Hopkins R. - Picture, Image

job:LAY00 6-8-1998 page:1 colour:0
Picture, Image and Experience

A Philosophical Inquiry
How do pictures represent? In this book Robert Hopkins casts new light
on an ancient question by connecting it to issues in the philosophies of
mind and perception. He starts by describing several striking features of
picturing that demand explanation. These features strongly suggest that
our experience of pictures is central to the way they represent, and
Hopkins characterizes that experience as one of resemblance in a
particular respect. He deals convincingly with the objections traditionally
assumed to be fatal to resemblance views, and shows how his own
account is uniquely well placed to explain picturing’s key features. His
discussion engages in detail with issues concerning perception in
general, including how to describe phenomena that have long puzzled
philosophers and psychologists, and the book concludes with an attempt
to see what a proper understanding of picturing can tell us about that
deeply mysterious phenomenon, the visual imagination.
Robert Hopkins is lecturer in philosophy at the University of Birmingham.

He has published a number of articles in journals including Philosophical
Review, Mind and Philosophical Quarterly.
MMMM
Picture, Image and Experience

A Philosophical Inquiry
Robert Hopkins
pu b l i sh e d by th e p r e s s s y n d i c a t e o f t h e un i v e r s i t y o f ca m b r i d ge
The Pitt Building, Trumpington Street, Cambridge cb2 1rp, United Kingdom
c am b r id g e un i ve r si t y p r e ss
The Edinburgh Building, Cambridge, cb2 2ru, United Kingdom
http://www.cup.cam.ac.uk
40 West 20th Street, New York, ny 10011–4211, USA http://www.cup.org
10 Stamford Road, Oakleigh, Melbourne 3166, Australia
© Robert Hopkins 1998
This book is in copyright. Subject to statutory exception and to the provisions of

relevant collective licensing agreements, no reproduction of any part may take
place without the written permission of Cambridge University Press.
First published in 1998
Printed in the United Kingdom at the University Press, Cambridge
Typeset in Plantin 10/12pt [vn]
A catalogue record for this book is available from the British Library
Library of Congress cataloguing in publication data
Hopkins, Robert.
Picture, image and experience / Robert Hopkins.
p. cm.
Includes bibliographical references and index.
is b n 0 521 58259 8 (hardback)
1. Representation (Philosophy) 2. Image (Philosophy)
3. Perception. 4. Experience. 5. Aesthetics. i. Title.
b105.r4h67 1998
[email protected] – dc21 98–4543 cip
is b n 0521 58259 8 hardback

Contents
List of illustrations page vii

Acknowledgements ix
Introduction 1
1 The question 7
1 Introduction 7
2 Resemblance 9
3 Goodman 13
4 Experiential accounts 14
2 Some features to explain 23

1 The importance of explanation 23
2 Explananda 24
3 Some explanations 36
4 Schier’s view 40
5 The way ahead 48
3 Outline shape 50
1 Experience and existence 50
2 Focussing the second problem 51
3 Outline shape 53
4 Perceiving outline shape 59
5 The views of others 63
4 A theory of depiction 71
1 Standards of correctness 71
2 Perspective 77
3 Minimum content 81
4 Other resemblances 84
5 Other resemblance-based views 88
5 Misrepresentation 94
1 A problem 94
2 A misplaced modesty 97
3 Degrees and parts 100
4 The solution – part one 104
5 A further worry: suYciency 106
v
vi Contents
6 The solution – part two: experience 107
7 A digression – determining the third dimension 114
8 The limits on misrepresentation 117
6 Indeterminacy and interpretation 122

1 The problem 122
2 A new opening 124
3 Separation: one diYculty overcome 128
4 Interpreting pictures 130
5 Separation and interpretation 137
6 Separation and some unWnished explanations 142
7 Accommodating diversity: resemblance, convention and drawing in
perspective 147
7 Visualizing 159
1 A new topic 159
2 Sartre – two views rejected 160
3 Sartre’s view 164
4 A simple alternative 168
5 Perspective in touch and vision 172
6 A little on some other senses 181
7 Visualizing versus seeing 184
8 Peacocke’s view 187
9 Some modest conclusions 192
Bibliography 201
Index 204
Illustrations
1 Pablo Picasso, Guernica (© Succession Picasso/DACS 1998) page 8

2 Elisabeth Vigée-Lebrun, Countess Golovine (reproduced by
permission of the Barber Institute of Fine Arts, University of
Birmingham) 12
3 Simone Martini, St John the Evangelist (reproduced by
Birmingham) 32
4 Pyramid without tracing 54
5 Pyramid with tracing 56
6 Pieter Molijn, Landscape with a Huntsman (reproduced by
Birmingham) 58
7 Gerald Scarfe, Kim Il Blair (reproduced by permission of the
artist) 95
8 Thomas Rowlandson, Dressing for a Masquerade (reproduced
by permission of the Barber Institute of Fine Arts, University
of Birmingham) 100
9 Necker Cube (reproduced by permission of Richard Gregory) 102
10 Cube 116
11 Impossible Triangle, closed (reproduced by permission of
Richard Gregory) 118
12 Impossible Triangle, open (reproduced by permission of
Richard Gregory) 119
13 Stick Wgure 123
14 Jacopo Tintoretto, Seated Man seen from Above (reproduced by
Birmingham) 126
15 Pierre Bonnard, The Dolls’ Dinner Party (reproduced by
permission of the Barber Fine Arts Institute, University of
Birmingham) 127
16 Alfred Stevens, Study of a Nude Figure (reproduced by permission
of the Barber Institute of Fine Arts, University of Birmingham) 140
vii
viii List of illustrations
17 Tsan Tsu Tsao, Mountainous Landscape (reproduced by

