Untitled

THE STRUCTURING BODY
A Critical Study in the Description &

Explanation of Perceptual Experience

Inauguraldissertation
zur Erlangung des Akademischen Grades
eines Dr. phil.,

vorgelegt dem Fachbereich 05 Philosophie und Philologie
der Johannes Gutenberg-Universita t Mainz

von
Adrian J.T. Alsmith
aus London

2011

| i

Table of contents

Tabl e of f i gures i i i
Abstract iv
Notes on the text v
General i ntroducti on 1
PART I 11
Chapter 1 Sense-data 12
1 . 0. Introducti on 1 2
1 . 1 . Common- sense, real i ty, and perceptual experi ence 13
1 . 2. Sense- data & the argument f rom perspecti val vari ati on 1 6
1 . 3. Indi rect percepti on 1 8
1 . 4. The metaphysi cs of sense- data 22
1 . 5. The neutral i ty of the di rect obj ects of perceptual experi ence 24
Chapter 2 Di sputi ng perceptual experi ence 29
2. 0. Introducti on 29
2. 1 . The pl anar proj ecti on vi ew 30
2. 2. The appeal to depi cti on 32
2. 3. The vi sual f i el d 35
2. 4. Indi rect perceptual experi ence 38
2. 5. Prostheti c percepti on 44
Chapter 3 Berkel ey & the l i mi ts of sensati on 49
3. 1 . Li mi tati on arguments & the one poi nt argument 50
3. 2. Berkel eys i deas of sense 53
3. 3. Three obj ects of i mmedi ate si ght 61
3. 4. Sensati on, percepti on, & perceptual experi ence 69
3. 5. Perceptual sensati on 73
PART II 79
Chapter 4 The neutral i ntenti onal i ty of perceptual experi ence 80
4. 1 . Neutral i ty and perceptual experi ence 81
4. 2. Intenti onal i ty 84
4. 3. Intenti onal i ty and perceptual experi ence 89
4. 4. The natural contents of perceptual experi ence 94
4. 5. Phenomenol ogy vs. neutral i ntenti onal i ty 99
ii |
Chapter 5 Seemi ng real 106
5. 0. Introducti on 1 06
5. 1 . Perf ecti on 1 07
5. 2. Immedi acy 1 20
5. 3. Transparency 1 28
5. 4. Presence 1 42
5. 5. Parti cul ari ty 1 46
Chapter 6 Husserl on the experi ence of parti cul ari ty 151
6. 1 . Husserl s Body 152
6. 2. Lei bhaf ti gkei t 1 54
6. 3. The Husserl i an account 157
6. 4. Husserl i an notes on Noan enacti vi sm 1 66
6. 5. Weak mi nd- i ndependence 173
PART III 183
Chapter 7 Structural affordances 184
7. 0. Introducti on 1 84
7. 1 . Propri etar y i nf ormati on 1 85
7. 2. The concept of a structural af fordance 202
7. 3. The empi ri cal real i ty of structural af f ordances 21 0
7. 4. A structural af f ordance theory of ki naestheti c experi ence 222
7. 5. Phenomenal grooves & ki naestheti c abi l i ti es 231
Chapter 8 On bodi ly sel f hood 236
8. 1 . Descartes hi msel f 237
8. 2. Cartesi an sel f hood 241
8. 3. Anti -Cartesi an sel f hood 244
8. 4. Bei ng one body 251
8. 5. Bei ng no body 256
Chapter 9 Cl eari ng the ground 263
9. 1 . Bodi ly sel f -model s 264
9. 2. Representi ng the body 272
9. 3. Neural vs. non-neural i ndi vi dual s 279
9. 4. Laws vs. mechani sm 286
9. 5. Representati on vs. dynami cs 295
Concl usi on 315
References 323

| iii

Table of figures
1 Perspectival views of a square ...................................................................................... 17
2 Planar projection in vision........................................................................................... 31
3 A String of Memories by J.D. Hillberry ........................................................................ 34
4 Mach's rendition of his left visual field ....................................................................... 35
5 The binocular visual field: deficits and neural correlates ............................................ 37
6 Prosthetic perception through planar tactile stimulus arrays ..................................... 47
7 Distance as a line directed end-wise to the eye ............................................................ 51
8 Berkeley's treatment of the Barrow Illusion ................................................................ 66
9 The feelSpace prosthesis ............................................................................................. 92
10 The Mller-Lyer illusion ............................................................................................ 122
11 Ascending and Descending by M. C. Escher. .............................................................. 124
12 Andrew Lipsons Lego construction of Ascending and Descending .......................... 124
13 A Necker cube ............................................................................................................ 138
14 Two views of a film-set faade ................................................................................... 156
15 Neurophysiology of muscle receptors ........................................................................ 186
16 The ambiguity of muscle spindle information ............................................................ 191
17 Forward and inverse modelling of sensorimotor processes ........................................ 195
18 Gibsonian proprioception .......................................................................................... 201
19 Redundancy in a four joint limb ................................................................................ 210
20 Humanoid walker prototypes at the Delft Technical University Biorobotics lab ........ 214
21 Wiping ref lex in the spinalised frog .......................................................................... 217
22 Stepping ref lex in the human infant .......................................................................... 219
23 Treadmill stepping patterns in pre-walking infants .................................................. 220
24 An experimental paradigm for sensorimotor control biases ......................................228
25 Autoscopic phenomena ............................................................................................. 269
26 - Global and local views on body representation .......................................................... 273
27 Two varieties of individual ......................................................................................... 283
28 Varieties of individualism .......................................................................................... 285
29 A pitchfork bifurcation ............................................................................................. 290
30 Cravers psi-phi model of mechanistic explanation ................................................... 293
31 James Watt's centrifugal governor ............................................................................ 299
32 Bechtel's representational governor.......................................................................... 304
33 Optimistically labelled comparator model ............................................................... 309
34 Simple schematic for emulation control ..................................................................... 311
35 Emulation and representation ................................................................................... 312
36 Articulated emulation ............................................................................................... 314
iv |
Abstract
The aims of the dissertation are to find the right description of the structure of per-
ceptual experience and to explore the ways in which the structure of the body might
serve to explain it.
In the first two parts, I articulate and defend the claim that perceptual experience
seems direct and the claim that its objects seem real. I defend these claims as inte-
gral parts of a coherent metaphysically neutral conception of perceptual experience.
Sense-datum theorists, certain influential perceptual psychologists, and early mod-
ern philosophers (most notably Berkeley) all disputed the claim that perceptual
experience seems direct. In Part I, I argue that the grounds on which they did so
were poor. The aim is then, in Part II, to give a proper appreciation of the distinctive
intentionality of perceptual experience whilst remaining metaphysically neutral. I
do so by drawing on the early work of Edmund Husserl, providing a characterisation
of the perceptual experience of objects as real, qua mind-independent particulars.
In Part III, I explore two possible explanations of the structure characterising the
intentionality of perceptual experience, both of which accord a distinctive explana-
tory role to the body. On one account, perceptual experience is structured by an
implicit pre-reflective consciousness of oneself as a body engaged in perceptual
activity. An alternative account makes no appeal to the metaphysically laden con-
cept of a bodily self. It seeks to explain the structure of perceptual experience by
appeal to anticipation of the structural constraints of the body. I develop this alter-
native by highlighting the conceptual and empirical basis for the idea that a first-
order structural affordance relation holds between a bodily agent and certain prop-
erties of its body. I then close with a discussion of the shared background assump-
tions that ought to inform disputes over whether the body itself (in addition to its
representation) ought to serve as an explanans in such an account.
Key words: Sense-datum theory, Berkeleys theory of vision, phenomenology of
transparency, phenomenology of reality, phenomenology of particularity, phenom-
enology of mind-independence, bodily selfhood, pre-reflective bodily self-
consciousness, structural affordances, bodily action, embodied cognition, cognitive
scientific explanation.

| v

Notes on the text
Where possible, I have followed the citation guidelines set out by the 6
th
edi-
tion of the Publication manual of the American Psychological Association
(2010). However, original publication dates are cited alongside contemporary
editions whenever there is significant disparity in pagination, format, con-
tent or date of publication ( 20 years). In cases where I have had the oppor-
tunity to access the original (or a facsimile), only that is cited.

All quotations from texts of originally non-Anglophone publication are from
the standard English translations. To facilitate cross referencing, in some
cases the original publication is cited alongside the first instance of the
translation cited in the main body of the text. In all these latter cases, for all
subsequent citations of the translation, pagination of the original will also
be cited between square brackets. Quotations over three lines in the body of
the text are at 10pt; quotations over three lines in footnotes are italicised.

Acronyms will be used for key phrases that recur several times during a
chapter. Each will be introduced in parentheses at its first occurrence in any
chapter in which such abbreviation is required.

The language of the text is British English. Hence, I have largely followed
British conventions of spelling and grammar in formal writing. However,
quotations will retain original orthography.
| 1

General introduction
Say you approached someone with an apple. You give it to them to feel, look at in
good light, maybe have a bite, etc. Then you ask them whether or not it seems like
they are perceiving the apple directly, and whether or not it seems real.
What answers might one expect?
Which would be correct?
How should we explain a given answer?
Much of the literature constituting contemporary philosophy of perception pre-
sumes that certain answers are to be expected of the confronted person. She will say
that it seems like she sees and touches and tastes the apple directly, and that the
apple she experiences is real. This conception of perceptual experience, the direct
realist conception, is presumed correct by many today. But it has not (at least, not
obviously) always been endorsed in the past. Moreover, it has yet to be given a sus-
tained articulation, let alone explained in any satisfactory manner. My aim in this
dissertation is to begin to fill this lacuna. The result is a description and explanation
of perceptual experience as structured by the constraints of the body. The project
might seem puzzling to some readers, and the result might well seem counter-
intuitive to some others. To begin to clarify, here is a part by part summary of what
is to come (a chapter by chapter summary follows at the end of the general intro-
duction).
In Part I, I discuss the claim that perceptual experience seems direct. A number of
influential figures in the recent history of ideas have claimed that the correct de-
scription of the direct object of perceptual experience forces us to recognise that
perception is indirect. Although this claim is framed variously as a metaphysical,
psychophysical and/or an epistemological thesis, it is interwoven with a supporting
account of perceptual experience as somehow impoverished with respect to the ob-
jects of perception. My hope is to tease apart claims that pertain to perceptual expe-
rience from claims that pertain to the metaphysics, psychophysics and/or episte-
mology of perception, in order to demonstrate that the impoverished conception of
perceptual experience is untenable.
In Part II, I discuss the claim that the objects of perceptual experience seem real.
Here the methodology developed in Part I becomes significant. I assume a position
of metaphysical neutrality on the objects of perceptual experience. From this stand-
point, I aim to construct a coherent description of how the intentionality of percep-
2 |
tual experience is conceived by direct realists. I do this by drawing on recent work
concerned to account for the phenomenon of transparency, the relation between
belief and perceptual experience, and the experience of mind-independent particu-
lars. In each case, one finds descriptions of aspects of perceptual experience driving
a metaphysical debate. In an effort to carry out a bona fide description of these as-
pects, whilst remaining true to the motivations of direct realism, I reconstruct
fragments of Husserls writings on the intentionality of perceptual experience. The
key claim developed in this reconstruction is that bodily experience structures the
intentionality of perceptual experience.
In Part III, I make programmatic attempts to provide an explanatory basis for sev-
eral positive features of the description evinced in Part I and Part II. My goal is to
provide a minimal theoretical basis for the claim that anticipations of continuity in
bodily movement structure the intentionality of perceptual experience. I consider,
only to discard, the more obvious strategy of claiming that this structuring role is
provided by the body qua bodily self. I lay the foundations for an alternative theory
by developing the concept of a first-order affordance relation holding between a
bodily agent and its parts, to be dubbed a structural affordance. Appeal to the no-
tion of a structural affordance presents an opportunity to develop novel accounts of
the representational strategies that might underlie the putative structuring role
played by bodily experience. In the final chapter, I discuss several issues that bear
upon the evaluation of whether the body itself (rather than only its representation)
ought to serve as an explanans in such an account.
Before I summarise the content of each chapter, some points of clarification are in
order. In contrast with many others that have written on the topic, I do not claim to
be faithful to any actual pre-theoretic, common-sense, folk conception of perceptu-
al experience if there is such a thing. Indeed, I would not be surprised if the poor
person you approached (apple in hand) wondered whether something sinister was
going on and treated you with a good deal of suspicion. Appealing to the opinions of
folk on matters that they would not usually have cause to question, matters on
which they might previously have had no opinion at all, is trickier than many might
like to admit. Moreover, arguably the class of creatures able to think about perceptu-
al experience is not co-extensive with the class of creatures able to undergo percep-
tual experience; limiting discussions of perceptual experience largely to homo sapi-
ens (qua human persons) potentially has its drawbacks.
Be all that as it may; it remains true that the appeal to a putatively pre-theoretic,
common-sense, folk conception of perceptual experience is pervasive in the study of
mind. Most notably, such an appeal is pervasive in the philosophy of perception and
in the philosophy of consciousness. A perspicuous case is Bertrand Russells proc-
| 3

lamation that [w]e are all in the habit of judging as to the real shapes of things,
and we do this so unreflectingly that we come to think we actually see the real
shapes (1912/1959, p. 10). Russell tells us what we all would be expected to say if
placed in good view of a rectangular table and asked what shape it looks to be: rec-
tangular, presumably. He moves on to assert another account of how it looks.
If our table is really rectangular, it will look, from almost all points
of view, as if it had two acute angles and two obtuse angles. If oppo-
site sides are parallel, they will look as if they converged to a point
away from the spectator; if they are of equal length, they will look as
if the nearer side were longer. (Russell, 1912/1959, pp. 10 - 11)
The suggestion of such an alternative account and its extrapolation to analogous
cases is the driving force of what has come to be known as the problem of percep-
tion, viz. the problem of whether reality can ever be perceived directly.
Russells line of argument is typical of those used to set up the problem. It is a classic
rendition of the argument from perspectival variation, sometimes described as a
version of the argument from perceptual relativity. Arguments of similar structure
form a family that also includes the argument from illusion and the argument from
hallucination. Each of these arguments has been held to yield a problem for the
claim that reality is perceived directly. And each of these arguments is either based
upon some observation of a case in which an object, property, relation, feature or
event is experienced as other than it is in reality, or a case in which it is experienced
despite the fact that it is not real or perhaps does not exist. Alfred Jules Ayer gives
voice to a battery of such cases:
It is remarked that a coin which looks circular from one point of view
may look elliptical from another; or that a stick which normally ap-
pears straight looks bent when it appears in water; or that to people
who take drugs such as mescal, things appear to change their colours.
The familiar cases of mirror images, and double images, and com-
plete hallucinations, such as the mirage, provide further examples.
[Similarly] it may be pointed out, for example, that the taste a thing
appears to have may vary with the condition of the palate; or that a
liquid will appear to have a different temperature according as the
hand that is feeling it is itself hot or cold; or that a coin seems larger
when it is placed on the tongue than when it is held in the palm of
the hand. (Ayer, 1973, p. 3)
It is easy to see how each of these examples might be employed as an antecedent in a
similar inference to Russells when he claims that what we see [of the table] is con-
stantly changing in shape as we move about the room and therefore that the sens-
4 |
es seem not to give us the truth about the table itself, but only about the appearance
of the table (1912/1959, p. 11).
Now, the fact that so many examples might be employed in a similar inference sug-
gests perhaps that the problem of perception is not so much a single problem as a
whole family of problems; but at the heart of all these lies a conception of perceptu-
al experience as a seemingly direct relation to reality. The problems raised by the
arguments above are not the only problems that perception poses in philosophy and
other disciplines; yet it might well be that most of other unmentioned problems will
also be haunted by the question of how to adequately characterise perceptual expe-
rience. If one is to understand the problems facing the study of perception, it seems
that one requires an appropriate grasp on perceptual experience.
Yet confrontation with such a daunting array of issues invites modesty in ones fo-
cus. Accordingly, I will attempt to frame a set of issues clustered around two claims:
that perceptual experience seems direct; and that its objects seem real. As regards
the arguments framing the problem of perception, I will only treat in any detail the
observations adduced in typical arguments from perspectival variation. Arguments
from perspectival variation focus on spatial properties and relations, such as size,
shape and distance. My assumption is that these can be usefully thought about in
isolation, at least for the present purposes. Certainly, I will mention the experience
of, e.g., light and colour, vibration, pain; but these will receive fairly cursory treat-
ment. I will mention hallucination only occasionally (as well as dreaming and imag-
ination), and there will be extended discussions of several illusions; but this will not
be in the context of addressing the arguments from hallucination and illusion per
se.
I also shy away from various questions regarding the distinction of the senses. Typi-
cal questions concern, e.g., whether we have the five senses commonly supposed,
and what is distinctive about a sense mode (or its objects) such that it is a specific
sense mode. A further question, bearing more directly on the dissertation, concerns
the extent to which claims about the structure of perceptual experience in one sense
mode might explain perceptual experience in other sense modes. As is a lamentable
tendency in contemporary philosophy of mind, the majority of my discussion per-
tains only to visual experience. The reader will find occasional reference to auditory
experience, tactile and olfactory experience; but these will be largely in the context
of discussing the views of others. I will not address the typical questions of how to
count the senses and how to distinguish them, nor will I address any further ques-
tions about the structure of experience across the senses. Put bluntly, I simply as-
sume that the nature of the senses is such that claims about the structure of experi-
encing intentional bodily movement can explain the structure of visual perceptual
| 5

experience. Although this assumption is elaborated upon, it is at no point given a
sustained defence. I regard it as one of the major issues raised by this dissertation
and hope to address it in future work.
Part III of the dissertation ought to be understood as a contribution to the project
of philosophical naturalism. As with the problem of perception above, it is perhaps
infelicitous to denote the project; anyone familiar with the expression will know that
naturalism comes in many varieties and degrees. Thus to be just slightly more defi-
nite, I allude to a naturalistic outlook as a matter of method, rather than as a stance
on whether, for instance, nature exhausts reality, or whether the latter is ultimately
reducible to the former. Moreover, I do not assume the authority of the natural sci-
ences over and against the claims and concepts (and peculiar problems) created and
dissected in philosophical discussion. The methodological naturalism I assume is
merely a conception of philosophy as broadly continuous with the sciences. More
specifically, I assume that the philosophical study of mind is broadly continuous
with the scientific study of mind.
Indeed, this dissertation is written with a studied indifference to disciplinary
boundaries and I believe that this reflects the plurality of methods in modern phi-
losophy of mind. I attempt to provide a discussion that is equally at home in the
philosophy of mind construed qua metaphysics, qua phenomenological study and
qua theoretical cognitive science. My attempts at the latter have required me to be a
somewhat humble student of psychology and the neurosciences. Having never prac-
ticed any of these sciences in a substantive sense, I have tried (where it is possible,
and not absolutely detrimental to style) to be cautious rather than cavalier in my
discussion. My hope is that in future collaboration and interactions I can correct the
errors likely to be present in my understanding and further develop any fruitful
lines of thought in concrete empirical study.
Part I
Chapt er 1
Sense-data
In the first chapter (and the two following) I discuss ways in which theorists have
attempted to distinguish the direct object of perceptual experience from the objects
of perception. In each case, this is putatively to be achieved on the basis of a certain
description of perceptual experience. This first chapter begins this discussion with
the sense-datum theory of perception; later I focus on early modern work in the
philosophy of perception, discussing more recent (circa. mid-to-late twentieth cen-
tury) work in the psychology of visual perception along the way. What all this work
has in common is an opposition to the direct realist conception of perceptual expe-
6 |
rience. This conception is widely considered to be the default opinion of those who
have not engaged in theoretical study of perception. Accordingly, the first section of
chapter 1 is dedicated to an outline of this default opinion. I make an initial clarifi-
cation of my stance on its putative pre-theoretical status. I then move on to intro-
duce the sense-datum theory, discussion of which proves useful in two ways. The
first is that the exercise of presenting the theory, in a manner that strips it of any
epistemological predilections, puts one in a better position to evaluate views that
characterise perceptual experience as indirect. The second is that close examination
of the metaphysical status of sense-data motivates a need to treat claims about the
direct of object of perceptual experience as ontologically neutral, i.e., in independ-
ence from any metaphysical determinations driven by a theory of perception.
Chapt er 2
Disputing perceptual experience
The second chapter begins to explore the question of whether or not perceptual
experience is indirect. Although the very idea might strike the reader as somewhat
perverse, my aim in this chapter and the next is to provide a fair trial for claims that
implicate such a description. I approach tentatively by extracting an account of per-
ceptual experience that does not hold it to be indirect per se, but impoverished with
respect to the spatial dimensions of the objects of perception. I consider two ways in
which this view can fail to find adequate justification on the very grounds on which
authors have attempted to provide justification. I then consider one way of conceiv-
ing of perceptual experience as indirect, found in the writings of the influential vis-
ual psychologist Irvin Rock. After showing the similar manner in which that account
also fails, I examine another strategy that receives more detailed discussion in the
next chapter.
Chapt er 3
Berkeley & the limits of sensation
In the final chapter of Part I, I further evaluate claims to the effect that perceptual
experience is indirect. I argue that Bishop George Berkeley held such a view, as re-
vealed in a study of his philosophy of perception. I begin by introducing a general
form of argument attributable to Berkeley, which I call a limitation argument. I then
show why Berkeley requires the use of limitation arguments to motivate his theory
of vision. Moreover, I examine three different ways in which he evidently used a
limitation argument to support the claim that we do not directly visually experience
distance. None of the proposals work because they exploit several misleading equiv-
ocations in their use of the notion of sensation. But Berkeleys failures are revealing,
for they enable further clarity on the ambiguous concept of sensation. In an attempt
| 7

to provide such clarity, I specify the manner in which a defendant of the common-
sense conception ought to treat the relationship between sensation and perceptual
experience.
Part I I
Chapt er 4
The neutral intentionality of perceptual experience
In the first chapter of Part II, I discuss the intentionality of perceptual experience
with a view to clarifying my neutral stance on the metaphysics of perceptual experi-
ence. Contemporary philosophers discussing perceptual experience are really dis-
cussing its metaphysical status, and typically this comes in discussion of the inten-
tionality of perceptual experience. I begin with an illustration of two instances of
the move that I want to avoid, moving on to discuss intentionality itself, giving brief
overviews of disputes over the history of intentionality and the problem of inten-
tionality as it is attributed to Franz Brentano. I then discuss ways in which inten-
tionality has been treated that are either inadequate in describing perceptual expe-
rience or inadequate in addressing the basic problem of intentionality. After that, I
move on to more contemporary discussions, specifically a research programme that
aims to solve the problem of intentionality by providing a naturalistic theory of rep-
resentational content. I end the chapter by distinguishing a neutral stance on inten-
tionality from any stance that involves adherence to a metaphysic of content, distin-
guishing the stance from several notions of phenomenology as I do so.
Chapt er 5
Seeming real
In the fifth chapter I move on to perhaps the most difficult task in this project, dis-
cerning the correct understanding of the claim that the objects of perceptual expe-
rience seem real. I attempt to convey such an understanding by critically reviewing
members of a family of concepts evoked in the literature, each of which has a bear-
ing on the claim. I begin with an outline of the deeper philosophical motivations of
direct realism, specifically with respect to the explanatory role that perceptual expe-
rience is supposed to play in fulfilling those motivations. The view requires a con-
ception of perceptual experience as an instance of a suis generis, simple relation to
the world, in order to maintain a simple connection between the subject and the
world she experiences. The most pressing and recurrent difficulty here is the as-
sumption that it is possible for a subject to be in states that are subjectively indistin-
guishable from perceptual experiences, in particular, the difficulty is that this might
serve to confound this connection. Whilst remaining impartial on the issue, I out-
8 |
line the assumptions about perceptual experience and its relation to other states
that gives rise to this difficulty, I then move on to provide a means of respecting
both the claimed suis generis nature of perceptual experience and the claimed im-
mediacy of perceptual experience, by distinguishing perceptual and theoretical be-
lief. In further efforts to distinguish description of perceptual experience from met-
aphysics, I clarify the implications of the claim that perceptual experience is trans-
parent. I then review several ways in which the objects of perceptual experience
have been described as seeming present, to conclude by arguing that, in the hands
of direct realists, perceptual presence is best construed as the claim that perceptual
experience seems to be of mind-independent particulars.
Chapt er 6
Husserl on the experience of particularity
In the last chapter of Part I, I offer a synthesis of the work in the first five chapters
of the thesis by drawing on the work of Edmund Husserl. The account presented
takes its initial motivation from the assumption that sensory experience is indeter-
minate and requires structure to yield perceptual experience. I begin by introducing
this point, noting that one of Husserls innovations is to claim that bodily experi-
ence provides such structure. A further innovation is the claim that the objects of
perceptual experience have a distinctive quality that he terms leibhaftige Gegeben-
heit. I state the manner in which I read the metaphor, as the experience of objects as
particulars in virtue of each exhibiting an extant unity of appearances. The core of
the chapter is an analysis of the sixteen inter-related claims that comprise the ac-
count. After presenting that, I contrast the account with a view recently developed
in the philosophy of perception by Alva No. Having hopefully thereby achieved a
fairly univocal understanding of the account, I extend its basic principles to a de-
scription of the experience of objects as mind-independent, by drawing upon and
sufficiently weakening concepts of objectivity found in the Oxonian tradition of the
mid-to-late twentieth century and the received Cartesian manner of distinguishing
between subjectivity and objectivity.
Part I I I
Chapt er 7
Structural affordances
The final chapter of Part II completes the description of the direct realist concep-
tion of perceptual experience that is a major aim of this thesis. A distinctive feature
of the account presented in that chapter is that kinaesthesia is described as provid-
ing a broadly continuous structure to perceptual sensations. Chapter 7, and the re-
| 9

maining two chapters of Part III, aim to pave a path towards the study of the natu-
ral basis of kinaesthesia. The chapter begins with a review of developments in neu-
rophysiology and psychology that suggest a general set of mechanisms that might
serve in a naturalistic explanation of kinaesthesia. In later sections, I provide a de-
tailed analysis of the concept of an affordance, in which I introduce the concept of a
structural affordance to provide a means for thinking about an agents relationship
to the constraints on its bodily movement. This is followed-up with a review of ways
in which agents actually deal with constraints upon bodily movement, in order to
show the extent to which structural affordances have been studied already. Struc-
tural affordances then serve as an element in a more general theory of the natural
basis of kinaesthetic experience. I introduce such an account by clarifying the ambi-
guities of my previous work on the spatial content of bodily experience, concluding
with a summary of how the theory can provide an explanation of the Husserlian
descriptions provided at the end of chapter 6, and other forms of kinaesthetic expe-
rience besides.
Chapt er 8
On bodily selfhood
Having laid out some basic ideas towards a naturalistic explanation of how kinaes-
thesia provides structure to the intentionality of perceptual experience, I then turn
to what might seem to be the elephant in the room. At no point in the previous
chapters do I describe kinaesthesia as a form of self-consciousness, but nevertheless
I do use the expression bodily subject, at several points. Although my use of this
term is essentially instrumental, the topic of bodily selfhood ought not to be dis-
missed entirely. Accordingly, I devote Chapter 8 to a discussion of what I under-
stand to be the core issues facing what I call the bodily self thesis, a positive meta-
physical thesis about the existence of bodily subjects qua bodily selves. I recount
famous arguments from the writings of Ren Descartes that are ostensibly against
the thesis, only to reach a stalemate. To facilitate further debate, I construct a slight-
ly more modern, Cartesian thesis that can be more straightforwardly addressed by
the bodily self thesis. After reviewing the core philosophical concerns of authors
who seek to defend the bodily self thesis, I draw upon Evans work in order to show
how they might establish a clear line of defence against the Cartesian. I then con-
struct a dilemma for defendants of the bodily self thesis, which leaves them with an
open challenge to specify the nature of the body that they claim to be the self. In
closing the chapter, I turn to the more general discussion of whether there is such a
thing as a self. My conclusion is that again the dispute is likely to remain unsettled,
welcoming the idea that the more fruitful topics of discussion are to be found else-
where.
10 |
Chapt er 9
Clearing the ground
In the final chapter, I lay out the issues I believe to be fruitful in pursuit of a natural-
istic explanation of how bodily experience might play a structuring role in perceptu-
al experience. The issues all cluster around a central dispute between two views on
body representation. These views are contrasted by their stance on the relationship
between properties of the body as a whole and representation of the body, where
one view posits a representation of the body as a whole, and the other does not.
Concomitantly, they are contrasted by their stance on whether body representation
is embodied (in a sense to be specifically stated). The impetus for distinguishing the
two views comes from recent work by Thomas Metzinger on the specific relation-
ship between representation of the body and the most minimal forms of subjectivi-
ty. The two views are developed as contrasting explanations of the body representa-
tional capacities involved. I go on to provide further means for distinguishing vari-
ants of each view and the differences between them. And I then argue for the virtues
of certain shared assumptions about what explanation implies in this context, and
certain shared constraints with respect to positing representations of and within a
dynamical physical process.
| 11

PART I

[If ] the ordinary man is making any mistake at all when he thinks that
he perceives physical objects [he] must be misinterpreting not just some
particular item but the general character of his perceptual experience.
But then what grounds do we have for thinking this is so?
(Ayer, 1973, p. 73)

12 |
Chapter 1 Sense-data
1. 0. Introduction
Most accounts of perceptual experience are presented in a way that grinds some
theoretical axe or another. I believe there is a reason for this: consideration of per-
ceptual experience itself has been a wellspring of distinctively philosophical prob-
lems concerning the metaphysical and epistemological status of perception. One of
the long-standing goals of the philosophical study of mind has been to reconcile the
presumed nature of perception with compelling arguments designed to force a de-
nial of that nature. And here the presumed nature of perception is simply that
which is supposed to seem pre-theoretically evident in perceptual experience.
This chapter and the next two chapters will survey some attempts to distinguish the
direct object of perceptual experience from the objects of perception, in each case
on the basis of a certain description of perceptual experience. This will involve dis-
cussion of several dated works in the study of perception. What these works all have
in common is their target, a conception of how perceptual experience is related to
reality, considered as the default opinion of those who have not engaged in theoreti-
cal study of perception. Accordingly, the first section of this chapter (1.1) is dedi-
cated to an outline of this default opinion (therein I make some initial clarification
of my stance on its putative pre-theoretical status). I then move on (in 1.2) to dis-
cuss the main topic of this chapter, the sense-datum theory of perception. The pur-
pose of discussing the theory is twofold. First, the majority of sense-datum theories
claimed that perception involved a mediatory relation or event; finding an appro-
priate notion of mediation and distinguishing this from the associated epistemolog-
ical consequences of many versions of the theory will put us in a better position to
construct and confront accounts that claim one ought to characterise perceptual
experience as indirect (see 1.3). Second, it will be instructive to see that close exam-
ination of the sense datum theory (in 1.4) in fact motivates (in 1.5) a need to treat
claims about the direct objects of perceptual experience in independence from met-
aphysical claims about the direct and/or ultimate objects of perception.
1.1 Common-sense, reality, and perceptual experience
1.2 Sense-data & the argument from perspectival variation
1.3 Indirect perception
1.4 The metaphysics of sense-data
1.5 The neutrality of the direct objects of perceptual experience
| 13

1. 1. Common-sense, reality, and perceptual experience
From the mouth of Socrates and the pen of Plato (388-367 BCE/1973) we are told
that perception is always of what is (152c, p. 16), and later that whenever I come to
be perceiving, I necessarily come to be perceiving something; because it's impossi-
ble to come to be perceiving, but not perceiving anything (160b, p. 29). In seeming
agreement, Thomas Reid writes in the 18
th
century that any man that perceives,
must perceive something (1785, Essay I, Chapter I, p. 20). And in the middle of the
century just passed, the same sentiment is expressed by a character of Wilfrid
Sellars invention when he exclaims that you cant see what isnt so! (1956, p. 270).
Admittedly, these quotes are taken out of context, and not all of them explicitly
speak of existence, and some of those that do might require supplementation to
constitute an affirmation of existence. But these wrinkles aside, they are sufficient
to suggest a commonly held ideal of perception:
Any instance of perception involves a relation to something that exists.
1

The ideal implies that if one perceives, then one perceives something that exists.
This is essentially a metaphysical statement. It states a fact about the nature of per-
ception as such. The ensuing metaphysical issues include how one treats the nature
of the relation and how one fills out the something that exists. Without taking the
trouble to address either, I will treat the concept of perception as in correspondence
with the ideal.
The ideal makes no claim about perceptual experience. More significantly, the ideal
is neutral on the validity of a conception of perceptual experience as
an immediate consciousness of the existence of things outside us (Strawson,
1979/2002, p. 97; latter emphasis mine). There are two aspects to this conception.
Both are crucial for understanding the basic issues of Part I and Part II. One as-
pect, the first, is the lack of mediation; perceptual experience seems direct. Second-
ly, and equally importantly, there is the commitment to reality in the experience
itself. The reality concerned is not (or at least not supposed to be) some theory of
what is real; indeed, thoughts about the metaphysical possibility of experiencing
reality need not be attributed to the subject. Rather this reality is simply the way in
which the objects of perceptual experience seem real; it is somewhat apposite that
the notion of reality P.F. Strawson appeals to is the rather unsophisticated existence
of things outside us. On this conception, perceptual experience is thought to reflect
a certain view of the world, as containing objects, variously propertied, located in a

1
I intentionally leave the nature of the relation opaque.
14 |
common space and continuing in their existence independently of our interrupted
and relatively fleeting perceptions of them (Strawson, 1979/2002, p. 94).
Strawsons remarks are an attempt to characterise a certain conception of reality that
is informed by the manner in which it seems to be experienced, which he calls
common-sense realism. He attempts to characterise a pre-theoretical notion of per-
ceiving which consists in an immediate awareness of things outside us (Strawson,
1979/2002, p. 99). This pre-theoretic notion is nevertheless closely associated with
both common-sense realism and a bona fide theory of perception known variously
as nave realism or direct realism. Occasionally these two labels (direct and nave
realism) refer to significantly different theories, and principally I will be concerned
with the latter.
2
For present purposes they highlight two related points of interest.
The first is that the direct realist theory of perception, in its compatibility with the
common-sense conception of reality, is intended to capture the sentiments of the
philosophically nave as regards their perceptual experience. It is assumed that, to
the philosophically nave, there just are objects properties and relations that exist
independently of our perceptions and that these are the things that we perceptually
experience. Strawson makes the point with passion:
I am talking of the ordinary non-philosophical man. I am talking of
us all before we felt, if ever we did feel, any inclination to respond to
the solicitations of a general scepticism, to regard it as raising a prob-
lem. I am saying that it follows from the character of sensible experi-
ence as we all actually enjoy it that a common-sense realist view of
the world does not in general have the status of a theory in respect of
that experience. (Strawson, 1979/2002, p. 95)
The second point of interest (highlighted by the name direct realism) indicates a
major point of contention: namely, whether this pre-theoretic conception of percep-
tual experience can either support (or be supported by) a theory that claims that
perceivers are directly related to reality in virtue of their perceptual experience. As the
claimed pre-theoretical conception of perceptual experience is essential to the
broader focus of the dissertation, I will say something more about where I stand on
its pre-theoretical status before moving on.

2
A difference sometimes urged between direct realism and nave realism is that pur-
portedly on the latter view we always and necessarily experience the world just as it is. By
contrast, direct realism is held to allow for the possibility of illusory perception, i.e. cases in
which a perceiver can genuinely perceive an object whilst experiencing it as having proper-
ties that it does not possess (A. D. Smith, 2002, pp. 21 - 23, 27 - 29). One question of interest
then, is whether direct realism is only attributable to subjects capable of distinguishing be-
tween their experiences and the objects of their experience (Mackie, 1976, pp. 67 - 68).
| 15

Firstly, I should make clear a point hinted at in the introduction: I will not be at-
tempting to claim that the perceptual experience of the ordinary non-philosophical
man (or woman, child, or creature for that matter) is one way rather than another.
Moreover, although my aim is to articulate what is often referred to as a pre-
theoretical conception of perceptual experience, I abrogate myself of any need to
remain true to any claims of its strictly pre-theoretical status. I will treat it as a theo-
retical description of perceptual experience in competition with others, competing
for the status of being the correct description of perceptual experience. My aim is
not to prove that it really is correct. My aim is to show that it has not been demon-
strated as false and that it is internally consistent, though occasionally I find it in
need of supplementation and theoretical apparatus that I will offer (in this regard,
see especially 3.4 and 6.3).
Secondly, Strawsons insistence is that the story he tells about the ordinary mans (or
woman's) perceptual experience is no more than a recapitulation of a story that man
(woman) must tell if he (she) is to remain true to the overall conceptual scheme
implicit in his (her) experience. His aim is to describe the metaphysical picture that
one is already committed to, if certain concepts are to even have the possibility of
application (cf. Macdonald, 2005, pp. 17 - 20; Strawson, 1959/2003, pp. 9 - 12). That
is not my aim. My aim is to describe the structure of perceptual experience that one
is already committed to in holding a certain conception of perceptual experience. In
Part 1, I will do so by critically engaging with the work of those who have developed
theories of perception that they believe provide cause to correct the direct realist
conception. In Part 2, I will do so by critically engaging with the work of those who
have sought to account for the direct realist conception by urging direct realism
proper, authors who offer a theory of the metaphysics and/or the natural grounding
of our capacity to think about the objective world in virtue of our perceptual experi-
ence. In all cases, the claims I will make of perceptual experience are of an entirely
conditional character. They are made in the service of articulating what I argue to be
the commitments of the direct realist conception, and can be thought of as an an-
swer to the question: If direct realism is true, how must perceptual experience be?
Nonetheless, from the outset I will be engaged in discussion of literature comprised
of claims and arguments that lack the conditional stance that I adopt. My hope is
that I am successful in presenting the discussion in such a way that the reader can
keep track of this difference.
16 |
1. 2. Sense-data & the argument from perspectival
variation
Strawson is not the only author to attribute the idea that perception is a direct rela-
tion to the ordinary non-philosophical man. Indeed, the following excerpt sug-
gests David Hume doing so with some confidence:
It seems evident that men are carried by a natural instinct or prepos-
session, to repose faith in their senses [] It seems also evident, that,
when men follow this blind and powerful instinct of nature, they al-
ways suppose the very images, presented by the sense, to be the ex-
ternal objects [] But this universal and primary opinion of men is
soon destroyed by the slightest philosophy. (1777/1975, XII, Part I,
pp. 151, 152)
Hume believes that this universal and primary opinion of men is soon destroyed
by the following line of reasoning:
The table, which we see, seems to diminish, as we remove further
from it: but the real table, which exists independent of us, suffers no
alteration: it was therefore nothing but its image, which was present
to the mind. (1777/1975, XII, Part I, p. 152)
But this slightest philosophy is a touch too slight as part of the argument has been
suppressed. The reader might recall an essentially similar but slightly more expand-
ed line of argument presented in the general introduction to this dissertation. For
convenience, and for the purposes of discussion, here is the passage in full:
We are all in the habit of judging as to the real shapes of things, and
we do this so unreflectingly that we come to think we actually see the
real shapes. But, in fact, as we all have to learn if we try to draw, a giv-
en thing looks different in shape from every different point of view. If
our table is really rectangular, it will look, from almost all points of
view, as if it had two acute angles and two obtuse angles. If opposite
sides are parallel, they will look as if they converged to a point away
from the spectator; if they are of equal length, they will look as if the
nearer side were longer. (Russell, 1912/1959, pp. 10 - 11)
A crucial step in the argument, suppressed in Hume's discussion, is the implication
that there is some conflict between the thought that we actually see the real
| 17

shapes of things and Russells suggestion that a given thing looks different in
shape from every different point of view (see Figure 1).
3

Figure 1 Perspectival views of a square

Charles Dunbar Broad identifies the proposed conflict more precisely, as a situation
in which two sets of properties apparently belong to the same object, and yet are
apparently incompatible with each other, where the apparent incompatibility is
between the apparent shapes and the supposed real shape, and between the change
in the appearances and the supposed constancy of the physical object (Broad, 1927,
pp. 85, 86). Russell, Broad and Hume (and many others) move from here to the
claim that this difference entails a distinction, i.e., that it is evident that the real
table, if there is one, is not the same as what we immediately experience by sight
(Russell, 1912/1959, p. 11). The reasoning implicitly appeals to Leibnizs law that if x
and y are identical then x and y cannot differ in any way; the assumption being that
this entails that if something is true of x and false of y, then x is not identical with y;
hence the distinction. Noting the appeal to Leibniz's law enables one to set out the
reasoning more plainly. The significance of demonstrating that the shapes that are
experienced differ from the shape of the table itself is in order to claim that these
shapes are attributable to something distinct from the table. But what that requires
is that there is something to which the shapes experienced are attributable in the

3
I leave aside for the moment the claim that in learning to draw we learn something
new about our visual experience (for which see the block quotation above). Just as the argu-
ment itself is not original to Russell, neither is the appeal to a speculative attitude assumed
in drawing and painting. This issue is taken up more fully in the next chapter, in 2.2.
A square viewed from different perspectives can look like a
trapezoid, a parallelogram, a rhombus, or some irregular
quadrilateral.

Reprinted from Palmer (1999, p. 327), with permission from
Massachusetts Institute of Technology 1999.
18 |
first place. For only then can it be said that it is true of something that it is a rhom-
bus, and true of something else (e.g. the table) that it is a square. The further as-
sumption here is what Howard Robinson calls the phenomenal principle:
If there sensibly appears to a subject to be something which possesses a
particular sensible quality then there is something of which the subject is
aware which does possess that sensible quality (1994, p. 32)
According to the phenomenal principle, if one experiences a table with the visible
shape of a rhombus, then there is something that one experiences which possesses
the property of rhomboid shape. If the principle is non-trivial, then it implies some
degree of reification with respect to the experience itself. Broad effectively states the
principle, and its presumed implication, in reference to another common example:
When I look at a penny from the side I am certainly aware of some-
thing; and it is certainly plausible to hold that this something is ellip-
tical in the same plain sense in which a suitably bent piece of wire,
looked at from straight above, is elliptical. (1927, p. 240)
The something is baptised by Broad as a sensum (1927, p. 240) and called by Rus-
sell (probably following the terminology of G. E. Moore) a sense-datum (1912/1959,
p. 12).
1. 3. Indirect perception
Sensa and sense-data are both signs of their time as much as Humes proposed im-
ages in the mind. Nevertheless, both have been posited as the necessary mediators
of perception. Thus, for instance, Hume claims that once the supposition that the
very images, presented by the sense [are] the external objects is demolished, it is
evident that the diminishing table is therefore nothing but its image [] present to
the mind (1777/1975, XII, Part I, pp. 151 - 152) Russell claims in summary that the
real table, if there is one, is not the same as what we immediately experience by
sight or touch or hearing and names as sense-data the things that are immediate-
ly known in sensation: such things as colours, sounds, smells, hardnesses, rough-
nesses, and so on (1912/1959, pp. 11, 12). Indeed, a number of influential figures in
early twentieth century British philosophy held that perception involves some form
of non-trivial mediation. And many held this view on the strength of the observa-
tion that careful description of the direct object of perceptual experience ought not
to refer to the real objects perceived; rather in describing the experience itself one
ought to only talk of sense-data, e.g., variously patterned colour patches, of particu-
lar size and two-dimensional shape (e.g. Broad, 1927; G. E. Moore, 1953; Russell,
| 19

1912/1959). These writers are collectively known today as the inaugurators of the
sense-datum theory (SDT) of perception. Although few accept the tenets of SDT
today, the view is notable as it forms a clear effort (though perhaps not the first ef-
fort) to systematically distinguish the direct object of perceptual experience from
the ultimate objects of perceptually based judgement. It is this focus on the direct
object of perceptual experience that makes SDT of interest, in particular where the
perceivers relation to the ultimate objects of perceptual judgement is thought to be
composite, qua mediated by the direct object of perceptual experience.
Sometimes advocates of SDT describe the mediatory process itself as an inference.
For instance, at one point Russell claims that the real table if there is one, is not
immediately known to us at all, but must be an inference from what is immediately
known (1912/1959, p. 11). The pressing issue here is that of the epistemological rela-
tion between the perceiver, the sense-datum, and the real table (or any other object
the perceptual judgement takes). In this connection, SDT is infamously associated
with an indefensible theory of knowledge, known as foundationalism. Briefly, there
are two classes of beliefs in a standard foundationalist system: one class consists of
beliefs that are self-justifying; another class consists of beliefs which gain justifica-
tion by recursive inferential links to the self-justified class. The canonical motiva-
tions for a foundationalist epistemology are the sceptical scenarios that frame Ren
Descartes method in his renowned Meditations (Descartes, 1642/1984). Perhaps we
can always go wrong about how things are beyond our immediate experience, but
(the thought continues) we cannot go wrong in our beliefs about how things seem
to us. Correspondingly, it was often thought that beliefs about sense-data might
thus serve as the constituents of a self-justifying class of beliefs, sometimes called
the given. And here lies the point of tension. For on a favoured conceptual analysis
of knowledge, knowledge involves (at least) a relation to a proposition: sense-data
are not propositions (as we will see in the next section, most likely, they are particu-
lars); beliefs about sense-data seem to be beliefs about the wrong kind of thing to
serve as the basis of knowledge (Roderick M. Chisholm, 1982; M. Williams, 2001, pp.
94 - 102).
Whether or not sense-data can be theoretically useful posits in a viable epistemolo-
gy, it is notable that many proponents of SDT nevertheless believe that reality is the
ultimate object of perceptual judgement. Indeed, positing sense-data need not nec-
essarily divorce genuine perception from the world. If one assumes, for instance,
that perception is often veridical precisely because a regular causal relationship
holds between sense-data and objects in the local environment, one is not then re-
quired to eschew the reality of what is perceived. And one could still maintain that
perception of reality is only ever indirect: in the first place what the subject per-
ceives is something else than the object; the subject perceives something causally
20 |
related to the object. Any view that holds that reality is perceptible, but only ever
indirectly perceived is aptly dubbed a form of indirect realism.
4
By contrast, any view
which holds that reality is directly perceived is a form of direct realism (hereafter,
DR). Direct perception is often defined negatively in contradistinction to indirect
perception (see, e.g., Pappas, 1989, pp. 156 - 157). One means of defining it thus is by
adopting Jacksons schema (adapted from Jackson, 1977, pp. 19 - 20):
S indirectly perceives x if and only if S perceives x in virtue of perceiving y,
and x and y are distinct.

S directly perceives x if and only if S perceives x, and S does not indirectly
perceive x
Jacksons schema is useful in that it provides a framework for discussing the issue of
whether or not perception is direct, whilst avoiding the unhelpful connotations of
mediation by inference (Jackson, 1977, pp. 7 - 11).
5
For Jackson, the issue of whether

4
SDT need not necessarily be formulated as committed to realism. If, for instance,
one holds not only that the objects of experience possess the properties that they are per-
ceived to have, but also that no objects have a distinct existence from their perception (per-
haps broadly construed), then one moves towards a form of idealism such as that associated
with Bishop George Berkeley (see Berkeley, 1710/2008, Part 1, 3, p. 84, for the canonical
textual support). Alternatively, one could hold that the existence of the objects perceived
depends upon their possible perception; thus one arrives at a modified form of idealism,
known as phenomenalism (see Roderick M. Chisholm, 1948 for critical analysis). Despite the
fact that idealists ground the existence of external objects in a peculiar manner, existence
beyond the subject is nevertheless vouchsafed. In the typical case they posit a Judeo-
Christian deity as an ultimate in their ontology (see Foster, 2008, pp. 220- 236 for discussion
of both Berkeley's views and Foster's own phenomenalism in this regard). So at a certain
granularity, a broad similarity between all these views is that an antecedently fixed fact
[either of the mind of God, or of reality otherwise conceived] determines truth and falsity in
the world as it presents itself (Lewis, 1955, p. 239).
5
Robert Schwartz provides a more extended deflationary treatment of debates over
the immediacy of perception framed by the issue of whether visual perception involves in-
ference (1994, pp. 84 - 124). Perhaps somewhat embarrassingly, he finds no less than five
different disputes in the philosophical and psychological literature that appeal to the notion
of inference. In light of these crosscutting debates he claims the question of whether vision
depends on inference is multiply ambiguous, largely because claims that vision does in-
volve inference are often vague and/or devoid of empirical content and accordingly he con-
cludes that no single yes or no answer to the question of visual inference is likely to
prove satisfactory (R. Schwartz, 1994, p. 111). For an alternative discussion of the various
senses of direct perception (not all of which appeal to the notion of inference) in contem-
porary and historical debates, which concludes far more optimistically, see Snowdon (1992).
| 21

and how one knows x on the basis of perception is orthogonal to determining the
nature of perception. As he puts it: If I see the friendly-looking dog and the friend-
ly-looking dog is about to attack me, then I see the dog who is about to attack me,
whether or not I am fortunate enough to know the fact (1977, p. 5). I am inclined to
agree. But this does leave the question of what the expression in virtue of means in
Jacksons schema. Jacksons suggestion is that the expression in virtue of is favour-
able because it expresses an asymmetric relation, according to which the direct ob-
ject of perception can be construed as having a primary status. True to his preferred
methodology, Jackson proposes that the asymmetry of the relation is best explicated
as a particular instance of analysability. Hence, one can perceive x in virtue of per-
ceiving y, but not vice versa, because perceiving x is analysable in terms of perceiv-
ing y, but not vice versa.
6
However, as Thomas Baldwin (1990, pp. 240 - 241) and Jos
Luis Bermdez (2000, pp. 356 - 357) note, the concept of indirect perception re-
quires a kind of asymmetry that captures the sense in which indirect perception is
dependent upon direct perception. And if that is right then an alternative is needed,
for there seem to be dependency relations that are not captured by analysability
(e.g., the relation between education and career path). Baldwin opts for explanation
as the genus, of which the in virtue of relation is a species (op. cit., 241). But given
famous cases of non-explanatory dependency, such as supervenience (for a review
see Horgan, 1993), Bermdez opts for the more general and opaque notion of "ob-
jective dependency" (op. cit., 357).
As general and opaque as it is, the idea that perceiving x is objectively dependent
upon perceiving y, but not vice versa, gives a lot of leeway that can account for a
variety of theoretical stances. It can encompass both Jacksons and Baldwins formu-
lations: the objective dependency relation between x and y might be such that per-
ception of x is definable in terms of perception of y; or it might be that perception
of x is explained by perception of y. Indeed, perhaps it can encompass any theoreti-
cal stance on indirect perception whatsoever, wherever an objective dependency
relation (or its lack) can be established between perception of x and perception of y.
SDT included: perception of x is objectively dependent upon perception of y, where
y is a sense-datum and y is the direct object of perceptual experience.

6
More fully:
An A is F in virtue of a B being F if the application of _ is F to an A is definable in terms of its
application to a B and a relation, R, between As and Bs, but not conversely. This gives us an
account for the indefinite case. We obtain an account for the definite case as follows: This A is
F in virtue of this B being F if (i) an A is F is true in virtue of a B being F (as just defined) (ii)
this A and this B are F, and (iii) this A and this B bear R to each other. (Jackson, 1977, p. 18)
This can be filled out mutatis mutandis for all adjustments of the schema discussed.
22 |
1. 4. The metaphysics of sense-data
Although it may be simple enough to establish a relatively uncontroversial frame-
work for SDT as an account of indirect perception, a key question remains: What are
these sense-data that the theory posits? In this section and the next, I will argue
that the abject failure of SDT in providing a satisfactory answer to this question re-
veals an inherent instability in equating sense-data with the proposed direct object
of perception. A moral that I will draw from the discussion is that sense-data are
better thought of as ontologically neutral posits that serve as place-holders for a
commonly agreed structural aspect of perceptual experience: namely, that percep-
tual experience has a direct object.
In the examples of perspectival variation, visual sensory properties are held to be
distinct from the actual properties of the object seen. The table is experienced as
trapezoidal though it is not; it looks smaller than it did when standing three feet
closer, yet its size remains the same. The properties of sense-data seem so closely
tied to sensory experience that they might be thought of as sensations. But the ma-
jor proponents of the theory showed clear resistance to identifying sense-data with
sensations. Hence, we find Russell offering some terminological legislation:
We shall give the name sensation to the experience of being imme-
diately aware of [things]. Thus, whenever we see a colour, we have a
sensation of the colour, but the colour itself is a sense-datum, not a
sensation. The colour is that of which we are immediately aware, and
the awareness itself is the sensation. (1912/1959, p. 12)
This legislation provides an initial indication of what sense-data are supposed to be,
and reveals a significant theoretical move made by certain proponents of SDT.
Sense-data are the objects of sensation, where these latter are understood as the
objects of sensory experience. The significant theoretical move is then to identify
the direct objects of perception with the objects of sensory experience, by means of,
e.g., the argument from perspectival variation presented in 1.2.
This move rests upon what has come to be known as an act/object metaphysics of
sensation. In this Russell is almost certainly following G. E. Moore, who urges a dis-
tinction between (indisputably mental) acts of consciousness from their objects
(whose mental or non-mental status could be disputed):
| 23

[W]hatever be its nature, the entity which is experienced must in all
cases be distinguished from the fact or event which consists in its be-
ing experienced; since by saying that it is experienced we mean that it
has a relation of a certain kind to something else [viz., the act of ex-
perience]. (1922, p. 169 emphases mine; see also 1903, p. 444)
Hence, both Moore and Russell (the latter, at least, at one point in his career) see
the need for a distinction between something being experienced and the something
which is experienced.
7
This distinction is deemed necessary to unmask a conflation
of arguably distinct relata, a conflation that might occur in employing the blanket
term sensation (see A. D. Smith, 2002, pp. 54 - 60 for discussion).
8
According to
this act/object metaphysics, sensations and their cognates are facts or events that
are not to be reified. They are certainly not the objects of sensory experience. Thus,
if sense-data are meant to be the objects of sensory experience, then sense-data are
not sensations.
But this only clarifies that sense-data, qua objects of sensory experience, are not
sensations. As yet there is no positive indication of their ontological status beyond
this negative claim. If one were looking for some positive indication of what it is
that sense-data are, such that the direct objects of experience can be identified as
sense-data, then one would be still be left wanting. Another (perhaps obvious) con-
strual is to regard sense-data as entities; and given that the properties of sense-data
seem so closely tied to the properties of experience, they might naturally be con-
ceived as mind-dependent entities. But this is certainly not the only conception of
their ontological status entertained by those who fostered the theory. Broad, for
instance, seems unconvinced by this simple inference. He writes that the objects he
terms sensa should be thought of as particular existents of a peculiar kind, peculi-
ar in so far as they are not physical and on his view there is no reason to suppose
that they are either states of mind or existentially mind-dependent (Broad, 1925, p.

7
Russell later relaxed his stance:
According to some authorsamong whom I was formerly includedit is necessary to distin-
guish between a sensation, which is a mental event, and its object, which is a patch of colour
or a noise or what not. If this distinction is made, the object of the sensation is called a
sense-datum or a sensible object. [] If it is not valid, the sensation and the sense-datum
are identical [] I have come to regard the distinction as not valid. (1914, p. 83)
Suffice to say, this hardly helps in discerning the metaphysical status of sense-data.
8
Russell famously expressed this relation as one of acquaintance: the relata being a
subject and object which need not have any community of nature. The subject is mental
and the object is not known to be mental except under introspection (1913/1984, p. 5; as
cited in A.D. Smith, 2002, p. 55); see also, Russell (1910-1911).
24 |
181). Though, he also notes that if one takes the phenomenal principle as support for
the existence of sense-data, then one seems logically bound (at least) to endorse
their qualitative mind-dependence, in that the properties that qualify a particular
sense-datum would seem to depend upon the properties of a particular experience
(Broad, 1927, pp. 107 - 108).
Let us recap on the somewhat mixed results here. Sense-data are apparently par-
ticular entities, but they seem to be neither mind-dependent, nor obviously mind-
independent. Even if one is clear that they should be construed as one or the other,
it is not then entirely clear which. Perhaps this lack of a clear ontological status for
the centre-piece of the theory is its deepest trouble. Take the case of Moore, perhaps
the first to employ the term sense-data, in his winter lectures of 1910 - 1911
(published as G. E. Moore, 1953). In extensive writings throughout his career, Moore
engaged himself in a complex and largely inconclusive dialectic, over whether and
how sense-data ought to be construed as (anything other than) parts of the surfaces
of mind-independent objects, and the consequences this might have for DR
(Baldwin, 1990, pp. 233 - 250). Indeed, Thomas Baldwin speaks of Moores utter
failure to develop his sense-datum theory into a satisfactory philosophy of percep-
tion (ibid., p. 261). In his eyes, Moores struggle seems maddeningly sisyphean:
[H]e found himself trapped within a conceptual treadmill which
condemned him to move from direct realism, to indirect realism and
thence to phenomenalism and then back again to direct realism and
on through the cycle, without any sign of a synthesis that would ena-
ble him to aufheben any stage of this dialectic. (ibid., p. 261) (ibid., p.
261)
1. 5. The neutrality of the direct objects of perceptual
experience
To say the least, these oscillations indicate that the metaphysical implications of
positing sense-data are unclear. Indeed, there is an inherent ambiguity in the argu-
ments with which they are associated:
| 25

These notorious arguments sometimes play a categorical, and some-
times a hypothetical role. Sometimes they are employed to show that
there are sense-data, and sometimes to show that if there are sense-
data then they are not identical with the surfaces of material objects.
(Jackson, 1977, p. 107)
As Jackson (1977, pp. 108 - 109) notes, this distinction is not always observed either
by advocates or dissenters of the view. But it ought to be observed, as it enables the
distinction of the contentious claims of sense-datum theorists from their mere
speculations.
The contentious claims stem from the phenomenal principle. Here it is once again:
aware which does possess that sensible quality (Robinson, 1994, p. 32)
In the context of the argument from perspectival variation, the phenomenal princi-
ple functions as a sense-datum inference. For once it is accepted, the question of
whether or not sense-data exist has already been given a positive answer (A. D.
Smith, 2002, p. 25). Accordingly, in making this inference one might be accused of
what Place calls (in a slightly different context) the phenomenological fallacy:
[] the mistake of supposing that when the subject describes his ex-
perience, when he describes how things look, sound, smell, taste, or
feel to him, he is describing the literal properties of objects and
events on a peculiar sort of [...] phenomenal field. (1956, p. 49)
9

Place is implicitly rejecting the consequent of the sense-datum inference. Perhaps
the most basic defence of the inference is stated simply by Broad, when he writes of
seeing a penny on his desk: If, in fact, nothing elliptical is before my mind, it is very
hard to see why the penny should seem elliptical rather than some other shape
(Broad, 1927, p. 240). The question of whether this presumed difficulty supports an
ontological conclusion (and which) is a vexed issue, one which SDT ought to ad-
dress; in the last section, we saw just how much trouble the proponents of SDT had
in addressing it.

9
A similar worry is that the something in the phenomenal principle is an instance
of what William Lycan calls the Banana peel: the tendency (of closet anti-materialists) to
tacitly describe sensory experiences as objects, rather than as events (1987, p. 17 ff.). If this
the case, one ought to be more frank and admit that in claiming that there are sense-data
one is claiming that there bloody well are phenomenal individuals (Lycan, 1990, p. 115).
26 |
For some, however, the historical lesson is that sense-data ought to hold a more
speculative status.
10
This would be to take the following remarks entirely seriously:
Thus the term sense-datum is meant to be a neutral term. The use of
it does not imply the acceptance of any particular theory. The term is
meant to stand in for [] something from which all theories of per-
ception ought to start, however much they diverge later. (Price, 1932,
p. 19)
11

Consider, in this regard, Moores attempt at a theory-neutral statement of sense-
data. First, he asks the reader to look at his [sic.] own right hand, and suggests that
at minimum he will be able to pick out something (G. E. Moore, 1925/1960, p.
276). Now with regard to that something he runs through two possibilities. One is
the possibility that it is a part of the surface of the hand, saying of a readers posi-
tive judgement of this, that it is a natural view to take that that thing is identical,
not, indeed, with his whole right hand, but with that part of his surface which he is
actually seeing (ibid., pp. 276 - 277). Another possibility is that sense-data are
something else, though he admits that they are only likely to be judged as anything
else on a little reflection. Perhaps, in considering the phenomena of perspectival
variation, Moores reader would be able to see that it is doubtful whether it can be
identical with the part of the surface of his hand in question (ibid., p. 277).
But this is not quite theory-neutral enough, as it masks what is essentially just a
further specification of the claim at the heart of the act/object metaphysics of sen-
sation, that an act of experience stands in a relation to its object (Baldwin, 1990, pp.
246 - 250; Crane, 2000, pp. 174 - 177). What ought to be left open is whether there is,
in any resolutely metaphysical sense, an object of sensory experience. It is not
enough to leave it an open question as to whether or not a sense-datum might be a
part of a seen hand or something else; for that begs the question as to whether a
sense-datum is really something at all, i.e. whether or not sense-data exist. Firths
gloss clearly illustrates this residual tendency to pre-emptively reify sense-data:

10
See, e.g., Tim Crane (2000, pp. 172 - 177).
11
A part of the text that I have omitted runs as follows: The term is meant to stand
for something whose existence is indubitable (however fleeting) (Price, 1932, p. 19). If the
line of argument pursued in the next few pages is correct, then Price ought to have omitted
this passage as well.
| 27

[Sense-data] are what we feel, sense, intuit, or immediately or ob-
serve, they are what is given to us, or what we are directly aware of, in
perception. And once we understand the meaning of sense-data as so
defined, we can presumably decide [...] just what kinds of entities are
properly called sense-data (Firth, 1949, p. 437)
What is exactly wrong about this, qua theory-neutral conception, is that it simply
does not hedge enough on the existence of sense-data. It is still prematurely as-
sumes that they are entities of some kind, and this should be a claim established
by a substantive theory.
Moore is a little closer to the mark later on in the passage lately cited, but makes the
same familiar mistake. Still speaking of the man asked to look at his hand, he ex-
plains:
Things of the sort (in a certain respect) of which this thing is, which
he sees in looking at his hand, and with regard to which he can un-
derstand how some philosophers should have supposed it to be the
part of the surface of his hand which he is seeing, while others have
supposed that it cant be, are what I mean by sense-data. (op. cit., p.
277, italics removed)
Although again there is a reificatory bias, I think Moore comes closer to expressing
something correct: whatever sense-data might be, they ought to be whatever is neu-
tral between the two possibilities he describes for the objects of direct perception.
However, these possibilities should not be thought as jointly exhaustive; the range
of possibilities is as massively disjunctive as there are theories of perception with
differing metaphysical commitments. Indeed, according to some, the bare nature of
sensation does not force one to posit any relation to an object. A clear instance of
this is the adverbial theory of perception. Adverbialists deny that it is a necessary
truth that, when a subject experiences certain properties, any object (mind-
dependent or independent) possesses the properties in question. On their view, the
properties of sensations themselves are modifications of ones sensory experience.
The idiom they suggest to capture this is obtained by transforming any construction
such as a appears f into a passive form such as being appeared to f-ly. Hence, an
adverbialist about visual sensation would hold that one does not have a visual sensa-
tion of something square; one senses squarely (R. M. Chisholm, 1957).
12
This is es-

12
In scaling up to a theory of perception, adverbialism famously has trouble with
grouping properties; in particular, it has trouble with the instantiation of properties that are
incompatible, and thus required to be separately grouped in one differentiated perceptual
event (Jackson, 1975). I leave these issues aside, as the point is simply to identify a theoreti-
cal option that does not analyse sensations as necessarily taking an object.
28 |
sentially a denial of the act/object analysis of sensation characteristic of much of the
work on SDT. More to the point, if sense-data are intended to be neutral between
theories of perception, then positing sense-data must not rule out a theory of per-
ception that employs an adverbial analysis of sensation. In order for that to be en-
sured, sense-data cannot be the objects of sensation; indeed, they cannot be objects
at all in any metaphysical sense. Under this guise, the struggles of the SDT do not
stem from the difficulty of moving from an entirely innocent premise to a contro-
versial conclusion (viz. the existence of sense-data). The troubles stem from finding
the correct characterisation of the innocent premise itself (viz. the pre-theoretical
status of sense-data).
To taking stock once again: Either sense-data are the categorical posits of SDT, in
which case their ontological status must be clarified (and perhaps cannot be). Or
sense-data are theory-neutral posits from which discussions of perception ought to
be begun, in which case their ontological status cannot be prejudged. In closing, I
suggest grabbing the second horn of the emerging dilemma. A natural way in which
sense-data might serve a theory-neutral role is by simply being a place-holder for
the objects of perceptual experience, where their status as objects, viz. their inclu-
sion in a particular ontology, remains unspecified and essentially neutral. Neverthe-
less, the claim that sensory experience qua perceptual experience ought to be de-
scribed as having an act/object structure would be consistent with a limited vindica-
tion of the act/object analysis of sensation. This would be to claim that perceptual
experience ought to be described as an experience of being related to something. A
commitment to this basic structure need not be accompanied by a commitment to a
specification of what this something is; an open question can remain as to whether
it is indeed really something at all in a metaphysically weighty sense. The appeal of
this stance on the objects of perceptual experience is that it is, as it were, a free as-
sumption. It can essentially be accepted by any theory of perception, and thus it
enables a discussion of perceptual experience that remains neutral on theories of
perception. However, a remaining point of interest and contention is that SDT
claims that, under the correct description, the (neutral) objects of perceptual expe-
rience differ in significant ways from the objects that are putatively perceived. In-
deed, this is the point from which the discussion began, and it is the point at which
the discussion will continue in the next chapter.
| 29

Chapter 2 Disputing perceptual
experience
2. 0. Introduction
A positive result that I take from the last chapter is the claim that the direct object
of perceptual experience ought to be construed as being ontologically neutral. This
clears the way for us to address controversies over the correct description of the di-
rect objects of perceptual experience, whilst leaving open their possible distinction
from the (direct or indirect) objects of perception.
One question that I will address in this chapter and the next is whether perceptual
experience involves a dependency of the kind posited by accounts of indirect per-
ception. That is, I will address the question of whether perceptual experience seems
indirect. Given the way in which indirect perception was defined in the previous
chapter, the very idea might strike the reader as somewhat perverse. It would in-
volve a rather odd iteration: one perceptually experiences x in virtue of perceptually
experiencing y; or so the story would go. However, my aim in this chapter (and in-
deed the next) is to provide a fair trial for claims such as these. In order to do so, I
will approach (in 2.1) by extracting an account of perceptual experience, not as
indirect, but as impoverished with respect to the spatial dimensions of the objects
of perception. Only after considering two ways in which this view can fail to find
adequate justification (in 2.2 & 2.3) will I consider ways of conceiving of percep-
tual experience as indirect per se. The first of these is found in the writings of the
influential visual psychologist Irvin Rock (2.4). After showing the ways in which
that account also fails, I will examine a second strategy (in 2.5) that will receive
more detailed discussion in the next chapter.
2.1 The planar projection view
2.2 The appeal to depiction
2.3 The visual field
2.4 Indirect perceptual experience
2.5 Prosthetic perception
30 |
2. 1. The planar projection view
Sense-datum theorists held that careful description of our visual experience ought
to be rather austere, involving only talk of shaped, patterned, coloured patches (see
e.g. Firth, 1949, p. 438). Characteristically, Moore holds up an envelope in one of his
lectures, and says with confidence afterwards that whilst all must agree that they
saw an envelope, there is a sense in which they did not all see the same thing:
Those of you on that side of the room will have seen a rhomboidal figure, while
those in front of me will have seen a figure more nearly rectangular (G. E. Moore,
1953, p. 33). And he volunteers a description of his own sense-data during the exhi-
bition: I saw a whitish patch of colour, of a particular size and shape (p. 32).
One can reconstruct the claims of the previous chapter in the same vein. For in-
stance, the sense-datum theory (SDT) claims when we indirectly perceive a coin, the
neutral description of our perceptual experience is of an elliptical patch; we visually
perceive a disk (the coin) in virtue of visually experiencing an elliptical shape. There
is an obvious relation between these two shapes, as Eric Schwitzgebel points out:
There are several ways to transform [the outline of a coin] into an el-
lipse, but the most natural in this context seems to be to project it
obliquely onto a two-dimensional plane presumably a plane per-
pendicular to the line of sight [...] The coin looks elliptical or has an
elliptical apparent shape because projecting it along the line of
sight onto a plane perpendicular to that line produces an elliptical
figure. (Schwitzgebel, 2006, p. 591)
Talk of projection can be taken as a device for picking out a claim about what we
directly experience in a given sense modality. The projections are planar in that they
retain only two dimensions of the three attributable to the objects (indirectly) per-
ceived. And they retain spatial organisation along only these dimensions, forming a
two-dimensional field or manifold of sensory properties (Firth, 1949, p. 436, also see
Figure 2). Assuming that these considerations have application in a sensory mode,
they would generalise to the perceptual experience involved in perceiving three-
dimensional spatial properties within that mode.
I will compress the above remarks into a simple claim that I will refer to as the pla-
nar projection view (PPV):
A subject indirectly perceives three-dimensional objects by perceptually ex-
periencing properties arranged along only two spatial dimensions.
| 31

Figure 2 Planar projection in vision

Several authors either reject PPV explicitly or would certainly be liable to do so. For
instance, when Strawson discusses the ordinary non-philosophical man (see 1.1)
he is denying the premise that a careful yet pre-theoretic description of perceptual
experience yields anything less than reference to real objects (commenting on Ayer,
1973; Strawson, 1979/2002, pp. 93 - 97). U.T. Place makes a similar point:
We describe our conscious experience not in terms of the mythologi-
cal phenomenal properties which are supposed to inhere in the
mythological objects in the mythological phenomenal field, but by
reference to the actual physical properties of the concrete physical
objects, events and processes which normally [...] give rise to the sort
of conscious experience which we are trying to describe (Place, 1956,
p. 49)
Others outright reject that the description, austere or not, tracks anything like our
experience. Gilbert Ryle, in setting up a discussion of claims now familiar, makes a
parenthetical aside, (though round plates, however steeply tilted, do not usually
look elliptical) as if reminding the reader that he is cognisant of the oddness of
the description in that regard (1949/2000, pp. 205 - 206). Similarly, A.D. Smith re-
marks sternly that the suggestion that pennies, for example, look elliptical when
seen from most angles is simply not true they look round (2002, p. 172). And
Schwitzgebel, an indefatigable sceptic, expresses unsurprising difficulty:
P & Q possess spatial properties in three dimensions.
P & Q possess the same spatial properties in two of these
dimensions, arranged according to a plane S.
Adapted from OShaughnessy (1980/2008, p. 209), 2008
Brian OShaughnessy.
32 |
For what its worth, as I stare at the penny now, Im inclined to say it
looks just plain circular, in a three dimensional space not elliptical
at all, in any sense or by any effort I can muster [...] I discern no ellip-
tical apparent shape. (Schwitzgebel, 2006, p. 590)
But the scepticism can run both ways. One can equally balk at the claim that the
immediate object of visual sensory experience is anything more than a planar pro-
jection. The authors cited above are troubled by the suggestion that the immediate
object of sensory experience can even seem to be less than what they intuitively
claim to perceive, or indeed, less than what they attribute explicitly to the common-
sense conception of perceptual experience. In violating this intuition, the most nat-
ural way of understanding the PPV is as corrective: ordinarily, perceptual experience
is taken to be of three-dimensional objects; but (perhaps) one ought to take it to be
of properties arranged along two dimensions.
2. 2. The appeal to depiction
As regards the spatiality of sight, PPV has an astonishing pedigree.
13
Most famous
amongst its progenitors are the British Empiricists. As a first instance, consider the
words of John Locke:
When we set before our eyes a round globe, of any uniform colour,
v.g. gold, alabaster, or jet, tis certain, that the idea thereby imprinted
in our mind, is of a flat circle variously shadowed, with several de-
grees of light and brightness coming to our eyes... [T]he idea we re-
ceive from thence, is only a plain [sic.] variously coloured, as is evi-
dent in painting (Locke, 1690/1997, Book II, Chapter IX, 8, pp. 143 -
144)
Lockes reference to painting as evidence is conspicuous. In a similar vein, the
reader might also remember Russells remark that as we all have to learn if we try to
draw, a given thing looks different in shape from every different point of view
(1912/1959, p. 10). An initial response, offered by Strawson, is that [the] frame of
mind in which we enjoy, if we ever do enjoy, this kind of experience is a rare and
sophisticated, not a standard or normal, frame of mind (1979/2002, p. 99). Howev-
er, nothing in Lockes remarks suggest he should be read as describing the mere
possibility of taking the painterly attitude. Rather he seems to be a claim about what
is actually experienced even when such an attitude is not taken, whenever we set

13
For a compact yet comprehensive critical and historical survey see A.D. Smith (2000,
pp. 481 - 482, 486 - 493).
| 33

something before our eyes so to speak. If so, then Locke would be one of the phi-
losophers Hume speaks of when he writes that it is commonly allowed by philoso-
phers that all bodies which discover themselves to the eye, appear as if painted on a
plain surface (Hume, 1739 - 1740/1960, Book I, Part II, 5, p. 56, emphasis mine).
Interpretive issues aside, Humes reference to a painted appearance naturally invites
the question: what support could the appeal to painting provide to the idea that
visual experience is two-dimensional?
First a brief point of clarification. There is a natural distinction to be made between
two senses of distance: absolute distance and relative distance (Grush, 2007a, p.
426). When I say that the cup is about half a meter away from my hand I am talking
about absolute distance. When I say that the handle of the cup is about five centi-
metres further away I am talking about relative distance. If I say that an object is at
some absolute distance, and upwards to the right of my hand, I imply that the ob-
ject is embedded in a three-dimensional space. If I say that the handle of the cup is
further away than its facing surface, I imply that the object embodies three-
dimensions. In more common parlance, absolute distance is simply distance and
relative distance is simply depth; both of these are tied to the third spatial dimen-
sion ruled out by the PPV.
The distinction clarifies that Locke and Hume, if they are indeed making claims
about perceptual experience when they speak of the objects before our eyes or
discovered to the eye, then they are claiming that the objects are experienced as
having no depth. But this leaves open the question of whether they are experienced
as being at some (absolute) distance away. In short, merely denying the experience
of relative distance would not amount to support for PPV, as it does not rule out the
third-dimension entirely (A. D. Smith, 2000, p. 487).
A more comprehensive denial is indicated by Bishop George Berkeleys remarks:
Suppose, for example, that looking at the moon I should say it were
fifty or sixty semidiameters of the earth distant from me. Let us see
what moon this is spoken of. It is plain it cannot be the visible moon,
or anything like the visible moon, or that which I see, which is only a
round, luminous plane of about thirty visible points in diameter.
(Berkeley, 1732/2008, 44, p. 21)
In effect, Berkeley makes two claims here: the visible moon that he sees is only a
round, luminous plane; and this visible moon cannot be identical with the moon
which he might judge to be a certain distance away from the earth. The first thing to
note is that there is no argument here. Berkeley does offer arguments for claims
such as these, but elsewhere; I will examine them in depth in the next chapter. A
second point (more to the present issue) is that any appeal to painted media in sup-
34 |
port of this more comprehensive denial cannot provide exclusive support. For cer-
tain artistic techniques have precisely the goal (and perhaps the effect) of producing
appearances of depth. Consider encountering an expertly drawn trompe-lil, such
as J.D. Hillberrys piece, A String of Memories (Figure 3). The expression trompe-lil
conveys an aim of the piece. It is intended to be indiscernible by sight alone that
certain aspects of the picture, such as the curl of the paper, the hanging string etc.,
do not have the depth that they would if they were not mere depictions.
14

Figure 3 A String of Memories by J.D. Hillberry

Accordingly, an appeal to painting cannot provide distinctive support for either a
two-dimensional or a three-dimensional account of visual experience. For there
would be no more call to postulate an immediate level of awareness of two dimen-
sional objects when seeing a trompe-lil, than there would be if the direct objects
of visual experience really seemed three dimensional, since the world, by hypothe-
sis, looks the same in the two cases (A. D. Smith, 2000, p. 487 emphasis mine).
The issue can be usefully thought of this way. A proponent of PPV might claim that
all visual experience involves a trick of the eye. Here the appeal to works like A
String of Memories might serve as a valid illustration. But a dissenter, who claims
that visual experience genuinely involves the experience of depth and distance, can

14
Or at least, to approach such a state of affairs, see 5.1 for further discussion.
In charcoal and pencil.
Adapted from:
http://www.jdhillberry.com/string_of_memories.htm
2008 J.D. Hillberry.
| 35

(equally validly) appeal to the very practice of trompe-lil as an attempt to mimic
ordinary visual experience of objects, in which the objects do seem to have depth
and seem to be some distance away. If the appeal provides support either way, it can
hardly provide conclusive support.
2. 3. The visual field
One way of interpreting the appeal to depiction is as a metaphorical description of
the visual field (VF). In his famous work, The Analysis of Sensation and the Relation
of the Psychical to the Physical, Ernst Mach writes: Thus, I lie upon my sofa. If I
close my right eye, the picture represented in the accompanying cut [see Figure 4] is
presented to my left eye (1896/1959, p. 18).

Figure 4 Mach's rendition of his left visual field

From Mach (1896/1959, p. 18), 1959 Dover Publications ltd.
36 |
Although Mach is a widely read author in what might be termed the history of phil-
osophical psychology, there is an ostensibly different sense of the term visual field
that is used in contemporary psychological science. Consider Figure 5, and the fol-
lowing entry from the Oxford dictionary of psychology:
In humans, the visual field of each eye subtends a visual angle of
about 17o degrees in the horizontal plane [and about 130 degrees in
the vertical] and there is almost complete [horizontal] overlap be-
tween the visual fields of the two eyes. The part of the visual field to
the left of the point on which gaze is fixated is called the left visual
field (lvf ) and the area to the right the right visual field (rvf ), each of
these areas being a hemifield. (Colman, 2003, p. 781)
In modern clinical psychology, there is an established procedure used to plot VFs
known as perimetry. In static perimetry, the patient fixes her gaze upon a target
while stimuli are presented on the vertical or horizontal plane. The patient reports
detection of the stimulus either verbally (in ordinary clinical settings) or by the push
of a button. By employing conventional algorithms and psychophysical methods,
detection thresholds can be determined for stimuli of various intensities presented
at various points, and at various degrees of removal from the central fixation point.
In kinetic perimetry, the patient is also required to fixate a central target. But in this
procedure, motion sensitivity at the boundary of a VF (and hence the boundary
itself ) is determined, by slowly moving stimuli from an extreme peripheral point
towards the focal point of gaze fixation (O. H. Schwartz, 2010, pp. 57 - 59).
Diagnoses of certain visual deficits seem to essentially require the notion of a (per-
imetrically determined) VF. For instance, some of the earliest articles in the study of
hemianopia (see Figure 5, H, I & L) define it as a selective deficit of form, colour
and/or lightness acuity in the left and/or right hemifield of each eye (Mackay, 1888;
Veasey, 1904). Today the range of hemianopic deficits are more varied and nuanced,
and though their organic bases are still unclear, the practice of using the VF in de-
fining and diagnosing visual deficits is no different today (Kerkhoff & Schenk, 2010).

| 37

Figure 5 The binocular visual field: deficits and neural correlates

So what is the relationship between the VF and visual perceptual experience? To get
clear on what is being asked: Call VF
1
the spatial extent of the frontoparallel plane in
which stimuli can elicit appropriate psychological responses in perimetric tests.
15

VF
1
is clearly essential to a range of neurological, ophthalmological and optometric

15
Cf. the external stimulus field, or field of view in Smythies (1996, p. 369). I am
supressing the fact that the notion of the visual field has raised some controversy in the
history of psychology (see Boring, 1952a; 1952b; Gibson, 1950, 1952). If my discussion is cor-
rect, the controversy stems from a failure to distinguish VF
1
from VF
2
.
The left and right visual fields (top, right) overlap partially to create a 120 central
binocular field, flanked on either side by a (conservatively estimated) 40 mo-
nocular crescent, creating a total binocular visual field of 160.
Particular lesion sites within the visual pathway (G L, right) are correlated with
particular field defects (G L. left). Thresholds (left) are indicated by shading;
darker shading indicates higher thresholds and thus lower stimulus sensitivity.
Reprinted from (Fauci et al., 2008, p. 184)
38 |
contexts. Call VF
2
a description of visual experience limited to the horizontal and
vertical dimensions. VF
2
is clearly what the PPV claims of the structure of visual
perceptual experience. It is perhaps illustrated by a literal apprehension of Machs
drawing, depicting the objects of his visual experience as seeming to be as flat as
presented on the page.
There are ways in which VF
1
and VF
2
are comparable; there is also a key sense in
which they are not. A paradigm that vividly illustrates a comparison has been devel-
oped by Stephen Kosslyn (1978). Kosslyn asked his subjects to walk towards fronto-
parallel positioned rectangular shapes of between 2 and 20 inches in height and
breadth, fixing their visual focus at the centre, and to stop walking when the edges
were no longer sharply in consciousness (1978, p. 383). With both size of and dis-
tance from the object known, the angle of the visual field could be determined.
Hence, the possibly indeterminate bounds of VF
2
(cf. Machs picture) could be ren-
dered (at least more) determinate as VF
1
.
But this does not help the PPV. It is not a significant point in the PPVs favour to
establish that visual experience involves the experience of planar properties within
certain bounds. Indeed, the fact that VF
1
is determined in perimetric tests on a regu-
lar basis has no bearing on whether or not PPV is correct. This ought to be clear by
noting the manner in which VF
1
is determined. What is conspicuously absent in
perimetry (apropos the PPV) is any test of sensitivity along the sagittal plane. The
lesson here is that the PPV requires a positive thesis about the absence of experienc-
ing properties and relations along the sagittal plane or indeed, the absence of any
experience of objects as having depth or seeming to be some distance away.
2. 4. Indirect perceptual experience
One way of constructing a positive thesis would be to limit the claim to certain
forms of perceptual experience and claim that they are actually composite. The
basic structure of the composition proposed could be along the lines Jacksons
schema, adjusted to provide a framework for potential accounts of indirect percep-
tual experience (IPE):
S indirectly perceptually experiences x if and only if S perceptually experi-
ences x in virtue of perceptually experiencing y, and x and y are distinct.
The following assertions offered by the eminent vision scientist, Irvin Rock, are
poignant in this regard:
| 39

[W]e cannot properly describe sensory experience without including
reference to the proximal mode of perception [...] Consider again the
circle seen at a slant [...] We do perceive the circle at a slant as circu-
lar, but we also are aware that its projected extensity relations are el-
liptical. The point is that we would be seriously distorting the phe-
nomenal facts if we chose to speak only of the constancy aspects of
perception [...] even when constancy is present a distant or slanted
object does not look exactly like a near or unslanted object of the
same size and shape (Rock, 1981, p. 263)
It ought to be relatively clear what Rock means by the proximal mode of percep-
tion. The projected extensity relations are the (problematically) variable relations
motivating the argument from perspectival variation in 1.2, but under a description
that ties the variation to a specific physiological basis. As Steven Palmer (a student
of Rocks) puts it in his influential textbook the proximal mode reflects mainly the
properties of the retinal image, or proximal stimulus (Palmer, 1999, p. 313).
The constancy aspects to be contrasted with the proximal mode are (in this case)
spatial properties that are experienced as remaining constant; or rather, they are the
spatial properties that are not experienced as varying, such as an objects perceived
size, shape, position and orientation.
16
These properties are discerned by means
some fixed standard. Hence, E.G. Boring asks What is perceptual size constancy?
and answers: It is the rule that perceived object size is invariant under the trans-
formation of tape-measured distance (1952b, p. 145); mutatis mutandis for other
spatial constancies. Perceptual constancy is thus thought to be a reflection of an
individuals capacity to experience the features of an object that actually are invari-
ant. And since antiquity, such experience has been thought to be the basis of an
individuals capacity to judge that she perceives the same object through the course
of some change in the relation between the object and the individual. As Thomas
Reid relates:

16
There are other (non-spatial) constancies of less obvious concern here. In vision for
instance, there is also lightness constancy and colour constancy. See Palmer (1999, pp. 125 -
136) for an introduction to both. For differing philosophical accounts of colour constancy see
No (2004, pp. 125 - 132) and Kelly (2004, pp. 82 - 88).
40 |
Let one look upon any familiar object, such as a book, at different dis-
tances and indifferent positions: is he not able to affirm, upon the
testimony of his sight, that it is the same book, the same object,
whether seen at the distance of one foot or of ten, whether in one po-
sition or another; that the colour is the same, the dimensions the
same, and the figure the same, as far as the eye can judge? This surely
must be acknowledged. The same individual object is presented to
the mind, only placed at different distances, and in different posi-
tions. (1764/1997, chapter VI, 3, p. 83)
In modern psychological science these constancies provide a rich puzzle that moti-
vates perceptual theory, both in their objective status and in the fact that they are
presumed to be experienced. A popular way of framing the issue sees the initial
challenge as one of understanding the mapping relations between optical events
and objects in the environment. For although each projection of light on the retina
maps to a unique point in the projected environment, the same does not hold vice
versa; as Palmer explains: The inverse mapping from image to environment goes
from two dimensions to three, and this is not a well-defined function: Each point in
the image could map into an infinite number of points in the environment. Conse-
quently, for every 2-D image on the back of our eyes, there are infinitely many dis-
tinct 3-D environments that could have given rise to it (Palmer, 1999, p. 23,
emphasis mine).
17

Broadly similar issues arise when turning to the mapping of optical events and the
perceptual experience of constancy. In part, the structure of this problem is already
familiar, for many of an objects actual properties will typically remain constant,
despite a potentially vast range of minute changes in the properties of the array pro-
jected onto the retina. However, as noted earlier, a key motivation here is the as-
sumption that constancy is perceptually experienced. Thus, as William Epstein de-
scribes:

17
This is a nice statement of what Fodor and Pylyshyn have dubbed the establish-
ment perspective (see e.g. 1981, pp. 162 -163). Such insights motivate the view that stages of
visual processing require inbuilt assumptions, constraining the range of possible environ-
ments inferred as giving rise to the immediate stimulus. J.J. Gibson famously dissents on this
point, seeing the basic issue as misconstrued and the puzzle (that I describe above) as an
otherwise avoidable result. On his view, the misconstrual is facilitated primarily by ignoring
the possibility that visual systems evolved to process information through movement. For
critical discussion of Gibsons views (and those of the establishment) on the implications of
retinal projection, see Schwartz (1994, pp. 128 - 133).
| 41

Changing conditions of viewing, for example, viewing distance, van-
tage point, lighting, bring with them changes in the local optical in-
put that is correlated with a distal object or event. Yet everyday expe-
rience and laboratory evidence demonstrates that perception is con-
stant over a wide range of optical variations that result from changes
of viewing conditions. (Epstein, 1995, p. 14)
18

Now, when Rock says that we would be seriously distorting the phenomenal facts if
we chose to speak only of the constancy aspects of perception (emphasis mine) he
is claiming that these optical variations correspond to an indispensible part of our
everyday experience. That is to say that on his view, these proximal properties are
not merely part of the process of perception, they are perceptually experienced. In-
deed, Rock argues that these proximal mode experiences are fundamental (op cit.,
p. 264). On his view, it is (at least partly) in virtue of these experiences that we indi-
rectly experience the constancy of the object, as constancy perception is based on a
stage of prior detection of a feature directly correlated with the proximal stimulus
(op cit., pp. 264 - 265).
19
The perceptual experience of constancy, according to Rock,
is indirect insofar as there is prior (if perhaps very brief) stage at which perceptual
experiences of proximal properties occur, and it is upon these prior experiences that
the experience of constancy depends.
Rock hinges his case for the primacy of the proximal mode on the fact that there is a
temporal stage at which the perceptual process can be disrupted, resulting in
judgements that conform to projected properties. As an illustration he cites work by
Epstein and Hatfield (1978a, 1978b), who found that judgements of the orientation
and shape of pieces of card (presented at a slant) could be disrupted by briefly pre-
senting a randomly patterned black and white board (at frontoparallel orientation)
between presentations of the target shapes (also see Epstein, Hatfield, & Muise,
1977). The recorded responses suggested to the authors that, for their subjects, the
targets all appeared to conform to projective shape and further, that the targets,

18
The notion of constancy makes most sense in application to situations where
viewing conditions change. But it is sometimes used to denote simply a tendency towards
veridical perception of size, shape, orientation etc. For instance, Epstein and Hatfield
(1978b, p. 501) report nearly perfect constancy judgments of shape following a static and
arrested viewing condition (described in their 1978a, pp. 138 - 139), reflecting a rather com-
mon extended usage in which actual optical variations do not matter. For ease of expression
I will employ this extended usage.
19
Partly because, on the kind of theory Rock proposes, perceptual systems take into
account several other cues that inform logical inferences at later stages of perceptual pro-
cessing (Rock, 1981, pp. 240 - 282).
42 |
which were rotated in depth as much as 65, tended to appear to have a frontoparal-
lel orientation (Epstein & Hatfield, 1978b, p. 501).
20

The effect of the interposed mask (the randomly patterned board) certainly sup-
ports the claim that perceptual processing can be analysed as information pro-
cessing. The information required for visual perception of shape and orientation
takes time to process; proceeding through stages of integration, storage, and re-
trieval etc. (cf. Sperling, 1963; Turvey, 1973). It is also likely that the mask interferes
with the process sometime after the crossing of the two optic nerves in the middle
of the head, at the optic chiasm, i.e. it interferes with later binocular vision of the
target. But one can accept this and yet deny Rocks inference. That is, one can deny
that it follows from the fact that the judgements conformed to projection infor-
mation (which is all that is available at the monocular stage of the process) that
therefore the subjects must have been very briefly conscious of such projections
prior to the (binocular) processing required for constancy.
A.D. Smith, also unconvinced, points out that masking can be employed to demon-
strate subliminal perception (2000, p. 507). In his dissent he draws on work by Tony
Marcel (1983, pp. 202 - 207, 223 -231) in which words like bread facilitated recogni-
tion of words like butter, despite being presented at so brief a time before a dis-
rupting mask as to fall below an established threshold for conscious detection (cf.
Eriksen, 1960). Smith sees this as setting up a reductio ad absurdum of Rocks view,
saying that: If Rock were correct, one could equally argue from the fact that mask-
ing can prevent conscious experience of an object altogether that the first conscious
stage of perception is subconscious! (2000, p. 507).
21
Unfortunately, Marcels study
is not comparable to Epstein & Hatfields in the way that Smith believes it to be.
After all, in Epstein & Hatfields study, there is little reason to think that the stimuli
were not consciously seen prior to the mask. Indeed, the effect of the masking
might well have been to influence the processing of stimuli received after the mask
(Epstein & Hatfield, 1978a, p. 143).
But there is equally little reason to endorse Rocks interpretation, especially in its
generalisation. Remember the thought was that perceptual experience involves two
sequential stages, first we perceptually experience the proximal mode, and then we

