 67, 69).
That immersion is what Merleau-Ponty later called the ‘flesh’ (chair)
common to ourselves and the world we perceive, inhabit, and are part of.
The perceptual environment is not the same as the space of reasons, and
our unthinking identification with it is precisely what pictures, unlike mere
signs, are able to evoke, so that we literally see in them—as opposed to read-
ing off, deciphering, or inferring from them—something other than them-
selves. It is what allows pictures to show rather than merely say.
Wittgenstein’s Tractatus, that austere blend of logic and mysticism,
seemingly worlds away from Merleau-Ponty’s phenomenological and
aesthetic essays, also invokes a distinction between saying and show-
ing. It is not exactly the distinction we find in Eye and Mind and The
Visible and the Invisible, but neither is it entirely different. Indeed, it
seems to me, Wittgenstein’s concept of showing in the Tractatus sheds
light on Merleau-Ponty’s denial that perception is a form of cognition,
and so too on his insight that the art of painting, both figurative and
abstract, involves an evocation of the world that differs fundamentally
both from discursive meaning and from the mere similarity of a picture
to what it depicts. Painting, as Merleau-Ponty conceives it, summons up
or makes manifest the world neither by describing nor by imitating it, but
by embodying its visibility—not unlike the way in which propositions,
according to Wittgenstein, mirror reality not by specifying or describing
it correctly, but by ‘displaying’ or ‘exhibiting’ (aufweisen) its logical form
(TLP §4.12).
An important clue to what Wittgenstein has in mind is the fact that he
does not merely distinguish showing from saying but insists further that
‘What can be shown, cannot be said’ (TLP §4.1212). The kind of ‘show-
ing’ he has in mind cannot therefore be our normal concept of depiction or
illustration—that is, our ordinary notion of the representational content of
an image, which often enough, after all, can also be stated or expressed. To
mirror, display, or exhibit in Wittgenstein’s sense is instead something for-
mal, something constitutive of the image’s being an image, its simultaneous
continuity and discontinuity with its object, its uncanny ability to partici-
pate in what it represents while yet standing far enough apart from it as to
be able to represent it. That distinctive perceptual manifestation of images,
their necessarily unstated surplus or residue of visual presence, I want to
suggest, is what Merleau-Ponty means by ‘promiscuity,’ or as he also says,
‘the metamorphosis of seeing and seen that defines both our flesh and the
vocation of the painter’ (PrP 168–9).
102  Taylor Carman
The Tractatus is famous for advancing (or seeming to advance) a ‘picture
theory’ of meaning.2 On Wittgenstein’s account, sentences (or propositions,
the German word Satz is ambiguous) have meaning, or say something, by
representing (darstellen) possible states of affairs, which in turn constitute
the world (‘The world is all that is the case’ (TLP §1)). True propositions
represent existent states of affairs, or what Wittgenstein calls ‘positive facts’
(TLP §2.06). Simply put, ‘A proposition is a picture of reality’ (TLP §4.021).
In addition to representing the world, however, propositions also display,
exhibit, or as Wittgenstein says, reflect or mirror (spiegeln) ‘what they must
have in common with reality in order to be able to represent it—­logical
form’ (TLP §4.12). ‘Propositions cannot represent logical form’ (TLP
§4.12), but instead ‘show the logical form of reality’ (TLP §4.12). What
some sentences—for example, A = A and S is p and not p—show in their
form is precisely that they cannot be showing the logical form of the world,
and so cannot be stating facts of any kind: ‘tautologies and contradictions
show that they say nothing’ (TLP §4.461). Only in this technical sense of
the word ‘show’ can Wittgenstein plausibly say, as he does emphatically,
that what can be shown ‘cannot be said’. Why not? What kind of showing
not only differs from saying, but excludes it?
In the ordinary sense of the word, a picture often ‘shows’ what can also be
said—for example, George Washington, or George Washington crossing the
Delaware, maybe even the fact that George Washington crossed the Dela-
ware. How does it do so? By being a painting of or about the person, the
deed, the fact. In depicting what it depicts, however, the picture necessarily
does something else, something in addition, namely, it presents itself as a
picture. To represent anything, that is, the picture must also exhibit itself
in such a way as to make manifest or plain that it is a picture (of whatever
it is a picture of). But that—what it ‘shows’ in Wittgenstein’s sense—is not,
indeed cannot be, what it represents. That it is a picture at all cannot be
included in what it depicts.
Or rather, more precisely, if a picture does somehow manage to depict that
it is a depiction of something, it can do so only by, again, manifesting itself,
‘showing’ itself in a way that is not contained in its representational content.
Artworks often exhibit sophisticated forms of self-reference. Andy Warhol’s
Brillo Box is a classic example, at least on Arthur Danto’s suggestion that it
poses a question about itself, namely, the question, What makes it a work of
art? (Danto 1964). Michelangelo’s ‘unfinished’ sculptures are another, since
they are arguably not just sculptures of partially revealed human forms, but
also sculptures of themselves revealing those forms. At a minimum, figura-
tive sculptures depict forms by being sculptures, by presenting themselves as
sculpted representations. If Michelangelo’s Awakening Slave can be said to
depict its own status as a sculpture, however, it must do so in some further
way, for example by remaining—in dramatic contrast, say, to the highly
polished Pietà—conspicuously unfinished, deliberately rough hewn so as
to expose its own sculptedness. Could we suppose that not just its being a
Painting and the Promiscuity of Vision  103
sculpture, but moreover its remaining conspicuously unfinished is also part
of what it depicts? Perhaps, but it can depict that only by manifesting or
presenting itself in some other way yet again, which in turn must fall outside
the scope of the depiction. If the sculpture comments on its own unfinished-
ness, for example, it does so (in part) simply by being made of marble—and
that fact is not something that marble can represent just by being marble.
A chunk of marble is not a sculpture of a chunk of marble, just as a piece of
paper is not a picture of a piece of paper. A picture, an image of any kind, is
always, cannot fail to be, more than what it depicts.
The bare fact of a representation’s being a representation at all, then, must
manifest itself and as it were, go without saying: ‘A picture cannot . . . depict
its pictorial form: it displays it’ (TLP §2.172). So too, in contrast to its con-
tent, or what it says, a proposition must simply display or exhibit its logical
form: ‘The rules of logical syntax must go without saying’ (TLP §3.334).
Representations, that is, carry a surplus of exhibition, inasmuch as there
is always more in how they manifest themselves than they can represent,
more in what they show of themselves than they can say. It is that surplus
of exhibition, that excess of manifestation over and beyond representation,
that Wittgenstein has in mind with his concept of showing. Propositions can
say something, can have meaning at all, only by showing themselves to be
propositions without saying that they are, just as pictures must simply show
themselves to be pictures without somehow representing—being pictures
of—that fact.
Keeping this essential fact about pictures in view helps stave off some
persistent confusions concerning the nature of pictorial representation.
One such confusion, as Descartes clearly understood, is the naïve (indeed
absurd) idea that depiction just is similarity. Another, the intellectualist
reaction that Merleau-Ponty finds in Descartes’s comments in the Optics
(Descartes 1985) concerning engraving, is the temptation to conflate images
with signs, the perceptual with the linguistic, to reduce visual depiction to
discursive description, token pictures to symbol types. Both sorts of con-
fusion, the mimetic and the symbolic conceptions of representation, are
intimately bound up with traditional theories of perception. Aristotelians
and empiricists sought to understand seeing on analogy with picturing: to
see something is to have an image in your mind that resembles the object
you see.3 Rationalists and intellectualists maintained, on the contrary, that
experience, vision included, is a form of judgment, its content conceptual.
According to contemporary theories of mind and language, perceptual con-
tent is literally symbolic, indeed computational.4
Although the temptation to think of perceiving as somehow confronting
(or perhaps drawing) a picture may be quite archaic, it gained much of its
historical momentum from the invention of linear perspective in Renais-
sance painting. Ancient theorists did not know, for example, that the only
thing that literally gets into our eyes in vision is light, not colours or sensible
forms peeling away from objects, as it were, traversing the distance between
104  Taylor Carman
them and us, and finally entering and inhabiting the soul. It therefore never
occurred to them that a perceptual picture might have to be projected—
somewhere, somehow—onto a flat surface, as opposed to the round (outer)
surface of the eye.5 This was not a trivial difference, for it meant that they
did understand, contrary to the prescriptions of linear technique centuries
later, that straight architectural lines often look curved, tapering or sagging
in the middle, and that an object twice as far away occupies half the angle
of what Euclid called the ‘visual cone,’ not half the distance along a flat
perpendicular surface.6
As Erwin Panofsky argued in his seminal essay Perspective as Symbolic
Form (Panofsky 1991), the invention and refinement of linear perspective
in the fifteenth century cannot therefore be understood simply as providing
a closer, truer approximation to perceptual experience. It was instead an
ingenious abstraction, a rationalization, a mathematization of visual space,
an idealized reconstruction that has come to seem natural, indeed transpar-
ent, to us only because our culture has so thoroughly absorbed and general-
ized the systematically unified conception of space that later flourished in
modern philosophy and science, especially Descartes and Kant. The reputa-
tion for phenomenological correctness that linear perspective has enjoyed
over the centuries, however, has wreaked havoc on theories of vision, most
egregiously perhaps in Berkeley’s argument that we literally see a two-­
dimensional mosaic of token images, and only then infer or impose upon it
an additional idea of depth.
Pictures, of course, are not mere copies of objects or simulations of visual
experience.7 According to Pliny the Elder, Zeuxis, in a contest with Parrha-
sius, painted grapes so realistically that birds flew down to eat them. Invited
to remove the curtain veiling Parrhasius’s painting, however, Zeuxis dis-
covered that it was in fact a painting of a curtain. He had been fooled, and
Parrhasius won the contest. Like the bird, Zeuxis failed to see the picture as
a picture. One dollar bill is not a picture of the other, nor, contrary to a com-
mon misconception, does pictorial representation normally aspire to trompe
l’oeil. Extremely schematic stick figures in children’s drawings are pictures,
even when they function more as conventional symbols than as visible like-
nesses: a circle is a person’s head or the sun, a triangle is a tree or the roof of
a house. Picturing, as Nelson Goodman argues, is at least in some minimal
sense a semantic phenomenon: ‘Denotation is the core of representation and
is independent of resemblance’ (1976: 5).
Goodman is surely wrong, though, to assimilate pictures to conventional
symbols as thoroughly as he does. Following Panofsky, he reminds us of the
many ways in which linear perspective departs from and distorts real visual
perception: it projects the scene onto a plane; it draws receding parallel
lines, but often neither verticals nor horizontals, as converging; it substi-
tutes a single, stationary, instantaneous point of view for our two eyes and
mobile bodies. And so on. And yet, even if Goodman is right, as I think he
is, to insist that linear (artificial) perspective cannot claim to get it uniquely
Painting and the Promiscuity of Vision  105
right about how things visually appear to us, it is hard to deny that it is at
least one way of getting it right about some of the structures inherent in real
(natural) perspective. As J. J. Gibson has said, ‘it is one thing to argue that
perspective is not necessary for a painting, but it is quite another to say that
perspective is a language’ (1986: 285). Artificial linear perspective is a con-
vention, but it is also a genuine innovation, since it does indeed mimic—if
only instantaneously and for a single, motionless eye—what Gibson calls
the ‘ambient array’ of light, which, according to his ‘ecological optics,’ pro-
vides the structural invariants that allow us to see (and more essentially, see
our way around in) the practical environment. Properly visual information,
Gibson insists, ‘is not explicit. The invariants cannot be put into words or
symbols’ (1986: 285).
Gibson, for his part, has been widely criticized for failing to recognize
the depth of the contribution we ourselves make to what we see, in contrast
to what the environment provides by way of lighting and scenery. That
criticism often comes from proponents of cognitivist theories of perception,
according to which what we contribute to vision is something like concep-
tual thought or information processing. But one needn’t be an intellectualist
or computationalist to suspect that Gibson has indeed underestimated the
role of our own spontaneity in perception, particularly what we might call
the imaginative dimension of visual experience, over and beyond its mere
responsiveness to given optical invariants.
Like Gibson, Arthur Danto rejects Goodman’s conventionalism by
observing that some of the same ‘ “perceptual pathways” are involved in
learning to recognize pictures of things the outlines of which are learned in
the perception of those things as such’ (Danto 1999: 114). Children who
grow up in a picture-free environment have little trouble recognizing things
in pictures on first viewing, and apparently even pigeons can sort some pic-
tures according to what they depict—or perhaps we should say, just as they
sort the objects depicted in the pictures.8 More to the point, Danto observes,
‘sentences, unlike pictures, are in a language,’ and languages, at least in lexi-
con and surface grammar, are conventional, whereas the ability to recognize
things depicted in pictures appears to be, at least in part, an innate capacity
(Danto 1999: 116).
Danto’s purpose is twofold. His main thesis is that while Wittgenstein
might well be right that there are things that can be shown but not said,
there are also things that can be said but not shown. His example is negative
facts (think of Sartre’s example of seeing that Pierre is not in the café), but
he could as easily have mentioned conjunction, disjunction, conditionals,
quantifiers, and tense. (Try drawing a picture of the fact that if Pierre is not
in the café, then I will have to borrow money from someone else.) But Danto
also wants to distinguish what we might call ‘brute pigeon perception’ from
our distinctively human ability to ‘read’ rather than merely see pictures, an
ability essential to our identifying some pictures as works of art, that is,
as artefacts demanding a special kind of interpretation and so enjoying a
106  Taylor Carman
special status as art works—as, in Danto’s words, transfigured rather than
merely commonplace (Danto 1983).
But the distinction between pigeon perception on the one hand, and the
interpretation of art works on the other, leaves out something crucial in
between: our mundane recognition of pictures as pictures. Danto writes, ‘it
is difficult to see in what way pictorial competence can differ from what one
might call perceptual competence’ (Danto 1999: 114). And yet what is dis-
tinctive of human beings, in contrast to pigeons, and even other higher apes,
is precisely our ordinary perceptual grasp of the duality of pictures, that is,
their both having representational content and being perceptible things in
their own right.9 This basic grasp of pictures as pictures is unique to human
beings, yet far more elementary than the specialized practice of interpreting
works of art. Basic pictorial understanding requires a kind of intelligence
that goes beyond the mere animal ability to recognize things depicted in pic-
tures, yet falls short of our hermeneutical understanding of things belonging
to that special ontological space that Danto calls ‘the artworld.’
It may well be that human beings have an innate capacity not only to
recognize things depicted in pictures, a capacity we share with pigeons, but
also to recognize pictures as pictures of something, even very schematic
combinations of lines and dots as faces, for example. Yet it is also surely
true that many pictures, both paintings and photographs, represent their
objects by means of stylized conventions, that there are culturally variable
vocabularies and grammars of depiction, and that we are not born with but
acquire, or even learn, the skills necessary for seeing them properly. Appre-
ciating Impressionism as a challenge to traditional concepts and techniques
of painting is an achievement of aesthetic understanding, but even recogniz-
ing Monet’s water lily paintings as paintings of water lilies might require
some prompting and demonstration, perhaps by the titles of the paintings
themselves—not to mention some prior familiarity with ponds and water
lilies. And yet Monet’s images are indeed visual images. Their intelligibility
is a specifically visual kind of intelligibility. They are not mere symbols. Like
linear perspective, Impressionism is a stylized convention, but not literally
a language.
What all this shows is that the dilemma we seem to face in conceiving of
pictures as either mere duplicates or reiterations of perception on the one
hand, or as thoroughly conventional symbol systems on the other—that is,
as either mechanical copies of visual experience or as arbitrarily contrived
languages—is a false dilemma. Merleau-Ponty, both in his phenomenologi-
cal account of perception and in his remarks on painting and language,
bypasses that dilemma and advances an original and plausible conception of
pictorial representation that recognizes its essential semantic basis and yet
maintains a robust distinction between images and signs.
Moreover, he maintains that both pictures and symbols occupy a con-
tinuum extending from direct, concrete involvement with the world to
more and more remote, transparent forms of depiction and description.
Painting and the Promiscuity of Vision  107
The world is at once perceptual and conceptual, conducive to both pictur-
ing and speaking, neither of which is essentially closer to or farther from the
world they mutually articulate. Each, however, may be relatively at grips
and expressively engaged with the world or—with increasing reflection and
calculation—more objective, either correct or incorrect.
In The Voices of Silence, the 1951 edition of his four volumes of art his-
tory and criticism originally published together under the title, The Psychol-
ogy of Art, André Malraux maintains that all artistic productions ‘speak
for the same endeavor; it is as though an unseen presence, the spirit of art,
were urging all on the same quest’ (1978: 46). In his great essay, Indirect
Language and the Voices of Silence, Merleau-Ponty replies that this can only
be half right, at best. For while it is true that all art forms are expressive,
visual, and linguistic expression differ not just in the means or materials
available to them, but in the very relations in which they stand to the world.
Malraux criticizes the ‘objectivist prejudice’ according to which meaning
simply adheres to things, awaiting reflection or imitation in art or language.
He is right that meaning is not an objective property of things, but neither
does it lie solely in the subjective moment of creative expression. Perhaps,
Merleau-Ponty suggests, Malraux ‘has not measured how deeply the preju-
dice is rooted. Perhaps he was too quick to concede the domain of the visible
world to it’ (S 47).
To get clear about the relation between images and words, and so between
visual art and literature, we must distinguish between two questions. First,
how is visual art rooted in the mute, nonsymbolic world of ordinary expe-
rience? Second, how does visual art nevertheless acquire at least some of
the expressiveness of language and literature? These two questions can be
collapsed into a single question, implicit in all of Merleau-Ponty’s essays on
painting—namely, how does visual art manage to speak to us? Merleau-
Ponty says, ‘language speaks, and the voices of painting are the voices of
silence’ (S 81). Is the phrase ‘voices of silence’ more than an oxymoron?
Insisting on a fundamental distinction between visual and symbolic con-
tent, Gibson writes, ‘There is no way of describing the awareness of being
in the environment at a certain place. Novelists attempt it, of course, but
they cannot put you in the picture in anything like the way the painter can’
(1986: 285). The nonsymbolic content of vision differs essentially from the
symbolic content of language. Gibson’s remark raises the question, how are
visual works of art able to capture our concrete sense of being somewhere,
given that they are neither reenactments of full-fledged visual experiences
nor mere signs?
I think Merleau-Ponty would agree with Gibson that there is a funda-
mental difference between pictures and symbols, for while linguistic signs
rely on general and abstract terms, images are bound to the concrete scenes
they depict. Language does not just reveal, it also refers arbitrarily: ‘a state-
ment purports to reveal the thing itself, it goes beyond itself toward what it
signifies’ (S 81). Language has something in common with ‘mute forms of
108  Taylor Carman
expression such as gestures or paintings’ (S 81), yet it differs from them in
allowing (at least approximate) synonymy, ‘the substitution of equivalent
meanings (sens)’ (S 81). Painting and perception, by contrast, are inarticu-
late, for they do not admit of the substitutivity of terms.
Yet, as Merleau-Ponty insists, painting does after all have a kind of voice,
somehow akin to literature, and he is, in the end, less interested in how
they differ than in what they have in common, namely, ‘the phenomenon of
expression’ (SNS 20). In exploring their rootedness in perception, he wants
to rescue them from the false dichotomy between the (supposedly) bare giv-
ens of sense experience and the (supposedly) pure abstractions of the intel-
lect. Perception is not a dumb confrontation with sensory input, and ‘no
thought ever detaches itself entirely from a support’ in our concrete relation
to the world (PrP 189). To see the wrongness of those distorted images
of perception and cognition, we need to appreciate the primitive expres-
sive intelligence at work in both vision and language, and in both art and
literature. We need to look beneath the difference between perceptual and
semantic content to see their common origins in expression and style.
Consider language. Language is not just an abstract system of signs.
According to the structural linguist Ferdinand de Saussure, signifiers signify
only in virtue of the system of differences among them, never by directly
expressing some discrete semantic content. And yet, Merleau-Ponty insists,
our experience of speaking and listening testifies to ‘the power speaking
subjects have of going beyond signs toward their meaning (sens). Signs do
not simply evoke other signs for us, on and on without end, nor is language
like a prison we are locked up in’ (S 81). We experience and understand
language as opening us onto a world, and no theory of syntax or semantics
should tempt us to dismiss that experience as an illusion.
Putting words together in speech is not just a matter of manipulating sym-
bols according to an algorithm but is more like painting. For both speaking
and painting are ways of evoking, ways of rendering things freshly visible.
The writer’s task, Merleau-Ponty says, is to apprehend and make the world
manifest through language, and in this sense, ‘his procedure is not so dif-
ferent from the painter’s’ (S 45). We might suppose that a painting is mere
colour and line and cannot say anything, unlike a poem or a novel, which
is composed of an established system of signs. Yet language is meaningful
not just as a function of the combination of signifiers, but thanks to quasi-­
perceptual effects such as mood, inflection, and silence. Speech emerges
against ‘a background of silence that does not cease to surround it and
without which it would say nothing’ (S 46). Spoken language is ‘simply
the highest point of a tacit and implicit accumulation of the same sort as
painting. . . . Like a painting, a novel expresses tacitly’ (S 76). The silent
significance of Julien’s actions in The Red and the Black, for example, is ‘not
in the words at all: it is between them, in the hollows of space, time, and
signification they mark out, just as movement at the cinema is between the
stationary images that follow one another’ (S 76). Literary language has a
‘halo of signification’ comparable to ‘the mute radiance of painting’ (S 78).
Painting and the Promiscuity of Vision  109
In addition to the explicit, articulate language of words and sentences,
Merleau-Ponty therefore maintains, ‘there is a tacit language, and painting
speaks in this way’ (S 47).
Indirect Language and the Voices of Silence is dedicated to Jean-Paul
­Sartre and can also be read as a reply to his essay, What Is Literature?
(Sartre 1988). There Sartre draws a sharp distinction between art and writ-
ing, poetry and prose. The artist is concerned only with appearances. ‘He
is therefore as far as he can be from considering colors and signs as a lan-
guage’ (Sartre 1988: 26). Like Merleau-Ponty, Sartre observes that linguistic
meaning permits the synonymy of different expressions: ‘the significance of
a ­melody—if one can still speak of significance—is nothing outside the mel-
ody itself, unlike ideas, which can be adequately rendered in several ways’
(Sartre 1988: 27). Language is an instrument for disclosing facts, truths
about the way the world is, whereas painting merely uncovers the appear-
ance of concrete particulars: ‘The writer can guide you and, if he describes
a hovel, make it seem the symbol of social injustice and provoke your indig-
nation. The painter is mute. He presents you with a hovel, that’s all’ (Sar-
tre 1988: 27). Sartre’s distinction, however, is not between linguistic and
visual representation as such, but between significative and aesthetic uses
of representation, between denotation and decoration, prose and poetry:
‘The empire of signs is prose; poetry is on the side of painting, sculpture,
and music. . . . Poets are men who refuse to utilize language,’ for the poetic
attitude ‘considers words as things and not as signs’ (Sartre 1988: 28–9).
Merleau-Ponty rejects Sartre’s distinction by insisting, first, that visual
arts like painting and sculpture do have a kind of voice of their own, that
they never merely display, but also (so to speak) speak of the things they
show; and second, that no use of language, no matter how artless or prosaic,
is literally without style, a mere transparent signifying instrument. What
vision and painting, art and literature, poetry and prose all share is a way of
seeing, a character, a style.
Even our normal ways of seeing and hearing, not to mention watching
and listening, are imbued with a certain character. Prior to any special effort,
‘perception already stylizes’ (S 54). How? By means of an ‘inner schema’
(S 53), a ‘system of equivalences’ (S 54) that coordinates one’s grip on things
and allows the world to reveal itself as coherent and intelligible. The body
schema is a bundle of flexible but enduring dispositions that organize ordi-
nary perception and behaviour. Likewise, there are more refined acquired
schemas that generate the styles immediately recognizable in artistic works.
So, for example,

Our handwriting is recognizable whether we trace letters on paper with

three fingers of our hand or in chalk on the blackboard at arm’s length,
for it is not a purely mechanical movement of our body . . . but a general
power of motor formulation capable of the transpositions that make up
the constancy of style.
(S 65)
110  Taylor Carman
Expression—affective, linguistic, or artistic—presupposes a world given in
perception, yet perception itself always already has expressive significance
of its own, for the body brings a distinct style of comportment to its appre-
hension of what it perceives: ‘if expression recreates and transforms, the
same was already true . . . of our perception of the world before painting,
since that perception already marked things with the trace of human elabo-
ration’ (S 59). We each find ourselves with subtly unique and individually
recognizable ways of walking, talking, and seeing things, and the creative
expression of artists is a further deliberate refinement of those characteristic
dispositions: ‘For each painter, style is the system of equivalences that he sets
up for himself’ (S 54).
Creative expression takes place not in the artist’s mind, but in her con-
crete engagement with the world:

