Psyche 9(04): 'Phenomenal Projection' by Zoltan Jakab
Patrick Wilken
PSYCHE: AN INTERDISCIPLINARY JOURNAL OF RESEARCH ON CONSCIOUSNESS
-----------------------------------------------------------------
PHENOMENAL PROJECTION
Zoltan Jakab
Department of Psychology
Rutgers University
152 Frelinghuysen Road
Piscataway, NJ 08854-8020
U.S.A.
Copyright (c) Zoltan Jakab 2003
PSYCHE, 9(04), January 2003
http://psyche.cs.monash.edu.au/v9/psyche-9-04-jakab.html
KEYWORDS: projection, sensory experience, representational
externalism, internalism, color vision.
ABSTRACT: In this paper I shall defend a projectivist view of sensory
experience. The case I shall focus on is that of color experience.
Projectivism has recently been criticized by some authors who claim
that it is unintelligible, or at least implausible, and that it makes
a severe category mistake. I shall argue that despite some prima facie
impressions of implausibility, projectivism can be made intelligible,
and plausible, if its details are spelled out in a reasonable way. In
addition, projectivism is ubiquitous in human psychology, and certain
cases of projection are reasonably viewed as making a category
mistake. Viewed from this perspective, sensory projection is just one
instance of projectivism, brought about by low-level perceptual
processing. Whether sensory projection is one that makes a category
mistake is not obvious. However, even if it does, this is perfectly
compatible with the evolutionary advantage of sensation and
perception.
1. INTRODUCTION
Projectivist views of color experience are typically linked to
subjectivist views of object color. Subjectivism, or eliminativism,
about object color denies that external objects are colored, and
proposes to understand the fact that external objects look to us colored
as a grand illusion or misrepresentation. On subjectivism, color
experience is a product of our brains, with its phenomenal character
determined by properties and processes of our brain. Objects look
colored to us, and what it is like for objects to look colored is
essentially determined by the phenomenal character of color experience.
When objects look colored to us, they look to have salient, spatially
homogeneous surface properties - properties that look to belong,
inherently and inseparably, to external surfaces located in space. But
it is exactly these surface properties that are illusory, or so the
subjectivist claims. Given this theoretical scenario, the subjectivist
has to tell us a story about how the phenomenal characters of our
sensory states "get out there", that is, how it is possible that these
phenomenal characters look to us to be inherent attributes of external
objects and surfaces. Projectivists contend that there is a plausible
story to be told in terms of phenomenal projections.
Roughly, such a projectivist account is proposed, albeit in a crude
form, by Boghossian and Velleman (1997, pp. 94-98). A more detailed
projectivist account is formulated by McGilvray (1994). In McGilvray's
view, visual projection is quite comprehensive: it actually creates its
own illusory, fictitious objects (he calls these illusory objects,
created by visual processing, *perceptual objects*). According to
McGilvray, the idea of assigning illusory color attributes to external,
physical objects is very difficult to make sense of, since (1) this idea
implies that colors (which, to him, are illusory and arise from
phenomenal color experinces) are located both in the head and on
external physical objects. Moreover, (2) organisms with different visual
systems perceive different colors on looking at the same stimulus, and
this makes it necessary to assign to physical objects a set of different
properties that it does not have (McGilvray, 1994, p. 226). For these
reasons McGilvray finds it a better alternative to assume that each
organism creates its own world of fictitious perceptual objects. These
visual fictions, or virtual objects act as interfaces between inner,
perceptual states and external physical objects (McGilvray, 1994, p.
227).
Georges Rey (1995) offers a rather different perspective on
projectivism. He discusses projectivism in general, drawing attention to
cases other than sensory experience - cases that are compelling examples
of mental projection.<1> For instance, people project psychological
mechanisms into themselves and others to explain their own and others'
behavior, and such projections of folk-psychological explanatory
theories are often entirely mistaken. As Nisbett and Wilson (1977)
pointed out, experimental subjects are unanimously mistaken about mental
factors that govern their choices and decisions, reasoning and problem
solving, or emotional reactions. The most typical error subjects made in
a large number of experiments is confusing memory recall and
introspective access to mental states and processes with applying or
generating causal theories about what could plausibly evoke one's own or
someone else's behavioral response. We tend to believe that we
successfully introspect or access reliable memory traces whereas we
"instinctively", that is, automatically, apply explanatory schemas.
In what follows I shall propose a projectivist view that is nevertheless
much less radical than McGilvray's. I do not find McGilvray's reasons
for embracing full-scale projectivism remotely convincing. First, colors
on his view are actually located in the head. But from the idea that
colors are illusory perceptual attributes of external objects the
subjectivist need not arrive at the (patently problematic) conclusion
that colors are actually located both on external objects and in the
head. Colors only look (illusorily) located on external objects and
surfaces. This looking located, or felt location, is a result of how our
perceptual systems work (more on this below), and it does not imply
anything like actual location (of the subjectivist's colors, i.e.,
phenomenal color experiences) on external objects. Second, attributing,
by different organisms, different illusory properties to the same
external object seems an innocent idea as well. The external object need
not actually have any of those illusory properties, so no double
location and multiple properties problems arise contrary to what
McGilvray seems to suggest (1994, p. 226). Therefore, the idea of
visually constructing whole virtual worlds seems unnecessary.
However, in this paper I shall still contend that we do need a subtle
form of projectivism to properly understand color perception. The rest
of this paper offers my reasons for endorsing specifically the
projectivist view. To save space I shall simply assume, and proceed
from, a more general approach: non-dispositionalist realism
(physicalism) about color and internalism about color experience. A
detailed defense of these views would go way beyond the limits of the
present paper. For such a defense see McLaughlin (2002a, 2002b). For
opposing views see Hilbert (1987), Dretske (1995), Tye (1995, 2000),
Byrne and Hilbert (1997), and Hilbert and Kalderon (2000). For a critique of
Dretske's, Tye's, Byrne's and Hilbert's views see Jakab, 2001. For some
critical remarks that are relevant to the Hilbert and Kalderon paper see
Jakab, 2002. To emphasize, the present paper argues simply that
projectivism is a coherent and unproblematic consequence of internalism
about color experience.