20
Incidentally, in Smiths discussion of Rocks appeal to these studies, he claims that
Rock goes well beyond the sentiments of the authors he draws upon (see A. D. Smith, 2000,
p. 508, n.30). The quotation cited above makes that somewhat doubtful.
21
This dispute does seem a little reminiscent of a puzzle elaborated by Dennett &
Kinsbourne, over the difficulties of choosing between Stalinesque and Orwellian ac-
counts of the stage of consciousness (Dennett & Kinsbourne, 2002, p. 158).
| 43

perceptually experience the shape, orientation etc. of the object itself. To this there
are at least two viable alternatives.
One asserts that both stages, the proximal mode and the experience of constancy,
are perceptually experienced simultaneously. This is compatible with Epstein & Hat-
fields results as it might be that the mask eliminates an experience of constancy
that occurred at the time of the original presentation. Such an account seems to be a
natural consequence of Alva Nos conciliatory treatment of the debate between
proponents of SDT and their opponents (2004, pp. 75 - 84). No argues, in sympa-
thy with Strawson (1979/2002), that we do not go beyond what we see when say that
an object appears to have roughly the three dimensional properties it does (op. cit.:
81). Yet he also holds that one and the same experience of a plate, e.g., must con-
tain within it the possibility of being described as the experience of something el-
liptical and as the experience of something circular tilted away from one (op.
cit.:165). On this view, the Epstein & Hatfield setup merely disambiguates an experi-
ence which is otherwise ambiguous.
Another account removes any mention of stages. So either:
(a). The subject experiences the shape and orientation of the stimulus.
Or:
(b). The subject experiences the proximal projection.
But never:
(c). Both (a) & (b) at the same time.
This would treat Epstein & Hatfields results as cases in which (b) occurs because (a)
is prevented from occurring. Such an account seems to be a natural consequence of
Sean Kellys views on perceptual constancy (see e.g. his 2004). Consider how he con-
trasts his account with Nos:
I agree with No that it is possible to experience the circularity of the
plate, and I agree also that it is possible to experience its apparent el-
lipticalness. What I disagree with him about, however, is that we al-
ways experience both of these at once. Try it yourself. (Kelly, 2008, p.
685)
22

22
Despite this final encouragement to the reader to see (what he believes to be) the
manifest truth of his view, Kelly does admit that the difference between him and No might
well have reached the real foot-stomping impasse that is the bane of all phenomenology
(2008, p. 685). He also expresses a hope that the issue will be resolved by a shape-priming
study, which apparently he was running at the time he wrote the article (ibid., pp. 687 - 689).
Sadly, as far as I can discern, the results have yet to be published.
44 |
As stated neither of these alternatives leads to the kind of indirect account of per-
ceptual experience that Rock proposes.
23
And both can accommodate the results
that he sees as supporting his view exclusively. There is, then, no compelling reason
to endorse the model of perceptual experience he proposes. The potential dispute
between his view of perceptual experience and its alternatives is underdetermined by
the psychophysical phenomena that he appeals to.
Before moving on, I want to note another interesting aspect of Rocks discussion of
the proximal mode. At one point, he insists that [p]roximal mode experiences are
best thought of as perceptions rather than as sensations (Rock, 1981, p. 264). This
was instrumental in my construal of him as an advocate of an indirect account of
perceptual experience. But it would also be possible to construe proximal mode ex-
periences as sensations, with the caveat that they are not sensations only, in that
they are (ex hypothesi) perceptually experienced in the course of indirect perceptual
experiences.
24
The significance of this strategy will be borne out in the next section
and the following chapter.
2. 5. Prosthetic perception
It is worth asking whether the planar projection model fits other senses. Denial can
come from unexpected sources. For instance, OShaughnessy, a self-admitted sense-
datum theorist, claims that in the sense of touch there is no analogue of the visual
field of visual sensations which mediates the perception of the environment
(O'Shaughnessy, 1989, p. 39). However, there seems to be a potential counter exam-
ple in the perception of distance through vibration. This might bring to mind the
hackneyed example of a blind man tapping his way around the world with a cane.
But there is a greater generality to the capacity. And it serves as an illustration of

23
Actually, there are at least two readings of Nos view. For instance, in his book Ac-
tion in Perception he states explicitly that [l]ooks, sounds, feels appearances generally
are perceptually basic (No, 2004, p. 81), which suggests more than sympathy with the
sense-datum theory (see e.g. Overgaard, 2010, pp. 274 - 277). But a contrary theme that also
runs through that book is expressed in his earlier work, namely that looks [...] are not men-
tal intermediaries or sense-data, but are themselves features of the environment, features of
how the space one occupies is structured around one (No, 2002, p. 61). It is presumably on
this basis that, in his response to Kelly, he claims it is no part of my view that our awareness
of appearances is basic or primitive (2008, p. 692); at least, not more basic or primitive than
perceptual experience of the constant properties of objects.
24
Needless to say, sensation would not be understood in the way that Moore and oth-
ers might prefer.
| 45

how touch can figure in forms of perception that might also fit the planar projection
model of perceptual experience.
The earliest general statement of the phenomenon came from the psychologist Da-
vid Katz. He is worth quoting at length:
By its nature the pressure sense is a proximal or near sense, that is,
the object must be in direct contact with the skin to be perceived.
Although we obtain certain information on the shape, hardness, and
other properties of objects not directly in contact with our body
when we touch them with a stick or a probe, that is most certainly
based on acquired associations in these and similar cases that are ex-
ceptions to the rule. The vibration sense is not a proximal sense in
the same sense. When we perceive vibrations on the floor of a room
from a tuning fork several meters away, or from a machine at a far
greater distance under some circumstances, then here, too, just as in
audition, oscillations serve as a mediator between us and the source
of stimulation. (Katz, 1925/1989, pp. 203 - 204)
So it is through this oscillatory mediation that the vibration sense enables touch to
free itself from its confinement, and to begin to conquer distance (ibid., p. 204).
But still, initially, it might be thought that the model of planar projection has lim-
ited application here. For instance, A.D. Smith argues that the very depth and mo-
bility of the bodys parts makes planar projection the exception rather than the rule:
with four movable limbs it is easy, if we experience a sensation in each one, to be
aware of four sensations occupying positions that cannot be located in a plane
(2002, p. 134). Similarly, Katz suggests that in considering the events when vibra-
tions might be experienced, such as a train ride, it is rather more typical to experi-
ence vibrations the way that Smith suggests presenting a complex of sensations at
assorted bodily locations such as the soles of the feet, the thigh, the buttocks, the
back, and perhaps even the resting hands and arms (Katz, 1925/1989, p. 205).
But then again, equally applicable cases would be resting ones hand on a washing
machine, holding a power drill or a sander etc. To use cases such as these for sup-
port, the PPV would need to distinguish between the experience of vibration on the
skin, and the experience of a body part (or indeed, the several parts or the whole
body) vibrating. The distinction is useful because the skin can be conceived as a
sheet wrapped over the contours of the body, and thus a plane that only contingent-
ly has curvature in the third dimension. Thus delimited surfaces of the body such as
the pad of the finger, the palm and back of the hand, the back, chest, stomach and
tongue could be construed as sufficiently planar.
46 |
Controlled vibrations on such delimited surfaces are natural candidates for exploita-
tion in creating prostheses for perceptual systems.
25
The blind mans stick is perhaps
an example, but more advanced technologies stem from the basic principles laid
down in the auditory sensory mode, by Kazimierz Noiszewskis invention of the
Elektroftalm (probably in 1889, see Spirkovska, 2004, p. 7). Noiszewskis device used
a substance called selenium, disposed to increase in electrical conductivity when
exposed to light, a cell of which is placed on the forehead. By exploiting changes in
the conductivity of the selenium cell (under the influence of varying light intensi-
ties) by means of a basic coupling system, the graded presence of light could be
converted to sound. Although acuity was limited, the device apparently helped the
visually impaired to discriminate light and dark areas on the basis of varying sound
intensities resulting through movement.
26

By the second half of the twentieth century, developments in electronics enabled
these principles to be extended. Most famously, since the sixties Paul Bach-y-Rita
has been engaged in the ongoing development of a series of devices exploiting the
possibilities of vibratory sensation. His basic design couples a low-resolution cam-
era to an array of stimulators, producing mechanical or electrical energy on the sur-
face of the skin. Through coordination of intensity, the array (placed on the back,
stomach, or tongue) produces a vibratory image projected from the camera via the
coupling device (Bach-y-Rita, 2002, 2004; Bach-y-Rita et al., 1969). After a crucial
learning phase of exploratory movement, users of early devices were able to bat balls
and perform assembly line tasks, though (disappointingly perhaps) not to the same

25
I use the term prosthesis (cf. Dennett, 1991, pp. 338 - 341) in place of Bach-y-ritas
(perhaps more familiar) term sensory substitution device in my agreement with Lenay et
al.s sentiment that the latter term is doubly misleading and in many ways unfortunate
(2003, p. 279). As they argue, it is unhelpful to call the devices merely sensory substitutions
as it suggests that their function is merely sensory, rather than perceptual (and perhaps, in a
more Gibsonian voice, masks the fact that all such devices require exploratory activity to be
of any use). It is also naive to call such devices substitutions. For that conveys the false
promise of replicating an organic sense in all its glorious detail, and downplays the various
distinctive possibilities afforded by the augmentation of extant sensorimotor capacities (see
Lenay et al., 2003, p. 280 ff.).
26
Fournier DAlbe developed a similar device later, called the Optophone. According
to a New York Times article, an early incarnation (a hand held cylinder containing selenium,
connected to a telephone receiver by a clockwork mechanism) enabled a blind man to dis-
cern windows from walls, and count the number of people between himself and a wall
(Marconi, 1912). Later versions apparently enabled users to read reasonably sized printed
letters (D'Albe, 1921).
| 47

degree of competency individuals operating the task by sight. However, significant
advances have been made over the years, as Bach-y-Rita reports:
In a very recent trial, within an hour [of use], a blind person was able
to discern a ball rolling on the floor to him; he was able to reach for a
soft drink on a table; and he was able to play the old game of paper,
scissors, rock. Later, he walked down a hallway, saw the door open-
ings, examined a door and its frame, actually noting that there was a
sign on the door. (Bach-y-Rita, 2004, p. 86)

Figure 6 Prosthetic perception through planar tactile stimulus arrays

These examples suggest the application of the PPV to other sensory modes than
vision. And, in doing so, they make a better case for the proposed correction than a
shaky appeal to painting. The suggestion would be that the subject perceptually
experiences a three-dimensional environment around them, in virtue of perceptual-
ly experiencing changes in the two dimensions of vibratory sensation produced by
the stimulus array. Certainly, those building and testing such prostheses typically
(A) An old dentist's chair with 400 vibrating solenoid stimulators on
the back rest, coupled to a TV camera mounted on a tripod.
Reprinted with permission from Macmillan Publishers Ltd. Bach-
y-Rita et al. (1969, p. 963) 1969.
(B) A flexible 144-pixel array, with 2.13 mm electrodes, designed spe-
cifically for tongue stimulation.
Printed with permission from Kurt Kaczmarek of the Tactile
Communication & Neurorehabilitation Lab, University of Wis-
consin-Madison.
(C) Schematic of a prosthetic perceptual system using the same elec-
trotactile array. Video data from the head-mounted camera are
transmitted to the tongue display unit (TDU), which converts the
video into a pattern of 144 low-voltage pulse trains each corre-
sponding to a pixel of the video image.
Adapted from Bach-y-Rita & Kercel (2003, p. 543), with permis-
sion from Elsevier 2003.
48 |
reject this kind of description. But the mere rejection of the description is not
enough to show that it is wrong.
To clarify the strategy further, note certain desiderata for an account of IPE. There
are at least three desiderata for such an account. Firstly, an account of IPE must mo-
tivate an internal structure to certain perceptual experiences, such that they can be
considered as indirect. Call this the structure desideratum. Secondly, to ensure
against an infinite regression, the structure claimed must pertain to a specific class
or form of perceptual experience, for instance, experiencing certain properties or
relations in a particular sensory mode. Call this the specificity desideratum. Thirdly,
the structure claimed must be amenable to generalisations concerning all members
of the specific class identified. Call this the generalisation desideratum.
The strategy would meet the both the structure and the specificity desiderata by
claiming that perceptual experience involves sensation, and that sensation involves
distinctive limitations. The specific class is sensations, though irrespective of senso-
ry mode: visual sensations, auditory sensations, vibratory sensations etc. What the-
se all have in common is both the fact that they are experienced and the fact that
they are held to be impoverished with respect to the (ex hypothesi indirect) object of
perceptual experience. In this way, the generalisation desideratum is met. For the
thought is that, in each case, pointing to this limitation directly supports a corre-
sponding claim about perceptual experience.
Accordingly, it is not enough to baldly assert that one really experiences something
more rich, in the face of the claim that first and foremost one experiences some-
thing more sparse. One needs to provide reasons for ones claim, in response to the
fact that reasons have been provided for its contrary. This strategy of appealing to
limitations in sensation needs to be developed further to see why it ultimately fails.
We will do just that in the next chapter with a focus on the work of Bishop George
Berkeley.
| 49

Chapter 3 Berkeley & the limits of
sensation
3. 0. Introduction
In the last chapter I covered several unsuccessful attempts to attack the direct realist
conception of perceptual experience. The lines of argument were twofold. One line
of argument held that the properties and objects experienced in vision are organ-
ised two-dimensionally, in contrast to the assumed three-dimensional nature of the
objects of perception. Another line of argument claimed that perceptual experience
of shape and size is composite: on this view, direct perceptual experience of project-
ed shape and size serves as a basis for indirect perceptual experience of actual shape
and size. At the end, I broached a strategy that was prima facie more promising
than any of those previously considered. It involved directly referring to the nature
of the sensations involved in a given perceptual experience, and emphasising their
limitations relative to what is perceptually experienced.
In this chapter, I will explore various ways of filling out the claim that sensation is
limited, and the claim that perceptual experience might be resultantly indirect. To
anchor the discussion in the views of an author who arguably held the relevant the-
ses, I will proceed via a study of Berkeleys philosophy of perception. I will begin by
introducing (in 3.1) a general form of argument attributable to Berkeley, which I
will call a limitation argument. I will then show (in 3.2) why Berkeleys theory of
vision requires limitation arguments, following up (in 3.3) with an examination of
three different ways in which he evidently used a limitation argument to support the
claim that we do not directly visually experience distance. As I will show (in 3.4),
none of the proposals work because they exploit several misleading equivocations in
their use of the notion of sensation. But their failures are revealing, for they allow
one to clarify ambiguities in concept of sensation, and to further specify the manner
in which a defendant of the common-sense conception ought to treat the relation-
ship between sensation and perceptual experience (3.5).
3.1 Limitation arguments & the one point argument
3.2 Berkeleys ideas of sense
3.3 Three objects of immediate sight
3.4 Sensation, perception, & perceptual experience
3.5 Perceptual sensation
50 |
3. 1. Limitation arguments & the one point argument
Consider the following claims:
(i). Visual sensations occur by means of the energetic impingement of light
upon the retina.
(ii). The only spatial properties one ought to attribute to the direct objects of
visual sensation are those licensed by the means identified in (i).
(iii). The only spatial properties one ought to attribute to the direct objects of
visual experience are those licensed by the means identified in (i).
(iv). Light travelling from any distance along a line perpendicular to the retina
projects to one point on the retina.
27

(v). One ought not to attribute relative or absolute distance to the direct ob-
jects of visual experience.
28

This is a prime instance of what I will call a limitation argument (LA). LAs are de-
fined by having a key premise (or premises) consisting in a claim concerning some
manner in which sensation is limited. The significant difficulties with this particular
LA come in considering the relationship between (i), (ii) & (iii). The argument, or
something much like it, is often attributed to Berkeley. That is, he is usually charged
with an endorsement of (iv) and (v), and implicitly (i) (iii), in connection with
passages such as the following:
29

27
There is a sense in which the optical situation is (importantly) more complicated
than this point suggests. What the static eye confronts is not merely a single ray of light, but
a pencil of rays, sampled from a unique ambient optic array i.e. the total quantity of light
converging from all directions on a given position (Gibson, 1966, pp. 12 - 14). The light itself
is reflected off surfaces, such as the ground extending away from the observer, projecting
rays to a variety of discriminable points on the retina (Gibson, 1950, pp. 59 - 62). The one-
point argument can be generalised to accommodate this case (see Figure 5), which indicates
the intractable nature of the problem when considering the optical situation of a static ob-
server.
28
All the oughts are intended to convey the kind of corrective or propaedeutic tone
that provides the sense-datum theorists an answer to their dissenters (see 2.1).
29
For a critical discussion of whether it is correct to attribute such an argument to
Berkeley, see Atherton (1990, pp. 9 - 15, 61 - 88). For a prominent example of charging Berke-
ley with (v) see Armstrong (1960, pp. 2 - 22), and for a more recent case see A.D. Smith (2000
passim). And for a brief sample of the range of differing views on the matter see Schwartz
(1994, pp. 4 - 6).
| 51

It is, I think, agreed by all that distance, of itself and immediately,
cannot be seen. For distance being a line directed end-wise to the
eye, it projects only one point in the fund of the eye, which point re-
mains invariably the same, whether the distance be longer or shorter.
(Berkeley, 1732/2008, 2, p. 7; see also Figure 7 below)

Figure 7 Distance as a line directed end-wise to the eye

The key premise of this one point argument would be the second sentence, which
clearly corresponds to (iv) above. The conclusion comes in the first sentence. But
whether it matches up with the conclusion (v) above turns on what Berkeley means
by immediately seen. In particular, it depends upon whether the range of spatial
properties one ought to attribute to the direct objects of visual experience is limited
in the sense claimed by (iii); and that depends upon the relation between visual
sensations in (i) and (ii) and visual experience in (ii).
In connection with (v), Berkeley is sometimes represented as endorsing the thesis
that we do not see distance at all, or if we do, that it is some kind of illusion. But in
fact, it is not at all clear he does hold that thesis in anything but a highly qualified
form. Again, the issue turns on what Berkeley means by immediately seen. For the
suggestion might be that distance is in fact seen, but in virtue of some mediation.
The interpretive line that I will assume is this: Berkeley proposes a form of indirect
visual perception. According to his theory, we do see distance, and the only manner
in which the sight of distance is illusory is in the impression that we see distance
immediately or by sight alone. But, to get an initial grasp on Berkeleys particular
(A) Light travelling from any point along the same line of sight
will project to the same point in the retina.
(B) Generalises from single rays to arrays: ABCD, IJKL, and
VWXY each project to abcd, respectively. Contra Gibson
(1950, pp. 59 - 62).
Both adapted from Schwartz (1994, pp. 21, 130), 1994 Rob-
ert Schwartz.

52 |
form of indirect perception, it is worth identifying the collective that Berkeley refers
to when he claims that it is agreed by all that distance is not immediately seen. As
Margaret Atherton clarifies, what was agreed by all in Berkeleys time was that the
immediate physical effects of light upon the eye could not be sufficient for various
features of visual perception (Atherton, 1990, pp. 62 - 68). This was a lively point of
interest in early modern perceptual theory, largely due to rapid advances in the the-
ory of physical and physiological optics: in particular, refinement of geometrical
principles and their application to the problems of mapping optical events with
their environmental sources. These advances gave rise to another common consen-
sus, one that Berkeley resisted, which was that the mind was innately disposed to
make rapid and unconscious use of the very axioms found to be so successful in the
optical theories of the time. This was also an account of indirect perception of sorts,
but one on which the notion of inference would be indispensable. For at its core was
the claim that when we see, we demonstrate, if perhaps unconsciously and auto-
matically, but nevertheless in just the manner in which one would be able to [geo-
metrically] demonstrate how we see (Turbayne, 1962, p. 161 see also pp. 159 - 171).
A key goal for Berkeley in writing his New theory of vision was to oppose the way in
which, as Colin Turbayne writes, geometry was imposed on the actual manner in
which we see (1962, p. 160) by a widely-read group of authors (Descartes and Male-
branche being notable figures) holding this view, whom I will refer to as the geome-
trists.
30
In doing so, Berkeleys intention was to create

a rival account of how it is
that we do see more than we seem to be licensed to in regard to (i), (ii) and (iv)
(Atherton, 1990, pp. 76 - 77). Hence, on this reading at least,

Berkeley is in agree-
ment with his contemporaries on the claim that distance can be seen, though it
cannot be immediately seen. However, it remains to be shown why it is that Berke-
ley required the use of LAs in supporting these claims.

30
Not to imply that there was a uniform geometrical theory: Broadly speaking, a
theory could employ geometry in an account of how we make judgements about spatial
properties of an external world, or more basically in an account of how we perceive spatially.
Berkeleys target was the latter. For discussion of Descartes and Malebranches views in this
regard, see (Atherton, 1990, pp. 16 - 57). William Molyneaux is also cited by Berkeley in con-
nection with his work on geometrical and physical optics (e.g. Berkeley, 1732/2008, 40, pp.
19- 20). Though it is not clear whether he is cited as an author giving an account of visual
perception per se (see Atherton, 1990, p. 17; and Turbayne, 1955, pp. 341 - 342 for contrasting
views).
| 53

3. 2. Berkeley s ideas of sense
In Berkeleys view the mind is populated by ideas. These fall into three categories:
ideas actually imprinted on the senses, or else such as are perceived by attending to
the passions and operations of the mind, or lastly ideas formed by help of memory
and imagination (Berkeley, 1710/2008, Part I, 1, p. 83). The first category is of pre-
sent interest, and he immediately goes on to say more about it:
By sight I have the ideas of light and colours with their several de-
grees and variations. By touch I perceive, for example, hard and soft,
heat and cold, motion and resistance, and of all these more and less
either as to quantity or degree. Smelling furnishes me with odours;
the palate with tastes, and hearing conveys sounds to the mind in all
their variety of tone and composition. (ibid., Part I, 1, 83)
Berkeley repeatedly flips between the terms idea of sense and sensation. But
clearly what he cares most about is this: whatever he is referring to with these terms,
it is that which is directly perceived. As he explains in the voice of Philonous: The
things, I say, immediately perceived are ideas or sensations, call them which you
will (Berkeley, 1734/2008, Dialogue II, p. 196 - 197). At least, then, Berkeley seems
unequivocal on two points: sensations are perceived; sensations are the direct ob-
jects of perception.
In fact, Berkeleys stance is more radical than this. He is famous for arguing that all
objects of perception are ideas and no more. This distinctive element can be re-
vealed in contrast to the traditional interpretation of Lockes view of perception. At
the heart of which is the claim that the visible properties of the ideas we directly see
resemble the visible properties of objects we indirectly see, in much the same way as
the patches of paint on a canvas resemble the objects depicted (Lowe, 1995, p. 38).
Crucially, this interpretation is bolstered by Lockes endorsement of a version of the
primary/secondary quality distinction (1690/1997, Book II, Chapter VIII, 9 - 10, p.
135). Primary qualities are physical properties of the actual objects we indirectly
perceive, such as solidity, extension, figure and mobility. And our ideas of them, on
the traditional interpretation, veridically resemble their indirect causes. By contrast,
ideas of secondary qualities deviate from their causes, as they are produced by pow-
ers of objects further reducible to their primary qualities. So, some ideas resemble
objects and some do not. In his denial of this distinction, Berkeley asserted what is
often called the likeness principle: that no idea could resemble some non-ideational
54 |
object, for an idea can be like nothing but an idea (1710/2008, Part I, 8, p. 85).
31
In
particular, it cannot be compared to something that is in essence distinct from an
idea.
This claim is dialectically related to Berkeleys famous dictum that the existence of
ideas consists in instances of perception; but as George Pappas (1999, pp. 143, fn. 113)
points out, it does not necessarily follow from it. For one could hold, as Locke might
have, that the existence of ideas of sense is dependent on acts of perception, and
still hold that they resemble objects which are not ideas. This gives further signifi-
cance to Berkeleys assertion of the likeness principle. For in doing so, Berkeley mo-
tivates something like the following disjunction: that either all perceivable objects
are ideas of sense, or no idea of sense resembles a perceivable object.
Berkeley accepts the first disjunct as what he sees to be the only alternative to the
sceptical implications of the second. Thus he asserts that all possible objects of per-
ception, not only the direct objects of perception, are collections of ideas. That is, on
his view, all possible objects of perception are collections of sensations and no more:
I see this cherry, I feel it, I taste it; and I am sure nothing cannot be
seen, or felt, or tasted. It is therefore real. Take away the sensations of
softness, moisture, redness, tartness, and you take away the cherry.
(1734/2008, Dialogue III, p. 229).
More to the point, they are a mass of associations between sensations, such that
ideas are united into one thing (or have one name given them) by the mind, be-
cause they are observed to attend each other (ibid., Dialogue III, p. 229) (ibid., Dia-
logue III, p. 229). Call this the association principle.
Reflection on the association principle provides crucial insight into Berkeleys phi-
losophy of perception. The key thought is that these associated ideas ought to be
properly distinguished if one is to understand how their relations enable us to per-
ceive.
32
In particular, as regards our direct perception we never see and feel one and

31
The argument for the likeness principle in the passage I cite is scanty. Though de-
spite this, when considered with related remarks in Berkeleys other writings, the argument
admits of several different interpretations, for discussion of which, see Ryan (2006).
32
Though, this is certainly not his only reason for distinguishing them. As George
Pappas points out, there are evidently stronger and weaker forms of the thesis that ideas of
sense are heterogenous (2002, pp. 56 - 57). The weaker form is simply the claim that the
direct objects of sight and touch are numerically distinct. An instance of this weaker claim is
quoted in the main text. The stronger form is that the direct objects of sight and touch are
distinct in kind (see Berkeley, 1732/2008, 121 - 123, pp. 52 - 53). This latter is related more
directly to the doctrine of abstract ideas that Berkeley abhorred (see next note).
| 55

the same object. That which is seen is one thing, and that which is felt is another [...]
the objects of sight and touch are two distinct things (Berkeley, 1732/2008, 49, p.
23).
33
Moreover, he sees the point to be entirely general:
Sitting in my study I hear a coach drive along the street; I look
through the casement and see it; I walk out and enter into it. Thus
common speech would incline one to think I heard, saw, and touched
the same thing, to wit, the coach. It is nevertheless certain, the ideas
intromitted by each sense are widely different and distinct from each
other; but having been observed constantly to go together, they are
spoken of as one and the same thing. (Berkeley, 1732/2008, 46, p. 23)
Just as the sound that we directly hear is distinct in kind from the light that we see,
so the ideas, perceivable by touch such as distance, tangible figure, and solidity
are distinct from what we directly see (ibid., 45, p. 22).
If Berkeley is to be interpreted as having some use for an LA, here is a natural entry
point. Distinct sensations, such as the touch, sight and taste of the cherry, are po-
tentially correlated over time. This correlation could result in various associations
that might explain our capacity to indirectly perceive properties which are ex hy-
pothesi beyond the range of a given sense, in virtue of perceiving properties associ-
ated with other senses that are not limited in that respect. The most pertinent case
would be in the association of sight and touch, where in perceiving certain ideas of
sight we anticipate which tangible ideas are, by the wonted ordinary course of na-
ture, like to follow (ibid., 45, p. 22) (ibid., 45, p. 22). And recognition of such
associations obtaining would provide principled reason for saying that there are
two sorts of objects apprehended by the eye, the one primarily and immediately,
the other secondarily and by intervention of the former (Berkeley, 1732/2008, 50,
p. 24). And if this were right, we should correct our description accordingly: When-
ever we say an object is at a distance, whenever we say it draws near or goes farther
off, we must always mean it of the latter sort, which properly belong to the touch

33
In De Anima Aristotle drew a distinction between special sensible properties, per-
ceptible exclusively within individual senses such as colour and heat, and common sensible
properties, perceptible between senses, such as motion, rest, number, shape, size and unity
(Aristotle, 350 BCE/2002, Book II, Chapter 6, 418a, pp. 25 - 26, Book III, Chapter 1, 425a - b,
pp. 46 - 47). Berkeley is essentially denying that the latter exist. In Lockes view most of the
common sensibles are primary qualities (Locke, 1690/1997, Book II, Chapter V, p. 128). In
particular, the objects of sight and touch might share a primary quality, like extension. And
this could be an empirical basis for abstraction to the idea of extension. Thus, in banishing
common sensibles, Berkeley is not only pre-empting a basis for a certain kind of abstract
idea, but he is also blocking any evidential ground for a primary/secondary quality distinc-
tion (Pappas, 2002, pp. 68 - 69) .
56 |
(ibid., 50, p. 24) (ibid., 50, p. 24). So for instance, when looking at an object, what
I directly see is a certain visible figure and colour, with some degree of faintness
and other circumstances. But in seeing an array of faint light I anticipate, from
what I have formerly observed [...] that if I advance forward so many paces or miles,
I shall be affected with such and such ideas of touch (ibid., 45, p. 22).
There is an equivocation in the reference of ideas of touch that ought to be clari-
fied to properly understand the mechanics of the theory. Certainly, the suggestion is
that when I see an object at a distance, I anticipate a sensation of bodily contact
with that object. That is a relevant meaning of the term touch today, and Berkeley
certainly held such sensations to fall into that category. But in Berkeleys idiom
touch includes a range of bodily sensations, including e.g. the sensations of bodily
movement resulting throughout an approach to the object (Berkeley, 1732/2008,
45, p. 22).
34
And these are of equal importance in the theorys general application.
On a modern view, these sensations would correlate with the organic registration of
the speed and extent of changes in muscle length, and perhaps with additional cues
from the stretching of the skin, the effort of generating movement, and the presence
of a motor signal (Proske & Gandevia, 2009).
35
And given Berkeleys proclivities, he
would likely prefer the thought that the primitive signals generated from these or-
ganic sources would be the basis of tangible ideas of bodily movement, rather
than geometric abstractions based upon primitive signals. Similarly, by association
with the sensation of accommodating the lens of the eye and/or turning the eyes
inward, blurriness could serve as a cue to seeing an object as (too) close (Berkeley,
1732/2008, 16 - 21, pp. 9 - 11; 27, p. 12; 36, p. 18). In this case, blurriness would be
correlated with registration of activity in the ciliary muscles affecting pressures in

34
Armstrong speculates that: Perhaps he is under the misleading influence of the
doctrine of the five senses, and so has to fit kinaesthetic experience under touch (1960, p. 71,
n.13). Actually, broader considerations suggest it is no accident that Berkeley conglomerates
a variety of sensations under touch. For if, e.g., kinaesthetic sensation is thought of as
genuinely different in kind from tactile sensation, then given Berkeleys aversion to common
sensibles (see the last note) he would need to assign distinct ideas assigned to distinct
modes, in order that there are no remaining ideas available to multiple sense modes (cf. also
Grush, 2007a, p. 427, n.32; Pappas, 2002). For arguments sake (though I do hold some
doubts) I will assume that either the conglomeration is innocent or that a suitable decompo-
sition can be executed.
35
Also in extensive locomotion approaching the object, non-visual sensory conse-
quences of bodily movement would not merely register changes in the configuration of ones
body, but also ongoing changes in spatial position, nowadays known as path integration
(see e.g. Klatzky, Loomis, & Golledge, 1997).
| 57

the tissue surrounding the eye lens, concomitantly affecting its roundness (Drake,
Vogl, & Mitchell, 2004, p. 853). And eye turning would be correlated with the rela-
tive lengths of extra-ocular muscles, primarily the medial and lateral recti that re-
spectively control the adduction and abduction of each eye(ibid., pp. 839 - 840)
(ibid., pp. 839 - 840). Again, the sentiment is that sensations derived directly from
these sources could be associated with the distance of objects. Arguably, this would
also be a reliable process, given that what essentially occurs are changes in the mo-
nocular and binocular angular relations to the light reflected from the object. And
in this regard, the interpretation would reflect the parity of Berkeleys reasoning.
For it would allow for the very same relations that the geometrical model might pos-
it as enabling the explicit computation of distance, to play an implicit role in the
association of distance with very minimal cues in visual sensation(op. cit., 22- 26,
pp. 11 - 12; 38 - 39, pp. 18 - 19) (op. cit., 22- 26, pp. 11 - 12; 38 - 39, pp. 18 - 19).
36

Furthermore, this account gives the simple visual sensations posited a powerful
function in guiding behaviour, simply by serving as cues to associated movements,
and ultimately contact that one might wish to facilitate or avoid (see e.g., Berkeley,
1732/2008, 59, p. 27; 87, p. 40).
I will not go any further in demonstrating how Berkeley could put LAs to use. Now it
remains to show that Berkeley in fact ought to do so. As Rick Grush notes, there is
certainly another kind of argument that Berkeley uses, which focuses on the contin-
gency of the associations formed on the basis of sensation:
[I]t might easily have been the case that faintness would correlate
with distance differently from the way in which it does [...] Similarly,
the correlation between specific felt lengths of the muscles control-
ling eye orientation and eye angles and the distance of seen objects is
also a contingent one. That having my eyes oriented such that their
angle to the foveated object is degrees feels like this rather than like
that is surely contingent, and could have been otherwise. (Grush,
2007a, p. 428)
Call this a contingency argument. Contingency arguments and LAs are not unrelat-
ed. For once it has been established that visual sensation is limited, the significance
of the limitation needs to be demonstrated. In the context of an associative theory
of perception like Berkeleys, one way of doing that would be to show that the cues

36
Berkeley also suggests that he could avail himself of divers other circumstances
that might contribute to frame our idea of distance. Thus, perhaps he would accept that
various pictorial cues and shading gradients could provide associative prompts for the per-
ception of relative depth (see Palmer, 1999, pp. 229 - 249 for a summary of modern work in
this area), in supplementation to the other cues discussed.
58 |
that do enable the perceiver to transcend said limitations could have done so in a
different way (or might not have done so at all). This is to say that an LA is sufficient
to at least motivate, if not justify, a contingency argument. On the other hand, the
very nature of such a contingency argument seems to rest upon there being no pri-
ma facie connection between what we mediately perceive and the media of our per-
ception, viz. sensations. The consequence of this is that it relies on an LA demon-
strating that there is no prima facie connection. For if sensation is not limited, then
the contingency argument either asserts something false: in assuming that there are
limitations to be transcended; or it asserts something trivial: that certain cues could
enable the transcendence of limitations, if sensation were limited, and if they did
then they could have been otherwise. This is to say that limitation arguments are
necessary to justify contingency arguments.
Grush rightly complains that Berkeley continually conflates these two arguments
when discussing visible size (or what he calls magnitude 1732/2008, 52 - 87, pp.
24 - 40), position and orientation (or what he calls situation, ibid. , 88 - 120, pp.
40 - 52).
37
In all these cases, only contingency arguments are provided, arguments
that depend upon unstated and potentially untenable LAs (Grush, 2007a, pp. 428 -
440).
38
Again, what this shows is the necessity of LAs for Berkeleys theory of per-
ception. To further see this, consider Berkeleys use of a familiar philosophical tool,
the thought experiment.
In his New Theory of Vision Berkeley employs two kinds of thought experiment: the
man [sic.] born blind, throughout; and, towards the end, the case of an unbodied
spirit, which is supposed to see perfectly well, i.e. to have a clear perception of the
proper and immediate objects of sight, but to have no sense of touch (Berkeley,
1732/2008, 153, p. 64). The man born blind thought experiment finds its origin in
correspondence between Locke and Molyneux, where the former published a ques-
tion posed to him by the latter: whether a man born blind, able to distinguish cubes
from spheres by touch, would be able to distinguish them by sight, when made to

37
The perceptive reader might have noticed that shape is not on this list. Actually,
Berkeley seems to assimilate two-dimensional shape (or figure) to size, by treating its
shape as its boundaries on breadth and height dimensions. Thus, figure is the termination
of magnitude (Berkeley, 1732/2008, 105, p. 46). His stance on three-dimensional shape viz.
relative depth is not quite as clear (but see last note).
38
For sympathetic accounts of Berkeleys views on situation and magnitude, see
Atherton (1990, pp. 108 - 171) and Schwartz (1994, pp. 58 - 83).
| 59

see (Locke, 1690/1997, Book II, Chapter IX, 8, p. 144)
39
Molyneuxs suggested an-
swer was negative, evidently (according to Locke at least) because the man born
blind could not have learnt that what affects his touch so [...] must affect his sight
so [...] or that a protuberant angle in the cube, that pressed his hand unequally, shall
appear to his eye, as it does in the cube (ibid., p. 144). The conditions of the test
vary under interpretation: Must the man be able to tell the difference at an instant
or over time? Might the man see a difference but be unable to express any judge-
ment of a difference? Respective answers from various eminent scholars were a lively
issue of dispute in the eighteenth century (see Degenaar, 1996, pp. 25 - 52 for a
review). And although the issue in the question that Locke published concerned
visuotactile relations (or their lack) in shape perception, similar questions could be
posed for any variety of sense modalities, and the perception of any spatial property.
Indeed, Molyneuxs first letter to Locke includes questions concerning the percep-
tion of distance, such as whether the congenitally blind man could know by his
Sight, before he stretchd out his Hand, whether he Could not Reach them, tho they
were Removed 20 or 1000 feet from him (as reproduced in Degenaar, 1996, p. 17).
Berkeleys answer is (of course) negative: a man born blind, being made to see,
would at first have no idea of distance by sight (1732/2008, 41, p. 20); and he takes
this to further imply the falsity of the geometric theory, given its predilection for
positing an innate geometry. But note that we only have course to believe Berkeley if
we accept the crucial premise that sensation is limited. For it is this and only this
that makes it false that one blind from his birth, being made to see, should stand in
need of no new experience in order to perceive distance by sight (42, p. 21). To put
the point another way, even if the geometrical theory were false, there could be

39
Opinions differ on the relevance and conclusiveness of actual cases of reverted con-
genital blindness. For instance, Armstrong condemns any such actual and potential cases as
having a tendency to either beg the question, or else to be capable of being interpreted to
the satisfaction of both [and presumably any] parties in the dispute (Armstrong, 1960, p.
64). By contrast, Gallagher is rather more optimistic, claiming that recent work in develop-
mental psychology and neuroscience suggest a definitively negative answer to the question,
though for slightly different reasons than Locke (and Molyneux) supposed (2005, pp. 153 -
172). During Berkeleys lifetime there was a famous case in which a boy who suffered from
either congenital or very early blindness had his cataracts removed by the innovative sur-
geon William Cheselden (Degenaar, 1996, pp. 53 - 56). Cheseldens report to the Royal Socie-
ty was subsequently cited by many (including Berkeley) as evidence in favour of an associa-
tive theory of perception; but the sentiment was far from unanimous, and scepticism
abounded as to what the report did or indeed could actually demonstrate (Degenaar, 1996,
pp. 87 - 88; von Senden, 1932/1960, pp. 219 - 221). I will treat the man born blind as a thought
experiment, as Berkeley seems to have done in his New theory of vision.
60 |
some brute fact about visual sensation (i.e. a hitherto unexplained fact) such that
we ought to hold that the congenitally blind man would see.
40

Similarly, when Berkeley claims later that such a one, if we suppose him made to
see, would not at first sight think that anything he saw was high or low, erect or in-
verted(95, p. 43), the very same constraints apply. It is of no consequence whether
or not, in this instance, either his associative theory or the geometrical theory is
correct. What matters is whether visual sensation is limited such that, e.g., neither
up nor down can be seen (cf. also Grush, 2007a, pp. 137 - 139). In this way, assess-
ments of the case of the unbodied spirit follow suit. When he describes the poor
being as not having any idea of distance (Berkeley, 1732/2008, 154, p. 64), again,
this can only be the correct assessment if visual sensation is thus limited.
In sum, a brief survey of Berkeleys philosophy of perception reveals the fundamen-
tal importance of direct objects of sight, sc. visual sensations, being limited. Upon
the assumption that this condition holds, Berkeley provides a parsimonious positive
account able to explain how we see spatial properties that ex hypothesi we are una-
ble to immediately see, through the mediative process of association with distinct
ideas of touch.
Throughout I have relaxed my qualms about the potential conflation of perception
and perceptual experience in the treatment of sensation. Now is the right point at
which to begin demonstrating that such conflation is rampant in Berkeleys work on
vision.

40
As Schwartz notes (1994, pp. 125 - 152), most modern work on vision amounts to a
development of Berkeleys ideas, a development of the work of the geometrists, some selec-
tive amalgamation of the two, or a descendent of Gibsons ecological optics as an alterna-
tive to both Berkeleys view and the geometrists (see e.g. Gibson, 1961, pp. 261 - 262). The
contrast in Gibsons work in this regard is in the claim that visual sensation is not limited
(see e.g. A. D. Smith, 2000, pp. 490 - 491). His essential insight was to note that as the point
of observation moves, the optic array exhibits constant change, resulting in a linear optic
flow field, where the bases of various optic cones (with apexes on the eye) will systematically
expand or contract. When two optic cones on the same visual plane are expanding at differ-
ent rates, the faster expanding cone will envelop the other, and if the corresponding features
are opaque, one perceived object will occlude another. See Lee (1993, pp. 44 - 47) for an in-
troduction and illustrations.
| 61

3. 3. Three objects of immediate sight
If what Berkeley means by immediately seen is merely the stimulation of light on
the eye, then this is the kind of physiological consideration that sets up problems to
be solved by an account of visual perception.
41
This seems to be what Berkeley aims
to bring to his readers attention when referring to what is agreed by all. After Gary
Thrane (1977, p. 246), let us make this explicit and call Berkeleys famous one point
argument, the physiological argument. Here it is again for the readers convenience:
It is, I think, agreed by all that distance, of itself and immediately,
cannot be seen. For distance being a line directed end-wise to the
eye, it projects only one point in the fund of the eye, which point re-
mains invariably the same, whether the distance be longer or shorter.
(Berkeley, 1732/2008, 2, p. 7)
As David Malet Armstrong points out: The argument seems to be valid only if we
assume that the immediate object of sight is the fund of the eye (1960, p. 9). As-
sume, for arguments sake, that it is valid on the physiological reading; could this
provide any support for claim (v) above, namely that one ought not to attribute rela-
tive or absolute distance to the direct objects of visual experience?
In this connection one might also seek to vindicate all the talk about painting in the
last chapter. For example, Molyneux wrote in his Dioptrica Nova: [T]he rays from
each point of the object [...] determine [...] on the retina, there painting distinctly
the vivid representation of the object (Molyneux, 1692, pp. 104 - 105; as cited in
Turbayne, 1955, p. 346, emphasis mine). One could reformulate those earlier re-
marks, as saying simply that the pictures of external objects are painted on the ret-
ina or fund of the eye; that we can see nothing which is not so painted (Berkeley,

41
Though, this emphasis blurs a broader issue. In ordinary viewing conditions, light
sampled by the eye is itself structured and distorted by the surfaces and media of the envi-
ronment in lawful ways (Gibson, 1966, pp. 208 - 220; 1986, pp. 65 - 93). As a consequence,
there are limits in what can be discovered about the immediate input to the visual system
through mathematically modelling the projection of rays to the back of the eye (Palmer,
1999, pp. 15 - 21).
62 |
1732/2008, 88, p. 40) But are the pictures or paintings on the retina the direct
objects of visual experience?
42

The relationship between visual experience and the retinal image is obviously com-
plicated by the inversion of the latter.
43
And Berkeley seems acutely sensitive to
this(ibid., 88, p. 40) (ibid., 88, p. 40). These two facts make the exegetical situa-
tion rather fuzzy. For instance, according to Turbayne, Berkeley only grants that
retinal images could be objects of experience (of any sort) for the sake of argument
(see Turbayne, 1955, p. 345ff.). On this reading, the only sensible construal of them
as such would be as derived from seeing the eyes of another body:
44
When we think
of the pictures in the fund of the eye, we imagine ourselves looking on the fund of
anothers eye, or another looking on the fund of our own eye, and beholding the
pictures painted thereon (Berkeley, 1732/2008, 116, p. 50).
45
The image would be
seen as an inversion of what it depicts. But this also sets up a contradiction when it
comes to the issue of whether we directly experience such an image, putatively pro-
jected to the back of our own eyes. For if we do experience the retinal image (direct-
ly), and if it is in virtue of this that we see an inverted image in the others eye (indi-
rectly), then the image that we experience cannot be inverted at all. Hence, on the
assumption that retinal images are inversions, and we do not directly experience

42
Hatfield and Epstein (1979) provide a nice history of the concept of a conscious
state with the phenomenal properties of the retinal image (emphasis mine), which they dub
the sensory core (p. 363). According to them, Descartes work on vision is the earliest ex-
plicit instance of a distinction between perception of the world and a sensory core. But the
notion seems operative in Keplers writings (though he did not employ such a distinction)
and may in essence be traced back to the Islamic philosopher Alhazen (p. 367 ff.).
43
The inversion of the retinal image should not be confused with the inverse prob-
lem of recovering the specific layout of the environment from the retinal projection (see
3.4 below).
44
Admittedly plausibility here is judged in the context of various of Berkeleys theoret-
ical commitments, such as his distinction between proper and improper ideas of sight, his
strongly held position on the heterogeneity of the senses, and the exclusive availability of
tangible ideas (amongst which ideas such as up and down, and magnitude are included) to
the sense of touch. See Turbayne (1955), and 3.2 above.
45
Gibson provides a nice description of how this might occur:
The formation of an image on the retina can be observed directly. If the excised eye of an
albino rabbit is fixed into a hole in a card and pointed toward a scene, by holding it in front of
one's own eye, we can actually see the inverted image on the curved rearward surface, looking
something like a miniature photographic transparency. (1950, p. 48)
| 63

anything inverted, we do not directly experience retinal images (cf. Berkeley,
1732/2008, 116, p. 50).
46

Turbayne cites textual evidence suggesting that such an argument would be availa-
ble to Berkeley. That is, on his interpretation Berkeley could deny that we visually
experience retinal images. But as Armstrong points out, this seems to lump Berkeley
with a prima facie contradiction of his own. For if Berkeley were to entertain such an
argument, he would no longer be in a position to run the physiological argument.
And if that is correct, then he would no longer be able to claim that distance cannot
be immediately visually experienced (Armstrong, 1956; 1960, pp. 45 - 52). Assuming
that is indeed Berkeleys aim, then the contradiction could be avoided only if he can
avoid making recourse to retinal events. Indeed, he seems to be running a more
direct version of the argument in the following dialogue:
Philonous: [I]s not distance a line turned endwise to the eye?
Hylas: It is.
Philonous: And can a line so situated be perceived by sight?
Hylas: It cannot.
Philonous: Does it now therefore follow that distance is not properly
and immediately perceived by sight?
Hylas: It should seem so. (Berkeley, 1734/2008, Dialogue I, p. 185)
Thrane (1977) calls this the phenomenological argument, for the purposes of explor-
ing a reading of perceived in this passage as perceptually experienced. It would
serve as an alternative route to (v), on the key assumption that the construal of dis-
tance as a line perpendicular to the eye is adequate to describe the experience of
depth.
A powerful attack on this assumption finds its source in some remarks made by
Maurice Merleau-Ponty. In characteristically ironic style, he remarks that we cannot
see the line because it is constituted by a series of points that one is perpetually
badly placed to see (Merleau-Ponty, 1945/1962, p. 255). The thought is presumably
based on the following passage: For by the distance between any two points noth-
ing more is meant than the number of intermediate points. If the given points are
visible, the distance between them is marked out by the number of the interjacent
visible points (Berkeley, 1732/2008, 112, p. 48). Merleau-Ponty comments that this
concession, that the distinction between points could be seen if they were viewed in
profile, reveals that depth is being treated as it is presented to a spectator standing

46
The retina is also concave. For contrasting views on the putative phenomenological
implications of projection onto a concave surface see Lehar (2003, pp. 39 - 42) and Schwitz-
gebel (2006, pp. 591 - 592).
64 |
at the side, in short as breadth (op. cit., p. 255; see also Armstrong, 1960, p. 15).
What needs to be demonstrated, if this construal is to do any work in the service of
the phenomenological argument, is that visible depth is not a suis generis visual
experience. The situation is reminiscent of the discussion of VF
1
& VF
2
in the last
chapter. A dissenter can admit that it is possible to restrict a description of the visual
field to two dimensions, as when describing a star as just to the left of a treetop
(Grush, 2007a, p. 425). But it is a step further to somehow reduce all phenomena of
visible depth to two dimensions, and Berkeleys proposal at least is unable to deliv-
er.
47
When faced with Merleau-Pontys objection, one might retreat slightly and
claim that the conception of distance as a line is an abstraction that is not meant to
be experienced per se, but merely feature in an explanation of ones experience. But
such a retreat would be unavailable to Berkeley, as he repeatedly insists that all fea-
tures of ones psychology should be at least potentially available to direct perception
(Berkeley, 1732/2008, 9 - 10, p. 8, and passim). And if perception here means per-
ceptual experience then the incompatibility is manifest.
In fact, reading perception as perceptual experience is crucial in Berkeleys rejec-
tion of the geometricians explanation of how vision is able to transcend the limits
of visual sensation. According to this received view, the operations of the mind in
computing distance from an impoverished sensation are like those of a blind man
computing distance by holding two crossed sticks in contact with an object. By
means of an innate, God-given capacity, the mind is supposed to achieve estimates
of absolute distance in part by determining the distance between the eyes (hands)
and the degree of each eyes (sticks) vergence angle. Berkeleys dissent on this point
is simple. He is willing to accept that this description might adequately describe the
optics of the situation, but not the psychology. In staunch opposition, he declares:
In vain shall any man tell me that I perceive certain lines and angles which intro-
duce into my mind the various ideas of distance, so long as I myself am conscious of
no such thing (, 12, p. 9). Though (no party claims that these reasoning processes
need to be conscious or attributable to the perceiver herself, Berkeleys issue is the
fact that they cannot be. His worry is that brutes and children could not even in

47
The account gets along fine if objects are separated at a certain breadth. But, for
instance, when two objects occupy the same positions by breadth and are yet seen as dis-
tinct, in most cases this is naturally described as experiencing one object occluding another
at a visible depth. Similarly, distinct objects revealed through dynamic changes in the envi-
ronment are also rather difficult to account for, such as the accretion and deletion of visible
surfaces through changes in vantage point and/or object location (Gibson, 1966, pp. 203 -
204; 1986, pp. 82 - 83).
| 65

principle be conscious of the operations that supposedly constituted their psychol-
ogy (Atherton, 1990, pp. 79 - 83).
48

To further underline the point, consider Berkeleys treatment of the Barrow illusion.
According to an account of monocular vision strongly tied to early modern optics, a
distant object will project rays closer to convergence than a nearer object. This sug-
gests the following prediction: the greater the convergence of the reflected rays
striking the eye, the further the object reflecting that light will appear to be. But
apparently in the Barrow illusion exactly the opposite happens: as the object is
drawn away, it appears closer (Berkeley, 1732/2008, 29, pp. 12 - 15). In one version of
the illusion, a double convex lens is interposed between the eye and the object,
causing the rays reflected from the object to converge and come to a focus before the
retina, resulting in a broad distribution of rays on the retina (see Figure 8).
49
This
result is some ways similar to the optical situation when objects are very close, the
rays that they reflect are broadly distributed on the retina, though in these situa-
tions this is because the rays come to a focus behind the retina(34 - 35, pp. 16 - 17)
(34 - 35, pp. 16 - 17).
50
As Berkeley recounts, the Barrow illusion is described as a
puzzle by the geometric theorists because despite the differing optical facts, in both
cases the psychological outcome is a blurry visual experience. He then takes this as
an opportunity to demonstrate the superiority of his associationist theory, claiming
that the usual and thus associated distal cause of a blurry visual experience is an
object that is very close. Hence it is no wonder that despite the fact that the object is

48
If this sounds strange to the reader it is because what Berkeley rejects is close to
common practice today. A number of conventions have been developed for attributing psy-
chologically salient states that are not required to be constrained by the conceptual reper-
toire of the subject, or any distinctive vocabulary that picks out their behaviour qua rational
agents. And interestingly several principles of Descartes psychology have clearly been inher-
ited in at least the orthodoxy of these (otherwise) more modern forms of psychological ex-
planation (see Wheeler, 2005, pp. 21 - 54). For a view which requires mental states to be con-
nected to the conscious states of a subject to which they are attributed, see Searle (1990,
1991). See Ludwig (1996) for a refinement, and similar considerations directly applied in the
context of modern vision science.
49
The other setup of the illusion uses a concave mirror, such as a shaving mirror, and
places the object behind the head of the viewer. So, upon retreating ones head away from
the image, one in fact moves closer and closer to an object that also appears closer and clos-
er. For a rich discussion of this version and a variety of historical insights, see Lennon (2007).
50
Naturally, this is resolved by adjusting the curvature of the lens to refocus the pencil
of rays on the retina, known as accommodation (Palmer, 1999, pp. 203 - 205). The need for
this arises less frequently (if at all) in small and primitive eyes. For discussion of the relevant
anatomy and physiology from an evolutionary perspective, see Gibson (1966, pp. 171 - 173).
66 |
being moved further and further away, by dint of the associations with its blurry
look, it is experienced as nearer and nearer(36, p. 18) (36, p. 18).
51

Figure 8 Berkeley's treatment of the Barrow Illusion

Berkeleys relationship with the phenomenological argument is then somewhat un-
stable. For on the one hand, a conception of visual depth experience as a line turned
endwise to the eye lacks motivation. And on the other hand, Berkeley cannot appeal
to the optical context where a conception of depth as a line turned endwise is well
motivated, as he rejects geometrical descriptions of vision as limited in their ex-
planatory reach, specifically with respect to visual experience.
Thrane (1977) helpfully uncovers yet another variant of the form of argument we
have been considering. This one is again in the form of a dialogue, and found in one

51
This point holds whether it might be admitted that it does not provide incontrovert-
ible support for his campaign against the geometrists (see Lennon, 2007, p. 39 ff.).
(A) Rays from a distant object come to a focus on the retina.
(B) Diverging rays from a near object come to a focus behind
the retina (before accommodation).
(C) The Barrow illusion produced with a double convex lens:
converging rays to come to a focus in front of the retina.
Adapted from Berkeley (1732/2008, 35, p. 17), 2008 Des-
mond M. Clarke.
| 67

of Berkeleys later works, Alciphron. It also incorporates several claims we have dis-
cussed already. First there is the claim that we do visually perceive distance, embod-
ied in Alciphrons prompted report:
Euphranor: Look, Alciphron, do you not see the castle upon yonder
hill?
Alciphron: I do.
Euphranor: Is it not at a great distance from you?
Alciphron: It is. (Berkeley, 1752/2008, Dialogue IV, 8, p. 278)
Then there is the claim that distance is a line perpendicular to the eye, and that such
a line would not project to more than one point on the retina, and that the appear-
ance of any distance will involve projection to the same point:
Euphranor: Tell me, Alciphron, is not distance a line turned end-wise
to the eye?
Alciphron: Doubtless.
Euphranor: And can a line, in that situation, project more than one
single point on the bottom of the eye?
Alciphron: It cannot.
Euphranor: Therefore the appearance of a long and of a short dis-
tance is of the same magnitude, or rather of no magnitude at all, be-
ing in all cases one single point.
Alciphron: It seems so (ibid.)
Finally, there is the claim that the preceding claims together motivate an indirect
account of the perception of distance:
Euphranor: Should it not follow from hence that distance is not im-
mediately perceived by the eye?
Alciphron: It should.
Euphranor: Must it not then be perceived by the mediation of some
other thing?
Alciphron: It must. (ibid.)
In the voice of Euphranor, Berkeley seems to move from an optical fact to a conclu-
sion that perception is indirect, via a claim about the optical basis of perceptual
68 |
experience. Thrane (1977, p. 246) accordingly refers to it as the physico-
phenomenological argument.
This argument has a disputed status amongst Berkeley scholars. For instance, Arm-
strong thinks that it is absurd to think that we visually experience retinal images,
and therefore that Berkeleys argument is invalid (Armstrong, 1960, p. 9). But Tur-
bayne seems to think that, because it is absurd to think that we see retinal images,
Berkeley must be only assenting to this claim for the sake of argument (Turbayne,
1955, p. 345ff.). Turbayne dubs this absurdity an instance of the psychologists falla-
cy. A putative error in reasoning that involves the supposition that we may be di-
rectly aware of things which we can only know about, a paradigm case of which
would be what occurs on the retinas of the optically and physiologically ignorant (p.
351). To the contrary, though, one might think that both the above are guilty of what
Thrane coins the philosophers fallacy, namely, the view that what we do not
know about cannot happen or exist. To illustrate he suggests that whilst we do not
know by seeing the retina that we see light impinging there [...] it scarcely follows
from this that we therefore do not see the light that (as it happens) does impinge on
the retina (Thrane, 1977, p. 259).
In fact, the interpretive disagreements and accused fallacies can all be seen as the
product of Berkeleys conflation of perception and perceptual experience. For even
if retinal events are a far cry from visual experience, one might still accept that per-
ception involves seeing light. And if that is true then the light that that hits the reti-
na is the light we see. In essence, the discussion reveals that Berkeley might have a
variety of intentions, and it is not clear which, when he says that what we immedi-
ately and properly see are only lights and colours in sundry situations and shades
and degrees of faintness and clearness, confusion and distinctness (Berkeley,
1732/2008, 77, p. 35). One is not particularly ambitious: he might mean simply to
point out that it is only light that hits the photoreceptors in the retina with varying
degrees of intensity and focus. But perhaps he means something less bland than
this. Perhaps he means to persuade us that because light is all that hits the retina, all
that we immediately visually perceive is light. Indeed, he might intend to imply this
and more. He might intend to persuade us that because all that we immediately
visually perceive is light, all that we immediately visually experience is light. This
would be exemplary of the strategy described at the end of the last chapter.
| 69

3. 4. Sensation, perception, & perceptual experience
In providing the grounds for his account of visible depth, Berkeley provides a variety
of LAs, and these arguments have their ambiguities. It is not clear whether Berkeley
means to persuade us that:
(a). all we immediately visually perceive is light;
or that:
(b). all we immediately visually experience is light;
or indeed:
(c). both the above.
It is also not clear whether Berkeleys intention is to persuade us of anything more
than the merely incidental fact that:
(d). all that hits the retina is light.
52

However, limiting Berkeleys purposes to (d) is certainly implausible. Berkeleys in-
tentions must accord to at least option (a), and more likely (b) and (c); and this is
the source of the trouble.
As I conveyed at the beginning of 3.2, a Berkeleian object is a collection of sensa-
tions, the existence of which is dependent upon perception. Indeed, [t]he things
[...] immediately perceived are ideas or sensations (Berkeley, 1734/2008, Dialogue II,
pp. 196 - 197, emphasis mine). This is a thesis that contrasts his view from one Locke
might have held, in claiming that sensations are the only objects of perception. But
it would be the most basic of logical errors to conclude from that the claim that all
sensations are perceived. It is still possible that, e.g., a visual sensation is merely the
unperceived event of light hitting the retina, in which case the existence of objects
would depend solely upon other ideas of sense that are perceived. However, what
gives us reason for thinking that Berkeley wishes to assert more than (d) is the ar-
chitecture of Berkeleys theory of visual perception. For if Berkeley does not identify
the immediate object of visual perception with the light that hits the retina, then he
cannot provide requisite motivation for his theory. Furthermore, discussion of the

52
Berkeley is certainly not the only one to equivocate on these matters. In more recent
times, OShaughnessy provides a discussion of the visual perception of light unhelpfully
equivocated with the visual experience of light (1984, compare pp. 193 and 200).
70 |
phenomenological version of the one-point argument gives reason for thinking that
Berkeley is being more ambitious than stating a brute physical fact. For apparently,
an explanation that posits operations upon lines and angles (in virtue of which per-
ceivers putatively compute distance) is offered in vain so long as Berkeley himself is
conscious of no such [lines and angles] (1732/2008, 12, p. 9). Similar examples
abound.
53
Unless Berkeley is employing a double standard, he is committed to
something like (b); and given reasons for attributing (a), then (c) seems equally
likely.
In fact, very early on Berkeleys ideas of sense were given such an interpretation. In
an essay critically reviewing Berkeleys Sentiments of Ideas Thomas Reid describes
an idea of sense (sensation by another name) as the act, or feeling of a sentient
being, the very essence of which consists in its being felt (1785, Essay II, Chapter
XI, p. 179). And in doing so, he explicates the claim that their esse is percepi as the
claim that their esse is sentiri: As there can be no notion or thought but in a think-
ing being, so there can be no sensation but in a sentient being (ibid., pp. 178 - 179)
(ibid., pp. 178 - 179).
54
This explicit appeal to sentience marks a phenomenological
reading of sensation, for which I will use the term sensory experience.
Reid uses the word feeling to mean the same as sensation in some cases (such as
above), but not in others. When feelings are not simply sensations, they denote the
perceptions we have of external objects, by the sense of touch (ibid., Essay I, Chap-
ter I, p. 34, emphasis mine). For instance, when feeling a body to be hard or soft,
rough or smooth, hot or cold; to feel these things, is to perceive them by touch

53
E.g., he is convinced of the falsity of the geometrical theorys postulation of
judgements concerning the convergence and divergence angles of the eyes by his own expe-
rience, since he is not conscious of employing the turn of the eyes to make such judge-
ments (Berkeley, 1732/2008, 19, p. 10).
54
By sentience Reid refers to what people today might call qualitative experience or con-
sciousness (cf. Nichols, 2007, p. 83). This latter term means something closer to introspec-
tion for Reid, namely that immediate knowledge which we have of our present thoughts
and purposes, and, in general, of all the present operations of our minds [...] and not of ex-
ternal things (1785, Chapter I, Essay I, pp. 17 - 18). Hence it is improper to say I am con-
scious of the table which is before me. I perceive it, I feel it, but do not say I am conscious of
it (ibid. p. 18). So he certainly has his own reasons for not using the word conscious here,
but none that conflicts with the interpretation. Sensations can be objects of consciousness
(in his sense) though not objects of perception (in his sense).
| 71

(ibid., p. 34).
55
But in contrast to a perception, a sensation proper is an act of mind
distinguished from all others by the fact that it hath no object distinct from the
act itself (ibid., p. 33). In short, Reid claims that the act/object analysis of sensa-
tion, exemplified in Berkeleys insistence in changing ideas into things (Berkeley,
1734/2008, Dialogue III, p. 225) is simply not applicable to sensation.
56
If the analysis
were applicable, then sensation and perception would be on a par in possessing ob-
jects, and indeed here lies the central difference between Berkeleys view and Reids.
Reid ventures: Suppose I am pricked with a pin [...] Is the pain I feel, a sensation?
Undoubtedly it is, and then asks Is the pin a sensation?, to indicate that un-
doubtedly it is not (Reid, 1785, Essay II, Chapter XI, p. 180). On Reids view, the pin
and every property it instantiates is necessarily independent of any sensation and
any property it can instantiate. For a sensation cannot have length and thickness,
and figure and weight etc. (ibid. p. 180). But for Berkeley a sensation can instantiate
all these properties. If objects are collections of sensations then by Leibnizs Law the
properties of objects are the same as the properties of sensations. On his view, any
act of sensation just is an act of perception.
Keeping in mind the ontological neutrality of the direct object of perceptual experi-
ence (see 1.5 above) can help focus on the fact that Berkeley and Reid propose con-
trasting accounts of both sensory experience and perceptual experience. On Reids
analysis a pain sensation itself is all feeling and no object, whereas the perceptual
experience of a pin is most definitely that of an object. On Berkeleys analysis the
experience of pain and the experience of a pin qua ideas of sense are on a par qua

55
Just as realism is superficially congruent with idealism, so Reids treatment of per-
ception is superficially congruent with Berkeleys: there is room for the reality of objects in
Berkeleys system, albeit an immaterial reality. For when Reid asserts that in sensation the
feeling and the thing felt are one and the same (1785, Essay I, Chapter I, p. 34), Berkeley
could reply (in agreement) that those immediate objects of perception, which according to
you are only appearances of things, I take to be the real things themselves (1734/2008,
Dialogue III, p. 225). Hence, the difference that makes a difference is not found in the meta-
physics of perception per se, but rather in the metaphysics of sensation; and, in the end (for
our purposes at least), the implicitly proposed phenomenology of sensation.
56
Berkeleys equivocation between sensations and ideas of sense is a little frustrating
here: Are the sensations objects (qua ideas)? Are there objects (qua ideas) of sensation?
Thankfully, Reid would answer both in the negative. Admittedly, in an earlier work, Reid
(1764/1997) does occasionally seem to say that sensations are perceived. But plausibly in
those cases he is merely falling in with familiar Berkeleyan [sic.] language in emphasising
their mind-dependent nature (Pappas, 1989, p. 158).
72 |
sensory experiences.
57
The issue the between the two views then, is whether we both
perceptually experience the pain and the pin, in the same way at the same time.
To claim that we do perceptually experience both the pain and the pin, in the same
way at the same time, is the kind of bold statement characteristic of an account of
perceptual experience as indirect. But it is not clear now how it could be motivated.
The difficulties Berkeley would have on this count can be brought out by attention
to some of his equivocations. Hold judgement for a moment on whether Reids in-
terpretation of Berkeleian sensation is correct; that is, entertain a reading of sensa-
tion that does not necessarily mean sensory experience. A key motivation for the
kind of theory of perception Berkeley provides is the proposal that visual sensation
is limited (see 3.2). In particular, the motivation is to show the need for an account
of the ways in which visual perception is not limited. We do see distance, and this is
a psychologically interesting fact that needs explaining (see 4.1). The claim that
sensations are perceived is a complication that can be transformed into a positive
aspect of the theory. To keep from claiming that sensations are both limited and
not, Berkeley needs to draw a distinction between sensations that are directly and
indirectly perceived, in which case the latter need not be limited in the same way as
the former. But one can accept all this, and still claim that
(e). Sensations of light on the retina are not sensory experiences
As I noted at the beginning of the section, and at more length in the last section,
there are reasons for thinking that Berkeleian visual sensations are sensory experi-
ences. But then, given that the objects of perception are sensations, and given that
the perception of depth is an explanandum for the theory, Berkeley seems to head
into a contradiction in his use of LAs: visual sensory experience is claimed to be
both limited and not so limited. Again, Berkeley could be rendered consistent here,
by marking a distinction between direct and indirect visual sensory experience. But
although that is reasonable as an account of visual perception, grounded in optical
and perhaps physiological facts, it has little to no motivation as an account of per-
ceptual experience. For why should we accept the premise that visual sensory expe-
rience is limited?
In fact, one is led to the claim that visual sensory experience is limited by accepting
the assumption that all sensations are perceptually experienced. To be maximally
charitable to Berkeley, in the course of a dispute over the correct description of per-

57
One way of making Berkeleys view more sensible would be to take a perceptual
model of bodily experience (cf., Ayers, 1993, pp. 162 - 163). But in that case, it is still arguable
that it is the body part that is perceptually experienced rather than the sensation itself.
| 73

ceptual experience, perhaps the appropriate question is: Why should one not think
that sensations are perceptually experienced? More plainly: Is there even a contrast
between sensation and perceptual experience? It is to this issue that we now turn.
3. 5. Perceptual sensation
At this point, it is worth explicitly stating the common-sense conception of percep-
tual experience as the direct perceptual experience (DPE) thesis. Stated as the nega-
tive of the indirect perceptual experience thesis (IPE), DPE is as follows:
S indirectly perceptually experiences x if and only if S perceptually experi-
ences x in virtue of perceptually experiencing y, and x and y are distinct.

S directly perceptually experiences x if and only if S perceptually experiences
x, and S does not indirectly perceptually experience x.
To ground the dispute between IPE and DPE, consider: What could the case of the
man born blind imply about visual sensory experience in general? A defender of
DPE might argue that it implies nothing. She could accept Berkeleys account of the
mans visual experience and still claim that it has no bearing upon ordinary visual
sensory experience. What Berkeley provides, and what his equivocations (and occa-
sionally his argumentative strategies) unhelpfully blur, is a powerful account of the
visual perception of depth and other spatial properties. Part of its power stems from
its ability to account for the visual experience of depth. And part of its intrigue is
appeal to the body, qua organ of touch. She could accept all this and yet deny that
there is (ordinarily at least) a form of visual sensory experience that involves any-
thing less than that which Berkeleys theory so elegantly explains, viz. the visual
experience of depth.
To further clarify, recall some earlier claims (from 4.1.):
(i). Visual sensation occurs by means of the energetic impingement of light
upon the retina
(ii). The only spatial properties one ought to attribute to the direct objects of
visual sensation are those licensed by the means identified in (i)
(iii). The only spatial properties one ought to attribute to the direct objects of
visual experience are those licensed by the means identified in (i)
74 |
The defendant of DPE above holds that (iii) is false in all cases excepting those simi-
lar to the man born blind, at that at any rate (iii) does not follow from (i) and (ii).
58

In a last gasp effort, a proponent of IPE could exploit an implicit concession. For
implicitly, the DPE advocate admits that if visual sensation is limited then it is lim-
ited as a form of sensation that is not usually conscious. And prima facie this seems
to trivialise the role of sensation in perceptual experience. For instance, consider
Reids example of smelling a rose. Roses have a light fragrant odour, and smelling
the odour just seems like smelling a rose. Reid used this as an example of a case in
which a sensory experience (of an odour) is inextricably interwoven with a percep-
tual experience (of a rose), and he intended to be an illustration of what is generally
the case. Now consider Berkeleys account of visible depth as mediated by tactile
sensations. A critic of the account might say that it is a wonder why visible depth
seems visual rather than tactile. But a ready response is that the occurrence of visual
sensations makes the sight of visible depth a visual experience rather than a tactile
experience. And in fact, the point has an even broader application. For it would en-
able a distinction between perceptually experiencing an object and conceiving it, in
that in the former case but not the latter there is a specific sensory character to the
experience.
There is a very obvious constraint here operating here, call it the sensory mode con-
straint:
A perceptual experience can only be a perceptual experience of a given mode
in virtue of the occurrence of sensations distinctive of that mode.
59

58
In fact, she need not even concede the point about the man born blind. She could
concede that the man born blind, upon first being made able to see, would certainly per-
ceive light (say, upon being placed in front of a sunlit window), and have a sensory experi-
ence of illumination. But she would not need to make the further concession that the man
would perceptually experience anything in sight; i.e., she could argue that this would be a
case of perception without perceptual experience. See A.D. Smith (2006, pp. 420 - 421) for a
discussion along these lines. Presently, I will construe the DPE theorists position as more
generous, as the way in which an IPE theorist might attempt to exploit the concession turns
out to be useful and illustrative.
| 75

What this shows is the need for the DPE advocate to clear up some further ambigui-
ty in the role of sensation in perceptual experience. First consider what I will call a
perceptual sensation:
A sensory experience of a given mode, in virtue of which a perceptual expe-
rience occurs and is an instance of that mode.
Perceptual sensations bridge sensory and perceptual experience. There are two
points of note here. One is that it is in virtue of perceptual sensations that sensory
experience seems perceptual. I will say a little more about this shortly and a lot more
in the next three chapters. A second is that it is in virtue of perceptual sensations
that perceptual experience seems sensory.
60
This extends the sensory mode con-
straint to state a further condition on perceptual experience, which translates to
Perceptual experience in mode x iff. perceptual sensation in mode x
If something like the sensory mode constraint is true then any adequate description
of perceptual experience ought to embrace perceptual sensation. And in doing so
the DPE theorist can resist the IPE theorists attempts to exploit the concession that
there is a form of sensation that is usually not conscious.
In fact, the concession can be turned into a positive claim by denying that the occur-
rence of perceptual sensation is a brute fact. Centuries of research on the psycholo-
gy and neuroscience of perceptual systems throw considerable doubt on that. More
to the point, insisting that the occurrence of perceptual sensation is simply a brute
fact rules out these avenues as a source of possible explanations. To accommodate
the idea that perceptual experience has a deep structure, what is needed is an ap-

59
An alternative would be fine, like: A perceptual experience can only be a perceptual
experience of a given mode in virtue of the occurrence of sensory experiences in that mode.
Here the necessary connection between such sensory and perceptual experiences is just
bluntly stated, rather than being suggested by the fact that sensory experiences can be dis-
tinctive of a mode. This has the virtue of avoiding the controversy over whether to distin-
guish the senses phenomenologically (after Grice, 1962/2002) or by their proper object
(perhaps after Aristotle, see Sorabji, 1971), or by some other means. But it does not make the
point as simply.
60
These comments are merely a reflection of a consensus in the literature. For in-
stance, OShaughnessy uses the term perceptual sensation to illustrate the first point
(though not with the same strength that I hold it) when he marks a distinctive type of sen-
sation that is involved in such perceptual experiences as seeings and hearings (1980/2008,
p. 204). It is also used in reference to the second point by A.D. Smith when asserts that per-
ceptual experience is sensuous in a way that mere thought is not because it involves percep-
tual sensation (2002, p. 61).
76 |
propriately neutral way of characterising the enabling conditions for perceptual
sensation. Consider in this regard the notion of a mere sensation:
A sensory event that is necessary for the occurrence of a perceptual sensa-
tion
This provides enough room for a variety of architectures of perceptual experience:
Mere sensation is one of potentially many individually necessary and jointly suffi-
cient conditions for perceptual sensation. Furthermore, the notion of mere sensa-
tion qua sensory event is general enough to accommodate the various notions of
visual sensation (VS) that we have discussed so far. By way of summary:
VS
1
: energetic impingement of light on the retina
VS
2
: unconscious sensation of light
VS
3
: sensory experience of light
All these could be mere sensations, and all these could be involved in ordinary visu-
al experience. And significantly, they are all forms of sensation that we have seen
invoked in the course of LAs (see 3.3). What the distinction between mere and
perceptual sensation unmasks is a fallacy of equivocation inherent in this kind of
argument as applied in account of perceptual experience. In particular, the infer-
ence from the limitation of mere sensations to the limitation of perceptual experi-
ence is invalid. For what is required is an inference from the limitation of perceptual
sensation to the limitation of perceptual experience.
The equivocation is understandable, as it is buried amongst a lot of cross-talk about
the immediate object of perception. Claims might be reasonable or trivial when
concerning perception. The same claims might be unreasonable and non-trivial
when concerning perceptual experience (and vice versa). But the dissociations are
sometimes subtle and easy to miss. Indeed, if what I have said above is correct, then
Berkeley does fail to note the relevant subtleties. And the reason is likely because he
fails to distinguish between mere and perceptual sensation. The visual sensations of
the man born blind and the unbodied spirit are not perceptual sensations. Sensa-
tions (of light) with the same organic bases might in some contexts be mere sensa-
tions and in other contexts be perceptual sensations. The difficulty is to not confuse
one for the other.
To illustrate, recall a discussion from the end of last chapter, where it was charitably
suggested that an IPE theorist could draw upon prosthetic perceptual experience as
a case in its favour. Paying attention to the distinction between perceptual and mere
sensation provides the means for rejecting an IPE description of prosthetic percep-
tual experience. When the prosthesis is actively used by a practiced user, the vibra-
tory stimuli are mere sensations that are inter alia necessary conditions for percep-
| 77

tual sensations. They are not perceptual sensations themselves, unless perhaps, the
user of the device is immobile.
To demonstrate just this point, Lenay et al. (2003) provide a vivid example of a study
by Lenay, Canu and Villon (1997) in which a single photoelectric cell with a receptive
field roughly corresponding to a 20 cone was affixed to a nger and coupled to a
single vibrator held in the other hand. Vibration was all-or-nothing according to a
threshold in the photoelectric cell. After a short period of freely moving the photoe-
lectric receptor and actively producing tactile stimulation, the authors claim that a
blind-folded subject typically becomes conscious of the presence of an object [sc., a
light source] situated in the space in front of him. Apparently it seems to the subject
that the temporal succession of sensations derive from different contacts with a
single distal object (Lenay et al., 2003, p. 282).
Moreover, this provides a direct illustration of the double province of sensations, in
this case sensation of vibration:
[A]rticial stimuli produced independently of the movements of the
nger on which the photoelectric cell is placed are not associated
with a distal perception, but continue to be perceived proximally at
the level of the skin. Similarly, if the movements cease, the distal spa-
tial perception disappears (Lenay et al., 2003, p. 282)
As seems to be intended, we should take perception here to mean perceptual ex-
perience. The contrast is between a case where a succession of vibratory sensations
results in the perceptual experience of a distal object and a case where the same
sensations result in perceptual experience of proximal object. In the former case,
the vibratory sensations are mere sensations that in the context of associated sensa-
tions of movement yield perceptual sensations in a nondescript prosthetic sensory
mode. Without the associated movement, they are perceptual sensations of vibra-
tion on the surface of the skin.
A further difference between the two cases, as described at least, is the object of the
experience. This tracks the change between vibratory sensations being perceptual
and mere sensations. But more to the point, it is conducive to the claim that percep-
tual experience always seems to be not merely direct, but directed. This is a familiar
theme in defences of direct realism, namely, that there clearly is some phenomeno-
logical distinction to be made between perceptual experience and mere sensation,
in that the former does indeed appear to be, in some sense, object-directed in a way
that the latter is not (A. D. Smith, 2002, p. 58).
This chapter and the last two chapters have provided a defence of the directness of
perceptual experience in the context of an assumed neutrality of the direct object of
78 |
perceptual experience. What we need now is a defence of the claim that perceptual
experience can be described as seeming to be directed upon reality, whilst maintain-
ing said neutrality. That is the aim of the next part of the dissertation.
| 79

PART II

The appearance of a thing (the experience) is not the thing that appears
(that supposedly stands over and against us). In the integration of con-
sciousness we experience the appearance as being in the phenomenal world,
things appear to us. The appearance itself does not appear, it is experienced.
(Husserl, 1900 & 1901, V, 2; as cited & translated by Hickerson, 2007, p. 79)
80 |
Chapter 4 The neutral intentionality
of perceptual experience
4. 0. Introduction
In 1.5., I drew a moral from the toils of sense-datum theorists: thou shalt remain
ontologically neutral in giving an account of perceptual experience. At this point it
is useful to make an addendum: making certain ontological assumptions is all well
and good if you are in the business of doing metaphysics (provided that you argue
for these assumptions at some point). I am not in that business, so I try and adhere
strictly to the moral. Yet a number of contemporary philosophers discussing percep-
tual experience are really discussing its metaphysical status, and typically this
comes in discussion of the intentionality of perceptual experience. In this chapter, I
discuss the intentionality of perceptual experience with a view to clarifying my neu-
tral stance on the metaphysics of perceptual experience.
I begin (in 4.1.) with an illustration of two instances of the move that I want to
avoid. The first will be already familiar from the first chapter, namely, the phenom-
enal principle. The second will be discussed further in the next chapter; it is the
claimed transparency of perceptual experience, of note here because it is typically
appealed to in claiming that perceptual experience is an intentional state. I then (in
4.2) move on to discuss intentionality itself, giving brief overviews of disputes over
the history of intentionality and the problem of intentionality as it is attributed to
Franz Brentano. To make a start on relating intentionality and perceptual experi-
ence, I discuss (in 4.3) ways in which intentionality has been treated that are either
inadequate for describing perceptual experience or inadequate for addressing the
basic problem of intentionality. After that, I focus on more contemporary discus-
sions (in 4.4), specifically a research programme that aims to solve the problem of
intentionality by providing a naturalistic theory of representational content. I close
(in 4.5) by distinguishing a neutral stance on intentionality from any stance that
involves adherence to a metaphysic of content, distinguishing the stance from sev-
eral notions of phenomenology as I do so.
4.1 Neutrality and perceptual experience
4.2 Intentionality
4.3 Intentionality and perceptual experience
4.4 The natural contents of perceptual experience
4.5 Phenomenology vs. neutral intentionality
| 81

4. 1. Neutrality and perceptual experience
There is a tendency in the philosophy of perception to run together issues concern-
ing the description of perceptual experience and issues concerning the metaphysics
of perception. As I mentioned at the beginning of chapter 1, this is essentially a mat-
ter of how the issues of the field are understood by those working within it. To give
an example, here is how one author characterises the distinctive questions, and
hence the distinctive problems, with which the philosophy of perception is con-
cerned:
rather than asking about how the mechanisms of perception work, it
asks what are the most general features of anything we could reason-
ably call perception or perceptual experience, and tries to give a
consistent and systematic description of them. It turns out that this
is a harder task than it might initially seem: quandaries, conun-
drums, and tensions within our conception of perception arise. I
claim that these problems are independent of the psychologi-
cal/scientific questions about perception, and also of the epistemo-
logical questions. The problems are phenomenological: they arise
from reflection upon perception as we experience it. (Crane, 2001a, p.
130)
But in fact these phenomenological problems are only problems because they lead
to conceptual problems, and at root these are problems in discerning the metaphys-
ical status of perception. To begin with an obvious example, take the phenomenal
principle:
aware which does possess that sensible quality (Robinson, 1994, p. 32)
When placed in the context of arguments such as the argument from perspectival
variation (see 1.2.), the phenomenal principle functions as an inference to the ex-
istence of sense-data. Stated on its own, it expresses a metaphysical thesis about
perceptual experience. M.G.F. Martin calls this actualism, describing it as the thesis
that whatever qualities one senses, some actual instance of those qualities and the
object which bears them must exist and be sensed (2000, p. 218). Actualism shares
its name with a more general metaphysical thesis, which denies that there are any
things which merely possibly exist.
61
What actualism about perceptual experience

61
At least: where the possible existence of x is in no way derivative of something actu-
al, x cannot exist. In a slogan: everything that exists actually exists. For a useful review of the
difficulties of unpacking this slogan (and similar) see Bennett (2005).
82 |
denies is that there are any things perceptually experienced that merely possibly
exist. According to Martins discussion, one way of making sense of why so many
otherwise very sensible individuals were lead to believe in the existence of sense-
data, is to understand them as attempting to do justice to the description of percep-
tual experience. As an example, he conveys a description of his experience as he
looks at a lavender bush and a straggling rose on the street of his North London
home:
[It] is evident to me that I am experiencing these things, and not
merely thinking about them, or imagining or remembering them.
The latter things I can do in the absence of the objects of perception,
but it does not seem to me that I can be this way, actually experienc-
ing, without the relevant objects or features present in my environ-
ment. This recommends Actualism to me, as defenders of the sense-
datum tradition have observed (2000, p. 219)
From the context of Martins discussion, it is clear that the sense in which things are
evident to him is not meant to hold any epistemological force. What he intends to
show is that direct realists and sense-datum theorists share an intuition, an intui-
tion that recommends actualism to them. In short, by embracing actualism is be-
ing true to the way the experience seems. But it is still a substantial issue whether
actualism is true. And relatedly, it is certainly controversial to claim that such ob-
servations can establish that Martin perceives a lavender bush and a straggling rose.
For one might deny that it follows from the fact that Martin perceptually experienc-
es a lavender bush and a straggling rose that there is a lavender bush and a strag-
gling rose which Martin perceives.
Another way in which a description is thought to recommend a metaphysical the-
sis is by reflection upon what has come to be known as the transparency of percep-
tual experience. Conveying the typical description, Martin writes:
When I stare out of the window I can see the lavender bush at the
end of the street, the straggling rose on my fence, and I can hear the
sound of traffic in nearby roads. When I reflect on what it is like for
me so to experience, these very same objects and features remain the
focus of attention as aspects of how I experience: this commits me to
Transparency with regards to perception (2000, pp. 218 - 219)
In the next chapter, I will examine two variants of the thesis that perceptual experi-
ence is transparent. At present it need only be noted the phenomenological claim
that perceptual experience is transparent has been taken to have implications for
the metaphysics of perception. In particular, it is thought by many to reveal a fun-
damental fact about perceptual experience: that the key to its metaphysical nature
| 83

lies in understanding its intentionality. This is a concept I will discuss further very
soon. But to give an initial grasp, what many of these authors have in mind is some-
thing along the lines that John Searle suggests (in appropriately vague terms) when
he says that: Intentionality is that property of many mental states and events by
which they are directed at or about or of objects and states of affairs in the world
(1983, p. 1). In the hands of some, the idea that perceptual experience is transparent
means both that introspection upon perceptual experience does not reveal any ele-
ments that are non-intentional and that this is to count against a theory that holds
that there are such elements to be found. Gilbert Harman, for instance, suggests
that insufficient attention to the intentional nature of perceptual experience as re-
vealed in noting its transparency, leads sense-datum theorists to identify the prop-
erties of the object of experience with the properties of something inner and men-
tal (2002, p. 665).
It is somewhat doubtful that the proponents of sense-datum theory in their adher-
ence to actualism were insensitive to the notion of intentionality. For instance, Rus-
sells (1905) paper On denoting, is widely recognised as a seminal work in the inten-
tionality of thought in the analytic tradition (see e.g. Jacob, 2010, 5; Siewert, 2008,
4). But my aim here is not to set the historical record straight.
62
For all I mean to
highlight is this: the idea that certain positions in the philosophy of perception are
motivated by reflection upon perceptual experience, exhibits a tendency to move
from claims about perceptual experience to claims about the metaphysics of percep-
tion that is of a rather general nature.
63
Indeed, the point of Martins (2000, pp. 218 -
219) discussion of actualism and transparency is to provide a straightforward char-
acterisation of direct realism as the combination of both theses. My discussion in
the two following chapters is an attempt to extract the phenomenological details of
these theses from the metaphysics that they putatively recommend. Part of that
requires a characterisation of perceptual experience as exhibiting intentionality, in
an appropriate sense of this rather philosophically rich concept.