The work is not brought to fulfillment far from things and in some
intimate laboratory to which the painter and the painter alone has the
key. . . . he always goes back to his world, as if the principle of the
equivalences by means of which he is going to manifest it had been
buried there since the beginning of time.
(S 55)

The cultivated body schema of the artist is a kind of second nature, a set of
acquired yet spontaneous skills, skills that come to feel natural though they
are in fact products of years of effort and practice. Carving out a unique
artistic style worthy of the name, over and beyond one’s everyday personal
style of moving and speaking, is like learning a second language. Merleau-
Ponty therefore refers to ‘the painter’s labor and study, that effort that is
so like an effort of thought and that allows us to speak of a language of
painting’ (S 55). If painting can be called a kind of language, it is a language
learned with reference to the more primitive means of expression inherent
in ordinary perceptual behaviour.
Vision is itself already essentially expressive, for it always has its own
bodily character, its own style. It is no more a brutally given natural fact
than any (so-called) ‘natural’ language, no more fixed and surveyable in all
its possible forms and purposes. Like speech, vision ‘moves itself, a means
that invents its own ends’ (PrP 165). This is why painting can never simply
duplicate the structure and content of visual experience. Even Renaissance
painters knew that their perspectiva artificialis was not simply a copy of
perception, ‘that no technique of perspective is an exact solution,’ that lin-
ear perspective is not a uniquely correct mode of representation, but rather
‘opened several pathways for painting’ (PrP 174). Painting is a mode of
creative expression, not a means of technically reproducing what we see
in our actual bodily engagement with the world: ‘Geometrical perspective
is no more the only way of looking at the sensible world than the classical
portrait is the only view of man’ (PW 53).
Painting and the Promiscuity of Vision  111
It is worth noting that since painting almost inevitably acquires and gener-
ates symbolic content of some kind, it can never simply capture the content
of visual perception with no additional symbolic import. Images and icons
are always more or less bound up with discourse in such a way that they
never merely reveal the world, but also always allude, refer, indicate, and
comment. Consequently, ‘no means of expression, once mastered, resolves
the problems of painting or transforms it into a technique, for no symbolic
form ever functions as a stimulus’ (PrP 175).
Malraux is therefore right to deny that meaning simply inheres in the
world, that there is a kind of natural language of things, which the arts sim-
ply echo or reflect. Unfortunately, he counters that crude objectivist preju-
dice, which he thinks defined classical art, with an equally crude subjectivist
interpretation of modern art as withdrawing from all concrete engagement
with the world and retreating into an inner sanctum of subjectivity, into ‘a
secret life outside the world’ (S 47).
Merleau-Ponty considers that dialectical reversal not just internally inco-
herent, but wrong even as a description of the works of art Malraux dis-
cusses. The difference between classical and modern painting cannot be
understood as the difference between objective and subjective. Abstract
painting, for example, is in no way detached or removed from the visible
world. Instead, visible features of a representationally indeterminate pic-
ture suggest or evoke a palpable visible presence—a hint of depth, space,
distance, light or shadow, coolness or warmth, perhaps weight, fragility,
softness, inflexibility, brittleness—yet without specifying any thing, prop-
erty, or even kind of thing as the object of depiction. Jackson Pollock’s drip
paintings, even Piet Mondrian’s austere black lines and coloured rectangles,
though they shed all trace of representational character and float free of spe-
cific objective reference, still nonetheless manifestly evoke and bring forth in
their own visible presence the visible presence of the external world. There
is nothing essentially ‘subjective’ about that. Or better, like Byzantine icon
painting and classical perspective, abstract art was always as objective as it
was subjective, as outwardly world-directed as it was spiritual and interior.
Not surprisingly, Malraux’s flawed notion of the subjectivity of modern
art is parasitic on a correspondingly flawed notion of the objectivity of clas-
sical styles. Malraux assumes that sensory input as such has remained more
or less constant through history and so conceives of the classical ideal as an
effort to reflect and reproduce that input. As Merleau-Ponty argues at length
in Phenomenology of Perception, however, the very idea of determinately
given sense data is confused, for it is meant to satisfy two competing, often
conflicting, identity criteria: that of the sensory stimulus and that of the
phenomenal appearance. The moon looks bigger on the horizon than at its
zenith, though its angular diameter, hence stimulus value, remains constant.
So, how big does it really look?
Classical painters cannot simply have been trying to mimic or duplicate
the world. Instead, Merleau-Ponty writes, following Panofsky, ‘classical
112  Taylor Carman
perspective is just one of the ways man has invented for projecting the per-
ceived world before him, it is not the copy of that world. It is an optional
interpretation of spontaneous vision’ (S 48–9). Granted, some paintings
approximate the formal and material properties of natural vision more
closely than others. But this is like saying that some particular pieces of
music express emotion better than others. That is true, but it does not imply
that there is a single musical form or tradition best suited to the expression
of emotion. Neither is there a single aesthetic style best suited to the evoca-
tion of visual experience. Given some artistic techniques and resources, one
can always do better or worse, but nature does not—indeed cannot—prefer
one style over others a priori.
Pictures depict; they do not literally describe. Nor do they depict by
describing, as language does. Instead, they express—or better yet, they pres-
ent and depict by expressing—the non-symbolic significance things have for
us in perception thanks to the stylized expressive habits of our bodily skills.
More precisely, pictures present and depict the visible world by transposing
perceptual meanings from their original sensorimotor terms into the largely
conventional idioms of pictorial practice. No doubt expression has natural
origins in instinctive behaviour, but any transposition of the spatial world of
perception onto the flat surface of a picture is also necessarily conventional,
beholden to its object yet at the same time, at least in part, arbitrary.

1 Published in Primacy of Perception, 159–90.
2 As Hans Sluga points out, the picture theory was Bertrand Russell’s before it was
(or seemed to be) Wittgenstein’s. I say ‘seeming’ and ‘seemed’ because, as Sluga
says, Wittgenstein would have rejected the word ‘theory’ as a characterization of
his own view (Sluga 2011: 37 n14).
3 Although Hume conceives of experience generally in terms of images, his use of
the word ‘impression’ is a vestige of Aristotelian theories of perception that took
the sense of touch rather than sight as their paradigm.
4 Jerry Fodor, for instance, candidly acknowledges the ‘Way of Ideas’ pedigree
of his own computational theory of intentionality as involving a ‘Language of
Thought’ syntactically realized in the brain: ‘To a first approximation, the idea
that there are mental representations is the idea that there are Ideas minus the
idea that Ideas are images’ (Fodor 1998: 8).
5 The round inner surface of the retina, which seemed to serve as a projection
screen inside the camera obscura of the eyeball, was unknown before Kepler.
6 Repeated doublings of, say, the width of a figure on a picture plane will not cor-
respond to doublings of the angle of the visual cone, since each enlargement of
the figure out to the edges of the (flat) picture will require a smaller and smaller
peripheral increase in the angle of view. The two values would correspond only
if the picture surface were a concave sphere. See Panofsky (1991).
7 Moreover, to argue that a picture is either a copy of an object or the simulation
of a visual experience would of course undermine the image-theory of percep-
tion, since that theory is predicated on the idea that we understand perception
better by referring it back to the supposedly more basic concept of a picture. If
pictures then turn out to be replicas of objects or perceptions, we are moving in
a circle.
Painting and the Promiscuity of Vision  113
8 On picture-deprived children, see Hochberg (1972: 70), cited in Danto (1999:
112). On pigeons, see Herrnstein et al. (1976), cited in Danto (1999: 113).
9 See Gibson (1986: 281–3). The distinction corresponds roughly to the distinc-
tion Descartes draws between the objective and the formal reality of an idea, that
is, its representational content and its intrinsic character.

Danto, A. 1964. The artworld. The Journal of Philosophy 61: 571–584.
Danto, A. 1983. The Transfiguration of the Commonplace: A Philosophy of Art.
Cambridge, MA: Harvard University Press.
Danto, A. 1999. Depiction and description. In his The Body/Body Problem: ­Selected
Essays. Berkeley: University of California Press, pp. 98–121. (Originally pub-
lished in 1982, in Philosophy and Phenomenological Research 43 (1): 1–18.)
Descartes, R. 1985. Optics, IV. In John Cottingham, Robert Stoothoff, and Dugald
Murdoch (eds. and trans.), The Philosophical Writings of Descartes, Volume 1.
Cambridge: Cambridge University Press, pp. 65–66.
Fodor, J. 1998. Concepts: Where Cognitive Science Went Wrong. Oxford: Claren-
don Press.
Gibson, J. J. 1986. The Ecological Approach to Visual Perception. Hillsdale: Law-
rence Erlbaum.
Goodman, N. 1976. Languages of Art: An Approach to a Theory of Symbols.
­Indianapolis: Hackett.
Herrnstein, R. J., Loveland, D. H. and Cable, C. 1976. Natural concepts in pigeons.
Journal of Experimental Psychology: Animal Behavior Processes 2 (4): 285–301.
Hochberg, J. 1972. The representation of things and people. In E. H. Gombrich,
J. Hochberg, and M. Black (eds.), Art, Perception, and Reality. Baltimore: Johns
Hopkins University Press, pp. 47–94.
Malraux, A. 1978. The Voices of Silence. Translated by S. Gilbert. Princeton: Prince­
ton University Press.
McDowell, J. 1992. Mind and World. Cambridge, MA: Harvard University Press.
Panofsky, E. 1991. Perspective as Symbolic Form. Translated by C. S. Wood. New
York: Zone Books.
Sartre, J. P. 1988. “What Is Literature?” and Other Essays. Edited by S. Unga. Cam-
bridge, MA: Harvard University Press.
Sluga, H. 2011. Wittgenstein. Oxford: Wiley-Blackwell.
7 The Recovery of Indeterminacy in
Merleau-Ponty and Wittgenstein1
David R. Cerbone

1.  Negative Indeterminacy: Quine

Within philosophy, the discovery or acknowledgement of indeterminacy2 is
generally regarded as something negative, a revelation often offered with the
aim of debunking some cherished idea or ideal and meant to be greeted with
disappointment.3 Consider the most famous example in the history of 20th
century analytic philosophy: Quine’s thesis of the indeterminacy of transla-
tion. As developed in the second chapter of Word and Object (Quine 1960)
and subsequently revisited and defended throughout his writings, Quine’s
thesis is that for any ‘target language’ it is possible to construct two or more
mutually incompatible manuals for translating that language into a second,
home (for the translator) language. Quine further claims that when it comes
to adjudicating between such rival manuals, there is no fact of the matter
as to which of the two or more rivals is the correct manual; each of the
manuals is equally compatible with all observed and observable data, and
so translation is in principle indeterminate. For Quine, this is not merely a
conclusion about translation, such that it is a problem only for field linguists
and the like; rather, the saga of the field linguist sheds light on the phenom-
enon of meaning as such. That translation is indeterminate shows more
generally that questions of meaning—of sameness of meaning—are not fully
or properly objective; or to put it a little differently, the extent to which such
questions can be answered objectively leaves a great deal in the way of slack.
The merits and import of Quine’s indeterminacy thesis are not my princi-
pal concern, although I will have more to say about Quine at the close of the
paper. I appeal to the thesis initially only to register the intended effect of the
thesis, namely to dislodge from their pedestal various ideas about semantic
determinacy, what Quine dubs ‘the museum myth of meaning.’ Insofar as
any version of such a ‘myth’ had been, prior to Quine’s argument, treated
as doctrine (or at least an intuition underwriting a more refined doctrine),
his pronouncement comes as a shock, as accepting it requires giving up ideas
to which one had previously been committed. The recognition of indetermi-
nacy means making do with less in terms of one’s conception of linguistic
meaning, and along with it various ways of thinking about speech, language
learning, understanding, and even mentality itself.
Indeterminacy in Merleau-Ponty and Wittgenstein  115
Rather than Quine, my interest in this paper lies in exploring various
claims to indeterminacy as they appear in Wittgenstein’s later philosophy
and Merleau-Ponty’s phenomenology. Although there is an element of demy-
thologization attached to these claims, especially in the case of Wittgenstein,
they are at the same time developed and insisted upon in a more positively
charged spirit. Merleau-Ponty says as much at the very outset of Phenom-
enology of Perception, when he declares that ‘we must recognize the inde-
terminate as a positive phenomenon’ (PhP 7).4 Rather than a disappointing
or demythologizing discovery, both Wittgenstein and Merleau-Ponty hope
to foster the recovery of indeterminacy, as something to be acknowledged
and appreciated, rather than something to be settled for in lieu of a prefer-
able, though perhaps unattainable, alternative.5
Both the particular notion of indeterminacy and the more general idea
that the goal of philosophy involves recovery mark an affinity between Witt-
genstein and Merleau-Ponty. Wittgenstein and Merleau-Ponty characterize
their respective philosophical projects as struggling to overcome a sense of
captivity and captivation (by what Wittgenstein calls ‘pictures’ and what
Merleau-Ponty calls ‘prejudices’) and a correlative quest for liberation,
which in each case involves a kind of recollection or return:

What we do is to bring words back from their metaphysical to their

everyday use.
(PI I §116)

[W]e must first awaken that experience of the world of which science is
the second-order expression.
(PhP lxxii)

And what is ‘reawakened’ or recovered is not something heretofore undis-

covered or arcane, but something familiar and proximate. As Wittgenstein

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.) The real foundations of their
inquiry do not strike people at all. Unless that fact has at one time
struck them.—And this means: we fail to be struck by what, once seen,
is most striking and powerful.
(PI I § 129)

Similarly, Merleau-Ponty wants to call attention to something that

serves as ‘an immediate source and as the final authority of our knowledge’
(PhP 24).6 This kind of immediacy—of something ‘always before one’s
eyes’—invites the question of just why such a task should be difficult. Witt-
genstein’s adducing of a kind of ‘paradox of proximity’ is no doubt part of
116  David R. Cerbone
the answer: we are so immersed in our language (our perceptual experience)
that we look right through or past it, but the difficulty is compounded by
the fact that what we look past it to is itself prepared for or projected by
language and experience in a way that conceals or covers over that projec-
tive process. Compare:

A picture held us captive. And we couldn’t get outside it, for it lay in
our language, and language seemed only to repeat it to us inexorably.
(PI I § 115)


Thus is formed ‘objective’ thought (in Kierkegaard’s sense)—the objec-

tive thought of common sense and of science—which in the end makes
us lose contact with the perceptual experience of which it is nevertheless
the result and the natural continuation.
(PhP 74)

Each of these passages registers a loss of ‘contact’ (with language, with per-
ception) as a source of the difficulty of their respective projects, where the
very phenomena to be recovered facilitate that loss in the first place: objec-
tive thought is the ‘natural sequel’ of perceptual experience, just as the cap-
tivating ‘pictures’ that concern Wittgenstein lie in ‘our language.’
What these programmatic remarks suggest is that properly appreciating
the place of indeterminacy in experience and language faces considerable
challenges, since language and experience themselves insinuate ideals of
determinacy and objectivity that need to be unmasked. The very ways in
which we see the world, follow rules, and apply concepts reinforce a sense
of completeness, a sense that ‘all the steps are really already taken’ (PI I
§ 219), and the perceived object is ‘the reason for all the experiences of it
that we have had or that we could have’ (PhP 69).
With these challenges in mind, let us examine how both Wittgenstein
and Merleau-Ponty work to bring the phenomenon of indeterminacy into
view. While not clearly separated in either Wittgenstein or Merleau-Ponty,
I want to explore the recovery of indeterminacy by separating a number of
moments or steps. The most basic are descriptions that serve as demonstra-
tions or recollections of indeterminacy as it is manifest in perceiving, speak-
ing, and understanding. This is what I call simply the fact of indeterminacy.
Once one begins looking for it, indeterminacy pervades our experience and
our use of language. The second, and more difficult, step is recalibrating
the response to this fact, so as not to be disappointed with its pervasive-
ness and so see it as something to be overcome or eliminated (at least in
principle). (It is one thing to be brought back to the ‘rough ground,’ but
quite another to be happy about such a return.) The second step divides
into two further steps or stages: first, a critique of the ideals against which
Indeterminacy in Merleau-Ponty and Wittgenstein  117
the fact of indeterminacy is to be found wanting, and second, reminders
concerning the positive dimensions of indeterminacy, its standing as a genu-
inely ‘positive phenomenon.’ In the final section of the paper, I will return
briefly to Quine’s indeterminacy thesis in order to illustrate just how little
can separate positive and negative receptions of the discovery (or recovery)
of indeterminacy.

2.  The Fact of Indeterminacy

I have already noted the early appearance of the theme of indeterminacy in
Merleau-Ponty, appearing as it does in the opening pages of Phenomenol-
ogy of Perception. This early insistence is due in part to the oppositional
structure of his examination, wherein his own phenomenology of percep-
tion is developed in conjunction with a critique of dominant views that
are in various ways oblivious to just those dimensions of perception (and
later embodiment) his phenomenology emphasizes. Gripped by objective
thought, none of the views Merleau-Ponty opposes can adequately accom-
modate the most basic dimensions of perceptual experience: ‘The perceptual
“something” is always in the middle of some other thing, it always belongs
to a “field” ’ (PhP 4). Such a field is informed by determinate and indeter-
minate dimensions, a determinate figure against a variously indeterminate
background. I look at the stapler on my desk to the right of my computer
screen. The surrounding objects and the region of my study still visible
do not simply disappear, but recede into the background, there to be seen
but only partially and indistinctly manifest while I focus on the stapler: ‘In
this region there is an indeterminate vision, a vision of something or other’
(PhP 6). Should I shift my gaze, one or more of the background objects will
become clearly delineated in my visual field, while the stapler quietly cedes
its foreground position. Avoiding my computer screen with its expanse of a
mostly blank Word document, I can idly explore the surrounding desk and
its denizens, moving effortlessly from one to the next with only slight move-
ments of my eyes and head. While doing so, the foreground-background
relationship is changing throughout, but in entirely unsurprising ways: what
comes into view is already seen as ‘waiting in the wings’ to then take centre
stage as I attend to it more closely.
This kind of push and pull of visual experience—of things emerging and
receding from view—exemplifies what I referred to above as the fact of inde-
terminacy, which any descriptively adequate account of perceptual experi-
ence must acknowledge. The fact of indeterminacy is likewise a pervasive
theme of the Philosophical Investigations. Indeed, I would contend that
the notion of indeterminacy appears in the very first section of the Inves-
tigations, immediately after the quotation from Augustine. At the close of
Wittgenstein’s first vignette involving the purchase of apples from a chart-
wielding shopkeeper, he remarks: ‘It is in this and similar ways that one
operates with words’ (PI § 1). This remark’s ‘and similar ways’ is a gesture
118  David R. Cerbone
toward indeterminacy in that the range of similar ways is in no way fixed or
fixable. This initial gesture is developed in subsequent remarks, especially in
the teens and twenties, for example in Wittgenstein’s likening of language
to a slowly evolving and expanding town in PI § 18 and his responding in
PI § 23 to the question, ‘But how many kinds of sentence are there?’ with,
‘There are countless kinds; countless different kinds of use of all the things
we call “signs”, “words”, “sentences”.’ Wittgenstein continues: ‘And this
diversity is not something fixed, given once for all; but new types of lan-
guage, new language-games, as we may say, come into existence, and others
become obsolete and get forgotten.’
These preliminary remarks, which begin to expose and challenge the
impulses and assumptions underlying various philosophical approaches to
language, sound themes that are taken up in more detail later in the text.
Most notably for our purposes, PI § 65 finds Wittgenstein returning to
the question of what constitutes language—its essence—and again appeal-
ing to an unfixed diversity of types of language: ‘Instead of pointing out
something common to all that we call language, I’m saying that these phe-
nomena have no one thing in common in virtue of which we use the same
word for all—but there are many different kinds of affinity between them.’
Wittgenstein continues by introducing the notion of ‘family resemblance.’
In doing so, the notion of indeterminacy emerges as a central concern:
family resemblance words such as ‘game’ are offered as easy or obvious
examples of linguistic indeterminacy. Lacking anything like ‘necessary and
sufficient conditions,’ the concept game is ‘not closed by a boundary’ (PI I
§68). Fluent speakers of English who by and large agree on what things are
games and what are not may also reasonably disagree about whether some
more peripheral activity is a game without thereby displaying any sort of
linguistic incompetence, while new activities will be called—understood to
be—games that were not, and perhaps could not have been, envisaged and
so anticipated by prior uses of the word: ‘What still counts as a game and
what no longer does? Can you say where the boundaries are? No. You can
draw some, for there aren’t any drawn yet. (But this never bothered you
when you used the word “game”.)’ (PI I § 68). Wittgenstein continues by
distinguishing between the rather hysterical charge that our use of the word
‘game’ is ‘unregulated,’ which suggests a kind of anything-goes arbitrari-
ness to its usage, and the calmer observation that ‘game’ is not ‘everywhere
bounded by rules’. The word ‘game’ has, we might say, a certain degree of
play in it.
That we have never been ‘bothered’ about the absence of boundaries with
respect to words like ‘game’ again exemplifies the fact of indeterminacy:
we carry on in our linguistic practices with words that are not ‘everywhere
bounded by rules,’ and for the most part, we manage to communicate with
one another without a great deal of stress and strain (the significance of our
so managing will be considered more closely later in our discussion). ‘One
can say that the concept of a game is a concept with blurred edges’ (PI I §71),
Indeterminacy in Merleau-Ponty and Wittgenstein  119
and having blurred edges does not thereby render the concept unusable or

Frege compares a concept to a region, and says that a region without

clear boundaries can’t be called a region at all. This presumably means
we can’t do anything with it.—But is it senseless to say ‘Stand roughly
here’? Imagine that I were standing with someone in a city square and
said that. As I say it, I do not bother drawing any boundary, but just
make a pointing gesture—as if I were indicating a particular spot. And
this is just how one might explain what a game is.
(PI I §71)

Although Wittgenstein’s introduction of the notion of family resemblance

and his extended discussion of ‘game’ suggests that he is gesturing toward,
and defending the legitimacy of, a particular kind or range of concepts—
family resemblance concepts, in distinction to some other varieties—how
the discussion develops instead indicates that the notion of not being ‘every-
where bounded by rules’ ramifies through our linguistic practices more gen-
erally. This is indicated by the subsequent discussion of proper names via the
example of ‘Moses,’ where there is no saying just which—or how many—
descriptions typically adduced to identify Moses I am ‘ready to substitute’
when one or another I have offered has been challenged. Instead: ‘I shall
perhaps say: By “Moses” I mean the man who did what the Bible relates of
Moses, or at any rate much of it. But how much? Have I decided how much
must turn out to be false for me to give up my proposition as false?’ Rather
than insisting that the ‘use of the name “Moses” is fixed and determined for
all possible cases,’ Wittgenstein suggests: ‘Isn’t it like this, that I have, so to
speak, a whole series of props in readiness, and am ready to lean on one if
another should be taken from under me, and vice versa?’ (PI I § 79).
The notion of ‘all possible cases’ is further interrogated in the next remark,
wherein Wittgenstein shifts from proper names to our common nouns for
common objects. Here we are asked to consider a curiously disappearing
and reappearing chair:

I say, ‘There is a chair over there’. What if I go to fetch it, and it suddenly
disappears from sight—‘So it wasn’t a chair, but some kind of illusion.’—
But a few seconds later, we see it again and are able to touch it, and so
on.—‘So the chair was there after all, and its disappearance was some
kind of illusion.’—But suppose that after a time it disappears again—or
seems to disappear. What are to say now? Have you rules ready for such
cases—rules saying whether such a thing is still to be called a ‘chair’?
But do we miss them when we use the word ‘chair’? And are we to say
that we do not really attach a meaning to this word, because we are not
equipped with rules for every possible application of it?
(PI I § 80)
120  David R. Cerbone
All three of the questions that close this remark are meant, I take it, to be
answered in the negative. We do not have rules at the ready for settling the
question of what to call the remarkable chair-like entity (or simulacrum)
Wittgenstein here describes, but the word ‘chair’ is not rendered meaning-
less because ‘we are not equipped with rules for every possible application’.
And of course there is nothing special about chairs: outlandish situations
can be adduced where the application of pretty much any common and
familiar word will become hazy, and we will either be at a loss for words or
disagree about what to say.7

3.  The Ineliminability of Indeterminacy

Even if there is something compelling at the descriptive level to both
­Merleau-Ponty’s and Wittgenstein’s appeals to indeterminacy, such that we
are prepared to acknowledge the fact of indeterminacy, someone committed
to the ideals of objective thought or to hew closer to Wittgenstein’s concerns,
a certain picture of logical structure, could nonetheless be unfazed by them.
That is, it remains open to claim that while we find such indeterminacies at
work in everyday experience and language, and while we might grudgingly
admit that we get along all right with such ‘blurry’ concepts and the like,
nonetheless such muddling through falls short of more exacting standards of
determinacy, clarity, and precision.8 In Wittgenstein’s case, even if we allow
that everyday concepts are not utterly meaningless, it may still be contended
that they are imperfect, and so could be replaced by concepts that, variously,
avoided or eliminated miscommunication, better captured the fine structure
of reality, and met these two desiderata by hewing to an ideal logical struc-
ture.9 For Merleau-Ponty, indeterminacy is primarily a perceptual, rather
than linguistic, phenomenon, but he likewise sees his appeal to the fact of
indeterminacy as liable to be met with the charge that this fact only shows
the failings of perception in relation to the project of coming to know an
objective, fully determinate world. As Merleau-Ponty notes:

In the world taken in itself, everything is determinate. There are of

course confused spectacles, such as a landscape in the fog, but even
so, one still admits that no real landscape is in itself confused—it is
only confused for us. Psychologists will contend that the object is never
ambiguous, that it only becomes so through inattention.
(PhP 6–7)

Whether construed linguistically or perceptually, indeterminacy appears to

be a decidedly negative phenomenon, to be overcome or eliminated.10 That
Wittgenstein or Merleau-Ponty is asking us to accept such ­indeterminacies—
to live with our blurry-edged concepts or ambiguous experience—is asking
us to settle for something less than ideal, and that is at best existentially
poignant and at worst intellectually deplorable.
Indeterminacy in Merleau-Ponty and Wittgenstein  121
If we consider further Wittgenstein’s example of the curiously disappear-
ing and reappearing chair, we may begin to discern a deeper point concerning
indeterminacy beyond noting the fact of it in many, if not all, our ordinary
uses of concepts. The first, though perhaps not the best, way to put this is
to suggest that indeterminacy is here to stay, in the sense that once we think
about what it would take to eliminate it, we see that it could not be done.
This becomes evident as we start to reflect on the phrase at the close of the
last sentence of § 80, where Wittgenstein asks after our possession of rules
for ‘every possible application’. Devising a set of rules to cover every pos-
sible application of a concept—past, present, and future—sounds initially
like a Herculean task of the first order. Just how many possible applications
are there and how does one know when ‘all’ of them have indeed been cov-
ered? But it is not just the enormity of the task that is at issue here; it is not
just a matter of being a task that may be daunting, practically unfeasible,
or even impossible. At issue is the clarity of the desideratum: what would it
even mean to have considered ‘every possible application’? Referring back
in § 84 to the § 68 discussion of the idea of the application of a word not
being ‘everywhere bounded by rules’, Wittgenstein goes beyond the initial
descriptively oriented claim to question the contrast that such a description
suggests: what exactly is the alternative to not being everywhere bounded
by rules? Wittgenstein writes:

Speaking of the application of a word, I said that it is not everywhere

bounded by rules. But what does a game look like that is everywhere
bounded by rules? whose rules never let a doubt creep in, but stop up
all the gaps where it might?—Can’t we imagine a rule regulating the
application of a rule; and a doubt which it removes—and so on?
(PI I § 84)

Wittgenstein’s ‘and so on’ here suggests a regress, which in turn suggests that
the problem with a complete list of rules—of complete determinacy—is not
one of impracticality or being more than we need to carry on with things.
A complete list of rules is not an ideal to which we might aspire and perhaps
poignantly fail to obtain; rather, as Wittgenstein suggests in the following
passage from Zettel, the very idea of such an ideal is what deserves scrutiny:

How should we have to imagine a complete list of rules for the employ-
ment of a word?—What do we mean by a complete list of rules for the
employment of a piece in chess? Couldn’t we always construct doubt-
ful cases, in which the normal list of rules does not decide? Think e.g.
of such a question as: how to determine who moved last, if a doubt is
raised about the reliability of the players’ memories?
The regulation of traffic in the streets permits and forbids certain
actions on the part of drivers and pedestrians; but it does not attempt to
guide the totality of their movements by prescription. And it would be
122  David R. Cerbone
senseless to talk of an ‘ideal’ ordering of traffic which should do that;
in the first place we should have no idea what to imagine as this ideal.
If someone wants to make traffic regulations stricter on some point
or other, that does not mean that he wants to approximate to such an
(Z § 440)

Here Wittgenstein challenges the very idea of ‘a complete list of rules,’ not-
ing the possibility of endlessly constructing ‘doubtful cases’ that would
require yet further rules to exclude or decide. Such concocted cases—like the
curious chair of Investigations § 80—are normally not at issue in our traffic
with concepts like chair and neither can be nor need be addressed before-
hand. The possibility of imagining a doubt is not the same thing as being
in doubt, so when someone suggests that ‘secure understanding is possible
only if we first doubt everything that can be doubted, and then remove all
these doubts’, the only reasonable response is the one Wittgenstein offers:
‘The signpost is in order—if, under normal circumstances, it fulfils its pur-
pose’ (PI I § 87).
In the example of traffic regulations in the Zettel passage cited above,
Wittgenstein notes that such regulations do ‘not attempt to guide the totality
of their movements by prescription’. One way of probing the idea of com-
plete determinacy would be to reflect on just such an attempt at guidance,
wherein there is no movement on the part of a person that is not prescribed
or regulated. Lacking in this person would be any kind of movement that
was, variously, erratic, eccentric, spontaneous, improvisational, or superflu-
ous, or at least where there is movement along these lines, it is to be judged
as falling short of what the rules prescribed. Everything about this person’s
actions would be, or aspire to be, determinate because everything about
them would be determined by rules. There are, I think, very particular and
special situations where we can understand striving, at least, for this kind of
total prescriptive governance—meticulously choreographed dances, march-
ing band formations, and other complex, highly stylized physical ­routines—
but two issues immediately arise: first, even restricting our attention to these
special circumstances, it is not clear if we can think of the prescriptions at
work as really guiding the totality of the participants’ bodily movements. If
the dance is successful, then the significant actions of the dancers will hew to
the instructions of the choreographer, but there will be many less obviously
noticeable movements and actions that the instructions do not cover: very
slight adjustments for maintaining balance, for example, or movements of
the eyes to anticipate the desired landing spot for a leap, and so on. While
these various movements may be in the service of the dancer’s efforts to
carry out the choreographer’s plans, it is not clear that there could be a
complete set of instructions that covered every possible movement on the
dancer’s part, as such a set would again have to avert every possible misun-
derstanding that could arise as the dancer tried to enact the dance.
Indeterminacy in Merleau-Ponty and Wittgenstein  123
That there are movements on the part of the participants that cannot be
readily understood as being prescribed brings us to the second issue, which
concerns generalizing beyond these very special circumstances: in non-­
choreographed circumstances (i.e., most of the time), our movements and
actions have considerable leeway or slack. As I sit at my desk, I shift and
fidget, cross and uncross my legs, lean in and back, and so on, none of which
either accords or fails to accord with any kind of rule; as my attention wan-
ders from the paper I’m trying to write, I shift my gaze about the desk in the
manner previously described, and in doing so, I take in visually the variety of
artefacts (the stapler, various old cameras, photographs of my children, my
coffee cup, etc.) that clutter my desk. The myriad slight adjustments I make
are essential to my visual field resolving into foreground and background,
and my field’s so resolving is what gives me a grip on a world of objects at all.
In Husserl (1989), Husserl had emphasized the importance of what he calls
‘functions of spontaneity’, such that any given perception may ‘disperse into a
‘possible’ series of perceptions’, to the constitution of spatiality and of a space
populated by materially real things (Husserl 1989: § 18). To be engaged per-
ceptually and actively with a world of objects is to be confronted with a realm
I can freely explore in an open-ended way. If my movements were severely
limited or rigidly prescribed, my experience and conception of the world
would be correspondingly impoverished to a considerable degree. Building
on Husserl’s insights, Merleau-Ponty contends that ‘freedom makes use of the
gaze and its spontaneous evaluations,’ and that ‘without these spontaneous
evaluations, we would not have a world, that is, a collection of things that
emerges from the formless mass by offering themselves to our body as things
“to be touched”, “to be taken”, or “to be climbed” ’ (PhP 465).
Merleau-Ponty contrasts the notion of a universe, which he characterizes
as a ‘completed and explicit totality,’ with the world—what we have—which
is ‘an open and indefinite multiplicity’ (PhP 73). While we may be able to
conceive in some attenuated way of such a universe, it is not something we
can understand ourselves as inhabiting, as engaging with actively and taking
a view upon from a particular location and perspective:

If the synthesis could be actual, if my experience formed a closed sys-

tem, if the thing and the world could be defined once and for all, if
spatio-temporal horizons could (even ideally) be made explicit and the
world could be conceived from nowhere, then nothing would exist.
I would survey the world from above, and far from all places and times
suddenly becoming real, they would in fact cease to be real because
I would not inhabit any of them and I would be nowhere engaged. If
I am always and everywhere, then I am never and nowhere.
(PhP 347)

What I take Merleau-Ponty to mean here is that the idea of a closed

and ‘defined once and for all’ system cancels the idea of a point of view.
124  David R. Cerbone
As attention to the fact of indeterminacy revealed, a point of view or a
perspective involves both determinate and indeterminate dimensions, a fore-
ground and a background replete with hidden or partially occluded ele-
ments and aspects. Nothing would exist in the sense that nothing would be
there for me, since I could not distinguish in any way between what is here
rather than there, in view rather than somewhere else. Such concepts and
contrasts presuppose my bodily presence in the world, which anchors and
informs my point of view on it. Hovering above the world is not a way of
being in the world.11

4.  The Vitality of Indeterminacy: The ‘Whole Hurly-Burly’

According to Wittgenstein, ‘game’ is a concept where there is no one feature
(or set of features) common to all games such that they are games. Wittgen-
stein further notes that any account I might give of my grasp of the concept
and any attempt I might make to instruct someone else so as to impart that
understanding would rely primarily on describing various exemplary games,
while noting that these and similar things are games (see PI I § 69). More
expansively, Wittgenstein says that explaining to someone what a game is
(or what ‘game’ means) would consist in ‘my describing examples of various
kinds of games, showing how all sorts of other games can be constructed
on the analogy of these, saying that I would hardly call this or that a game,
and so on’ (PI I § 75). Despite the reliance on a small handful of examples
and despite the rather vague gesture toward ‘similar things,’ the remarkable
thing (or at least Wittgenstein wants us to see it as remarkable) is the general
ease with which the concept of game is circulated and imparted. As with
Wittgenstein’s amiable mathematicians at PI I § 240, disputes rarely break
out over whether some activity or other is a game; children learning English
are generally able to cotton on to the concept and use it relatively fluently
quite early on. If we linger on this perfectly ordinary and yet remarkable
fact, our attention may be drawn to the dependency of the ease and success
of maintaining and passing on our life with ‘game’ on a kind of shared ori-
entation or sensibility, what Wittgenstein refers to shortly after appealing to
the largely peaceful scene in mathematics as ‘agreement in judgments’ (PI I
§ 242). Immediately prior to registering this ‘queer’ formulation, he refers
equally to agreement in ‘form of life.’
The forms of agreement Wittgenstein refers to in these passages should
not be understood in terms of pacts or conventions; the agreement upon
which our life with the concept ‘game’ depends is not a matter of various
things we have collectively decided amongst ourselves. Talk of agreement
(the German here is Übereinstimmung, which may also be translated as,
among other things, conformity, harmony, unison, and even synchroniza-
tion) registers the ways in which we are alike in numerous and various ways
and to various degrees: eating, sleeping, walking, talking, laughing, smiling,
crying, gesturing, playing, and so on.12 We are not necessarily like ‘two peas
Indeterminacy in Merleau-Ponty and Wittgenstein  125
in a pod’, but our growing up into these and far more complex activities
generally make for a small and mostly manageable range of variations on
common themes. I like broccoli, you do not; I find these sorts of jokes funny,
you do not, but we still have in common the idea of having a favourite
food or finding something funny to hold us together to a far greater degree
than to a creature for whom having favourites of anything or reacting with
amusement was unfathomable. Similarly, we do not all play or enjoy the
same games, but we can generally recognize what others play or enjoy as
games; they strike us as sufficiently similar, even if we cannot always say
how. Our being so struck bespeaks the kind of common orientation Witt-
genstein has in mind here that runs as deep as our most basic capacities for
pleasure and pain, amusement and sorrow, entertainment and adventure.
Our being able to talk of games with one another and pass on the concept
to new generations enlists these myriad capacities. Were they to go missing,
our life with the concept ‘game’ would either change considerably or fade
away altogether.13
In one of his later manuscripts, Wittgenstein writes:

We judge an action according to its background within human life, and

this background is not monochrome, but we might picture it as a very
complicated filigree pattern, which, to be sure, we can’t copy, but which
we can recognize from the general impression it makes.
(RPP II § 624)

I read the appeal to a background here as along the same lines as his talk of
‘agreement in judgments’ and ‘form of life’ in the Investigations. What Witt-
genstein notes here of an action applies equally to games: we judge some-
thing to be a game, not according to a fixed list of features, but ‘according
to its background within human life’, as something that it makes sense for
someone or some of us to play, where the range of what counts as playing
is equally dependent on this background. Note that Wittgenstein describes
this background in terms of a ‘complicated filigree pattern’ that we can-
not copy, which registers a new level or layer of indeterminacy. To be sure,
the background in question makes a ‘general impression’, but it eludes any
kind of precise demarcation or summary. In the immediately subsequent
remarks in this manuscript, Wittgenstein further elaborates on this kind of

The background is the bustle of life. And our concept points to some-
thing within this bustle.
(RPP II § 625)

The notion of a ‘bustle’ (the German here is Getriebe) is itself a deliber-

ately vague notion (at RPP II § 622 immediately preceding this stretch of
remarks, Wittgenstein discusses the paradigmatic vague14 concept, ‘heap’);
126  David R. Cerbone
there is no saying just how much activity or what variety constitutes a bus-
tle. Instead, it is a matter of the impression it makes. As Wittgenstein imme-
diately notes:

And it is the very concept ‘bustle’ that brings about this indefiniteness.
For a bustle comes about only through constant repetition. And there is
no definite starting point for ‘constant repetition’.
(RPP II § 626)

‘Bustle’ is only one of several terms that Wittgenstein uses in remarks from
this late period in his writings. Immediately after invoking the notion of a
bustle, he adds another:

How could human behavior be described? Surely only by showing the

actions of a variety of humans, as they are all mixed up together. Not
what one man is doing now, but the whole hurly-burly [Gewimmel], is
the background against which we see an action, and it determines our
judgment, our concepts, and our reactions.
(RPP II § 629; Z § 567)

This last remark also appears in Zettel, where it is followed by the following
pair of remarks:

Seeing life as a weave, this pattern (pretence, say) is not always com-
plete and is varied in a multiplicity of ways. But we, in our conceptual
world, keep on seeing the same, recurring with variations. That is how
our concepts take it. For concepts are not for use on a single occasion.
(Z § 568)

And one pattern in the weave is interwoven with many others.

(Z §569)

The idea of a pattern appears in the opening remarks of Part II of the Philo-
sophical Investigations, most notably where Wittgenstein says that ‘ “grief”
describes a pattern which recurs, with different variations, in the tapestry
of life’ (PI II § i). What all of these notions—weave, pattern, bustle, hurly-
burly—appear to be driving at is not just the way particular actions (as
well as uses of a word, applications of a concept, and so on) are embedded
in broader surroundings, but the way those surroundings form a kind of
indefinite backdrop. As indefinite, there is room for variation, multiplic-
ity, leeway for innumerable differences. Although Wittgenstein insists on a
kind of fundamental agreement underwriting our uses of concepts and our
managing to communicate with one another via such concepts, this should
not be understood as any kind of lockstep uniformity; indeed, Wittgenstein
suggests that were there such homogeneity, many of our familiar concepts
would lose their grip.
Indeterminacy in Merleau-Ponty and Wittgenstein  127
This last point deserves further development, as it points in the direc-
tion of the most significant way in which indeterminacy is for Wittgenstein
a ‘positive phenomenon.’ Moreover, attention to his point redoubles the
ideas canvassed in the previous section, namely, that indeterminacy is not
just a descriptive feature of the ways we make sense of the world and one
another but in some way essential to the activity of making sense. In another
later manuscript, Wittgenstein writes: ‘If a pattern of life is the basis for the
use of a word then the word must contain some amount of indefiniteness.
The pattern of life, after all, is not one of exact regularity’ (LW I § 211).
Although much of Wittgenstein’s late discussions involving indeterminacy
are devoted to questions of expression, emotion, and feeling, there would
appear to be a more general moral here, for is there any word that we speak
that does not have as its basis a ‘pattern of life’? The whole motivation for
Wittgenstein’s introduction of the notion of a language-game at the out-
set of the Philosophical Investigations—viz. its third intended sense as ‘the
whole, consisting of language and the activities into which it is woven’ (PI
I §7—my emphasis)—is to suggest a negative answer to this last question.
Later in the same manuscript, Wittgenstein refers back to the opening of the
Investigations and the Builders of §2 in order to emphasize this point: ‘If
the language-game, the activity, for instance, building a house (as in No 2),
fixes the use of a word, then the concept of use is flexible, and varies along
with the concept of activity. But that is in the essence of language’ (LW
I § 340). This is an especially interesting remark when we recall that the
theme of indeterminacy becomes central in the Investigations via Wittgen-
stein’s discussion of family resemblance, which itself is introduced as a way
of refusing to say what is essential to language. Now it would appear that
an essential feature of language is precisely to lack any kind of fixed essence.
Consider a remark from within the stretch recently considered but which
I initially passed over, immediately after adducing the notion of a bustle,
Wittgenstein remarks:

Variability itself is a characteristic of behaviour without which behav-

iour would be to us as something completely different. (The facial fea-
tures characteristic of grief, for instance, are not more meaningful than
their mobility.)
(RPP II §627)

The pairing here of meaning and mobility underscores the significance of

the latter to the former: immobilizing or otherwise limiting the variability of
behaviour and expression robs them of their expressiveness, thereby under-
cutting their being the kinds of meaningful gestures and displays of feeling
that they are. It is hard to know precisely how to secure this point. After
all, counterexamples of various kinds appear to be available: we are able to
‘read’ the expressions of very schematic, simplified figures, in cartoons, for
example; less cartoonishly, people do suffer debilitating conditions, paraly-
sis of the facial muscles, for example, yet we still count them as possessed of
128  David R. Cerbone
feeling. I do not think these count as genuine counterexamples: cartoonish
expressions are themselves a kind of variation, one which our understand-
ing of expressions allows, but their appearing that way in cartoons is part
of what secures their intelligibility. Transferred to real people, such expres-
sions take on a more terrifying aspect. The paralysis case is also disturbing,
which attests to Wittgenstein’s point: a person suffering from such a condi-
tion thereby becomes more enigmatic, but again the leeway in our concepts
allows for continuation of their ‘patterns’ in such cases as well: the muscles
around the mouth may be frozen, say, but the widening of the eyes, the
shake of the head, a clenched fist may be sufficient for that person to express
herself. Wittgenstein notes that ‘given much less variety, a sharply bounded
conceptual structure would have to seem natural,’ but then goes on to ask:
‘But why does it seem so difficult to imagine the simplified case?’ by which
I take him to mean: imagining such cases as the rule, rather than as variously
tragic or cartoonish variations on our own familiarly unbounded structure.
The final paragraph of the same remark offers an attempt:

Is it as if one were trying to imagine a facial expression not susceptible

of gradual and subtle alterations; but which had, say, just five positions;
when it changed it would snap straight from one to another. Would
this fixed smile really be a smile? And why not?—I might not be able
to react as I do to a smile. Maybe it would not make me smile myself.
(RPP II § 614)

A face that snapped from one expression to another, as though with the turn-
ing of a click-wheel, would admit of considerably more determinacy than
any of our own. There would be no question as to what ‘position’ such a
person’s face was presently in—Position 1, say, rather than 3—but that kind
of determinacy would not lend to those expressions anything more by way
of intelligibility. Indeed, quite the opposite: the limited range of expressions
and the rigid transitions among them make it hard to see them as expres-
sions at all. There would instead be something mechanistic about such a
physiognomy, which is part of why Wittgenstein says it would likely ‘not
make me smile myself.’ Such a face would not invite a reciprocal response,
as a smiling or otherwise friendly mien typically does. Wittgenstein contin-
ues by noting that ‘a facial expression that was completely fixed couldn’t be
a friendly one. Variability and irregularity are essential to a friendly expres-
sion. Irregularity is part of its physiognomy’ (RPP II § 615).
Wittgenstein’s stress on the significance of variability and irregularity is
accompanied in the later manuscripts with an approving use of the notion
of the soul, most famously when he says in the philosophy of psychology
fragment: ‘My attitude toward him is an attitude towards a soul,’ which is
quickly qualified with, ‘I am not of the opinion that he has a soul’ (PI II § iv).
I read Wittgenstein’s refusal of the category of opinion here to register
Indeterminacy in Merleau-Ponty and Wittgenstein  129
the depth of the attitude in question: an opinion is something that can be
weighed and appraised, altered or deleted in response to evidence, and so
on, whereas the attitude Wittgenstein reports has a kind of immediacy that
is not responsive to evidence. Indeed, it is only in light of such an attitude
that certain kinds or categories of evidence become available: that someone
displays evidence of melancholy or hurt feelings, shame or regret presup-
poses, rather than supports, the notion of a soul. Hence Wittgenstein’s blunt
rejection of any attempt to reduce the soul to the body: ‘Am I saying some-
thing like, “and the soul itself is merely something about the body”? No.
(I am not that hard up for categories.)’ (RPP II § 690). For Wittgenstein,
the notion of a soul is bound up with a variety of other notions: expression,
thought, and feeling, but also being unpredictable and even unknowable.15
Moreover, Wittgenstein contrasts the notion of a soul—of having a soul—
with that of the mechanical—of being a machine: ‘ “A dog is more like a
human being than a being endowed with a human form, but which behaved
‘mechanically’.” Behaved according to simple rules?’ (RPP II § 623) And
elsewhere he writes:

But the opposite of being full of soul is being mechanical. If you want
to act like a robot—how does your behaviour deviate from our ordi-
nary behaviour? By the fact that our ordinary movements cannot even
approximately be described by means of geometrical concepts.
(RPP I §324)16

Something that is ‘full of soul’ displays a kind of variability, flexibility,

and indeterminacy radically unlike the workings of a machine, whose move-
ments may be more precise but also more rigidly parameterized.17 Else-
where, Wittgenstein explicitly connects this idea of being full of soul with
the notions of unforseeability and indeterminacy: ‘But with a human being,
the assumption is that it is impossible to gain an insight into the mechanism.
Thus indeterminacy is postulated’ (RPP II §666). No such ‘assumption’ is
involved in the idea of a machine or a mechanism: ‘Looking at it one might
think: if I knew what it looked like inside, what was going on right now,
I would know what to expect’ (RPP II §665). While the laws of its opera-
tion may be very complicated, such that predicting its movements might
elude the casual observer, there is certainly no ‘postulated’ indeterminacy
at work. But this lack of indeterminacy is precisely what underwrites the
opposition between being mechanical and being full of soul: mechanized,
rigid motions are marked by an absence of the kind of expressiveness and
meaning conveyed in the fluid and variable actions of living beings. In the
Investigations, Wittgenstein asks, almost despairingly, ‘How could one so
much as get the idea of ascribing a sensation to a thing?’ What dispels the
sense of despair is the image of an organic, animalistic kind of movement,
albeit of a very primitive kind: ‘And now look at a wriggling fly, and at once
130  David R. Cerbone
these difficulties vanish, and pain seems to get a foothold here, where before
everything was, so to speak, too smooth for it’ (PI I §284). Wriggling is ‘full
of soul’ in the sense that it expresses and conveys the organism’s suffering,
its plight as something trapped, perhaps even doomed. And wriggling is
not a sharply bounded kind of movement: if the fly only flipped one way
and another in a perfectly rigid way, without any variation, its movements
would likely remain ‘too smooth’ for the ascription of any kind of pain.18
Wittgenstein’s wriggling fly finds a counterpart in Merleau-Ponty’s men-
tion of a drowning one, purportedly contemplated by Spinoza:

Spinoza would not have spent so much time considering a drowning fly
if this behavior had not offered to the eye something other than a frag-
ment of extension; the theory of animal machines is a ‘resistance’ to the
phenomenon of behavior.
(SB 127)

Merleau-Ponty’s invocation of the notion of resistance in its psychoana-

lytic sense suggests that he regards considering the body—even the body
of an animal—in strictly mechanical terms as indicative of a latent hostility
toward the unruly but expressive antics of the living organism. Approached
without resistance, such antics bespeak not the workings of a machine but
of a soul. As with Wittgenstein, for Merleau-Ponty the expressiveness of the
face registers this idea most emphatically:

A face is a center of human expression, the transparent envelope of the

attitudes and desires of others, the place of manifestation, the barely
material support for a multitude of intentions. This is why it seems
impossible for us to treat a face or a body, even a dead body, like a thing.
They are sacred entities, not the ‘givens of sight.’
(SB 167)