Birmingham) 148
18 Jan Davidsz de Heem, Still Life with a Nautilus Cup
(reproduced by permission of the Barber Institute of Fine Arts,
University of Birmingham) 149
19 Fernand Léger, Composition with Fruit (reproduced by
Birmingham) 150
Acknowledgements
I am grateful to the following journals for permission to reproduce

material: Philosophical Review for parts of my paper ‘Explaining Depiction’
(Philosophical Review 104, 1995. Copyright 1995 Cornell University,
reprinted by permission of the publisher); Mind for parts of ‘Resemblance
and Misrepresentation’ (Mind vol. 103, no. 412, 1994, by permission of
Oxford University Press); Philosophical Quarterly for a section of ‘El
Greco’s Eyesight: Interpreting Pictures and the Psychology of Vision’
(Philosophical Quarterly vol. 47, no. 189, 1997). I am also grateful to the
Barber Institute of Fine Art, University of Birmingham, for its generous
help in Wnding and reproducing suitable illustrations; and to Churchill
College, Cambridge, for providing indispensable Wnancial support and a
stimulating working environment.
Many people have helped shape this book. Malcolm Budd, a doctoral
supervisor as insightful and conscientious as one could wish for, has
remained just as supportive in subsequent years. David Owens has read
almost all the material here, as well as much that, thanks to him, is not
included. His advice on matters tactical is worth almost as much as his
criticism on issues of substance. Barrie Falk has helped enrich my sense of
the legitimate concerns of philosophical inquiry, thereby opening the door
to some of the questions this book addresses. Many others there is only
space to list. They include Josep Corbi, Nick Dent, Christel Fricke, Chris
Hookway, Dom Lopes, Colin Lyas, Greg McCulloch, Penelope Mackie,
Derek Matravers, Mike Martin, Alex Miller, Paul Noordhof, Tom Pink,
Michael Tanner, Ken Walton, Joss Walker and José Zalabardo. I thank
my friends and family for their constant support in a project rather distant
from their concerns. Lastly I thank Marion Thain, for sustenance both
intellectual and otherwise.
ix
For Marion
1 The question
1 Introduction
How do pictures represent? Consider for a moment Picasso’s Guernica
(Wgure 1). It is a large Xat surface covered with paint, broken by diVering
areas of colour into angular shards of blue-grey, brown and black. This
surface has a history, having been painted in 1937 in response to the
German bombing of the Basque town; and a location, now residing, after
many years of controversy, in Madrid. In these respects – having a
distinctive composition, history and location – it is like many objects of the
sort which Wll our world. Unlike many others, however, the painting also
represents. It is a powerful evocation of a terrible event. It shows a scene of
suVering, carnage and mutilation, scattered with the bodies of man and
beast alike. Thus there are two aspects to the picture’s nature. It is on the
one hand a material object, on the other a representation. One way to
present our problem is to ask how one thing can Wll both these roles. How
can a paint-covered surface represent other objects and scenes at all?
We can ask a parallel question about language. A written description of
a scene is on the one hand a set of marks on a surface, and on the other a
representation of absent objects and events. Here too we may wonder how
one item is able to play both roles. Yet there is a diVerence between the
two cases. We may begin to suspect this if we note some obvious contrasts
between examples of the two. A written description has to be read in a
certain order, but the eye is free to roam over a picture without confusing
the viewer. The colour of the marks rarely matters to what a description
says, but often aVects what a picture represents. The relative location of
diVerent bits of the picture dictates the spatial relations between the
objects they stand for, but the same does not seem true for the words in the
description, not at least in any very direct way. And one might go on.
These diVerences suggest that pictures and words represent in diVerent
ways, that the form of representation involved is diVerent in the two cases.
Thus one way to reformulate the question with which we began is this.
What is pictorial representation? What is the form of representation which
7
Figure 1 Pablo Picasso, Guernica

The question 9
pictures display, and how does it diVer from representation in language?