62
For a deeper discussion that places the concerns and motivations of sense-datum
theory in the context of the dominant metaphysics of its heyday, see Martin (2000).
63
For an article that exhibits this tendency somewhat rampantly, see Schellenberg
(2010, esp. the "phenomenological problems" she raises on pp. 33 - 34).
84 |
4. 2. Intentionality
The study of intentionality is a minefield of several crosscutting and occasionally
orthogonal controversies.
64
It is perhaps unsurprising then, that one of these con-
troversies is over the roots of the modern concept of intentionality, and indeed how
modern it is; that is, how long the concept of intentionality itself has been the sub-
ject of intellectual enquiry.
65
Elisabeth Anscombe claims that the etymology of in-
tentionality as understood in the 20
th
century is a metaphorical extension of the
latin phrase intendere arcum in, which from intendo (to hold out, stretch or strain),
arcus (an arc or a bow) and in (at or on) yields to stretch a bow at. The metaphor is
applied by introducing animus (mind or soul) to yield intendere animum in. Hence,
the directedness of intentionality is metaphorically described as a mind stretching
towards something (Anscombe, 1965, p. 4). By contrast, Crane traces the noun in-
tentio (a stretch or tension), derived from intendere, to the medieval scholastic
translation of Aristotles term for thought, , as a product of their engagement
with the writings of Aristotles Arabian commentators. In the hands of the scholas-
tics, an intentio was a concept, and prima intentio and secunda intentio referred to
concrete objects and logical categories of concrete object concepts respectively, the
latter being the main topic of interest. Another line Crane identifies, runs through
St. Thomas Aquinas extensions of Aristotles hylemorphic theory of sensory percep-
tion to an account of thought, according to which mind assumes the form of the
object thought, but whereas the form of the object has an esse natural, the form of
the object in the mind has an esse intentionale (intentional existence). A yet further
line he identifies is Ren Descartes distinction between the formal reality of an idea,
which consists in facts about its reality tout court, and its objective reality, which
consists in facts about what the idea concerns; objective reality (in this sense) being
closer to what is meant by intentionality in contemporary times (Crane, 2001a, pp. 8
- 10).
However, Victor Caston traces yet another alternative path between Aristotles phi-
losophy and the scholastics, via St. Augustines supplementation of Aristotles dyad-

64
For extensive bibliographic material (in German, but of mostly Anglophone texts),
see Barz (2002) and the service sections in Metzinger (2010). Jacob (2010) and Siewert (2008)
also provide general introductions, each with a comprehensive range of bibliographic mate-
rial.
65
For a collection of essays on the history of intentionality, see Perler (2001). For a
dedicated monograph on the history of intentionality in the Brentanian tradition (particu-
larly Brentano himself and two of his students, Twardowski and Husserl), see Hickerson
(2007).
| 85

ic analysis of perceptual relations with a third relatum, the intentio, which he treat-
ed as the striving or willing of the mind in the perceptual act. This, Caston argues, is
something of theme in Augustines work. And as the theme is revealed in other are-
as, one discovers Augustine engaging with extromission theories of perception de-
veloped by the stoics at least as early as the third century B.C.E. One of these theo-
rists, Chrysippus, describes vision as related to an object by the (tension,
straining) of a visual cone. , moreover, is derived from (to stretch,
strain tight) a cognate of the latin verb intendo that Anscombe claims to be the basis
of a key metaphor operative in the modern understanding of intentionality (Caston,
2001, pp. 24 - 25; 2008, 1).
Whatever its actual roots may be, Franz Clemens Brentano is typically credited with
introducing the concept of intentionality to modern philosophy in the following
passage:
Every mental phenomenon is characterized by what the Scholastics
of the Middle Ages called the intentional (or mental) inexistence of
an object, and what we might call, though not wholly unambiguous-
ly, reference to a content, direction toward an object (which is not to
be understood here as meaning a thing), or immanent objectivity.
Every mental phenomenon includes something as object within it-
self, although they do not all do so in the same way. In presentation
something is presented, in judgement something is affirmed or de-
nied, in love loved, in hate hated, in desire desired and so on.
This intentional in-existence is characteristic exclusively of mental
phenomena. No physical phenomenon exhibits anything like it. We
can, therefore, define mental phenomena by saying that they are
those phenomena which contain an object intentionally within
themselves (1874/1973, p. 68)
86 |
There are three key theses to be found in this passage. In order of occurrence, they
are:
(i). Every mental phenomenon is characterised by the intentional inexistence
of an object
(ii). Every mental phenomenon is directed toward an object
(iii). Mental phenomena are defined and distinguished from physical phenom-
ena by the fact that they contain an object intentionally within themselves
Thesis (ii) is closest to my concerns. But a lot of the issues I hope to leave aside are
due to the fact that much of the contemporary interest in intentionality concern the
way in which (ii) is bound up with interpretations of (i) and (iii).
Begin with the basic lesson drawn from Brentano in mainstream philosophy of
mind, aptly summarised by Dan Dennett:
Brentanos point about direction upon an object is this: One cannot
want without wanting something, imagine without imagining some-
thing, hope without hoping for something, and yet the object in all
these cases does not, or need not, exist in the fashion of objects of
physical actions, such as lifting, touching, sitting upon. (1969/1983, p.
21)
Hoping, wanting, imagining, etc. are all states of mind that characterise a subjects
attitude towards something. But the rub comes in the contrast with the fact that,
whereas when I lift something, it ought to follow that there is something that I lift,
if I want a wife:
it not only does not follow that there is a wife I want, but also does
not follow that there is a woman whom I want to marry, any more
than it would follow from the fact that I want a spaceship that there is
a spaceship that I want. (ibid., p. 21)
The peculiarity of intentionality is in the peculiarity of intentional relations, the
prima facie absurdity of which is captured by Wittgenstein when he writes:
I can look for him when he is not there, but not hang him when he is
not there.
One might want to say: But he must be somewhere there if I am
looking for him. Then he must be somewhere there too if I dont
find him and even if he doesnt exist at all. (1953/2001, 462, 113e)
| 87

Wittgensteins quip and Dennetts rendition of the lesson to be learnt from Brenta-
no both illustrate a reading of inexistence in thesis (i), and containing an object
intentionally within in thesis (iii), which serves to construct what is often called the
problem of intentionality. The problem of intentionality is the problem of naturalis-
ing the mind. In particular, the difficulty is the minds putative relation to queer
objects that seem to have either a diminished existence within the mind or no viable
existence at all in their mere inexistence, whilst nevertheless construing mind as a
part of the real world and thus causally related to that world.
Brentano is often credited equally with introducing both the notion of intentional
inexistence and the problem of intentionality to philosophy of mind. But some
doubt that Brentano introduced the idea of intentionality at the same time as intro-
ducing the problem:
In Brentanos earlier interpretation of intentionality, he understood
the intentional object to be a mentally immanent object and the in-
tentional relation to be an immanent relation between the mind and
its contents [treating] the intentional object as purely immanent and
as non-real. (D. Moran, 2000, p. 49)
This reading draws support from early parts of Brentanos Psychology from empirical
standpoints (1874/1973, see e.g. p. 19) that seem to express a view on which all ob-
jects of mind are phenomena, and all phenomena, mental and physical alike, are
held to exist in the mind, the latter as appearances distinct from reality (Crane,
2006a, pp. 27 - 30; see also D. Moran, 1996, pp. 17 - 22). Thus, a thought about a
physical phenomenon merely relates the mind to an object within the mind, an in-
existent object, in so far as it is intentionally contained within the mind. No conti-
nuity with reality is assumed on this view. But later, upon the publication of the
second edition of that work (now widely available in English translation) Brentano
apparently came to change his stance, suggesting that in some cases intentionality is
a relation to reality. If one bears in mind the notion of inexistence introduced in
the context of his earlier views, then this simply states the problem of intentionality;
certainly, it is of little worth in resolving it (Crane, 2006a, pp. 31 - 33).
For the sake of completeness, I ought to note that according to another exegesis,
Brentanos real problem was the problematic implications of his empiricism, rather
than the problem of intentionality as classically conceived. Ryan Hickerson (2007)
notes that of the three basic concepts in the passage from Brentano above, physical
phenomena have been given a conspicuous lack of attention. Drawing upon Brenta-
nos publications previous to the Psychology Hickerson argues that Brentanos con-
cern in demarcating the mental from the physical stems from his work on Comtean
positivism. It is from Comte (on this reading) that Brentano inherits his concept of a
88 |
physical phenomenon, qua physical and qua phenomenon. Phenomena are the psy-
chophysical causes of sensations according to Brentano, and all mental phenomena
are thus presentations or based on presentations (1874/1973, pp. 63 - 65 ).
66
But
mental phenomena include not only presentations, but also judgements and nor-
mative evaluations that are supposed equally to be based upon presentations. It is
then something of a puzzle to see how Brentano could give an account of a wide
range of judgements, for [mental] phenomena are frequently about more than
merely physical phenomena, such as when one judges that a centaur does not exist
(Hickerson, 2007, p. 41). Hence, Hickerson sees the following dilemma in the offing:
Brentano must either restrict himself to physical phenomena as
physical facts revealed via sensations [] in which case they cannot
exhaust the contents of all judgements; or Brentano must expand his
lists to include things like putative entities and fictional objects, in
which case the contents of mental phenomena are no longer merely
the physical facts revealed via sensations []. (ibid., p. 41)
I leave aside the question of which of these problems was the real problem that
Brentano faced (if not indeed both), and the further question of whether focus on
Brentanos work can provide insight into the deepest problems posed by the concept
of intentionality.
67
Suffice to say that these problems of intentionality are generated
by the thought that intentional objects need not exist in the same sense as ordinary

66
This involves him claiming that feelings (e.g. of pain) are based on presentations
(e.g. of a body part), and later on, when he gets to the famous definition of mental phenom-
ena as intentionally containing an object, he relies on his earlier remarks to defend that the-
sis from a potential criticism along the lines of Hamilton's thesis that feelings do not have an
act-object structure (Brentano, 1874/1973, pp. 68 - 69 ).
67
For yet another problem of intentionality, see Barz (2008); and for arguments that a
complete analysis of intentionality creates insurmountable problems for materialism, see
Bealer (1996).
| 89

concrete objects.
68
This is effectively to treat inexistence as tantamount to possible
non-existence, allowing the more general worry to be put in the form of a trio of
inconsistent claims:
(1) Intentional objects are ordinary existing entities.
(2) Intentional objects are what intentional states are about.
(3) Intentional states can be about things which do not exist.
(Crane, 2001b, p. 337)
Call this the basic problem of intentionality. Claim (2) can be seen as simply analyt-
ic and as setting up the problem, the relation between (1) and (3). In order to frame
the basic problem and its relationship to accounts of perceptual experience, I will
now examine two modern classic treatments of intentionality. Both aim to provide
an answer to the problem; both desire to give an account of the intentionality of
perceptual experience; neither is satisfactory.
4. 3. Intentionality and perceptual experience
Elizabeth Anscombe made an early attempt to capture the intentionality of percep-
tion, suggesting that the verbs of sense-perception [such as to see, to touch, to hear
etc.] are intentional or essentially have an intentional aspect (1965, p. 11). On such a
view, seeing x (for example) often takes a material object, such as in: You cant
have seen a unicorn, unicorns dont exist. But this is merely derivative of the general
case in which seeing takes an intentional object. And ex hypothesi the intentional
and material object need not coincide. Such claims are usually corroborated by ref-
erence to the possibility of hallucination; Anscombes rhetoric is a case in point:

68
A related issue that incensed Russell (1905) (apparently upon his encounter with
Alexus Meinongs work (1904/1981)) is that of expressions which seem to refer to non-
existent objects. I leave aside discussion of this issue, and indeed several other issues in the
philosophy of thought and language (though for a useful reference, see Luntley (1999)). My
omissions include developments of the work of Gottlob Frege (1892/1960), though I do dis-
cuss Gareth Evans and John McDowells ideas on the reference of singular terms. Although
the latter saw themselves as neo-Fregeans, for the purposes of our discussion it will suffice
to make a comparison and contrast with certain Russellian ideas that Evans draws upon.
Similarly, although many representationalist (or sometimes intentionalist) views in the
philosophy of perception draw upon the Fregean notion of a mode of presentation (see
Chalmers, 2004; and Crane 2010, 3.3 for a review), this feature of such views is not essential
for the issues of interest here.
90 |
if a man aimed at that dark patch against the foliage, and that patch
was his fathers hat with his fathers head in it, then his father was the
material object of his aim; but if he aimed at some patch in a totally
hallucinatory scene, and hit his father, you could not say that (1965,
p. 17)
The suggestion is that qua hallucinatory, the patch in the scene does not exist. As-
suming that it is only the class of intentional objects that are characterised by the
possibility of their non-existence, the man therefore aims at an intentional object;
mutatis mutandis for seeing, hearing, touching, etc.
Anscombes suggestion gets us as far as it does by reducing the notion of an inten-
tional object to the linguistic notion of an intentional verb: a verb that exhibits
intensionality (with an s) and occurs transitively (cf., 1965, pp. 3 - 4). The hallmark
of an intensional linguistic context is when either of the following two conditions
hold:
Failure of substitution of co-referring terms salve veritate:
e.g., I might believe that the former President of the United States was a
man of sound judgement, but I do not thereby believe that of George W.
Bush.
Or:
Failure of existential generalisation:
e.g., I might believe that God can answer all my questions, but it does not
follow that God exists.
Accordingly, these belief statements are intensional.
69
The features that they exhibit
in this respect seem to provide a natural account of the possible non-existence of
intentional objects. And the directedness of intentionality is then treated as a con-
sequence of the fact that transitive verb phrases require direct objects. In this sense
there is an object to my believing even when I say that, I believe nothing of what
you say.

69
For a famous attempt to account for certain features of such contexts, such as the
fact that the identity of the objects of co-referents can be informative, see Frege (1892/1960).
For arguments to the effect that the real problem of intentionality lies in reconciling our
intuitions about intentional contexts with our intuitions about Leibnizs law, see Barz
(2008).
| 91

Despite its simplicity, there is a natural suspicion that Anscombes proposal gets
things somewhat backward. The suspicion, more plainly, is that the intensionality
of belief sentences is a consequence of the intentionality of the beliefs themselves
(Crane, 2003, p. 35, italics original). If that is right then Anscombe is guilty of con-
fusing properties of reports with [the intentionality of ] the things reported (Searle,
1983, p. 24).
70

One way of showing this would be to demonstrate that beliefs qua mental states can
exhibit extensionality (the opposite of intensionality), by demonstrating that either
co-referent substitution salva veritate or existential generalisation holds. Searle at-
tempts the latter. On Searles view, one can validly infer from the (unexpressed)
belief that King Arthur killed Sir Lancelot to there is a unique x such that x =
King Arthur, and there is a unique y such that y = Sir Lancelot, and x killed y (1983,
p. 24).
71
This approach captures directedness straightforwardly enough. But that
comes at the cost of denying the possible non-existence of intentional objects.
More on that in a moment suffice to say for now that one does not need to pay the
cost of denying the possible non-existence of intentional objects in order to demon-
strate the poverty of Anscombes linguistic reduction. For an alternative (indeed, in
the present context, better) strategy is to note cases of perceptual experience exhib-
iting directedness despite a perspicuous lack of apposite verbs of sense perception.
One such is a case of prosthetic perceptual experience induced by extended use of a
device called feelSpace, developed by the Magnetic Perception group at the Univer-
sity of Osnabrcks Institute of Cognitive Science (S. K. Nagel et al., 2005). Their
prosthesis consists of a battery powered compass and CPU coupled to vibrating pads
lining a belt worn round the waist (see Figure 9). The motors creating the vibration
are individually and exclusively active in an all-or-nothing fashion, and the individ-
ual activity corresponds to magnetic north in every instance. After a six-week train-
ing period, users exhibit physiological responses consonant with altered interac-
tions between visual and vestibular systems specific to rotation of the body, and are
significantly more competent in basic navigation tasks in urban environments. Per-
formance benefits are lesser and more variable for complex tasks in virtual envi-
ronments (presumably because these deviate somewhat from training conditions).
But interestingly, the most active users report dramatic (though individually vari-

70
Searles position on this point is unequivocal but not without subtlety, as his own
framework for discussing intentionality is based upon the model of speech acts (1983, pp. 4 -
13).
71
It may not be true there is a unique x and y bearing that relation. But that is orthog-
onal to the point, as the validity of the inference is merely supposed to express the condi-
tions under which the belief would be true iff. the conditions described obtain.
92 |
ous) changes in their experienced relation to spatial properties of their environ-
ment.

Figure 9 The feelSpace prosthesis

In some instances, these changes in the experience of spatial properties could be
construed as enhancements of spatial thinking: increased spatial memory; in-
creased capacity to order places on a two dimensional plane; all objects and land-
marks become easily conceivable as relative to magnetic north etc. But the actual
indication seems to be that all these changes are grounded on the development of a
spatial feeling that is not identical to but enabled by the systematicity of the tac-
tile stimulation (S. K. Nagel et al., 2005, R22). Hence, subjects report being sponta-
neously and intuitively aware of the direction of intuitively present but spatially
distant places (ibid., R22, italics specify subjects' reports, translated from German by
the authors). Most poignantly, the authors of the study claim that the resultant ex-
perience is ineffable in an important respect:
Articulating the perceptual quality they accessed and the qualitative
experience arising from the different kind of spatial perception was
hard. The observer got the impression that they lack concepts for
what happens, such that they could only use metaphors and compar-
isons to come closer to an explanation. (S. K. Nagel et al., 2005, R23)
Sometimes consciousness is held to be essentially ineffable, and constituted by in-
trinsic (i.e., non-relational) and primitive (i.e., unanalysable) properties called
The box at the front contains the main control unit and a 3DM-G compass. The
belt is lined with flat vibratory motors, using Velcro fastening for variable po-
sitioning. The CPU outputs to the vibratory motors and houses a socket for op-
tional attachment to a personal computer.
Adapted from
http://feelspace.cogsci.uni-osnabrueck.de/en/technology_01.html
and Nagel, Carl, Kringe, Mrtin, & Knig (2005, R15)

| 93

qualia.
72
But one need not claim the existence of qualia (defined as such) to under-
stand why the experiences enabled by feelSpace are ineffable. The more austere ob-
servation is simply that natural languages (in this case English and German) do not
have the requisite verbs of sense perception to describe their experience. Moreover,
users of feelSpace seem subject to a range of experiences broadly describable as be-
ing perceptually directed towards their environment in new and interesting ways,
their experience cannot be adequately described as touching our planets magnetic
field, or indeed seeing it.
Shedding the vestiges of Anscombes linguistic reduction leaves one with a bit of a
puzzle. At least on that view, the possible non-existence of certain intentional phe-
nomena could be accommodated fairly innocently, as a feature of intensional con-
texts. And this sits happily with another central feature of intentionality, namely,
that one who thinks must think something, one who touches must touch something,
one who loves must love someone, etc. In each of these cases the italics denote the
intentional object toward which the mind is directed. On Anscombes analysis this
metaphysical necessity is reduced to the structure of certain natural languages, as
the direct object of transitive verbs. By contrast, Searles account identifies inten-
tional objects with material objects. Intentional objects are not features of language;
they are material things.
[A]n Intentional object is just an object like any other; it has no pecu-
liar ontological status at all [...] Thus, for example, if Bill admires
President Carter, then the Intentional [sic.] object of his admiration
is President Carter, the actual man and not some shadowy interme-
diate entity between Bill and the man. (Searle, 1983, pp. 16 - 17)
Although Searle is admirable in refusing any skeletons in his metaphysical closet, he
does so simply by skirting the trilemma mentioned at the end of the last section:

72
Although I will not speak much of qualia, the reader is certainly owed reference
material for this well-worn (some might say worn-out) concept. For a historical account of
qualia as a philosophical concept, and a comparison to the history of sense-data, see Crane
(2000). For general arguments against the existence of qualia, see Dennett (2002b) and
Harman (2002). For an argument against the existence of the original referent of the con-
cept, and an argument for its elimination from the study of consciousness, see Metzinger
(2003a, pp. 83 - 86).
94 |
(1) Intentional objects are ordinary existing entities.
(2) Intentional objects are what intentional states are about.
(3) Intentional states can be about things which do not exist.
(Crane, 2001b, p. 337)
Searle essentially glosses over (3). But in doing so he limits the notion of intention-
ality with questionable warrant. To adapt an example of Anscombes: poems from
the early ages of human history describe a king of the gods, a father of the sky, who
(according to some) lived on mount Olympus, had a penchant for young ladies and
throwing lightning bolts etc. He was worshipped in ancient Greece, under the name
Zeus. Presumably, Zeus does not (and did not) exist in the way that President Carter
did. In fact, the least hackneyed way of stating the matter is to say that Zeus has
never existed at all. Yet, the Greeks built statues depicting the object of their wor-
ship, the god that they believed to be thus and so. So if believing and worshipping
are intentional states, then intentional states can be about things that do not exist.
Denying that beliefs are intentional simply because they can be about things that do
not exist (a consequence of Searles view) seems to get things backward in its own
way. It constitutes an attempt to make the phenomena fit the theory, rather than the
other way round. The difficulties stem from the task of providing a theory of inten-
tionality that meshes with the appealing metaphysical intuitions that objects both
exist and do so as the ordinary concrete objects we are familiar with. Providing a
theory that can meet this requirement is a difficult task; it is a worthwhile task, but
it is not one to be attempted here. In fact, it is a task I attempt to specifically avoid,
by appeal to the development of the concept of intentionality in the early work of
Brentanos student, Edmund Husserl. But before I elaborate on that further I will
convey the more common manner in which the intentionality of perceptual experi-
ence is treated in contemporary philosophy of perception.
4. 4. The natural contents of perceptual experience
Something like Searles view is often posited as an antecedent assumption in setting
up the basic problem. Here is a representative passage:
| 95

When we think about a real object, such as Paris, it seems that the
object of our thought is that real object, the city, with all its lights.
Moreover, it seems that a mental state's capacity to relate in this way
to real objects is important in the role that they play in the explana-
tion of action [...] So it looks as though intentionality must involve a
real relation among real-world objects. (Segal, 2005, p. 284)
As remarked earlier, thoughts can be directed at objects that cannot be real-world
objects. In lieu of this, the tension comes in the second sentence. To see how it
might be motivated in the context of perceptual experience, consider a case where
A friend shouts from behind. You turn. It looks as if a rock is flying at your face
(Sturgeon, 2000, p. 8). Assuming that you do not wish to be hit, the likely result of
you seeing a rock flying towards your face would be an evasive manoeuvre. Crudely
perhaps, this suggests that the object of your visual experience would play a causal
role in your behavioural response. But presumably something can only bear causal
relation to something else if both things exist. And as has been repeatedly suggest-
ed, if the object of your experience is an intentional object, then it need not exist.
The modern approach to the problem of intentionality is to reduce the notion of
intentionality to a naturalised form of content. This latter term is certainly polysem-
ic, in fact some go so far as to say that it is deliberately vague (Brown, 2007, 1).
This is not to say that the philosophy of content is a sloppy enterprise, but rather
that there are many and various accounts of content. Mere use of the term itself
should not stack the deck in favour of one of these particular views. What holds for
the general case holds for perception also:
A perceptual state might have the content that there is a green object
in front of me [...] We can be neutral on just what sort of content this
is: it might involve a set of possible worlds, a complex of objects and
properties, a complex of Fregean senses, or something else again.
(Chalmers, 2004, p. 155; see also, Siegel, 2010, 3.ff)
73

The neutral use of the term is neutral on what the metaphysics of content will
turn out to be. But it is not metaphysically neutral per se. For what is agreed by all in
discussion of the nature of content, is that contents have conditions of satisfaction:

73
These remarks are certainly not idiosyncratic to Chalmers work. Searle gives a clear
statement of the canonical thought here:
The argument that visual experiences are intrinsically Intentional, in sum, is that they have
conditions of satisfaction which are determined by the content of the experience in exactly the
same sense that other Intentional states have conditions of satisfaction which are determined
by [their] content. (1983, p. 40)
96 |
they are the sort of thing that can be satisfied or can fail to be satisfied by states of
the world (Chalmers, 2004, p. 155, italics original).
These conditions of satisfaction are sometimes called accuracy conditions. If a state
has specifiable accuracy conditions, then that is often taken to suffice for claiming
that the state is representational, and in this sense content is often just shorthand
for representational content. Against the background of the problem of intentional-
ity, the main issue in the study of perceptual experience might seem to be that of
discerning the natural means by which perceptual states have representational con-
tent. That is, how do perceptual states naturally come to represent the contents of
perceptual experience? The most popular answers this question appeal in some way
or other to causal covariation.
For instance, here is how Michael Tye defines the relation S represents that P: If
optimal conditions obtain, S is tokened in x if and only if P and because P (1995, p.
101). In Tyes work, the content that P is construed as a proposition; the content of
perceptual experience is whatever would be expressed by whatever followed a that-
clause in a sentence describing the experience. Nevertheless, the content is held to
be a natural phenomenon, specified by causal covariation in optimal conditions.
The content of the state is the state of affairs that would cause the state to occur
under optimal conditions. It is in reference to the particular states of affairs ex-
pressed by the proposition, together with other factors, that the theorist is able to
determine what the state in question represents (cf. Stampe, 1977).
This is a similar approach to one taken by Fred Dretske. In his (1981) Dretske devel-
ops an information-theoretic analysis of content, based roughly around the simple
idea that every effect carries information about, and thus indicates, its cause (cf.
Grice, 1971). But a crucial feature of the view is that the content of a state is not iden-
tified with the present cause of the state. Rather the content is identified with in-
formation about the conditions the state would indicate if the conditions that were
present during learning obtained. Dretskes thought on the nature of these learning
conditions has seen some development. In his earlier work, Dretske describes these
learning conditions as those which ought to eliminate any equivocation in the signal
carrying the information (Dretske, 1981, pp. 194 - 195). In his later work, learning
conditions are those that serve the purpose of promoting to representational sta-
tus one of the many things indicated, in virtue of explaining a causal chain that
leads to behaviour (ibid., 1988, p. 84, n4). But in both cases, what is essentially re-
quired is either the right soldering by Mother Nature (ibid., 1998, p. 98; cf., Milli-
kan, 1989a), or favourable ontogenetic conditions for associative learning (ibid.,
1998, p. 98 - 107). The existence of these two options, i.e., favourable phylogenetic or
ontogenetic history, becomes important when applying the view to an account of
| 97

perceptual experience. In his (1995) Dretske distinguishes between systemic repre-
sentations, whose contents are fixed phylogenetically, and acquired representations,
whose contents are fixed ontogenetically. He says the following of the distinction:
[Sensory] experiences are to be identified with states whose represen-
tational properties are systemic. Thought (conceptual states in gen-
eral), on the other hand, are states whose representational properties
are acquired. As a result, experiences have their representational con-
tent fixed by the biological functions of the sensory systems of which
they are states (Dretske, 1995, p. 9)
Dretske thus endorses a version of the thesis that sensory perceptual experience has
non-conceptual content.
74
However, the relationship between these two kinds of
content is made clearer in his earlier work, in which he distinguishes between two
ways in which information can be encoded. A particular passage is worth quoting at
length:
The contrast between an analog and a digital encoding of infor-
mation [...] is useful for distinguishing between sensory and cogni-
tive processes. Perception is a process by means of which information
is delivered within a richer matrix of information (hence in analog
form) to the cognitive centers for their selective use. Seeing, hearing,
and smelling are different ways we have of getting information about
s to a digital-conversion unit whose function it is to extract pertinent
information from the sensory representation for purposes of modify-
ing output. It is the successful conversion of information into (ap-
propriate) digital form that constitutes the essence of cognitive activ-
ity. (Dretske, 1981, pp. 141 142)
75

This develops a thesis, due perhaps originally to Gareth Evans (1982, p. 229), that is
common amongst a number of recent attempts to specify the content of perceptual
experience: that perceptual experience has a fineness of grain that outstrips our
cognitive capacities.
76

74
For different versions of the thesis that perceptual experience has non-conceptual
content see Bermdez and Cahen (2008), and Heck (2000).
75
In case there are any doubts as to whether he is discussing sensory experience here,
he then continues, to note that: The traditional idea that knowledge, belief, and thought
involve concepts while sensation (or sensory experience) does not is reflected in this coding
difference (Dretske, 1981, p.142).
76
Held notably by Tye (1995) and Peacocke( 1992), denied most notably by McDowell
(1994) and Brewer (1999). See also Heck (2000) and Kelly (2001) for reflections on the de-
bate.
98 |
Dretskes work reflects a research programme in modern philosophy of mind devot-
ed to grounding the intentionality of the mind in non-intentional relations, whilst
nevertheless attempting to recognise that which is not only distinctive of the mind,
but distinctive about perceptual experience and consciousness more generally.
With respect to consciousness, Dretske motivates a familiar distinction between
thought and sensation, or sensory and cognitive processing. By way of a link to
the next section, let me mention another way of motivating the distinction: arguing
that sensation is radically indeterminate in a way that thought is not. This would be
exactly contrary to the staple of early modern empiricism, that the mind in part
comprises simple and ostensibly identifiable sensory elements (viz. ideas of sense,
or sensory ideas). In a contrary frame of mind, William James writes:
[The] view of these philosophers [the empiricists] has been called lit-
tle into question, for our common experience seems at first sight to
corroborate it entirely. Are not the sensations we get from the same
object, for example, always the same? Does not the same piano-key,
struck with the same force, make us hear in the same way? Does not
the same grass give us the same feeling of green, the same sky the
same feeling of blue, and do we not get the same olfactory sensation
no matter how many times we put our nose to the same flask of co-
logne? It seems a piece of metaphysical sophistry to suggest that we
do not; and yet a close attention to the matter shows that there is no
proof that the same bodily sensation is ever got by us twice. (James,
1890/1950, p. 231)
If James is right that it is impossible to show that the same sensation is involved in a
perceptual experience of the same object, this motivates a bifurcation of content in
a different manner than suggested by the non-conceptualists. More radically, one
might conclude that the stream of consciousness consists only in contentless sensa-
tions, perhaps only beliefs and other more cognitive states of mind have content.
Most would not move to this conclusion.
77
In fact the more typical bifurcation in the
contemporary literature is one that does not bifurcate between mental states that
do and do not have content. But rather one that bifurcates between the kinds of
content mental states can have. This will see further discussion in the next section.

77
Siewert (2008, 4) ascribes this view to Putnam (1981), though without detailing any
specific page or chapter. Accordingly, I am (sadly) unable to corroborate the reference.
| 99

4. 5. Phenomenology vs. neutral intentionality
Studies in the metaphysics of perceptual content are rife with inferences from
claims about perceptual experience to claims about metaphysics. But in order to see
them clearly, we need some terminology. Consider, to begin, Martin Davies state-
ment that perceptual content is a matter of how the world seems to the experi-
encer (1992, p. 26). This seems to be a fairly innocuous notion of content, hardly a
substantive metaphysical position. He continues, explaining that perceptual con-
tent is, in this sense, phenomenological content (ibid., p. 26). Now, what does that
mean? Dennett can offer a helping hand:
Philosophers and psychologists often use the term phenomenology
as an umbrella term to cover all the items the fauna and flora, you
might say that inhabit our conscious experience: thoughts, smells,
itches, pains, imagined purple cows, hunches, and all the rest. (1991,
p. 44)
Call this phenomenology
U
(note the subscript, for umbrella). Prima facie phenom-
enology
U
is metaphysically neutral enough, as long as one discharges the metaphor
of inhabiting. But then Davies carries on to reveal what he is really saying, which is
that where there is no phenomenological difference for the subject, there is no dif-
ference in perceptual content (op. cit., p. 26). This is to say that perceptual content
supervenes on phenomenology. To clarify, supervenience is an asymmetric de-
pendency relation between some A and some B such that a difference in x with re-
spect to A could not occur without some difference with respect to B. The concept is
typically employed in thinking about which parts of the world can vary whilst some
state or property of interest remains the same.
78
For instance, views that espouse
local supervenience with respect to the content properties of states will hold that
these states and/or their properties supervene on the intrinsic properties of an indi-

78
The concept of supervenience originates in discussions of metaethics, where Moore
held that a moral property such as goodness was a suis generis non-natural property depend-
ent upon the natural. He construed this dependence as such that:
[I]f a given thing possesses any [property] in a certain degree , then not only must that same
thing possess it, under all circumstances, in the same degree, but also anything exactly like it,
must, under all circumstances, possess it in exactly the same degree (G. E. Moore, 1922, p.
261)
Later Hare apparently coined the term supervenience in print (in his 1952), for pretty much
the relation Moore identified, though Hare was of the view that there are no moral proper-
ties (or indeed facts) to supervene. For reviews of the subtleties in the origins and scope of
the concept of supervenience, see Horgan (1993) and Kim (1984).
100 |
vidual. Hence, for instance, if a state A of an individual i locally supervenes on its
intrinsic property B, then a duplicate of that individual with respect to B would also
be in state A. As a subvenient upon which the content depends, phenomenology
seems rather closer to indicating a metaphysics of content than the umbrella term
that Dennett gestures at. Call all any view on which perceptual experience is tied to
a metaphysical stance on content or anything else, phenomenology
M
(again, note the
subscript, for metaphysical). Phenomenology
M
is the subject matter of several of
the views of the relationship between perceptual experience and intentionality dis-
cussed in this chapter, and in many other chapters of this thesis.
One way of conceptualising phenomenology
M
has been provided by Terence Horgan
and John Tienson in their discussion of phenomenal intentionality: a pervasive
kind of intentionality which is determined by phenomenology alone (2002, p. 524).
Their move from description to metaphysics is most clear when they argue by way
of introspective description of human experience that any two possible phenome-
nal duplicates have exactly similar intentional states vis-a-vis such content (ibid.,
pp. 521, 524) (ibid., pp. 521, 524). Phenomenal duplicates are identical with respect
to their total phenomenology
M
, where this can be as multisensory, cognitive and
long-lasting as one likes (ibid., pp. 521 524) (ibid., pp. 521 - 524). The fact that their
views on content are bifurcatory becomes clear when they start to talk about the
correctness conditions that purportedly comprise phenomenal intentionality. One
thing that they are very clear about is that a certain class of the accuracy conditions
of phenomenal intentionality (which they call, according to convention, narrow
correctness conditions) are not derived from the causal relations grounding natural-
ly arising intentionality, such as in the theories developed and appealed to by
Dretske (1981) and Tye (1995), and Millikan (1989a) (Horgan & Tienson, 2002, pp.
527 - 529). As an example, they suggest we think about the total phenomenology
M

of perceptually interacting with a picture that seems to hang crooked on a wall. As I
mentioned, they countenance two sorts of correctness conditions. One narrow
sort is common between phenomenal duplicates, whilst they engage in the very
same interactions. They are duplicates in so far as they experience seeing and doing
the same things as they try to get the picture to look straight, and the manner (if
any) in which their experiences are identical, is the manner in which their experi-
ences share narrow correctness conditions. The other sort, wide correctness condi-
tions, contrasts the duplicates, in virtue of both the fact that the pictures (presuma-
bly existent particulars) are themselves not identical, and that the phenomenal du-
plicates are numerically distinct. Thus the pictures, in virtue of being numerically
distinct from one another, each determine the correctness of the respective phe-
nomenology
M
of the numerically distinct phenomenal duplicates interacting with
them (ibid., pp. 526 529) (ibid., pp. 526 - 529). In sum,
| 101

There are thus two ways of thinking of truth conditions: as deter-
mined wholly by phenomenology, and as determined in part by items
in the experiencer's environment that satisfy the experiencer's phe-
nomenology. (Horgan & Tienson, 2002, p. 525)
The bifurcation assumes that the properties of the phenomenology
M
that determine
the status of phenomenal duplicates as such are general properties that are common
between the two individuals.
79
This allows that, with respect to narrow accuracy
conditions, one could have a phenomenal duplicate experiencing a (perfectly) hal-
lucinatory (or perhaps dreamt) Cartesian world (Horgan & Tienson, 2002, pp. 524,
527). Narrow accuracy conditions are thus fundamentally distinct from the wide
accuracy conditions which serve to determine which object in the world (if any) the
subject experiences (ibid., p. 529) (ibid., p. 529).
According to Horgan & Tienson, proper consideration of phenomenology
M
reveals a
metaphysical barrier between perceptual experience and reality. Allowing for the
possibility of the same phenomenology
M
across radically different situations dic-
tates that phenomenology
M
must be characterised in a manner that distinguishes it
from any real object existing in a certain situation. For it is precisely those objects
that may not exist in certain situations in which phenomenology
M
is held to be the
same.
There are two main reasons for adopting the neutral intentionality thesis. The first
is to avoid using the term phenomenology in an ambiguous fashion. The second is
to unambiguously adopt a neutral metaphysical stance on the issues in question.
This stance is apt as my aim is to provide a neutral description of a certain concep-
tion of perceptual experience. This requires being impartial on the question of
whether or not the same phenomenology
M
could occur in radically different situa-
tions, finessing any need to mention the presence or absence of the objects of expe-
rience.

To describe perceptual experience in this way is to describe perceptual expe-
rience qua experiential, i.e., how an experience seems such that it seems perceptual.
What discussions of phenomenology
M
often rule out, by metaphysical fiat, is the

79
An issue, which I leave aside here, is that these properties must certainly be deter-
minable as somewhat specific. For instance the crooked picture can be such it instantiates a
particular angular relation to the horizon, has a centre of gravity such that it swings to one
side on the hook when poorly placed in just a certain way etc. That is, a specific general way,
F, which such objects may be, and which infinitely many qualitatively distinct possible ob-
jects are (Brewer, 2006, p. 173). If the James passage presented at the end of the last section
is taken to have any metaphysical implications, then any view which attributed such content
to conscious sensations (as Horgan & Tienson, and a number of others seem to) would be in
trouble.
102 |
possibility of making appeal to the object of an experience in explaining how it
seems to be a perceptual experience. Such a restriction is not imposed upon an ac-
count which has no concern to demonstrate a stance on whether or not (and indeed
how) the correctness conditions of experience changes in accordance with the actu-
al presence or absence of an object.
There is a sense in which discussions of phenomenology
M
are orthogonal to the
motivations for attributing intentionality to perceptual experience. Certainly, if per-
ceptual experience is a natural phenomenon, then a metaphysical analysis of per-
ceptual experience might best proceed by understanding whether it has content,
and if so then what that content is. And indeed, it ought to be noted, that an analy-
sis of the content of perceptual experience is not limited to construing content as
propositional. Content could plausibly be e.g. digital, analogue, pictorial, or a model
of a limited range of possibilities.
80
But what I will argue at points in the next chap-
ter is that it is mistaken to then assert that the intentional objects of perceptual ex-
perience seem like propositions, digits, analogies, pictures, models etc. If such as-
sertions are made, it is likely because the normal understanding of the phrase
intentional object of experience is effectively the representational content of expe-
rience (Crane, 2006b, p. 136). I make no claim as to what the normal understanding
of an intentional object is. Likewise I offer no theory of intentionality, and thus no
answer to the problem of intentionality. But it would be remiss to deny that describ-
ing the common-conception of perceptual experience appealed to by philosophers,
requires appeal to some notion of intentional object. My remit is not to deny this,
but to maintain an adherence to what I will call the neutral intentionality thesis,
which amounts to the claim that one ought to treat the objects of perceptual experi-
ence as neutrally intentional objects, or simply neutral objects for short.
An appeal to neutral intentionality minimally involves a very literal appeal to the
features of intentionality discussed in this section. The first is its directedness as
expressed by Husserl when he says that the word intentionality signifies nothing
else than [...] to be conscious of something (1950, p. 72; 1950/1999, p. 33). The se-
cond is an otherwise clumsy reading of the expression inexistence, as simply a de-
nial of the claim that being an intentional object as such is being an entity of any

80
For a defence of the view that the content of perceptual experience need not be
propositional, and could be (even on fairly orthodox contemporary principles) pictorial or
otherwise, see Crane (2009). For an account of how the content of various (oft neglected)
aspects of conscious experience might issue from modelling processes, see Metzinger
(2003a, pp. 13 - 61). For an outline of a modelling approach to perception, see Grush (2004,
pp. 390 - 393), and for applications to the spatial content of perceptual experience, see Grush
(2007b).
| 103

kind (Crane, 2001b, p. 340, italics original). Most pressingly, this last amounts to an
abject denial of any need to identify intentional objects with representational con-
tent for the purposes of discussing perceptual experience.
It is worth saying a few brief words on using Husserl to provide an account of per-
ceptual experience as conceived by proponents of direct realism. The worry might
stem from passages like this:
The attempt to conceive the universe of true being as something ly-
ing outside the universe of possible consciousness, possible
knowledge, possible evidence, the two being related to one another
merely externally by a rigid law, is nonsensical. (Husserl, 1950/1999,
p. 84 [117])
This kind of remark hardly sounds metaphysically neutral, and indeed, hardly
friendly to a realist metaphysic. Furthermore, it suggests that Husserl was engaged
in rather different project than mine. Call this project phenomenology
T
(again, sub-
script, this time for transcendental).
81

But then a look at Husserls earlier work, such his lectures of 1907, finds him saying
things like:
To perceive a house means to have the consciousness, to have the
phenomenon, of a house standing there in the flesh. How matters
stand with the so-called existence of the house, with the true Being
of the house, and what this existence means about all that nothing
is said (1973, p. 15; 1973/1997, p. 12)
Summarising this Husserl, one commentator writes along the lines of how I have
been treating phenomenology and neutral intentionality, as follows:

81
For a comprehensive introduction to Husserls transcendental philosophy, see Zaha-
vi (2003a, pp. 44 ff.). I also leave aside the complex issues surrounding the proper interpreta-
tion of Husserls reduction(s), but for a good introduction see Moran (2000, pp. 124 - 163).
104 |
the distinction between existent and non-existent objects is one that
turns up within the intentions phenomenology describes: some ob-
jects, for example, those of our fictional imagination or fantasy, are
intended as not existing, while objects of perception, say, are intend-
ed as existing. But this will be true whether or not these objects actu-
ally exist independently of the intentions in question [...] one can do
complete justice descriptively to the experiences treated by phenom-
enology and drop all reference to actual existencethough not, of
course, to intended existence. (Carr, 1999, p. 74)
However, the first quote (and many others of its kind) indicates that it would be
pretty hard to make the interpretative case that Husserl intended the scope of phe-
nomenology to metaphysically neutral (Zahavi, 2003b), or indeed that he was not a
full-blown Idealist (A. D. Smith, 2003, pp. 71 - 110). I do not want to dispute this. The
neutral intentionality thesis is probably closest to the descriptive phenomenology of
the early Husserl and not at all to the transcendental (idealist) phenomenology that
he arguably developed in the course of his later pontifications. That makes the exe-
getical situation sound awfully neat, which even the briefest encounter with Husserl
demonstrates to be unlikely. So I ought to say flatly that in my exploration of Hus-
serls ideas below I read them all roughly in the light of the descriptive project.
It is worth noting that this descriptive project, call it phenomenology
D
, is rather a
different notion from Dennetts phenomenology
U
. Actually, Dennett has two uses of
the term, one we have already come across: all the items [...] that inhabit our con-
scious experience: thoughts, smells, itches, pains, imagined purple cows, hunches,
and all the rest (Dennett, 1991, p. 44). Given the previous items on the list, and the
metaphor of inhabiting I do not hold much confidence that all the rest includes
e.g. particular things perceptually experienced as immediately present, real and
mind-independent. Dennetts second usage involves explicit reference to Husserl
(and his teacher, Brentano) as Phenomenologists (with a capital P) who propose
that one can get to one's own notional world (sc., which in other places Dennett
calls a subjective world), by some special somewhat introspectionist bit of mental
gymnastics called, by some, the phenomenological reduction (Dennett, 1987, p.
153). This he calls auto-phenomenology, as contrasted with his own project of hetero-
phenomenology, determining the notional world of another, from the outside
(ibid. , p. 153).
This might be roughly the right contrast if drawn against introspectionist psycholo-
gy. But what Dennetts heavy-handed treatment misses is that the descriptive focus
in the phenomenology
D
is not simply upon the subjects experience but also the
manner in which the world is experienced, i.e., the world just as experienced
| 105

(Drummond, 2007, p. 61). This makes for a much larger and more difficult project
than might be obvious:
Phenomenology is not, as certain analytical philosophers seem to
think, just a matter of cataloguing how things seem to you. Phenom-
enological research is difficult because it is, in large part, a matter of
uncovering ... implicit intentionality. (A. D. Smith, 2008, p. 324)
Here Smith is quoting and paraphrasing his own translation of a passage in which
Husserl (alternately translated) describes phenomenology as everywhere a matter
of uncovering the intentionality implicit in the experience (1950/1999, p. 64 [98]).
82

On this gloss, phenomenology
D
is something I have been arguing about since the
beginning of the thesis. Indeed, sense-datum theorists, Berkeley, all the empiricists
who made questionable comparisons between visual experience and painting, all
those enthusiastic about the phenomena that recommend actualism (and in the
next chapter) transparency, presence and the particularity of experience, might all
occasionally count as doing phenomenology
D
in this sense. Occasionally. But most
often, they are doing phenomenology
M
with respect to phenomenology
U
. The result
is a somewhat fragmented account of perceptual experience. My task is to bring
together these fragments under the conception of perceptual experience traceable
through the work of direct realists. The next two chapters constitute a culmination
of that project.

82
Although Dennetts description is fairly typical of those working in the philosophy
of mind and cognitive science, there are some exceptions, e.g. Tim Van Gelder captures the
point nicely when he says that phenomenological description does not mean (as it some-
times does elsewhere) describing one's own introspectively accessible ideas, but rather the
careful study and description of [...] things that are usually somehow hidden from view
(1998, p. 135 n10).
106 |
Chapter 5 Seeming real
5. 0. Introduction
In the last chapter I outlined the neutral intentionality thesis, the function of which
is enable me to provide a description of the direct realist conception of perceptual
experience without defending direct realism as such. Perhaps the most difficult task
in this project is discerning the correct understanding of the claim that the objects
of perceptual experience seem real. In this chapter, I will convey such an under-
standing by critically reviewing members of a family of concepts evoked in the liter-
ature that each has a bearing on the claim.
I begin (in 5.1) with an outline of the deeper philosophical motivations of direct
realism and the explanatory role that perceptual experience is supposed to play in
fulfilling those motivations. The central motivation is to maintain a simple connec-
tion between the subject and the world she experiences by appealing to perceptual
experience as an instance of a suis generis, simple relation to the world. The recur-
ring difficulty here is the assumption that there are possible states that are subjec-
tively indistinguishable from perceptual experiences that might serve to confound
this connection. I then (in 5.2.) provide a means of respecting both the claimed
suis generis nature of perceptual experience and the claimed immediacy of percep-
tual experience, by distinguishing perceptual and theoretical belief. I then move on
to clarify the implications of the claim that perceptual experience is transparent (in
5.3), and the claim that the objects of perceptual experience seem present (in 5.4).
In the final section (5.5.), I argue that in the hands of direct realists perceptual
presence is best construed as the claim that perceptual experience seems to be of
mind-independent particulars.
5.1 Perfection
5.2 Immediacy
5.3 Transparency
5.4 Presence
5.5 Particularity
| 107

5. 1. Perfection
An obvious motivation driving direct realism (DR) is to do justice to the way ordi-
nary perceptual experience is conceived to be, or rather, how it is claimed that ordi-
nary perceptual experience seems to be. Strawsons pre-theoretic man is a case in
point, and the need to frame sense-datum theory as offering a corrective account of
perceptual experience is certainly suggestive. As I indicated in the introduction to
this dissertation, I think that appealing to the opinions of non-philosophers on
what are essentially matters of philosophical discussion is difficult, and more so
than has been adequately recognised in discussions of perceptual experience. I also
wondered whether people do have any opinions on these matters. This is perhaps an
especially delicate point where a theory might like folk have an opinion in concord-
ance (or in some cases discordance) with what it suggests.
But having said all that, I do not mean to claim that perceptual experience somehow
actually does deviate from the DR model. At most, I mean to express the second-
order distance that I have assumed. Indeed, if perceptual experience does conform
to the DR model then this would be something that would seem rather strange to
question, and the importance of the question in that case would be suspected to be
elsewhere. This last is what I believe to be generally the case for disputes over DR.
For what is philosophically at issue is not so much how perceptual experience seems
to be (at least not since the heyday of sense-datum theory), but rather more general
concerns about the relation between the mind and what is thought to be independ-
ent of it. As discussed in 1, DR is sometimes called Commonsense Realism and oth-
er times Nave Realism. The labels are applied in part because a great deal is made of
commonsense and theoretically nave (i.e., theoretically untutored) conceptions of
perceptual experience. But the reason for all the fuss is not so much a defence of the
apple-receivers world-view so strangely brought into question in the introduction.
It is a reaction to various assumptions built into theories of perception that are
thought to have the result of making the relation between mind and world prob-
lematic, particularly when common-sense apparently suggests that there was never
a problem here in the first place.
Amongst contemporary DR theorists, the common-sense preserving theme is most
evident in the work of Bill Brewer (1999, 2006) and John Campbell (2002). Though
both draw on themes from Evans (1982) and Strawsons (1959/2003) work, the latter
being an explicit proponent of a DR theory of perception (see his 1979/2002). The
preservation of commonsense principles about the minds relation to the world was
also a key motivation in Reids (1764/1997, 1785) philosophy of mind, and the more
contemporary work of Armstrong (1960, 1961), Searle (1983), Martin (2002a, 2002b),
John McDowell (1994, 1998), Paul Snowdon (1980 - 1981, 1992, 2005), and A.D. Smith
108 |
(2000, 2002). In all cases the appeal to common sense is touted as the most plausi-
ble way to make sense of how we can think about mind-independent objects at all,
precisely because it gives such a simple explanation of how we experience objects in
our local environment. For if ex hypothesi experience of an object is a simple rela-
tion holding between perceiver and object then, as Campbell says (with some un-
derstatement) it ought to be possible [...] to extract the mind independent world
from an experience which has the mind-independent world as its constituent
(2002, pp. 115, 121).
Such a glib statement perhaps obscures the issue. To make it clearer, I will make a
slightly slower approach at one of the more popular ways of establishing the claim.
Then I will consider certain problems that are generated from principles that the DR
conception accepts. These problems belong to the class of phenomenological prob-
lems broached at the beginning of the last chapter. They are of interest both be-
cause of their putative phenomenological
M
consequences and because they seem to
motivate the much more blunt approach bruited by Campbell above.
Consider, first, a programmatic but typical account of DR:
if an experience E is a genuine perception by subject S of object O
then the occurrence of E places S in such a relation to O that were S
able to entertain demonstrative thoughts [...] then S could entertain
the true demonstrative thought that is O. (Snowdon, 2005, p. 138)
Call a perceptual demonstrative thought any instance in which a demonstrative
thought is yielded by perceptual experience. The significance of perceptual demon-
stratives according to DR is that they ground our capacity to refer to objects at all,
precisely because they are held to be object-dependent. To say that demonstrative
thoughts are object-dependent is to say that the thought
that is O
involves O as a constituent. This is an idea that Gareth Evans accredits to Russells
work on singular terms. Evans takes the consequence somewhat further than Rus-
sell might have been happy with, to express a kind of epistemic distance on the part
of the subject as to what the nature of their own thought is.

Such that it may be, for
a subject, exactly as though he were thinking about a physical object (say) which he
can see, and yet that, precisely because there is no physical object he is seeing, he
may fail to have a thought of the kind he supposes himself to have (Evans, 1982, p.
| 109

45).
83
Hence, in order for such a thought to actually refer to O, at least two condi-
tions must be met. The first is that O must exist. Any condition in which this were
not met would be one in which demonstrative reference is impossible. For instance,
consider a snippet from the famous soliloquy:
Is this a dagger which I see before me,
The handle toward my hand? Come, let me clutch thee.
I have thee not, and yet I see thee still.
(Macbeth, Act 2 Scene 1 of William Shakespeares Macbeth)
Grant, for illustrative purposes, that demonstrative thought is object-dependent,
and that the objects of sufficiently delusory hallucination do not exist. In this in-
stance, in thinking the thought,
that is a dagger
Macbeth fails to refer at all. Not even a demonstrative mode of reference can make it
such that Macbeth is thinking about something in particular rather than nothing at
all.
This brings us to the second condition, which is that if a thinker is to refer to a par-
ticular O there must be some available means of discriminating O from all other
particulars. Evans refers to this as Russells principle: it is not possible for a person
to have a thought about something unless he knows which particular individual in
the world he is thinking about (1982, p. 44). Pressure to meet this condition forms
the basis of the argument for the primacy of perceptual demonstratives.
In fact, giving primacy to certain forms of reference is not a strategy unique to DR.
Russell thought there to be two ways of meeting the condition embedded in the
principle that Evans attributed to him. They are commonly known, due to an influ-
ential article, as knowledge by acquaintance and knowledge by description (Russell,
1910-1911). For Russell, a fan of sense-datum theory, the class of what is known by
acquaintance is limited to sense-data and ones own mind (ibid., p. 109). In all other
cases, indeed most cases, a subject knows which thing they refer to by means of
some definite description (ibid., pp. 112 115). Though these descriptions may be at
various stages of removal from knowledge by acquaintance, they must be composed
of constituents that the subject is acquainted with, if the subject is to be said to
properly understand the propositions that the thoughts express (ibid., pp. 115 - 117).

83
Russell would not have countenanced such a conclusion on the basis that he re-
stricted such thoughts to refer to a private language that labelled instances of sense-data
(e.g., 1918/2010, pp. 29 - 30), a famous target of criticism in the later work of Ludwig Witt-
genstein (1953/2001239 - 277, pp. 74 - 82).
110 |
In forming the referential bedrock for thought about the mind-independent world,
acquaintance plays much the same role for Russell as it does for DR. The clear dif-
ference is that for DR, perceptual demonstratives refer to the mind-independent
world in the first place. A strong motivation for this is that if this connection is left
to description alone, then thought about particular objects will not be possible. The
key requirement is that one must be able to pick out one particular in contradistinc-
tion to all others. The difficulty is that certain metaphysical possibilities suggest
that this condition cannot be met by description alone. Strawson describes such a
situation as follows:
[O]ne may be very well informed about a particular sector of the uni-
verse. One may know beyond any doubt that there is only one partic-
ular thing or person in that sector which answers to a certain general
description. But this [...] does not guarantee that the description ap-
plies uniquely. For there might be another particular, answering to
the same description, in another sector of the universe [...] However
much one adds to the description of the sector one knows aboutits
internal detail and its external relationsthis possibility of massive
reduplication remains open. (1959/2003, p. 20)
Such a dramatic situation clearly creates problems for a view that both requires
thoughts to have a unique referent and requires said referent to be identified by
description. But despite the drama, Strawson thinks that reference sans description
can be secured in just a few steps. The first, is to assert (presumably with Russell)
that: All identifying description of particulars may include, ultimately, a demon-
strative element (Strawson, 1959/2003, p. 22). Second, is to assert a certain concep-
tion of the mind-independent world: For all particulars in space and time, it is not
only plausible to claim, it is necessary to admit, that there is [...] a system [...] of
spatial and temporal relations, in which every particular is uniquely related to every
other (ibid., p. 22). Third, is to assert that reference to any particular is anchored by
demonstrative reference to a particular at a point in space and time:
The universe might be repetitive in various ways. But this fact is no
obstacle in principle to supplying descriptions of the kind required.
For by demonstrative identification we can determine a common ref-
erence point and common axes of spatial direction; and with these at
our disposal we have also the theoretical possibility of a description
of every other particular in space and time as uniquely related to our
reference point (ibid., p. 22)
This last step is the establishment of an egocentric conception of the world, involv-
ing essentially a co-construction of a subject of experience as distinct from a world
that the subject experiences (Strawson, 1959/2003, p. 69). This solution is not quite
| 111

complete though. For what is needed is not just the construction of distinct relata,
subject and world. What is needed is some final and clear conception of how it is
that they are related. And given the previous steps this must amount to an explana-
tion of how a subject can refer to the particulars in space and time. It is precisely
here that DR gets its motivation. For what it urges, at least in its modern forms, is
that experience of objects has an explanatory role to play: it explains our ability to
think demonstratively about perceived objects (Campbell, 2002, p. 114).
This last point has to be taken with some care. In fact, there has been some in
fighting in the vicinity of how it is exactly the perceptual experience can play this
explanatory role. Consider certain problems with one such proposal, according to
which the explanatory role is secured precisely because the content of perceptual
experience is constituted by demonstrative concepts.
84
These would be expressed by
sentences such as:
that thing (there) is thus
Where, as Bill Brewer puts it, the content is doubly world-dependent in that it
involves demonstrative reference to both the object and its properties (2006, p.
166).
85
Following Brewers terminology, I will refer to this as the content view. The
content view might not help DR much; it obscures the putative explanatory role
that perceptual experience is held to play. To see the difficulty, we need to consider
certain cases of non-genuine or (as I shall call them) faux perception. A first such
example can be approached by recounting Macbeths struggles further:

84
I am sidelining a controversy here over whether the content of perceptual experi-
ence is exhaustively conceptual. For instance, McDowell (1994, esp. pp. 162 - 174) and Pea-
cocke (1998) represent opposite sides of a controversy concerning whether demonstrative
concepts can adequately specify the fine-grainedness required of the content of perceptual
experience. Both of these authors see Evanss work (1982, pp. 154 - 160, 229) as an important
early instance of the idea.
85
It ought to be noted that Brewer is speaking of his earlier view as developed in his
(1999, esp. 184 - 215). His aim in the article cited in the main text is to confess that he now
thinks that all of this is too little too late, as it were: the whole framework [of assigning con-
tent to perceptual experience] has to be rejected (Brewer, 2006, p. 166). His worry is that the
view cannot describe perceptual experience adequately as related to particular objects (as
opposed to some general way an object might be) as they are, e.g., actually distributed in
space (as opposed to how they are represented as distributed in space). At least, he argues, it
cannot do so without making worryingly ad hoc assumptions about the determination of
content, which suggests that it cannot do so at all.
112 |
Art thou not, fatal vision, sensible
To feeling as to sight? or art thou but
A dagger of the mind, a false creation,
Proceeding from the heat-oppressed brain?
I see thee yet, in form as palpable
As this which now I draw. (op cit.)
Here Macbeth struggles to question how genuine his experience is. A step further is
to remove Macbeths struggle, or even his disposition to struggle, and have him per-
fectly convinced and liable to assert (for whatever insane purpose) the proposition:
that is a dagger which I see before me
Both DR proponents and dissenters typically assume that one can build a lot into
such a case. One could, for instance, follow Campbells lead and suggest that the
dagger seems heavy, substantial, does not shimmer unduly etc. (2002, p. 117). The
point of these augmentations is to ensure that the experience corresponds in every
detectable manner to how the dagger would seem would that it were there. Call this
a perfect hallucination. Here perfection is judged according to its match with its
counterpart, i.e., the extent to which the hallucination is subjectively indistinguish-
able from ordinary perceptual experience. If the hallucination is subjectively indis-
tinguishable from ordinary perceptual experience then Macbeth cannot tell, simply
by being involved in the experience, that the experience is not an instance of per-
ception.
Call such states as Macbeths, faux perceptual states. As they are defined here, all
that is required for a subject to experience a faux perceptual state is that they be in
some sensory state that instantiates a grade at the apex of which is subjective indis-
tinguishability relative to perceptual experience. Faux perceptual states are many,
and variously found in discussions of perception and the nature of mind. When
construed as perfect, they are often simultaneously construed as realistic. John Mi-
chael Hinton provides an obvious instance, when he suggests that a perfect illusion,
would be an illusion of [...] reality, such that if you are involved in the illusion then
you cannot tell, simply by being involved in it, that it is not that [sic.] reality (1973,
p. 145). Perfect illusions are hard to come by. Indeed, J.J. Gibson even wrote once
that The illusion of reality is a myth (Gibson, 1986, p. 281). Part of his point is that
the actual cases where one might look for support, such as photorealistic painting or
trompe-lil, do not cut the mustard. Strictly speaking, he would argue, the pictures
only look real when viewing is subject to unnatural constraints, such as very brief,
staccato viewing, often through an aperture, and these cases are exceptions that
prove the rule that:
| 113

When no constraints are put on the visual system, we look around,
walk up to something interesting and move around it so as to see it
from all sides, and go from one vista to another. That is natural vision
(Gibson, 1986, p. 1)
Despite the increasingly impressive work in virtual reality since Gibsons time of
writing, it might be a bit much to suggest that even when natural vision is in play
the virtual objects and environments are subjectively indistinguishable from the real
thing. However, in the current context Gibsons reaction is overly harsh. For what is
at issue here is not so much whether perfect illusions actually do exist, but rather
whether they could not conceivably exist as a matter of metaphysical principle. In
phrasing the issue that way I take Humes dictum, that whatever is conceivable
without contradiction is metaphysically possible, as a token standard of metaphysi-
cal principle (1777/1975, IV, Part I, p. 25). After watching an episode of Star Trek
that made extensive use of the Holodeck, I think it would be fairly easy to conceive
of a situation that satisfactorily met the constraints of natural vision, but would
nonetheless be an illusion on most counts.
Another faux perceptual state that has had significant discussion is that of dream-
ing. Famously, the claim that dreaming involves a realistic experience is the basis for
the Cartesian sceptical scenarios that were the wellsprings of modern philosophy.
Making perspicuous use of marginally exotic experiences potentially enjoyed by the
reader, the scenarios envisaged by Descartes would appeal to some version of the
thesis that the experiences we take as dreams are (at their best) qualitatively similar
to what we take as waking (Newman, 2010, 3.1). In extending this qualitative simi-
larity to its limit one arrives at the notion of a perfect dream.
Recent interdisciplinary work on dreaming has developed the idea that it is an im-
mersive form of spatio-temporal hallucination (see Windt, 2010). Part of the moti-
vation of understanding dreams this way would perhaps be their allegedly passive
character (at least in pre-lucid cases). But the issue is certainly controversial, as oth-
er proposals have alternatively conceptualised dreaming as a species of imagination
(Ichikawa, 2009; O'Shaughnessy, 2002). The idea of a perfectly imagined scene or
object might seem to violate a classic Humean distinction between perception,
memory and imagination, where perception (surely, qua perceptual experience) has
a greater force and vivacity or liveliness than memory, and still greater than im-
agination (1739 - 1740/19602, p. 17).). But clearly this is a difference in degree rather
than a difference in kind (cf. Reid, 1785Essay I, Chapter I, pp. 28 - 30). Force, vi-
vacity and the like merely indicate the gradient along which the imagining must
ascend. In order to count as perfect, imagining must be conceived as having at least
the force and vivacity of perceptual experience.
114 |
The dispute over how to conceptualise dreaming turns largely on how one charac-
terizes the metaphysics and mechanism of hallucination and imagination. Without
passing judgement on these issues, it is sufficient for our purposes to conceive of
dream experience as ambiguous when approaching perfection. Let me illustrate by
stepping out of the arena of academic philosophical literature for the moment. It is
precisely the ambiguity in question that the Portuguese writer Jose Luis Borges ex-
ploits in his The Circular Ruins, a fictional tale of a mystic who strove to dream a
man into existence, to dream him completely, in painstaking detail, and insert him
into reality (Borges, 1998, p. 97). Naturally, the very idea of dreaming an object as
so real it becomes real is fantastic nonsense. But of present interest is the obstacle
Borges places in the path of his protagonist. Initially approaching the task by as-
sembling an audience of potential candidates, gradually the mystic whittles the fan-
tasy group down to one promising individual. But, just when it seemed success was
near, catastrophe ensued: [he] realized he had not dreamed [] He swore to put
behind him the vast hallucination that had first drawn him off the track, and he
sought another way to approach his task (Borges, 1998, pp. 97 - 98). The alternative
approach was to dream his subject part by part, beginning with the heart. Here Bor-
ges provides a compelling description of a perfect dream (perhaps qua imagination):
He dreamed the heart warm, active, secret ... Each night he perceived
it with greater clarity, greater certainty... from many angles, many dis-
tances. On the fourteenth night he stroked the pulmonary artery
with his forefinger, and then the entire heart, inside and out. (Borges,
1998, p. 98)
Borges aim, of course, is not to argue for or against any particular philosophical
view. Indeed, one can leave it as a beautifully open question whether this is another
hallucination, a mental image, or a dream as a suis generis state. All I want to em-
phasise is this: it is precisely because the earlier hallucination was so perfect that it
was mistaken for another faux perceptual state that the mystic had antecedent rea-
sons for believing he was in. At the height of perfection, all faux perceptual states
and perceptual states alike are to be conceived as subjectively indistinguishable on
the basis of their experience alone.
Returning now to Macbeths faux perceptual state. Although it is hallucinatory, it is
meant to be perfect, such that it seems like a perceptual experience in all admissible
respects; at least, it does so as far as Macbeth can tell. Indeed, the key is that in both
cases, if Macbeth were of the disposition to do so, he would be liable to assert
that is a dagger which I see before me
as an expression of the content of his experience. However, in only one of these cas-
es is there a dagger before him. Hence, DR is presented with a prima facie challenge.
| 115

DR holds that perceptual experience can explain how (most basically) the mind is
connected to mind-independent things. The chief explanans is the fact that its con-
tent is demonstrative content. But since the same contents can be entertained in the
presence or absence of the mind-independent object, that does not seem to do
much explaining at all.
DR theorists are certainly sensitive to this. In reaction, many contemporary defend-
ers argue that too much is being drawn from the assumption of indiscriminability.
This is usually framed as the assumption that instances of genuine perception and
perfectly faux counterparts are all conjoined under a common or fundamental kind
of experience (e.g., Martin, 2002b, p. 404; Snowdon, 2005, p. 136). Rejection or refu-
tation of this assumption involves denying the conjunction, and asserting a disjunc-
tion of kinds of experience. Adopting a slightly different language, one could say
that a perfect experience, on the common kind assumption, can be equally charac-
terised as an instance of perception, an instance of perfect illusion, an instance of
perfect hallucination, and an instance of perfect imagination. Someone who rejects
(or seeks to refute) the common kind assumption will hold that there is some fea-
ture of perceptual experience, such that it ought to be characterised as either a gen-
uine instance of perception, or an instance of perfect illusion, or hallucination, or
imagination. Authors that disjoin kinds of experience in this way are contemporarily
called disjunctivists.
86

To set disjunctivism in relief, consider briefly the other extreme. There is a family of
contemporary theories of consciousness that are adequately dubbed virtual-reality
theories of consciousness (Revonsuo, 2010, pp. 197 - 200). Virtual reality theorists
take perfection to be a significant explanandum for any adequate theory of con-
sciousness. Antti Revonsuo characterises the situation as follows:

86
See Haddock and Macpherson (2008) and Byrne and Logue (2009) for recent collec-
tions of classic and contemporary papers on disjunctivism. The earliest to outline (if some-
what obscurely) a disjunctive theory of perception was Hinton in his (1973). Snowdon was
perhaps the first state the view with clarity in his (1980 - 1981); and see his (2008) for a useful
discussion of Hintons work. Particularly influential more recent discussions are McDowell
(1998, pp. 228 - 259) and a series of publications by Martin (2002b, 2004); see Nudds (2009)
for a useful review of critical responses to the latters work. For completeness, it ought to be
noted that although the interests of DR mesh nicely with disjunctivism, it does not neces-
sarily entail realism. For instance, A.D. Smith (2008) argues that Husserl is readily described
as a disjunctivist, despite equally valid reasons for thinking he was an idealist (for which see
A.D. Smith (2003, pp. 71 - 110)).
116 |
Our experiences seem to happen in a real world, a spatially extended
physical world beyond our bodies. The visual qualities we see are lo-
cated distally, not in here but over there, on the surfaces of external
objects. Our bodies seem to be in the centre of this world, and inter-
acting directly with the physical objects of the world. One thing is for
sure: Our experiences do not seem to happen somewhere, or indeed
anywhere, inside our brains! (Revonsuo, 2010, p. 197)
However, these phenomena are treated as tantamount to pervasive hallucination:
out-of-the-brain experiences generated by (or at least, explained by positing, what
might be metaphorically understood as) a virtual-reality simulation created by the
brain (Revonsuo, 2000, p. 65). The parenthetical qualifications are necessary in or-
der to see that positing a metaphor, virtual reality simulation, and a theoretical con-
struct, representation, is meant to provide significant purchase in modelling and
explaining the phenomena in question. Hence, virtual-reality theorists typically
advocate qualified use of the metaphor as a model for consciousness, and unquali-
fied use of the theoretical construct in explanation of consciousness (Revonsuo,
2000, pp. 65 - 66). The metaphor is qualified in so far as the limitations of contem-
porary virtual-reality technology are recognised and crucial difference are highlight-
ed: Flight simulators drive phenomenal models of reality, but they do not yet create
them (Metzinger, 2003a, p. 556).
87
The unqualified use of the theoretical construct
is simply that virtual-reality theorists conjoin states of perception, illusion, halluci-
nation and imagination under a common kind of state, that is, under a common
kind of neural representational process. Thus according to a leading advocate of the
view, Thomas Metzinger:

87
And a further qualification that is crucial for (and distinctive of ) Thomas Metz-
ingers theory is the claim that human brains simulate the pilot as well (Metzinger, 2003a,
p. 556, emphasis original).
| 117

[A] fruitful way of looking at the human brain [...] is as a system
which, even in ordinary waking states, constantly hallucinates at the
world, as a system that constantly lets its internal autonomous simu-
lational dynamics collide with the ongoing ow of sensory input, vig-
orously dreaming at the world and thereby generating the content of
phenomenal experience. (Metzinger, 2003a, p. 52)
88

This unabashed conjoining of kinds of faux and perceptual states places virtual-
reality theories neatly amongst the family of views that disjunctivists would want to
reject, which we can call conjunctivist theories. And just as disjunctivist theories
frequently involve an endorsement and defence of DR, conjunctivist theories just as
frequently do not. Hence it is unsurprising that virtual-reality theorists are indirect
realists at most (e.g. Lehar, 2003, pp. 1 - 31).
The main strength of the conjunctivist position is that, at least in the guise of a vir-
tual-reality theory, it has tools to explain how faux perceptual states might be per-
fect. In this regard, the conjunctivist might see disjunctivism as guilty of an embar-
rassing silence on the matter.
89
But if the motivations of DR are what I have de-
scribed them to be, then the issue between the two theories is more complex than
that. For the aim of DR is not so much to offer an account of perceptual experience,
but to use an account of perceptual experience to restore a common-sense picture

88
Two is hardly a family (of theorists). But others include Steven Lehar (2003), and
Rick Grush, who claims that understanding representational processes in the brain as (what
he calls) emulation processes explains how it is that the same process that can operate
offline as imagery, can operate online as the provider of expectations that fill in and inter-
pret bare sensory information(2004, p. 392), where the role of sensory input from the envi-
ronment is merely ...to constrain the configuration and evolution of this representation. In
motto form, perception is a controlled hallucination process (393). Note, that these are
claims about perception, imagination and hallucination. So, whilst it might be true that
Grush is a virtual-reality theorist the thought that he intends to make claims about the na-
ture consciousness per se might be stretching his intentions too far; but, see his (2007b)
where the same architecture is extended to account for spatial phenomenology (or at least
what he calls spatial purport).
89
Disjunctivists on the other hand put an alternative spin on the matter, suggesting a
sort of respectful quetism in recognition of the mutually agreed limits of self-knowledge
with respect to subjectively indistinguishable states. Thus the line of argument is this: The
fact that genuine and perfectly faux states are subjectively indistinguishable, tells us no more
than that they are subjectively indistinguishable. In particular, it does not and cannot tell us
that they are the same kind of state. If this conclusion is secured, it opens up the possibility
of appeal to corollary issues to argue that DR should be favoured on reflective equilibrium.
See Martin (2004) for arguments along these lines.
118 |
of how the mind is related to things independent of it. In this regard, consider again
what a conjunctivist view seems to be committed to, as we have set things up. Qua
conjunctivists, they are clearly advocating mass error of an interesting kind. For the
less experience seems perfect, and thus the less experience (for all intents and pur-
poses) seems perceptual, the less error the subject is subjected to. In contrast, the
main strength of the disjunctivist position is that it allows for the possibility of
thoughts actually referring to the mind-independent world, and indeed knowledge
of that world, through perceptual experience as a suis generis state. Hence, the
question that the disjunctivist might see as an embarrassment to the conjunctivist is
this: How then, can it be that consciousness makes knowledge of the world possi-
ble? The thrust of the question can be put more directly as a criticism as follows:
There is no sense we make of thought and knowledge reaching out to
the world in a way that transcends what is available to the subject
from within the perspective of consciousness. This form of magic re-
alism involves depriving the concepts of knowledge and thought of
their normal meaning precisely by denying constitutive connections
with consciousness. (Eilan, 1997, p. 236)
However, that does not obviously conclude the issue. The safest conclusion here I
think is that any view that does not allow for the possibility of perceptual experience
itself being in an appropriate relation to mind-independent reality will end up hav-
ing to say something counter-intuitive about how we can think about reality, and
know certain states of affairs. But one thing learnt from the history of inquiry is that
a theory being counter-intuitive does not make it more likely to be false.
Meanwhile, DR is left in a position where it is not obvious how it meets its purport-
ed aim. If the model of the conjunctivist position is to be appropriately contrasted,
two things need to be done. The first is to find a metaphor (as the virtual-reality
theorists have done) that exhibits an appropriate sensitivity to the ongoing discov-
ery of causal relations underlying perceptual states. For as things stand, it is hard to
think of perceptual experience of objects on the DR model, when
[...] there is a sense in which cognitive processing is a 'common fac-
tor', an element found in both veridical and hallucinatory processing.
Whether or not there is an external object being seen, the same fea-
tures may be located on just the same feature maps, and they may be
bound together in just the same way. (Campbell, 2002, p. 118)
What is needed is a way of conceiving of DR such that it is compatible with any par-
ticular discovery concerning the biology, computational, functional architecture etc.
of perceptual systems. Furthermore, what is needed is a metaphor of a similar gen-
erality to the virtual-reality metaphor in play in conjunctivist accounts. Campbell,
| 119

following up a discussion of these issues, suggests the following as adequate to the
purpose:
Suppose we have a medium which, like glass, can be transparent. But
suppose that, unlike glass, it is highly volatile, and needs constant ad-
justment and recalibration if it is to remain transparent in different
contexts. Suppose, in fact, that the adjustment required is always
sensitive to the finest details of the scene being viewed. The upshot
of the adjustment, in each case, is still not the construction of a rep-
resentation on the medium of the scene being viewed; the upshot of
the adjustment is simply that the medium becomes transparent. You
might think of visual processing as a bit like that. It is not that the
brain is constructing a conscious inner representation whose intrin-
sic character is independent of the environment. It is, rather, that
there is a kind of complex adjustment that the brain has to undergo,
in each context, in order that you can be visually related to the things
around you; so that you can see them, in other words. (2002, p. 119)
90

Secondly, what is needed is a way of reinstating the explanatory role of conscious
experience. The proposal noted earlier was that perceptual experience is held to
have a certain kind of content that secures a direct-relation between mind and
world. But the trouble with that attempt is that it makes the relation one of success
(qua successful reference) and the difficulty with that is that the relation might just
as well fail to hold, as in cases of perfect faux perceptual experience (Brewer, 2006,
pp. 166 - 168). In a situation like this, it is hard to then press the explanatory role of
perceptual experience. Indeed, what it shows is that on such an account, it was nev-
er perceptual experience that was playing an explanatory role in the first place.
The simplest response in a situation like this is to get back to basics. Campbell offers
a path towards reconciling the formulation of DR with its general aims, by noting
the (retrospectively obvious) fact that, if experience of objects is to explain our abil-
ity to think about objects:

90
This indicates the extent to which metaphors only hold as much implication as their
theoretical context; cf. Metzinger (2005b, p. 13).
120 |
That means that we cannot view experience of objects as a way of
grasping thoughts about objects. Experience of objects has to be
something more primitive than the ability to think about objects, in
terms of which the ability to think about objects can be explained.
(2002, p. 122)
With these points in mind, the importance of a certain account of perceptual expe-
rience becomes more evident. For if perceptual experience is construed as anything
but a simple and primitive relation between mind and world, then it cannot be ap-
pealed to as a distinctive grounding condition for thought and knowledge. If DR is
to appeal to perceptual experience in this way as a grounding for the possibility of
conceiving the world as mind-independent, then it cannot build a concept of the
world as thus into the experience. What is required is an account of perceptual ex-
perience that does not suggest any identification or any form of dependency be-
tween a subjects experience of the world and their thoughts concerning the world.
And what is tricky is providing such an account, whilst simultaneously providing an
account of how it perceptual experience can seem to be of things that are mind-
independent.
5. 2. Immediacy
Having said all that (and I am aware that it was a lot) there is a sense in which the
relation between DR and the account of perceptual experience it appeals to is actu-
ally quite simple. This is especially clear when put in contrast with conjunctivist
views of the ilk presented in the last section. For whilst those views are not required
to take the common sense conception of perceptual experience at face value, so to
speak, DR certainly is thus required. Witness for instance, the following passage
from Mike Martin:
[The] phenomenological character of all perceptual experience re-
quires us to view the transparency and immediacy of perceptual ex-
perience as involving actual relations between the subject and the
objects of perception and their features. In just the case of veridical
perception, the experience is a matter of certain objects being pre-
sented as just so (Martin, 2002b, p. 402)
The next section will be devoted to the transparency of perceptual experience, but
for now focus on the question: what is the immediacy of perceptual experience sup-
posed to be? In a review article, Matt Soteriou characterises it as follows:
| 121

[Our] perceptual experiences are non-neutral about the actual pres-
ence of the mind-independent objects and features in our environ-
ment they are directed upon. They appear to have a coercive power
and authority over our beliefs. (2009, 3.3.)
There are obvious epistemological connotations to the phenomenon. J. L. Austin,
for instance, in discussing a refined form of Foundationalism, on which there is
supposed to be some special kind of sentences whose function it is to formulate the
evidence on which other kinds are based (1962, p. 115), has his audience consider a
case in which such indirect justification is superfluous:
If I find a few buckets of pig-food, thats a bit more evidence, and the
noises and the smell may provide better evidence still. But if the an-
imal then emerges and stands there plainly in view [...] its coming in-
to view doesnt provide me with more evidence that its a pig, I can
now just see that it is, the question is settled (ibid., p. 115, italics orig-
inal)
The point can easily be abstracted from its epistemological context, for whether or
not the epistemological question is settled, there is no sense in Austin merely enter-
taining the belief that there are pigs in his part of Oxford when visually confronted
with the honking beast.
In this section, I will attempt to navigate through the several pitfalls DR must avoid
even with a basic connection between belief and perceptual experience. Consider
David Armstrongs analysis of perception as a form of belief acquisition. Arm-
strongs thesis is that perceptions are acquirings of beliefs. A first worry is that it
might be thought to follow from this that the intentionality of perception reduces
to the intentionality of the beliefs acquired (Armstrong, 1968, pp. 210 - 211); or in-
deed that perceptions [...] are nothing but the acquiring of beliefs about the world
by means of our senses (Armstrong, 1961, p. 132).

The difficulty is that it seems that
simply believing something can occur whilst not perceptually experiencing the
thing believed to be so. Austin might have believed there were pigs in the vicinity
because someone convincing told him that there were. But this ought to be con-
ceived as a rather different experience to a head-on confrontation with a pink-
skinned, pig-shaped, stinking, snorting beast. Searle expresses the basic point as
follows:
122 |
I see a yellow station wagon in front of me, the experience I have is
directly of the object. [...] The experience has a kind of directness,
immediacy and involuntariness which is not shared by a belief I
might have about the object in its absence. (1983, pp. 45, 46)
What is needed is an understanding of how perceptual experience can involve the
experience of believing something in virtue of perceptually experiencing the object
with respect to which the belief is held. And if the goals of DR as described earlier
are correct, this needs to be achieved without assuming that a perceptual experience
is simply a belief, or that it is in some way dependent upon belief. From this angle of
approach, Searles remark serves as a challenge, one that might seem almost impos-
sible to meet. In fact, I will argue later that an appropriately weaker version of this
challenge can be met. But first it will be useful to consider a more common contro-
versy over the relation between perceptual experience and belief.
Remarks like the following are typical: there are times when belief and perception
come apart. Optical illusions are a case in point. Knowing that one is witnessing an
optical illusion does not make the illusion go away (Bermdez 1995, p. 335). Take
the optical illusion in Figure 10, for instance. Looking at the lines on the left, the
history of perceptual psychology is testament to the fact that most people say that
the top line looks longer than the bottom. But looking at the lines on the right, it
can perhaps be seen that the endpoints of the two lines (sans wings) are parallel. So
looking back to the left, it is certainly rational to believe that the two lines are of
equal length (and they are), but most might still say that one looks longer.

Figure 10 The Mller-Lyer illusion
But perhaps as regards the relationship between perceptual experience and belief,
there are illusions and then there are illusions (so to speak):
| 123

Sitting in the perceptual psychologists laboratory it is easy to tell
myself that the things I know I am being induced to see are mere il-
lusions [...] in such a situation I do disbelieve my senses [...] What if,
however, the psychologist made it seem to me really seem that a
six-foot spider was making its way relentlessly toward me? I doubt I
should remain in my seat. (A. D. Smith, 2001, p. 291)
Smiths anecdote illustrates the need for a conception of belief that can absorb both
kinds of case. He suggests that perceptual and theoretical belief are distinct in so far
as perceptual belief can exist alongside a contradictory theoretical belief (A. D.
Smith, 2001, p. 292). This distinction would allow the DR theorist to meet a
weakened version of the challenge described earlier. Schematically put, that
challenge was to show how:
(i). perceptual experience can involve the experience of believing something
in virtue of perceptually experiencing the object with respect to which the
belief is held;
whilst denying that:
(ii). perceptual experience is simply a belief, or it is in some way dependent
upon belief.
The weakening of the challenge would come in reading (ii) as:
(iii). perceptual experience is simply a theoretical belief, or it is in some way
dependent upon theoretical belief;
and this could be denied by claiming that:
(iv). perceptual experience is simply perceptual belief, or it is in some way de-
pendent upon perceptual belief.
The latter disjunct of (iv) would then satisfy (i).
This could do with some further illustration. For that purpose consider Eschers
Ascending and Descending (Figure 11).
124 |

Figure 11 Ascending and Descending by M. C. Escher.

Figure 11 is a famous depiction of an impossible object. But it is a depiction, and one
might quibble whether one can experience a picture itself as embodying an impos-
sible object. So, perhaps closer to the mark, consider a Lego version created by
Andrew Lipson (Figure 12).

Figure 12 Andrew Lipsons Lego construction of Ascending and Descending

Printed with permission, 2010 M.C. Escher Company BV

Images on the right present alternate angles on the labelled
image on the left. All retrieved from:
http://www.andrewlipson.com
| 125

The DR theorist could illustrate the distinction between perceptual and theoretical
belief as follows. Say that one is actually confronted with Lipsons Escher. By attend-
ing to different relative heights one can discern two opposite directions of ascen-
sion. Observing the labels at the corners of the staircase, on can ask a series of rela-
tive height judgements. I will give some presumptive answers to make the case. Is
(b) higher than (a)? Presumably yes. Is (c) higher than (b)? Yes, also. So one can
infer that (c) is higher than (a). But then ask: Is (a) higher than (d)? Yes. Is (d) high-
er than (c)? Yes, again. But now one can infer that (a) is higher than (c).
91
It is incon-
sistent to believe both that (c) is higher than (a) and that (a) is higher than (c). But
there is no need for the perceptual belief that e.g., (c) seems higher than (a) to be
consistent with the theoretical belief that (a) is higher than (c).
A feature of Smiths example that gave it persuasive power was that the seeming
presence of menacing arachnid with presumed ill intentions ought to result in eva-
sive behaviour. Now whether one sees the experience as a reason or a cause for the
behaviour, it treads dangerously close to the metaphysical territory leading to the
lands of the problem of intentionality. To rid the analysis of some artefacts of the
example, note that I could just as well have made the distinction with a look at Lip-
sons Lego Escher. There the perceptual beliefs accrued were required to play nei-
ther rational nor causal role. All that was required to illustrate was the contrast and
lack of any need for consistency between two beliefs.
This divide allows one to properly parse out the perceptual experience itself, from
the theoretical beliefs that the subject might have concerning the admissible con-
tents of reality. It allows one to separate the immediacy of the subjects perceptual
experience from the subjects conception of what she experiences. Certainly, percep-
tual experience can take the same (neutral) intentional object as thought; one can
think about what one experiences. But the key is to treat immediacy in such a way
that it does not become a determination of ones thoughts.
To outline such a treatment I will consider some alternative accounts of a difficult
case. Consider Oliver Sacks narration of the first of a recurring series of hallucina-
tory episodes suffered by a Hungarian man of the arts, named Frigyes Karinthy:

91
I owe this manner of illustration to Jeroen Smeets. See also Smeets and Brenner
(2008) where they use instances of the perception of impossible objects to provide a counter
example to a Helmholtzian/Bayesian account of perception (e.g. Krding and Wolpert
2008). On such a view, perception is understood as involving unconscious inference of the
most likely stimulus to have caused the retinal image. An impossible object is not likely at
all, and yet we can experience objects as being in impossible configurations.
126 |
He was having tea at his favorite caf in Budapest one evening when
he heard a distinct rumbling noise, followed by a slow, increasing
reverberation [...] a louder and louder roar [...] only to fade gradually
into silence. He looked up and was surprised to see that nothing was
happening. There was no train; nor, indeed, was he near a train sta-
tion. What were they playing at? Karinthy wondered. Trains run-
ning outside? [...] Some new means of locomotion? It was only after
the fourth train that he realized he was having a hallucination. In
his memoir, A Journey Round My Skull, Karinthy reflects on how he
has occasionally heard his own name whispered softlywe have all
had such experiences. But this was something quite different. The
roaring of a train [was] loud, continuous, and insistent. It was power-
ful enough to drown real sounds [...] After a while I realized to my
astonishment that the outer world was not responsible [...] the noise
must be coming from inside my head (Karinthy, 2008, pp. vii, viii,
cited in Farkas, forthcoming, ms p. 10)
In a forthcoming paper on the sense of reality in hallucination, Katalin Farkas con-
tinues where Sacks leaves off:
Karinthy reports that the next day same time, same cafe the
'trains' started to come again, but, unlike on the first occasion, he
didn't even look up: he knew it was something happening inside. The
experience repeated itself every day, and, together with the other de-
veloping symptoms of his tumour, became part of his daily routine.
(Farkas, forthcoming, ms p. 10)
She then offers an interpretation of the case:
This process, it seems to me, illustrates how the experience lost its in-
itial sense of reality. This can hardly be explained by a change in the
phenomenological features of the experience. What changed was Ka-
rinthy's knowledge that the event was not happening. (ibid., ms p. 10)
The problem with this interpretation is that it assumes the point at issue. The issue
is whether the experience itself changes, or merely Karinthys response to it. The
gradual familiarity that Karinthy exhibits can just as well be seen as the effects of his
theoretical beliefs on his behaviour. Given his evolving knowledge about his condi-
tion (he suffered from a brain tumour that was later operated) it would be no won-
der that his behaviour evolved accordingly. But none of this suggests that the expe-
riences themselves were affected by his theoretical beliefs.
Dreams provide another apposite case. They are often reported as being bizarre.
They are an interesting topic of artistic (particularly cinematic) study, because they
are conceived as being potentially wildly inconsistent and yet simultaneously realis-
| 127

tic. It is important to recognise, incidentally, that a perfect dream is not necessarily
what is known as a lucid dream. By definition a lucid dream involves insight upon
the fact that one is dreaming (Revonsuo, 2010, p. 247). Consequently, the famous
Cartesian sceptical scenario cannot be one that involves such lucidity. In passing,
Mike Martin makes just the error I am warning against:
Cartesian sceptical hypotheses such as lucid dreaming and malign
demons generalise cases of delusive experience: perfect hallucina-
tions. Given the manner in which Descartes treats dreaming, a lucid
dream may as well be an example of extended hallucination (Martin,
2009, p. 107)
The attribution of lucidity to the poor Cartesian subject seems to be straightfor-
wardly mistaken.
92
But still, the two are perhaps not mutually exclusive; a dream can
apparently be realistic despite the dreamers lucidity, as indicated in the following
narrative:
I was about to enter the house when, on glancing casually at [the
pavement] stones, my attention became riveted by a passing strange
phenomenon, so extraordinary that I could not believe my eyes
they had seemingly all changed their position in the night [...] Then
the solution flashed upon me: though this glorious summer morning
seemed as real as real could be, I was dreaming! (Oliver Fox, as cited
in Metzinger, 2009, p. 140)
Typically, insight is achieved by means of performing reality checks (see
Metzinger, 2009, p. 141 for a handy list). For the most part these are checks on con-
sistency, particularly when they pertain to the environment experienced. The guid-
ing assumption is that reality is consistent, and governed by constant principles,
principles that a dream might violate without ones notice. The purpose of the vari-
ous tests upon consistency is to determine what one ought to (theoretically) believe
about the experienced environment, i.e., whether one ought to (theoretically) be-
lieve that it is real. Again, according to the definition of lucid dreaming (and the
report above) this can be in spite of the fact that the objects of the dream environ-
ment are experienced as realistic. The purpose of said reality checks is not to deter-

92
In fact, it seems that what Martin is referring to is closer what I am calling a perfect
dream. In particular, a perfect dream that does not involve lucidity. I am allowing a perfect
dream to involve experience of an inconsistent environment. This serves to highlight how
dramatic the assumptions of the dream scenario are. Without even the possibility of discern-
ing ones predicament through reality checks etc., the scenario moves beyond even perfect
dreaming to involve what Andy Clark calls industrial strength deception (Clark, 2005).
128 |
mine how a dream environment looks or feels, or indeed how realistic it looks or
feels. It is to determine whether they are as they look and feel.
The distinction between theoretical and perceptual belief also enables one to ac-
complish the tricky feat that I set as an aim in the introduction and in the first sec-
tion of this chapter. It allows one to make initial sense of the idea that perceptual
experience can seem real, whilst resisting an assimilation of that experience to the
subjects thought about the objects of experience. Furthermore, it allows all this to
be done in a manner that is compatible with the neutral intentionality thesis.
5. 3. Transparency
In the course of an attack on an idealist metaphysics of perception in 1903, Moore
devoted some time to an analysis of sensation.
We have then in every sensation two distinct terms, (1) conscious-
ness, in respect of which all sensations are alike; and (2) something
else, in respect of which one sensation differs from another. It will be
convenient if I may be allowed to call this second term the object of
a sensation: this also [as with consciousness] without yet attempting
to say what I mean by the word. (1903, p. 444)
His ensuing complaint was that an idealist would identify consciousness with the
object, such that an expression sensation of blue denoted a single term. For if a
sensation of blue differs from a sensation of green in virtue of a second term,
namely, the particular blue in the former sensation and the particular green in the
latter sensation, then this second term is distinct from the first, namely, conscious-
ness, which is claimed to be universally present in all cases of sensation. By identi-
fying the two terms, arguably the idealist winds up identifying particulars with uni-
versals (G. E. Moore, 1903, p. 445).
93

Moore then embarks on one of two famous and influential digressions, suggesting
that idealists come to believe that the two are one and the same, rather than two and
different because:

93
For more on this interpretation of Moores argument and aims, see Hellie (2007, pp.
341, 358 n17).
| 129

when we refer to introspection and try to discover what the sensation
of blue is, it is very easy to suppose that we have before us only a sin-
gle term. The term blue is easy enough to distinguish, but the other
element which I have called consciousnessthat which sensation of
blue has in common with sensation of greenis extremely difficult
to fix [...] in general, that which makes the sensation of blue a mental
fact seems to escape us: it seems, if I may use a metaphor, to be
transparentwe look through it and see nothing but the blue; we
may be convinced that there is something but what it is no philoso-
pher, I think, has yet clearly recognized. (1903, p. 446)
Notice that Moore is talking about introspection of sensation. The nature of intro-
spection is a controversial topic as much as any other in philosophy of mind.
94
But
assume as a minimal characterisation, that when Moore engages in act of what he
calls introspection, he is doing something different to what is ordinarily the case in
an occurrence of a sensation of blue. The attempt, in any case, is to introspect upon
the act of experiencing, to catch the act itself, albeit in unusual circumstances. If
that is right, then what is striking is the extent to which nothing beyond the ordi-
nary is easily revealed, that the relationship between the act of experiencing and the
object of experience is so intimate. Moore is suggesting that it is understandable
that the idealist goes astray because it is hard to discern any distinction between the
two. In particular, introspection (whatever that is) can fail to reveal a distinction
between a mental element (consciousness) and a non-mental element (the object)
of perceptual experience. And this failure, on the part of idealists, is precisely that of

94
What one means by introspection, and the controversies one engages in, can vary
with ones theoretical interests. For instance, if one is interested in self-knowledge: Intro-
spection could be understood as ones ordinary capacities for rational and intelligent delib-
eration of consistency amongst beliefs (Shoemaker, 1996, pp. 224 - 246). Or even, simply
making ones mind up on a matter, by determining what the relevant state of the world is
(Evans, 1982, pp. 224 - 230; R. Moran, 2001, pp. 35 - 99). Alternatively, it could be understood
as a special form of inner sense-perception (Armstrong, 1968, pp. 323 - 338; Lycan, 2002). If
one is more interested in methods of studying the mind and consciousness, introspection
could be understood as a means of producing first-person reports, however unreliable
(Hurlburt & Schwitzgebel, 2007). If one is more interested in cognitive architectures under-
lying consciousness, the term could refer to the differential allocation of cognitive and atten-
tional resources to processes within an organisms boundaries, according to which some
state of affairs may be experienced as internal or external (Metzinger, 2003a, pp. 32 - 38).
Although the extension of Moores introspection must be one of these, he operates with a
vaguer notion: whatever it is that one does when asked to attend to the look of something,
how it tastes, feels etc. In the current context, the only substantial point required is that
whatever this is, it is something different from what occurs in ordinary perceptual experi-
ence.
130 |
failing to hold [consciousness] and blue before their minds and to compare them,
in the same way in which they can compare blue and green (G. E. Moore, 1903, p.
450). One could fail to do something because it is impossible, or one can fail to do
something that is possible but simply too difficult. Moore is of the opinion that ide-
alists fail in the second sense, that consciousness and its object can be distin-
guished if we look enough, and if we know that there is something to look for
(ibid., p. 450).
The other famous passage from that article contains a description of the failure it-
self:
the moment we try to fix our attention upon consciousness and to see
what, distinctly, it is, it seems to vanish: it seems as if we had before
us a mere emptiness. When we try to introspect the sensation of
blue, all we can see is the blue: the other element is as if it were di-
aphanous (G. E. Moore, 1903, p. 450)
Call this failure simpliciter, the transparency phenomenon. A great deal of the litera-
ture making reference to the transparency or diaphanousness of perceptual expe-
rience is bound up with issues concerning what the metaphysics of consciousness
should involve.
95
But this is in a rather different sense than concerned Moore. The
contemporary point of issue is whether (and for some, simply, how it is that) con-
sciousness is reducible to representational content.
96
The main challenge in this
regard is the claim that qualia are intrinsic properties of experience, to be found by
Moorean introspection, and that these are irreducible to representational content.
Occasionally, the challenge is not stated as such, but more generally as either the
claim that we experience non-intentional properties inherent in consciousness,
such as the intrinsic nature of our mental states (Harman, 2002, p. 664). Against

95
Discussion of transparency has been rather popular of late. Benj Hellie (2007, p. 356
n1) lists over eighty articles referring to or making use of some relevant notion of transpar-
ency published between 1990 and 2007. Although such an output is rather meagre in com-
parison to rates of publication in the natural sciences, it is quite outstanding for a single
topic in a subdiscipline of philosophy.
96
However, the discussion of transparency is just one perspicuous case of a more gen-
eral debate, in which no appeals to transparency need be made. Also the claim that con-
scious experience reduces to representational content is one of the stronger variants of the
relationship between the two (see Chalmers, 2004). Possibly the weakest relationship would
be that of supervenience. For a good reference on various conceptions of the relationship,
and a defence of the view that consciousness supervenes on representational content, see
Byrne (2001). For explorations of the complexities of that content and several constraints on
lower levels of description, see Metzinger (2003a, pp. 116 - 197).
| 131

this broader background, there are actually various kinds of transparency thesis
developed. I will discuss these theses to argue that, despite occasional suggestions
to the contrary, a considered view on transparency should not entail that perceptual
experience is an experience of representational content, or indeed the vehicle prop-
erties of that representation.
To begin, here is the most typical manner of appeal to transparency in the litera-
ture:
97