Late in Part One of Phenomenology of Perception, Merleau-Ponty

describes ‘man’ as having ‘a genius for ambiguity’ (PhP 195). The ‘genius’ at
issue is a kind of resistance to the possibility of sharply bounded ­categories—
most notably, the biological versus cultural, but also the physiological and
the mental—such that everything we do and say can be traced to both
kinds of categories but without any neat bifurcation of different aspects of
what we say and do into one category or the other. There is instead a kind
of ‘escape’ across the boundaries. Merleau-Ponty goes so far as to say that
this ‘genius for ambiguity might serve to define man.’ His appeal to ambi-
guity here, to the way human activity cannot be neatly decomposed into
a number of mutually exclusive categories, recalls Wittgenstein’s appeal to
the ‘hurly-burly’ that forms the backdrop of human life, a ‘filigree pattern’
that makes an impression but that we cannot copy. ‘Irregularity is part of
the physiognomy’ of human life, and both Merleau-Ponty and Wittgenstein
Indeterminacy in Merleau-Ponty and Wittgenstein  131
emphasize that this kind of indeterminacy is not a defect in our knowledge
or understanding of ourselves. We cannot copy the pattern because the pat-
tern is not something settled or fixed, but shifting and evolving over time.
In the chapter preceding the one where he appeals to our ‘genius for
ambiguity,’ at the culmination of his discussion of sexuality, Merleau-Ponty
declares that in human existence:

There is a principle of indetermination, and this indetermination does

not merely exist for us, it does not result from some imperfection in
our knowledge, and we must not hold that a God might sound out our
hearts and minds and determine what comes from nature and what
comes from freedom.
(PhP 172–3)

He goes on to say that:

Existence is indeterminate in itself because of its fundamental structure:

insofar as existence is the very operation by which something that had
no sense takes on sense, by which something that only had a sexual
sense adopts a more general signification, by which chance is trans-
formed into reason, or in other words insofar as existence is the taking
up of a de facto situation.
(PhP 173)

The ‘meaningless takes on meaning’ in virtue of this ‘principle of inde-

terminacy.’ The indeterminate, the unbounded: such notions are not for
Merleau-Ponty and Wittgenstein obstacles or impediments to meaning but
constitutive of it. For Merleau-Ponty, as for Wittgenstein, ‘ambiguity is not
an imperfection of consciousness or existence, it is their very definition’
(PhP 347). If this is correct, then any attempt to overcome or eliminate such
ambiguity and irregularity is bound eventually to come to grief. But this does
not render such attempts impossible. Merleau-Ponty and Wittgenstein both
see their thinking as up against dominant views wherein such indeterminacy
both can and should be eliminated. Nor does it mean that every attempt to
eliminate any bit of ambiguity is bound to have untoward effects. We ‘can
draw a boundary—for a special purpose,’ and it is an open question whether
doing so counts as an improvement. Elsewhere Wittgenstein considers the
possibility of sharpening up our language-games, replacing them with ones
that are more exact. Not surprisingly, his response toward the consequences
of such an endeavor itself registers an attitude of indeterminacy:

And now the question remains whether we would give up our language-
game which rests on ‘imponderable evidence’ and frequently leads to
uncertainty, if it were possible to exchange it for a more exact one
which by and large would have similar consequences. For instance, we
132  David R. Cerbone
could work with a mechanical ‘lie detector’ and redefine a lie as that
which causes a deflection on the lie detector.
So the question is: Would we change our way of living if this or that
were provided for us?—And how could I answer that?
(LW II 95)

5.  Quinean Coda: Intentionality and the ‘Dramatic Idiom’

Toward the end of Chapter 2 of Word and Object (Quine 1960), Quine
conjectures in a footnote that ‘perhaps the doctrine of the indeterminacy
of translation will have little air of paradox for readers familiar with Witt-
genstein’s remarks on meaning’ (Quine 1960: 77). The question of the
relation between Quine and Wittgenstein is a difficult one. Burton Dreben
has referred to them as the ‘odd couple’ (Quine and Merleau-Ponty would
no doubt be an odder pairing still), while Peter Hacker has invoked a
notion of ‘proximity at great distance,’ which bespeaks a lingering ‘air
of paradox’ despite Quine’s suggestion to the contrary.19 Without trying
to settle entirely the issue of their interrelation, we might say something
about this. There would be little in the way of paradox or surprise for
either Wittgenstein or Merleau-Ponty if confronted with Quine’s indeter-
minacy thesis, at least in this one respect: neither of them would balk
especially at the contention that rival manuals of translations ‘are com-
patible with all the same distributions of states and relations over elemen-
tary particles’ (Quine 1981: 23). That questions of meaning would not be
settled by determinations regarding the particles revealed and explored
by physics is hardly to be wondered at, since that is entirely the wrong
place to look, and so Merleau-Ponty for one puts little stock in an attempt
to ‘construct the picture of this world, life, perception, or mind’ starting
from the ‘physicist’s atoms’ (PhP 24). Any such reconstructive projects
are doomed to failure, as the phenomenon of meaning or significance has
already been lost from view.
Whence then the distance despite the great proximity? The difference lies
in the lessons to be gleaned from the failure of the notion of meaning to
be amenable to our more scientific aspirations. For Quine, the difference
between what is recognized from the standpoint of ‘the strictest scientific
spirit’ and what is allowed for when using the intentional idiom is the dif-
ference ‘between literal theory and dramatic portrayal’ (Quine 1960: 219).
In so relegating the intentional idiom, Quine thereby registers a kind of
agreement with Brentano concerning the relation of the intentional to the
non-intentional: just as the dramatic cannot be rendered without loss in lit-
eral terms, so too the intentional cannot be reduced to the non-intentional.
Again, this is in keeping with Quine’s central indeterminacy thesis: since
conflicting manuals of translation are ‘physically equivalent’ (Quine 1981:
23), there is no hope of reducing them to, or identifying them in terms of,
Indeterminacy in Merleau-Ponty and Wittgenstein  133
properly scientific (i.e., physical) theories. As Quine notes, ‘Brentano’s thesis
of the irreducibility of intentional idioms is of a piece with the thesis of the
indeterminacy of translation’ (Quine 1960: 221). Quine’s agreement with
Brentano only goes so far, however, since Quine draws an altogether differ-
ent conclusion from the basic point about irreducibility:

One may accept the Brentano thesis either as showing the indispensabil-
ity of intentional idioms and the importance of an autonomous science
of intention, or as showing the baselessness of intentional idioms and
the emptiness of a science of intention. My attitude, unlike Brentano’s,
is the second.
(Quine 1960: 221)

Quine means baselessness quite literally here, since there is nothing in phys-
ics to settle the rivalry between competing manuals of translation: such
rivalries outstrip, and so cannot be based upon, matters of physical fact.
However, this passage about the lessons of Brentano’s thesis contains
something of a fudge on Quine’s part, since each of the alternatives con-
sists of a conjunction of two claims and Quine really only rejects one of
the two conjuncts he attributes to Brentano. While Quine certainly does
not share Brentano’s opinion on ‘the importance of an autonomous sci-
ence of intention,’ what is less clear is whether he regards intentional idi-
oms to be dispensable. That is, while he clearly sees no hope of bringing
the intentional into line with the strict standards of scientific theorizing, at
the same time, he concedes the importance of such idioms, and in various
contexts. While Quine maintains that ‘intentional idioms’ have no role
to play when ‘limning the true and ultimate structure of reality’ and that
‘all traits of reality worthy of the name can be set down’ in the austere
idiom of the natural sciences, he nonetheless does not ‘forswear daily use
of intentional idioms, or maintain that they are practically dispensable’
(Quine 1960: 221).20
Quine’s perhaps grudging acknowledgement of the importance of the
‘intentional idioms’ renders questionable his condemnation of ‘a science of
intention’ as empty. While perhaps such an enterprise could not and should
not aspire to the austerity of the natural sciences, something so fundamen-
tal and indispensable to our lives merits investigation. Ought we not to be
interested in something more than what is available from the perspective of
the natural sciences, if, as Quine himself acknowledges, they ‘conceal from
us the “cultural world” or the “human world” in which almost our entire
life happens’ (PhP 25)? Although Quine emphatically rejects the notion of
an autonomous ‘science of the mental,’ the fact that the ‘austere idiom’ of
the natural sciences is insufficient to capture life’s ‘drama’—what Merleau-
Ponty calls ‘the homeland of our thoughts’ (PhP 26)—speaks compellingly
in favour of just such an enterprise.
134  David R. Cerbone
1 An earlier version of this paper was presented at the annual meeting of the
American Society for Existential Phenomenology in Berkeley, California, in
March 2013. I would like to thank the participants in that conference—­especially
Taylor Carman, Hans Sluga, and Mark Wrathall—for their comments and criti-
cisms. I would also like to thank Randall Havas, Edward Minar, and especially
Komarine Romdenh-Romluc for their comments and discussion of earlier drafts
of this paper.
2 I should note at the outset that throughout the discussion various senses of
indeterminacy will be in play. For Quine, indeterminacy means the absence of
any fact of the matter with respect to a rivalry between two (or more) alterna-
tives. I think something like this notion of indeterminacy carries over to Witt-
genstein and Merleau-Ponty, but often the term suggests something more along
the lines of being indefinite, vague, unbounded, open-ended, or unresolvable.
I will not endeavor in this discussion to render the notion of indeterminacy more
3 This is not true of philosophy alone. For many at the time—most notably
­Einstein—the postulation of quantum indeterminacies via Bohr’s Copenhagen
interpretation of quantum mechanics was hardly a cause for celebration.
4 In thinking about the place of indeterminacy in Merleau-Ponty, I have profited
greatly from Kelly (2005), as well as Siewert (2005). Both Kelly and Siewert
offer careful analyses and interpretations of the meaning and significance of
indeterminacy in perceptual experience. As Siewert notes, his discussion avoids
ascending prematurely to the ‘grander themes’ of Merleau-Ponty’s existential
philosophy, favoring instead the ‘humble sensory case’ that must first be properly
understood. At the risk of undue haste, I do try below to consider some of those
broader themes. I have also profited from Sapontzis (1978).
5 The theme of indeterminacy in Wittgenstein’s philosophy has not, to my mind,
been sufficiently explored. The most thorough treatment thus far is Meyer (1998).
I have profited greatly from Meyer’s discussion, although I don’t know if I agree
with him on all points of interpretation. Meyer conceives of indeterminacy in
terms of judgments and their determination (or not) by rules. Indeterminate
judgments form a special class or kind, in contrast to garden variety determi-
nate judgments. While this may well be one kind of indeterminacy at issue in
Wittgenstein, I do not think it is the only one. As I try to show below, there is
a kind of indeterminacy in language (or concepts) that does not form a special
sub-category. See also Hertzberg (1983) and ter Hark (2004), which rightly notes
how the notion of indeterminacy gains prominence in Wittgenstein’s last writ-
ings on the philosophy of psychology. Recent work by Theodore Schatzki has
also foregrounded the notion of indeterminacy in Wittgenstein’s later philosophy
and in doing so, emphasizes the affinities between Wittgenstein and another cen-
tral figure in 20th-century phenomenology, Martin Heidegger. See, for example,
Schatzki (2013). There is, finally, one precedent for bringing Merleau-Ponty and
Wittgenstein together on this topic: Marsh (1975). Although I agree with Marsh
in terms of the general congruence of Merleau-Ponty and Wittgenstein on this
score, his discussion does not, in my view, properly emphasize the positive dimen-
sions of indeterminacy in both cases: its ‘triumph’ is mainly one of debunking
traditional Cartesian pictures of mind and world. Indeed, Marsh reads Witt-
genstein as primarily, and even only, a negative philosopher concerned only to
establish a thoroughgoing anti-essentialism, in contrast to a kind of enlightened
essentialism in Merleau-Ponty. My discussion should make clear my disagree-
ment with Marsh on this front.
6 Merleau-Ponty is here referring to experience. In the older Smith translation,
experience is described as ‘what stares us in the face’.
Indeterminacy in Merleau-Ponty and Wittgenstein  135
7 The last time I taught the Investigations, the class was pretty evenly divided
over whether what Wittgenstein describes in PI I § 80 ought to be called a chair,
which seemed to demonstrate his point, especially as there was little in the way
of resources to settle such a dispute.
8 For a spirited defense of Frege’s demand for sharp concepts, see Weiner (1997).
See also Diamond (1995).
9 There are contexts where claims of this sort are perfectly in order, where greater
precision in categorization or discrimination is necessary for various purposes.
Wittgenstein addresses this point at PI I § 88 of the Investigations, noting that
this kind of goal-dependent notion of determinacy or exactitude does not license
its generalization beyond any and all particular contexts: ‘No single ideal of
exactness has been envisaged; we do not know what we are to make of this
10 There is, I believe, a parallel between this kind of response to the fact of indeter-
minacy and the perceived inadequacy of ‘pragmatic’ responses to skepticism to
the effect that our everyday epistemic standards are good enough for the kinds of
knowledge claims we make and exchange outside of any specialized, distinctively
philosophical context. That both Merleau-Ponty and Wittgenstein address the
canvassed response to indeterminacy is suggestive for how they might address
both skepticism and the kind of impatience that often greets pragmatic responses
to it. I cannot, however, follow up these parallels here.
11 Even the notion of hovering suggests some particular location, and so would still
involve the notion of perspective, along with various forms of indeterminacy and
12 See PI I § 25.
13 I am aware in my putting things this way of my tremendous debt to the work
of Stanley Cavell. I should note especially here Cavell (1976) and its appeal to
the ‘whirl of organism’ (formulated without, to my knowledge, acquaintance
with the late manuscripts on the philosophy of psychology that I discuss below)
within which our concepts have their life. I am particularly indebted as well
to Cavell’s remarks on learning to speak as requiring the projection of words
into new contexts as involving what Cavell refers to as an ‘inner constancy’ and
‘outer variance.’ See Cavell (1982), Part Two, Chapter VII: ‘Excursus on Witt-
genstein’s Vision of Language.’
14 Elsewhere, Wittgenstein writes: ‘The greatest difficulty in these investigations is
to find a way of representing vagueness’ (LW I § 347).
15 ‘I think unforeseeability must be an essential property of the mental. Just like the
endless multiplicity of expression’ (LW II 65).
16 See also LW II 66: ‘One could also put it this way: How would a human body
have to act so that one would not be inclined to speak of inner and outer human
states? Again and again, I think: “like a machine” ’.
17 This paragraph is taken almost verbatim from Cerbone (2013). In that discus-
sion, I am considering Wittgenstein’s (and Heidegger’s) relation to modernist
architecture, more specifically, to Le Corbusier’s reconception of the house as ‘a
machine for living in.’
18 Again using the notion of a pattern and the imagery of weaving, Wittgenstein
in one later manuscript writes: ‘ “Grief” describes a pattern which recurs in the
weave of our life. Now a process is also part of this pattern. If a man’s bodily
expressions of sorrow and joy alternated, say with the ticking of a metronome,
then this would not result in the pattern of sorrow or joy. (This does not mean
that joy or grief are kinds of behaviour.)’ (LW I § 406).
19 See Dreben (1996) and Hacker (1996).
20 I discuss these tensions in Quine’s attitude towards the ‘intentional idiom’ at
greater length in Cerbone (2012), parts of which I have made use of in this
136  David R. Cerbone
Cavell, S. 1976. The availability of Wittgenstein’s later philosophy. In his Must We
Mean What We Say? Cambridge: Cambridge University Press, pp. 44–72.
Cavell, S. 1982. The Claim of Reason: Wittgenstein, Skepticism, Morality, and Trag-
edy. Oxford: Oxford University Press.
Cerbone, D. 2012. Lost belongings: Heidegger, naturalism, and natural science. In
T. Glazebrook (ed.), Heidegger on Science. Albany: SUNY Press, pp. 131–153.
Cerbone, D. 2013. Dwelling on rough ground: Heidegger, Wittgenstein, architec-
ture. In D. Egan, S. Reynolds, and A. Wendland (eds.), Wittgenstein and Hei-
degger. London: Routledge, pp. 245–260.
Diamond, C. 1995. Frege against fuzz. In her The Realistic Spirit: Wittgenstein, Phi-
losophy, and the Mind. Cambridge, MA: MIT Press, pp. 145–178.
Dreben, B. 1996. Quine and Wittgenstein: The odd couple. In R. Arrington and H-J.
Glock (eds.), Wittgenstein and Quine. London: Routledge, pp. 39–61.
Hacker, P.M.S. 1996. Wittgenstein and Quine: Proximity at great distance. In
R. ­Arrington and H-J. Glock (eds.), Wittgenstein and Quine. London: Routledge,
pp. 1–38.
Hertzberg, L. 1983. The Indeterminacy of the mental. Proceedings of the Aristote-
lian Society, Supplementary Volumes 57: 91–109.
Husserl, E. 1989. Ideas Pertaining to a Pure Phenomenology and to Phenomeno-
logical Philosophy: Second Book. Translated by R. Rojcewicz and A. Schuwer.
Dordrecht: Kluwer Academic Publishers.
Kelly, S. D. 2005. Seeing things in Merleau-Ponty. In T. Carman and M. Hansen
(eds.), The Cambridge Companion to Merleau-Ponty. Cambridge: Cambridge
University Press, pp. 74–110.
Marsh, J. L. 1975. The triumph of ambiguity: Merleau-Ponty and Wittgenstein.
Philosophy Today 19 (3): 243–255.
Meyer, T. A. 1998. Rule: A Study in Indeterminacy. Doctoral dissertation, Depart-
ment of Philosophy, University of Pennsylvania.
Quine, W.V.O. 1960. Word and Object. Cambridge, MA: MIT Press.
Quine, W.V.O. 1981. Things and their place in theories. In his Theories and Things.
Cambridge, MA: Harvard University Press, pp. 1–23.
Sapontzis, S. F. 1978. A note on Merleau-Ponty’s ‘ambiguity’. Philosophy and Phe-
nomenological Research 38 (4): 538–543.
Schatzki, T. R. 2013. Human activity as an indeterminate social event. In D. Egan,
S. Reynolds, and A. Wendland (eds.), Wittgenstein and Heidegger. London:
Routledge, pp. 179–194.
Siewert, C. 2005. Attention and sensorimotor intentionality. In D. Smith and
A. Thomasson (eds.), Phenomenology and the Philosophy of Mind. Oxford:
­Oxford University Press, pp. 270–294.
ter Hark, M. 2004. ‘Patterns of life’: a third Wittgensteinian concept. In D. Moyal-
Sharrock (ed.), The Third Wittgenstein: The Post-Investigations Works. Hamp-
shire: Ashgate, pp. 125–144.
Weiner, J. 1997. Has Frege a philosophy of language? In W. W. Tait (ed.), Early Ana-
lytic Philosophy: Frege, Russell Wittgenstein. Chicago: Open Court, pp. 249–273.
8 Wittgenstein and Merleau-Ponty
on Knowledge and Certainty
Thomas Baldwin

Anyone familiar with Wittgenstein’s notes On Certainty and Merleau-­

Ponty’s Phenomenology of Perception will have noticed that there are
similarities between Wittgenstein’s account of the certainty of our common-
sense picture of the world (OC §93) and Merleau-Ponty’s description of the
‘faith’ that ‘places us in the world prior to every science’ (PhP 359). My aim
in this paper is to explore these similarities and also some of the differences
between their positions. I start by identifying a few points from Wittgen-
stein’s position, since his notes On Certainty are readily accessible, and then
use these points as the basis for an initial comparison with Merleau-Ponty’s
position which has to be brought together from passages spread throughout
the five hundred pages of Phenomenology of Perception.

1.  Introduction—Some Quick Comparisons

Wittgenstein begins his notes On Certainty with his critical response to
what he takes to have been G. E. Moore’s ‘dogmatic’ response to skepti-
cism: ‘from his utterance “I know . . .” it does not follow that he does know
it.1 That he does know takes some shewing’ (OC §§13–4). This response
is, however, combined with a positive appropriation of the truisms which
Moore set out in his paper ‘A Defence of Common Sense’ (Moore 1993:
107–9) as ‘hinge’ propositions which provide a fixed, secure, picture of the
world for the networks of ordinary beliefs which pivot around them (OC
§341). As such, these hinge propositions are certainties which are ‘exempt
from doubt’ (OC §341). To take his position further Wittgenstein intro-
duces a categorial distinction between knowledge and certainty (OC §308)
which leads him into an inquiry into the significance of claims about knowl-
edge. The main point of these claims, he suggests, is to provide assessments
of the reliability of potential informants, so that to claim to know some-
thing ‘might be to indicate where I can be relied upon’ (OC §575). In one
of his final notes (OC §622), Wittgenstein then remarks that this way of
thinking about knowledge enables one to understand how one might accept
Moore’s characteristic claims to knowledge, for example that circumstances
are such that he knows ‘this is a hand’, without taking it that Moore’s claim
138  Thomas Baldwin
is ‘philosophically astonishing’, that is to say, that it provides by itself a
decisive counterexample to skeptical objections to the possibility of such
knowledge. For accepting Moore’s claim to knowledge is not the same as
explaining why Moore can be relied upon in this instance. Wittgenstein
holds that such an explanation can be provided once one recognises that
the possibility of mistakes and reasonable doubts arises only where certain
conditions are satisfied and that under normal circumstances these condi-
tions do not apply to Moore’s claim. So Moore can be relied upon; he does
know that this is a hand. But showing this was not just a matter of taking
his word for it.
I say more below about Wittgenstein’s On Certainty and the positions
advanced there. But it is now time to turn to Merleau-Ponty’s Phenom-
enology of Perception. Merleau-Ponty remarks that ‘In a phenomenological
conception, this dogmatism and this skepticism are simultaneously over-
come’ (PhP 418), and a position of this kind, lying between dogmatism and
skepticism, is on the face of it similar to that suggested by Wittgenstein. The
‘dogmatism’ Merleau-Ponty has in mind here is a view which he attributes
to Spinoza, that under ordinary circumstances the world is just ‘evident’ to
us in a way which makes doubt or error unreasonable, which is similar to
the position attributed by Wittgenstein to Moore. Merleau-Ponty’s ‘skeptic’
by contrast is someone who regards what is evident as merely evident to
us, and thus as an appearance which cannot be assumed to be a reliable
guide to what is really the case. Merleau-Ponty’s response is that the skeptic
here betrays a misunderstanding of experience and reality as terms which
are in principle absolutely independent of each other, instead of recognising
that they are essentially related, so that, despite the fallibility of experience
which the dogmatist overlooks, it is appropriate ‘to define being as what
appears to us’ (PhP 419). This comment appears to have an idealist implica-
tion which would be alien to Wittgenstein, who treats idealism as a form
of skepticism (OC §37), and I shall return in the final part of this paper to
the issue this raises. In the present context, however, a different aspect of
Merleau-Ponty’s response to skepticism merits attention, namely his insis-
tence that doubt makes sense only where it draws on beliefs that themselves
commit the doubter to the world—‘neither error nor doubt ever cut us off
from truth, because they are surrounded by an horizon of the world’ (PhP
419). So, like Wittgenstein, Merleau-Ponty argues that the position of the
radical skeptic is undermined by the fact that the possibility of doubt rests
on an antecedent commitment to the existence of the world.
For Wittgenstein these commitments are manifested in the role and status
of Moore’s truisms as hinge propositions which are exempt from doubt.
In his Phenomenology of Perception Merleau-Ponty begins by affirming in
his ‘Preface’ that ‘the certainties of common sense’ are ‘the constant theme
of philosophy’ (PhP lxxvii). This sounds Moorean, but Merleau-Ponty
qualifies his attachment to these certainties by saying that phenomenol-
ogy requires one to suspend a commitment to them precisely in order to
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  139
reawaken an understanding of the ways in which as ‘presuppositions of
every thought’ they are liable to pass unnoticed (PhP lxxvii). In a rather
similar way Wittgenstein remarks in his Philosophical Investigations that
one of his aims is to bring to our attention ‘[t]he aspects of things that are
most important for us’ which ‘are hidden because of their simplicity and
familiarity’ (PI I §129); and he then adds the parenthetical comment ‘(One
is unable to notice something—because it is always before one’s eyes)’ which
might suggest that some way of distancing oneself from these aspects, and
thus a kind of ‘epoché’, will then be needed to bring them to one’s attention.
But Wittgenstein does not in fact recommend anything that resembles the
phenomenological epoché; instead, in the next paragraph, he remarks that
the ‘clear and simple language-games’ introduced at the beginning of his
Philosophical Investigations are intended as ‘objects of comparison which
are meant to throw light on the facts of our language by way not only of
similarities, but also of dissimilarities’ (PI I §130). So, and this is, I think, a
significant ‘dissimilarity’ in philosophical method, Wittgenstein takes it that
reflection on ‘clear and simple language-games’ should enable us to notice
the aspects of things that are normally hidden because they are too famil-
iar, whereas for Merleau-Ponty the phenomenological epoché is intended to
achieve much the same result by a kind of detached reflection which ‘steps
back in order to see transcendences spring forth’ as it ‘loosens the inten-
tional threads that connect us to the world in order to make them appear’
(PhP lxxvii).2
Returning now to the issue of certainty, the claim that is characteristic
of Merleau-Ponty’s position is that ‘There is an absolute certainty of the
world in general, but not of any particular being’ (PhP 311). He does not,
however, make it clear how this absolute certainty of the world in general
sits alongside the certainties of common sense. One might just take it that,
for Merleau-Ponty, these latter certainties concern ‘particular beings’ and
thus lack the ‘absolute’ certainty of the world itself. But it is also possible
to propose a Moorean interpretation of them in terms of something like
the list of ‘truisms’ which Moore took to comprise the ‘common sense
view of the world’ which he claimed to ‘know for certain’ (Moore 1993:
118). For Moore, these truisms characterise fundamental aspects of the
world, such as space, time (Moore’s list here includes Wittgenstein’s favou-
rite example: ‘the earth had existed for many years before my body was
born’—Moore (1993: 107); cf. (OC §84) ), substance (‘things’), other peo-
ple, and so on, and these are all surely to be included in the conception of
the ‘world in general’ of which Merleau-Ponty declares us to be absolutely
certain. So it seems reasonable to take it that Merleau-Ponty’s ‘absolute
certainty’ includes ‘the certainties of common sense’, understood along the
lines of Moore’s list of truisms; indeed Moore’s list begins with a proposi-
tion which is especially appropriate in the context of a comparison with
Merleau-Ponty: ‘There exists at present a living human body, which is my
body’ (Moore 1993: 107).
140  Thomas Baldwin
To complete this first, brief comparison between the positions of Wittgen-
stein and Merleau-Ponty, I turn to the question of the relationship between
certainty and knowledge. As I have indicated, Wittgenstein places them in
different categories: where certainty applies to the propositions which pro-
vide the frame of our picture of the world, attributions of knowledge arise
within the context of discussions about who can be relied upon as a source
of information. Merleau-Ponty also distinguishes certainty from knowledge,
but in rather different ways. He takes it that the certainty of the world pro-
vides the basis for ‘thetic’ knowledge about it: ‘the certainty of the thing and
of the world precedes the thetic knowledge of their properties’ (PhP 402)—
where ‘thetic knowledge’ is explicit knowledge that is typically expressed in
language. But he also writes of a ‘primordial knowledge of all things’ that is
awakened by ‘the actual contact with the thing’ (PhP 388), that is, the kind
of perception which is ‘originary knowledge’ (PhP 45) in which ‘we merge
with this body which knows more than we do about the world’ (PhP 248).
This kind of knowledge cannot be the thetic knowledge of the properties of
things that is preceded by the certainty of the world; instead it is an implicit
‘non-thetic’ knowledge of the world provided by perception and implicit in
that certainty of the world. So Merleau-Ponty’s account of the relationship
between certainty and knowledge is mediated by his distinction between
these two kinds of knowledge, between, for example, our ‘objective and
detached knowledge of the body’ and ‘this other knowledge that we have of
it because it is always with us and because we are bodies’ (PhP 213). There
is more to be said about this distinction, and I will come back to it. But on
the face of it, this distinction between two kinds of knowledge is rather dif-
ferent from Wittgenstein’s position, so that the agreement between them on
the existence of a distinction between certainty and knowledge is combined
with disagreement concerning its basis.
In order to proceed further, I need to lay out some more features of the
positions advanced by Wittgenstein and Merleau-Ponty. As before, I start off
from Wittgenstein since his discussion is the more approachable although,
of course, it also gives rise to difficult questions.