Much of this book will be spent trying to understand this form of
representation, which I shall for variety’s sake consider under several
names – ‘pictorial representation’, ‘depiction’ and ‘picturing’.
Now, we will get nowhere with this issue unless we immediately take
note of something. It is that pictures themselves represent in a variety of
ways. I do not mean by this that pictorial representation may itself
fragment into a largely disparate collection of subspecies. That may
indeed be so; but if it is, it should be a conclusion of our enquiry, not a
preliminary to it. Rather, I mean that pictorial representation may be only
one of several forms which pictures exhibit. An example helps make this
suggestion more plausible.
Many pictures from the religious art of the West represent the Holy
Spirit by depicting a dove. I suggest that there is not one form of
representation here, but two. The dove is depicted, but the Holy Spirit is
represented in some other way. After all, the Spirit is only represented by
virtue of the fact that the dove is, but the converse is not true. This
suggests that the representation of the Holy Spirit is a more complex,
more derived phenomenon than the representation of the dove. Further, a
description of the scene which mentioned a dove might represent the
presence of the Holy Spirit in a similarly derived manner. This provides at
least some reason for thinking that the description and the picture
represent the Spirit in the same way, a way that will not therefore be
distinctively pictorial. In contrast, they represent the dove in very diVerent
ways, and the diVerence is precisely that between pictorial and linguistic
representation.
If pictures can represent in several diVerent ways, we will never be clear
about pictorial representation unless we take care not to confuse it with
those other forms. We must, then, be prepared Wrst to isolate depiction
and then to attempt to understand it. Before we engage in either task for
ourselves, however, we should see what answers to our question are
already available.
2 Resemblance
There is a natural thought with which to begin consideration of these
matters, and which has appealed to many. It is that depiction is
intimately connected to resemblance. Pictures, the thought runs, look
like or resemble what they represent. A photograph of someone re-
sembles that person, and an oil painting of a well-groomed horse looks
like a horse. In contrast, written descriptions of these things, whatever
tongue they are in, do not resemble them. For how could a series of
10 Picture, Image and Experience
letters on a page resemble a Xesh-and-blood person or stallion? Here,

then, is the crucial diVerence between representation by pictures and by
words: the former depends on resemblance and the latter does not.
Here, indeed, is the essence of depiction: one thing depicts another only
if the Wrst resembles the second.
Attractive as this idea may seem, it has been subject to copious and
powerful criticism.… From this onslaught three problems emerge as
suYciently diYcult to require tackling by any serious version of the
resemblance view. I will brieXy expound each.
Before I can outline the Wrst problem, we must consider the logical
framework on which the fabric of depiction rests. Some depiction is
depiction of a particular item. For example, Elisabeth Vigée-Lebrun’s
Portrait of Countess Golovine (Wgure 2) depicts that woman, the
eighteenth-century Russian aristocrat. However, not every picture de-
picts a particular. Although Guernica depicts a mother weeping over her
dead child, there is no particular woman represented here. Pictures such
as this depict some, but no particular, thing with certain properties – in
this case the property of being a woman, being a mother, being wracked
by grief, holding a child and so on. However, if we leave matters at that we
implicitly exaggerate the diVerences between the two sorts of picture. For
even those pictures which do depict particular things must depict them as
having certain properties. The Vigée-Lebrun portrait, for example,
depicts the Countess as alert, as holding her shawl to her chest in a
striking way, as wearing a distinctive scarf in her tousled hair. Vigée-
Lebrun did not have to attribute these properties to the Countess – she
might have depicted her without her shawl, dozing peacefully. But what
she could not do was to depict Golovine without attributing any
properties to her. Thus the distinction we want is really between pictures
representing some set of properties, though not some particular which
possesses them; and pictures representing some set of properties as
possessed by a particular thing.
Against this background we may formulate the Wrst problem for the
resemblance view. It is that it seems only to cope with the depiction of
particulars. The problem is that resemblance is a relation between two
particulars – one resembling the other. It is hard to know how to make
sense of resemblance between a particular thing and some, but no
… Criticism goes back at least to Descartes. See the selection from the fourth discourse of his
Optics (1637) in Cottingham et al. 1985, pp. 164–66. For more recent scepticism, see
Goodman 1969, chapter 1; Black 1972, pp. 117–25; Walton 1973, footnote 23; Schier 1986,
pp. 2–9, 179–88.
The locus classicus for discussion of these issues is Goodman 1969, chapter 1, sections 5–6.
For a related distinction see Kaplan 1969, pp. 225–28. For discussion see Howell 1974;
Novitz 1977; Phillips 1978; and Wollheim 1987, chapter 2, note 16.
The question 11
particular, item of a certain sort – a horse, say. For resemblance is