Suppose you have a visual experience of a shiny, blood-soaked dagger
[...] Now try to become aware of your experience itself, inside you,
apart from its objects. Try to focus your attention on some intrinsic
feature of the experience that distinguishes it from other experiences,
something other than what it is an experience of. The task seems im-
possible (Tye, 1995, pp. 135 - 136)
Tye claims it is impossible to focus on any intrinsic feature of the experience that
distinguishes it from other experiences. Implicitly, it is assumed that what the expe-
rience is an experience of cannot count as such an intrinsic feature, given that ex-
perience-of denotes a relation, an intentional relation. The transparency thesis is
that perceptual experiences have no introspectible features over and above those
implicated in their intentional contents (Tye, 1995, p. 136). In this context, the act
of introspection (along Moores vague lines) is to be construed as a test for meta-
physical theories of consciousness, i.e., some means of discerning a lesson to be
drawn from reflection on the phenomenology
M
of transparency. A theory which
claims that consciousness is identical to, reduces to, or supervenes on representa-
tional content, need predict no more than the following: When one introspects up-
on a perceptual experience, all one will experience is the object one looks at, touch-
es, etc.; there are no properties that experiences are held to have over and above the
properties of the objects represented. A view which claims that we directly experi-
ence intrinsic properties of our mental states over and above the properties of the
objects represented, would seem to require that these would be revealed in intro-
spection. But for some that does not seem to happen:
one's awareness seems always to slip through the experience to the
redness and shininess, as instantiated together externally. In turning
one's mind inward to attend to the experience, one seems to end up
scrutinizing external features or properties. (Tye, 1995, pp. 135 - 136)

97
For prototypical instances of the kind of discussion I am referring to, see Block
(2007a, pp. 533 - 570), Harman (2002), Tye (2002), Shoemaker (1990). For a deflationary
perspective on the importance of the debate, see Crane (2006b).
132 |
when one drinks in the blue of a summer sky or the red of a ripe
tomato [...] these are experienced, perceptually rather than introspec-
tively, as located outside one, in the sky or in the tomato, not as fea-
tures of one's experience (Shoemaker, 1990, p. 112)
98

And we find Martin, once again gazing out of the window of his North London
home:
When my attention is directed out at the world, the lavender bush
and its features occupy centre stage [...] when my attention is turned
inwards instead to my experience, the bush is not replaced by some
other entity [...] I attend to what it is like for me to inspect the laven-
der bush through perceptually attending to the bush itself while at
the same time reflecting on what I am doing. So it does not seem to
me as if there is any object apart from the bush for me to be attend-
ing to or reflecting on while doing this. (2002b, pp. 380 - 381)
99

The transparency phenomenon is held to reveal a reason for the failure in question.
One cannot distinguish consciousness and its objects, because consciousness is no
more than the representation of its objects (Tye, 2002). Call this transparency thesis
1 (TT
1
). TT
1
decomposes to two claims.
100
One is that we perceptually experience ob-
jects as external. By this it is meant that the objects of experience seem to be mind-
independent. So we have claim ():

98
It ought to be noted that later on in the article, Shoemaker gives several arguments
for the existence of qualia.
99
Although the procedure of Martins description is roughly that which others would
use to support an account of the representational nature of perceptual experience, his con-
sidered stance is neutral in that paper.
100
As far as I can tell, Martin was the first to note this explicitly (2002b, p. 378)
| 133

Perceptual experience is as of mind-independent objects, surfaces, vol-
umes, properties, relations etc.
101

Both the transparency thesis discussed in this section are committed to () or some-
thing sufficiently close to it. What distinguishes TT
1
is a second claim, that all we
experience are these mind-independent objects as they are represented to be. So we
have claim ():
Perceptual experience is exclusively of objects (surfaces, volumes, proper-
ties, relations etc.) represented as mind-independent.
102

Despite the open inclusion provided in the parantheses, () might be considered as
rather too strong.
103
But note that whether or not that is the case, endorsing claim
() already blocks a certain way of reading (), or even a certain way of reading such
claims as phenomenal character is the same as representational content (Tye,
2000, p. 45). What it blocks is the rather literal reading, that one perceptually expe-
riences representational content. To see this, consider Cranes remarks, whilst bear-
ing in mind the neutral notion of an intentional object discussed in the last chapter:

101
The expression as of here can be seen as more or less loaded. For a loaded in-
stance, probably intended by proponents of TT
1
, witness Martin Davies when speaking of
the notion of perceptual content immediately adding the clarifying parenthesis (of expe-
riences as of ... ) (1992, p. 23). A less loaded understanding is neutral on the issue of wheth-
er a perceptual experience is in fact a relation to mind-independent reality. This would be to
say the experience is as of not because of e.g., the possibility of cases in which it might not
involve a relation to mind-independent reality, but simply to state that the experience does
seem to be so.
102
Incidentally, I leave aside issues as to whether or not some or other higher-order
theory of consciousness (such as the higher-order perception theory) is correct. Although in
some ways connected to the issue of whether perceptual experience is transparent, disputes
concerning higher-order theories have little to no bearing on the broader project of articu-
lating a DR conception of perceptual experience. For critical reviews of higher-order theo-
ries of consciousness, see Gzeldere (1995) and Zahavi (2005, pp. 11 30).
103
For various counter-examples, see Block (2007a, pp. 539 - 543). For other counter-
examples in the course of the development of an alternative transparency thesis (TT
2
, be-
low), see Metzinger (2003b, pp. 359 - 362) and the discussion below.
134 |
If I see a rabbit, the rabbit is the intentional object of my experience.
My experience may also have the propositional content that there is a
rabbit running through the eld. But I do not see such propositional
contents or propositions; I see rabbits and elds. For this reason,
even if perceptual experience should be analysed in terms of rela-
tions to propositions, it is at best misleading and at worst absurd to
say that these propositions are the intentional objects of experience.
(2006b, p. 136)
Cranes point is insightful, but perhaps unoriginal, as it is notably similar to a point
made by Mike Martin:
It is true that if one endorses a Russellian conception of proposi-
tional content, one might make some sense of taking the objects and
properties that the content is about as its literal constituents. In that
case, one might endorse the thought that one delighted in [sc. per-
ceptually experienced] an aspect of content, but unless we also as-
sume that in being aware of a constituent of a content one is thereby
aware of the content, it still wont follow from being aware of the
ocean and its blueness that the experiences content is immediately
accessible to ones consciousness. (Martin, 2002b, p. 382)
As far as I am aware, all proponents of TT
1
hold that the content of perceptual expe-
rience is propositional. But whether they hold some other view of content is irrele-
vant. The key point is this: despite the way people sometimes speak about the rela-
tionship between the intentionality of perceptual experience and representational
content, it is inconsistent to claim () and () and to claim that one experiences
representational content. It is one thing to hold that experience has correctness
conditions. It is quite another to hold that we experience the correctness conditions
themselves.
Consider next, transparency as a claim about the architecture of systems that meet
certain minimal conditions for enabling conscious experience. Putatively, certain
states of these systems (in virtue of which the system is conscious, dubbed phe-
nomenal states) exhibit a feature which Thomas Metzinger (2003a, p. 169) dubs
phenomenal transparency.
104
The feature is characterised by claim ():

104
Metzinger once upon a time called this semantic transparency (2000, p. 299). Van
Gulick! But later in reviewing various ways in which a notion of transparency could be em-
ployed he refined his usage and adopted the term phenomenal transparency for essentially
the same phenomenon (Metzinger, 2003a, p. 165; but see also pp. 386 387).
| 135

For any phenomenal state, the degree of phenomenal transparency is in-
versely proportional to the introspective degree of attentional availability
of earlier processing stages (Metzinger, 2003a, p. 165).
Whilst Moores use of the word introspection is vague, Metzingers usage is well-
defined, but theory-laden; the latter, largely because he is a staunch internalist
about both the vehicles and contents of conscious mental states.
105
To be clear on
what Metzinger intends by (), some brief remarks are in order. On Metzingers
view, all phenomenally conscious states are exhaustively constituted by neurocom-
putational properties. All are thus in some sense internally directed, and thus in-
trospective in various different senses of internal. Metzinger distinguishes four
different notions of introspection on the basis of these principles (2003a, p. 36). The
first is the most relevant. By Metzingers lights, Introspection
1
(note the subscript)
is
a phenomenal process of attentionally representing certain aspects of
an internal system state, the intentional content of which is consti-
tuted by a part of the world depicted as external. The accompanying
phenomenology is what we ordinarily describe as attention or the
subjective experience of attending to some object in our environ-
ment. (2003a, p. 36)
Introspection
1
draws upon a functional notion of attention:
attention is a process of subsymbolic resource allocation taking place
within a representational system exhibiting phenomenal states. It is a
form of non-conceptual metarepresentation operating on certain
parts of the currently active, internal model of reality, the conscious
model of the world. (2003b, p. 356)

105
For his distinction between phenomenal and intentional content (the former a sub-
set of the latter) see Metzinger (2003a, p. 173). For the distinction between internalism and
externalism about the contents and vehicles of intentional and phenomenal content, see
Hurley (2010).
136 |
Clearly there is a lot built into the account. But the centrepiece assumption is that
representational systems are conscious in virtue of internal modelling.
106
This as-
sumption and others are made against the background of a constraint-satisfaction
approach to the study of consciousness (Metzinger, 2003a, pp. 116 - 208, and passim;
2005b; Weisberg, 2005). That is, it constitutes an answer to the question: What con-
ditions must a representation satisfy if it is to enable conscious experience? One of
these conditions is that in standard situations and from a rst-person perspective
the contents of phenomenal states always are in a world, and furthermore that
This world is presented in the mode of naive realism (Metzinger, 2003a, p. 131).
Metzinger does not mean anything particularly complex in this regard, merely that
whatever is transparently represented is experienced as real and as undoubtedly
existing by this system (Metzinger, 2003b, p. 364). Accordingly, Metzinger endorses
something near enough to ().
107
Together with () this amounts to transparency
thesis 2 (TT
2
).
It ought to be clear that TT
2
is a substantial thesis in the way that the mere transpar-
ency phenomenon is not. For the particular what is not introspectively accessible on
Metzingers view is the representational character of the contents of conscious ex-
perience (2003a, p. 169). By Metzingers lights, naive realism ought to be construed
as a phenomenal property and it ought to be taken as a literal instance of naivet
concerning the actual facts of the matter underlying subjective experience (under

106
Several other items to clarify are resource allocation, sub-symbolic, metarepre-
sentation, and non-conceptual. Resource allocation is allocation of information-
processing resources. It is sub-symbolic because the information is in a non-linguistic and/or
non-symbolic format (pace Fodor, 1975). A metarepresentation is a representation of a repre-
sentation. The label nonconceptual takes at least two meanings, and Metzinger probably
intends both. Firstly, a natural consequence of Metzingers enthusiasm for an account of
consciousness that it is sensitive to biological history (e.g. 2003a, p. 178; 2003b, p. 362) is that
one should not presuppose an organism need possess concepts to be in (me-
ta)representational states, whether or not specification of their content requires the use of
concepts (Bermdez, 1998, pp. 49 - 101). But also, given that their format is sub-symbolic,
Metzingers (meta)representations are non-conceptual in another sense, in that their con-
tent is not structured in a manner amenable to conceptual articulation (this would be a
version of 'content' non-conceptualism, see Heck, 2000, pp. 482 - 488). What is important
in any case is that the resultant phenomenology is that of sensorily experiencing something,
and not that of conceiving something to be a certain way. In attending to an object we in-
crease our degree of alertness and orient towards the object in question. However, we do so
without mentally forming a concept for this object (Metzinger, 2003b, p. 356).
107
Nowhere do I find Metzinger wishing to distinguish between perceptually experi-
encing something as real and perceptually experiencing something as mind-independent.
Though there is some nuance to his endorsement of () that will become apparent shortly.
| 137

the slogan appearance is not knowledge). Indeed, Metzinger claims that employ-
ing phenomenal transparency as a conceptual constraint forms a decisive step in a
thorough-going error theory of self-knowledge: a view of the nature of subjectivity
and consciousness as a fundamentally neural process of a deep complexity that is
ordinarily beyond the reach of introspection (Metzinger, 2000, p. 299; 2003a, pp.
330 - 340, passim). But ordinarily holds some weight here. For note that claim ()
holds that phenomenal transparency is a graded phenomenon. Hence, the possibil-
ity that phenomenal states can exist at the other end of the scale is built into TT
2
.
Accordingly, Metzinger, following in the footsteps of Moore, provides cases in which
he suggests one can be aware of the representational nature of ones conscious expe-
rience:
Sensory experience is the paradigmatic example of fully transparent
phenomenal content. There are, however, examples of sensory opaci-
ty, for instance, during extremely short transition phases in bistable
phenomena, e.g., if a consciously experienced Necker cube switches
from one interpretation to the next and back, or during the phenom-
enon of binocular rivalry (2003b, p. 360)
In part, this echoes the appeals that others make to binocular rivalry as a window
into the deep structure of consciousness. For example, Leopold and Logothetis
claim that [p]henomena like binocular rivalry demonstrate that, although our
brain can simultaneously analyse two conflicting representations, a selection pro-
cess [...] ensures that we are only aware of one (1999, p. 262). But in fact Metzingers
claim is much stronger. His claim is that this phenomenon (and others) reveals to
the subject, within the occurent sensory experience itself, that phenomenal pro-
cesses are representational processes (Metzinger, 2003a, pp. 170 - 175, 178).
Thus () certainly puts TT
2
in conflict with (). And as endorsement of () figured
in defence of the neutral intentionality thesis, it looks like I need to either attack ()
or show why TT
2
is compatible with the neutral intentionality thesis after all. To be
safe, I will do both.
Take a Necker cube such as that in Figure 13.
138 |

Figure 13 A Necker cube
Necker cubes are of interest because they are thought to be a paradigm instance of
an ambiguous image. The experience of a Necker cube as ambiguous can be de-
scribed very basically as follows: when looking at the cube, it seems to flip between
B looking closer than H, and H looking closer than B. There is a further unambigu-
ous possibility, which would be to experience the cube as merely seven abutted
two-dimensional shapes: a square, two triangles, and four irregular quadrilaterals.
Now, say that representational processes enabling consciousness are phenomenal
representations, and that phenomenal representations represent by modelling pos-
sibilities. Further, leaving aside the two-dimensional possibility, assume that the
bistability is enabled by two competing models. These models are of two mutually
exclusive and jointly exhaustive possibilities with respect to the same cube (experi-
enced as a cube): one in which B is closer than H, and another in which H is closer
than B. These possibilities are crude statements of the correctness conditions for the
models, and in turn the phenomenal representations. What does one experience, in
each case when the Necker cube flips? Metzinger claims that because the Necker
cube occasionally flips the subject experiences the cube as a representation. If his
representational analysis is along the lines sketched here, the claim is really that the
subject experiences the cube (as it flips) as a model: a model of B being closer than
H, or H being closer than B.
108
But why say this? Why not simply say that the subject
experiences either B being closer than H, or H being closer than B?
By my lights, someone who endorsed () & () ought to say just this. Earlier, I drew
on Cranes suggestion that describing the objects of perceptual experience as propo-
sitions is misleading and even absurd. I argued further that such a description is
inconsistent with claiming both () & (). But it was really just an extension of
Cranes (and perhaps originally Martins) point, which I take to be that subjects do
not perceptually experience contents themselves. This same point cannot be applied

108
See 8.5 below.
| 139

to (), at least not straightforwardly. For whilst Cranes point is that we do not expe-
rience contents, () suggests it is possible that properties of the vehicles, the pro-
cesses that enable contents, are experienced; namely, that the objects of experience
are representational processes. The gap between Cranes point and () amounts to
the conceptual distance between vehicles and contents, or less metaphorically, the
vehicle-content distinction.
The vehicle-content distinction is between the carriers of content and the content
carried. It is widely used. For instance, expressing the most common usage in phi-
losophy of mind, Hurley writes: We should distinguish properties represented in
content from properties of the vehicles of content: the subpersonal states and pro-
cesses that carry content (1998, p. 17, see also pp. 32 - 41, 144). Subpersonal states
and processes are such in so far as they are mechanical and thus not normatively
constrained in the manner that states typically attributed to persons are, such as
beliefs and desires etc. (see Dennett, 1969/1983, pp. 90 - 98; 1987, pp. 43 - 68). De-
pending on ones proclivities, these might be the processes studied in cognitive psy-
chology, cognitive neuroscience, neurophysiology etc.
109

However, the distinction itself is broader in kind than that between personal and
sub-personal explanation, or indeed non-mechanical and mechanical. One can
easily find people saying that experience represents, a phrase which (oddly
enough) suggests that the experience is itself the vehicle.
110
In other quarters, one
finds the claim that organisms are the vehicles for replicators, alias genes (Dawkins,
1976/2006, p. 19ff.).
111
And in still others, a distinction [...] between using an expres-
sion and mentioning it (A. W. Moore, 1986, p. 173): the distinction employed when
I say that I pity the fool that uses the expression I pity the fool in a piece of aca-
demic writing. With respect to organisms and genes, perhaps one might not have
much trouble finding distinct entities. But with respect to the use-mention distinc-
tion, a distinction on the basis of reification might be misleading. The same ought
to hold in general with respect to vehicles and contents, at least amongst those nat-
uralistically inclined.

109
With respect to mental content, a distinction along these lines goes back at least as
far as early modern work on perception (Yolton, 1996, pp. 183 - 214).
110
For a fairly randomly chosen but nevertheless typical selection, see Tye (1996, pp.
120, 121, 124).
111
For issues concerning content attribution in biology, see Maynard Smith (2000) and
Griffiths (2001). For a comparison of these with respective issues in the philosophy of
cognitive science, see Wheeler and Clark (1999).
140 |
Reification worries seem to be exactly what motivates Metzingers violation of the
vehicle-content distinction. Speaking of () he says that
this denition departs from the classic vehicle-content distinction.
The standard way of dening transparency would be to say that only
content properties of the phenomenal representata are introspective-
ly available to the system, and not vehicle properties. The vehicle-
content distinction is a highly useful conceptual instrument, but it
contains subtle residues of Cartesian dualism in that it always tempts
us to reify the vehicle and the content, by conceiving of them as dis-
tinct, independent entities. (Metzinger, 2003a, p. 166)
He then goes on to suggest a solution
A more empirically plausible model of representational content will
have to describe it as an aspect of an ongoing process and not as some
kind of abstract object. What we need is embodied content, as it
werean ongoing and physically realized process of containing, not
a content. (ibid., p. 166)
But this can be recognised whilst still being sanguine about a distinction between
vehicles and contents. Admittedly, the format of a representation, the choice of ab-
straction, is closely tied to properties of the vehicle. Representing everything being
illuminated (a content) by means of ink on paper (a vehicle) is probably easiest
done in a natural language (a linguistic format). Ones choice of abstraction may be
more or less distant from a propositional format, depending on how close ones the-
ory of content is to biological reality. For instance, Paul Churchland (inspired by
advances in the study of sensorimotor coordination in the 1970s and 80s) suggests
that the right form of abstraction is a position in a suitable state space of a dynam-
ical system potentially implemented by cortical and cerebellar brain activity (1986,
p. 280, emphases original). Perhaps further along these lines, Metzinger muses that
increased attention to evolutionary constraints in cognitive science could motivate a
view of content as a dynamical physical process pulsating across the boundaries of
the system, in virtue of ongoing interaction between an agent and its environment
(2003a, p. 113, emphasis original, see also pp. 13 n1, 111 - 116).
112

All of that suggests that there might be more biologically plausible views of content
than those tied to propositional formats: views that posit complex interaction be-
tween content (or containing), and the material processes that enable it. That pos-
sibility is perfectly compatible with the vehicle and the content being distinct, qua

112
For early articulation of thoughts in this direction, see Churchland and Churchland
(1983), and Clark (1989, pp. 61 - 80).
| 141

having distinct properties. For instance, a vehicle such as paper might have proper-
ties such as weight, selective permeability, light reflectance etc. The content ena-
bled, everything being illuminated, need not have any of these properties. Indeed
none of these properties need make a difference to a vehicles capacity to enable the
content that it does.
113

Hence, even though () violates the vehicle-content distinction, the motivations
adduced in its favour are really motivations for greater sophistication in format,
greater sophistication in the theory of content, and greater sophistication in how
one treats the relationship between vehicle and content. All that is certainly wel-
come. But as far as that goes, the vehicle-content distinction is still in place. So ()
violates the vehicle-content distinction without reason.
114

But say some reason could be found: would TT
2
be incompatible with the neutral
intentionality thesis? Remember that I mentioned earlier that reference to ordinary
conditions holds some weight indicating that phenomenal transparency is graded.
At the transparent (viz. non-opaque) end, that is in standard configurations,
Metzinger admits one's phenomenal experience has an untranscendably realistic
character (Metzinger, 2005b, p. 2). I suspect such standard configurations to be
the ordinary conditions of perceptual experience. To see why, consider another ex-
ample Metzinger uses to illustrate phenomenal opacity:

113
I am willing to concede though, that some of these properties might make all the
difference in certain cases. Even if the carrier of content is readily distinguished from con-
tent itself, complex inter-relations in a proper understanding of the two is a live option.
Metzingers hopes for closer attention to the dynamic embodiment of representational pro-
cesses might well be satisfied by Mark Rowlands recent work on representational deeds (see
Rowlands, 2006, p. 201 ff.). Though it is notable that Rowlands (as with others interested in
the dynamics of embodiment, e.g. Hurley, 1998, 2010) maintains strict adherence to the ve-
hicle-content distinction (op. cit. 2006, p. 28), and occasionally uses it in arguments to sup-
port his case (op. cit. 2006, pp. 76, 181).
114
If the view then comes to claiming that content properties are introspectively avail-
able (claim () perhaps) then it would be an instance of what Millikan calls a content inter-
nalizing and then content-externalising move, viz. projecting properties claimed in or by
the visaging onto the inner vehicle of the visaging and then projecting properties of the
vehicle of the visaging into the visagings content (1991, p. 441ff.). Millikan uses the term
visaging to stand in relation to perceiving in the way believing stands to knowing. I would
just as well use perceptually experiencing. For perhaps the opposite view, see Byrne (2009).
142 |
If a subject in a laboratory experiment and under the influence of a
hallucinogenic psychoactive substance (say, LSD or 2-CB) observes
abstract geometrical patterns on the wall, breathing, and slowly
evolving into deeper and ever deeper forms of ineffable beauty, then
she will frequently be aware of the representational character of her
visual experience in that sub-region of her phenomenal space. Typi-
cally, the subject will immediately have doubts about the veridicality
of her experiential state, cognitively bracket it and take back the
existence assumption, something which effortlessly goes along with
visual experience in standard situations. (2003b, p. 360)
Neutrally intentional perceptual experiences, both those that occur in standard
situations and those that arise from perfect faux perceptual experiences, are at the
transparent end of Metzingers scale. The neutral intentionality thesis is compatible
with the possibility of phenomenal opacity with the proviso that a phenomenally
opaque experience would not be a perceptual experience. Note that this compatibil-
ity is not conditional on the arguments above failing. If the arguments go through, it
simply calls for a re-conception of the scale. But what then is the scale, if it is no
longer the introspective degree of attentional availability of earlier processing stag-
es? The answer is in the examples themselves. The Necker cube seems to flip, and
the experience is that of an illusion. The walls seem to start breathing, and the expe-
rience is that of an hallucination. Each of these deviations from perceptual experi-
ence could be given a more extended treatment, but as such they suffice to reveal
what I take to be the key motivation of TT
2
: sensory experience can instantiate vari-
ous points on a phenomenological spectrum; faux perceptual states, if possible,
would be clustered within a zone of that spectrum. Alternatively, the range of senso-
ry experience could be conceived as graded on a scale, at the apex of which sensory
experience seems perceptual, whereas at lower grades it seems like something else.
In any case, the discussion highlights the importance of having a clear conception of
perceptual experience, for it is this conception which dictates ones judgement con-
cerning the extent to which perceptual experiences might be similar to other states.
Moreover, it allows one to better understand what is meant by the claim that the
neutral intentionality thesis concerns only experiences that have the quality of be-
ing perceptual.
5. 4. Presence
I have been treating the concepts applied to perceptual experience largely separate-
ly. But in fact they are interconnected and interanimating. The notion of immedia-
cy is a case in point. It expresses the idea that perceptual experience seems direct
| 143

(and immediate in that sense). But just as often it is claimed that the objects of that
experience seem to be immediately present. The immediacy of the experience af-
fects what the subject believes concerning the object, in a manner that I hope to
have articulated in 5.2. But this leaves the task of explicating the notion of pres-
ence, to which I turn now.
Presence can be taken to have a temporal and a spatial meaning. Speaking of the
former, Metzinger refers equivalently to the de nunc character of experience and the
phenomenological constraint of presence (2003a, p. 128, quotation marks
removed), going on to remark that:
Simple phenomenal content always is temporal content. This is to say
that it always contains temporal information and this information is
depicted as invariant: the state of affairs in question is holding exact-
ly now. (ibid., p. 128)
Similarly, Armstrong proclaims that [p]erception normally gives us up-to-the-
moment information about the physical world (1961, p. 109), and this characterises
the special, 'flavour' of perception (cf., ibid., pp. 111, 112).
115
Others express the same
sentiment by appeal to other metaphysical notions. For instance, J.J. Valberg, in
searching for an appropriate definition of the objects of experience, suggests that
they invariably exhibit experiential presence, and that experiential presence is lim-
ited to the category of particulars. What is present, directly available, is always a
particular (1992, p. 20). He continues the line of thought to say that,
In the tradition, particulars are temporal [...] objects. That is, they are
objects of which it makes sense to ask how long they have existed or
lasted, when they began to exist, and so on [...] Hence [...] an object
present in experience is always a temporal object (ibid., p. 20)
Others focus their discussion on spatial presence. For instance, Alva No discusses
what he calls the problem of perceptual presence (2004, p. 60), which he takes to
be the problem of explaining phenomena such as the following:

115
A somewhat evolved descendent of this idea is Metzingers notion of presentational
content, inspired by Brentanos notion of Prsentation (Metzinger, personal communica-
tion). Presentations are highly stimulus-driven perceptual states, whose content is either too
ephemeral or in the wrong format to inform conceptual thought, but can potentially inform
intentional action (Metzinger, 2003a, pp. 86 - 93). According to Metzinger, presentational
states are those which generate the experience of presence, and furthermore they are trans-
parent states par excellence (ibid., pp. 98 - 99).
144 |
consider, as an example, a perceptual experience such as that you
might enjoy if you were to hold a bottle in your hands with eyes
closed. You have a sense of the presence of the whole bottle, even
though you only make contact with the bottle at isolated points
(ibid., p. 60, note removed)
116

However, the notion of presence is also sometimes contrasted with imagining an
object in its absence. Mohan Matthen, focussing on the visual mode, expresses the
difference between imagining or dreaming a visual scene and experiencing it
through normal visual perception, where in the latter case the scene is not simply
imaged, but seems to present the perceivers own surroundings as so which he re-
fers to as a feeling of presence (2005, p. 305). He discerns the difference as due to
the presence of motion-guiding processes in ordinary visual perception, and their
absence in visual imagination.
117
This, he believes, broadly corresponds to a differ-
ence between experiencing the space that is within a picture and experiencing a
picture as located in space, a subtle difference that is nevertheless discernable:

116
It ought to be noted that No discusses presence in more general terms that encom-
pass any instance in which perceptual experience seems to outstrip occurrently available
sensation. Generally stated as such, this is an issue that runs back to the 17
th
Century at least,
aspects of which we have discussed in depth in Chapters 2 & 3. In recent times, rather dra-
matic demonstrations can be found in recent work on attention and visual memory, for a
discussion of which see No (2001; 2004, pp. 49 - 65).
117
In support of this claim he draws upon work in cognitive psychology, neuropsychol-
ogy and neurophysiology that suggests a functional dissociation between visual processing
in dorsally and ventrally located areas of the brain (Milner & Goodale, 1995). Matthens mo-
tion-guiding vision is enabled by the faster, more ephemeral and more ancient dorsal
stream, repeatedly demonstrated as insensitive to pictorial and descriptive features of a visu-
al scene. For a slightly contrasting discussion of the implications that the dorsal-ventral
model might have for visual experience, see Jacob & Jeannerod (2003, pp. 135 - 176). A simple
account of the model posits a dissociation between perceived size and maximum grip aper-
ture in a grasping action, assumed to be based upon visual information about object size
(see, e.g., Aglioti, DeSouza, & Goodale, 1995). There is some debate over whether data
demonstrating behavioural differences in response to illusions are best explained by this
assumption, as visually informed action might be based on other spatial attributes, such as
the position of points on an objects surface. For a review along these lines see Smeets,
Brenner, de Grave, & Cuijpers (2002).
| 145

many observers claim [...] that objects in a picture look a determinate
distance away. I am inclined to think this response reects a misun-
derstanding [...] a confusion between the picture and the things it
depicts [...] Suppose that a depicted man in a particular picture looks
from various visual cues to be 10 feet away from you. Now, close your
eyes and move three steps closer. Open your eyes. How close does he
now look? No closer than before. (ibid., pp. 316 - 317)
Martin (2002b, p. 402ff.) contains an extended discussion of the differing accounts
of imaginative experience that would be given by a DR theory and an account based
on the claim that perceptual experience is representational. Assuming that, there is
an to visualise an apple is to imagine a visual experience of an apple (ibid., pp. 402
403), the DR theorist (ostensibly Martin himself ) would claim that in imagining a
visual experience, the imaginative project is non-neutral with respect to the pres-
ence of imagined objects in the scene:
Having the visual experience of water puts one in a position which is
not neutral with respect to the actual environment as to whether blue
water is present or not: that is how we have to characterise what our
visual experience is like. Visualising the water puts you in a position
of not being neutral with respect to the imagined situation. In visual-
ising an expanse of water, one is not non-committal whether the im-
agined scene contains a blue expanse of water. Furthermore, visualis-
ing this way can have consequences for what one accepts about the
imagined situation and hence what one comes to believe possible. In
the furniture shop I might visualise a table being turned on its side
and passing through a doorway, and on that basis decide that it is
possible to move such a table into the house without too much effort.
(Martin, 2002, p. 414)
By contrast, Martin claims that a representationalist account of perceptual experi-
ence would need to distinguish perceptual representation from imaginative repre-
sentation, where in the former case the object is represented as present, and in the
latter case the object is represented as absent. Hence, on Martins view, although
considerations concerning perceptual experience might reach an impasse (e.g. be-
cause the transparency phenomenon speaks equally in favour of both views) DRs
stance on imaginative sensory experience, is apparently more adequate to the expe-
rience than a representationalist view.
I have my doubts that the representationalist needs to accept all the assumptions
required to set up Martins critique. But I think one of these assumptions amounts
to a statement of the DR conception of perceptual presence in genuine perception;
in his words: a visual experience of an expanse of water which is the veridical per-
146 |
ception involves the patch of water as a constituent of the experience (ibid., p. 414).
This relates directly to the issue that I brought to the fore at the beginning of the
chapter as the basic concern of DR theorists, namely the preservation of a simple
relation between the mind and the world. According to this line of thought, recall,
experience of objects has an explanatory role to play: it explains our ability to think
demonstratively about perceived objects (Campbell, 2002, p. 114). Fore-fronting the
importance of such a capacity puts certain constraints on how to treat perceptual
experience in a DR framework, and in the current context, perceptual presence.
5. 5. Particularity
As we saw above, presence can have both a spatial and a temporal meaning. Also, as
noted at the end of 5.1., one of the key steps in DR purportedly securing the rela-
tion of mind to world is to overcome the dangers inherent in massive reduplication
of the world, by establishing the theoretical possibility of a description of every
other particular in space and time as uniquely related to our reference point
(Strawson, 1959/2003, p. 22). If one construes the subject herself as a particular in
space and time, this at least lays the ground for demonstrative thought about the
world itself. It is here that the explanatory role of perceptual experience is meant to
come in, as Martin observes: If experience makes reasonable such demonstrative
judgments, surely how things are presented as being must reflect the fact that it is
one particular thing rather than another that one perceives (2002a, p. 182).
Although we have used the notion in a few places, it is worth having in front of us
the basic idea of what particulars are (leaving aside the metaphysical puzzles they
enter into) on this understanding. The following suffices:
Particulars are distinctive in being individuated by their spatio-
temporal locations, so that no two particulars (or, at least, no two
particulars of the same kind) can exist in two wholly distinct places at
the same time. (Lowe, 1995, p. 348)
With this in mind, one can interpret Martins last remark with a fair degree of
strength. What is being suggested is that a perceptual experience stands in a propri-
etary relation to its direct object. In a very broad sense, this is a similar idea to one
held by sense-datum theory. Proponents of that view also thought that perceptual
experiences were proprietary to their direct objects. But the key difference is that
according to DR perceptual experiences are proprietary to their direct objects, and
their direct objects are mind-independent. In the final section of the next chapter I
will develop the idea that perceptual experience can seem to be of mind-
| 147

independent objects. In the course of that chapter I will also develop an account of
how perceptual experience can seem to be of particular objects, one that is amena-
ble to the idea that perceptual experiences are proprietary to their objects.
But first, what do I mean by describing the view as holding that perceptual experi-
ences stand in a proprietary relation to their objects? With respect to thought, I
already provided an application of this idea in 5.1, though it can be made a little
more precise than it was expressed there. The most notable proponents of this view,
namely, Gareth Evans (1982, pp. 30 -31, 40 - 42) and John McDowell (1998, pp. 229 -
232) equivocate between two closely related options. One is the claim that a certain
thought content is dependent upon the existence of the object of the thought, in
that the thought can only be determined as correct or not relative to that particular
object (cf. Burge, 1977). The thinker would suffer an illusion of a thought being
about one particular object rather than another. Another is the stronger claim that
the thought itself would be unavailable to the thinker in the absence of its particular
object, because the thought involves that particular object as a constituent (Martin,
2002a, pp. 176 - 178). The thinker would suffer from an illusion of thinking that
thought rather than another thought, or none at all. We can apply thus mutatis mu-
tandis to perceptual experience, where the second claim serves to fill out the idea of
perceptual experience as a simple relation:
A perceptual experience is object-dependent iff. its content can only be de-
termined as correct or not relative to the particular object it is dependent
upon.

A perceptual experience is object-involving iff. it can only be instantiated in
the presence of the particular object it involves.
Both are metaphysical statements qua phenomenology
M
. Indeed, there is a conspic-
uous absence of a description of particularity in the literature, and by that I mean
simply a description of the experience, rather than an account of phenomenology
M
.
These two are (as I hope to have shown above) often run together. Indeed, a prime
instance can be found in a forthcoming paper by Michelle Montague, in which she
complains that none of the proposals she considers in the literature, capture the
phenomenological particularity fact. For instance, she argues that in their implicit
appeal to a general content shared between phenomenal duplicates, Horgan and
Tienson (see 4.5) arguably miss the particularity of perceptual experience. That
much is clear on all hands. Even in their appeal to wide correctness conditions,
where the content is determined by the particular object itself, this only due to the
brute presence of the object. As Montague sees it, this would only constitute an ex-
planation of the experience of particularity by conflating the fact that a perceptual
experience, when veridical, is of a particular object and the fact that it seems to a
148 |
subject that she is perceiving a particular object (forthcoming, ms p. 12).
118
The wor-
ry here is the (now familiar) possibility of a perfect hallucination which seems like a
perceptual experience of a particular object, scene etc. This misses the more ex-
treme claim broached earlier that the perceptual experience would not be of the
same type or might not even occur in absence of the object. But in any case, by Mon-
tagues lights, that option would not do much better in the task of providing a
wholly phenomenological account of the phenomenological fact; for a wholly
phenomenological account is defined precisely as one that does not give any role
to actual objects (forthcoming, ms p. 18). By stipulating over phenomenology
thus, she runs together phenomenology
D
, with phenomenology
M
. For it is only ac-
cording to a certain account of phenomenology
M
(one which, moreover, ought to be
contested by DR) that a phenomenological account ought not give any role to ac-
tual objects, namely, the kind of narrow (alias internal or internalist) account
that Horgan & Tienson provide.
With this in mind, consider Montagues positive proposal that perceptual experi-
ences involve, as part of their basic structure, a bare demonstrative thought-form
which can be represented as [that (thing) ] (forthcoming, ms p. 16). This is a
thought form, rather than a thought with content, in order to account for the ob-
ject-dependency of the thought.
119
Given that it is a thought, the experience involved
is non-sensory, and given the contrastive name cognitive phenomenology (in the
sense of Dennetts phenomenology
U
). But this cognitive aspect is not meant to ex-
haust the experience. It is meant to describe a structural feature which she dubs
object-positing. In complement to the various sensory aspects of a total experi-
ence, she claims that the thisness of an experience is due to the mind positing a
bare demonstrative that (thing) upon which the experienced properties of the
object are hung (Montague, forthcoming, ms. 16 ff.).

118
Some care needs to be taken as regards to what it supposed to be explained here.
Sturgeon puts a similar point even more harshly, saying that: Disjunctivisms celebrated
explanatory punch springs from ideology [...] pulled from thin air, in that it relies on phe-
nomenal notions in its approach to phenomenology [but] does not explain their origin, ap-
plicability or explanatory role (2000, p. 13). The context of these shortcomings is somewhat
different from present concerns. For as Sturgeon clarifies on the next page, the lesson he
means to draw is the ubiquity of the explanatory gap faced by explanations of how con-
sciousness can arise from material properties and relations (Levine, 1983). At least for the
purposes that I attributed to DR at the beginning of the chapter, such an explanation of
consciousness is not an aim of the theory in the first place.
119
It might alternately be described as having gappy propositional content, with a slot
that is filled by a property instantiation in the veridical case; see, e.g., Tye (2007).
| 149

Montague expresses the theme of her object-positing proposal as broadly Kantian,
and similar ideas have rippled through recent psychology and neuroscience. For
instance, Metzinger & Gallese, giving a gloss on the emerging picture they see in
contemporary neuroscience, say that: An interesting, but frequently overlooked
fact is that the human brain possesses an ontology [...] it makes assumptions about
what the relevant [...] building blocks of external reality are (2003, p. 557).
120
Simi-
larly, a deeply entrenched idea in cognitive psychology is that visual perception re-
lies on a psychological mechanism sensitive to Spelke objects: bounded, coherent,
three-dimensional stimuli that move together.
121
According to Fei Xu (1997) such a
mechanism is responsible for a maximally general sortal concept, physical object,
which both infants and adults use to individuate and identify the objects of thought
and interaction under more specific sortals. Though this would be in certain re-
spects different from the demonstratively specified bare particular that Montague
suggests, its function in this context would largely be the same (Xu, 1997, pp. 368 -
370).
122

If any of these proposals are along the same lines as Montagues, then there is per-
haps one sense in which Montague is not obviously doing metaphysics at all, but
rather positing capacities that explain a certain experience. This is a grey area in
which there are certainly assumptions made about a subjects psychological consti-
tution, viz. the ontology of mind, but no assumptions necessarily need to be made
about what ultimately grounds those capacities, or whether those capacities are pos-
ited as anything more than an instrumental grasp on the mind. Incidentally, I take it
that this is something that the neutral intentionality thesis would allow, for in posit-
ing said capacities, even if e.g. they are the subvenients of other capacities, one is
not thereby taking a stance on the metaphysics of content. Having said that, it is
clear that in defining a phenomenological account in the way she does, the capaci-

120
They go on to discuss the idea that the brain builds a representational ontology of
agents qua individuals acting upon their environment. However, the more general idea of
the brain as an ontology engine is broached in several places in Metzingers (2003a), with the
tentative suggestion that the mastery of natural language might have a feedback effect re-
sulting in a tendency to mentally refer to things (see e.g., pp. 67 68 n40).
121
After Spelke (1990), whom Montague references in a footnote.
122
See especially Xu and Carey (1996) for psychological research to back up Xus pro-
posal. For a critical response see Wiggins (1997). For an alternative, non-conceptual ap-
proach to how and why the visual system preferentially tracks certain stimulus features see
Pylyshyn (2000). For a review of these and other proposals see Casati (2004). And for a more
general state of the art review of the psychological literature that draws explicit links to met-
aphysical discussions of object-hood, see Scholl (2007).
150 |
ties that she posits have strong implications for the metaphysics of content by de-
fault. And the substantial metaphysical claim here is roughly the opposite of that
which an object-involving view would take: thoughtful minds, whose experiences do
not fundamentally involve their mind-independent objects, are somehow responsi-
ble for the mind-independent object-hood that they experience.
As a description of the experience this is at most just a start. Montague certainly
admits that the sensory aspects of the experience are necessary for the particularity
of what is experienced, and her paper is (moreover) work in progress. But as the
account develops, it is hard to guess how it could describe sensory experience of an
object as the presence of a particular. Furthermore, it would be hard to describe the
seeming presence of a mind-independent particular by means of object-positing.
For nothing seems to require that such objects amount to more than the sense-data
Russell thought we were acquainted with (1910-1911, pp. 111 - 112). One of the goals of
the next chapter is to develop an alternative proposal by drawing on the early ideas
of Edmund Husserl.
| 151

Chapter 6 Husserl on the experience
of particularity
6. 0. Introduction
The goal of the last five chapters has been to articulate the direct realist conception
of perceptual experience. This has involved both defence of the conception and res-
olute distinction of the conception from typically accompanying claims concerning
the metaphysics of perception and its intentionality. I now offer synthesis of the
work in these chapters by focussing on the experience of particular objects as mind-
independent, by drawing upon the work of Edmund Husserl.
The account to be presented consists of a series of inter-related claims which to-
gether provide a detailed description of perceptual experience as of particular mind-
independent objects. It takes its initial motivation from the assumption that senso-
ry experience is indeterminate and requires structure to yield perceptual experience.
In the first section (6.1), I introduce this point, noting there that one of Husserls
innovations is to claim that bodily experience provides such structure. The status of
the body in Husserls philosophy is bound up with other controversies that I leave
aside, but I note the controversies before moving on to describe another of Husserls
innovations (in 6.2), namely the claim that the objects of perceptual experience
have a distinctive quality that he terms leibhaftige Gegebenheit. I state the manner
in which I read this expression, illustrating the central idea, before moving on to a
reconstruction of the account itself (in 6.3). After presenting an analysis of the
sixteen inter-related claims that comprise the account, I then (in 6.4) contrast the
account with a view recently developed in the philosophy of perception by Alva No.
Having distinguished the Husserlian account thus, I extend its basic principles to a
description of the experience of objects as mind-independent (in 6.5), by suffi-
ciently weakening concepts of objectivity found in the Oxonian tradition of the
twentieth century and in the Cartesian notions of accessibility and incorrigibility.
6.1 Husserls Body
6.2 Leibhaftigkeit
6.3 The Husserlian account
6.4 Husserlian notes on Noan enactivism
6.5 Weak mind-independence
152 |
6. 1. Husserl s Body
A distinctive feature of the Husserlian account that I am about to sketch, is that it
gives a distinctive role to bodily experience in structuring perceptual experience.
This is important because the account begins from the premise that sensory experi-
ence is in a certain manner indeterminate. By way of illustration, consider a few
lines more of the passage that I cited from James Psychology in chapter 4:
a close attention to the matter shows that there is no proof that the
same bodily sensation is ever got by us twice
What is got twice is the same OBJECT. We hear the same note over
and over again; we see the same quality of green, or smell the same
objective perfume, or experience the same species of pain [] how in-
veterate is our habit of not attending to sensations as subjective facts,
but of simply using them as stepping-stones to pass over to the
recognition of the realities whose presence they reveal [] We take
no heed, as a rule, of the different way in which the same things look
and sound and smell at different distances and under different cir-
cumstances. The sameness of the things is what we are concerned to
ascertain; and any sensations that assure us of that will probably be
considered in a rough way to be the same with each other. (James,
1890/1950, p. 231; also partially cited in Hickerson, 2008. p. 103)
The Husserlian account I describe below can be read as an attempt to deepen James
description. Indeed, I adopt what Hickerson (2008) calls a Jamesian reading of the
texts that I cite. James claim is that perceptual experience involves the experience of
sameness through changes in sensation. But what remains to be further described is
the special relationship between the sensations, and the manner in which this aris-
es. It is only against this background that the claimed structuring of perceptual ex-
perience through bodily experience can be understood. However, I ought to note
that it is not so clear what status the body has in Husserls philosophy more general-
ly.
123
For the sake of clarity, before moving on to the account of particularity, I will
say a little on the dispute.
It is often thought (especially by those working in the philosophy of cognitive sci-
ence) that Husserls work ought to be seen as the culmination of a (faulty) Cartesian
paradigm of disembodied philosophy of mind. That is, although worthy of consid-
erable credit for inventing the method of phenomenological reduction, Husserl

123
For discussion of whether it makes sense to speak of Husserls philosophical thought
as a whole, see Hickerson (2008, pp. 121 124).
| 153

systematically failed to recognise the importance of the body in perception.
124
For
example, Shaun Gallagher provides the following diagnosis: In Husserl's transcen-
dental analysis, the body is reduced to a perceived object and appears to have no
role in the production of perceptual experience, whereas by contrast, Merleau-
Ponty sets out to show that the body itself is doing the perceiving (1995, p. 223, see
also pp. 231 235).
125
However, many manuscripts published posthumously (most
conspicuously, his 1973, 1973/1997) clearly emphasise central roles for the body in
Husserls phenomenology. Given this ever expanding corpus it is an understatement
that there is textual support for more than one reading of Husserl. In the present
context, then, consider two emphases. One emphasis on which [t]he Body [...] is
necessarily involved in all perception (Husserl, 1952, p. 56; 1952/1989, p. 61,
emphasis original). And another emphasis on which the subject has an abiding
sense of self, logically prior to and independent of anything outside my conscious-
ness, including my body, namely my awareness of myself as the pure or transcen-
dental I(Carman, 2000, p. 214). Whether these are both valid readings of Husserls
account of subjectivity in perceptual experience, and whether either is thus sound,
will partly depend upon how one ought to understand the metaphysical implica-
tions of Husserlian intentionality.
126
But more to our present concerns, it will cen-
trally depend upon how one ought to understand Husserls concept of Leib, ren-
dered as Body in a quoted translation above.
The most neutral account of the difference between body and Body here is in a dis-
tinction between the body as an (abstracted) object of attention, scientific investi-
gation, or introspective scrutiny (Krper) and the body as the non-perceptually (i.e.,
merely) experienced subject of perception and action (Leib). This is often rendered
as a distinction between the living and the lived body.
127
But what this distinction
connotes, especially in relation to Husserls various notions of an ego (and particu-
larly in relation to his Ur-Ich), is a matter open for some dispute. Minimally then, I
will take Leib/Body to refer to the experiential subject of perception and action;

124
See Dreyfus & Hall (1982), and Varela, Thompson & Rosch (1991) for prime exam-
ples. But see Thompson (2007, Appendix I) for some second thoughts given the wisdom of
hindsight.
125
Gallagher (personal communication) admits that this opposition was merely for
rhetorical purposes, and not in fact his considered view of Husserl (compare e.g., Gallagher,
2005, pp. 133 - 152).
126
For a review of the range of disputes, see Drummond (2003, pp. 71 - 81, 90 - 91 n.32).
127
See and cf. the translators notes to Husserl (1952/1989, pp. xiv xv) and (1973/1997,
p. xvi). Also see pp. 137 - 138 [161 - 163] of the latter.
154 |
though for many this can simply be thought of as the experience of being a bodily
subject engaged in perceptual activity.
Lastly, I want to unequivocally declare that issues concerning the continuity of Hus-
serls thought with Merleau-Pontys (and even Heideggers) are also not within my
remit. Given the substantial development and popularity of Merleau-Pontys
thought on the body qua perceptual agent, the account presented might bring his
work to mind. But these themes are originally found in Husserl. Accordingly, it is
certainly better to dub the account Husserls (or at least Husserlian) rather than
otherwise.
128

6. 2. Leibhaftigkeit
It will be useful to begin by illustrating the tenor of the claims to come. We can do
that by looking back at chapters past (1 3) where we took great pains to dispose of
accounts of IPE that develop from one of the possibilities that Moore considered for
the nature of sense-data. But the reader will remember, there was at least one other
possibility that Moore was concerned with, namely, that sense-data were parts of
the surfaces of material objects (see 1.5). From this one could derive the object sur-
face view:
A subject indirectly perceptually experiences an object by directly perceptu-
ally experiencing a part of its surface.
In fact, Moore thought that construing sense-data as parts of the surfaces of materi-
al objects was tantamount to a form of direct realism (see Baldwin, 1990, pp. 242 -
246). This initiated a long-running debate (in part, between Moore and himself)
over whether in perceiving the surface of a material object one perceived the object
itself or something non-trivially distinct.
129
But leaving that issue aside, we can see
that if it is meant to indicate indirect perceptual experience, then the object surface
account of perceptual experience certainly does not rule out alternatives. Consider
another take on perspectival variation. Initially, it was suggested that experiencing
something as elliptical is incompatible with experiencing a circular region of a
three-dimensional space. But an alternative is to claim that perceptually experienc-
ing a circular region of three-dimensional space from a particular angle of regard

128
For supplementary analysis and exegesis, see Drummond (1979, 2003), Madary
(2010), Mensch (2001), A. D. Smith (2007, 2008), and Zahavi (1994, 2002, 2003a).
129
See Bermdez (2000, p. 358 ff.) for a recent discussion.
| 155

just is to perceptually experience that very same thing as elliptical. Now apply that
line of thought to experiencing the surface of opaque objects. Perceptually experi-
encing the surface of an opaque three-dimensional object is not incompatible with
perceptually experiencing a three-dimensional object; it is a crucial feature of per-
spectivally experiencing that particular object as three-dimensional. If it were not
for the manner in which we have been led to this point, I think it would seem obvi-
ous to the point of triviality. The claim then is this: The way a particular thing is
experienced to be from one perspective is bound up with how it is experienced to be
beyond that perspective.
130

To illustrate a similar point, Sean Kelly (2004) argues that there is clearly a differ-
ence between the experience of seeing a faade once its illusory nature has been
revealed (see Figure 14), and that of an object that simply looks like the object it
turns out to be. One could understand this as simply a fact of further knowledge
concerning the object in question. But the point of interest is the claim that it just
looks different. In attempting to deal with this (experiential) difference, phenome-
nologists since Husserl have been concerned to provide some account of the way in
which the hidden aspects of an experienced object the backside it is experienced
to have, for instance are present to me in my experience of it (Kelly, 2004, p. 78).
Kelly motivates their toils by saying that, on the experiential level at least,
[w]ithout such an account, we have no resources to distinguish between the case in
which it looks to be a faade and the case in which it looks to be an object (ibid.).

130
Funnily enough, the claim is captured neatly by a sense-datum theorist:
Even the most elementary perceptual act, the barest acceptance, not only is inexhaustive and
indefinite, but (so to speak) knows that it is. The unspecified more is brought before the mind
in the very act itself [] what we take to exist is a spatially complete material thing, fully de-
terminate in all its characteristics, and continuing through time [] what form its spatial
completeness takes, the present act does not tell us. (Price, 1932, p. 179)
Price also remarks earlier that even in the single perceptual act the material thing is pre-
sented bodily, leibhaft [sic.] gegeben, footnoting on that last expression: The phrase is
Professor Edmund Husserls (ibid., 1932, p. 152 n2). This is the only reference to Husserl
amongst what are evidently a stream of Husserlian ideas.
156 |
Figure 14 Two views of a film-set faade

What is interesting about cases such as these is that there are so many ways in
which the visual sensations could be thought of (pace James) as similar, both on an
experiential level and on a more basic organic level (i.e. Berkeleyan sensations of a
VS
1
type). But crucially, the manner in which those sensations are held to yield per-
ceptual experience of particular things differs. As I will show later, by drawing on
the distinction between perceptual and mere sensations the account presented has
resources to describe the difference that makes a difference.
One of Husserls key innovations was a particular characterisation of experiencing
an object perceptually.
131
On his view, this necessarily involves a characteristic that

131
Oftentimes translators of Husserl distinguish between Object as a translation for
Husserls use of the German word Objekt, and the lowercase object as a translation for
Husserls use of the German word Gegenstand. Generally it is thought that, after Cairns who
introduced the convention (see Husserl, 1950/1999, p. 3, n. 2), these denote two different
senses. In particular, some (e.g. Rojcewicz & Schuwer (Husserl, 1952/1989, p. xv), and Ro-
jcewicz (Husserl, 1973/1997, p. xvi)) take it that although all Objects are objects, not all ob-
jects are Objects, precisely because Objects are intersubjectively accessible objects, whereas
objects simpliciter need not be (such as dream or phantasy objects). Occasionally though, it
seems that Husserl does use the two interchangeably (e.g., in his Husserl, 1966/2001).
Though I will often quote from texts where the distinction is employed, it is not central for
my purposes, and I will render the particular objects that I speak of persistently in the lower
case. In all instances it should be clear that I take Husserls points as concerning particular
objects as transcendent in only the sense presented in the main text.
(A) A faade at close-viewing.
(B) A faade in full-view. The train depot from the Gone with the
Wind film-set can be seen in the background behind the girders
supporting the faade itself.
Adapted from:
http://www.retroweb.com/40acres/40acres_flat_facade.jpg
| 157

he considers to be the hallmark of perception, namely [the] Leibhaftigkeit of the
object in perception, that is, the fact that the object always appears ... as bodily pre-
sent" (Pacherie, 1999, p. 148). In what follows, I will take the phenomenon that the
expression leibhaftige Gegenbenheit metaphorically marks to be the experience of
objects as particulars in virtue of each exhibiting an extant unity of appearances.
That is, it is in virtue of exhibiting a unity of appearances that the object stands in
perception as there in the flesh, it stands, to speak still more precisely, as actually
present (Husserl, 1973, p. 14; 1973/1997, p. 12).
132

Elsewhere, Husserl describes this feature of our experience of objects as of them
being transcendant. In this context, I take this to mean the experience of objects as
existing beyond our experience of them. There is a strong (though fairly common)
claim that could be made here: that something only counts as an object of percep-
tion if it in fact has a mind-independent existence. But here is a weaker claim: an
object is only experienced as an object if it is experienced as mind-independent
(Husserl, 1976/1982, pp. 88 - 92 [75 - 78]). In contemporary philosophy, the term
mind-independence functions as synonymous with reality. For the present purpos-
es, the value of an account of perceptual experience which gets us closer to under-
standing how objects can seem to be mind-independent is then obvious. In the next
section a weak notion of mind-independence will be evident in fragments, but will
not be made explicit until the section after. In order to have a label for particular
objects experienced as (weakly) mind-independent, I will use the term particular
object with this connotation.
6. 3. The Husserlian account
To begin we need a substantial definition of particular objects. On my reconstruc-
tion, the proposal is that in order for an object to seem particular there must be an
inherent indeterminacy in perceptual experience. The conceptual apparatus Husserl
provides to characterise this indeterminacy is a distinction between modes of expe-
rience; a narrow or proper (read, determinate) mode and a broad or improper (read,

132
Suffice to say that I take perception in all these quotes to be referring to perceptual
experience. The in virtue of claim that I am ascribing to Husserl here does not commit him
(or the exegesis I will provide forthwith) to a form of indirect perceptual experience. What
would be required for that would be the kind of bizarre iteration of perceptual experience
that we have been considering in previous chapters. In this case, rather, the in virtue of
merely denotes the objective dependency of perceptual experience on certain other factors
(see 1.3). I will often use the word requires to mean the same. An aim of this chapter is to
articulate these requirements.
158 |
indeterminate) mode. The proper mode of a particular object accords to the region
of its surface in view from a given perspective. So, in an instance where one sees a
house, Husserls claim here is that, at a certain perspective, properly we only see the
front side (Husserl, 1973/1997, p. 42 [49], emphasis mine). By contrast, the improp-
er mode of the object accords to surface regions beyond that perspective. I will refer
to these simply as aspects of the object which are yet to be determined. Hence par-
ticular objects are experienced as having aspects that have yet to be determined by
our experience of them. To state this point explicitly:
(a). Definition: Particular objects have determinate and indeterminate aspects.
In distinguishing determinate and indeterminate modes of appearance, if the view
is to claim anything contrary to the Object Surface View, then it must deny that in
the course of experiencing an opaque particular object we experience the presented
side as an autonomous particular object.
(b). Determinate aspects are not autonomous particular objects.
The denial can be motivated by a contrast against experiencing pieces of an object.
A side of an object (as it is to be understood here) is an aspect of the whole: it cannot
be experienced as having an independent existence and yet also be experienced as a
side (cf., Husserl, 1973/1997, pp. 43- 44 [50 - 52]). In order for a side to look like a
side, rather than a potentially autonomous object, it must seem to be part of an ob-
ject that has aspects that are yet to be determined. Determinate and indeterminate
aspects are thus inter-related. Hence, although the account holds that the phenom-
enology is somewhat neutral on the nature of hidden surface aspects, it cannot hold
that it is neutral to the extent that those aspects are taken to not exist. Accordingly:
(c). Hidden aspects are experienced as determinable precisely in so far as pre-
sented sides are experienced as extant sides.
The interrelation between aspects is such that the particular object does not consist
in any single determinate sensory experience, or any conglomeration of such. It is to
be identified with the higher order unity to be revealed through its progressively
determined aspects.
(d). Particular objects are experienced as such in the progressive determination
of their indeterminate aspects.
Together (b) (d) clarify the proposed relationship between determinate and inde-
terminate aspects:
| 159

(e). For every determinate aspect of a particular object there is a range of implied
indeterminate aspects of the same particular object.
The foregoing claims are pretty Husserlian: [A] consciousness of ever new possibili-
ties of appearance constantly persists over against the factual, limited courses of
appearance, transcending them (Husserl, 1966/2001, p. 39 [33]), with the result
that, no delimitation whatsoever lies in the appearances as such; their essence pre-
scribes possibilities of unitary progression (Husserl, 1973/1997, p. 108 [130]).
133
Cru-
cially then, on the account being presented, perceptual experience is holistic, in-
volving a series of inter-animating (that is, mutually implying) perceptual sensa-
tions unfolding in the cycle of determinacy and indeterminacy. Most basically, this
sets a constraint on perceptual sensations. If perceptual sensations are of the same
particular object then they ought to coincide somehow (Husserl, 1973/1997, pp. 77 -
80 [91 - 95]). There ought to be continuity across a multiplicity of possible sensa-
tions. I discuss this further in 7.5. But there is a prior issue that arises from claim
(c). If an object is a particular object, then it ought to seem to have determinable
aspects. How might the experience of determination proceed? Husserls answer is a
series of perceptual steps, where in each step there is both a determinate sensation
and the potential for further determinations (Husserl, 1973/1997, p. 81 [96]).
To even make sense of this requirement there must be some basic change across
minimally distinct sensations. And, at least as regards spatial experience, change in
sensation occurs through movement. Movement of the subject, the object, or both,
enables change in sensation (Husserl, 1973/1997, pp. 132 [155 - 156]). The actual role
of movement (and, in particular, endogenous movement of the subject) in the man-
ifestation of change needs further explication. But it will be useful to immediately
address the following question: what of the situation in which neither the subject
nor the object moves; is this a case in which the object is not experienced as a par-
ticular? There is some metaphysical bite to the question, and in that sense it is or-
thogonal. But running through both a positive and negative answer will help clarify
the description.
A positive answer (i.e. that this would be a case in which there is no perceptual expe-
rience) follows naturally from claims (a) (e). If it were truly possible for there to be
an appearance in a situation which excluded perceptual steps entirely (where one
reads entirely as strong as possible), then this would not be an appearance of a
particular object. For, by definition and by implication, particular objects are just
those objects upon which alternative perspectives seem possible. If a positive answer
sounds strange, this is probably the reason.

133
cf. also Husserl (1952/1989, p. 39 [36]).
160 |
Finding the positive answer strange might seem like a reason to reject one or all of
the earlier claims. But note that, upon further specification, the account proposed is
not so strong as to be without the resources to provide a negative answer (i.e. that
this would not be a case in which there is no particular object experienced) provided
that perceptual steps are not ruled out entirely.

For it is not the actual instantiation
of perceptual steps itself that is required, but rather the structure that pertains to
the seeming possibility of such steps. This structure falls under what translators of
Husserl often render as the inner horizon of the experience (see Husserl, 1966, pp.
6 - 7, 7 - 8; 1966/2001, pp. 43, 44 and passim). In this case the structure is that of
anticipation and fulfilment (Husserl, 1973/1997, pp. 86 - 87 [103]). We can character-
ise perceiving an object as having hidden aspects roughly as follows:
(f). Subjects anticipate that indeterminate aspects will be revealed as perspec-
tive on the particular object varies.
(g). The conditions for anticipations being fulfilled are just those conditions
where there are further perceptual steps involving a series of sensations of
the same particular object.
(h). A series of sensations must be anticipated but need not be manifest for the
subject to experience an object qua particular object.
(f) is a further statement of determinability. (g) will be elaborated upon shortly, for
clearly the bare notion of anticipation is a place-holder for a more substantial de-
scription, viz. a description of what is being anticipated. The general structure of
anticipation and fulfilment (as relevant to the present account) pertains to the dis-
cussion of the coincidence of sensation that will follow shortly. But note presently
that (h) is crucial for the possibility of a negative answer to the question posed earli-
er. That is, one needs only to anticipate that there would be fulfilment if the viewing
conditions were to change. One need not actually go through the perceptual steps in
order to anticipate them.
By construing cases where change is ruled out entirely as limit cases, we can move
on to further characterise the anticipation/fulfilment structure. This will be to ex-
plicate the coinciding determinability mentioned earlier, as continuity.
The potential fulfilment of anticipations requires both a distinction between and
continuity across sensations. The distinction between appearances is precisely the
difference between determinations, viz. the way that an object looks at different
orientations, and/or from different angles. Continuity across appearances pertains
to particular relationships between determinate and indeterminate modes of an
objects appearance, and between indeterminate appearances themselves. So it is a
little more complex.
| 161

At each perceptual step there is necessarily a broad, and contingently a more nar-
row, continuity in the range of potential sensations that correspond to the occurrent
sensory experience of the object (Husserl, 1973/1997, pp. 78 - 83 [92 - 98]). To cap-
ture the difference, Husserl employs a metaphor that serves as useful aid to exposi-
tion. A series of sensations can exhibit a certain harmony in its continuation, pre-
cisely because there is no conflict between perceptual steps (Husserl, 1973/1997, pp.
79 - 80 [93 - 95]). This holds true, for example, when an orb that looks perfectly
spherical from one angle continues to be experienced as such during the unveiling
of previously hidden aspects. There is continuity between distinct sensations: new
sensations on the one hand, and the retention of previous sensations on the other.
Hence, continuity is made possible precisely by:
A gradually receding retention of a previous sensation, and;
A previous anticipation being fulfilled by an occurrent sensation.
Ex hypothesi, () & () yield an overall harmony between appearances (Husserl,
1966/2001, pp. 44 - 46, 49 - 50 [47 - 10, 12 - 13]; 1973/1997, pp. 83 - 87 [99 - 104]).
134

Construe such cases of experiential harmony as definitive of (and thus () & () as
defining) narrow continuity.
By distinguishing between narrow and broad continuity, I want to bring out a pow-
erful feature of Husserls account of particular objects. For the distinction allows the
account to leave the determinability of indeterminate aspects open in such a way
that a non-harmony is also possible despite the identity of the appearing object
(Husserl, 1973/1997, p. 81 [95]). That is, there might be cases in which a conflict does
obtain, where nevertheless a series of sensations are experienced as providing per-
ception of the same object. To elaborate on an earlier example, an orb that looks like
a perfect sphere at first sight might later be revealed to have a gaping cavity. Yet,
despite the cavity the experience is nevertheless of the same particular object, pre-
viously experienced as a perfect sphere. Thus, even if, in the course of sensation a
conflicting determination manifests itself it might still be the case that a particular
object is experienced as a unitary body, into whose unitary form the deformation
fits, precisely as a partial structure into a total structure (Husserl, 1973/1997, p. 82
[97]). In such a case, there is a certain lack of continuity between retained and oc-
current experience, producing a violation of anticipation (or, if you like, a dishar-

134
It is no accident if this brings to mind Husserls famous analysis of time-
consciousness. A crucial difference, however, between the spatial and temporal experience
of objects, is that with the former but not the latter there is the possibility of reversing the
series of presentations (Husserl, 1973/1997, pp. 87 - 88 [104 - 105]). See Drummond (2003,
pp. 75 - 81) for further discussion.
162 |
monious conflict). But, there is nevertheless a sufficient broad continuity such that
the experience is one of simply becoming more familiar with that particular object
(Husserl, 1973/1997, pp. 58 - 59 [70 - 71]).
Already we can begin to make intelligible what Kelly was claiming earlier: the puta-
tive experiential difference which the object surface view failed to capture; viz. the
difference between seeing a faade once its illusory nature had been revealed, and
seeing an object that simply looks like the object it turns out to be (where this is a
more basic difference than a change in ones knowledge of propositions concerning
the object in question). If what has just been said is right, then in both cases the
perceptual experience involves anticipated perceptual steps. In the case of the pres-
ently illusory object the anticipated steps pertain to precisely the particular object
the illuded subject takes it to be. But, in the case of the formerly illusory object the
experience has changed, precisely because the anticipation has changed. The antici-
pation is now of perceptual steps which no longer correspond to the former illusion.
Rather the experience is of hidden aspects that are a radical departure from the as-
pects presented. But the departure is not so radical that there is no broad continuity
in the series of sensations; once the illusory object is revealed as such, this broad
continuity is all that is anticipated.
So narrow continuity is too stringent a constraint on particular experience. For any
two sensations to be of the same particular object they must involve experiences of
the same object as determined, or determinable in the same [broad or narrow] way
(Husserl, 1973/1997, p. 79 [93]). And for that to obtain there must be at least a broad
continuity experienced across change, as perceptual steps progress. Thus we have a
further proposition of the view:
(i). Experience of a particular object in a series of sensations requires experienc-
ing broad continuity across change in sensations.
As broached earlier, (i) requires (at least) the seeming possibility of bringing about
change. For the object must be experienced as having indeterminate (yet determi-
nable) aspects, disclosable through movement of the subject, or the object, or both.
In addition, we said earlier that the limit case would be one in which all movements
were barred, for by definition this would not be a case in which one were experienc-
ing a particular object. A similar case is that in which the only possible source of
change is in the presentation of a mobile particular object. Seeing this as anything
but a limit case would render the view as absurd. To see why, consider what I take to
be the main motivation of such an idea: the fact that the possibility to manifest
change in aspect has so far been taken as the central necessity. Such a bare state-
ment leaves open the question of where change does in fact come from (Husserl,
| 163

1973/1997, p. 76 [89]). But the sense of possibility involved cannot be merely contin-
gent on the potential for objects to move, for that would require that every particu-
lar object be experienced as a mobile object. Or, conversely, it would imply that ob-
jects that are not currently in motion cannot be experienced as particulars, unless
they are somehow experienced as potentially mobile.
I do not want to deny that most interesting objects of perceptual experience can and
do move. And I am not claiming that, in the case of an object that did change its
position and orientation, corresponding changes in sensation would be irrelevant.
Rather, I am claiming that the account cannot require this, even as a disposition of
the object. The focus, then, is on the simplified case of an immobile particular ob-
ject. A virtue of the simplified case is that it applies pressure: if the account is to
require potential for movement, it must find the source of that potential somewhere
else than in the particular object itself. It is here that the eponymous structuring
body takes centre stage. For rather than requiring potential for movement in the
particular object, the account instead requires potential for movement in the bodily
ability of the subject. Hence:
(j). For every series of perceptual sensations, there must be an anticipation or
manifestation of an effecting series of bodily movements.
But as has been repeatedly stressed, the need for change is within sensation. Thus as
a source of change, the potential for movement must also be systematically con-
nected to particular series of sensations. For Husserl, both of these requirements are
captured in the concept of kinaesthesia, the corresponding fields of kinaesthetic
sensations, and their enabling kinaesthetic systems, the operations of which form a
kinaesthetic situation (Husserl, 1973/1997, pp. 131 253 [154 - 293]).
The concept of kinaesthesia (from the greek (movement) and (sen-
sation)), in this context, certainly should not be understood as the sensation of bod-
ily movement per se, as its familiar meaning might suggest. Rather, primarily it
ought to denote the ability to make certain bodily movements (Husserl, 1952/1989,
p. 323 [310]). It is here that the account posits the endogenous source of change in
sensation. In execution, every bodily movement produces kinaesthetic sensations.
But in a similar manner to (h), these need only be anticipated.
(k). For every series of bodily movements, there must be an anticipation or man-
ifestation of a proprietary series of mere kinaesthetic sensations.
Kinaesthetic sensations are systematically coupled to respective series of visual sen-
sations, each of which poises anticipations of further (as yet indeterminate) series of
164 |
visual sensations, also coupled to kinaesthetic sensations (Husserl, 1973/1997, pp.
136, 144 [160 - 161, 170 - 171]).
Kinaesthesia (in general) and kinaesthetic sensations (in particular) are enabled by
various nested kinaesthetic systems that define kinaesthetic fields, the points of
which potentially correspond to particular kinaesthetic sensations. For example,
oculomotor systems enable sensations of eye-movement, and the range of move-
ment enabled by these systems defines a field of potential sensations of eye-
movement (Husserl, 1973/1997, pp. 144 - 147 [170 - 175]). These are dependent upon
superseding cephalomotor systems, the range of movement of which defines a field
of potential sensations of head-movement. These are further dependent on super-
seding systems governing torso rotation, flexion and extension, and finally locomo-
tor systems governing whole-body translation (Husserl, 1973/1997, pp. 173 176 [204
- 209]).
The kinaesthetic systems are thus not necessarily discrete, for instance movements
of the eye can be augmented along the same trajectory by movements of the head.
In a similar regard, they can systematically replace one another in yielding the same
visual sensations (Husserl, 1973/1997, pp. 144, 176 [170 - 171, 208 - 209]). But each
system has a particular range of spatial determination, according to the dimension-
ality enabled by both its own degrees of freedom and those of the kinaesthetic sys-
tems that it nests. Thus for instance the oculomotor field is two-dimensional in only
allowing for movement along vertical and horizontal axes (Husserl, 1973/1997, pp.
283 284 [330 - 331]). Whereas the superseding cephalomotor systems enable free
leaning of the head, by pitching forwards or backwards, or rolling from side to side;
and free approach or retreat, by extension or retraction of the neck; thus enabling a
further range of otherwise unavailable depth determinations (Husserl, 1973/1997,
pp. 284 285 [331 - 332]). These are further elaborated by movements of the torso,
and finally the locomotor systems that enable not only the greatest range of depth
determinations such as expansion, retraction, and covering, but also orbital move-
ments around particular objects.
135
All of which would otherwise be only partially
available through a combination of eye, head, and torso movements (Husserl,
1973/1997, pp. 175 176, 192 205, 209 - 217 [206 - 209, 227 - 243, 246 - 255]).
The holistic convolution of these systems forms a total system, to which pertains a
kinaesthetic situation, setting the perceptual circumstances of a visual sensation
and the initial conditions for the anticipation of ensuing sensations. Drawing on

135
Furthermore, it might be that it is only through these further determinations that
the perceiver is able to disjoin a series of appearances of one object from those of another
(Husserl, 1973/1997, pp. 207 209 [243 - 246]).
| 165

earlier claims we can characterise the nature of this situation, and thus the role of
kinaesthesia in perceptual experience, as follows. Perceiving objects requires poten-
tial for change in sensation. At minimum said changes are possible through bodily
movement. Bodily movement produces kinaesthetic sensations. Kinaesthetic sensa-
tions are systematically coupled to visual sensations. Therefore, anticipating a
change in visual sensation involves implicitly anticipating a series of kinaesthetic
sensations consequent to the movements that bring about said change (Husserl,
1952/1989, pp. 135 - 136 [127 - 129]).
The crucial claims are that kinaesthetic sensations are systematically coupled to
visual sensations and that it is as a result of this coupling that visual sensations are
perceptual sensations. Although the kinaesthetic sensations are not perceptual sen-
sations they enable perceptual sensations to be such. In our earlier terminology they
are mere sensations.
(l). For every series of perceptual sensations, there must be an anticipation or
manifestation of a respective affecting series of mere kinaesthetic sensa-
tions.
By means of mere kinaesthetic sensation, the subject is able track the necessary
continuities in perceptual sensation and thus experience (potentially) revealed as-
pects as parts of an improperly experienced whole.
In fact Husserl puts the point rather strongly:
consciousness of the identity of the form within the continuous
change of its modes of givenness, which we are calling here its as-
pects, essentially presupposes the continuous unfolding, played out
in the back-ground of attention, of the concomitant kinaesthetic
sensation complexes (Husserl, 1952/1989, p. 136 [128])
We can make best sense of this by understanding the role of kinaesthesia as ena-
bling a kinaesthetic cycle of effecting movements and affecting kinaesthetic sensa-
tions:
the running off of the optical and the change of the kinaesthetic do
not occur alongside each other, but rather proceed in the unity of an
intentionality that goes from the optical datum to the kinaesthetic
and through the kinaesthetic leads to the optical, so that every opti-
cal is a terminus ad quem and, at the same time, functions as a termi-
nus a quo (Husserl, Manuscript C 16, IV, 40b, as cited in (and
translated by) Mensch, 2001, p. 42) (see also Husserl, 1952/1989, pp.
62 - 63 [57 - 58])
166 |
Hence:
(m). Definition: a kinaesthetic cycle is a cycle of perceptual sensations and mere
kinaesthetic sensations.
Given this cyclical connection, the account must imply that (perceptual) visual sen-
sation depends upon (mere) kinaesthetic sensation in yielding perceptual experi-
ence.
(n). For every perceptual sensation, there must be an anticipation or manifesta-
tion of a kinaesthetic cycle.
This is to say that insofar as the continuity of perceptual sensation is dependent
upon continuity in change of sensation, continuity of perceptual sensation is de-
pendent upon the continuity of a series of mere kinaesthetic sensation (Husserl,
1952/1989, pp. 70 - 71, 135 - 136 [165 - 166, 127 - 129]; 1973/1997, pp. 149 155, 157 - 159,
316 - 317 [176 - 184, 186 - 189, 356 - 357]). Thus each perceptual sensation indicates
not only a further (broadly continuous) perceptual sensation, but also an affecting
mere kinaesthetic sensation. And this itself indicates both a further (broadly con-
tinuous) mere kinaesthetic sensation and a perceptual sensation, where the latter is
experienced as continuous with the previous perceptual sensation in virtue of the
kinaesthetic cycle.
(o). Broad continuity in a series of perceptual sensations requires broad continu-
ity in mere kinaesthetic sensations throughout a kinaesthetic cycle.
Given (a) (i), (j) (o) yield:
(p). Experience of particular objects requires anticipation of a broadly continu-
ous kinaesthetic cycle.
6. 4. Husserlian notes on Noan enactivism
Claim (p) completes the Husserlian account in its essentials. There are more details
to work out concerning how particular objects can seem mind-independent. They
will occupy us in the next section. There are also more details to be worked out con-
cerning what it is to anticipate a broadly continuous kinaesthetic cycle. That will
occupy us in places in the following chapters. But before moving on, here is as good
a place as any to note that the reader might already find a striking similarity here
with a contemporary view of perception attributed to No, a recent defendent of
direct realism. No argues that perception is a bodily activity, through many an ap-
| 167

peal to phenomenology
D
. In this respect, he is often called an enactivist and his
view is referred to as a species of enactivism. So one might wonder (leaving aside the
fact that Husserl predates No) is this Husserlian account a species of enactivism?
Enactivism was introduced by the eminent biologist Fransisco Varela, his student
Evan Thompson, and the renowned psychologist Eleanor Rosch, to usher in a new
paradigm in cognitive science. Their aim in coining the term was to emphasise the
importance of cycles of activity in a general theory of cognition, and to thus evoke a
view of cognition as the enactment of a world and a mind on the basis of a history
of the variety of actions that a being in the world performs (Varela et al., 1991, p. 9).
Despite the important differences in the original reference (viz. its explicit stress on
history, its aims as a general account of mind) in the consciousness and perception
literature enactivism is most immediately associated with Nos work. Other con-
temporary views are similar enough in their core principles (such as taking sen-
sorimotor engagement to be primitive to mentality in some way) to be called enac-
tivist (e.g. Hurley, 1998), and indeed some proponents self-identify as such
(Gallagher, 2005; Hutto, 2005; Thompson, 2005, 2007). Others who nevertheless
ground perceptual experience in dispositions for bodily movement resist the
(somewhat unnecessary) conflation (Grush, 2007b). Indeed, a common label does
not imply a common thesis, or the inverse. Hence, one could call Husserl an enac-
tivist if one wants, but it would not serve much purpose. Indeed, one could call
Berkeley an enactivist, but it might not be particularly illuminating. In any case, let
me dedicate a few pages to clarifying the similarities and differences between Nos
theory of perception and the Husserlian account of perceptual experience.
A central posit in Nos theory of perception is that of a sensorimotor profile,
sometimes also referred to as a sensorimotor contingency or sensorimotor de-
pendency. I will use these terms interchangeably, as No seems to. If there is a rhe-
torically significant difference between these notions it has certainly not been made
clear by No in any work of his of which I am aware.
136
Indeed, they all seem to have
the same reference: a pattern that unfolds in the changing sensory stimulation re-
sultant from either movement of the sensors or the object, although movement of
the sensors is taken to be more significant. Andy Clark summarises the relevant
ideas with a remarkable clarity that makes it worth quoting him at length:

136
The interested reader should see ORegan and No (2001, p. 940ff.) for the original
statement of sensorimotor contingencies, but it should be noted that my comments are
directed at Nos more recent solo efforts, as represented in the texts that I cite.
168 |
Sensorimotor dependencies are relations between movement or
change and sensory stimulation. Such relations may be of many
kinds, but what they all have in common is that they concern a kind
of loop or cycle linking real-world objects and properties with sys-
tematically changing patterns of sensory stimulation. These chang-
ing patterns of sensory stimulation may be caused by the movements
of the subject (this is the central case), as when we use head and eye
movements to scan a visual scene. Or they may be caused by move-
ments of the object itself or be due to other elements in the environ-
mental frame (e.g., changes in illumination or light source). In addi-
tion, some features of these various kinds of changing patterns will
be due to properties of the objects themselves (e.g., the self-similarity
of a straight horizontal line along its length, giving rise to an un-
changing pattern of retinal stimulation as the eye tracks along the
line). Other features [] will be due to the idiosyncrasies of the
human visual apparatus. For example, the same straight line project-
ed onto the retina distorts dramatically as the eye moves up and
down due to the curvature of the eyeball. (2008, p. 171)
Consider the Husserlian kinaesthetic cycle and Noan sensorimotor contingency as
conceptual tools in isolation from the account in which each is introduced. Even the
casual reader can note significant similarity, such that one might regard the latter as
a (perhaps unwitting) precisification of the former. But note also that this is only in
so far as one considers the central case, where changing patterns of sensory stimu-
lation are brought about by the perceiver. In having application beyond the central
case sensorimotor contingencies are more general than kinaesthetic cycles. This is
one significant difference between the two accounts. But it is inert as things stand.
Another significant difference is in the scope of what is to be potentially explained.
The Husserlian account aims to provide a description of the implicit intentionality
of our perceptual experience of particular objects, and accordingly introduces the
notion of a kinaesthetic cycle. By contrast, No employs the theoretical tool of a
sensorimotor contingency time and again to provide explanations of many and vari-
ous qualities of perceptual experience, such as perceiving the particular object one
does in a specific way with both its specific shape and its specific colour etc. Again,
this difference is largely inert. I say largely, because in the details of Nos account of
the spatial experience of object perception one finds claims of unnecessary strength
and even manifest incoherence. It is here that, to the extent that Nos account is a
mutation of the Husserlian account, Nos account (to extend the metaphor) ought
to be seen as maladaptive. At least, so I shall argue. To see the relevant differences
and to see why they are thus inert, we need to explore Nos account of spatial per-
ception a little further.
| 169

For a No-type enactivist, what perception amounts to is an understanding of the
ways in which we must explore the world to grasp its detail. In the case of spatial
perception, what we understand in perceiving an object as three dimensional is its
visual potential. The visual potential of an object is the way its aspects change as a
result of movement. So to experience an object as three dimensional on the basis of
how it looks is to understand its visual potential. Take an earlier example: we expe-
rience the plate as elliptical from a certain perspective, yet we also experience it as
circular. On Nos account, our experience of it as circular is explained by our under-
standing of how its ellipse changes as we move, or as it moves relative to us. So the
visual potential of an object is just the way in which sensory stimulation varies due
to movement induced change. Unsurprisingly, then, our understanding of visual
potentials reduces to an understanding of the sensorimotor profiles objects will
display: [w]hen we experience something as a cube in perception, we do so because
we recognize that its appearance varies (or would vary) as a result of movement, that
it exhibits a specic sensorimotor prole (No, 2004, pp. 101-102). To see the plate
as elliptical and yet see it as circular is to bring to bear upon it ones knowledge and
understanding of its sensorimotor profile. The sensorimotor relations that distin-
guish said profiles are markedly systematic, indeed mathematically formalisable:
Tomatoes no less than cubes, have determinate visual potentials, or,
as we can think of the visual potential, sensorimotor profiles. As we
get to more complicated forms, such as animal bodies, plants, and so
forth, the mathematics needed to determine the sensorimotor prole
of an object gets more complicated. Our visual perceptual skills,
however, are that sophisticated, encompassing these complex (but
ultimately manageable) relationships (No, 2004, p. 78)
In short, there is nothing ambiguous about a sensorimotor profile, or understanding
a sensory-motor profile. Understanding a sensorimotor profile involves anticipating
ones ongoing experience to unfold in the specific way specified by that sensorimo-
tor profile (ibid., pp. 75 - 122).
This highlights an important difference between the Husserlian account presented
and Nos account: On the latter view, but not the former, perception is constituted
by ones knowledge of something specific vis vis expected patterns of stimulation.
The emphasis on a certain understanding that the perceiver brings to her contact
with the object (which gives perception of the object its content) seems to require
that there be an anticipation of the specific patterns of sensory stimulation that one
already knows how to make the object exhibit.
More schematically: One understands objects of perception insofar as one under-
stands the relevant sensorimotor profiles. One understands sensorimotor profiles
170 |
thanks to ones learned experience of them. And ones learned experience of sen-
sorimotor profiles is due to ones history of exploratory engagement with objects
that exhibit said profiles for only through self-movement can one test and so
learn the relevant patterns of sensorimotor dependence (No, 2004, p. 13, emphasis
original). Therefore, [t]o perceive something, you must understand it, and to un-
derstand it you must, in a way already know it, you must have already made its ac-
quaintance (No, 2009, p. 474).
This is a far more radical thesis than is necessary, on at least two counts. On the first
count it is too strong to provide a negative answer to an appropriately reframed ver-
sion of the earlier question posed to the Husserlian view: what of the situation in
which neither the perceiver nor the object moves, is this a case in which one does
not perceive at all? A positive answer would be forced, but not because of the defini-
tion of what particular objects are qua our experience of them (as potentially ex-
plorable objects); rather, because of a definition of perception in general (as an ex-
ploratory activity).
Perhaps this can be avoided by emphasising the role of what No calls sensorimotor
knowledge. Occurrent exploratory activity is not necessary, as the perceiver simply
anticipates what such exploration would yield, in the sense that the perceiver knows
how to bring about certain patterns of stimulation by means of certain sensorimotor
routines. This brings us to the second count. The obvious difficulty is that there is
no way to account for novel objects as objects of perception. For even if we grant
that one must grasp certain stimulus pattern relationships between circles and ap-
parent ellipses, in order to perceive an apparent ellipse as a circular region of three
dimensional space, it is not then clear how the very same account can provide an
explanation of how we see a novel opaque object consisting of a novel complex of
shapes. For that object cannot be seen all at once, it must by de facto necessity be an
object that has hidden aspects.
No might respond that to claim that the aspects are hidden is a misnomer, for they
are virtually present as readily accessible to the perceiver through exploratory activi-
ty (2004, pp. 59 - 65). But to claim that is not enough. It cannot stand without the
further claim that the perceiver perceives in virtue of its sensorimotor knowledge,
viz. its knowledge of exactly how the patterns of sensory stimulation will unfold. So
Nos final response might be that one does not see an entirely novel object at all
(where, again, one reads entirely as strong as possible), at least not until one learns
and understands (through practice) the requisite patterns of sensorimotor stimula-
tion and thus acquires perceptual content; as merely to be given visual impressions
is not yet to be made to see. To see one must have visual impressions that one un-
derstands (2004, pp. 4 - 5, emphasis original). But again, that would be to claim
| 171

that the objects sensorimotor profile had already been learned. In short, it is to
claim that the object was not novel after all.
137

Clearly then, Nos account is far more radical than Husserls account, indeed per-
haps too radical. It might be thought that this is because Nos account takes it that
our experience of objects hidden aspects and all is determinate through and
through. As he says in a recent paper, speaking of the familiar philosophical tomato:
The nature of our access to the front is of a kind with that of our access to the back
(No, 2009, p. 475). By grasping the determinate visual potential of the object in
virtue of our knowledge of how the sensory stimulation will evolve, the tomato
(front and back) is there for us, present, in reach (ibid., p. 475) in an equally de-
terminate manner in all aspects.
But this rather natural reading actually points to significant tension within the view.
In fact, No repeatedly asserts that all perceptual experience is radically indetermi-
nate, precisely in virtue of being virtual all the way in, elaborating the idea as fol-
lows:
Notice that although the whole facing surface of the tomato is pre-
sent to you, in contrast with the far side, which is out of view, you can
no more embrace the whole of the facing side at once in conscious-
ness than embrace the whole tomato in consciousness all at once.
(No, 2004, p. 134, emphasis original)
And he reiterates the point elsewhere:

137
This claim might be harmless when made in the context of a multi-level account of
perception. That is, when supplemented with an account of how the preconscious matching
of sensory and motor processes enables the sensorimotor understanding putatively consti-
tuting perceptual experience. But that would make Nos frequent appeals to phenomenolo-
gy rather redundant. For prima facie, without an illuminating sub-personal story to tell, the
overall account will be found wanting. Note, this is not a problem for others in favour of a
sensorimotor theory of consciousness, such as Susan Hurley (1998) who explicitly makes a
distinction between circular causality in sub-personal sensorimotor dynamics and the con-
ceptual interdependence of perception upon action in order to build explanatory links be-
tween the two levels. It is not obvious that this is what No has in mind, not only because he
makes frequent appeal to phenomenology, but also because he has a tendency to slip be-
tween personal and sub-personal level explanation without principle (see e.g. 2004, pp. 30 -
31). For a criticism of No in this regard, see Block (2007b).
172 |
Experience in the large, and in the small, is complex and manifold; it
is always an encounter with hidden complexity. Experience is fractal
in this sense [...] In a way, for perception, everything is hidden. Noth-
ing is given. (No, 2009, p. 473)
This, it ought to be said, is in reaction to the view that there are primitive qualitative
states, often known as qualia, to be construed as the simple atoms of experience. By
contrast, Nos phenomenological
D
assertions are broached in the service of
demonstrating that experience is ineliminably holistic in a sense that can be made
tolerably precise. What you are given in experience is always a structured field. You
are never given individual, atom-like qualia. (2004, p. 134).
This last point is (or at least ought to be) well taken. Indeed, Husserl makes much
the same point when characterising the inner horizon of a particular object, claim-
ing that even what is already seen is laden with an anticipatory intention. It what
is already seen is constantly there as a framework prefiguring something new; it is
an x to be determined more closely (Husserl, 1966/2001, p. 43 [7]). But one could
recognise the holistic character of perceptual experience without subscribing to the
proclamation that there is never anything specific about the way in which a particu-
lar object is experienced. And one could do this precisely by distinguishing between
determinate and indeterminate modes of object perception. Indeed, this is what the
Husserlian account does. Essentially, determinacy is a function of a perceivers per-
spective, a point that we will explore further in the next section. For now all we need
to note is that, in embracing the distinction, one can recognise that the determinacy
of the presented aspect could well be more determinate, whilst nonetheless also
recognising that it has a determinacy that is lacking in the openness of the hidden
aspects of the object.
By contrast, as a consequence of not making a consistent distinction, the sensorimo-
tor theory seems almost obfuscating. On the one hand we are told that we know
something (in the sense of knowing how mentioned above) about the object that
we perceive, namely, its determinate visual potential (No, 2004, p. 78), and on the
other we are told that all experience has a complexity that never bottoms out as con-
crete in any way, in that it is unendingly fractal (No, 2004, p. 135; 2009, p. 473).
Sometimes No seems to have not made up his mind on the matter: I said before
that nothing is given. We might say: if anything is given, everything is. If the front of
the tomato is given, then so is that back. (2009, p. 475). As profound as these com-
ments sound, they reveal the basic dilemma the sensorimotor theory must face.
Either it must give up the claim that experience is fractal, in which case the parity
between hidden and presented aspects is false. Or it must give up the view that
knowledge of how to bring about particular patterns of sensory stimulation is con-
| 173

stitutive of perceptual experience, in which case the heart of the view might well be
lost.
Indeed, if both these bullets are bitten, then (at least qua an account of perceptual
experience) the view loses its significance, for it is thus weakened to something vir-
tually indistinguishable from the Husserlian position espoused here. Though, I am
fine with a new dog playing some old tricks. Especially if it helps us see those old
tricks in a new light.
6. 5. Weak mind-independence
In this last section, I want to clarify how various elements of the Husserlian account
indicate a weak-notion of mind-independence that is nevertheless amenable to the
sentiments of DR. It will be useful to draw on certain themes found in literature that
is concerned with epistemological and metaphysical issues. In that literature, mind-
independence tends to imply objective existence viz. dependence upon objective
states of the world. This is in contrast to dependence upon conscious states of a sub-
ject viz. dependence upon subjective states of the mind. So it is best to begin by
looking at some ways of drawing the subjective/objective distinction.
Objective existence is often understood as strongly as possible: as possessing a kind
of extreme context-independence; as existing in a way that is only adequately char-
acterised when one describes things from no point of view, a manner of descrip-
tion that somehow supersedes and encompasses individual points of view. By con-
trast, when one aims to understand the subjective existence of some state, one
must persistently take into account certain facts that vary systematically with the
subjects context. These are often further dependent upon other states of the sub-
ject, and are taken to exist in a way that is only adequately characterised when we
describe things from that subjects point of view (A. W. Moore, 1997, pp. 1 - 18; T.
Nagel, 1989, pp. 13 - 15 and passim; B. Williams, 1978/2005, pp. 48 - 49, 228 - 230).
Unfortunately the notion of a point of view here is rather obscure. The fundamental
distinction is between how the world seems in a subjects thought and experience of
the world, and how the world is without such restrictions, as it is there anyway,
independent of thought and experience (B. Williams, 1978/2005, p. 48). A typical
reason for using the notion of a point of view in drawing this distinction is to exploit
the use of the same term in stating the two sides of the divide, and emphasise that
understanding either side requires understanding both. That is, in order to under-
stand the subjective one must be able to contrast it with the objective, and vice ver-
sa. In the spirit of an Oxonian tradition that begins with Strawson (1959/2003; and
174 |
esp. Strawson's interpretation of Kants Transcendental Deduction in his 1966, pp.
57 - 67), we could take that last sentence as descriptive of the actual mental capaci-
ties that need to be in place in order to ascribe a grasp of objectivity to a particular
subject. On Strawsons view, one cannot possess such capacities unless one also has
a basic grasp on the world as both spatially and temporally extended beyond the
bounds of occurrent experience. For instance, he discusses the idea of a series of
experiences which belong to a unitary consciousness as follows
Not all the members of such a series are in fact self-ascribed: a man
may be more prone to forget himself in contemplation of the world
(or in transactions with it) than he is to be conscious of, or to think
of, himself as perceiving (or doing) what he perceives (or does). But it
is a shining fact about such a series of experiences, whether self-
ascribed or not, that its members collectively build up or yield,
though not all of them contribute to, a picture of a unified objective
world through which the experiences themselves collectively consti-
tute a single, subjective, experiential route [...] conceptualized as to
determine a distinction between the subjective route of his experi-
ences and the objective world through which it is a route (Strawson,
1966, p. 62)
Strawsons worry, it seems, is that without such a hefty conceptualisation of a se-
quence of experiences, ones experience will be all subject and no world. Elsewhere
he describes these conditions as those required for non-solipsistic consciousness,
which he elaborates as
the consciousness of a being who has a use for the distinction be-
tween himself and his states on the one hand, and something not
himself or a state of himself, of which he has experience, on the oth-
er; and by a solipsistic consciousness, the consciousness of a being
who has no use for this distinction. (Strawson, 1959/2003, p. 69)
In order to think about objects as being there anyway, one must be able to think of
them as extant in certain places that one does not currently perceive. And in order
to do that one must have certain capacities for spatial thinking; one must be able to
conceive of a spatial system of objects, through which oneself, another object,
moves, but which extends beyond the limits of one's observation at any moment, or,
more generally, is never fully revealed to observation at any moment (Strawson,
1959/2003, p. 74). This would be a spatial system that was importantly not exclusive-
ly defined in terms of relations to the subject. Rather in thinking of objects as locat-
ed in such a system, the subject ought to be able to entertain, and potentially ex-
change, several distinct items simultaneously occupying distinct places defined by
their spatial relations to one another (Evans, 1985, pp. 151 - 152).
| 175