2.  Wittgenstein and On Certainty

The notes published as On Certainty are among Wittgenstein’s last writ-
ings. The first part (OC §§1–65) was written on twenty loose sheets of
paper which Wittgenstein left in his room in Elizabeth Anscombe’s house
in Oxford where he lived from April 1950 until February 1951. She reports
(OC ‘Preface’, vi) that she formed the impression that they were written in
Vienna in early 1950, shortly before he came to stay with her. The next part
(OC §§66–299) comes from notebooks which Wittgenstein wrote while
staying with her.3 In March 1951 Wittgenstein moved back to Cambridge,
and while living there he wrote the final part (OC §§300–676) in more note-
books; the last entry is dated 27 April 1951, just two days before his death.
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  141
When reading these notes, therefore, one must realise that they are quite
unlike the text of part I of Philosophical Investigations (the text of part II
is a different matter) which comprises a series of paragraphs carefully
selected from longer manuscripts and meticulously organised for publica-
tion by Wittgenstein himself. In the case of On Certainty, the text comprises
unrevised notes which were not prepared for publication by Wittgenstein; as
published they are ordered only by the date of composition and have been
numbered by the editors (G.E.M. Anscombe and G. H. von Wright), not by
Wittgenstein himself. The editors do, however, add that in his notebooks
Wittgenstein himself had marked off these notes as a topic separate from his
notes on the other topics discussed there (colour, ‘the inner’, and culture). So
they felt themselves to be justified in publishing these notes by themselves as
a single, sustained treatment of one topic, to which they gave the title Über
Gewissheit (On Certainty). Concerning this choice of title, it is worth add-
ing that alongside the familiar German words for ‘certain’ and ‘certainty’,
‘gewiss’ and ‘Gewissheit’, Wittgenstein often uses the words ‘sicher’ and
‘Sicherheit’, which in many contexts one would translate as ‘secure’ and
‘security’ but which here have an epistemic application so that the transla-
tions ‘sure’ and ‘being sure’ are generally more appropriate. However, the
difference between these idioms is not indicated in the English translation in
which the single pair of English words, ‘certain’ and ‘certainty’, is employed
to translate both pairs of German terms, despite the fact that for Wittgen-
stein these idioms do not appear to be interchangeable.
The immediate stimulus for Wittgenstein’s reflections on this topic seems
to have been discussions with his friend Norman Malcolm. In the sum-
mer of 1949, before his return to Vienna, Wittgenstein spent three months
in the United States with Malcolm, and Malcolm reports (Malcolm 1977)
that during this visit they discussed his recently published paper ‘Defending
Common Sense’ (Malcolm 1949) in which he criticised Moore’s defence of
common sense. Malcolm does not suggest that Wittgenstein wrote anything
on this topic while he was in the United States; but it seems likely that he
had been sufficiently stimulated by their discussions to write up his thoughts
on the topic while he was in Vienna. So although there are no references to
Malcolm’s paper in Wittgenstein’s notes, the notes start with his own criti-
cal response to Moore. The very first note (OC §1) begins with a reference
to Moore’s ‘Proof of an External World’ (Moore 1993: 165–6): ‘If you do
know that here is one hand, we’ll grant you all the rest. . . .’, and shortly
afterwards (OC §6) there is a reference to ‘A Defence of Common Sense’
(Moore 1993: 106 ff.) when he writes: ‘Now, can one enumerate what one
knows (like Moore)? Straight off like that, I believe not. . . .’ But it should
also be acknowledged that some of the themes of On Certainty are antici-
pated in notes which Wittgenstein wrote in 1937 under the title ‘Cause and
Effect: Intuitive Awareness’ (PO 370–405). The title of these notes alludes
to the fact that Wittgenstein begins by commenting on a paper by Russell
in which Russell discusses knowledge of causal facts (Russell 1935); but
142  Thomas Baldwin
most of Wittgenstein’s notes in fact deal with questions concerning doubt,
certainty and knowledge which are not discussed in this paper by Russell.
Instead there is a clear, if implicit, reference to Moore (‘A philosopher who
protests, “We KNOW there’s a chair over there” is simply describing a
game’ (PO 381)), and the discussions of doubt and certainty can be read as
responses to Moore’s approach to skepticism. Indeed, the similarity between
the positions suggested in these notes and some of those advanced later in
On Certainty indicates that the role of Wittgenstein’s discussions with Mal-
colm was in fact as much to remind Wittgenstein of what he had previously
thought about these issues as to stimulate him to new lines of thought.4
In the notes which comprise the first part of the text Wittgenstein is
primarily concerned to lay out what he takes to be unsatisfactory about
Moore’s response to skepticism, as exemplified by his characteristic claims
to knowledge. Wittgenstein takes it that Moore’s view was that in obvious
situations (‘Here is one hand’ etc.) we can tell just by introspection that we
know what is the case, as if knowledge is ‘an extremely important mental
state’ (OC §6) which we can identify in these obvious cases. Wittgenstein’s
objection to this view is that no such introspectively identifiable mental state
can guarantee the truth of its content (OC §§12–13); but of course ‘x knows
that p’ entails ‘p’. So knowledge cannot be a mental state of that kind, and
to get a better understanding of what it is we need to understand the role
of attributions of knowledge in our language-games (OC §18), a point to
which Wittgenstein returns in much more detail in the last part of his notes.
Having rejected what he takes to have been Moore’s refutation of skepti-
cism, Wittgenstein himself suggests the different strategy which I mentioned
earlier: he argues that the possibility of making a mistake about something
requires that one have some understanding of the subject in question, and
that the possession of such an understanding implies that one’s judgments
in straightforward cases be correct. Wittgenstein then uses this approach to
construct a new version of Moore’s famous ‘proof of an external world’:

This situation is thus not the same for a proposition like ‘At this dis-
tance from the sun there is a planet’ and ‘Here is a hand’ (namely my
own hand).
  For it is not true that a mistake merely gets more and more improb-
able as we pass from the planet to my own hand. No: at some point it
has ceased to be conceivable.
(OC §§ 52–4)

As we saw earlier Merleau-Ponty makes a similar claim, that there are limits
to error and doubt ‘because they are surrounded by an horizon of the world’
(PhP 419). It is not, I think, fanciful to suggest that these lines of argu-
ment are similar, in that the kind of understanding which is a prerequisite
of the capacity to make a mistake is one which provides one with at least
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  143
‘an horizon of the world’. Indeed, given Merleau-Ponty’s emphasis on the
fundamental importance of one’s embodiment to one’s grasp of the world, it
seems appropriate to attribute to him the thesis that this basic understand-
ing includes an awareness of the presence before one of one’s own hand
when it is easily visible.
In the next part of On Certainty Wittgenstein develops further this thesis
that there are limits to doubt by reinterpreting Moore’s common sense tru-
isms as providing a ‘picture of the world’ which is ‘the inherited background
against which I distinguish between true and false’ (OC §94). As discussed
earlier, these truisms are the ‘hinge’ certainties which provide a secure basis for
our ordinary empirical methods of inquiry. This status comes only from their
internal role within our system of beliefs (OC §136). They are not a priori
principles of the kind that rationalist philosophers aim to provide; nor are they
externally established by empirical investigations. Insofar as we have reasons
for believing them, therefore, these reasons can come only from our acceptance
of the understanding of the world and of ourselves which they facilitate.

When we first begin to believe anything, what we believe is not a single

proposition, it is a whole system of propositions. (Light dawns gradu-
ally over the whole).
(OC §141)

The child learns to believe a host of things. I.e. it learns to act according
to these beliefs. Bit by bit there forms a system of what is believed, and
in that system some things stand unshakeably fast and some are more
or less liable to shift. What stands fast does so, not because it is intrinsi-
cally obvious or convincing; it is rather held fast by what lies around it.
(OC §144)

In other passages Wittgenstein qualifies his description here of the hinge

propositions as propositions which ‘stand unshakeably fast’. For when com-
paring them to the banks of a river which facilitate the flow of water (i.e.,
ordinary empirical beliefs), he allows that some of these banks move from
time to time (OC §§96–9). His thought here seems to be that we modify our
ordinary beliefs all the time, and when a sufficient body of the beliefs which
‘lie around’ our core beliefs has altered we may find that we need to modify
some of these core beliefs too.5 Nonetheless, he maintains, revisions of this
kind are not just a matter of learning from experience since ‘experience is
not the ground for our game of judging’ (OC §131). Instead these revisions
involve a kind of ‘conversion’ in which one is ‘brought to look at the world
in a different way’ (OC §92).
These remarks introduce the issue of epistemic pluralism, according to
which there can be different systems of belief, involving different hinge
144  Thomas Baldwin
propositions, such that adherents of one system may not be in a position
to provide persuasive reasons in favour of their system to those who accept
a different one. Where we think of radical differences of belief, such as
those between ourselves and say, the ancient Greeks, this possibility does
not seem especially threatening. For in these cases the starting-points of
the two sides are so far apart that it is scarcely surprising that appeals
to experience and reasoning are unlikely to be persuasive. But the ques-
tion remains whether there is reason to think that, given unlimited time,
resources and goodwill, it should be possible to arrive at a reasonable con-
sensus involving beliefs which cluster around a single picture of the world.
For if this is not the case, epistemic pluralism threatens to lead to a rela-
tivist skepticism which rejects the possibility of objective knowledge. And
at one point in On Certainty Wittgenstein expresses some sympathy for a
conclusion of this kind:

But what men consider reasonable or unreasonable alters. At certain

periods men find reasonable what at other periods they found unrea-
sonable. And vice versa.
But is there no objective character (Merkmal) here?
Very intelligent and well-educated people believe in the story of cre-
ation in the Bible, while others hold it as proven false, and the grounds
for the latter are well known to the former.
(OC §336)

Although the tone of this passage is rather equivocal, and the case of bibli-
cal beliefs is a difficult one to adjudicate since the evidence, for and against,
theological beliefs is notoriously contested, this passage indicates Wittgen-
stein’s acceptance that this type of skeptical challenge to the ‘objective char-
acter’ of beliefs needs to be taken seriously.6 Despite one’s own picture of the
world being sure (sicher), the skeptical challenge to it cannot be dismissed
by a simple Moorean affirmation of knowledge of its truth. And yet Witt-
genstein was not a skeptic.
A central feature of his response is that we have no choice but to endorse
the standards of rationality that are inherent in our own system.

All testing, all confirmation and disconfirmation of a hypothesis takes

place already within a system. And this system is not a more or less
arbitrary and doubtful point of departure for all our arguments: no, it
belongs to the essence of what we call an argument. The system is not
so much the point of departure, as the element in which arguments have
their life.
(OC §105)

The thought here is that we cannot, and should not, attempt to respond to
the skeptic’s challenge to our system of beliefs with a complete epoché of
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  145
our standards for reasoning, in which we look for further reasons for choice
between different systems. For although we can and do reflectively assess
and modify our beliefs all the time, there can be no challenges to our beliefs,
no ‘arguments’ which lie altogether beyond our system; we have no choice
but to work from within. At first this may appear unsatisfying as we con-
sider the hypothesis of radically alien systems of belief. But Wittgenstein’s
claim is that this hypothesis turns out to be empty, in that any substantial
challenge to our system of beliefs requires arguments which we can under-
stand; but for each of us ‘the element in which arguments have their life’ is
precisely our own system of beliefs.
Wittgenstein’s line of thought here can be linked to his discussion of
rule-following in Philosophical Investigations. He argues here that in any
assessment of what we should think or do, we are bound to run out of rea-
sons: ‘the chain of reasons has an end’ (PI I §326); and ‘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” ’ (PI I§217). Hence, as he
famously concludes: ‘When I obey a rule, I do not choose. I obey the rule
blindly’ (PI I §219). So, returning now to On Certainty, despite the fact that
‘[t]he difficulty is to realise the groundlessness of our believing’ (OC §166),7
we should come to terms with the fact that this is an inescapable feature of
our lives, so that the groundlessness of the way in which we are sure of our
picture of the world is not ‘akin to hastiness or superficiality’, but is instead
part of our ‘form of life’ (OC §358). Wittgenstein then continues:

But that means I want to conceive it as something that lies beyond being
justified or unjustified; as it were, as something animal.
(OC §359)

This reference to what is ‘animal’ is then taken further in a well-known, but

difficult, later note:

I want to regard man here as an animal; as a primitive being to which

one grants instincts but not ratiocination.8 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
(OC §475)

The interpretation of this passage is not straightforward, but I think it should

be read in the light of Wittgenstein’s comment in Philosophical Investiga-
tions that ‘What we are supplying are really remarks on the natural history
of human beings’ (PI I§415). For this suggests that his claim in On Certainty
is not that our picture of the world is the product of our own ‘instincts’, our
own ‘primitive’ non-rational motivations. Instead what he has in mind is a
kind of natural history, according to which the antecedents of our inherited
146  Thomas Baldwin
picture of the world lie in the picture of the world acquired instinctively by
our primitive ancestors, and then slowly enriched and modified by subse-
quent generations who, by developing ‘a primitive means of communica-
tion’, eventually made it possible for us to have a language in which we can
express our system of beliefs and reason critically concerning it. One might
want to question the attribution of this speculative natural history to Witt-
genstein by bringing into play a later comment from Philosophical Investi-
gations: ‘we are not doing natural science; nor yet natural history—since we
can also invent fictitious natural history for our purposes’ (PI II §xii, 230e).
But in fact, I see no reason to regard the natural history sketched above as
wholly fictitious, though of course it is massively inexplicit in its positive
assumptions concerning the emergence of language. Indeed, it is supported
by some remarks by Wittgenstein from his 1937 discussion of doubt and
certainty to which I referred earlier, in particular by the role he assigns here
to certainty in the development of language:

The origin and the primitive form of the language game is a reaction;
only from this can more complicated forms develop.
Language—I want to say—is a refinement. ‘In the beginning was the
First there must be firm, hard stone for building, and the blocks are laid
rough-hewn one on another. Afterwards it’s certainly important that
the stone can be trimmed, that it’s not too hard.
The primitive form of the language game is certainty (Sicherheit), not
uncertainty. For uncertainty could never lead to action.
I want to say: it is characteristic of our language that the foundation on
which it grows consists in steady ways of living, regular ways of acting.
Its function is determined above all by action, which it accompanies.
(PO 394–7)

In the light of all this, then, how do things stand concerning the relativist
challenge to the objectivity of our beliefs which arises from the hypothesis
of radically different systems of belief? The situation is, I think, similar to
that which applies to his account of rule-following (i.e., we believe ‘blindly’,
without grounds or external reasons). But once we appreciate why this is
so, why reasons and justifications are inherently internal to our systems of
belief, we should also understand why the search for a further external vin-
dication of the system is pointless, so that in pressing such a question ‘you
are already going round in a circle’:

Well, if everything speaks for a hypothesis and nothing against it—is it

then certainly true? One may designate it as such—But does it cer-
tainly agree with reality, with the facts?—With this question you are
already going round in a circle.
To be sure there is justification; but justification comes to an end.
(OC §§ 191–2)
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  147
If further reassurance is sought, then one can also bear in mind the ‘natural
history’ of one’s beliefs, for this indicates that their roots lie in our practices,
in our ‘form of life’, and basic beliefs which are unhelpful or worse are likely
to have been discarded. At one point Wittgenstein remarks that his line of
thought ‘sounds like pragmatism’ (OC §422)—he probably has William
James in mind; but as if to head off the attempt to attribute to him anything
like Carnap’s notion of a pragmatic ‘external reason’, he adds:

This game proves its worth. That may be the cause of its being played,
but it is not the ground.
(OC § 474)

So although the appeal to natural history can be reflectively incorporated

into our system of belief, where it provides some further reason for confi-
dence in the system, pragmatic considerations by themselves do not provide
an external ‘ground’ for the system.

3.  Merleau-Ponty on Certainty

In turning now to continue my comparison between the positions advanced
by Wittgenstein and Merleau-Ponty, I begin with a few comments about
Merleau-Ponty’s Phenomenology of Perception.10 The book was published
in 1945, having been written during the war. Merleau-Ponty’s project in the
book was to fulfil a project which his earlier work, The Structure of Behav-
iour, had suggested to him. His conclusion there, reached after an inquiry
into the psychology of perception which included a sympathetic assessment
of the work of the Gestalt psychologists, had been that natural science, so
far from being a way of revealing the basis of perception, is itself dependent
upon the fact that ‘laws have meaning only as a means of conceptualising
the perceived world’ (SB 145). Merleau-Ponty takes this to imply ‘that the
universe of naturalism has not been able to become self-enclosed, and that
perception is not an event of nature’ (SB 145). I shall say more about these
striking claims later, but what is important here is that in the light of this
conclusion Merleau-Ponty proposes that in order to understand the role of
the senses, and the body generally, one should ‘return to perception as a type
of original experience in which the real world is constituted in its specificity’
(SB 220); and he then adds in a footnote to this proposal the remark ‘We
are defining here the “phenomenological reduction” in the sense given to
it in Husserl’s final philosophy’ (SB n. 56, 249). This ‘return to perception’
by means of a ‘phenomenological reduction’ is then the project of his next
work, his Phenomenology of Perception.
A central theme of this book concerns the role of the body as ‘the ­subject
of perception’ (PhP 213) which sustains our relationship with the world
since ‘we are in the world through our bodies and we perceive the world
through our bodies’ (PhP 213). In the following dense passage Merleau-
Ponty summarises the complex position he takes himself to have established:
148  Thomas Baldwin
The subject is being-in-the-world and the world remains ‘subjective’,
since its texture and its articulations are sketched out by the subject’s
movement of transcendence. Thus, along with the world—as the cradle
of significations, as the sense of all senses, and as the ground of all
thoughts—we also discovered the means of overcoming the alternatives
between realism and idealism, between contingency and absolute rea-
son, and between non-sense and sense. The world, such as we have
attempted to reveal it—as the primordial unity of all of our experiences
on the horizon of our life and as the unique term of all our projects—is
no longer the visible unfolding of a constituting thought, nor a fortu-
itous collection of parts, and certainly not the operation of a directing
thought upon an indifferent matter; rather, the world is the homeland
of all rationality.
(PhP 454)

There is much here that invites further discussion, and I shall come back to
it later. But in the present context it is significant that if we take ‘the world’
to be the intentional object of a Wittgensteinian system of beliefs, i.e., to be
the world as it is believed to be, then the thought that the world so conceived
is ‘the cradle of significations’, ‘the sense of all senses’, ‘the ground of all
thoughts’, and ‘the homeland of all rationality’ can be seen to be not so far
removed from some of the central themes of Wittgenstein’s position, though
I shall argue at the end that there are also significant differences here.
Before going into more of these questions and comparisons, however, there
are a few mundane details about the relations, direct and indirect, between
Merleau-Ponty and Wittgenstein to lay out. First, although ­Merleau-Ponty’s
Phenomenology of Perception was published in 1945, well before Wittgen-
stein commenced his notes On Certainty in 1949, these works were writ-
ten at quite different stages in their lives (Wittgenstein was born in 1889,
­Merleau-Ponty in 1908) and there was no reason for Wittgenstein to have
heard of Merleau-Ponty, let alone read his Phenomenology. Equally one
could not expect Merleau-Ponty to have heard of Wittgenstein in 1945.
One might hope that Merleau-Ponty would subsequently have read and
appreciated some of Wittgenstein’s writings (though of course he died in
1961, long before On Certainty was published in 1969). But the only refer-
ence to Wittgenstein I know of in Merleau-Ponty’s writings is enigmatic; it
occurs in the course notes from his 1959–60 lectures on Husserl’s paper ‘On
the Origin of Geometry’. At one point where Merleau-Ponty is discussing
Husserl’s account of language, he notes:

Husserl takes [11language seriously, gives it an ontological function,

makes it bear a page of Being, precisely because he is not enclosed (Witt-
genstein) (the British) in the immanence of language, conceived as a
thing, but follows its implications of sense . . .].
(HLP 43; translation modified)
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  149
It is not clear here which writings by Wittgenstein Merleau-Ponty has in
mind here (or who ‘the British’ are); nor whether the reference to Wittgen-
stein and the British should be taken to imply that they, unlike Husserl,
are ‘enclosed in the immanence of language’—or that they are like Husserl
and not so enclosed. The former interpretation seems more likely since it
would imply that Merleau-Ponty is saying that Husserl is to be esteemed
for an insight concerning language which is missing from the discussions
of language by Wittgenstein and the British, which is a more interesting
statement than one which simply treats them all on a par. As far as the ref-
erence to Wittgenstein is concerned, one might then take the remark about
being ‘enclosed in the immanence of language’ to be a comment on the
position advanced by Wittgenstein in the Tractatus Logico-Philosophicus,
as exemplified perhaps by the famous proposition that ‘The limits of my
language mean the limits of my world’ (TLP §5.6). But it would in fact be
a serious misunderstanding of the Tractatus to suppose that Wittgenstein
here conceives of language ‘as a thing’; instead he takes it to be a system
of propositional signs, which are facts, not things. And he certainly does
‘follow its implications of sense’ in one sense, in that he takes the require-
ment that sense be determinate to imply that objects are simple. If one were
to turn instead to Wittgenstein’s later writings, then one would encounter
immediately his conception of a language-game and it is absolutely clear
that when discussing language-games he is not ‘enclosed in the immanence
of language, conceived as a thing’, since the point of the conception of a
language-game is to show how the meaning of language is inseparable from
its use in the context of activities such as house building; and he does follow
‘its implications of sense’ by emphasising its presuppositions, such as agree-
ment in judgment (PI I§242). Indeed, Wittgenstein makes much more of this
aspect of language than Husserl does. So whatever interpretation one gives
to this passing reference, it is disappointing. The conclusion must be that,
for reasons unknown, Merleau-Ponty never acquired a proper understand-
ing and appreciation of Wittgenstein’s writings.12
I now return to the main topic of this section, Merleau-Ponty’s account
of certainty, doubt and knowledge, and the similarities and differences
between this account and Wittgenstein’s position as discussed in the previ-
ous section. A good place to start is with Merleau-Ponty’s affirmation of
the ‘absolute certainty of the world in general, but not of any particular
being’ (PhP 311). I suggested earlier that there are significant similarities
between this claim and Wittgenstein’s remarks about the certainty of the
general ‘picture of the world’ provided by the Moorean hinge propositions.
In considering these similarities further, however, it is useful to start from an
assessment of the connection between the positions of Merleau-Ponty and
Husserl concerning this issue. Merleau-Ponty uses phrases such as ‘originary
opinion’ (PhP 358), ‘primordial opinion’ (PhP 359) and ‘primordial faith’
(PhP 431) to describe our attitude to the world which is absolutely certain,
and in a footnote he alludes to Husserl’s ‘Urdoxa or Urglaube’ as his source
150  Thomas Baldwin
for these phrases (PhP 359 fn. 96, see PhP 552). He does not refer here to
any of Husserl’s texts, but in a later footnote (PhP 382 fn. 14, see PhP 553),
he refers to the way in which in Experience and Judgment Husserl ‘grounds
knowledge upon an originary doxa’. This note shows that Merleau-Ponty
must have found Experience and Judgment, Husserl’s last book, among
the papers which Father van Breda brought to Louvain in 1938 and which
Merleau-Ponty was able to study there before the German invasion in 1940.
It is in this book that Husserl writes about the ‘Urdoxa’ (Husserl 1973:
59), the ‘originary doxa’, and Merleau-Ponty’s references to it indicate his
acceptance of something close to Husserl’s doctrine that all justifications
of claims to knowledge lead back to ‘the original experience of the life-
world’ (Husserl 1973: 45), and thus to a kind of ‘pre-predicative experience’
(as Husserl calls it; I discuss the conception of ‘pre-predicative experience’
further below) which provides a ‘justification of doxa which is the realm
of ultimately original self-evidence’ (Husserl 1973: 46). But it is, I think,
open to question quite how far Merleau-Ponty’s acceptance of Husserl’s
position extends. While Merleau-Ponty does agree that the intentionality of
judgment depends ‘as its very condition of possibility’ upon the ‘operative
intentionality’ (PhP 453) inherent in our experience, it does not follow that
he agrees that, in providing justifications for our ordinary beliefs, this expe-
rience needs the status of ‘ultimately original self-evidence’. Indeed, when
discussing how it is that hallucinations can deceive us, he remarks that it is
necessary to ‘strip perception of its apodictic certainty’ (PhP 359; cf. ‘Pref-
ace’ lxxx). So I take it that Merleau-Ponty qualifies Husserl’s conception of
the epistemic status of experience: the ‘absolute certainty of the world in
general’ does not require a foundation in self-evident experience. Instead the
existence of the world is a matter of ‘primordial faith’ (PhP 431) because it
is a condition of the possibility of the inherently fallible process of obtaining
knowledge about the ‘particular beings’ that we perceive.
Despite his rejection of Husserl’s conception of the original self-evidence
of experience, however, Merleau-Ponty does agree with Husserl that our
beliefs get their meaning and warrant from perception, which is our ‘origi-
nary knowledge’ (PhP 45) and is therefore ‘the background against which
all objects stand out and is thus presupposed by them’ (PhP lxxiv). By con-
trast Wittgenstein’s view (as we saw earlier) is that ‘experience is not the
ground for our game of judging’ (OC §131). So whereas Merleau-Ponty
takes it that perception is the (back)ground for our judgments, for Witt-
genstein there is no such external ground. Grounds, like reasons and argu-
ments, are essentially internal to our system of beliefs, which include our
perceptual judgments, though not the perceptions themselves which cause
these judgments and which are not grounds for them.
However, there are two significant complications which affect this com-
parison. The first starts from the fact that Merleau-Ponty holds that the alter-
natives ‘reason’ and ‘cause’ are not exhaustive: he takes it that there is also
an intermediate category of ‘motive’ (PhP 51) which connects the sensible
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  151
phenomena given in experience in such a way that they are perceived as fea-
tures such as the colours, shape and size of objects located before one—for
example, as the colour, shape and size of a glass of water on a table before
one. The connections here are not reasons, since no reasoning is involved;
nor, he maintains, are they causes since perception is interpretive—sensible
phenomena motivate a unified interpretation of the scene before one as ‘the
sense it offers’ (PhP 51). The assumption here that ‘sense’ (i.e., meaning)
and causation are incompatible is characteristic of Merleau-Ponty’s posi-
tion, and I shall briefly come back to it later. But continuing now with the
account of his position, he holds that it is not only the unity of the perceived
world that is motivated in this way, our perceptions also motivate the belief
that things are as they appear to us—for example that that there is indeed
a glass of water on the table before one. For this connection is not a matter
of reasoning; although when we reflect we can bring reasoning to bear on
our beliefs about whether things are indeed as they appear, the connection
between unreflective perception and belief is immediate and unreasoned.
Equally, Merleau-Ponty holds, this connection between perception and
belief is not simply causal, for it is crucial that the perception transmits its
intentional content to the belief, and this involvement of intentional content
in the relationship cannot, he thinks, be just a matter of causation since the
content itself is not a causal structure. So when towards the end of the book
he affirms, concerning the ways in which ‘things take shape’ and in which
we understand ourselves and others, that ‘All that remains is to recognise
these phenomena that ground all of our certainties’ (PhP 431), we should
not take his claim to be that the phenomena are ‘reasons’ for our certainties;
instead his claim is that they are the ‘motives’ for them.
Once Merleau-Ponty’s position is understood in this way, it is clear that it
offers a significant alternative to Wittgenstein’s position in On Certainty.13
Nonetheless, the fact that Merleau-Ponty uses the term ‘ground’ to charac-
terise his position is also indicative of the fact that he accepts that there is a
sense in which perception provides the foundation for our beliefs, including
‘our certainties’. On this issue, therefore, there remains, in the face of it, a
significant disagreement between Merleau-Ponty and Wittgenstein. But at
this point a second complication needs to be addressed, which starts from
Wittgenstein’s position. As I emphasised earlier, Wittgenstein distinguishes
the secure certainty of the core propositions in our system of belief from
our knowledge, and maintains that claims to knowledge typically occur in
discussions as to whether someone is a reliable source of information about
some question, whether ‘he is in a position to know’ (OC §555). Since
someone is in a position to know something where they have good evidence
for what they think, it follows that ‘Whether I know something depends
on whether the evidence backs me up or contradicts me’ (OC §504). So
Wittgenstein certainly does hold that knowledge involves evidence, and thus
grounds, although, he thinks, we cannot have grounds for our system of
belief and its core certainties as a whole. Thus, one may object, the contrast
152  Thomas Baldwin
suggested earlier between Wittgenstein and Merleau-Ponty is overstated,
since Wittgenstein and Merleau-Ponty at least agree that objective knowl-
edge needs grounds. Admittedly, they then disagree about these grounds:
for Wittgenstein the grounds are supplied by evidence which belongs within
systems of belief (which include perceptual judgments) that lack grounds,
whereas for Merleau-Ponty it is perception that is the ground of all of our
judgments. Despite this difference, and the point emphasised earlier, that
Merleau-Ponty distinguishes between two types of knowledge, ‘thetic’ and
‘non-thetic’, in a way which is not found in On Certainty, however, there
is more similarity between their positions than the previous comparison
allowed for.
To pursue further the questions raised by these complications it is neces-
sary to attend to some further features of Merleau-Ponty’s position.

4.  ‘Thetic’ and ‘Operative’ Intentionality

Merleau-Ponty’s distinction between reasons and motives connects with the
distinction I mentioned before between two kinds of knowledge, ‘thetic’ and
‘non-thetic’, since reasons connect the thetic judgments such as those which
enter into our ‘objective and detached knowledge of the body’ (PhP 213)
whereas motives connect the non-thetic states which provide ‘this other
knowledge that we have of it because it is always with us and because we are
bodies’. Early on in Phenomenology of Perception Merleau-Ponty explains
this ‘thetic/non-thetic’ distinction in the following way;

Now, as we have just seen, the perception of one’s own body and exter-
nal perception offer us the example of a non-thetic consciousness, that
is, of a consciousness that does not possess the full determination of its
objects, the example of a lived logic that does not give an account of
itself, . . . Objective thought cannot assimilate these phenomena.
(PhP 50–1)

So, as applied to knowledge, the contrast is one between a kind of cer-

tain knowledge that ‘does not possess the full determination of its objects’
and ‘does not give an account of itself’, and a different kind of knowl-
edge, typically objective knowledge, which does have these characteristics.
Wittgenstein does not propose any such distinction between two kinds of
knowledge in his discussions of knowledge in On Certainty; hence in com-
paring their accounts of knowledge we need to consider what the distinction
amounts to.
When one first encounters it, Merleau-Ponty’s distinction seems reminis-
cent of Russell’s distinction between propositional knowledge by description
and non-propositional knowledge by acquaintance of things and properties
(Russell 1912: ch. v).14 Knowledge by description fits the specification of
thetic knowledge since by describing its object it provides both ‘an account’
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  153
of what is known and ‘the full determination of its objects’ in the sense
that the description of the object implies its existence and identity. Equally,
since knowledge by acquaintance is supposed to involve no more than the
immediate presence of an object such as a specific colour to a subject, it
appears to satisfy Merleau-Ponty’s requirements for non-thetic knowledge
in that it ‘does not give an account of itself’ or determine its object. So
Russell’s description/acquaintance distinction seems to fit Merleau-Ponty’s
thetic/non-thetic distinction reasonably well. However, this interpretation
of Merleau-Ponty’s distinction would expose it to the criticisms of knowl-
edge by acquaintance advanced by Wittgenstein in the early sections of his
Philosophical Investigations. He argues here that knowledge of the colour
one is seeing, for example, requires an understanding of what it would be to
see the same colour on another occasion, and that this understanding is not
provided just by the simple demonstration of one sample of the colour since
‘an ostensive definition can be variously interpreted in every case’ (PI I §28).
So although visual acquaintance can contribute to a person’s knowledge of
some colour where the person already has a general understanding of the
role of colour words in the language (PI I §30), there can be no pure knowl-
edge by acquaintance alone, let alone the ‘perfect’ and ‘complete’ knowl-
edge Russell supposed it to be (Russell 1912: 73).
Hence, if the Russellian interpretation of Merleau-Ponty’s thetic/non-
thetic distinction were correct, Merleau-Ponty’s distinction, and the doc-
trines associated with it, would be vulnerable to a lethal criticism from
Wittgenstein. But in truth this interpretation is mistaken. Although the
association between thetic knowledge and knowledge by description is rea-
sonable, Merleau-Ponty’s conception of non-thetic knowledge, and non-
thetic consciousness generally, is far removed from Russellian knowledge by
acquaintance, which Merleau-Ponty also emphatically rejects. Right at the
start of Phenomenology of Perception he writes:

In beginning the study of perception, we find in language the seemingly

clear and straightforward notion of sensation: I sense red or blue, hot
or cold. We will see, however, that this is the most confused notion
there is . . .
(PhP 3)

Merleau-Ponty’s reason for rejecting anything like Russellian acquaintance

with colours is that he takes it that perception, which includes sensation,
always involves a ‘field’ with a spatial figure/ground structure in which
colours are manifest as the colours of things. So—

The red patch I see on the rug is only red if the shadow that lies across it
is taken into account; its quality only appears in relation to the play of
light, and thus only as an element in a spatial configuration.
(PhP 5)
154  Thomas Baldwin
This inescapable complexity in perception is compounded by the indetermi-
nacy which arises from the open structure of the perceptual field, whereby
changes in light, distance, and orientation induce changes in what is per-
ceived. So there can be no pure acquaintance with a region whose colour
is wholly determinate all by itself; even the perceived colour of standard
colour samples varies with lighting and context.
Thus there can be no question of interpreting Merleau-Ponty’s non-thetic
knowledge as Russellian knowledge by acquaintance. What is nonetheless
worth remarking here is the difference between the criticisms advanced by
Merleau-Ponty and those of Wittgenstein. Where Merleau-Ponty concen-
trates on the complexity of the perceptual field, Wittgenstein emphasizes the
substantive assumptions inherent in our understanding of colour words and
other terms which describe potential objects of acquaintance. Are there con-
nections here? Wittgenstein’s key point, that ‘an ostensive definition can be
variously interpreted in every case’, is supported by Merleau-Ponty’s thesis
of the indeterminacy of the perceptual field inherent in an ostensive defini-
tion. But whether one might similarly argue from Wittgenstein’s point to
Merleau-Ponty’s indeterminacy thesis depends on whether our non-thetic
knowledge of the world is conceived as in some way dependent on the use
of language. On the face of it this is not the case insofar as non-thetic knowl-
edge ‘does not give an account of itself’. But first we need a better account
of Merleau-Ponty’s conception of non-thetic knowledge.
A helpful way into this question is to start from another familiar dis-
tinction between two kinds of knowledge, that between ‘knowing that’
and ‘knowing how’. Thetic knowledge is obviously knowledge that; but
is ‘know-how’ non-thetic knowledge? It does in a way satisfy all too eas-
ily the requirements of not possessing ‘the full determination of its objects’
and exemplifying ‘a lived logic that does not give an account of itself’ (PhP
50–1). One who knows how to ride a bicycle does not need to ‘give an
account’ of this know-how, of how one adjusts one’s posture to be able
to negotiate corners and so on. But on reflection, it is clear that Merleau-
Ponty’s conception of the non-thetic knowledge of the world and ourselves
provided by perception and employed in action is not itself simply a mat-
ter of ‘know-how’; instead, it is the practical understanding of the world
and oneself which is manifested in one’s ability to do the things which one
knows how to do, such as, indeed, riding a bicycle to work. This practical
understanding is primarily a matter of knowledge of one’s location, ori-
entation, situation and context, but also of the intentions and thoughts of
one’s companions which is essential to the ability to engage with others in
shared activities. As such it also meets the twin requirements of not requir-
ing ‘the full determination of its objects’, e.g., a detailed knowledge of one’s
geographical location, or ‘an account of itself’, e.g., an account of how we
maintain our normal orientation while moving. By contrast, then, thetic
knowledge will typically be the detached, objective knowledge of the world,
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  155
ourselves and others which is not immediately provided by perception and
used in action, though of course it is an invaluable resource for them.
It is uncontentious that one can distinguish in this way between prac-
tical understanding and objective knowledge. The difficult issue concerns
the significance of this distinction. Wittgenstein’s examples in On Certainty
include both the objective knowledge that 12 × 12 = 144 (OC §43) and the
practical knowledge inherent in agency, as in his comment:

My life shews that I know or am certain that there is a chair over there,
or a door, and so on.—I tell a friend e.g. ‘Take that chair over there’,
‘Shut the door’, etc. etc.
(OC §7)

But Wittgenstein treats both cases in much the same way: because they con-
cern obvious truths, there is no easy way in which one can be mistaken
about them, since understanding them includes the ability to judge correctly
about them in normal situations (OC §81). By contrast, for Merleau-Ponty,
the non-thetic practical understanding provided by ‘originary perception’
(PhP 252) has a foundational role with respect to thetic knowledge since it
provides the ‘certainty of the thing and the world’ which ‘precedes the thetic
knowledge of their properties’ (PhP 402). Admittedly, for Merleau-Ponty
non-thetic perception does not provide reasons for thetic beliefs; instead it
provides ‘motives’ for them. But this relationship is one of one-way depen-
dence; as Merleau-Ponty puts it towards the end of Phenomenology of Per-
ception, ‘We uncovered, beneath act or thetic intentionality—and in fact
as its very condition of possibility—an operative intentionality’ (PhP 453)
which, we are told at the start of the book in the ‘Preface’ is ‘the intentional-
ity that establishes the natural and pre-predicative unity of the world and of
our life’ (PhP lxxxii). So for Merleau-Ponty the distinction between objec-
tive thetic knowledge and practical non-thetic understanding is central to
the structure of human knowledge.
To explore this further, we need to consider another distinction which
Merleau-Ponty takes over from Husserl and associates with the thetic/non-
thetic distinction, that between ‘predicative’ and ‘pre-predicative’ knowl-
edge. ‘Predicative’ knowledge involves acts of predication, i.e., judgment
(PhP lxxxii), and is typically expressed in language (PhP lxxix); so it is
thetic knowledge. ‘Pre-predicative’ knowledge (PhP 77), by contrast, is
the non-thetic practical knowledge inherent in perception which provides
for ‘the natural and pre-predicative unity of the perceived world’ (PhP
lxxxii). We encountered the conception of ‘pre-predicative experience’ ear-
lier in the context of Husserl’s hypothesis of an Urdoxa, an ‘original faith’
involving pre-predicative experience (Husserl 1973: 45). I observed that
­Merleau-Ponty does not follow Husserl in taking pre-predicative experi-
ence to be self-­evident; but it remains central to his account of knowledge.
156  Thomas Baldwin
What makes it ‘pre-predicative’ is that it ‘precedes’ predication, which
implies not only that it does not include an act of judgment but also that
its intentional content is in some sense antecedent to the predications which
are characteristic of language. Just what this sense amounts to, however, is
difficult to pin down. At one point Merleau-Ponty expresses the distinction
between predicative thetic and pre-predicative non-thetic intentionality in
the following way:

Husserl’s originality lies beyond the notion of intentionality; rather it

is found in the elaboration of this notion and in the discovery, beneath
the intentionality of representations, of a more profound intentionality,
which others have called existence.
(PhP 124 n. 57, see p. 520)

This passage indicates that pre-predicative intentional content is not con-

stituted by ‘representations’, which include both the supposed bodily rep-
resentations Merleau-Ponty criticises in his discussion of the phantom limb
(PhP 82ff.) and the intellectual representations hypothesised by rationalist
philosophers and psychologists which Merleau-Ponty also rejects (PhP 140
n. 99, see PhP 523–5). His positive account of this ‘more profound inten-
tionality’ is then suggested in the following remark:

The gesture of reaching one’s hand out toward an object contains a

reference to the object, not as a representation, but as this highly deter-
minate thing towards which we project ourselves, beside which we are
in anticipation and which we haunt.
(PhP 140, translation modified)

So we find here, in the context of a discussion of bodily movement as ‘origi-

nal intentionality’ (PhP 139), a conception of intentional content as inher-
ently relational, directed to ‘this highly determinate thing towards which
we project ourselves’. What then needs to be added is that intentional
content so conceived, even if it is ‘pre-predicative’, is not altogether ‘pre-­
conceptual’ (or ‘non-conceptual’). Instead the ways in which objects and
events are referred to in gestures such as reaching will typically draw on
indexical ­spatio-temporal indications of them as, so to speak, ‘over there’ or
‘a moment ago’, where the indications are gestural, although one can subse-
quently use spoken language to identify and discuss them.
At this point we have reached the core of Merleau-Ponty’s position in
Phenomenology of Perception—his conception of our non-thetic, pre-­
predicative, understanding of the perceived world which underpins our
ordinary thetic, predicative, knowledge of things in such a comprehen-
sive way that he is led to write that ‘the being of the perceived is the pre-­
predicative being toward which our total existence is polarised’ (PhP 336).
Indeed, it is worth quoting in full a passage from the end of Phenomenology
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  157
of Perception to which I have already referred in which Merleau-Ponty pres-
ents the unqualified dependence of thetic intentionality on non-thetic inten-
tionality as a conclusion which he has established:

We uncovered, beneath act or thetic intentionality—and in fact as its

very condition of possibility—an operative intentionality already at
work prior to every thesis and every judgment.
(PhP 453)

There is no sign here of any qualification; this fundamental operative inten-

tionality is said to be at work ‘prior to every thesis and every judgment’.
On this issue, then, there is, it would seem, a clear difference between
­Merleau-Ponty’s foundationalism and Wittgenstein’s holistic conception of
our systems of beliefs in On Certainty, despite the complications introduced
in the previous section. But there are two ways in which one might argue
that even this apparent disagreement turns out, on further investigation, to
be a misconception. First, by dissipating the foundationalist appearance of
Merleau-Ponty’s position; second, by identifying a foundationalist theme
in Wittgenstein’s position comparable to that attributed to Merleau-Ponty.
I take these suggestions in turn.
In his first book, The Structure of Behavior, Merleau-Ponty writes:

Man is not a rational animal. The appearance of reason and mind does
not leave intact a sphere of self-enclosed instincts in man.
(SB 181)

If one associates ‘reason and mind’ with language and predicative judg-
ment, and ‘instincts’ with pre-predicative practical understanding, Merleau-
Ponty’s claim here implies that the emergence of a capacity for predicative
judgment ‘does not leave intact a sphere of self-enclosed’ pre-predicative
practical understanding of the world. This does not show that the former
capacity does not continue to depend on the latter; but once the depen-
dence between these two types of capacity, between the act intentionality of
judgment and the pre-predicative intentionality of practical understanding
is regarded as a two-way interdependence, the significance of the claim that
the pre-predicative intentionality of practical understanding is ‘already at
work prior to every thesis and every judgment’ has been qualified, since
the priority thesis has to be understood to allow for the fact that this pre-
predicative intentionality is not a ‘self-enclosed sphere’, but is also informed
by predicative knowledge of the world. In Phenomenology of Perception
Merleau-Ponty provides examples of this relationship in his discussion of
the acquisition of skills (‘habits’), such as learning to type at a keyboard or
to play the organ (PhP 145–8). For in these cases, predicative knowledge of
the keyboard (etc.) is transformed into a new ‘habit’ which enters into our
practical understanding of ourselves and the world:
158  Thomas Baldwin
The body, then, has understood and the habit has been acquired when
the body allows itself to be penetrated by a new signification, when it
has assimilated a new meaningful core.
(PhP 148)

Thus the capacity for assimilating a new signification shows that non-thetic,
operative intentionality can be informed by a ‘signification’ which is initially
presented in thetic, predicative knowledge. So the basic foundationalist the-
sis that predicative knowledge depends, in every case, on pre-predicative
understanding has to be understood to allow that this pre-predicative under-
standing can be enriched, or extended, by significations which are drawn, in
the first instance, from predicative knowledge.
Yet while this qualification is significant, there is no sign in Phenomenology
of Perception of Merleau-Ponty withdrawing his claim that pre-­predicative
intentionality is ‘prior to every thesis’. Thus if the apparent disagreement here
between Merleau-Ponty and Wittgenstein is to be overcome it will have to
be by identifying a foundationalist theme in Wittgenstein’s position compa-
rable to that affirmed by Merleau-Ponty.
To consider this suggestion, I want to go back to the passage from Witt-
genstein’s 1937 notes, part of which I repeat here:

Language—I want to say—is a refinement. “In the beginning was the

. . .
I want to say: it is characteristic of our language that the foundation on
which it grows consists in steady ways of living, regular ways of acting.
Its function is determined above all by action, which it accompanies.
(PO 394–7)

The claim here that language has a ‘foundation . . . in steady ways of living,
regular ways of acting’ employs a distinction between language and action
which is comparable to Merleau-Ponty’s predicative/pre-predicative distinc-
tion; and the fact that Wittgenstein writes here of action as a ‘foundation’
implies, on the face of it, that this distinction has precisely the foundational
significance that Merleau-Ponty attributes to it. However, for Wittgenstein
this applies only to the initial emergence of language. Once language-games
are up and running, the need for ‘steady ways of living, regular ways of
acting’ will be met by ways of living and acting which are informed by lan-
guage, in practices which, to borrow Merleau-Ponty’s idioms, are penetrated
by new significations so that they assimilate ‘a new meaningful core’. And
now the question is whether there is good reason to continue to distinguish
between predicative judgment and pre-predicative knowledge and give the
latter a foundationalist role with respect to the former. Merleau-Ponty, as we
have seen, seems to think that this is the case. But once one understands that
the point of the conception of a language-game is precisely to emphasise the
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  159
interdependence between language and action in our practices, there seems
no good reason to take it that the predicative component of these practices
which is manifest in what is being said is dependent upon our practical
understanding of our situation in a way in which the latter is not equally
dependent on the former. Indeed, for Wittgenstein the core presupposition
of our language-games is that we are able to sustain agreement in judgment
concerning particular cases; so if anything, it is the predicative component
of our practices which he takes to be fundamental. However, what seems
central to his position is really the interrelatedness of action and language.
Thus, going back now to a passage cited earlier from On Certainty—

My life shews that I know or am certain that there is a chair over there,
or a door, and so on.—I tell a friend e.g. ‘Take that chair over there’,
‘Shut the door’, etc. etc.
(OC § 7)

—we see that Wittgenstein illustrates here the connection between knowl-
edge and action by reference to what he tells his friend. There is no
­predicative/pre-predicative distinction in play here.
It might be suggested that the ‘rule-following’ considerations provide a
better way of identifying a foundationalist theme in Wittgenstein’s position
that resembles Merleau-Ponty’s position. For in his Remarks on the Founda-
tions of Mathematics Wittgenstein famously states that