presumably a matter of shared properties – A resembles B provided both
have a certain property F. Yet only what exists has properties, and thus
can share them with anything else. Since there are no horses which are not
particular horses, there are no properties enjoyed by some, but no
particular, horse. So there can be no resemblance between a picture and
such a horse, and thus no prospect for understanding the depiction of a
(no particular) horse in terms of resemblance. It seems that what we have
here is at best a way of understanding some depiction, not even the
beginnings of an account of it all.À
The second problem strikes at the heart of the resemblance view. The
view is motivated by intuitions that particular pictures do indeed re-
semble what they depict, but there is reason to doubt that any such
intuition is sound. Resemblance must be resemblance in certain re-
spects. If two things resemble, they must do so in respect of some
property or other, perhaps in respect of many. Unfortunately, when we
ask in what respect picture and object resemble, it is easier to Wnd
diVerence than likeness. Consider a case as likely to motivate the view as
any, that of a photograph of someone. Picture and person are not the
same shape, and need not be the same colour (the photograph might be
black and white). It is hard to suggest that the two resemble in texture,
and while patterns of light and dark might here be shared, they would
not be in the case of an outline pencil sketch of the person. Moreover,
photograph and person are clearly made of diVerent materials, and are
patently diVerent in their capacities and proclivities – the latter being, for
example, animate and mobile, the former not. In short, for any respect
on which we alight, resemblance is lacking and diVerence is plentiful.
Thus we must question the intuition from which we began. Unless we
can say something about the points of similarity, we must abandon the
idea that there are any.
The third problem consolidates the attack begun by the second. The
latter undermined the idea that any intuition concerning resemblance is
sound. The former suggests that for some pictures it is not even initially
tempting to say that they resemble what they depict. What drives the
resemblance view is consideration of photographs and other realistic
depictions. But the range of pictures is far wider than that, and so also is
the range of depiction itself. Is it even superWcially plausible that
À What of depiction of Wctional entities, such as Pegasus? The problem is just as acute here,
since again there is no possession of properties to sustain resemblance. For it is not the case
that Pegasus has properties, though it is Wctional that he does. However, the depiction of
Wctional entities is both complex and parasitic upon non-Wctive depiction. In what follows,
I will not be oVering an account of Wctive depiction, although I hope that what I do say
could provide the basis for such an account.
Figure 2 Elisabeth Vigée-Lebrun, Countess Golovine

resemblance holds between picture and object in the case of caricatures,
or stick Wgure drawings, or even of post-cubist pictures such as Guernica?
These depict, but do not, apparently, resemble what they depict. Unless
these examples can be accommodated, resemblance can provide at best
the core of realistic depiction, not of picturing itself.
The question 13
3 Goodman
One way to react to these diYculties would be to reject, not simply the
resemblance view, but the attitude to depiction which it embodies.
Resemblance draws the contrast between pictorial and linguistic repre-
sentation very sharply, the one turning on resemblance, and therefore on
some relation which might be thought to hold independently of human
contrivance; the other presumably depending on conventions, deriving its
power to represent entirely from our decisions and practices. Perhaps
instead picturing is no less conventional than describing, but simply
exploits diVerent conventions. The problem this presents is to say what is
special about the conventions governing depiction. The Wnest exponent of
this approach is Nelson Goodman.Ã
Goodman begins by developing the means to describe many diVerent
forms of representation. The fundamental notion is that of the symbol
system. A symbol system consists of a set of characters correlated with a
Weld of reference. Characters are ways of grouping marks on surfaces.
Thus the following three marks fall into two groups: a a d. The Wrst two
both fall into the group constituted by inscriptions of the Wrst letter of the
Roman alphabet, the third falls into the group consisting of inscriptions of
its fourth letter. A set of characters is just that – the Roman alphabet
provides one example, its Wrst eight letters provide another. A Weld of
reference is a set of items the characters refer to – the notes of the musical
octave, or the stars, for instance. Combining a set of characters with a Weld
of reference yields a symbol system. Examples are the names for the notes
of the octave, or, at a higher level of complexity, written English.
What is special about symbol systems which are pictorial? Goodman
identiWes three features which are important. First, such systems are
syntactically dense. That is, they ‘provide for inWnitely many characters so
ordered that between any two there is a third’ (1969 iv, p. 2). Second, they
are semantically dense. The Weld of reference of the characters is so ordered
that between any two referents there is a third (1969 iv, p. 5). Finally, they
are relatively replete – a relatively wide range of properties of the mark aVect
which character it inscribes (1969 vi, p. 1). Roughly, what all this amounts
to is the following. Pictorial systems are ones in which, for a wide range of
properties of the mark on the surface, the tiniest diVerences in that
property matter to what is represented. For any such diVerence may alter
which character the mark inscribes, and thus which item in the Weld of
reference is being represented. Thus the precise colour of a patch of paint
Ã See Goodman 1969 and, for some minor clariWcations and revisions, Goodman and Elgin
1988, chapter 8. For some of the widespread discussion of this view, see Wollheim 1977;
Schier 1986, chapter 1, section 6; Peacocke 1987; Walton 1990, chapter 8, passim.
on the surface of a portrait may aVect exactly what colour the sitter’s coat
is represented as being. In contrast, the exact shape of the marks on a page
of script is irrelevant to what they describe – provided that the words
remain the same, we could change the style of script without altering its
meaning.
Now, an example shows that Goodman’s conditions cannot be suY-
cient for depiction.Õ We might use a graph to track the temperature of a
quantity of colourless gas over time. With time elapsed along the x-axis,
various features of the plotted line might feed, in a weighted manner, into
the temperature represented. These features might include the line’s
height against the y-axis, its thickness, its hue, its saturation, its bright-
ness, and so on. The graph would be a symbol in a system which is both
syntactically and semantically dense, and relatively replete. Yet, as has
been noted by those who oVer such examples, it would not depict
anything.
This claim would not by itself alarm Goodman. He accepts that he has
not deWned depiction. His claim is instead that the three features he
describes are all exhibited by pictorial systems, and that they are not all
exhibited by those which represent linguistically. So although depiction
has not been deWned, it has been diVerentiated from description (1988
viii, section 5). However, I think that the modesty of Goodman’s
ambitions may not entirely dispense with the unease the gas example
created. For the graph seems, as it were, very distant from depiction; the
way it represents seems at best only peripherally related to the way that
pictures do. Since the graph has all the features Goodman identiWes as
distinctive of depiction, we may wonder whether he has provided what we
were looking for. After all, we did not merely want to separate depiction
from description, but to understand the nature of the former. The
example encourages the idea that Goodman’s view somehow fails to do
this. However, this can remain only an unfocussed anxiety until the next
chapter, where we will Wnd out more about depiction.
4 Experiential accounts
One source of worry about Goodman’s account might be that it takes
insuYcient account of the visual nature of picturing. Goodman does not
require that symbols in pictorial systems be seen, and certainly does not
insist that the things they represent be visible – as the gas graph
demonstrates. Yet it seems no accident that there are no spoken pictures,
the way that a description can be spoken as well as written; and that
Õ I have adapted this example from Peacocke 1987, p. 405.

