You are on page 1of 368

The Ways Things Look

Richard Price

All Souls College, University of Oxford

Submitted for the degree of Doctor of Philosophy

Trinity Term, 2006

Word Count: 74,984 (excluding bibliographical references)


76,084 (including bibliographical references)

1
The Ways Things Look
(Abstract)

Richard Price, All Souls College, Oxford University


D.Phil, Trinity Term, 2006

Philosophers have often raised the question what kind of information is available to
vision. For instance, Berkeley argued that one could not see depth, Hume argued that one could
not see necessary connections and, according to Paul Guyer, Kant held that there is no perception
of change, but only change of perception (Guyer 2004).

Recently philosophers have asked what kinds of properties visual experience represents.
According to sparse views, visual experience represents a sparse range of properties, for instance
just colours, shapes, positions and sizes (see McGinn 1982, Burge 2003, Millar 2000). According
to rich views, visual experience represents a rich range of properties, for instance properties such
as being a tomato, and being sad (see Peacocke 2003, Siegel 2006, Searle 1983, McDowell
1998b).

In this thesis I discuss some questions arising from the above debate. Instead of using the
terminology of what properties visual experience represents, I define a kind of looking,
phenomenal looking, which is individuated in terms of differences in visual phenomenal
character, and I raise the question what kinds of properties objects phenomenally look to have.

A summary of the thesis is as follows:

(1) I argue that objects phenomenally look to have only colours and positions.
Furthermore, I argue that the colour properties include only shades of colour, and not
determinables such as being red. Thus I defend a very sparse view about the
properties that objects phenomenally look to us to have.

(2) I argue that objects do not phenomenally look to the left and to the right of me, and
that objects do not phenomenally look to bear visual angles to me.

(3) (2) raises a puzzle about what kinds of position properties objects do phenomenally
look to have. I defend primitivism, according to which the position properties that
objects phenomenally look to have are sui generis, absolute, properties, and thus are
not relations to observers or other objects.

(4) I argue that necessarily, objects do not have the position properties that they
phenomenally look to us to have, and that necessarily, objects do not have the colour
properties that they phenomenally look to us to have.

(5) I argue that the position properties that objects phenomenally look to us to have
include x, y and z coordinates, and not merely x and y coordinates. That is, I argue
that we see in 3D rather than in 2D.

(6) I raise the question what kinds of colour and location properties objects would
phenomenally look to beings with more coarse-grained discriminatory abilities than

2
us. I reach no firm conclusions about this topic, but discuss some ways in one might
approach it.

3
Acknowledgements

There are many people and many institutions to whom I owe thanks. This thesis

built upon my B.Phil thesis, and I am grateful to the AHRC, which gave me funding for

the B.Phil. I am grateful to Corpus Christi College and the Faculty of Philosophy, which

jointly funded me for the first year of my D.Phil. And I am grateful to All Souls College,

which has funded me for the two years since then.

My greatest intellectual debt of gratitude is to my supervisor, Tim Williamson.

The claims and principles in this thesis are far clearer and more precise as a result of his

influence. I am extremely grateful to him for instilling in me a sense of the standards of

precision and clarity that one should aim for when doing philosophy. Furthermore, the

content of almost every argument in this thesis has been influenced by his comments and

suggestions.

There are several people who have listened patiently as I struggled to articulate a

fledgling idea, and without these people many of the arguments below would not exist.

There is one person in this category whom I would like to single out for special thanks:

Hemdat Lerman. I have discussed most of the ideas in this thesis with Hemdat Lerman,

and I am very grateful to her for her comments and encouragement.

There are very many others with whom I have discussed the issues in this thesis,

and whom I would like to thank: Brian Ball, Helen Beebee, Wylie Breckenridge, Bill

4
Brewer, Myles Burnyeat, Jeremy Butterfield, Stephen Butterfill, John Campbell, David

Chalmers, Tim Crane, Adrian Cussins, Heather Demarest, Imogen Dickie, Fred Dretske,

Naomi Eilan, Ciara Fairley, Jane Friedman, Anil Gomes, John Hawthorne, John Hyman,

Susan Hurley, Stephen Kearns, William Kilborn, Thomas Kroedel, Rae Langton, Maria

Lasonen, Geoffrey Lee, Rory Madden, Ofra Magidor, Julia Markovits, Mike Martin,

Sarah Moss, Katie Myers, Anders Nes, Matthew Nudds, Christopher Peacocke, Ian

Phillips, Hanna Pickard, Daniel Rothschild, Nick Shea, Susanna Siegel, Nico Silins, Julia

Simon, Declan Smithies, Matthew Soteriou, Robert Stalnaker, Nick Zangwill, audiences

at the Ockham Society, the 2004 Warwick Graduate Conference, the 2004

Columbia/NYU Graduate Conference and the 2005 Harvard/MIT Graduate Conference.

Finally, I would like to thank my parents, David Price and Mary Price. They made

education their number one priority, and encouraged my brother, my sister and me to

think for ourselves at a young age. Dinner time at the Price family frequently involved an

extended debate over a contentious proposition, with all of us children taking different

points of view. These dinner-time debates opened my eyes to the fun that could be had

from arguing, and sowed the seeds of my subsequent interest in philosophy. I dedicate

this thesis to my mother and father.

Richard Price

3rd August, 2006

5
Contents
(Expanded Contents Overleaf)

Introduction
Historical and Contemporary Views On The Visible/Non-Visible
Distinction, and What the Visible/Non-Visible Distinction Is Not 10

1 A Sparse View About The Properties That Objects Phenomenally Look


To Have 74

2 The Tension Between Exportation And The View That Objects


Phenomenally Look To The Left And Right Of Me 144

3 Two More Arguments Against The View that Objects Phenomenally


Look To The Left and Right Of Me, And Arguments Against Some
Other Views About Phenomenal Position Properties 183

4 Primitivism About Phenomenal Position Properties 229

5 An Argument That Objects Phenomenally Look To Have z Coordinates 266

6 Coarse-Grained Vision and New Kinds of Phenomenal Character 292

6
Expanded Contents

Introduction

Historical and Contemporary Views On The Visible/Non-Visible Distinction, and


What the Visible/Non-Visible Distinction Is Not...............................................11
1 The Project.................................................................................................................11
2 History of the Debate.................................................................................................16
3 What The Issue Is Not................................................................................................20
3.1 The Observation/Theory Distinction...............................................................20
3.2 Conceptual/Nonconceptual Content...............................................................25
3.3 The Given.......................................................................................................26
3.4 Inferential vs. Immediate Belief......................................................................27
3.5 What is Seen vs. What is Believed.................................................................28
3.6 The Sense-Datum Theory of Perception.........................................................29
4 Similar Problems In Other Areas of Philosophy........................................................30
4.1 Visual Imagination..........................................................................................30
4.2 Aesthetics........................................................................................................31
4.3 Semantics/Pragmatics......................................................................................33
5 The Debate in Contemporary Philosophy..................................................................35
5.1 Epistemological motivations..........................................................................35
5.1.1 Other Minds.....................................................................................35
5.1.2 Moral Properties..............................................................................37
5.1.3 Properties of Objects-Too-Small-To-See.........................................37
5.1.4 Theological Properties.....................................................................38
5.1.5 Semantic Properties..........................................................................38
5.2 Non-epistemological motivations...................................................................39
6 Tests For The Visible/Non-Visible Distinction...........................................................42
6.1 The Ascent Routine Test.................................................................................43
6.2 Lewis’s Test.....................................................................................................48
6.3 The Same Look Test........................................................................................53
6.4 The Veridicality Test........................................................................................54
6.5 Behavioural Tests.............................................................................................58
6.6 Neuroscientific Data.......................................................................................60
6.7 The Concept Test.............................................................................................62
6.8 Epistemological Tests......................................................................................66
6.9 Phenomenological Tests..................................................................................69
7 The Significance Of The Question.............................................................................70
8 Chapter Survey...........................................................................................................71
A Sparse View About The Properties That Objects Phenomenally Look To Have. . .75
1 Phenomenal Looking.................................................................................................75
1.1 Externalist-looking..........................................................................................78
1.2 Epistemic Looking...........................................................................................82
1.3 Nonconceptual Looking...................................................................................83

7
1.4 Theory-laden Looking....................................................................................84
1.5 The Relata of Phenomenal Looking...............................................................85
2 The Phenomenal Difference Test...............................................................................93
2.1 Being a tomato................................................................................................93
2.2 Being a table, being expensive.......................................................................98
2.3 Depth properties..............................................................................................99
2.4 Change...........................................................................................................102
2.5 Causal properties...........................................................................................105
2.6 The property of having a back.......................................................................107
3 The Arguments of Searle and Siegel.........................................................................109
3.1 Aspect-switching...........................................................................................111
3.1.1 Patterns of attention.......................................................................113
3.1.2 How one takes the object to be......................................................115
3.1.3 Residual Visual Phenomenal Differences and Visual Imagination 116
3.2 House/façade.................................................................................................119
3.3 Language acquisition and recognitional capacities......................................120
3.4 Phenomenally looking a new way................................................................122
4 Determinabilism......................................................................................................125
4.1 Colour Illusions.............................................................................................125
4.2 The Vagueness Argument..............................................................................127
4.3 Phenomenal Looking To Have Determinables Alone...................................130
4.4 Kinds of Colour Phenomenal Character........................................................133
4.5 The Red Anti-Entailment Principle................................................................135
5 Other Properties.......................................................................................................136
5.1 Being the same shade as................................................................................136
5.2 Being square.................................................................................................138
5.3 Having a certain length..................................................................................139
6 Conclusion................................................................................................................143
The Tension Between Exportation And The View That Objects Phenomenally Look
To The Left And Right Of Me...........................................................................145
1 Exportation For Belief.............................................................................................147
1.1 Kripke’s Example.........................................................................................149
2 Exportation For Phenomenal Looking.....................................................................153
2.1 Propositional Contents...................................................................................155
2.2 Conceptual Representation............................................................................156
2.3 Knowledge Which........................................................................................159
2.4 Consequences of the phenomenal looking exportation principle.................161
3 Responses To The Argument...................................................................................164
3.1 Modes of Presentation..................................................................................164
3.2 Temporal Properties.......................................................................................174
3.3 Definite Descriptions....................................................................................177
3.4 Exporting Predicates......................................................................................180
Two More Arguments Against The View that Objects Phenomenally Look To The
Left and Right Of Me, And Arguments Against Some Other Views About
Phenomenal Position Properties.......................................................................185
1 The Observer-Relative View Of Phenomenal Position Properties...........................186

8
1.1 The Problem of the Second Relatum.............................................................186
1.2 The Functionality of Front and Back and Top and Bottom...........................195
1.2.1 Front and Back...............................................................................196
1.2.2 Top and Bottom..............................................................................203
1.3 Visual Angles................................................................................................211
2 Other Views About Phenomenal Position Properties..............................................218
2.1 Field-Of-View Relationalism.........................................................................218
2.2 Leibnizian Relational Properties....................................................................223
2.3 Substantivalism..............................................................................................226
Primitivism About Phenomenal Position Properties..................................................231
1 The Perspective Problem.........................................................................................232
2 The Colour Inversion Problem................................................................................236
2.1 No Problem View..........................................................................................237
2.2 The Many Colours View...............................................................................240
2.3 The No Colours View...................................................................................244
2.4 The Unique Colour View..............................................................................246
3 Necessary Colour and Position Scepticism..............................................................254
3.1 Necessarily Coextensive Properties..............................................................256
3.2 Phenomenal Positions As Particulars.............................................................258
4 Consequences of Primitivism...................................................................................260
4.1 Phenomenal Spatial Relations......................................................................260
4.2 The Shifting Hypothesis...............................................................................262
4.3 The Scaling Hypothesis.................................................................................265
5 Conclusion...............................................................................................................267
An Argument That Objects Phenomenally Look To Have z Coordinates................268
1 The z Coordinate Constraint....................................................................................268
2 The Possibility of Objects Phenomenally Looking To Have z Coordinates............272
3 Objects Phenomenally Looking To Have z Coordinates to Us...............................289
4 Conclusion...............................................................................................................293
Coarse-Grained Vision and New Kinds of Phenomenal Character..........................294
1 Berkeley on Abstract Ideas......................................................................................294
2 Some Initial Definitions...........................................................................................295
3 The Colour Phenomenal Character Problem...........................................................298
4 Pointy and Gunky Colour Phenomenal Character...................................................301
5 The Location Phenomenal Character Problem........................................................307
5.1 The Apparent Asymmetry Between The Colour and Location Phenomenal
Character Problems..............................................................................................311
6 Responses To The Colour Phenomenal Character Problem....................................317
6.1 The Being Red* Principle.............................................................................317
6.2 The Identity View.........................................................................................318
6.3 The Asymmetric Entailment View................................................................320
6.4 The Disjunctive View...................................................................................322
6.5 The Determinable View................................................................................323
7 Wang’s Paradox.......................................................................................................324
8 The Disjunctive View Discussed.............................................................................327
8.1 Phenomenal Character..................................................................................329

9
8.1.1 Multiple Kinds of Colour Phenomenal Character like Red*-
phenomenal character .............................................................................335
8.1.2 Phenomenally Looking not F..........................................................341
9 The Similarity View.................................................................................................344
9.1 The One-Many Similarity Relation..............................................................346
9.2 The One-Many Similarity Relation Involving Non-colour Properties.........349
9.3 Vertical/Horizontal Relations........................................................................351
9.4 Weaker Starting Assumptions........................................................................352
10 Conclusion..............................................................................................................362

10
Introduction

Historical and Contemporary Views On The Visible/Non-

Visible Distinction, and What the Visible/Non-Visible

Distinction Is Not

1 The Project

In this thesis I define a kind of looking, phenomenal looking, which is

individuated in terms of differences in visual phenomenal character. I discuss some issues

concerning the properties that objects phenomenally look to have.

The first issue I discuss concerns the range of properties that objects

phenomenally look to have. I argue that objects phenomenally look to have only colours

and positions.

The second issue concerns the kinds of position properties that objects

phenomenally look to have. I call the position properties that objects phenomenally look

to have phenomenal position properties. I argue that phenomenal position properties are

11
not observer-relative properties such as being to the left of me and being to the right of

me, but instead absolute, non-relational properties. I call this view concerning

phenomenal position properties primitivism.

The third issue concerns whether objects have phenomenal position properties.

There is some similarity between the arguments over whether objects have phenomenal

position properties and the arguments over whether objects have the colour properties

that they phenomenally look to have. I argue that necessarily objects do not have the

colour and position properties that they phenomenally look to have.

The fourth issue concerns whether the position properties that objects

phenomenally look to have consist of x, y and z coordinates, or only x and y coordinates.

I argue that the position properties that objects phenomenally look to have consist of x, y

and z coordinates.

The fifth issue concerns how objects would phenomenally look to beings with

more coarse-grained discriminatory capacities than us. For instance, suppose that patch1

and patch2 phenomenally look red1 and red2 respectively to one. Consider a being, say a

dog, which cannot discriminate the colour that patch1 phenomenally looks from the

colour that patch2 phenomenally looks. One hypothesis is that both patch1 and patch2

phenomenally look red1 to the dog. Another is that both patch1 and patch2 phenomenally

look red2 to the dog. The fifth issue I discuss in the thesis concerns what alternative

hypotheses there may be regarding how patch1 and patch2 phenomenally look to the dog.

12
Broadly speaking, this thesis is an investigation into the kinds of properties that

objects phenomenally look to have.

Many philosophers have been interested in questions such as which properties are

visible, or which properties visual experience represents. Consider the quotation below

from Mackie:

‘If I see someone else peel a potato, I see the relative movement of
the potato and the knife, I see the gradually lengthening strip of peel come
up from the surface of the potato, each new portion coming up just as the
leading edge of the knife reaches it; but, as Hume would rightly protest, I
don’t see the knife making the peel come up. And what I most obviously
fail to see, though I do judge, is that each bit of the peel would not have
come up if the knife had not moved in there.’ (Mackie, 1999, p133).

And consider this quotation from Siegel:

‘Consider the content:

(#) that Ms. Elfenbein is out of town.

‘This is the sort of content that it would be normal to believe on the


basis of perception…But (#) does not seem to be the sort of content that is
ever properly attributable to perception itself. Even if one perceived Ms.
Elfenbein, despite her absence—perhaps by talking to her on the telephone
—one still wouldn’t perceive her being out of town. If one cannot sense
that someone is out of town, then (#) is not a content properly attributable
to perception.’ (Siegel, 2006, p481).

And consider this quotation from Dummett:

13
‘The notion of an observation report is a very loose one… The
following conditions appear to be required. First, the making of the
observation report must not rest on any extraneous inference (must not
represent ‘a conclusion of the witness’), as e.g., in ‘I see that the Smiths
forgot to cancel their newspapers’.’ Dummett, 1976, p95).

Mackie, Siegel and Dummett are all appealing to an intuitive distinction between

visible and non-visible properties. Mackie’s claim that one cannot see the knife making

the peel come up, Siegel’s claim that one cannot perceive Mrs Elfenbein being out of

town, and Dummett’s claim that one cannot observe that the Smiths forgot to cancel their

newspapers all appeal to the intuitive distinction between visible and non-visible

properties.

One of the aims of this thesis is to argue for a constraint on the visible/non-visible

distinction. A property is visible, I argue, iff it is one that an object can phenomenally

look to have. Phenomenal looking is defined as the kind of looking that satisfies the

following principle:

The Phenomenal Character Principle: Necessarily, for all objects, x, y and z, all

properties F and G, all times t1 and t2, and all

worlds w1 and w2, if x looks F to z at t1 in

w1, y does not look F to z at t2 in w2, and y

looks G to z at t2 in w2, then there is a visual

phenomenal difference between the way that

x looks to z at t1 in w1 and the way that y

looks to z at t2 in w2.

14
In chapter 1 I will explain the phenomenal character principle in more detail and

distinguish phenomenal looking from other kinds of looking. I assume that visual

experience can represent a property iff that property is visible, and therefore that visual

experience can represent a property iff that property is one that objects can phenomenally

look to have.

Two points of clarification are in order. The first is that the notion of a visible

property suggests that some properties can be seen, in the same way that objects can be

seen. However, as I use the expression ‘visible’, a property is visible iff an object can

phenomenally look to have it. I do not assume that properties can be seen.

The second point of clarification concerns the modality involved in the

expressions ‘visible’ and ‘can phenomenally look to have’. I am interested in what

properties objects phenomenally look to have in the actual and nearby worlds. I do not

consider what properties it is metaphysically possible for objects phenomenally to look to

have.

One of the issues discussed in this thesis concerns how rich the range of visible

properties is. According to views on which the range of visible properties is very rich, the

range of visible properties includes moral properties, the property of being in a particular

mental state, theological properties, semantic properties, and the property of belonging to

a particular natural kind.

15
According to sparse views, the only visible properties are those such as colour,

shape, size and position (the ‘rich/sparse’ terminology is from Lee, 2004). I argue for a

very sparse view, according to which the only visible properties are colours and positions.

Defenders of relatively rich views of the range of visible properties include Susanna

Siegel and Christopher Peacocke, and defenders of relatively sparse views of the range of

visible properties include Colin McGinn, Alan Millar and Fred Dretske.

In this introduction I discuss the views of some historical figures concerning the

visible/non-visible distinction. Then I distinguish the debate over the visible/non-visible

distinction from certain other debates, such as the debate over whether observation is

theory-laden or not. Lastly, I discuss some tests that contemporary philosophers have

proposed that are designed to distinguish visible from non-visible properties.

2 History of the Debate

In the 2nd Meditation, Descartes writes as follows:

‘But were I then perchance to look out my window and observe


men crossing the square, I would ordinarily say I see the men themselves
just as I say I see the wax. But what do I see aside from hats and clothes,
which could conceal automata? Yet I judge them to be men. Thus what I
thought I had seen with my eyes, I actually grasped with the faculty of
judgement, which is in my mind.’ (Descartes, 1993, p22).

16
On the most natural interpretation of this passage, Descartes is investigating

which of the properties of the men in the square before him are available to vision, and

which are available only to judgement. The conclusion is that whilst he initially thought

that the property of being a man was one that ‘I thought I had seen with my eyes’, in fact

it is one that ‘I actually grasped with the faculty of judgement.’

According to this interpretation, the implication of the question ‘But what do I see

aside from hats and clothes, which could conceal automata?’ is that it is consistent with

what Descartes sees to be the case that the objects in the square are automata, and

therefore that the property of being a man is not one that, on this occasion, Descartes sees

to be instantiated.

There was considerable interest in the 18th and 19th centuries concerning whether

depth properties are visible; that is, whether or not an object’s distance away from one is

a visible property of it. In A New Theory of Vision, Berkeley argued that depth is not

visible.

‘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 remains
invariably the same, whether the distance be longer or shorter.’ (Berkeley,
1975, p9).

In Of the Principles of Human Knowledge, Berkeley describes his conclusions

from A New Theory of Vision as follows:

17
‘… in strict truth the ideas of sight…, do not suggest or mark out
to us things actually existing at a distance, but only admonish us that ideas
of touch will be imprinted in our minds at such and such distances of time,
and in consequence of such and such actions.’ (Berkeley, 1975, p89).

John Stuart Mill endorsed Berkeley’s argument that depth is not visible, holding

that it ‘proves conclusively that distance from the eye is not seen, but inferred’ (Mill,

1875, p95, cited in Smith, 2000, p488).

Hume also agreed with Berkeley about depth being non-visible:

‘Tis commonly allowed by philosophers that all bodies which


discover themselves to the eye, appear as if painted on a plain surface, and
that their different degrees of remoteness from ourselves are discover’d
more by reason than by the senses.’ (Hume, 1978 p56, cited in Smith,
2000, p481).

Hume also denied that we can observe necessary connections between objects.

‘…we are led astray by a false philosophy… when we transfer the


determination of the thought to external objects, and suppose any real
intelligible connection betwixt them… [I] have observed, that objects bear
to each other the relations of contiguity and succession… But if we go any
further, and ascribe a power or necessary connexion to these objects, this
is what we can never observe in them.’ (Hume, 1978, p168-9).

Reid also held that depth was not visible, and used this premise to argue that the

geometry of the shape properties that objects look to have is non-Euclidean. He gives his

argument that depth is not visible in the following paragraph:

18
‘Supposing the eye placed in the centre of a sphere, every great
circle of the sphere will have the same appearance to the eye as if it was a
straight line. For the curvature of the circle being turned directly toward
the eye, is not perceived by it. And for the same reason, any line which is
drawn in the plane of a great circle of the sphere, whether it be in reality
straight or curve, will appear straight to the eye.’ (Reid, 2000, p103).

Reid shows that it follows from this that any plane triangle, that is, a triangle

drawn on a plane, will look the same as some spherical triangle, that is, a triangle drawn

on the surface of a sphere. He appears to argue that it follows from this point that a plane

triangle looks spherically triangular.

‘Hence it is evident, that every visible right-lined triangle, will


coincide in all its parts with some spherical triangle. The sides of the one
will appear equal to the sides of the other, and the angles of the one to the
angles of the other, each to each; and therefore the whole of the one
triangle will appear equal to the whole of the other. In a word, to the eye
they will be one and the same, and have the same mathematical properties.
The properties therefore of visible right-lined triangles, are not the same
with the properties of plain triangles, but are the same with those of
spherical triangles.’ (Reid, 2000, p104).

On a natural interpretation of this paragraph, the argument seems to be that since

plane and spherical triangles look the same, they both look spherically triangular. If this is

the correct interpretation of the argument, then the argument is invalid, since plane and

spherical triangles looking the same is consistent with them both looking to have the

property of being a plane triangle. Van Cleve has argued that in fact Reid’s conclusion

that both triangles look spherically triangular rests on more than the claim that they look

the same (Van Cleve 2002).

19
In the Second Analogy Kant aims to establish the claim that every event has a

cause. Kant’s argument has been interpreted in different ways, but at least one

interpretation links Kant’s discussion with the visible/non-visible distinction. Guyer

writes as follows:

‘In the Second Analogy, Kant argues for a further condition for
making judgments about change in objects: because even when we
undergo a sequence of perceptions, there is nothing in their immediate
sensory content to tell us that there is an objective change, let alone what
particular sequence of change there is, we can only distinguish a
‘subjective sequence of apprehension from the objective sequence of
appearances’ (A 193/B 238) by judging that a particular sequence of
objective states of affairs, a fortiori the sequence of our perceptions of
those states, has been determined in accordance with a rule that states of
the second type can only follow states of the first type - precisely what we
mean by a causal law.’ (Guyer, 2004, §7).

If Guyer is right, then the Second Analogy raises the question whether there are

perceptions of change, or merely changes in one’s perceptions. This is equivalent to the

question whether the property of having changed is a visible property. According to

Guyer’s interpretation of the Second Analogy, it is a premise of the Second Analogy that

there are no perceptions of change, but instead only changes in one’s perceptions.

3 What The Issue Is Not

3.1 The Observation/Theory Distinction

20
Putnam introduces the observation/theory distinction as follows:

‘The basis for the division appears to be as follows: the


observation terms apply to what may be called publicly observable things
and signify observable qualities of these things, while the theoretical terms
correspond to the remaining unobservable qualities and things.’ (Putnam,
1962, p240).

Thus, for Putnam, the observation/theory distinction is at least the distinction

between observable and unobservable properties. Putnam also claims that it is a

distinction between observable and unobservable things, but I will not focus on this latter

distinction. At one point in ‘Observation Reconsidered’, Fodor seems to agree that the

distinction between observable and non-observable properties is the central distinction in

the observation/theory debate:

‘The obvious suggestion would be, on the one hand, that what
makes a term observational is that it denotes, by independent criteria, an
observable property.’ (Fodor, 1983, p30).

However, at the beginning of ‘Observation Reconsidered’, Fodor characterizes his

project as follows:

‘Several arguments are considered which purport to demonstrate


the impossibility of theory-neutral observation. The most important of
these infers the continuity of observation with theory from the presumed
continuity of perception with cognition, a doctrine widely espoused in
recent cognitive psychology.’ (Fodor, 1984, p23).

Theory-neutral observation is defined by Fodor as follows:

21
‘As a first shot at what the theory-neutrality of observation comes
to: given the same stimulations, two organisms with the same
sensory/perceptual psychology will quite generally observe the same
things, and hence arrive at the same observational beliefs, however much
their theoretical commitments may differ.’ (Fodor, 1984, p25).

It seems that there are two issues in the debate over the observation/theory

distinction. One is whether a distinction between observable properties and non-

observable properties can be drawn, and the second is whether there is theory-neutral

observation. In Fodor’s paper these two issues are run together; the argument in the paper

suggests that Fodor thinks that there is a distinction between observable and non-

observable properties only if there is theory-neutral observation. I will argue that this is

incorrect, and that there may still be a distinction between observable and non-observable

properties even if all observation is theory-laden.

Let us suppose that, if a particular observation can be laden with a theory about

Fs, then being F is an observable property. This assumption fits with how philosophers

have used the notion of theory-laden observation: if one’s observation is laden with a

theory one has about gamma-ray tubes, then the property of being a gamma-ray tube can

be observed. Suppose that observation can be laden by any theory. This supposition

would entail that all properties are observable. It would not undermine the notion of an

observable property, nor would it undermine the notion that there is a distinction between

observable properties and unobservable properties. It would show simply that there are

no properties on one side of this distinction.

22
Consider an analogy with desire. Let us suppose that, if one’s desires can be laden

with a theory about Fs, then the property of being F can feature in the content of one’s

desires. Suppose that one’s desires can be laden with any theory. This supposition would

entail that any property could feature in the contents of one’s desires. It would not

undermine the notion of a property that can feature in the content of one’s desires, nor

would it undermine the distinction between properties that can feature in the contents of

desires and those that cannot.

The following is one way in which one might try to defend the claim that if

observation is theory-laden, then there is no distinction between observable and

unobservable properties. The claim that observation is theory-laden is often accompanied

by the claim that vision is continuous with judgement. Suppose that there are simply

judgements, some of which are more observational than others. One might think that if

this is the case, then there is no difference between observable and unobservable

properties.

However, this would be a mistake. The fact that some judgements are more

observational than others does not mean that no judgements are observational,

simpliciter. Analogously, the fact that some cars are faster than other cars does not mean

that no cars are fast, simpliciter. It is incorrect to suppose that whenever there is a

continuum of Fness, there is no distinction between being F and not being F. On this

view, the observable properties would be the ones that can feature in the observational

judgements.

23
If there are simply judgements, some of which are more observational than others,

then it may be conceded that, in some sense, there is no qualitative difference between

observable properties and unobservable properties, just as, in some sense, there is no

qualitative distinction between fast and slow cars. One might think that if there is no

qualitative distinction between observable and unobservable properties, and therefore that

observability is a matter of degree, then there would be less interest in the notion.

However, there does not seem to be any reason why this would be the case. The fact that

justification and goodness come in degrees does not make them any philosophically less

interesting notions.

There is a question about whether philosophers writing about the

observation/theory distinction were really interested in whether observation was theory-

laden, or whether they were really interested in whether a distinction can be drawn

between observable properties and unobservable properties. At least part of the interest in

the observation/theory distinction was epistemological. Fodor writes:

‘Second, the observational fixation of belief plays a special role in


the adjudication and resolution of clashes of opinion. When observation is
not appealed to, attempts to settle disputes often take the form of a search
for premises that the disputants share… None of this applies, however,
when the beliefs at issue are observational. Since observation is not a
process in which new beliefs are inferred from old ones, the use of
observation to resolve disputes does not depend on a prior consensus as to
what premises may be assumed.’ (Fodor, 1984, p24).

24
Fodor’s argument is that if observation is theory-neutral, then it can provide

evidence for or against a theory. By contrast, if observation is theory-laden, it is less clear

that it can be provide evidence for or against a theory. This suggests that the real interest

in the debate over the observation/theory distinction is in whether observation is theory-

laden, rather than in whether there is a distinction between observable and unobservable

properties.

In this section I have argued that the issue of whether there is an

observable/unobservable distinction is independent of the issue of whether observation is

theory-laden.

However, there may well be a connection between the question whether

observation is theory-laden and the question what range of properties is observable. For

instance, many philosophers who argue for a rich view about the range of visible

properties hold that observation is theory-laden, and that observation can be laden with a

rich range of theories. I will discuss these views in more detail below.

3.2 Conceptual/Nonconceptual Content

A state is conceptual iff a subject’s being in that state entails that that subject

possesses concepts that characterize the content of that state. A state is nonconceptual iff

25
it is not conceptual. For this definition of conceptual content, see Cussins 1992, Brewer

1999, Evans 1982.

The question whether there is a distinction between visible and non-visible

properties is independent of the question whether visual experience is a conceptual state.

One may hold that visual experience is a conceptual state and also think that there is a

distinction between properties that are represented by visual experience and properties

that are not.

3.3 The Given

In ‘Empiricism and the Philosophy of Mind’, Wilfrid Sellars launched an attack

on the ‘myth of the given’ (Sellars 1997). Some commentators have argued that Sellars is

not entirely clear about what the myth of the given is. Daniel Bonevac writes as follows:

‘Sellars never specifies precisely what the Myth of the Given is.
Tracing Sellars’s dialectic, the reader gets the sense that the target
repeatedly shifts. The thesis that there is something independent of
acquired conceptual capacities given in experience remains entangled with
the role the thesis plays in foundationalist theories of knowledge.’
(Bonevac, 2002, p1).

McDowell takes Sellars to be interested in the question whether perceptual

experience has nonconceptual content:

26
‘My thinking starts from a central element in Wilfrid Sellars’s
attack on the Myth of the Given. Sellars means to undermine more than
traditional empiricism. But in his classic ‘Empiricism and the Philosophy
of Mind’ he concentrates on the idea of something given in experience
independently of acquired conceptual capacities.’ (McDowell, 1998a,
p365).

If McDowell’s interpretation is correct, then whether there is a given or not is a

separate question from whether there is a distinction between visible and non-visible

properties.

3.4 Inferential vs. Immediate Belief

It is tempting to think of the distinction between visible properties and non-visible

properties in terms of the distinction between properties that one immediately forms

beliefs about, and properties which one forms beliefs about only inferentially. Consider

the following dialogue:

Jody: I work on the distinction between visible and non-visible properties.

Derek: Oh, what is that all about?

Jody: Well, suppose that you are looking at wet pavements, and you judge that it has

just rained. Is it literally part of what you see that it has just rained?

27
Derek: No, clearly I infer that it has just rained on the basis of my visual experience and

some background beliefs about the most likely cause of the pavements being wet.

On the other hand, I directly, non-inferentially believe that the pavements are wet,

so presumably that the pavements are wet is part of what I see.

Derek’s last reply to Jody is a very natural one to make. However, the visible/non-

visible distinction is not the same as the distinction between properties that one can form

beliefs about non-inferentially and properties that one can form beliefs about only

inferentially.

Firstly, it may be that when one forms a belief, about some object, that it is red,

this belief is inferred from the belief that the object looks red, and that things are the way

they look. In addition, one’s belief that the object looks red may be an inference from the

belief that it does not look green, the belief that it does not look blue, and so on. Whether

these inferences occur or not is independent of whether being red is a visible property.

Secondly, a subject could be wired to believe, whenever they see wet pavements,

that it has just rained. When such a subject sees wet pavements, the belief about its

having just rained that would be formed would be non-inferential. However, this would

not show that the property of its having just rained would be visible by this subject.

3.5 What is Seen vs. What is Believed

28
One might think that the distinction between visible and non-visible properties is

the same as the distinction between properties that one can see and properties that one

can only have mental states such as beliefs and desires about. Thus, for instance, colour

properties can be seen, but perhaps the property of having forgotten to cancel the papers

is one that one can have only mental states such as beliefs and desires about.

However, for all that has been said, visual experience may be a kind of belief. If

visual experience was a certain kind of belief, there would still be a distinction between

visible and non-visible properties, namely a distinction between properties that can

feature in the contents of this kind of belief and properties that cannot. The existence of a

distinction between visible and non-visible properties is independent of the question

whether visual experience is a kind of belief state.

3.6 The Sense-Datum Theory of Perception

Occasionally, in conversation, philosophers have wondered whether defenders of

the sparse view, who hold that the range of visible properties includes only properties

such as colour and shape, are defending a version of the sense-datum view of perception.

The sense-datum view is a certain view about the metaphysics of the perceptual relation.

It holds that visual perception consists in the subject’s being aware of internal mental

images that are caused in the right way by external objects.

29
A defender of the sparse view of visible properties is not committed to any

particular view of the metaphysics of the perceptual relation. One can think that objects

phenomenally look to have only a restricted range of properties and be neutral on which

account of the metaphysics of the perceptual relation is correct.

4 Similar Problems In Other Areas of Philosophy

4.1 Visual Imagination

There is a question about what properties one’s visual imagination can represent.

Call the imaginability thesis the claim that, for any proposition p, being able visually to

imagine that p entails that it is possible that p. Suppose that one can visually imagine that

water is not H2O. If this was possible, then the imaginability thesis would be false, since

necessarily, water is H2O.

However, one might argue that one cannot imagine that water is not H2O, despite

it seeming as though one can. According to this argument, one really imagines that a

transparent liquid that falls from the sky and fills the lakes is not called ‘H2O’ by

scientists, and one misdescribes the content of one’s visual imagination as being that

water is not H2O. If the content of one’s visual imagination is that a transparent liquid that

falls from the sky and fills the lakes is not called ‘H2O’ by scientists, then the

30
imaginability thesis remains intact, since it is possible that a transparent liquid that falls

from the sky and fills the lakes is not called ‘H2O’ by scientists.

Thus, what range of properties one can visually imagine makes a difference to

arguments about whether the imaginability thesis is true.

Wittgenstein was also interested in the question as to what properties visual

imagination represents. Here he asks whether visual imagination can represent the

property of being King’s College, or whether it can represent only properties such as

being a building.

‘Someone says, he imagines King’s College on fire. We ask him:


“How do you know that it’s King’s College you imagine on fire? Couldn’t
it be a different building, very much like it? In fact, is your imagination so
absolutely exact that there might not be a dozen buildings whose
representation your image could be?”—And still you say: “There’s no
doubt I imagine King’s College and no other building.”’ (Wittgenstein,
1958, p39).

4.2 Aesthetics

Kendall Walton argues that what distinguishes the way a picture represents a mill,

say, from the way a novel represents a mill is that, when looking at a picture, one

imagines that the object of one’s visual perception is the mill, whereas when looking at a

verbal description of a mill, one does not imagine that the object of one’s visual

perception is the mill. (Walton, 1990).

31
Walton says that a depiction is a prop used in the activity of imagining that the

object of one’s perception is something other than the actual object of one’s perception.

When one looks at a depiction of a mill, Walton says that one fictionally sees the mill.

Walton discusses what restrictions there may be on the properties that can be

depicted:

‘Pictures—representations that are depictive in some respects—can


portray (generate fictional truths about) non-visual phenomena, of course.
The question is whether this portrayal is depiction. The indirect
representation of the sound of a shot by means of the sudden rising of a
flock of birds in The Docks of New York is clearly not depiction. Seeing
the images of the rising birds is not, fictionally, perceiving the shot. What
about the portrayal in cartoons of sounds and patterns of thermal radiation
or smells by means of concentric arcs or wavy lines emanating from the
source—a gong or a campfire or a garbage heap? It is probably not
fictional that we see sounds or smells or heat when we see the picture, nor
is our looking at the picture fictionally a looking at such non-visual
phenomena.’ (Walton, 1990, p331-2).

Given Walton’s definition of a depiction, the range of possible contents of a

depiction is constrained by what one can imagine to be the object of one one’s visual

perception. If it is not possible to imagine that one is visually perceiving a sound, then no

representation can depict a sound.

Even if one disagrees with Walton’s theory of depiction, one may still ask whether

the way in which a picture can represent a sound or a smell is the same as the way in

which a picture can represent a colour or a shape.

32
4.3 Semantics/Pragmatics

Suppose that Tom says ‘I have not had a bath’. Certainly, in making this utterance,

Tom implies that he has not had a bath today. However, there is a question about whether

Tom’s utterance expresses the proposition that he has not had a bath today, or whether it

expresses the proposition that has not had a bath, which entails that Tom has never had a

bath. The question concerns where to draw the line between the semantic content of an

utterance and what is pragmatically implied by the utterance.

Suppose that a sales assistant in the supermarket says ‘all the chocolate has sold

out’. Certainly, in making this utterance, the sales assistant implies that all the chocolate

in the supermarket has sold out. However, there is a question about whether the sales

assistant’s utterance expresses the proposition that all the chocolate in the supermarket

has sold out, or whether it expresses the proposition that all the chocolate in the universe

has sold out.

An utterance expresses a certain semantic content. Someone who hears that

utterance will acquire a bundle of information on the basis of hearing it. There is a

question about how much of the bundle of information is part of the semantic content

expressed by the sentence and how much is due to other factors. The analogy with the

visible/non-visible distinction is as follows. A visual experience has a certain content.

33
Someone who has that visual experience will acquire a bundle of information on the basis

of having it. There is a question about how much of the bundle of information is part of

the content of the visual experience, and how much is due to other factors.

One test for distinguishing semantic content from pragmatic content is as follows.

If there are two utterances, u1 and u2, which express the same proposition, and there is

some proposition q that is conveyed by u1 that is not conveyed by u2, then q is not part of

the semantic content of u1 or u2.

The analogue of this test for the visible/non-visible distinction is as follows. If

there are two objects, O1 and O2, which phenomenally look the same, and for some

property F, O2 does not phenomenally look F, then O1 does not phenomenally look F

either.

Suppose that an object x phenomenally looks red and round, and there is

disagreement about whether x phenomenally looks to be a tomato. If there is an object y

that phenomenally looks the same as x, but does not phenomenally look to be a tomato,

then x does not phenomenally look to be a tomato.

I will be using a different test in chapter 1, but this test and my test raise some of

the same issues.

34
5 The Debate in Contemporary Philosophy

5.1 Epistemological motivations

The following is the structure of one traditional kind of sceptical argument:

1.) Experience does not represent properties of kind F.

2.) If experience does not represent properties of kind F, then experience does not

justify our beliefs about properties of kind F.

3.) If experience does not justify our beliefs about properties of kind F, then

nothing justifies our beliefs about properties of kind F.

4.) Therefore, our beliefs about properties of kind F are not justified.

This argument has been applied to a number of different kinds of property,

including the mental properties of others, properties of objects that are too small to see,

moral properties, semantic properties and theological properties. A common response to

this kind of argument has been to deny 1.), and argue that experience does represent

properties of the kind in question.

5.1.1 Other Minds

McDowell’s response to the problem of other minds is as follows.

35
‘In the view of this different realist, we should not jib at, or
interpret away, the common-sense thought that, on those occasions that are
paradigmatically suitable for training in the assertoric use of the relevant
part of a language, one can literally perceive, in another person’s facial
expression or his behaviour, that he is in pain, and not just infer that he is
in pain from what one perceives.’ (McDowell, 1998b, p305).

It seems that in this passage, McDowell is arguing that one’s perceptual

experiences can represent that someone is in pain.

In an earlier work, McDowell and Evans write:

‘Seeing cheerfulness in a face is not inferring that its owner is


cheerful from the way his face looks.’ (Evans and McDowell, 1976,
pxxii).

In her paper ‘Perceiving Intentions’, Joelle Proust argues that one’s experiences

can represent the mental properties of others.

‘It is a fact of experience that when a perceiver observes someone


else’s bodily movements, she directly perceives these movements as goal-
directed and intentional.’ (Proust, 2003, p6).

Peacocke also argues that the property of being sad is visible:

‘To describe, when seeing the face of a person, the experience in


which they look sad in non-emotional terms is not to capture its distinctive
representational content. There is no kind, described without reference to
the emotions, of which one can say that the facial expression appears to be
of that kind, and it is merely an additional judgment on the part of the
person that people looking that way are sad.’ (Peacocke 2003, p66).

36
5.1.2 Moral Properties

In Moral Vision, David McNaughton argues that moral beliefs are justified by our

experiences representing moral properties. He writes:

‘I do not see an expanse of coloured cloud, which is not itself seen


as beautiful, and then experience a thrill of pleasure to which I give the
name of beauty. The beauty of the sunset is woven into the fabric of my
experience of it. I see the sunset as beautiful…

‘Reflection on the aesthetic case can serve to make us less certain


of Hume’s claim in the moral case. If I see several children throwing
stones at an injured animal I may claim that I can just see that what they
are doing is cruel. Similarly, the insolence of a drunken guest’s behaviour
seems no less observable than the cut of his suit.’ (McNaughton, 1988,
p56).

5.1.3 Properties of Objects-Too-Small-To-See

A sceptical argument of the above kind has been applied to properties such as

electrons. How can our beliefs about electrons be justified if our experience does not

represent anything as an electron?

Consider the following passage from Robert Brandom.

‘It is important to understand that under the appropriate


circumstances, which include the presence of a bubble chamber or similar
device, and for the right community of observers, mu-mesons are literally
observable—non-inferentially reportable in the same sense in which red
things are for the rest of us. It is a mistake to think that what is really non-

37
inferentially observed is only the vapour trail and that the presence of mu-
mesons is only inferred.’ (Brandom, 1994, 223).

Brandom seems to be arguing in this passage that the property of being a mu-

meson is observable because mu-mesons are non-inferentially reportable.

5.1.4 Theological Properties

A sceptical question arises about the epistemological status of religious belief.

William Alston has argued that religious beliefs can be justified by perceptual

experiences which represent theological properties. He writes:

‘The central thesis of this book is that experiential awareness of


God, or as I shall be saying, the perception of God, makes an important
contribution to the grounds of religious belief. More specifically, a person
can become justified in holding certain kinds of beliefs about God by
virtue of perceiving God as being or doing so-and-so. The kinds of beliefs
that can be so justified I shall call ‘M-beliefs’ (‘M’ for ‘manifestation’).
M-beliefs are beliefs to the effect that God is doing something currently
vis-à-vis the subject—comforting, strengthening, guiding, communicating
a message, sustaining the subject in being—or to the effect that God has
some (allegedly) perceivable property—goodness, power, lovingness.’
(Alston, 1991, p1).

In ‘The Myth of Religious Experience’, Nick Zangwill argues, against Alston, that

it is not possible for perceptual experiences to represent theological properties (Zangwill,

2004).

5.1.5 Semantic Properties

38
Gareth Evans and John McDowell raise the question how one comes to know

what someone means by an utterance. In a passage partly quoted above, they claim that

one can perceive the meaning in someone’s utterance:

‘Seeing cheerfulness in a face is not inferring that its owner is


cheerful from the way his face looks…. Just so, we suspect, with
language: that is, it is essential to language as we know it that our
understanding of meanings should normally be perception of meaning,
and hence precisely not a matter of inference. (Evans and McDowell,
1976, pxxii).

5.2 Non-epistemological motivations

Not all philosophers are interested in the visible/non-visible distinction for

epistemological reasons.

As we saw at the beginning of this introduction, Mackie claims that causation is

not visible:

‘If I see someone else peel a potato, I see the relative movement of
the potato and the knife… but… I don’t see the knife making the peel
come up.’ (Mackie, 1999, p133).

Strawson argues:

‘[T]here are countless true specifications or descriptions of the


things we, for example, see… which do not, and perhaps could not, figure
in any truthful account of, for example, how those things look to us.

39
Examples would be the descriptions ‘married five times’, as applied to a
man, or ‘once shaken by the Queen’, as applied to a hand.’ (Strawson,
1974b, p67)

Burge claims that the visual system can detect a sparse range of properties:

‘There is considerable empirical evidence that perceptual systems


per se make reference to properties within a relatively confined range. The
human visual system makes reference to particular colours, shapes,
positions, motions, textures—and to certain abstractions from them—as
such, but not to tomatoes, mercury, cancer, cars, presidents, sonnets, as
such.’ (Burge, 2003, pp545-546).

Pollock claims that an object does not visually look to have a back of one kind

rather than another:

[C]onsider the easiest case, which is the case in which it turns out
that the wall is all that is left of an old church— there is nothing behind it.
When we take ourselves to be seeing a church we are making a mistake,
but it is not a perceptual error…. The information encoded by perception
is not wrong — it is the judgment we make on the basis of it that is
wrong.’ (Pollock, forthcoming, p18).

Colin McGinn argues for a sparse view in the following paragraph:

‘[Since] it is possible for things to have the same appearance as


tigers, yet not fall under the concept of tiger, we should therefore restrict
the concepts invoked to characterize content to those that relate to the
appearances of things: concepts of colour, superficial texture, shape etc. In
this way we limit our ascription of content to how things seem to the
perceiver.’ (McGinn, 1982, p40).

40
In the following passage David Smith picks up on the distinction between

perception of change and change of perception that is salient in Kant’s argument in the 2nd

Analogy:

‘The crucial point to recognize in all this is that a succession of


appearances does not of itself amount to the appearance of succession…
just as a changing appearance does not entail an appearance of change, a
persisting appearance does not entail an appearance of persistence.’
(Smith, 2002, p124).