A subject that is able to grasp such a system is able to entertain thoughts about par-
ticular places they occupy by understanding the connectedness of those places. And
this is most basically achieved by simply moving around and understanding the
practical implications of getting from one perceived place to another, i.e. under-
standing that the space of ones perception and action is embedded in a space that
exists beyond it. Understanding these practical implications would require con-
straining ones thought about the perception of places by means of certain temporal
restrictions. Such as, for instance: that sequences of perceptions relate to spatially
adjacent locations; that thoughts about perceiving a location at a time must be con-
strained by thoughts about speed of travel over time (Cassam, 1997, pp. 38 39;
Evans, 1982, pp. 153, 243).
All of this is required according to Strawson and Evans, in order to ascribe an under-
standing of objectivity to a subject. Once a subject has such a grasp on objectivity,
they also have a certain grasp of their own subjectivity. For the subject then has use
for a distinction between himself and his states on the one hand, and something
not himself or a state of himself of which he has experience on the other (Strawson,
1959/2003, p. 69). That is to say, that the conditions for ascribing the capacity to
think about mind-independent objects are also the conditions for ascribing what
Strawson calls non-solipsistic self-consciousness. Thus we might ascribe the subject
the capacity to think about herself as having a point of view upon the world, on the
very same conditions that we ascribe the subject the capacity for thinking about the
world as independent of that point of view.
But all of this is too strong if ones aim is to describe a simple perceptual experience
of a particular object. To put it another way, the objective existence that the Straw-
sonian line on objectivity requires is the object of the subjects theoretical beliefs
about the world that they experience. In order to capture the more immediate sense
in which particular objects are claimed to be experienced as mind-independent I
will draw a distinction between three levels at which an object might seem to be
independent of the subject:
(i). A neutral object experienced as having aspects that are not occurrently ac-
cessed
(ii). A neutral object experienced as having continued mind-independent ex-
istence
(iii). A neutral object experienced as having an objective existence
The weak sense of mind-independence that the Husserlian account describes takes
(i) to be sufficient for (ii). (iii) is putatively available to a subject whose capacities for
176 |
spatio-temporal thinking meet the Oxonian conditions outlined above. In order for
the Husserlian account to be coherent with what has been said in previous chapters,
it cannot make an appeal to the third level in order to account for the first two lev-
els.
We can get a handle on how this can be achieved by considering another way of
drawing the subjective/objective divide. On a view attributable to Descartes, mental
states exhibit two distinctive features that distinguish them as subjective. Firstly, if a
subjects mind instantiates a state with a certain content, then they will believe that
it does. That is, all mental states are evident to the subject.
138
Secondly, if a subject
believes that their mind instantiates a state with a certain content, then their mind
does in fact instantiate that state. that is, the subject is incorrigible with regard to its
mental states (B. Williams, 1978/2005, pp. 64 68). Both of these involve the im-
plicit assumption that the subject has access to its mental states, and the first can be
understood as claiming that the subject has guaranteed access to its mental states.
Now, without drawing any verdict on whether or not a subjects access to its mental
states is in fact evident (or guaranteed) and/or incorrigible, one can illustrate the
subjective/objective divide by stating the following: It is not the case that a subjects
access to (putatively non-mental) states of the world exhibits either evidence or
incorrigibility. That is to say, it is not the case that if the world instantiates a certain
state, then the subject will believe that it does; and it is not the case that if a subject
believes that the world instantiates a certain state, then the world does in fact in-
stantiate that state. On this way of drawing the subjective/objective distinction,
objective states are states that exhibit neither evidence nor incorrigibility.
Now to apply this analysis to drawing the link between (i) and (ii). States of particu-
lar objects are experienced as states that are not evident, in the sense that hidden
aspects of objects are experienced as states that are not evident. It is not the case
that if a hidden aspect the object instantiates a state with a given content, then the
subject will experience it as such. This gives further substance to the notion of inde-
terminacy in perception, and the sense in which the content of hidden aspects is left

138
Witness e.g. that Descartes defines thought to include everything that is within us in
such a way that we are immediately aware of it. Thus all the operations of the will, the intel-
lect, the imagination and the senses are thoughts (1642/1984, p. 113). This is often taken to
be crucial to a Cartesian theory of self-knowledge. In fact it is not obvious that Descartes
considered view is that simply having a thought is sufficient for knowing what thought one
is having, viz. that self-knowledge is evident (see Broughton, 2008, pp. 187 192). A weaker
form of the idea of mental states as evident holds that all mental states must in principle be
accessible to their subject, in the sense of being brought into the focus of conscious atten-
tion.
| 177

open on the Husserlian account. Consider again the case of perceiving a faade for
the first time. The claim is that it is not the case that when one experiences the hid-
den aspects of the faade as being determinable in a certain way, that the hidden
aspects of the faade are in fact thus determinable. Consider again, also, the notions
of broad and narrow continuity as a means of delineating the extent to which the
determination of hidden aspects is left open. These can be further understood as
delineating the extent to which a subject is corrigible in experiencing the hidden
aspects of an object. The experience of a subject is always corrigible concerning any
narrow continuity exhibited in a series of sensation, for particular objects of certain
kinds (faades being a perspicuous case) might well violate the anticipation of nar-
row continuity. However, if the object is experienced as a particular object, then the
only circumstances under which the experience of a subject is corrigible with re-
spect to broad continuity are circumstances under which the subject no longer ex-
periences the same object.
By employing a Cartesian distinction between mind and world, we provided a nega-
tive contrast between the presented and hidden sides of a three-dimensional object.
However, the account also trades on a notion of perspective that is also usefully clar-
ified in reference to the Oxonian account. On that account, a subject can only be
accredited with a perspective if they can have a thought about a place p in a frame of
reference somehow centred upon the subject, where p is also conceived as being
within a space whose frame of reference is not centred upon the subject. In order to
have such a thought, the subject must be able to think objectively about space,
they must know what is for = p to be true where is a stand-in for an arbitrary
fundamental [...] holistic, identification of a place (Evans, 1982, p. 162). As I said
earlier, the Husserlian account cannot appeal to such capacities, so it needs a much
weaker notion of perspective. In any case, it has no need for the grand scale of
thought well beyond the particular object experienced. Where the two accounts
make contact, though, is in the idea of a frame of reference centred upon the sub-
ject. Consider for instance, the following quotation from Husserl:
each Ego has its own domain of perceptual things and necessarily
perceives the things from a certain orientation. The things appear
and do so from a certain side, and in is this mode of perceiving is in-
cluded irrevocably a here and its basic directions. All spatial being
necessarily appears in such a way that it appears either nearer or far-
ther, above or below to the right or to the left (Husserl, 1952/1989, pp.
165 - 166 [158])
Husserl uses various spatial terms: here, nearer, farther, above, below, to the
right, to the left. One might add terms such as there, in front, behind etc. In
contemporary literature, these terms are often referred to as egocentric terms.
178 |
What the Husserlian account describes are certain stable limitations in the relation-
ship between the subject and the object of experience. This relationship can be de-
scribed in egocentric terms, where the determinate aspects of an object are experi-
enced as here and the indeterminate aspects as there. The Oxonian tradition re-
quires that if one is to attribute to subject an understanding of what it is for both
here and there to be in the same fundamental frame of reference, one must attribute
to that subject a sophisticated conception of an objective world. By contrast, the
Husserlian account merely wants to claim that all that is thingly-real in the sur-
rounding world of the Ego has its relation to the Body (Husserl, 1952/1989, p. 61
[56], italics removed).
It might seem that an obvious candidate for this rather minimal grasp on spatial
relations is provided in the bare notion of an egocentric frame of reference: a refer-
ence frame structured in reference to a privileged point of origin. This latter is at
least supposed to be Husserls Body, Der Leib, the relatum that all particular objects
of experience stand in relation to. I express some hesitation because a set of spatial
relations structured around point of origin does not require any reference to a
body as such a point of origin (Cassam, 1997, pp. 45 - 46). To make this relation
more than coincidental one needs a sense of a point of origin that makes an intrinsic
reference to bodies. For instance, one could understand the notion of an egocentric
frame of reference as a special case of an object-centred frame of reference, viz. a
body-centred frame of reference. Using certain described axes of the body one could
set up a system of spatial relations according to which one might locate objects in
reference to the body as a privileged locus (Campbell, 1994, pp. 9- 10; Peacocke,
1992, pp. 61 - 63).
Take a human body, for instance. It can be described as exhibiting certain axes. One
could cast a horizontal plane extending perpendicular to its length, splitting it into
arbitrary superior and inferior regions. A frontal plane might extend vertically along
its length, dividing it into anterior and posterior regions. A sagittal plane could also
extend vertically, dividing the body into left and right regions. These spatial planes
would enable us to give substantive definition to the axes of orientation up/down,
left/right, forward/backward. As each of these three planes lie orthogonal to each
other one can rotate the body around a point at which they cross somewhere in
the torso and still keep these axes of orientation constant. That is, the spatial
planes form the fundamental axes for the egocentric space, in the sense that the
spatial relations they characterise cannot be reduced to more fundamental axes of
another frame of reference. So even if for instance the body in question is inverted
relative to an extrinsic frame of reference (such as that defining the gravitational
field) there is still an intrinsic frame of reference, defined by the axes we have de-
scribed, according to which there is still a determinate up/down, left/right, for-
| 179

ward/backward (thus it is upside down so to speak). Having set up these bodily
axes and a bodily origin we are then able to locate objects relative to that origin.
Though this might seem very neat, it might not seem so any more when one at-
tempts to describe the spatiality of perceptual experience enjoyed by means of any
of the special senses. The problem is that the point of origin is in an odd place, and
the same place regardless of the sense involved. For on the present definition the
point of origin is described as lying somewhere in the torso, rather than e.g. just
between and behind the eyes. One could suggest an ad hoc amendment, such as
having the very same spatial planes run through a point in the head. But an obvious
difficulty with this is just the fact that it is an ad hoc solution. Such a solution might
be admissible if vision were the only sense modality through which a subject could
conceivably enjoy spatial experience, but then that would rule out e.g. tactile expe-
rience. This might motivate yet another ad hoc definition of an egocentric reference
frame, perhaps centred upon the hand. But this only leaves us with the problem of
how the account would describe a subject as simultaneously touched in another
part of her body, let alone when she simultaneously sees and touches the same ob-
ject.
In fact, in the specific context of the Husserlian account even more problems stem
from the idea that egocentric spatial relations ought to be defined in reference to a
point of origin. For the central claim is that a subject experiences particular objects
in virtue of the poised activity of her kinaesthetic systems affecting and effecting
perceptual sensations, enabling anticipations of continuity and unity. For each kin-
aesthetic system a proprietary frame of reference could be described, centred upon
the eye, or the head, the torso etc. At an information processing level, assuming
such frames of reference is almost unavoidable, and the co-ordination of said frames
of reference is an important problem with interesting implications. But this does
not manifest itself as such at the experiential level. Consider the following example:
there is an object directly in front of a subjects torso; her head is turned to the right
and her eyes turned to the left. On the basis of various postulated body-centred
frames of reference one could say both that the object is experienced as in front of
her and to her left. Thus the object occupies two separately specifiable egocentric
locations at once. Again, this might be right insofar as, at some level of description,
multiple frames of reference need to be accounted for. But if the subject experiences
an object as seeming to be in two places at once, then they are simply not experienc-
ing a particular object.
It is possible to resolve such difficulties in a way consistent with the commitments of
the Husserlian account, simply by taking bodily movement to be intrinsic to an ac-
180 |
count of egocentricity. The approach would be to endorse what I will call the behav-
ioural egocentricity claim:
Behavioural egocentricity: there is nothing more to experiencing x as at ego-
centric location e, beyond a subject being disposed to perform potential ori-
entating actions with regard to x
Thus in specifying objects as egocentrically located, egocentric terms such as up, or
down, to the right or to the left, in front, behind, over there etc. derive their mean-
ing from their connection with bodily action (Evans, 1985, p. 384). For Evans this
was first and foremost an informational connection between perceptual input and
behavioural output, such that having the perceptual information at least partly
consists in being disposed to do various things (Evans, 1985, p. 383).
139
Though this
is a powerful analysis on that level of description (see Grush, 1998, 2007b), what we
are after is the experiential significance of this connection, which is simply that the
egocentric spatial relations are structured as a field of potential action (Taylor,
1978, p. 155).
140
This avoids all the problems of the earlier analyses rather simply: by
not identifying the egocentric perceptual field with any particular egocentric frame
of reference; or indeed any frame of reference at all. It brings a new set of constraints
also. For the concerned are channelled through the overall structure and mobility
of the body. The situation of individual parts constrains their movement dynamics.
And these constraints on movement dynamics set the range of .
Now, with all that in mind, cast a critical eye over the following statements:

139
Admittedly, Evans was not always consistent in his unfinished work; witness the various
uses of the term information in his (1982), and see Campbell (2005).
140
Although, as already stated, Evans aim was to establish the necessity of an informational
connection between sensory and motor capacities, Evans (1982, p. 156; 1985, pp. 384 - 385)
quotes Taylor (1978, p. 154) seemingly in order to illustrate the view on phenomenological
grounds. Taylor himself attributes the view to Merleau-ponty (1978, pp. 152 -153). Thus as a
supplement to the Husserlian account, one might consider this way of substantiating the
notion of egocentric space to be rather Merleau-Pontian. To a certain extent that may be
right, however the emphasis placed upon bodily structure is more in the spirit of Samuel
Todes (2001) phenomenology
T
. And, in fact, it goes somewhat further in emphasising not
only the front-back asymmetry of bodily structure, but also the interlocked nature of the
parts.
| 181

The aspect of the moment is necessarily related to the zero-point of
orientation, to the absolute here, and to the concomitant system of
the depth dimension (fore-back), and of the breadth and height di-
mensions (right-left, above-below). (Husserl, 1952/1989, p. 135[127 -
128])
The sense-things are thereby constituted in the subjective way of
orientation and are constituted for us ... such that a distinctive
sense-thing, the Body, is given as the constant bearer of the center
of orientation. (Husserl, 1952/1989, p. 70 [65])
It ought not to be Husserls considered view that one takes talk of a zero-point or
an absolute here too literally. Similarly, I advise rejection of the claim that this
centre of orientation is something that is actually borne by, or located within the
bodily subject. I have explored reasonable options as to what these might mean be-
yond vaguely gesturing at the body; none of them seem to work.
So in what sense do both the here of the presented side and the there of the hid-
den side stand in a spatial relationship to the body? And how do these relate to the
idea that subjects anticipate a broadly continuous cycle of kinaesthetic sensations?
Employing a useful piece of terminology adapted from Evans, I will call the space of
egocentric relations a behavioural space, structured by the abilities of the subject
(Evans, 1982, p. 160). The positive proposal is that if one treats the notions of cen-
tering, or anchoring, or any point of origin as superfluous, one will not be throw-
ing out the baby with the bathwater. For discarding any substantive notion of an
egocentric frame of reference does not force one to discard the idea that the egocen-
tricity of perceptual experience is systematic. This is precisely because the structure
of behavioural space is still systematic in a way that admits of egocentric terms:
up/down, left/right, in front/behind are still the natural axes of orientation. But this
is not a consequence of the described axes of the body per se, but rather a conse-
quence of the situation and mobility of the perceptual apparatus that these axes
help us track.
141

In this regard, the body exhibits symmetry along the lateral plane in contrast with
asymmetry along the frontal plane, forming systematic constraints on : e.g. the
situation of the subjects eyes dictate that in order for her perceptual sensation of an

141
In some regards this account is similar to Campbells gloss on what is necessary in
providing what he refers to as a psychologically substantive conception of egocentric space
(1994, pp. 10 - 15). An important difference is that Campbell (somewhat oddly) takes re-
course to actual perception of the body to be necessary in accounting for the relation be-
tween egocentric perception and action. The account presented here need not make such an
appeal.
182 |
object to be broadly continuous as she passes it on her right she will eventually need
to move her head, and then her torso, turn her hips, and eventually her whole body.
Experiencing the object as on the right in any one of these instances consists in be-
ing disposed to in this manner, whether or not one does in fact . But all that
amounts to is the claim that experiencing the object as on the right involves antici-
pating a broadly continuous cycle of kinaesthetic sensations that would occur, were
the subject to .
This also allows for the possibility that there is a genuine experiential difference
between experiencing an object that is directly in front of the torso and directly in
front of the head and eyes, and experiencing an object that is to the right or left of
the torso and yet in front of the head and eyes. Furthermore, it provides a key step
in clarifying the systematic relationship between perceptual sensation and kinaes-
thesia. For anticipations concerning ensuing perceptual sensations, will in part be
anticipations of ensuing arrangements of various kinaesthetic systems, bringing a
new comportment of the body.
| 183

PART III

An age which demands free inquiry, pushed without fear or compromise
to its legitimate conclusions []
(T. L. Peacock, The epicier (1836); as cited in Reed, 1994, p. 297)

184 |
Chapter 7 Structural affordances
7. 0. Introduction
The previous chapter completed the description of a certain conception of percep-
tual experience that is a major aim of this thesis. A distinctive feature of the account
presented in that chapter was the idea that kinaesthesia provides a broadly continu-
ous structure to perceptual sensations. The remaining three chapters, including the
present, aim to provide a theory of the natural basis of kinaesthesia, and clarify the
terms of its evaluation.
I begin (in 7.1) with a review of various developments in neurophysiology and psy-
chology that suggest a general set of mechanisms that might serve in a naturalistic
explanation of kinaesthesia. These mechanisms are studied in the effort of under-
standing the flow of information within and between certain physiological struc-
tures of an organism, information concerning both the movement of parts of the
body relative to other parts, and the body as a whole relative to its environment. The
guiding thought is that in understanding these informational transactions, one
might move towards an understanding of how subjects come to experience bodily
movement, comportment and spatial situation. In later sections, I will build upon
this review, supplementing it (in 7.2) with a detailed analysis of the concept of an
affordance, in which I will introduce the concept of a structural affordance as
providing a way of thinking about an agents relationship to the constraints on its
bodily movement. This will be followed (in 7.3) by a review of ways in which agents
deal with constraints upon bodily movement, in order to show the extent to which
structural affordances have been studied already. Structural affordances will then
serve as an element in a more general theory of the natural basis of kinaesthetic
experience. I will introduce such an account (in 7.4) by clarifying the ambiguities
of my previous work on the spatial content of bodily experience. I then conclude (in
7.5) with an application of the theory as an explanation of the Husserlian descrip-
tions provided at the end of chapter 6, and other forms of kinaesthetic experience
besides.
7.1 Proprietary information
7.2 The concept of a structural affordance
7.3 The empirical reality of structural affordances
7.4 A structural affordance theory of kinaesthetic experience
7.5 Phenomenal grooves & kinaesthetic abilities
| 185

7. 1. Proprietary information
Neurophysiologists have been aware of organic processes that might serve to pro-
vide information about bodily movement for over a century:
The deep tissues underlying the surface sheet [of the body] have re-
ceptors specific to [...] changes going on in the organism itself, par-
ticularly in its muscles and their accessory organs (tendons, joints,
blood vessels, etc.). Since in this field the stimuli to the receptors are
given by the organism itself, their field may be called the proprio-
ceptive field. (Sherrington, 1906, p. 131)
All proprioceptive receptors are mechanoreceptors, in that they produce nerve im-
pulses when deformed mechanically, and in this case particularly by being
stretched. Charles Sherringtons (1894) work on cerebrated animals provided early
evidence for an abundance of sensory receptors embedded within skeletal muscle.
These structures, dubbed muscle spindles are now commonly thought to have
adapted for registering limb movement, through the detection of changes in muscle
length (see Figure 15). Muscle spindles are integrated with the (intrafusal) fibres of
skeletal muscle and surrounded by the (extrafusal) bulk of the muscle itself (see
Figure 15A). So when the muscle is stretched the spindles are stretched also. They
house two types of stretch-sensitive nerve-endings, primary and secondary endings.
In both types of nerve-ending the frequency of impulses increases with stretch, as
one might expect. But in primary endings frequency also significantly increases with
velocity of stretch. Their afferent projections run to the cell body of a ganglion near
the spinal cord, whose projections take several different paths (see Figure 15B D).
186 |

Figure 15 Neurophysiology of muscle receptors

(A) A muscle spindle and a Golgi tendon organ. For the purpose of illus-
tration the muscle spindle is exaggerated in size, as compared to the
surrounding extrafusal muscle fibres.
Adapted from Vander et. al. (2001, p. 338)
2001 McGraw-Hill Companies
(B) (D)
Afferent nerve channels running from stretch receptors (see text for
details), based on information from Latash (2008a, pp. 261 - 262) and
Vander, Sherman, & Luciano (2001, p. 339).
| 187

Some of the muscle spindles paths synapse directly on motor neurons in the spinal
cord controlling the bulk of the very same muscle. The function of these paths is to
retain muscle tension; for in the absence of muscle tension, sensory signals from the
spindles would drop out. However, thanks to a spinomuscular loop, when the mus-
cle contracts, tension upon intrafusal fibres (and thus the spindles) is retained
through innervation by dedicated gamma motor neurons, coactivated with alpha
motor neurons that control the bulk of the muscle (see Figure 15B). Other paths
synapse on motor neurons controlling muscles that act in synchrony, or on inter-
neurons that in turn serve to inhibit the activity of motor neurons controlling an-
tagonistic muscles in sympathy with the activity of the muscle projected from (see
Figure 15C). And finally, still others run up the spinal cord, through the medulla
oblongata and the sensory relay stations in the thalamus, terminating in the parie-
tal cortex (see Figure 15D).
142

Muscle spindles are the largest, fastest and easily the most complex of the classic
proprioceptors. But there are also several other potential sources of information
about joint movement. Golgi tendon organs are embedded at the junction between
skeletal muscle fibres and their tendons, just prior to their insertion into the bone
(see Figure 15A, bottom-right). Their location places them poorly to indicate muscle
length, but they do function as highly efficient force detection mechanisms. Their
nerve endings wrap around the collagen bundles in the tendon, and the consistency
of tendons makes them springy. So as the extrafusal muscle contracts, the tension
within the tendon increases. Increase in tension stretches the embedded nerve end-
ings producing an increased frequency of impulses. Changes in tension effectively
correspond to changes in muscle force, so the embedded receptors produce nerve
impulses indicative of the entire range of muscle force.
143

To make appropriate contact with the material of the foregoing chapter, we need a
way of discussing the relationship between these structures and bodily experience,

142
For good reviews of the function, anatomy and physiology of muscle spindles and
Golgi tendon organs (below) see Vander et al. (2001, pp. 337 - 340) and Latash (2008a, pp. 39
- 47, 258 - 263).
143
Another important contribution comes from Sherringtons surface sheet. Although
this violates the initial contrast that he used to mark out the proprioceptive field, recent
work indicates that skin-stretch contributes to (at least) the sense of finger, arm and leg
position (D. F. Collins, Refshauge, Todd, & Gandevia, 2005). Receptors found in joint cap-
sules were also long thought by many (including Sherrington) to provide salient information
about the whole range of joint angles. However, evidence that they might only respond to
extreme joint angles (Burgess & Clark, 1969) sparked some debate, for a review of which see
Proske, Schaible, & Schmidt (1988).
188 |
or more specifically, the relationship between these structures and the experience of
bodily movement. As an operational marker for that kind of experience, I will focus
on whether these structures determine self-ascriptions of bodily movement.
144
Here
determination means just that the output of the structure is necessary and suffi-
cient for the self-ascriptive behaviour in question. As it turns out, there are good
reasons to think that these sensory structures are neither necessary nor sufficient.
Consider a case in which all of these sources are lost, dubbed by Oliver Sacks as
proprio-blindness.
145
Jonathon Cole describes the early difficulties of Ian Water-
man, soon after he suffered specific loss of the sensory structures we have just de-
scribed (1995, pp. 24 - 34):
He could not feel anything with his arms, his legs or his body. That
was frightening enough, but he had no awareness of their position ei-
ther [...] [H]e could make an arm move. But he had no ability to con-
trol the speed or direction of the movement [...] He realised that was
why he had fallen out of bed [...] he was losing the internal feedback
of limb position (1995, p. 12)
146

These dramatic effects might suggest that such internal feedback is necessary for
the self-ascription of bodily movement. To see this, consider a (much more com-
mon) situation in which all internal feedback from a limb is lost, precisely because
the limb itself has been lost, and yet it is felt to move position. Such cases have been
documented since (at least) the 16
th
century CE.
147
In the 19
th
century they were

144
I realise, however, that the notion of self-ascription brings its own problems, but for
a discussion of those I ask the readers patience until the next chapter.
145
See Sacks preface to Coles book (1995, p. x). Sacks also discusses a very similar case
of his own, Christina: the disembodied lady in his (1986, pp. 47 - 58).
146
At points Coles narrative gives the impression that Ian knew what was missing un-
der such descriptions as loss of internal feedback. However, the tenor of the autobiography
is a tale of Ians difficulty in comprehending and managing his own situation, and especially
in expressing what was wrong to family and friends, who were even more clueless than the
doctors and nurses treating him (Cole, 1995, p. 22 ff.). For more on Ian Watermans deaf-
ferented life see Gallagher (2005, pp. 40 64). For a comparison with an even more extreme
case (G.L.) see Cole and Paillard (1995).
147
The first well-documented description of phantom limb sensation is due to the
progressive 16th century surgeon Ambroise Par. Several major (and minor intellectual)
figures consistently documented the phenomenon in the centuries leading up to Mitchells
article, including Descartes in his letters and (most notably) in his 6th Meditation
(1642/1984, p. 53). For a detailed review of each of these descriptions and biographical de-
tails of their authors, see Finger & Hustwit (2003).
| 189

dubbed phantom limbs by an American surgeon named Silas Weir Mitchell. Re-
porting his experience treating casualties of the American civil war, Mitchell speaks
of amputees being haunted, as it were, by a constant or inconstant fractional phan-
tom of so much of himself as had been lopped away an unseen ghost of the lost
part (Mitchell, 1871, p. 565, as cited in Finger & Hustwit, 2003, p. 683). Phantom
sensations are often painful, and often this seems to be the result of the body part in
question assuming a determinate posture that is highly uncomfortable. For in-
stance, the fingers of a phantom hand may seem to be twisted in anatomically im-
possible positions, or the thumb might seem to be pushed through the palm of the
hand (Bailey & Mrsch, 1941, pp. 38 - 39). Furthermore, both spontaneous and vol-
untarily induced sensations of phantom limb movement are typical. As to the for-
mer, a phantom hand might seem to spontaneously clench, or adopt a new position
(Ramachandran & Hirstein, 1998, p. 1607). Mitchell describes a case in which a vol-
untary movement of the phantom hand had unfortunate effects:
A very gallant fellow, who had lost an arm at Shiloh, was always
acutely conscious of the limb as still present. On one occasion, when
riding, he used the lost hand to grasp the reigns, while with the other
he struck his horse. He paid for his blunder with a fall. (op. cit., p.
566)
In sum, it would be hard to claim that the output of sensory structures embedded in
body parts is necessary for self-ascriptive behaviour concerning the parts in ques-
tion, when the self-ascriptive behaviour can occur in absence of the parts in which
those structures are embedded.
Leaving this matter aside for the moment, consider now a gesture towards sufficien-
cy. Both muscle spindles and Golgi tendon organs can be tricked by artificial vibra-
tion. The former are far more sensitive than the latter and when tricked they can
produce interesting illusions. If the muscle belly is vibrated at 100 Hz with a physio-
therapy vibrator, low amplitude changes in muscle-length drive the stretch sensors
within. A train of nerve impulses relayed through the spinal cord drives a sympa-
thetic response in the antagonist, resulting in an involuntary movement. If this
movement is gently checked until halted, the spindles will continue to send infor-
mation up the spinal-cord, consonant with lengthening of the muscle, despite the
fact that the muscle is not getting any longer. Typically, a subject in this condition,
required to track the angle of the stimulated arm by matching it with the non-
stimulated arm, will continue to extend the latter well after the formers movement
has been checked. Indicating that the angle of the arm that they ascribe to them-
selves is based upon an illusory experience of arm position, itself likely based upon
an influx of false information due to the effects of vibration (Goodwin, McCloskey,
& Matthews, 1972a, 1972b). Moreover, with such finely placed stretch receptors,
190 |
muscle spindles might be thought of as ideal detectors of muscle length. And if a
system could be appraised of the relative lengths of muscles either side of a joint, it
ought to be able to discern its angle, or its movement etc.
However, Mark Latash raises several reasons for blocking the claim that the output
of these structures provides sufficient information about bodily movement to be the
source of the self-ascriptive behaviour in studies such as these (2008a, pp. 258 -
260). Firstly, though they are sensitive to stretch, nerve impulses from primary and
secondary endings will only correlate with extrafusal muscular activity. Say one is
moving an arm (as depicted in Figure 16A), in such a case an increase in muscle ac-
tivity will result in a decrease in muscle length. If the movement is obstructed (as in
Figure 16B) then there will not be the same decrease, but there would still be an
increase in alpha motor neuron signalling (to move the arm) and thus there would
also be an increase in coactivated gamma motorneuron signalling, driving the sen-
sory nerves wrapped about the intrafusal muscles. This could result in a comparable
frequency of nerve impulses running up the spinal cord in two situations of differ-
ing muscle length. In any case, the information provided should concern not only
the length of the muscle itself, but the combined length of the muscle and its ten-
dons. Again, in a situation where a movement is blocked, activation from motor-
neurons will increase the tension of the muscles. But the same increase in tension
will not occur in the tendons. They will retain the same degree of elasticity, and un-
der the force of the movement their length will extend (see Figure 16B).
| 191

Figure 16 The ambiguity of muscle spindle information

So there seem to be several physiological reasons why muscle spindles alone cannot
be sufficient for self-ascription of changes in position. That might merely indicate
that signals from these receptors need to be integrated with signals from other so-
matic sources, such as Golgi tendon organs and impulses from cutaneous receptors
resulting from patterns of skin stretch. However, several converging lines of evi-
dence indicate that central processes might play a more active role than merely in-
tegrating sensory information from the periphery.
To appreciate this one needs to take a detour through the development of a contem-
porarily popular control-theoretic analysis of motor control. And in turn, to appre-
ciate that one needs at least a qualitative grasp on control systems analysis. At bot-
tom, all control systems have a basic common structure. All have at least two com-
ponents: a controller and a plant; the latter being the target/controlled system. The-
se two components, controller and plant, map goal states and control signals in
opposite directions. Thus the controller implements the inverse mapping: it takes
goal states as input and produces control signals as output. The plant implements
the forward mapping: it takes control signals as input and produces goal states. To-
gether they produce the identity mapping of goal states to goal states (Grush, 2004,
pp. 378 - 379). Control in the system is either open-loop, in the case of feed-
forward system, or closed-loop, in the case of a feed-back system. Say the controller
(A) An obstructed movement: muscle fibres contract and afferent
impulse correlates with contraction.
(B) An obstructed movement: muscle fibres contract much less and
the tendon stretches; nevertheless, comparable afferent impulse
to (B).
Adapted from Latash (2008a, p. 259) 2008 Mark Latash
192 |
receives input for a goal state; it sends control signals to the plant, which imple-
ments forward mapping from control signals to the output of a goal state. So far this
is a feed-forward system. For example, when one sets a microwave oven to defrost
for time x, buttons are pressed, the process unfolds for time x and halts when ini-
tially commanded to, regardless of the state of (or any feedback from) the item to be
defrosted.
148
By contrast, feed-back systems form closed-loops, where the signal
flow is bi-directional: the controller receives signals from the output of the plant
and sends control signals according to the relative disparity between the output and
the desired goal state. So feedback from the plant modifies the control signal driv-
ing the plant, forming a closed-loop and ideally keeping everything on track. Not
only is the plant regulated by the behaviour of the controller, but if all is set up
properly the controller is regulated by the behaviour of the plant. A conventional
toilet illustrates how such a system might break down: if the feedback mechanism,
i.e., the toilets ball-cock lever, fails to modify (inhibit) the control signal, the cistern
overflows. Other examples are thermostat regulators, vehicle cruise control systems
etc.
For some time, it was common consensus that most skilled movements were exe-
cuted open-loop; that is, without the ongoing influence of sensory signals. This
seemed especially plausible in cases of often repeated, somewhat ballistic move-
ments, and insights from these quarters were generalised into the concept of a mo-
tor program, a set of muscle commands that are structured before a movement
sequence begins, and that allows the entire sequence to be carried out uninfluenced
by peripheral feedback (Keele, 1968, p. 387)
149
A prima facie source of support for
such a purely feed-forward account of motor control is the fact that incoming pe-
ripheral sensory signals cannot travel fast enough to influence the ongoing stream

148
The example is Clarks, see his (2008, pp. 150 - 151).
149
Keeles (1968) review article evidently captured the zeitgeist, receiving over a 150
citations in the next 15 years, winning it the accolade of being a citation classic from Cur-
rent Contents (no. 37, Sept. 12, 1983, see http://garfield.library.upenn.edu/classics.html).
Desmurget & Grafton (2000, p. 425ff.) provide a nice survey of the development of different
concepts of motor control following Keeles article.
| 193

of motor signals at the right time.
150
Although this fact is problematic for any feed-
back analysis of motor control that might restrict feedback to peripheral signals,
the mere fact that a signal delay exists does not exclusively support a feed-forward
analysis. Furthermore, cases of decreased motor performance after peripheral deaf-
ferentation (such as Ian Waterman above) indicate that the control system must be
dependent upon afferent signals in some fashion.
The framework for a hybrid analysis was inspired by ideas from a slightly different
quarter. Since at least the 19
th
century, perceptual psychologists and sensory physi-
ologists have been developing models of how efferent motor signals and afferent
sensory signals interact.
151
Much of the work has stemmed from Wilhelm Wundts
idea of a sensation of innervation. The concept (and its disrepute in the 1940s) is
captured by Boring as follows:
All through the second half of the nineteenth century there persisted
a confusion between muscular sensations and sensations of innerva-
tion. Most [...] believed that the innervation of efferent tracts in vol-
untary action establishes sensations that arise wholly within the
brain and not by way of afferent fibres from the activated muscles.
These sensations of innervation, forming the sensory conscious basis
of the experience of volition, were relinquished, along with the will,
only reluctantly by the psychology of the twentieth century. In the
nineteenth century it was never certain whether an experience of ef-
fort was a muscular sensation or a sensation of innervation (as cited
in Ross & Bischof, 1981, p. 319)
152

Boring opted for the former option, and many others also took Sherringtons discov-
eries of the sensory properties of muscle spindles (and the fast reflexes that their

150
And this is despite the fact that the axons running from muscle spindles are insulat-
ed such that they propagate the fastest signals of all axons in the body. Their primary sensory
axons carry a signal at roughly 120 metres per second, and secondary sensory axons are
smaller and carry a signal at between 20 to 60 metres per second (Latash, 2008a, p. 42).
However, this only accounts for the speed at which each impulse travels between synaptic
terminals. Furthermore, efferent signalling will implicate similar delays, viz. time-lag be-
tween the central generation of motor signals and their actually resulting in sensory conse-
quences to be fed-back (Grush, 1997, p. 21 n13; Latash, 2008b, pp. 70 - 72).
151
In fact similar ideas run back to the pre-Socratics, right through Descartes time
until they were refuted in their ancient form in the mid-eighteenth century. For a review of
how the concepts and methods evolved over the ages, see Grsser (1986).
152
Ross and Bischof (1981) provide a useful review of the fate of Wundts sensation of
innervation in his own writings.
194 |
afferent paths enable) to be responsible for the experience of endogenously generat-
ed movement. However, in the 1950s, Roger Sperry and Erich von Holst each inde-
pendently revived Wundts earlier theory in accounting for certain visual perceptual
phenomena. One of these amounts to a contrast described between the illusory
displacement of the visual field in man when the eyeball is moved passively as when
tapped with the finger tips and the lack of such displacement when the same
movement of the eye in space is brought about through an active response (1950, p.
488). According to Sperry, what explains the difference is the presence of a corol-
lary discharge from the oculomotor system in the second case, and its lacking in
the first case. Using the term efference copy Von Holst described a closely related
phenomenon. In a situation where an eye is fixed in place and the ocular muscles
are paralysed:
[] every intended but unfulfilled eye movement results in the per-
ception of a quantitative movement of the surroundings in the same
direction. Since here nothing happens on the afferent pathways, this
false perception can only result from the activity, originated the in-
tention of the eye movement [...] the unmatched efference-copy
causes the perception (von Holst, 1954/1996, p. 418)
A simple account of the way in which this common idea has been developed is that
there is a central mechanism that uses a copy of the motor signal to compute the
likely sensory feedback generated by the motor signal (see Figure 17). This mecha-
nism functions as model of the causal process in which the motor behaviour pro-
duces sensory consequences. It is often called a forward model. The motor com-
mand itself is generated by a system that takes sensory measurement as input and
assembles an appropriate series of compensatory motor signals to move the body
from its starting state to a target state. This mechanism functions as a model of the
inverse of the causal process occasioned by the motor behaviour. It is often called an
inverse model.
| 195

Figure 17 Forward and inverse modelling of sensorimotor processes
This forms a powerfully simple analysis of motor control. Return to the dispute be-
tween feed-forward and feed-back accounts. A forward modelling process for a giv-
en movement can be trained by taking a copy of the motor signal as input and com-
paring predicted sensory consequences with actual sensory consequences.
153
Once
adequately trained, it then provides a candidate solution to the problem of signal
latency. In the execution of a movement, certain areas of the brain take a copy of the

153
For a nice introduction to the application of forward modelling analyses to issues on
motor learning, see Wolpert, Ghahramani & Flanagan (2001).
A simplified schema of inverse and forward modelling processes (see text
for explanation). Meant as illustration only: for more detailed hypotheses
about the measurement, estimation and prediction processes that might
adequately inform inverse and forward models, see Wolpert, Ghahramani,
& Jordan (1995) and Grush (2004, pp. 380 - 382).
196 |
motor signal and run through a forward modelling process.
154
A prediction of any
resultant discrepancy between current state and target state can then be generated
and used to adjust the inverse modelling process, and as both processes are hypoth-
esised to occur within the brain, signal latency is far less of an issue. Call this the
forward modelling analysis (FMA) of sensorimotor control.

This kind of analysis is closely associated with the comparator approach to agentive
experience, a view according which the system(s) responsible for agentive experi-
ences are nested within the very mechanisms responsible for motor production
(Bayne & Pacherie, 2007, p. 485).
155
This is typically contrasted with the narrative
approach (2007), which treats agentive experience as a form of rationalising inter-

154
A common consensus is that the cerebellum (which receives afferent proprioceptive
channels from the arms and trunk) is implicated in encoding forward models (D M Wolpert,
Miall, & Kawato, 1998). In their (2000) review, Desmurget and Grafton consider a range of
evidence pointing to various divisions of labour between the anterior parasagittal cortex of
the cerebellum and the intra parietal sulcus of the posterior parietal cortex. Recent work also
seems to be converging on a specific role for the basal ganglia in discerning when appropri-
ate interventions to an ongoing sensorimotor process are required (Tunik, Houk, & Grafton,
2009). Another recent development is the application of similar principles to the observa-
tion of actions (Miall, 2003; D M Wolpert, 2003).
155
For instance a typical study in this line of research, conducted by Shergill et al.
(2003), had subjects sit opposite one another and take turns in applying force to one anoth-
ers finger tips. Although in each case the subjects were required to apply the same amount
of force as they received, the forces escalated turn after turn. In further manipulations it was
demonstrated that the estimation of self-generated forces is roughly 50% weaker than the
estimation of any force of an external origin of the same magnitude. The explanation sug-
gested is that the comparison of the predicted sensory consequences attenuates the incom-
ing sensory signal if that signal is close to what has been predicted. From here it is typical
(though not uncontroversial) to make a simple extension of the ideas of Sperry and von
Holst, and claim that the comparison of predicted and incoming signal actually provides a
means of distinguishing between endogenous and exogenously generated movement. For
discussion of issues surrounding such an inference, see de Vignemont & Fourneret (2004).
For a critical discussion of the comparator approach more generally, see Synofzik, Vosgerau,
& Newen (2008).
| 197

pretation that it is potentially (or perhaps continually) confabulatory.
156
The opposi-
tion is certainly not absolute. For instance, Bayne and Pacherie, provide several rec-
onciliatory suggestions (2007, p. 485ff.). In a similar spirit, Synofzik, Vosgerau, &
Newen (2008), suggest a two-step account in which feelings of agency (produced by
flexible sub-personal weighting of sensorimotor cues) interact with deliberations
and judgements in a context sensitive manner, producing a variety of dimensions of
agentive experience as a result of both bottom-up and top-down influences. Pacher-
ie (2008) proposes a model of equal generality and flexibility, but one which iterates
an FMA structure. On the whole, reconciliatory models are to be favoured in light of
the fact that the term agentive experience can cover a rather broad range of phe-
nomena with several elusive subtleties (cf. Bayne, 2008). In that regard, a point of
issue is whether or not a distinction between the experience of agency and the expe-
rience of ownership is robust (Gallagher, 2000). And this issue might well turn on
whether the experience is dependent upon the voluntary nature of the movement
(whatever that might mean) as contrasted with what it is the agent is trying to do
(Gallagher, 2007), or perhaps more broadly on conceptions of self-consciousness (de
Haan & de Bruin, 2010). Furthermore, if agentive experience is conceived as a form
of self-consciousness (which it usually is) then a further variable is the extent to
which contents attributed to such states are of normative significance, particularly
with regard to the rational concerns of the subject (Bayne & Pacherie, 2007).
Suffice to say that by appealing to an FMA, in an account the experience of bodily
movement, I do not mean to thereby advocate the comparator approach mentioned

156
A key line of support for the narrative approach comes from some classic neuropsy-
chological studies on split-brain patients who have undergone a commisurotomy, a surgical
operation performed on severe epileptics which removes the functional connections be-
tween two hemispheres of the brain. Gazzaniga provides a typical illustration:
For example, give a command to the silent, speechless right hemisphere: Take a walk. Then
see how a subject typically pushes back her chair from the testing table and starts to walk
away. You ask, Why are you doing that? The subject replies, Oh, I need to get a drink.
(1998, p. 133)
Observations like this lead Gazzaniga and colleagues to suggest that the left-hemisphere,
though functionally isolated from the original stimulus information, is disposed to interpret
the actions of the subject such that they form a coherent story (Gazzaniga, 2000; Roser &
Gazzaniga, 2004). And the line of argument has been extended to the claim that the very
experience of being the agent of a particular action involves such an interpretive process,
even in individuals with intact colossal structures. Thus, for instance, Wegner, Sparrow and
Wienerman claim that the feeling that we consciously will our own actions is traceable to
an inference we make from the match between our conscious thoughts and observed action
(2004, p. 839). For critical discussion, see Bayne & Pacherie (2007).
198 |
above. Moreover, the liberty taken in talking about self-ascriptive behaviour, rather
than proprioceptive awareness, or indeed a sense of agency, agentive experience,
agentive self-awareness, etc., is to ensure that issues such as these are not prejudged.
I would be happy to concede that a bare FMA would need to be supplemented in
order to account for these many and various phenomena.
Nevertheless, FMA does provide a useful insight into the sensations of movement
that apparently occur in the experience of phantom body parts. In their (1973),
Melzack & Bromage were able to induce sensations that resemble those of phantom
limb sensations in healthy subjects by anaesthetising the afferent and efferent
nerves of their arms. Limb positions were reported that deviated significantly from
that of the position of their actual arm, as well as spontaneous changes in position.
By contrast, when subjects attempted to voluntarily move their arms, changes in
position occurred only in correlation with retained muscle activity. To further
explore, whether experienced changes in position bear a regular relationship to
attempted yet physiologically suppressed movements, Gandevia, Smith, Crawford,
Proske, & Taylor (2006) induced rapid paralysis of the hand by means of modified
sphygmomanometer cuff and an intravenous injection of lignocaine. To test
estimations of wrist angle, they placed a large protactor on a board obscuring the
subjects hand, which itself was mounted (thumb-side up) in a manipulable
apparatus that allowed the wrist to extend away from and flex towards the body.
Subjects were first asked to indicate the angle of their arm after a movement, then
later to make the same movements, with the same degree of force after paralysis.
Despite complete paralysis of all relevant afferent and efferent nerves, both the
estimates and testimony of the subjects indicated movements of a phantom hand in
the direction of the attempted movement. Moreover, these movements exhibited
patterns of variation consonant with the degree of force applied. This lead the
authors to suggest that the occurrence of attempted movements produced illusions
of position change in the direction of intended movement (Gandevia et al., 2006, p.
709).
Let me gather the ideas together for a moment. I have been arguing that internal
modelling of the sensorimotor process (characterised by FMA) supplements
incoming information from the peripheral sensory structures classically known as
the proprioceptors. At this stage, a little clarification is in order. For the general idea
of proprioception can be understood in many ways:
| 199

First, there are internal proprioceptive systems, namely those chan-
nels of information [] whose source is the body. Second, there is
proprioceptive information, understood to include all the information
available about the body []. Third, there is proprioceptive aware-
ness, where this is taken to be conscious experience of the body []
(1995, p. 14)

The term internal proprioceptive system indicates a demarcation amongst propri-
oceptive systems, a demarcation between internal proprioceptive systems and pro-
prioceptive systems more generally. A distinction worth making here, precisely to
allow for the possibility that proprioceptive information, viz. information available
about the body (moreover, sensory information) might come from through chan-
nels within a variety of systems. This is precisely Gibsons sentiment, when he writes
that:
The eyes, ears, and skin can register the behaviour of the individual
as well as external events. The eyes, for example, register the move-
ments of the head forward, backward or rotary by way of the mo-
tions of ambient light [...] The ears register the sounds of locomotion,
and of vocalization. The skin registers manipulation. (1966, p. 34)
Although these claims are not particularly controversial today, they had some signif-
icance at Gibsons time of writing. The idea that all the special senses can register
the same information (sc. the behaviour of the individual) runs against the grain
of the doctrine of specific nerve energies, that each organ of sense is provided with
a capacity for receiving certain changes to be played upon it, as it were, yet each is
utterly incapable of receiving the impression destined for another organ of sensa-
tion (Charles Bell, 1811, as cited in Norrsell, Finger, & Lajonchere, 1999, p. 457).
157
In
fact, Gibsons point ran against the combination of the specific-nerve energy doc-
trine with the received wisdom from Charles Sherrington, that there are two pri-
mary distributions of [...] receptor organs, each a field in certain respects funda-
mentally different from the other (1906, p. 131). Sherrington continues,

157
The idea, though nascent in Bells work was developed by Johannes Mller and ap-
parently established as orthodoxy by Helmholtz (Hergenhahn, 2009, pp. 235 - 236; Norrsell
et al., 1999, pp. 457 - 458). The doctrine further suggested that subjects are conscious only of
sensations and not the environmental source of stimulation, a thesis that Gibson opposed
for most of his career (see notes 27, 47, & 40).
200 |
The surface field [...] extero-ceptive as it may be called, is rich in
number and variety of receptors to which adaptation has evolved in
it. [The receptors of ] the proprio-ceptive field in contradistinction
from those of the extero-ceptive [...] receive their stimulation by
some action e.g. muscular contraction, which was itself a primary re-
action to excitation of a surface receptor by the environment. (1906,
p. 131)
The contrast that Gibson wants to draw is clear:
all the perceptual systems are propriosensitive as well as exterosensi-
tive [...] information that is specific to the self is picked up as such,
no matter what sensory nerve is delivering impulses to the brain [...]
information about the self is multiple [...] all kinds are picked up
concurrently (1986, p. 115)
Gibsons most famous discovery in this regard was that patterns of retinal stimula-
tion could specify the movement of an observer relative to surfaces in the observed
environment (Gibson, Olum, & Rosenblatt, 1955). The basic idea is that the effects
of the observers motion can be extracted from invariant features of the (superficial-
ly chaotic) optic flow. The optic flow field is a dynamic property of the optic array,
which is itself constituted by the set of optic cones convergent on the point of ob-
servation. Optic cones should not be confused with colour photoreceptors, rather
they are theoretical posits corresponding to (ex hypothesi) real structural features of
the light reflected by textural elements in the incident environment; or as Lee puts
it, light reflected from each element to the point of observation forms a thin optic
cone, with the base of the cone on the element and the apex at the point of observa-
tion (Lee, 1993, p. 44). As the point of observation moves, the optic array exhibits
constant change, resulting in a linear optic flow field, where various optic cones will
systematically expand or contract and texture elements will delete or accrete (see
Figure 18A). To give an example, say you are rushing towards a tree with eyes fo-
cussed upon it. At some point, you stop before immanent impact with the tree
causes you a mischief. According to the above framework, it is the systematically
changing patterns in the optic array itself, as the bases of various optic cones expand
and the trees visual profile begins to loom, that directly inform you of your location
relative to the object (Lee & Kalmus, 1980). To explore this phenomenon and its
relation to postural regulation, Roly Lishman & David Lee (1973) used an apparatus
that would allow two facing walls of a room to lean forward and backward whilst the
floor remained stationary. When fixating a wall that is gradually tilting forwards,
optic flow is produced in peripheral vision that is consistent with the head moving
backwards (see Figure 18). In this condition adults were observed to sway back and
forth in compensation, keeping their alignment with wall in sympathy with its
| 201

movement and infants could be (perhaps rather cruelly) induced to fall over when
the tilting was rapid (Lee & Aronson, 1974).

Figure 18 Gibsonian proprioception

In a homely remark, Gibson suggests that such processes might also determinants
of orientation relative to the ground:
(A) Optic flow relative to a focus of expansion
Reprinted from Gibson (1986, p. 124)
1986 Lawrence Erlbaum, Inc.
(B) The moving room paradigm (see text)
Adapted from Steven M. Bokers website:
http://people.virginia.edu/~smb3u/NASPSPA9506a/node1.html
202 |
An amusement park device called the Haunted Swing used to be
popular. A couple entered what appeared to be an ordinary room and
were seated in a swing hanging from a bar running horizontally
across the room. The room, not the seat, then began to swing on the
shaft from which the seat was suspended. When the room eventually
made a complete revolution, the occupants felt themselves go head
over heels. What a sensation! (1986, p. 186)
Gibsonians are as confident as Gibson that the principles of optic flow will general-
ise across all the special sense-perceptual systems. The confidence is well motivated.
Accounts of visual, auditory and haptic proprioception can all exploit FMA. As we
saw already, FMA itself was exapted from certain hypotheses developed in an ac-
count visual perception. Nothing (in principle, at least) prevents the extension of
the framework to understanding multi-sensorimotor processes.
158

7. 2. The concept of a structural affordance
Another of Gibsons novel contributions to perceptual psychology is the notion of
an affordance, simple examples of which are seeing a ditch as leapable and feeling a
substance as malleable. Gibson took this (largely phenomenological
D
) insight, and
developed the idea that:
An affordance [...] points two ways, to the environment and to the
observer. So does the information to specify an affordance [...] the in-
formation to specify the utilities of the environment is accompanied
by information to specify the observer himself [...] This is only to re-
emphasize that exteroception is accompanied by proprioception
(1986, p. 141)
Here is the first example he gives:

158
For a good example, Mark Wexlers lab is engaged the application of a broadly FMA
approach to Gibsonian depth perception (Wexler, Panerai, Lamouret, & Droulez, 2001;
Wexler & van Boxtel, 2005).
| 203

If a terrestrial surface is nearly horizontal (instead of slanted), nearly
flat (instead of convex or concave), and sufficiently extended (relative
to the size of the animal) and if its substance is rigid (relative to the
weight of the animal), then the surface affords support. (Gibson,
1986, p. 127, emphasis original)
The status of affordances is slightly ambiguous. At first blush, when Gibson says
that the surface affords support, it would suggest that he wishes to pick out certain
properties of the environment. Indeed, when he introduces the term itself he says
that the affordances of the environment are what it offers the animal (ibid., p. 127,
emphasis mine). But then he says this: The verb to afford is found in the dictionary,
but noun affordance is not. I have made it up. I mean by it something that refers to
both the animal and the environment in a way that no existing term does (ibid., p.
127, emphasis mine). How could a term refer to both animal and environment? He
then offers a third option (which I think is the most sensible) saying what he wants
to do refer to is something that implicates their relationship, for he insists that the
term implies the complementarity of the animal and the environment (ibid., p.
127, emphasis mine). So we have a few options for interpreting the term. Indeed,
there are two extremes. At one end, the interpretation is that affordance denotes a
property of the environment, which bears a special and particular relationship to an
animal. At the other end, affordance denotes the (special, particular) relation it-
self.
159
Although this is a significant matter of theoretical choice, I opt for the latter
interpretation; but not entirely without motivation. Treating affordances as rela-

159
There is of course the possibility that affordances could be both. Norman (1988)
aims to provide a distinction between perceived and real affordances. In effect, this is a dis-
tinction between the perceived properties of a thing (viz. properties that bear particular
relations to particular animals) and the actual relation that an animal bears to that thing.
Now, if a proponent of DR (ostensibly Gibson himself ) wants to give a unitary account of
affordances, what is at stake is whether the properties things are perceived to have are really
properties that they do have (and whether one should call those properties affordances). An
alternative possibility would be to dub affordances as the interdependent relationship be-
tween animal and environment. It should be noted that much theoretical literature on af-
fordances assumes that affordances are environmental properties that bear particular rela-
tions to animals, and occasionally this assumption is made without argument (see Chemero,
2009, pp. 135 - 147 for a review). However, others have argued for the view that affordances
are properties of an animal-environment system (Stoffregen, 2003), or relations between the
abilities of animals and features of their environments (Chemero, 2003). As the reader can
surmise, my preference is for the latter.
204 |
tions is the simplest way of capturing the reality of affordances: they are as real as
their relata.
160

Interpretive issues aside, there are several things about affordances that are clear.
The relations are certainly conditional. They only occur within a certain range of
possibility, denoted here by the explicit contrasts used (nearly horizontal vs. slanted
etc.). In this case, and presumably others, multiple conditions must simultaneously
obtain in a particular situation for the instantiation of a particular affordance. If a
terrestrial surface were nearly horizontal over one period of time, but only suffi-
ciently extended at another, and rigid at neither time, then there would be no time
during which the affordance held at that place.
161

Identifying affordances requires identifying relational properties of the environ-
ment that are inter-dependent with relational properties of the animal. And they
must be identified as such; for only this would give us a handle on the complemen-
tary relationship that is putatively at hand in the instantiation of an affordance. But
it certainly does not follow from the fact that a theory of affordances must identify
certain relational properties, that they must be known to the perceiver under that
description. In fact this is just what Gibson wished to avoid. If there is anything that
is clear about what it means to perceive an affordance, it is that it is precisely not
supposed to convey the idea that seeing a surface as affording support involves en-
tertaining (or endorsing) a thought such as:
The surface I see that is nearly horizontal (instead of slanted), nearly flat
(instead of convex or concave), and sufficiently extended (relative to the size
of my body) and its substance is rigid (relative to my body) is in such-and-
such a spatiotemporal relation to me that the relation affords support.

160
An analysis of affordances as properties of the environment is bound to get some
motivation from linguistic practice. Here the objection might run that it is the ditch that is
leapable, not the relation. I would be happy to talk about a relation of leapability if needs be.
But as a term of art, affordance need not be hostage to linguistic practice. Writing style
probably ought to be though, and for this reason I will occasionally use the simple construc-
tion x affords y for a. In all cases though, this can be reconstructed into the rather more
awkward construction, x stands in affordance relation y to a.
161
And the same point holds, of course, for conditions holding in different places at the
same time. It is also clear that the conditions are not necessarily binary phenomena, but are
potentially graded; a surface that is nearly flat may do just as well as a surface that is com-
pletely flat. This gradation may further dictate the extent to which an affordance holds (and
even in part its evaluation as positive or negative). A slippery surface, for instance, might be
some grade between an affordance of support and falling; and this might be an instability to
be exploited by the right morphology (cf. Pfeifer & Bongard, 2007, p. 99).
| 205

If anything is meant to follow from Gibsons claim that affordances seem to be per-
ceived directly because they are perceived directly, it is that the perceiver need not
be appraised of affordance relations under any description of them as such in order
to perceive them (1986, p. 140).
Nevertheless, in order to study affordances one needs to identify the relational
properties of the animal that are supposed to stand in an affordance relation with
the environment. In the quoted case, size and weight (at least) seem to be im-
portant properties. These may be determined in a number of ways, for instance,
relative to conventions of measurement designed to approximate objective physical
units, or relative to a particular contextual relation with no such design. The crucial
difference here is that the contextual relation is symmetric, whereas the conven-
tional relation is asymmetric. The girl being six feet tall depends upon conventional
measurements of feet and inches, but conventional measurements of feet and inch-
es do not depend upon the height of any particular individual. There have been
some elegant studies designed to demonstrate the importance of contextual (rather
than conventional) relations in the instantiation of an affordance. For instance,
Warren and Whang (1987) sought to investigate the extent to which an aperture
affords walking from one side of a partition to the other by measuring the frequency
of shoulder turning in a group of smaller-than-average participants and a group of
larger-than-average participants. Unsurprisingly, a decreased frequency of shoulder
turning was found to be positively correlated with an increase in aperture width. But
when aperture width was determined relative to conventional physical units, the
psychophysical functions of the two groups were fairly dissimilar. It was only when
aperture width was determined relative to the distance between the participants
shoulders that the psychophysical functions of the two groups became comparable
(indeed strikingly similar).
I now want to argue that there is a particular kind of affordance relation pertaining
to the structure of that body, one that can be distinguished from all others. The ini-
tial impetus comes from Jose Luis Bermdezs discussion of a candidate solution to
the problem of how to segment the body into parts:
206 |
Let me now introduce the technical concept of a hinge. The intuitive
idea that I want to capture with this term is the idea of a body part
that allows one to move a further body part. Examples of hinges are
the neck, the jaw socket, the shoulders, the elbows, the wrists, the
knuckles, the nger joints, the leg sockets, the knees, and the ankles.
The distinction between moveable and immoveable body parts, to-
gether with the concept of a hinge, creates the following picture of
how the human body is segmented. A relatively immoveable torso is
linked by hinges to five moveable limbs (the head, two legs, and two
arms), each of which is further segmented by means of further hing-
es. (1998, p. 155)
Bermdez suggests that his talk of hinges provides a nonarbitrary way of segment-
ing the body that accords pretty closely with how we classify body parts in everyday
thought and speech (1998, p. 156). This is not quite right. It may be non-arbitrary,
but it does not exactly capture reference to noses and ears, for example. Still, I want
to extract something positive from the idea by noting another point he makes,
which is that awareness of the location of the hinges, as well as of the possibilities
for movement that they afford, can plausibly be viewed as an inevitable concomitant
of learning to act with ones body (ibid., p. 156). A look around his book would sug-
gest that Bermdez is using the notion of an affordance here in the Gibsonian sense
(cf. Bermdez, 1998, pp. 103 - 129). According to the account of affordances adopted
above, affordances are relations. Here the relationship is between a bodily agent and
its body. Call this kind of an affordance a structural affordance (SA), to repeat:
SA A structural affordance is a relation between a bodily agent and its body.
This in turn enriches the idea of an (ordinary) affordance relation between an agent
and environment. On this analysis, agent-environmental affordances (leapability,
malleability etc.) are second-order relations, prior to which are an agents first-order
SAs. Call agent-environmental affordances AEs for the sake of the distinction. One
can define the relationship between SAs and AEs as follows; reading Q and R as spe-
cific relations, x as an unspecified relation, and b as an individual:
(1). If there are affordances, then there are SAs.
(2). There are affordances that are not SAs.
(3). If x is an affordance in which b as a relatum, then either x is an SA relation R,
or else x is an AE relation Q, and b is a relatum in Q in virtue of being a rela-
tum in R.
The definition of the relationship between SAs and environmental affordances is
similar to a seminal definition of the notion of a basic action (cf. Danto, 1965, p.
| 207

142). Here I will understand basic action to mean any instance of an agent trying to
maintain or change its bodily comportment. This is nevertheless meant to be along
the lines of the typical understanding. A basic action can be in the service of some
further (instrumental) action that an agent is trying to perform, but it need not be.
Furthermore, trying to act requires the following condition: only when actions are
initiated (and usually performed) under the impression that the actual performance
is possible might we say that the agent tries or is trying to perform that action. This
is one side of the philosophical coin paid in a trying-analysis of action, the other
side of which is the metaphysical possibility of the actions total failure.
OShaughnessy provides a description of this latter:
There is a perfectly genuine sense, suppressed by philosophers of a
common-sensical orientation, in which no event, including intended
act-events, can be foretold as an absolute certainty. That sense is this:
the world is known to have harbored freak happenings; this is a per-
manent potential of the world, and of no situation can it be said:
This situation bears a charmed life, it is guaranteed not to harbor
such a freak. (1973, pp. 365 - 366)
For our purposes the point here is not (or at least ought not to be) the mere sugges-
tion of the metaphysical possibility of a radically indeterminate world. Rather the
significant insight is that when we try to perform some basic action, such trying is
always engaged in spite of the possibility that reality may not co-operate. But never-
theless, trying entails the presumption on the agents part that success is at least a
remote possibility (O'Shaughnessy, 1973, p. 367). If this is correct then there ought
to be some way of specifying the constraints on the action in question, and some
way of ascribing sensitivity to such constraints to the agent that licenses genuine
ascription of trying. And there also ought to be a general term that encompasses the
veridical case and the non-veridical case, in which the agent tries under a false im-
pression of those constraints. I will use the word anticipation as such a general
term: one only tries to do what one anticipates to be possible. I will use the term
practical knowledge as a label for the case in which an agents anticipations are ve-
ridical: one only knows what one can do in virtue of anticipating what is possible.
This possibility is far narrower than the general metaphysical possibility that
OShaughnessy entertains. An agents possibilities for action are set by an agents
SAs. Hence, anticipations constitute practical knowledge relative to the SAs in-
volved in the basic action an agent tries to perform.
The ascription of practical knowledge to those that try and perform basic actions
also enriches a canonical theory of the intentionality of action, originally advanced
by John Searle (1983, pp. 8, 85, 94). On Searles view intentionality in general has
208 |
two salient features, it has a world-to-mind direction of fit, and a mind-to-world
direction of causation, each of which he describes as follows:
When I try to make the world be the way I intend it to be, I succeed if
the world comes to be the way I intend it to be (world-to-mind direc-
tion of fit) only if I make it be that way (mind-to-world direction of
causation). (Searle, 1983, p. 96, emphases original)
This match between what the agent tries to do and what actually happens is only
intelligible given the possibility of a mismatch. Part of Searles concern here is the
possibility of causal exclusion (cf., 1983, pp. 89 - 90). The general platitude is that if
an agent is not causally involved in the occurrence of an event, the event is either
not an action of hers or not an action at all. But equally important is to note that, if
the agent is taken to be successful in trying to make the world the way they intend it
to be, a background assumption is the possibility of failure. Once again, what I
mean to indicate here is not merely the metaphysical possibility of a total failure,
the permanent potential of an uncooperative world. Far more significant (in basic
action at least) are the possibilities for an agents failure that are simply beyond the
specific bounds of the basic action she tries to perform. We can only get a grip on
such a failure if we attribute to an agent a false anticipation of their SAs. That is, the
failure in question is a failure to practically know the SAs of her body, where the SAs
in question are specific to the basic actions that an agent tries to perform.
One way in which in the discussion so far has been deficient is that it wholly ne-
glects that fact that the body itself has structure. Indeed, the reader might recall
that the concept was initially discussed as a corollary to an analysis of body par-
thood, namely, Bermdezs hinge analysis. A positive feature of the hinge analysis is
that it forefronts the fact that parts of the body have the relational property of being
interlocked with one another. Furthermore, by employing the idea of a body part
that allows one to move a further body part in the analysis, Bermudez likely intends
to indicate the importance of this property in this regard. The assumption being
that it is in virtue of the fact that its body parts are interlocked with one another
that a bodily agent is able to act. But this cannot be sufficient for the possibility of
basic action. It is not so much the interlocking of parts that is important, but rather
(as a consequence) the manner in which in which the parts causally interact, in vir-
tue of their properties. Moreover, in absence of reasons to think otherwise, one
ought to allow for the possibility of part-whole causal interaction, in addition to the
possibility of part-part interaction fore-fronted by Bermudezs hinge analysis.
Here then is a fuller definition unpacking the notion of a structural affordance fur-
ther:
| 209

SA* A structural affordance is a relation between a bodily agent b and the prop-
erties I
P
and I
W
of parts of its body, in virtue of which a basic action is pos-
sible for b.
o I
P
is the property of being causally interactive with other parts x
1
, x
2
,
x
3
, ..., x
n
in virtue of their properties P
1
, P
2
, P
3
, P
n
.
o I
W
is the property of being causally interactive with the whole body
in virtue of its properties W
1
, W
2
, W
3
, W
n
.
This is closer to being satisfactory.
162
Now the main work is in showing that reality
of structural affordances is plausible. I will do this by means of a potted review of
research in motor coordination.
163
The aim is to make some piecemeal progress by
picking out at least a few candidate Ps and Ws (marking them as (P) and (W) as I go
along). In doing so, I hope to highlight some of the causal processes on which SAs
are dependent. I will also minimally address how something might count as a bodily
agent, and thus how something can count as a basic action of a bodily agent (when I
do that it will be obvious I am doing so). My treatment of these will be somewhat
cursory, but sufficient to indicate the direction I think a treatment of SAs should
take.

162
An alternative (suggested to me by Sascha Fink) would be to describe structural
affordances literally as structures with sets of relations defined upon them. This would be
something like the following:
SA** A structural affordance is a relation between a bodily agent b and the relations I and
J governing the set of bs body parts, in virtue of which a basic action is possible
for b
o I is the relation of part x causally interacting with other parts x
1
, x
2
, x
3
, ..., x
n

o J is the relation of part x causally interacting with the whole body
An interesting line of future research would be to explore the implications of this and other
alternative ways of defining structural affordances.
163
A caveat: Here follows a trail of phenomena employed to illustrate the path to an
empirically real concept of a structural affordance. It should not be treated as a rigorous
argument to demonstrate that it ought to lead there and not somewhere else. It also should
not be treated as a rigorous defence of a dynamical systems theory of motor coordination.
That task has been undertaken by several authors far more qualified to speak to the relevant
issues (Kelso, 1995, 2002; Thelen & Smith, 1994; Turvey, 1990). The science of coordinated
bodily movement (like any science) is a constantly evolving field, full of its own controver-
sies. It is not hard to see that even fairly comprehensive review articles hardly manage to
hide the axe that they are grinding (cf. Turvey, 1990). The devil is ever in the details of theo-
retical interpretation, and partly as a result I have tried to down-play the heavier aspects of
the theories of the researchers whose work I draw upon. Though for some discussion of
those heavier aspects, see 9.5.
210 |
7. 3. The empirical reality of structural affordances
A significant property of multi-jointed agents is motor redundancy (see Figure 19).
Latash summarises a favourite example, which he attributes to the father of the con-
cept, Nikolai Bernstein:
Touch your nose with your right index finger. Now try to move the
arm without losing the contact between the fingertip and the nose.
This is easy to do. This means that one can touch the nose with very
many combinations of arm joint angles. Nevertheless, when the task
was presented, you did it with a particular joint combination.
(Latash, 2008b, p. 35)
164

Figure 19 Redundancy in a four joint limb

His point is that your arm has the (P) property of being highly redundant in a cer-
tain sense. A glance at Figure 19 can help clarify. Assume an idealisation of your arm
as having only four joints (shoulder, elbow, wrist and first knuckle), where each
joint has only one degree of freedom. In order to reach endpoint E, you could as-
sume angles a, b, c & d. But you could also reach E by assuming a, b, c & d. Indeed,
you could also reach E by assuming a, b, c, d etc.
Redundancy thus described, gives rise to a very simplified instance of what is often
known as Bernsteins problem. As Michael Turvey notes, when considered as a (W)
property of the whole body, motor redundancy is all the more complex:

164
Latash makes something of an understatement here. He is well aware that there are
not just very many combinations of arm joint angles that will bring the finger to the nose.
As he notes later, potentially the number of combinations is infinite (Latash, 2008b, p. 36).
See main text.
Adapted from Latash (2008b, p. 36).
| 211

As characteristic expressions of biological systems, coordinations
necessarily involve bringing into proper relation multiple and differ-
ent component parts (e.g. 10
14
cellular units in 10
3
varieties), defined
over multiple scales of space and time. The challenge of properly re-
lating many different components is readily illustrated [...] There are
about 792 muscles in the human body that combine to bring about
energetic changes at the skeletal joints. Suppose we conceptualize the
human body as an aggregate of just hinge joints like the elbow. It
would then comprise about 100 mechanical degrees of freedom each
characterizable by two states, position and velocity, to yield a state
space of, minimally, 200 dimensions. (1990, p. 938)
The term Turvey uses at the end, state space, denotes a concept from the dynamical
analysis of physical systems. It refers to an abstraction of all possible states that a
system can assume, where each point in the space marks a potential state of the
system being modelled. Suffice to say that a 200-dimensional space of possibilities
(which itself is a simplification) is of mind-boggling complexity. Biological move-
ments regularly trace a path through the high-dimensional spaces that Bernsteins
problem reveals. But interestingly, they regularly exploit the variety of options; they
trace different paths to reach the same goal. And this is the case even in stereotyped
movements performed by highly trained individuals. As Latash conveys, Bernstein
discovered this himself in a careful study of labour workers (Latash, 2008b, p. 31).
He attached small light-bulbs at key points on the bodies of blacksmiths, as well as
on their familiar hammer. Then he photographed their movements using an innova-
tive high-speed shutter, whilst they performed a typical hammer-strike. What the
photos revealed was that across strikes there was much less variability in the move-
ment of the tool than in the various individual joints moving it. This suggests a sim-
ple and smart solution to Bernsteins problem: treat the system as if it had fewer
degrees of freedom by lumping components together. J. A. Scott Kelso illustrates the
idea nicely:
During a movement, the internal degrees of freedom are not con-
trolled directly but are constrained to relate among themselves in a
relatively fixed and autonomous fashion. Imagine driving a car or a
truck that had a separate steering mechanism for each wheel instead
of a single steering mechanism for all the wheels. Tough to say the
least! Joining the components into a collective unit, however, allows
the collective to be controlled as if it had fewer degrees of freedom
than make up its parts, thus greatly simplifying control. (1995, p. 38)
This solution motivates a fairly autonomous conception of components of a motor
task, where elements of a system are not controlled individually, like segments of a
marionettes body by attached strings, but united into task-specific [...] structural
212 |
units (Latash, 2008b, p. 53, emphasis original).
165
According to this gloss, task-
specific structural unity is a (P) property of the blacksmiths arms. Parts of the
blacksmiths arm work in cooperation, stabilising and compensating for one anoth-
ers variable behaviour to consistently reach the target.
Structural unity can also be seen as a (W) property of the whole body in an activity
like locomotion. To give an intuitive illustration, consider how one might approach
the design of systems capable of human-like walking. Humans, like all three-
dimensional objects, have a minimum of three degrees of freedom: a body can pitch
forward and back, roll side-to-side, and turn (yaw) left or right. If treated as struc-
tural unit then it might be possible to keep the number of degrees of freedom in a
locomotive body fairly close to three. Pitching and rolling are less technically known
as leaning. Too much leaning can lead to falling over, so leaning whilst walking
ought to somehow be stabilised. The problem can be conceived by imagining the
walker as an inverted pendulum on either the sagittal or frontal plane. The aim is to
keep the ball of the pendulum roughly perpendicular to a flat walking surface (ide-
alized as having no incline or decline). What is required is to achieve a sustained
natural pendular motion by instantiating the (W) property of stably coupled com-
ponents.

One way of doing this would be to make sure that pressure exerted upon
the walking surface is kept within a safe area inside the foot edges, by making sure
that the foot of the stance leg remains flat on the floor whilst the swing leg is
brought forward (Wisse, 2005, pp. 113- 114).
166
But this is harder than it sounds. All

165
Structural units are sometimes also referred to as synergies, and under that guise
their study dates back to Bernsteins work. In particular, Israel Gelfand (a mathematician)
and Michael Tsetlin (a physicist), both of whom worked closely with Bernstein (see Latash,
2008b, pp. 34 - 35) developed several principles to demarcate the essence of structural
units/synergies. A detailed inspection of these is not required for present purposes, but it is
worth at least providing the reader with Latashs summary
Axiom 1. The internal structure of a structural unit is always more complex than its inter-
action with the environment (which may include other structural units).
Axiom 2. Part of a structural unit cannot itself be a structural unit with respect to the same
group of tasks.
Axiom 3. Elements of structural unit that do not work with respect to a task [either]
Axiom 3a are eliminated and a new structural unit is formed or
Axiom 3b find their own places within the task. (2008b, pp. 58 - 59)
166
As the expression flat-footed connotes, this is not in fact how most humans ought
to place their feet. As Wisse & van Frankenhuyzen (2006, p. 144) point out, the shape of typi-
cal human feet actually channels the centre of pressure forwards in the progression of the
step cycle. This fact has been reflected and exploited in the design of prosthetic feet, and
passive dynamic walkers (see next page and next note) since McGeers early prototypes.
| 213

potentially destabilising elements would require continuous monitoring and con-
trol, which adds considerably to the basic three-dimensions.
Enter the wild-walker.
167
The Delft Biorobotics lab have developed a simple con-
struction that they call Mike, to demonstrate that the task specified above can be
more simply executed by coupling components to one another such that they collec-
tively produce a self-stabilising behaviour.
168
Mike has two symmetrical pairs of legs,
each leg in a pair is fused to the other, one pair moves outside the other, and each
pair has a knee joint (see Figure 20A). To facilitate movement on a flat (rather than a
declining sloped) surface, Mike has oscillatory pneumatic actuators either side of
his outer hip joints, so called McKibben muscles. There is also a muscle at every knee
joint extending the leg, counteracted by a spring on the other side (see Figure 20B).
Muscle activity is regulated by manually tuned timing and a simple switching
mechanism in the inner and outer feet that antagonistically couples the hip mus-
cles. So, for example, when the inner foot mechanism switches, the outer knee mus-

167
Otherwise known as a passive-dynamic walker, or sometimes referred to as a limit-
cycle walker as ideally it would implement self-sustaining oscillations which would resist
deviations from a limit point trajectory (D. G E Hobbelen & Wisse, 2007). See Collins, Wisse
& Ruina (2001, p. 608) for brief description of the ancestors of modern-day designs: clever
tinker-toy ramp walkers that were purely passive, but were not designed to mimic natural
human gait and so exhibited rather comical waddling motions. Tad Mcgeer (1990), likens
the design approach to that taken in the development of early flying machines. Wilbur &
Orville Wright (and their predecessors) initially mastered the dynamics and control of free
gliding before adding a fairly unsophisticated motor (Pershey, 2005). Similarly, research in
wild/passive-dynamic walking primarily involves the design, simulation and test of walking
machines with minimal (if any) motor resources, making passive-dynamic ramp walkers
[] the gliders of walking robots (S. Collins et al., 2001, p. 609). Rather than using endoge-
nous power to control, maintain, stabilise and co-ordinate locomotion, an appropriately
designed passive dynamic walker is set in stable motion by a deft push. Gravitational forces
and the inherent dynamics of physically coupled components with human-like geometry do
the rest. Obviously this requires a sloped surface in purely passive setups, but flat surface
locomotion is possible with the addition of very simple oscillatory motors to an appropriate
structural context for a limit cycle (S. Collins, Ruina, Tedrake, & Wisse, 2005; S. Collins et al.,
2001; D.G.E. Hobbelen & Wisse, 2008; Iida, Gmez, & Pfeifer, 2005; Iida & Pfeifer, 2004).
168
Two other robots were built upon similar design principles. One capable of walking
on level floors, called Latch Walker, so-called because its energy input is regulated by a
wound spring and a latch at the hip joint alternating power transfer. The other only walks a
sloped declining surface, and is named Museon Walker after the science museum in the
Hague, possibly a consequence of its persistent selection as a demonstration of passive dy-
namic walking at the Delft Biorobotics lab. (Mike is probably called Mike because of his
muscles). Videos are available at http://dbl.tudelft.nl.
214 |
cles are deactivated, the muscles at front of the outer hip are activated. The knee
muscle is reactivated just under half a second later, and as the outer foot strikes, the
mechanism switches the power from the front hip muscles to those at the back. As a
result, when set to walking Mike exhibits a robust gait, with a fairly fast leg motion
of even length (see Wisse, 2005, pp. 116 - 122; Wisse & Frankenhuyzen, 2006). In fact
this is the key, as Mikes intrinsic dynamics essentially implement the following two
principles: You will never fall forward if you put your swing leg fast enough in front
of your stance leg. In order to prevent falling backward the next step, the swing leg
shouldnt be too far in front (Wisse, 2005, p. 122, italics removed).

Figure 20 Humanoid walker prototypes at the Delft Technical University Biorobotics lab

Mikes structure presents a solution to a tendency for unstable pitching that keeps
degrees of freedom low. In fact, he is built to have only three degrees of freedom,
one at his hip, and one for each pair of knees. As a result Mike only moves in the
sagittal and horizontal plane. He looks like someone marching from the side. But
from the front he looks like someone shuffling on crutches, for the lateral stability
inherent in his inner-outer leg design foregoes the problem of lateral leaning. A
(A) Mike, a quadruped two-dimensional walker with actuated hips and knees.
(B) McKibben muscles situated on Mikes hips and knees (see text).
(C) Denise, a biped two-dimensional walker with unactuated knees, a passive upper body,
counterswinging arms and an ankle joint modelled on a skate-board truck.
(D) Video-stills of Denise in motion (video available at http://dbl.tudelft.nl).
All pictures from Wisse (2005), reprinted with permission from the author.
| 215

slightly more humanoid solution to this would be lateral foot-placement; but a po-
tential drawback is that reducing lateral leaning in this way could lead to an in-
crease in pitching instability. Martijn Wisse suggests a compromise:
Similar to skateboards and bicycles, one could use steering (yaw) to
stabilize lean, at least as long as the system is moving forward with
sufficient velocity. The same principle is applicable to walking, and
can be implemented in walking robots with an ankle joint that kine-
matically couples lean to yaw. (2005, p. 125)
Working on these principles the Delft team built Denise, a three-dimensional walk-
er. Denise only has two feet, she also has a passive upper body, two counter-
swinging arms rigidly coupled to hip angle, and a (skateboard-like) lateral leaning
ankle joint, resulting in five degrees of freedom (see Figure 20C). Her ankles and
knees are entirely passive, and her hip joint reciprocally couples her two legs to one
another. Her hip motion is controlled in a similar manner to Mikes, but with less
actuation. At each step, a switch in the striking foot activates the contralateral hip
muscle, and releases a latch at the knee joint, allowing the leg to bend and then
swing back to extension. Because of the ankle joint, each of Denises steps is slightly
laterally displaced. But each step drives her forward to another step: when the right
foot strikes, the left knee is released, and the hip muscles start pulling the left leg
forward, Denises weight shifts forwards, the leg extends and the foot lands, repeat-
ing the process for the contralateral leg. Consequently, she swaggers along stably at
around 2.5 km/h (see Figure 20D).
To the extent that Mike and Denise provide models of actual walking, they provide
nice examples of whole body structural unities that could be constituent properties
of an SA. From my phrasing the reader might guess that there is a caveat in the of-
fing. It would be overly permissive to call these delicately balanced machines agents.
Not because they are machines, but because they do not exhibit enough of the right
216 |
kind of task-specific flexibility.
169
My aim is not to impose any particularly firm leg-
islation here. Agency is a graded phenomenon, full of subtleties that deserve their
own extended treatment. But it will be useful to at least note that Mike & Denise are
at the wrong end of the scale. Something I glossed over above was the fact that alt-
hough Mike is a two-dimensional walker, he does have a tendency to gradually steer,
and this can lead him to happily crash into a wall or onto an ill-affording surface. I
also glossed over the fact that Denise falls on her front in roughly one in twenty
steps, usually due to an irregularity in the surface. The solutions to these problems
are fairly simple (though doubtless not so simple to implement). What Mike needs
are some surface and obstacle sensors, and some way of changing his yaw angle.
Denise simply needs to swing her leg forward faster when her upper body tips to far
over her centre of gravity. The point here is that they do not exhibit this kind of flex-
ibility, not that they could not.
It ought to be clear also that I am not ruling against fairly automated processes be-
ing instances of agency. Indeed, I am happy to allow that reflexes of various kinds
are actions, provided that they exhibit enough task-specific flexibility. Let me illus-
trate something closer to the agency end of the scale by recounting the wonders of
the decapitated frog. In the 18
th
and 19
th
centuries, German scientists made signifi-
cant advances in vertebrate physiology by studying what these creatures could do
with only their spinal and peripheral nervous system. In a brief survey of this work,
Latash describes Eduard Friedrich Wilhelm Pflger (1829 1910) hanging a headless
frog on a frame and applying a small patch of paper soaked in weak acid solution to

169
Cf. Metzinger (2003a, pp. 40 - 41, n20) for hints at a similar deflationary gesture. I
fully intend the criterion that I advance to be significantly weaker than criteria usually ad-
vanced as necessary for a voluntary action. Hence, by stipulation in order to perform a
basic action, an agent need not be a person, in e.g. Harry Frankfurts (1971) sense of the term.
An agent need not be a person in Strawsons (1959/2003, pp. 94 - 98) or in Dennetts
(1969/1983, pp. 90 - 98) sense either. However, given my sympathies with a trying analysis of
basic action, I do require that a basic action performed by an agent be an instance of trying.
This has the consequence that both anticipation and practical knowledge can be sub-
personally attributed; though they need not be necessarily so attributed, they can also be
states of a person if the agent meets the relevant conditions for personhood. For instance, if
one of the sufficient conditions for personhood is the attribution of conscious states to a
material object (cf. Strawson above), and the anticipations are of e.g. certain continuities in
perceptual sensation, then this is would be an instance in which anticipations (and practical
knowledge in the veridical case) are attributed at the personal level. However, if a system
does not meet requisite conditions for personhood, and/or the anticipations are of e.g. a
predicted match between sensory input and motor output, then this would be an instance in
which anticipations (and practical knowledge in the veridical case) are attributed at a sub-
personal level.
| 217

a spot on its back. Soon after application, the ipsilateral hindlimb produced a
smooth wiping motion, immediately removing the offending substance (see Figure
21A). Pflger then chopped that leg off, reapplied the solvent, and observed that the
other leg did the same but after longer delay (Latash, 2008b, p. 23).

Figure 21 Wiping reflex in the spinalised frog

In more recent years, work on spinalised frogs (surgically incised just above the me-
dulla oblongata) has revealed the wiping reflex to have considerable flexibility.
Berkinblit, Feldman, & Fukson (1986) applied various perturbations to the hindlimb
to explore whether and how it would adapt. In one study they restricted rotation of
the knee-joint to about 5, and although usually the joint would move through
about 50, the very first wipe induced was successful. After that they placed a cast on
its ankle, again a successful wipe, first time. They then applied a bracelet to the
segment between the two most distal joints, equal to the weight of the entire limb
itself, and the wiping reflex was still successful (see Figure 21B). Incidentally, this
provides a nice example of how the very redundancy that Bernsteins problem high-
lights provides an essential source of versatility (Kelso, 1995, p. 87). But more to
the present point, observing these movements must have exhibited a striking telos.
(A) Stages of the wiping reflex, adapted from Berkinblit, Feldman, &
Fukson (1986).
(B) Two perturbations of the wiping reflex studied by Berkinblit et
al. (1986), adapted from Latash (2008b, p. 45).
218 |
Apparently Pflger was so astonished at the perfect nature of this reaction that he
supposed the existence of a spinal soul (Fukson, Berkinblit, & Feldman, 1980, p.
1261). I would not go so far. But I would go so far as to say that these frogs are far
further towards the agent end of the scale than Mike and Denise.
Despite this relatively straightforward (if permissive) criterion for agency, the pres-
ence or absence of an affordance for basic action can be rather labile. An interesting
case of this is found in the development of infant locomotion. Most newborn in-
fants exhibit alternate stepping movements if their upper body is supported, though
these seem to disappear after a couple of months (Zelazo, Zelazo, & Kolb, 1972).
Beyond this, Esther Thelen and colleagues made several further observations. First
is that these movements are not random thrashings of the legs, but rather orga-
nized movements with a recognisable structure in time and space (Thelen & Smith,
1994, p. 11). Second is that these babies actually exhibit increased movement of their
legs in the putative non-stepping period: they become rather fond of kicking their
feet in the air as they lie on their back. These are rather striking when combined
with a third: that these supine kicking movements exhibit the same kinematic pro-
file (viz. structure in time and space) and a similar pattern of muscular activity to
their precocial stepping movements (Thelen & Fisher, 1982). From these and other
data points, Thelen & Smith conclude that what had previously been considered as
distinct and separate behaviors are in fact manifestations of the same motor out-
put performed in two different postures (op. cit., p. 11).
So what had changed? Why do these kids stop stepping when they can perform the
movement involved perfectly well? A further line of Thelens research indicates that
the affordance fades as they lose the (P) property of a delicately balanced force-to-
mass ratio. Newborns get fatter faster than their muscles can handle, they are simply
not strong enough to keep their (perfectly healthy) chub in check. With this in
mind Thelen, Fisher, & Ridley-Johnson (1984) compared stepping frequency with
the rate of change in newborn infants body mass over the first month of life. They
found that infants who had gained the most stepped less than infants who were
gaining weight more slowly, suggesting that it was the rate at which body mass in-
creased rather than simply body mass itself that caused the stepping reflex to recede
(ibid., p. 485). To further probe the ways in which muscle strength is an index of the
bounds of movement, they investigated the extent to which stepping could be in-
fluenced by adding and relieving weight. And unsurprisingly, after being loaded
with the equivalent of an extra two weeks worth of body mass, infant stepping de-
creased significantly in frequency and height, whilst submerging them in a bath of
warm water (see Figure 22) produced the opposite effect (ibid., p. 489).
| 219

Figure 22 Stepping reflex in the human infant

However, if this is to be attributed as an action, it needs to exhibit the requisite task-
specific flexibility. Some indication of this again comes from Thelens work. She
discovered that 7-month-old infants (i.e. infants who might otherwise be thought to
be at a non-stepping age), if provided with upper body support, produced pro-
nounced and smoothly coordinated stepping movements as soon as they were
placed on a treadmill (Thelen, 1986). Soon after (inspired by studies on quadruped
gait) Thelen, Ulrich, & Niles (1987) placed infants of the same age on a split-belt
treadmill (see Figure 23A & B), and observed smooth adjustments of swing and
stance when the belts were run at different speeds. This work has been followed-up
more recently in an elegant study by Yang, Lamont, & Pang (2005). From their sam-
ple of 5 -12-month old infants they report that the majority could not only rapidly
adapt to belts running at 2:1 ratios, but could also produce coordinated stepping on
belts running in opposite directions (see Figure 23E)!
(A) 3-month old tested for stepping with feet on table; infrared light-
emitting diodes are visible on the hip, knee, ankle and toe joints.
(B) Stepping reflex induced in the same 3-month old when submerged in
warm water.
Both from Thelen & Smith (1994, p. 12).
220 |

Figure 23 Treadmill stepping patterns in pre-walking infants
I persist in describing this as an instance of stepping (qua basic action), rather than
walking, in order to be sensitive to certain reservations concerning the latter. For
instance, in a review of the relation between prenatal movements and post-natal
development Heinz Prechtl makes the following remarks:
Not all movements observed in the fetus belong to the category of
anticipation of post-natal functions. The human fetus frequently
changes its position in the uterus. There are specific movement pat-
terns to bring this about [such as] alternating leg movements which,
if proper contact is made with the uterine wall, lead to somersaults.
The latter pattern lingers on for the first two months after birth and
is known as neonatal stepping; it plainly represents a fetal adaptation
and not a post-natal locomotor pattern. (1997, p. 49)
If Prechtls point is valid then it is as likely that in stepping the infants are trying to
walk as that they are trying to somersault! But given the fact that the infants tread-
mill stepping is qualitatively more similar to adult stepping than the simple neona-
tal reflex (Thelen, 1986, pp. 1500 - 1501), it might be that Prechtls ruling does not
(A) Infant being supported on a split-belt treadmill.
(B) Split-belt treadmill filmed from four angles with infrared cameras (see curtain rails)
and an ordinary video camera (positioned on the floor beside the treadmill).
Both from Thelen & Ulrich (1991, p. 49).
(C) Toe-strike stepping.
(D) Flat-foot stepping.
Both from Ulrich (1997, p. 326)
(E) Bidirectional split-belt stepping: left leg steps forward, whilst right leg steps back-
ward. Solid arrows indicate belt-motion and foot-motion during stance. Open ar-
rows indicate foot-motion during swing.
From Yang, Lamont, & Pang (2005, p. 6874)
| 221

speak to the complexities of the developmental trajectory. Typical adult-stepping
involves a rolling foot motion in which the heel-strikes the ground first, moving
through the flat of the foot to the toe. By contrast, neonatal stepping often involves
toe-first strikes (see Figure 23C) and the occasional flat-footed contact (see Figure
23D). To investigate the intermediate stages more closely, Thelen & Ulrich (1991, p.
36ff.) looked for individual differences that might predict the emergence of stable
treadmill stepping in the first year of life. They discovered a trend of poor treadmill
performance associated with toe-first striking and/or inward foot rotation, whereas
good swing and stance came with flat-footed contact. Only in the latter posture can
the stance leg be pulled back far enough for a sufficient frequency of muscle spindle
impulses to loop through the spine and activate the antagonists, so the leg can then
swing far enough in front for a stable step (recall Figure 15 in 7.1, and Mikes step-
ping principles above). And, as Thelen & Smith explain, the flat-footed postural
arrangement is only available given the (P) property of well-balanced relative ten-
sion of antagonistic muscles.
170
In particular, the tension of flexors needs to be with-
in a certain range:
[If ] the tension [in flexors] is too loose, the treadmill will not impart
sufficient stretch to overcome the inertia of the leg and it will not
swing forward [alternatively, if the flexors are too tight] the treadmill
will not impart enough pull to stretch it. In neither case will the
stretch receptors be sufficiently activated for reciprocal phasing.
(Thelen & Smith, 1994, p. 112)
With that I hope to have sufficiently illustrated the delicate balance of (P) and (W)
properties in virtue of which basic actions are possible, and thereby given a rudi-
mentary sense of what structural affordances actually are. This is little more than a
beginning of the study of structural affordances, for although (as reviewed above)
the relational properties which they typically involve are well understood, the rela-
tions themselves are rarely identified as such.