The question 15
pictures seem especially suited to representing the visible world. Thus in

some way depiction seems bound to the visual, although this thought
certainly needs clarifying. One way to do that is to claim that depiction
essentially involves a special visual experience on the part of the viewer.
What might this option amount to, and what could its justiWcation be?
Sometimes, when you look at a picture, you do not at Wrst see what it
depicts. You can see the patterned surface before you, and can see quite
clearly what marks lie where. You may even be able to tell that something is
depicted, from the obvious care with which the surface has been marked,
but not what is. Then, in a moment, the way you see the surface is
transformed. You can now see, let us suppose, that the picture is of a
horse, that the strange shaped lump that had puzzled you depicts its head,
those straggly lines of colour its legs, and so forth. It need not be that you
now see any of those marks as lying diVerently from how they seemed to
before. You need not think that you previously mistook or overlooked the
position of some mark. It is simply that you now see the marks as
organized in a particular way.
There seem to be two diVerent experiences here, one preceding
understanding the picture, the other accompanying it. The thought
behind the approach to depiction we are now considering is that every
picture admits of an experience akin to this latter experience. Moreover,
the idea is that this second sort of experience holds the key to depiction. If
we can discover what is special about it, in particular how it diVers from
the ‘before’ experience, then, the thought runs, we can analyse picturing. I
call this the experiential approach to the topic. Its claims will seem more
plausible if we make some further observations about the ‘after’ experi-
ence.
Let us follow common practice and dub this experience seeing-in, since
it is both convenient and natural to describe you as seeing a horse in the
picture.Œ There are four features of seeing-in which are of immediate
relevance.
First, seeing-in is an experience with a distinctive phenomenology. By
this I mean that there is something it is like to have that experience, and
that what it is like to have it diVers from what it is like to have other
experiences. One crucial diVerence in this respect is with the ‘before’
experience above, since the move from that to seeing-in involves a change
in phenomenology. But there are other important contrasts too, and in
fact they fall into two groups. On the one hand, seeing a horse in the
picture diVers from visual experiences of other kinds. It diVers, in its
Œ ‘Seeing-in’ is Wollheim’s term, at least in his later writings (cf. 1980, supplementary essay
v). Its use is now widespread, and I follow custom in using it without adopting Wollheim’s
own account of the experience.
phenomenology, from failing to see a horse therein, but also from the
experience of seeing a horse in the Xesh, or the experience of visualizing a
horse. So seeing-in is a special kind of experience, one marked out by what
it is like to have it. On the other hand, particular experiences of seeing-in
also diVer from one another in this way. To see a horse in a picture is to
have an experience with a diVerent phenomenology from seeing a donkey,
a man or anything else therein.
Second, seeing-in is an experience whose content somehow includes
the picture’s object. That is, it is in some way experience as of something
other than the picture – in this case a horse. The thought (or some such) of
a horse enters your experience of the picture.
Third, despite the last point, seeing-in remains a way of seeing the
picture. It is not merely an experience induced and sustained by exposure
to the patterned surface. That would be all that was involved in, for
example, a case where seeing a familiar picture sends you into reverie, a
series of vivid visualizings of past or fantasized events. In seeing-in, in
contrast, the patterned surface is seen to be before you. Seeing-in is in
important part an experience which represents the surface of the picture –
as is shown by the crucial fact that it represents it as organized in a special
way.
The second and third features are connected to the Wrst. That the
phenomenology of seeing-in diVers from that of other types of experience
is due both to the thought of an absent object and to the awareness of the
marked surface before one. The former is not necessarily present in
ordinary seeing, i.e. seeing something in the Xesh; nor in seeing a picture
without seeing anything in it. The latter does not hold for visualizing
something. Furthermore, diVerences between speciWc experiences of
seeing-in also turn on these two features. In fact, if one such experience
diVers phenomenologically from another, it is either because it involves
the thought of a diVerent absent object (a cow, rather than a horse,
perhaps); or because it involves an awareness of a diVerently marked
surface (e.g. of an etching rather than the oil painting it copies); or both.
Fourth, seeing-in is an integrated whole. The thought of an absent
object and the awareness of the marked surface before one are both
present, but not in such a way as to fragment the experience. To see the
point here, consider another example. Suppose that while looking at a
castle you visualize a horse. This complex experience lacks a feature I want
to ascribe to seeing-in. It is a whole composed of two clear parts – the
seeing and the visualizing. Each component could occur in isolation from
the other – you might have seen the castle without visualizing the horse, or
vice versa. Had either component occurred singly, its phenomenology
would be just as it is in the case where it occurs with the other. Thus what
The question 17
we have here is at most an experience formed by the simple concatenation