The question whether depth is visible is still discussed. Price claimed that depth is

visible:

‘It is obvious that all visual sense-data have the characteristic of


depth, or ‘outness’. This characteristic is just as much given as colour or
shape, whether we can explain it or not.’ (Price, 1950, p218).

O’Shaughnessy has endorsed Berkeley’s argument that depth is not visible

(O’Shaughnessy 1980 vol.1 p172, cited in Smith, 2000). Smith considers a number of

arguments for the view that depth is not visible and finds them wanting (Smith 2000).

In The Bounds of Sense, Strawson discusses the geometry of the shapes that

objects look to have. He defines phenomenal geometry as the geometry of the figures that

we represent in visual imagination, and states that ‘if there is such a thing as phenomenal

geometry, then we could reasonably say that it would primarily be the geometry of the

spatial appearances of physical things and only secondarily, if at all, the geometry of

physical things themselves.’ (Strawson, 1966, p282).

41
Strawson argues that phenomenal geometry is Euclidean:

‘Consider the proposition that not more than one straight line can
be drawn between any two points. The natural way to satisfy ourselves of
the truth of this axiom of phenomenal geometry is to consider an actual or
imagined figure. When we do this, it becomes evident that we cannot,
either in imagination or on paper, give ourselves a picture such that we are
prepared to say of it both that it shows two distinct straight lines and that it
shows both these lines as drawn through the same two points.’ (Strawson,
1966, p283).

Hopkins has argued in response to Strawson that for small objects of the kind we

see, a Euclidean triangle would look the same as a non-Euclidean triangle, and that

phenomenal geometry is neutral or indeterminate between being Euclidean and being

non-Euclidean (Hopkins, 1973, p22-23).

Many philosophers have argued that the geometry of the shape properties that

objects look to have is non-Euclidean. Van Cleve (2002), Yaffe (2002) and Belot (2003)

have defended updated versions of Reid’s argument for this view, and Angell (1974) and

Lucas (1969) have offered their own arguments for this view.

6 Tests For The Visible/Non-Visible Distinction

We need a test in order to decide which properties are visible and which are non-

visible. We cannot take true sentences containing perceptual verbs at face-value, since we

42
often use perceptual verbs to refer to non-perceptual states. For instance, consider the

following dialogue which, we may suppose, takes place in the dark:

Martha: It looks to me as if you are in agreement with Russell.

Susan: I see what you mean.

Both sentences uttered in this dialogue contain perceptual verbs, and yet Martha

and Susan are not referring to perceptual states. One might argue that Martha and Susan

are using the perceptual verbs metaphorically, but then a test would be needed to

distinguish metaphorical from non-metaphorical uses of perceptual verbs.

I shall now consider some tests to distinguish visible from non-visible properties.

6.1 The Ascent Routine Test

When discussing how a subject can gain knowledge of the contents of his

perceptual experiences, Evans writes as follows:

‘[A] subject can gain knowledge of his internal informational


states in a very simple way: by re-using precisely those skills of
conceptualization that he uses to make judgements about the world. Here
is how he can do it. He goes through exactly the same procedure as he
would go through if he were trying to make a judgement about how it is at
this place now, but excluding any knowledge he has of an extraneous kind.
(That is, he seeks to determine what he would judge if he did not have
such extraneous information.). The result will necessarily be closely

43
correlated with the content of the informational state which he is in at that
time. Now he may prefix this result with the operator ‘It seems to me as
though…’ (Evans, 1982, p227-8).

In a footnote, Evans writes: ‘For ‘extraneous’, see Dummett, ‘What is a Theory of

Meaning (II)?’, p95. The relevant passage in Dummett has been quoted above, and is

reproduced below:

‘The notion of an observation report is a very loose one. Without


attempting to go into all the problems which arise if one wishes to make it
sharper… the following conditions appear to be required. First, the making
of the observation report must not rest on any extraneous inference (must
not represent ‘a conclusion of the witness’), as e.g., in ‘I see that the
Smiths forgot to cancel their newspapers’.’ Dummett, 1976, p95).

The above remarks suggest the following test for determining what properties our

visual experiences represent:

Ascent Routine Test: Necessarily, for all propositions p and subjects s, if s would

judge, ignoring any information extraneous to his visual

experience, that p, then s’s visual experience represents that

p.

Robert Gordon endorses an analogue of the ascent routine test for belief. Later he

argues that this test can also allow a subject to gain knowledge of whether they are in

pain, so it is possible that he would endorse the ascent routine test itself. Gordon writes:

44
‘I argue that such self-ascription relies instead on what I call
ascent routines. For example, the way in which adults ordinarily
determine whether or not they believe that p is simply to ask themselves
the question whether or not p. Thus, if someone were to ask me, (Q1) ‘do
you believe Mickey Mouse has a tail?’ I would ask myself, (Q2) ‘Does
Mickey Mouse have a tail?’ (with certain constraints on how I obtain the
answer to Q2). If the answer to Q2 is Yes, then the presumptive answer to
Q1 (the best I can do without taking into consideration possible conflict
between verbal and non-verbal behaviour) is Yes (or, ‘Yes, I do believe
that Mickey Mouse has a tail’). The answer to Q1 is No if either the
answer to Q2 is no or no answer is available within the constraints.’
(Gordon, 1996, p15).

In the following passage Peacocke employs a stronger version of the ascent

routine test which includes both a necessary and a sufficient condition for an experience

to have a particular content:

‘We see things as tomatoes, and not as anything weaker. If the


representational content of experience is given by what someone would
judge, taking that experience at face value, then our ordinary experiences
have a content concerning tomatoes, and not tomato-like objects.’
(Peacocke, 1983, p93).

Peacocke’s test as follows:

Peacocke’s Test: Necessarily, for all propositions p and subjects s, s’s experience

represents that p iff s would judge that p, taking his experience at

face value.

I will assume that the expressions ‘ignoring any information extraneous to his

visual experience’ and ‘taking his experience at face value’ specify the same condition.

Peacocke’s test differs from the ascent routine test only in that Peacocke holds that it is a

45
necessary condition on one’s experience representing that p that one would judge, taking

one’s experience at face value, that p.

If perceptual experience has a non-conceptual content, then this necessary

condition will have to be rejected. For arguments that experience does have a

nonconceptual content, see Evans 1982, Bermudez 1998, Cussins 1990, Heck 2000 and

Peacocke 2001.

I will focus on the weaker ascent routine test, rather than Peacocke’s test. Evans

himself suggests that the ascent routine test is not correct:

‘The procedure I have described does not produce infallible


knowledge of the informational state, for mistakes of the kind that occur
when the subject makes judgements about the world can also produce
inaccuracies when the same procedure is re-used for this different purpose.
Consider a case in which a subject sees ten points of light arranged in a
circle, but reports that there are eleven points of light arranged in a circle,
because he has made a mistake in counting, forgetting where he began.
Such a mistake can clearly occur again when the subject re-uses the
procedure in order to gain knowledge of his internal state: his report ‘I
seem to see eleven points of light arranged in a circle’ is just wrong.’
(Evans, 1982, p229).

Evans claims that such mistakes are limited:

‘However, when the subject conceptualizes his experience in terms


of some very elementary concept, such as a simple colour concept like
‘red’, it is not easy to make sense of his making a mistake.’ (Evans, ibid,
p229).

46
This latter claim seems questionable. One can change one’s mind about the colour

an object looks. For instance, consider looking at a rock face which is brown with a

purple tinge. Because rock faces with a purple tinge are unusual, one might be

particularly struck by the purple tinge of the rock face, and judge, taking one’s experience

at face value, that the rock face is purple. On second thoughts, one might realize that the

rock face is not purple, but merely brown with a purple tinge. It seems quite possible that

one’s visual experience stays the same whilst one changes one’s mind about the colour of

the rock face. This point echoes Tim Williamson’s point that a hypochondriac may judge

falsely of an itch that it is a pain (Williamson, 2000, p24).

A second criticism concerns the condition of ‘taking one’s experience at face

value’, or ‘ignoring information extraneous to the experience’. On one understanding,

‘taking one’s experiences at face value’ means ignoring any collateral information or

background beliefs that one has. However, this condition seems too strong. Suppose that

an object looks red, and that one seeks to determine what one would judge on the basis of

one’s visual experience, ignoring all of one’s background beliefs, including such beliefs

as one’s belief that red objects look like that, the belief that there are many shades of red,

and the belief that objects are the way they look. In such a situation it does not seem that

one would form any judgement. Similarly, if an object looks square, and one seeks to

determine what one would judge, ignoring all of one’s background beliefs, including such

beliefs as the belief that squares have four sides, then it does not seem that one would

form any judgement.

47
On a second understanding of ‘taking one’s experiences at face value’ and

‘ignoring information extraneous to the experience’, these expressions mean ‘assuming

that all and only the properties represented by the experience are instantiated’. However,

on this understanding, to apply the ascent routine test is circular; one can apply it only if

one first knows what properties are represented by one’s experience.

Thus it seems that the conditions of taking one’s experiences at face value and

ignoring information extraneous to the visual experience are either circular, or are so

strong that they would rule out any property from being represented by one’s experiences.

6.2 Lewis’s Test

In the following passage, Lewis proposes a test for determining the content of

experience:

‘Visual experience depends on the scene before the eyes, and the
subject’s beliefs about that scene depend in turn partly on his visual
experience. The content of the experience is, roughly, the content of the
belief it tends to produce.

‘The matter is more complicated, however. The same visual


experience will have a different impact on the beliefs of different subjects,
depending on what they believed beforehand. (And on other differences
between them, e.g. differences of intelligence.) Holmes will believe more
on the basis of a given visual experience than Watson; and Watson in turn
will believe more than someone who suspects that he has fallen victim to a
field linguist no less powerful than deceitful. We should take the range of
prior states that actually exist among us, and ask what is common to the
impact of a given visual experience on all these states. Only if a certain
belief would be produced in almost every case may we take its content as

48
part of the content of the visual experience. (The more stringently we take
‘almost every’, the more we cut down the content of the visual experience
and the more of its impact we attribute to unconscious inference; for our
purposes, we need not consider how that line ought to be drawn.)

‘… Not all the content of visual experience can be characterized in


terms of the beliefs it tends to produce. It is part of the content that the
duck-rabbit look like a duck or a rabbit, but the belief produced is that
there is no duck and no rabbit but only paper and ink’ (Lewis, 1980, p239-
240).

Let us formulate Lewis’s test as follows:

Lewis’s Test: Necessarily, for all propositions p, and experiences e, if e would produce

the belief that p in almost every case, then e represents that p.

The above test provides a sufficient condition for an experience to represent a

proposition. It is not clear whether Lewis also intended his test to provide a necessary

condition for experience to represent a proposition. He writes above ‘Only if a certain

belief would be produced in almost every case may we take its content as part of the

content of the visual experience.’ This suggests that the test is intended to provide a

necessary condition on an experience representing a property. However, at the start of the

final paragraph quoted above, he writes ‘Not all the content of visual experience can be

characterized in terms of the beliefs it tends to produce’. This suggests that the test is not

intended to provide a necessary condition on visual experience representing a

proposition.

49
In what follows I will assume that Lewis intends to provide only a sufficient

condition for experience to represent some proposition p.

One problem with the ascent routine test that Evans pointed out is that subjects

are not infallible about how things look to them. Certain judgements that subjects make,

taking their experiences at face value, will not reflect the contents of those experiences. It

seems that Lewis’s test faces this problem also, since it seems as conceivable that

everyone may make a mistake about a particular experience as it is that one person may

make a mistake about a particular experience. For instance, consider Evans’s example of

someone who miscounts the number of lights before them, and judges that there are

eleven lights in a circle when in fact there are ten. Suppose that everyone makes this

mistake. It still seems that in this case everyone’s experiences represent ten lights, not

eleven. Lewis’s test would entail that their experiences represented eleven lights.

Similarly, suppose that, when looking at the rock face described above, everyone

was struck by its purple tinge and judged, taking their experiences at face value, that the

rock face was purple, and then changed their minds and decided that it was brown with a

purple tinge. It seems that everyone may change their mind about the colour of the rock

face without their visual experiences changing.

Lewis’s test assumes that errors such as these will be local and not widespread.

Whilst he is right that it is less likely that everyone will make a mistake than that one

person will make a mistake, it still seems conceivable that everyone will make a mistake.

50
A second problem concerns the use of extraneous information. When forming

judgements on the basis of their experiences, subjects may rely on background

information as well as the information that their experiences represent in order to form

their judgement, with the effect that the judgement goes beyond the content of the

experience.

The ascent routine explicitly rules out the use of such extraneous information.

Lewis’s test does not. It seems that Lewis’s appeal to what beliefs an experience would

produce in almost every case is an attempt to isolate those beliefs that do not rest on

information extraneous to the experience. The thought may be the following. Suppose

that a subject S forms a belief B, partly on the basis of an experience E, and partly on the

basis of information I which is extraneous to E. The suggestion may be that it is unlikely

that almost everyone has I, and therefore unlikely that E will produce B in almost every

case.

However, it seems conceivable that everyone has I. Consider Dummett’s example

above of a judgement that is based on an experience and also on information extraneous

to the experience: ‘I see that the Smiths forgot to cancel their newspapers’. Let us assume

that the subject in Dummett’s case sees a pile of newspapers outside the Smiths’ house,

and when forming the judgement that the Smiths forgot to cancel their newspapers, relies

on the information, extraneous to the experience, that a pile of newspapers outside

51
someone’s house is a reliable indication of the inhabitants of the house having forgotten

to cancel their newspapers.

It seems quite possible that everyone has this extraneous piece of information.

Perhaps there are no cultures in which people do not believe that a pile of newspapers

outside someone’s house reliably indicates that the inhabitants of the house have

forgotten to cancel their newspapers, and perhaps this is also the case in nearby worlds. If

this is the case, the experience of newspapers piling up in front of someone’s house will

produce in almost everyone the belief that the inhabitants of the house have forgotten to

cancel their newspapers. Thus Lewis’s condition that an experience produce the same

belief in almost every case does not succeed in isolating those beliefs that are based only

on an experience, and not on information extraneous to the experience.

A third problem with Lewis’s test is that it entails that the contents of say, F-type

experiences in a subject S are dependent on what beliefs F-type experiences produce in

others, which is a counter-intuitive result. Take any proposition which, at t1, F-type

experiences do not represent. Suppose that, at t2, F-type experiences produce the belief

that p in almost every case. Lewis’s test entails that, at t2, F-type experiences represent

that p. Lewis’s test allows for the following possibility: that S’s F-type experiences

change from not representing p to representing p entirely because of cognitive changes in

others. This situation could occur either if S’s F-type experiences produce the belief that p

at both t1 and t2, or if S’s F-type experiences produce the belief that p at neither t1 nor t2.

That this is a possibility seems to be a counter-intuitive consequence of Lewis’s test.

52
A fourth problem with Lewis’s test concerns sole inhabitants of worlds. Suppose

that S is the only subject in the world, and suppose that S is very persuasive, so that, were

anyone else to exist, S would persuade them to believe the same propositions on the basis

of their experiences as she does on the basis of hers. Lewis’s test has the consequence

that, for any proposition p such that an F-type experience in S produces the belief that p,

an F-type experience of S represents that p. This seems a counter-intuitive result.

6.3 The Same Look Test

Alan Millar, Fred Dretske and Colin McGinn propose a test for distinguishing

visible properties from non-visible properties in the following passages:

‘[I]f there is a possible situation in which it is false that A is G, yet


everything (including A), looks the same as in the actual situation [in
which A is G], then clearly the fact that A is G is not a fact merely about
the look of A.’ (Millar, 2000, p4).

‘If you can get a B (e.g., fool's gold) that looks (in the phenomenal
sense--assuming there is such a sense) the same as A (gold), then
experience does not represent something as A (or B).’ (Dretske,
correspondence).

‘[Since] it is possible for things to have the same appearance as


tigers, yet not fall under the concept of tiger, we should therefore restrict
the concepts invoked to characterize content to those that relate to the
appearances of things: concepts of colour, superficial texture, shape etc. In
this way we limit our ascription of content to how things seem to the
perceiver.’ (McGinn, 1982, p40).

53
We can formulate the test that Millar, Dretske and McGinn propose as follows.

The Same Look Test: Necessarily, for all properties F, if there is an object that is

F that looks the same as an object which is not F, then

being F is not a visible property.

The same look test is stronger than some of its proponents have noticed. No

properties that are standardly thought of as visible pass the same look test. Consider the

properties of being square and being triangular. Suppose that one sees a square in normal

lighting conditions, and a triangle placed under a prism in such a way it looks square. In

this example a square looks the same as an object which is not square, and therefore,

according to the same look test, being square is not a visible property. A similar style of

argument would entail that colour properties and position properties are non-visible.

The problem with the same look test is that, if an object that is F looks the same

as an object which is not F, this is consistent with them both looking F. Therefore, the fact

that gold looks the same as fool’s gold does not entail that being gold is a non-visible

property; it may be that both gold and fool’s gold look gold. Similarly, the fact that a

square looks the same as a triangle that is under a prism is consistent with both the square

and the triangle looking square.

6.4 The Veridicality Test

54
Consider the following quotations from Alex Byrne, John Pollock and Susanna

Siegel.

Alex Byrne writes:

‘Imagine someone with normal vision looking at an object that is


shaped and coloured exactly like a red tomato. She might characterize the
scene before her eyes by saying that there seems to be a red ripe bulgy
tomato before her. Presumably the content of her experience at least
concerns the colour and shape of the object. But does it also specify the
object before her as ripe, or as a tomato?... Is her experience some kind of
illusion if the object is a red but unripe tomato, or if the object is made of
papier-mâché?’ (Byrne, 2001, p202).

John Pollock writes:

[C]onsider the easiest case, which is the case in which it turns out
that the wall is all that is left of an old church— there is nothing behind it.
When we take ourselves to be seeing a church we are making a mistake,
but it is not a perceptual error…. The information encoded by perception
is not wrong — it is the judgment we make on the basis of it that is
wrong.’ (Pollock, forthcoming, p18).

Susanna Siegel writes:

‘The views I’ve just mentioned differ on what the veridicality


conditions of visual experience are. The less committal the contents of
visual experience, the less misperception there is. For instance, suppose
you and your brother come across a bowl full of expertly designed wax
fruits. Your brother is fooled into thinking that there are ripe juicy peaches
and pears in the bowl: he believes that there are peaches and pears in the
bowl, and this belief of his is false. The scene doesn't fool you, let’s
suppose, but only because you already believed on some non-perceptual

55
basis—for instance, from reading your daily horoscope's predictions—that
you would see some fake fruits today. Because you have this background
belief, you suspect trickery, and unlike your brother, you don’t end up
believing that there are peaches and pears in the bowl. Might there be in
such a case some sort of error in your visual experience, even if not in
your belief? A perceptual error would be one from which not even your
suspicion protects you: if you misperceive, then your visual experience’s
content is false: your visual experience tells you that there are peaches and
pears on the table, and that is incorrect. In contrast, if no perceptual error
is involved in this case, then the contents of your visual experience are less
committal, but correct: they tell you, for instance, that the contents of the
bowl have certain colours and shapes.’ (Siegel, 2006, p483).

These passages suggest the following test.

The Veridicality Test: Necessarily, for all objects x and properties F, if:

(i) x is the best or an equal best candidate for a case of

an object looking F, and

(ii) x is not F and

(iii) (i) and (ii) do not provide a reason for thinking that

x is being misperceived,

then being F is not a visible property.

Suppose that one is looking at a wax tomato, O, that looks the same as a paradigm

tomato, and is thus an equal best candidate for a case of an object looking to be a tomato.

If O’s not being a tomato would not provide a reason to think that an illusion is occurring,

then, according to the veridicality test, being a tomato is not a visible property.

56
One problem with the veridicality test is that intuitions differ as to whether an

illusion is occurring in cases such as the above. Some will say that no illusion is

occurring, and others will say that an illusion is occurring. It is hard to see how one will

determine whether an illusion is occurring without resorting to an independent test about

which properties are visible and which are not.

A second problem with the veridicality test is that it delivers incorrect results for

essential properties. For instance, suppose that being shaped is an essential property of

any object in which it is instantiated. Condition (ii), therefore, will not be true in any

possible world. From a necessarily false antecedent anything follows, including the

proposition that being shaped is visible and the proposition that being shaped is non-

visible.

A test similar to the veridicality test provides a sufficient condition for a property

to be visible:

The Veridicality Test*: Necessarily, for all objects x and properties F, if:

(i) x is the best or an equal best candidate for a case of

an object looking F, and

(ii) x is not F and

57
(iii) (i) and (ii) provide a reason for thinking that x is

being misperceived,

then being F is a visible property.

A problem for the veridicality test* is that it delivers the wrong results for

determinates and determinables. Suppose that being coloured is not an essential property

of any object in which it instantiated (for instance, suppose that any object could have

been transparent, and therefore colourless). Consider a case of an object, O, that looks

red; O is an equal best candidate for an object looking coloured. Suppose that O is not

coloured. Since not being coloured entails not being red, and since O looks red, then there

is reason to think that an illusion is occurring. The veridicality test* therefore entails that

being coloured is a visible property. In effect, the veridicality test* takes it to be trivial

that if being red is visible, then being coloured is visible. Even if this conditional is true,

it does not seem to be trivially true.

6.5 Behavioural Tests

In conversation philosophers sometimes appeal to behavioural data in order to

determine what properties visual experience represents. According to one argument, if the

way a dog responds when his master walks through the front door is different from the

way the dog responds when someone else walks through the front door, this is evidence

58
that the dog’s visual experience can represent that his master has walked through the front

door.

A premise of the argument is that the dog’s selective behavioural disposition

towards his master requires some psychological explanation. One explanation of the

dog’s behavioural dispositions is that the dog believes that his master has walked through

the front door.

Presumably it is an implicit assumption in the argument that dogs are not mentally

sophisticated enough to have beliefs, and therefore that an explanation in terms of the

dog’s beliefs is not possible. Once this explanation is rejected, it seems that only the

perceptual experiences of the dog can explain the dog’s behavioural dispositions.

Even if dogs are not mentally sophisticated enough to have beliefs, and therefore

if the psychological explanation of the dog’s behaviour cites only the dog’s perceptual

experiences, these perceptual experiences need not have rich contents. It is possible, for

instance, that the dog is responding to a particular smell when he is near his master, or a

particular pattern of colours and shapes that his visual experience represents when he sees

his master. The explanation of the dog’s behaviour does not require that the dog’s visual

experience represent the property of being a master.

Secondly, the proponent of the above argument seems committed to the view that

some of the dog’s behavioural dispositions are not explained by any psychological state

59
of the dog. It is natural to explain a given dog’s behaviour when it sees food by saying

that the dog wants food. A proponent of the above argument would presumably reject this

explanation on the grounds that dogs are not mentally sophisticated enough to have

desires. An explanation of the dog’s behaviour in terms of the contents of the dog’s visual

experiences would be incomplete. Even if the dog’s visual experience does represent the

existence of food, such a state would not explain the dog’s eagerly moving towards the

food. Thus it seems that a proponent of the above argument is committed to saying that

there is no psychological state of the dog that explains its behavioural disposition in the

presence of food. Once this is conceded, however, it is not clear why a proponent of the

above argument would insist that there needs to be a psychological state of the dog that

explains the dog’s behaviour when it sees its master.

6.6 Neuroscientific Data

One might wonder whether certain neuroscientific data provide constraints on

which properties are represented by visual experience. A scientist may seek to determine

which parts of the brain are active when a subject has a visual experience, and she may

try to identify these parts by examining a subject’s responses to his experiences. For

instance, a subject may report when he is having a certain kind of experience, and the

scientist can correlate that report with a certain pattern of activity in the brain.

Alternatively, a brain-damaged subject may not be capable of making certain reports,

60
such as reports about the location of seen objects, and the scientist may be able to use this

data to determine the function of the brain-damaged part of the brain.

Whether the scientist uses verbal or non-verbal responses of the subject, it seems

that she is restricted to these kinds of data when aiming to identify the neural basis of

experience. This means that the scientist is identifying, in the subject’s brain, those parts

that are responsible for the states of the subject when the subject is responding to his

experiences. Call the parts of the brain that the scientist identifies X. If we assume that

the existence of responses of the observed kind indicate the existence of visual

experiences, then it can be inferred that X is also responsible for the subject having visual

experiences.

A scientist may also seek to determine in which parts of the brain information

about natural kinds is processed, and may try to identify these parts by considering which

parts of the brain are active when a subject recognizes a natural kind, and which parts of

the brain are such that, when damaged, the subject is unable to recognize a natural kind.

Call the parts of the brain that the scientist identifies in this way Y. For all that has been

said so far, Y may be a part of X.

One might wonder whether the existence or not of channels of information from

Y to X could provide evidence that the visual experience of a given subject S represents

natural kind properties. Suppose that there is a channel of information from Y to X. It

does not seem that this could provide evidence that S’s visual experience represents

61
natural kind properties. It is possible, after all, that the states that S is in when he

responds to his experiences, rather than the visual experiences themselves, represent

natural kind properties.

Suppose that there is no channel of information from Y to X. It is not clear that

this shows that S’s visual experience does not represent natural kind information. Y was

identified as being responsible for S recognizing a natural kind. It is possible that S’s

visual experiences are pre-recognitional states that represent natural kind properties.

My argument has been that neuroscientific data of the above kind will not shed

light on what properties visual experience represents. I have not argued that it is

impossible in principle for neuroscientific data to shed light on what properties visual

experience represents.

6.7 The Concept Test

If one has a concept, it is natural to wonder how one acquired the concept. For

some concepts, it is intuitive that one acquires them because one’s visual experience

represents the properties picked out by those concepts. The concept red21 seems to be

such a concept. If it turns out that visual experience does not represent the property of

being red21, and yet one does have the concept red21, then there is a prima facie challenge

to explain how one acquired this concept.

62
One may wonder how one came to acquire concepts such as tomato and water.

One explanation is that one’s visual experience represents the properties of being a

tomato and being water respectively. However, this explanation is not available to

someone who holds a relatively sparse view about the properties that visual experience

represents, according to which visual experience does not represent the properties of

being a tomato and being water. Thus one might think that there is a prima facie

challenge to a defender of a sparse view of this kind to explain how one does come to

acquire the concepts tomato and water.

It seems that this is a challenge not only for sparse views about what visual

experience represents, but for any view that allows that some of our concepts pick out

properties that visual experience, or any other kind of perceptual experience, does not

represent. Some account of how these concepts are acquired is necessary, and one would

need to consider these accounts before deciding whether they would apply to concepts

such as tomato and water.

In the literature on externalism there is a popular account of how we fix the

reference of terms such as ‘water’ and ‘tomato’, and therefore understand sentences

containing these terms. This account is related to accounts of concept acquisition since it

seems that understanding sentences containing the words ‘water’ and ‘tomato’ is

sufficient for having the concepts water and tomato. According to one account within the

literature on externalism, we fix the reference of ‘water’ with a definite description. This

63
definite description may either be one such as ‘the stuff in our environment that falls from

the sky, fills the lakes and comes out of taps’, or one such as ‘the chemical structure that

is common to w1, w2… w10’ where w1, w2… w10 are ostensively demonstrated samples of

water. Understanding that the reference of ‘water’ is the reference of these definite

descriptions enables us to understand sentences containing the word ‘water’, and thereby

acquire the concept water. This is not to say that the concept water is equivalent to the

concept expressed by either of the definite descriptions in question, but only that grasping

the concepts expressed by the definite descriptions enables us to acquire the concept

water. The popularity of this account of how we acquire concepts as water suggests that

most philosophers do not think that we acquire these concepts on the basis of our visual

experiences representing the properties that they pick out.

As it turns out, philosophers who argue that visual experiences represent, say,

natural kinds often do so by arguing that visual experience is theory-laden. According to

this argument, visual experiences come to represent natural kinds as a result of the subject

having certain beliefs about natural kinds. Let us call the argument that a subject’s visual

experience represents a certain property as a result of the subject having certain beliefs

about that property the theory-laden argument. Clearly a philosopher who employs the

theory-laden argument to argue that visual experiences come to represent natural kinds

may not also hold that we acquire the concepts of these natural kinds on the basis of our

visual experiences representing natural kinds.

64
The theory-laden argument is the standard defence of the claim that visual

experience represents a rich range of properties, including properties such as being water.

Susanna Siegel writes:

‘First, I will discuss some cases in which a perceiver is disposed to


recognize a K-property on the basis of visual experience. I'll argue in each
sort of case that such sensitivity makes a difference to the phenomenology
of visual experience. Furthermore, I'll suggest, its making a difference to
visual phenomenology is a reason to think that visual experiences
represent the K-property to which the subject is sensitive.’ (Siegel, 2006,
p483-4)

For Siegel, K-properties are properties ‘other than colour, shape, illumination,

motion, and the property of being an object’ (Siegel, 2006, p482). Siegel thinks that a

subject’s visual experience comes to represent K-properties because the subject has

recognitional abilities for K-properties.

Helen Beebee holds that possessing the concept causation is what makes

perceptual representation of causation possible:

‘[T]here is some evidence to suggest that infants as young as six


months can ‘perceive causation’ in a simple kinematic sequence… [in a
footnote] ‘Perceive causation’ is in scare quotes because the claim should
not be taken to be the claim that six-month-olds have thick experience in
the sense that they represent kinematic sequences as being causal—since
they lack anything like the full-blown concept of causation that would
make such representation possible.’ (Beebee, 2003, p266).

The following quotation suggests that Peacocke would endorse the theory-laden

argument:

65
‘We see things as tomatoes, and not as anything weaker. If the
representational content of experience is given by what someone would
judge, taking that experience at face value, then our ordinary experiences
have a content concerning tomatoes, and not tomato-like objects.’
(Peacocke, 1983, p93).

Given that most philosophers who defend the view that visual experiences have

rich contents employ the theory-laden argument, these philosophers are not better-placed

at explaining how we come to have the concepts characterizing these contents than the

philosophers who argue that visual experiences do not have these rich contents.

6.8 Epistemological Tests

Let us call a judgement formed at least partly on the basis of a visual experience a

visual judgement. One might wonder about how visual judgements are justified. There

are some propositions p such that one’s visual judgements that p are normally justified by

one’s visually representing that p. For instance, if one visually judges, of some object O,

that O is red21, then normally, this kind of visual judgement is justified by one’s visual

experience representing that O is red21.

One can also visually judge, of some object O, that O is a tomato, and one may

wonder how a visual judgement of this kind is justified. On one account, such a visual

judgement is justified by one’s visual experience representing that O is a tomato.

However, if one thinks that visual experiences do not represent the property of being a

66
tomato, then one needs to provide an alternative account of how the visual judgement that

O is a tomato is justified.

As with the concept test, this is not a challenge only to the sparse view, but to any

view on which one can visually judge that p without one’s visual experiences

representing that p. For instance, looking at a lamp, Jill may visually judge that it would

make a good present for Eleanor, and few proponents of rich views of the content of

visual experience would wish to say that Jill’s visual experience represents that the lamp

would make a good present for Eleanor.

Thus most participants in the debate over the richness of the content of visual

experience would agree that some account is required of how, for some proposition p,

one’s visual judgement that p can be justified without one’s visual experience

representing that p. One would need to consider such accounts before one can decide

whether they apply to visual judgements about tomatoes.

One possibility is that one’s visual judgement that p, for some p, can be justified

by one’s ability to produce an inference whose conclusion is p. On this account, Jill’s

visual judgement, based on seeing a certain lamp, that it would make a good present for

Eleanor, is justified by Jill’s ability to produce an inference such as the following:

1.) The lamp has a striking shape.

67
2.) If the lamp has a striking shape, then it would make a good present for

Eleanor.

3.) The lamp would make a good present for Eleanor.

This account does seem to apply to the property of being a tomato. It could be that

Jill’s judgement, of some seen object O, that O is a tomato, is justified by Jill’s ability to

produce an inference such as the following:

4.) O is that way.

5.) If O is that way, then O is a tomato.

6.) O is a tomato.

In this inference, ‘that way’ may refer to a pattern of colours and shapes that Jill’s

visual experiences represent.

As we saw above in the discussion of the concept test, standard defences of rich

views of the content of experience employ the theory-laden argument. That is, they argue

that the ability of a subject S visually to represent a particular object as a tomato is related

to S’s disposition to judge that something is a tomato. It is not always clear, on these

accounts, whether this relation is causal or constitutive; that is, whether it is proposed that

S’s visual experience’s representing an object O as a tomato consists in S’s disposition to

judge that O is a tomato, or is a causal consequence of S’s disposition to judge that O is a

tomato. We can consider each option in turn.

68
Suppose that the fact that S’s visual experience represents that O is a tomato

consists in S’s disposition to judge that O is a tomato. On this supposition, it does not

seem that S’s visual experience can justify S’s judgement that O is a tomato, since it does

not seem plausible that one’s disposition to judge that p can justify one’s judgement that

p.

Suppose that the fact that S’s visual experience represents that O is a tomato is a

causal consequence of S’s disposition to judge that O is a tomato. There does not seem

any reason to think that the judgement that S is disposed to make is any more likely to be

true because S’s disposition has this effect on his experiences, and therefore, on this

supposition, it does not seem that S’s visual experience can justify S’s judgement that O

is a tomato.

Thus it seems that there are no epistemological advantages of standard defences

of the rich view over the sparse view. On neither view are judgements about properties

such as being a tomato justified by visual experiences that represent those properties.

6.9 Phenomenological Tests

Siegel appeals to phenomenological considerations in order to argue that visual

experiences have relatively rich contents. She defines ‘K-properties’ to be properties

69
‘other than colour, shape, illumination, motion, and the property of being an object’

(Siegel, 2006, p482). She writes:

‘The thesis I will defend is

Thesis K: In some visual experiences, some K-properties are represented.

My defence of Thesis K goes as follows. First, I will discuss some cases in


which a perceiver is disposed to recognize a K-property on the basis of
visual experience. I'll argue in each sort of case that such sensitivity makes
a difference to the phenomenology of visual experience. Furthermore, I'll
suggest, its making a difference to visual phenomenology is a reason to
think that visual experiences represent the K-property to which the subject
is sensitive.’ (Siegel, ibid, p482-4).

Siegel holds that the investigation of what phenomenological differences occur in

visual experience can help determine what properties are represented in visual

experience. I agree with Siegel, and argue for this view in chapter 1, where I also

consider Siegel’s argument.

7 The Significance Of The Question

One might wonder about the significance of the distinction between visible and

non-visible properties. Some have thought that part of the interest in the distinction is

epistemological. Siegel holds that what range of properties visual experience represents

will determine the range of propositions that visual experiences justify. She writes:

70
‘It would be odd if … what contents visual experiences had was
totally irrelevant to what propositions the experience (together with any
other epistemically relevant factors) provided justification for believing.
(Siegel, 2006, p488).

As we saw in the section on epistemological tests, given the way standard

versions of the rich view are argued for, including the way Siegel argues for a version of

the rich view, it does not seem that they have the epistemological consequences that

Siegel suggests they have.

The distinction between visible and non-visible properties seems to me to have

considerable intrinsic interest. Its interest does not lie only in its possible ramifications

for other areas of philosophy. Part of what makes the distinction interesting is that it is

intuitive, and yet hard to grapple with and hard to make precise. Making it precise

involves finding a constraint on the notion of a visible property, arguing that it is better

than rival constraints and defending it against objections. It is satisfying to think about

this problem and to try to develop a clear understanding of it.

8 Chapter Survey

In chapter 1 I define a kind of looking, phenomenal looking. Phenomenal looking

is defined by the phenomenal character principle, which holds that differences in visual

phenomenal character constrain differences in the properties that objects phenomenally

look to have. In the chapter I argue that objects phenomenally look to have only colours

and positions, and that the colour properties do not include properties such as being red.

71
Although, in this introduction, I have made use of the notion of the content of

visual experience, in chapter 1 I explain why, in the rest of the thesis, I will be using only

the notion of the properties that objects phenomenally look to have.

In chapter 2 I argue for a constraint on phenomenal looking, the phenomenal

looking exportation principle, and I argue that the phenomenal looking exportation

principle entails that when subjects do not see themselves, objects do not phenomenally

look to them to have observer-relative properties, such as being to the left of them and

being to the right of them.

In chapter 3 I provide three more arguments against the view that the position

properties that objects phenomenally look to have are observer-relative properties. These

arguments raise the question of what kinds of position properties objects do

phenomenally look to have. I consider various other accounts of what kinds of position

properties objects phenomenally look to have, and argue that they face serious objections.

In chapter 4 I defend a positive view about the position properties that objects

phenomenally look to have. According to the view I defend, the position properties that

objects phenomenally look to have, which I call phenomenal position properties, are

absolute properties, in the sense that they are not relations to any other objects.

72
Furthermore, I argue that necessarily, objects do not have the position properties

they phenomenally look to have. Thus, even if objects actually have absolute position

properties, these are not phenomenal position properties. My argument is structurally

similar to spectrum inversion arguments in the philosophy of colour, and I discuss these

arguments. I argue that the most plausible response to these spectrum inversion

arguments is to hold that objects do not have the colour properties that they

phenomenally look to have.

Traditionally philosophers who have been attracted to a sparse view about the

properties that objects phenomenally look to have held that objects phenomenally look to

have only two kinds of position coordinates: a left/right (x) coordinate and an up/down

(y) coordinate. In spite of defending a very sparse view about the properties that objects

phenomenally look to have, I hold that objects phenomenally look to have z coordinates

in addition to x and y coordinates. In chapter 5 I present an argument for this view.

In chapter 6 I discuss how objects would phenomenally look to beings with more

coarse-grained discriminatory capacities than us. For instance, suppose that patch1 and

patch2 phenomenally look red1 and red2 respectively to one. Consider a being, say a dog,

which cannot discriminate the colour that patch1 phenomenally looks from the colour that

patch2 phenomenally looks. A question arises as to how patch1 and patch2 phenomenally

look to the dog. One hypothesis is that both patch1 and patch2 phenomenally look red1 to

the dog, or that both patch1 and patch2 phenomenally look red2 to the dog. In chapter 6 I

73
discuss what alternative hypotheses there may be regarding how patch1 and patch2

phenomenally look to the dog.

74
Chapter 1

A Sparse View About The Properties That Objects

Phenomenally Look To Have

1 Phenomenal Looking

In this chapter I identify a certain kind of looking, which I call phenomenal

looking, and I explore what properties objects phenomenally look to have. I argue that

objects phenomenally look to have only colours and positions, and that the colour

properties do not include determinables such as being red.

I identify phenomenal looking by arguing for a constraint on it: a constraint that

states a necessary condition on a kind of looking. My methodology is similar to that of

someone who wishes to identify, say, a particular kind of justification, and does so by

identifying a constraint on a particular kind of justification.

The following principle is a preliminary formulation of the constraint:

The Synchronic

75
Phenomenal Character Principle: Necessarily, for all objects x, y and z, and all

properties F and G, if x looks F to z, y does

not look F to z, and y looks G to z, then

there is a visual phenomenal difference

between the ways that x and y look to z.

I intend to apply the constraint diachronically and across worlds. Therefore the

full constraint, the phenomenal character principle, quantifies over times and worlds, and

is as follows:

The Phenomenal Character Principle: Necessarily, for all objects, x, y and z, all

properties F and G, all times t1 and t2, and all

worlds w1 and w2, if x looks F to z at t1 in

w1, y does not look F to z at t2 in w2, and y

looks G to z at t2 in w2, then there is a visual

phenomenal difference between the way that

x looks to z at t1 in w1 and the way that y

looks to z at t2 in w2.

I assume that only one kind of looking satisfies the phenomenal character

principle, and I call this phenomenal looking. What it means to say that there is a visual

phenomenal difference between the ways that two objects, A and B, look to S is that what

76
it is visually like for S for A to look the way it does to S is different from what it is

visually like for S for B to look to the way it does to S.

The phenomenal character principle is phrased in terms of how things look to a

particular subject. Sometimes I refer to the properties that objects phenomenally look to

have and leave it implicit that there is some particular subject to whom these objects

phenomenally look to have the properties in question.

The phenomenal character principle uses the locution ‘an object looks F’, where

‘F’ is to be replaced by an adjective. In English, some properties can be expressed by

predicates of the form ‘is + adjective’. For instance, the property of being red can be

expressed by the predicate ‘is red’. However, some properties, for instance, the property

of being a tomato, are not expressed by predicates of the form ‘is + adjective’. There is

no predicate ‘is tomatoey’ which expresses the property of being a tomato.

In order to express the question whether an object can stand in the phenomenal

looking relation to the property of being a tomato, we could invent an adjective,

‘tomatoey’, stipulate that being tomatoey is identical with being a tomato, and then ask

whether objects can phenomenally look tomatoey. Instead, however, I will simply ask

whether an object can phenomenally look to be a tomato, or whether an object can

phenomenally look to have the property of being a tomato. Phenomenally looking to be

an F and phenomenal looking to have the property of being an F obey the same constraint

as phenomenally looking F; that is, if an object x phenomenally looks to be F, and

77
another object y does not phenomenally look to be F, but phenomenally looks to be G,

then there is a visual phenomenal difference between the ways that x and y phenomenally

look to be. The stronger principle which quantifies over times and worlds also applies to

phenomenally looking to be F.

Some philosophers have argued that the expression ‘looks to be’ refers to a more

epistemic kind of looking than the expression ‘looks’. However I do not use

‘phenomenally looks to be’ in a different sense from ‘phenomenally looks’; they both

obey the same constraint. Using the locution of ‘phenomenally looking to be an F’ is

merely a way of avoiding having to introduce new adjectives such as ‘tomatoey’.

1.1 Externalist-looking

One way of understanding phenomenal looking is to consider kinds of looking

that are not phenomenal. Suppose that everything that looks red to Joe looks green to Fay.

And suppose that Joe and Fay are looking at a tomato, which looks red to Joe and green

to Fay. Some externalists about perception, whom we shall call moderate externalists,

have suggested that there are two kinds of looking, internalist-looking and externalist-

looking. Internalist-looking is explained as the kind of looking involved in the tomato

looking different colours to Joe and Fay. Moderate externalists allow that tomatoes

internalist-look different to Joe and to Fay, even though cases of tomatoes internalist-

78
looking the way they do to Joe are normally caused by the same surface reflectance

properties as cases of tomatoes internalist-looking the way they do to Fay.

Externalist looking is introduced as follows:

Externalist-Looking: Necessarily, for all objects x and y, and all properties F and

G, if x internalist-looks F to y, and cases of objects

internalist-looking F to y are normally caused by those

objects being G, then x externalist-looks G to y.

The tomato internalist-looks different colours to Joe and to Fay. However, the

tomato may well externalist-look to have the same properties to Joe and Fay. If cases of

objects internalist-looking red to Joe are normally caused by those objects being G, and

cases of internalist-looking green to Fay are normally caused by those objects being G,

then the tomato will externalist-look G to Joe and to Fay.

Externalist-looking does not satisfy the phenomenal character principle. Suppose

that Joe is looking at a tomato, which internalist-looks red to him. Say that the tomato

externalist-looks G to him. Suppose that in world w1, cases of internalist-looking red to

Joe are normally caused by objects being F, where being F is distinct from being G. In

that world, when the tomato internalist-looks red to Joe, it will externalist-look F to him,

i.e. different from how it actually externalist-looks to him, even though there is no visual

phenomenal difference between the way the tomato looks to Joe in the actual world and

79
the way it looks in w1. It follows that externalist-looking does not satisfy the phenomenal

character principle.

Internalist-looking, by contrast, does seem to satisfy the phenomenal character

principle. It seems to follow from the tomato internalist-looking different to Joe and Fay

that there is a visual phenomenal difference between the way the tomato looks to Joe and

the way it looks to Fay.

Some externalists will reject the claim that there is internalist-looking. Let us call

such an externalist a radical externalist. Since externalist-looking was introduced in

terms of internalist-looking, a radical externalist will deny that there is externalist-

looking. A radical externalist will acknowledge that there is externalist-looking*, a

constraint on which is as follows:

Externalist-looking*: Necessarily, for all objects x and y, and all

properties F, x externalist-looks* F to y only if there

is a mental state s of y such that s is normally

caused by the presence of F-ness.

Thus, an object externalist-looks* red to Joe only if there is a mental state of Joe

which is normally caused by redness. A radical externalist will account for the intuition

that everything that looks red to Joe might look green to Fay as follows. They will hold

that, for some surface reflectance property F, what it is visually like for Joe for objects to

80
externalist-look* F to him is different from what it is visually like for Fay for objects to

externalist-look* F to her.

The disagreement between the radical externalist and the moderate externalist

concerns whether there is a kind of looking that is individuated in terms of visual

phenomenal character. The moderate externalist holds that there is such a kind of looking,

and the radical externalist denies this.

In this chapter I claim that there is a kind of looking, phenomenal looking, that is

individuated in terms of visual phenomenal character, and, in particular, one that obeys

the phenomenal character principle. By contrasting phenomenal looking with other kinds

of looking, I hope to make the claim that there is phenomenal looking plausible.

There is a third kind of externalist position which would reject an assumption

common to moderate and radical externalism. We will call this kind of externalist a

phenomenal externalist. A phenomenal externalist endorses the following claim:

Phenomenal Externalism: Necessarily, for all subjects x and y, if the mental states of x

are normally caused by the same properties as the mental

states of y, then what it is like for x to have x’s mental

states is the same as what it is like for y to have y’s mental

states.

81
According to a phenomenal externalist, if the mental states of Joe and Fay are

normally caused by the same properties of objects, then what it is like for Joe to see

tomatoes is the same as what it is like for Fay to see tomatoes. Both the moderate and the

radical externalist reject this claim. In this chapter I do not argue against phenomenal

externalism, but merely note that it seems counter-intuitive. It seems plausible that even if

the mental states of Joe and Fay are normally caused by the same properties of objects,

what it is like for Joe and Fay to see certain objects may differ. In what follows I will

assume that phenomenal externalism is false.

1.2 Epistemic Looking

It is intuitive that some looks-statements refer to a state of a subject that is at least

partly epistemic. For instance, if Joan is looking at a DVD cover, she may say ‘this film

looks intriguing’, and, intuitively, this looks-statement refers to an epistemic kind of

looking.

It seems that epistemic looking is not phenomenal looking. There can be cases in

which the way an object epistemically looks changes between t1 and t2 without there

being any visual phenomenal difference between the way the object looks at t1 and the

way it looks at t2. Suppose that, at t1, one is told that one is in a room where all and only

red things look green. When one is asked to pick out the red objects, one may well find

an object that looks green, and say ‘that object looks red to me’. At t2 one is told that the

82
room is such that all and only blue things look green. When asked to pick out blue things,

one may well find an object which looks green and say ‘that object looks blue to me’.