170
The developmental story perhaps runs more fully as follows:
Newborn infants have a characteristic flexor bias in their limbs; legs and arms are held tightly
to the body [...] probably partially as a result of the tightly packed fetal position [...] the limbs
are relaxed only over many months, and indeed extensor strength in the lengths lags behind
flexor strength throughout the first year. (Thelen & Smith, 1994, p. 112)
It is likely that it is for these reasons that [h]ighly flexed individuals and those who did not
have sufficient extensor strength could not put their foot flat on the belt, and consequently
did less well in treadmill stepping (ibid., p. 112).
222 |
7. 4. A structural affordance theory of kinaesthetic
experience
In this section, I want to set the background for a structural affordance theory of
kinaesthetic experience. Call this SA2, so-called because in one of my only two pub-
lications to date (A. J. T. Smith, 2009), I attempted to develop a structural af-
fordance theory of the spatial content of bodily experience. Call that SA1. SA1 has
been usefully criticised by Frdrique de Vignemont (2009). As the reader might
surmise, I do believe that SA1, when properly understood, can avoid the criticisms
levelled against it so far. But I do not intend to defend SA1 here. Indeed, attention to
certain aspects of de Vignemonts criticism, and the general worries they convey can
help clarify the commitments of SA2. In particular, they help to show why said wor-
ries do not bear on the latter thesis (whether or not they actually do bear on the
former). In the next section, I will show how SA2 both encompasses and serves to
explain the Husserlian phenomenology of chapter 6, and certain other perceptual
phenomena, through the notion of a kinaesthetic ability.
It will be useful to begin by explaining what I intended to do with SA1. I wanted to
find a balanced resolution of certain puzzles and complications in ascribing spatial
content to the experienced location of a bodily sensation. One of these complica-
tions is that any account of the spatial content of bodily experience ought to recog-
nise the fact that parts of the body are differentially mobile, and yet structurally
interlocked (see e.g., Bermdez, 2005, pp. 310 -311). Another is that the sensorimotor
cells in frontal and parietal lobes causally implicated in responding to stimuli on
and around the body, employ multiple distinct egocentric frames of reference indi-
vidually anchored to body parts (Graziano & Cooke, 2006). Together these suggest
that the project of discerning a global frame of reference for bodily experience might
well be in error. I did not give this particularly recalcitrant problem a label, so I will
do so now and call it the partiality problem.
The critique offered by de Vignemont did not bear upon the partiality problem.
Instead, she focussed on certain undesirable features of my positive proposal. The
strategy of the positive proposal was to somehow employ the fact that whole-body
movement (by definition) involves the body as a unity. My aim was to somehow im-
port this essentially unifying feature into an account of how attention to a part of
the body involves the experience of that part against the background of the body as
a whole. However, as a result, my account bore a striking similarity to an enactive
theory of bodily experience (de Vignemont, 2009, p. 101). In her more recent work,
de Vignemont has clarified that there are at least three potential aims for such a
theory, with respect to e.g. tactile experience:
| 223

(i) to individuate the tactile modality (e.g., a tactile experience, by
contrast with a visual one); (ii) to explain the phenomenology of tac-
tile sensations (e.g., the conscious sensation of pressure on my skin);
(iii) to explain the spatial content of tactile experiences (e.g., the sen-
sation that there is an object touching me on my palm that is round).
(de Vignemont, forthcoming-b)
My aim was something in the region of (iii). But I had no intention of explaining
phenomena such as the experience something round touching ones hand. Rather
my aim was to provide an account of what it is to experience ones hand (attention to
which might be drawn by a tactile experience of contact) against the background of
the body as a whole. In fact, this would be just one instance of a general case in
which experience of part of the body involves experience of a part of an integrated
whole. This deserves a label too; call it body-mereological (BM) experience.
171
My aim
was to try and resolve the conflicts generated by BM experience and the partiality
problem, when they are conjoined with claims like the following:
the intrinsic spatiality of bodily awareness is sustained by its present-
ing certain parts of the subjects body as filling particular egocentric
locations [...] this presentation of the body as determinately extended
in physical space depends in turn on the bodily sensation itself ap-
parently being set in that body part there (Brewer, 1995, p. 300)
What suggested to de Vignemont that I was offering an enactive theory to achieve
that aim, were my claims that:
(a) at the sub-personal level sensory information channelled through the tem-
poral and parietal lobes achieves egocentric spatial content in virtue of
providing detailed specifications for movement

(b) at the personal level sensory information achieves egocentric spatial content
in virtue of priming the selection of particular types of action
(cf. de Vignemont, 2009, p. 101; A. J. T. Smith, 2009, pp. 93 - 94).
As far as I can see these alone do not commit my earlier view to a species of enactiv-
ism. However, here (and in other places) the account suffered from ambiguous
presentation. I was not clear at the time (or at least not sufficiently explicit) that by
distinguishing (a) and (b) I meant to distinguish processes that are not part of the
structure of experienced (i.e., (a)), from processes that are (i.e., (b)). It was certainly

171
The term body mereology is due to de Vignemont, Tsakiris & Haggard (2006).
Though they restrict their discussion to part-part relations in neural representation of the
body, they state that the term applies more generally to part-part and part-whole relations,
and presumably of the body simpliciter.
224 |
not clear what bearing these remarks were supposed to have on issues I was discuss-
ing at the time. Thus it is perhaps unsurprising that from (a) and (b), de Vignemont
infers that my intention is the rather vague one of grounding bodily experience in
action (2009, p. 102ff.). The ensuing counter-examples that she brings to bear, indi-
cate that one of her most serious worries is that I might be guilty of assuming an
analogue to (what Andy Clark has called) the assumption of experience-based con-
trol (EBC). Here it is appropriately modified:
[Bodily experience] presents the [body] in a richly textured way, a way
that presents fine detail and that is, in virtue of this richness, espe-
cially apt for, and typically utilized in, the control and guidance of fi-
ne-tuned action (cf. Clark, 2001, p. 496)
For the sake of simplicity, let me stipulate that any enactive theory of bodily experi-
ence is committed to EBC. Here are two problematic observations for such a theo-
ry.
172
The first is the existence of numbsense, in which cortical lesions result in an-
aesthesia of large areas of the body yet sensorimotor processing is retained
(Rossetti, Rode, & Boisson, 1995). Jacques Paillard, speaking of his own patient R.S.,
recounts what is striking about individuals in this situation:
Unable to detect and to perceive any tactile stimulation delivered at
various sites on her right hand when vision was prevented, this pa-
tient showed, to her own surprise, a spontaneous ability to point her
left finger toward stimulated places on her deafferented right hand
(2005, p. 102)
A second source of problems for EBC comes from a phenomenon recently revealed
in a variation of an experimental paradigm used in the study of multi-sensory con-

172
These are both found (along with others) in de Vignemonts (forthcoming-b) in
which she aims to block various options for any potential enactive theory. Clarks aim in his
article is to present similar difficulties for EBC with respect to visual experience (2001, p.
499ff.) by appeal to work in the cognitive psychology and neuropsychology of vision per-
formed under the dual-streams paradigm (see footnote 117 above). On the same basis, Jacob
& Jeannerod (2003, p. 171) provide a succinct critique of ORegan & Nes (2001) sensorimo-
tor theory of visual perception, the direct descendent of which is the enactive approach to
vision advanced by Ne (2004). Pierre Jacob (2006) follows up the critique, raising similar
and more general worries directly targeted at Ne. And more recently, in his (2006; 2008, pp.
169 - 195) Clark reviews a spate of similar arguments against the strong sensorimotor mod-
el (essentially Nes view), concluding in favour of an account on which visual experience
depends upon implicit knowledge of poise over a rather coarse-grained action space (2008,
p. 191; cf. Ward, Roberts, & Clark, forthcoming). This last is a descendent of the hypothesis
of experience based selection that Clark advances in his (2001, p. 512).
| 225

flicts, known as the rubber-hand illusion (Botvinick & Cohen, 1998). In a typical
induction of the illusion, a subject looks at a rubber hand that lies in an anatomical-
ly congruent orientation to their own occluded hand. After around 10-15 seconds of
watching the hand being stroked in synchrony with strokes applied to their unseen
hands the illusion takes effect. It manifests in various measurements. Those typical-
ly recorded are a variety of qualitative changes in perceptual experience of the rub-
ber hand, measured by questionnaire, and errors in perceptual judgement of the
location of the hidden hand, biased towards the location of the rubber hand. In the
original Botvinick & Cohen (1998) study, the latter was measured by a non-ballistic
reaching task, which involved tracing the finger of the un-stimulated hand along a
ruler under the table occluding view of the stimulated hand. In an elegant refine-
ment of the paradigm, Kammers, de Vignemont, Verhagen, & Dijkerman (2009)
employed a passive perceptual response in which the participant would watch the
experimenter trace two fingers along a table covering both the participants hands
until they seemed to be in line with the felt location of their own hands. They then
compared the measured degree of perceptual bias with performance of various bal-
listic and bimanual motor responses. They found that in both bimanual grasping
and ballistic reaching participants were largely accurate, whereas passive perceptual
responses exhibited a typical pattern of errors.
In fact, I was well aware of the dangers of EBC, and this was precisely the reason why
I made the distinction between (a) and (b).
173
But I glossed over a rather thorny is-
sue. What I claimed to be expressing in (a) and (b) was an Evansian view on egocen-
tric spatial content. However, Evans is singled out by Clark as guilty of assuming
EBC (2001, p. 498; cf. Evans, 1985, pp. 383 - 385). As I said, it is not my intention to
defend my earlier view here. But prima facie the issue is pressing, as I have devel-
oped a phenomenology of egocentric perceptual experience on broadly Evansian
ideas. It is worth showing what it would be to be committed to EBC, and worth
demonstrating at least briefly why the account I endorse at present is EBC-innocent.
A further corollary of this exercise is that it will also be useful for setting the context
for SA2 as an account of kinaesthetic experience that clearly does not assume EBC.
To my knowledge the first person to employ Evans ideas in an account of the spati-
ality of bodily experience was Brewer, saying things like the following:

173
I referenced Clarks article in the main text and footnoted that I intended to adopt a
view of the relationship between bodily experience and the guidance of action broadly in
line with Clarks (A. J. T. Smith, 2009, pp. 93 - 94 n15), for which, see the previous footnote.
226 |
When I feel a sharp pain in the back of my left hand or an itch at the
end of my nose, the appropriateness of action concerning these actu-
al bodily locations is written into the very nature of the experience it-
self [...] Which bodily location is involved is given at least in part as a
kind of practical demonstrative: there, said as one reaches for the
place in question (1995, p. 298)
He goes on to cash the experienced appropriateness of the practical demonstra-
tive as the typical form of practical knowledge, knowing-how:
Knowledge of how to point to, protect, rub, or scratch the location of
a sensation is not so much discovered by some comparison of the
consequences of ones flailing movements with an external picture of
a bodily target as immediately present as what it amounts to for the
sensation to seem to be where it is (ibid., p. 302)
The relevant question with regard to EBC is what such know-how is supposed to
involve. In particular, does it involve the guidance of the action (as the contrast with
flailing movements suggests)? It is not my intention to legislate on the details of
how we ought to conceive of such knowing-how in general. Stanley & Williamson
(2001) have argued that all know-how reduces to propositional knowledge under a
practical mode of presentation. No (2005) demurs. Hybrid positions are possible
besides these. In any case, I think it is safe to assume the default in discussion of
knowing-how, which is that it consists in a set of dispositions (Ryle, 1949/2000).
Grush has usefully identified two kinds of dispositions applicable in this context,
detail-specifying dispositions and type-selecting dispositions.
A detail-specifying disposition is a disposition that, for any given be-
havior type (such as a grasp or foveation, or whatever), specifies the
details of how that behavior type will be executed if it is executed. So
for example will my intended grasp (behavior type) be implemented
by moving my hand like this, or like that? (Grush, 2007b, p. 393)
This could potentially amount to asserting EBC, if detail specification is understood
as being fine-grained specification of movement, and experiencing an object just
amounts to the possession of a disposition for such fine-grained specification.
Detail-specifying dispositions are to be contrasted with type-specifying disposi-
tions:
| 227

A type-selecting disposition is something about the stimulus that
motivates the execution of this or that behavior type, as opposed to
nothing or some other behavior type. For instance, a bright ash
might motivate a head turn and foveation, but not a grasp; an itch
might motivate an arm and hand movement and scratch, but not any
eye movement (Grush, 2007b, p. 393)
In 6.5, I argued that in order to make sense of particular objects being experienced
egocentrically within the constraints of the Husserlian account, it would be neces-
sary to endorse the behavioural egocentricity claim:
Behavioural egocentricity: there is nothing more to experiencing x as at ego-
centric location e, beyond a subject being disposed to perform potential ori-
entating actions with regard to x
The important point is that are to be defined as types of action, or rather that the
disposition to is not to be defined as the disposition to perform a movement of a
specifically fine level of granularity. However, in order to clarify that, I will need to
appeal to more than the intuitive contrast drawn by Grush.
Consider some experimental work on motor control biases. Fourneret & Jeannerod
(1998) asked 13 right-handed subjects to draw a straight line, straight ahead, on a
graphics tablet, with their right hand. They were also asked to keep their hand in
contact with the surface of the tablet, and to move it at an even speed toward a tar-
get presented on the sagittal plane. The subjects could not see their hands directly;
rather they were presented with a mirror situated over their arm and hand, above
which was a computer interfaced with the graphics tablet (see Figure 24A & C). This
allowed the experimenters to introduce a subtle bias on some trials, creating a devi-
ation in the output of the tablet, in either the left or right direction, by 2, 5, 7 or 10.
So, for instance, if a leftward bias was introduced, a subject would have to adjust
their movements rightward in order to compensate and produce a line on the screen
that would reach the target. No matter the magnitude of the bias introduced, sub-
jects were able to adjust their movements accordingly (Figure 24B). That in itself is
interesting, but it is even more interesting in conjunction with their reports.
Subjects were asked, According to your own impression, which line corresponds to
the actual trajectory drawn by your hand? They had to give a number in response,
corresponding to the range of options presented on Figure 24D. All responses clus-
tered around line no. 7. In short, subjects were not able to verbally report on the
extent to which compensatory changes were made. For a non-verbal estimate, in a
session before the start of the first trial, subjects were instructed to draw a line that
they judged to be straight ahead. This was repeated 5 times in order to gain an esti-
mate of each individuals subjective straight ahead. Then after each trial they had
228 |
their hands passively placed at the original start-point and were asked to panto-
mime the movement they had just made, but with their eyes closed. The overall
tendency in pantomimed movements was very close to the subjective straight
ahead. Again, subjects were not able to report on the extent to which compensatory
changes were made.

Figure 24 An experimental paradigm for sensorimotor control biases

Anti-EBC interpretations of these results typically run as follows:
although proprioceptive and internally generated action-related sig-
nals correctly operated at an automatic and unconscious level, they
were not available for conscious monitoring (Fourneret & Jeannerod,
1998, p. 1138)
(A) Schematic of the experimental set-up and apparatus
(B) A typical compensatory movement after the introduction of a 10 bias
(C) Overhead view of (A) and description of the task
(D) A chart used for verbal indications of subjects movement trajectories
Reprinted from Fourneret & Jeannerod (1998) with permission. 1998
Elsevier.
| 229

normal subjects appear to be unable to monitor the signals generated
by their own movements [] Subjects, when they are requested to
make a conscious judgment about what they did, simply tend to ad-
here to the goal, but cannot access the way it has been achieved
(Jeannerod, 2003, p. 134)
I think the results above weigh pretty heavily against an EBC account of kinaesthetic
experience. There is a temptation to infer that the movements themselves are not
experienced at all. However, it does not follow from the fact that one does not expe-
rience x in a fine-grained way, that therefore one does not experience x at all. For
example, even if all I can make out of an object in the distance is a blur, or all you
can hear of my voice is a faint squeak, it would still be true that I experience seeing
something and that you experience hearing something. A similar point holds for the
experience of movement. The movement occurred, and we have no antecedent rea-
son for thinking that the subject is not experiencing any movement. If one is to
demonstrate that, it is not sufficient to demonstrate that the subject does not expe-
rience movement in a fine-grained manner.
Here is about the right place to illustrate the kind of explanation that would be pro-
vided by SA2. To take an example, recall the antics of the blacksmiths arm from
7.3. As an agent his possibilities for action are set by his SAs. As he brings down
each strike, he anticipates the possibility of the same type of action. His arm has the
(P) properties of motor redundancy and structural unity. As he repeatedly strikes
the same spot, he employs a slightly different combination of joint angles each time,
but to the same effect. SA2 claims that he does not thereby experience the move-
ment as different. In anticipating the possibility of the action he tries to perform, he
anticipates a broadly continuous pattern of kinaesthetic sensory experience corre-
sponding to that action. There are two key moves here. The first is to claim that
subjects anticipate phenomenal grooves:
Phenomenal groove: A broadly continuous pattern of kinaesthetic sensory
experience corresponding to a type of basic action
This is essentially an extension of the Husserlian description, as I will clarify below.
The second is to posit kinaesthetic ability on the part of the subject:
Kinaesthetic ability: the ability to anticipate phenomenal grooves in virtue
of anticipating SAs.
The Fourneret & Jeannerod study provides an opportunity to clarify these concepts.
In order to adequately accommodate the bias introduced by the experimenter, we
need to distinguish between the experience of an action-token and the phenomenal
groove of an action-type. An action-token corresponds to the actual movement per-
230 |
formed by the subject, in its finest granularity. This token can be an instance of a
right-arm straight-ahead type, or it can be an instance of another type. Now apply
the notion of broad continuity. Just as sensations of a particular object can violate
narrow continuity and nevertheless be experienced as the same particular object,
kinaesthetic sensations can violate a certain narrow continuity and nevertheless be
experienced as the same action-token. This would be in virtue of being within the
phenomenal groove of a particular movement type. With this gloss, the Fourneret &
Jeannerod study provides a positive insight into the experience of action. For it pro-
vides reason for thinking that the experience of movement corresponds to the antic-
ipated experience of an action-type, viz. the anticipation of a phenomenal groove.
A follow-up study by Slachevsky et al. (2001) provides the opportunity to indicate
the measurement of phenomenal grooves. They ran a very similar study to
Fourneret & Jeannerod (1998), except that all biases introduced were rightward, and
the extent to which biases could be introduced was considerably larger (up to 40).
At an average of around 14 normal subjects were able to report that their move-
ments deviated from straight ahead; indicating that sensations of movement can
only deviate up to 14 if they are to be experienced as being of the same action-type.
The phenomenal groove for a right-arm-straight-ahead action extends to about 14
on the sagittal plane.
It is worth explicitly stating why phenomenal grooves and kinaesthetic abilities are
EBC-innocent. Here is the modified version of EBC again:
[Bodily experience] presents [the body] in a richly textured way, a
way that presents fine detail and that is, in virtue of this richness, es-
pecially apt for, and typically utilized in, the control and guidance of
fine-tuned action (cf. Clark, 2001, p. 496)
In SA2, no claim is made about a relationship of controlling or guiding between
experience and action. Furthermore, phenomenal grooves are defined in terms of
broad continuity. This is effectively the opposite of experiencing fine-detail and
richness.
Additionally, the term allows one to compress certain elements of the Husserlian
account. The conclusion of 6.4 was this:
(p). Experience of particular objects requires anticipation of a broadly continu-
ous kinaesthetic cycle.
Conjoining the behavioural egocentricity claim with (p) produces:
| 231

experiencing x at egocentric location e involves anticipating a broadly con-
tinuous cycle of kinaesthetic sensations that would occur, were the subject
to with regard to x
And given the definition of a phenomenal groove, the previous bullet-point trans-
lates to:
experiencing x at egocentric location e involves anticipating a phenomenal
groove that would occur, were the subject to with regard to x
7. 5. Phenomenal grooves & kinaesthetic abilities
SA1 and SA2 are distinguished by their respective explanandum and explanantia
(see Table 1).
Table 1 - Structural affordance theories of bodily experience

Explanandum

Explanantia
SA1
Body mereological
experience.
Representation of parts of the body;
exploiting structural unity of the body.

SA2
Structuring role of
kinaesthesia in
perceptual experience.

Kinaesthetic ability.

I can now briefly show SA2 in application: In 6.5 it was claimed egocentric loca-
tions are systematically arranged in virtue of being constrained by the structure
and mobility of the perceptual organs. These constraints are precisely the relevant
(P) and (W) properties of each kinaesthetic system. SA2 will claim that:
In experiencing x at egocentric location e a subject is able to anticipate the
phenomenal groove that would occur, were she to with regard to x, in vir-
tue of anticipating the instantiation of the SAs that enable .
This generalises to any case in which the agent might anticipate or try to make a
type of basic action. And if the Husserlian account is on track, then this will hold for
any instance in which a particular object is experienced. To convey the point, let me
cherry pick two of the other claims made in chapter 6.
232 |
(k). For every series of bodily movements, there must be an anticipation or man-
ifestation of a proprietary series of mere kinaesthetic sensations.
174

(o). Broad continuity in a series of perceptual sensations requires broad continu-
ity in mere kinaesthetic sensations throughout a kinaesthetic cycle.
Unless claim (o) is to be absolutely empty of content, there needs to be a principled
explanation of why there are bounds to a subjects anticipation, when she antici-
pates broad continuity in a series of kinaesthetic sensations (sc. a phenomenal
groove). These bounds are set by her anticipation of SAs (sc. her kinaesthetic abil-
ity). In short, broad continuity in a series of perceptual sensations is explained by
the subjects kinaesthetic ability to anticipate phenomenal grooves.
If practical knowledge is a candidate for explaining the egocentricity and particular-
ity of perceptual experience, what about other perceptual phenomena? I think it
would be unwise to claim that practical knowledge can explain every aspect of per-
ceptual experience, or indeed every case in which one might anticipate having cer-
tain (even sensory) experiences. But there is least one more interesting phenome-
non with respect to which the idea is of some importance, which found its first ex-
pression in the 20
th
century in OShaughnessys (1989), but has been traced by A.D.
Smith (2002, pp. 153, 304 n137) through 19
th
century German and French philosophi-
cal writings. Smith gives an initial description as follows:

174
There is a sense in which claim (k) is false, as it is too generally stated. Cases such as
Ian Watermans (see 8.3.) provide obvious counterexamples:
A common problem in patients without feedback is that when they are not looking at their
limbs the arms and legs start to move on their own. The fingers dont remain still, but writhe
in small movements, and the arms may start moving uncontrollably. If he turned his gaze
away for a few seconds, his arm would often come up and hit him, or someone sitting close by.
(Cole, 1995, p. 16)
Accordingly, claim (k) should actually be something like:
(k*) For every series of bodily movements that a subject with a healthy peripheral nerv-
ous system tries or anticipates trying to perform there must be an anticipation or
manifestation of a proprietary series of mere kinaesthetic sensations.
But this is merely claim (k) with italicised cetera that had previously been assumed to be
paribi.
| 233

Although no mere impact on a sensitive surface as such will give rise
to perceptual consciousness, we certainly feel objects impacting on
us from without. This fact needs to be recognized in any adequate
perceptual theory [...] the phenomenon that is central here [...] is
that of a check or impediment to our active movement: an experi-
enced obstacle to our animal striving, as when we push or pull against
things. (2002, p. 153)
Smiths suggestion is that the latter is somehow basic to tactile experience.
OShaughnessy comes to a similar conclusion from a different tack. He wonders why
so many and various phenomena are all described as instances of tactile perceptual
experience, offering the following sample.
The simplest case of all is momentarily feeling something contact a
point on one's body. The next most simple case is feeling something
extensive momentarily contact one. Complexity appears when time,
motion, and activity appear as additional variables. Thus, one might
passively feel an insect wander across one's chest. Or actively run
one's hand around the outside of a box. Finally, one might feel one's
way around a room in total darkness; or, if blind, navigate one's way
across an entire city (O'Shaughnessy, 1989, p. 41)
The intuition he takes as obvious is that they all count as cases of the sense of touch
because in all cases the sensation of contact is used as access to the perceived
item (ibid., p. 41). But this he immediately rejects in the face of a counter-example:
One's right hand might be completely numb, and yet by discovering that we cannot
push it through something it evidently is touching we discover the presence of that
something (ibid., p. 41). At which point he asks (rhetorically): Then do we not
perceive that object, and through touch? (ibid., p. 41). Perhaps we do, but there is a
frustrating clumsiness about OShaughnessys exposition. After all, he leaves open
the question of how it is evident that one is touching something. From all he says
it could be evident to a subject because she just had a look at what impeded her
movement. This equivocation obscures the main point, which is that the experience
of movement being impeded is in some way primitive to tactile experience.
A. D. Smith is more careful. To illustrate the phenomenon, he considers the experi-
ence of a sightless being whose first experience is of movement through empty
space:
234 |
Here experience would be, if I may so put it, of sheer kinesis in a
void, not movement in a space that is appreciated as a realm poten-
tially holding alien entities. There would be no sense, here, of any
bodily limits beyond which things might be located. Such a limit,
and hence a phenomenal body itself, and hence a tactile distinction
between inner and outer, and hence a genuine felt space in which we
are located along with possible other things, emerges only when con-
tact is made with something. Such contact [...] is our sole mode of ac-
cess in this modality to spatially located objects. [It] simultaneously
reveals both an outer object and our own body by establishing a space
in which both are located, and, thereby, confers spatial meaning on
what would otherwise be meaningless kinesis. (op. cit., p. 158)
Leaving aside the potential complications introduced by vestibular sensation and
somatic graviception, there is something right about this. But it is worth noting a
distinction between there being limits to bodily experience, and the various ways in
which the body is experienced as having limits. To begin to see this consider Bill
Brewers interim conclusion in his (1995) battle with the modern Cartesian. Trium-
phantly, he claims that: Bodily awareness is intrinsically spatial: the apparent loca-
tion of a sensation is as essential to its very nature as its purely qualitative feel
(Brewer, 1995, p. 300). This observation can be supplemented with Martins remark
that the sense one has of the location of a sensation brings with it the sense that
the location in question falls within one of ones apparent boundaries (1995, p.
271).
175
But the kind of boundary that Martin has in mind brings with it a modal
contrast [...] between regions where one could currently feel sensation and those
where one couldnt, a contrast that is to be drawn within the content of the experi-
ences themselves (Martin, 1995, pp. 271-272). And he clearly intends this to be a
fully universal claim.
Any region in which it seems to one that one could now be feeling
sensations will thereby feel to one to fall within one of ones bounda-
ries; at the same time, one has the sense that there are locations out-
side of ones boundaries, whatever these happen to be, since the
space one feels these locations to be part of feels as if it extends be-
yond whatever one does feel (Martin, 1995, p. 271)
But this is precisely what the experience of Smiths first-time mover would lack. For
ex hypothesi, it is only once a region of space that a part of her body occupies is

175
Brewer echoes the point, saying that this spatial location of sensation comes to
light only as one is aware of ones body as determinately extending into and filling certain
regions of the perceived physical world (1995, p. 300). See also OShaughnessy (1980/2008,
pp. 198 - 202).
| 235

forcefully abutted with a region of space occupied by another body (or another part
of her body) through her own movement, that she will then have an experience of
occupying space. However, an open question remains. For despite the lack of any
experience of boundaries in the sense that Martin proposes, presumably she could
still have some experience of the fact that her movements are constrained. Indeed,
in order for there to be any significance in the fact that she experiences her body as
being able to move no further, she must know (or come to know) how far her body
would be able to move were it not thus impeded. In order for there to be an experi-
enced obstacle to [her] animal striving, she needs to have (or come to have) some
experienced anticipation of the bodily structure that affords such striving in the first
place, and the extent to which it does so. She needs, in short, to have some minimal
kinaesthetic ability.
236 |
Chapter 8 On bodily selfhood
8. 0. Introduction
In the previous chapter, I laid out some basic ideas towards a naturalistic explana-
tion of how kinaesthesia provides structure to the intentionality of perceptual expe-
rience, as conceived in earlier chapters. I have not described kinaesthesia as a form
of self-consciousness. This might seem unprincipled, as at several points I have used
the word subject and indeed the expression bodily subject. However, the reader
may be unsurprised to hear that my use of this term is not intended to connote that
subjects are things to be found on a final metaphysical analysis; nor do I subscribe
to the contrary of that proposition. Nevertheless, the topic of bodily selfhood ought
not to be dismissed entirely, and I devote this chapter to a discussion of what I un-
derstand to be the core issues.
The chapter begins by presenting the bodily self thesis, a positive metaphysical the-
sis about the existence of bodily subjects qua bodily selves. I confront the thesis with
arguments found in the Discourse and the Meditations in which Descartes seeks to
establish the real distinction of himself from his body. An inevitable stalemate is
reached. In order to move the dispute forward, I construct a Cartesian thesis which
can be more readily addressed by the bodily self thesis. I then review the core con-
cerns of authors who seek to defend the bodily self thesis, drawing upon Evans
work in order to show how they might do so. Having established a clear line of de-
fence against the Cartesian, I then construct a dilemma for defendants of the bodily
self thesis, which I do not believe to be so easily overcome. The dilemma leaves the
bodily self thesis with an open challenge: to specify the nature of the body that is
claimed to be the self. In closing the chapter, I turn to the more general discussion
of whether subjects of experience exist at all, i.e., whether there is such a thing as a
self. I conclude that again the dispute is likely to remain unsettled, and welcome the
idea that the more fruitful topics of discussion are to be found elsewhere.
8.1 Descartes himself
8.2 Cartesian selfhood
8.3 Anti-Cartesian selfhood
8.4 Being one body
8.5 Being no body
| 237

8. 1. Descartes himself
My treatment of the intentional objects of experience is such that I do not take a
stance on their existence; only their seeming existence is deemed relevant. Thus far
I have managed to gloss over questions concerning the metaphysical nature of the
subject of experience. Although the issues are rather different, my stance here is the
same. Just as intentional objects are not posited as entities of any kind, so also sub-
jects of perceptual experience intentionally directed at such objects enjoy the same
neutral status as far as this dissertation is concerned.
But so far we have not come across any background against which such neutrality
could be claimed. For instance, puzzles concerning the existence of intentional ob-
jects concerned cases of hallucination and the worship of Ancient Greek gods: what
corresponding issue could there be as regards subjects of experience? After all, in
such cases, it is never at issue whether the subject suffering hallucination or en-
gaged in worship has some queer non-actual status. Indeed, given remarks at the
end of 6.3, a dense (but fairly accurate) gloss on the content of 6.4 and 6.5 is
this: A bodily subject experiences particular objects in virtue of anticipating a cycle
of kinaesthetic sensations affecting and effecting visual sensations and enabling
anticipations of continuity and unity across experiences of the object. It seems like
the account posited is committed in some way to the idea of a bodily subject. Why
not embrace its existence wholeheartedly? Why not, in short, embrace the bodily-
self thesis (BST):
Bodily subjects of experience exist as bodily selves
From the way I frame the thesis, the reader can surely glean that I will call any sub-
ject of experience that is held to exist, a self. The metaphysical status of selves is an
issue that is just as problematic as the status of intentional objects; and the exist-
ence of bodily subjects more so. This is not to state that the problems cannot be
solved (or dissolved), merely to state that they are somewhat vexed, and arise on not
one but two fronts. So the question that many begin with is not: what is a bodily
self? But rather, what is a self, such that it is (or is not) bodily?
Here is a characteristic remark that suggests that attributing BST to Husserl is not
too far off track:
238 |
In seeing, the eyes are directed upon the seen and run over its edges,
surfaces etc. When it touches objects, the hand slides over them.
Moving myself, I bring my ear closer in order to hear [...] given with
the localization of the kinesthetic series in the relevant moving
member of the Body is the fact that in all perception and perceptual
exhibition (experience) the Body is involved as freely moved sense
organ, as freely moved totality of sense organs (Husserl, 1952/1989, p.
61 [56], italics removed)
Another source of quotes that hint at a metaphysical account of BST is to be found
in the French philosophical tradition. For instance, witness Maurice Merleau-Ponty
declaring that I am a body which rises toward the world (1945/1962, p. 75), and
Jean-Paul Sartre asserting that I am my body to the extent that I am; I am not my
body to the extent that I am not what I am (1943/1992, p. 326). Whatever the in-
tended connotations of these last remarks, they stand in thorough contrast with an
inference inscribed earlier by another French author, Ren Descartes: the famous
ego cogito, ergo sum. The reasoning is found in his Discourse on Method
(1637/1985); having discovered what later came to be known as the cogito as a solid
ground from which to proceed, he immediately follows with an explicit rejection of
BST:
Next I examined attentively what I was. I saw that while I could pre-
tend that I had no body and that there was no world and no place for
me to be in, I could not for all that pretend that I did not exist. I saw
on the contrary that from the mere fact that I thought of doubting
the truth of other things, it followed quite evidently and certainly
that I existed; whereas if I had merely ceased thinking, even if every-
thing else I had ever imagined had been true, I should have had no
reason to believe that I existed. From this I knew I was a substance
whose whole essence or nature is simply to think, and which does not
require any place, or depend on any material thing, in order to exist.
Accordingly this I, - that is, the soul by which I am what I am - is en-
tirely distinct from the body, and indeed is easier to know than the
body, and would not fail to be whatever it is, even if the body did not
exist. (Descartes, 1637/1985, p. 127)
I should decompose some of the moves made here. Begin with from the mere fact
that I thought of doubting [...] it followed quite evidently and certainly that I exist-
ed. This expresses the claim that thinking provides sufficient warrant for belief in
the existence of a self. And then, whereas if I had merely ceased thinking [...] I
should have had no reason to believe that I existed expresses the claim that think-
ing is a necessary condition on warranted belief in the existence of a self, at least in
so far as one is rational. Neither of these suggests a claim about the identity or non-
| 239

identity of the self with the body. However, Descartes then claims to know that he is
a substance whose whole essence or nature is simply to think, and which does not
require any place, or depend on any material thing, in order to exist. And from the-
se he draws the conclusion that the self is entirely distinct from the body [...] and
would not fail to be whatever it is, even if the body did not exist. Thus, far from
being neutral on questions of identity, he asserts the existence of the self as an im-
material substance, wholly distinct from the body.
176

This argument moves from various epistemological premises (that all hinge on the
assumption that there is something epistemically special about thought), to an on-
tological distinction between selves and bodies. The natural response for a fan of
BST is to deny that the epistemological premises matter, or rather to say that they
are irrelevant. She could thus assert the reality of the bodily self, claiming that it is
real in precisely the sense that its status is determined in independence of our epis-
temic capacities. As McDowell remarks, ones bodily being need not always impress
itself on one (1994, p. 104 n25). This would be to say that it does not follow from the
fact that I do not know at present (or even, could never know) that I am a body, that
I am not a body.
However, Descartes of the Meditations has a response available. In particular, he
appeals to a methodological principle in asserting the real distinction of mind and
body, namely: the fact that I can clearly and distinctly understand one thing apart
from another is enough to make me certain that the two things are distinct
(Descartes, 1642/1984, p. 54). And he continues:
[Simply] by knowing that I exist and [clearly and distinctly] seeing at
the same time that absolutely nothing else belongs to my nature or
essence except that I am a thinking thing, I can infer correctly that
my essence consists solely in the fact that I am a thinking thing.
(ibid., p. 54)
The 6
th
meditation, from whence this quotation comes, is also where Descartes be-
gins to rebuild the world on the firm (or perhaps, merely, systematic) principles he
has established in the course of his reflections. The first of these is the existence of
his body in which he experiences sensations. Not wishing to now deny this, he con-
tinues:

176
See Hatfield (2003, pp. 103 - 106) for a discussion of these various elements as they
appear in the 2
nd
Meditation.
240 |
It is true that I may have (or, to anticipate, that I certainly have) a
body that is very closely joined to me. But nevertheless, on the one
hand I have a clear and distinct idea of myself, insofar as I am simply
a thinking, non-extended thing; and on the other hand I have a dis-
tinct idea of body, insofar as this is simply an extended, non-thinking
thing. (ibid., p. 54)
On this basis, he concludes that it is certain that I am really distinct from my body,
and can exist without it (ibid., p. 54). This fuller conclusion certainly has the inten-
tion of ruling out the possibility suggested by the retort above. I cannot be such that
my bodily being is unknown (the conclusion urges), for it is certain, by the method
of clear and distinct perception, that I can exist without a body.
There are many, many points at which one could disagree with Descartes reasoning.
One potential weak-point is the conflation of metaphysical truth with certainty, and
indeed the experience of certainty, in using clear and distinct ideas to discern meta-
physical insights.
177
Another is the ground (or lack thereof ) of clear and distinct
ideas.
178
But these would both be disputes of method. Much better would be to es-
tablish BST on common ground. For that I want to focus on his seeming concession
that It is true that I may have (or, to anticipate, that I certainly have) a body that is
very closely joined to me. This is also usually construed as a fundamental weak-
point in his argument. That is, in large part the issues here are with the fundaments
of Descartes dualism, namely, whether properties of mind and properties of matter
(such as extension) are incompatible, and how they might nevertheless causally
interact. Again, this would not be a debate so much on the point at issue, but the
broader metaphysical background.
179

177
I am grateful to Jenny Windt for this suggestion (see also Metzinger, 2003a, p. 132).
Janet Broughton has also argued resourcefully that Descartes employs a form of argument
persistently throughout the Meditations (and several other writings) according to which
indubitability establishes truth (2002, p. 166, see esp. pp. 97 - 107).
178
This is something of cottage industry in Descartes scholarship, see Rickless (2005)
for a good discussion and reference on the topic.
179
But for a nice expose of Descartes interrogation on the topic by Princess Elizabeth
of Bohemia, see Hatfield (2003, pp. 266 - 269). And for discussion of Descartes oscillating
position, see Wilson (1978, pp. 179 - 193).
| 241

8. 2. Cartesian selfhood
In order to facilitate a more direct confrontation, I want to switch from talking
about Descartes method, or what he may or may not have been justified in claim-
ing, to talking about the received Cartesian view of the nature of selfhood. In par-
ticular, consider the following adapted quotation as a fairly Cartesian view of em-
bodiment:
part of what it means to means to say that a [self ] is embodied in a
particular body [...] leaves it a purely contingent fact that the causal
connections between [selves] and bodies are such that [selves] count
as being embodied in the bodies. (Shoemaker, 2003, p. 114)
180

One can also extract a broad picture of body-self interaction from Shoemakers re-
marks as follows:
a [self ] is volitionally embodied in a certain body to the extent that
volitions of the [self ] produce in that body movements that conform
to them or fulfil them, that is, movements that the [self ] is trying to
produce or which are constitutive of actions he is trying to perform
[...]
a [self ] is sensorily embodied in a certain body to the extent that
the interactions of that body with its surroundings produce in the
[self ] sense-experiences corresponding to, and constituting veridical
perceptions of, aspects of those surroundings. (ibid., p. 117)
With this framework in place, the Cartesian can prima facie accommodate all the
phenomena discussed in chapter 6. But this gives no reason to favour one over the
other. If one leaves aside methodological issues, and issues concerning the coher-
ence of dualism, one has something of an impasse. If the Cartesian is going to cut
any ice against BST she needs to find an edge in discerning the nature of selfhood.
To further make her case, the Cartesian could stake a claim to the true reference of
the first-person singular concept I, in distinguishing the true nature of the self.
Wittgenstein seems to express a helpful thought in some notes for his students,
known as The blue book, saying that the idea that the real I lives in my body is con-
nected with the peculiar grammar of the word I, and the misunderstandings this

180
I have exchanged Shoemakers references to persons for selves in order to facilitate
the point. It should accordingly be noted that in the article Shoemaker is concerned with the
logical possibility of the disembodiment of persons (and indeed minds more generally) not
selves per se.
242 |
grammar is liable to give rise to (1958, p. 66). To illustrate this peculiar grammar
one might take expressions like I see, I speak, I move, I feel pressure. In so far as
reference to the body is required these can all be decomposed. Again, Shoemakers
remarks illustrate:
My body is the body from whose eyes I see, the body whose mouth
emits sounds when I speak, the body whose arm goes up when I raise
my arm, the body that has something pressing against it when I feel
pressure, and so on. (Shoemaker, 1968, p. 567)
An anti-Cartesian way of reading this sentence would be to say that in referring to
my eyes, I thereby refer to a part of myself. But here is a deliberately Cartesian way
of reading the sentence: the body (and each of its parts) is a relatum in a relation of
ownership. It is the other relatum, I, that is the subject proper; the two terms refer
distinctly.
The Cartesian might further elaborate, by claiming that there are two patterns of
use of the word I (or my), which refer distinctly. There is that which Wittgenstein
calls use as object e.g. My arm is broken, I have grown six inches, I have a bump
on my forehead, The wind blows my hair about. And another which he calls use as
subject e.g. I see so-and-so, I hear so-and-so, I try to lift my arm, I think it will
rain, I have a toothache (1958, pp. 66 - 67, italics original). The problem with all
uses of I (or indeed my) as object is that they involve a potentially erroneous pro-
cess of identifying the referent by description.
181
This would not be a problem for the
notion of a bodily subject, but for the fact that uses of I as subject do not seem to
involve such a process.
182

181
Descriptive reference is not the only mode of reference that can fail to adequately
capture self-reference. Hector-Neri Castaeda (1966/2001), argues extensively for the irre-
ducibility of reference to oneself as oneself (to signify this distinctive manner of reference he
uses the device * applied to personal pronouns). John Perry (1979) adduces a range of simi-
lar arguments to argue for the irreducibility of certain I-thoughts (viz. those that express
beliefs about where, when and who one is) to propositions, or de re attributions of belief.
Castaeda indicates that his interest was first drawn to the issue by some observations made
by Peter Geach (1957) on the attribution of beliefs to others about themselves.
182
Shoemaker remarks: Ordinarily a question of the form Which body is Ss body?
would ask for the identification of a corpse (2003, p. 117 n7). Existential phenomenologists
(cf. also Husserls notion of Leib) are concerned to reject any straightforward identification
of the body as an object for closely related reasons. But it is not clear that this would be a
fruitful line to take against the Cartesian.
| 243

Take a case such as I have been seen. If all goes well I picks out myself as a body.
An instance where I might actually think a thought expressible so, would be upon
seeing someone pointing and thinking of the thing that she is pointing at call
this a and believing that thing to have been seen call this property F. If based
upon such thoughts, the judgement
I am F
is based upon an identification of a as b, and b as myself. This is a process of identi-
fication. In this process there are really three judgements
a is F, b is F, & I am b
In order to yield the conclusion that I am F, these must rest upon two identity
propositions
a b, & b myself
But mix-ups happen. The person may be pointing at something far more interesting
behind my head. As a mere product of vanity (or fear), the identification fails, and
so the self-reference fails. And we cannot resolve the possibility of reference failure
by a recursive process of identification, as each identification may be inadequate (it
may have latched onto the wrong object or none at all) and thus require a further
identification, which itself might require a further identification, and so on ad in-
finitum (Russell, 1910-1911, pp. 110 - 111, 119 - 120; Shoemaker, 1968, p. 561; 1996, p.
13).
183

Hence, in such cases the possibility of an error has been provided for
(Wittgenstein, 1958, p. 67). Following Wittgenstein, Sydney Shoemaker describes
this as the possibility of error relative to the first-person pronoun (Shoemaker,
1968, pp. 556 - 557). Such error is possible iff. one judges some a to be F, but may be

183
One could assume that in descriptive singular reference, the subjects knowledge of
which object is in question is ultimately determined by a definite description, such as The
X uniquely satisfied by a. Thus every judgement of the form a is F would rest upon only the
pair of judgements a = The X & The X is F. It is still possible that I realise that X is
without realising that I am X and therefore that I am F (see Brewer, 1995, pp. 293 - 294).
In fact, as many have noted, this model allows for cases in which speaking of or referring
to oneself is compatible with not knowing that the object one speaks of is oneself
(Anscombe, 1975/1991, p. 72). For instance, I may be standing in line at my local shop and
give an idle glance at a closed circuit television screen presenting an image of the shop floor.
In doing so I may notice that the bald man (the X) is getting his pocket picked (is ), but
yet not realise it is my pocket being picked (I am F) as I had forgotten that I just had my
hair cut (see Perry, 1979, p. 3 for a similar example).
244 |
mistaken, for the exclusive reason that one believes a to be oneself.
184
By contrast,
Shoemaker and Wittgenstein hold this kind of error to be impossible when one re-
fers to oneself as a subject. If I say I smell a rat it would be absurd for someone to
ask me Are you sure it is you who smells a rat? One can ask of course, but a degree
of irrationality is implicated (on the part of the subject in question) in the expecta-
tion of any possible negative answer. For when I refer to myself as a subject my ref-
erence is immune to error through misidentification relative to the first person pro-
noun.
Encouraged, the Cartesian can continue to urge that on this principle the body and
the subject ought to be distinguished. Take the statement:
I see with my eyes
BST ought to take this as equivalent to:
I visually experience as a body
But in light of the contrasting uses of I above, the Cartesian can claim it is a com-
plex statement that ought to be further decomposed. One component is an ego
statement declaring that I see. Another, distinct, component is a body statement
declaring that my body has eyes. She might then triumphantly pronounce that ego
statements are (strictly speaking) the only correct instantiation of self-reference, on
the basis that all and only Ego statements cannot fail to refer to the subject, precise-
ly because they only attempt to ascribe properties to the subject.
185

8. 3. Anti-Cartesian selfhood
This line of argument attempts to achieve, on the part of the Cartesian account, the
status of inference to the best explanation: our patterns of reference in using the
term I to refer to our bodies are precarious, precisely because we cannot know our-
selves as bodies; for ex hypothesi we are not our bodies at all. But in fact the propo-

184
So one might attribute F to another and not commit the error in question. Or one
might misattribute F when one should have attributed G and not commit the error in ques-
tion. Or one could misattribute the attitude or mode of ones relation (believing instead of
hoping, hearing instead of seeing) and still not commit the error in question.
185
See Brewer (1992, pp. 295 - 297). The Cartesian theme is also inspired by Anscombes
worry that if I is a referring expression, then Descartes was right. But now the troubles
start (1975/1991, p. 78). However, see Evans (1985, pp. 224 235) for a rather different moral
that might be drawn from such a discussion.
| 245

nent of BST could easily appeal to cases in which I is used as subject in the self-
ascription of bodily properties. And these could be bolstered by a positive account
of a mode of reference that guarantees against reference failure.
A key feature of the potential reference failure in the case above was that self-
ascription of the predicate F occurred through a process of identification. In partic-
ular, the process involved establishing the identity of descriptions. At each stage the
identity could fail, and thus the reference could fail. In contrast, Evans proposes a
model of self-identification roughly along the lines of his account of demonstrative
thought.
186
Instead of ascribing F as part of a mediate description of the object in
question, discrimination by demonstrative is direct, in that it does not involve any-
thing like the process of identification described above. Evans calls it identifica-
tion-free in the sense that the subject does not come to know that a predicate is
being instantiated and only then indirectly identify an object by means of some
description involving that predicate. Rather the very means by which the subject
comes to know that a predicate is instantiated is exactly the means by which the
subject comes to know directly which object instantiates that predicate (Evans, 1982,
pp. 179 - 181).
The directness characteristic of identification-free reference necessitates a particu-
lar relationship between referrer and referred: an informational link. As Evans puts
it

186
He suggests that we see self-identification on the model of here-thoughts, rather
than this-thoughts, largely to accommodate for cases of potential lag in the receipt of in-
formation updating the thought. Though this suggestion is not meant to bar there being
distinctive ways of gaining information about oneself that inform bodily self-ascription
(1982, pp. 151 - 170, 215 - 216).
As a side note, although Evans work has been influential in this area, the literature on im-
munity to error through misidentification relative to the first person pronoun is now rather
large. For a good reference on recent work, and an exploration of how a (roughly) Evansian
account fares in the light of various illusions and delusions, see de Vignemont (forthcoming-
a).
246 |
distinctively, demonstrative thoughts take place in the context of a
continuing informational link between subject and object: the sub-
ject has an evolving conception of the object, and is so situated vis-{-
vis the object that the conception which controls his thinking is dis-
posed to evolve according to changes in the information he receives
from the object. (1982, p. 146)
This requires further comment. In order for the presence of an information link to
be sufficient for thought about an object, the subject must be able to locate the ob-
ject in space on the basis of the information received (ibid., p. 149). The idea of such
a disposition is broad enough to allow for the limit case in which there is a lag in the
receipt of information. In such a case, the subjects thinking nevertheless rests upon
an information-link such that his thinking [...] will be controlled by the infor-
mation yielded by the link if any [information] emerges (ibid., p. 153 n20). Such a
relationship (between informational input and behavioural output) is best specified
as functioning in independence from any description that the subject believes to
hold concerning that relationship. This is on pain of the object being presented by a
description such as the object being presented via such and such an information-
link. To capture this constraint, we will name such a function from information to
behaviour as non-descriptive. For a mechanism to be non-descriptive it is sufficient
that it is sub-personal, though this is not necessary. What is necessary is that the
process must not involve (or indeed be) a process of identification. And, finally, in
any case this process must determine the subjects conscious thoughts concerning
that object (ibid., p. 146 n8).
The result of such an information-link (and an appropriate set of dispositions)
would be that if a subject claimed that
a is F
on the basis of information received via this link, there can be no question of
whether it is a that her thought concerns.
Applying this model to bodily self-ascription requires that one be able to specify an
information link that meets the specified conditions.

In particular, the input-output
process must be:
(i). non-descriptive;
(ii). provide ongoing spatial information;
(iii). be proprietary to the object (the body); and,
(iv). determine the subjects conscious self-ascriptive behaviour.

| 247

Evans evidently thought that it would be possible to find information-links that met
such conditions:
it seems equally not to make sense for a subject to utter Someones
legs are crossed, but is it I whose legs are crossed?, when the first
component is expressive of knowledge which the subject has gained
about the position of his limbs, available to him in the normal way.
(1982, p. 216)
187

we are able to know our position, orientation, and relation to other
objects in the world on the basis of our perceptions of the world. In-
cluded here are such things as [...] knowing that one is moving in a
train by seeing the world slide by; knowing that there is a tree in
front of one, or to the right or left by seeing it; and so on. Once again,
none of the following utterances appears to make sense when the
first component expresses knowledge gained in this way [...] Some-
one is moving, but is it I?; Someone is standing in front of a tree, but
is it I? (1982, p. 222)
188

Amongst those who have been most explicit about BST, there is a recurrent worry
about conceptualising the bodily self as an object. Call this the body-as-object (BO)
problem. Worries about BO are expressed in a famous statement of Merleau-pontys
I observe external objects with my body, I handle them, examine
them, walk around them, but my body itself is a thing which I do not
observe: in order to do so, I should need the use of a second body
which itself would be unobservable. (1945/1962, p. 91)
Similarly Sartre remarks

187
Here the reference to ownership ought to be given the anti-Cartesian reading
broached in the last section.
188
There is a thorny issue that I am suppressing here. Evans thought that for an infor-
mational-link embodied in a perceptual system to yield consciousness, it must input to a
concept-using system. As he puts it
we arrive at conscious perceptual experience when sensory input is not only connected to
behavioural dispositions in the way I have been describing perhaps in some phylogenetically
more ancient part of the brain but also serves as input to a thinking, concept-applying, and
reasoning system; so that the subjects thoughts, plans, and deliberations are also systemati-
cally dependent on the informational properties of the input (Evans, 1982, p. 158)
To avoid taking a stance on the issue of whether or not this is the right way to distinguish
between perceptual content and non-conscious informational content, I invite the reader to
figure their preferred account into the dispositional background.
248 |
It is true that I see and touch my legs and my hands. Moreover noth-
ing prevents me from imagining an arrangement of the sense organs
such that a living being could see one of his eyes while the eye which
was seen was directing its glance upon the world. But it is to be noted
that in this case [...] I apprehend it as a sense organ constituted in the
world in a particular way, but I cannot see the seeing; that is, I can-
not apprehend it in the process of revealing an aspect of the world to
me. Either it is a thing among other things, or else it is that by which
things are revealed to me. But it cannot be both at the same time.
(1943/1992, p. 304)
Consider a similar worry in the present context. Construing the body as an object of
perceptual experience opens up the way to a Cartesian distance between the subject
and the body. Along these lines, Bill Brewer expresses concerns about a Cartesian
dualist strategy which
[] transposes Evans general account of perceptual-demonstrative
reference [] into an explanation of the immunity of certain bodily
self-ascriptions to error through misidentification. The basic idea
would be that [body] judgments like I am being prodded just above
the right knee are effectively of the form That knee is being prod-
ded, where the reference-fixing quasi-perceptual information-link is
provided by bodily experience. (1995, p. 297)
What the BST requires is some means of discerning an account of bodily-self ascrip-
tion that is both immune to error relative to the first person pronoun and insulated
from Cartesian riposte described by Brewer. A strategy like this would involve find-
ing an Evansian information link, and demonstrating that it can yield bodily-self
ascription on some non-perceptual basis. Shaun Gallagher, summarising the work
of Brian OShaughnessy (1980/2008, pp. 179 - 202; 2000, pp. 628 - 655) and the phe-
nomenological authors lately cited, marks such a perceptual/non-perceptual con-
trast as the difference between the experience of attending to something and the
experience of the (unattended) means of that attending:
I touch with my ngers, I have tactile experience, and I perceive the
shape of the object in my hand, precisely when I am not perceiving
my ngers. The tactile perception of an object is not accomplished
through my perceptual awareness of the changing spatial properties
of my ngers; awareness of my ngers is not equivalent to my tactile
awareness of the object (2003, p. 61)
In the terminology that we have developed in earlier chapters, the difference here is
none other than the difference between mere sensations and perceptual sensations.
In these terms, Gallaghers point is that in the act of perceptually experiencing
| 249

something, the body as an organ enabling that experience is itself only experienced
in mere sensation. With a similar sentiment Dan Zahavi declares that when he is
directed at and occupied with objects, his perceptual act and its bodily roots are
generally passed over in favor of the perceived, elaborating to say that his body
tends to efface itself on its way to its intentional goal (1999, p. 98). Another simple
way of describing this situation is to say that the body, as a subject, is non-
intentionally experienced. It is experienced but does not count as an object of expe-
rience in any sense.
What makes for the underlying tension is that the body is an object like many oth-
ers, such as other bodies of the same species. Along these lines Dorothe Legrand
(also inspired by the French phenomenological tradition) coins the term pre-
reflective immune bodily self-consciousness (2006, p. 102). She means two things
by this. Firstly she means to denote a non-perceptual experience of the body as a
subject. In particular, she wants to mark out kinaesthetic experience of the body as a
form of self-consciousness (Legrand, 2006, p. 95ff.). Secondly, the very point of
specifying self-consciousness of the body as a subject is to open up the possibility
that the body could be experienced in a manner that is immune to errors of identifi-
cation relative to the first-person pronoun.
Husserls concepts of Leib and kinaesthesia take on an obvious significance here.
Kinaesthetic sensations are the mere sensations that enable perceptual sensations,
and kinaesthesia is the anticipation of those sensations unfolding in broadly con-
tinuous manner. Der Leib is the bodily subject of experience, and in metaphysical
status, the bodily self. Connecting Leib and kinaesthesia renders the latter as an
ongoing source of self-consciousness in episodes of perceptually experiencing par-
ticular objects; it is anticipations of continuity in such self-consciousness that ena-
ble the requisite continuity in perceptual sensation.
Legrand is particularly sensitive to the BO problem, which she expresses as a tension
between (1) a description of the body as a physical object and (2) a determination
of a bodily form of self-consciousness that does not imply the identication of the
body as an object (Legrand, 2006, p. 96). But arguably the two are not ultimately
incompatible. In fact, an Evansian information-link underlying pre-reflective bodily
self-consciousness is precisely what might deliver (2). The first condition is not a
necessary condition for all accounts which would claim the BST. But something like
it is a necessary condition for any naturalistic account that hopes to give an account
of the biological mechanisms that realise an information-link that delivers (2). So
one must find mechanisms that do double duty; they must both enable a general
account of kinaesthesia and ground an Evansian information link.
250 |
As the reader might have surmised already, a defendant of BST taking this strategy
could readily appeal to some of the research reviewed in the last chapter. Take, for
instance, the internal proprioceptive systems (from 7.1) which might consist of
central predictive processes supplied by various information channels finding their
origin in muscle spindles, golgi tendon organs etc. These would fulfil the criteria as
follows:
(i). They are non-descriptive in that they are embedded in the deep structure
of the mammalian peripheral and central nervous system.
(ii). They would provide ongoing spatial information about the position of
limbs and their movement.
(iii). They are proprietary to the body, as they are enabled by dedicated afferent
channels combined with central processes running predictions about the
ongoing state of the very same body.
(iv). And collectively these mechanisms determine the subjects conscious self-
ascriptions of limb position and movement.
| 251

Moreover, the various Gibsonian proprioceptive sources (at the end of 7.1), includ-
ing agent-environmental affordances (at the beginning of 7.2), would fulfil the
criteria as follows:
(i). They are non-descriptive in that they are either sub-personal or instanti-
ated by relations that need not be descriptively identified.
(ii). They would provide ongoing spatial information about the location and
orientation of the body, its movement, and/or its possibilities and limita-
tions for action.
(iii). They are proprietary to the body, as they provide information about spe-
cific relations between that particular body (the body of the perceiver)
and the environment.
(iv). And collectively these mechanisms determine the subjects conscious self-
ascriptions of location, orientation, movement, and/or possibilities and
limitations for action in the environment.
8. 4. Being one body
If these claims hold, then with these resources BST looks to be on strong grounds.
However, now I want to consider various lines of scepticism concerning BST that do
not simply stem from a Cartesian account of selfhood.
To set the stage, consider some recent work on the development of brain-machine
interfaces. Carmena, et al. (2003) trained macaque monkeys to perform a virtual
reaching task involving a robot-manipulator arm, initially controlled by a joystick in
the monkeys hand. As monkeys learnt to perform the task they used recordings of
frontoparietal activity in conjunction with recordings of hand position, grip force
and hand velocity to generate predictive models that were later used to allow the
monkey to manipulate the robotic arm in a brain control mode. Eventually, in this
latter condition, the monkeys stopped moving their real arm in performing the task
(ibid., pp. 194 - 197). Extrapolate from this, add in corresponding somatosensory
feedback (the monkeys were controlling by visual feedback from a display), and well
trained forward and inverse models. Now imagine yourself in the situation. Fur-
thermore, stipulate that you have been anaesthetised and are otherwise disposed to
take the faux somatosensory feedback to be genuine. You try and move your arm,
somewhere a robotic arm moves, and you, via an information-link are disposed to
judge
My arm is moving
252 |
Is this judgement immune to error through misidentification relative to the first-
person pronoun? Evans has the material for an initial answer
we cannot think of the kinaesthetic and proprioceptive system as
gaining knowledge of truths about the condition of a body which
leaves the question of the identity of the body open. If the subject
does not know that [his arm has moved] on this basis (because he is
in the situation described), then he does not know anything on this
basis. (To judge that someone has his legs bent would be a wild shot
in the dark.) (1982, p. 221)
Evans point is that a hidden premise behind the question is that someones arm is
picked out by the statement, i.e. it is assumed that the expression my arm is mov-
ing is genuinely a judgement of some state of affairs. Evans denies this premise. But
this dodge misses the deeper point at issue, which is expressed by the more direct
question: what is it in the good case that makes it such that when someone express-
es such a thought, they pick out a part of their body as a part of themselves?
Quassim Cassam (1997) discusses alien limb cases with roughly this question in
mind.
189
In doing so he develops and rejects what he calls an idealist account of
body-ownership attributable to Locke. His developments of Lockes view make use
of OShaughnessys discussion of proprioceptively experienced body parts as being
immediately present. So I will introduce that briefly before turning to Cassam.
OShaughnessys phrase immediately present stems from his rejection of what he
calls a volition theory of action, a theory that moves from the claim that all action
is a producing by means of voliting to the conclusion that there can merely be a
difference of degree between voliting limb movement and the mysterious putative
voliting of movement in extra-bodily objects (1980/2008, p. 98). This allowance for
the conceptual possibility of voluntary movement of extra-bodily objects stems from
an ultimately Cartesian origin (cf. the volitional embodiment discussed in 8.1.), in
that it sees the relationship between willed movement and the particular body that
moves to be ultimately contingent. Now it is one thing to admit that the possibility
that in trying to move, ones trying might be a total failure. But this kind of contin-

189
Alien limbs in this context ought not to be confused with the phenomenon of
Alien-hand syndrome, a condition in which ones hand simply feels foreign, which Tony
Marcel (2006) usefully distinguishes from the Anarchic-hand sign, more akin to the strug-
gles exhibited by Peter Sellars character Dr. Strangelove. Marcel does seem to think that
cases of Anarchic-hand syndrome provide counter examples to immunity to error through
misidentification relative to the first-person pronoun (ibid., p. 81). He is usefully corrected
on this count by Christopher Peacocke (2006, p. 109),
| 253

gency comes with the assumption that one only tries to do what one sincerely be-
lieves is possible. This latter assumption rules out (at least in those that are rational)
the belief that one can will action upon non-bodily objects (as if by magic). The
question then is how to characterise the special relationship that holds between a
trying agent and its body, such that its body is deemed to be exclusively and pri-
marily the means of basic action. His solution is that the one wills and hence also
the limb in which that occurs must in a special sense [be] immediately present to
the willing subject (ibid., 1980/2008, p. 181). He soon spells out various conditions
on the immediate presence of a body part, two of which are of most interest; firstly,
one has feeling in it; and secondly, the part is for one a differentiated part of the
totality, ones body (ibid., 1980/2008, p. 182).
Return now to Cassams enriched idealist account of body-ownership. It consists in
the following two conditions being met:
(a*) to experience a limb or a body-part as a part of one is for it to be
immediately present to one, and (b*) the immediate presence to one
of a limb or body-part is necessary and sufficient for it to be a part of
one (Cassam, 1997, p. 65)
His main worry with (b*) is that it does not seem adequate to rule out the good case
from the bad case, namely, that ascriptions made on this basis
190

cannot be immune [to error through misidentification relative to the
first-person pronoun] because it is logically possible that the imme-
diately present limb actually belongs, say, to ones twin on the other
side of the room, and that it is immediately present simply because
ones brain is registering information from it. (ibid., p. 65)
However, his solution (for characterising the good case) seems rather confused. He
remarks that the failure of the idealist account is that it fails to respect the intuition
that a limb with which one is not materially united cannot be a part of one (ibid., p.
65). He then indicates that the importance of material unity in this case is spatial
proximity in virtue of local conjunction. But at the end of that very sentence he
footnotes a reference to Michael Ayers discussion of the importance of keeping ma-
terial unity and local conjunction conceptually distinct. Nevertheless, he ploughs on

190
Cassam does have another worry, namely that not all of our body parts seem to be
immediately present to us. However, this is hardly a concern when it comes to applying the
model to scenarios in which the subject is supposedly self-ascribing a property in virtue of
being disposed to respond to the outputs of an information-link. More plainly: a simple
response to his worry is that a bodily subject only self-ascribes livers, hearts and stomachs as
such on the basis of a description.
254 |
to suggest his own realist alternative: for a limb or a body-part to be a part of one,
it is necessary and sufficient that one is materially united with it (ibid., p. 66). He
continues by noting an interesting feature of the view. In its stress upon material
unity there is an important sense in which it is the bodily self rather than individu-
al body-parts which is the presented subject, which is to say that it is the material-
ly united whole which is, as it were, the ultimate presented subject (ibid., p. 66).
Perhaps there is something right about the material unity constraint. But even if so,
note that it gives rise to a serious instance of the BO problem. For what it seems to
require is simply that the bodily self as a materially united object is the ultimate
presented subject.
I am not sure how this issue could be resolved, but to tease out some further prob-
lems, consider another sci-fi scenario, originally designed by Tim Bayne.
191
I will let
Bayne recount the thought experiment for the most part. He asks us to imagine a
creature with three bodies and a single mind, called Borgy:
Borgys mind is (or is implemented in) his three brains [] Borgys
brains communicate by means of miniature radio transmitters, and
are functionally integrated to the same degree that our two neural
hemispheres are. Because he has three bodies, Borgy can be in three
non-contiguous places at once; he can sunbathe on Bondi beach in
Sydney, hike in an Arizonan desert, and have breakfast in a Parisian
caf all at the same time. Borgys perceptual experiences are struc-
tured around each of his three bodies: he sees a dog as standing thus-
and-so with respect to body A, he hears a car as moving away from
body B, and so on. (Bayne, 2004, p. 225)
One objection that might be raised (by a BST proponent) already at this point is that
Borgy is being described in manner that suggests he is distinct from his bodies. In
order to accommodate such a worry, I will dub his three bodies as Borgy A, Borgy B
and Borgy C respectively and modify the original text accordingly. With that said,
here is the real problem for BST:

191
Baynes concerns are admittedly slightly orthogonal to ours here. In his paper, he is
attempting to evaluate what kinds of constraints, if any, that self-consciousness might place
upon the unity of consciousness. The scenario to follow shortly was originally designed to
put pressure on what he calls the weak ecological constraint, which amounts to the claim
that co-conscious experiences cannot be ecologically structured around different ecological
selves if they are to be unified (2004, p. 223).
| 255

Although Borgy has three bodies and is thus either a scattered
physical object or a set of physical objects he is a single subject of
thought and action. (ibid., p. 225)
Even under a non-question-begging description of the scenario, a proponent of BST
seems forced to say that either the bodily self can be a single scattered physical ob-
ject, or it can be a set of physical objects. The former is incompatible with any reso-
lution to the alien limb issue that requires material unity. To see why the latter is
problematic consider a little more detail:
Borgys perceptual states feed into a unitary cognitive system, and he
has direct control over each of his three bodies (this, in part, is what
makes these three bodies his bodies). He acts with one (or more) of
his bodies in the way that you and I act with one (or more) of our
limbs. It is Borgy (and not Borgy A), that sees the tree that is visually
apparent to Borgy A; it is Borgy (and not Borgy B) that plays tennis
with Borgy B. Information derived from each of Borgys three bodies
is non-inferentially available via his unified cognitive system for
actions involving any one of his three bodies. (ibid., p. 225)
This kind of global availability is important for Baynes purposes, for his own pur-
pose in constructing the scenario is to evaluate the claim that co-conscious experi-
ences cannot be ecologically structured around different ecological selves if they are
to be unified. In short, Bayne wants to suggest that Borgys perceptual experiences
could be phenomenally unified despite the fact that they are structured around
distinct bodies (ibid., p. 225). Now although it is not the unity of consciousness per
se that we are interested in here, this is certainly a difficult case for BST. For if the
case at hand is coherent, then a single subject, Borgy, could instantiate various spa-
tial locations, and thus various perceptual perspectives. As Bayne claims:
Borgy experiences the world from three discontinuous spatial per-
spectives [] subsumed by a single phenomenal perspective. The
unity of this phenomenal perspective would enable him to integrate
the contents of states that occur in distinct perceptual fields. (ibid.,
pp. 225 - 226)
Borgy could, on the basis of the visual sensing of Borgy A, self-ascribe the property
of being in front of a tree. But would he be self-ascribing the property to himself as a
bodily self? If so then it would require that Borgy A be understood as simply a part
of Borgy, which leaves following dilemma for BST proponents: either they must pro-
vide an account of the unity of the body that can rule out cases of alien limbs (in
which case Borgy stands as a counter-example) or they must provide an account of
256 |
the unity of the body that can rule out Borgy (in which case alien limbs stand as a
counter-example).
192

8. 5. Being no body
I now turn to another source of problems for BST. The entire discussion so far has
assumed that selves exist, and that the main point of issue for BST is in defending
the claim that the self can be coherently conceived as bodily. However, there is a
famous line of dissent on the very first assumption. Call this the no-self thesis (NST):
Subjects of experience do not exist.
193

Prima facie NST is rather hard to swallow. How does one even take seriously being
told that what they take to be themselves does not exist? It seems rather strange to
even consider the idea, or even what such an idea might amount to. Naturally then,
when Metzinger asserts that: No such things as selves or subjects of experience
exist in the world (2003a, p. 577), he is often met with an incredulous stare. But, of
course, he is not the only one to hold a view on which the existence of the self is
denied. Indeed Varela, Thompson and Rosch claim that all of the reflective tradi-
tions in human history philosophy, science, psychoanalysis, religion, meditation
have challenged the naive sense of self. (1991, p. 59).
194
Still, there are a number of
things it is important to understand upon encountering versions of NST. Most im-
portant to understand is that NST ought to be aimed at theoretical definitions of

192
Here I have been raising the possibility of multiple bodies to a self, but there is a
further issue (not explored here, but equally problematic for BST) of the possibility of mul-
tiple selves to a body. For some discussion, see Dennett (1991, pp. 419 - 426).
193
This statement can be read as an ellipsis of a more complex statement. I keep the
simpler statement for the sake of generality.
194
It is important to separate no-self theses from the thesis that selves are elusive with
regard to perception. This latter thesis is expressed by David Hume, when he says:
For my part, when I enter most intimately into what I call myself, I always stumble on some
particular 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 (1739 - 1740/1960, Book I, Part IV, VI, p. 252)
It is perfectly consistent with realism about selfhood to claim that the self is something that
persistently evades our perception in some peculiar way (Howell, forthcoming).
| 257

selfhood, as (chiefly) contrasted with descriptive definitions. Ruth Garret Millikan
describes the contrast as follows:
Descriptive definitions are thought to describe marks that people ac-
tually attend to when applying terms. Conceptual analysts take
themselves to be attempting descriptive definitions. [...] A theoreti-
cal definition is the sort the scientist gives you in saying that water is
HOH, that gold is the element with atomic number 79 or that con-
sumption was, in reality, several varieties of respiratory disease, the
chief being tuberculosis, which is an infection caused by the bacte-
rium bacillus tuberculosis. (1989b, p. 291)
On this gloss, a theoretical definition is aimed giving an account of something in
the world in virtue of a theory about its underlying structure. I do not mean to rule
out the possibility that we can get at the structure of the world by analysing the
structure of our concepts. But a theoretical definition ought to stand as true or fall
as false quite apart from how we think. This is important, as NST is often posed as
either or both of the following:
(a). nothing exists that possesses the properties of selfhood folk psychologically
understood;
and/or
(b). nothing exists that possesses the properties of selfhood classically con-
ceived.
I use an inclusive-or as it is not always clear which of these is being advocated, if not
both. In claiming to illustrate (a) Metzinger provides the following examples:
I experience my leg subjectively as always having belonged to me; I
always experience my thoughts, my focal attention, and my emotions
as part of my own stream of consciousness; Voluntary acts are initiat-
ed by myself [...] I am someone; I experience myself as being identical
through time (2003a, p. 302, emphases original, quotation marks
removed)
And as regards (b), Albahari (forthcoming) lists said properties as follows:
258 |
Thinks thoughts, owns experiences, initiates actions.
The self exists prior to the thoughts, etc. it thinks the thoughts; it is
not constructed by the thoughts.
The thoughts, experiences and actions belong to the one subject so
the self is a focus of unity.
Every self is a unique and bounded individual, a thing with bounda-
ries that separate it from other things there could only be one of it
The self endures over time when you wake up, it the very same you
as the one who went to sleep.
The self observes many things but, although possessing phenome-
nal character, it cannot directly observe itself. It is curiously elusive in
this way.
(Albahari, forthcoming)
Many of these properties seem to overlap with those listed above as (a) properties
(cf. Metzinger, 2003a, p. 626). But whether or not the folk psychologically under-
stood and the classically conceived notions of selfhood ascribe different properties,
the key is that if they are to be worthwhile targets for NST, they must do so under a
theoretical definition. In short, it is only worthwhile for NST to claim (a) or (b) to
the extent that there is genuinely a theoretical definition which NST can demon-
strate to be false.
It is also important to distinguish NST from any kind of anti-realism about subjec-
tive experience. NSTs are often borne out of theories of consciousness that aim to
provide substantive accounts of subjective experience (cf. MacKenzie, 2008;
Metzinger, 2003b). They can thus accept the sentiment that it makes no sense to
say that the subjectivity of [...] experience is something that can be detached or iso-
lated from the experience, or to say that it is something it can lack (Zahavi, 2005, p.
126). The claim is that despite there being a compelling sense of self somehow in-
herent in the subjectivity of experience, this sense (it is argued) does not imply the
existence of anything real that one has a sense of. At their best, NST theorists are
more careful on this point than their opponents:
| 259

There is a distinct phenomenology of singularity, a non-sensory phe-
nomenology of thisnessfor example, in the phenomenology of
meditation, but also in bodily self-consciousness. If we look closely
enough, we can discover the phenomenology of primitive thisness in
our own subjective experience [...] But phenomenological structure
per se will never determine metaphysics (Metzinger, 2011)
That last sentence is the key point of issue, as a fan of selfhood is bound to respond
that:
If we were to wholeheartedly endorse such a restrictive metaphysical
principle, we would declare illusory most of the world we live in and
know and care about. Why not rather insist that the self is real if it
has experiential reality and that the validity of our account of the self
is to be measured by its ability to be faithful to our experience, by its
ability to capture and articulate (invariant) experiential structures?
(Zahavi, 2005, p. 128)
There are three points made here, and all help to illustrate the stalemate between
NST and this particular kind of opponent. Let us take them in reverse. Firstly, as we
just saw, proponents of NST do indeed have it as their aim to capture and articulate
(invariant) experiential structures.
195
But they will claim that these structures are
no more than appearances, they have no purchase on reality. Again, the slogan is
that phenomenological structure per se will never determine metaphysics. This
brings us to the second point. Zahavis suggestion is that the self is real if it has
experiential reality. If the suggestion is endorsed, this would be the claim that self-
consciousness is sufficient for the existence of a self (Zahavi, 2005, pp. 115 - 132).
More explicitly, on this view, the self is claimed to possess experiential reality, is
taken to be linked to the first-person perspective, and is, in fact, identified with the
very first-personal givenness of the experiential phenomena (Zahavi, 2005, p. 106).
Given that NST wishes to provide an account of the very same experiential phenom-
ena, to identify the self with such phenomena might seem like begging the ques-
tion. But then if all parties agree to the existence of the experiential phenomena,
and one party wishes to define the self as nothing over and above said phenomena,
it is hard to see how there is anything more than a semantic disagreement. Thirdly,
although the declaration that most of the world we live in and know and care
about is illusory might be seen as a motivation for NST, what is required is some-
thing even stronger. Certainly, it would be a small step from concluding that what

195
Witness e.g. Metzingers remark that what makes consciously experienced selfhood
special, and different from all the other forms of experiential content, is the fact that [...] it is
highly invariant. It is always there (2003a, p. 626).
260 |
we take to be reality is different from how it appears, to the claim that what we take
to be the self is different from how it appears. But even this is not quite enough for
NST, what would really need to be claimed is not that the self is an illusion, but ra-
ther more akin to an hallucination. For an illusion involves merely being mistaken
about the properties that are ascribable to something; it is only hallucination that
properly involves being mistaken as to the existence of something.
196

All of this suggests the importance of NST being targeted at a bona fide theoretical
definition. Identifying the self with an experiential dimension is likely no more than
a descriptive definition. As such it simply cannot be a target of NST, and it cuts no
ice against NST. Any dispute in this territory is bound to involve talking at cross-
purposes, certainly not a substantial debate.
But then what would count as a theoretical definition of selfhood? One way forward
would be to demonstrate that folk psychology (or perhaps, folk phenomenology)
genuinely constitutes a theoretical definition of self-hood and to work out in detail
what that theory is. As Stephen Stich remarks: we could hardly hope to show that a
theory is false (or to show that it is true) unless we have a reasonably accurate ac-
count of what the theory says (Stich, 1991, p. 236). However, the difficulties here are
likely to parallel those of a view which famously sought to deny the reality of mental
states by construing folk psychology as a theory (P. M. Churchland, 1981). The basic
problem is that the answer to questions such as
Do subjects of experience exist?
will either be indeterminate or idiosyncratic. This at least is Stichs thesis, and it
rests on the claim that any theory of how and whether terms refer will actually be no
more than a description of the psychology of those doing the referring (1991, p.
236ff.). The result is that the question boils down to a matter of intuition, and that
the answer either is indiscernible (i.e. statements such as x exists have no deter-

196
Metzinger is careful to make exactly this stronger claim: e.g. all selves are either
hallucinated (phenomenologically), or elements of inaccurate, reicatory phenomenological
descriptions; and later he thus points out that on his view there is no one whose illusion
the conscious self could be, no one who is confusing herself with anything (2003a, pp. 462,
634, emphases original).
| 261

minate truth-value) or trivial (i.e. statements such as x exists are only false accord-
ing to the idiosyncrasies of a restricted class of psychological subjects).
197

According to some proponents of the view, the most thoroughly worked out ver-
sions of NST ought to take it as their remit to explain why the thesis is counter-
intuitive (see Metzinger, 2011, 2). Metzingers explanation is strikingly close to the
sources of indeterminacy and triviality that Stich mentions. For instance, on his
view, NST is counter-intuitive in virtue of being a phenomenal impossibility
(Metzinger, 2003a, p. 626). Indeed, he defines intuitive plausibility as phenomenal
possibility, the latter being a property of every thought or idea which we can suc-
cessfully transform into the content of a coherent phenomenal simulation (ibid., p.
55). And phenomenal possibilities can be both indeterminate and idiosyncratic
The result of a certain thought experiment, say, of Swampman travel-
ing to Inverted Earth [...] may intuitively appear as plausible to us,
whereas a dream, in retrospect, may look bizarre [...] the reverse is
possible as well []
phenomenal possibility is always relative to a certain class of concrete
representational systems (ibid., p. 55)
In respect of this last point, it is no wonder that Metzinger confesses that the head-
line-grabbing thesis that selves do not exist is only an answer to the crude tradi-
tional metaphysics of selfhood, and that the concept of being no one (or in respect
to BST, the concept of being nobody) is a rather trivial one (2003a, p. 626). In a
slightly more optimistic vein, he closes a symposium on his (2003a) with the follow-
ing suggestion:
It may be more fruitful to investigate the representational and func-
tional deep structure of phenomenal self-consciousness from a truly
interdisciplinary perspective, to focus on systematic detail questions
and fine-grained, empirical bottom-up constraints, than to get lost in
a competition of theories about what the appropriate weak notion of
self is [...] Ontology and the big picture will eventually fall into
place all by themselves. (Metzinger, 2006b, p. 4)
There is a sentiment here that is exactly in line with the official stance of this disser-
tation: complete neutrality on selfhood, i.e. complete neutrality on the actual exist-

197
Note that a proponent of NST can claim that I is a referring expression. For on
their view, I might refer to conventionally defined entities that are not real (MacKenzie,
2008, pp. 253 - 254). But the possibility might still remain that the real referent of the term
might be indeterminate or idiosyncratic.
262 |
ence of subjects. Clearly Metzingers scepticism about the self stems from a deep
scepticism about whether phenomenology can determine metaphysics. And from
here, against certain other assumptions, he concludes that selves do not exist. I am
not embracing that conclusion, for I intend to remain neutral on the question. Part
of the motivation for this neutrality is simply to remain on topic. If the rest of this
dissertation were concerned with whether or not selves exist and whether or not
they are bodily, I would not be talking about the bodily nature of perceptual experi-
ence anymore, I would be talking about the metaphysics of selfhood. That is an
interesting question in its own right, but it ought to be treated elsewhere.
| 263

Chapter 9 Clearing the ground
9. 0. Introduction
I concluded the last chapter by noting that the specific issue of the existence of sub-
jects of experience (bodily or otherwise) is not a fruitful one to pursue. In this chap-
ter, I want to lay out the issues I do believe to be fruitful in pursuit of a naturalistic
explanation of how bodily experience might play a structuring role in perceptual
experience. The issues are all clustered around a central dispute between two views
on body representation. These views take a contrasting stance on the relationship
between properties of the body as a whole and the representation of the body. And
potentially, they take a contrasting stance on whether the body itself is indispensi-
ble in an explanation of how certain representational capacities are achieved.
I begin with an overview of a theory of subjective experience whose architect has
recently identified the specific relationship between representation of the body and
the most minimal form of subjectivity as a fruitful research target. I use the theory
as a foil for discerning certain general questions about the relationship between the
body and its representation, viz. the extent to which body representations are em-
bodied. In the next section, I outline two kinds of view, a local view and a global
view, that are distinguished by their contrasting stance on whether or not the body
is represented as a whole, and potentially also by their contrasting stance on wheth-
er or not body representation is embodied. I provide a further means for distin-
guishing variants of each view, and the differences between them, by developing
various forms of individualistic cognitive scientific explanation. In the last two sec-
tions, I argue for the virtues of certain shared assumptions about what explanation
implies in this context, and certain shared constraints with respect to positing rep-
resentations of and within a dynamical physical process.
9.1 Bodily self-models
9.2 Representing the body
9.3 Neural vs. non-neural individuals
9.4 Laws vs. mechanism
9.5 Representation vs. dynamics
264 |
9. 1. Bodily self -models
Since the publication of his magnum opus, Being no one (2003a), Metzinger has
turned to several detail questions. These have ranged from study of the representa-
tional, functional and neurobiological mechanisms involved in rudimentary social
cognition (Metzinger and Gallese, 2003), to corresponding questions concerning the
experience of volition (Metzinger, 2006a). But by far he has paid the most attention
to discerning the nature of the simplest form of self-consciousness, and its relation
to the body (Blanke & Metzinger, 2009; Lenggenhager, Tadi, Metzinger, & Blanke,
2007; Metzinger, 2005a; Metzinger, Lenggenhager, Tadi, & Blanke, 2007).

This shows
Metzinger keeping in lockstep with advances in contemporary science, specifically
the body representation boom that has occurred in just the last decade.
198
However,
this work is also in part an extension of ideas to be found in his (2003a). So, before
moving on to the detail questions, I will give a brief and somewhat superficial ac-
count of these ideas, largely relying on the words of Metzinger himself to minimise
errors of exposition.
Metzinger develops a multi-level theory of subjective experience as a complex repre-
sentational process, on the basis of three theoretical constructs: the phenomenal
self-model (PSM); the world-model (WM); and the phenomenal model of the inten-
tionality relation (PMIR). In particular, he urges that the PSM and PMIR will not
only be discovered by the sciences of mind, but furthermore, that forging a path to
their discovery is the only way in which a science of subjectivity can get any pur-
chase (Metzinger, 2003a, pp. 9, 427). Whilst the idea of a WM can be traced to a
general representationalist approach to consciousness, the PSM and the PMIR are
the distinctive posits of Metzingers theory, the self-model theory of subjectivity
(SMT). He provides a succinct definition of a self-model as follows:

198
As testament to the boom, several special issues have recently been published on
body representation and related topics: Neuropsychologia The Sense of Body (48, no. 3,
February 2010); Experimental Brain Research, "Body Representation" (203, no. 3, July 2010);
Cognitive Neuroscience, The Body in the Brain: Body Representations, Processes and Neural
Mechanisms" (to be published). And these come in addition to edited volumes and mono-
graphs on the same (Gallagher, 2005; Klatzky, MacWhinney, & Behrmann, 2008; Knblich,
Thornton, Grosjean, & Shiffrar, 2006).
| 265

[A] self-model is an integrated model of the very representational
system, which is currently activating it within itself, as a whole. Typi-
cally it will possess a bottom-up component driven by sensory input
(self-presentation). This input perturbs or modulates the incessant
activity of top-down processes continuously generating new hypoth-
eses about the current state of the system (self-simulation), thereby
arriving at a functionally more or less adequate internal image of the
systems overall, actual situation (self-representation). (Metzinger,
2005b, p. 18)
Metzinger uses the notion of a model (2003a), to describe perceptual experience in
representational terms, e.g., [i]f you now look at the book in your hands, you are
not aware of the highly complex neural process in your visual cortex, but of the con-
tent of a phenomenal mental model (Metzinger, 2003a, p. 22, and passim until p.
208 where the concept is officially introduced). But this is actually the product of
the development of several concepts of representation, a subset of which, phenom-
enal representations, have contents that characterise a rich variety of conscious
states. First there is the basic notion of mental/phenomenal representation, which
involves a three-place relationship between a representandum, represented by a rep-
resentatum (a time-slice of an information process) for an information processing-
system (ibid., pp. 21 - 30). Two other notions derive from this. The first is men-
tal/phenomenal presentation, the traditional philosophical notion of a quale ren-
dered scientifically respectable as a stimulus correlated internal state with working
memory constraints characteristic of early sensory processing (ibid., pp. 86 - 93).
The second is mental/phenomenal simulation, a virtual process of modelling poten-
tially measurable, observable states of a given target. Simulation is largely driven by
spontaneous activity of the system, and its function is to enable the planning, imag-
ining and/or dreaming of various possibilities (ibid., pp. 43 - 62). Phenomenal men-
tal models (inspired by Philip Johnson-Lairds work on mental models) bridge the
dreamt-up simulations with anchored presentations, as structures that integrate
presentational content. This process itself accords to a variety of constraints on how
the information carried must be transformed and processed to yield the variety of
conscious contents comprising the WM (ibid., pp. 117 - 211). To understand self-
modelling as a particular variety of self-representation requires a further conceptual
tool:
266 |
A simple and clear way to explicate the notion of self-representation
consists in introducing the notion of emulation while distinguishing
three cases. An information-processing system S can represent a
physical object in its environment. Call this process object represen-
tation. In some cases, the object component will be another infor-
mation-processing system: an information-processing system S
1
can
represent another information-processing system S
2
in its environ-
ment. Call this emulation. If S
1
[is identical to] S
2
, then we have a
case in which one information-processing system internally emulates
itself. Call this process self-representation. (ibid., p. 269)
Emulation is distinguished from simulation in just those cases in which a system
reproduces not only directly measurable behaviour of the target system but also the
more abstract processes that give rise to that behaviour. Although much of Metz-
ingers work could be related to material on many pages of this dissertation, the
concept of PSM, qua self-emulative self-representation, leads Metzinger to a range
of interesting conclusions that bear directly on discussions above. Bearing in mind
worries about the body as object, for instance, consider:
A self-modeling information-processing system internally and con-
tinuously simulates its own observable output as well as it emulates
abstract properties of its own internal information processingand
it does so for itself. [] In human beings, it is particularly interesting
to note how the self-model simultaneously treats the target system
as an object (e.g., by using proprioceptive feedback in internally
simulating ongoing bodily movements) and as a subject (e.g., by
emulating its own cognitive processing in a way that makes it availa-
ble for conscious access). This is what embodiment means []
(ibid., p. 301)
The breadth of the concept is such that it enables speculation on virtually every as-
pect of self-consciousness as philosophically conceived, where in each case this is
achieved by transforming that aspect into a principle or a constraint on the self-
modelling process. It also enables speculation on the bases of a number of patholog-
ical cases which ostensibly present missing or deviant partitions of the PSM, allow-
ing further differentiation of the concept. But more to the present point, it leads
Metzinger to make some striking claims about the nature of bodily experience:
| 267

You are never in contact with your own bodyas an embodied, con-
scious entity you are the contents of an image, a dynamical image
that constantly changes in a very high number of different dimen-
sions. However, this image is at the same time a physical part of your
body, as it invariably possesses a true neurobiological description.
(ibid., pp. 301 302)
The very process of modelling the states of the body in which the PSM is realised,
drives a conceptual distance between the output of the modelling process and its
presentational input. According to Metzinger, an almost inescapable result is a
special form of inner darkness which consists in the fact that the representational
character of the contents of self-consciousness is not accessible to subjective experi-
ence (ibid., p. 331). The result being that the self-model undergoing subjective ex-
perience is in a state of perpetual error, or as Metzinger puts it, a nave realistic self-
misunderstanding (ibid., p. 279).
The basis for this claim (i.e. the basis upon which self-consciousness is held to be
epistemically impoverished) is chiefly to be found in the endorsement of a distinc-
tion between phenomenal and intentional content.
The intentional content of the state in your head depends on the fact
of your hands really existing or on the state being a reliable instru-
ment for gaining knowledge about yourself in general. If the selfrep-
resentational vehicle [] currently is a good and reliably functioning
instrument of self-knowledge, then it allows you, precisely due to its
transparency, to directly look and feel through it, onto your own
hands. It makes the information carried by it globally available in an
integrated fashion without you having to care about how this actually
happens. (ibid, p. 334)
This idea, that knowledge is made possible by means unknown to the knower, is
exactly the kind of analysis that direct realists wish to resist. Nevertheless, Metz-
ingers reasons for making the claim are plain:
The phenomenal content of your hand representation, however, is
one component which stays invariant, regardless of whether your
hands exist or not []. It does not matter if all this is a dream or an
out-of-body experience, or if you are using two phantom limbs to
hold a hallucinated book right now, because the phenomenal content
of your bodily self-representation can in principle stay the same.
(ibid., p. 334)
Indeed, take any aspect of the intentionality of perceptual experience as conceived
by direct realists. With the full-blown SMT to hand Metzinger can both accommo-
268 |
date said aspect, whilst distinguishing between phenomenal and intentional con-
tent. This is precisely due to the final component of the SMT, the PMIR:
The central defining characteristic of phenomenal models of the in-
tentionality-relation is that they depict a certain relationship as cur-
rently holding between the system as a whole, as transparently repre-
sented to itself, and an object-component. [A PMIR would thus al-
low] a system to consciously experience itself as being not only a part
of the world, but of being fully immersed in it through a dense net-
work of causal, perceptual, cognitive, attentional, and agentive rela-
tions. (Metzinger, 2005, p. 5)
The content of a PMIR is a special form of phenomenal content, but phenomenal
content nonetheless. Accordingly, the PMIR could represent an intentional relation
to a concrete object even where no such relation obtained. One such case is that of
an out of body experience (OBE). Metzinger translates a report of one such case for
his readers:
I awoke at nightit must have been at about 3 a.m. and realized
that I was completely unable to move. I was absolutely certain I was
not dreaming, as I was enjoying full consciousness. Filled with fear
about my current condition, I had only one goal, namely to be able to
move my body again. I concentrated all my will-power and tried to
roll over to one side: Something rolled, but not my body some-
thing that was me, my whole consciousness including all of its sensa-
tions. I rolled onto the oor beside the bed. While this happened, I
did not feel bodiless, but as if my body consisted of a substance in
between the gaseous and the liquid state. To the present day, I have
never forgotten the combination of amazement and great surprise
that gripped me when I felt myself falling onto the oor, but without
the expected thud. Had the movement actually unfolded in my nor-
mal physical body, my head would have had to collide with the edge
of my bedside table. Lying on the oor, I was overcome by terrible
fear and panic. I knew that I possessed a body, and I only had one
great desire to be able to control it again. With a sudden jolt, I re-
gained control, without knowing how I managed to get back into it.
(Waetli, 1983 p. 25, as cited & translated by Metzinger 2003a, p. 491)
What Waelti describes through the medium of Metzinger is just one of many auto-
scopic phenomena: hallucinatory episodes in which a subject sees his or her own
body (Blanke & Mohr, 2005). With respect to seeing the whole body, these are clas-
sified under three types, The first of these are autoscopic out of body experiences,
such as described by Waelti, in which a subject reports having the feeling of being at
an elevated location some distance from what they believe to be their actual body,
| 269

whilst they view that body from that location (see Figure 25A). Another form is the
basic autoscopic hallucination, in which a subject reports seeing a double, a body
that looks like their own, but without accompanying distortions in their spatial rela-
tions to the seen environment (see Figure 25B). And a third is heutoscopy, in which
the subject reports oscillating between seeing a double and seeing a body which
they believe to be their actual body, changing their orientation respectively (see
Figure 25C).