of two elements, and perhaps only the co-occurrence of two diVerent
experiences. In seeing-in, by contrast, there are no elements which can be
torn from the whole without doing violence to them. If either the thought
of the absent object or the awareness of the marked surface occurred in
isolation, the phenomenology of each would be diVerent. Thus the
experience cannot be broken down into elements which could stand
alone.
These four features of seeing-in help us make sense of the experiential
approach to depiction. The idea is that every depiction may be experi-
enced in this distinctive way. The resulting experience will be an integral
whole, with its own distinctive phenomenology. It will owe that phenom-
enology to the fact that it is a way of seeing the picture, but also that it
somehow involves the thought of something else, the picture’s object.
Pictures sustain such experiences, and they depict by virtue of sustaining
them. Moreover, the precise nature of the experience determines what
each picture depicts. If it sustains an experience permeated by the thought
of a horse, it depicts a horse. If it sustains one permeated by the thought of
the Countess Golovine, it depicts the Countess. Thus we have the
beginnings of an account of what picturing is, and of what is required to
depict one thing rather than another.
However, there are only the beginnings of such an account here. For
one thing, what is in many ways the crucial feature, the second, remains
hopelessly vague. I presented it as the claim that experience of the picture
somehow involves the thought, or some such, of something else. This is
really only the sketch of a claim. Until we have been told more about what
is involved in the experience, and the nature of its involvement, it is far
from clear what the second feature amounts to. For another thing, as
matters stand it is not obvious that the four features suYce to distinguish
seeing-in from other experiences. Consider, for example, that of seeing a
horse through a dirty window pane. Here there is awareness of a
‘patterned’ surface, and of a second thing (the horse), producing an
experience which has a distinctive phenomenology and is a phenomenally
integrated whole. Yet clearly this is not the same experience as seeing-in.
For, to mention just one important diVerence, seeing the horse through
the marks on the window is not to see them as organized in a special way.
The upshot is that we need to say more. We need a characterization of
seeing-in, something to take us beyond the four features above. And we
need that characterization to help clarify the second feature, the nature of
the involvement of the absent object in the experience. And this means
that there will be several diVerent views which embody the experiential
approach. For the four features take us to the boundaries of the territory
within which agreement is to be found. From now on all will be

controversy and discord.œ
It is helpful to begin with the simplest account of these matters. That is
to suggest that pictures are experienced in exactly the same way as their
objects. In other words, when you see a horse in the picture, your
experience is identical, in its phenomenal nature, to the experience you
have when you see a horse in the Xesh. Note that this view makes no claims
about the beliefs to which seeing-in gives rise. In particular, it does not
claim that you are led to believe that a horse is before you when you see the
picture. I will call this position the illusionist view.–
Illusionism is a useful position to bear in mind when thinking about
depiction, as I hope will become clear as we proceed. Its utility is,
however, somewhat surprising in the light of the obvious fact that the view
is false. It is simply untrue that looking at a picture is just like looking at the
picture’s object. This is plain for pictures which are not realistic –
caricatures, pencil sketches, cave paintings, and the like. But it is also true
of almost all other pictures too. For if illusionism were right, seeing-in
would not have the third feature identiWed above. Seeing something in a
picture would not then be a way of seeing the picture, but merely a matter
of having an experience induced and sustained by exposure to the
picture’s surface. For as we have deWned the view, it does not allow the
experience to represent the picture as before the viewer at all. If it did so, it
could hardly match in phenomenology the experience of seeing a horse in
the Xesh, an experience for which there is no painted surface to be seen,
and which in no way represents the environment as containing such a
surface.—
A more promising line is oVered by Richard Wollheim.…» In his view, the
primary feature of seeing-in is what he calls ‘two-foldness’, the property of
being a single experience comprised of two ‘distinguishable but also
inseparable’ aspects (1987, p. 46). One aspect, the ‘conWgurational’, is
analogous to the experience of seeing the picture without seeing anything
in it. The other, ‘recognitional’, aspect is analogous to seeing the picture’s
object face-to-face.
œ For accounts of seeing-in other than those discussed below, cf. Schier 1986, chapter 10;
Peacocke 1987; Budd 1993.
– Although I do not know of anyone who has held a position as strong as the one here
described, there are many who are tempted to try and weaken it. Cf. Gombrich 1977.
— Are there any pictures we see in the way the illusionist describes? It is certainly the goal of
trompe l’oeil painting to generate that experience, and no doubt that goal is attainable,
although probably only if the picture is seen in rather special circumstances. However, if
the third feature is right, this illusionist experience is not seeing-in.
…» Wollheim’s view has altered over the years in both substance (compare his 1980 with
1987) and terminology (compare the text of his 1980 with its supplementary essay v). I
discuss the latest, and best, version.
The question 19
It is clear what role two-foldness is to play. It allows Wollheim to