Assuming that one’s looks-statements in this context refer to epistemic looking,

the way that the objects epistemically look between t1 and t2 has changed, and yet there

need be no visual phenomenal difference between the ways that the objects look at t1 and

the ways that they look at t2; in the phenomenal sense of ‘looks’, the objects look green at

t1 and at t2. Thus epistemic looking does not satisfy the phenomenal character principle.

1.3 Nonconceptual Looking

Nonconceptual looking is constrained by the following principle:

The Nonconceptual Principle: Necessarily, for all objects x and y and all properties

F, x nonconceptually looks F to y only if x’s looking

F to y does not entail that y has a concept of F.

A kind of looking is conceptual iff it is not nonconceptual. One might wonder

whether there are any connections between the notion of phenomenal looking and the

notions of conceptual and nonconceptual looking. One might wonder, for instance,

whether phenomenal looking would have to be nonconceptual.

83
Prima facie there are no connections between phenomenal looking and

nonconceptual looking. The central notion in the constraint on phenomenal looking was

the notion of a visual phenomenal difference, and it does not seem that there is any

property of visual phenomenal differences which suggests that phenomenal looking

should be either nonconceptual or conceptual.

1.4 Theory-laden Looking

There is a position according to which all seeing is theory-laden. According to this

position, what theories one has can influence the way things look to one. Nelson

Goodman defends this position in the following passage:

‘The myths of the innocent eye and the absolute given are unholy
accomplices. Both derive from and foster the idea of knowing as a
processing of raw material received from the senses, and of this raw
material as being discoverable either through purification rites or by
methodical disinterpretation. But reception and interpretation are not
separable operations; they are thoroughly interdependent… Content
cannot be extracted by peeling off layers of comment.’ (Nelson Goodman,
1976, p8.)

And David Lewis seems to take the same view in the following passage:

‘Hintikka uses causal relations for cross-identification between the


actual world and its perceptual alternatives, but not for cross-identification
between these alternatives… I think my use of causal relations even in
cross-identifying between alternatives has its uses in the perceptual case…
and is indispensable in the doxastic and epistemic cases. It has its price:
suitably ordinary causal relations must prevail in the perceptual

84
alternatives, making causal information part of the content of perceptual
experience. But I think that objectionable only given the forlorn hope that
we can speak sensibly of the pure content of perceptual experience,
separated from all collateral information.’ (David Lewis, 1999, p380, n.7).

The view being defended in the above two passages seems to be that all looking is

theory-laden looking; that is, that all looking is such that what theories one has can affect

what properties objects can look to one to have. One might wonder whether there are any

connections between theory-laden looking and phenomenal looking. For instance, one

might wonder whether phenomenal looking would have to be looking that is not theory-

laden.

Prima facie there is no connection between phenomenal looking and theory-laden

looking. The visual phenomenal differences referred to by the phenomenal character

principle may be generated and influenced by the theories that one has, or they may not

be. Susanna Siegel defends a view on which the theories that one has can introduce visual

phenomenal differences into the ways that objects look; we shall consider her argument

below. There does not seem to be any obvious requirement that phenomenal looking is

either theory-laden or non-theory-laden.

1.5 The Relata of Phenomenal Looking

The property view is as follows:

85
The Property View: Phenomenal looking is a relation between two objects and a

property.

The two objects in question are the perceived object and the perceiving subject.

Thus, according to the property view, A’s phenomenally looking F to S is a matter of A’s

standing in the phenomenal looking relation to S and to the property of being F.

The propositional view is as follows:

The Propositional View: Phenomenal looking is a relation between a subject and a

proposition.

The relation in question is that of visual representation, and the proposition in

question may be singular or general. Let us call the view on which the proposition in

question is a singular proposition the singular propositional view, and let us call the view

on which the proposition in question is a general proposition the general proposition

view. According to the singular propositional view, A’s phenomenally looking F to S is a

matter of S’s standing in the visual representation relation to the proposition that A is F,

and according to the general propositional view, A’s phenomenally looking F to S is a

matter of S’s standing in the visual representation relation to the proposition that

something is F.

86
It seems that there are some grounds for favouring the property view over the

propositional view. Consider (1) and (2):

(1) A is the colour that it phenomenally looks to S.

(2) A is the colour that B phenomenally looks to S.

A defender of the singular propositional view is likely to endorse (3) as a

specification of the truth-conditions of (1):

(3) (1) is true iff the proposition about the colour of A that S stands in the visual

representation relation to is true.

It seems that the truth-conditions of (2) should be similar in structure to the truth-

conditions of (1). However a defender of the singular propositional view who endorses

(3) cannot accommodate this fact. Suppose that A phenomenally looks red, and B

phenomenally looks green. A defender of the singular propositional view will hold that S

stands in the visual representation relation to the propositions that A is red and that B is

green. However, the truth of neither of these propositions is relevant to the truth of (2).

On the supposition in question about the colours that A and B phenomenally look,

whether (2) is true is determined by whether or not A is green.

The argument against the singular propositional view is that it seems that a

defender of the view would endorse (3) as a specification of the truth-conditions of (1).

87
Intuitively, the truth-conditions of (2) have a similar structure to the truth-conditions of

(1). However, a defender of the singular propositional view cannot hold that anything like

(3) specifies the truth-conditions of (1).

A defender of the general propositional view does not seem to be better off than a

defender of the singular propositional view in this respect. Consider all the ways that A

phenomenally looks, and call the proposition that something has those ways P1. Consider

all the ways that B phenomenally looks, and call the proposition that something has those

ways P2. A defender of the general propositional view will hold that S visually represents

P1 and P2. Although the truth of P1 is relevant to the truth of (1), neither P1 nor P2 is

relevant to the truth of (2).

One might accept the property view, but hold the propositional property view:

Propositional Property View: Phenomenal looking is a relation between two

objects and a propositional property.

A propositional property is, for some proposition p, the property of being such

that p is true. A non-propositional property is a property that is not a propositional

property. Thus, the property of being such that 2 + 2 = 4 is a propositional property, and

the property of being red is a non-propositional property.

88
We should reject the propositional property view for the same reason that we

rejected the propositional view. Suppose that A phenomenally looks red to S, and suppose

that a defender of the propositional property view holds that A phenomenally looks red to

S iff A stands in the phenomenal looking relation to S and to the property of being such

that A is red. Then it seems that (1) will be true iff A is such that A is red. Suppose that B

phenomenally looks green. Assuming that the truth-conditions of (1) and (2) have a

similar structure, then (2) will be true, on the propositional property view, iff A is such

that B is green. However, given the suppositions that we have made concerning the

colour that A and B phenomenally look to have, intuitively (2) is true iff A is green.

In the argument just given we assumed that the propositions in question are

singular propositions. However, as in the case of the propositional view, the propositional

property view is no better off if the propositions are general.

Thus, it seems that we should accept the non-propositional property view:

The Non-Propositional Property View: Phenomenal looking is a relation between

two objects and a non-propositional

property.

There may be some reasons for thinking that S has perceptual states with

propositional contents. For instance, some think that, when S is hallucinating, and

therefore when nothing phenomenally looks any way to S, S is nevertheless in a

89
perceptual state with a certain propositional content. None of the above points challenges

this particular thought. The aim of the above points has been to show that there is no

obvious entailment from A’s phenomenally looking red to S to S’s being in a perceptual

state with a propositional content.

The non-propositional property view, and the property view in general, might be

challenged in a different way. Consider the looking/talking principle:

The Ways of Looking/Talking Principle: Ways of looking are analogous to ways of

talking, dancing, and eating, and ways of

performing actions generally.

We report ways of performing actions using adverbs. For instance, we might say

(1) or (2):

(1) Jenny talks quickly

(2) Chris dances well.

If the ways of looking/talking principle is correct, then, instead of saying (3), we

should say (4):

(3) This apple looks red

(4) This apple looks redly.

90
The ways of looking/talking principle challenges the property view in the

following way. We do not think that Jenny’s talking quickly involves Jenny standing in

the talking relation to the property of being quick, nor that Chris’s dancing well involves

Chris standing in the dancing relation to the property of being good. If the ways of

looking/talking principle is correct, then, by analogy, we should not think of object O’s

looking red to subject S as involving O standing in the looking relation to S and to the

property of being red. By extension, we should not think of O’s phenomenally looking

red to S as involving O standing in the phenomenal looking relation to S and to the

property of being red.

The motivation for the ways of looking/talking principle is that we are inclined to

say both (5) and (6):

(5) Jenny talks a certain way.

(6) Jenny looks a certain way.

Similarly, we may refer to the way that Jenny talks, and the way that Jenny looks.

However, the linguistic similarity between (5) and (6) seems misleading. (5) seems to be

short for (7), whilst (6) does not seem to be short for (8):

(7) Jenny talks in a certain way.

(8) Jenny looks in a certain way.

91
Similarly, rather than referring to the way that Jenny talks, it seems more correct

to refer to the way in which Jenny talks. Consider sentence (9).

(9) Jenny was described a certain way.

(9) seems to have two readings. On the property reading, (9) means that there is a

property that Jenny was described as having. For instance, she might have been described

as beautiful. On the adverbial reading, (9) means that the describing was performed in a

certain way. For instance, the describing might have been done concisely. (6) seems to be

more analogous to the property reading of (9) than it is to the adverbial reading of (9).

The first argument that the ways of looking/talking principle is false is that (5)

seems short for (7), but (6) does not seem short for (8). The second argument that the

ways of looking/talking principle is false is that (10) seems assertable but (11) does not

seem assertable:

(10)Jenny is the way she looks.

(11)Jenny is the way she talks.

If Jenny talks quickly, then (11) is true iff (12) is true, but (12) is not grammatical:

(12)Jenny is quickly.

92
Thus, it does not seem that (11) is assertable. By contrast (10) does seem to be

assertable. The fact that (10) is assertable suggests that we should say (3) rather than (4)

above. Thus, it seems that we should reject the ways of looking/talking principle, and

retain the non-propositional property view.

2 The Phenomenal Difference Test

In this section I discuss a test, the phenomenal difference test, that is designed to

help determine which properties objects phenomenally look to have. I will first apply it to

the property of being a tomato. In the presentation of the argument, it will be a

simplifying assumption that objects phenomenally look to have shape properties. I will

later argue that objects do not phenomenally look to have shape properties, but in the

argument that follows the assumption is harmless. I also assume, in the arguments that

follow, that objects phenomenally look to have shades of colour, and position properties.

2.1 Being a tomato

Consider all the specific colour and shape properties that a particular tomato

phenomenally looks to a subject to have. Call this highly specific set of colour and shape

properties the tomato colour and shape properties. The first part of my argument is that it

is possible for an object phenomenally to look to have the tomato colour and shape

93
properties without the object phenomenally looking to be a tomato. That is, the following

principle is true:

The Tomato Anti-Entailment Principle: For all objects x and y, x’s phenomenally

looking to y to have the tomato colour and

shape properties does not entail that x

phenomenally looks to y to be a tomato.

Suppose that in another possible world, w1, there are twin-tomatoes. Twin-

tomatoes are not tomatoes, but there is no visual phenomenal difference between the way

that tomatoes and twin-tomatoes look. Suppose that a given tomato phenomenally looks

to Oscar, an inhabitant of the actual world, to have the tomato colour and shape

properties, and a given twin-tomato phenomenally looks to twin-Oscar, an inhabitant of

w1, to have the tomato colour and shape properties as well.

It would be very implausible to say that this twin-tomato phenomenally looks to

twin-Oscar to be a tomato, and therefore that the twin-tomato is not the way it

phenomenally looks to him. Twin-Oscar has as much right to say that the tomato in the

actual world phenomenally looks to Oscar to be a twin-tomato, and therefore is not the

way it phenomenally looks to Oscar. The correct description of w1 seems to be that the

twin-tomato phenomenally looks to have the tomato colour and shape properties without

phenomenally looking to be at tomato. This example establishes the tomato anti-

entailment principle.

94
We are now in a position to apply the phenomenal difference test. The test can be

applied in two ways, a diachronic and a synchronic way. I will start with the diachronic

way.

Suppose that, at t1, O phenomenally looks to have the tomato colour and shape

properties, but does not phenomenally look to be a tomato. Suppose that at t2, O comes

phenomenally to look to be a tomato in addition to phenomenally looking to have the

tomato colour and shape properties. What kind of visual phenomenal difference might we

expect to notice between the way O looks at t1 and t2?

The kind of possibility envisaged in the paragraph above is stronger than that

entailed by the tomato anti-entailment principle, but it does not seem that there are

grounds for ruling it out, once the tomato anti-entailment principle has been granted.

Suppose that we applied the phenomenal difference test to two shades of colour.

At t3, O phenomenally looks red56, and then at t4, O phenomenally looks green32. It seems

clear what the visual phenomenal difference is between the way that O looks at t3 and the

way O looks at t4. This is the kind of visual phenomenal difference we are expecting to

notice between the way O looks at t1 and the way O looks at t2. It seems that we cannot

imagine a visual phenomenal difference between the way O looks at t1 and the way O

looks at t2, and this suggests that objects do not phenomenally look to be tomatoes.

95
A different application of the diachronic phenomenal difference test is as follows.

Suppose that at t1, O phenomenally looks to have the tomato colour and shape properties,

as well as the property of being a tomato. And suppose that at t2, O still phenomenally

looks to have the tomato colour and shape properties, but phenomenally looks to be a

banana instead of a tomato. What kind of visual phenomenal difference might we expect

to notice between the ways that O looks at t1 and t2? It seems that we cannot imagine a

difference, and this suggests that objects do not phenomenally look to be tomatoes, or

bananas.

Susanna Siegel has proposed an argument that suggests that she would hold that

there is a visual phenomenal difference between the way that tomatoes look before and

after one acquires a recognitional capacity for tomatoes. John Searle has argued that there

is a visual phenomenal difference between the way a house looks depending on whether

one thinks of the house as a mere façade, or as a whole house. I consider these arguments

in section 3.

Another possible response to the above argument is as follows. Suppose that

phenomenal externalism is true, and therefore that the phenomenal character of one’s

mental states is determined by what properties those mental states are normally caused

by. It might be that, at t1, the state of affairs of an object’s phenomenally looking to have

the tomato colour and shape properties is normally caused by that object’s being a

tomato, whilst at t2, the state of affairs of an object’s phenomenally looking to have the

tomato colour and shape properties is normally caused by that object’s being a banana.

96
The phenomenal externalist might argue that this difference in the causal facts suffices

for a visual phenomenal difference between the way O looks at t1 and at t2. This is a

possible response to the argument. In this chapter I am assuming that phenomenal

externalism is false, but I acknowledge that some may find it an attractive way out of the

above argument.

Now I will apply the phenomenal difference test in a synchronic way. Suppose

that we have two objects in front of us, A and B. Suppose that both A and B

phenomenally look to have the tomato colour and shape properties, and that in addition, A

phenomenally looks to be a tomato, while B does not.

Admittedly, this kind of possibility is also stronger than that entailed by the

tomato anti-entailment principle, but, again, it does not seem that there are grounds for

ruling it out, once the tomato anti-entailment principle has been granted.

What kind of visual phenomenal difference might there be between the ways that

A and B look? Again, suppose that we had applied the synchronic phenomenal difference

test to two shades of colour. If we had said that C and D phenomenally look the same

shape but C phenomenally looks green32 and D phenomenally looks red21, it would be

clear what the visual phenomenal difference would be between the ways that C and D

look to us. Returning to the application of the synchronic phenomenal difference test to

the property of being a tomato, it seems that we cannot imagine what visual phenomenal

97
difference there might be between the ways that A and B look to us, and this suggests that

objects do not phenomenally look to be tomatoes.

A different application of the synchronic phenomenal difference test is as follows.

Suppose that A and B phenomenally look to have the same colour and shape properties,

and A phenomenally looks to be a tomato, whilst B phenomenally looks to be a banana.

What kind of visual phenomenal difference might there be between the ways that A and B

look? It seems that we cannot imagine such a difference, and this suggests that objects do

not phenomenally look to be tomatoes or bananas.

I now apply the phenomenal difference test to the following properties, and I

argue that objects do not phenomenally look to have them.

• being a table

• being expensive

• for any properties F and G, having changed from F to G

• for some event E, causing E

• being red

• for any object x, being the same colour as x

• being square

2.2 Being a table, being expensive

98
Let us apply the phenomenal difference test to the property of being a table.

Suppose that we have two objects, A and B, which phenomenally look to have all the

same colour and shape properties, and A phenomenally looks to be a table and B does not

phenomenally look to be a table. What kind of visual phenomenal difference might there

be between the ways that A and B look? It seems that we cannot imagine such a visual

phenomenal difference, and this suggests that objects do not phenomenally look to be

tables.

Suppose that A and B phenomenally look to have the same colour and shape

properties, but A phenomenally looks expensive and B does not. What kind of

phenomenal difference might there be between the ways that A and B look? It seems that

we cannot imagine such a visual phenomenal difference, and this suggests that objects do

not phenomenally look expensive.

2.3 Depth properties

Let us now consider depth properties. It seems that at least objects phenomenally

look to have positions on an x, or left-right, axis, and on a y, or up-down, axis. Do objects

phenomenally look to have positions on a z, or forwards-backwards, axis? I shall call

positions on each of these axes ‘coordinates’, and I shall argue that it is prima facie

implausible that objects phenomenally look to have z coordinates.

99
For simplicity, I will assume in this argument that objects phenomenally look to

have certain sizes. Later in this chapter I will argue that objects do not phenomenally look

to have size properties, but the assumption in the following argument to the contrary is

harmless.

Consider the following anti-entailment principle:

The Depth Anti-Entailment Principle: Necessarily, for all objects x and y, there is

no size property F and no z coordinate G

such that x’s phenomenally looking F to y

entails that x phenomenally looks G to y.

Suppose one is looking at a red square kite in a blue sky, and one does not know

what the object of one’s perception is. The kite phenomenally looks a certain size. If the

kite increases in size between t1 and t2, the kite will phenomenally look a greater size at t2

than at t1. Although the kite phenomenally looks a certain size at t1, it does not seem that

there is any particular z coordinate which the kite phenomenally looks. And this seems

true regardless of what size property the kite phenomenally looks to have. Thus the size

anti-entailment principle seems intuitive. We are now in a position to apply the

phenomenal difference test.

100
Let us assume, for the sake of argument, that objects do phenomenally look to

have z coordinates. Suppose that an object, O, phenomenally looks the same size, shape,

and colour properties, and the same x and y coordinates from t1 to t2. At t1, O

phenomenally looks to have z coordinate z15, and then at t2 O phenomenally looks to have

z coordinate z20, so O phenomenally looks further away at t2 than at t1. We are imagining

an isolated change between t1 and t2 in the z coordinate that O phenomenally looks to

have. None of the other properties that O phenomenally looks to have at t1 changes. What

kind of visual phenomenal difference might we expect to notice between the way that O

looks at t1 and t2?

Suppose that instead of focussing on a change in the z coordinate that O

phenomenally looked to have between t1 and t2, we considered a change in the x

coordinate that O phenomenally looked to have between t1 and t2. That is, O

phenomenally looks the same colour, size, shape, y and z coordinate between t1 and t2,

but it phenomenally looks to have a different x coordinate at t1 from the one it

phenomenally looks to have at t2. The position O phenomenally looks to have at t2 is

either further to the right or further to the left of the position that O phenomenally looks

to have at t1. It seems that there would be a clear visual phenomenal difference between

the way that O looks at t1 and the way that O looks at t2.

Returning to the case of the change in the z coordinate that O phenomenally looks

to have, it seems we cannot imagine what kind of visual phenomenal difference there

101
might be between the way that O looks at t1 and the way that O looks at t2. This is prima

facie evidence that objects do not phenomenally look to have z coordinates.

In chapter 5 I argue that the possibility of 360 degree vision suggests that the

phenomenal difference test in this case can be met, and that objects do phenomenally

look to have z coordinates. The aim of this section is merely to show that there is a prima

facie challenge to the idea that objects phenomenally look to have z coordinates.

2.4 Change

Let us now consider the property of having changed some property or other, for

instance, the property of having changed position, or changed colour. Let us address the

question whether a given object, O, ever phenomenally looks to have changed colour

between t1 and t2. Certainly it is possible that there are two different colours, F and G, and

O phenomenally looks F at t1, and phenomenally looks G at t2. However, what is at issue

is whether it is ever the case that O phenomenally looks to have changed from red to

green, where ‘changed from red to green’ is within the scope of the ‘phenomenally

looks’.

Historically this problem has been usually put by asking whether there is

perception of change, as opposed to mere change of perception. Change of perception is

simply a matter of having different perceptions at different times. Perception of change,

102
as we shall interpret it, is a matter of objects phenomenally looking to have changed,

where, for two properties F and G, ‘was F and is now G’ is within the scope of

‘phenomenally looking’.

Suppose that at t1, object A phenomenally looks red, and at t2, A phenomenally

looks green. To say, at t2, that A phenomenally looks to have changed from being red to

being green is to say that A phenomenally looks to have been red earlier and green now.

Let us apply the phenomenal difference test to this case.

Suppose that there are two objects, A and B. At t1, A and B phenomenally look

red. At t2, A and B phenomenally look green. Whilst, at t2, B phenomenally looks to have

been red earlier and green now, A, at t2, does not phenomenally look to have been red

earlier and green now; A, at t2, simply phenomenally looks green. What kind of visual

phenomenal difference might we expect to notice between the ways that A and B look

between t1 and t2? It seems hard to imagine what visual phenomenal difference there

might be, and this suggests that objects do not phenomenally look to have been red earlier

and green now.

One could respond by arguing that the state of affairs in which at t1 A

phenomenally looks red, and at t2, A phenomenally looks green, entails that A

phenomenally looks to have been red earlier and green now. If this was the case, then we

could not apply the phenomenal difference test. The anti-entailment principle that is

relevant to this argument is as follows:

103
The Change Anti-Entailment Principle: For all objects x and y, properties F and G

and times t1 and t2, the facts that, at t1, x

phenomenally looks F to y, and, at t2, x

phenomenally looks G to y, do not entail

that, at t2, x phenomenally looks to y to have

been F previously and G at t2.

The change anti-entailment principle seems very intuitive. For instance if a

particular object, O, phenomenally looks red to S on Monday, and phenomenally looks

green to S on Tuesday, it seems implausible to say that it follows that, on Tuesday, O

phenomenally looks to S to have been red previously and green on Tuesday.

One might argue that, if one is continuously looking at O, and the colour that O

phenomenally looks changes from red to green, then it follows that O phenomenally

looks to have been red earlier and green now. However, this claim does not seem

plausible either. It seems intuitive that there are cases in which one continuously looks at

some object O between two times t1 and t2, and, at t1, O phenomenally looks red, and, at

t2, O phenomenally looks green, and one is not inclined to say that, at t2, O phenomenally

looks to have been red previously and green now. If there is an entailment principle that

concerns phenomenally looking to have been red earlier and green now, it is not clear

what it is.

104
The argument above concerning change generalizes to rule out objects

phenomenally looking to move. An object phenomenally looking to move is a matter of

the object phenomenally looking to be changing its position properties, which, in turn,

seems to be a matter of the object phenomenally looking to have been there before and

here now.

We can apply the phenomenal difference test to this case as well. Suppose that we

are looking at two balls, A and B, travelling the same distance at the same speed. Each

ball phenomenally looks the same shape and colour and size, and the positions that each

ball phenomenally looks to have are changing at the same rate. Moreover, whilst A

phenomenally looks to be moving, B does not phenomenally look to be moving. B is

such that the position it phenomenally looks to have is different at each time, but it does

not phenomenally look to be moving. What kind of visual phenomenal difference might

we expect to notice between the way that A looks and the way that B looks? It seems that

there will be no visual phenomenal difference, and this suggests that objects do not

phenomenally look to move.

2.5 Causal properties

Let us consider whether events phenomenally look to cause other events. Suppose

that there are two objects, A and B, and that the following facts obtain:

105
t1: Both A and B phenomenally look red.

t2: A phenomenally looks green; B phenomenally looks red.

t3: A phenomenally looks green; B phenomenally looks green.

Therefore, the colour that A phenomenally looks changes from red to green

between t1 and t2, and the colour that B phenomenally looks follows suit from t2 to t3. Let

us now ask whether, at t3, A phenomenally looks to have caused it to be the case that B is

green.

The answer to this question seems to be ‘no’. If one swapped B for a qualitative

duplicate, say C, the result would be that A looks to bear a relation to C rather than B,

which is a different way that A would look, and yet there would be no visual phenomenal

difference in the way that A looks.

However, we can remove B from within the scope of the ‘phenomenally looks’

and rephrase the original question by asking whether, at t3, A phenomenally looks to have

caused it to be the case that there is some green object. We can apply the phenomenal

difference test to this case. Suppose that, in addition to A and B above, there is another

pair of objects, C and D. The facts about the colours that C and D phenomenally look

over time are the same as with A and B, namely:

t1: Both C and D phenomenally look red.

t2: C phenomenally looks green; D phenomenally looks red.

106
t3: C phenomenally looks green; D phenomenally looks green.

The difference between A and B on the one hand, and C and D on the other hand,

is the following:

t3: A phenomenally looks to have caused it to be the case that there is some

green object; C does not phenomenally look to have caused it to be the

case that there is some green object.

What might the visual phenomenal difference be between the way A looks and the

way that C looks? It seems that we cannot imagine what this visual phenomenal

difference would be, and this suggests that objects do not phenomenally look to cause it

to be the case that there is something that is green.

There is also the question whether objects phenomenally look to cause it to be the

case that an object is at such and such location. The argument given above could be

adapted to imply that this property is not one that objects phenomenally look to have.

2.6 The property of having a back

107
Let us now consider whether objects phenomenally look to have backs. Suppose

we are looking at a particular tomato. One question is whether the tomato phenomenally

looks to have a back. A second question is whether there is a certain kind of back, say a

semi-spherical one, which the tomato phenomenally looks to have.

Before we apply the phenomenal difference test to this case, we should consider

whether the two anti-entailment principles in question hold. For instance, one might hold

that it follows from a particular object phenomenally looking red and round that it

phenomenally looks to have a back of a particular kind. This seems implausible. Suppose

that there is a world, w1, in which tomatoes are semi-spherical, and that in w1 a subject, S,

is looking at the curved side of a tomato. Intuitively, the tomato’s not having a semi-

spherical back would not provide a reason for thinking that the tomato is not the way it

phenomenally looks, and this suggests that in w1, the tomato does not phenomenally look

to have a semi-spherical back. This establishes that phenomenally looking red and round

does not entail phenomenally looking to have a semi-spherical back.

One might hold that phenomenally looking round entails phenomenally looking to

have some back, and one might defend this view by arguing that the properties that

objects phenomenally look to have are closed under obvious entailment. Later I argue

that phenomenally looking red21 does not entail phenomenally looking red, even though

the entailment between being red21 and being red is obvious. If this argument is sound, a

special argument would have to be given for thinking that phenomenally looking round

entails phenomenally looking to have a back.

108
In what follows I shall assume the anti-entailment principle that phenomenally

looking to be red and round does not entail phenomenally looking to have a back.

Suppose that objects A and B phenomenally look the same colour and shape properties,

and A phenomenally looks to have a back whilst B does not. What kind of visual

phenomenal difference might we expect to notice between the ways that A and B look? It

seems that we cannot imagine such a difference, and this suggests that objects do not

phenomenally look to have backs.

3 The Arguments of Searle and Siegel

Susanna Siegel and John Searle argue that there are visual phenomenal

differences that are not accounted for in terms of different colour and position properties

that objects phenomenally look to have. Searle gives two examples:

‘Consider the following figure:

‘This can be seen as the word ‘TOOT’, as a table with two large
balloons underneath, as the numeral 1001 with a line over the top, as a
bridge with two pipelines crossing underneath, as the eyes of a man
wearing a hat with string hanging down each side, and so on…..

‘Consider, for example, the difference between looking at the


front of a house where one takes it to be the front of a whole house, and
looking at the front of a house where one takes it to be a mere façade,

109
e.g. as part of a movie set. If one believes one is seeing a whole house,
the front of the house actually looks different from the way it looks if
one believes one is seeing a false façade of a house….It is part of the
content of my visual experience when I look at a whole house that I
expect the rest of the house to be there if, for example, I enter the house
or go round to the back.’ (Searle, 1983, p54-55).

Siegel also gives two examples:

‘Consider a page of Cyrillic text. The way it looks to someone


before and after she learns to read Russian seems to bring about a
phenomenological difference in how the text looks. (Christopher Peacocke
makes a similar phenomenological claim in ch. 3 of A Study of Concepts).
When you are first learning to read the script of a language that is new to
you, you have to attend to each word, and perhaps to each letter,
separately. In contrast, once you can easily read it, it takes a special effort
to attend to the shapes of the script separately from its semantic properties.
You become disposed to attend to the semantic properties of the words in
the text, and less disposed to attend visually to the orthographic ones.

The second example involves a different recognitional disposition.


Suppose you have never seen a pine tree before, and are hired to cut down
all the pine trees in a grove containing trees of many different sorts.
Someone points out to you which trees are pine trees. Some weeks pass,
and your disposition to distinguish the pine trees from the others improves.
Eventually, you can spot the pine trees immediately: they become visually
salient to you. Like the recognitional disposition you gain, the salience of
the trees emerges gradually. Gaining this recognitional disposition is
reflected in a phenomenological difference between the visual experiences
had before and after the recognitional disposition was fully developed.’
(Siegel, 2006, p490-491)

We can put the challenge posed by the above passages as follows: there exist

visual phenomenal differences that are best explained in terms of objects phenomenally

looking to have properties other than colours and positions.

110
In sections 3.1, 3.2 and 3.3 I will offer alternative explanations of the phenomena

described by Searle and Siegel. In section 3.4 I will argue that even if the phenomena that

Searle and Siegel describe do require that objects phenomenally look to have properties

other than colours or positions, these properties do not include natural kind properties,

such as the property of being a tomato.

I will consider three different kinds of case: firstly, Searle’s ‘TOOT’ aspect-

switching case, secondly, Searle’s house/façade case and thirdly, Siegel’s language and

pine tree cases.

3.1 Aspect-switching

Searle’s ‘TOOT’ figure below is one of a variety of ambiguous figures: figures

which one can see under different aspects, or which one can see as different types of

object.

111
Other examples of ambiguous figures include the duck/rabbit picture:

(from http://mathworld.wolfram.com/Rabbit-DuckIllusion.html)

And also the young girl/old woman picture:

(from http://mathworld.wolfram.com/YoungGirl-OldWomanIllusion.html)

I will argue that the phenomenal shifts in these examples are explicable in terms

of the following differences:

• differences in patterns of attention

• differences in how one takes the objects to be

112
• differences in how one visually imagines the objects to be

Searle and Siegel argue that the best explanation of the visual phenomenal

differences is that, in their terminology, visual experiences represent properties in

addition to colours and positions. Let us call this account of aspect-switching the content

view. Let us call the account I shall defend, which appeals to the above three factors, the

non-content view. The non-content view is so-called because it does not appeal to

differences in the properties that the objects in question phenomenally look to have.

I will use the expressions ‘seeing as’ and ‘seeing under a certain aspect’ in such a

way that it is not controversial that one can see ambiguous figures as different kinds of

objects, or under various aspects. I take the substantive issue to concern the question what

the best explanation of seeing as is.

3.1.1 Patterns of attention

Normally, aspect-switching on an ambiguous figure is accompanied by a shift in

one’s patterns of attention towards the figure, though these shifts need not be the same for

different individuals. My patterns of attention change in the following way when I see the

above ambiguous figures under different aspects. When seeing the duck/rabbit as a rabbit,

I tend to look at the picture from left to right, and when seeing it as a duck, I tend to look

at it from right to left. Also, when seeing it as a rabbit, I attend to the rabbit’s mouth and

eye together, when I see it as a duck, I attend to the duck’s eye and beak together. When

113
seeing ‘TOOT’ as the word ‘TOOT’, I tend to attend to the whole word at once, whereas

when I see it as a man wearing a hat with strings hanging down each side, I tend to attend

to the eyes first and then the hat. When I see the young girl/old woman as an old woman,

I tend to attend to the mouth and the eyes first, and when I see it as a young girl, I attend

to her cheek and left shoulder first.

Changing one’s patterns of attention towards a figure can cause a phenomenal

difference to occur. Consider the following picture:

Initially, one’s attention is evenly distributed over the shapes in the picture. After

one sees the Dalmatian in the middle of the picture, one attends to the specific outline of

the Dalmatian. There seems to be a phenomenal difference associated with this shift in

one’s pattern of attention.

Consider figure A below. One can see figure A as being composed of a white

triangle superimposed on black circles, and one can see figure A as composed of three

black circles with wedges cut out of them. When one sees figure A as composed of a

white triangle superimposed on black circles, it seems that one attends to the straight

lines between the three black circles that would be the edges of the white triangle. When

114
one sees figure A as composed of three black circles with wedges cut out of them, then

one attends to these circles alone without attending to the straight lines that would be the

edges of the white triangle.

3.1.2 How one takes the object to be

Cognitive states can have phenomenal character. There can be something it is like

to understand a proposition. Switching between one’s cognitive states can thus involve

phenomenal shifts. For instance, one experiences a phenomenal shift when thinking of

the two different meanings of the sentence ‘visiting royalty can be boring’. This

phenomenal shift is not perceptual: one can be thinking about the different meanings of

this sentence with one’s eyes shut and without one’s other sense modalities being

stimulated. It is natural to think that there is some shift in cognitive phenomenal character

between thinking of the duck/rabbit figure as a duck and thinking of it as a rabbit. Bill

Brewer relies exclusively on this factor in his account of aspect-switching:

115
‘[When] I see it as a duck, say, this is again a phenomenological
change, but one of conceptual classificatory engagement with the very
diagram presented to me. Similarly, when I shift aspects, and see it as a
rabbit, there is an alteration in this phenomenology of the categorization of
what is presented. (Brewer, 2006, p18).

3.1.3 Residual Visual Phenomenal Differences and Visual Imagination

One might argue that changes in one’s patterns of attention towards the

duck/rabbit figure, and changes in how one takes the figure to be, do not fully explain the

phenomenal shift that one experiences. One might argue that one can prevent one’s

patterns of attention from changing, perhaps by looking only at the duck/rabbit’s eye and

beak/ears, and still experience a visual phenomenal shift when one aspect-switches on the

figure.

If it is possible to aspect-switch on ambiguous figures without any change in

one’s patterns of attention, and if there is a residual specifically visual phenomenal

difference between seeing the figure under the two aspects, then the above two factors

that we have discussed are not sufficient to account for the phenomenon of aspect-

switching.

We may appeal to states of visual imagination in order to explain the residual

visual phenomenal difference. In particular, when one is looking at ambiguous figures,

one may well be imagining unseen aspects of the figure. Unlike ordinary objects,

116
ambiguous figures tend to be quite abstract, or lacking in detail. Thus it is not implausible

to suppose that one’s visual imagination might ‘fill in’ some detail when one looks at

ambiguous figures.

For instance, when one sees the ‘TOOT’ picture as a man wearing a hat with two

strings hanging down each side, one may visually imagine the boundary of a face below

the eyes. This may be all that one visually imagines. One may not visually imagine

features of the face such as a nose or a mouth. Merely visually imagining a small detail

such as the boundary of the face would be sufficient to generate a visual phenomenal

difference between seeing the picture as a man and seeing it as, say, a bridge. When one

sees the figure as a bridge with two pipes passing underneath it, one may visually

imagine the pipes as going off in such and such a direction. Again, this may be all that

one visually imagines. Merely visually imagining brief lines going off in a certain

direction would be sufficient to generate a visual phenomenal difference that would

distinguish seeing the picture as a bridge from seeing it as a man.

The aim of appealing to visual imagination to explain any residual visual

phenomenal differences that occur when one aspect-switches on an ambiguous figure,

having fixed one’s patterns of attention, is that the phenomenal character of visual

imaginative states is similar to (although not identical to) the phenomenal character of

visual experiences. There seems to be a reason why visual imagination is called visual

imagination. Thus shifts in one’s visual imaginative states may be able to account for the

residual visual phenomenal differences in question.

117
I shall now consider three objections to the above appeal to visual imagination.

Firstly, it does not seem, when we aspect-switch on ambiguous figures, that we are aware

of vivid imaginative states concerning the unseen aspects of the picture. This seems

correct. However, visual imaginative states need not be vivid. If one visually imagines

the back of the door that one is seeing as green, then the visual imaginative state that one

is in need not be especially vivid.

The second objection is that it is implausible to suppose that we imagine detailed

images of unseen aspects of the picture. We addressed this objection above: it is sufficient

to generate a visual phenomenal difference between seeing different aspects of an

ambiguous figure that one visually imagines very few unseen details.

The third objection is as follows.

1.) Seeing a normal duck as a duck is the same phenomenon as seeing the duck/rabbit

picture as a duck.

2.) When we see a normal duck as a duck, we are not visually imagining unseen

aspects of the duck.

Therefore,

3.) When we see the duck/rabbit picture as a duck, we are not visually imagining

unseen aspects of the picture.

118
My reply to this objection is to reject 1.). We noted above that many ambiguous

figures are abstract and lacking in detail. It seems plausible that one’s visual imagination

is more active when one is looking at more abstract drawings; one’s imagination is more

likely to fill in detail. If this is right, then, since the duck/rabbit drawing is relatively

abstract, it is plausible to suppose that seeing it as a duck is different from standard cases

of seeing ordinary ducks as ducks.

3.2 House/façade

Searle claims that a house will look different to one depending on whether one

thinks of it as a house or as a mere façade. Strawson has also claimed that a bush against

a wall will look different to one depending on whether one thinks of it as a bush or as a

painting on the wall (Strawson, 1974a, p46). Let us suppose that the claims are that the

house and the bush phenomenally look different depending on how one thinks of them.

By ‘façade’, Searle could have in mind the front wall of a house, or merely a large

painting of a house. Let us first suppose that he has in mind a painting of a house. On this

assumption, the case is that one is actually looking at a house, but one thinks one is

looking at a painting of a house.

When thinking of the house as a painting, one’s patterns of attention may change.

For instance, one may start to attend to the shadows that the window ledges cast, thinking

119
that the painter has drawn these particularly well to convey an impression of depth. One

may also attend to the different shades of colour of the front of the house, thinking that

the painter has skilfully captured the way the apparent colour of that object can be

affected by light falling unevenly on the surface of an object. Thus, when thinking of the

house as a façade, there may be features of it that one attends to that one did not attend to

when one thought of it as a house.

In addition to one’s patterns of attention changing, one may visually imagine the

house as having the back that a painting would have, and this shift in one’s imaginative

states would contribute to the phenomenal shift that one experiences. It may also be that

thinking of the house as a façade is phenomenally different from thinking of it as a house,

and that this difference partly explains the phenomenal shift that one experiences when

thinking of the house as a façade.

Let us now suppose that, by ‘façade’, Searle has in mind the front wall of a house.

In this case if there is a phenomenal difference between thinking of the house as a house

and thinking of it as a façade, it is unlikely that the above account of shifting patterns of

attention would help explain it. To explain the phenomenal difference in this case, I

would appeal to shifts in how one visually imagines the object to be, as well as shifts in

how one thinks of the object. Thus, my account of Searle’s example draws on the same

resources as my account of aspect-switching above.

3.3 Language acquisition and recognitional capacities

120
Siegel argues that the acquisition of certain conceptual abilities, such as the ability

to read Russian and the ability to recognize pine trees, can make certain kinds of objects,

such as Russian sentences and pine trees, phenomenally look different.

Siegel argues that there is a phenomenal difference between looking at a page

before and after one learns Russian. This claim seems plausible. Inevitably some

phenomenal shift will result just because, after one has learnt Russian, one can read the

text on the page, whereas before one could not, and understanding the text on the page

does have a certain kind of phenomenal character.

However, this account would not explain the visual phenomenal shift that occurs

in this case. Siegel argues that, after one has learnt Russian, one ‘become[s] disposed to

attend to the semantic properties of the words in the text.’ (Siegel, 2006). It is not clear

that one needs to appeal to the idea of attending to semantic properties of the words in

order to explain the visual phenomenal shift in question. An alternative explanation is that

one becomes disposed to attend to the linguistically significant aspects of certain letters;

aspects that distinguish those letters from other letters. For instance, there are two letters

in the Cyrillic alphabet that differ only by the presence of a hook on the bottom of one of

them. In Chinese the width, height and thickness of a stroke are relevant to determining

which character it is part of. Learning Russian and Chinese will cause one to be disposed

to attend to such features of the letters as these, and these altered patterns of attention

towards texts in Russian and Chinese will produce a visual phenomenal shift.

121
If this explanation is correct, then the largest phenomenal shifts will occur when

one learns a language which is written in an alphabet with which one is not familiar. A

phenomenal shift may still occur when one learns a language which is written in an

alphabet that one does know. The fact that one can read sentences written in the language

will alter one’s patterns of attention to sentences written in that language. For instance,

depending on the language, one may come to look at newly intelligible sentences in a

systematic way from left to right, whereas before one learnt the language one might not

have looked at the sentences in a systematic way from left to right, because one might not

have been attempting to read them.

A similar explanation may be offered for the phenomenal difference that occurs

when one learns to recognize pine trees. After one learns to recognize pine trees, one will

start to attend to those features of pine trees that distinguish them from other trees, for

instance the colour or thickness of the bark. Acquiring a recognitional disposition for pine

trees will cause one’s patterns of attention to shift when one looks at a grove containing

pine trees and other sorts of trees.

3.4 Phenomenally looking a new way

Suppose that acquiring a recognitional disposition for pine trees does cause pine

trees phenomenally to look different. Suppose also that acquiring a recognitional

122
disposition for tomatoes, a disposition which I shall henceforth call the concept tomato,

causes tomatoes phenomenally to look different. I will now discuss the question what

new property tomatoes phenomenally look to one to have after one acquires the concept

tomato.

Someone might query at this stage the inference from ‘acquiring the concept

tomato makes tomatoes phenomenally look different’ to ‘there is some new F such that

acquiring the concept tomato makes tomatoes phenomenally look F’. However, this

inference seems plausible. If x phenomenally looks different from y, then the way x

phenomenally looks is different from the way y phenomenally looks. Given that any

object x phenomenally looks a certain way iff x phenomenally looks to have a certain

property, and given that it is doubtful that acquiring the concept tomato makes tomatoes

phenomenally look to have fewer properties, then the inference in question is valid.

I will now argue that, supposing that there is a property F such that acquiring the

concept tomato causes tomatoes to look F, being F is not the property of being a tomato.

On twin-earth there are fruits that look and taste just the same as tomatoes do, but

which are not tomatoes, as they are made from different molecules from tomatoes. Call

these fruits on twin-earth ‘twin-tomatoes’. There is no visual phenomenal difference

between the way that twin-tomatoes look to inhabitants of twin-earth and the way that

tomatoes look to inhabitants of earth. Here we are assuming phenomenal internalism,

which is as follows:

123
Phenomenal Internalism: Necessarily, for all subjects s1 and s2, if s1 is a molecule-

for-molecule duplicate of s2, then s1 does not differ from

s2 with respect to the phenomenal character of the

mental states of s1 and s2.

Suppose that Oscar is an inhabitant of earth, and twin-Oscar is an inhabitant of

twin-earth. Developing the assumption we made above, just as Oscar’s acquisition of the

concept tomato causes tomatoes phenomenally to look some new F to him, so twin-

Oscar’s acquisition of the concept twin-tomato causes twin-tomatoes phenomenally to

look some new F to him. It seems that the acquisition of the concept tomato will bring

about the same kind of visual phenomenal shift as the acquisition of the concept twin-

tomato. The argument for this claim is that Oscar and twin-Oscar do not know the

difference between tomatoes and twin-tomatoes, and they are thus in the same types of

brain state, narrowly construed. Given phenomenal internalism, the acquisition of the

concepts tomato and twin-tomato by Oscar and twin-Oscar will produce the same kind of

visual phenomenal shift in Oscar and twin-Oscar.

Let us accept the claim that Oscar’s acquiring the concept tomato and twin-

Oscar’s acquiring the concept twin-tomato bring about the same kind of visual

phenomenal shift for Oscar and twin-Oscar. It follows that there is some new F such that

Oscar’s and twin-Oscar’s acquisition of their respective concepts causes tomatoes and

twin-tomatoes phenomenally to look F to Oscar and twin-Oscar respectively. If being F is

124
the property of being a tomato, then twin-tomatoes will not be the way they

phenomenally look to twin-Oscar. This is counter-intuitive, since twin-Oscar has as much

right to say that being F is the property of being a twin-tomato, and that tomatoes are not

the way they phenomenally look to Oscar. To avoid an asymmetric treatment of the cases,

it seems that the only option is to hold that being F is neither the property of being a

tomato nor the property of being a twin-tomato.

4 Determinabilism

I call the view that objects phenomenally look to have properties such as being

red, being green and being blue, determinabilism. In this section I offer three arguments

against determinabilism, and one argument which raises an issue concerning

determinabilism, but which is ultimately inconclusive.

4.1 Colour Illusions

Suppose that the following facts obtain:

(i) A is red21, and phenomenally looks red21.

(ii) A is square, and phenomenally looks triangular.

Given that (i) and (ii) obtain, it seems appropriate to say (1):

125
(1) A is partly the way it phenomenally looks, but not entirely the way it

phenomenally looks .

Suppose that the following fact obtains:

(iii) B is red21, and phenomenally looks red22.

If determinabilism is true, then if B phenomenally looks red22, then, in addition, B

phenomenally looks red. Assuming determinabilism, then if (iii) is true, it should be

appropriate to assert (2).

(2) B is partly the way it phenomenally looks in respect of colour, but not entirely the

way it phenomenally looks in respect of colour.

After all, if B phenomenally looks red, and if B is red, then it should follow that B

is partly the way it phenomenally looks in respect of colour. However, (2) seems false.

(3) seems a better description of the state of affairs.

(3) B is roughly the way it phenomenally looks in respect of colour, but not exactly

the way it phenomenally looks in respect of colour.

In reply, the determinabilist might defend the maximally specific principle:

126
The Maximally Specific Principle: Expressions such as ‘the way that x looks in respect

of colour’ are normally understood as concerning

the most specific colour properties that x looks to

have, rather than the determinable colour properties

that x looks to have.

If the maximally specific principle is correct, then the determinabilist would have

an explanation of why (2) seems false: we naturally understand the phrase ‘the way it

phenomenally looks in respect of colour’ as referring to the most specific colour property

that B phenomenally looks to have.

However, it is not clear that we should accept the maximally specific principle. In

a context in which one is separating out green, red and blue objects, it seems that

expressions such as ‘the way that x looks in respect of colour’ concern determinable

colours such as green, red and blue, rather than specific shades of those colours.

Similarly, in such a context expressions such as ‘the way that x is believed to be in

respect of colour’ intuitively concern determinable colours such as green, red and blue

rather than specific shades of those colours.