Following according to the rule is FUNDAMENTAL to our language-

game. It characterises what we call description.
(RFM 330)15

To connect this remark with Merleau-Ponty’s position, however, it needs

to be appropriate to take it that ‘our language-game’ exemplifies predica-
tive, thetic judgments, whereas ‘following according to the rule’ involves
the exercise of non-thetic, pre-predicative understanding, so that the latter
is being said to be ‘FUNDAMENTAL’ to the former. Since it is obvious
that language-games involve predicative judgment, the issue here resolves
around the significance of the phrase ‘following according to the rule’. To
obtain a proper understanding of this phrase it is, I think, essential to place
the passage in which it occurs in the context of the preceding remark:

Someone asks me: What is the colour of this flower? I answer: ‘red’.—
Are you absolutely sure? Yes, absolutely sure! But may I not have
been deceived and called the wrong colour ‘red’? No. The certainty
with which I call the colour ‘red’ is the rigidity of my measuring-rod,
it is the rigidity from which I start. When I give descriptions, that is
not to be brought into doubt. This simply characterises what we call
160  Thomas Baldwin
(I may of course even here assume a slip of the tongue, but nothing
(RFM §28)

Two points are immediately noticeable here: first, Wittgenstein is writing

about descriptions of colour, and not, for example, proofs; second, the
emphasis here is on certainty. Wittgenstein claims that our capacity for
describing colours starts out from our certainty concerning straightforward
cases (e.g., that this flower is red, a certainty which constitutes ‘the rigid-
ity of my measuring rod’). The way in which Wittgenstein then ends this
first remark implies that what ‘characterizes what we call describing’, at
least when one is describing colours, is precisely this certainty concerning
straightforward cases which provides us with, as it were, ‘the rigidity of
my measuring rod’. The next remark ends with a very similar comment
‘It characterizes what we call description’; and although ‘it’ here refers to
‘Following according to the rule’, the meaning of this comment must be
similar to that of the earlier one. The obvious way to connect them, there-
fore, is to regard this second remark as a generalisation of the first, so that
in the case of describing colours ‘following according to the rule’ is a matter
of relying on our certainty concerning straightforward cases (‘this flower
is red’) to describe more complex colours. The general point here, there-
fore, concerns an epistemological foundation of our language-game: what
is ‘FUNDAMENTAL’ to our language-game is the certainty concerning
obvious cases which provides us with ‘rigid measuring rods’ which we can
apply to situations where the facts are not obvious. As such, therefore, this
claim simply anticipates the position developed in much more detail in On
Certainty. Indeed, there is an extended discussion of our understanding of
colour words in On Certainty (§§522–30) in which Wittgenstein develops
and explores further the position he has put forward in these two remarks.
This account of the significance of Wittgenstein’s remark is reminiscent of
Merleau-Ponty’s claim that ‘the certainty of the thing and the world precedes
the thetic knowledge of their properties’ (PhP 402). However, the details do
not match up in such a way that this discussion of rule-following provides a
basis for attributing to Wittgenstein a foundationalist position comparable
to that advanced by Merleau-Ponty. For there is nothing non-thetic or pre-
predicative in the conception of rule-following as reliance on the certainty
of our judgments in straightforward cases such as ‘This flower is red’. Fur-
thermore, as the more extensive discussion in On Certainty indicates, it was
not Wittgenstein’s view that our certainty concerning these straightforward
cases provides a ground for our judgments in more difficult ones; instead the
system of beliefs (judgments), which includes the distinctive ‘rigid’ certainty
of ‘This flower is red’, just needs to fit together as a whole (cf. OC §144).
Thus what is ‘FUNDAMENTAL’ to our language-game about following
according to a rule is the general fact that ‘the truth of certain empirical
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  161
propositions belongs to our frame of reference’ (OC §83). It is not that these
empirical certainties provide a foundation for our system of beliefs.16

5.  The Subjective World

The conclusion of this discussion is, therefore, that, despite all the com-
plications and qualifications discussed in the previous section, there is a
clear difference between Wittgenstein’s holistic conception of our systems
of beliefs in On Certainty and Merleau-Ponty’s thesis that there is a kind of
pre-predicative practical understanding (an ‘operative intentionality’) which
is ‘already at work prior to every thesis’ (‘act intentionality’) and thus in
particular prior to predicative, propositional knowledge of the world. In
this final section I want to explore Merleau-Ponty’s reasons for this position
and then consider how far Wittgenstein might be sympathetic to them.
In the course of the preceding discussion, it will have become apparent
that Merleau-Ponty connects the predicative/pre-predicative distinction with
an objective/subjective distinction, where the subjectivity of a state such as
perception is understood as arising from its essential dependence upon a rela-
tionship with the subject as such, so that knowledge is objective where no
such dependence upon the subject is involved. Objective knowledge is cer-
tainly predicative, and even though there is subjective predicative knowledge,
expressed through indexical judgments, Merleau-Ponty implies that what is
important about predicative judgment is that it brings with it a capacity for
objective knowledge. By contrast pre-predicative knowledge is inherently sub-
jective since it is structured by non-representational indexical relationships
between the subject and the world. Thus Merleau-Ponty’s epistemological the-
sis of the priority of the pre-predicative over the predicative indicates that he
assigns priority to experiences informed by subjective relationships over objec-
tive knowledge. But more needs to be said about the objective/subjective dis-
tinction and the types of priority that can be brought into play in this context.
It is obvious from the perspectival character of perception that perception
is inherently subjective in Merleau-Ponty’s sense; there is no objective ‘view
from nowhere’. Nonetheless, it is also characteristic of perception that we
perceive things as objects whose properties are not intrinsically perspectival
but are instead objective. It is, for example, immediately apparent to us that
things do not change shape and size just because we move in relation to them;
instead, their properties appear to us as causes of their perspectival appear-
ance, which changes as we move in relation to them and vice-versa—as
Merleau-Ponty acknowledges:

Our perception ends in objects, and the object, once constituted,

appears as the reason for all the experiences of it which we have had or
could have.
(PhP 69)
162  Thomas Baldwin
Thus although perception is subjective, it presents the perceived world as
objective. This yields the common sense position that while there is an ines-
capable subjective aspect to experience, and thus to our evidence for our
beliefs, the world we perceive, and as we believe it to be, is objective. We
can provide subjective descriptions of many features of the world; but these
descriptions draw upon relationships, typically spatio-temporal, between
ourselves and these features which are not essential to these features. So
whereas what is subjective has priority in the order of knowledge,17 the
objective conception of the world has priority in the order of being.
This common sense position is too simple: self-knowledge, for example,
is knowledge of facts that are themselves essentially subjective; and the old
debates about whether colours are ‘secondary’ qualities suggest that some
features of the perceived world are neither straightforwardly subjective nor
fully objective, but have an intermediate ‘anthropocentric’ status.18 It is not,
however, necessary to attempt to present here a full account of these mat-
ters; instead, switching attention back to Merleau-Ponty, the point to grasp
is that he holds that the objective conception of the world, which he associ-
ates with the natural sciences as well as common sense, is inherently deriva-
tive and incomplete. Thus at the start of Phenomenology of Perception, in
the ‘Preface’, he writes:

The entire universe of science is constructed upon the lived world, and if
we wish to think science rigorously, to appreciate precisely its sense and
its scope, we must first awaken that experience of the world of which
science is the secondary expression.
(PhP lxxii)

A key stage in ‘awakening’ this experience comes from a critical examina-

tion of the objective conception of the body:

Let us, then, consider objective thought at work in the constitution of

our body as an object, since this is a decisive moment in the genesis of
the objective world. We will see that in science itself, one’s own body
evades the treatment which it wanted to impose upon it. And since the
genesis of the objective body is but a moment in the constitution of the
object, the body, by withdrawing from the objective world, will carry
with it the intentional threads that unite it to its surroundings . . . .
(PhP 74)

The conclusion, then, is that just as the objective conception of the body is
only ‘an impoverished image’ of the body as we experience it (PhP 456), the
objective conception of the world is inherently ‘derivative’ (PhP 393). Hence:

The world, in the full sense of the word, is not an object; it has an enve-
lope of objective determinations, but also fissures and lacunae through
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  163
which subjectivities become lodged in it, or rather which are these sub-
jectivities themselves.
(PhP 349)

To support this thesis, Merleau-Ponty provides many rich and illumi-

nating considerations which I have no space to examine here.19 But as he
develops his position, it becomes increasingly clear that what motivates his
commitment to it is a distinctive version of transcendental idealism whose
subject is not the abstract non-empirical self of classical rationalism but is
instead the engaged, embodied agent whose perceptions and movements
give meaning (sens) to the perceived world, especially its spatio-temporal
form. Thus, to repeat a passage quoted earlier:

The subject is being-in-the-world and the world remains ‘subjective’,

since its texture and its articulations are sketched out by the subject’s
movement of transcendence.
(PhP 454)

In developing this position, Merleau-Ponty emphasises that it differs from

that presented by Husserl in writings such as Ideas I in which Husserl writes
of ‘a sense-bestowing consciousness which, for its part, exists absolutely’
(Husserl 1982: 129). Husserl’s position gives primacy to predicative acts
which ‘bestow’ meaning on experience, whereas the conclusion of Merleau-
Ponty’s long discussion of motricity, the capacity for bodily movement and
agency, is that:

The experience of the body leads us to recognize an imposition of sense

that does not come from a universal constituting consciousness, a sense
that adheres to certain contents. My body is this meaningful core that
behaves as a general function and that nevertheless exists and that is
susceptible to illness.
(PhP 148)

Thus it is through our embodiment that sense is ‘imposed’ on things in

such a way that ‘we are condemned to meaning’ (PhP 1xxxiv).20 One might
wonder what the difference is supposed to be between the ‘bestowal’ (don-
ner) of sense and the ‘imposition’ (imposition) of sense? In the former case
Merleau-Ponty almost always just uses Husserl’s term Sinngebung, indicat-
ing that he has in mind Husserl’s conception in Ideas I, which Merleau-
Ponty takes to be a predicative act. By contrast Merleau-Ponty’s conception
of the imposition of sense is of something which informs pre-predicative
experience. So the ‘bestowal’/‘imposition’ distinction is a marker for the
predicative/pre-predicative distinction. Indeed, looking back to Merleau-
Ponty’s thesis of priority of pre-predicative experience, one can see that it is
the fundamental sense-imposing role of the body which implies the priority
164  Thomas Baldwin
of this type of experience since its pre-predicativity expresses the way in
which our embodiment brings meaning to our lives.
One might still wonder whether this aspect of Merleau-Ponty’s position
is correctly regarded as a form of transcendental idealism, since he himself
describes his position as one which stands between the alternatives of real-
ism and idealism (PhP 454). But this description takes it that ‘idealism’ is
the view that the world is ‘the visible unfolding of a constituting thought’
(PhP 454), the kind of idealism exemplified by Husserl’s sense-bestowing
consciousness which Merleau-Ponty explicitly repudiates (PP lxxv). None-
theless, as the passage quoted above shows, he affirms a bodily ‘imposition
of sense’, which is fundamentally idealist in conception; and he describes
this imposition of sense as ‘transcendental’ in the following passage:

Along with the natural world and the social world, we have discovered
that which is truly transcendental, which is not the collection of consti-
tutive operations through which a transparent world, without shadows
and without opacity, is spread out in front of an impartial spectator, but
rather the ambiguous life where the Ursprung of transcendences takes
place . . . .
(PhP 382)

This ‘ambiguous life where the Ursprung of transcendences takes place’ is

engaged perception, perception which is informed by a bodily imposition of
sense. Its transcendental status is reflected in a memorable passage in which
Merleau-Ponty invokes Valéry’s poetry21 to characterise his conception of

All knowledge is established within the horizons opened up by percep-

tion. Since perception is the ‘flaw’ in this ‘great diamond’, there can be
no question of describing it as one of the facts that happen in the world,
for the picture of the world will always include this lacuna that we are
and by which the world itself comes into existence.
(PhP 215)

It is the status of perception, as the Ursprung ‘by which the world comes
into existence, which implies that ‘there can be no question of describing it
[perception] as one of the facts that happens in the world’—though, more
accurately, Merleau-Ponty’s claim should be that ‘there can be no question
of describing it [perception] as just one of the facts that happens in the
world’ for it is central to his conception of the ‘ambiguous’ status of per-
ception that it is ‘simultaneously creating (naturant) and created (naturé)’.
The message here is the same as that which we encountered earlier, in the
conclusion of The Structure of Behavior, ‘that the universe of naturalism
has not been able to become self-enclosed, and that perception is not an
event of nature’ (SB 145). Furthermore, it is not only perception’s bodily
‘imposition of sense’ that is transcendental; the ‘sense’ thus imposed has
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  165
the same status since it carries an essential reference back to the bodily
subject. Indeed, it is this sense which constitutes the ‘fissures and lacunae
through which subjectivities become lodged in [the world], or rather which
are these subjectivities themselves’ (PhP 349). For Merleau-Ponty, the most
important aspect of these subjectivities is time since ‘The ideal of objective
thought is simultaneously grounded upon and left in ruins by temporality’
(PhP 349), a comment which anticipates the conclusion of his chapter on
‘Temporality’, that ‘time is neither a real process nor an actual succession
that I could limit myself simply to recording. It is born of my relation with
things’ (PhP 434). A final aspect of this transcendental idealism concerns
causation. I mentioned earlier in connection with Merleau-Ponty’s interest-
ing conception of the ‘motives’ which connect and unify sensible phenomena
that he holds that this type of connection is not causal, because it concerns
the way in which diverse phenomena are connected as features of a unified
scene which expresses ‘the sense it offers’ (PhP 51). So again, some transcen-
dental ‘imposition of sense’ is conceived to be at work in the ‘motivated’
connection of phenomena, and Merleau-Ponty takes this to exclude a causal
account of this connection. While this line of thought is understandable, it
is worth noting that it is being assumed here that causation itself does not
belong among the types of sense that are imposed by perception. If the con-
trary assumption were made—and one can think of Kant’s transcendental
idealism here—it might well appear legitimate, on the contrary, to regard
Merleau-Ponty’s motivated connections as a type of causal connection that
is appropriate for sensible phenomena.
At this point I should add that just as Kant combined his transcen-
dental idealism with empirical realism (indeed the former is supposed to
provide a warrant for the latter), Merleau-Ponty also combines his ver-
sion of transcendental idealism with his own form of empirical realism,
despite the fact that he rejects the ‘realism’ of those who hold that, for
example, perception is just ‘one of the facts that happen in the world’.
Merleau-Ponty certainly holds that the objects of perception are typically
real objects with real properties; it is only those aspects of the perceived
world which manifest the body’s imposition of sense, such as spatial ori-
entation and temporal order, which are essentially subjective and thus not
transcendentally real. What is nonetheless distinctive of Merleau-Ponty’s
empirical realism is his insistence that empirical reality is constrained by
the Berkeley-inspired condition that ‘we cannot ultimately conceive of a
thing that could be neither perceived nor perceptible’ (PhP 334), which
leads him to the notorious claim that ‘Nothing will ever lead me to under-
stand what a nebula, which could not be seen by anyone, might be’ (PhP
456). The resulting position has some similarities with the ‘anti-realism’
of those who hold that there cannot be truths, for example concerning
the past, whose truth transcends all possible verification because all the
relevant evidence has been destroyed.22 But in both cases the position looks
to be based on a misguided belief that the limits of perception (or evidence)
are the limits of the world.
166  Thomas Baldwin
To pursue this issue further, however, would lead one from epistemol-
ogy into metaphysics and thus beyond the limits of this paper. Returning
instead to transcendental idealism, one familiar anxiety concerns the appar-
ent implication that for each subject there is their own ‘subjective’ world,
essentially dependent on relationships with that subject. ‘Metaphysical’ ver-
sions of this position respond to this anxiety by holding that in some sense
there is just one ultimate subject.23 That option is plainly not available to
­Merleau-Ponty. Instead his response starts from the thesis that although
there are indefinitely many embodied subjects of perception, the way in
which perception imposes sense is essentially the same for everyone. Per-
ception, he writes, ‘takes place within an atmosphere of generality and is
presented to us as anonymous’ (PhP 223), so that ‘if I wanted to express
perceptual experience with precision, I would have to say that one per-
ceives in me, not that I perceive’ (PhP 223). Thus, he suggests, although the
sense (e.g., the visual perspective) imposed by each subject is not the same,
because these senses are all of the same type they can be readily harmonised.
Just as one person’s changing perceptions over time can be integrated into
an extended experience of one changing situation, the perceptions of differ-
ent people can be shared in such a way that their senses are harmonised as
perceptions of just one world. The crucial assumption Merleau-Ponty intro-
duces here is that the experiences of different people can be harmonised in
this way; this is the intersubjectivity of experience. Whereas the subjectivity
of experience implies that its content is essentially subject-related, its inter-
subjectivity implies that this content can be fitted alongside that of other
subjects. Of course, differences and disagreements remain possible; but
the intersubjectivity of experience implies that it is these differences which
require explanation, and not the normal expectation of agreement. Hence,
Merleau-Ponty concludes, the unity and uniqueness of the world is both
sustained and vindicated by intersubjectivity:

The phenomenological world is not pure being, but rather the sense that
shines forth at the intersection of my experiences and at the intersection
of my experiences with those of others through a sort of gearing into
each other. The phenomenological world is thus inseparable from sub-
jectivity and intersubjectivity, which establishes their unity through the
taking up of my past experiences into my present experiences, or of the
other person’s experience into my own.
(PhP lxxxiv)

Merleau-Ponty discusses intersubjectivity at length in the chapter on

‘Others and the Social World’ in part II of Phenomenology of Perception.
But one of the odd features of the book is that he does not here confront
a challenge for which he has himself prepared the ground by his earlier
discussions of the famous case of Johann Schneider, the former German sol-
dier with brain damage whose handicaps had been studied and described in
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  167
detail by Kurt Goldstein and Adhémar Gelb. Merleau-Ponty proposes that
‘all of Schneider’s disorders can be reduced to a unity’ (PhP 137) which he
describes in the following famous passage:

Let us say . . . that the life of consciousness—epistemic life, the life of

desire, or perceptual life—is underpinned by an ‘intentional arc’ that
projects around us our past, our future, our human milieu, our physical
situation, our ideological situation, and our moral situation, or rather,
that ensures that we are situated within all these relationships. This
intentional arc creates the unity of the senses, the unity of the senses
with intelligence, and the unity of sensibility and motricity. And this is
what ‘goes limp’ in the disorder.
(PhP 137)

Merleau-Ponty does not state that this ‘intentional arc’ is the fundamental
pre-predicative operative intentionality discussed earlier. But this identifica-
tion fits with its role as ‘underpinning’ the life of consciousness, such as
‘epistemic life’. One might then wonder how ‘our ideological situation, and
our moral situation’ can be altogether pre-predicative; but an implication
of the discussion of habits mentioned above is that the body has the capac-
ity to assimilate new meanings (PhP 148), and it is a familiar fact that we
are prone to assimilate the ideological and moral attitudes we encounter in
childhood in such a way that we ‘project’ them unthinkingly along with the
more fundamental assumptions which provide for the unity of the senses. So
it looks plausible to take it that this intentional arc is an aspect of the fun-
damental operative intentionality, and I shall assume that this identification
is correct, although this point is not crucial to the issue I want to discuss.
This issue concerns the implications of Schneider’s profound handicaps, and
in particular the question of whether the normal assumption of intersubjec-
tivity does not apply in his case. The case for an affirmative answer is that
once the ‘intentional arc’ which underpins Schneider’s ‘personal core and
his power of existing’ (PhP 136) has become severely attenuated, with the
result that his capacities, interests, and concerns (as described by Gelb and
Goldstein) are very different from those of a normal person, it is not appro-
priate to speak of ‘the intersection of [his] experiences with those of others
through a sort of gearing into each other’ or of our ‘taking up . . . [his]
experience into [our] own’. Yet these are the marks of intersubjectivity. But
if intersubjectivity fails, the uniqueness implied by talk of ‘the intersubjec-
tive phenomenological world’ is not warranted, and instead a pluralism of
subjective phenomenological worlds will have to be accepted.
The issue that arises here is reminiscent of that discussed earlier in con-
nection with the hypothesis of epistemic pluralism in On Certainty, arising
from the possibility of radically different systems of belief with different
hinge propositions, and the question of there being a mark of the objectiv-
ity of any one of them. But for the moment, I shall postpone a comparison
168  Thomas Baldwin
between the ways in which Wittgenstein and Merleau-Ponty deal with this
issue in order to concentrate on Merleau-Ponty’s treatment of it. In thinking
about the challenge to Merleau-Ponty, one should recognise that the issue
for him is not primarily one of our not having grounds for belief in an objec-
tive world, given his view that the objective world is inherently derivative
and incomplete. It is rather whether it is legitimate for him to write of ‘the
world’ (i.e., the intersubjective phenomenological world) ‘as the primordial
unity of all our experiences on the horizon of our life and as the unique
term of all of our projects’ and thus as ‘the homeland of all rationality’ (PhP
454). For if there is a pluralism of worlds which cannot be harmonised
by the assumption of intersubjectivity, then it looks as though there is, for
example, no single ‘homeland of all rationality’, in which case there is here
a significant skeptical challenge.
Insofar as Merleau-Ponty’s discussion of Schneider suggests a response to
this challenge, it is that the account given by Gelb and Goldstein of Schnei-
der’s situation implies that his world is a truncated version of ours, of the
world of a normal human adult; indeed, Merleau-Ponty reports, this is also
how things seem to him: ‘Schneider would still like to form political and
religious opinions, but he knows that it is useless to try’ (PhP 136). Hence
even if the assumption of intersubjectivity is not warranted and Schneider’s
world is different from ours, this pluralist conclusion does not after all pose
a significant skeptical threat; for if Schneider’s world is just a cut down ver-
sion of our world, then our world can remain the ‘homeland of all rational-
ity’ since Schneider’s world does not include a serious alternative system of
rationality. The merits of this response obviously depend on whether the
account provided by Gelb and Goldstein of Schneider’s situation is satisfac-
tory, and this is not a matter that can now be assessed, though we have now
learnt to be much more sensitive and receptive to the voices of the disabled
than used to be the case. A similar point arises in the context of Merleau-
Ponty’s discussion of a schizophrenic subject who suffers hallucinations.
Merleau-Ponty describes the situation as one in which the subject’s life has
been disrupted by the intrusion of deceptive hallucinations which are not
perceptions but masquerade as such within the subject’s consciousness (PhP
351), and his diagnosis of this case is that the ‘faith’24 which normally ‘car-
ries us beyond subjectivity’ and ‘places us in the world’ has been truncated
by the hallucinations, so that the subject ‘becomes bogged down in [their]
private appearances’ (PhP 358–9). As a result, it seems clear that the normal
assumption of intersubjectivity has broken down—we cannot incorporate
‘[his] experience into [our] own’. But as with the discussion of Schneider,
Merleau-Ponty’s discussion does not suggest that the fact that the life-world
of the schizophrenic is very different from that of a healthy adult poses a
significant skeptical challenge to the assumption that the latter’s world is
‘the homeland of all rationality’.
This assumption is one that subsequent discussions of schizophrenia have
called into question, and Merleau-Ponty’s discussion in Phenomenology of
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  169
Perception of the challenge posed by those who are labelled ‘schizophrenic’
now appears rather imperceptive.25 This impression is reinforced by the
position he takes in some popular radio talks which he gave in 1948 and
which have been published as The World of Perception. On the one hand,
Merleau-Ponty argues that we need to recognise the different worlds of
‘children, primitive people, the sick, or more so still, animals’ (WP 72);
but on the other hand, he claims that these worlds lack coherence, whereas
the world of the ‘healthy, civilised, adult human being strives for coher-
ence’ (WP 72). Hence ‘Adult thought, normal and civilized, is better than
childish, morbid or barbaric thought’ (WP 73)—so again, the very idea of
a skeptical challenge to it is banished by an assertion of its superiority. It
is hard not to be disappointed by Merleau-Ponty’s discussion of this issue.
He simply fails to consider that where intersubjectivity fails and there is a
plurality of worlds, there is a potential for a skeptical challenge to the pre-
eminence of the world of those who are described as ‘healthy, civilised, adult
human being[s]’. Indeed, it was precisely the insight of one of his early crit-
ics, the young Michel Foucault, that the history of social responses to ‘folly’
shows that there is a serious challenge here that needs to be taken seriously
(as he demonstrated in his masterpiece Folie et Déraison—Foucault 1961).
At this point I return to Wittgenstein and his discussion of these issues in
On Certainty. Two points can be made immediately: first, as indicated ear-
lier, Wittgenstein did recognise the skeptical challenge that arises from epis-
temic pluralism (cf. OC §336); second, there is a marked difference in the
tone of the two discussions. Merleau-Ponty’s remark that ‘Adult thought,
normal and civilized, is better than childish, morbid or barbaric thought’
is very alien to Wittgenstein’s non-judgmental use of examples from other