Figure 25 Autoscopic phenomena

Cases such as these are of interest to Metzinger in his recent work because he be-
lieves they provide a window into understanding the simplest form of self-
consciousness, which he dubs minimal phenomenal selfhood (MPS) (Blanke & Metz-
inger, 2009, p. 7; Metzinger 2009b, p. 257). MPS is held to consist in a weak first
person perspective (weak 1PP) anchored by a global representation (GB-rep) of the
spatially situated body experienced as a whole. Although my focus is not on the
weak 1pp per se, some points I have made earlier have some bearing. So I will note
them briefly.
(A) Out of body experience
(B) Autoscopic hallucination
(C) Heutoscopy

See text for description.
All based on Blanke & Mohr (2005, pp. 186 187).
270 |
Subjective experience is often held to be perspectival experience. In a strong sense of
that expression, perspectival experience connotes several capacities: to conceive of
oneself as oneself, as distinct from others, and as distinct from the world that one
perceives; to be capable of entertaining a variety of attitudes to a variety of objects,
and to iterate that capacity as applied to ones own thoughts, states and relations,
and those of others (Baker, 1998, 2000; Metzinger, 2003a, pp. 395 400, 411 412
and passim; Newen & Vogeley, 2003) The notion of a strong 1PP (and each of its
various connotations) is a research target of a much higher complexity than MPS.
MPS is associated with a weak sense of the expression perspectival experience,
connoting merely that the immediate spatial contents of perceptual experience are
organised in an egocentric field. The specific notion of a weak 1PP that MPS is asso-
ciated with in Metzingers work is a certain phenomenological
D
claim, namely that
perceptual experience is perspectival; though it is thought to have functional impli-
cations. As Blanke & Metzinger put it: A weak 1PP is a purely geometrical feature of
a perceptual or imagined model of reality possessing a point of projection function-
ing as its origin in sensory and mental processing (2009, p. 7; cf. Metzinger 2009b,
p. 257).
As I discussed in 6.5, the spatiality of perceptual experience is typically understood
in terms of an egocentric frame of reference, not just by Husserl, but by several con-
temporary authors (e.g. Peacocke 1992, Campbell 1994, Cassam 1997). I argued
above that this understanding cannot reconcile the claim that the objects of percep-
tual experience are located by reference to a point of origin, and the claim that said
point of origin falls within the body of the perceiver. In the case of MPS, the corre-
sponding claim would be that the point of origin is functionally anchored in a rep-
resentation of the bodys global spatial properties. Although this is a rarely disputed
point, I believe that it likely rests upon the assumption that the spatiality of percep-
tual experience can be adequately conceived unimodally. It is an open question
whether, if that is wrong, that spatiality ought to then be conceived multi-modally
or supramodally. But in any case, I presented lines of argument above against the
unimodal conception. These were to the effect that the postulated anchoring of the
frame of reference would either be ad hoc, or that it would lead to claiming the pos-
sibility of there being mutually exclusive egocentric terms applicable in description
of one perceptual experience (i.e., both in front and to the right). I will not convey
the point further here; instead I will move on to discuss the other component of
MPS, GB-rep.
| 271

The notion of a GB-rep has some ambiguities that are not frequently noted.
199
My
purpose in the remainder of this section will be to explore these ambiguities by de-
veloping an opposing stance. The hope is that in doing so, I can both highlight the
future research issues that need to be better understood, and develop a novel con-
ception of the relationship between body representation and the body. Let me be
clear though, I do not intend this exercise to be seen as an attempted critique on
SMT; I intend to treat the claims of MPS in conceptual isolation. Indeed, what I
hoped to have expressed above is the deep complexity of SMT. Although, prima fa-
cie, the theory can be summarised in a fairly accessible fashion, its actual specifica-
tion confronts one with a plethora of technical devices that Metzinger uses to fur-
ther his theoretical aims. One of these is the distinction between phenomenal and
intentional content. Phenomenal content is phenomenology
M
with regard to this
distinction. Hence it is a distinction that is unavailable to me in the neutral stance
that I have taken on the metaphysics of content.
This makes a number of points of dispute somewhat orthogonal. In particular,
Metzingers variously placed claims to the effect that the experience of being a body
is epistemically unjustified will not be evaluated. I believe these claims validly fol-
low from the specific way in which phenomenal content is defined; hence evalua-
tion of the claim would lead to evaluation of the definition, which is not my remit.
Nevertheless, I do believe that an interesting point of issue can be raised with re-
spect to the extent to which explanations of body representational capacities require
appeal to their embodiment. That is a somewhat overused term, and in fact I only
have use of the term with respect to body representation, so let me stipulate a condi-
tion that is both necessary and sufficient for the embodiment of body representa-
tion:
A body representation is embodied iff. reference to the actual body (and/or
its properties and/or its relations) is indispensible in an explanation of how
the representation achieves its function
The sense in which reference to the body is held to indispensable is something I
hope to convey in the course of discussing the questions I want to focus on for the
remainder of this chapter. Those questions are these:
What would lead one to question the embodiment of body representation?
And how might one evaluate the embodiment of body representation?

199
One ambiguity already might be the term global representation. Although I am
adopting the term from Blanke & Metzinger (2009, pp. 10, 12), in the interest of disambigua-
tion I note what they clearly mean it convey: a global representation is a representation of
global properties of the body (cf. 2009, p. 9).
272 |
9. 2. Representing the body
The embodiment of body representation might naturally be questioned by appeal to
OBEs. Shortly, I hope it will be clear why it is this kind of case, and not cases in
which only specific parts of the body are illusorily experienced or are the objects of
hallucination, that is required to motivate the disembodiment of body representa-
tion. However, the question to ask first is this:
What justification is there in positing a GB-rep of the body in order to ex-
plain OBEs?
An initial worry might stem from a basic denial of the phenomenon, in suggesting
(perhaps) that the reports, such as the one translated by Metzinger that I quoted
above, are in fact confabulations. Confabulation is a perfectly general epistemic de-
ficiency, to which individuals are particularly liable when employing autobiograph-
ical memory (Hirstein, 2005, pp. 43 - 68, 177 - 212). But this line of thinking can be
readily halted through appraisal of the fact that out of body experiences can be re-
ported as they occur. For instance, Blanke et al. (2002) stimulated the right angular
gyrus of an epilepsy sufferers, who then reported of a visual experience of her body
from above. Although the authors do not convey the extent of the time-delay be-
tween stimulation and report, it is not a wild thought to assume that it occurred
fairly soon after.
In any case, the issue of whether the reports are valid and not corrupted by unrelia-
ble memory mechanisms is not one that I want to dispute. Rather, what I want to
resist the claim that the reported experiences themselves justify positing a GB-rep.
More specifically in fact, what I want to resist is the move from the claim that MPS
is associated with global, fully embodied self-consciousness and is experienced as a
single feature to the claim that there is a coherent representation of the whole,
spatially situated body (Blanke & Metzinger, 2009, p. 9). To some this might be
thought as just a matter of conceptual truth: if experiencing is representing, then
whatever is experienced is represented; if global properties are experienced, then
they must be represented. What I want to resist is the intuition that prompts the
question: How could it be any other way?
This might sound like I am setting up an attack on the very idea that the brain rep-
resents the body. I am not. In fact, what I am claiming is that there is no need to
employ a representation of the body as a whole when adequate representation of
parts will serve in its stead. Call this a local view on body representation. The local
view I have in mind is the most ambitious variant: a thoroughly general view on
which only representations of parts and their properties are required for any given
task in which representation of the body is required at all. Call these representa-
| 273

tions of parts LB-reps. The local view contrasts with any view that posits GB-reps.
Call such a view a global view on body representation. Rudimentary versions of the
global view come in at least two varieties. The first simply posits a GB-rep. The se-
cond posits both GB-reps and LB-reps.

Figure 26 - Global and local views on body representation

If the second of these simplified global views were the only formulation available,
and if the ambitious local view achieved its ambitious aims, then the local view
ought to be ruled the status of best explanation, in virtue of its parsimony relative to
the others. However, that is a piece of idle speculation at this point. I do not assume
that the two simplified global views exhaust the options in this area. Indeed, one
ought to leave the possibility of different kinds of GB-rep, differentiated by content,
format etc., posited for different kinds of task; and overall a rather more complex
and precise theory than either of these global views amount to. Moreover, an issue
Representational relations indicated by arrows: between GB-rep and global prop-
erties of the body by open arrow; between LB-rep and body parts and their prop-
erties shown by filled arrows. Potential causal interactions between LB-reps
shown by dotted lines.
(A) A global view that only posits a GB-rep.
(B) A global view that posits both LB-reps and a GB-rep.
(C) A local view that posits only LB-reps.
274 |
which I will not address is that of potential causal relationships between internal
representations of parts of the body and the implications this may or may not have
for positing GB-reps. I am focussed on the very first hurdle here: getting the discus-
sion going, in order to prompt further specification of possible global views, and
open up the possibility of exploring local views.
I ought to note, in the interest of disambiguation, that various manifestations of the
global view enjoy a fairly established and orthodox status. Two concepts, often con-
flated, are closely associated with the idea of a GB-rep: the body schema, and the
body image. Summarising his own and other comprehensive reviews, Shaun Gal-
lagher notes that in the literature of disciplines ranging from medicine to robotics
[] the schema or image of the body is alternately characterized as a
physiological functioning, a conscious model or mental representa-
tion, an unconscious image, a manner of organizing bodily experi-
ences, an artificially induced reflection, a collection of thoughts, feel-
ings, and memories, a set of objectively defined physical positions, a
neuronal map or cortical representation, and an ideation-
al/conceptual activity. (2005, p. 22)
And in light of all this he remarks
[...] it does not seem possible that the body schema/body image is at
once or even at different times all of these things. The variety of pro-
posals considered here suggests that, rather than one complex con-
cept with a plurality of names, there is, over and above the termino-
logical ambiguity, a basic conceptual confusion. (Gallagher, 2005, p.
22)
Despite this apparent confusion, it is still reasonable to say that the terms implicate
GB-reps for some (see Paillard, 2005 for a particularly clear case). This would be
roughly the kind of orthodoxy that a local view would serve to resist, by taking it as
an open question whether tasks that are claimed to require GB-reps actually do re-
quire them.
200

Here is some prima facie support for the local view. The local view claims that parts
of the body are represented in the nervous system. In fact, this is probably the most
secure claim to emerge from the body representation literature. In particular, the
local view can make appeal to the study of spatial representation of body parts. A
strong basis to begin with is the fact that single neuron recording studies have re-

200
For a state of the art review of the body image and body schema distinction, and
other potential taxonomies of body representation, see de Vignemont (2010).
| 275

peatedly documented both cortical and subcortical neural populations with multi-
modal receptor fields centred upon particular body parts. For example, Fogassi et al.
(1996) report that 56% of the cells in area F4 of the ventral premotor cortex have
both visual and somatosensory receptor fields centred upon the arms, face, neck,
and torso. And likewise, Graziano & Gross (1993) report that 20-30% of cells in area
7b of the posterior parietal cortex, and 24% of cells in the putamen, have similar
properties.
There is a range of psychophysical data that is convergent with the neurophysiologi-
cal data in this regard. For instance, Spence et al. (2004) asked subjects to hold mul-
tisensory stimulus cubes containing vibrotactile stimulators, placed where they
gripped the cube with their forefinger and thumb respectively, with LEDs visible at
respective forefinger and thumb locations. There were thus four possible stimulus
locations, for two possible stimulus modalities. While holding the cubes partici-
pants were required to fixate a central point situated midway between their two
hands, held equidistant from the midline. Their task was to determine the location
of a single tactile stimulus delivered either to the thumb or the forefinger of either
the left or right hand, while ignoring a visual distracter presented simultaneously at
one of the four locations. This setup reliably produced a crossmodal congruency
effect: when visual stimuli appeared at the same location as tactile stimuli, reaction
times were drastically faster and fewer tactile discrimination errors occurred (see
Spence et al., 2004). Again this suggests, at the very least, a dense integration of vis-
ual and tactile information concerning the relevant region.
The point is also illustrated by some excellent neuropsychological studies. Certain
patients with right-hemisphere lesions are able to detect individual stimuli (light
flashes, touches) on either side of their bodies, on or near their hands for instance.
But when presented with simultaneous stimuli on symmetrically opposite sides of
the body they are only able to detect a stimulus on their right (ipsilesional) side, a
phenomenon known as extinction. Extinction may occur within a particular modali-
ty, such as vision or touch known as unimodal extinction or it can occur across
modalities, such as vision and touch known as crossmodal extinction, in this par-
ticular case visuotactile extinction. Visuotactile extinction has been reported in case
studies involving visual stimuli presented in spatial regions proximal to the hands
(di Pellegrino, L{davas, & Farn, 1997), and also in regions surrounding the arms,
the face and the head (L{davas, 2002; Mattingley, Driver, Beschin, & Robertson,
1997). All of which suggests that the brain systematically integrates tactile infor-
mation, concerning the superficial surface and/or orientation of a particular body
part, with visual information concerning the space surrounding that part. Generally
speaking, bodily proximity is an important variable in these studies (roughly 70cm:
see L{davas, Farn, Zeloni & di Pellegrino (2000) for some investigation).
276 |
This is all only to show that the very idea that parts of the body are represented is
not without basis. A more general theory of representation is required to marshal
studies like these into the service of a local view. Indeed, the nature of the LB-reps
posited by an ambitious local view awaits further specification. But with the possi-
bility of LB-reps in mind, consider now the full-body illusion paradigm that is
standardly thought to involve manipulation of a GB-rep. In a series of studies Olaf
Blanke and his colleagues present participants with a virtual image of a full-size
body (or a body-sized cube in control conditions) being visibly stroked on the rear
surface of the torso (i.e., its back), whilst they are synchronously or asynchronously
stroked at the same location on their own torso (Aspell, Lenggenhager, & Blanke,
2009; Lenggenhager, Blanke, & Mouthon, 2009; Lenggenhager et al., 2007).
The aim of the paradigm is to introduce a conflict between vision and touch, in or-
der to induce a sense of ownership for the virtual body that is seen being touched
and to create an error of perceived location relative to the subjects actual location.
A questionnaire is used in order to measure the extent to which a sense of owner-
ship is induced. In the synchronous condition, answers are consistent with the claim
that subjects experience a sense of ownership for the virtual body that they see be-
ing stroked. The measurement of perceived location is more implicit. For instance,
in one study participants are blindfolded and passively walked backwards a few
steps, and then asked to return to what they belief to be their original location. In
the synchronous condition participants perceptual judgements of their original
location are significantly closer to the location of the virtual body than their actual
location during the experiment.
The difficulty here is that there is little assurance that this is a manipulation of a GB-
rep. There is no measure of whether, or even the extent to which the effect is local-
ised to particular parts. This is pertinent, because to claim that the illusion is a ma-
nipulation of a GB-rep implicates the assumption that it is not localised to particu-
lar parts (cf. Blanke & Metzinger, 2009, p. 9). The difficulty becomes vivid when
confronting recent work on the rubber-hand illusion (RHI), of which the full-body
illusion paradigm is a modification. As described in 7.4, in the RHI paradigm,
multi-sensory conflict is established by stroking a presented rubber or virtual hand
and synchronously stroking a participants own hand (hidden from sight). An im-
plicit measure of perceived location shift is conducted by passively moving the par-
ticipants hand (still hidden) and asking them to replace it to the original location.
Now, the point is not merely that the RHI paradigm plausibly studies LB-reps. The
point is that the RHI turns out to be (in some cases) not in fact an illusion localised
to the hand, but an illusion localised to the finger. With essentially these issues in
mind, Tsakiris, Prabhu & Haggard hypothesised that:
| 277

[If] RHI is entirely localised, then the perceptual shift for a non-
stimulated finger should be zero. If the RHI transfers fully from the
stimulated finger to other non-stimulated fingers, then the perceptu-
al shift should be equal for both fingers. (2006, p. 427)
And indeed, Tsakiris and Haggard (2005) and Tsakiris et al. (2006) found a pattern
of localised perceptual shifts, where
[] the perceived location of the stimulated` finger shifted towards
the video image location in the same way as in the original RHI [...]
when participants judged the proprioceptive position of another fin-
ger which had not been stimulated, the shifts were significantly re-
duced. (Tsakiris et al., 2006, p. 427)
The issue then is whether a similar pattern of partial localisation would be found in
the full-body illusion paradigm. The questions here are straightforward and likely to
be answered soon enough. For instance: if it were not the torso, but perhaps the arm
that were stroked, would the results of the self-localisation measure be the same or
different? And would there be the same or different reports illicited? Suffice to say
that until it can be incontrovertibly shown that the illusion does not localise to parts
of the body, talk of the fullness of the body illusion is misleading, and the name of
the paradigm a misnomer.
But actually, I think the issue between local and global views is more complicated
than the above suggests. For what I see to be the fundamental difference between
the two views is that the global view takes global properties of the body to be prop-
erties that need representing, whereas the local view (at least a potential local view)
takes them to be properties that can be exploited. The notion of exploitative repre-
sentation is owed to Robert A. Wilson. To convey the sense of exploitation here,
consider his example, in which he compares two ways of building a device for keep-
ing a tally of how far a car has travelled:
278 |
One way it could do this would be for the assumption that 1 rotation
= x meters to be part of its calculational machinery [] Another way
of achieving this end would be to be built simply to record x meters
for every rotation, thus exploiting the fact that 1 rotation = x meters.
In the first case it encodes a representational assumption, and uses
this to compute its output. In the second, it contains no such encod-
ing but instead uses an existing relationship between its structure
and the structure of the world (Wilson, 2004, p. 163)
201

Wilsons second device might not employ any representations at all. However, I am
suggesting that a representational system can represent aspects of its target and also
use a non-representational strategy along the lines of the second device Wilson de-
scribes. But in this case, it is one that uses an existing relationship within the holis-
tic structure of the body. Hence, the exploitable properties can be indicated with
reference to the rudimentary definition of a structural affordance from 7.2:
SA* A structural affordance is a relation between a bodily agent b and the prop-
erties I
P
and I
W
of parts of its body, in virtue of which a basic action is pos-
sible for b.
o I
P
is the property of being causally interactive with other parts x
1
, x
2
,
x
3
, ..., x
n
in virtue of their properties P
1
, P
2
, P
3
, P
n
.
o I
W
is the property of being causally interactive with the whole body
in virtue of its properties W
1
, W
2
, W
3
, W
n
.
The claim is that a system could, in each case in which it represents the body, repre-
sent individual parts of the body and either the property I
P
or some property P and
in doing so exploit the fact that those very same parts possess the property I
W
. Such
a system would employ the holistic structure of the body itself in achieving a given
task, precisely by not representing that structure as an interconnected whole, but
only in each case as a connected part. This would amount to an explanation of a
body representational capacity that required appeal to its embodiment. For in such
a case, the actual body and its properties and relations would be indispensable
terms in an explanation of how these capacities achieve their function. A con-
trasting global view would claim that a system requires a GB-rep of either some W
or the property I
W
to achieve certain tasks. If W properties are represented, then the
closest of the three views pictured is Figure 26A. If either the property I
W
is repre-
sented, or the property I
W
and some property W, then the closest is Figure 26B.

201
Cf. also the discussion of minimal memory strategies in Ballard, Hayhoe, Pook and
Ras (1997, p. 732)
| 279

9. 3. Neural vs. non-neural individuals
Similar ideas to that espoused in the local view detailed above are readily found.
This allows me to clarify what it would be to deny the embodiment of body repre-
sentation and claim that reference to the body is dispensable in an explanation of
certain representational capacities. It would be, in effect, to claim that structural
affordances present us with a case of causal spread, but not explanatory spread. Alt-
hough these two concepts will take a little unpacking, they are neatly summarised
by Matteo Mameli as follows:
Causal spread occurs when we discover some new factor causally in-
volved in the occurrence of a phenomenon [...] explanatory spread
occurs when we realize that some factor that was not taken to be part
of a sufficient explanation of a phenomenon needs to be included in
such explanation. (2005, p. 388 emphases modified)
So what is the relationship between the two? Mameli elaborates:
Since the fact that something is causally required does not entail that
it is also explanatorily required, causal spread does not necessarily
lead to explanatory spread. But in cases where the newly discovered
causal factor is deemed to be an important one, causal spread is likely
to generate the inclusion of the newly discovered factor in any suffi-
cient explanation of phenomenon to which this factor causally con-
tributes. That is, in these cases, causal spread leads to explanatory
spread (2005, p. 388 emphasis mine)
Causal spread is necessary for explanatory spread, but not sufficient. To achieve
sufficiency, it seems like importance is the difference that makes the difference. In
fact, the best way of grasping what is deemed to be important is to get a handle on
what the talk of spread amounts to. Mameli adopts the terms (causal and explana-
tory spread) from an article by Mike Wheeler & Andy Clark, where they attempt to
convey the purported significance of cases in which causal contributions made to
online adaptive success by environmental and (non-neural) bodily happenings are
on a par with those made by neural happenings (Wheeler & Clark, 1999, p.108). The
moral, they believe, is that in such cases no aspect of the extended brain-body-
environment causal system should be explanatorily privileged in the cognitive-
scientific understanding of the observed behaviour (ibid., p. 108 emphasis origi-
nal).
These claims come directly after an exegesis of Thelen & Smiths treatment of the
development of infant locomotion (see 7.3); and in particular, frequently occurring
rhetoric like the following:
280 |
Without a context, there is no essence of leg movements during the
first year. Leg coordination patterns are entirely situation dependent
[...] There is, likewise, no essence of locomotion either in the motor
cortex or in the spinal cord. Indeed, it would be equally credible to
assign the essence of walking to the treadmill than to the neural
structure, because it is the action of the treadmill that elicits the
most locomotor-like behaviour. (Thelen & Smith, 1994, pp. 16 - 17
emphasis original)
What Thelen & Smith are railing against is a certain style of explaining infant devel-
opment that dominated for much of the 20
th
century. In one form, this style mani-
fests as describing the infant as passing through stages of a genetically determined
developmental plan which unfolds as a consequence of brain maturation alone
(McGraw, 1940). In another form, this style manifests as describing the changes in
the infants endogenous activity (i.e. the disappearance of stepping) as a direct
consequence of global changes in the infant brains information processing capacity
(Zelazo, 1984). A common thread between these accounts, according to Thelen &
Smith, is the widespread and seductive, but ultimately illusory thought that be-
haviour can be reduced to an essential that is represented in some privileged form
within the organism (1994, p. 8). It is this rejection that Wheeler & Clark attempt to
express when they say that
This sharing-out of the explanatory weight call it explanatory
spread is manifestly at odds with cognitive scientific tradition,
which is to seek explanations of intelligent behaviour that appeal
fundamentally, and usually exclusively, to strictly agent-internal neu-
ral phenomena (Wheeler & Clark, 1999, p. 108)
A key point here, though not emphasised as such, is that it is only relative to an ap-
propriate contrast with cognitive scientific tradition that the causal nexus can be
seen as spreading. And as causal spread is necessary for explanatory spread, the
same goes for that too. Furthermore, what is regarded as important in the theoreti-
cal framework in which the spreading is identified must not have been identified as
important in the framework relative to which the spreading occurs. There are hints
of this in yet another variant of the spreading idiom, when Mike Wheeler has his
reader consider the nontrivial causal spread that occurs when causally spread fac-
tors reveal themselves to be at the root of some distinctive target feature of the
phenomenon of interest (Wheeler, 2005, p. 207 emphasis original). Explanatory
spread qua non-trivial causal spread, feeds off controversy, or at the very least con-
trast.
I want to move towards a minimal sense of rigour in identifying the relevant con-
trast between the global and local views mentioned. I think it would be best under-
| 281

stood as a dispute over the nature of the individual in a given cognitive scientific
explanation. That is, the issue can be fruitfully understood as a dispute between two
theories that are committed to individualistic cognitive scientific explanation of
some kind. Individualism in explanation of mind in general has seen a lot of discus-
sion. Wilson expresses the intuitions behind it as follows:
Where does the mind begin and end? We think of the mind as tied to
and delimited by individuals. Minds do not float free in the air or be-
long to larger, amorphous entities, such as groups, societies, or cul-
tures. No they are tightly coupled with individuals (2004, p. 3)
Individualism can have application with respect to the classification or the location
of mental states. With respect to classification, a famous instance is Jerry Fodors
recommendation that psychology ought to abstract away from details of the envi-
ronment in classifying mental states.
202
The classification thesis is usually (though
need not be) run together with a location thesis which assumes that states of mind
are located within the boundaries of some canonical object of scientific study. It is
the location thesis that is the more useful focus here.
There are three assumptions to individualism about location. The first is that there
is an individual: some well-defined entity, perhaps a clearly demarcated system
(Sterelny, 1990, pp. 82, 104 - 108). The second is that it is to the individual alone
which the properties of interest are attributable. These assumptions are of a very
general kind, and have application in various domains of scientific inquiry.
203
And to
the extent that these assumptions are normative, they employ the ought of good
practice, rather than a priori deduction, stemming from a history of influential sci-
entific research (Rupert, 2009, pp. 38 - 41; R. A. Wilson, 2004, pp. 40 - 44). A third
assumption is also usefully expressed by Wilson:

202
This is a large and complicated topic initiated originally by Fodors (1980) article.
Burge (1986) and Stich (1980, 1983) provide very different kinds of response, the former re-
jecting individualism, the latter accepting individualism to the detriment of the representa-
tional theory of mind. For some the debate has reached the stage of scholasticism, and
should be left to the history of philosophy in favour of a philosophy of mind qua cognitive
science (Chemero & Silberstein, 2008). In any case, see Kim (1996, pp. 184 - 201) for a nice
introduction and bibliography.
203
Another area where individualistic theses have seen rich discussion is in the philos-
ophy of biology. See Sterelny & Griffiths (1999, pp. 53 - 111) for an overview. See Wilson
(2005) for the advancement of anti-individualist theses in that context.
282 |
Minds exist inside individuals [...] we might quibble about how far
throughout the brain and central nervous system the mind extends
spatially. But again, the boundary of the mind is no greater than the
boundary of the individual (2004, p. 3)
Individualism then, is the sum of three assumptions:
(a). There is an individual i.
(b). Causal explanatory properties of kind x are exclusively attributable to i.
(c). Causal explanatory properties of kind x are spatially located within i.
The historical contingency underlying the first two assumptions can make the idea
of the boundaries of the individual (and thus the extent of its location) somewhat
fluid and perhaps accidental. Again, as regards individualism about the mind, Wil-
son characterises the received view as one of two options: If it doesnt stop further
in, in the brain, then it at least stops at the skin (2004, p. 3). Perhaps there are many
more varieties of individual than just these two. But these are the two that we will
consider. The same three options are possible for cognitive scientific explanation. In
fact, it is in this context that individualism about the mind has seen most discus-
sion. From now on I will restrict discussion to this arena. Here, very briefly, are
some examples of these three options from the recent literature.
An intracranial construal of the boundaries of cognition identifies the boundary
of the cognitive individual (the cognitive system) with the cranium (Adams &
Aizawa, 2001). The individual within those bounds is the brain. Call this a neural
individual. Global views such as briefly described above prima facie propose a neu-
ral individual. An alternative conception is to see the individual as individuated by
organismic boundaries (Rupert, 2004, p. 393). Call this an organismic individual.
The local view mentioned above proposes an organismic individual. Perhaps a use-
ful way of thinking about the difference between the two is provided by analogy
with the spinalised frogs described in 7.3. The task-specific flexibility enabled by
the (P) properties of these otherwise helpless creatures could motivate positing ei-
ther a spinal individual, or simply an organismic individual, depending on ones
theoretical proclivities. Analogously, tasks involving the W properties of an organ-
isms body could motivate positing either a neural or an organismic individual, de-
pending on ones penchant for exploitative representation.
The discussion has provided two ways of filling out assumption (a), corresponding
to two varieties of individual (see Figure 27).
| 283

Figure 27 Two varieties of individual

There are also, then, at least two rudimentary varieties of individualism at hand,
neural individualism and organismic individualism, where the latter constitutes
explanatory spread relative to the former. I call them rudimentary varieties in the
same vein that I called the global views above rudimentary, to indicate that they
could be and indeed ought to be more sophisticated, and complicated as a result.
Nevertheless, before saying why and how that would be so, I will briefly state the
two views. Reading b as the brain, neural individualism would be:
(a). There is an individual b.
(b). Causal explanatory properties of kind x are exclusively attributable to b.
(c). Causal explanatory properties of kind x are spatially located within b.
And reading o as the organism, organismic individualism would be:
(a). There is an individual o.
(b). Causal explanatory properties of kind x are exclusively attributable to o.
(c). Causal explanatory properties of kind x are spatially located within o.
But one might want to hold that things can get more complicated when filling out
assumptions (b) and (c). One reason for doing so would depend on the endorse-
ment of a particular view of realisation. Realisation is often understood as a deter-
minative relation running in the opposite direction to supervenience, i.e. from real-
isers to supervenients. Wilson (2004, pp. 100 - 143) complains that the standard
284 |
views of realisation (usually implicit in discussions of supervenience) conflate the
claim that realisers are metaphysically sufficient for the properties or states that
they realise with the claim that realisers of states or properties are wholly constitut-
ed by the individual that bears them. By contrast, on what Carl Gillett (2002) calls a
dimensioned-view of realisation, the realisers of properties can vary independently
of the bearers of the properties realised. Wilson develops a similar view, on which
the realisers of properties extend beyond the individual that possesses said proper-
ties in cases of what he calls context-sensitive realisation. If dimensioned accounts
of realisation such as Wilsons context-sensitive view are correct then one ought to
make room for answers to (b) & (c) to differ. One might deny (c), claiming that the
realisers of the properties in question extend beyond the bounds of the individual
identified in (a), to which the properties are attributed in (b).
One way of doing so, but preserving individualism about location, would be to call
for a distinction between strong and weak forms of individualism. A strong form has
all three assumptions line up together. A weak form replaces (c), to allow that one
might attribute the properties in question to some bearer located within the bound-
aries of a broader system identified in locating the properties in question. Hence,
weak individualism would require the possibility of replacing (c) with something
like the following:
(d). Realisers of properties of kind x are located within a system S
This would allow one to distinguish between the location of the individual to whom
the states or properties are attributed, and a potentially broader system within
which the individual is located. It would likely involve the claim that whilst the
properties belong to the individual, they do not in the same sense belong to the
system that is metaphysically sufficient for their realisation.
Alternatively, one might want to analyse individuals as systems. But then one might
still want to locate the properties of interest beyond the bounds of the individual
held to possess them, because the individuals boundaries (qua system) are some-
what irregular. To give a sense of how this might occur, here is a handy primer on a
form of systems analysis in which irregular boundaries of a system are somewhat
innocent:
First, a system is defined by its organizationthat is, the functional
relations among its elements. These relations cannot be changed
without changing the identity of the system. Next, systems can be
described as either facultative or obligate. Facultative systems are
temporary, organized for a particular occasion and disbanded readily.
Obligate systems, on the other hand, are more or less permanent, at
least relative to the lifetime of their parts (M. L. Wilson, 2002, p. 630)
| 285

Figure 28 Varieties of individualism

Although this is a distinct motivation for dissociating (b) and (c), I see it to be to the
same effect. A view on which the individual is facultative might well have trouble
locating the boundaries of the individual in order to locate the properties within it.
Obligate individuals would not present such trouble. For taxonomic purposes I treat
the latter as entity-like systems, equivalent to determinate regions of space that re-
main stable over time, or are clearly delineated at a given time. On a final analysis,
this stipulation might need to be rejected, along with the assumption that if locat-
Attributions can be made to any putative individual within the boundary indicated by the
attribution (att.) arrow. Locations are identified in reference to determinate spatiotem-
poral regions. Solid lines indicate obligate systems, dotted lines (on the right) indicate
facultative systems (see text).
286 |
ing x is possible, the location of x must be such a determinate spatio-temporal re-
gion (see Figure 28, right-hand side).

A last thing to note is that the taxonomy is fully intended to pay little to no heed to
the extant terminology that refers to cognition variously as embodied and/or em-
bedded and/or extended. These terms now have fully achieved the status of buzz-
words. They are hardly of much use in identifying viable theoretical options, by
means of which one can draw the basic contrasts required for marking explanatory
spread systematically. Having said that, here is a gesture at locating these terms
within the framework, moving from the easiest to locate to the hardest: Embedded
cognition, in the hands of Rob Rupert is most certainly a form of strong organismic
individualism (2004, 2009). Extended cognition in the hands of Andy Clark moti-
vates a new kind of individual, an extended individual, and espouses thus either a
strong or a weak extended individualism (2003, 2008). Embodied cognition is hard-
est to locate. The variety of uses of the term embodiment can place it anywhere
from weak neural individualism, to weak extended individualism (see Prinz, 2009
for useful suggestions; and M. L. Wilson, 2002 for a good review).
9. 4. Laws vs. mechanism
I have been talking about local and global views, and indeed, neural and organismic
individualisms, as contrasting approaches to cognitive scientific explanation. How-
ever, it is crucial that these approaches can agree upon what cognitive scientific ex-
planation should involve. I want to promote a view on scientific explanation in gen-
eral that has seen some fruitful developments recently. But first I will outline a ra-
ther more canonical model of explanation and its shortcomings.
Explanations are given in many different contexts. As Robert Wilson and Frank Keil
note, they are ubiquitous, pervading our activities from the most simple and mun-
dane (Just going to the store to buy some milk) to the most sophisticated and unu-
sual (2000, p. 87). Sophistication is certainly an aim of philosophical explanation,
and unusualness is often an outcome (perhaps even by design). But for all that,
philosophical work on explanation has tended to focus on scientific explanation. In
some form or other, such work stems back to the birth of natural philosophy in the
discussions of the Ancients.
204
But in the modern era, much of the discussion has

204
E.g. for a brief survey of Aristotles account of causal explanation see Psillos (2007,
pp. 100 - 102). And for a comparison of its similarities to the DN model (discussed in the
main text above), see Jeffrey (1971, pp. 19 - 20).
| 287

been held captive by the deductive-nomological (DN) model of explanation champi-
oned most notably by Carl Hempel. The DN model is structured around two basic
ideas. The first is that nature is characterised by regular patterns. The second is that
any natural event is sufficiently explained by demonstrating its occurrence as a logi-
cal consequence of one of these patterns. In slightly more detail, a DN explanation
involves two elements
Explanans: a sentence (or class of sentences) adduced to account for the
phenomenon
Explanandum: a sentence (or class of sentences) that describes some phe-
nomenon to be explained (Hempel & Oppenheim, 1948, pp. 136 - 137)
From here there is all the work to do. If (*) is to sufficiently explain () it ought to
meet certain further conditions. In its classic statement, DN provides several further
conditions that are indeed characteristic of the model. Firstly, any () must be a
logical consequence of the (*) and the sentences constituting (*) must be true. Ac-
cordingly, the ensuing explanation would be a sound deduction. Secondly, and cru-
cially, the sentence (or class) in (*) must do more than simply state antecedent
background conditions: it must state a law of nature in virtue of which the deduc-
tion is valid (ibid., p. 137). On a classical DN conception, natural laws are no more
than summaries of regularly observed patterns of events. Thus a summary of a
causal explanation looks like this:
If E describes a particular event, then the antecedent circumstances
described in the sentences C1, C2, ..., C
k
may be said jointly to cause
that event, in the sense that there are certain empirical regularities,
expressed by the laws L1, L2, ..., L
r
, which imply that whenever condi-
tions of the kind indicated by C1, C2, ..., C
k
occur, an event of the
kind described in E will take place. (Hempel & Oppenheim, 1948, p.
139)
205

These empirical regularities are assumed to pick out the form of nature, so scien-
tific explanation ought to be a rational reconstruction of the latter in the form of an
argument, i.e.:

205
There is a Humean strand that is implicit in Hempel & Oppenheim (1948) but nev-
ertheless present. Indeed, according to some the only justification the model can find is its
correspondence with an empiricist conception of causation, assumed to be antecedently
correct (e.g., Boyd, 1991, pp. 355 - 359). Though I recognise the (especially historical) im-
portance of concepts of causation in accounts of explanation, it would take us to far afield to
explore them in any depth. For more on Humes views and subsequent influence in this area,
see Psillos (2007, pp. 110 - 114).
288 |
L1, L2, ..., L
r

C1, C2, ..., C
k

___________________________________________________________________________________________________________
E
Thus if one follows the DN model, the real conceptual work in scientific explanation
is in giving a precise characterization of the logical relation between the Ls and Cs
on the one hand and E on the other (Kim, 1963, p. 286). If one has that, then one
can argue that the phenomenon is to be expected, given that the law holds true. In
respect of this basic advertence to laws in explaining phenomena, the DN model is
often called a covering law model of explanation; indeed, Hempel & Oppenheim
talk briefly of the subsumption of the explanandum under a set of laws (1948, p.
139). But as Robert Cummins complains, this is a potential source of problems for
the DN model:
No laws are explanatory in the sense required by DN. Laws simply tell
us what happens; they do not tell us why or how. Molire, satirizing
the appeals to occult properties and virtues, tweaks the doctors of
his time for explaining that opium puts people to sleep because it has
a dormitival virtue. But isnt this just what subsumption under law
always amounts to? (2000, p. 119)
A defender of DN might say that what Molires doctors refer to are non-explanatory
laws, and that the business of scientific explanation is in finding necessary and suf-
ficient conditions for laws to be genuinely explanatory. But if they did, they would
miss his point, which is simply that this cannot be done. In any case, what might be
of more interest is the fact that it is not being done. That is to say, it does not look
like scientists themselves are undertaking the task of finding explanatory laws, and
yet they seem to be engaging in fruitful explanation. Indeed, in psychophysics for
instance, where regular patterns (often labelled effects) are discovered, they are
explandanda rather than explanantia. Cummins underlines the point in reference to
a famous auditory illusion (McGurk & Macdonald, 1976):
no one thinks that the McGurk effect explains the data it subsumes.
No one not in the grip of the DN model would suppose that one
could explain why someone hears a consonant like the speaking
mouth appears to make by appeal to the McGurk effect. That just is
the McGurk effect (op. cit., p. 119)
Another respect in which the DN model might be in error is in the fact that it treats
explanation and prediction as entirely symmetrical. As Hempel & Oppenheim note
| 289

the same formal analysis, applies to scientific prediction as well as to
explanation. The difference between the two is of a pragmatic charac-
ter. If E is given, i.e. if we know that the phenomenon described by E
has occurred, and a suitable set of statements C1, C2, ..., C
k
, L1, L2, ...,
L
r
, is provided afterwards, we speak of an explanation of the phe-
nomenon in question. If the latter statements are given and E is de-
rived prior to the occurrence of the phenomenon it describes, we
speak of a prediction (1948, p. 138)
But explanation can occur without prediction and vice versa. Indeed, by the lights of
the DN model, if one has all the relevant Ls, one ought to be able to derive any given
E that is observed as a logical consequence, and thus provide the requisite explana-
tory argument. But without any Cs, it would be hard to predict which E would occur
and when. Hempel & Oppenheim seem wise to this, and thus they stipulate that an
explanation is not fully adequate unless its explanans, if taken account of in time,
could have served as a basis for predicting the phenomenon under consideration
(1948, p. 138). However, it is certainly possible to provide some explanation of a phe-
nomenon without predicting it. For a text-book case, consider the fact that a hori-
zontal layer of fluid heated from below will exhibit convection rolls at a certain
temperature. Here is an explanation: the temperature applied rises, naturally liquid
at the bottom of the layer becomes warmer than at the top; the cooler liquid (of a
greater density) begins to fall whilst the warmer liquid (of a lesser density) rises,
and the competing forces result in a rolling motion, ordering the elements of the
liquid on a macroscopic level into what are known as Bnard cells (see Figure 29).
This is a shallow explanation (but an explanation nonetheless) of what is often
called a pitchfork bifurcation in a self-organising system.
206
And, as Kelso relates,
these are full of surprises:

206
I borrow the term shallow explanation from Wilson & Keil (2000). They argue that
all explanations are shallow when unsupplemented by theory, as situation which they dub
the theoretical abyss (ibid., pp. 98 - 100). However, they argue that this is the norm in most
explanatory practice and develop an externalist account of explanatory knowledge:
Although we, as individuals, are faced with the theoretical abyss as the norm, the theoretical
knowledge we lack exists in others, just not in us [...] explanation and theories that under-
write its depth, are wide, that is, they do not supervene on an individuals intrinsic, physical
properties. (R. Wilson & Keil, 2000, p. 103)
I think this is right, but I cannot argue the point here, and it strays from the main issue in
any case.
290 |

Figure 29 A pitchfork bifurcation

If we look at the arrangement of the convection rolls we see that their
motion can be either in one direction or the other [...] Once this di-
rection of motion begins in a particular cell, it will not change. But
how does the fluid decide which way to go? The answer is Lady Luck
herself. (1995, p. 10)
Kelso is being a bit cryptic here, but what he means is that if one had a perfect sys-
tem it would be equally probable that it would go in one direction or the other
(1995, p. 10). If this is right, then it would be impossible to predict which direction
the rolls would take, whilst in either case we can nevertheless explain why the roll-
ing motion occurred.
Cumminss point above about the vacuity of the suggestion that the McGurk effect
explains the data it subsumes is a simple statement of the fact that prediction can
occur without explanation. One can predict that if I look like I am making a gaga
sound, and if I sound like I am making baba sound, then you will hear a dada
R represents the temperature gradient (control parameter). q
represents the state of the fluid (collective variable). Solid
lines and balls represent stable potential states, and the bro-
ken line and open ball represent unstable potential states. At
a critical point in temperature increase, the system moves
from stability (grey box left) to instability (grey box right-
middle), and can then move into a particular pattern of roll-
ing motion (right, grey boxes right top and bottom).
Adapted from Kelso (Kelso, 1995)
| 291

sound. That just is the McGurk effect. More generally, appraised of both L and C one
ought to be able to deduce and thus predict E. But that will be no more than a de-
scription of one instance of L (the McGurk effect), and it is just that that needs ex-
plaining. Accordingly, it is hard to see how the DN model would meet Hempel &
Oppenheims own dictum that an explanation ought [t]o explain the phenomena in
the world of our experience, to answer the question why? rather than only the
question what? (1948, p. 135).
But perhaps the most damning argument against the DN model, certainly from the
current perspective, is that it can license explanatory inferences that are trivial, due
to the inclusion of trivial premises in the inferential chain. Wesley Salmon provides
a classic example:
John Jones avoided becoming pregnant during the past year, for he
has taken his wife's birth control pills regularly, and every man who
regularly takes birth control pills avoids pregnancy. (1971, p. 34)
This can be reconstructed in a manner that makes the steps of the argument more
obvious, and more obviously in accordance with the DN model:
L: Every male (that regularly takes birth control pills) avoids pregnancy
C: John Jones is male (and has taken birth control pills regularly)
___________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ ____
E: John Jones avoids pregnancy
The parathentical statements indicate irrelevant laws and conditions. They point to
the deepest problem for DN, that of specifying which laws and conditions are rele-
vant to which phenomena. Indeed, the issue generalises beyond this rather obvious
case. Unless the DN model can provide principled conditions that rule cases like
these out, it will forever be plagued by the possibility of explanations that are more
subtly trivial, but trivial nonetheless.
Perhaps these criticisms are not fatal. In so far as they stem from a failure to ade-
quately describe scientific practice, they could probably be avoided by weakening
the aim of a philosophical account of scientific explanation. That is, the aim might
simply be to serve as a rational reconstruction of scientific explanation, a context
of justification distinct from the context of discovery (Reichenbach, 1938/1961, pp. 6
- 7). But with respect to adjudicating over disputes between rival theories, such as
that between local and global views, DN is irrevocably flawed. The flaw is in the fact
that DN explanations are arguments. For the status of an argument is wholly unaf-
fected by including any number of premises that are wholly irrelevant (whether or
not a law is included as a premise). For instance, Carl Craver (discussing his favour-
ite example) writes that it is perfectly legitimate to wonder
292 |
[] whether an action potential was caused by the injection of cur-
rent, or the injection of a volume of uid, or the physical disruption
of the membrane [all theoretical options at one point in the history
of studying the action potential]. Such considerations are part of the
reason why neuroscientists run controlled experiments. They try to
determine which factors make a difference and which do not. (2007,
p. 37)
It is certainly not clear how the DN model would allow one to thus adjudicate which
factors make a difference and which do not, and thus which of the competing ex-
planations offered by rival theories ought to be endorsed.
I now want to turn positive, to describe a model of explanation that I believe to be
promising and the most suitable for present purposes: the compositional mechanis-
tic model of explanation, roughly as developed by Craver (2007).
207
The CM model
does not appeal to the structure of arguments as an appropriate structure for an
explanation. Rather it appeals to the structure of mechanism. Certainly, mechanis-
tic explanation can take the form of a deductive argument. For instance, Descartes
thought that one could describe the earth and indeed the whole visible universe as
if it were a machine by appeal to principles which are known to all and admitted
by all, namely the shape, size, position and motion of particles of matter and in
doing so one would realise that
[] there is nothing in the whole of nature (nothing, that is, which
should be referred to purely corporeal causes, i.e. those devoid of
thought and mind) which is incapable of being deductively explained
on the basis of these selfsame principles. (1644/1985, p. 279)
208

However, a modern compositional mechanistic approach proceeds instead by ap-
peal to structures that are doubly hostage to empirical fortune, both as regards their
explanandum and their explanans. The explanandum of a mechanistic explanation
is what Carl Craver (2007) calls the explanandum phenomenon. For our purposes,
this can be judged as a holistic description of the behaviour of a mechanism. A
mechanistic explanation can fail at this level, simply by failing to pick out a real
phenomenon. For instance, Craver writes:

207
I am more confident that the compositional mechanistic model is suitable for my
purposes than that it is, generally speaking, the best of all the options. For a fuller defence of
a modern mechanistic view of explanation, see Craver (2007, pp. 28 - 62).
208
For a summary of Descartes views on mechanistic explanation, see Psillos (2007, pp.
102 - 105).
| 293

Descartes [...] believed that nerves are hollow conduits for the ow of
animal spirits, and that they activate muscles by inating them.
Starting with this idea, one would be led to search for mechanisms
that explain, for example, how nerves shunt the ow of animal spirits
into this nerve or that, how they activate muscles, and how light or
auditory stimuli impact upon this hydraulic machine (2007, p. 123)
209

A concomitant of this failure would be the individuation of a fictional set of ex-
planantia, viz. the components of the mechanism by means of which the behaviour
is to be explained. In this case these would be, e.g., components whose activity is
responsible for the inflation of muscles, or elastic sensory nerve fibres that open
cavities in the brain which influence the flow of animal spirits through the pineal
gland.
210

Figure 30 Cravers psi-phi model of mechanistic explanation

That the failure occurs at both levels is testament to the fact that the components of
a mechanism are mereologically related to its behaviour (as parts to the whole).
Slightly more formally, the behaviour of a mechanism S is explained by the pro-

209
Other related ways in which a mechanistic explanation can fail are by the mi-
characterising or only partially characterising the explanandum phenomenon (2007, pp. 123
- 128).
210
For a sympathetic reconstruction of Descartes views on mechanism, that provides a
reading on which the fictional components that Descartes posits are mere artefacts of his
time that do not bear upon the structure of the phenomena he identifies, see Wheeler (2005,
pp. 30 - 38).
From Craver (2007, p. 121)
294 |
ductive activities (
1
,
2
, ...
n
) of Ss components (X
1
, X
2
, ... X
m
) in producing
(Craver, 2007, p. 138 and passim) as depicted in Figure 30. Activity here means both
positive causal interactions such as activation and negative causal interactions such
as inhibition of a process. And the productivity of components activity indicates
that their causal interactions are potentially exploitable for manipulation and con-
trol (ibid., pp. 6, 63 - 106). In modelling or describing the mechanism, one will thus
be describing the causal relations between components in terms of their relation-
ships of manipulability, corresponding to the arrows in between smaller ellipses, in
the lower major ellipse in Figure 30. Furthermore, in order to ensure that a compo-
nent is genuinely a component i.e. genuinely a part of the mechanism both part
and whole ought to stand in a symmetrical relation of mutual manipulability. That
is to say that something like the following two conditionals should hold (Craver,
2007, pp. 155, 159):
211

If is set to the value 1 in an intervention, then takes on the value f(1)
If is set to the value 1 in an intervention, then takes on the value f(1).
Colloquially, Craver expresses the idea as follows:
a component is relevant to the behavior of a mechanism as a whole
when one can wiggle the behavior of the whole by wiggling the be-
havior of the component and one can wiggle the behavior of the
component by wiggling the behaviour as a whole. (Craver, 2007, p.
153)
Though the idea itself is simple, it is general enough to accommodate various kinds
of experimental intervention. Moreover, and this is the key, it is general enough to
provide a framework for both local and global views to investigate which compo-
nents are actually relevant to the mechanism that enables the representational ca-
pacities in question. If there is to be progress on the potential disputes between
these views, I believe that it will need to be conducted under the mutually agreed
assumption that explanation is mechanistic.

211
I say something like, as Cravers discussion is highly sensitive to both the condi-
tions relative to which an explanation is required, and the ideal conditions under which an
intervention is made (2007, pp. 153 - 160).
| 295

9. 5. Representation vs. dynamics
The dispute between global and local views is a dispute within the auspices of a rep-
resentationalist approach cognitive scientific explanation. I have in mind the theo-
retical construct of representation that plays a central explanatory role in artificial
intelligence, cognitive psychology, cognitive neuroscience, cognitive ethology, social
psychology and linguistics. Distinguish this from representationalism in the philos-
ophy of language, the view that the essence of language is its capacity to be directed
at the way things are. Also distinguish this from representationalism in the philoso-
phy of perception and consciousness (discussed in Chapters 4 & 5): the view that
the phenomenal character of an experience just is its representational content,
where that representational content can itself be understood and characterized
without appeal to phenomenal character (Block, 2007a, p. 611). Block calls this
representationism: again, this is not the view that is discussed here.
Representationalism (R), as discussed here, is the view that the behaviour of certain
systems is best explained as achieving certain crucial functions by computation over
representations. But given the fairly liberal use of the term (as indicated above),
perhaps Searle is right when he remarks: There is probably no more abused a term
in the history of philosophy than representation (1983, p. 11).
212
Accordingly, it
might be unsurprising that a number of authors that I have drawn upon in illustrat-
ing the empirical reality of structural affordances have been particularly reticent to

212
Representationalism has been around a while, though the ancestry of the term rep-
resentation is unclear, or at least there is some room for debate; see for instance:
Modern cognitive science was launched when the claim that cognitive processes are computa-
tional in character was annexed to the representational theory of mind. The latter doctrine
(which goes back at least as far as Platothe term idea was the precursor to the term rep-
resentation) is the view according to which mental states are, for the most part, conceived as
inner representational states (Wheeler, 2005, p. 6)
And compare:
The rst safely documented occurrence of the concept in the Western history of ideas can be
found in Cicero, who uses repraesentatio predominantly in his letters and speeches and less in
his philosophical writings. A Greek prototype of the Latin concept of repraesentatio, which
could be clearly denoted, does not exist. However, it seems as if all current semantic elements
of representation already appear in its Latin version. For the Romans repraesentare, in a
very literal sense, meant to bring something back into the present that had previously been
absent (Metzinger, 2003a, p. 20 n7).
So if you think that Platonic Ideas correspond to a viable concept of representation, then the
concept goes back at least as far as Platos time. If you do not, then Ciceros era is a good bet.
296 |
posit representations in their explanations. Thus, for instance, Thelen & Smith pose
the question: How does the organism continually adapt and create new solutions to
new problems? To which they respond: The answer we present [] makes no ap-
peal to the use of representations (Thelen & Smith, 1994, p. 42). What I believe a
local view should embrace is the idea that the intrinsic dynamics of the body (re-
viewed in 7.3) can be exploited in representation of the body. But what that re-
quires is a dynamic theory of representation, a theory of representation on which the
targets of a representation can be dynamical processes and the representations
themselves are sub-systems in a broader dynamical system including the target sys-
tems represented. This ought to be available to a global view also. But on such a view
the contrasting stance would be that every dynamical process relevant to the execu-
tion of a task must be represented. Again, in order for such a dispute to move for-
ward, a shared basis is required. But it is important to have a minimal degree of care
in positing representations of and within dynamical systems. So I urge the adoption
of a dynamical theory of representation somewhat tentatively, and in doing so I
hope to clarify some guidelines for positing representations at all.
As noted above, representation is a polysemic term, and R has been challenged in
several ways that correspond to the various notions of representation that have been
developed in the short history of cognitive science. However, in the mid-nineties,
Tim Van Gelder developed a line of criticism that he thought applied against R
across the board. He saw his arguments to be
[] not based on an unduly restrictive definition of the notion of rep-
resentation; they go through on pretty much any reasonable charac-
terization, based around a core idea of some state of a system which,
by virtue of some general representational scheme, stands in for
some further state of affairs, thereby enabling the system to behave
appropriately with respect to that state of affairs (Van Gelder, 1995, p.
351)
Van Gelder gets this core idea from some passing remarks made by John Hauge-
land (1992/1998, pp. 172 - 173).
213
Before I get to his arguments against R, I should
clarify certain aspects of Haugelands treatment. In fact, the most useful outcome

213
Haugeland devotes just two pages of a twenty-seven page piece to providing a few
sketchy and dogmatic remarks of a family resemblance concept of representation
(1992/1998, p. 172). Nevertheless, his remarks have been adopted by several authors includ-
ing Clark (1997), Grush (2003), Van Gelder (1995), and Wheeler (2005). Given the brevity of
Haugelands treatment, some commentators find this rather odd (e.g. Chemero, 2009, p. 55).
Though equally, it may indicate how little one needs to say if one says the right thing; an
admirable skill indeed.
| 297

Van Gelders critique results from an insight buried therein, and indeed it actually
shows how to accommodate his worries. The gist of Haugelands criteria is usefully
summarised by Clark, as saying that a system employs representations internal to
the system itself just in case:
(1) It must coordinate its behaviours with environmental features
that are not always reliably present to the system.
(2) It copes with such cases by having something else (in place of a
signal directly received from the environment) stand in and
guide behaviour in its stead.
(3) That something else is part of a more general representational
scheme that allows the standing in to occur systematically and
allows for a variety of related representational states. (1997, p.
144)
Collectively these are held to be sufficient conditions for representations. In fact, it
is claim (1) that does the real work here, and not simply because (2) and (3) indirect-
ly refer back to it. Rather (1) is essential for an instance of representation because it
is from this that it follows that system makes use of some state as a representation.
214

It is only this that allows an R style explanation to meet what William Ramsey calls
the job-description challenge: the challenge of explaining how a physical state actu-
ally fulfils the role of representing [...] accounting for the way something actually
serves as a representation in a cognitive system (2007, p. xv). Haugeland provides a
way of meeting the challenge. Filling out the details of (1) he draws a contrast to add
a little clarity:

214
Clark (1997, pp. 144 - 145) disagrees, arguing that the decouplability that (1) and (2)
require is too strong. (3) he thinks is enough, urging that a warranted ascription of an in-
stance of representation requires only that we confront some kind of encoding system
(Clark, 1997, p. 145 emphasis original), such as found, for example, in a pool of cells in the rat
hippocampus that respond to changes in head-direction (Taube, Muller, & Ranck, 1990). The
issue here is a microcosm of the broader issue discussed below. But let me say briefly that
the fact that the cells stand in systematic causal relations to other states in the rats neural
economy can be neatly captured by simply admitting that the cells are causal mediators. The
claim that the systematicity is in virtue of an encoding scheme is what does all the work
here. But as urged below, what really needs to be demonstrated is that the state is being used
or consumed for its informational properties, rather than its causal properties alone.
298 |
[Plants] that track the sun with their leaves neednt represent it or its
position, because the tracking can be guided directly by the sun it-
self. But if the relevant features are not always present (manifest),
then they can, at least in some cases, be represented; that is, some-
thing else can stand in for them, with the power to guide behavior in
their stead. That which stands in for something else in this way is a
representation; that which it stands in for is its content; and its stand-
ing in for that content is its representing it (Haugeland, 1992/1998, p.
172)
According to some, remarks like these imply an overly restrictive condition on rep-
resentation. I will come back to that. But first I return to Van Gelders line of argu-
ment. It hinges around his central example of James Watts solution to the engineer-
ing problem of regulating the emissions of a steam valve in order to keep the speed
of a steam-powered engine constant, despite variation in fuel and load (Van Gelder,
1995, pp. 349 - 350). His solution was a beautifully simple system that exploited me-
chanical coupling of centrifugal and gravitational forces, known as the Watt or cen-
trifugal governor (see Figure 31). It consisted of two spherical weights at the end of
two scissor arms connected to a central spindle via sliding collar. The spindle was
geared to an input shaft, at the base of which is a pulley, belted to a flywheel trans-
ferring power from the steam engine. The weighted arms were also connected to a
throttle valve modulating steam input to the engine. As the flywheel speed in-
creased, the weights flew upward (according to centrifugal forces), sliding the collar
downwards, closing the throttle valve, reducing steam input, reducing power to the
engine. As a consequence, flywheel rotation speed would decrease, spindle speed
would decrease, the weights would move downwards according to gravitational
forces, and the throttle valve would open. With the right balance of these compo-
nents at the right settings, the governor proved to be a highly effective solution.
| 299

Figure 31 James Watt's centrifugal governor

Van Gelder discusses another kind of solution to the governing problem inspired by
R. If one were to build a device, a computational governor, that solved the problem
by computation over representation, it would likely implement an algorithm similar
to the following:
1. Measure the speed of the flywheel.
2. Compare the actual speed against the desired speed.
3. If there is no discrepancy, return to step 1. Otherwise
a. measure the current steam pressure;
b. calculate the desired alteration in steam pressure;
c. calculate the necessary throttle valve adjustment.
4. Make the throttle valve adjustment.
Return to step 1. (Van Gelder, 1995, p. 348)
This solution has all the hallmarks of a classic R style explanation. It involves steps
of measurement, computation and resultant action. Corresponding to each of these
there would be subsystems which individually perform specific operations, and col-
lectively compose the regulating system. The subsystems would achieve their tasks
by creating tokens that stand in for, i.e., represent flywheel speed, steam pressure,
throttle valve adjustments etc. And to achieve the task, they would communicate by
passing said tokens in a sequential fashion, enabling the next subsystem to execute
300 |
the next stage in the process according to the rule of their cyclic organisation (Van
Gelder, 1995, pp. 348, 350 - 351).
215

Van Gelder argues that the fact that the engineering problem was not solved in this
way ought to give us pause for thought. When we confront the actual governor, we
do not find any components passing representations to perform the functions pro-
posed by the computational governor. Instead, he suggests, the key to explaining the
behaviour of the centrifugal governor is to understand the basic principles accord-
ing to which the arms angular velocity changes over time. As this depends upon the
rates of change in the speed of the flywheel, one needs to understand the rates of
change in the former as a function of the latter. However, the rates of change in
speed of the flywheel are a function of the rates of change in the arms angular ve-
locity. The regular behaviour arises due to the fact that the two are coupled to one
another and mutually modulate one anothers behaviour over time.
Van Gelder points out that a powerful way of further specifying such a closely cou-
pled relationship is by applying the tools of dynamical systems theory (DST). Certain
key concepts are of especial importance for expressing the rudimentary ideas, all of
which I have already presented in one form or another. Recall the concept (briefly
expressed in 9.3.) of a state space: the set of all the possible states that a system can
assume. The dimensions of this space are determined by its state variables. Bern-
steins problem arises because the state space for even the most rudimentary move-
ments made by real biological systems will involve so many variables (e.g. joint an-
gle rotations) as to make the state space extraordinarily high-dimensional. A third
key concept is that of a control parameter. The control parameters of a system drive
it through its state space. For instance, in the Rayleigh-Bernard convection (the be-
ginning of which I described in 9.4, see Figure 29), the temperature gradient is the
control parameter: when the control parameter reaches a certain critical value, the
system meets a bifurcation point at which it is unstable and disposed to transform
into one of two states. Another key concept is that of a collective variable. Collective
variables characterize the way in which global patterns of a system evolve over time.
Again, in the Rayleigh-Bernard convection, all the molecules of the liquid are drawn
into an ordered global pattern, and this behaviour can be described as a collective
variable of the system. A further concept is that of an attractor, a region of state
space towards which the trajectory of the system is disposed to evolve. The simple
robots built at Delft Technical University provide nice examples of attractors mani-
fest in the observable behaviour of a system. Ideally the balance between their com-
ponents is such that they implement a sustained oscillatory activity. This would be

215
For a useful discussion of this mode of explanation as generally applied in cognitive
psychology, see Haugeland (1981/1998 esp., pp. 17 - 28).
| 301

describable as a trajectory in state space, where that trajectory moves towards a pe-
riodic cycle around a particular point in that space, observable as a stable walking
pattern.
216

In a full analysis, these and other concepts are formally denoted as the elements and
products of differential equations that mathematically describe the evolution of the
system modelled. These equations of evolution will specify the rate of change at a
given moment of time of some variable (on one side of the equation) as a function
of the contemporaneous values of other variables and parameters (on the other
side). This analysis is highly general and can very readily be applied to the centrifu-
gal governor. If considered independently of its coupling with the throttle valve, on
one side of the equation there will be parameters and variables such as the speed of
the engine, the gravitational constant, the gearing constant, the friction constant
and the length of the spindle arm. And the correct calculation of these will describe
the instantaneous acceleration of the angular velocity of the spindle arms. In the
regulation of engine speed the governor is connected to the throttle valve, modulat-
ing the engines speed, and thereby modulating the angular velocity of the arms.
Accordingly, engine speed will be a parameter in a rather more complex equation
determining the rate at which the governor changes its state. A specification of the
engines speed will also involve a number of further variables, including the setting
of the throttle valve, the parameter of which is the angular velocity of spindle arms.
Hence, angular velocity is a variable in one equation in which engine speed is a pa-
rameter, and engine speed is a variable in another equation in which angular veloci-
ty is a parameter (Van Gelder, 1995, pp. 355 - 357).
So how much ice does a dynamical systems analysis of the centrifugal governor cut
against R? Van Gelder sees it to be a noteworthy fact about standard explanations
of how the governor works is [...] that they never talk about representations. Citing
his own account, and classic and contemporary treatments of the governor as in-
stances of non-representational explanation, he claims that the reason why repre-
sentations are absent in the discussion, is simply because the governor contains no
representations (Van Gelder, 1995, p. 352). And later he elaborates to say that be-
cause the arm angle and engine speed are at all times both determined by, and de-
termining, each others behaviour their relationship is much more subtle and
complex than can be adequately explained by the notion of representation. Hence,
when we fully understand the relationship between engine speed and arm angle,

216
For useful introduction to these concepts (though not the formal mathematical
analysis itself ), see part I of Abraham & Shaw (1992). Also see Thelen & Smith (1994, pp. 45 -
69) for an overview of applications and further references.
302 |
we see that the notion of representation is just the wrong sort of conceptual tool to
apply (Van Gelder, 1995, p. 353).
However, some certainly do not see that the notion of representation is misplaced
here. For instance, William Bechtel writes:
The fact that the angle of the spindle arms represents the speed of
the flywheel becomes more clear when we consider why it was in-
serted into the mechanism to begin with. The flywheel turns at a
specific speed, but there is no way to use this directly to open and
close the valve. The spindle and arms were inserted so as to encode
information about the speed in a format that could be used by the
valve opening mechanism (1998, p. 305)
And speaking to the lack of representation-talk that Van Gelder highlights, he offers
the following diagnosis:
The reason no one has to comment explicitly on the fact that the arm
angles stand in for and thus represent the speed of the flywheel is
that this system is very simple, and most people see the connection
directly. But if someone does not understand how the governor
works, the first thing one would draw attention to is how the spindle
arm angle registers the speed of the flywheel (Bechtel, 1998, p. 305)
Bechtel leans on a weak-point in Van Gelders discussion. Van Gelders implication
is that the centrifugal governor cannot be given the kind of componential analysis
that an explanation of the computational governors capacities must trade in. But in
fact, his first explanation of how the governor works (and my own, above) trades in
rather simple engineering talk of causal relations between the flywheel, the spindle,
the arms, the collar, the throttle valve etc. Although modelling the system by means
of DST provides a more precise and detailed description of the governor, the ex-
planatory payoff is not that large. As William Ramsey remarks, had [Van Gelder]
offered only the DST account of the governors operations, I suspect most readers
would be scratching their heads wondering how the system actually works (2007,
p. 210).
Van Gelder does certainly allow for the possibility that elements of a dynamical sys-
tem could be called representations if one so wishes. Indeed, in line with the senti-
ments of Dennett (1987), it should be noted that it is always possible to describe
something as a representation. But doing so might not be worth much. Here Van
Gelder offers a simple way of avoiding vacuous triviality (or indeed complication):
| 303

A useful criterion of representation a reliable way of telling whether
a system contains them or not is to ask whether there is any explan-
atory utility in describing the system in representational terms. If you
really can make substantially more sense of how a system works by
concretely describing various identifiable parts or aspects of it as rep-
resentations [...] that is the best evidence you could have that the sys-
tem really does contain representations (1995, p. 352)
Bechtels description does not meet this simple criterion. To show as much, I need
to outline a little more detail on what his description is. The basics of his account
draw upon insights made by Dretske (1981) and Millikan (1993).
217
What Bechtel
takes from Dretske is the claim that any effect can carry information about (or more
strongly, can be about) its cause provided that it is nomically dependent upon that
cause.
218
In more deflationary terms, information is cheap and ubiquitous. But in
more optimistic terms (at least for the naturalist), there is something in nature (not
merely in the minds that struggle to comprehend nature), some objective, observer-
independent fact or set of facts, that forms the basis of one thing's meaning or indi-
cating something about another (Dretske, 1988, p. 58). Bechtel (1998, p. 298; 2001,
p. 336) does not yet take this to be sufficient for representation, he draws Millikan
for a supplementary analysis. Though he is somewhat selective about what he takes
from Millikans view, important elements of which are succinctly summarised by
Tony Chemero as follows:
A feature R
0
of a system S is a Representation for S if and only if:
(R1) R
0
stands between a representation producer P and a representa-
tion consumer C that have been [designed] to fit one another.
(R2) R
0
has as its function to adapt the representation consumer C to
some aspect A
0
of the environment, in particular by leading S to be-
have appropriately with respect to A
0
, even when A
0
is not the case.
(R3) There are (in addition to R
0
) transformations of R
0
, R
1
... R
n
, that
have as their function to adapt the representation consumer C to cor-
responding transformations of A
0
, A
1
... A
n
(2009, p. 54)

217
See Bechtel (1998, 2001). Bechtel (2008, pp. 159 - 200) contains all the core elements
of his previous treatments, as well as a discussion of the notion of representation more gen-
erally. All contain (especially the latter) a typically rich treatment of the development of the
concept of representation in the history of studying the sensorimotor cortex. However, to
recount these virtues would (sadly) be redundant here.
218
See Dretske (1981, pp. 63 - 82).
304 |
Bechtel takes two analytic requirements from this. The first is that representation
requires a triadic analysis.
219
Something, such as a Dretskian indication, becomes a
representation because of how it is used by a system in controlling its behaviour
(Bechtel, 2001, p. 336). The second is that representation requires a teleological
analysis. It is because a system is designed to use a state to inform it about some-
thing distal that makes the state a representation (ibid., 2001, p. 336). Putting the
pieces together then:
Representations are found only in those systems in which there is
another process (a consumer) designed to use the representation in
generating its behaviour (in the simplest cases, to coordinate its be-
haviour with respect to what is represented) and where the fact that
the representation carried that information was the reason the down-
line process was designed to rely on the representation (Bechtel,
2001, p. 337)
And these principles are then applied to the operation of the centrifugal governor
(see Figure 32).

Figure 32 Bechtel's representational governor

219
The triadic analysis was originally advanced by Charles Sanders Peirce (1839 1914).
For a good discussion see Von Eckardt (1993, pp. 145 - 160).
A schematic description of the spindle arms representing the
speed of the flywheel for the valve.
From Bechtel (2008, p. 191)
| 305

With this description in mind Bechtel muses:
Someone examining Watts invention might ask: What do the spindle
arms do in the governor? The answer would be: They represent in-
formation as to whether the flywheel is moving too fast or too slow
and pass this information to the valve that controls the steam flow.
(2008, p. 342)
Now to the problems, all of which follow from a lack of respect for the job-
description challenge.
220
What Bechtel provides is at most several necessary condi-
tions for representation, each of which he applies to the description of the governor.
But what is missing is any reason for thinking that these conditions are even jointly
sufficient. Essentially there are three conditions involved in Bechtels account. One
is that a state carries information about another state in virtue of being nomically
dependent upon it. Although Bechtel does not explicitly trade on the representa-
tional connotations of information talk, he does so implicitly in describing the
spindle arms as representing the flywheel speed, and then passing this infor-
mation on to the valve. But this adds nothing to the original non-DST, non-R de-
scription. The real answer to what the spindle arms do in the governor is that they
cause the throttle valve to either open or close. They might carry information, but it
is a further issue as to whether that information is being used as such. For instance,
perhaps Nelsons Column in Londons Trafalgar Square carries information about
(and thus represents) the life-achievements and glorious death of the naval hero.
Perhaps the direction he faces might well indicate (represent) South, to the lost
traveller, or the cardinal direction in which Nelsons last flagship HMS Victory lies,
to a lover of trivia. But none of this is demonstrated by me putting the column to
use as shade from the sun on a hot summers day.
This brings me to the second and third conditions. It is not sufficient for representa-
tion that a nomically dependent state be selected for its nomic dependence. A state
could be selected for its function as a nomically dependent causal mediator alone.
What is supposed to do the extra work here is a consumer putting the state to use.
However, again, the state could be (and, in the case of the governor, actually is) put
to use for any number of its non-representational properties. To see the point, we
can adapt the example above (adopted from Ramsey (2007, p. 136)). Suppose my
grandfather wants shade in his garden at around afternoon-tea time. Instead of
building a shade, he plants a tree whose branches are full of bushy and easily-
pruned leaves. By paying attention to the location of the sun in the sky, and pruning
the branches appropriately, he has the right length and breadth of shade at the right

220
This paragraph and the next are inspired by Ramsey (2007, pp. 118 - 150, 210 - 217).
306 |
time. Sadly, a fire spreads from the neighbours house, destroying the tree. But never
mind, using seeds from the first, he plants another tree of the same species in the
same place. This tree, like the first, has been selected because of the height it grows
to and the manageability of its branches. It has been selected for a certain job, to
provide shade. Now, a glance at the trees shadow could represent the fact that it is
(or is not) tea time. Indeed, one could consume the shadow to tell any time of day.
But that is not the use to which it is being put. The point, of course, is that represen-
tations must be put to use qua representations; they must be put to use for their
representational properties; they must meet the job-description challenge.
Bechtels representational description adds nothing to a non-representational de-
scription. Worse still, it fails to actually identify a state that is unambiguously a rep-
resentation. As Ramsey points out the angle of the spindle arms stands in for the
speed of the engine as much as the angle of their joints does. As does the speed of
the spindle, the flywheels rotation speed, the height of the sliding collar etc.
All of these elements go into states that depend upon the engine
speed, and all of these are specifically designed to have a causal in-
fluence on the valve flap. In fact, many of these elements would ac-
tually qualify as playing a triple role, serving as a state that is repre-
sented, a representation of some other state, and a representation us-
er, all at once. (Ramsey, 2007, p. 212)
Before evincing some positive considerations concerning the sufficient conditions
for representation, I want to draw some general lessons here. The beginning of one
is appositely expressed by Grush when he writes:
It is one thing to have a system such that representational explana-
tions of its behavior are natural and fruitful, and then be in a position
of trying to provide a theory of representation in order to make sense
of such explanations. It is quite another when there is a system whose
operation admits of fruitful explanation without representation talk,
and then to grasp at straws concerning how one might be able to fi-
nagle a representational analysis of some of its states anyway because
one is ideologically committed to the idea that all problem solving
requires representation. (2003, p. 66)
The same point should be well taken for DST also. Ideology should not lead theory.
But the more general point, at least as concerns representation, is that the issue is
not so much whether a system does represent at all, or whether it could represent.
The point is whether, on a case by case basis, one is justified in identifying a particu-
lar state as a representation, or in explaining a particular instance of behaviour by
positing a representation. Every representation posited must have an explanatory
| 307

payoff for the case in hand. Call this the modesty constraint. Whilst the job-
description challenge serves as a means to evaluate whether a state or process is ac-
tually functioning as a representation, the modesty constraint serves to curb the
positing of representations in the first place. It is an axiological constraint, similar to
the more general value of parsimony, in that it dictates that caution against positing
unjustified representations is a virtue.
Some insightful thoughts along the lines of the modesty constraint are expressed by
Brian Cantwell Smith in his (1996) discussion of inscription errors or instances of
pre-emptive registration. The error comes in both a metaphysical and a methodo-
logical version. The second is closest to my concern, but as the methodological ver-
sion apes the metaphysical, I will cover that first. It is
[] a tendency for a theorist or observer, first, to write or project or
impose a set of ontological assumptions onto a computational system
(onto the system itself, onto the task-domain, onto the relation be-
tween the two, and so forth) and then, second, to read those assump-
tions or their consequences back off the system, as if that constituted
an independent ontological discovery or theoretical result (B. C.
Smith, 1996, p. 50)
As Smith has put it (in conversation, see Clapin, 2002, pp. 266 - 267), the concern is
any piggy-backing of assumptions about representations on assumptions about
object identity (see also Smith, B.C., 1996, pp. 50 - 54). Hence, in metaphysical ver-
sion, the real gripe is with assuming object identity.
221
However, despite these very
general worries, Smith does license a shred of inscribing practice, when he says in a
footnote that inscription per se is not an error. In any situation, more or less writ-
ing on the world is invariably necessary. This opens the way for a more methodo-
logical take on the error, which is then a failure to recognize a particularly blatant
dependence of a claim or result on the original inscription, especially when that
inscription is not appropriately warranted e.g., in a case of chauvinism, projection,
or naivet (ibid., p. 50). Thus, whilst accepting that we may inevitably commit the
metaphysical inscription error, there might still be some hope that we can avoid its
methodological ancestor.

221
As Dennett describes: The Grandaddy case of pre-emptive registration is imagining
we can parse the universe primordially into objects (2002a, p. 225). The difficulty is that the
token object assumed to have been identified at one point and reidentified at another, may
not have been. It might equally be the reinstantiation of some type. Indeed, it might be that
any project to discern individuals is doomed to grim failure, as [t]here may not be any com-
pelling reason to believe that there is even a metaphysical fact of the matter (Smith, B.C,
1996, p. 55).
308 |
To give some examples of the ancestor, witness Drew McDermott engaging in some
friendly ridicule of a number of his colleagues (and to some degree himself ) for the
attribution of wishful mnemonics:
A major source of simple mindedness in AI programs is the use of
mnemonics like understand or goal to refer to programs and data
structures. The practice has been inherited from more traditional
programming applications, in which it is liberating and enlightening
to be able to refer to program structures by their purposes. Indeed,
part of the thrust of the structured programming movement is to
program entirely in terms of purposes at one level before implement-
ing them by the most convenient of the (presumably many) alterna-
tive lower-level constructs (McDermott, 1976, p. 4)
The illicit slide is to identifying some part of the model as some mental state or oth-
er. For in doing so, one jumps the gun on the explanation of that mental state. As
McDermott complains with a pinch of sarcasm:
If a researcher tries to write an understanding program, it isnt be-
cause he has thought of a better way of implementing this well-
understood task, but because he thinks he can come closer to writing
the first implementation. If he calls the main loop of his program
UNDERSTAND, he is (until proven innocent) merely begging the
question. He may mislead a lot of people, most prominently himself,
and enrage a lot of others (ibid.)
As McDermott advises: What he should do instead is refer to this main loop as
G0034, and see if he can convince himself or anyone else that G0034 implements
some part of understanding (ibid.). The error is an artefact of overenthusiastic the-
orising. It is still very easily found in contemporary literature, especially where box-
and-arrow diagrams are concerned, as these naturally invite optimistic labelling. For
instance, the literature on agentive experience is riddled with examples like Figure
33:

| 309

Figure 33 Optimistically labelled comparator model

Again, calling whatever is going on in those boxes desired, or saying that the predic-
tions made by such and such component of the model concerns the relation be-
tween movement and perception, is getting a little ahead of oneself.
222

However, all of the above should not obscure (in fact it should motivate) the fact
that some minimally sufficient ground for representation is needed. Several authors
have been inspired by a simple faux-evolutionary story that is somewhat suggestive
about the conditions under which a system would require the use of representa-
tions. For instance, Gerard OBrien and Jon Opie (2004) muse upon various ways in
which Nature might furnish an organism with the capacity to seek the good and
avoid the bad in its environmental niche. The simplest means would be to blindly
bounce in complete hostage to fortune. A slightly more sophisticated and slightly
less dangerous means would be to exploit ambient radiant or mechanical energy to

222
In Synofzik et al.s defence, they have clearly inherited these optimistic conventions
from the diagrams in Frith et al.s papers (2000a, p. 1773; 2000b, p. 359).
Optimistic labels within solid boxes are surrounded with
dashed boxes. From Synofzik et al. (2008)
310 |
provide cues about the state of the environment prior to immediate contact. How-
ever, unreliable or latent receipt of such cues would expose the organism to many of
the dangers inherent in blindly bumping into potentially nasty elements of its envi-
ronment. A third strategy proves much more effective:
Instead of responding directly to the information impacting on its
sensory surfaces, a creature can use it to construct internal models of
the environment on-board states that stand in for or represent
external objects, relations and states of affairs. These internal repre-
sentations, rather than the information-laden signals from which
they were constructed, can then be pressed into service to shape be-
haviour. Such a decoupling of behaviour from direct environmental
control confers great benefits, since internal models of the environ-
ment can have a stability and definiteness that is lacking in the sig-
nals that impact on a creatures sensory surfaces. They also enable a
creature to respond to features of the world that are not immediately
present, to use past experiences to shape present behaviour, to plan
for the future, and, in creatures such as ourselves, to be sensitive to
very abstract features of the world. (O'Brien & Opie, 2004, pp. 1 - 2)
223

The authors suggestion is that in this last case, the organisms adaptive engagement
with its environment is facilitated by representation. The clear pay-off in terms of
behavioural flexibility is intended to motivate the very idea of systems representing
at all. In several articles, Rick Grush has developed ideas from control-theoretic
approaches to sensorimotor control (see 7.1) to provide a general account of repre-
sentation in respect of roughly this motivation.
224
To recap some key ideas, recall
that control systems have at least two components: a controller and a plant. In a
feed-back system, the controller receives signals from the output of the plant and
sends control signals according to the relative disparity between output and target
state. However, incoming peripheral sensory signals are ambiguous (Latash, 2008a,
pp. 258 - 260) and cannot travel fast enough to influence the ongoing stream of mo-
tor signals at the right time (Grush, 1997, p. 21 n13; Latash, 2008b, pp. 70 - 72). Nev-

223
Cf. Dennett (1996, pp. 88 - 90), Clark & Grush (1999, p. 7), Metzinger (2003a, pp. 47
- 49). For an especially rich discussion, see B.C. Smith (1996, pp. 213 - 276).
224
I say roughly, because a distinctive feature of Grushs theory of representation is
that states of the environment are represented as interacted with. See e.g. Grush (1997, pp. 18
- 19).
| 311

ertheless, smooth sensorimotor control is achieved.
225
Grush argues, on the same
lines as described in 7.1, that the reason for this is because the system runs an emu-
lation of the plant. Emulators enable the system to not only accommodate signal
delay and unreliability, but also to discover how the plant will change over time,
without any obligation to implement such changes (Grush, 1997, pp. 10 - 11). These
capacities come in virtue of driving a faux control process alongside (or in absence
of ) the actual control process. Hence, the controller can interact with an emulation
of the plant, or it can interact with the plant itself (see Figure 34).

Figure 34 Simple schematic for emulation control

Emulation fulfils the basic criterion of having a representational job-description.
Furthermore, it does so modestly (there is a clear explanatory justification), by
avoiding pre-emptive registration (as a methodological folly) and optimistic label-
ling. Speaking more generally, the emulation theory can be described as a resem-
blance theory of representation.
226
A famous (if short) critique of resemblance

225
Grush sees forward models as subordinate to emulators as a superordinate category
(2007b, p. 410). Forward models, as I described them earlier, model the sensory-
consequences of particular motor signals. In Grushs terms, described thus, they are modal
emulators. However, forward models could conceivably model changes in the state of the
plant as abstracted from any sensory consequence measured. They would thus be function-
ing as amodal emulators (Grush, 2004, pp. 387 - 389).
226
It ought to be noted that Grush does not ever explicitly describe the theory as a re-
semblance theory. It might be that he intends emulation to be understood as a suis generis
form of representation. However, resemblance is certainly broader category than emulation,
and the latter can be usefully brought under the former to clarify at least a version of emula-
tive representation.
Double arrows indicate causal interactions.
Dashed double arrows indicate intermittent causal interactions.
312 |
comes from Nelson Goodman.
227
He pejoratively refers to it as the most naive theo-
ry of representation, where A represents B if and only if A appreciably resembles
B or A represents B to the extent that A resembles B (Goodman, 1969, p. 3). He
continues unsympathetically:
Some of the faults are obvious enough. An object resembles itself to
the maximum degree but rarely represents itself; resemblance, unlike
representation, is reflexive. Again, unlike representation, resem-
blance is symmetric: B is as much like A as A is like B, but while a
painting may represent the Duke of Wellington, the Duke doesn't
represent the painting. Furthermore, in many cases neither one of a
pair of very like objects represents the other: none of the automobiles
off an assembly line is a picture of any of the rest; and a man is not
normally a representation of another man, even his twin brother.
Plainly, resemblance to any degree is no sufficient condition for rep-
resentation. (Goodman, 1969, p. 4)
Goodmans points are well taken. No one should claim that resemblance itself is
sufficient for representation, any more than one should claim that nomic-
dependence is sufficient for representation. The extra required is provided by
adopting a triadic analysis, according to which the emulator produces representa-
tions of the plant to be consumed by the controller (see Figure 35).

Figure 35 Emulation and representation

Now, to finally return to the broader context: I think that this is the basic framework
for representation that global and local views should share. From this very basic
sketch the emulation theory could be developed in various ways, each of which may

227
For a more general treatment, see Cummins (1989, pp. 27 - 34).
The triadic analysis applied to the emulation schematic
| 313

serve a global or local views purposes. But let me leave the reader with some further
specification to indicate the general line. According to one formulation, the emula-
tor functions as an articulated model. The components of an articulated model,
called articulants, are functionally organised such that each articulant corresponds
to a significant element (a state variable or a parameter) of the system modelled,
such that the interaction of the articulants is isomorphic to the interaction between
the variables of the MSS (2004, p. 379).
228
Grush describes the basic idea with con-
siderable clarity in one of his early papers (see also Figure 36). To begin with he
characterises the system to be modelled as a dynamical system with N parameters,
e
1
, e
2
, ..., e
n
. A proportion of these, e
1
, e
2
, ..., e
i
will be input parameters, in that
their equations of evolution will include one or more parameters of the controller.
Others, labelled e
i+1
, e
i+2
, ..., e
i+k
will be output parameters, in that their parame-
ters show up in the equations of evolution for one or more of the parameters of the
controller. An articulated emulator for this system would take e*
1
, e*
2
, ..., e*
n
as
inputs, and e*
i +1
, e*
i +2
, ..., e*
i +k
as outputs. On the assumption that the dynamics
of the emulator and plant are formally equivalent, then any given articulant e*
x
, rep-
resents parameter e
x
in the plant (Grush, 1997, pp. 16 - 17).