The fact that (2) seems false is therefore evidence that determinabilism is false.

4.2 The Vagueness Argument

127
Suppose that there is a sequence of colour patches, P1-P100, where P1

phenomenally looks red1, a clear shade of red, and subsequent patches in the sequence

phenomenally look progressively less clear shades of red, and P100 phenomenally looks

yellow20, a clear shade of yellow. This situation is displayed in figure 1.

Figure 1

P1 P2 P3 P100
……………………

Phenomenally
looks: Red1 Red2 Red3 Yellow20

According to the determinabilist, P1 phenomenally looks red in addition to

phenomenally looking red1. Suppose that P50 phenomenally looks red50, and suppose that

it is vague whether red50 is a shade of red. The determinabilist will presumably hold that

it is vague whether P50 phenomenally looks red. After all, if P1 phenomenally looks red, it

does not seem that there is a clear cut-off point in series at which the patches stop

phenomenally looking red.

By ‘it is vague whether P50 phenomenally looks red’ one might loosely understand

(1):

(1) It is vague whether the shade that P50 phenomenally looks is a shade of red.

128
However, (2) is the correct reading of ‘it is vague whether P50 phenomenally looks

red’, and it is to claims such as (2) that the determinabilist is committed.

(2) It is vague whether the phenomenal looking relation holds between P50 and the

property of being red.

(1) itself is not a controversial claim: it is a claim that an anti-determinabilist may

accept.

If it is clear that P1 phenomenally looks red, and it is vague whether P50

phenomenally looks red, then the determinabilist seems committed to there being a

varying level of vagueness in respect of the colours that P1-P50 phenomenally look. This

seems counter-intuitive. It does not seem that the level of vagueness in how P1 and P50

phenomenally look in respect of colour is different. We will call the phenomenon of a

varying level of vagueness with respect to the colour properties that objects

phenomenally look varying phenomenal vagueness.

The determinabilist could avoid being committed to varying phenomenal

vagueness only by saying that there is some colour determinable D such that it is vague

whether red1 is a shade of D, and vague whether P1 phenomenally looks D. For instance,

D may be reddish-brownish and the determinabilist may hold that it is vague whether red1

is a shade of reddish-brownish, and therefore that it is vague whether P1 phenomenally

129
looks reddish-brownish. Moreover, the determinabilist will have to specify such

determinables for every other patch in the sequence in order to avoid the consequence of

varying phenomenal vagueness.

The intuitive support for determinabilism is that we say that objects look red.

However, once it is pointed out that determinabilism, if it is to avoid being committed to

varying phenomenal vagueness, entails that there are comparatively arbitrary

determinables that objects phenomenally look, such as ones that begin three-quarters of

the way through the red part of the colour spectrum and end half-way through the blue

area of the colour spectrum, then determinabilism seems to be a less attractive view.

Determinabilism captures some of our pretheoretic intuitions but at the same time has

some slightly counter-intuitive consequences.

Thus, the fact that determinabilism seems to entail that objects phenomenally look

to have comparatively arbitrary determinables is a minor cost of the view.

4.3 Phenomenal Looking To Have Determinables Alone

Suppose that there are one hundred colour patches, P1-P100, and each of P1-P100

phenomenally looks a different shade of red to one. Consider now a dog which is looking

at P1-P100, and which cannot discriminate the colour any patch phenomenally looks from

the colour any other patch phenomenally looks.

130
Two hypotheses about how the patches phenomenally look to the dog are as

follows. On the first hypothesis, there is a shade, F, such that one of the patches P1-P100

phenomenally looks F to us, and which is such that all of the patches P1-P100

phenomenally look F to the dog. For instance, F might be red56, or red17. On the second

hypothesis, each of P1-P100 simply phenomenally looks red to the dog without

phenomenally looking any particular shade of red to the dog.

If objects phenomenally look red to us, then this lends some support to the second

hypothesis. One objection to the second hypothesis is that objects cannot phenomenally

look red, and this objection would be countered if objects phenomenally look red to us.

Let us assume that the second hypothesis is at least metaphysically possible. It seems

very difficult to imagine how the patches would phenomenally look to the dog on the

second hypothesis. That is, it seems very difficult to imagine what it would be like for an

object phenomenally to look red without phenomenally looking any particular shade of

red.

Consider the following argument against the determinabilist, formulated as a

reductio ad absurdum. The argument is that, given 2.) and 3.), we should reject 1.).

1.) Objects phenomenally look red to us.

131
2.) If objects phenomenally look red to us, then we can visually imagine a red object

that is no specific shade of red.

3.) We cannot imagine a red object that is no specific shade of red.

The argument for 2.) depends on the following two claims.

(i) It is metaphysically possible that objects phenomenally look red

without phenomenally looking a specific shade of red (the second

hypothesis above).

(ii) We have reliable introspective access to the ways that objects

phenomenally look to us.

Suppose that an object O phenomenally looks red21 and red to us. Given (i) and

(ii), it seems that we should be able to introspect on the ways that O phenomenally looks,

which ways include the property of being red and the property of being red21, and visually

imagine an object that phenomenally looks red but does not phenomenally look any

particular shade of red. If we can do this, then we should be able visually to imagine an

object that is red but no particular shade of red.

The determinabilist is likely to reject 2.). The determinabilist is likely to argue

that, since objects that phenomenally look red always phenomenally look some specific

shade of red, there is a good explanation of why we cannot visually imagine an object as

red without imagining it as a specific shade of red.

132
Suppose that it had been the case that objects that phenomenally look green

always phenomenally look circular. A defender of 2.) will argue that in such a world we

would be able to imagine an object as green without imagining it as circular. A

determinabilist is likely to reject that claim.

It is not obvious whether 2.) is true or not. Therefore, this argument against the

determinabilist is inconclusive.

4.4 Kinds of Colour Phenomenal Character

Given that (1) and (2) quantify over only actual objects, both seem plausible. An

anti-determinabilist will consider (2) to be vacuously true.

(1) For all objects x, and for all human subjects y, if x phenomenally looks red21 to y,

then there is some location property F such that x phenomenally looks F to y.

(2) For all objects x, and for all human subjects y, if x phenomenally looks red, then

there is a shade of red F such that x phenomenally looks F to y.

Suppose that an object A phenomenally looks red21, and that A phenomenally

looks at location l23. It seems that we are able to distinguish two kinds of phenomenal

character that our visual experience of A has: a particular kind of colour phenomenal

133
character and a particular kind of location phenomenal character. We can call the kind of

colour phenomenal character red21-phenomenal character, and the kind of location

phenomenal character l23-phenomenal character.

I will now discuss two cases that seem to count against determinabilism. The first

is as follows. Suppose that an object B phenomenally looks red21. According to a

determinabilist, B will also phenomenally look red. If the determinabilist is correct, it

seems that we should be able to distinguish two kinds of colour phenomenal character

that our visual experience of B has: red21-phenomenal character, and also a kind of colour

phenomenal character possessed by all and only visual experiences of objects that

phenomenally look red; we may call this kind of colour phenomenal character red-

phenomenal character. However, when an object, x, phenomenally looks red21, it seems

that we are not able to distinguish two kinds of colour phenomenal character that our

visual experience of x has, namely red21-phenomenal character and red-phenomenal

character. It seems that, when an object, x, phenomenally looks red21, we can distinguish

just one kind of colour phenomenal character, namely red21-phenomenal character. This

seems to be evidence that determinabilism is false.

The second case that seems to count against determinabilism is as follows.

Suppose that B phenomenally looks red21 and that C phenomenally looks red22. One’s

visual experience of B has red21-phenomenal character, and one’s visual experience of C

has red22-phenomenal character. Thus one’s visual experiences of B and C have very

similar but distinct kinds of colour phenomenal character.

134
According to a determinabilist, B and C also phenomenally look red. If this is

correct, then we should expect that one’s visual experience of B and one’s visual

experience of C have a common kind of colour phenomenal character, namely red-

phenomenal character. However, whilst one’s visual experience of B is very similar in

respect of colour phenomenal character to one’s visual experience of C, it does not seem

that there is a kind of colour phenomenal character with respect to which one’s visual

experiences of B and C are identical. This seems to be evidence that determinabilism is

false.

4.5 The Red Anti-Entailment Principle

The red anti-entailment principle is as follows:

The Red Anti-Entailment Principle: For all objects x and y, x’s phenomenally

looking red21 to y does not entail that x

phenomenally looks red to y.

Anti-entailment principles have played a significant role in the arguments of this

chapter. Perhaps the most controversial anti-entailment principle is the red anti-

entailment principle. If the arguments above against determinabilism are sound, then the

135
red anti-entailment principle is correct, and if the red anti-entailment principle is correct,

this bolsters the case for the other anti-entailment principles considered in this chapter.

5 Other Properties

5.1 Being the same shade as

I shall now consider whether it is ever the case that, for some x and y, x

phenomenally looks the same shade as y, where ‘the same shade as y’ is within the scope

of the ‘phenomenally looks’. I shall assume this reading of ‘x phenomenally looks the

same shade as y’ in what follows.

Suppose that one sees two objects, A and B. It is a consequence of the

phenomenal character principle that the property of being the same shade as B is not one

that A phenomenally looks to have. This is because if A phenomenally looks the same

shade as B, and if B is swapped for a duplicate C, and comes phenomenally to look the

same shade as C, there would have to be a visual phenomenal difference between the

ways that A looks at the different times. In fact there will be no such visual phenomenal

difference, and it follows that A does not phenomenally look the same shade as B.

One might wonder whether a given object A may phenomenally look the same

shade as some F, G, H object, where ‘some F, G, H object’ is within the scope of

136
‘phenomenally look’. In what follows I shall argue that A does not phenomenally look the

same shade as some F, G and H object.

Consider three objects, A, B, and C, and suppose that they all phenomenally look

red56. Furthermore A phenomenally looks the same shade as an F, G and H object, whilst

C does not phenomenally look the same shade as an F, G and H object. What visual

phenomenal difference might there be between the way that A looks and the way that C

looks? It seems that we cannot imagine what kind of visual phenomenal difference there

would be, and this suggests that A does not phenomenally look the same shade as an F, G

and H object.

The relevant anti-entailment principle is as follows:

The Same Shade Anti-Entailment Principle: For all objects x, y and z, there are

no properties F, G and H such that

x’s phenomenally looking red21 to z

and y’s phenomenally looking red21

to z entail that x phenomenally looks

to z the same shade as some F, G and

H object.

Someone might respond to the above argument by denying the same shade anti-

entailment principle. However, if any case of phenomenally looking F entails

137
phenomenally looking G, where F and G are different, one would have thought that

phenomenally looking red21 entails phenomenally looking red. The fact that the latter

entailment does not hold means that there would have to be a special reason to deny the

same shade anti-entailment principle, and it is not clear that there is such a reason.

We can apply the same argument to other relations, such as being the same length

as. For instance, suppose there are three lines, A, B and C, and each line phenomenally

looks length F. A phenomenally looks the same length as an F, G and H object, but C does

not phenomenally look the same length as an F, G and H object. What visual phenomenal

difference might we expect to notice between the way that A looks and the way that C

looks? It seems that we cannot imagine such a visual phenomenal difference, and this

suggests that A does not phenomenally look the same length as an F, G and H object.

5.2 Being square

Suppose that A and B are squares. Furthermore, suppose that the positions that A’s

boundary points phenomenally look to have are arranged in a square, and the positions

that B’s boundary points phenomenally look to have are arranged in a square.

Furthermore, A phenomenally looks square and B does not phenomenally look square.

What visual phenomenal difference might we expect to notice between the ways that A

and B look? It seems that we cannot imagine what visual phenomenal difference there

might be, and this suggests that objects do not phenomenally look square.

138
The relevant anti-entailment principle here is:

The Square Anti-Entailment Principle: For all objects x and y, the fact that the

positions that x’s boundary points

phenomenally look to y to have are arranged

in a square does not entail that x

phenomenally looks square to y.

One could argue that the square anti-entailment principle is false. However, if the

red anti-entailment principle is true, then one would have to have a special reason for

denying the square anti-entailment principle. It is not clear what such a reason would be.

Therefore it seems we have reason to think that objects do not phenomenally look square.

In the discussion of the tomato anti-entailment principle I assumed that objects

phenomenally look to have shape properties. However, this assumption was not essential.

It could have been replaced with the assumption that the position properties that various

parts of the boundaries of the objects phenomenally look to have are arranged in a certain

shape. I assume that the position properties that the boundary of a given object

phenomenally look to have may be arranged in the shape, for example, of a square or a

circle.

5.3 Having a certain length

139
Suppose that A and B are straight lines. A consists of dots A1-A10 and B consists of

dots B1-B10. Furthermore, suppose that one sees dots A1-A10 and dots B1-B10, and that the

position that A1 phenomenally looks to be in is a distance of L from the position that A10

phenomenally looks to be in, and the position that B1 phenomenally looks to be in is L

from the position that B10 phenomenally looks to be in. Furthermore, suppose that A

phenomenally looks L long and B does not phenomenally look L long. What visual

phenomenal difference might we expect to notice between the ways that A and B look? It

seems hard to imagine what such a visual phenomenal difference might be, and this

suggests that neither A nor B phenomenally looks to have a length.

The relevant anti-entailment principle here is:

The Length Anti-Entailment Principle: For all objects x and y, if x is a straight line

composed of visible dots x1-x10, where x1 is

at the beginning of x and x10 is at the end of

x, and where x1-x10 are all the same size, the

fact that the position that x1 phenomenally

looks to y to be in is L from the position that

x10 phenomenally looks to y to be in does

not entail that x phenomenally looks to y to

have a length of L.

140
The argument for the length anti-entailment principle is similar to the argument

for previous anti-entailment principles: that if the red anti-entailment principle is true,

then there would have to be a special reason to think that the length anti-entailment

principle is false. If the argument above is sound, then objects do not phenomenally look

to have length properties. Since length properties are one kind of size property, the

argument suggests that objects do not phenomenally look to have size properties. A very

similar argument to the one above would establish that objects do not phenomenally look

to have size properties.

In the argument against the claim that some lines phenomenally look the same

lengths as other lines, I assume that lines phenomenally look certain lengths. However,

this assumption is not essential. It could be replaced by the assumption that various parts

of the line phenomenally look to have various positions.

One might object to my argument that objects do not phenomenally look to have

length properties by endorsing the following two principles:

The Unrestricted Seeing Principle: Necessarily, for all objects w, x, y and z, if x

and y compose z, then w sees x and y only if

w sees z.

141
The Length Object-Property Principle: Necessarily, for all objects x and y, x sees y

only if there is a length property L such that

y phenomenally looks L to x.

The length object property principle is so-called because it posits a connection

between the objects one sees and the kinds of properties those objects phenomenally look

to have.

It is not obvious why one should accept the length object-property principle. It is

not obvious why, for instance, it is more plausible than the determinable colour object-

property principle which holds that a subject S sees an object O only if there is a

determinable colour D such that O phenomenally looks D to S. There may be some

correct object-property principles. For instance, consider the following object-property

principles:

The Colour Object-Property Principle: Necessarily, for all objects x and y, x sees y

only if there is a colour property C such that

y phenomenally looks C to x.

The Position Object-Property Principle: Necessarily, for all objects x and y, x sees y

only if there is a position property P such

that y phenomenally looks P to x.

142
Even these principles are not self-evidently true. Suppose that A and B compose

C, and that subject S sees A, B and C. As long as A and B phenomenally look to have

colours and positions, prima facie it seems that S might be able to see C without C

phenomenally looking to have a colour or position to S.

Furthermore, if one accepted the unrestricted seeing principle, then one might

pause before accepting the position object-property principle. Suppose that any two

objects compose a third, and therefore that the computer screen and the mug before me

compose an object. Suppose I see both the computer screen and the mug. If the

unrestricted seeing principle is correct, I see the composite screen-mug object. If the

position object-property principle is correct, then the screen-mug object phenomenally

looks to have a certain position. Presumably this position property is the property of

being, for two disjoint positions L* and L**, partly in L* and partly in L**. But it is not

obvious that the property of being partly in L* and partly in L** is one that objects

phenomenally look to have. Certainly this is a substantive claim.

Thus, it does not seem that there is any reason to accept the length object-property

principle, and there is some controversy concerning whether any object-property

principle is correct.

6 Conclusion

143
In this chapter I have argued, using a combination of anti-entailment principles

and applications of the phenomenal difference test, that, objects do not phenomenally

look to have properties such as being a tomato, having changed, and being square. The

arguments that I have employed are of a general form, and I believe that they apply,

mutatis mutandis, to any properties that are not colour and position properties. Thus, I

believe that arguments of the form used in this chapter show that objects phenomenally

look to have just colour and position properties. In addition, I have also argued that the

colour properties do not include determinable properties such as being red, or being

coloured.

144
Chapter 2

The Tension Between Exportation And The View That Objects

Phenomenally Look To The Left And Right Of Me

In this chapter I will defend the following principles:

The Belief Exportation Principle: Necessarily, for all subjects S and objects x, if a

singular term referring to x occurs within the scope

of ‘believes that’ in a true ascription of belief to S,

then x is believed by S to be some way.

The Phenomenal Looking

Exportation Principle: Necessarily, for all subjects S and objects x, if a

singular term referring to x occurs within the scope

of ‘phenomenally looks’ in a true ascription of a

state of phenomenal looking to S, then x

phenomenally looks some way to S.

145
The Seeing Principle: Necessarily, for all objects x and y, and all

properties F, x phenomenally looks F to y only if y

sees x.

The phenomenal looking exportation principle and the seeing principle together

entail the relational phenomenal looking principle. It is the relational looking principle

with which I am principally concerned in this chapter.

The Relational Phenomenal

Looking Principle: Necessarily, for all objects x, y and z, all relations

R, if x phenomenally looks to y to bear R to z, then

y sees z.

Suppose that one is looking at a particular building, A, and that one does not see

the Eiffel Tower, and that one says ‘A looks to me the same height as the Eiffel Tower’.

Supposing that ‘the Eiffel Tower’ is a singular term, then the relational phenomenal

looking principle entails that A does not phenomenally look to one the same height as the

Eiffel Tower, where ‘the same height as the Eiffel Tower’ is within the scope of

‘phenomenally look’.

Suppose that I am looking at an object, B, that is to the right of me, and I say ‘B

looks to the right of me’. Supposing that I do not see myself, then the relational

146
phenomenal looking principle entails that the object I see does not phenomenally look to

the right of me, where ‘to the right of me’ is within the scope of ‘phenomenally look’.

The relational phenomenal looking principle entails that, when I do not see

myself, objects do not phenomenally look to have the following properties:

• Being to the left of me

• Being in front of me

• Being far away from me

• Being circular at a slant from me.

1 Exportation For Belief

In ‘Quantifiers and Propositional Attitudes’, Quine argued that if (1) is true, then

(2) is true (Quine, 1956):

(1) Ralph believes that Ortcutt is a spy.

(2) Ralph believes z(z is a spy) of Ortcutt.

147
That is, if Ralph believes that Ortcutt is a spy, then one can export the name

‘Ortcutt’ from within the scope of the ‘believes that’. By contrast, (4) does not follow

from (3).

(3) Ralph believes that there exists a spy.

(4) There exists a spy such that Ralph believes z(z is a spy) of that spy.

In general, it seems that if a subject S believes that T is F, where ‘T’ is a singular

term, and ‘F’ is a predicate, then it is possible to export ‘T’ from within the scope of the

‘believes that’. Thus, in general, if S believes that T is F, then T is believed by S to be F.

Let us express the intuition here in its most general form:

The Belief Exportation Principle: Necessarily, for all subjects S and objects x, if a

singular term referring to x occurs within the scope

of ‘believes that’ in a true ascription of belief to S,

then x is believed by S to be some way.

I will assume that a singular term is one that contributes only its reference to the

proposition expressed by sentences containing it. I will also assume that proper names,

demonstratives and indexicals are singular terms. A Fregean would deny that proper

names, demonstratives and indexicals are singular terms, as we have defined ‘singular

148
term’. Later I will discuss a version of the belief exportation principle that a Fregean

could accept.

1.1 Kripke’s Example

There is an example, attributed to Kripke, that could be used to challenge the

belief exportation principle (thanks to Ofra Magidor for bringing this example to my

attention). Soames discusses the example in (Soames, 2003, p411-412), and below I

paraphrase his description of the example.

Suppose that I have one false belief that I express using the sentence ‘S’. I

introduce a name, ‘PU’ via the following reference-fixing description:

(5) The object x which is such that, if S, then x is Princeton University, and if not-S,

then x is Saul Kripke’s left thumbnail.

I think ‘S’ is true, so I think ‘PU’ refers to Princeton University. However, ‘S’ is

false, and ‘PU’ in fact refers to Saul Kripke’s left thumbnail. Thinking that Princeton

University is an institution of higher learning, I form the belief that PU is an institution of

higher learning.

149
Suppose that (6) is true. (6), together with the belief exportation principle, entails

(7), and (7) entails (8).

(6) I believe that PU is an institution of higher learning.

(7) PU is believed by me to be an institution of higher learning.

(8) Saul Kripke’s left thumbnail is believed by me to be an institution of higher

learning.

Soames thinks that (8) is false, and his solution is to reject (6). Soames does not

challenge the inference from (6) to (7), and therefore he does not himself take Kripke’s

example to undermine the belief exportation principle. In another work he takes the belief

exportation principle to be ‘intuitively compelling’ (Soames, 2005, p261).

Soames’s own solution is to reject (6), and to argue that (5) is not sufficient to

introduce the name ‘PU’. He uses Kripke’s example to argue that there are some

restrictions on the use of descriptions to fix the referents of names. (Soames, 2003, p416).

Robin Jeshion has discussed Kripke’s example, and she regards (8) as acceptable.

That is, she does not think that the example undermines the belief exportation principle,

nor that (5) is an unacceptable way of introducing a name. (Jeshion, forthcoming).

However, one could use the example to challenge the belief exportation principle.

One might hold that it is intuitive that (6) is true and intuitive that (7) is false.

150
I will consider two responses to this challenge to the belief exportation principle.

The first response is that (9) follows from (6):

(9) I have a belief about PU.

If (9) does follow from (6), then it no longer seems objectionable to suppose that

(7) follows from (6). However, in the context of the current challenge to the belief

exportation principle, it seems question-begging to suppose that (9) follows from (6).

After all, in (9) ‘PU’ has been exported from within the scope of the belief-ascription in

(6), so the move from (6) to (9) employs a variant of the belief exportation principle.

The second response is as follows. We are assuming that names are singular

terms. We have said that singular terms contribute only their referents to the propositions

expressed by sentences containing them. Let us suppose that ‘Bob’ is a very well-known

name for Saul Kripke’s left thumbnail: speakers know that ‘Bob’ refers to Saul Kripke’s

left thumbnail. If ‘PU’ is a singular term, then (6) is true iff (10) is true:

(10)I believe that Bob is an institution of higher learning.

If (10) is true, then it does not seem so objectionable to suppose that (11) is true:

(11)Bob is believed by me to be an institution of higher learning.

151
And (11) is true iff (8) is true, since Bob is Saul Kripke’s left thumbnail.

Let us call the view that names are singular terms Russellianism. If Russellianism

is correct, then (6) entails (10). One might think that this consequence is implausible, and

thus reject Russellianism in favour of Fregeanism, the view that names contribute modes

of presentations of referents to propositions expressed by sentences containing those

names. According to a Fregean, (6) does not entail (10), since the names ‘Bob’ and ‘PU’

are associated with different modes of presentation.

Thus, in reply to Kripke’s example it seems that there are three options. Firstly,

one could, as Soames does, argue that (6) is false. Secondly, one could, as Jeshion does,

argue that (7) is acceptable. Thirdly, one could endorse Fregeanism and argue that (6)

does not entail (7). Endorsing Fregeanism does not involve abandoning the belief

exportation principle, but it involves the denial that there are any singular terms. This

denial removes the theoretical interest of the belief exportation principle.

A Fregean who denies that (6) entails (7) may accept that, ordinarily, it is possible

to export proper names from within the scope of attitude ascriptions. A Fregean may say

that (6) does not entail (7) for a reason that is peculiar to the situation. For instance, a

Fregean may say that (6) does not entail (7) because I do not know which thing ‘PU’ is,

or because I am not acquainted with Saul Kripke’s left thumbnail as ‘PU’.

152
Let us suppose that a Fregean holds that names, demonstratives and indexicals

belong to the same semantic category. Let us say that they are all singular terms*. A

Fregean may accept the following principle:

The Belief Exportation Principle*: Necessarily, for all subjects S and objects x, if there

is a singular term* ‘T’ such that

(i) ‘T’ refers to x;

(ii) ‘T’ occurs within the scope of ‘believes that’

in a true ascription of belief to S;

(iii) S knows which thing T is;

then x is believed by S to be some way.

The belief exportation principle* is like the belief exportation principle, except

that the belief exportation principle* concerns singular terms*, and a condition of

knowledge-which has been added to the antecedent of the conditional in the principle.

In what follows I will assume the belief exportation principle, though my

argument does not depend on the assumption of this principle instead of the belief

exportation principle*.

2 Exportation For Phenomenal Looking

153
Analogues of the belief exportation principle apply for desires, hopes, fears and

assertions. If a subject S desires that John open the door, then John is desired by S to

open the door. If S said that John opened the door, then John was said by S to have

opened the door.

Phenomenal-looks statements ascribe mental states to subjects: they ascribe the

mental state of being phenomenally looked to by a certain object. Given that analogues of

the belief exportation principle seem to apply in general to ascriptions of mental states to

subjects, it seems that there is a prima facie reason to think that an analogue of the belief

exportation principle applies to states of being phenomenally looked to. This analogue of

the belief exportation principle is the phenomenal looking exportation principle.

The Phenomenal Looking

Exportation Principle: Necessarily, for all subjects S and objects x, if a

singular term referring to x occurs within the scope

of ‘phenomenally looks’ in a true ascription of a

state of phenomenal looking to S, then x

phenomenally looks some way to S.

One might argue that there is a reason why the belief exportation principle holds,

and why analogues of it hold for hopes, fears and desires, and that this reason does not

154
support the phenomenal looking exportation principle. I shall now consider some

arguments of this form.

2.1 Propositional Contents

One might argue that the belief exportation principle and analogues of it apply

only to states with propositional contents, such as beliefs, desires, hopes and fears. As we

argued in chapter 1, phenomenal looking is a relation between two objects and a property.

Therefore states of being phenomenally looked to a certain way do not have propositional

contents.

It might be argued that this difference between states of being phenomenally

looked to a certain way, and states of belief, desires, hopes and fears, explains why the

belief exportation principle and analogues of it for desires, hopes and fears are true

without the phenomenal looking exportation principle being true.

However, it is not clear why the fact that beliefs have propositional content would

explain why the belief exportation principle is true. David Lewis held that beliefs were

relations to properties rather than propositions, but there seems nothing about this theory

in itself that rules Lewis out from accepting the belief exportation principle. (Lewis,

1979). The claim that (13) below follows from (12) is supposed to be pretheoretically

plausible, and is not based on a particular theory about the belief relation.

155
(12)John believes that Sally is tall.

(13)Sally is believed by John to be tall.

Thus, this argument does not seem to undermine support for the phenomenal

looking exportation principle.

2.2 Conceptual Representation

Let us introduce the term ‘represents’ as follows. If one believes, desires, fears or

hopes that Sally is tall, then one is representing the property of being Sally and the

property of being tall. If Sally phenomenally looks tall to one, then one is also

representing the property of being tall.

We can introduce ‘conceptual representation’ as follows:

Conceptual Representation: For all subjects s, states w and properties F, w of s

conceptually represents Fness only if s’s being in w

entails that s has a concept of being F.

A state represents a property Fness nonconceptually iff it represents Fness, but

does not conceptually represent Fness.

156
Let us suppose that beliefs, desires, fears and hopes represent properties in a

conceptual way, and states of being phenomenally looked to represent properties in a

nonconceptual way. One might defend the following principle:

The Conceptual Representation Principle: For all kinds of states K, if K

represents properties conceptually,

then this fact explains why the belief

exportation principle, or an analogue

of it, applies to K.

If the conceptual representation principle is correct, then the argument for the

phenomenal looking exportation principle is undermined.

According to one account of the nature of concepts, the conceptual representation

principle is false. On this account of concepts, the concept F is the ability to think about F

things as F things. Thus, the concept dog is the ability to think about dogs as dogs. It

follows from this account of concepts that thoughts by definition represent properties in a

conceptual way.

A thought is a mental state to which an analogue of the belief exportation

principle applies. If S thinks that John opened the door, then John is thought by S to have

opened the door. However, if thoughts represent properties conceptually by definition,

157
then this fact cannot explain why an analogue of the belief exportation principle applies

to thoughts. Therefore, if the conceptual representation principle is correct, a different

account of concepts from the one suggested above is required.

I will now present an additional problem for the conceptual representation

principle. Intuitively, the state of saying that p is one to which an analogue of the belief

exportation principle applies. However, I shall argue that the state of saying that p

represents properties nonconceptually. If there should be a common explanation of why

the belief exportation principle is true, and of why an analogue of the belief exportation

principle for the state of saying that p is true, then we should reject the conceptual

representation principle.

It seems that one can say that p without having concepts that characterize the

proposition that p. If an eight year old child, say, wanting to impress her chemistry

teacher, utters the sentence ‘water has two hydrogen atoms and one oxygen atom’, a

sentence which she has committed to memory from a chemistry textbook, it seems

intuitive that she says that water has two hydrogen atoms and one oxygen atom, even

though she may not have the concepts hydrogen atom and oxygen atom. The basis for this

claim is that it seems legitimate for the teacher to reply ‘what you just said is quite right,

but do you understand what you just said?’. That is, one can say that p without being able

to entertain the proposition that p.

158
If the child had said ‘water has two oxygen atoms and one hydrogen atom’ it

would have been legitimate for the teacher to say ‘what you said is not quite right’ which

presupposes that the child said that water has two oxygen atoms and one hydrogen atom.

Thus, the state of saying that p, for some p, seems to represent properties

nonconceptually. It also seems that an analogue of the belief exportation principle applies

to the state of saying. If Jack said that Sally is tall, then it seems to follow that Sally was

said by Jack to be tall. If these points are correct, then it seems that we should reject the

conceptual representation principle: there seems to be no connection between a state’s

representing properties conceptually and a state’s satisfying an analogue of the belief

exportation principle.

2.3 Knowledge Which

An objector might defend the following principle:

The Knowledge Which Principle: For all subjects S, if ‘T’ is a singular term, and if

one can export ‘T’ from within the scope of

‘believes that’ in a true belief ascription to S, then

this possibility is explained by the fact that S knows

which thing T is.

159
Consider again (12) and (13):

(12)John believes that Sally is tall.

(13)Sally is believed by John to be tall.

According to the knowledge which principle, if (12) entails (13), then the

explanation of this fact is that John knows who Sally is.

Consider now this principle:

The Nonconceptual

Representation Principle: For all states K, and all subjects S, if K is a state of

being phenomenally looked to, then since K

represents properties nonconceptually, what S

knows is irrelevant to the question what properties

K represents.

We now consider an objector who claims that the combination of the knowledge

which principle and the nonconceptual representation principle entails that there is no

reason to think that the phenomenal looking exportation principle is true.

160
The objection is that the possibility of exportation of singular terms in cases of

belief reports is explained by states of knowing which, but states of knowing which are

irrelevant to the ways that an object can phenomenally look to a subject.

It is not clear what the argument is for the knowledge which principle. The fact

that John believes that Sally is tall does not seem explanatory of the fact that Sally is

believed by John to be tall. Thus if the objection is to stand, further argument for the

knowledge which principle is required.

Secondly, we argued above that the state of saying that p, for some p, represents

properties nonconceptually, and that an analogue of the belief exportation principle

applies to the state of saying that p, for some p. If this argument is sound, and if the

explanation for the ability to export singular terms from within the scope of ‘says that’ in

true ascriptions of what is said is the same as the equivalent ability in the case of belief

reports, then it seems we should reject the knowledge which principle.

Therefore, this objection to the phenomenal looking exportation principle is not

convincing.

2.4 Consequences of the phenomenal looking exportation principle

The seeing principle is as follows:

161
The Seeing Principle: Necessarily, for all objects x and y, and all

properties F, x phenomenally looks F to y only if y

sees x.

The seeing principle seems to be intuitively plausible. The seeing principle and

the phenomenal looking exportation principle together entail the relational phenomenal

looking principle:

The Relational Phenomenal

Looking Principle: Necessarily, for all objects x, y and z, and all

relations R, if x phenomenally looks to y to bear R

to z, then y sees z.

The relational looking phenomenal looking principle entails that, when I do not

see myself, objects do not phenomenally look to bear relations to me, such as being to the

left of me, or being to the right of me.

Suppose that I see two objects, A and B, and I say (14):

(14)A phenomenally looks to me to the right of B.

162
One might think that, since I see B, (14) satisfies the relational phenomenal

looking principle. However, A’s being to the right of B is a three-place relation between

A, B and me: it is a matter of A’s being further to the right of me than B is. According to

the relational phenomenal looking principle, A phenomenally looks to stand in this three-

place relation to B and to me only if I see myself.

The relational phenomenal looking principle rules out objects phenomenally

looking a certain distance from me when I do not see myself. Suppose that I see two

objects, A and B, and I say (15):

(15)A phenomenally looks to me further away than B.

One might think that (15) satisfies the relational phenomenal looking principle

since I see B. However, A’s being further away than B is a three-place relation between A,

B and me: it is a matter of A’s being further away from me than B. According to the

relational phenomenal looking principle, A phenomenally looks to stand in this three-

place relation to B and to me only if I see myself.

Some have argued that shape constancy is a feature of visual perception.

According to some philosophers who endorse this claim, when a circular coin is at a slant

from me, it phenomenally looks circular and at a slant from me. However, it is a

consequence of the relational phenomenal looking principle that, if I do not see myself,

then the coin does not phenomenally look at a slant from me.

163
Some philosophers have argued that the property of being red is an observer-

relative property. According to these philosophers, an object’s phenomenally looking red

is a matter of the object’s phenomenally looking to cause a certain kind of experience in

me. However, the relational phenomenal looking principle has the consequence that, if I

do not see myself, then objects do not phenomenally look to cause a certain kind of

experience in me.

3 Responses To The Argument

In this section I consider two kinds of fall-back positions that one might defend in

the light of the above argument, and also an objection to the above argument.

3.1 Modes of Presentation

Consider the following sentence:

(16)S believes that A is to the left.

According to view that we can call propositionalism, belief is a relation between a

subject and a proposition. I will assume propositionalism in what follows. According to a

Russellian, a proposition is a structured entity containing objects, properties and relations.

164
Thus the proposition that Sally is tall contains Sally and the property of being tall.

According to Russellianism, if (16) is true, then what S believes is a proposition

containing A, the relation of being to the left of, and S. Thus, according to Russellianism,

(16) is elliptical for (17):

(17)S believes that A is to the left of S.

According to a Fregean, a proposition is composed of modes of presentations of

objects and properties. The same object may have different modes of presentation. Thus,

according to a Fregean, the propositions that Hesperus is a star and that Phosphorus is a

star are distinct. A Fregean could argue that there is an indexical mode of presentation to

the left which picks out the relation of being to the left of whatever subject is entertaining

that mode of presentation.

If (16) is true, a Fregean could argue that the proposition that S believes is

composed of the modes of presentation A and to the left. A Fregean could thus deny that

(16) is elliptical for (17).

Consider the following view:

The Property View: Phenomenal looking is a relation between two objects and a

property.

165
If the property view is true, then (18) entails (19).

(18)A phenomenally looks to me to the left.

(19)A phenomenally looks to me to the left of me.

The phenomenal looking exportation principle clearly applies to (19).

However, one could defend the Fregean view:

The Fregean View: Phenomenal looking is a relation between two objects and a

Fregean mode of presentation of a property.

According to the Fregean view, if (18) is true, then A stands in the phenomenal

looking relation to me and to the mode of presentation to the left, and thus (18) need not

entail (19). I will discuss the Fregean view below.

Fregean modes of presentation are most naturally thought of as concepts, in the

non-Fregean sense of ‘concept’. Thus, if a subject is entertaining the Fregean mode of

presentation to the left, then she has the concept to the left.

Let us now define the nonconceptual mode of presentation view.

The Nonconceptual Mode

166
Of Presentation View: Phenomenal looking is a relation between two objects and a

nonconceptual mode of presentation of a property.

Peacocke has argued that, in addition to Fregean modes of presentation, there are

non-Fregean modes of presentation:

‘We shall not do justice to the fine-grained phenomenology of


experience if we restrict ourselves to those contents which can be built up
by referring to the properties and relations that the perceived objects are
represented by the experience as possessing. We must, in describing the
fine-grained phenomenology, make use of the notion of the way in which
some property or relation is given in experience. The same shape can be
perceived in two different ways, and the same holds for the shape
properties, if we regard them as within the representational content of
experience. Ernst Mach’s example of one and the same shape that can be
perceived either as a square or as a regular diamond is a familiar example.’
(Peacocke, 2001, p240-241).

In an earlier work, Peacocke elaborates on his example of the shape that can be

perceived either as a square or as a diamond:

‘Intuitively, the difference between perceiving something as a


square and perceiving it as a (regular) diamond is in part a difference in
which symmetries are perceived. When something is perceived as a
diamond, the perceived symmetry is a symmetry about the bisector of its
angles. When something is perceived as a square, the perceived symmetry
is a symmetry about the bisector of its sides.’ (Peacocke, 1992, p76).

Peacocke argues that the content of visual experience is nonconceptual, and so we

should not think of the ways in which properties are given in experience that he refers to

as Fregean modes of presentation. Rather, we should take Peacocke to be introducing the

notion of a nonconceptual mode of presentation of a property.

167
The view that emerges from the above two quoted passages is a combination of

the property view and the nonconceptual mode of presentation view.

The Property and Nonconceptual

Mode of Presentation View: Phenomenal looking is a relation between two

objects, a property and a nonconceptual mode of

presentation of that property.

This view emerges from the passages as Peacocke suggests that the content of

experience consists both of properties and nonconceptual modes of presentation of those

properties.

A consequence of the property view is that (18) entails (19). This is also a

consequence of the property and nonconceptual mode of presentation view. According to

the property and nonconceptual mode of presentation view, (18) expresses the proposition

that A stands in the phenomenal looking relation to me, to the property of being to the left

of me, and to a nonconceptual mode of presentation of the property of being to the left of

me.

It seems that there are two views which will allow the inference from (18) to (19)

to be blocked: the Fregean view and the nonconceptual mode of presentation view.

Consider this passage from Gareth Evans:

168
‘It is not necessary… that the subject possess the egocentric
concept ‘to the right’ if he is to be able to have the experience of a sound
as being to the right.’ (Evans, 1982, p159).

Many who defend the view that (18) can be true even when I do not see myself

find Evans’s point above compelling. Given that Evans’s claim seems inconsistent with

the Fregean view, I shall focus on the nonconceptual mode of presentation view below.

There is a question about how one should decide whether or not there are two

nonconceptual modes of presentation of the same property. For instance, suppose that an

object phenomenally looks red1. According to the nonconceptual mode of presentation

view, the object stands in the phenomenal looking relation to some nonconceptual mode

of presentation of the property of being red1; let us call this nonconceptual mode of

presentation red1*. There is question about what the criterion is for determining whether

there is a second nonconceptual mode of presentation of being red1, say F*, which the

object may stand in the phenomenal looking relation to.

Frege provided a sufficient condition for introducing Fregean modes of

presentation. If it is possible rationally to doubt that A is B, then ‘A’ and ‘B’ have distinct

Fregean modes of presentation. However, it is not clear that this condition can be

appealed to here: there is no reason to think that what one can rationally doubt is a

constraint on nonconceptual modes of presentation.

169
Peacocke’s remarks in the passages quoted above suggest the following necessary

condition on there being two nonconceptual modes of presentation of the same property.

The Phenomenal Condition On

Nonconceptual Modes of Presentation: Necessarily, for all objects x, y and z, and all

properties F, if there are two nonconceptual

modes of presentation of being F, F* and

F**, then there is a visual phenomenal

difference between x’s phenomenally

looking F* to z and x’s phenomenally

looking F** to z.

According to Peacocke, there is a phenomenal difference between an object

phenomenally looking square and an object phenomenally looking diamond. If this is

correct, since being square is being diamond, the claim that there are two nonconceptual

modes of presentation of being square satisfies the phenomenal condition on

nonconceptual modes of presentation.

The way that Peacocke characterizes these phenomenal differences is in terms of

which properties are perceived. He says that ‘[w]hen something is perceived as a

diamond, the perceived symmetry is a symmetry about the bisector of its angles. When

something is perceived as a square, the perceived symmetry is a symmetry about the

bisector of its sides.’ (Peacocke, 1992, p76).

170
The natural view that emerges from this quotation is that an object may

phenomenally look to have a single shape property from t1 to t3, but phenomenally look

symmetrical along one axis from t1 to t2 and phenomenally look symmetrical along a

different axis from t2 to t3. It is not clear why Peacocke rejected this view and preferred to

account for the phenomenal difference in question by introducing nonconceptual modes

of presentation.

However, there might be other reasons to introduce nonconceptual modes of

presentation. John Campbell, for instance, argues that cases of objects looking to the left

and to the right provide sufficient motivation to introduce modes of presentation.

‘Although vision provides a great deal of information about


oneself, information whose most direct articulation uses the first person,
this is not because the visual information itself already employs the first
person. Rather, the egocentric frame used in vision employs monadic
spatial notions, such as ‘to the right’, ‘to the left’, ‘above’, ‘in front’ and so
on, rather than relational notions, such as ‘to my right’, ‘above me’, ‘in
front of me’, and so on.’ (Campbell, 1994, p119).

Let us assume the phenomenal condition on nonconceptual modes of presentation.

Suppose that an object can phenomenally look tallerTB, where tallerTB* is a

nonconceptual mode of presentation picking out the property of being taller than Tony

Blair. Suppose that, at t1, one sees Bill Clinton and one does not see Tony Blair, and

suppose that Bill Clinton phenomenally looks tallerTB to one.

171
Suppose that at t2 one sees Bill Clinton and Tony Blair, and that Bill Clinton

phenomenally looks taller than Tony Blair to one. Let us suppose that at t1 Bill Clinton

stands in the phenomenal looking relation to you and to the nonconceptual mode of

presentation tallerTB*, and let us suppose that at t2, Bill Clinton stands in the phenomenal

looking relation to the nonconceptual mode of presentation taller than Tony Blair*.

What kind of visual phenomenal difference might one notice between Bill Clinton

phenomenally looking tallerTB and Bill Clinton phenomenally looking taller than Tony

Blair? Certainly, when one comes to see Tony Blair, there will be visual phenomenal

differences owing to Tony Blair phenomenally looking to have a colour and a position.

However, what kind of visual phenomenal difference might one expect to notice that is

due in particular to Bill Clinton standing in the phenomenal looking relation to the

nonconceptual mode of presentation taller than Tony Blair* as opposed to the

nonconceptual mode of presentation tallerTB*? It is not obvious that there is such a

visual phenomenal difference, and, given the phenomenal condition on nonconceptual

modes of presentation, it would follow from this that there are not two different

nonconceptual modes of presentation of the property of being taller than Tony Blair.

Similar problems apply when one considers nonconceptual modes of presentation

of the property of being to the left of me; let us call one such nonconceptual mode of

presentation to the left*. Suppose that, at t1, subject S does not see himself, but does see

an apple, and the apple phenomenally looks to S to the left. Suppose that, at t2, S sees

himself and the apple, and the apple phenomenally looks to S to the left of S. What kind

172
of visual phenomenal difference might S notice between t1 and t2? Certainly, when S sees

himself, there will be visual phenomenal differences owing to S phenomenally looking to

S to have a certain colour and a position. But what visual phenomenal difference might S

expect to notice that is due in particular to S standing in the phenomenal looking relation

to the nonconceptual mode of presentation to the left of S* as opposed to the

nonconceptual mode of presentation to the left*? It is not obvious that there is such a

visual phenomenal difference, and, given the phenomenal condition on nonconceptual

modes of presentation, it would follow that there are not two different nonconceptual

modes of presentation of the property of being to the left of S.

This result is consistent with there only being the nonconceptual mode of

presentation to the left*. However, if the defender of nonconceptual modes of

presentations allows that an object can ever phenomenally look to bear some relation to

another object, it is not clear why they would deny that an object can phenomenally look

to S to be to the left of S when S does in fact see himself. It seems more plausible that the

lack of visual phenomenal difference identified in the above paragraph between t1 and t2

is explained by there being no nonconceptual mode of presentation to the left*.

The above considerations suggest that phenomenal looking is not a relation

between objects and nonconceptual modes of presentations. Thus we may assume that

(18) is elliptical for (19):

(18)A phenomenally looks to me to the left.

173
(19)A phenomenally looks to me to the left of me.

3.2 Temporal Properties

This objection is best phrased in the terminology of the content of experience as

opposed to the terminology of the ways that objects phenomenally look. The objector

argues as follows. There must be a time index in the content of experience. After all, if

my experience represents that this apple is green, my experience is not made veridical by

this apple being green at some time in the past or future. My experience is veridical only

if the apple is green now. Therefore my experience represents that this apple is green

now. And ‘now’ picks out a relation to a time. If t1 is the time of the experience, the

content of my experience is: this apple is green at t1. So we can say that ‘this apple

phenomenally looks green at t1’ reports the content of my experience.

But, the objection continues, the sentence ‘this apple phenomenally looks green at

t1’ does not satisfy the phenomenal looking exportation principle. Assuming that ‘t1’ is a

singular term, we cannot export it and say ‘t1 phenomenally looks to S to be some way’,

since S does not see t1.

One way to respond to this challenge would be to allow that the apple

phenomenally looks green now, but deny that ‘now’ picks out a relation to a time. One

174
could defend nominalism and continue to describe events as happening ‘now’, but deny

that there are entities, times, that events stand in the ‘occurring at’ relation to.

However, I will not pursue this nominalist line of response here. My answer to

this challenge is to deny that objects phenomenally look to have properties such as being

green now. Suppose that, at t2, we ask the following question about some apple A:

(20)Was A, at t1, the way it now phenomenally looks?

According to the objector, if, at t2, A phenomenally looks green, (21) reports the

way A phenomenally looks at t2:

(21)A phenomenally looks green now.

The objector holds that ‘now’ in (21) is within the scope of the ‘phenomenally

looks’. Since ‘now’ in (21) picks out time t2, the objector holds that if (21) is true, then

the way A phenomenally looks is being green at t2.

However, when we ask (20), we are not asking whether the apple is green at t2,

which is the time at which we ask (20). If, at t2, the apple phenomenally looks green, then

(20) is asking whether the apple was green at t1. Thus the natural reading of (20) is

incompatible with the objector’s claim that, when an object phenomenally looks green, it

175
really phenomenally looks green now, where ‘now’ is within the scope of the

‘phenomenally looks’.