My principal aim here is to argue that, despite some fierce criticism of
the projectivist view (Shoemaker, 1997, 228-233; Tye, 2000, 165-167),
this view remains in good shape. Shoemaker and Tye argue that
projectivism is unintelligible or at the very least profoundly
implausible; in addition, it commits a category mistake by assuming that
what look to us to be surface properties of external objects are really
properties or states of our nervous systems. In response, I shall argue
that the idea of projections made at different levels of cognition is
not at all unintelligible. What is more, there are undeniable actual
cases of such projections - cases where we attribute to the external
world properties, states and entities that it does not actually have. It
is also arguable that some (but not all) cases of projection make an
inherent category mistake. That is, it is certain actual psychological
mechanisms themselves that make a category mistake, not our theory of
them. Placed in Rey's broader perspective, I shall primarily focus on
sensory projection, trying to solve the problems that have been raised
specifically against that kind of projection.
In the rest of this paper I shall first support my position that
internalism about color experience needs to assume a subtle form of
sensory projection. Then I attempt to answer at least a selection of the
most important worries that are commonly raised against the idea of
projectivism.
2. IF COLOR PHYSICALISM, THEN WHY PROJECTIVISM?
As I argued elsewhere (Jakab, 2001, 2002; see also McLaughlin, 2002b),
representational externalist views of phenomenal color experience (as I
will call them, *phenomenal externalist* views) face serious (in my
opinion, fatal) problems. Some of these problems arise from the
externalist assumption that the so-called reflectance theory of object
color (Hilbert, 1987; Byrne and Hilbert, 1997; Tye, 1995, chap. 5, 2000,
chap. 7; Hilbert and Kalderon, 2000; Bradley & Tye, 2001) is correct -
since this theory may well be incorrect (see Jakab, 2001). Other
problems for phenomenal externalism are independent of this assumption.
In what follows, based on my own and others' arguments, I shall take an
anti-externalist stand about phenomenal color experience. On this view,
the phenomenal character of color experience is supervenient on (i.e.,
roughly, is determined by, in any theoretically interesting sense of the
term) properties of our visual system.
The core idea of phenomenal, or sensory projection is this: *object
colors look to us the way the phenomenal character of color experience
is*. Traditionally, phenomenal character is understood as the way things
look to us, so, on the traditional reading, this definition is circular:
object colors look to us the way they look to us. I propose to escape
from this circle in the following way. In my understanding, perceptual
states, or what is the same, (perceptual) experiences, are
psychological states of something looking a certain way. Phenomenal
characters, that is, the what-it-is-like- to-undergo-them aspect of
perceptual states are neuro-computational properties of perceptual
states, or so I assume here. I make this assumption on the basis of what
I see an inference to the best explanation, even though I shall not
offer a defense of this view of phenomenal character here. In general,
the neuro-computational properties that are the phenomenal characters
include the complex functional property of *being related to the rest of
one's brain in the appropriate way* (e.g., being a state of activation
of one's color-vision system; see Jakab, 1999 for a little more). The
specific neuro-computational properties that constitute phenomenal color
character derive largely from the opponent processing mechanisms of
color perception (see DeValois and DeValois, 1997; Hardin, 1988, pp. 34-
35; Werner and Wooten, 1979; Hunt, 1982). Perceptual states are neuro-
computational states themselves; by assumption, some of their neuro-
computational properties are crucial to instantiating phenomenal
characters - phenomenal characters are identical with (or, perhaps,
merely supervene on<2>) these key neuro-computatational properties.<3>
In general, projection is systematic misattribution: taking the world to
have a property that is really an illusory, non-existent property
generated by some real properties of our brain states. Sensory
projection is a perception-generated form of intentional inexistence.
The illusory properties created in sensory projection are not identical
with sensory experience (or its phenomenal character), just as Santa
Claus is not identical with mental representations of Santa Claus. For
the illusory properties are by definition nonexistent, whereas the
experiences that generate them exist. Still, the systematically illusory
properties supervene on (are generated by) phenomenal color character.
Perceptual states and their phenomenal characters are the result of
mandatory, encapsulated processing that takes place in perceptual
modules. Phenomenal experience is the final state in these modules,
their output that is then passed on to central, non-mandatory processing
- processing that involves concept application. I take it that
automatic, encapsulated, modular perceptual processing<4> does not
involve concept application, merely analog transformations of transducer
outputs (i.e., outputs of sensory cells in the retina, the cochlea, the
skin, and other places). These analog transformations are describable by
mathematical functions (often by non-linear ones).
Corresponding to this distinction I wish to distinguish *modular*
(automatic, encapsulated) versus *central*, non-mandatory,
'transparent', concept-involving perceptual processing. The latter is
built on, or includes (necessarily), the former, plus adds to it concept
deployment. The former can exist without the latter: perception is
possible without concept deployment - arguably it is possible without
having any concepts<5>.
In what follows, by the contexts 'looks__' (e.g., looks red), 'looks
like__' (e.g., shapes look like types of spatial distribution), 'looks
to belong to', and 'looks to have__' (e.g., looks to have an attribute)
I will refer to perception that is non-concept-involving, or purely
modular. I will use 'looks to be' and 'is perceived as' to refer to
perception that is concept-involving. This distinction is admittedly
close to Chisholm's and Jackson's distinction between *phenomenal* and
*epistemic* uses of 'looks' (Chisholm, 1957; Jackson, 1977), though I
prefer to formulate it in a different, "levels-of-processing"
terminology.
However, since I endorse a non-dispositional realist (physicalist) view
of object color, the following question arises immediately: I accept
that objects are colored, and they look colored to us. So where's the
projection?
Reply. It is not the existence of object colors with regard to which
color perception is illusory. Color perception is illusory with respect
to what kind of properties object colors are. In general, what it is
like to see colors is largely undetermined by what object colors
themselves are like. Object colors look to us in ways in which they are
not. There is an interesting list of ways in which colors look to us -
even though they, that is, the relevant physical surface properties that
are the object colors - aren't that way. Here it is.
First, even though most of the instances of the color red are instances
of a particular type of surface reflectance, it (i.e., the object color
red) does not look that way - it does not look like surface reflectance
of some sort. This sounds uncontroversial despite the fact that I have
no idea what a surface would look like if it looked like a surface
reflectance. By the way, nor do we perceive colors as surface
reflectances. We have to independently learn, from empirical science,
that the tomato's relevant color property is its surface reflectance.