228
Grush uses the word analogous where I have used isomorphic. Isomorphism is
sometimes used as a means to precisify the concept of resemblance between structures.
Cummins provides a nice outline of what it is for two structures to be isomorphic:
The idea is that there is a mapping between the two structures such that (1) for every object in
the content structure C, there is exactly one corresponding object in the representing struc-
ture R; (2) for every relation defined in C, there is exactly one corresponding relation defined
in R; and (3) whenever a relation defined in C holds of an n-tuple of objects in C, the corre-
sponding relation in R holds of an n-tuple of objects in R. (1996, p. 96)
Isomorphism as described here is a form of what OBrien and Opie call second-order resem-
blance, a situation in which the relations among a system of representing vehicles mirror
the relations among their objects (2004, p. 10). On some views, this results in an extra-
ordinary degree of representational capacity. For instance, according to Cummins (1996, p.
96), any element of a representing structure can represent an element of the represented
structure, and the same for relations and states of affairs that hold true of multiple elements
of that structure.
314 |

Figure 36 Articulated emulation

Furthermore, application of certain points of earlier discussion provides a strong
rationale for ruling out the claim that dynamic elements of the controller represent
the plant. Say the controller implements a mapping from e
i +1
, e
i +2
, ..., e
i +k
, to e
1
,
e
2
, ..., e
n
. In such a case, e
x
does not represent e
x
, for several reasons. Firstly, it
would be puzzling to work out how e
x
is being employed as a representation. And
what would the consumer be? Not the plant. Perhaps the emulator, but the key issue
here is what justification there would be for claiming that e
x
functions as any more
than an element in a causal process. Furthermore, any attempted justification is
likely to violate the modesty constraint. For unless it were denied that e*
x
represents
e
x
, the claim must be that e
x
is being represented twice over, but with the very same
justification in each case promiscuous indeed.
General schematic for a control process employing an articu-
lated emulator. Elements e*
n
represent elements e
x
whereas
elements e
n
do not.
| 315

Conclusion
I conclude in two sections. First, I present what I believe to be the relatively novel
results of each chapter. Second, I discuss what I believe to be the most important
issues that the dissertation has left open.
Some resul t s of t he di s cussi on
Chapt er 1
The ontological neutrality of the direct object of experience: Either sense-data
are the categorical posits of sense-datum theory, or their ontological status qua di-
rect objects of experience should not be prejudged. I argued that there is much rea-
son to think that the ontological status of sense-data cannot be sufficiently clarified
to meet the first option.
Chapt er 2
Two-dimensionality & the visual-field: Several famous means of arguing that the
objects of visual experience are only ordered on two-dimensions and/or that visual
experience is indirect are unsupported by the very sources that they appeal to. This
is despite the fact that a description of the spatiality of visual experience can be arti-
ficially limited to two-dimensions, and that the psychology of vision fruitfully em-
ploys the concept of a two-dimensional visual field.
Chapt er 3
The nature of sensation in Berkeleys theory of vision: Berkeleys theory of vi-
sion depends upon the assumption that visual sensation is impoverished in convey-
ing spatial dimensions. However, if Berkeley thereby intends to make claims about
visual experience on this basis, he does so only by conflating perceptual sensation
and mere sensation.
Perceptual sensation: A sensory experience of a given mode, in virtue of
which a perceptual experience occurs and is an instance of that mode.
Mere sensation: A sensory event that is necessary for the occurrence of a
perceptual sensation. In vision mere sensations come in at least three varie-
ties
energetic impingement of light on the retina
unconscious sensation of light
sensory experience of light
316 |
Chapt er 4
Four senses of the word phenomenology:
Phenomenology
D
: A description of conscious experience or the act or pro-
ject of describing it.
Phenomenology
U
: An umbrella term for conscious experience.
Phenomenology
M
: A conscious experience conceived as having some defi-
nite metaphysical status.
Phenomenology
T
: Transcendental phenomenology, associated with the
transcendental reduction and transcendental idealism.
Chapt er 5
A spectrum or scale of sensory experiences: Illusions, hallucinations, dreams,
and sensory imaginings subjectively indistinguishable from instances of perception
can be conceived either as situated on a spectrum or at the apex of a scale, where
that situation or height is distinctive of perceptual experience.
Chapt er 6
A description of the experience of particular objects as mind-independent:
Although it is crucial to the direct realist conception of perceptual experience that
the objects of perceptual experience seem to be particular and mind-independent,
there is a distinctive lack of further specification in the literature. By drawing mostly
on Husserls early work, particularly his Thing & space lectures of 1907, it is possible
to articulate the claim that perceptual experience seems to be of mind-independent
particular objects in sixteen inter-related claims.
Chapt er 7
The concept of a structural affordance: Affordances can be analysed as relations.
The canonical affordance relation between an agent and its environment can be
further analysed as a second-order relation prior to which is a first-order relation,
dubbed structural affordance. This latter is an affordance between a bodily agent
and the properties of its body in virtue of which basic actions are possible.
Structural affordance theories of kinaesthetic experience:
SA1: A theory of the spatial content of bodily experience; specifically, the experience
of parts of the body as parts of an integrated whole, dubbed body-mereological
experience. Body mereological experience is to be explained by representation of
parts of the body that exploit the structural unity of the body.
| 317

SA2: A theory of the structuring role of kinaesthesia in perceptual experience. Kin-
aesthesia is re-described as the anticipation of a pattern of kinaesthetic experience
associated with a type of basic action, dubbed phenomenal grooves. Phenomenal
grooves are to be explained by kinaesthetic ability, the ability to anticipate the phe-
nomenal grooves of a type of action in virtue of anticipating the structural af-
fordances of a type of action.
Chapt er 8
A dilemma for the bodily self thesis: The bodily self thesis holds that bodily sub-
jects of experience exist qua bodily selves. Two imagined scenarios force incompati-
ble responses from a defendant of the bodily self thesis. The first claims the experi-
ence of disconnected parts of the body is possible. The second claims that simulta-
neous perceptual experience of one subject from several disconnected perspectives
is possible. If response to the first involves stipulation that the bodily self is materi-
ally unified, then the second case is unanswerable. If response to the second case
involves stipulation that the bodily self can be a scattered material object, then the
first case is unanswerable.
Chapt er 9
Embodiment of body representation: A body representation is embodied if and
when reference to the actual body and/or its properties and/or its relations is indis-
pensable in an explanation of how the representation achieves its function.
Local representation of the body through exploitation of structural af-
fordances: A local theory of body representation claims that a system could, in
each case in which it represents the body, represent individual parts of the body
only and their properties of being causally interactive with other parts of the body.
Such a theory would thereby forego the need to represent the body as a whole, by
exploiting the fact that those very same parts possess property of being causally
interactive with the body as a whole.
318 |
Some i ssues l ef t open
As with any finite study, there are many issues left open by this dissertation, and I
will not cover them all. But I will cluster two that stand out particularly into two sets
of remarks.
Percept ual experi ence & percept i on
The descriptions of perceptual experience discussed at various points in this disser-
tation have been treated as the conceptions of professional philosophers engaged in
discussion of the nature of perception. I have taken on the task of describing, de-
fending and clarifying a particular conception that manifests itself in a set of shared
ideals ensconced within the philosophy of direct realism. However, there is a clear
sense in which that task is incomplete. For what has at no point been established is
whether that conception has any bearing on what is putatively being thought about
and discussed: perceptual experience. Indeed, this is just an aspect of a more gen-
eral issue left wide open. Do claims made about perceptual experience have any
bearing upon perceptual experience itself?
Naturally, the question presupposes that perceptual experiences do exist. Indeed,
few go so far as to deny this. I think a fairly standard view is that this point is not
open to real dispute, neatly expressed by Searle when he says:
It is a bit difficult to know how one would argue for the existence of
perceptual experiences to someone who denied their existence. It
would be a bit like arguing for the existence of pains: if their exist-
ence is not obvious already, no philosophical argument could con-
vince one. (1983, p. 44)
For some though, denial of the existence of perceptual experiences is actually moti-
vated by the very claims that philosophers make about them. For instance, Alex
Byrne (2009) notes that according to some versions of the transparency thesis at-
tention to perceptual experience itself is impossible, for all one ends up attending to
is the world. This might not immediately suggest reason to question the existence of
perceptual experience in general. But if all one has to individuate experiences is at-
tention to the world, it might suggest scepticism about the possibility of distin-
guishing distinct, punctate occurrences of perceptual experience. Developing this
worry into a positive claim, Tye suggests the simplest hypothesis to be that for
each extended period between states of unconsciousness, there is just one experi-
ence unified over time (2003, pp. 85 108). To which a sceptic might respond: One
experience too many, perhaps: a simpler hypothesis is that there are no experiences
(Byrne, 2009, p. 435).
| 319

However, if being a sceptic means denying the existence of perceptual experiences
tout court, then Byrne does not count. His point is rather that accounts of the met-
aphysics of perception can get along fine without positing the philosophical notion
of a perceptual experience as a particular mental event:
Obviously there are experiences: watching the nal inning was a
thrilling experience, and eating the crackerjack was an unpleasant
one, for example. However, to conclude from this that there are visu-
al experiences and gustatory experiences in the special philosophi-
cal sense is just to ignore the fact that experience in its philosophical
use is not a harmless extension of ordinary usage. (op. cit., p. 433)
But then again, perception is hardly a harmless extension of ordinary usage either.
And on some views, these harmful extensions are absolutely intertwined. Witness,
for instance, Dretskes proclamation that: Seeing a piano being played is constitut-
ed, in part, by a visual experience, hearing by an auditory experience. Until these
experiences occur one has not seen or heard the piano (Dretske, 1995, p. 9). Now
there is certainly a use of the term perception which allows for the possibility that it
might occur without an accompanying experience (cf. Marcel, 1983), but if percep-
tion in its harmful philosophical usage is a term of art in the first place, then the
issue is merely in a choice of words.
Say then, that one admits that perceptual experience exists only insofar as percep-
tion does. This brings one directly back to the first issue of whether claims made
about perceptual experience have any bearing upon perceptual experience itself.
Here again it is not hard to find claims to the effect that there are certain points at
which dispute cannot be taken seriously, even amongst those somewhat sympathet-
ic to expressions of scepticism. An example offered by Bayne & Spener is that of
looking at a page such as the one you are seeing now:
Is it reasonable to doubt the deliverance of introspection in this in-
stance, i.e. that it visually appears to you as if there is a white surface
with black marks before you? Surely not: this introspective judge-
ment might not be infallible, but it is one in which a high degree of
confidence would be entirely warranted. What holds here holds of
countless other introspective judgements that one might make []
introspective judgements concerning which serious doubt seems
perverse (2010, pp. 7 - 8)
Although the prima facie unassuming nature of their description might suggest
something that can hardly be disputed, no reason is given for why one should favour
this over the description that, e.g., it seems just as if you are looking at the words; or
the words on the page; or at the dissertation you have in front of you. A reason could
320 |
probably be provided, roughly along the same reasons provided by sense-datum
theorists, but denial in this case would perhaps not be so perverse.
I think that this is a major outstanding issue, and I cannot see how it would be re-
solved. Admittedly, in some ways it is just an aspect of the more general and thor-
oughly treated problem of the relationship between consciousness and introspec-
tion (see e.g. Hurlburt & Schwitzgebel, 2007). But what makes this particular issue
especially pressing is the fact that if the claims are radically misguided, the conse-
quence is that an entire range of disputes in the philosophy of perception are ren-
dered insignificant. Indeed, the methodology of many of the sections in Part I and
Part II was to extract claims about perceptual experience by understanding their
broader philosophical significance for the project of direct realism and its defence.
If it turned out that those claims are mere fictions (as Dennett (1991) might say),
then the proclamation that direct realism is not merely nave but false (or alterna-
tively, that it vindicates the world view of the nave) becomes simply hot air in the
seminar room.
Local & gl obal vi ews on ki naest het i c abi l i t y
Although I believe the problem of mapping conceptions of perceptual experience
onto actual perceptual experience to be rather vexed, in Part III I largely ignored
that issue. I turned instead to a discussion of how one might explain a core feature
of that conception (the structuring role of bodily experience) by identifying natu-
rally occurring mechanisms. It might be a glaringly obvious fact to the reader that,
despite somewhat optimistic remarks at the end of chapter 7, it has certainly not
been established how this would be done in the end. But this is merely to
acknowledge the fact that there is plenty of interesting work to be done here.
Structural affordances, phenomenal grooves and kinaesthetic abilities are defined in
such a way that in identifying their realisers, one would be identifying the realisers
of Husserlian kinaesthesia as detailed in chapter 6. The discussion of body repre-
sentation in chapter 9 is intended to lay the groundwork for that project. The hope
is that a fruitful controversy could emerge over whether the existence of structural
affordances results in exploitative representational strategies. Potentially, I believe
that both the local view and the global views discussed there could (upon further
specification) serve in explanations of kinaesthetic ability. However, in concluding
the thesis I want to acknowledge some of the problems faced by a defendant of the
local view.
In 7.4 I gave an example of a study by Fourneret & Jeannerod which provided a
proof of concept for phenomenal grooves, despite the fact the idea of a phenomenal
groove is perhaps slightly contrary to their claims. A similar situation confronts us in
| 321

discussion of a recent study by Kannape, Schwabe, Tadi, & Blanke, (2010). By draw-
ing on an impressive variety of tools, these researchers were able to study the effects
of angular biases introduced during the execution of full-body movements towards
visually specified targets. Their participants were adorned with infra-red markers to
facilitate motion capture, and presented with a projection of a three-dimensional
avatar (sharing their skeletal structure) whose movements were coupled to their
own. Several targets were randomly projected in the room, and the task set was to
simply move to the location of the target. However, in 75% of the trials the experi-
menters would introduce a deviation of 5 - 30 in the movements of the avatar, re-
sulting in various degrees of compensatory movement according to the extent of the
bias and the orientation of the avatar. As with the Fourneret & Jeannerod study, the
sensorimotor bias in itself is interesting. But the bias is even more interesting in
conjunction with the pattern of participants responses to the question posed to
them in each case: Did the movement shown on the screen correspond to the
movement you just performed? (Kannape et al., 2010, p. 1631).

In their results the authors report that deviations of 5, 10, and 15 lead to many
erroneous self-attributions but these were found to be decreasing in magnitude
with increasing angular deviation (Kannape et al., 2010, p. 1631). However, as
broached above, one could alternatively claim that deviations below 15 all fall (to a
greater or lesser degree) within the phenomenal groove of the action specified by
the task.
(A) Participant in tracking arena with avatar and target visi-
ble in projected virtual room
(B) Virtual room and avatar with target, deviation and com-
pensatory movement shown
Figure 37 - Motor biases induced by movement of an avatar
322 |
Now consider how global and local views might explain what is going on here. Each
might claim that representation of the body enables kinaesthetic abilities. Thus the
agent would be able to anticipate the phenomenal grooves of the types of action she
performs in moving towards a particular target. An important feature of the study is
that the actions that the subject is required to perform involve translatory move-
ments of the body as a whole. On the global view the parameters of these move-
ments can be represented straightforwardly; at least, if it is possible to represent any
global properties of the body, it must be possible to represent these. However, the
local view would have to tell a rather more complex story. Somehow the activity of
each part would be represented such that the system exhibited the same behaviour
as it would if it were representing global properties of the body. Again, what I mean
to point out is that there is all the work to do here. What needs to be shown is how
structural affordances actually fill in the gaps left by LB-reps. In short, it needs to be
shown how LB-reps can only represent the P properties of a given structural af-
fordance and yet enable the anticipation of W properties.
But note, even if that could be shown, it is not at all clear that this would be a suffi-
cient explanation of phenomenal grooves. For even if we assume that structural af-
fordances exist and that their W properties can be anticipated by the agent without
being represented, there is the further task of explaining how kinaesthetic abilities
are realised. Perhaps this could be resisted by claiming that anticipation of structur-
al affordances just is anticipation of phenomenal grooves. But that would just leave
a mystery as to why this is so. This is a problem for any local view on body represen-
tation that tries to explain the putative structuring role of Husserlian kinaesthesia;
it is also a problem for any view that tries to explain body-mereological experience.
But a global view need not face the same kind of problem, for it can readily draw
upon the resources of an extensive theory of subjective experience, the self-model
theory of subjectivity (Metzinger, 2003a). To my mind, this opens up tantalising
prospects for future research into the viability of both global and local views on
body representation, for the implications of structural affordances for either is yet to
be understood.
| 323

References
Abraham, R., & Shaw, D. (1992). Dynamics: The geometry of behaviour.
Redwood city, CA: Addison-Wesley Publishing Co.
Adams, F., & Aizawa, K. (2001). The bounds of cognition. Philosophical
Psychology, 14(1), 43 - 64.
Aglioti, S., DeSouza, J. F. X., & Goodale, M. A. (1995). Size-contrast illusions
deceive the eye but not the hand. Current biology : CB, 5(6), 679-685.
Albahari, M. (forthcoming). Why we should preserve the classical concept of
self. In M. Siderits, E. Thompson & D. Zahavi (Eds.), Self, no self?
Perspectives from analytical, phenomenological and Indian traditions.
Oxford: Oxford University Press.
American Psychological Association. (2010). Publication manual of the
American Psychological Association. Washington, DC: American
Psychological Association.
Anscombe, G. E. M. (1965). The intentionality of sensation Metaphysics and
the philosophy of mind: Collected philosophical papers, Volume III (pp.
3-20). Oxford: Blackwell.
Anscombe, G. E. M. (1975/1991). The first person. In D. M. Rosenthal (Ed.),
The nature of mind (pp. 71 - 81). Oxford: Oxford University Press.
Aristotle. (350 BCE/2002). De Anima, Books II and III (with passages from
Book I): Translated with introduction and notes by D.W. Hamlyn. With
a report on recent work and a revised bibliography by Christopher
Shields (D. W. Hamlyn, Trans.). Oxford: Clarendon Press.
Armstrong, D. M. (1956). Discussion: Berkeley's New Theory of Vision.
Journal of the History of Ideas, 17(1), 127-129.
Armstrong, D. M. (1960). Berkeley's theory of vision. Melbourne: Melbourne
University Press.
Armstrong, D. M. (1961). Perception and the physical world. New York:
Humanities Press.
Armstrong, D. M. (1968). A materialist theory of the mind. London:
Routledge, Kegan and Paul.
Aspell, J. E., Lenggenhager, B., & Blanke, O. (2009). Keeping in Touch with
One's Self: Multisensory Mechanisms of Self-Consciousness. PLoS
ONE, 4(8).
Atherton, M. (1990). Berkeley's revolution in vision. Ithaca: Cornell University
Press.
Austin, J. L. (1962). Sense and sensibilia. Oxford: Clarendon Press.
Ayer, A. J. (1973). The central questions of philosophy. London: Weidenfeld
and Nicholson.
Ayers, M. (1993). Locke: Epistemology and ontology. London: Routledge.
324 |
Bach-y-Rita, P. (2002). Sensory substitution and qualia. In E. Thompson & A.
No (Eds.), Vision and mind: Selected readings in the philosophy of
perception (pp. 497 - 514). Cambridge, Mass: MIT Press.
Bach-y-Rita, P. (2004). Tactile sensory substitution studies. Annals of the
New York Academy of Sciences, 1013, 83 - 91.
Bach-y-Rita, P., Collins, C. C., Saunders, F. A., White, B., & Scadden, L.
(1969). Vision substitution by tactile image projection. Nature,
221(5184), 963-964.
Bach-y-Rita, P., & W. Kercel, S. (2003). Sensory substitution and the human-
machine interface. Trends in Cognitive Sciences, 7(12), 541 - 546.
Bailey, A. A., & Mrsch, F. P. (1941). Phantom limb. Canadian Medical
Association journal, 45(1), 37-42.
Baker, L. R. (1998). The first-person erspective: A test for naturalism.
American Philosophical Quarterly, 35(4), 327-348.
Baker, L. R. (2000). Persons and bodies: A constitution view. New York:
Cambridge University Press.
Baldwin, T. (1990). G. E. Moore. London, New York: Routeledge.
Ballard, D. H., Hayhoe, M. M., Pook, P. K., & Rao, R. P. (1997). Deictic codes
for the embodiment of cognition. Behavioral and Brain Sciences,
20(4), 723 - 742. doi: citeulike-article-id:731546
Barz, W. (2002). Bibliographie zum Thema Intentionalitt. Retrieved from
http://www.uni-
bielefeld.de/philosophie/personen/barz/BibliographieIntentionalitaet.pdf
Barz, W. (2008). The real trouble with intentionality. Philosophical
Explorations, 11(2), 79 92.
Bayne, T. (2004). Self-consciousness and the unity of consciousness. Monist,
87(2), 219-236.
Bayne, T. (2008). The phenomenology of agency. Philosophy Compass, 3(1),
182-202.
Bayne, T., & Pacherie, E. (2007). Narrators and comparators: the architecture
of agentive self-awareness. Synthese, 159(3), 475-491.
Bayne, T., & Spener, M. (2010). Introspective humility. Philosophical Issues,
20(1), 1 - 22.
Bealer, G. (1996). Materialism and the logical structure of intentionality. In
H. Robinson (Ed.), Objections to Physicalism (pp. 101 - 126). New
York: Oxford University Press.
Bechtel, W. (1998). Representations and cognitive explanations: Assessing
the dynamicist challenge in cognitive science. Cognitive Science, 22(3),
295-317.
Bechtel, W. (2001). Representations: From neural systems to cognitive
systems. In W. Bechtel, P. Mandik, J. Mundale & R. S. Stufflebeam
| 325

(Eds.), Philosophy and the neurosciences (pp. 332 - 349). Malden, MA:
Blackwell.
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on
cognitive neuroscience. New York: Taylor & Francis Group LLC.
Bennett, K. (2005). Two axes of Actualism. The Philosophical Review, 114(3),
297 - 326.
Berkeley, G. (1710/2008). A treatise concerning the principles of human
knowledge. In D. M. Clarke (Ed.), Berkeley: Philosophical writings (pp.
67 - 150). Cambridge: Cambridge University Press.
Berkeley, G. (1732/2008). An essay towards a new theory of vision. In D. M.
Clarke (Ed.), Berkeley: Philosophical writings (pp. 1 - 66). Cambridge:
Berkeley, G. (1734/2008). Three dialogues between Hylas and Philonous. In
D. M. Clarke (Ed.), Berkeley: Philosophical writings (pp. 151 - 242).
Cambridge: Cambridge University Press.
Berkeley, G. (1752/2008). Alciphron: or, the minute philosopher. In D. M.
Clarke (Ed.), Berkeley: Philosophical writings (pp. 269 - 314).
Berkinblit, M. B., Feldman, A. G., & Fukson, O. I. (1986). Adaptability of
innate motor patterns and motor control mechanisms. Behavioral and
Brain Sciences, 9(04), 585-599.
Bermdez , J. L. (1995). Nonconceptual content: From perceptual experience
to subpersonal computational states. Mind & Language, 10(4), 333-
369.
Bermdez, J. L. (1998). The paradox of self-consciousness. Cambridge, MA:
MIT Press.
Bermdez, J. L. (2000). Naturalized sense data. Philosophy and
Phenomenological Research, 61, 353.
Bermdez, J. L. (2005). The phenomenology of bodily awareness. In A. L.
Thomasson & D. W. Smith (Eds.), Phenomenology and philosophy of
mind (pp. 295 - 316). Oxford: Oxford University Press.
Bermdez , J. L., & Cahen, A. (2008). Nonconceptual mental content. The
Stanford encyclopedia of philosophy (Spring 2010 Edition). Retrieved
from <http://plato.stanford.edu/archives/spr2010/entries/content-
nonconceptual/>
Blanke, O., & Metzinger, T. (2009). Full-body illusions and minimal
phenomenal selfhood. Trends in Cognitive Sciences, 13, 713.
Blanke, O., & Mohr, C. (2005). Out-of-body experience, heautoscopy, and
autoscopic hallucination of neurological origin: Implications for
neurocognitive mechanisms of corporeal awareness and self-
consciousness. Brain Research Reviews, 50(1), 184-199. doi: DOI:
10.1016/j.brainresrev.2005.05.008
326 |
Blanke, O., Ortigue, S., Landis, T., & Seeck, M. (2002). Stimulating illusory
own-body perceptions. Nature, 419, 269 - 270.
Block, N. (2007a). Consciousness, function and representation: Collected
papers (Vol. 1). Cambridge, MA: MIT Press.
Block, N. (2007b). Review of Alva No, Action in perception. In N. Block
(Ed.), Consciousness, function, and representation: Collected papers
(Vol. 1, pp. 363 - 375). Cambridge, MA: MIT Press.
Borges, J. L. (1998). Collected fictions (A. Hurley, Trans.): Penguin Classics.
Boring, E. G. (1952a). The Gibsonian visual field. Psychological Review, 59,
246 - 247.
Boring, E. G. (1952b). Visual perception as invariance. Psychological Review,
59, 141 - 148.
Botvinick, M., & Cohen, J. (1998). Rubber hands 'feel' touch that eyes see.
Nature, 391, 756756.
Boyd, R. (1991). Observations, explanatory power and simplicity: Toward a
non-Humean account. In R. Boyd, P. Gasper & J. D. Trout (Eds.), The
philosophy of science (pp. 349 - 378). Cambridge, MA: MIT Press.
Brentano, F. (1874/1973). Psychology from an empirical standpoint (A. C.
Rancurello, D. B. Terrell & L. L. McAlister, Trans.). London:
Routledge.
Brewer, B. (1992). Self-location and agency. Mind, 101, 1734.
Brewer, B. (1995). Bodily awareness and the self. In N. Eilan, A. Marcel & J. L.
Bermdez (Eds.), The body and the self (pp. 291251). Cambridge, MA:
MIT Press.
Brewer, B. (1999). Perception and reason. Oxford: Oxford University Press.
Brewer, B. (2006). Perception and content. European Journal of Philosophy,
14(2), 165-181.
Broad, C. D. (1925). The mind and its place in nature. London: Kegan Paul.
Broad, C. D. (1927). Scientific thought. London: Harcourt, Brace and
Company, inc.
Broughton, J. (2002). Descartes' method of doubt. Princeton: Princeton
University Press.
Broughton, J. (2008). Self-knowledge. In J. Broughton & J. Carriero (Eds.), A
companion to Descartes (pp. 179 - 195). Malden, MA: Blackwell.
Brown, C. (2007). Narrow mental content. The Stanford encyclopedia of
philosophy (Spring 2007 Edition). Retrieved from
http://plato.stanford.edu/entries/content-narrow/
Burge, T. (1977). Belief de re. The Journal of Philosophy, 74(6), 338-362.
Burge, T. (1986). Individualism in psychology. Philosophical Review, 95(3 -
46).
Burgess, P. R., & Clark, F. J. (1969). Characteristics of knee joint receptors in
the cat. The Journal of Physiology, 203(2), 317-335.
| 327

Byrne, A. (2001). Intentionalism defended. The Philosophical Review, 110(2),
199-240.
Byrne, A. (2009). Experience and content. The Philosophical Quarterly,
59(236), 429-451.
Byrne, A., & Logue, H. (Eds.). (2009). Disjunctivism: Contemporary readings.
Cambridge, MA: MIT Press.
Campbell, J. (1994). Past, space, and self. Cambridge, MA: MIT Press.
Campbell, J. (2002). Reference and consciousness. Oxford: Oxford University
Press.
Campbell, J. (2005). Information processing, phenomenal consciousness and
Molyneux's question. In J. L. Bermdez (Ed.), Thought, reference and
experience: Themes from the philosophy of Gareth Evans (pp. 195 -
219). Oxford: Oxford University Press.
Carman, T. (2000). The body in Husserl and Merleau-Ponty. Philosophical
Topics, 27, 205226.
Carmena, J. M., Lebedev, M. A., Crist, R. E., O'Doherty, J. E., Santucci, D. M.,
Dimitrov, D. F., . . . Nicolelis, M. A. L. (2003). Learning to control a
brain-machine interface for reaching and grasping by primates. PLoS
Biology, 1(2), 193-208.
Carr, D. (1999). The paradox of subjectivity. Oxford: Oxford University Press.
Casati, R. (2004). Is the object concept formal? Dialectica, 58(3), 383-394.
Cassam, Q. (1997). Self and world. Oxford: Oxford University Press.
Castaeda, H.-N. (1966/2001). 'He': A study in the logic of self-consciousness.
In A. Brook & R. DeVidi (Eds.), Self-reference and self-awareness.
Amsterdam: John Benjamins Publishing Co.
Caston, V. (2001). Connecting traditions: Augustine and the Greeks on
intentionality. In D. Perler (Ed.), Ancient and Medieval theories of
intentionality (pp. 23 48). Leiden: Brill.
Caston, V. (2008). Intentionality in ancient philosophy. The Stanford
encyclopedia of philosophy (Fall 2008 Edition). Retrieved from
http://plato.stanford.edu/archives/fall2008/entries/intentionality-ancient/
Chalmers, D. (2004). The representational character of experience. In B.
Leiter (Ed.), The future for philosophy (pp. 153 - 181). Oxford: Oxford
University Press.
Chemero, T. (2003). An outline of a theory of affordances. Ecological
Psychology, 15, 181 - 195.
Chemero, T. (2009). Radical embodied cognitive science. Cambridge, MA:
MIT Press.
Chemero, T., & Silberstein, M. (2008). After the philosophy of mind:
Replacing scholasticism with science. Philosophy of Science, 75(1), 1-
27.
328 |
Chisholm, R. M. (1948). The problem of empiricism. The Journal of
Philosophy, 45(19), 512-517.
Chisholm, R. M. (1957). Perceiving: A philosophical study. Ithaca, NY: Cornell
University Press.
Chisholm, R. M. (1982). The foundations of knowing. Minneapolis: University
of Minnesota Press.
Churchland, P. M. (1981). Eliminative materialism and the propositional
attitudes. The Journal of Philosophy, 78, 67-90.
Churchland, P. M. (1986). Some reductive strategies in cognitive
neurobiology. Mind, 95(379), 279-309.
Churchland, P. S., & Churchland, P. M. (1983). Stalking the wild epistemic
engine. Nos, 17, 5.
Clapin, H. (Ed.). (2002). Philosophy of mental representation. Oxford: Oxford
University Press.
Clark, A. (1989). Microcognition: Philosophy, cognitive science, and parallel
distributed processing. Cambridge, MA: MIT Press.
Clark, A. (1997). Being there: Putting brain, body, and world together again.
Clark, A. (2001). Visual experience and motor action: Are the bonds too
tight? The Philosophical Review, 110(4), 495-519.
Clark, A. (2003). Natural-born cyborgs: Minds, technologies and the future of
human intelligence. Oxford: Oxford University Press.
Clark, A. (2005). The twisted matrix: Dream, simulation or hybrid. In C.
Grau (Ed.), Philosophers explore The Matrix (pp. 177 - 198). Oxford:
Oxford University Press.
Clark, A. (2006). Cognitive complexity and the sensorimotor frontier.
Aristotelian Society Supplementary Volume, 80, 43 65.
Clark, A. (2008). Supersizing the mind: Embodiment, action, and cognitive
extension. Oxford: Oxford University Press.
Clark, A., & Grush, R. (1999). Towards a cognitive robotics. Adaptive
Behavior, 7(1), 5 -16.
Cole, J. (1995). Pride and a daily marathon. Cambridge, MA: MIT Press.
Cole, J., & Paillard, J. (1995). Living without touch and peripheral information
about bodily movement: Studies with deafferented subjects.
Collins, D. F., Refshauge, K. M., Todd, G., & Gandevia, S. C. (2005).
Cutaneous receptors contribute to kinesthesia at the index finger,
elbow, and knee. J Neurophysiol, 94(3), 1699-1706.
Collins, S., Ruina, A., Tedrake, R., & Wisse, M. (2005). Efficient bipedal
robots based on passive-dynamic walkers. Science, 307(5712), 1082 -
1085.
| 329

Collins, S., Wisse, M., & Ruina, A. (2001). A three-dimensional passive-
dynamic walking robot with two legs and knees. The International
Journal of Robotics Research, 20(7), 607-615.
Colman, A. M. (2003). A dictionary of psychology. Oxford: Oxford university
Press.
Crane, T. (2000). The origins of qualia. In T. Crane & S. Patterson (Eds.), The
history of the mind-body problem (pp. 169 - 194). London, New York:
Routeledge.
Crane, T. (2001a). Elements of mind: An introduction to the philosophy of
mind. Oxford: Oxford University Press.
Crane, T. (2001b). Intentional objects. Ratio, 14(4), 336 - 349.
Crane, T. (2003). The mechanical mind: A philosophical introduction to minds
machines and mental representation (2nd ed.). London: Routledge.
Crane, T. (2006a). Brentano's concept of intentional inexistence. In M. Textor
(Ed.), The Austrian contribution to analytic philosophy (pp. 20 - 35).
London: Routledge.
Crane, T. (2006b). Is there a perceptual relation? In T. Gendler & J.
Hawthorne (Eds.), Perceptual experience (pp. 126 - 146). Oxford:
Crane, T. (2009). Is perception a propositional attitude? The Philosophical
Quarterly, 59(236), 452-469.
Craver, C. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Oxford: Oxford University Press.
Cummins, R. E. (1989). Meaning and mental representation. Cambridge, MA:
MIT Press.
Cummins, R. E. (1996). Representations, targets, and attitudes. Cambridge,
MA: MIT Press.
Cummins, R. E. (2000). "How does it work?" versus "What are the laws?":
Two conceptions of psychological explanation. In F. Keil & R. Wilson
(Eds.), Explanation and cognition. Cambridge, MA: MIT Press.
D'Albe, F. (1921). Amplifying the Optophone. Nature, 107, 43-43.
Danto, A. C. (1965). Basic actions. American Philosophical Quarterly, 2(2), 141
- 148.
Davies, M. (1992). Perceptual content and local supervenience. Proceedings
of the Aristotelian Society, 92, 21-45.
Dawkins, R. (1976/2006). The selfish gene. Oxford: Oxford University Press.
de Haan, S., & de Bruin, L. (2010). Reconstructing the minimal self, or how to
make sense of agency and ownership. Phenomenology and the
Cognitive Sciences, 9(3), 373-396.
de Vignemont, F. (2009). Bodily spatial content. Psyche, 15(1), 100 - 108.
de Vignemont, F. (2010). Body schema and body image - Pros and cons.
Neuropsychologia, 48, 669-680.
330 |
de Vignemont, F. (forthcoming-a). Bodily immunity to error. In S. Prosser &
F. Recanati (Eds.), Immunity to error through misidentification.
de Vignemont, F. (forthcoming-b). A mosquito bite against the enactive
approach to bodily experiences. Journal of Philosophy.
de Vignemont, F., & Fourneret, P. (2004). The sense of agency: A
philosophical and empirical review of the Who system.
Consciousness and Cognition, 13(1), 1 - 19.
De Vignemont, F., Tsakiris, M., & Haggard, P. (2006). Body mereology. In G.
Knblich, M. Thornton, M. Grosjean & M. Shiffrar (Eds.), Human
body perception from the inside out (pp. 147-170). New York: Oxford
University Press.
Degenaar, M. (1996). Molyneux's problem: Three centuries of discussion on
the perception of forms. Dordrecht: Kluwer Academic Publishers.
Dennett, D. C. (1969/1983). Content and consciousness (2nd ed.). London:
Routledge.
Dennett, D. C. (1987). The intentional stance. Cambridge, MA: MIT Press.
Dennett, D. C. (1991). Consciousness explained. New York: Back Bay Books.
Dennett, D. C. (1996). Kinds of minds. New York: Basic Books.
Dennett, D. C. (2002a). Brian Cantwell Smith on evolution, objectivity and
intentionality. In H. Clapin (Ed.), Philosophy of mental representation
(pp. 222 - 236). Oxford: Oxford University Press.
Dennett, D. C. (2002b). Quining qualia. In N. Block, O. Flanagan & G.
Gzeldere (Eds.), The nature of consciousness: Philosophical debates
(5th ed., pp. 619 - 642). Cambridge, Mass: MIT Press.
Dennett, D. C., & Kinsbourne, M. (2002). Time and the observer. In N. Block,
O. Flanagan & G. Gzeldere (Eds.), The nature of consciousness:
Philosophical debates (pp. 141 - 174). Cambridge, Mass: MIT Press.
Descartes, R. (1637/1985). Discourse on method (J. Cottingham, R. Stoothoff
& D. Murdoch, Trans.) The philosophical writings of Descartes.
Descartes, R. (1642/1984). Meditations on first philosophy (J. Cottingham, R.
Stoothoff & D. Murdoch, Trans.) The philosophical writings of
Descartes (Vol. 2). Cambridge: Cambridge University Press.
Descartes, R. (1644/1985). Principles of philosophy (J. Cottingham, R.
Stoothoff & D. Murdoch, Trans.) The philosophical writings of
Descartes (Vol. 1, pp. 177 - 292). Cambridge: Cambridge University
Press.
Desmurget, M., & Grafton, S. (2000). Forward modeling allows feedback
control for fast reaching movements. Trends in Cognitive Sciences,
4(11), 423 - 431.
| 331

di Pellegrino, G., L{davas, E., & Farn, A. (1997). Seeing where your hands
are. Nature, 380, 730.
Drake, R. L., Vogl, W., & Mitchell, A. W. M. (2004). Gray's anatomy for
students: Elsevier Inc.
Dretske, F. (1981). Knowledge and the flow of information. Cambridge, MA:
MIT Press.
Dretske, F. (1988). Explaining behavior: Reasons in a world of causes.
Dretske, F. (1995). Naturalizing the mind. Cambridge, MA: MIT Press.
Dreyfus, H. L., & Hall, H. (1982). Husserl, intentionality and cognitive
science. Cambridge, MA: MIT press.
Drummond, J. J. (1979). On seeing a material thing in space: The role of
kinaesthesis in visual perception. Philosophy and Phenomenological
Research, 40, 19-32.
Drummond, J. J. (2003). The structure of intentionality. In D. Welton (Ed.),
The new Husserl: A critical reader (pp. 65 - 92). Bloomington &
Indianapolis: Indiana University Press.
Drummond, J. J. (2007). Phenomenology: Neither auto- nor hetero- be.
Phenomenology and the Cognitive Sciences, 6(1-2), 57-74.
Eilan, N. (1997). Objectivity and the perspective of consciousness. European
Journal of Philosophy, 5, 235-250.
Eilan, N., Marcel, A. J., & Bermdez , J. L. (1995). Self-consciousness and the
body: An interdisciplinary introduction. In J. L. Bermdez , N. M.
Eilan & A. J. Marcel (Eds.), The body and the self (pp. 1 - 28).
Epstein, W. (1995). The metatheoretical context. In W. Epstein & S. Rogers
(Eds.), Perception of space and motion (pp. 1 - 22). San Diego:
Academic Press.
Epstein, W., & Hatfield, G. C. (1978a). Functional equivalence of masking
and cue reduction in perception of shape at a slant. Perception and
Psychophysics, 23(2), 137 - 144.
Epstein, W., & Hatfield, G. C. (1978b). The locus of masking shape-at-a-slant.
Perception and Psychophysics, 24(6), 501 - 504.
Epstein, W., Hatfield, G. C., & Muise, G. (1977). Perceived shape at a slant as a
function of processing time and processing load. Journal of
Experimental Psychology: Human Perception and Performance, 3(3),
473 - 483.
Eriksen, C. W. (1960). Discrimination and learning without awareness: A
methodological survey and evaluation. Psychological Review, 67(5),
279 - 300.
Evans, G. (1982). The varieties of reference. Oxford: Oxford University Press.
Evans, G. (1985). Collected papers. Oxford: Oxford University Press.
332 |
Farkas, K. (forthcoming). A sense of reality. In F. Macpherson & D. Plaitchas
(Eds.), Hallucination. Cambridge, MA: MIT Press.
Fauci, A. S., Braunwald, E., Kasper, D. L., Hauser, S. L., Longo, D. L.,
Jameson, J. L., & Loscalzo, J. (2008). Harrison's principles of internal
medicine (17th ed.). New York: McGraw Hill.
Finger, S., & Hustwit, M. P. (2003). Five early accounts of phantom limb in
context: Pare, Descartes, Lemos, Bell, and Mitchell. Neurosurgery,
52(3), 675-686.
Firth, R. (1949). Sense-data and the percept theory - Part 1. Mind, 58(232),
434-465.
Fodor, J. (1975). The language of thought. Harvard: Harvard University Press.
Fodor, J. (1980). Methodological solipsism considered as a research strategy
in cognitive psychology. Behavioral and Brain Sciences, 3(1), 63 - 73.
Fodor, J., & Pylyshyn, Z. W. (1981). How direct is visual perception? Some
reflections on Gibsons ecological approach. Cognition, 9, 139-196.
Fogassi, L., Gallese, V., Fadiga, L., Luppino, G., Matelli, M., & Rizzolati, G.
(1996). Coding of peripersonal space in inferior premotor cortex (Area
F4). Journal of Neurophysiology, 76, 141-157.
Foster, J. (2008). A world for us: A case for phenomenalistic idealism. New
York: Oxford University Press.
Fourneret, P., & Jeannerod, M. (1998). Limited conscious monitoring of
motor performance in normal subjects. Neuropsychologia, 36, 1133
1140.
Frankfurt, H. G. (1971). Freedom of the will and the concept of a person. The
Journal of Philosophy, 68(1), 5-20.
Frege, G. (1892/1960). On sense and reference (M. Black, Trans.). In M. Black
& P. Geach (Eds.), Translations from the philosophical writings of
Gottlob Frege (pp. 56 - 78). Oxford: Basil Blackwell.
Frith, C. D., Blakemore, S. J., & Wolpert, D. M. (2000a). Abnormalities in the
awareness and control of action. Philosophical Transactions of the
Royal Society of London. Series B, Biological sciences, 355, 1771-1788.
Frith, C. D., Blakemore, S. J., & Wolpert, D. M. (2000b). Explaining the
symptoms of schizophrenia: Abnormalities in the awareness of
action. Brain Research Reviews, 31(2-3), 357-363.
Fukson, O. I., Berkinblit, M. B., & Feldman, A. G. (1980). The spinal frog
takes into account the scheme of its body during the wiping reflex.
Science, 209(4462), 1261-1263.
Gallagher, S. (1995). Body schema and intentionality. In J. L. Bermdez, A.
Marcel & N. Eilan (Eds.), The body and the self (pp. 225244).
Gallagher, S. (2000). Philosophical conceptions of the self: Implications for
cognitive science. Trends in Cognitive Sciences, 4(1), 14-21.
| 333

Gallagher, S. (2003). Bodily self-awareness and object perception. Theoria Et
Historia Scientarum, 7(1), 53 - 68.
Gallagher, S. (2005). How the body shapes the mind. Oxford: Oxford
University Press.
Gallagher, S. (2007). The natural philosophy of agency. Philosophy Compass,
2(2), 347-357.
Gandevia, S. C., Smith, J. L., Crawford, M., Proske, U., & Taylor, J. L. (2006).
Motor commands contribute to human position sense. The Journal of
Physiology, 571(3), 703-710.
Gazzaniga, M. S. (1998). The mind's past. Berkeley, CA: University of
California Press.
Gazzaniga, M. S. (2000). Cerebral specialization and interhemispheric
communication. Brain, 123(7), 1293-1326.
Geach, P. T. (1957). On beliefs about oneself. Analysis, 18(1), 23-24.
Gibson, J. J. (1950). The perception of the visual world. Boston: Houghton
Mifflin.
Gibson, J. J. (1952). The visual field and the visual world: A reply to Professor
Boring. Psychological Review, 59, 149 152.
Gibson, J. J. (1961). Ecological optics. Vision Research, 1(3-4), 253-262.
Gibson, J. J. (1966). The senses considered as perceptual systems. Boston:
Houghton Mifflin.
Gibson, J. J. (1986). The ecological approach to visual perception. Hillsdale,
NJ: Lawrence Erlbaum.
Gibson, J. J., Olum, P., & Rosenblatt, F. (1955). Parallax and perspective
during aircraft landings. The American Journal of Psychology, 68(3),
372-385.
Gillett, C. (2002). The dimensions of realization: A critique of the Standard
view. Analysis, 62(276), 316-323.
Goodman, N. (1969). Languages of art. Oxford: Oxford University Press.
Goodwin, G. M., McCloskey, D. I., & Matthews, P. B. (1972a). The
contribution of muscle afferents to kinaesthesia shown by vibration
induced illusions of movement and by the effects of paralysing joint
afferents. Brain, 95(4), 705-748.
Goodwin, G. M., McCloskey, D. I., & Matthews, P. B. (1972b). Proprioceptive
illusions induced by muscle vibration: contribution by muscle
spindles to perception? Science, 175(28), 1382-1384.
Graziano, M. S., & Cooke, D. F. (2006). Parieto-frontal interactions, personal
space, and defensive behavior. Neuropsychologia, 44, 845-859.
Graziano, M. S., & Gross, C. (1993). A bimodal map of space: Somatosensory
receptive fields in the macaque putamen with corresponding visual
receptive fields. Experimental Brain Research, 97, 96-109.
334 |
Grice, H. P. (1962/2002). Some remarks about the senses. In E. Thompson &
A. Ne (Eds.), Vision and Mind: Selected readings in the philosophy of
perception (pp. 35-55). Cambridge, MA: MIT Press.
Grice, H. P. (1971). Meaning. In J. Rosenberg & C. Travis (Eds.), Readings in
the Philosophy of Language (pp. 436 - 444). New York: Prentice Hall.
Griffiths, P. E. (2001). Genetic information: A metaphor in search of a theory.
Philosophy of Science, 68(3), 394-412.
Grush, R. (1997). The architecture of representation. Philosophical
Psychology, 10(1), 5-23.
Grush, R. (1998). Skill and spatial content. Electronic Journal of Analytic
Philosophy, 6.
Grush, R. (2003). In defense of some 'Cartesian' assumptions concerning the
brain and its operation. Biology and Philosophy, 18, 53-93.
Grush, R. (2004). The emulation theory of representation: Motor control,
imagery, and perception. Behavioral and brain sciences, 27, 377-396.
Grush, R. (2007a). Berkeley and the spatiality of vision. Journal of the History
of Philosophy, 45(3), 413 - 442.
Grush, R. (2007b). Skill theory v2.0: Dispositions, emulation, and spatial
perception. Synthese, 159, 389416.
Grsser, O. J. (1986). Interaction of efferent and afferent signals in visual
perception: A history of ideas and experimental paradigms. Acta
Psychologica, 63(1), 3-21.
Gzeldere, G. (1995). Is consciousness the perception of what passes in one's
own mind? In T. Metzinger (Ed.), Conscious experience (pp. 335 -
359). Schningh: Imprint Academic.
Haddock, A., & Macpherson, F. (Eds.). (2008). Disjunctivism: Perception,
action, knowledge. Oxford: Oxford University Press.
Hare, R. M. (1952). The language of morals. Oxford: Oxford University Press.
Harman, G. (2002). The intrinsic quality of experience. In N. Block, J. R.
Flanagan & G. Gzeldere (Eds.), The nature of consciousness:
Philosophical debates (5th ed., pp. 663 - 675). Cambridge, MA: MIT
Press.
Hatfield, G. C. (2003). Routledge philosophy guidebook to Descartes and the
Meditations. London: Routledge.
Hatfield, G. C., & Epstein, W. (1979). The sensory core and the medieval
foundations of early modern perceptual theory. Isis, 70(3), 363-384.
Haugeland, J. (1981/1998). The nature and plausibility of cognitivism. In J.
Haugeland (Ed.), Having thought (pp. 9 - 46). Cambridge, MA:
Harvard University Press.
Haugeland, J. (1992/1998). Representational genera. In J. Haugeland (Ed.),
Having thought: Essays in the metaphysics of mind (pp. 171 - 206).
Cambridge, MA: Harvard University Press.
| 335

Heck, R. G. (2000). Nonconceptual content and the "space of reasons". The
Philosophical Review, 109(4), 483-523.
Hellie, B. (2007). That which makes the sensation of blue a mental fact:
Moore on phenomenal relationism. European Journal of Philosophy,
15(3), 334-366.
Hempel, C. G., & Oppenheim, P. (1948). Studies in the logic of explanation.
Philosophy of Science, 15(2), 135-175.
Hergenhahn, B. R. (2009). An introduction to the history of psychology (6th
ed.). Belmont, CA: Wadsworth, Cengage Learning.
Hickerson, R. (2007). The history of intentionality: Theories of consciousness
from Brentano to Husserl. London: Continuum.
Hinton, J. M. (1973). Experiences: An inquiry into some ambiguities. Oxford:
Clarendon Press.
Hirstein, W. (2005). Brain fiction: Self-deception and the riddle of
confabulation. Cambridge, MA: MIT Press.
Hobbelen, D. G. E., & Wisse, M. (2007). Limit cycle walking. In M. Hackel
(Ed.), Humanoid robots: Human-like machines (pp. 277 - 294).
Vienna: Itech.
Hobbelen, D. G. E., & Wisse, M. (2008). Ankle actuation for limit cycle
walkers. The International Journal of Robotics Research, 27(6), 709-
735.
Horgan, T. E. (1993). From supervenience to superdupervenience: Meeting
the demands of a material world. Mind, 102(408), 555-586.
Horgan, T. E., & Tienson, J. L. (2002). The intentionality of phenomenology
and the phenomenology of intentionality. In D. J. Chalmers (Ed.),
Philosophy of Mind: Classical and contemporary readings (pp. 520 -
533). Oxford: Oxford University Press.
Hume, D. (1739 - 1740/1960). A treatise of human nature, Being an attempt to
introduce the experimental method of reasoning into moral subjects
(Reprint of the first edition in three volumes). Oxford: Oxford
University Press.
Hume, D. (1777/1975). An enquiry concerning human understanding. In L. A.
Selby-Bigge & P. H. Nidditch (Eds.), Enquiries concerning human
understanding and concerning principles of morals: Oxford University
Press.
Hurlburt, R. T., & Schwitzgebel, E. (2007). Describing inner experience?
Proponent meets sceptic. Cambridge, MA: MIT Press.
Hurley, S. L. (1998). Consciousness in action. Cambridge, MA: Harvard
University Press.
Hurley, S. L. (2010). Varieties of externalism. In R. Menary (Ed.), The
extended mind (pp. 101 - 154). Cambridge, MA: MIT Press.
336 |
Husserl, E. (1900 & 1901). Logische Untersuchungen. Halle an der Saale: Max
Niemeyer.
Husserl, E. (1950). Cartesianische Meditationen und Pariser Vortrge. The
Hague: Martinus Nijhoff Publishers.
Husserl, E. (1950/1999). Cartesian meditations: An introduction to
phenomenology (D. Cairns, Trans.). Dordrecht: Kluwer Academic
Publishers.
Husserl, E. (1952). Ideen zur einer reinen Phnomenologie und
phnomenologischen Philosophie (Zweites Buch): Phnomenologische
Untersuchungen zur Konstitution. The Hague: Martinus Nijhoff
Publishers.
Husserl, E. (1952/1989). Ideas pertaining to a pure phenomenology and to a
phenomenological philosophy (Second book): Studies in the
phenomenology of constitution (R. Rojcewicz & A. Schuwer, Trans.).
Dordrecht: Kluwer Academic Publishers.
Husserl, E. (1966). Analysen zur passiven Synthesis. Aus Vorlesungs- und
Forschungsmanuskripten, 1918-1926. The Hague: Martinus Nijhoff
Publishers.
Husserl, E. (1966/2001). Analyses concerning passive and active synthesis:
Lectures on transcendental logic. Dordrecht: Kluwer Academic
Publishers.
Husserl, E. (1973). Ding und Raum. Vorlesungen 1907. The Hague: Martinus
Nijhoff Publishers.
Husserl, E. (1973/1997). Thing and space: Lectures of 1907 (R. Rojcewicz,
Trans.). Dordrecht: Kluwer Academic Publishers.
Husserl, E. (1976/1982). Ideas pertaining to a pure phenomenology and to a
phenomenological philosophy (First book): General introduction to a
pure phenomenology (F. Kersten, Trans.). The Hague: Martinus
Nijhoff Publishers.
Hutto, D. D. (2005). Knowing what? Radical versus conservative enactivism.
Phenomenology and the Cognitive Sciences, 4, 389405.
Ichikawa, J. (2009). Dreaming and imagination. Mind and Language, 24(1),
103-121.
Iida, F., Gmez, G., & Pfeifer, R. (2005). Exploiting body dynamics for
controlling a running quadruped robot. 12th International Conference
on Advanced Robotics, 229-235.
Iida, F., & Pfeifer, R. (2004). Cheap rapid locomotion of a quadruped robot:
Self-stabilization of bounding gait. Intelligent Autonomous Systems,
8, 642649.
Jackson, F. (1975). On the adverbial anaylsis of visual experience.
Metaphilosophy, 6(2), 127-135.
| 337

Jackson, F. (1977). Perception: A representative theory: Cambridge University
Press.
Jacob, P. (2006). Why visual experience is likely to resist being enacted.
Psyche, 12(1), 1-12.
Jacob, P. (2010). Intentionality. The Stanford encyclopedia of philosophy (Fall
2010 Edition). Retrieved from
http://plato.stanford.edu/archives/fall2010/entries/intentionality/
Jacob, P., & Jeannerod, M. (2003). Ways of seeing: The scope and limits of
visual cognition. Oxford: Oxford University Press.
Jeannerod, M. (2003). Consciousness of action and self-consciousness: A
cognitive neuroscience approach. In J. Rssler & N. M. Eilan (Eds.),
Agency and self-awareness: Issues in philosophy and psychology (pp.
128 - 149). Oxford: Oxford University Press.
Jeffrey, R. C. (1971). Statistical explanation vs. statistical inference. In W.
Salmon, R. C. Jeffrey & J. G. Greeno (Eds.), Statistical explanation and
statistical relevance. Pittsburgh: University of Pittsburgh Press.
Kammers, M. P. M., de Vignemont, F., Verhagen, L., & Dijkerman, H. C.
(2009). The rubber hand illusion in action. Neuropsychologia, 47(1),
204-211.
Kannape, O. A., Schwabe, L., Tadi, T., & Blanke, O. (2010). The limits of
agency in walking humans. Neuropsychologia, 48(6), 1628-1636.
Karinthy, F. (2008). A journey round my skull (V. D. Barker, Trans.). New
York: New York Review of Books Classics.
Katz, D. (1925/1989). The world of touch (L. E. Krueger, Trans.). New Jersey:
Lawrence Erlbaum Associates.
Keele, S. W. (1968). Movement control in skilled motor performance.
Psychological bulletin, 70(6), 387 - 403.
Kelly, S. D. (2004). Seeing things in Merleau-Ponty. In T. Carman (Ed.),
Cambridge companion to Merleau-Ponty (pp. 74110). Cambridge:
Kelly, S. D. (2008). Content and constancy: Phenomenology, psychology, and
the content of perception. Philosophy and Phenomenological
Research, 76(3), 682 - 690.
Kelso, J. A. S. (1995). Dynamic patterns: The self-organisation of brain and
behaviour. Cambridge, MA: MIT Press.
Kelso, J. A. S. (2002). The complementary nature of coordination dynamics :
Self-organization and agency. Nonlinear Phenomena in Complex
Systems, 4, 364 - 371.
Kerkhoff, G., & Schenk, T. (2010). Line bisection in homonymous visual field
defects - Recent findings and future directions. Cortex, In Press,
Corrected Proof.
338 |
Kim, J. (1963). On the logical conditions of deductive explanation. Philosophy
of Science, 30(3), 286-291.
Kim, J. (1984). Concepts of supervenience. Philosophy and Phenomenological
Research, 45(2), 153 - 176.
Kim, J. (1996). Philosophy of mind. Oxford: Westview Press.
Klatzky, R. L., Loomis, J. M., & Golledge, R. G. (1997). Encoding spatial
representations through nonvisually guided locomotion: Tests of
human path integration. The Psychology of Learning and Motivation,
37, 42 - 84.
Klatzky, R. L., MacWhinney, B., & Behrmann, M. (Eds.). (2008).
Embodiment, ego-space and action. New York: Psychology Press.
Knblich, G., Thornton, I. M., Grosjean, M., & Shiffrar, M. (Eds.). (2006).
Human body perception from the inside out. Oxford: Oxford
University Press.
Kosslyn, S. M. (1978). Measuring the visual angle of the mind's eye. Cognitive
Psychology, 10(3), 356 - 389.
L{davas, E. (2002). Functional and dynamic properties of visual peripersonal
space. Trends in Cognitive Sciences, 6, 17-22.
L{davas, E., Farn, A., Zeloni, G., & di Pellegrino, G. (2000). Seeing or not
seeing where your hands are. Experimental Brain Research, 131, 458
467.
Latash, M. (2008a). Neurophysiological basis of movement. Champaign, IL:
Human Kinetics.
Latash, M. (2008b). Synergy. Oxford: Oxford University Press.
Lee, D. N. (1993). Body-environment coupling. In U. Neisser (Ed.), The
perceived self (pp. 43 - 67). Cambridge: Cambridge University Press.
Lee, D. N., & Aronson, E. (1974). Visual proprioceptive control of standing in
human infants. Perception and psychophysics, 15(3), 529.
Lee, D. N., & Kalmus, H. (1980). The optic flow field: The foundation of
vision. Philosophical Transactions of the Royal Society of London. B,
Biological Sciences, 290(1038), 169-179.
Legrand, D. (2006). The bodily self: The sensori-motor roots of pre-reflective
self-consciousness. Phenomenology and the Cognitive Sciences, 5(1),
89-118.
Lehar, S. (2003). The world in your head: A gestalt view of the mechanism of
conscious experience. Mahwah, NJ: Lawrence Erlbaum Associates.
Lenay, C., Canu, S., & Villon, P. (1997). Technology and perception: The
contribution of sensory substitution systems. Paper presented at the
Proceedings of the Second International Conference on Cognitive
Technology, Aizu, Japan, Los Alamitos.
Lenay, C., Gapenne, O., Hanneton, S., Marque, C., & Genoulle, C. (2003).
Sensory substitution: Limits and perspectives. In Y. Hatwell, A. Streri
| 339

& E. Gentaz (Eds.), Touching for knowing: Cognitive psychology of
haptic manual perception (pp. 275 - 292). Amsterdam: John
Benjamins.
Lenggenhager, B., Blanke, O., & Mouthon, M. (2009). Spatial aspects of
bodily self-consciousness. Consciousness and Cognition, 18, 110117.
Lenggenhager, B., Tadi, T., Metzinger, T., & Blanke, O. (2007). Video ergo
sum: Manipulating bodily self-consciousness. Science, 317(5841), 1096-
1099.
Lennon, T. M. (2007). The significance of the Barrovian Case. Studies In
History and Philosophy of Science Part A, 38(1), 36-55.
Leopold, D. A., & Logothetis, N. K. (1999). Multistable phenomena:
Changing views in perception. Trends in Cognitive Sciences, 3(7), 254-
264.
Levine, J. (1983). Materialism and qualia: The explanatory gap. Pacific
philosophical quarterly, 64(October), 354-361.
Lewis, C. I. (1955). Realism or phenomenalism? The Philosophical Review,
64(2), 233-247.
Lishman, J. R., & Lee, D. N. (1973). The autonomy of visual kinaesthesis.
Perception, 2(3), 287-294.
Locke, J. (1690/1997). An Essay concerning Human Understanding. London:
Penguin Classics.
Lowe, E. J. (1995). Locke on human understanding. London: Routeledge.
Ludwig, K. (1996). Explaining why things look as they do. In K. Akins (Ed.),
Perception. New York: Oxford University Press.
Luntley, M. (1999). Contemporary philosophy of thought. Oxford: Blackwell.
Lycan, W. G. (1987). Consciousness. Cambridge, Mass: MIT Press.
Lycan, W. G. (1990). What is the "subjectivity" of the mental. Philosophical
Perspectives, 4, 109-130.
Lycan, W. G. (2002). Consciousness as internal monitoring. In N. Block, J. R.
Flanagan & G. Gzeldere (Eds.), The nature of consciousness:
Philosophical debates (pp. 755 - 772). Cambridge, MA: MIT Press.
Macdonald, C. (2005). Varieties of things: Foundations of contemporary
metaphysics. Oxford: Blackwell.
Mach, E. (1896/1959). The analysis of sensation and the relation of the
psychical to the physical (C. M. Williams, Trans.). New York: Dover
Publications.
Mackay, G. (1888). A discussion on a contribution to the study of
Hemianopsia, with special reference to acquired colour-blindness.
British Medical Journal, 2(1454), 1033-1037.
MacKenzie, M. (2008). Self-awareness without a self: Buddhism and the
reflexivity of awareness. Asian Philosophy, 18(3), 245 - 266.
340 |
Madary, M. (2010). Husserl on perceptual constancy. European Journal of
Philosophy, 18(2).
Mameli, M. (2005). The inheritance of features. Biology and Philosophy,
20(2), 365-399.
Marcel, A. J. (1983). Conscious and unconscious perception: Experiments on
visual masking and word recognition. Cognitive Psychology, 15(2), 197-
237.
Marcel, A. J. (2006). The sense of agency: Awareness and ownership of
action. In J. Rssler & N. M. Eilan (Eds.), Agency and self-awareness:
Issues in philosophy and psychology (pp. 48 - 93). Oxford: Oxford
University Press.
Marconi, T. W. T. (1912, June 27th). Makes light audible: Test of optophone
made at London conference, New York Times.
Martin, M. G. F. (1995). Bodily awareness: A sense of ownership. In J. L.
Bermdez, A. Marcel & N. Eilan (Eds.), The body and the self (pp.
267289). Cambridge, MA: MIT Press.
Martin, M. G. F. (2000). Beyond dispute: Sense-data, intentionality and the
mind-body problem. In T. Crane & S. Patterson (Eds.), History of the
mind-body problem (pp. 195-231). London: Routeledge.
Martin, M. G. F. (2002a). Particular thoughts and singular thought. In A.
O'Hear (Ed.), Logic, Thought, and Language (Vol. 51, pp. 173-214).
Martin, M. G. F. (2002b). The transparency of experience. Mind and
Language, 17, 376-425.
Martin, M. G. F. (2004). The limits of self-awareness. Philosophical Studies,
120(1 - 3), 37 - 89.
Martin, M. G. F. (2009). The reality of appearances. In A. Byrne & H. Logue
(Eds.), Disjunctivism: Contemporary readings (pp. 91 - 116).
Matthen, M. (2005). Seeing, doing, and knowing: A philosophical theory of
sense perception. Oxford: Oxford University Press.
Mattingley, J. B., Driver, J., Beschin, N., & Robertson, I. H. (1997). Attentional
competition between modalities: Extinction between touch and
vision after right hemisphere damage. Neuropsychologia 35, 867880.
Maynard Smith, J. (2000). The concept of information in biology. Philosophy
of Science, 67(2), 177-194.
McDowell, J. H. (1994). Mind and world. Harvard: Harvard University Press.
McDowell, J. H. (1998). Meaning, knowledge and reality. Cambridge, MA:
Harvard University Press.
McGeer, T. (1990). Passive dynamic walking. The International Journal of
Robotics Research, 9, 62 - 82.
| 341

McGraw, M. B. (1940). Neuromuscular development of the human infant as
exemplified in the achievement of erect locomotion. The Journal of
Pediatrics, 17(6), 747 - 771.
McGurk, H., & Macdonald, J. (1976). Hearing lips and seeing voices. Nature,
264(5588), 746-748.
Meinong, A. (1904/1981). The theory of objects. In R. M. Chisholm (Ed.),
Realism and the background of phenomenology (pp. 76 117).
Atascadero, CA: Ridgeview.
Melzack, R., & Bromage, P. R. (1973). Experimental phantom limbs.
Experimental Neurology, 39(2), 261-269.
Mensch, J. R. (2001). Postfoundational phenomenology: Husserlian reflections
on presence and embodiment. University Park, PA: Pennsylvania State
University Press.
Merleau-Ponty, M. (1945/1962). Phenomenology of perception (C. Smith,
Trans.). London: Routledge and Kegan Paul.
Metzinger, T. (2000). The subjectivity of subjective experience: A
representationalist analysis of the first-person perspective. In T.
Metzinger (Ed.), Neural correlates of consciousness: Empirical and
conceptual questions (pp. 285 - 306). Cambridge, MA: MIT Press.
Metzinger, T. (2003a). Being no one: The self-model theory of subjectivity.
Metzinger, T. (2003b). Phenomenal transparency and cognitive self-
reference. Phenomenology and the Cognitive Sciences, 2, 353-393.
Metzinger, T. (2005a). Out-of-body experiences as the origin of the concept
of a "soul". Mind & Matter, 3(1), 57 - 84.
Metzinger, T. (2005b). Prcis: Being no one. Psyche, 11, 1-35.
Metzinger, T. (2006a). Conscious volition and mental representation: Toward
a more fine-grained analysis. In N. Sebanz & W. Prinz (Eds.),
Disorders of volition (pp. 19 - 48). Cambridge, MA: MIT Press.
Metzinger, T. (2006b). Reply to Zahavi: The value of historical scholarship.
Psyche, 12(2), 1 - 4.
Metzinger, T. (2009). The ego tunnel: The science of mind and the myth of the
self. New York: Basic Books.
Metzinger, T. (2011). The no-self alternative. In S. Gallagher (Ed.), The Oxford
handbook of the self (pp. 279 - 296). Oxford: Oxford University Press.
Metzinger, T. (Ed.). (2010). Grundkurs Philosophie des Geistes in drei Bnden:
Band 3: Intentionalitt und mentale Reprsentation. Paderborn:
Mentis.
Metzinger, T., & Gallese, V. (2003). The emergence of a shared action
ontology: Building blocks for a theory. Consciousness and Cognition,
12(4), 549-571.
342 |
Metzinger, T., Lenggenhager, B., Tadi, T., & Blanke, O. (2007). Video ergo
sum: Manipulating bodily self-consciousness. Science, 317(5841), 1096
- 1099.
Miall, R. C. (2003). Connecting mirror neurons and forward models.
NeuroReport, 14(17), 2135-2137.
Millikan, R. G. (1989a). Biosemantics. The Journal of Philosophy, 86(6), 281-
297.
Millikan, R. G. (1989b). In defense of proper functions. Philosophy of Science,
56(2), 288-302.
Millikan, R. G. (1991). Perceptual content and fregean myth. Mind, 100(4),
439-459.
Millikan, R. G. (1993). White queen psychology and other essays for Alice.
Milner, A. D., & Goodale, M. A. (1995). The visual brain in action. Oxford:
Mitchell, S. W. (1871). Phantom limbs. Lippincott Magazine, 8, 563-569.
Molyneux, W. (1692). Dioptrica nova: A treatise of dioptricks, in two parts,
wherein the various effects and appearances of spherick glasses, both
convex and concave, single and combined, in telescopes and
microscopes, together with their usefulness in many concerns of
humane life, are explained. London: Printed for Benj. Tooke.
Montague, M. (forthcoming). The phenomenology of particularity. In T.
Bayne & M. Montague (Eds.), Cognitive phenomenology. Oxford:
Moore, A. W. (1986). How significant is the use/mention distinction?
Analysis, 46(4), 173-179.
Moore, A. W. (1997). Points of view. Oxford: Oxford University Press.
Moore, G. E. (1903). The refutation of idealism. Mind, 12, 433 - 453.
Moore, G. E. (1922). Philosophical studies. London: Harcourt, Brace and Co.
Moore, G. E. (1925/1960). A defence of common sense. In R. M. Chisholm
(Ed.), Realism and the background of phenomenology (pp. 255 - 282).
Illinois: The Free Press.
Moore, G. E. (1953). Some main problems of philosophy. London: George
Allen & Unwin.
Moran, D. (1996). The inaugural address: Brentano's thesis. Proceedings of
the Aristotelian Society, Supplementary Volumes, 70, 1-27.
Moran, D. (2000). Introduction to phenomenology. London: Routledge.
Moran, R. (2001). Authority and estrangement: An essay on self-knowledge.
Princeton: Princeton University Press.
Nagel, S. K., Carl, C., Kringe, T., Mrtin, R., & Knig, P. (2005). Beyond
sensory substitution - learning the sixth sense. Journal of Neural
Engineering, 2(4), R13 - R26.
| 343

Nagel, T. (1989). The view from nowhere. Oxford: Oxford University Press.
Newen, A., & Vogeley, K. (2003). Self-representation: Searching for a neural
signature of self-consciousness. Consciousness and Cognition, 12, 529
- 543.
Newman, L. (2010). Descartes epistemology. The Stanford encyclopedia of
philosophy (Summer 2010 Edition). Retrieved from
<http://plato.stanford.edu/archives/sum2010/entries/descartes-
epistemology/>
Nichols, R. (2007). Thomas Reid's theory of perception. Oxford: Oxford
University Press.
No, A. (2001). Experience and the active mind. Synthese, 61(1), 41-60.
No, A. (2002). On what we see. Pacific philosophical quarterly, 83, 57 - 80.
No, A. (2004). Action in perception. Cambridge, MA: MIT Press.
No, A. (2005). Against intellectualism. Analysis, 65(288), 278-290.
No, A. (2008). Reply to Campbell, Martin, and Kelly. Philosophy and
Phenomenological Research, 76, 691 - 706.
No, A. (2009). Conscious reference. The Philosophical Quarterly, 59, 470
482.
Norman, D. (1988). The pyschology of everyday things. New York: Basic
books.
Norrsell, U., Finger, S., & Lajonchere, C. (1999). Cutaneous sensory spots and
the "law of specific nerve energies": history and development of ideas.
Brain Research Bulletin, 48(5), 457-465.
Nudds, M. (2009). Recent work in perception: Naive realism and its
opponents. Analysis, 69, 334 - 346.
O'Brien, G., & Opie, J. (2004). Notes toward a structuralist theory of mental
representation. In H. Clapin, P. Staines & P. Slezak (Eds.),
Representation in mind: New approaches to mental representation (pp.
1 - 20). Amsterdam: Elsevier.
O'Regan, J. K., & No, A. (2001). A sensorimotor account of vision and visual
consciousness. Behavioral and brain sciences, 24, 939973.
O'Shaughnessy, B. (1973). Trying (as the mental "pineal gland"). The Journal
of Philosophy, 70(13), 365-386.
O'Shaughnessy, B. (1980/2008). The will: A dual aspect theory (Vol. 1).
O'Shaughnessy, B. (1984). Seeing the light. Proceedings of the Aristotelian
Society, 85, 193-218.
O'Shaughnessy, B. (1989). The sense of touch. Australasian Journal of
Philosophy, 67, 37-58.
O'Shaughnessy, B. (2000). Consciousness and the world. Oxford: Oxford
University Press.
344 |
O'Shaughnessy, B. (2002). Dreaming. Inquiry: An Interdisciplinary Journal of
Philosophy, 45(4), 399 - 432.
Overgaard, S. (2010). On the looks of things. Pacific philosophical quarterly,
91(2), 260 - 284.
Pacherie, E. (1999). Leibhaftigkeit and representational theories of
perception. In J. Petitot, F. J. Varela, B. Pachoud & J.-M. Roy (Eds.),
Naturalizing phenomenology: Issues in contemporary phenomenology
and cognitive science (pp. 148 - 160). Stanford: Stanford University
Press.
Pacherie, E. (2008). The phenomenology of action: A conceptual framework.
Cognition, 107, 179-217.
Paillard, J. (2005). Vectorial versus configural encoding of body space: A
neural basis for a distinction between body schema and body image.
In H. De Preester & V. Knockaert (Eds.), Body image and body
schema: Interdisciplinary perspectives on the body (pp. 111 - 126).
Amsterdam: John Benjamins Publishing.
Palmer, S. (1999). Vision science: From photons to phenomenology. Boston:
MIT Press.
Pappas, G. S. (1989). Sensation and perception in Reid. Nos, 23(2), 155-167.
Pappas, G. S. (1999). Berkeley and scepticism. Philosophy and
Phenomenological Research, 59(1), 133-149.
Pappas, G. S. (2002). Abstract ideas and the new theory of vision. British
Journal for the History of Philosophy, 10(1), 55 - 69.
Peacocke, C. (1992). A study of concepts. Cambridge, MA: MIT Press.
Peacocke, C. (1998). Review: Nonconceptual content defended. Philosophy
and Phenomenological Research, 58(2), 381-388.
Peacocke, C. (2006). Action: Awareness, ownership, and knowledge. In J.
Rssler & N. M. Eilan (Eds.), Agency and self-awareness: Issues in
philosophy and psychology (pp. 94 - 110). Oxford: Oxford University
Press.
Perler, D. (Ed.). (2001). Ancient and medieval theories of intentionality.
Leiden: Brill.
Perry, J. (1979). The problem of the essential indexical. Nos, 13(1), 3-21.
Pershey, E. J. (2005). Celebrating the invention of flight in a hands-on way:
Replicating the 1902 experimental glider flights of the Wright
brothers. In M. D. Bowles & V. P. Dawson (Eds.), Realizing the dream
of flight: Biographical essays in honor of the centennial of flight, 1903-
2003 (pp. 257 - 271). Washington, D.C.: NASA History Division, Office
of External Relations.
Pfeifer, R., & Bongard, J. (2007). How the body shapes the way we think: A
new view of intelligence. Cambridge, MA: MIT Press.
| 345

Place, U. T. (1956). Is consciousness a brain process? British Journal of
Psychology, 47, 44 - 50.
Plato. (388-367 BCE/1973). Theaetetus (J. H. McDowell, Trans.). Oxford:
Prechtl, H. F. R. (1997). The importance of fetal movements. In K. J. Connolly
& H. Forssberg (Eds.), Neurophysiology & neuropsychology of motor
development (pp. 42 - 53). London: Mac Keith Press.
Price, H. H. (1932). Perception. London: Methuen.
Prinz, J. (2009). Is consciousness embodied? In P. Robbins & M. Aydede
(Eds.), The Cambridge handbook of situated cognition (pp. 419 - 437).
New York: Cambridge University Press.
Proske, U., & Gandevia, S. C. (2009). The kinaesthetic senses. The Journal of
Physiology, 587(17), 4139-4146.
Proske, U., Schaible, H. G., & Schmidt, R. F. (1988). Joint receptors and
kinaesthesia. Experimental Brain Research, 72(2), 219-224.
Psillos, S. (2007). Past and contemporary perspectives on explanation. In T.
Kuipers (Ed.), Handbook of the philosophy of science: General
philosophy of science - Focal issues (pp. 97 - 173). Amsterdam:
Elsevier.
Pylyshyn, Z. W. (2000). Situating vision in the world. Trends in Cognitive
Sciences, 4(5), 197-207.
Ramachandran, V. S., & Hirstein, W. (1998). The perception of phantom
limbs. The D. O. Hebb lecture. Brain, 121(9), 1603-1630.
Ramsey, W. M. (2007). Representation reconsidered. Cambridge: Cambridge
University Press.
Reed, E. S. (1994). The separation of psychology from philosophy: Studies in
the sciences of mind 1815-1879. In C. L. Ten (Ed.), The nineteenth
century: Routledge history of philosophy (Vol. 7, pp. 297 - 356).
London: Routledge.
Reichenbach, H. (1938/1961). Experience and prediction (5th ed.). Chicago:
University of Chicago Press.
Reid, T. (1764/1997). An inquiry into the human mind on the principles of
common sense. Edinburgh: Edinburgh University Press.
Reid, T. (1785). Essays on the intellectual powers of man. London: (British
Library) John Bell, Parliament Square, and G. G. J. & J. Robinson.
Revonsuo, A. (2000). Prospects for a scientific research program on
consciousness. In T. Metzinger (Ed.), Neural correlates of
consciousness: Empirical and conceptual questions (pp. 57 - 76).
Revonsuo, A. (2010). Consciousness: The science of subjectivity. Hove, New
York: Psychology Press.
Rickless, S. C. (2005). The cartesian fallacy fallacy. Nos, 39(2), 309 - 336.
346 |
Robinson, H. (1994). Perception. London: Routeledge.
Rock, I. (1981). The logic of perception. Cambridge, MA: MIT Press.
Roser, M., & Gazzaniga, M. S. (2004). Automatic brains & interpretive minds.
Current Directions in Psychological Science, 13(2), 56-59.
Ross, H. E., & Bischof, K. (1981). Wundt's views on sensations of innervation:
A reevaluation. Perception, 10(3), 319 - 329.
Rossetti, Y., Rode, G., & Boisson, D. (1995). Implicit processing of
somaesthetic information: a dissociation between where and how?
NeuroReport, 6(3), 506-510.
Rowlands, M. (2006). Body language: Representation in action. Cambridge,
MA: MIT Press.
Rupert, R. (2004). Challenges to the hypothesis of extended cognition.
Journal of Philosophy, 101, 389428.
Rupert, R. (2009). Cognitive systems and the extended mind. New York:
Russell, B. (1905). On denoting. Mind, 14(56), 479 - 493.
Russell, B. (1910-1911). Knowledge by acquaintance and knowledge by
description. Proceedings of the Aristotelian Society, New Series, v(XI),
108 - 128.
Russell, B. (1912/1959). The problems of philosophy. Oxford: Oxford
University Press.
Russell, B. (1913/1984). Theory of knowledge: The 1913 manuscript. In misc.
(Ed.), The collected papers of Bertrand Russell (Vol. 7). London:
George Allen & Unwin.
Russell, B. (1914). Our knowledge of the external world. London: George Allen
& Unwin.
Russell, B. (1918/2010). The philosophy of logical atomism. London:
Routledge.
Ryan, T. (2006). A new account of Berkeley's likeness principle. British
Journal for the History of Philosophy, 14(4), 561 - 580.
Ryle, G. (1949/2000). The concept of mind. London: Penguin Classics.
Sacks, O. (1986). The man who mistook his wife for a hat. London: Picador.
Salmon, W. (1971). Statistical explanation. In W. Salmon, R. C. Jeffrey & J. G.
Greeno (Eds.), Statistical explanation & statistical relevance (pp. 29 -
88). Pittsburgh: University of Pittsburgh Press.
Sartre, J.-P. (1943/1992). Being and nothingness: A phenomenological essay on
ontology. New York: Washington Square Press.
Schellenberg, S. (2010). The particularity and phenomenology of perceptual
experience. Philosophical Studies, 149, 19 - 48.
Scholl, B. J. (2007). Object persistence in Philosophy and Psychology. Mind &
Language, 22(5), 563-591.
| 347

Schwartz, O. H. (2010). Visual perception: A clinical orientation (4th ed.).
New York: McGraw Hill.
Schwartz, R. (1994). Vision: Variations on some Berklean themes. Cambridge,
Mass: Blackwell.
Schwitzgebel, E. (2006). Do things look flat? Philosophy and
Searle, J. (1983). Intentionality: An essay in the philosophy of mind.
Searle, J. (1990). Consciousness, explanatory inversion, and cognitive science.
Behavioural and Brain sciences, 13, 585 - 642.
Searle, J. (1991). Consciousness, unconsciousness and intentionality.
Philosophical Issues, 1, 45-66.
Segal, G. (2005). Intentionality. In F. Jackson & M. Smith (Eds.), Oxford
handbook of contemporary philosophy (pp. 283 - 288). Oxford: Oxford
university press.
Sellars, W. (1956). Empiricism and the philosophy of mind. In H. Feigl & M.
Scrivens (Eds.), Minnesota Studies in the Philosophy of Science (Vol. 1,
pp. 253-329). Minneapolis: University of Minnesota Press.
Shergill, S. S., Bays, P. M., Frith, C. D., & Wolpert, D. M. (2003). Two eyes for
an eye: The neuroscience of force escalation. Science, 301(5630), 187.
Sherrington, C. (1894). On the anatomical constitution of nerves of skeletal
muscles; with remarks on recurrent fibres in the ventral spinal nerve-
root. The Journal of Physiology, 17(3-4), 210-258.
Sherrington, C. (1906). The intergrative action of the nervous system. New
Haven, CT: Yale University Press.
Shoemaker, S. (1968). Self-reference and self-awareness. Journal of
Philosophy, 65(19), 555-567.
Shoemaker, S. (1990). Qualities and qualia: What's in the mind? Philosophy
and Phenomenological Research, 50, 109-131.
Shoemaker, S. (1996). The first person perspective and other essays.
Shoemaker, S. (2003). Embodiment and behaviour. In S. Shoemaker (Ed.),
Identity, cause and mind (pp. 113 - 138). Oxford: Oxford University
Press.
Siegel, S. (2010). The contents of perception. The Stanford encyclopaedia of
philosophy (Fall 2010 edition). Retrieved from
http://plato.stanford.edu/archives/fall2010/entries/perception-contents/
Siewert, C. (2008). Consciousness and intentionality. The Stanford
encyclopedia of philosophy (Fall 2008 Edition). Retrieved from
http://plato.stanford.edu/archives/fall2008/entries/consciousness-
intentionality/
Slachevsky, A., Pillon, B., Fourneret, P., Pradat-diehl, P., Jeannerod, M., &
Dubois, B. (2001). Preserved adjustment but impaired awareness in a
348 |
sensory-motor conflict following prefrontal lesions. Journal of
Cognitive Neuroscience, 13, 332-340.
Smeets, J., Brenner, E., de Grave, D., & Cuijpers, R. (2002). Illusions in action:
Consequences of inconsistent processing of spatial attributes.
Experimental Brain Research, 147(2), 135-144.
Smith, A. D. (2000). Space and sight. Mind, 109(435), 481-518.
Smith, A. D. (2001). Perception and belief. Philosophy and Phenomenological
Research, 62(2), 283-309.
Smith, A. D. (2002). The problem of perception. Harvard: Harvard University
Press.
Smith, A. D. (2003). Husserl and the Cartesian Meditations. London:
Routledge.
Smith, A. D. (2006). In defence of Direct Realism. Philosophy and
Smith, A. D. (2007). The flesh of perception: Merleau-Ponty and Husserl. In
T. Baldwin (Ed.), Reading Merleau-Ponty: On the phenomenology of
perception (pp. 1 - 23). London: Routledge.
Smith, A. D. (2008). Husserl and externalism. Synthese, 160(3), 313 - 333.
Smith, A. J. T. (2009). Acting on (bodily) experience. Psyche, 15(1), 82-99.
Smith, B. C. (1996). On the origin of objects. Cambridge, MA: MIT Press.
Smythies, J. (1996). A note on the concept of the visual field in neurology,
psychology, and visual neuroscience. Perception, 25(3), 369 - 371.
Snowdon, P. (1980 - 1981). Perception, vision, and causation. Proceedings of
the Aristotelian Society, 81, 175-192.
Snowdon, P. (1992). How to interpret 'direct perception'. In T. Crane (Ed.),
The contents of experience: Essays on perception (pp. 48-78).
Snowdon, P. (2005). The formulation of disjunctivism: A response to Fish.
Proceedings of the Aristotelian Society, 105, 129 - 141.
Snowdon, P. (2008). Hinton and the origins of disjunctivism. In A. Haddock
& F. Macpherson (Eds.), Disjunctivism: Perception, action, knowledge
(pp. 35 - 57). Oxford: Oxford University Press.
Sorabji, R. (1971). Aristotle on demarcating the five senses. The Philosophical
Review, 80(1), 55-79.
Soteriou, M. (2009). The disjunctive theory of perception. The Stanford
encyclopedia of philosophy (Spring 2009 Edition). Retrieved from
http://plato.stanford.edu/archives/fall2009/entries/perception-disjunctive/
Spelke, E. S. (1990). Principles of object perception. Cognitive Science, 14(1),
29-56.
Spence, C., Pavani, F., Maravita, A., & Holmes, N. P. (2004). Multisensory
contributions to the 3-D representation of visuotactile peripersonal
| 349

space in humans: evidence from the crossmodal congruency task.
Journal of Physiology, 98, 171-189.
Sperling, G. (1963). A model for visual memory tasks. Human Factors, 5, 19-
31.
Sperry, R. W. (1950). Neural basis of the spontaneous optokinetic response
produced by visual inversion. Journal of Computational Physiological
Psychology, 43(6), 482-489.
Spirkovska, L. (2004). Summary of tactile user interfaces techniques and
systems. National Aeronautics and Space Administration. Ames
Research Center, Moffett Field, California. Retrieved from
http://ti.arc.nasa.gov/m/pub/archive/0905.pdf
Stampe, D. W. (1977). Toward a causal theory of linguistic representation.
Midwest Studies in Philosophy, 2(1), 42 - 63.
Stanley, J., & Williamson, T. (2001). Knowing how. The Journal of Philosophy,
411-444.
Sterelny, K. (1990). The representational theory of mind: An introduction.
Oxford: Basil Blackwell.
Sterelny, K., & Griffiths, P. E. (1999). Sex and death: An introduction to the
philosophy of biology. Chicago: University of Chicago Press.
Stich, S. (1980). Paying the price for methodological solipsism. Behavioral
and Brain Sciences, 3(1), 97 - 98.
Stich, S. (1983). From folk psychology to cognitive science. Cambridge, MA:
MIT Press.
Stich, S. (1991). Do true believers exist? Proceedings of the Aristotelian
Society: Supplementary volume, 65, 229 - 244.
Stoffregen, T. A. (2003). Affordances as properties of the animal-environment
system. Ecological psychology, 15, 115 - 134.
Strawson, P. F. (1959/2003). Individuals: An essay in descriptive metaphysics.
London: Routledge.
Strawson, P. F. (1966). The bounds of sense: An essay on Kant's Critique of
Pure Reason. London: Methuen.
Strawson, P. F. (1979/2002). Perception and its objects. In E. Thompson & A.
Ne (Eds.), Vision and Mind: Selected readings in the philosophy of
perception (pp. 91 - 111). Cambridge, MA: MIT Press.
Sturgeon, S. (2000). Matters of mind. London: Routledge.
Synofzik, M., Vosgerau, G., & Newen, A. (2008). Beyond the comparator
model: A multifactorial two-step account of agency. Consciousness
and Cognition, 17(1), 219-239.
Taube, J., Muller, R., & Ranck, J., Jr. (1990). Head-direction cells recorded
from the postsubiculum in freely moving rats. I. Description and
quantitative analysis. J. Neurosci., 10(2), 420-435.
350 |
Taylor, C. (1978). The validity of transcendental arguments. Proceedings of
the Aristotelian Society, 79, 151-165.
Thelen, E. (1986). Treadmill-elicited stepping in seven-month-old infants.
Child Development, 57(6), 1498-1506.
Thelen, E., & Fisher, D. M. (1982). Newborn stepping: An explanation for a
"disappearing" reflex. Developmental Psychology, 18(5), 760-775.
Thelen, E., Fisher, D. M., & Ridley-Johnson, R. (1984). The relationship
between physical growth and a newborn reflex. Infant Behavior and
Development, 7(4), 479-493.
Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the
development of cognition and action. Cambridge, MA: MIT Press.
Thelen, E., & Ulrich, B. D. (1991). Hidden skills: A dynamic systems analysis
of treadmill stepping during the first year. Monographs of the Society
for Research in Child Development, 56(1), 1 - 98.
Thelen, E., Ulrich, B. D., & Niles, D. (1987). Bilateral coordination in human
infants: Stepping on a split-belt treadmill. Journal of Experimental
Psychology: Human Perception and Performance, 13(3), 405 410.
Thompson, E. (2005). Sensorimotor subjectivity and the enactive approach
to experience. Phenomenology and the Cognitive Sciences, 4, 407427.
Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences
of mind. Cambridge, MA: The Belknap Press of Harvard University
Press.
Thrane, G. (1977). Berkeley's "proper object of vision". Journal of the History
of Ideas, 38(2), 243-260.
Todes, S. (2001). Body and world. Cambridge, MA: MIT Press.
Tsakiris, M., & Haggard, P. (2005). Experimenting with the acting self.
Cognitive Neuropsychology, 22, 387-407.
Tsakiris, M., Prabhu, G., & Haggard, P. (2006). Having a body versus moving
your body: How agency structures body-ownership. Consciousness
and cognition, 15, 423-432.
Tunik, E., Houk, J. C., & Grafton, S. T. (2009). Basal ganglia contribution to
the initiation of corrective submovements. NeuroImage, 47(4), 1757-
1766.
Turbayne, C. M. (1955). Berkeley and Molyneux on retinal images. Journal of
the History of Ideas, 16(3), 339-355.
Turbayne, C. M. (1962). The myth of metaphor. New Haven, CT: Yale
University Press.
Turvey, M. T. (1973). On peripheral and central processes in vision: Inferences
from an information-processing analysis of masking with patterned
stimuli. Psychological Review, 80, 1 - 52.
Turvey, M. T. (1990). Coordination. American Psychologist, 45, 938-953.
| 351

Tye, M. (1995). Ten problems of consciousness: A representational theory of
the phenomenal mind. Cambridge, MA: MIT Press.
Tye, M. (1996). Perceptual experience is a many-layered thing. Philosophical
Issues, 7, 117-126.
Tye, M. (2000). Consciousness, color, content. Cambridge, MA: MIT Press.
Tye, M. (2002). Representationalism and the transparency of experience.
Nos, 36(1), 137-151.
Tye, M. (2003). Consciousness and persons: Unity and identity. Cambridge,
MA: MIT Press
Tye, M. (2007). Intentionalism and the argument from no common content.
Philosophical Perspectives, 21(1), 589-613.
Ulrich, B. D. (1997). Dynamic systems theory and skill development in
infants and children. In K. J. Connolly & H. Forssberg (Eds.),
Neurophysiology & neuropsychology of motor development (pp. 319 -
345). London: Mac Keith Press.
Valberg, J. J. (1992). The puzzle of experience. In T. Crane (Ed.), The contents
of experience: Essays on perception (pp. 18 - 47). Cambridge:
Van Gelder, T. (1995). What might cognition be, if not computation? The
Journal of Philosophy, 92(7), 345-381.
Van Gelder, T. (1998). The roles of philosophy in cognitive science.
Philosophical Psychology, 11(2), 117 - 136.
Vander, A. J., Sherman, J. H., & Luciano, D. S. (2001). Human physiology: The
mechanism of body function (8th ed.). New York: McGraw-Hill.
Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitive
science and human experience. Cambridge: MIT Press.
Veasey, C. A. (1904). Observations of a case of bi-temporal hemianopsia with
some unusual changes in the visual fields. Transactions of the
American Ophthalmological Society, 10(2), 383-387.
Von Eckardt, B. (1993). What is cognitive science? Cambridge, MA: MIT
Press.
von Holst, E. (1954/1996). Relations between the central nervous system and
the peripheral organs. In L. D. Houck & L. C. Drickamer (Eds.),
Foundations of animal behavior: Classic papers with commentaries
(pp. 414 - 419). Chicago: University of Chicago Press.
von Senden, M. (1932/1960). Space and sight: The perception of space and
shape in the congenitally blind before and after operation (P. Heath,
Trans.). London: Methuen & co. Ltd.
Ward, D., Roberts, T., & Clark, A. (forthcoming). Knowing what we can do:
actions, intentions, and the construction of phenomenal experience.
Synthese.
352 |
Warren, W., & Wang, S. (1987). Visual guidance of walking through
apertures: Body-scaled information for affordances. Journal of
Experimental Psychology: Human Perception and Performance, 13(3),
371 - 383.
Wegner, D. M., Sparrow, B., & Winerman, L. (2004). Vicarious agency:
Experiencing control over the movements of others. Journal of
Personality and Social Psychology, 86(6), 838 - 848.
Weisberg, J. (2005). Consciousness constrained: A commentary on Thomas
Metzinger's Being no one. Psyche, 8(17).
Wexler, M., Panerai, F., Lamouret, I., & Droulez, J. (2001). Self-motion and
the perception of stationary objects. Nature, 409(6816), 85-88.
Wexler, M., & van Boxtel, J. J. A. (2005). Depth perception by the active
observer. Trends in Cognitive Sciences, 9(9), 431-438.
Wheeler, M. (2005). Reconstructing the cognitive world: The next step.
Wheeler, M., & Clark, A. (1999). Genic representation: Reconciling content
and causal complexity. British Journal for the Philosophy of Science,
50(1), 103-135. doi: 10.1093/bjps/50.1.103
Wiggins, D. (1997). Sortal concepts: A reply to Xu. Mind & Language, 12(3-4),
413-421.
Williams, B. (1978/2005). Descartes: The project of pure enquiry. Oxford:
Routledge.
Williams, M. (2001). Problems of knowledge: A critical introduction to
epistemology. Oxford: Oxford University Press.
Wilson, M. D. (1978). Descartes. London: Routledge.
Wilson, M. L. (2002). Six views of embodied cognition. Psychonomic bulletin
& review, 9(4), 625 - 636.
Wilson, R., & Keil, F. (2000). The shadows and shallows of explanation. In F.
Keil & R. Wilson (Eds.), Explanation and cognition (pp. 87 - 114).
Wilson, R. A. (2004). Boundaries of the mind: The individual in the fragile
sciences. Cambridge: Cambridge University Press.
Wilson, R. A. (2005). Genes and the agents of life: The individual in the fragile
sciences. Cambridge: Cambridge University Press.
Windt, J. M. (2010). The immersive spatiotemporal hallucination model of
dreaming. Phenomenology and the Cognitive Sciences, 9(2), 295 - 316.
Wisse, M. (2005). Three additions to passive dynamic walking; actuation, an
upper body, and 3D stability. 4th IEEE/RAS International Conference
on Humanoid Robots, 1, 113 - 132.
Wisse, M., & Frankenhuyzen, J. (2006). Design and construction of MIKE; a
2-D autonomous biped based on passive dynamic walking. In H.
| 353

Kimura, K. Tsuchiya, A. Ishiguro & H. Witte (Eds.), Adaptive motion
of animals and machines (pp. 143-154). Tokyo: Springer.
Wittgenstein, L. (1953/2001). Philosophical investigations (G. E. M.
Anscombe, Trans. 3rd ed.). Malden, MA: Blackwell.
Wittgenstein, L. (1958). The blue and brown books. Oxford: Blackwell.
Wolpert, D. M. (2003). A unifying computational framework for motor
control and social interaction. Philosophical Transactions of the Royal
Society of London. B, Biological Sciences, 358, 593-602.
Wolpert, D. M., Ghahramani, Z., & Flanagan, J. R. (2001). Perspectives and
problems in motor learning. Trends in Cognitive Sciences, 5(11), 487-
494.
Wolpert, D. M., Ghahramani, Z., & Jordan, M. (1995). An internal model for
sensorimotor integration. Science, 269(5232), 1880-1882.
Wolpert, D. M., Miall, R. C., & Kawato, M. (1998). Internal models in the
cerebellum. Trends in Cognitive Sciences, 2, 338-347.
Xu, F. (1997). From Lot's wife to a pillar of salt: Evidence that physical object
is a sortal concept. Mind & Language, 12(3-4), 365-392.
Xu, F., & Carey, S. (1996). Infants' metaphysics: The case of numerical
identity. Cognitive Psychology, 30(2), 111-153.
Yang, J. F., Lamont, E. V., & Pang, M. Y. C. (2005). Split-belt treadmill
stepping in infants suggests autonomous pattern generators for the
left and right leg in humans. Journal of Neuroscience, 25(29), 6869-
6876.
Yolton, J. W. (1996). Perception & reality: A history from Descartes to Kant.
Ithaca, NY: Cornell University Press.
Zahavi, D. (1994). Husserl's phenomenology of the body. Etudes
Phnomnologiques, 10, 6384.
Zahavi, D. (1999). Self-awareness and alterity: A phenomenological
investigation. Evanston, Illinois: Northwestern University Press.
Zahavi, D. (2002). Merleau-Ponty on Husserl: A reappraisal. In L. Embree &
T. Toadvine (Eds.), Merleau-Ponty's reading of Husserl (pp. 329).
Dordrecht: Kluwer Academic Publishers.
Zahavi, D. (2003a). Husserl's phenomenology. Stanford, CA: Stanford
University Press.
Zahavi, D. (2003b). Phenomenology and metaphysics. In D. Zahavi, S.
Heinmaa & H. Ruin (Eds.), Metaphysics, facticity and interpretation:
Phenomenology in the Nordic countries (pp. 3 - 22). Dordrecht:
Kluwer Academic Publishers.
Zahavi, D. (2005). Subjectivity and selfhood: Investigating the first-person
perspective. Cambridge, MA: MIT Press.
Zelazo, P. R. (1984). The development of walking: New findings and old
assumptions. Journal of Motor Behaviour, 15, 99 - 137.
354 |
Zelazo, P. R., Zelazo, N. A., & Kolb, S. (1972). "Walking" in the newborn.
Science, 176(4032), 314-315.

Untitled

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Untitled

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

Untitled

Uploaded by

Copyright:

Available Formats

THE STRUCTURING BODY

A Critical Study in the Description &

You might also like

Untitled

Uploaded by

Document Informationclick to expand document informationA critical study in the description and explanation of perceptual experience. The author argues that perceptual experi ence is a'sense-data' experience.

Document Informationclick to expand document information

Copyright:

Available Formats

Untitled

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

Untitled

Uploaded by

Copyright:

Available Formats

THE STRUCTURING BODY

A Critical Study in the Description &

You might also like