The objector argued that a time index should be included within the content of

experience because otherwise the experience could be made ‘accidentally veridical’ by

the apple being green at some point in the past or future. Call this the accidental

veridicality problem. The accidental veridicality problem is removed if we use the

terminology of the ‘is the way it phenomenally looks’ instead of ‘veridical’. Suppose that

we ask (22):

(22)Is the apple the way it phenomenally looks?

In asking (22), we are clearly asking whether the apple now has the properties that

it phenomenally looks to have, and so there is no danger of how the apple was yesterday

or two years ago affecting the answer to this question. One can avoid the problem of

accidental veridicality by distinguishing between (22) and (23) and (24):

(23)Was the apple the way it phenomenally looks?

(24)Will the apple be the way it phenomenally looks?

Each question will receive an answer that is unaffected by the answers to the other

questions, and thus the accidental veridicality problem is avoided.

176
3.3 Definite Descriptions

The belief exportation principle and the phenomenal looking exportation principle

are constraints on singular terms occurring within the scope of certain mental state

ascriptions. Given our definition of a singular term, on which a singular term contributes

only its reference to the proposition expressed by a sentence in which it occurs, it is

plausible that definite descriptions are not singular terms.

Moreover, it seems plausible that definite descriptions cannot always be exported

from within the scope of mental state ascriptions. It does not seem, for instance, that (25)

entails (26):

(25)S believes that the F is G.

(26)The F is believed by S to be G.

Someone might exploit the fact that definite descriptions are not exportable from

within the scope of attitude ascriptions and argue that whilst (27) might not be true, (28)

may well be true:

(27)O phenomenally looks to me to the left of me.

(28)O phenomenally looks to me to the left of the subject here.

177
Call the appeal to definite descriptions to avoid the main argument of this paper

the definite description fall-back option. ‘Here’ in (28) is a singular term, and, by the

phenomenal looking exportation principle, is exportable. If ‘here’ refers to the exact

region that I occupy, then it seems that if I do not see myself, then I will also not see the

reference of ‘here’, in which case ‘here’ will not be exportable from within the scope of

‘phenomenally looks’ in (28). If ‘here’ refers to a larger region than the one that I exactly

occupy, then there will be occasions in which another subject exists in that region, and

the uniqueness condition implied by the definite description ‘the subject here’ will not be

satisfied.

A defender of the definite description fall-back option might refine their position

and argue that (29) may be true:

(29)O phenomenally looks to me to the left of the subject at the centre of this region

of space.

We suppose that the reference of ‘this region of space’ is large enough so that, if

the subject can see any region of space, she can see this one. We also assume that the

subject can see regions of space.

There is initially an issue of refinement. ‘The subject at the centre of this region of

space’ presumably means ‘the subject centred on the central point in this region of space’.

However, there is no point which is the exact centre of a given human being, and

178
therefore no region of space which has a human being centred on its central point. A

defender of the definite description fall-back option could solve this problem by arguing

instead that (30) may be true:

(30)O phenomenally looks to me to the left of the subject part of which occupies the

point at the centre of this region of space.

Suppose that I am a woman and am pregnant with a baby developed enough to be

a subject. Suppose that the reference of ‘this region’ in (30) happens to be one whose

central point is occupied by a part of my baby. If the central point is one that is also

occupied by a part of me, then the uniqueness implied by ‘the subject part of which

occupies the point at the centre of this region of space’ will not be satisfied.

If the central point is one that only a part of my baby occupies, then the

uniqueness of ‘the subject part of which occupies the point at the centre of this region of

space’ will be satisfied, but whether O is the way it phenomenally looks to me will

depend on which direction my baby is facing in, which is implausible.

To avoid these problems, what region is picked out by ‘this region’ would have to

be sensitive to whether one is pregnant or not, how big one’s baby is if one is pregnant,

and so on. It seems implausible to suppose that facts about the position properties that

objects phenomenally look to have are sensitive to facts about whether one is pregnant,

179
and facts about the size of one’s baby if one is pregnant. Thus the definite description

fall-back option is an unattractive view.

3.4 Exporting Predicates

Suppose that (31) expresses the same proposition as (32):

(31)A is red21.

(32)Red21ness is a property of A.

If (31) expresses the same proposition as (32), then, given propositionalism,

which we introduced above, (33) expresses the same proposition as (34):

(33)S believes that A is red21.

(34)S believes that red21ness is a property of A.

Since ‘red21ness’ is a singular term, then by the belief exportation principle, (34)

entails (35):

(35)Red21ness is believed by S to be a property of A.

180
If (33) entails (35), then one can export predicates from within the scope of

‘believes that’ in true belief ascriptions. If this is true, then one should be able to export

predicates from within the scope of ‘phenomenally looks’ in true ascriptions of states of

phenomenal looking. That is, if (31) expresses the same proposition as (32), then it seems

that (36) expresses the same proposition as (37):

(36)A phenomenally looks red21 to S.

(37)A phenomenally looks to S to have red21ness.

Since ‘red21ness’ is a singular term, then, by the phenomenal looking exportation

principle, (37) entails (38):

(38)Red21ness phenomenally looks to S to be a property of A.

By the seeing principle, (38) entails (39):

(39)S sees red21ness.

The argument against the phenomenal looking exportation principle is as follows.

The phenomenal looking exportation principle together with (36) entail (39). However,

(39) is false. Therefore we should reject the phenomenal looking exportation principle.

181
One might try to avoid this argument by arguing that (39) is true. There is a sense

of (39) which is true, the sense in which (39) is true iff some object phenomenally looks

red21 to S. However, this is not the sense of ‘sees’ in sentences such as ‘S sees that apple’,

and it is the latter sense of ‘sees’, which we may call the literal sense of ‘sees’, that the

seeing principle employs. Thus this way of avoiding the argument is not available.

One might argue that, even in the literal sense of ‘sees’, (39) may be true.

Suppose that ‘red21ness’ in (37), (38) and (39) refers to a trope that A has, or to a universal

which is wholly present in A. One might think that on this supposition, (39) would be

true. Some philosophers have claimed that one can see property instances as well as

objects.

However, this option does not seem viable, since it seems that (36) can be true

without A being red21. In such a situation there would be no property instance of red21ness

in A to be seen. If an object can phenomenally look red21 without being red21, then (36)

does not entail (39).

Another way of responding to the argument against the phenomenal looking

exportation principle is to endorse ostrich nominalism, and hold that properties do not

exist (Armstrong 1978 introduces the term ‘ostrich nominalism’). If properties do not

exist, then it would seem that (31) does not express the same proposition as (32), and that

(36) does not express the same proposition as (37). Ostrich nominalism would therefore

block the entailment from (36) to (39).

182
Hence, if ostrich nominalism is true, then we can defend the phenomenal looking

exportation principle, and the conclusions of this chapter stand.

In the section on modes of presentation above we discussed the property view and

the nonconceptual mode of presentation view:

The Property View: Phenomenal looking is a relation between two objects and a

property.

The Nonconceptual Mode

Of Presentation View: Phenomenal looking is a relation between two objects and a

nonconceptual mode of presentation of a property.

If ostrich nominalism is true, then neither the property view nor the

nonconceptual mode of presentation view is true. An objector may wonder how one

would argue that (18) entails (19) other than by defending the property view:

(18)A phenomenally looks to me to the left.

(19)A phenomenally looks to me to the left of me.

Our response to this objector is as follows. The phenomenal character principle,

introduced in chapter 1, entails that, if an object can phenomenally look to the left

183
without phenomenally looking to the left of me, then what it is like for me for an object

phenomenally to look to the left is different from what it is like for me for an object

phenomenally to look to the left of me. In the section above on modes of presentation, we

argued that there were certain constraints on this phenomenal difference, and we argued

that it does not seem that these constraints are met.

184
Chapter 3

Two More Arguments Against The View that Objects

Phenomenally Look To The Left and Right Of Me, And

Arguments Against Some Other Views About Phenomenal

Position Properties

Let us call the position properties that objects phenomenally look to have

phenomenal position properties. In the last chapter I argued that phenomenal position

properties are not observer-relative position properties. In section 1 of this chapter, I offer

some further arguments for this conclusion, and in section 2 I present some arguments

against some other conceptions of phenomenal position properties.

Many of my arguments against the observer-relative view have the following

form. I argue that the observer-relative view entails that whether objects are where they

phenomenally look to be depends on the wrong kinds of facts. That is, the observer-

relative view entails, for some reason R, that R is a reason for thinking that objects are

not where they phenomenally look to be, when, intuitively, R is not a reason for thinking

that objects are not where they phenomenally look to be.

185
In chapter 4 I defend the view that, necessarily, objects do not have the position

properties that they phenomenally look to have. However, there are good and bad reasons

for thinking that objects are not where they phenomenally look to be. Suppose that R1 is a

reason for thinking that objects are not where they phenomenally look to be, and R2 is not

a reason for thinking that objects are not where they phenomenally look to be. If a view V

entails that R2 is a reason for thinking that objects are not where they phenomenally look

to be, then one can reject V, even if one holds, because of R1, that objects are not where

they phenomenally look to be.

1 The Observer-Relative View Of Phenomenal Position Properties

1.1 The Problem of the Second Relatum

In the last chapter our working definition of the observer-relative view was this:

Observer-Relative View: Phenomenal position properties are relations to observers;

that is, objects phenomenally look to bear spatial relations

to observers.

For instance, a defender of the observer-relative view may say that a particular

object phenomenally looks to S to be to the left of S, or to be at some angle from S.

186
In this section I discuss how plausible it is that objects phenomenally look to bear

relations to the observer, as opposed to certain aspects of the observer, such as the

observer’s head or eyes. The question I discuss is what the most plausible value of x is in

the claim that objects phenomenally look to observers to bear a relation to x. I call this

the problem of the second relatum.

I raise a problem for the observer-relative view as defined above, on which the

second relatum is the observer, and I argue that the only way of avoiding this problem is

to amend the observer-relative view so that the second relatum is the observer’s pupils.

However, I argue that a different kind of problem arises for the amended view. Thus both

the original observer-relative view and the amended observer-relative view face

problems.

Suppose that, at t1, a cup is on my right, and the proponent of the observer-relative

view says that the cup phenomenally looks to the right of me. Figure 1 represents this

situation from a bird’s eye perspective. Arrows next to terms indicate the directions in

which the referents of those terms are pointing.

187
Figure 1

cup
t1

eyes
nose

head

body

Suppose that, at t2, I rotate my body to the right by forty degrees, so that my body

faces the cup, but I keep my head facing in the same direction as it was facing in at t1; my

head is no longer perpendicular to my body. This is illustrated in figure 2:

Figure 2

t2 cup

eyes
nose

head

body

Assuming that I do not see my body, it seems very plausible that there will be no

visual phenomenal difference between the way that the cup looks to me at t1 and the way

it looks to me at t2; after all, my head is pointing in the same direction at both times, and

188
the visible scene has not changed. There may well be a proprioceptive phenomenal

difference between t1 and t2, owing to the angle that my body bears to my head being

different at the two times, but intuitively this proprioceptive phenomenal difference is not

a visual phenomenal difference. So that we may focus on what purely visual phenomenal

differences there are, let us suppose that one’s proprioceptive sense has been numbed.

According to the phenomenal character principle developed in chapter 1, if there

is no visual phenomenal difference between the way the cup looks at t1 and t2, then if the

cup phenomenally looks to my right at t1, it phenomenally looks to my right at t2 as well.

Since the cup is straight ahead of me at t2, the proponent of the observer-relative view is

committed to saying that, at t2, I am having an illusion: the cup is not where it

phenomenally looks to me to be.

This seems an implausible consequence. It seems implausible that objects are

where they phenomenally look to be only when one’s head is perpendicular to one’s

body.

This consequence can be avoided if we amend the observer-relative view so that

the second relatum is the observer’s head, rather than the observer: so, according to the

amended observer-relative view, at t1, the cup phenomenally looks to the right of my

head. This solves the problem mentioned in the paragraph above. When, at t2, I rotate my

body to the right by forty degrees, but keep my head facing in the same direction as it

was facing in at t1, the cup remains to the right of my head. Thus, if we say that the cup

189
phenomenally looks to the right of my head at t1 and t2, we can say that at both t1 and t2 it

is where it phenomenally looks to be; so no illusion occurs.

However, the same problem arises in a different form for the observer-relative

view when we consider the possible movement of my eyes. Let us suppose that at t3, I

rotate my head to the right by forty degrees, so now it is facing in the same direction as

my body, but I keep my eyes facing in the same direction as they were facing in at t2 and

t1, so my eyes are facing in a different direction from my head. This is illustrated in figure

3.

Figure 3

cup
t3

eyes

head

body

The situation now is that the cup is straight ahead of my head, and it is to the right

of my eyes, since my eyes are facing in the same direction as they were at t1, and the cup

was to the right of my eyes at t1. If we assume that I do not see my nose, then it seems

that there will be no visual phenomenal difference between the way the cup looks at t2

and t3. In normal circumstances, there would be a proprioceptive phenomenal difference

190
between t2 and t3, but in the example we are supposing that my proprioceptive sense is

numbed.

Given that there is no visual phenomenal difference between the way the cup

looks at t2 and t3, if the cup phenomenally looks to the right of my head at t2, then it

phenomenally looks to the right of my head at t3. However, the cup is straight ahead of

my head at t3, and so it follows that I am having an illusion: the cup is not where it

phenomenally looks to be.

This seems an implausible consequence; it seems odd to say that objects are

where they phenomenally look to be only when one’s eyes are facing in the same

direction as one’s head.

One can avoid the objection by amending the observer-relative view so that the

instead of the observer’s head being the second relatum, the observer’s eyes are the

second and third relata. Therefore, in the example above, the cup phenomenally looks to

the right of my eyes, and since it is to the right of my eyes from t1 to t3, the cup is where it

phenomenally looks to be from t1 to t3.

However, we can now consider a fourth possible moving part: one’s pupils.

Although one’s pupils cannot in fact move independently of one’s eyes, it seems that they

could have done. Consider the following world, w1:

191
w1: A world which is identical in all respects to the actual world, except that,

in w1, our pupils can move independently of our eyes.

So, the events that occur in w1 between t1 to t3 are the same as the events that

occur between those times in the actual world.

Suppose that, at t4, I move my eyes forty degrees to the right so that the cup is

straight ahead of my eyes, but I keep my pupils facing in the same direction as they were

facing in from t1 to t3. This is illustrated in figure 4.

Figure 4

cup
t4 pupils

eyes

head

body

It seems that there need be no visual phenomenal difference between the way the

cup phenomenally looks at t3 and t4. Or course, our eyes could be so constructed so that

there is a visual phenomenal difference, but they need not be. If we say that the cup

phenomenally looks to the right of my eyes at t3, then it phenomenally looks to the right

192
of my eyes at t4. Since the cup is not to the right of my eyes at t4, it follows that the cup is

not where it phenomenally looks to be. Hence an illusion is occurring. However, it seems

odd to say that, in otherwise normal circumstances, objects are where they phenomenally

look to be only when one’s pupils are facing in the same direction as one’s eyes. It seems

just as odd to say this as it does to say that objects are where they phenomenally look to

be only when the direction of one’s head is perpendicular to one’s body.

To avoid this, we should amend the observer-relative view so that the second and

third relata are the observer’s pupils, rather than the observer’s eyes. Thus, a proponent of

the observer-relative view will say that the cup phenomenally looks to the right of my

pupils from t1 to t4. Since the cup is to the right of my pupils from t1 to t4, no illusion

occurs.

However, this formulation of the observer-relative view now faces a different

objection. Situation S1 is as follows:

S1: A normal situation in which we are looking at a cup that is straight ahead,

and our eyes are facing the cup.

Consider the world w2:

193
w2: A world in which, for objects phenomenally to look to us the way they do

in S1 in the actual world, our eyes need to be angled twenty degrees to the

right, rather than facing straight ahead, as they are in the actual world.

To make w2 vivid, we can assume that in w2 our pupils have photo-sensitive cells

on the sides of our pupils.

Suppose that, in the actual world, a cup is positioned straight ahead of me, such

that a proponent of the amended observer-relative view would say that the cup

phenomenally looks straight ahead of my pupils. In w2, the cup is positioned straight

ahead of me, and my pupils are angled at twenty degrees to the right. Given how our eyes

work in w2, the cup phenomenally looks the same to me in w2 as it does in the actual

world. That is, there is no visual phenomenal difference between the way the cup

phenomenally looks to me in w2, and the way it phenomenally looks to me in the actual

world.

It follows from the phenomenal character principle that if the cup phenomenally

looks straight ahead of my pupils in the actual world, it phenomenally looks straight

ahead of my pupils in w2. However, since the cup is not straight ahead of my pupils in w2,

but rather to the left of my pupils, given that my pupils are angled to the right in w2, then

the cup is not where it phenomenally looks to be in w2.

194
However, this seems an implausible consequence. If there is an illusion occurring

in one of the actual world and w2, there does not seem any reason why it should be

occurring in w2. The inhabitants of w2 have as much right to say that the illusion is

occurring in the actual world as we have right to say that the illusion is occurring in w2.

The observer-relative view is committed to the view that the inhabitants of one of these

two worlds are having illusions, while the inhabitants of the other world are not, and this

seems an implausible commitment.

The original formulation of the observer-relative view was that objects

phenomenally look to bear spatial relations to the observer, and the final formulation of

the observer-relative view was that objects phenomenally look to bear relations to the

observer’s pupils. The aim of this section has been to show that both formulations of the

observer-relative view face problems.

1.2 The Functionality of Front and Back and Top and Bottom

Intuitively, the right/left distinction is dependent on the front/back distinction and

the top/bottom distinction. If an object has neither a front nor a back, then it has neither a

right nor a left; and if an object has neither a top nor a bottom, then it has neither a right

nor a left. For instance, a perfectly homogeneous sphere has neither a front nor a back,

nor a top nor a bottom; consequently it has neither a right nor a left. Even if the sphere

195
had a front and a back, if it had neither a top nor a bottom, then it would have neither a

right nor a left.

1.2.1 Front and Back

It seems that the concept of the front of an object is a functional one; an object’s

having a front is a matter of one of the sides of the object’s having a certain functional

role. That an object has a front or a back is not intrinsic to that object. For instance, a

chair has a front and a back; and yet we can imagine a world in which intrinsic duplicates

of chairs do not have fronts and backs. For instance, in a world in which intrinsic

duplicates of chairs are plants, and these plants are not used in any particular way, then it

seems these duplicates of chairs have neither fronts nor backs.

By altering the functions of an object, it seems possible to turn an object that has a

front and a back into an object that does not have a front and a back. For instance, we are

inclined to say that a microwave has a front and a back; the front is the side with the door

and the dials on. However, if we put a television screen on the back of the microwave, so

that the same object can be used to watch television on and to cook with, then we would

be inclined to say that the object no longer has a front nor a back. There would be no

basis for deciding which side the front of the object was and which side the back was.

It seems that this process of changing an object so that it has neither a front nor a

back could be applied to a human being. The functionality of a human being happens to

196
be heavily dominated by one of its sides. Its eyes, mouth, nose are all on the same side of

its body, and its arms and legs naturally move in the direction that this side faces.

Suppose that these functions were more evenly distributed on two sides of the human

body. For instance, suppose that the mouth was on the actual back of the head, and a

human’s legs naturally worked in the direction that the actual back of the human body

faces. We may suppose that, in other respects, the human body is the same as it actually

is. This is illustrated in figure 5:

Figure 5

Eyes

Mouth nose

Direction of movement of arms

Direction of movement of legs

It seems that we would say that a being of the kind illustrated in figure 5 would

have neither a front nor a back; it is not obvious on what basis we would decide which

side is the front and which side is the back. Given that the right/left distinction is based

on the front/back distinction, it follows that a being of the above kind would have neither

a right nor a left. Let us call such a being a front-and-backless being.

197
Now, let us suppose that a subject, S, turns, between t1 and t2, from being a normal

human being into a front-and-backless being. Let us suppose that, all along, S is looking

at an apple, which, at t1, is to its right. According to the original formulation of the

observer-relative view, at t1, the apple phenomenally looks to S to be to S’s right. There is

no visual phenomenal difference between the ways that objects phenomenally look to S

between t1 and t2, so it follows, on the original observer-relative view, that, at t2, the apple

phenomenally looks to the right of S.

However, at t2, S has neither a front nor a back, and therefore neither a right nor a

left. It follows that the apple is not where it phenomenally looks to S to be, and so S is

having an illusion. This seems implausible. Consider the following world, w3:

w3: A world in which all beings are originally formed like S is at t2, and

therefore are front-and-backless.

Call the beings in w3 originally front-and-backless beings. Suppose that objects

phenomenally look the same to these beings as they do to S at t2. It seems implausible to

say that the beings in w3 have illusions about where objects phenomenally look to them to

be.

One way of responding to the above argument is to hold that terms such as ‘right’

and ‘left’ are rigid designators. According to this view, certain descriptions are used to fix

198
the reference of ‘right’ and ‘left’, but these descriptions are not part of the meaning of

‘right’ and ‘left’ (for a defence of this view, see Pooley 2002). Call this the rigid view.

Above we argued that, intuitively, the right/left distinction is dependent on the front/back

distinction. That is, an object without a front or back intuitively does not have a right and

left. If we are to keep this connection between the notions of front and back and the

notions of right and left, then it should be part of the rigid view that ‘front’ and ‘back’ are

rigid designators. Consider a world, w4, which is as follows:

w4: A world just like the actual world, except that our legs and arms naturally

move in the actual backwards direction, and all the features characteristic

of our actual front sides are on our actual back sides, and all the features

characteristic of our actual back sides are on our actual front sides.

In w4, according to the rigid view, we walk backwards. This seems a counter-

intuitive part of the rigid view. It seems more natural to say that, in w4, we walk forwards,

and the side that, in w4, the expression ‘my front’ picks out is different from the side that,

in the actual world, ‘my front’ picks out. Nevertheless, I shall consider how the rigid view

affects the above argument.

If the rigid view is correct, then if S has a right side and a left side at t1, then S

also has a right side and a left side at t2. However, the rigid view is consistent with the

possibility of originally front-and-backless beings, as they were described above, who

would have neither right nor left sides. After all, the rigid view allows that descriptions

199
are used to fix the references of terms such as ‘right’ and ‘left’. Thus, if such descriptions

do not pick out any sides, as, intuitively, they would not in the case of originally front-

and-backless beings, then terms such as ‘right’ and ‘left’ would not refer.

The argument above was that the observer-relative view has two implausible

consequences: the first is that it provides a reason for thinking that, at t2, S is having an

illusion; the second is that it provides a reason to think that originally front-and-backless

beings suffer illusions. If the rigid view is correct, then the observer-relative view does

not have the first consequence, but it does still have the second consequence.

In what follows I will be assuming that the rigid view is false. I will be appealing

to temporal as well as modal differences in the right and left sides of beings. The rigid

view is not consistent with either of these kinds of differences. However, there is a way of

amending the modal arguments below to render them consistent with the rigid view.

When I write that an object which actually has a front and a back might have been

originally front-and-backless, we can rephrase the envisaged possibility as being that

there might have been an originally front-and-backless being. Nothing in the arguments

below depends on the over-time or cross-world identity of the originally front-and-

backless being with some object which does have a front and a back.

A second way in which a proponent of the observer-relative view could respond to

the above argument is by holding that the apple phenomenally looks to S to the right of

S’s pupils at t1 and t2, and, at least in the above example, S’s pupils do, plausibly, have a

200
front and a back, and therefore a right and a left. However, we can apply the same

procedure to S’s pupils that we applied to S’s body: we can ensure that S’s eyes do not

have a front and a back by assigning an important function to the back of S’s pupils in

such a way that we have no basis for deciding which side of S’s pupils is the back, and

which side is the front.

One slightly far-fetched, but nonetheless possible, example is the following.

Suppose that S’s head was a ring with a hole in the middle, and in the hole was a round

object containing a pupil. Through one end of the pupil light comes in, enabling S to see;

through the other end, air comes in, enabling S to smell. We can assume that air can only

enter this end of the pupil owing to there being a lens at the other end of the pupil. Thus

the pupil has two functions, and different ends of the pupil are responsible for each of the

two functions. A vertical cross-section of S, viewed from the side, is illustrated in figure

6.

Figure 6

S’s head

S’s eye
Light into S’s pupil
Air into S’s pupil

S’s pupil

201
In figure 6, it does not seem that S’s pupil has a front and a back; after all, it is not

clear on what basis we could decide which end the front of the pupil was, and which end

the back of the pupil was. Let us specify world w5 as follows:

w5: A world in which S is as figure 6 illustrates her to be.

Let us suppose that, at t2 in w5, S is looking at an apple as she is in the actual

world at t2. There need be no visual phenomenal difference between the way the apple

phenomenally looks to S at t2 in the actual world, and the way the apple phenomenally

looks to S at t2 in w5. If the apple phenomenally looks to the right of S’s pupils at t2 in the

actual world, it follows from the phenomenal character principle that the apple

phenomenally looks to the right of S’s pupils at t2 in w5. Since S’s pupils at t2 in w5 have

neither a right nor a left, then S is having an illusion in w5. This seems to be an

unattractive consequence of the observer-relative view.

Thus, the functionality of the notions of front and back seems to pose a problem

for the observer-relative view. The crux of the problem is that one can change those

functions of a given human being that are relevant to determining which side is the front,

and which side is the back, without changing how things phenomenally look to that

human being. There is no necessary connection between these functions and the way

things phenomenally look to the human being. Thus cases arise in which the observer-

relative view is committed to saying that an object phenomenally looks to S to be to the

right of S (or a part of S) even when S (or the relevant part of S) has no right or left.

202
1.2.2 Top and Bottom

A similar problem arises when one considers the notions of top and bottom.

Consider the following world, w6:

w6: A world in which our normal posture is a handstand. In this world it is

normal for us to live standing on our hands, and to walk by moving our

arms, much as we walk by moving our legs in the actual world.

Because the handstand is the normal posture for us in w6, it seems that our tops

and bottoms in w6 are different from our tops and bottoms in the actual world. Whilst in

the actual world, the bottoms of our bodies are the ends with our legs on, and the tops of

our bodies are the ends with our heads on, in w6 the situation seems to be reversed. In w6

it seems that the bottoms of our bodies are the ends with our arms and heads on, and the

tops of our bodies are the ends with our legs on. Thus, our top in the actual world is our

bottom in w6, and our bottom in the actual world is our top in w6. As suggested above, if

one held the rigid view, then one should consider a world just like w6, except that the

beings in question are distinct from us.

Let us assume that the sides of me that are my front and back are the same in w6

and the actual world. Nonetheless, since the ends of me that are my top and bottom are

different in w6 and the actual world, the sides of me that are my right and left are different

203
in w6 and in the actual world. For instance, in the actual world my heart is on my left,

whereas in w6 my heart is on my right.

w6 creates a problem for the observer-relative view. Suppose that, at t1 in the

actual world, I am looking at a cup to my right, and, according to the observer-relative

view, the cup phenomenally looks to the right of me (or to the right of a part of me; it will

not matter which for the purposes of the argument). Then at t2 that I do a hand-stand,

bearing the same angle to the cup as I did at t1. The states of affairs at t1 and t2 in the

actual world are illustrated in figures 7 and 8:

Figure 7

cup
t1 in the actual world

eyes
nose

head

My left body My right

204
Figure 8

The following is a birds-eye view of the handstand.

cup
t2 in the actual world

Eyes, Upside down body


head

Soles of feet

My right My left

Body

Since my hand-stand, illustrated in figure 8, is only a temporary posture, and not

my normal one, it does not seem that, at t2, my legs have become the top of my body and

my head has become the bottom. Rather it seems that, when doing the handstand, my legs

remain at the bottom of my body, and my head remains at the top of my body. It is only

when a handstand posture is the norm for a population that there is pressure to say that

the legs are at the tops of the bodies in that population, and the heads are at the bottoms

of the bodies in that population. Given that the ends that are my top and bottom do not

change when I temporarily do a handstand, it does not seem that the sides that are my

right and my left change when I temporarily do a handstand. Intuitively, my heart remains

on the left-hand side of me even when I temporarily do a handstand.

Because I am upside down at t2, the proponent of the observer-relative view will

say that the cup phenomenally looks to my left at t2; and since it is to my left at t2, it is

205
where it phenomenally looks to be. However, suppose that I am in w6, that I am standing

on my head, being supported by my arms, looking at the cup, and that I bear the same

relation to it as I do in the actual world. In other words, the situation is exactly the same

as in figure 8, except there is one difference: my right and left have switched around. This

is because my top and bottom are different in w6 from my top and bottom in the actual

world, and so my right and left are different in w6 from my right and left in the actual

world. When I am looking at the cup at t2 in w6, the cup will be to the right of me,

whereas when I am looking at the cup at t2 in the actual world, the cup is to the left of me.

This is illustrated in figure 9. Figure 9 is identical to figure 8 except that the arrow

pointing to my left in figure 8 points to my right in figure 9, and the arrow that points to

my right in figure 8 points to my left in figure 9.

Figure 9

cup
t2 in w6

Eyes, Upside down body


head

Soles of feet

My left My right

Body

There need be no visual phenomenal difference between the way the cup looks to

me at t2 in the actual world, and the way it looks to me at t2 in w6. It follows from the

phenomenal character principle that if the cup phenomenally looks to the left of me at t2

206
in the actual world, it phenomenally looks to the left of me at t2 in w6. Since it is not to

the left of me at t2 in w6, the cup is not where it phenomenally looks to be. This seems an

implausible consequence, since I, in w6, seem to have as much reason to say that I, in the

actual world, am the one having an illusion, rather than vice versa.

One can consider a temporal version of this objection, where worlds are replaced

by times. At t2, when I am doing a handstand, the top of my body is touching the ground,

so the cup is to the left of me. Between t2 and t10, it becomes normal for me to assume the

handstand posture, and thus my top and bottom at t10 are different from my top and

bottom at t2. Thus, in this temporal version of the objection, figure 8 can be used to

illustrate the state of affairs at t2, and figure 9 can be used to illustrate the state of affairs

at t10. The rest of the argument is the same as above: if the cup phenomenally looks to me

to be to my left at t2, then it phenomenally looks to me to be to my left at t10. Since the

cup is not to my left at t10, the cup is not where it phenomenally looks to be at t10, which is

an unattractive consequence of the observer-relative view.

A proponent of the observer-relative view might object to this temporal argument

by pointing to psychological evidence which suggests that after a subject wears up/down

inverting lenses for a certain amount of time, perceived objects phenomenally look to the

subject to have the orientation that they phenomenally looked to the subject to have

before the subject starting wearing the inverting lenses; that is, that inverting lenses have

only a short-lived effect. If this was correct, then we cannot assume that there would be

no visual phenomenal difference between the way that objects phenomenally look at t2

207
and at t10. As it turns out, this psychological evidence is unclear (see O’Regan and Noe

2001). However, if it does turn out to the be the case that inverting lenses have only a

short-lived effect, then the temporal argument would have to be rejected. The modal

argument would be unaffected, however, as the psychological evidence in question only

concerns what happens in the actual world. The evidence does not suggest that a world

such as w6 is not possible, and all that is required for the modal argument above is that w6

is possible.

A proponent of the observer-relative view may reply to the modal argument as

follows. They may argue that, when I am upside-down at t2 in the actual world, as

illustrated in figure 8, the cup does not phenomenally look to my left, despite the fact that

it is to my left. They may argue that, in fact, the cup phenomenally looks to my right at t2

in the actual world, and therefore I am having an illusion. The motivation for this reply is

as follows: since there is no visual phenomenal difference between how things

phenomenally look at t2 in the actual world and at t2 in w6, i.e. the situations illustrated in

figures 8 and 9 respectively, if the proponent of the observer-relative view can say that

the cup phenomenally looks to my right at t2 in the actual world, then he can say that the

cup phenomenally looks to my right at t2 in w6. The background assumption to the reply

is that, given that the observer-relative view is committed to there being an illusion in

either the actual world or w6 at t2, it is better that the illusion occur in the actual world,

since at least I am assuming an abnormal posture at t2 in the actual world, and this latter

fact can be used to reduce the implausibility of holding that an illusion is occurring.

208
However, this reply is not likely to work, since there is pressure on the proponent

of the observer-relative view to say that, at t2 in the actual world, in the situation

illustrated in figure 8, the cup does phenomenally look to my left. This is because there is

no visual phenomenal difference between the way the cup phenomenally looks to me at t2

in the actual world and the way the cup phenomenally looks to me in a situation which is

as good a candidate as any to be described as one in which the cup phenomenally looks to

my left. This situation will be described below. The situation described below is

envisaged as occurring in the actual world.

At t1, I am looking at the cup, which is to my right. The situation is illustrated in

figure 7. Suppose that t3 is just like t1, except that the scene that I am viewing has been

rotated by 180 degrees around a central point. The rotation affects only the objects I see; I

stay upright at my normal orientation. The situation at t3 is illustrated in figure 10.

Figure 10

t3

Cup (part of the visible scene which has been rotated around a central point)

eyes
nose

head

My left body My right

209
Recall the descriptions of t1, t2 and t3 in the actual world:

t1 in the actual world: A normal situation in which I am looking at a cup,

which is to my right, illustrated in figure 7.

t2 in the actual world: The same situation as at t1, except that I am doing a

handstand, illustrated in figure 8.

t3 in the actual world: The same situation as at t1, except that the visible

scene has been rotated by 180 degrees around a

central point, illustrated in figure 10. I have the

posture I have at t1, and am thus not doing a

handstand.

It will be easy to find the right point around which the scene is rotated by 180

degrees such that there is no visual phenomenal difference between the way the cup

phenomenally looks to me at t3, and the way the cup phenomenally looks to me at t2. At

both t2 and t3 it would be natural to describe things as looking upside down. Admittedly,

there will be a proprioceptive phenomenal difference between the situation at t3 and the

situation at t2, but as we have claimed before, this does not seem to be a visual

phenomenal difference.

210
It seems that the proponent of the observer-relative view will want to say that, at

t3, the cup phenomenally looks to my left. Given that there is no visual phenomenal

difference between the way the cup phenomenally looks at t3 and the way it

phenomenally looks at t2, the proponent of the observer-relative view is committed to

saying that the cup phenomenally looks to my left at t2 in the actual world.

As with the objection concerning the notions of front and back, the crux of the

objection concerning the notions of top and bottom is that one can change the factors that

determine which ends of a being are its top and bottom, and therefore which sides of the

being are its left and right, without changing how things phenomenally look to that being.

Thus cases arise in which the observer-relative view is committed to saying of two

apparently symmetrical cases that one involves an illusion and the other does not, and

this is an implausible consequence of the observer-relative view.

1.3 Visual Angles

In this section I shall discuss another observer-relative view, the visual angle

view:

Visual Angle View: Phenomenal position properties are angles from the observer.

211
We often say that one line looks at a certain angle from another line. For instance,

in figure 11 below, we may say that line A looks at 90 degrees from line B. For the sake

of argument, I will assume that the looking involved here is phenomenal looking.

Figure 11

A B

According to the visual angle view, a given object, say a cup, may phenomenally

look to be at 150 degrees from the subject, S. However, there is a question about what it

means to say that the cup phenomenally looks at a certain angle from S. Similarly, there

is a question about what it means to say that the cup is at a certain angle from S. Lines

and planes bear angles to each other, not items such as the cup and S.

When a proponent of the visual angle view says that a given cup phenomenally

looks to be at 150 degrees from the subject, one interpretation of this claim is that the line

connecting the cup with the bridge of S’s nose phenomenally looks at a certain angle

from the line that originates at the bridge of S’s nose and goes off straight to S’s left; i.e. a

line that is tangential to the bridge of S’s nose. Let us call the bridge of S’s nose B, and let

us call this interpretation of the visual angle view the pretheoretic interpretation:

212
The Pretheoretic Interpretation: ‘O phenomenally looks to be at 150 degrees from S’

is true iff the line connecting O and B

phenomenally looks to S to be at 150 degrees from

the line that originates at B and goes off straight to

S’s left.

The pretheoretic interpretation of the visual angle view is illustrated below in

figure 12. Line L2 is the line connecting the bridge of S’s nose with the cup, and line L1 is

the line originating at the bridge of S’s nose and going off straight to the left.

Figure 12

cup
L2
150˚
L1

head
eyes

My left body My right


Bridge of nose (B)

In this section I will argue that the pretheoretic interpretation of the visual angle

view faces a difficulty. There is an alternative interpretation of the visual angle view that

213
avoids the difficulty, but I will argue that this alternative interpretation faces a separate

problem.

In figure 11 above, what angle A phenomenally looks to bear to B depends on

what perspective one is viewing the lines from. For instance, consider the point where

line A meets line B; call this point X. Consider the arc that starts 10cm above X and

finishes 10cm below it. 0 degrees on the arc is the beginning of the arc 10cm above X.

One’s perspective on the angle between A and B can be from different angles on this arc.

If one’s perspective on the angle is from 90 degrees on the arc, i.e. directly above X, then

A phenomenally looks to bear an angle of 90 degrees to B. If one’s perspective on the

angle is from 150 degrees on the arc, then A phenomenally looks to bear a much wider

angle to B. If one’s perspective on the angle is from 180 degrees on the arc, then A

phenomenally looks to bear an angle of 180 degrees to B. That is, A and B phenomenally

look to be a straight line. Another way of putting this is that if S is looking at lines A and

B from the side, rather than from above, then they will phenomenally look to bear an

angle of 180 degrees to each other.

According to the pretheoretic interpretation of the visual angle view, L2

phenomenally looks to S to be at 150 degrees from L1. However, there is a problem with

this claim. The problem is that S’s perspective on lines L1 and L2 is from the side, rather

than from above. As we saw in the case of figure 11, if one’s perspective on two lines is

from the side, then they will phenomenally look to be at 180 degrees from each other.

Thus L1 and L2 phenomenally look to be at 180 degrees to each other. It is only if S’s

214
perspective on L1 and L2 was from directly above that L1 and L2 would phenomenally

look to be at 150 degrees from each other.

To make this vivid, we can suppose that L1 and L2 are black rods. If S is viewing

these rods from the side, then they will phenomenally look to bear 180 degrees to each

other, whatever angle they in fact bear to each other.

There is an alternative interpretation of the visual angle view which avoids the

above problem. Let us call this the propositional property interpretation of the visual

angle view.

The Propositional Property

Interpretation: ‘O phenomenally looks to be at 150 degrees from S’

is true iff there exists a property of being such that the

line that connects O and B is 150 degrees from the

line that originates at B and goes off straight to S’s

left, and O phenomenally looks to have this property.

In chapter 1 we drew the distinction between propositional properties and non-

propositional properties. A propositional property is, for some p, the property of being

such that p is true. A non-propositional property is a property that is not a propositional

property.

215
According to the pretheoretic interpretation of the visual angle view, objects

phenomenally look to have non-propositional properties. According to the propositional

property interpretation of the visual angle view, objects phenomenally look to have

propositional properties, such as the property of being such that the line that connects O

and B is 150 degrees from the line that originates at B and goes off straight to S’s left.

In chapter 1 we argued that phenomenal looking is a relation between two objects

and a non-propositional property. The argument was as follows. I will first give the

argument using colour properties, and then give the argument using position properties.

Suppose that phenomenal looking is a relation between two objects and a propositional

property. Suppose that the conditions are normal, that A is red, and that B is green, and

that A phenomenally looks such that A is red, and that B phenomenally looks such B is

green. It seems intuitive that A is not the way that B phenomenally looks in respect of

colour. However, if ‘the way that B phenomenally looks in respect of colour’ refers to the

property of being such that B is green, then A is the way that B phenomenally looks in

respect of colour, since A is such that B is green. If, on the other hand, ‘the way that B

phenomenally looks in respect of colour’ refers to the property of being green, then it

does follow, as intuitively it ought to, that A is not the way B phenomenally looks in

respect of colour.

Suppose that the propositional property interpretation of the visual angle view is

correct. Assuming that B is the bridge of my nose, suppose that object O1 phenomenally

looks such that the line connecting O1 and B is 150 degrees from the line that originates

216
at B and goes off to my left. And suppose that O2 phenomenally looks such that the line

connecting O2 and B is 100 degrees from the line that originates at B and goes off to my

left. This situation is illustrated in figure 13.

Figure 13

150˚ O2
100˚

O1

head
eyes

My left body My right


Bridge of nose (B)

O1 and O2 phenomenally look to have different positions. Suppose that O1 is the

way it phenomenally looks in respect of position, and that O2 is the way it phenomenally

looks in respect of position. It seems we want to say that O1 is not the way that O2

phenomenally looks in respect of position. However, if ‘the way that O2 phenomenally

looks in respect of position’ refers to the property of being such that the line connecting

O2 and B is 100 degrees from the line that originates at B and goes off to my left, then O1

will be the way that O2 phenomenally looks in respect of position. After all, O1 is such

that the line connecting O2 and B is 100 degrees from the line that originates at B and

goes off to my left.

217
Thus, if the argument that we developed in chapter 1 against the view that

phenomenal looking is a relation between two objects and a propositional property is

sound, then we should reject the propositional interpretation of the visual angle view.

2 Other Views About Phenomenal Position Properties

All of the views considered so far have been varieties of the observer-relative

view. They have thus all been inconsistent with the phenomenal looking exportation

principle developed in chapter 2, which entails that if x phenomenally looks to S to bear

R to S, then S sees S, and if x phenomenally looks such that y bears R to S, then S sees S.

In this section I discuss some non-observer-relative views about the nature of

phenomenal position properties. With the exception of the Leibnizian relationalist view,

the views below are consistent with the phenomenal looking exportation principle.

2.1 Field-Of-View Relationalism

According to the field-of-view relationalism, the position properties that objects

phenomenally look to have are relations between the perceived objects. For instance,

suppose that one is looking at the following sequence of letters:

A B C

218
The field-of-view relationalist will say that the position properties that A, B and C

phenomenally look to have are spatial relations between them, for instance, the relations

of B being halfway between A and C, A being twice as far from C as it is from B, C being

twice as far from A as it is from B, and so on. The relations in question are restricted to

non-observer-relative ones, since field-of-view relationalism is designed to be consistent

with the phenomenal looking exportation principle. Thus, the relations do not include

relations such as being to the left of, since A’s being to the left of B is a matter of A’s

being further to the left of the perceiving subject than B.

One immediate problem faces field-of-view relationalism. Suppose that one is

simply looking at one object alone, say a kite against a blue sky. Intuitively, if the kite

moves enough between t1 and t2, the position that the kite phenomenally looks to be in at

t2 will be different from the position that the kite phenomenally looks to be in at t1. This

situation is illustrated in figure 14.

Figure 14

The kite at t1 The kite at t2

219
It is not obvious how field-of-view relationalism accommodates this possibility.

According to field-of-view relationalism, the position that an object phenomenally looks

to have may change only if the relations that the object phenomenally looks to bear to

other seen objects changes. However, in the case of the kite, the kite is the only object

one sees, and therefore there are no objects to which the kite may phenomenally look to

bear different relations at t1 and t2.

It seems that the response which the field-of-view relationalist must make here is

that the kite is not the only object that one sees. The field-of-view relationalist must say

that one also sees parts of the sky. If one does see parts of the sky, then there will be a

particular part of the sky to which the kite phenomenally looks closer at t2 than at t1, and

therefore the field-of-view relationalist can account for the fact that the kite

phenomenally looks to be in different positions at t1 and t2.

A more pressing problem facing field-of-view relationalism is that it cannot

accommodate the possibility that if a set of objects are inverted around some axis, then

they will come phenomenally to look in different positions. Suppose that, at t1, a subject

S1 is looking at a red rectangle to the left of a green rectangle, which together take up S1’s

entire field of view. This situation is illustrated in figure 15.

220
Figure 15

t1

Red Green
rectangle rectangle

S1

In describing where the two rectangles phenomenally look to S1 to be, the field-

of-view relationalist cannot say that the red rectangle phenomenally looks to the left of

the green rectangle, since the relation of x being to the left of y is an observer-relative

relation. All that the field of view relationalist can say is that the red rectangle

phenomenally looks next to the green rectangle.

Suppose that the red and the green rectangles swap positions between t1 and t2.

The situation at t2 is illustrated in figure 16.

221
Figure 16

t2

Green Red
rectangle rectangle

S1

According to the field-of-view relationalist, the positions that the rectangles

phenomenally look to have between t1 and t2 will not change, since at t2 the red rectangle

still phenomenally looks next to the green rectangle. The relations that the rectangles

phenomenally look to bear to each other are the same before and after they swap

positions. However it seems clear that the red and green rectangles phenomenally look in

different positions at t1 and t2.

The objection also applies when one considers the subject changing position.

Suppose that, at t1, S1 is looking at the red and green rectangles before her. At t2, S1 walks

round the rectangles and looks at them from the opposite perspective. This situation is

illustrated in figure 17.

222
Figure 17

S1 at t2

Red Green
rectangle rectangle

S1 at t1

At t2, the rectangles clearly phenomenally look in different positions to S1.

However, according to the field-of-view relationalist, the rectangles phenomenally look

in the same position at t1 and t2, since for every relation R that the rectangles

phenomenally look to bear to each other at t1, they phenomenally look to bear R to each

other at t2. This example shows that it is possible for the position that an object

phenomenally looks to change without the relations that the objects phenomenally looks

to bear to other objects changing, and therefore that field-of-view relationalism is false.

2.2 Leibnizian Relational Properties

223
One might try to avoid the problems that face field-of-view relationalism by

holding that phenomenal position properties are Leibnizian relational properties, i.e.

relations to every other object in the universe. When the red and green rectangle swap

positions, although no relation between them changes, relations between them and other

objects in the universe change. Thus, if objects phenomenally look to have Leibnizian

relational properties, then it is possible to explain how the rectangles phenomenally look

to be in different positions after they are swapped.

Similarly, when S1 adopts the opposite perspective on the two rectangles, although

the relations between the two rectangles do not change, their relations to other objects in

the universe, for instance S1, change, and so the Leibnizian relationalist can account for

the fact that the rectangles phenomenally look to be in different positions after S1 adopts

the opposite perspective on them.

However, the Leibnizian relational view is inconsistent with the phenomenal

looking exportation principle, since one does not see every object in the universe, and a

consequence of the phenomenal looking exportation principle is that x phenomenally

looks to S to bear R to y only if S sees x an y.

The Leibnizian relationalist also faces an objection similar to the one facing the

field-of-view relationalist. Instead of considering a single subject adopting the opposite

perspective on two rectangles, let us suppose that there are two subjects, S1 and S2, sitting

224
opposite each other, with the red and green rectangle between them. This situation is

illustrated in figure 18.

Figure 18

S2

Red Green
rectangle rectangle

S1

The rectangles clearly phenomenally look to have different positions to S1 and to

S2, but the Leibnizian relationalist does not have the resources to account for this.