Men have believed that they could make rain; why should not a king be
brought up in the belief that the world began with him? And if Moore
and this king were to meet and discuss, could Moore really prove his
belief to be the right one? I do not say that Moore could not convert the
king to his view, but it would a conversion of a special kind; the king
would be brought to look at the world in a different way.
(OC §92)

The more profound point, however, is whereas Merleau-Ponty takes it that

he has shown that there is one coherent, intersubjective, phenomenological
world which normal adults share and which is the homeland of all rational-
ity (PhP 454), Wittgenstein does not think that anything of this kind can,
or indeed needs to, be shown. That is, to put the point in his terms, we
should not seek to show that our system of beliefs is ‘the homeland of all
rationality’. Any reasons that we can make sense of already have a place
within our system of beliefs since it is ‘the element in which arguments have
their life’ (OC §105). So once we have reflectively satisfied ourselves about
170  Thomas Baldwin
our beliefs and our reasons for them, there is nothing more we can do to
justify them. Instead we must learn to live with the ‘groundlessness of our
believing’ (OC §166). As a consequence, the reasons we have are inherently
non-exclusive; they do not exclude there being other systems of belief with
reasons of which we can make no sense as reasons. So even though we try to
satisfy ourselves that, so far as we can tell, our beliefs are rational, we have
no reason for holding that our system of beliefs is all-inclusive, that it is the
homeland of all rationality.
Having set out this comparison of the responses of Wittgenstein and
Merleau-Ponty to the skeptical challenge arising from epistemic pluralism,
I turn finally, and briefly, to some more speculative comparisons concerning
the positions I have attributed to Merleau-Ponty concerning the imposition
of sense and transcendental idealism. I have suggested here that Merleau-
Ponty’s thesis of the priority of pre-predicative knowledge as compared
with predicative knowledge derives from his account of the fundamental
pre-predicative imposition of sense that informs perception and action. For
Wittgenstein, by contrast, meaning arises in the context of the language-
games within which people engage with each other in a great variety of
activities. So for Wittgenstein meaning is typically an ingredient of pred-
icative acts—questions, answers, statements, and commands. Hence while
there is no reason for Wittgenstein to deny that we have the kind of non-
thetic pre-predicative practical understanding which Merleau-Ponty empha-
sizes, there is no reason for him to give it any special priority as compared
with our familiar predicative knowledge of the world, both subjective and
objective; and there is equally no reason for him to prioritise subjective
descriptions of the world as compared with objective ones, except insofar
as the Moorean certainties which comprise the core of our systems of belief
are typically subjective.
Finally, there is the issue of transcendental idealism. I have argued that
Merleau-Ponty presents a novel form of transcendental idealism, founded
upon the role of perception in giving form to the perceived world. Can
one say anything similar about Wittgenstein’s position? Does the funda-
mental role of language-games in the constitution of meaning bring with it
a form of ‘linguistic idealism’? Does the possibility of language impose form
on what can be said? In the Tractatus Wittgenstein had maintained that
a proposition (Satz) represents a possible situation in virtue of the fact it
shares the logical form of that which it represents; and he held that because
logic is in this way a mirror-image (Spiegelbild) of the world, it is ‘tran-
scendental’ (TLP §6.13). So there is here a kind of transcendental idealism,
though Wittgenstein emphasizes logic is a method of inference and not a
system of fundamental truths. But I take it to be a feature of the change
in his position in Philosophical Investigations, with the new emphasis on
the multiplicity of language-games which involve a variety of speech acts
of which representation (Darstellung) is but one (PI I §§ 23–4), that he
does not seek to exhibit some definite system of analysis or inference as a
transcendental condition of the possibility of language-games in general.26
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  171
He does of course emphasize that language-games involve practices which
sustain the rules that guide the language and that these practices require
general agreement in judgment. But although he argues that these require-
ments exclude a language whose rules are purely ‘private’, nothing specific
is prescribed concerning the general form of language. So there is here no
trace of transcendental idealism.
It is, therefore, no cause for concern that in On Certainty Wittgenstein
just treats idealism as a form of skepticism (OC §37). But what is more
important is the difference between Wittgenstein and Merleau-Ponty con-
cerning the role for any reference to what is ‘transcendental’ in epistemology
(and philosophy in general). I discussed earlier the emphasis on naturalism
in Wittgenstein’s later writings, which is exemplified by his statement in
On Certainty ‘I want to regard man here as an animal (Tier)’ (OC §475).
Although the significance of talk of what is ‘natural’ (or ‘animal’) is always
contestable, the terms ‘natural’ and ‘transcendental’ are generally used as
contraries, and this certainly applies to Wittgenstein’s later writings. Thus,
Wittgenstein’s later naturalism stands in sharp contrast to Merleau-Ponty’s
claim to have discovered that which is ‘truly transcendental’ (PhP 382).
Admittedly Merleau-Ponty then mitigates and complicates his claim by
maintaining the ‘ambiguous life’ which is truly transcendental is ‘ambigu-
ous’ because it combines what is naturé (formed by nature) with what is
naturans (gives form to nature) (PhP 383), which suggests that he does not
take it that what is transcendental is not also, in some sense, natural. But it
is the claim that because perception gives form to nature it is not ‘an event
of nature’ (SB 145) or ‘one of the facts that happens in the world’ (PhP
215), that is alien to Wittgenstein’s later philosophy. It manifests the wish
for a fundamental philosophical explanation of facts, when, for Wittgen-
stein, ‘Philosophy simply puts everything before us, and neither explains nor
deduces anything’ (PI I §126).

1 Wittgenstein is probably alluding here to a passage in Moore’s ‘Proof of an
External World’; see Moore (1993: 166). I do not myself think that Moore is
here offering a response to skeptical arguments, but it is not important in this
context whether or not this is correct.
2 However, one might add here that in fact many of the considerations which
­Merleau-Ponty advances arise from comparisons between the everyday experi-
ences which the reader is assumed to enjoy and the abnormal experiences of those
with serious handicaps (a phantom limb, brain damage, aphasia, etc.). So there is
a comparative aspect to Merleau-Ponty’s method of inquiry too which is similar
in this respect to Wittgenstein’s method of inquiry, though Merleau-Ponty does
not explain how his references to the abnormal experiences of others ‘loosen the
intentional threads that connect us to the world in order to make them appear’.
3 The editors divide this part of the text into two parts (§§66–192, §§193–299),
but there is no obvious reason for this division since it just reflects the fact that
one notebook ends with §192 and the next one starts off with §193 (see the
‘Addendum to “The Wittgenstein Papers” ’ (PO 509)). However, the dates
172  Thomas Baldwin
attached to notes §287 onward show that there was a six-month gap between
§299 and §300, so it is right to treat §§300–676 (which is an uninterrupted
sequence of notes) as a separate part.
4 There are a few brief remarks about certainty at the end of part II xi of Philo-
sophical Investigations (224–8). These remarks are taken from notes which Witt-
genstein wrote in Dublin in 1949 before going to the United States, and which
have been published as his Last Writings on the Philosophy of Psychology vol-
ume I; see esp. §§ 882–92. But these comments primarily concern self-knowledge
and thus do not directly anticipate the main themes of On Certainty.
5 Wittgenstein’s remark that: ‘Everything that I have seen or heard gives me the
conviction that no man has ever been far from the earth. Nothing in my picture
of the world speaks in favour of the opposite’ (OC §93) exemplifies one type of
case in which a hinge proposition has to be abandoned. What Wittgenstein wrote
in 1950 was fair enough then, but it is no longer true for us, given the success of
the Apollo programme in 1969.
6 Wittgenstein’s phrase here, ‘objectives Merkmal’, might also be translated as
‘mark of objectivity’, which is not quite what is suggested by the translation
‘objective character’.
7 There is striking reversal in Wittgenstein’s metaphors here: the fact that, as he
puts it in Philosophical Investigations, one reaches ‘bedrock’ in the justifications
for one’s beliefs becomes, in the idiom of On Certainty, the fact that these beliefs
are ‘groundless’. ‘Bedrock’ is not, so to speak, a ‘ground’.
8 This is an odd translation of the French word ‘Raisonnement’ (reasoning) which
Wittgenstein uses.
9 Wittgenstein is here quoting the famous line from Goethe’s Faust: ‘Im Amfang
war der Tat’, which he also quotes in On Certainty §402.
10 It may be asked whether my concentration here on Phenomenology of Percep-
tion is appropriate; in particular whether one should not also take into account
some of Merleau-Ponty’s later texts, such as The Visible and the Invisible. My
main reason for not giving these texts much attention here is that they do not
include significant discussions of the topic of this paper. At the time of his death
Merleau-Ponty was working towards new accounts of perception and language,
such as those sketched in Chapter 4 ‘The Intertwining—The Chiasm’ of The Vis-
ible and the Invisible. I believe that these accounts do have some implications for
debates about certainty and knowledge; but these implications are not explicit in
these late texts, and it would require a good deal of interpretive work to identify
them with confidence. So although that is a task worth undertaking, it is one for
a separate paper.
11 The square bracket indicates that the text which follows was inserted by
­Merleau-Ponty as a comment.
12 One might also ask whether Wittgenstein had any familiarity with Husserl’s
writings which would provide an indirect point of contact between Wittgenstein
and Merleau-Ponty. So far as I am aware, there are no references to Husserl in
Wittgenstein’s writings; but after he returned to the study of philosophy in 1929,
he wrote for a time about ‘phenomenology’, as in the following remarks from the
start of his 1930 Philosophical Remarks:

A recognition of what is essential and what inessential in our language if

it is to represent, a recognition of which parts of our language are wheels
turning idly, amounts to the construction of a phenomenological language.
Physics differs from phenomenology in that it is concerned to establish
laws. Phenomenology only establishes the possibilities. Thus, phenomenol-
ogy would be the grammar of the description of those facts on which phys-
ics builds its theories.
(PR 51)
Wittengenstein and Merleau-Ponty on Knowledge and Certainty  173
The references here to phenomenology are striking; but are they to phenom-
enology as conceived by Husserl? Husserl certainly intended to identify ‘what is
essential and what inessential’ (see, for example, the opening pages of Ideas I)—
but not, on the face of it, by means of an inquiry into ‘what is essential and what
inessential in our language if it is to represent’. So even if Wittgenstein had some
familiarity with Husserl’s phenomenology, perhaps through discussions with
Moritz Schlick who was familiar with (though very critical of) some of Husserl’s
work, Wittgenstein does not appear to have been interested in the characteristic
phenomenological method of inquiry. Instead for a time he seems to have used
the term ‘phenomenology’ to describe an inquiry into the conditions of linguistic
representation which would avoid the problems arising from his exclusive reli-
ance in the Tractatus on logical analysis (hence the reference to ‘grammar’, and
not ‘logic’, in the second passage above).
13 There is some overlap between the topics which Merleau-Ponty deals with as cases
of ‘motivation’ and the points Wittgenstein discusses as ‘seeing an aspect’ in Philo-
sophical Investigations part II xi. But Wittgenstein seems more concerned to char-
acterise the ways in which seeing is a ‘tangled’ matter (PI II 200e) than to develop
anything comparable to Merleau-Ponty’s concept of phenomenological motivation.
14 Russell suggests (1912: 70) that this distinction is manifested in French by
the distinction between ‘savoir’ and ‘connâitre’. But Merleau-Ponty’s use of these
verbs does not follow this rule: he uses ‘savoir’ for both the kinds of knowledge
of the body mentioned above (PhP 213), and elsewhere uses ‘connâitre’ both
for objective, thetic, knowledge of the world (PhP lxxxii) and for the non-thetic
originary knowledge of things inherent in perception (PhP 45). Indeed, if there
is any significance to the ways in which Merleau-Ponty uses ‘savoir’ and ‘con-
nâitre’, I have not been able to identify it.
15 This remark comes from Remarks on the Foundations of Mathematics, part VI,
which is dated ‘c. 1943/1944’.
16 One might suggest that there is another strand of Wittgenstein’s rule-following
discussion which offers a potential connection with Merleau-Ponty’s position,
namely the thesis cited earlier that ‘When I obey a rule, I do not choose. I obey
the rule blindly’ (PI I §219). For, one might propose, ‘blind’ rule-following can be
understood to involve a pre-predicative understanding of the world that grounds
the exercise of predicative judgment. This proposal, however, rests on a misun-
derstanding: Wittgenstein’s thesis concerns judgment; ‘blind’ rule-following is
making a judgment in a situation in which there is no conclusive guidance to be
followed (and the generality of the thesis comes from the claim that all situations
are of this kind). So there is nothing pre-predicative about ‘blind’ rule-following.
17 It is notable that Moore’s common-sense truisms are subjective in this sense.
18 cf. D. Wiggins ‘Truth, Invention, and the Meaning of Life’ in Wiggins (1987);
esp. 107.
19 I examine this aspect of Merleau-Ponty’s position in Baldwin (2013).
20 I revert here to Colin Smith’s translation of ‘sens’ as ‘meaning’, since Landes’s
translation ‘condemned to sense’ makes little sense.
21 ‘Mes repentirs, mes doutes, mes contraintes/ Sont le défaut de ton grand dia-
mant’ from ‘Le cimetière marin’ by Paul Valéry.
22 cf. M. Dummett ‘The Reality of the Past’, in Dummett (1978); esp. 362–3.
23 A classic instance is T. H. Green’s single ‘eternally complete consciousness’; see
Green (1890: 72ff).
24 It is here that he alludes in a note to Husserl’s Urdoxa, which I discussed earlier.
25 The classic text here is Laing and Esterson (1964).
26 It is worth noting that the passages quoted in note 11 from Wittgenstein’s 1930
Philosophical Remarks start out with precisely the question of recognising ‘what
is essential and what inessential in our language if it is to represent’. It is that
question which he come to reject as misconceived during the 1930s.
174  Thomas Baldwin
Baldwin, T. 2013. Merleau-Ponty’s phenomenological critique of natural science.
In H. Carel and D. Meacham (eds.), Phenomenology and Naturalism (Royal
Institute of Philosophy supplement 72). Cambridge: Cambridge University Press,
pp. 189–219.
Dummett, M. 1978. Truth and other Enigmas. London: Duckworth.
Foucault, M. 1961. Folie et Déraison. Paris: Gallimard.
Green, T. H. 1890. Prolegomena to Ethics. Oxford: Clarendon Press.
Husserl, E. 1973. Experience and Judgment. Edited by L. Landgrebe, Translated by
J. Churchill and K. Ameriks. London: Routledge.
Husserl, E. 1982. Ideas I. Translated by F. Kersten. The Hague: Martinus Nijhoff.
Laing, R. D. and Esterson, A. 1964. Reason, Sanity and Madness. London: Penguin.
Malcolm, N. 1949. Defending common sense. Philosophical Review 58: 201–220.
Malcolm, N. 1977. Moore and Wittgenstein on the sense of ‘I know’. In his Thought
and Knowledge. Ithaca: Cornell University Press, pp. 170–179.
Moore, G. E. 1993. Selected Writings. Edited by T. Baldwin. London: Routledge.
Russell, B. 1912. The Problems of Philosophy. London: Williams and Norgate.
Russell, B. 1935. The limits of Empiricism. Proceedings of the Aristotelian Society
36: 131–150.
Wiggins, D. 1987. Needs, Values and Truth. Oxford: Wiley-Blackwell.

Thomas Baldwin is an emeritus professor of philosophy at the University

of York. From 2005 until 2015 he was editor of the UK’s leading phi-
losophy journal, Mind. Recent work includes forthcoming papers about
Russell on truth and on modality.
Chantal Bax is an NWO Veni researcher at Radboud University Nijme-
gen. In her dissertation, which was awarded the Praemium Erasmianum
Research Prize and has been published by Continuum under the title Sub-
jectivity after Wittgenstein, she placed Wittgenstein’s later work in the
context of the continental debate about the so-called death of man. She is
currently working on a project about Levinas, Nancy, and the rethinking
of community.
Taylor Carman is the author of Heidegger’s Analytic (2003) and Merleau-
Ponty (2008; 2nd edition forthcoming) and has coedited The Cambridge
Companion to Merleau-Ponty (2005). He has published articles on a
number of topics in phenomenology and is currently writing a book on
David R. Cerbone is professor of philosophy at West Virginia University. He
is the author of Understanding Phenomenology, Heidegger: A Guide for
the Perplexed, and Existentialism: All That Matters, as well as numerous
articles on Heidegger, Wittgenstein, and the phenomenological tradition.
Kathleen Lennon is a professor of philosophy at the University of Hull,
UK. Her recent publications include the 2015 monograph, Imagination
and the Imaginary (Routledge). She is also co-editor of the 2012 col-
lection  Embodied Selves (Palgrave). Kathleen has published numerous
articles on topics including: the imagination, the philosophy of embodi-
ment, phenomenology, gender theory, and old age.
Katherine J. Morris is a fellow in philosophy at Mansfield College, Oxford
University, UK. Her books include Descartes’ Dualism (with Gordon
Baker, Routledge, 1996), Sartre (Blackwell Great Minds series, 2008), and
Starting with Merleau-Ponty (Continuum Starting With series, 2012), and
176 Contributors
she has published widely on Descartes, Wittgenstein, Sartre, and Merleau-
Ponty. She also co-edits the series International Perspectives in Philosophy
and Psychiatry for Oxford University Press.
Søren Overgaard is associate professor of philosophy at the University of
Copenhagen. His main research topics are perception, social cognition,
and philosophical methodology. He is the author of Husserl and Heidegger
on Being in the World (2004) and Wittgenstein and Other Minds (2007),
co-author of An Introduction to Metaphilosophy (2013), and co-editor
of The Routledge Companion to Phenomenology (2011) and The Cam-
bridge Companion to Philosophical Methodology (forthcoming).
Komarine Romdenh-Romluc is a senior lecturer in philosophy at the Uni-
versity of Sheffield, UK. She is the author of the Routledge GuideBook
to Merleau-Ponty and the Phenomenology of Perception (2011). She has
also published articles on various aspects of Merleau-Ponty’s philosophy,
and other issues at the intersection of phenomenology and the philoso-
phy of mind.

ambiguity 76 – 7, 130 – 1, 136 epoché 139, 144; see also

analytical attitude 16, 18 – 20, 22 phenomenological reduction
anonymity 72, 75 expression 3 – 4, 7 – 9, 24, 31 – 46, 53 – 4,
art 4, 22, 41 – 5, 101 – 2, 104 – 12; 56 – 7, 59 – 61, 76, 84, 107 – 12, 115,
see also painting; pictures 127 – 30, 162
aspect-seeing see seeing-as
family resemblance 118 – 19, 127
background 2 – 5, 9, 13, 19 – 20, 22, flesh 101
33 – 4, 40 – 2, 44, 69 – 70, 72 – 3, 108, freedom 66 – 7, 71 – 6, 123, 131
117, 123 – 6, 143, 150
behaviour 1 – 2, 4, 8 – 9, 35 – 6, 49 – 51, Gallagher, Shaun 4 – 5, 49 – 57, 59, 61 – 2
53 – 4, 57 – 62, 72, 96 – 7, 109 – 12, Gelb, Adhémar, and Goldstein, Kurt
127, 129 167 – 8
behaviourism 12, 28, 63 geometry 40, 148
body, the 2 – 9, 17 – 20, 23, 26, 32 – 3, Gestalt 2 – 4, 11 – 27, 31, 33, 35 – 6, 39,
36, 38 – 9, 41 – 2, 45, 57, 68, 82, 42, 44, 46, 53; Gestalt psychology
85 – 6, 88 – 97, 101, 109 – 11, 123, 2 – 3, 11 – 27, 147
129 – 30, 139 – 40, 147, 152, 158, gesture 3 – 5, 15, 17, 21, 31 – 3, 35 – 40,
162 – 3, 165, 167; body schema 7, 28, 42 – 5
91, 94 – 7, 109 – 10; corporeality 71; Goodman, Nelson 104 – 5
embodiment 72, 91, 117, 143, 163 – 4
Brentano, Franz 132 – 3 habit 24 – 5, 95 – 6, 112, 157 – 8, 167;
see also skill
causation 3, 5 – 6, 18 – 19, 21, 23, 26, hands touching 92 – 3, 95
74, 141, 147, 151, 161 Heidegger, Martin 32, 41, 45, 66
certainties 5 – 6, 9, 68 – 72, 77 – 8, 137 – 9, hinge propositions 5, 9, 137 – 8, 143,
143, 151, 161, 170 149, 167; see also certainties
children 5, 10, 37, 51, 67, 69 – 70, 80, history 27, 53, 74 – 7, 111, 145 – 7, 149
104 – 5, 124, 169 horizon 74, 111, 123, 138, 142 – 3, 148,
communism 73 – 5, 77; class 164, 168
consciousness 6, 73 – 5 Husserl, Edmund 11, 76, 85, 123,
culture 26 – 7, 31, 34, 44, 59, 71, 104, 147 – 50, 155 – 6, 163 – 4
141, 169
indeterminacy 8 – 9, 77, 114 – 33, 154
Descartes, Réne 87, 100, 103 – 4 intentional arc 167

eidetic reduction 85 know-how 154

emotion 31 – 3, 37 – 8, 41, 49 – 50, 53, Köhler, Wolfgang 2 – 3, 11 – 12, 14 – 18,
55 – 61, 87, 94, 112, 127 21 – 2, 27
language 1 – 2, 7 – 9, 18, 24 – 5, 40, 60 – 1, practice 2, 31 – 2, 37, 40 – 1, 44 – 6, 66,
83, 88, 103, 105, 106 – 12, 114, 116, 70 – 1, 77 – 8, 84 – 5, 89, 106, 110,
118, 120, 127, 140, 145 – 6, 148 – 9, 112, 118 – 19, 147, 158 – 9, 171
153 – 60, 170 – 1; language-game pre-predicative 9, 150, 155 – 61, 163,
1, 71, 118, 127, 131, 139, 142, 167, 170
146, 149, 158 – 60, 170 – 1; private private language 68, 85
language 24, 68, 85, 171
Quine, W.V.O. 8, 114 – 15, 117, 132 – 3
Malcolm, Norman 141
Malraux, André 42, 107, 111 reciprocity 36
maximum grip 20 representation 7 – 8, 86, 88, 100 – 4, 106,
mind, nature of 2 – 8, 18, 31 – 2, 35, 53, 109 – 11, 156, 170
57, 60, 82, 110, 157; other minds rules, rule-following 8 – 9, 41, 44, 71,
5 – 6, 49 – 62, 68, 82; see also soul 103, 116, 118 – 23, 128 – 9, 145 – 6,
Moore, G. E. 68 – 9, 137 – 9, 141 – 4, 159 – 60, 171
149, 169 – 70
motives/motivation 6, 18 – 19, 21 – 3, 38, Sartre, Jean-Paul 73, 86, 105, 109
42, 145, 150 – 2, 155, 165 scepticism 68
Schilder, Paul 94 – 5
naturalism 10, 147, 164, 171 schizophrenia 7, 89 – 90, 97, 168
no-ownership 87 Schneider, Johann 22, 72 – 3, 166 – 8
science 27, 53, 70, 104, 115 – 16, 133,
objective thought 2, 25, 73, 116 – 17, 136 – 7, 146 – 7, 162
120, 152, 162, 165 see: duck-rabbit 3 – 4, 14; seeing-as 3,
operative intentionality 20, 150, 14 – 17, 22, 24, 34; sense-data 35,
152 – 61, 167 111; see also perception
sensation 12, 32, 49, 88, 92 – 3, 95,
pain 33 – 4, 36, 40, 60, 86, 125, 130 101, 129, 153
painting 21, 41, 44 – 5, 100 – 12; see also skill 96, 106, 110, 112, 157; see also
art; pictures habit
perception 3 – 5, 7 – 12, 17 – 23, 35, social 5, 51, 66 – 78, 109, 164, 166, 169
37 – 40, 43, 46, 49 – 54, 56, 60 – 2, 72, solipsism 6, 81 – 97
76, 85 – 6, 88, 93, 96, 100 – 1, 103 – 6, soul 2, 8, 23, 32, 37, 57, 59, 104, 128 – 30
108 – 12, 117, 120, 123, 132, 140, speech 24 – 5, 37 – 8, 40, 108, 110,
147, 150 – 5, 161 – 6, 168, 170 – 1; 114, 170
perceptual faith 9, 149 – 50, 155 style 27, 38, 42, 56, 108 – 12
perspective 6 – 8, 41, 74, 78, 82 – 3,
85 – 93, 97, 103 – 6, 110 – 12, thetic vs. non-thetic distinction 9,
123 – 4, 166 18 – 19, 21, 140, 152 – 60, 170
phenomenological reduction 147; time, experience of 8, 33 – 5, 94 – 5, 108,
see also epoché 123, 139, 165 – 6
pictures 7 – 8, 15, 21, 46, 100 – 1, transcendental idealism 9 – 10, 163 – 71
103 – 7, 112, 115 – 16
pluralism 143 – 4, 167 – 70 world, lived 9, 74, 82 – 5, 90, 95, 140,
politics 67, 72 154, 162