accommodate the second and third features of seeing-in, that its content
somehow involves the picture’s object, and that it represents the picture’s
surface. Seeing a horse in the Xesh clearly involves awareness of a horse.
Equally, seeing the picture without understanding it clearly involves
experience of its surface. So if seeing a horse in the picture involves
elements analogous to each of these two experiences, it will both somehow
involve a horse and include awareness of the picture.
However, we might wonder whether this way of accommodating the
two features is satisfactory. For one thing, it looks like it might repeat the
errors of the illusionist view. It would do so if it claimed that one fold in
seeing-in were phenomenally identical to seeing, for example, a horse in
the Xesh. For those pictures for which illusionism was not at all plausible –
caricatures, pencil sketches, and the like, do not seem better suited even to
the weaker claim that part of the experience matches that of seeing the
picture’s object. Certainly, as I argued in discussing the fourth feature,
there is no part of the experience which can simply be excised from the
whole; yet if some part matched ordinary visual experience of the object,
that would surely prove separable in just that illicit way. Perhaps this
merely shows that we should not understand Wollheim’s ‘analogous’ as
‘phenomenally identical’. But if not, how are we to take it?
Wollheim anticipates this objection, and builds into his account a
response to it. He makes it clear that the two aspects are to be analogous to
their corresponding experiences in a way that does not involve their being
phenomenally identical to them. Moreover, it is a mistake to ask in what
way either fold is analogous to the relevant experience:
We get lost once we start comparing the phenomenology of our perception of the
boy when we see him in the [surface], or the phenomenology of our perception of
the [surface] when we see the boy in it, with that of our perception of the boy or
[surface] seen face-to-face . . .. The particular complexity that one kind of
experience has and the other lacks makes their phenomenology incommensurate.
(1987, pp. 46–47)
Unfortunately this line is not one we are able to accept. Above we rejected
illusionism as fundamentally implausible. That left us needing an alterna-
tive way to understand seeing-in. Wollheim now enters, conWrming that it
would be wrong to do this by identifying any part of seeing-in with seeing
face-to-face, but refusing to oVer an alternative. This is not in the least
informative. It does not answer our earlier question about the way in
which the notion of a horse enters the experience of seeing one in a picture.
For all we are told is that some element in seeing-in is somehow analogous
to seeing a horse. Equally, since all we are told about the other element is
that it is somehow analogous to ‘before’ experience of the picture, we are no

wiser in trying to incorporate the third feature of seeing-in.
Now, all this caution might be tolerable were Wollheim’s reason for it a
good one. The reason we are oVered is that the complexity of seeing-in as a
whole renders the phenomenology of the two folds ‘incommensurate’
with that of their analogues. Unfortunately, it is not clear why this should
be. After all, the complexity of the whole is just that, a feature of the
complete experience. Why should it aVect our ability to grasp the nature of
one of the components of that complex whole?…… True, I myself have
argued that seeing-in’s integrated nature means that it does not break
down into separable independent components. But it does not follow
from this that we should see it as composed of elements which are
indescribable. Perhaps we should understand it as a whole, rather than
attempting to identify strands in it which cannot be teased apart from the
rest. Since Wollheim just ignores this possibility, his caution begins to look
like defeatism.
One view which does attempt to say more about seeing-in is Kendall
Walton’s.… He makes use of the attractive idea that seeing-in involves the
visual imagination. The appeal to imagination is what enables Walton to
accommodate seeing-in’s second feature, the involvement of the absent
object: in seeing a horse in a picture, one imagines seeing a horse. But one
does this by virtue of seeing the picture, and hence the third feature,
awareness of the marked surface. Now, simply imagining one thing while
seeing another will not produce an experience with the distinctive, and
integrated, phenomenology of seeing-in, as was demonstrated by the case
of visualizing a horse while looking at a castle. So Walton adds the Wnal
requirement that one is to imagine of one’s seeing the picture that it is
one’s seeing the horse (1990, p. 293).
One might wonder whether this Wnal requirement is up to the job.
Visualizing a horse is an experience with its own distinctive phenomenol-
ogy. So is seeing a marked surface as before one. How can these two
combine to form an integrated whole?…À There are really two problems
here. First, how is there room, as it were, in a single experience for two
elements each of which has a phenomenology suYciently rich to consti-
tute an experience in its own right? Combining two such elements in the
castle case merely produced a composite experience. How is the experi-
ence described by Walton better placed? In the case of the castle
integration would not be achieved if the subject acquired the belief that
her seeing the castle caused her to imagine seeing a horse. Why should it
…… This and many other incisive criticisms of Wollheim’s view are made in Budd 1992.
… Walton 1990, chapter 8. See also Walton 1973.
…À Walton stresses that seeing-in is indeed integrated (1990, p. 295).
The question 21
be any more help, in the case of the picture, for her to imagine seeing the
marks to be her seeing the horse? Second, if the two elements are to be
integrated, they presumably must be transformed in the process. If not,
the resulting experience must be composite as the experience of the castle
was, and as seeing-in is not. But if the elements are transformed in their
union, what is their phenomenology once changed? Unless Walton can
say something about them in their altered state, he has not succeeded in
characterizing seeing-in in terms of its phenomenology, as he hoped to do;
rather, he has at most characterized it in terms of the resources which go
into it.
However, these diYculties only hold if we take Walton to be saying that
seeing-in involves a particular sort of imagining, visualizing. For visualiz-
ing has a distinctive phenomenology of its own, and this generates the
problems above.…Ã But there are other sorts of imagining to which Walton
might appeal; imagining has propositional, as well as experiential, forms.
Walton himself notes that imagining covers many diVerent sorts of activity
(1990, pp. 13, 43), and considers the diYculty of distinguishing proposi-
tional imagining from merely entertaining a proposition (pp. 19–20).
Perhaps, then, we would do better to take the imaginative seeing which,
on his account, seeing-in involves to be something other than visualizing.
There seems no reason to deny that one can imagine, in these other ways,
seeing something, and since other forms of imagining lack the rich
phenomenology of visualizing, we might thereby hope to avoid the
problems encountered above. Better still, when challenged to clarify his
position on seeing-in, Walton suggests that he may indeed intend
something other than visualizing (1991, pp. 423–27).
Unfortunately, this second reading of the view faces troubles of its own.
The account now certainly avoids the diYculty of capturing the fourth
feature of pictorial experience, that is the diYculty of integrating the
imagining and the seeing. The trouble now is from the Wrst feature,
providing an experience which is phenomenologically distinctive at all.
To see this, consider the making of an episode of Star Trek. The cast are
rehearsing a scene in which the covers on a window on the ship’s bridge
are pulled back to reveal the inert wreck of a friendly spacecraft. All they
see is a plain blue screen, onto which Wlm of a model of the ruined hulk will
later be projected. Finding the actors’ performances unusually wooden,
the director tells them to concentrate on imagining their reaction to the
sight as it is unveiled. He explicitly instructs them to imagine that, as the
covers are slowly pulled back, their seeing this bit of the screen is their
catching sight of the ship’s engines, their seeing that bit their making out
…Ã For much more on this phenomenology, see chapter 7 below.