Suppose that the red rectangle stands in the set of spatial relations R1 to the other objects

in the universe, and suppose that the green rectangle stands in the set of spatial relations

R2 to the other objects in the universe. According to the Leibnizian relationalist, the red

rectangle phenomenally looks to have R1 to S1 and to S2, and the green rectangle

phenomenally looks to have R2 to S1 and to S2. Thus the Leibnizian relationalist cannot

225
accommodate the fact that the red and green rectangles phenomenally look to have

different positions to S1 and S2.

Let us call the above problem, as illustrated in figure 18, the perspective problem.

The Perspective Problem: The problem of accounting for the fact that, in normal

conditions, if there are four objects, w, x, y and z, such that:

(i) w and x are side by side

(ii) y and z are looking at w and x

(iii) y and z have opposite perspectives on w and x

then:

(iv) the position that w phenomenally looks to y is different from the

position that w phenomenally looks to z.

(v) the position that x phenomenally looks to y is different from the

position that x phenomenally looks to z.

2.3 Substantivalism

One might take the view that a given object phenomenally looks to be in that

position, where that position picks out some absolute, substantivalist location. Thus,

when S1 is looking at the two rectangles, and the rectangles swap positions between t1 and

t2, the substantivalist can explain the fact that they come phenomenally to look in

226
different positions. Before they swap position, the red rectangle phenomenally looks to be

in that position1, and after they swap position, the red rectangle phenomenally looks to be

in that position2, and that position1 is different from that position2.

However, the substantivalist view does not have a solution to the perspective

problem. If the red rectangle, as illustrated in figure 18, is in that position1, and the green

rectangle is in that position2, then the red rectangle will phenomenally look to both S1 and

S2 to have that position1, and the green rectangle will phenomenally look to both S1 and

S2 to have that position2. Substantivalism entails that the positions that the two rectangles

phenomenally look to S1 to have are the same as the positions that the two rectangles

phenomenally look to S2 to have, and intuitively this is not the case.

Substantivalism may deny the intuition that the perspective problem exists. That

is, the view may deny the intuition that the positions that the red and green rectangle

phenomenally look to S1 to have are different from the positions that the red and green

rectangle phenomenally look to S2 to have. However, this denial is inconsistent with the

phenomenal character principle. Recall figures 15 and 16.

227
Figure 15

t1

Red Green
rectangle rectangle

S1

Figure 16

t2

Green Red
rectangle rectangle

S1

Figures 15 and 16 illustrate the red and green rectangles swapping position

between t1 and t2. Suppose that, between t2 and t3, the red and green rectangles swap

positions again, thereby returning to the positions that they had at t1, and S1 walks around

228
the two rectangles, coming to view them from the opposite perspective. This is illustrated

in figure 19.

Figure 19

S1 at t3

Red Green
rectangle rectangle

It seems that there will be no visual phenomenal difference between the way that

the rectangles phenomenally look to S1 at t2 and at t3. It follows from the phenomenal

character principle that the rectangles phenomenally look to S1 to have the same positions

at t2 and t3. We established already that the rectangles phenomenally look to have

different positions at t1 and t2. It follows that the rectangles phenomenally look to have

different positions at t1 and t3. This establishes that the perspective problem exists.

One of the considerable advantages of observer-relative views is that they have a

solution to the perspective problem: when S1 adopts the opposite perspective on the two

rectangles between t2 and t3, a proponent of the observer-relative view can say that, at t2,

229
the green rectangle phenomenally looks to the left of the red rectangle, and at t3, the green

rectangle phenomenally looks to the right of the red rectangle.

However, as we have seen in the last chapter and in this one, there are reasons to

reject the observer-relative view. In the next chapter, I shall discuss the perspective

problem further and offer an account of phenomenal position properties which provides a

solution to the perspective problem.

230
Chapter 4

Primitivism About Phenomenal Position Properties

In the last two chapters we have considered a number of views about the nature of

the position properties that objects phenomenally look to have: the observer-relative

view, the field-of-view relationalist view, the Leibnizian relationalist view, and the

substantivalist view. All of these views seem to face serious objections. In this chapter I

shall sketch a positive view about the nature of phenomenal position properties, which I

call primitivism.

There is a view known as primitivism within the philosophy of colour, and the

primitivism that I defend is similar in some ways to this view. I will refer to primitivism

within the philosophy of colour as colour primitivism (see Campbell 1997, Chalmers

2006). According to colour primitivism, colour properties are sui generis, primitive

properties; they are not dispositions to bring about certain states of affairs, nor are they

molecular grounds of those dispositions, nor are they non-dispositional relations to the

observer. Some hold that colour primitivism entails colour scepticism, the view that

objects do not have colour properties (e.g. Chalmers 2006).

231
According to primitivism about phenomenal position properties, phenomenal

position properties are absolute, sui generis position properties. These properties are not

relations to the observer, nor are they relations to other seen or unseen objects. I will

argue also that primitivism entails position scepticism, the view that objects do not have

the position properties that they phenomenally look to have. I am presupposing a

conception of properties according to which properties can exist uninstantiated.

1 The Perspective Problem

Both primitivism and substantivalism, a view about phenomenal position

properties which we considered in the last chapter, hold that phenomenal position

properties are absolute, non-relational, properties. However, primitivism does not have

the same commitments as substantivalism. According to substantivalism, a given object

may phenomenally look to have that position, where that position picks out an absolute

position in space. Thus, it is part of the substantivalist view that phenomenal position

properties are the kinds of position properties that objects have. The difference between

primitivism and substantivalism is that primitivism does not have this commitment; the

view itself is silent on whether objects do or do not have the position properties that they

phenomenally look to have.

However, I shall argue that, given primitivism, the best solution to the perspective

problem discussed at the end of the previous chapter involves denying that objects have

232
the position properties that they phenomenally look to have. The perspective problem is

as follows:

The Perspective Problem: The problem of accounting for the fact that, in normal

conditions, if there are four objects, w, x, y and z, such that:

(i) w and x are side by side

(ii) y and z are looking at w and x

(iii) y and z have opposite perspectives on w and x

then:

(iv) the position that w phenomenally looks to y is different from the

position that w phenomenally looks to z.

(v) the position that x phenomenally looks to y is different from the

position that x phenomenally looks to z.

Let us consider an instance of the perspective problem: S1 and S2 are sitting

opposite each other, looking at a red rectangle and a green rectangle side-by-side on a

table before them. This situation is illustrated in figure 1.

233
Figure 1

S2

Red Green
rectangle rectangle

S1

According to the primitivist, the two rectangles phenomenally look to have

different primitive positions to S1 and to S2. Now we must ask: which of S1 and S2 is

right? Do the rectangles have the position properties that they phenomenally look to S1 to

have or the position properties that they phenomenally look to S2 to have? The options

seem to be: both of S1 and S2 are right; one and only one of them is right; neither of them

is right.

Let us call the view that one and only one of them is right the unique position

view. The problem with the unique position view is that it seems arbitrary to pick out

either S1 or S2 as the one that is right. Indeed, the rectangles will phenomenally look to

234
have different primitive position properties for infinitely many different perspectives that

may be had on the rectangles, and the unique position view holds that the rectangles in

fact have just one of these position properties. The arbitrariness of picking out two

particular position properties as the ones that the rectangles actually have makes this

position unattractive.

Let us call the view that both of S1 and S2 are right the many positions view. Since

there are infinitely many different perspectives on the rectangles that other subjects may

adopt, each of which involves the rectangles phenomenally looking at different positions,

the many positions view is committed to holding that every object has infinitely many

distinct position properties at once. This commitment makes the many positions view

curious. After all, an object cannot have more than one physical position of the same size

at once. It is highly counter-intuitive that there is any kind of position which is such that

an object can have many equally sized positions of that kind at once. Therefore the many

positions view is unattractive.

Let us call the view that neither of S1 and S2 is right the no positions view. On the

no positions view, objects never have the position properties that they phenomenally look

to have. The primitive position properties that objects phenomenally look to have are not

the kinds of properties that objects have. While this view is far from intuitive, I shall

argue that it is the most intuitive out of the three options.

235
The above style of argument is very similar to one in the philosophy of colour. In

the philosophy of colour, the case that generates the problem is the colour inversion

problem, which I define below. The colour inversion problem and the perspective

problem raise the same kind of difficulty for colour properties and phenomenal position

properties respectively. The colour inversion problem has been discussed a great deal in

the literature, and not all of the responses to it have involved accepting the analogue of

the no positions view. Some responses have involved accepting analogues of one of the

other views, and some have challenged the possibility of colour inversion. Below I shall

consider some of these responses to the colour inversion problem to see what light they

may shed on the perspective problem.

2 The Colour Inversion Problem

The colour inversion problem is as follows:

The Colour Inversion Problem: The problem of accounting for the fact that, for two

halves of the population, H1 and H2, it is possible

that tomatoes phenomenally look red to H1 and

green to H2.

Suppose that, for two halves of the population, H1 and H2, tomatoes

phenomenally look red to H1 and green to H2. Which half is right? That is, are tomatoes

the colour they phenomenally look to H1 or the colour they phenomenally look to H2?

236
The options are: both halves are right; one and only one half is right; and neither half is

right.

Let us call the view that one and only one half of the population is right the

unique colour view. The problem with the unique colour view is that it seems arbitrary to

pick out one of them rather than the other. Since tomatoes could phenomenally look a

different colour to every member of the population, the unique colour view is committed

to saying that, in such a scenario, there would be one member of the population to whom

the tomato phenomenally looks the colour that it has. The arbitrariness of picking out

which member of the population this is makes this view unattractive.

Let us call the view that both halves are right the many colours view. Since

tomatoes could phenomenally look to have a different colour to every member of the

population, the many colours view is committed to saying that objects have a possibly

infinite number of colours at once. This view is highly counter-intuitive. Intuitively,

colours such as red and green are incompatible, in the sense that nothing can be red and

green all over at once.

Let us call the view that neither half is right the no colours view. According to the

no colours view, objects do not have the colours that they phenomenally look to have.

Colour properties are not the kind of properties that objects have.

2.1 No Problem View

237
According to the no problem view, there is in fact no colour inversion problem.

Different defences may be given of the no problem view. According to the externalist

defence, tomatoes look the same colour to both halves of the population because the

experiences of both halves of the population are normally caused by the same surface

reflectance property of tomatoes (see Block 1996).

However, in chapter 1 we distinguished between phenomenal looking and

externalist looking, and the description of the colour inversion problem is that the

tomatoes phenomenally look different colours, not that they externalist-look different

colours. Thus the externalist argument does not support the no problem view.

However, an externalist may challenge our description of the colour inversion

problem by denying that phenomenal looking exists. Such an externalist will grant the

notion of there being some objects x and y, and some property F, such that there is

something it is like for y for x to externalist-look F to y, but they will deny that this

notion corresponds to a particular kind of looking, phenomenal looking. This externalist’s

description of what we have called ‘the colour inversion problem’ is that the tomatoes

look the same colour to both halves of the population, but what it is like for one half for

tomatoes to look that colour is different from what it is like for the other half for tomatoes

to look that colour. According to this view, there is not a kind of looking, phenomenally

looking, which is such that tomatoes phenomenally look different colours to different

halves of the population.

238
It seems that the pretheoretic, natural description of the colour inversion problem

involves the claim that tomatoes look different colours to different halves of the

population. Pretheoretically, this seems the default position to hold. If this description is

to be replaced, then there must be some theoretical argument for why: perhaps there is a

constraint on the notion of looking that phenomenal looking fails to satisfy. But it is not

clear what this constraint might be. Certainly this is a difficult and foundational issue.

However, it seems that there is an intuitive case for phenomenally looking, and without a

theoretical argument to the contrary, we should let this case stand.

Michael Thau has proposed an argument which, on the surface at least, seems to

be another defence of the no problem view. Thau has argued that objects do not

phenomenally look to have colour properties, and a fortiori, tomatoes do not

phenomenally look different colours to the two halves of the population. (Thau 2002).

This may seem to be a defence of the no problem view. However, Thau does

accept that tomatoes phenomenally look to have different properties to the two halves of

the population. His view is that these properties are not colour properties, but some other

kind of property. For our purposes, it does not matter whether these different ways that

tomatoes phenomenally look to the two halves of the population are colour properties or

not. We can introduce new terms, red* and green*, which refer to the different properties

that tomatoes phenomenally look to the two halves of the population, and then we can

239
express the colour inversion problem in terms of whether tomatoes are red* or green*.

Thus Thau’s argument is not a genuine defence of the no problem view.

2.2 The Many Colours View

Earlier our objection to the many colours view was that it seems that the colour

properties that objects phenomenally look to have to different halves of the population are

incompatible. That is, it seems that nothing can be red and green all over at once. The

current line of response holds the colour properties in question are observer-relative,

which explains why a single object can have two colour properties at once (see

Shoemaker 1994). Thus the intuition that nothing can be red and green all over at once is

an illusion.

On one way of formulating this observer-relative view, a colour property is the

property of normally causing a certain kind of experience in a particular subject. Thus,

when an object x phenomenally looks red to S, x phenomenally looks to S to have the

property of normally causing a certain kind of experience in S. This observer-relative

view of colour properties will conflict with the phenomenal looking exportation principle,

developed in chapter 2. A consequence of the phenomenal looking exportation principle

is that an object x phenomenally looks to S to bear R to S only if S sees S. If our

argument for the phenomenal looking exportation principle is sound, then since an object

240
can phenomenally look to have a colour property to a subject without the subject seeing

herself, we should reject the observer-relative view of colour properties.

According to Brad Thompson, the experiences of the two halves of the population

represent the tomato as having the same property, but they represent this property in

different ways, or under different modes of presentation (Thompson 2005). It is the

difference in the way the property is represented that explains the phenomenal difference

between the experiences of the two halves of the population. Let us call this the two-

dimensionalist view. According to the two-dimensionalist view, the content of a given

experience has a primary and a secondary intension, and the property represented by the

experience is part of the secondary intension of the content of the experience, and the

way that property is represented is part of the primary intension of the content of the

experience.

Thompson introduces modes of presentation as follows:

‘Modes of presentation can be thought of as corresponding to


primary intensions, functions from centred worlds considered as actual to
properties. A centred world is a possible way the world could be, with a
marked centre indicating an individual and a time. The same function,
given a different centred world, can return a different property.’
(Thompson, p25, 2006).

One way of thinking of primary intensions is as the semantic values of

descriptions. For instance, we might take the primary intension of ‘water’ to be the

semantic value of the description ‘the watery stuff around here’. In different worlds this

241
description picks out different kinds of stuff. The secondary intension of ‘water’ in a

given world w is the stuff in w that is picked out by the description ‘the watery stuff

around here’.

On this construal, primary intensions are properties. For instance, the primary

intension of ‘water’ might be the property of being the watery stuff around here. On this

construal, the claim that the content of visual experience contains a primary intension and

a secondary intension is equivalent to the claim that visual experience represents two

kinds of properties. Thompson appeals to primary intensions to solve the colour inversion

problem. The relevant primary intensions are properties such as the property of normally

causing a red-type experience in me, and the property of normally causing a green-type

experience in me. Suppose that there are two subjects, S1 and S2 and, as we might say,

tomatoes look red to S1 and green to S2. According to Thompson, the visual experiences

of S1 and S2 represent tomatoes as having the same colour property, say some surface

reflectance property, but the content of S1’s visual experience contains the primary

intension of normally causing a red-type experience in S1, and the content of S1’s visual

experience contains the primary intension of normally causing a green-type experience in

S2.

If primary intensions are properties, then we can express the two-dimensionalist

solution to the colour inversion problem as follows. For some surface reflectance

property F, both S1 and S2 visually represent tomatoes as F. But S1 visually represents

242
tomatoes as causing red-type visual experiences in her, and S2 visually represents

tomatoes as causing green-type visual experiences in him.

In the terminology of phenomenal looking, the two-dimensionalist claim is that a

given tomato stands in the phenomenal looking relation to S1 and both to the property of

being F, and to the property of causing red-type visual experiences in S1. Therefore, the

two-dimensionalist’s account of objects phenomenal looking to have colour properties

involves the claim that objects phenomenally look to have observer-relative properties. If

our argument for the phenomenal looking exportation principle is sound, then given that

objects can phenomenally look to have colour properties to subjects without those

subjects seeing themselves, we should reject the two-dimensionalist view.

Thompson introduced modes of presentation as primary intensions, and it is

natural to think of these primary intensions as properties. However, one might wonder

whether an object x can phenomenally look to have some property F under a mode of

presentation G, where G is not a property that x phenomenally looks to have. As we saw

in chapter 2, Christopher Peacocke has defended the view that the content of visual

experience includes both properties and ways in which those properties are perceived.

‘We shall not do justice to the fine-grained phenomenology of


experience if we restrict ourselves to those contents which can be built up
by referring to the properties and relations that the perceived objects are
represented by the experience as possessing. We must, in describing the
fine-grained phenomenology, make use of the notion of the way in which
some property or relation is given in experience. The same shape can be
perceived in two different ways, and the same holds for the shape
properties, if we regard them as within the representational content of

243
experience. Ernst Mach’s example of one and the same shape that can be
perceived either as a square or as a regular diamond is a familiar example.’
(Peacocke, 2001, p240-241).

However, in elaborating on this example, Peacocke seems to characterize these

ways in which properties are given in experience in terms of properties which one

perceives the object to have:

‘Intuitively, the difference between perceiving something as a


square and perceiving it as a (regular) diamond is in part a difference in
which symmetries are perceived. When something is perceived as a
diamond, the perceived symmetry is a symmetry about the bisector of its
angles. When something is perceived as a square, the perceived symmetry
is a symmetry about the bisector of its sides.’ (Peacocke, 1992, p76).

In the above paragraph, Peacocke claims that which symmetries are perceived

determines whether one sees a shape as a square or a diamond. Thus this example does

not illustrate how objects could phenomenally look to have properties under modes of

presentation, where these modes of presentation are not additional properties that the

objects phenomenally look to have.

2.3 The No Colours View

In ‘Perception and the Fall from Eden’ David Chalmers argues that experiences

have three types of content: Edenic content, Fregean content and Russellian content

(Chalmers, 2006). The Russellian content consists of the properties that objects

externalist-look to have: this content is fixed by what certain kinds of experiences are

244
normally caused by. The Fregean content consists of the functions or primary intensions

appealed to by two-dimensionalism in the discussion above. When introducing Edenic

content, Chalmers writes:

‘The view of content that most directly mirrors the phenomenology


of colour experience is primitivism. Phenomenologically, it seems to us as
if visual experience presents simple intrinsic qualities of objects in the
world, spread out over the surface of the object… [Colours] are simple
intrinsic qualities whose nature we seem to grasp fully in perceptual
experience. For the world to be exactly the way that my phenomenology
seems to present it as being, the world would have to be an Edenic world
in which these properties are instantiated.’ (Chalmers, 2006, p66)

Thus Edenic content consists of primitive colour properties: sui generis properties

that are not surface reflectance properties nor dispositions to cause certain types of

experiences in subjects.

Chalmers’s view of the colour inversion problem is that the two halves of the

population have experiences with different Edenic contents. The two halves of the

population represent in experience different primitive colour properties. His argument for

the no colours view proceeds by elimination.

His argument against the many colours view is that it entails that every object has

every colour at any particular time, since, as he says ‘for any phenomenal colour, it seems

that there is a community in which the apple normally causes experiences with that

phenomenal colour’ (Chalmers, 2006, p68). Chalmers holds that this is a bad result since:

245
‘It follows that no colour experience of an object can be illusory
with respect to colour. Whatever the phenomenal colour of the experience,
the object will have the corresponding primitive property, so the
experience will be veridical. This conclusion is perhaps even more
counterintuitive than the conclusion that all colour experiences are
illusory.’ (Chalmers, 2006, p68).

His argument against the unique colour view is that it ‘imposes an asymmetry on

what otherwise seems to be a quite symmetrical situation’ (Chalmers, 2006, p68). Both

the lighting conditions and the perceptual mechanisms are working normally for the two

halves of the population.

With regards to the no colour view, Chalmers argues as follows:

‘Once these options are ruled out, the reasonable conclusion is that
neither experience is veridical: the apple is neither perfectly red not
perfectly green. Generalizing from this case, this reasoning suggests that
primitive properties are not instantiated at all. I think that this is clearly the
most reasonable view for a primitivist should take: on this view,
experiences attribute primitive properties, but their objects never possess
these properties.’ (Chalmers, 2006, p69).

2.4 The Unique Colour View

Some colour primitivists have argued that the correct response to the colour

inversion problem is the unique colour view, according to which one and only one half is

right (Bill Brewer in conversation).

246
Our objection to the unique colour view above was that it seemed arbitrary to pick

one half of the population as the one that had veridical experiences. All the factors that

seem relevant to determining whether an experience is veridical seem to be symmetrical

with respect to the two halves of the population. Let us call this type of objection the

arbitrariness objection.

An arbitrariness objection takes the following form. One is confronted with two

items, A and B, and for some property F the question arises as to whether they are both F,

one and only one of them is F, or neither of them is F. It seems arbitrary to say that one

and only one of them is F since every factor that seems relevant to determining whether

one of them is F is symmetrical with respect to A and B. That is, the only relevant

symmetry-breakers between A and B do not seem relevant to the question whether A or B

is F. Let us call this the no relevant symmetry-breakers condition.

Another common feature of arbitrariness objections is that if we consider A and B

on their own, there is normally prima facie evidence that each of them is F. Let us call

this the individual case plausibility condition. Admittedly, the individual case plausibility

condition is not a necessary condition for the arbitrariness objection to apply. For

instance, if we are debating whether the number 2 or the number 3 is green, and consider

the view that one and only one of 2 and 3 is green, then it would indeed seem arbitrary to

pick one number over the other as the one that is green, even though the individual case

plausibility condition is not met. But in this case, and in general when the individual case

247
plausibility condition is not met, the debate would never progress far enough for it to be

appropriate to introduce the arbitrariness objection.

Clearly defenders of the unique colour view are not persuaded by the arbitrariness

objection. One way of assessing the force of the arbitrariness objection is to consider

other debates in which it is applied.

One problem to which the arbitrariness objection applies is that of fission in the

personal identity debate. In the fission problem, A’s brain is severed into two parts, and

the two parts are placed in the bodies of B and C. The options with regards to A’s

persistence through time are: A is identical to B and C; A is identical to one and only one

of B and C; A is identical to neither B nor C.

Let us call the second option, the view that A is identical to one and only one of B

and C, the unique survivor view. The unique survivor view has been subject to the

arbitrariness objection. Firstly, the individual case plausibility condition is met in the

fission case. If we consider B on its own, then there is prima facie plausibility to the idea

that B is identical with A. Secondly the no relevant symmetry-breakers condition is met

in the fission case; the only symmetry-breakers are factors that are not relevant to A’s

persistence through time. For instance, one symmetry-breaker may be that B’s brain

contains molecule56 which C’s brain does not. But this symmetry-breaker is intuitively

not relevant to the question of which of them is identical with A.

248
We appealed to the arbitrariness objection when discussing the observer-relative

view in chapter 3. For instance, when comparing human beings who normally stand on

their feet and human beings who normally stand on their hands the other way up, we

established that the observer-relative view was committed to saying that human beings of

either one kind or the other were having illusions about where objects phenomenally look

to be. The arbitrariness objection was that it seemed arbitrary to pick out one kind of

human being and say that human beings of that kind were the ones having illusions.

Let us return now to the case of fission. Let us call someone who is not persuaded

by the arbitrariness objection a symmetry myope; i.e. someone who does not see the

symmetry that others see. A symmetry myope is very unlikely to proceed by identifying a

symmetry-breaker and arguing that that symmetry-breaker is a relevant condition to A’s

persistence over time. B and C can be qualitatively identical, and thus the only symmetry-

breakers will be differences such as which particular molecules B’s brain contains, and

these factors do not seem to be relevant to A’s persistence over time. It seems more likely

that a symmetry myope will argue as follows. It is a brute fact that A persists through the

fission process with one half of his brain or the other. It is a mistake to think that every

fact about A’s persistence through time has to be constitutively explained by other

differences; that way of thinking leads to an infinite regress. There is simply a primitive,

ungrounded fact of the matter concerning which of half of A’s brain A persists with.

249
Let us say that a concept F is ungrounded when there are no conditions which

constitutively explain why Fness obtains. According to the above argument, the concept

persistence through time is ungrounded.

Let us say that a concept F is unanalysable if it is not possible to develop a

principled and non-circular set of necessary and sufficient conditions for the application

of F. A symmetry myope may argue that if persistence through time is unanalysable, it is

ungrounded. And they may argue that persistence through time is unanalysable, pointing

out that the arguments that have been offered in support of the claim that concepts such

as goodness and knowledge are unanalysable, for instance philosophers’ inability to

develop counter-example-free analyses, apply equally strongly to the concept persistence

through time.

It seems that the last argument is invalid. It does not follow from a concept being

unanalysable that it is ungrounded. Suppose that goodness is unanalysable. One can

accept this and maintain that one can specify the conditions that constitutively explain the

quantity of goodness present in any given situation. For instance, even though goodness

is unanalysable, this itself does not stop one from knowing the set of conditions that

goodness supervenes on. One may know that in a given situation S the quantity of

suffering in S, but not the quantity of salt in the sea, is relevant to determining how much

goodness is present in S. Furthermore, if one asks why the quantity of suffering in S is

relevant to determining the quantity of goodness in S, intuitively one answer is that the

quantity of suffering in S partly constitutively explains the quantity of goodness in S.

250
Thus, it seems that there is no conflict between a concept F being unanalysable and there

being conditions which constitutively explain why Fness obtains.

Let us now consider the question whether the facts concerning persistence

through time supervene on the facts that do not concern persistence through time. For

instance, in the fission case suppose that one specifies all the facts about the fission

process that do not mention A’s persistence over time. The fission process occurs between

t1 and t2. There is human being A at t1. We introduce Left and Right as names for the left

half and the right half of A’s brain at t1. Between t1 and t2 A has her brain split into two

parts, Left and Right, and these two parts are placed into two bodies. Thus at t2 we have

the person with Left and the person with Right. We can now ask whether A persists at t2

as the person with Left or as the person with Right. Suppose that in the actual world A

persists as the person with Left. Is there a possible world which is identical to the actual

world in all the respects that can be specified without presupposing A’s persistence over

time, and in which A persists as the person with Right? Let us express this question in

terms of whether the persistence facts about A supervene on the non-persistence facts

about A.

Let us suppose that we accept that the persistence facts about A supervene on the

non-persistence facts about A. It is not clear whether the symmetry myope can account

for why we accept this thesis. In what follows I assume the following definition of

supervenience:

251
Supervenience: For all types x and y, the x-type facts supervene on the y-type facts

iff for any set of y-type facts, W, there is some set of x-type facts Z

such that necessarily, if the facts in W obtain, then the facts in Z

obtain.

Let us suppose that one wishes to defend the thesis that, for some particular x and

y, the x-type facts supervene on the y-type facts. One way of defending this thesis would

be to argue that the y-type facts constitutively explain the x-type facts. Let us express this

idea in the following principle:

The Supervenience-Explanation Principle: If the y-type facts constitutively

explain the x-type facts, then the x-

type facts supervene on the y-type

facts.

The defence of the supervenience-explanation principle is that if the y-type facts

constitutively explain the x-type facts, then the y-type facts are sufficient for the x-type

facts; and, given the definition of supervenience above, it follows the x-type facts

supervene on the y-type facts. Arguing that the y-type facts constitutively explain the x-

type facts is one way of defending the claim that the x-type facts supervene on the y-type

facts.

252
According to the symmetry myope, the non-persistence facts about A do not

constitutively explain the persistence facts about A. This commitment blocks one way for

the symmetry myope to defend the view that the persistence facts about A supervene on

the non-persistence facts about A. Since we find the view that the persistence facts about

A supervene on the non-persistence facts about A very plausible, the above commitment

of the symmetry myope blocks one way for them to account for why we take this view to

be very plausible. There may be other ways for the symmetry myope to defend the view

that the persistence facts about A supervene on the non-persistence facts about A, but it is

not immediately obvious what these other ways are.

The same point applies to the colour inversion case. According to the symmetry

myope, when tomatoes phenomenally look red to half the population and green to the

other half of the population, tomatoes may be either red or green. Let us suppose that

tomatoes are red. Does a given tomato’s colour supervene on any of the other properties

of this tomato? It is not clear that the symmetry myope could argue that it does. Since the

symmetry myope argues that the non-colour properties of the tomatoes do not

constitutively explain why the tomato has the colour that it has, it is not clear how they

could defend the view that the tomato’s colour supervenes on the non-colour properties of

the tomato.

It is no part of the argument above that every fact about an object needs to be

constitutively explained in terms of another fact about the object. For instance, there may

be no properties of an object x that constitutively explains x’s position. However, equally

253
x’s position properties do not seem to supervene on x’s non-position properties. The

argument above was that if we take x’s F-type properties to supervene on x’s G-type

properties, then some defence of this view is required. One possible defence of the view

is that x’s G-type properties constitutively explain x’s F-type properties. The symmetry

myope is not able to make use of this kind of defence of the supervenience claims in

question in the above examples, and thus to the extent that they wish to defend the

supervenience claims in question, they need to develop alternative defences of these

claims. The main challenge to the symmetry myope is that it is not obvious what these

alternative defences would be.

This concludes our discussion of the colour inversion problem. We have

considered a number of views that are responses to the problem of colour inversion.

Although the no colours view is not very intuitive, it seems more intuitive than the

alternative views. We raised the problem of colour inversion in order to see whether the

responses to it may shed some light on the perspective problem. Since our response to the

colour inversion problem is the analogue of our response to the perspective problem,

consideration of the former problem has not persuaded us to change our mind about the

best response to the latter problem.

3 Necessary Colour and Position Scepticism

We have argued that objects do not have the colour or position properties that they

phenomenally look to have. Let us now consider whether objects could have had the

254
colour and position properties that they phenomenally look to have. Let us call the views

that objects do not have in the actual world the colour and position properties they

phenomenally look to have actual-world colour scepticism and actual-world position

scepticism respectively, and let us call the views that, necessarily, objects do not have the

colour and position properties that they phenomenally look to have necessary colour

scepticism and necessary position scepticism respectively. I shall argue that the

considerations that count in favour of actual-world colour and position scepticism also

count in favour of necessary colour and position scepticism.

For the sake of argument, let us suppose that actual-world colour scepticism is

true, and necessary colour scepticism is false. Then there is a possible world in which

objects have the colours that they phenomenally look to have. Let us call this world w1. It

seems easy to bring about a state of affairs in w1 in which tomatoes phenomenally look

one colour to half the population and another colour to the other half of the population.

Assuming that objects can have only one colour at a time, it follows that the unique

colour view is correct in w1; that is, that one and only one half of the population may be

right. But if we make this judgement about w1, the question arises as to what gives us the

confidence to rule out the unique colour view in the actual world. It seems that unless

there is some contingent feature of the actual world that explains why the unique colour

view is false in the actual world, then there is no reason to treat w1 and the actual world

asymmetrically. Therefore, to the extent that there is reason to believe actual-world

colour scepticism, there is reason to believe necessary colour scepticism.

255
The same argument applies to the relation between actual-world position

scepticism and necessary position scepticism. Suppose, for the sake of argument, that

actual-world position scepticism is true, and necessary position scepticism is false. Let us

call the world in which objects have the position properties they phenomenally look to

have w2. It seems easy to bring about a state of affairs in w2 in which subjects view an

object from multiple perspectives, and the object phenomenally looks to have different

positions to each of them. Assuming that an object can have only one position at a time, it

follows that the unique position view is true in w2. But if we make this judgement about

w2, the question arises as to what gives us the confidence to rule out the unique position

view from being true in the actual world. Unless we can point to some contingent feature

of the actual world that explains why the unique position view is false in the actual world,

it seems that there is no reason to treat w2 and the actual world asymmetrically. Therefore,

to the extent that there is reason to believe actual-world position scepticism, there is

reason to believe necessary position scepticism.

3.1 Necessarily Coextensive Properties

There is a problem that emerges from necessary colour and position scepticism.

Consider the following three claims:

The Property View: Phenomenal looking is a relation between two

objects and a property.

256
Necessary Coextension: For all properties F and G, being F is identical with

being G iff necessarily, for all objects x, x is F iff x

is G.

Necessary Colour Scepticism: Necessarily, objects do not have the colour

properties that they phenomenally look to have.

Necessary colour scepticism and necessary coextension together entail that being

red is identical with being green. This conclusion, together with the property view, entails

that for all x, x phenomenally looks red iff x phenomenally looks green, which is false.

One response to this problem is to abandon necessary colour scepticism. If we

accept this option, we would have to investigate where the argument for necessary colour

scepticism that we considered in section 2 went wrong. It is not clear that either the

property view or the necessary coextension view give any indication of where that

argument might have gone wrong.

A second response to the problem is to abandon the necessary coextension

principle. If we do this, then we need to identify an alternative method of individuating

properties.

257
A third response to the problem is to abandon the property view. Ostrich

nominalism about properties is the view that properties do not exist (Armstrong 1978

introduces the term ‘ostrich nominalism’). An ostrich nominalist would deny the property

view. According to ostrich nominalism, an object O’s phenomenally looking red to a

subject S is not to be understood in terms of O’s standing in a certain relation to S and to

the property of being red.

We have considered what is involved in rejecting necessary colour scepticism.

Proper consideration of this problem would require an investigation into what is involved

in rejecting the property view and the necessary coextension view. However, this would

take us beyond the scope of this chapter. In what follows I will assume that one of the

property view and the necessary coextension view should be given up.

Even though endorsing ostrich nominalism is one of the options, I will not always

write in a nominalistically acceptable way, in the sense that I frequently write about ‘the

position properties that objects phenomenally look to have’. However, I assume that

ostrich nominalist paraphrases are available for sentences such as these, and therefore

that the ostrich nominalist option suggested in this section is not closed off by such

descriptions.

3.2 Phenomenal Positions As Particulars

258
There is a problem that emerges from the following three views:

Necessary Position Scepticism: Necessarily, objects do not have the position

properties that they phenomenally look to have.

Phenomenal Looking

Exportation Principle: Necessarily, for all subjects S and objects x, if a

singular term referring to x occurs within the scope

of ‘phenomenally looks’ in a true ascription of a

state of phenomenal looking to S, then x

phenomenally looks some way to S.

The Phenomenal Positions

As Particulars Principle: There are objects x and y, and a phenomenal

position l, such that x phenomenally looks at l to y.

The phenomenal positions as particulars principle together with the phenomenal

looking exportation principle entail that there is a phenomenal position that

phenomenally looks some way to someone. Necessary position scepticism entails that

nothing has phenomenal positions. If nothing has phenomenal positions, it seems that

phenomenal positions do not phenomenally look any way to subjects (thanks to Stephen

Kearns for pointing out this problem).

259
My reply to this problem is to abandon the phenomenal positions as particulars

principle. Rather than holding that objects phenomenally look to occupy phenomenal

positions, we should hold that there are phenomenal position properties that objects

phenomenally look to have. For instance, an object might phenomenally look to have

position property l1. We should think of objects phenomenally looking to have position

properties in the same way in which we think of objects phenomenally looking to have

colour properties. That is, just as we do not think of colour properties as entities that

objects phenomenally look to stand in a certain relation to, nor should we think of

position properties as entities that objects phenomenally look to stand in a certain relation

to.

4 Consequences of Primitivism

In this section I develop some consequences of primitivism about phenomenal

position properties. I discuss some relations between phenomenal position properties, and

discuss ways in which the same objects might phenomenally look to have different

position properties to different subjects.

4.1 Phenomenal Spatial Relations

260
First I will introduce some names for relations that hold between phenomenal

position properties. Consider a normal situation in which one is looking at an object, O,

and O moves to the right between t1 and t2. This situation is illustrated in figure 2:

Figure 2

O (at t1) O (at t2)

The phenomenal position property that O phenomenally looks to have at t2 bears a

certain relation to the phenomenal position property that O phenomenally looks to have at

t1. We shall call this relation being to the phenomenal right of. Thus, if O phenomenally

looks to have position property l1 at t1, and it phenomenally looks to have position

property l2 at t2, then l2 is to the phenomenal right of l1. Conversely, l1 is to the

phenomenal left of l2.

Consider a normal situation in which one is looking at an object O, and O moves

upwards between t1 and t2. This situation is illustrated in figure 3:

Figure 3

O (at t2)

261
O (at t1)

Suppose that O phenomenally looks to have l1 at t1, and l3 at t3. We shall say that l3

is phenomenally above l1, and that, conversely, l1 is phenomenally below l3.

Consider a normal situation in which one is looking at three objects, A, B and C,

and B lies on the straight line between A and C. This situation is illustrated in figure 4:

Figure 4

A B C

If A phenomenally looks to have l1, B phenomenally looks to have l2, and C

phenomenally looks to have l4, then we shall say that l1 is phenomenally closer to l2 than

it is to l4. Conversely, l4 is phenomenally further away from l1 than it is from l2. We shall

also say that the phenomenal distance between l1 and l4 is equal to the phenomenal

distance between l1 and l2 added to the phenomenal distance between l2 and l4. Let us call

the relations that we have defined above phenomenal spatial relations.

4.2 The Shifting Hypothesis

262
Suppose that there are two disjoint sets of phenomenal position properties, SP1

and SP2, such that the phenomenal spatial relations that members of SP1 bear to other

members of SP1 are the same as the phenomenal spatial relations that members of SP2

bear to other members of SP2. Suppose, furthermore, that every member of SP1 is

phenomenally above every member of SP2. We can define the shifting hypothesis as

follows:

The Shifting Hypothesis: There could be two beings, x and y, such that:

(i) x and y are looking at the same set of objects W

(ii) The members of W fill the fields of view of x and y

respectively.

(iii) The members of W phenomenally look to x to have

the members of SP1, and no other position

properties.

(iv) The members of W phenomenally look to y to have

the members of SP2, and no other position

properties.

SP1 and SP2 are represented in figure 5. Phenomenal position properties occupy

coordinates on the graph, and the further apart two phenomenal position properties are on

the x axis, the further to the phenomenal right one of them is from the other. The further

apart two phenomenal position properties are on the y axis, the further one of them is

phenomenally above the other.

263
Figure 5

SP1

SP2
x axis

y axis

According to field-of-view relationalism, phenomenal position properties are

relations between seen objects. On this view, necessarily, if S1 and S2 see the same set of

objects, and those objects phenomenally look to bear the same spatial relations to each

other, then the position properties that the objects phenomenally look to S1 and to S2 to

have are the same. According to the shifting hypothesis, there could be two subjects, S1

and S2, say, who see the same set of objects, and those objects phenomenally look to bear

the same set of spatial relations to each other, but the objects phenomenally look to have

different sets of position properties to S1 and to S2. Thus, field-of-view relationalism is

inconsistent with the shifting hypothesis. However, in chapter 3 we argued that there

were reasons to reject field-of-view relationalism.

There is a positive reason to accept the shifting hypothesis. Call the position

properties in the upper semi-circle of SP2 SP2U, and call the position properties in the

264
lower semi-circle of SP2 SP2L. If primitivism is true, then there seems no reason why it

should not be possible for the following conditions to be met:

(i) There are two beings, x and y, such that x and y are looking at the same set of

objects W.

(ii) The members of W fill the fields of view of x and y respectively.

(iii) The members of W phenomenally look to x to have the members of SP2U, and

no other position properties.

(iv) The members of W phenomenally look to y to have the members of SP2L, and

no other position properties.

In a world in which the above conditions to be met, the position properties that

objects phenomenally look to x to have are phenomenally above the position properties

that objects phenomenally look to y to have. If it is possible that (i)-(iv) are met, then it

seems that it would be possible for the conditions in the shifting hypothesis to be met.

4.3 The Scaling Hypothesis

Consider two sets of phenomenal position properties, SP3 and SP4, where SP3 is a

subset of SP4. Let us suppose that SP3 and SP4 are as they are illustrated to be in figure 6.

265
Figure 6

SP4
SP3
x axis

y axis

We can now define the scaling hypothesis:

The Scaling Hypothesis: There could be two beings, x and y, such that:

(i) x and y are looking at the same set of objects W.

(ii) The members of W fill the fields of view of x and y

respectively.

(iii) The set of position properties that the members of

W phenomenally look to x to have is SP3.

(iv) The set of position properties that the members of

W phenomenally look to y to have is SP4.

That is, for every pair of members of W that x and y see, m1 and m2, the position

properties that m1 and m2 phenomenally look to x to have are phenomenally closer

together than the position properties that m1 and m2 phenomenally look to y to have.

266
An alternative formulation of the scaling hypothesis which captures a similar idea

is that there could be two beings, x and y, such that x and y see the same objects, and

these objects fill the fields of view of x and y respectively, and the objects phenomenally

look smaller to x than they do to y. That is, x sees the same set of objects as y, but sees

them in a smaller scale than y does. However, this formulation of the scaling hypothesis

is not open to me, since, in chapter 1, I argued that objects do not phenomenally look to

have size properties.

The members of W may phenomenally look to x and to y to bear the same set of

spatial relations to each other. Thus the fact that the scaling hypothesis allows that the

members of W phenomenally look to have different sets of position properties entails that

the scaling hypothesis is inconsistent with field-of-view relationalism. However, this is

not a problem for the scaling hypothesis, as we found reason to reject field-of-view

relationalism in chapter 3. If primitivism is true, then it seems that there is no reason why

the scaling hypothesis should not also be true.

5 Conclusion

In this chapter I have argued that phenomenal position properties are sui generis,

absolute position properties. I have also argued for necessary colour and position

scepticism, the views that, necessarily, objects do not have the colour and position

properties that they phenomenally look to have.

267
Chapter 5

An Argument That Objects Phenomenally Look To Have z

Coordinates

1 The z Coordinate Constraint

In chapter 4 I argued for primitivism, the view that the position properties that

objects phenomenally look to have are absolute, primitive properties, which are such that

it is not metaphysically possible for objects to have them. In this chapter I shall address

the question whether these position properties consist of two coordinates or three

coordinates.

It is uncontroversial that objects phenomenally look to have horizontal and

vertical coordinates, or x and y coordinates. It is intuitive that there may be three objects,

A, B and C, such that the position that B phenomenally looks is further along in a

horizontal direction from the position that A phenomenally looks, and such that the

position that C phenomenally looks is further along in a vertical direction from the

position that A phenomenally looks. It is controversial, however, whether objects

phenomenally look to have coordinates on a forwards/backwards axis, or a z axis.

268
Traditionally this question has been expressed in terms of whether one can visually

perceive depth properties, for instance whether an object can phenomenally look at a

certain point on an axis extending forwards of one.

In chapter 1, I argued that there is the following constraint on the possibility of

objects phenomenally looking to have z coordinates:

The z Coordinate Constraint: If it is possible that objects phenomenally look to

have z coordinates, then it is possible that there are

three objects, o1, o2, o3, such that o1 and o2

phenomenally look to o3 to have the same x and y

coordinates, but distinct z coordinates.

Analogues of the z coordinate constraint for x and y coordinates are easily met.

For instance, consider looking at a red cube at t1 against a white background. Suppose

that, between t1 and t2, the cube moves to the left by a small distance. The x coordinates

that the cube phenomenally looks to have will be different at t1 and t2, and the y

coordinates and, if applicable, the z coordinates, that the cube phenomenally looks to

have will be the same at t1 and t2.

Suppose that between t2 and t3 the cube moves up a small distance. The y

coordinates that the cube phenomenally looks to have will be different at t2 and t3, and the

269
x coordinates and, if applicable, the z coordinates, that the cube phenomenally looks to

have will be the same at t2 and t3.

It is less easy to imagine how the z coordinate constraint might be met. For

instance, suppose that between t3 and t4, the cube moves further away from one. This

situation is illustrated in figure 1. Figure 1 is a bird’s eye perspective of a horizontal cross

section of oneself and the cube at two different times, t3 and t4.

Figure 1

The cube at t4

C1 C2

The cube at t3

C1 C2

The x and y coordinates that the cube phenomenally looks at t3 and at t4 will be

different. Indeed, the set of x and y coordinates that the cube phenomenally looks to have

at t4 will be a subset of the set of x and y coordinates that the cube phenomenally looks to

have at t3. C1 and C2 are the two ends of the top of the side of the cube that one is looking

at. Suppose, for instance, that, at t3, C1 phenomenally looks to have x coordinate x5, and

270
C2 phenomenally looks to have x coordinate x10. At t4, C1 and C2 will phenomenally look

to have x coordinates that are phenomenally closer together than x5 and x10 are. For

instance, at t4, C1 may phenomenally look to have x6, and C2 may phenomenally look to

have x8.

Thus, this example does not provide a way of satisfying the z coordinate

constraint. To satisfy the z coordinate constraint, we need to ensure that the object

phenomenally looks to have the same x and y coordinates at t4 as it phenomenally looks

to have at t3.

One way to ensure this, given that the cube has moved away from one, is to

enlarge the cube by an appropriate amount. Let us suppose that some enlarging of the

cube takes place between t4 and t5. This enlarging is such that, at t5, the facing side of the

cube is at the same distance away from one as it is at t4, and that the facing side of the

cube phenomenally looks to have the same x and y coordinates that it phenomenally

looks to have at t3. This situation is illustrated in figure 2. As with figure 1, figure 2 is a

bird’s eye perspective on a horizontal cross-section of oneself and the cube.

271
Figure 2

The cube at t5, having been appropriately enlarged.

The cube at t4

The cube at t3

Will the cube phenomenally look to have different z coordinates at t3 and t5? It

seems that there will in fact be no visual phenomenal difference between the way the

cube phenomenally looks at t3, and the way the cube phenomenally looks at t5. Given the

phenomenal character principle, it follows that there is no difference in any z coordinate

that the cube phenomenally looks to have at t3 and at t5.

2 The Possibility of Objects Phenomenally Looking To Have z

Coordinates

In this section I will argue that, by considering the possibility of 360 degree

vision, one can show that two objects can phenomenally look to some being to have the

same x and y coordinates, but different z coordinates, and therefore that the z coordinate

272
constraint can be met. In the next section, I will argue that objects phenomenally look to

have z coordinates to us. If objects do phenomenally look to us to have z coordinates,

then the correct response to the argument in the paragraph above is to say that the cube

phenomenally looks to have the same z coordinate at t3 and t5.

Consider a cubical being, which we shall call Cube, that has eyes covering all of

its six sides. Figure 3 is a bird’s eye perspective of a horizontal cross-section of Cube:

Figure 3

Cube has 360 degree vision, and the information from its various eyes is all

integrated into one global visual experience. Just as the information from our two eyes is

integrated into one visual experience, so the information from all of Cube’s eyes is

integrated into one global visual experience.