This feature of color perception is termed by Johnston lack of
revelation (Johnston, 1997, pp. 138-142).
The positive side here is that colors look simple, primitive, or
"atomic" features of the world; they do not look like (nor are they
perceived as) dispositions or complexes of other causal features
(Harman, 1997, pp. 253, 260; McGinn, 1996, p. 542).<6> But of course,
in no sense are surface reflectances or other surface color properties
atomic or primitive attributes of objects. Reflective object colors
(surface reflectances) visually look atomic or primitive; they are
perceived as atomic or primitive; these ideas seem to make sense. But
the idea that they are atomic or primitive seems to lack sense.
Prima facie, the fact that colors do not look like reflectances is not
extremely surprising. However, notice the following contrast. (i) Shapes
are types of spatial distribution (of matter), and (ii) we do perceive
them as types of spatial distribution. What is more, shapes arguably
look like types of spatial distribution. Argument: Our very concept of
shape is largely perceptual in origin. Just think of how long it took
from the conceiving of Euclidean geometry to formulating non-euclidean
geometries - the latter make assumptions that contrast with
perception-driven intuition (e.g., changing Euclid's fifth axiom). So,
if we assume that shapes look like spatial distributions, that helps to
explain how shape perception gave rise to our concept of shape.
Second, even though object colors are, as a matter of fact, highly
derivative, anthropocentric, "uninteresting" properties (Tye, 2000, p.
161; Hilbert, 1987, pp. 13-5; 115; 119-120; Gibbard, 1996), they do not
look derivative or uninteresting. To the contrary, colors look salient,
attractive, interesting, often enjoyable properties of objects.
Third, certain object colors look binary (they look like mixtures of two
other chromatic colors) whereas others look unique (they do not look
like chromatic mixtures). The unique-binary distinction is not underlain
by any corresponding structural feature in the object colors themselves.
Nor is unity, the perceptual similarity relations of the colors. Purple
looks more similar to red than to yellow; however, at the level of
emitted wavelengths or surface reflectances, one can find no measureable
relations that parallel such perceptual similarity judgements.<7>
Fourth, the opponent organization of perceived colors is not underlain
by any physical attributes of object colors. For instance, red and green
as surface reflectance or light emission profiles can physically mix and
result in intermediate patterns. So can red and yellow ones. However,
most physical (additive or subtractive) mixtures of red and yellow will
look reddish yellow (i.e., orange), whereas no physical mixture of red
and green (indeed, no surface ever) will look reddish green. Additive
mixtures of red and green typically look yellow (i.e., neither reddish
nor greenish); subtractive mixtures of red and green can look brownish,
yellowish or achromatic gray. Red and green, just as red and yellow, mix
physically; but only red and yellow mix psychologically, at the level of
color experience. Psychological color mixing has features arising from
opponent processing that are arbitrary in the sense that they do not
indicate any corresponding features (like compatibility vs.
incompatibility) in the object colors themselves.
In sum, there is a whole list of color attributes that derive from the
phenomenal character of color experience. Next point: these attributes
are illusory attributes of object colors. Why is this so? First, we can
perceptually detect, hence perceive, these attributes of colors. Based
on color perception we can also judge (i.e., form beliefs about) the
color attributes of surfaces. We can, for instance, perceive whether a
particular color is binary or unique; how similar it is (perceptually)
to other colors; or that it is a salient color that pops out of many
backgrounds. But, as we have just seen, object colors do not,
inherently, or observer-independently, have these attributes. These
attributes arise from how we perceptually react to object colors - that
is, from the phenomenal character of color experience. Yet, for
instance, binariness looks to us to inseparably belong to object colors;
in general, the ways object colors look to us look to belong to the
surfaces of spatially located objects. So, these illusory attributes,
arising from phenomenal color experience, are, by mechanisms of
perception, projected onto the surfaces that have the corresponding
object colors. It is by means of such sensory projections that we see
surfaces at all. This is the mechanism by which what it is like to see
colors arises.
It is important to note that what we perceive are object colors and
their illusory attributes, but not the phenomenal character of color
experience. That we do not perceive. Object colors look binary or unique
(and do so by means of the way the phenomenal character of color
experiences is), but the phenomenal character of color experience does
not look that way (indeed, it does not look any way). In general, 'A
looks B', or even 'A is perceived as B', need not imply that one
perceives B. Therefore, 'object colors look to us the way the phenomenal
character of color experience is' need not imply that we perceive
phenomenal characters themselves or the ways in which they are in a
particular situation. A quick example is as follows.
I am drugged and (mis)perceive my dog as Pegasus. This need not imply
that I perceive, or misperceive, Pegasus. What I (mis)perceive is my
dog. The *perceiving A as B* relation is asymmetric: A causally affects
my perceptual systems, B does not. I am successful in perceiving the dog
in the sense that its presence causally affects my perception - though
it does so via a misidentification.
There is arguably a non-concept-involving counterpart of this
phenomenon: my dog looks like a horse with wings (or *a horsey-wingy
thing*) to me. This can be the result of a hallucination - a distortion
of my perceptual (non-conceptual) representation - there need be no
conceptual involvement here.<8>
3. DETAILS OF THE PROJECTIVIST VIEW
3.1. THE ANALOGY WITH GAUGES
Phenomenal color characters are modes of presentation that are intenally
supervenient. There are two possible approaches at this point. The first
is to say that phenomenal character is part of the representational
vehicle, hence it is not content of any sort. There are obvious examples
of non-representational differences in modes of presentation: for
instance, purely syntactic differences in two names that have the same
referent (i.e., two concepts that refer directly, not via a
description). In a Fodorian vein, we can generalize this idea to
concepts and other kinds of mental representation (Fodor, 1998, chap.