what remains of the hull, and so on. It seems to me that in obeying the
director’s instructions the cast will meet all the conditions imposed by the
second reading above. But they will surely not see a spacecraft in the
screen. They simply continue to see a plain blue, undiVerentiated surface.
Thus this second version of Walton’s view is also in trouble. At least this
time the analysans does not have features inconsistent with the analysan-
dum, but the analysis oVered does seem to be insuYcient.…Õ
Thus neither Walton’s view nor any of the others discussed above is free
from diYculties. On the other hand, none is entirely without appeal. What
is needed now is some way of structuring the debate between these various
positions, to allow us to weigh the merits of each. That structure is what
the next chapter is intended to provide.
…Õ Walton mentions two other requirements on seeing-in to which he might appeal here. He
states that for seeing-in to occur the games of imagination played with the surface must be
suYciently rich and suYciently vivid (1990, p. 296). Richness amounts to there being a
wide range of visual actions such that the viewer really performs that action on the picture
and imaginatively performs it on the object depicted. I do not see that this will help deXect
the Star Trek example. As for vividness, it is hard to make sense of in the context of
imagining which is not visualizing.

Hopkins R. - Picture, Image

Uploaded by

Copyright:

Available Formats

Hopkins R. - Picture, Image

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hopkins R. - Picture, Image

Uploaded by

Copyright:

Available Formats

job:LAY00 6-8-1998 page:1 colour:0

Picture, Image and Experience

Robert Hopkins is lecturer in philosophy at the University of Birmingham.

Picture, Image and Experience

© Robert Hopkins 1998

This book is in copyright. Subject to statutory exception and to the provisions of

First published in 1998

Printed in the United Kingdom at the University Press, Cambridge

Typeset in Plantin 10/12pt [vn]

Library of Congress cataloguing in publication data

is b n 0521 58259 8 hardback

List of illustrations page vii

2 Some features to explain 23

6 Indeterminacy and interpretation 122

1 Pablo Picasso, Guernica (© Succession Picasso/DACS 1998) page 8

viii List of illustrations

17 Tsan Tsu Tsao, Mountainous Landscape (reproduced by

I am grateful to the following journals for permission to reproduce

Figure 1 Pablo Picasso, Guernica

pictures display, and how does it diVer from representation in language?

10 Picture, Image and Experience

letters on a page resemble a Xesh-and-blood person or stallion? Here,

particular, item of a certain sort – a horse, say. For resemblance is

12 Picture, Image and Experience

Figure 2 Elisabeth Vigée-Lebrun, Countess Golovine

14 Picture, Image and Experience

Õ I have adapted this example from Peacocke 1987, p. 405.

pictures seem especially suited to representing the visible world. Thus in

16 Picture, Image and Experience

we have here is at most an experience formed by the simple concatenation

18 Picture, Image and Experience

within which agreement is to be found. From now on all will be

It is clear what role two-foldness is to play. It allows Wollheim to

20 Picture, Image and Experience

that it is somehow analogous to ‘before’ experience of the picture, we are no

…Ã For much more on this phenomenology, see chapter 7 below.

22 Picture, Image and Experience

You might also like