There is good evidence that certain animals, such as woodcocks, have 360 degree

vision (Waldvogel 1990). It is less easy to demonstrate that information from the eyes of

such animals is integrated into one visual experience, but there does not seem any

obvious obstacle to this being the case. If we reflect on our own experiences, we can

273
imagine that our field of vision, which is approximately 140 degrees, might be increased,

to 200 degrees, say by our acquiring a third eye on one side of our heads. There does not

seem any obstacle to the information received from this additional eye being integrated

into the same visual experience as the information from our existing two eyes. And the

same point seems to apply to fourth, fifth and sixth eyes that we might acquire, which

would give us 360 degree vision. Thus it seems reasonable to assume that the information

from all of Cube’s eyes is integrated into one global visual experience.

Let us suppose that, as illustrated in figure 4, Cube sees two apples, apple1 and

apple2. Apple1 is on one side of Cube, and apple2 is on the opposite side of Cube. My

central argument will be that apple1 and apple2 phenomenally look to Cube to have

exactly the same x and y coordinates. Since apple1 and apple2 clearly do not

phenomenally look to Cube to have exactly the same position, apple1 and apple2 must

phenomenally look to have different coordinates on some third axis, a z axis.

Figure 4

Apple1

z axis

Apple2 x axis

y axis

274
I have labelled the axis going across the page the x axis, the axis going up the

page the z axis, and the axis going into the page the y axis. The axes are labelled as such

since figures 3 and 4 are how a horizontal cross-section of Cube would look from above.

These are the axes of physical space that Cube is in; for clarity I will henceforth refer to

them as physical axes.

If our treatment of the perspective problem in chapter 4 is correct, then the

position properties that objects phenomenally look to Cube to have are not coordinates on

these physical axes. That is, the position properties that objects phenomenally look to

Cube to have are not ones that they actually have. The perspective problem applies to

Cube as well as to us. Suppose that Cube, and another similar being, Cube2, are sitting

opposite each other, looking at two adjacent circles, one red and one green. The two

circles phenomenally look to be in different positions to Cube and Cube2. Since the

circles presumably cannot have two different position properties at once, for the reasons

mentioned above, the option that seems most plausible is that the circles do not have the

position properties that they phenomenally look to Cube and Cube2 to have.

From now on, to avoid confusion, I will refer to the coordinates that objects

phenomenally look to have as phenomenal coordinates, and the coordinates that objects

in fact have as physical coordinates. Thus, the question we are addressing at present is

whether objects phenomenally look to have phenomenal z coordinates.

275
In my argument I will also be appealing to the relations introduced in section 4.1

in chapter 4 of being to the phenomenal right/left of, and being phenomenally

above/below.

Let us suppose that apple1 phenomenally looks to Cube to have phenomenal x and

y coordinates, x5 and y5. A central premise in my argument that objects phenomenally

look to Cube to have phenomenal z coordinates is that apple2 may also phenomenally

look to Cube to have phenomenal x and y coordinates, x5 and y5. That is, a central

premise in my argument is that apple1 and apple2 may phenomenally look to Cube to have

the same phenomenal x and y coordinates.

The following is an argument for this premise. Suppose that apple2 phenomenally

looked at a different phenomenal x coordinate from apple1. Would the position property

that apple1 phenomenally looks to have be to the phenomenal right or to the phenomenal

left of the position property that apple2 phenomenally looks to have? If apple2

phenomenally looks to have a different phenomenal x coordinate from apple1, then the

phenomenal x coordinate that apple2 phenomenally looks to have must be either to the

phenomenal right or to the phenomenal left of the phenomenal x coordinate that apple1

phenomenally looks to have. Suppose, for the sake of argument, that apple2

phenomenally looks at phenomenal x10, that is, that the phenomenal x coordinate that it

phenomenally looks to have is to the phenomenal right of the phenomenal x coordinate

that apple1 phenomenally looks to have.

276
Let us now suppose that, between t1 and t2, Cube expands to the right and left, so

that it becomes an oblong. And suppose that it acquires enough new eyes that it continues

to have eyes all over the surface of its body. The expanded Cube is illustrated in figure 5.

Figure 5

Apple1 Apple3

z axis

Apple2 x axis

y axis

Figure 5 illustrates Cube’s shape after the expansion at t2. We assume that, at t2,

apple1 still phenomenally looks to have phenomenal x5. And we also assume that, at t2,

apple2 still phenomenally looks to have the same phenomenal x coordinate that, at t1, it

phenomenally looked to have, which, by hypothesis, is x10.

It seems that the effect of expanding Cube to the right and left is that objects

moving along the side marked ‘1’ in figure 5 can now phenomenally look to Cube to have

a wider range of phenomenal x coordinates. For instance, let us suppose that, at t1, an

object moving along side 1 could phenomenally look to have phenomenal x coordinates

277
ranging from x3 and x7. It seems plausible to suppose that, at t2, an object moving along

side 1 might phenomenally look to have phenomenal x coordinates ranging from x0 and

x10.

Given this, it seems plausible to suppose that, at t2, apple3 phenomenally looks to

have phenomenal x10. After all, the effect of the expansion is that objects can

phenomenally look further to the phenomenal right than they could before the expansion.

So it seems plausible to suppose that apple3 may phenomenally look to Cube to have

phenomenal x coordinate x10.

Since, by hypothesis, apple2 also phenomenally looks to have x10, at t2, apple3 and

apple2 phenomenally look to have the same phenomenal x coordinate. It is this kind of

claim that forms a central premise in my argument. That is, what is central to my

argument is that Cube sees two objects on opposite side of it, and these two objects

phenomenally look to it to have the same phenomenal x and y coordinates.

For the purposes of my argument, it does not matter whether it is apple2 and

apple3 that phenomenally look to have the same phenomenal x coordinate, or whether it is

apple1 and apple2 that phenomenally look to have the same phenomenal x coordinate. We

began supposing that apple1 and apple2 phenomenally look to have the same phenomenal

x coordinate. When we considered a challenge to this premise, we showed that if apple2

and apple1 do not phenomenally look to have the same phenomenal x coordinate, then at

least apple2 and apple3 phenomenally look to have the same phenomenal x coordinate.

278
Since the aim is to show that two objects on opposite sides of Cube can phenomenally

look to Cube to have the same phenomenal x coordinate, it does not matter whether this

claim holds for apple3 and apple2, or apple1 and apple2. For simplicity, I will assume that

it holds for apple1 and apple2, as we originally assumed. Thus apple1 and apple2

phenomenally look to Cube to have the same phenomenal x coordinate.

Mutatis mutandis, the same argument as above will show that it is safe to take as a

premise that apple1 and apple2 phenomenally look to have the same phenomenal y

coordinate. Thus, we assume that apple1 and apple2 phenomenally look to Cube to have

the same phenomenal x and y coordinates. It is clear that apple1 and apple2 phenomenally

look to Cube to have different position properties. If apple1 and apple2 phenomenally

looked to Cube to have the same position property, then there would be no visual

phenomenal difference between Cube seeing apple1 on its own, and Cube seeing apple1

and apple2 together. In fact, however, there is a significant visual phenomenal difference

between Cube seeing apple1 on its own, and its seeing apple1 and apple2 together. The

visual phenomenal difference is due in part to the fact that when Cube sees apple2 in

addition to seeing apple1, apple2 phenomenally looks to have a different position property

from apple1.

There seems to be only one way to accommodate the fact that apple1 and apple2

phenomenally look to Cube to be in different positions, and that is to allow that apple1

and apple2 phenomenally look to have different phenomenal z coordinates. We started off

279
this section by considering the following constraint on its being possible for objects

phenomenally to look to have phenomenal z coordinates.

The z Coordinate Constraint: If it is possible that objects phenomenally look to

have z coordinates, then it is possible that there are

three objects, o1, o2, o3, such that o1 and o2

phenomenally look to o3 to have the same x and y

coordinates, but distinct z coordinates.

It seems that the state of affairs illustrated in figure 4 meets this constraint. Apple1

and apple2 phenomenally look to Cube to have the same x and y coordinates but different

z coordinates.

The point that originally gave us grounds for suspicion that objects phenomenally

look to us to have z coordinates was that it does not seem straightforward to bring about a

situation in which there is an isolated change in the phenomenal z coordinate that objects

phenomenally look to us to have. That is, it does not seem straightforward to bring about

a situation in which an object phenomenally looks to us to have the same phenomenal x

and y coordinates between t1 and t2, but distinct phenomenal z coordinates. By contrast, it

does seem straightforward to bring about an isolated change in the phenomenal x and y

coordinates that objects phenomenally look to have. It does seem straightforward, for

instance, to bring about a situation in which an object phenomenally looks to have the

280
same phenomenal y and z coordinate between t1 and t2, and a different phenomenal x

coordinate.

The same restriction may well apply to Cube. It might well be the case that, at any

given time, only two of the three phenomenal coordinates that an object phenomenally

looks to Cube to have can change. For instance, suppose that, between t1 and t2, apple1

moves further away from Cube on the physical z axis, as illustrated in figure 6.

Figure 6

Apple1 (at t2)

Apple1 (at t1)

z axis

x axis
Apple2
y axis

The phenomenal z coordinate that apple1 phenomenally looks to have may not

change. In fact, if Cube is like us, all that will happen is that, at t2, apple1 will

phenomenally look to have phenomenal x and y coordinates that are closer together than

the phenomenal x and y coordinates that apple1 phenomenally looks to have at t1.

281
Suppose that, between t2 and t3, apple1 moves down the left-hand side of Cube.

This is illustrated in figure 7.

Figure 7

Apple1 (at t2)

Apple1 (at t1)

z axis

Apple1 (at t3)

x axis
Apple2
y axis

Between t2 and t3 the phenomenal z coordinate that apple1 phenomenally looks to

have will change. However, it is not so clear that, whilst this is happening, the

phenomenal x coordinate that apple1 phenomenally looks to have can change. For

instance, suppose that, between t3 and t4, apple1 moves away from the left-hand side of

Cube on the physical x axis. This is illustrated in figure 8.

282
Figure 8

z axis

Apple1 Apple1
(at t4) (at t3)

x axis
Apple2
y axis

If Cube is like us, it seems that when apple1 moves away from Cube on the

physical x axis between t3 and t4, the phenomenal y and z coordinates that apple1

phenomenally looks to Cube have will become closer together (here I am using the

description ‘the phenomenal y and z coordinates that apple1 phenomenally looks to Cube

to have’ non-rigidly; similar descriptions below are also used non-rigidly). It does not

seem that the phenomenal x coordinate that apple1 phenomenally looks to have need

change between t3 and t4.

There is still reason to believe, however, that as apple1 is travelling down the left-

hand side of Cube between t2 and t3, there is some phenomenal x coordinate that apple1

phenomenally looks to Cube to have. After all, suppose that, at t3, apple2 is travelling up

the right-hand side of Cube, and that it phenomenally looks to have the same phenomenal

z and y coordinates as apple1. This is illustrated in figure 9.

283
Figure 9

Apple1 (at t2)

Apple1 (at t1)

z axis

Apple1 Apple2
(at t3) (at t3)

x axis

y axis

Clearly, apple1 and apple2 phenomenally look to have different position properties

to Cube, and to accommodate this we must allow that, at t3, apple1 and apple2

phenomenally look to Cube to have different phenomenal x coordinates.

It may be that, as apple1 is rounding one of Cube’s corners, it can phenomenally

look to Cube to change all three of its phenomenal coordinates. For instance, suppose that

apple1 moves to the top left corner of Cube, as in figure 10.

284
Figure 10

Apple1
z axis
2

1 3

Apple2 x axis

y axis

It could be that Cube can see apple1 out of the eyes on side 1 and out of the eyes

on side 2. When apple1 moves along side 1, the phenomenal z coordinate it phenomenally

looks to have changes, and when it moves along side 2, the phenomenal x coordinate it

phenomenally looks to have changes. Thus, it is possible that when apple1 moves

diagonally in the directions of the double-headed arrow in figure 10, both the x and z

phenomenal coordinates that it phenomenally looks to have change. And, by moving

upwards at the same time, the phenomenal y coordinate that it phenomenally looks to

have can change too.

Another way in which the phenomenal x, y and z coordinates that an object

phenomenally looks to a subject to have could change all at once is if the subject was L-

shaped, as in figure 11.

285
Figure 11

z axis

1 Apple1

2
x axis

y axis

Let us call the being in figure 11 L. As apple1 moves along side 1 of L, the

phenomenal z coordinate that it phenomenally looks to have changes. As apple1 moves

along side 2 of L, the phenomenal x coordinate that it phenomenally looks to have

changes. Thus, if apple1 moves in a diagonal direction from sides 1 and 2, in the

directions of the double-headed arrow, then both the phenomenal z and the phenomenal x

coordinate that apple1 phenomenally looks to have may change at the same time. And if

apple1 moves upwards at the same time, then the phenomenal x, y and z coordinates that

apple1 phenomenally looks to have may change at the same time.

There is a slight difficulty when considering apple1’s movement away from L’s

sides. Suppose that apple1 moves towards side 2 between t1 and t2, as in figure 12, whilst

remaining the same distance from side 1.

286
Figure 12

z axis

Apple1 at t1
1
Apple1 at t2

2
x axis

y axis

When objects move closer to us, the phenomenal x and y coordinates that the

boundary points of those objects phenomenally look to have become further apart. If we

were simply considering how things phenomenally look to side 2 of L, as opposed to L as

a whole, it would be natural to suppose that at t2, the phenomenal x and y coordinates that

the boundary points of apple1 phenomenally look to have become further apart. However,

apple1 is the same distance from side 1 at t2 as it was at t1, and thus, if we were

considering how things phenomenally look from the perspective of side 1, we would not

expect the phenomenal x and y coordinates that the boundary points of apple1

phenomenally look to have to change (thanks to Maria Lasonen for this point).

However, since we are supposing that the information from L’s eyes is integrated

into one global visual experience, we cannot suppose that the phenomenal x and y

coordinates that the boundary points of apple1 phenomenally look to have both change

and do not change. Thus L’s visual system must somehow reconcile the conflicting

information that it is receiving from its eyes. At t2, the image that apple1 is projecting onto

287
the retina of L’s eyes on side 2 is larger than the image that apple1 is projecting onto the

retina of L’s eyes on side 1. The information that L’s visual system is receiving from its

eyes on side 2 suggests that the visual system should make apple1’s boundary points

phenomenally look further apart at t2 than at t1; and the information that L’s visual system

is receiving from its eyes on side 1 suggests that the visual system should not make

apple1’s boundary points phenomenally look further apart at t2 than at t1.

It seems an empirical question how L’s visual system would resolve this conflict.

Perhaps side 1 dominates, and L’s visual system would resolve the conflict by not making

the boundary points of apple1 phenomenally look further apart at t2 than at t1. Or perhaps

L’s visual system would average out the information from the eyes on side 1 and side 2,

and make the boundary points of apple1 phenomenally look slightly further apart at t2

than at t1. Both of these options seem possible.

One might think that all of the questions that we have been considering in this

section are ones for empirical science, and not ones that are amenable to armchair

analysis. How can we know how things would phenomenally look to Cube? However, the

assumptions that we made in the example of Cube were quite weak. They were 1.) 360

degree vision is possible, 2.) the information from the eyes on Cube would be integrated

into one visual experience. The first assumption is well supported by empirical evidence.

The second assumption seemed quite plausible when we considered how 360 degree

vision might be realized in our own case; I will not repeat the argument here. Thus it does

288
not seem too ambitious to think that the armchair approach can deliver results in the case

of examples such as Cube.

3 Objects Phenomenally Looking To Have z Coordinates to Us

We have established that objects phenomenally look to Cube to have z

coordinates. It remains to be shown that objects phenomenally look to have phenomenal z

coordinates to us. The intuitive challenge to the hypothesis that objects phenomenally

look to have phenomenal z coordinates to us that we considered earlier was that, although

there can be isolated changes in the phenomenal x coordinates that objects phenomenally

look to us to have, and isolated changes in the phenomenal y coordinates that objects

phenomenally look to us to have, it is not straightforward that there can be isolated

changes in the phenomenal z coordinates that objects putatively phenomenally look to us

to have.

However, this challenge can be answered. Apple1 phenomenally looks to Cube to

have a phenomenal z coordinate even though, when apple1 restricts its movements to the

side of Cube that it is on in figures 4 and 5 (and does not go around the corners of Cube),

there cannot be any changes in the phenomenal z coordinate that apple1 phenomenally

looks to have; apple1 phenomenally looks to have the same phenomenal z coordinate

throughout. Our answer to the intuitive challenge, then, is that the fact that there are no

changes in the phenomenal z coordinate that objects putatively phenomenally look to us

289
to have is compatible with there being some phenomenal z coordinate that those objects

phenomenally look to have.

I shall now argue that there is some reason to think that objects phenomenally

look to us to have phenomenal z coordinates. Consider the situation illustrated in figure

13.

Figure 13

C D

Supposing that I am the subject represented in figure 13, object A is at the left-

hand edge of my field of view. Put informally, my argument that objects phenomenally

look to us to have phenomenal z coordinates is as follows. D phenomenally looks as far

from C as B phenomenally looks from A. However, D phenomenally looks further to the

phenomenal right of C than B does from A. A, B, C and D phenomenally look to have the

same phenomenal y coordinate. These points are consistent only if A and B phenomenally

look to have different phenomenal z coordinates.

290
Put more precisely, the argument is as follows. We will call the position properties

that A, B, C and D phenomenally look to have LA, LB, LC and LD respectively.

1.) There is some phenomenal distance w, such that LA is w from LB, and LC is w

from LD.

2.) Whilst the phenomenal x coordinate in LD is some phenomenal distance w from

the phenomenal x coordinate in LC, phenomenal x coordinate in LB is less than w

from the phenomenal x coordinate in LA.

3.) A, B, C and D phenomenally look to me to have the same phenomenal y

coordinate.

Therefore:

4.) LA contains a distinct phenomenal z coordinate from LB.

The argument for 1.) is that it seems as though A phenomenally looks as far apart

from B as C phenomenally looks from D. The argument for 2.) is that whilst LB may be a

little to the phenomenal right of LA, it does not seem as much to the phenomenal right of

LA as LD seems to be from LC.

The argument for 3.) is that A, B, C and D are all at the same height relative to

me, and thus they all phenomenally look to have the same phenomenal y coordinate.

291
It follows from 1.), 2.) and 3.) that LA contains a distinct phenomenal z

coordinate from LB, and therefore that A phenomenally looks to have a distinct

phenomenal z coordinate from B.

1.) and 2.) are the crucial premises in this argument. When one introspects on

one’s visual experiences, they come to seem plausible. For instance, at the moment my

computer processor and my printer are arranged as A and B are respectively, and the left-

hand and right-hand edges of the screen of my computer are arranged as C and D are

respectively. That is, my computer processor and printer are to the left of me, and my

computer screen is in front of me. My computer processor phenomenally looks as far

from my printer as the right-hand edge of my computer screen phenomenally looks from

the left-hand edge of my computer screen. But my printer does not phenomenally look as

far to the phenomenal right of my computer processor as the right-hand edge of my

computer screen phenomenally looks from the left-hand edge of my computer screen. All

of the objects in question phenomenally look to have the same phenomenal y coordinate.

The only way of reconciling the above points is to suppose that my computer processor

phenomenally looks to have a different phenomenal z coordinate from my printer. This

seems an intuitive result: in this situation, it does seem natural to say that the printer

phenomenally looks forward of the computer processor.

In our discussion of Cube above, we established that if we came to have 360

degree vision, then objects would phenomenally look to have phenomenal z coordinates.

Furthermore, we showed that, as we acquired eyes on the sides of our heads, then objects

292
moving up and down our left and right hand would phenomenally look to have changing

phenomenal z coordinates. That is, it is the objects moving in certain directions at the

left-most and right-most extremities of our field of view that, in such a situation, would

phenomenally look to have changing phenomenal z coordinates. This observation, based

on thought experiment, is the same as the introspective observation that we made when

considering figure 13 above: it is at the left-most and right-most extremities of our field

of view, where the angle of vision is the greatest, that perceived objects can

phenomenally look to us to have different phenomenal z coordinates.

4 Conclusion

In this chapter I have argued that it is metaphysically possible for objects

phenomenally to look to have phenomenal z coordinates, and, furthermore, that there is

some introspective evidence to suggest that objects do phenomenally look to have

phenomenal z coordinates.

293
Chapter 6

Coarse-Grained Vision and New Kinds of Phenomenal

Character

1 Berkeley on Abstract Ideas

Berkeley characterises Locke’s view about abstract ideas as follows:

‘So likewise the mind, by leaving out of the particular colours


perceived by sense that which distinguishes them one from another, and
retaining only that which is common to all, makes an idea of colour in
abstract which is neither red, nor blue, not white, nor any other
determinate colour.’ (Berkeley, 1975, p67).

Berkeley’s view is that he can form ideas of particular colours, but not an idea of

colour in general:

‘Whether others, have this wonderful faculty of abstracting their


ideas, they best can tell: for myself, I dare be confident I have it not…
But…whatever hand or eye I imagine, it must have some particular shape
and colour. Likewise the idea of a man that I frame to myself must be
either of a white, or a black, or a tawny, a straight, or a crooked, a tall, or a
low, or a middle-sized man. (Berkeley, 1975, p68).

294
In his introduction to Berkeley’s Philosophical Writings, David Armstrong

disagrees with Berkeley:

‘It is perfectly possible… to have a mental image of a piece of


crimson cloth of no particular shade of crimson.’ (Berkeley, 1965, p28).

In this chapter I try to make precise intuitions of the kind that Armstrong has, and

which Berkeley doesn’t have, and I try to make some progress towards evaluating them

(thanks to Ciara Fairley for pointing out the similarity between the issues I discuss in this

chapter and Berkeley’s views).

2 Some Initial Definitions

Visual Experience: Necessarily, for all objects x and y, and all properties F, x

has a visual experience of y as F iff y phenomenally looks

F to x.

Phenomenal Character: Necessarily, for all visual experiences e1 and e2, what it’s

like to have e1 is the same as what it’s like to have e2 iff e1

and e2 have the same kinds of phenomenal character.

There are different kinds of phenomenal character that a visual experience may

have. At the least a visual experience has colour phenomenal character and location

295
phenomenal character. We will say that the kinds of phenomenal character that visual

experiences have are all kinds of visual phenomenal character.

The correspondence principle identifies a relation that holds between certain kinds

of phenomenal character and certain properties that objects phenomenally look to have.

The Correspondence Principle:

(1) For all colour properties F, if it is possible that there are two objects, x and y, such

that x phenomenally looks F to y, then there is a kind of colour phenomenal

character K such that:

(i) Necessarily, for all objects w and z, if w phenomenally looks F to z,

then z’s visual experience of w has K.

(ii) For all kinds of colour phenomenal character L, if L is not K, and if,

necessarily, for all w and z, if w phenomenally looks F to z, then z’s

visual experience of w has L, then K is more specific than L.

(2) For all location properties F, if it is possible that there are two objects, x and y,

such that x phenomenally looks F to y, then there is a kind of location phenomenal

character K such that:

(i) Necessarily, for all objects w and z, if w phenomenally looks F to z,

then z’s visual experience of w has K.

296
(ii) For all kinds of location phenomenal character L, if L is not K, and if,

necessarily, for all w and z, if w phenomenally looks F to z, then z’s

visual experience of w has L, then K is more specific than L.

‘Specificity’: For all kinds of phenomenal character K and L, K is more specific than

L iff:

(i) Necessarily, for all visual experiences e, if e has K, then e has L.

(ii) It is not the case that, necessarily, for all experiences e, if e has L, then

e has K.

When there is, for example, a colour property F and a kind of colour phenomenal

character K which meet condition (1) of the correspondence principle, I will say that K

corresponds to F. Similarly when there is a location property F and a kind of location

phenomenal character K which meets condition (2) of the correspondence principle, I

will say that K corresponds to F.

The correspondence principle states that, for every colour and location property

that an object can phenomenally look to have, there are unique kinds of phenomenal

character that correspond to those properties.

If an object can phenomenally look red21, then the correspondence principle

entails that there is a unique kind of phenomenal character that corresponds to being

red21. We shall call this kind of phenomenal character red21-phenomenal character. In

297
general, for all F, if there is a unique kind of phenomenal character that corresponds to

being F, then that kind of phenomenal character will be called F-phenomenal character.

3 The Colour Phenomenal Character Problem

Suppose that one is looking at two patches, patch1 and patch2. Patch1

phenomenally looks red1 to one, and patch2 phenomenally looks red2 to one. One can

discriminate red1 from red2.

Suppose now that a dog is looking at patch1 and patch2, and the dog is unable to

discriminate the colour that patch1 phenomenally looks to it from the colour that patch2

phenomenally looks to it. What might the colour phenomenal character of the dog’s

visual experience be?

Let us introduce some terms.

S: The set of all the kinds of colour phenomenal character that one’s visual

experiences ever have.

S+: S together with all the kinds of phenomenal character that are between the

members of S.

298
S-: The set containing all the kinds of colour phenomenal character from red1-

phenomenal character to red2-phenomenal character inclusive.

Assuming that one’s visual experiences do have red1-phenomenal character and

red2-phenomenal character, S- is a subset of S+.

The description of S+ appeals to the notion of betweenness. Intuitively, some

colours are between other colours. For instance, it seems intuitive that red21 is between

red20 and red22. It is also intuitive that some kinds of colour phenomenal character are

between other kinds of colour phenomenal character. For instance, it is intuitive that

red21-phenomenal character is between red20-phenomenal character and red22-phenomenal

character.

One might be tempted by the following principle.

The Betweenness Principle: For all properties F, G and H, if there are kinds of

phenomenal character that correspond to being F,

being G and being H respectively, namely F-

phenomenal character, G-phenomenal character and

H-phenomenal character respectively, then F-

phenomenal character is between G-phenomenal

character and H-phenomenal character iff being F is

between being G and being H.

299
However, I will not rely on the betweenness principle. Later in this chapter I will

discuss a view on which the colour properties that objects phenomenally look to us to

have are disjunctions of properties. On this view, supposing that an object can

phenomenally look red2 to us, being red2 is a disjunction of other properties. It is not clear

whether there is an intuitive sense in which one disjunction of properties may be between

two others. Nevertheless, even if being red1, being red2 and being red3 are disjunctive

properties, say, it seems intuitive that there are betweenness relations between the kinds

of phenomenal character that correspond to these properties. The intuition that there are

such betweenness relations between these kinds of phenomenal character is independent

of the intuition that there are betweenness relations between the kinds of properties these

kinds of phenomenal character correspond to.

Later in this chapter I will define the relation of x being in between y and z in

terms of certain similarity relations. At this stage, however, I will rely on an intuitive

understanding of which kinds of phenomenal character are between which.

We can now formulate two hypotheses about the kinds of colour phenomenal

character that the dog’s visual experience has.

300
The Common Colour

Phenomenal Character View: The only kinds of colour phenomenal

character that the dog’s visual experience

has are members of S+.

The New Colour

Phenomenal Character View: The dog’s visual experience has a kind of

colour phenomenal character, red*-

phenomenal character, which is such that:

(i) it is not a member of S+.

(ii) it is more similar to any member of

S- than it is to any member of S+

that is not a member of S-.

There is an intuition, I think, that the new colour phenomenal character view is

worth exploring, and this chapter will be devoted to exploring this view.

4 Pointy and Gunky Colour Phenomenal Character

In this section I apply the notions of gunk and pointiness to colour phenomenal

character. I will argue that the new colour phenomenal character is plausible only if

colour phenomenal character is pointy.

301
I shall first apply the notions of gunk and pointiness to physical space.

Gunky space: Space is gunky iff every part of space has a proper part.

Pointy space: Space is pointy iff not every part of space has a proper part.

To understand the application of these notions to colour phenomenal character, we

need to understand the notion of a range of colour phenomenal character. Consider figure

1.

Figure 1

Figure 1 phenomenally looks to have a range of colour, and, correspondingly,

one’s visual experience of figure 1 has a range of colour phenomenal character.

According to one view, the pointy view of colour phenomenal character, the range

of colour phenomenal character that one’s visual experience of figure 1 has is composed

of kinds of colour phenomenal character that are not ranges, but which are homogeneous.

A homogeneous kind of colour phenomenal character is a kind of colour phenomenal

character that is not a range of colour phenomenal character.

302
A consequence of the pointy view is that, although one’s visual experience of

figure 1 as a whole has a range of colour phenomenal character, there are parts of figure

1, perhaps very small parts, such that one’s visual experiences of those parts have

homogeneous kinds of colour phenomenal character.

According to another view, the gunky view of colour phenomenal character, every

kind of colour phenomenal character is a range of colour phenomenal character. On this

view, there are no homogeneous kinds of colour phenomenal character; one’s visual

experience of any part, no matter how small, of figure 1, has a range of colour

phenomenal character. On this view, the range of colour phenomenal character that one’s

visual experience of figure 1 has overall is composed only of sub-ranges of colour

phenomenal character.

The definitions of these views are as follows.

Gunky colour phenomenal character: Colour phenomenal character is gunky iff

every kind of colour phenomenal character

is a range of colour phenomenal character.

Pointy colour phenomenal character: Colour phenomenal character is pointy iff

not every kind of colour phenomenal

character is a range of colour phenomenal

character.

303
Homogeneous colour

phenomenal character: A kind of colour phenomenal character is

homogeneous iff it is not a range of colour

phenomenal character.

Consider figure 2.

Figure 2

One might think that it is clear that colour phenomenal character is pointy. After

all, it does not seem that one’s visual experience of figure 2 has a range of colour

phenomenal character. It seems that one’s visual experience of figure 2 has a

homogeneous kind of colour phenomenal character.

A proponent of the gunky colour phenomenal character view will hold that this

intuition is not decisive against it. They will hold that there are ranges of colour

phenomenal character the beginnings and ends of which are so similar that one is not able

to discriminate them. That is, one will not be able to tell, of these ranges of colour

phenomenal character, that they are ranges of colour phenomenal character as opposed to

homogeneous kinds of colour phenomenal character. Given that one cannot discriminate

the beginnings from the ends of these ranges of colour phenomenal character, one is

304
liable to take them, incorrectly, to be homogeneous kinds of colour phenomenal

character.

A proponent of the gunky colour phenomenal character view will account for

one’s visual experience of figure 2 firstly by positing a range of colour phenomenal

character, say grey26-phenomenal character, the beginning and end of which one is not

able to discriminate, and secondly by holding either that one’s visual experience of figure

2 as a whole has grey26-colour phenomenal character, or that one’s visual experience of

parts of figure 2 have grey26-phenomenal character.

On the first hypothesis, on which one’s visual experience of figure 2 as a whole

has grey26-colour phenomenal character, one’s visual experience of the left half of figure

2 will have one half of the grey26-colour phenomenal character range, and one’s visual

experience of the right half of figure 2 will have the other half of the grey26-phenomenal

character range. On the second hypothesis, one’s visual experiences of parts of figure 2

have grey26-phenomenal character. Thus, there may be particular sub-regions of figure 2

of some area, say, 1mm2, such that one’s visual experiences of these sub-regions each

have grey26-phenomenal character. According to the gunky view, on either hypothesis we

are liable to think that our visual experience has a homogeneous kind of colour

phenomenal character even though it does not.

Suppose that colour phenomenal character is gunky. Then red*-phenomenal

character, red1-phenomenal character and red2-phenomenal character are ranges of colour

305
phenomenal character. What might red*-colour phenomenal character be? It could be the

range from red1-phenomenal character to red2-phenomenal character inclusive. However,

this range is in S+. After all, red0-phenomenal character and red3-phenomenal character

are both in S+, and intuitively the range from red1-phenomenal character to red2-

phenomenal character is between red0-phenomenal character and red3-phenomenal

character.

If red*-phenomenal character is not the range from red1-phenomenal character to

red2-phenomenal character, then it is hard to conceive what red*-phenomenal character

could be.

Suppose that colour phenomenal character is pointy. Suppose furthermore that

red*-phenomenal character, red1-phenomenal character, and red2-phenomenal character

are homogeneous kinds of colour phenomenal character. It seems easier to imagine that

there is a red*-phenomenal character on these assumptions. It seems that when one first

considers the new colour phenomenal character view, and considers it to be worth

exploring, one has in mind the view that the kinds of colour phenomenal character in

question are homogeneous kinds of phenomenal character.

The distinction between gunky colour phenomenal character and pointy colour

phenomenal character is fairly abstract, and one might wonder what the significance of it

is for the issues discussed in this chapter. The aim of this section has been to show that

the distinction is relevant to the issues discussed in this chapter. I have argued that the

306
new colour phenomenal character view is plausible only if colour phenomenal character

is pointy.

5 The Location Phenomenal Character Problem

I shall now argue that one can also draw a distinction between the common

location phenomenal character view and the new location phenomenal character view.

Suppose that one is looking at patch1 and patch2, and patch1 phenomenally looks to have

position property l1, and patch2 phenomenally looks to have position property l2. One’s

visual experience of patch1 has l1-phenomenal character, and one’s visual experience of

patch2 has l2-phenomenal character. Suppose that one can discriminate l1 from l2.

Suppose that the dog, looking at patch1 and patch2, cannot discriminate the

position property that patch1 phenomenally looks to it to have from the position property

that patch2 phenomenally looks to it to have. What kind of location phenomenal character

might the dog’s visual experience have?

Let us introduce some terms:

LS: The set of all the kinds of location phenomenal character that one’s visual

experiences ever have.

LS+: LS together with all the kinds of location phenomenal character that are between

the members of LS.

307
LS-: The set of all the kinds of location phenomenal character between l1-phenomenal

character and l2-phenomenal character.

We can now formulate two hypotheses:

The Common Location

Phenomenal Character View: The only kinds of location phenomenal character

that the dog’s visual experience has are members of

LS+.

The New Location

Phenomenal Character View: The dog’s visual experience has a kind of location

phenomenal character, l*-phenomenal character,

which is such that:

(i) it is not a member of LS+.

(ii) it is more similar to any member of LS- than

it is to any member of LS+ that is not a

member of LS-.

Let us apply the terms ‘gunky’ and ‘pointy’ to location phenomenal character. To

understand the application of these terms to location phenomenal character, we need to

understand the notion of a region of location phenomenal character. Consider again figure

2:

308
Figure 2

Different parts of figure 2 phenomenally look to have different position

properties. One’s visual experience of figure 2 has a region of location phenomenal

character.

According to one view, the pointy view of location phenomenal character, the

region of location phenomenal character that one’s visual experience of figure 2 has is

composed of kinds of location phenomenal character that are not regions of colour

phenomenal character, but rather are points of location phenomenal character; a point of

location phenomenal character is a kind of location phenomenal character that is not a

region of location phenomenal character. This view is analogous to the view on which a

region of physical space is composed of zero-dimensional spatial points.

According to another view, the gunky view of location phenomenal character, the

region of location phenomenal character that one’s visual experience of figure 2 has is

composed only of sub-regions of location phenomenal character. This view is analogous

to the view on which a region of physical space is composed only of sub-regions of space

that are non-zero-dimensional.

309
Gunky location phenomenal character: Location phenomenal character is gunky iff

every kind of location phenomenal character

is a region of location phenomenal character.

Pointy location phenomenal character: Location phenomenal character is pointy iff

not every kind of location phenomenal

character is a region of location phenomenal

character.

Point of location

phenomenal character: A point of location phenomenal character is

a kind of location phenomenal character that

is not a region of location phenomenal

character.

Suppose that location phenomenal character is gunky. Then l*-phenomenal

character, l1-phenomenal character and l2-phenomenal character are regions of location

phenomenal character. On this assumption, what might l*-phenomenal character be? It

could be the region consisting of l1-phenomenal character, l2-phenomenal character, and

every kind of location phenomenal character in between l1-phenomenal character and l2-

phenomenal character. However, it follows from our definition of LS+ that this region is

in LS+. On the assumption that location phenomenal character is gunky, it is hard to

conceive of what l*-phenomenal character might be.

310
Suppose that location phenomenal character is pointy. Suppose further that l*-

phenomenal character, l1-phenomenal character and l2-phenomenal character are points of

location phenomenal character. It is not obvious that we can conceive of an l*-

phenomenal character on these assumptions.

Thus, unlike the case of colour phenomenal character, the assumption that

location phenomenal character is pointy does not seem to make the new location

phenomenal character view more plausible.

5.1 The Apparent Asymmetry Between The Colour and Location

Phenomenal Character Problems

There are various reactions that we might have to the apparent asymmetry

between the colour phenomenal character problem and the location phenomenal character

problem.

One might argue that, if there are homogeneous kinds of colour phenomenal

character, then we can imagine such kinds of colour phenomenal character. By contrast, if

there are points of location phenomenal character, we cannot imagine points of location

phenomenal character, and this explains why we cannot imagine l*-phenomenal character

as a location phenomenal character point. According to this argument, if we could

311
imagine points of location phenomenal character, then we could imagine l*-phenomenal

character as a point of location phenomenal character.

A second reaction is that one doubts how plausible the new colour phenomenal

character view is, on the assumption that colour phenomenal character is pointy. I will

discuss this view below.

Thirdly, one might accept the asymmetry between the colour phenomenal

character problem and the location phenomenal character problem. That is, one accepts

that the new colour phenomenal character view is plausible on the assumption that colour

phenomenal character is pointy, and that the location phenomenal character view is not

plausible on the assumption that location phenomenal character is pointy.

If one was to develop the second reaction, one would try to explain away the

intuition that the new colour phenomenal character view is plausible. One might appeal to

the jumbled-up colour hypothesis. The terms used in the formulation of this hypothesis

will be explained below.

The Jumbled-Up Colour Hypothesis: When we conceive of red*-phenomenal

character, we are conceiving of an region of

location phenomenal character which is

occupied by a jumbled-up collection of

different kinds of colour phenomenal

312
character, such as red1-phenomenal

character, red56-phenomenal character, and

so on.

In order to explain the jumbled-up colour hypothesis, we need to introduce the

notion of a kind of colour phenomenal character occupying a kind of location

phenomenal character, and we need to define what it is for a kind of location phenomenal

character to be occupied by a jumbled-up collection of kinds of colour phenomenal

character.

I will start by introducing the notion of a kind of colour phenomenal character

occupying a kind of location phenomenal character.

Suppose that, at t1, the following facts obtain:

t1

(i) A particular cup, A, phenomenally looks red1 to one.

(ii) The background to A phenomenally looks white to one.

(iii) A phenomenally looks to have location l1 to one.

Between t1 and t2, A moves so that, at t2, the following facts obtain:

313
t2

(iv) A phenomenally looks red1 to one.

(v) The background to A phenomenally looks white to one.

(vi) A phenomenally looks to have location l2 to one.

Certainly one’s visual experience at t1 has a different kind of phenomenal

character from one’s visual experience at t2. However, both one’s visual experience at t1

and one’s visual experience at t2 have red1-phenomenal character. Furthermore, both one’s

visual experience at t1 and one’s visual experience at t2 have l1-phenomenal character and

l2-phenomenal character. After all, consider the part of the background that one sees at t1,

and that A moves into and occludes at t2. One’s visual experience, at t1, of this part of the

background has l2-phenomenal character. And one’s visual experience, at t2, of the part of

the background that, at t1, A occluded has l1-phenomenal character.

Hence, it is not possible to describe exhaustively the phenomenal character of a

visual experience by specifying simply the kind of colour phenomenal character and the

kind of location phenomenal character that the visual experience has. One needs to

introduce a relation of occupation between kinds of colour phenomenal character and

kinds of location phenomenal character. The difference between t1 and t2 is that, at t1, the

red1-phenomenal character of one’s visual experience occupies the l1-phenomenal

character of one’s visual experience, and at t2, the red1-phenomenal character of one’s

visual experience occupies the l2-phenomenal character of one’s visual experience.

314
The example we have just considered is a motivation for the occupation principle:

The Occupation Principle: Kinds of colour phenomenal character occupy kinds of

location phenomenal character.

We will define ‘jumbled-up’ as follows:

‘Jumbled-Up’: For all regions of location phenomenal character I, I is

occupied by a jumbled-up collection of kinds of colour

phenomenal character iff there are distinct sub-regions of I,

I1, I2, I3, and there are distinct kinds of colour phenomenal

character, C1, C2, C3, such that:

(1) I2 is between I1 and I3

(2) C2 is not between C1 and C3

(3) C1 occupies I1

(4) C2 occupies I2

(5) C3 occupies I3.

According to the jumbled-up colour hypothesis, when we take ourselves to

conceive red*-phenomenal character, in fact we are conceiving a region of location

phenomenal character that is occupied by a jumbled-up collection of kinds of colour

315
phenomenal character. For instance, we may be conceiving of a region of location

phenomenal character I, which has sub-intervals I1, I2 and I3, and which is such that: I2 is

between I1 and I3; I1 is occupied by red57-phenomenal character; I2 is occupied by red12-

phenomenal character, and I3 is occupied by red36-phenomenal character. This situation is

represented in figure 3:

Figure 3

Red57-phenomenal character Red12-phenomenal character Red36-phenomenal character

I1 I2 I3

A proponent of the jumbled-up colour hypothesis is likely to hold that, when we

take ourselves to conceive of red*-phenomenal character, we are conceiving of a region

of location phenomenal character whose sub-regions are occupied by very many different

kinds of colour phenomenal character, as opposed to the three kinds in figure 3.

In this chapter I wish to explore the new colour phenomenal character view, so I

will not be pursuing the jumbled-up colour hypothesis further.

Above it seemed that the assumption that location phenomenal character is pointy

did not seem to make it any easier to conceive of l*-phenomenal character. The first

reaction was that, despite this, l*-phenomenal character does exist. The second reaction

was that we should abandon the idea that red*-phenomenal character exists. The third

316
reaction was that we should acknowledge that red*-phenomenal character exists, but

deny that l*-phenomenal character exists.

In what follows I will remain neutral between the first reaction and the third

reaction. I will devote the rest of this chapter to exploring the new colour phenomenal

character view on the assumption that colour phenomenal character is pointy.

6 Responses To The Colour Phenomenal Character Problem

6.1 The Being Red* Principle

If the new colour phenomenal character view is correct, then, when the dog looks

at patch1 and patch2, its visual experience of both patches has red*-phenomenal character.

In what follows I will assume the following principle:

The Being Red* Principle: There is a unique property of being red* which

red*-phenomenal character corresponds to.

In the rest of this chapter I will explore what relationship holds between being

red* and being red1 and being red2. The aim is to find a relationship between being red*

and being red1 and being red2 that might explain the relationship that red*-phenomenal

character bears to red1-phenomenal character and red2-phenomenal character.

317
According to the asymmetric entailment view, being red* is entailed by either of

red1 or red2, but does not entail either of red1 or red2. According to the disjunctive view,

being red* is a disjunction of being red1 and being red2. According to the similarity view,

being red* bears a special relation of one-many similarity to being red1 and being red2.

These are the main views that I will discuss. However, before I discuss them, I will

discuss a fourth view, the identity view.

6.2 The Identity View

The Identity View: Being red* is identical with being red1 and with

being red2.

I discuss the identity view because it follows from the following two principles,

one of which we have argued for, and one of which many find plausible.

Necessary Colour Scepticism: Necessarily, objects do not have the colour

properties that they phenomenally look to have.

Necessary Coextension View: For all objects x and properties F and G, if,

necessarily, x is F iff x is G, then being F is

identical to being G.

318
According to necessary colour scepticism, being red*, being red1 and being red2

are necessarily coextensive. According to the necessary coextension view, necessarily

coextensive properties are identical.

The identity view, the being red* principle and the correspondence principle

together entail that the new colour phenomenal character view is false. The identity view

and the being red* principle together entail that red*-phenomenal character and red1-

phenomenal character correspond to the same property. The correspondence principle

entails that, for every property F that an object can phenomenally look to have, there is a

unique kind of phenomenal character that corresponds to being F. It follows that, if red*-

phenomenal character and red1-phenomenal character correspond to the same property,

then red*-phenomenal character is identical with red1-phenomenal character. This claim

is inconsistent with the new colour phenomenal character view.

We have independent reason to reject the necessary coextension view, and

therefore we have no reason to accept the identity view. Necessary colour scepticism and

the necessary coextension view together entail that being red1 is identical with being

green56. This claim entails the false conclusion that for all objects x and y, x

phenomenally looks red1 to y iff x phenomenally looks green56 to y. If the argument that

we gave in chapter 4 for necessary colour scepticism is sound, then we should give up the

necessary coextension view.

319
6.3 The Asymmetric Entailment View

The asymmetric entailment view is as follows:

Asymmetric Entailment View:

(i) Necessarily, for all x, if x is either red1 or

red2, then x is red*.

(ii) It is not the case that necessarily, if x is red*,

then x is red1, and it is not the case that

necessarily, if x is red*, then x is red2.

Necessary colour scepticism entails that (ii) is false. Necessary colour scepticism

entails that, necessarily, if an object is red*, then it is red1.

One might argue that, if necessary colour scepticism is correct, then the new

phenomenal character view is implausible in any case, and thus the question of what

relationship might hold between red* and being red1 and being red2 does not arise.

However, this seems incorrect. The independence intuition seems plausible:

The Independence Intuition: It is epistemically possible that both necessary

colour scepticism and the new colour phenomenal

character view are true.

320
The independence intuition is that the issue of whether or not the new colour

phenomenal character view is true is independent of the issue of whether or not objects

could have had the colour properties that they phenomenally look to have. If the

independence intuition is correct, then coming to believe necessary colour scepticism

should not affect our credence in the new colour phenomenal character view.

The objection above to the asymmetric entailment view depended on necessary

colour scepticism, which we defended in chapter 4. Necessary colour scepticism is

certainly a controversial thesis, and many readers may prefer to give up necessary colour

scepticism in order to defend the asymmetric entailment view. However, if the

independence intuition is correct, then, even if necessary colour scepticism is false, it

seems that we should reject the asymmetric entailment view. According to the

independence intuition, whether the new colour phenomenal character view is true is

independent of whether necessary colour scepticism is true. According to the asymmetric

entailment view, whether the new colour phenomenal character view is true is not

independent of whether necessary colour scepticism is true. Therefore, if the

independence intuition is correct, then the asymmetric entailment view makes the new

colour phenomenal character view depend on the wrong kind of issue, namely whether or

not necessary colour scepticism is true. This a reason to reject the asymmetric entailment

view, and furthermore, it is a reason that is independent of whether or not necessary

colour scepticism is in fact true.

321
One might argue that the relationship between being red* and being red1 and

being red2 is being more coarse-grained than. I will be assuming that being F is more

coarse-grained than being G iff being G entails being F, and being F does not entail being

G. Hence, I am assuming that the relation of being more coarse-grained than is captured

by the asymmetric entailment view. I called this chapter ‘Coarse-Grained Vision and New

Kinds of Phenomenal Character’ not because I think that the phenomenon I discuss is

correctly described in terms of the notion of coarseness of grain, but because I have

noticed that this phenomenon is naturally thought of by many philosophers in terms of

the notion of coarseness of grain.