1). The second approach is to hold that phenomenal character is an
aspect of content (or aboutness) that is nevertheless internally
supervenient. Though aboutness itself is an organism-environment
relation, in this particular case it supervenes on what's within the
skin. This is a version of narrow intentionalism about phenomenal
character. I prefer this second option, for the following reason. As I
shall argue below, when we perceive colors, we are directly or primarily
aware of surfaces and their properties (even though some of those
properties are systematically illusory). It is the external surfaces
themselves that look to us in ways the phenomenal character of color
experience is.<9> There is, in addition, a second, externalist aspect of
aboutness of color experience: this arises from the lawlike covariation
between colors and color experience types. Color experience still
reliably indicates color, that is, carries information about color. What
I am denying, however, is that this externalist component of content
plays any role in determining the phenomenal character of color
experience. Phenomenal color character is not identical with externalist
content, nor does it supervene on externalist content (see Jakab, 2001
for argument). For instance, the experience of violet reliably indicates
the object color violet. However, it makes violets look more similar to
reds than to greens, whereas at the level of the colors - the causally
effective surface properties that normally elicit experiences of red,
violet, and green - quite the opposite similarity relations obtain
(i.e., violets are more similar to greens than to reds, in terms of,
say, surface reflectance).
Think of the problem thus: phenomenal experiences are the needle
positions of our sensory gauges: as such, they are supervenient on the
internal constitution of the organism. Just like the needle positions of
a fuel gauge: the gauge can point to "FULL" even if there's water in the
tank, or the tank of the wreck has long been removed, and the needle is
stuck in the FULL position. Needle position itself is not a relational
property of the gauge (only the information it conveys is). Similarly,
the same fuel level can be indicated by different fuelgauges that have
differently looking needle positions (analog, digital, mechanical,
electronic, etc.) that nevertheless convey the same information.
Now comes the key worry. How can it be that, although phenomenal
character belongs to our perceptual states (it is an attribute of our
perceptual states), yet we experience that phenomenal character as
located in outer space, as intrinsic properties of stimuli?<10> (Ross,
2000, p. 52n11, p. 54; Tye, 2000, pp. 165-166; Shoemaker, 1994, p. 25;
McGilvray, 1994, pp. 226-227).
General answer: *felt location* is a key part of the act of projecting.
Felt location amounts to the taking (by non-concept-involving
perception) of the internalist content of our perceptual states to be a
state of affairs in the external world. The systematically illusory
color-attributes that arise from the phenomenal color characters look to
belong to external objects and surfaces exactly by virtue of the
phenomenal color characters' being coupled with felt locations.
Detailed answer, part 1: the reason why the phenomenal character of
color experience comes with a felt location, that is, why it gives rise
to illusory attributes that look to belong, inherently, or inseparably,
to external objects, is that this is evolution's solution to the problem
of assigning these needle position analogs to what they are about: the
corresponding stimuli. We perceive the object colors by undergoing the
needle-position-analog phenomenal experiences that in turn come with
felt location.
Detailed answer, part 2: how is it possible for experiences to have felt
location? Our color experiences are perceptual states within our heads;
they are experiences as of spatially located things. They convey
information about spatial location and are, as particular perceptual
states, interpreted by the rest of the visual system as indicating some
spatial location at which a perceived stimulus is located. This
interpretation-as- indicating-location results in the felt location<11>
(at the level of phenomenal character) that in turn represents physical
location. Here is a little more detail; for the sake of simplicity I
speak about the perception of depth instead of location in general.
Color experiences are representationally atomic. Perceptual states that
are particular color experiences either do not have constituent (or
syntactic) structure at all, or they have only a rather minimal one.<12>
Color experiences reliably indicate object colors, but do not map, or
depict, any complexity that characterizes particular color stimuli in
terms of physical properties. Color experience simply scales object
colors in a 3D sensory space (i.e., color space). On the contrary,
visual experiences (perceptual representations) of shapes have rich
constituent structure, and this constituent structure is there to
systematically map, or depict (in the form of analog representations
like symbol-filled arrays), particular shapes and spatial patterns that
obtain in the perceived spatial layouts.
Depth cues are typically relational: the depth from the perceiver of a
particular item (e.g., an object O) within a scene is estimated from O's
perceived relations to the background, and other parts of the scene.
Monocular depth cues like relative size, partial occlusion, or movement
relative to other objects in the scene are relational attributes. In
order to extract monocular depth information about the depth of a
particular object O, O's relation to the rest of the scene has to be
represented. This happens by building up complex perceptual
representations with constituent structure that are interpreted by
processes that operate on them as analog maps of the perceived spatial
layouts. Similarly for binocular depth cues (i.e., small differences in
the two retinal projections of the same object). Such differences obtain
only in sufficiently complex retinal projections. In a Ganzfeld-like
perceptual situation the single color experience that we have has no
determinate felt depth, because in the retinal projections that a
Ganzfeld stimulation gives rise to no monocular or binocular depth cues
are available. To summarize, felt depth arguably arises only when we
undergo sufficiently complex perceptual states that encode, in their
constituent structure, relational information (spatial relations) about
different entities in the perceived scene. This sounds like a standard
representationalist account of depth perception. I offer it as an
account of how felt depth (felt location) can possibly arise. Felt depth
(and, in general, felt location) arises from processing complex
perceptual representations in our visual system. These complex states
encode information about depth, and are interpreted by the processes
that operate on them as conveying such information. Whether this account
is externalist about felt location (a particular aspect of phenomenal
character) is another question.<13>
Object colors look to us the way the phenomenal character of color
experience is. They - the colors - also look to us spatially located,
quite veridically. Somewhat less veridically, the illusory perceptual
attributes of the colors that are products of our color vision system
also look to belong to, or look like attributes of, external objects and
surfaces - they also look spatially located. For instance, it is a
purple surface - the surface itself - that looks more similar to a red
surface than to a yellow one, even though, in terms of
perceiver-independent properties (i.e., excluding relational properties
with perceivers as a relata; in terms of stimulus properties that are
causally responsible for color perceptions), these similarity relations
need not obtain. We see the colors by means of undergoing phenomenal
color experiences that come with felt location - and phenomenal color
experiences with their felt locations are products of our brain. But
from this it does not follow that phenomenal color experience is
actually located in outer space - so no double location problem arises.
All that's said is that the phenomenal character of color experience
most often (though not necessarily) includes a felt location. Object
colors are located in external space. Phenomenal color experiences are
located in the brain. Colors as we perceive them, that is, illusory
properties produced by phenomenal color experience (but not identical
with color experience) appear located in external space. What is located
in the brain (states and events) create the perceptual impressions that
we call felt depth, felt redness, and so on. These impressions
successfully, that is, veridically, indicate conditions in the
environment, namely actual distance from the perceiver, and object
colors.
I propose to understand the story I have told so far in terms of an
adverbial account, as opposed to a sense datum theory. Sense datum
theories have it that sense impressions are the direct objects of
perception. They are the only objects of which we are aware when we
hallucinate, for instance. On the adverbial theory, sense impressions
are ways or modes of perceiving, or ways of being appeared to. They are
not in any sense objects of perception. Rather, sense impressions are
perceptual reactions to the objects of perception - the external
stimuli. Sense impressions are states and events in our brain. I endorse
the adverbial theory. In my view, the only objects of perception are
external physical objects.
3.2. HOW ABOUT CATEGORY MISTAKES?
What I have said so far, may seem to imply this: color perception
attributes properties of our experience (its phenomenal character) to
external objects. Moreover, as certain authors emphasize (Shoemaker,
1994; Tye, 2000, pp. 165-167), in this assumption there is already a
severe category mistake. For this view simply confuses properties of
mental states with those of stimuli. Felt redness, the phenomenal
character of red experiences (i.e., a property of mental states), cannot
possibly be instantiated in external objects. There is no room for such
a property in the external world - how could the external world host a
property of a mental state? The whole projectivist idea smells like
nonsense. I have three points in reply to this objection.
1. We can immediately introduce a twist and say that if there is a
category mistake somewhere around here, then it is a category mistake
inherent in low level perception itself, not in my theory of it. Assume
that there is indeed such a category mistake inherent in sensation. But
why shouldn't color perception present to us the world in a way in which
it cannot literally be, given that this inherently misleading
presentation has its own evolutionary advantages? Color perception still
endows us with a powerful means of discriminating objects and surfaces
that are not (or not easily) discriminable via other perceivable
aspects.
2. It is also arguable that sensory projection commits no category
mistake at all - even though it still commits a systematic mistake of
some sort (though one that does not reduce fitness). Here is some
argument. Phenomenal projection would be ruled out if phenomenal
externalism were right, that is, if there were ordinary stimulus
properties that played a key role in the determination of the phenomenal
characters of color experiences. But, as I argued elsewhere (Jakab,
2001), there are no such stimulus properties, and this effectively
falsifies phenomenal externalist positions. However, what phenomenal
color experience might be taken to mistakenly suggest to us, upon
reflection, is that there exist such stimulus properties in the physical
objects we perceive. What sort of stimulus properties should these be?
Well, they should be natural kind essences or at least physical types of
some sort. They should be inherent, non- disjunctive, causally effective
properties of surfaces and volumes that are specifically causally
responsible for our color perceptions in ordinary circumstances. They
should, in terms of causally effective stimulus properties that are
specifically causally responsible for our color perceptions, exhibit
exactly those similarity relations, unique-binary division, and
compatibility-incompatibility relations which colors as we perceive them
do. But if all that color experience suggests to us is that object
colors are such physical types, then it seems that there is no category
mistake involved in phenomenal projection at all. Consider the parallel
with shape perception. What our visual (and tactile) experience of
shapes suggests to us is that there are causally effective,
non-disjunctive properties, or physical types (namely the shapes) out
there such that they play a key role in determining our phenomenal
(visual or tactile) experiences of shapes. The key difference is that
this suggestion is correct in the case of shape perception whereas it is
incorrect in the case of color perception.
Here is another line of argument to support the "no category mistake"
response. The phenomenal character of color experience gives rise to
illusory surface properties, alright, but these illusory properties are
not identical with the phenomenal characters. Therefore it does not
follow that the illusory properties that we perceive surfaces as having
are themselves mental. For one thing, the illusory properties are, by
definition, non-existent, whereas the phenomenal characters that
generate them exist. So the two can't be the same. What sounds a lot
better is the idea that the illusory properties supervene on the
phenomenal characters. Obvious cases of systematic perceptual illusion
are the best analogies here. In the case of the Müller-Lyer illusion,
for instance, the illusory attribute (the length difference between the
two segments) is brought about by perceptual processing in the brain;
arguably, I think, this illusory attribute is supervenient on what's
happening in the brain. Still, the illusory length difference and the
perceptual processing that creates it are not the same thing. To
summarize, object colors exist, but the way in which they look to us
includes a systematic illusion (or system of illusions).
3. Projectivism, that is, attributing to the external world properties
and entities that it does not actually have, is ubiquitous in human
psychology. Paranoid subjects attribute hostility and malevolence to
others. Children (and adults in numerous cultures) attribute
psychological states to inanimate objects - an example of animistic
thinking. Paranoia is a form of projection that does not make a category
mistake (even though it does make a systematic mistake of some sort).
Even if, in particular cases, some other person is not actually hostile
to the paranoid subject, he might be so.<14>
Animistic thinking, on the other hand, is a case of projection that does
seem to commit a category mistake. Animistic projections are often quite
literal. For instance, when a native person thinks that the Holy
Mountain is angry at her tribe, that's why it is sending lightning bolts
(coughes up lava and smoke, etc.) for over three days now, she is
arguably not using the concept of anger in a metaphorical sense, but
rather, in a quite literal one. Instances of such animistic thinking in
tribal cultures involve a literal attribution of mental states to
inanimate beings, not just a metaphorical one. The native would reject
the idea that the Holy Mountain does not really have any emotional
states like anger at all. The opposite sounds like a contradiction:
false attribution cannot include the insight that there is a false
attribution going on. Animistic projection is more than mere metaphor
application (e.g., "Funny, it looks like the Holy Mountain is angry at
us - but of course that is an absurd idea, taken literally.")
It is arguable that such phenomena are not limited to tribal cultures.
For instance, it might occur to an atheist that the idea of an
omnipotent, omniscient creator of the world is just a projection - a
manifestation of animistic thinking widespread in western culture adult
population. However, those who do believe in God think that God
literally exists, that is, our world literally includes such an
omnipotent creator (who is actually the origin of everything else in the
world) - there is no metaphoric sense involved here.<15> Staying with
the example of the native woman for a moment, assuming (at the
conceptual level) that a mountain is literally angry is no less nonsense
than assuming (at the stage of low level perception) that surfaces
actually have attributes that are, as a matter of fact, illusory, and
are products of our color vision system. Or, perhaps, neither case is
nonsense.