6.4 The Disjunctive View

The disjunctive view is an attempt to develop the new colour phenomenal

character view in the light of necessary colour scepticism.

The Disjunctive View: Being red* is a disjunction of being red1 and being red2.

Necessary colour scepticism entails that being red1 or red2 is necessarily

coextensive with being red1. However, having rejected the necessary coextension view,

we are not under pressure to identify the disjunctive property of being red1 or red2 with

being red1.

322
Later I will discuss the disjunctive view in more detail. I will briefly discuss a

further view.

6.5 The Determinable View

The determinable view is as follows:

The Determinable View: Being red* is a determinable of the properties being red1

and being red2.

According to one view, being F is a determinable of being G only if being G

entails being F, and being F does not entail being G. On this view, the determinable view

would entail the asymmetric entailment view, which we have argued is false.

According to another view, being F is a determinable of being G only if being F is

a disjunction, one of whose disjuncts is being G. On this view, the determinable view

entails the disjunctive view, which we shall consider below.

On another view, the relation of being a determinable of is primitive. This is a live

option. However, before we consider taking it, it is worth exploring other responses to the

colour phenomenal character problem. Before discussing the disjunctive view, I will

discuss Dummett’s argument in ‘Wang’s Paradox’ (Dummett 1978).

323
7 Wang’s Paradox

There is some inclination to think that one of the arguments that Dummett

develops in ‘Wang’s Paradox’ bears on the question whether the new colour phenomenal

character view is correct (Dummett 1978). In this section I will argue that Dummett’s

argument in fact addresses a different issue.

Dummett is considering an example in which one is looking at the minute-hand of

a clock. He writes as follows:

‘For let us suppose that space and time are continua, and let us
change the example so that the minute-hand now moves at a uniform rate.
Let us further suppose that whether or not the minute-hand occupies
discriminably different positions at different moments depends uniformly
upon whether or not the angle made by the two positions of the minute-
hand is greater than a certain minimum. It will then follow that, however
gross our perception of the position of the minute-hand may be, there is a
continuum of distinct phenomenal positions for the minute-hand: for, for
any two distinct physical positions of the minute-hand, even if they are not
discriminably different, there will be a third physical position which is
discriminably different from the one but not from the other.’ (Dummett,
1978, p266-267).

Let us express Dummett’s argument as follows.

1.) For all objects x and y, there is a distance D such that whether or not a given

subject S can discriminate x’s position from y’s position depends uniformly on

whether or not x is more than D from y.

324
2.) For all objects x and y, if x and y are in distinct positions, then there is an object z

such that z is D from y and z is more than D from x.

3.) For all objects x and y, if x and y are in distinct positions, then there is an object z

such that S can discriminate x’s position from z’s position and S cannot

discriminate y’s position from z’s position.

4.) For all objects x and y, if x and y are in distinct positions, then if x and y

phenomenally look to S to have position properties, then the position properties

that x and y phenomenally look to have are distinct.

This argument has a very striking conclusion: that however small a distance that

an object moves between t1 and t2, the position property that the object phenomenally

looks to have at t1 is different from the position property that the object phenomenally

looks to have at t2.

There is an analogue of Dummett’s argument for colour properties. 1.)* is as

plausible as 1.):

1.)* For all objects x and y, there is a determinable property N and a distance D

between the determinates of N such that whether or not a given subject S can

discriminate the colour x phenomenally looks from the colour y phenomenally

looks depends uniformly on whether or not x’s determinate of N is more than D

from y’s determinate of N.

325
N might be the property of reflecting some wavelength of light, and the distance

between the determinates of this determinable will be the difference between wavelengths

of light. A parallel argument to the one above will show that however small the difference

between x’s determinate of N and y’s determinate of N, the colour properties that x and y

will phenomenally look to have are distinct.

Since we are discussing the new colour phenomenal character view in this

chapter, I will discuss the analogue of Dummett’s argument for colour properties. If this

analogue of Dummett’s argument is sound, then there can be arbitrarily small differences

between the colour properties that objects phenomenally look to us to have. A small

extension to the argument would establish that there can be arbitrarily small differences

between the kinds of colour phenomenal character that two visual experiences of a given

subject have.

One might think that this latter consequence counts against the new colour

phenomenal character view. However, I shall argue that the new colour phenomenal

character view is consistent with this consequence.

Suppose that the disjunctive view is correct, and that patch1 phenomenally looks

red1 or red2 to the dog. Consider the property of being red1` or red2`, where red1` is

arbitrarily similar to red1, and red2` is arbitrarily similar to red2`. It may be that some third

patch, patch3, phenomenally looks red1` or red2` to the dog. If patch3 phenomenally looks

red1` or red2` to the dog, then it seems plausible that the dog’s visual experience of patch3

326
would have a kind of colour phenomenal character that is distinct from, but arbitrarily

similar to, red*-phenomenal character.

This example shows that there is no inconsistency between the new kinds of

colour phenomenal character that the new colour phenomenal character view postulates,

and arbitrarily small differences between those new kinds of colour phenomenal

character. That is, the existence of arbitrarily small differences between kinds of colour

phenomenal character does not support the common colour phenomenal character view

over the new colour phenomenal character view. Therefore Dummett’s argument in

‘Wang’s Paradox’ does not bear on the question whether or not the new colour

phenomenal character view is correct.

8 The Disjunctive View Discussed

In this section I explore the disjunctive view. To begin with, I consider two

arguments that objects do not phenomenally look to have disjunctive properties.

One might argue that objects do not phenomenally look to have disjunctive

properties on the grounds that, when, for some object A and some properties F and G, we

say (1), what we mean is (2).

(1) A phenomenally looks F or G

(2) Either A phenomenally looks F or A phenomenally looks G.

327
According to this argument, an object never phenomenally looks F or G, where ‘F

or G’ is within the scope of ‘phenomenally looks’.

It may be the case that whenever we say that an object phenomenally looks F or

G, what we mean is that either the object phenomenally looks F or it phenomenally looks

G. But it does not follow from this that objects do not phenomenally look to have

disjunctive properties.

It may be the case that whenever we say that some object phenomenally looks my

favourite colour, what we mean is that the colour that the object phenomenally looks is

my favourite colour. But this fact by itself does not show that objects do not

phenomenally look my favourite colour, where ‘my favourite colour’ is within the scope

of the ‘phenomenally looks’. More argument would be required to show that objects

cannot phenomenally look to have the property of being my favourite colour.

Some have argued that there is a conflict between the claim that the properties

that objects phenomenally look to have vary gradually and the claim that objects

phenomenally look to have disjunctive properties. According to this argument,

disjunctive properties do not vary gradually.

328
It is not clear why one might think that disjunctive properties cannot vary

gradually. In the section above on Wang’s paradox, it seemed plausible that disjunctive

properties could vary gradually.

8.1 Phenomenal Character

The disjunctive view was offered as an explanation of the new colour phenomenal

character view. The proposal was that the explanation of the dog’s visual experience of

patch1 having red*-phenomenal character was that patch1 phenomenally looks red1 or red2

to the dog. In this section I consider whether the disjunctive view is an adequate

explanation of the dog’s visual experience of patch1 having red*-phenomenal character.

If patch1 phenomenally looks red1 or red2 to the dog, then the colour phenomenal

character of the dog’s visual experience may be either compositional or non-

compositional. Compositional phenomenal character is defined as follows.

Phenomenal Compositionality: For all x and y, x’s visual experience of y has a

compositional phenomenal character iff for all

properties F, if F is either a property, or a

constituent of a property, that y phenomenally looks

to x to have, then x’s visual experience has F-

phenomenal character.

329
I mean the notion of a constituent of a property as follows: being F is a constituent

of the properties being F or G, being F and G, being not F, being F if G, and so on.

If patch1 phenomenally looks red1 or red2 to the dog, and if the dog’s visual

experience of patch1 has compositional phenomenal character, then the dog’s visual

experience of patch1 will have both red1-phenomenal character and red2-phenomenal

character.

Suppose that patch1 is the only object one sees, and that every facing part of it

phenomenally looks red1 to one. One’s visual experience of patch1 will have red1-

phenomenal character. Intuitively, red1-phenomenal character will be the only kind of

colour phenomenal character that one’s visual experience has. That one’s visual

experience has red1-phenomenal character seems to exhaust what it is like colour-wise for

one to have the experience. In such circumstances, it does not seem that one’s visual

experience could also have red2-phenomenal character.

Suppose that patch1 is the only object that the dog sees, and that every part of

patch1 facing the dog phenomenally looks red1 or red2 to the dog. The combination of the

disjunctive view and the commitment to phenomenal compositionality entails that the

dog’s visual experience of patch1 has both red1-phenomenal character and red2-

phenomenal character. It seems that we have good reason to reject the combination of the

disjunctive view and the commitment to phenomenal compositionality.

330
A defender of the disjunctive view is likely to argue that, when patch1

phenomenally looks red1 or red2 to the dog, then the dog’s visual experience of patch1 has

non-compositional phenomenal character. Let us define the non-compositional view as

follows:

The Non-Compositional View: Necessarily, for all objects x and y, and all

properties F and G, if x phenomenally looks F or G

to y, then y’s visual experience of x has non-

compositional phenomenal character.

Non-compositional phenomenal character is phenomenal character that is not

compositional. There is a question as to what might explain the non-compositional view.

After all, phenomenal compositionality seems to hold when the property that an object

phenomenally looks to have is a conjunctive property. That is, the principle of

conjunction compositionality seems very plausible.

Conjunction Compositionality: Necessarily, for all objects x and y, and properties F

and G, if x phenomenally looks F and G to y, then

y’s visual experience of x has F-phenomenal

character and G-phenomenal character.

331
For instance, if an object x phenomenally looks red1 and at position l1 to y, then it

is intuitive that y’s visual experience of x has both red1-phenomenal character and l1-

phenomenal character.

One could explain the non-compositional view by arguing for the following

principle:

The Neutrality Principle: Necessarily, for all objects x and y, and all properties F and

G, if one of the properties that x phenomenally looks to y to

have is neutral between being F and being G, then y’s

visual experience has non-compositional phenomenal

character.

One definition of neutrality is the following:

The Entailment View of Neutrality: For all properties F, G and H, being F is

neutral between being G and being H iff for

all objects x,

(i) necessarily, if x is either G or H, then

x is F;

(ii) it is not the case that, necessarily, if x

is F, then x is G;

332
(iii) it is not the case that necessarily, if x

is F, then x is H.

Since the properties in question are colour properties, and since the disjunctive

view is intended to be consistent with necessary colour scepticism, I assume that a

proponent of the neutrality principle would not accept the entailment view of neutrality. I

assume that they will take it to be a primitive fact that the disjunctive property of being F

or G is neutral between being F and being G. Thus the property of being either such that 2

+ 2 = 5 or such that 2 + 2 = 6 will count as neutral between the property of being such

that 2 + 2 = 5 and the property of being such that 2 + 2 = 6.

It seems that the following principle is plausible:

The Disjunction Principle: Necessarily, for all properties F and G, if an object can

phenomenally look F or G, then

(i) there are properties H and I such that an object can

phenomenally look H and I.

(ii) an object can phenomenally look not F

(iii) an object can phenomenally look not both not F and

not G

(iv) an object can phenomenally look F if G.

333
The argument for the disjunction principle is that, for all F and G, if it is possible

for an object phenomenally to look F or G, then it would be ad hoc to deny any of (i) to

(iv) above. If an object can phenomenally look F or G, and some of (i) to (iv) above are

false, then some special reason for this would have to be given.

For instance, a proponent of the view that, for some F and G, objects can

phenomenally look F or G, may appeal to ordinary language in support of their view. A

proponent of this view might hold that we say such sentences as ‘he looks either

courageous or foolish to me’. However, we also say such sentences as ‘she looks tall and

brown-haired to me’, ‘she looks not interested’ and ‘the policy looks fine if there are no

objections from the backbenchers’. It also seems that there will be some context in which

one might say, for some x and for some F and G, that x looks not both not F and not G.

Thus, if ordinary language is intended to support the view that, for some F and G, objects

can phenomenally look F or G, then it seems that proponents of the view should endorse

(i) to (iv) above too.

I will now argue that the combination of the neutrality principle and the

disjunction principle has two counter-intuitive consequences. My arguments will employ

assumptions such as that, for some x and some F and G, x phenomenally looks not F, and

F if G, and not both not F and not G. I do not take a stand on whether these assumptions

are true. My claim is that these assumptions follow from the disjunctive view, and thus it

is legitimate to make them when assessing the disjunctive view.

334
8.1.1 Multiple Kinds of Colour Phenomenal Character like Red*-phenomenal
character

A large part of the motivation for the non-compositional view is that we do have

some idea of a kind of colour phenomenal character like red*-phenomenal character. That

is, we do have an idea of a kind of colour phenomenal character that satisfies the

conditions in the new colour phenomenal character view, which is, recall, as follows:

The New Colour

Phenomenal Character View: The dog’s visual experience has a kind of

colour phenomenal character, red*-

phenomenal character, which is such that:

(i) it is not a member of S+.

(ii) it is more similar to any member of

S- than it is to any member of S+

that is not a member of S-.

In this section I argue that the disjunctive view, together with the neutrality

principle, the disjunction principle and a further plausible principle together entail that

there are multiple kinds of colour phenomenal character that satisfy conditions (i) and

(ii). This is a counter-intuitive consequence, and is therefore a cost to accepting the

disjunctive view.

335
According to DeMorgan’s law, being F or G is equivalent to not being both not F

and not G. Thus, it seems that if being F or G is neutral between being F and being G,

then not being both not F and not G is also neutral between being F and being G. The

disjunction principle entails that, if an object can phenomenally look F or G, then an

object can phenomenally look not both not F and not G. The neutrality principle entails

that, if an object O phenomenally looks not both not F and not G to S, then S’s visual

experience of O has non-compositional phenomenal character.

One might think that, if object O phenomenally looks F or G to S1, and O

phenomenally looks not both not F and not G to S2, then the non-compositional

phenomenal character of S1’s visual experience is the same as the non-compositional

phenomenal character of S2’s visual experience. However, this possibility is ruled out by

the following principle:

The Phenomenal Distinctness Principle: Necessarily, for all objects x, y and z, and

for all distinct properties F and G, if x can

phenomenally look F to z and y can

phenomenally look G to z, then the

phenomenal character that corresponds to

being F is distinct from the phenomenal

character that corresponds to being G.

336
One might wonder what the relationship is between the phenomenal distinctness

principle and the correspondence principle. A consequence of the correspondence

principle is that, for every property that an object can phenomenally look to have, there is

a unique kind of phenomenal character that corresponds to that property. We could define

the reverse correspondence principle as follows: for every kind of phenomenal character,

there is a unique property that that kind of phenomenal character corresponds to. The

reverse correspondence principle is similar to, though not equivalent to, the phenomenal

distinctness principle. The phenomenal distinctness principle allows that there may be

kinds of phenomenal character that correspond to no property that an object could

phenomenally look to have, whereas the reverse correspondence principle does not allow

this.

The argument for the phenomenal distinctness principle is that it holds for the

standard properties that we think objects phenomenally look to have. For instance, if x

phenomenally looks red1 to z, and y phenomenally looks red2 to z, then it is intuitive that

the phenomenal character of z’s visual experience of x will differ from the phenomenal

character of z’s visual experience of y.

One might argue that, given that, in section 6.2, we abandoned the necessary

coextension view, and thus committed ourselves to a fine-grained way of individuating

properties, the phenomenal distinctness principle is not plausible. This objection rests on

the following principle:

337
The Anti-Phenomenal Distinctness

Principle: Necessarily, for all objects x and y, and all

properties F and G, if being F is necessarily

coextensive with being G, and if x can

phenomenally look F to y, and x can

phenomenally look G to y, then the same

kind of phenomenal character corresponds

to being F and being G.

If object O1 phenomenally looks F or G to S, and object O2 phenomenally looks

not both not F and not G to S, then the anti-phenomenal distinctness principle entails that

the kind of phenomenal character that corresponds to being F or G is identical with the

kind of phenomenal character that corresponds with being not both not F and not G.

However, necessary colour scepticism gives us a reason to reject the anti-

phenomenal distinctness principle. Necessary colour scepticism entails that being red1 is

necessarily coextensive with being green1. However, the kind of colour phenomenal

character that corresponds to being red1 is distinct from the kind of colour phenomenal

character that corresponds to being green1. Therefore, if necessary colour scepticism is

true, one cannot object to the phenomenal distinctness principle by appealing to the anti-

phenomenal distinctness principle.

338
The main premises of the first argument against the disjunctive view are as

follows:

1.) By the disjunction principle, if an object can phenomenally look F or G, then an

object can phenomenally look not both not F and not G.

2.) By the neutrality principle,

a. if an object x phenomenally looks F or G to a subject S, then S’s visual

experience of x has non-compositional phenomenal character.

b. If an object x phenomenally looks not both not F and not G to S, then S’s

visual experience of x has non-compositional phenomenal character.

3.) By the phenomenal distinctness principle, the phenomenal character that

corresponds to being F or G is distinct from the phenomenal character that

corresponds to being not both not F and not G.

Let us assume that the non-compositional phenomenal character that corresponds

to the property of being red1 or red2 is red*-phenomenal character. Let us call the non-

compositional phenomenal character that corresponds to the property of being not both

not red1 and not red2 red**-phenomenal character.

Red*-phenomenal character was introduced as the kind of phenomenal character

that satisfies conditions (i) and (ii) in the new colour phenomenal character view above.

If red**-phenomenal character exists, it seems plausible that it also satisfies (i) and (ii) in

339
the new colour phenomenal character view. It is not clear what reason there would be for

denying this claim.

This establishes that the disjunctive view is true only if there are at least two kinds

of colour phenomenal character that satisfy conditions (i) and (ii) in the new colour

phenomenal character view. Since we do not intuitively think that there are two such

kinds of colour phenomenal character, this raises doubts about the adequacy of the

disjunctive view in explaining the new colour phenomenal character view.

We can also establish that the disjunctive view is true only if there is a third kind

of colour phenomenal character, red***-phenomenal character, which satisfies conditions

(i) and (ii).

Being F or G is equivalent to being F if not G. Hence, if being F or G is neutral

between being F and being G, then, intuitively, being F if not G is neutral between being

F and being G. By the disjunction principle, an object can phenomenally look F or G only

if an object can phenomenally look F if not G. By an argument similar to 1.) to 3.) above

we can establish that there is a kind of colour phenomenal character that corresponds to

being F if not G that is a), distinct from red*-phenomenal character and red**-

phenomenal character, and b), non-compositional. We can call this kind of phenomenal

character red***-phenomenal character, and, if it exists, it seems plausible that it would

satisfy conditions (i) and (ii) in the new colour phenomenal character view.

340
The fact that the disjunctive view is true only if there are at least three kinds of

colour phenomenal character that satisfy conditions (i) and (ii) in the new colour

phenomenal character view intuitively seems to be a cost of the disjunctive view.

8.1.2 Phenomenally Looking not F

Consider these two principles:

Conjunction Compositionality: Necessarily, for all objects x and y, and properties F

and G, if x phenomenally looks F and G to y, then

y’s visual experience of x has F-phenomenal

character and G-phenomenal character.

Negation Compositionality: Necessarily, for all objects x and y, and all

properties F, if x phenomenally looks not F to y,

then y’s visual experience of x has F-phenomenal

character.

Conjunction compositionality and negation compositionality together entail

negated conjunction compositionality:

Negated Conjunction Compositionality: Necessarily, for all objects x and y, and all

properties F and G, if x phenomenally looks

341
not both not F and not G to y, then y’s visual

experience of x has both F-phenomenal

character and G-phenomenal character.

If negated conjunction compositionality is true, then it is possible that an object x

phenomenally looks not both not F and not G to some subject S, and for S’s visual

experience of x to have compositional phenomenal character. This is not consistent with

the neutrality principle, reproduced below, given that not being both not F and not G is

neutral between being F and being G.

The Neutrality Principle: Necessarily, for all objects x and y, and all properties F and

G, if one of the properties that x phenomenally looks to y to

have is neutral between being F and being G, then y’s

visual experience has non-compositional phenomenal

character.

The neutrality principle was offered as an explanation of the non-compositional

view. Thus if it is false an alternative explanation of the non-compositional view is

required.

Conjunction compositionality seems very plausible. It seems plausible that if an

object x phenomenally looks red1 and at position l1 to subject S, then S’s visual

experience of x has both red1-phenomenal character and l1-phenomenal character.

342
It seems more likely that a defender of the disjunctive view will reject negation

compositionality. There is a minor cost to doing this. Recall the non-compositional view.

The Non-Compositional View: Necessarily, for all objects x and y, and all

properties F and G, if x phenomenally looks F or G

to y, then y’s visual experience of x has non-

compositional phenomenal character.

The non-compositional view did not strike us as implausible as we already had a

pretheoretic idea of a non-compositional phenomenal character, namely red*-phenomenal

character. Thus the non-compositional view fitted with our pretheoretic intuitions.

However, in giving up negation compositionality, the disjunctive view is committed to

holding that, when an object x phenomenally looks not red1 to a subject S, then S’s visual

experience of x has non-compositional phenomenal character, which we can call not-

red1-phenomenal character. There is no pretheoretic support for the existence of not-red1-

phenomenal character, and indeed it is hard to conceive what it might be. The fact that the

disjunctive view is committed to the existence of not-red1-phenomenal character is thus is

a cost of the disjunctive view.

In this section I have argued for two claims. Firstly, I argued that the disjunctive

view is plausible only if the non-compositional view is correct. Secondly, I argued that

there are some costs of accepting the non-compositional view.

343
9 The Similarity View

In the last few sections we have assumed the being red* principle, reproduced

below, and we have been considering what relation might hold between being red* and

being red1 and being red2.

The Being Red* Principle: There is a unique property of being red* which

red*-phenomenal character corresponds to.

In this section I discuss the similarity view, which is as follows.

The Similarity View: Being red* bears the relation of one-many similarity to

being red1 and being red2.

The aim of this section is to define the relation of one-many similarity, and to

explore the similarity view.

By appealing to similarity relations between being red*, being red2 and being red2,

the similarity view mirrors the way in which new colour phenomenal character view

explains the relation between red*-phenomenal character, red1-phenomenal character and

red2-phenomenal character.

344
The New Colour

Phenomenal Character View: The dog’s visual experience has a kind of

colour phenomenal character, red*-

phenomenal character, which is such that:

(i) it is not a member of S+.

(ii) it is more similar to any member of

S- than it is to any member of S+

that is not a member of S-.

The relevant sets are as follows:

S: The set of all the kinds of colour phenomenal character that one’s visual

experiences ever have.

S+: S together with all the kinds of phenomenal character that are between the

members of S.

S-: The set containing all the kinds of colour phenomenal character from red1-

phenomenal character to red2-phenomenal character inclusive.

The central idea in the new colour phenomenal character view is that red*-

phenomenal character is not a member of S+, but that it is more similar to red1-

phenomenal character, red2-phenomenal character, and all the kinds of phenomenal

character between red1-phenomenal character and red2-phenomenal character, than it is to

any other kind of colour phenomenal character in S+.

345
Just as red*-phenomenal character is introduced in terms of the similarity

relations that it bears to red1-phenomenal character and to red2-phenomenal character, the

similarity view explains the relation between being red* and being red1 and being red2

also in terms of a special kind of similarity relation, the one-many similarity relation.

9.1 The One-Many Similarity Relation

The one-many similarity relation is as follows:

1.) x is one-many similar to y and z iff

a. x is equally similar to y and to z.

b. there is no w such that

i. w is between y and z

ii. x is more similar to w than it is to either of y or z.

c. x is distinct from y and z.

The simplest way of defining ‘betweenness’ involves assuming that colour space,

i.e. the set of colour properties, is a metric space. This assumption is as follows. I assume

that, where x and y are colour properties, d(x,y) is the degree of dissimilarity between x

and y.

346
The Metric Space Assumption: For all colour properties x, y and z:

(i) d(x,y) ≥ 0.

(ii) d(x,y) = 0 iff x = y.

(iii) d(x,y) = d(y,x).

(iv) d(x,z) ≤ d(x,y) + d(y,z).

The metric space assumption holds that there are degrees of dissimilarity between

colour properties that can be added together. If the metric assumption is correct, then we

can define ‘betweenness’ as follows:

2.) x is between y and z iff d(x,y) + d(x,z) = d(y,z).

‘x is equally similar to y and to z’ is defined as follows:

3.) x is equally similar to y and to z iff d(x,y) = d(x,z).

‘x is more similar to y than to z’ is defined as follows:

4.) x is more similar to y than to z iff d(x,y) < d(x,z).

A formal model of the one-many similarity relation is as follows. Consider a set

containing three elements, a, b and c. Let us stipulate the following to be true:

347
(i) d(a,b) = 1

(ii) d(a,c) = 1

(iii) d(b,c) = 0.5

In this model, a is one-many similar to b and c:

(i) a is equally similar to b and to c.

(ii) There is no x in between b and c which is such that a is more similar to x

than it is to either of b and c.

(iii) a is distinct from b and c.

The similarity view holds that being red* is one-many similar to being red1 and to

being red2. Figure 4 below is a representation of a set of colour properties, where colour

properties at a higher level are one-many similar to a set of colour properties at a lower

level. Being red1 is one-many similar to being red1.1, being red1.2 and being red1.3. Figure 4

is an abstract representation of certain colour properties. Although we have represented

these colour properties using lines, this should not be taken to imply anything about the

metaphysical nature of the colour properties. In particular, the figure should not be taken

to suggest that the colour properties are ranges of colour in the way that, according to the

gunky view of colour phenomenal character, kinds of colour phenomenal character are

ranges of colour phenomenal character.

Quite possibly there is a densely ordered set of colour properties between being

red1.1 and being red1.3 to which being red1 is one-many similar. Being red1.3 is one-many

348
similar to being red1.31, being red1.32 and being red1.33. Being red1.1 and being red1.2 are also

one-many similar to various colour properties, but these colour properties are not

represented in figure 4.

Figure 4

Red1
Red1.1
Red1.2
Red1.3 Red1.31
Red1.32
Red1.33
Red1.331
Red1.332
Red1.333
Red1.3331
Red1.3332
Red1.3333
Red1.33331
Red1.33332

…….

9.2 The One-Many Similarity Relation Involving Non-colour Properties

It seems that the following principle is plausible:

Negation Invariance: Necessarily, for all properties F, G and H, being F is more

similar to being G than to being H iff being not F is more

similar to being not G than to being not H.

349
Negation invariance entails that being red* is one-many similar to being red1 and

being red2 iff being not red* is one-many similar to being not red1 and being not red2.

The similarity view was advanced as an explanation of the new colour

phenomenal character view. Suppose that objects could phenomenally look to have

properties such as being not red*, being not red1, and being not red2. It might be thought a

consequence of that supposition that not-red*-phenomenal character bears a similar

relationship to not-red1-phenomenal character and not-red2-phenomenal character as

red*-phenomenal character does to red1-phenomenal character and to red2-phenomenal

character.

It is not clear that, given the supposition that objects can phenomenally look to

have properties such as being not red*, that the above consequence of this supposition is

a cost of the similarity view. Furthermore, it is not clear whether this supposition is

indeed correct. Thus, without further argument, we should not take negation invariance to

present a problem for the similarity view.

It seems that being red1 is one-many similar to having position l1 and having

position l2 for the following reason:

(i) being red1 is equally similar to having l1 and having l2;

(ii) there is no w such that

i. w is between having l1 and having l2

350
ii. being red1 is more similar to w than to either of having l1 or having

l2.

(iii) being red1 is distinct from having l1 and having l2.

That being red1 is one-many similar to being at l1 and being at l2 does not seem to

be a cost of the similarity view. The similarity view holds only that being red* is one-

many similar to being red1 and being red2, and that this fact explains why the new colour

phenomenal character view is correct. This explanatory project does not seem to be

jeopardized by the fact that being red1 is one-many similar to being at l1 and being at l2.

9.3 Vertical/Horizontal Relations

In this section I will argue for the following constraint on the relation of one-

many similarity.

The Horizontal/Vertical Principle: Necessarily, for all x, y and z, if x is one-

many similar to y and z, then

d(y,z) < 2d(x,y).

The proof of this principle is as follows. If x is one-many similar to y and z, then

x is not between y and z. This follows from condition 1.)b of the definition of one-many

similarity. Suppose that the following are true:

351
(i) d(x,y) = 1

(ii) d(x,z) = 1

(iii) d(y,z) = 2

It follows from (i)-(iii), together with the definition of betweenness above, that x

is between y and z. Furthermore, it is not possible that (i), (ii) and (iv) be true:

(iv) d(y,z) > 2

This establishes the horizontal/vertical principle. The principle above is called

‘the horizontal/vertical principle’ because when x is one-many similar to y and z, it can

help to think of the similarity relations between y and z as horizontal relations, and the

relations between x and y, and between x and z, as vertical relations. The

horizontal/vertical principle imposes a constraint on the relation between horizontal

similarity relations and vertical similarity relations.

9.4 Weaker Starting Assumptions

We stated that the similarity view is as follows:

352
The Similarity View: Being red* bears the relation of one-many similarity to

being red1 and being red2.

We defined ‘one-many similarity as follows:

2.) x is one-many similar to y and z iff

b. x is equally similar to y and to z.

c. there is no w such that

i. w is between y and z

ii. x is more similar to w than it is to either of y or z.

d. x is distinct from y and z.

We made the metric space assumption, the assumption that colour space is a

metric space, and defined ‘betweenness’ as follows:

2.) x is between y and z iff d(x,y) + d(x,z) = d(y,z).

If colour space is not a metric space, then we would have to reject the similarity

view. However, we could accept a view similar to the similarity view, namely the

comparative similarity view:

The Comparative Similarity View: Being red* is comparatively one-many

similar to being red1 and being red2.

353
‘Comparative one-many similarity’ is defined in the same way as ‘one-many

similarity’, except that ‘betweenness’ is replaced with ‘comparative betweenness’:

5.) x is comparatively one-many similar to y and z iff

a. x is equally similar to y and to z.

b. there is no w such that:

i. w is comparatively between y and z

ii. x is more similar to w than it is to either of y or z.

c. x is distinct from y and z.

‘Equal similarity’ and ‘comparative betweenness’ are defined in terms of the

three-place similarity relation of x being more similar to y than to z:

6.) x is comparatively between y and z iff

a. y is more similar to x than it is to z.

b. z is more similar to x than it is to y.

c. There is no w such that

either:

i. w is as similar to y as x is to y, and w is more similar to z than x is

to z.

or

354
ii. w is as similar to z as x is to z, and w is more similar to y than x is

to y.

‘x is as similar to y as it is to z’ is defined as follows:

7.) x is as similar to y as it is to z iff

a. x is not more similar to y than it is to z.

b. x is not more similar to z than it is to y.

Hopefully 7.) is plausible and does not stand in need of much explanation. For

instance, in figures 5 and 6, x is as similar to y as it is to z:

Figure 5

y x z

Figure 6

y z

355
The reasoning behind conditions 6.)a and 6.)b is as follows. If we consider figure

5, it is clear that a constraint on x being between y and z is that z is closer to x than it is to

y, and y is closer to x than it is to z.

The reasoning behind condition 6.)c is as follows. Consider the following case:

Figure 7

y w z

In this case, conditions 6.)a and 6.)b are met, but in no intuitive sense of ‘between’

is x between y and z. What ensures that x is not comparatively between y and z is

condition 6.)c. There is a w such that w is as similar to z as x is to z, but which is more

similar to y than x is to y.

Betweenness, as defined by 2.), and comparative betweenness come apart in two

ways. It follows from 2.) that for all x and y, x is between x and y. However, it follows

from 6.) that, for all x and y, x is not between x and y. 6.)a blocks y from being identical

with z, and 6.)b blocks y from being identical with x. This difference, however, between

betweenness and comparative betweenness is small; it can be removed by adding a

condition on betweenness, as defined by 2.), that x be distinct from y and z.

356
Betweenness and comparative betweenness come apart if colour space is metric

and if it contains gaps. A gap in colour space is defined as follows:

8.) There is a gap in colour space iff there are two colours x and y such

that:

a. d(x,y) > 0

b. there is no colour z such that:

i. z is distinct from x and y.

ii. z is between x and y.

Recall figure 7:

Figure 7

y w z

Suppose that colour space is metric, and that w does not exist. This situation is

represented in figure 8:

Figure 8

y z

357
Conditions 6.)a and 6.)b are met. Furthermore, we may assume that there are

enough gaps so that condition 6.)c is met. Therefore, although x is not between y and z, x

will be comparatively between y and z.

This is a counter-intuitive instance of comparative betweenness: comparative

betweenness applies when no intuitive notion of betweenness applies. One might take this

to be a problem for our appeal to the notion of comparative betweenness. However, we

appeal to comparative betweenness only on the assumption that colour space is not a

metric space. If colour space is not a metric space, then there will be no gaps as we have

defined them, and therefore problematic instances of comparative betweenness of the

above kind will not arise. If colour space is a metric space, then we will not appeal to

comparative betweenness, as we will have no need for it, but rather to betweenness, as

defined by 2.).

The horizontal/vertical principle, reproduced below, can still be defended even if

colour space is not a metric space.

The Horizontal/Vertical Principle: Necessarily, for all x, y and z, if x is one-

many similar to y and z, then

d(y,z) < 2d(x,y).

We will understand ‘d(y,z) < 2d(x,y)’ as ‘y is less than twice as dissimilar from z

as x is from y’.

358
Before defining ‘x is less than twice as dissimilar from z as x is from y’, we will

define ‘x is twice as dissimilar from y as z is from y’:

9.) x is twice as dissimilar from y as z is from y iff there is a w such that

a. w is as similar to y as z is to y.

b. w is between x and y

c. w is as similar to x as it is to y.

In figure 9, x is twice as dissimilar from y as z is from y:

Figure 9

w z

9.) explains why x is twice as dissimilar from y as z is from y. There is a w such

that w is as similar to y as z is to y; w is between x and y; and w is as similar to x as it is

to y.

‘x is less than twice as dissimilar from y as z is from y’ can be defined as follows:

359
10.) x is less than twice as dissimilar from y as z is from y iff there is a w

such that:

a. w is twice as dissimilar from y as z is from y.

b. x is more similar to y than w is to y.

The proof of the horizontal/vertical principle, using only the notion of

comparative similarity, is as follows.

1.) Condition 5.)b in the definition of ‘comparative one-many similarity’

entails that if x is comparatively one-many similar to y and z, then x is

not comparatively between y and z.

2.) If x is equally similar to y and to z, and if y is not less than twice as

dissimilar from z as x is from y and z, then x is comparatively between

y and z.

Suppose that the following are true:

(i) x is equally similar to y and to z.

(ii) y is twice as dissimilar from z as x is from y.

It follows from (i) and (ii) that x is comparatively between y and z. Recall the

definition of ‘comparative betweenness’:

360
6.) x is comparatively between y and z iff

a. y is more similar to x than it is to z.

b. z is more similar to x than it is to y.

c. There is no w such that

either:

i. w is as similar to y as x is to y, and w is more similar to z than x is

to z.

or

ii. w is as similar to z as x is to z, and w is more similar to y than x is

to y.

It follows from (i) and (ii) that 6.)a and 6.)b are met. Furthermore, 6.)c is met.

Take any w such that w is as similar to y as x is to y. Since y is twice as dissimilar from z

as x is from y, it follows that w is not more similar to z than x is to z. The same applies to

any w such that w is as similar to z as x is to z. And the same reasoning applies when we

replace (ii) with (iii):

(iii) y is more than twice as dissimilar from z as x is from y.

This establishes the horizontal/vertical principle.

Initially we defined the ‘one-many similarity’ relation on the assumption that

colour space is a metric space. The purpose of the last section has been to show that there

361
is a notion of comparative betweenness, and, if colour space is not a metric space, then

one can use the notion of comparative betweenness to define a relation similar to the one-

many similarity relation, namely the comparative one-many similarity relation.

10 Conclusion

We started by considering the new colour phenomenal character view. It seemed

that this view was plausible only if colour phenomenal character is pointy. We did not

pursue the new location phenomenal character view, as it was not clear whether that view

is plausible either on the assumption that location phenomenal character is gunky or on

the assumption that location phenomenal character is pointy.

The bulk of the chapter was devoted to exploring the new colour phenomenal

character view, and, in particular, to giving an account of the relationship between being

red* and being red1 and being red2 that would explain the new colour phenomenal

character view. We rejected the asymmetric entailment view, and discussed some costs of

the disjunctive view. We did not discover any obvious problems with the similarity view,

but we must acknowledge that the claim that there is a colour property, being red*, which

is one-many similar to being red1 and being red2, is speculative.

We have not reached a conclusion about which solution to the colour phenomenal

character problem is correct. My aim has been to describe the problem, and to discuss

some solutions to it.

362
Bibliography

Alston, W., 1991, Perceiving God. Ithaca NY: Cornell University Press.

Angell, R., 1974, ‘The Geometry of Visibles.’ Nous, 8: 87-117.

Armstrong, D., 1978, Nominalism and Realism. Cambridge: Cambridge University Press.

Bach, K., 2004, ‘Getting a Thing into A Thought.’ http://online.sfsu.edu/~kbach/

Beebee, H., 2003, ‘Seeing Causing.’ Proceedings of the Aristotelian Society, 103: 257-80.

Belot, G., 2003, ‘Remarks on the Geometry of Visibles.’ Philosophical Quarterly, 53:

581-586.

Berkeley, B., 1965, Berkeley’s Philosophical Writings, ed. D. Armstrong. London:

Collier-Macmillan.

Berkeley, B., 1975, Philosophical Works, ed. M. Ayers. London: Dent.

Bermudez, J., 1998, The Paradox of Self-Consciousness. Cambridge, Massachusetts:

MIT Press.

Block, N., 1996, ‘Mental Paint and Mental Latex’ in E. Villanueva, ed. Philosophical

Issues 7: Perception: 19-49.

Bonevac, D., 2002, ‘Sellars vs. the Given.’ Philosophy and Phenomenological Research,

64: 1-30.

Brandom, R., 1994, Making It Explicit. Cambridge, Massachusetts: Harvard University

Press.

Brewer, B., 1999. Perception and Reason. Oxford: Oxford University Press.

Brewer, B., 2006, ‘How To Account for Illusion’, at

http://www2.warwick.ac.uk/fac/soc/philosophy/staff/brewer/

363
Burge, T., 2003, ‘Perceptual Entitlement.’ Philosophy and Phenomenological Research,

67: 503-548.

Byrne, A., 2001, ‘Intentionalism Defended.’ The Philosophical Review, 110: 199-240.

Campbell, J., 1997, ‘A Simple View of Colour’ in Readings on The Philosophy of Colour,

eds. Byrne and Hilbert. Cambridge, Massachusetts: MIT Press.

Campbell, J., 1994, Past, Space and Self. Cambridge, Massachusetts: MIT Press.

Chalmers, D., 2005, ‘Perception and the Fall from Eden’, in Perceptual Experience, eds.

Hawthorne and Gendler. Oxford: Oxford University Press.

Crane, T., 1988, ‘The waterfall illusion.’ Analysis 48: 142–147.

Cussins, A., 1991, ‘The Connectionist Constructionist of Concepts’, in Readings on The

Philosophy of Artificial Intelligence, ed. M. Boden. Oxford: Oxford University Press.

Davidson, D., 2001, Subjective, Intersubjective, Objective. Oxford: Clarendon Press.

Descartes, R., 1993, Meditations on First Philosophy. Indianapolis: Hackett Publishing

Company.

Dummett, M., 1976, ‘What is a Theory of Meaning (II)?’ in Truth and Meaning, eds. G.

Evans and J. McDowell. Oxford: Oxford University Press.

Dummett, M., 1978, ‘Wang’s Paradox’, in Truth and Other Enigmas. London:

Duckworth.

Evans, G., 1982, The Varieties of Reference. Oxford: Oxford University Press.

Evans, G., and McDowell, J., 1976, Truth and Meaning. Oxford: Oxford University

Press.

Fodor, J., 1984, ‘Observation Reconsidered.’ Philosophy of Science, 51: 21-43.

Goodman, N., 1976, Languages of Art. Indianapolis: Hackett Publishing Company.

364
Gordon, R., 1996, “‘Radical’ Simulationism” in P. Carruthers and P. Smith, eds., Theories

of Theories of Mind. Cambridge: Cambridge University Press.

Guyer, P., 2004. ‘Kant, Immanuel’, in Routledge Encyclopedia of Philosophy, ed. E.

Craig. London: Routledge.

Heck, R., 2000. ‘Nonconceptual Content and the ‘Space of Reasons.’’ The Philosophical

Review, 109: 483-522.

Hopkins, R., 1973, ‘Visual Geometry.’ The Philosophical Review, vol. 82: 3-34.

Hume, D., 1978, A Treatise of Human Nature, ed. P. Nidditch. Oxford: Oxford University

Press.

Jeshion, R., 2006, ‘Soames on Descriptive Reference-Fixing.’ Philosophical Issues, 16:

120-140.

Kant, I., 1929, Critique of Pure Reason, translated by Norman-Kemp Smith. Houndmills:

Palgrave.

Lee, G., 2004, ‘Reply to Price.’ Columbia/NYU Graduate Philosophy Conference.

Lewis, D., 1979, ‘Attitudes De Dicto and De Se.’ The Philosophical Review, 88: 513-543.

Lewis, D., 1980, ‘Veridical Hallucination and Prosthetic Vision.’ Australasian Journal of

Philosophy, 58: 239-249.

Lewis, D., 1999, ‘Individuation by Acquaintance and by Stipulation’ in Papers in

Metaphysics and Epistemology, Cambridge: Cambridge University Press.

Lucas, J., 1969, ‘Euclides ab omni naevo vindicatus.’ British Journal for the Philosophy

of Science, 20: 1-11.

McDowell, J., 1994, Mind and World. Cambridge, Massachusetts: Harvard University

Press.

365
McDowell, J., 1998a, ‘Précis of Mind and World.’ Philosophy and Phenomenological

Research 58: 365-368.

McDowell, J., 1998b, Meaning, Knowledge and Reality. Cambridge, Massachusetts:

Harvard University Press.

McGinn, C., 1982, The Character of Mind. Oxford: Oxford University Press.

Mackie, J., 1999, The Cement of the Universe. Oxford: Oxford University Press.

McNaughton, D., 1988, Moral Vision. Oxford: Blackwell.

Mill, J.S., 1875, ‘Bailey on Berkeley’s Theory of Vision’ in his Dissertations and

Discussion, vol. 2. London: Longmans, Green, and Dyer.

Millar, A., 2000. ‘The Scope of Perceptual Knowledge.’ Philosophy, 75: 75-88.

O’Regan, J. K. and Noe, A., 2001, ‘A Sensorimotor Account of Vision and Visual

Consciousness.’ Behavioral and Brain Sciences, 24: 939-1011.

O’Shaughnessy, B., 1980, The Will. Volume 1, Cambridge: Cambridge University Press.

Peacocke, C., 1983. Sense and content : experience, thought, and their

relations. Oxford: Oxford University Press.

Peacocke, C., 1992, A Study Of Concepts. Cambridge, Massachusetts: MIT Press.

Peacocke, C., 2001. ‘Does Perception Have a Nonconceptual Content.’ Journal of

Philosophy, 98: 239-264.

Peacocke, C., 2003, The Realm of Reason. Oxford: Clarendon Press.

Pollock, J., forthcoming, ‘Perceptual Representation’, in Epistemology: New Essays, ed.

Quentin Smith. Oxford: Oxford University Press, available at

http://oscarhome.soc-sci.arizona.edu/ftp/PAPERS/Perceptual-representation.pdf

Price, H., 1950, Perception. London: Methuen.

366
Price, R, 2005, ‘Content Ascriptions and the Reversibility Constraint.’ Philosophical

Perspectives, 19: 353-374.

Proust, J., 2003, ‘Perceiving Intentions’, in Agency and Self-Awareness: Issues in

Philosophy and Psychology, eds. J. Roessler and N. Eilan. Oxford: Oxford University

Press.

Putnam, H., 1962, ‘What Theories are Not’ Logic, Methodology and Philosophy of

Science, eds. Ernest Nagel, Patrick Suppes, and Alfred Tarski. Stanford, California:

Stanford University Press.

Quine, W. V. O., 1956, ‘Quantifiers and Propositional Attitudes.’ Journal of Philosophy,

53: 177-186.

Reid, T., 2000, An Inquiry into the Human Mind on the Principles of Common Sense.

Edinburgh: Edinburgh University Press.

Searle, J., 1983, Intentionality. Cambridge: Cambridge University Press.

Sellars, W., 1997, Empiricism and the Philosophy of Mind. Cambridge, Massachusetts:

Harvard University Press.

Shoemaker, S., 1994, ‘Self-knowledge and “inner-sense”.’. Philosophy and

Phenomenological Research 54: 249-314.

Siegel, S., 2006, ‘Which Properties Are Represented in Perception?’, in Perceptual

Experience, eds. T. Gendler-Szabo and J. Hawthorne. Oxford: Oxford University Press.

Smith, D., 2000, ‘Space and Sight.’ Mind, 109: 481-518.

Smith, D., 2002, The Problem of Perception. Cambridge, Massachusetts: Harvard

University Press.

367
Soames, S., 2003, Philosophical Analysis in the Twentieth Century. Volume 2:

The Age of Meaning. Princeton: Princeton University Press.

Soames, S., 2005, Reference and Description. Princeton: Princeton University Press.

Strawson, P. F., 1966, The Bounds of Sense. London: Routledge.

Strawson, P. F., 1974a, ‘Imagination and Perception’, in Freedom and Resentment.

London: Methuen.

Strawson, P. F., 1974b, ‘Causation in Perception, in Freedom and Resentment. London:

Methuen.

Thau, M., 2002, Consciousness and Cognition. New York: Oxford University Press.

Thompson, B., 2006, ‘The Spatial Content of Experience’, available at

http://faculty.smu.edu/bthompso/papers.html

Tye, M., 2005, ‘On the Nonconceptual Content of Experience.’ M. E. Reicher, J. C.

Marek (Eds.), Experience and Analysis. 221-239. Wien: öbv&hpt, 2005.

Van Cleve, J., 2002, ‘Thomas Reid’s Geometry of Visibles.’ The Philosophical Review,

111: 373-416.

Waldvogel, J.A., 1990, ‘The Bird's Eye View.’ American Scientist, 78: 342-353.

Williamson, T., 2000, Knowledge and Its Limits. Oxford: Oxford University Press.

Wittgenstein, L., 1958, The Blue and Brown Books. Oxford: Blackwell.

Yaffe, G., 2002, ‘Reconsidering Reid’s Geometry of Visibles.’ Philosophical Quarterly,

52: 602-620.

Zangwill, N., 2004, ‘The myth of religious experience.’ Religious Studies, 40: 1-22.

368