In sum, it seems that projection appears at different levels of
psychological organization. Whether or not it involves an inherent
category mistake, it is a phenomenon of psychology whose existence is
difficult to deny. If there is also a category mistake inherent in such
a low (modular) level of mental organization as sensation, it is
certainly adaptive: it has endowed us with a powerful discriminatory
capacity. The fact that sensory projection is completely resistant to
intellectual insight (i.e., object colors just look to us the way the
phenomenal character of color experience is, no matter how we reflect
upon this phenomenon) is well explained by cognitive impenetrability.
Low level perceptual processes are largely uninfluenced by conceptual
ones.
3.3. EXERCISES: FORMS OF AWARENESS
3.3.1. PERCEPTUAL AWARENESS
Perceptual awareness is direct or primary; it is awareness of external
stimuli. We are perceptually aware of object colors by means of their
being reliably indicated to us by color experience. Object colors look
to us the way the phenomenal character of color experience is. However,
color perception by no means makes this projective identity relation
epistemically transparent to us. This is my view of transparency:
sensory projection (and, plausibly, other sorts of projection) are not
self-revelatory. For example, it is typically not obvious to paranoid
subjects that they merely (i.e., falsely) attribute hostility and
malevolence to others in particular cases - if this were obvious, then
there would be no projection.<16> Sensory projection serves to give us
some sort of direct access to stimulus properties. In the case of object
color, this "some sort of" direct access includes reliable indication,
but it does not include veridicality in every aspect, with respect to
the properties indicated (see Akins, 1996, p. 364, for a similar
suggestion).
3.3.2. INTROSPECTIVE AWARENESS
Introspective awareness is secondary: it is awareness of what is going
on in one's mind. It arises from "extra" processing of our occurrent
perceptual states; processing that is not necessarily part of
perception. This may mean belief formation (e.g., "I am currently
undergoing such and such a perceptual experience"), but it can mean
other things as well. For instance, it can mean constructive imagination
(e.g., Jakab, 2000, Sec. 5), or focusing attention (processing efforts)
on construcive imagination; the recall of perceptual memory traces, or
the maintaining of traces of perception formed a moment ago, as opposed
to information gathering through the senses.
For a typical example, introspective awareness can take the following
form: (1) undergoing an experience of type E; (2) forming the belief
that one is undergoing an experience of type E; (3) knowing or believing
that there is no corresponding external stimulus present that is
responsible for the occurrence of E. Here (2) and (3) are "extra"
processes not necessarily involved in perception. Therefore, this sort
of awareness - a version of introspective awareness - is secondary, in
comparison with perceptual awareness. By (3) the "usual immediate
epistemic import" of the experience E is secondarily suspended, as when
we learn that there's no tank in the wreck, just the needle of the fuel
gauge is stuck in the FULL position. This sketchy account of
introspection is in some respects similar to representationalist
accounts (e.g., Dretske, 1995, chap. 2, esp. p. 63; Tye, 2000, pp.
51-54), even though I do not think that in the case of introspective
visual awareness our attention goes outside, onto external stimuli (Tye,
2000, p. 51).
3.3.3. HALLUCINATION
When we hallucinate there is no object, or relevant, causally effective
stimulus property (the normal causal antecedent of appearance), in
addition to the systematic illusion that's inherent in normal,
successful instances of color perception. Hallucination is mistaken
perceptual awareness: awareness of an object where there is no
corresponding object to be aware of.
4. CONTRAST AND SUMMARY: VIEWS OPPOSING PROJECTIVISM
As I said earlier, the point of this paper is made in a framework of
specific assumptions. One group of assumptions is about perception in
general (i.e., early vision and the idea that much of perceptual
processing is uninfluenced by top down, concept-driven mechanisms, and
does not presuppose concept deployment). I take it that these ideas are
quite firmly established in the literature. The other key assumption,
namely internalism about color experience, has a more controversial
status: many, including leading experts, believe, and have argued
thoroughly, that phenomenal externalism is the right view to take. Yet I
assume that phenomenal externalism is incorrect and internalism is
correct, because at other places I made detailed arguments to this
effect (Jakab, 2001, 2002). Moreover, I am not alone in holding such
views (among others see McLaughlin, 2002b; Block, 1997, 1999; Kirk,
1994).
However, as a means of summarizing, it is worth looking out of this
framework to briefly consider those views that oppose projectivism.
There are at least three main anti-projectivist approaches to color
experience: representationalism (Dretske, 1995; Tye, 1995, 2000; Byrne
and Hilbert, 1997; Hilbert and Kalderon, 2000), revelationism (Campbell,
1993; see also Byrne, 2001, p. 245; McLaughlin, 2002a, 2002b; Atherton,
2002), and Shoemaker's view of phenomenal character (Shoemaker, 1994).
Shoemaker's approach is, in a way, representationalist, though it is
importantly different from Dretske's and Tye's (see Tye, 2000, chap. 5
for a critical discussion). Shoemaker attempts to avoid projectivism
despite accepting an internalist approach to color experience.
Revelationism is quite close in spirit to representationalism. Both
views hold that object colors themselves crucially determine what it is
like to see them (i.e., the phenomenal character of color experience),
therefore on these views there is no need to posit a sensory-level
projection of internally generated phenomenal characters. Note however,
that Shoemaker denies this determination relation (see Shoemaker, 1994,
pp. 35-36, and other places). Still, Shoemaker wants to avoid
projectivism by saying that color experience veridically represents some
properties of perceived objects other than their colors. Shoemaker calls
these properties *phenomenal properties*.
Campbell's Simple View (Campbell, 1993) is a paradigmatic exposition of
revelationism (see also Byrne, 2001, p. 245; McLaughlin, 2002a, 2002b;
Atherton, 2002). Revelationists hold that color perception gives us
access to the essential nature of colors and whatever conceptual
knowledge we might acquire about color is only secondary. This
conceptual knowledge cannot affect or correct the knowledge by
acquaintance that perception gives us about color; color perception is
the best guide to the very nature of object color. For critique of the
Simple View see Smith, 1993; Tye, 2000, p. 149, and Jakab, 2001, pp.
22-26, 159-161. For a more general critical discussion of revelationism
see McLaughlin, 2002b. For some defending lines see Atherton, 2002. For
a brief reflection on Atherton's view see Jakab, 2002, note 4.
The key point of all these views is that (1) there's no need for
assuming sensory projection - a kind of systematic misrepresentation -
because phenomenal color experience veridically represents object colors
(or, in Shoemaker's view, his phenomenal properties).
Representationalists hold that color experience represents colors in
such a way that its representational content can explain the attributes
of phenomenal color character (unity, the unique-binary distinction,
opponent organization, and so on). Revelationists agree with
representationalists that unity, the unique-binary distinction, and
opponent organization in color experience veridically represent
corresponding relational properties of the colors. In addition, some
representationalists also claim that (2) projectivism is an inherently
problematic view. I reject both (1) and (2). As against (1), I argue in
Jakab, 2001, 2002. In the present paper I have focused on arguing
against (2).<17>
NOTES
<1> Rey's own definition of projection is this: "...*we expect there to
be phenomena in the world correlative to stable psychological states in
ourselves, but there turn out not to be any*" (Rey, 1995, p. 136; his
italics). I will offer a similar definition of projectivism later in
this paper.
<2> I'd rather prefer the identity version. Again, I do not claim to
offer here a defense of a neuro-computational account of phenomenal
character (although the present paper as a whole can be taken as part of
such a defense).
<3> I use 'experience' inclusively, as referring to a neural/perceptual
state that carries its phenomenal character. For a brief discussion of
this terminological issue, see Tye, 2000, p. 15. I follow S. Kosslyn
(1980, 1990, 1994) and others like Marr (1982) and I. Biederman (1990)
in assuming that visual perceptual representations are maplike, analog
ones, built out of some basic set of primitive symbols.
<4> In the case of vision, I think of processing up to Marr's 3D
representations (Marr, 1982). Alternatively, later stages of modular
processing are perhaps better understood in terms of Biederman's model
(Biederman, 1990).
<5> The distinction between modular, nonconceptual, and central,
concept- involving representations is a well-supported one. Think of the
notion of early vision, or low-level vision (Marr, 1982; Stillings et
al., 1995, pp. 464-490). As a phenomenon closely related to early (or
intermediate-level) visual processing, think of random dot stereograms
(Julesz, 1971). On looking at the stereograms in the stereoscope we
suddenly see shape and depth, merely on the basis of binocular disparity
cues. Very likely indeed, the binocular integration underlying this
phenomenon does not include concept deployment, yet it results in
conscious perceptual experience (as of shape and depth). Binocular
integration happens in the visual cortex, so early (or intermediate)
visual processing is by no means equivalent to pre- cortical processing.
However, in the case of color processing, some steps that are of key
importance in shaping our experience of color happen pre- cortically.
Opponent recombination of the cone signals is an example: it happens in
the LGN of Thalamus (DeValois and DeValois, 1997). See note 8 for
further support of the distinction between non-conceptual and conceptual
levels of representation.
<6> There is no general agreement on this issue, however. For instance,
Langsam (2000) argues that colors look like dispositions (in the non-
conceptual, or phenomenal, sense of 'looks': see Byrne, 2001, p. 239).
Byrne, however, finds Langsam's reasoning inconclusive, and argues that
one interesting reading of his claim is false: it does not appear that
colors are dispositions (Byrne, 2001, pp. 242, 243; note that 'appears'
is used by Byrne in the same sense as conceptual - or epistemic -
'looks'). Tye (2000, pp. 55-57) argues that colors look like types of
reflectance (again, in the non-conceptual sense of 'looks'). In claiming
this, Tye assumes the reflectance theory of color that he defends in
other parts of his book. This latter issue is critical for my view, so I
have to take a stand on it. I flatly deny Tye's idea that colors look
like reflectances on the following grounds. In my view, phenomenal color
characters are internally generated, and they are modes of presentation
(of object colors), though non-conceptual ones. Therefore, on the
phenomenal internalist approach, the non-conceptual (phenomenal) 'looks'
context is hyperintensional (because its operand includes internally
supervenient modes of presentation to which the context is sensitive,
thus the principle of substitutivity breaks down in this context). That
is, contra Tye, even if one assumes that redness in objects is surface
reflectance such-and-such, a red object, in looking red, does not look
like surface reflectance such-and- such. Just like we can think that
water is wet without thinking that H2O is wet.
<7> Some authors (Byrne and Hilbert, 1997, pp. 279-281; Tye, 2000, pp.
162-165; Bradley and Tye, 2001) argue that the unique-binary distinction
is parallelled by physical attributes of object colors. Tye's proposed
solution (2000, pp. 162-165) is, on empirical grounds, badly mistaken
(Jakab, 2001, pp. 68-81; 229-230). Bradley and Tye (2001) repeat the
same proposal and offer some clues of how it might be fixed (p. 482).
However, these authors do not even mention the problem of unity. Unity
and the unique-binary distinction are closely linked phenomena, but even
assuming that Bradley and Tye are right about the objective bases of the
unique-binary distinction, it is by no means obvious that their proposal
generalizes to the much more complex issue of unity. Matthen (1999, pp.
65-66) argues thus "Because violet looks reddish, it looks more similar
to orange with which it shares a component, than to greenish yellow,
with which it shares no similarity. However, violet is at the opposite
end of the visual spectrum from orange. In actuality, it is closer to
greenish yellows than to orange. Thus, opponent processing distorts the
ordering of colors by wavelength". Hilbert and Kalderon (2000) argue
that representational externalism about color experience can handle
unity. I agree with Matthen (and Thompson, 1995, pp. 122-133; 2000) that
it cannot (see Jakab, 2002 for my argument).
<8> The arguments supporting this case are well known. Perceiving shapes
does not presuppose having shape concepts. Think of Marr's (1982) or
Biederman's (1990) models of shape perception. For instance, geons and
their complexes are not shape concepts; they are non-conceptual
perceptual representations. This is because geons are not available for
reasoning (as opposed to concepts), and the representation of shapes at
this level seems uninfluenced by whatever beliefs or conceptual
knowledge one might have about shapes. (Think of figural aftereffects,
the Müller- Lyer illusion, and other shape illusions). Therefore one can
represent the shape of one's dog at this level, without deploying any
shape concepts. This perceptual representation can also be severely
distorted, due to drug influence. For another example, think of
perceiving very complex shapes like a fractal picture. The richness and
complexity of the shape percept here is not likely to be fully captured
by any concept we might recall to represent such shapes - therefore it
is reasonable to posit a separate level of perceptual representation to
account for this richness. See Kirk, 1994, pp. 124-125; Tye, 1995, pp.
137-143; 2000, p. 11; Raffman, 1995, for similar points on the richness
of perceptual representations. Note also that philosophers holding
different views of the phenomenal seem to unanimously endorse the
distinction between conceptual and nonconceptual (perceptual)
representations (Tye, 1995, Chs. 4-5; 2000, pp. 56, 62 - Tye is a
representational externalist about the phenomenal; see however Kirk,
1994, Chs. 4-5, esp. pp. 128, 130, 136 for a very similar distinction in
an anti-externalist account of the phenomenal). Of course, a primary
source for understanding this distinction is Fodor (1983, 1990).
<9> I am grateful to Brian McLaughlin for illuminating this point to me.
<10> By 'experiencing' I mean undergoing (an experience), but not
perceiving. I do not like this formulation of the problem; I only intend
to paraphrase Ross, Tye and Shoemaker here. The view just cited in the
main text is called by Shoemaker (1994, p. 25) literal projectivism. The
view I'm defending in this paper is closer to figurative projectivism
(Shoemaker, 1994, pp. 25-26).
<11> That is, felt location comes from the interpretation by the rest of
the system of a token perceptual state (narrowly individuated), and not
from the relation of that perceptual state to environmental stimuli. By
'interpretation' here I mean, for instance, some (implemented)
computational process - simply some causal interactions between the
narrowly individuated perceptual state and the rest of the system. That
is, theoretically, there is room even for the claim that felt location,
as an aspect of phenomenal character, is supervenient on internal
constitution, even though this aspect of phenomenal character still
systematically correlates with information picked up by vision about
actual location.
<12> The experience of a unique hue with maximum saturation is, in my
opinion, an example of a perceptual state with no constituent structure.
The experience of a binary hue is a perceptual state with a rather
minimal constituent structure. For more on this topic, see Jakab, 2000,
sections 3 and 4.
<13> It need not be: one might argue that as long as the same internal
states and processes obtain in the subject (starting at the retina), she
undergoes the same perceptual experience with the same felt location, no
matter what external circumstances produce those internal states. Brains
in vats too have experience with felt location (or so the intuition
goes), even though their visual representations do not covary with
anything like real depth of perceived objects from the subject.
<14> Paranoid subjects sometimes make other people become angry of them
- an instance of self-fulfilling prophecy.
<15> See Rey (1995, p. 136) for more on projection and religious
experience.
<16> This case is a little more complex than my presentation suggests.
Psychotherapists know well the phenomenon when a patient has insight of
his/her paranoia, yet on leaving the session and returning to "real
life" continues to make paranoid attributions spontaneously. An insight
of paranoia at a certain point in time does not immediately "pervade"
the subject's personality, obliterating paranoid thinking. After the
first catharsis of insight on the analyst's couch is gone, it remains
difficult for the patient to continue to believe, "deep down", that
other people are indeed so different from what he has thought about them
all along. (I think even rational insight of paranoia and emotion-based
paranoid attributions that in turn influence behavior can continuously
coexist in subjects.) Personality and (rational) intellect are quite
separate aspects of human minds, however, both influence the ways we
think about the world.
<17> I wish to thank Andrew Brook, Brian McLaughlin, and two anonymous
reviewers for their thoroughgoing commentaries on earlier versions of
this paper.
REFERENCES
Akins, K. (1996). Of Sensory Systems and the Aboutness of Mental States,
*Journal of Philosophy, 93,* 337-372.
Atherton, M. (2002) Asking about the Nature of Color, In Mausfeld R.,
and Heyer, D. (Eds.): *Colour Vision: From Light to Object.* Oxford:
Oxford University Press.
Biederman, I. (1990). Higher-Level Vision. In D. N. Osherson, S. M.
Kosslyn, & J. M. Hollerbach (Eds.), *An Invitation to Cognitive
Science* (Vol. 2: Visual Cognition and Action, pp. 41-72).
Cambridge, MA: MIT Press.
Block, N. (1997). Inverted Earth. In Ned Block, Owen Flanagan, Güven
Güzeldere (Eds.), *The Nature of Consciousness*, Cambridge, MA:
MIT Press
Block, N. (1999). Sexism, Racism, Ageism, and the Nature of
Consciousness. *Philosophical Topics, 26,* (1-2), 39-70.
Boghossian, P., A., & Velleman, D., J. (1997). Color as a Secondary
Quality. In Alex Byrne and David R. Hilbert (Eds.), *Readings on
Color* (Vol. 1: The Philosophy of Color, pp. 81-103). Cambridge,
MA: MIT Press
Bradley, P., & Tye, M. (2001). Of colors, kestrels, caterpillars and
leaves. *Journal of Philosophy, 98,* 469-487.
Byrne, A. (2001). Do colours look like dispositions? *The Philosophical
Quarterly, 51,* 238-245.
Byrne, A., & Hilbert, D., R. (1997). Colors and Reflectances. In Alex
Byrne and David R. Hilbert (Eds.), *Readings on Color* (Vol. 1: The
Philosophy of Color, pp. 263-288). Cambridge, MA: MIT Press
Campbell, J. (1993). A Simple View of Color. In John Haldane and Crispin
Wright (Eds.): *Reality, Representation and Projection.* Oxford
University Press
Chisholm, R. (1957). *Perceiving: A Philosophical Study.* Ithaca, New
York: Cornell University Press.
DeValois, R., L., & DeValois K., K. (1997) Neural Coding of Color. In
Alex Byrne and David R. Hilbert (Eds.): *Readings on Color* (Vol. 2:
The Science of Color, pp. 93-140), Cambridge MA: MIT Press.
Dretske, F. (1995). *Naturalizing the Mind.* Cambridge, MA: MIT Press
Fodor, J., A. (1983). *The Modularity of Mind.* Cambridge, MA: MIT
Press
Fodor, J. A. (1990). Why should the mind be modular? In J. Fodor: *A
theory of content and other essays.* Cambridge, MA: MIT Press
Fodor, J., A. (1998). *Concepts.* Cambridge, MA: MIT Press
Gibbard, A. (1996). Visible Properties of Human Interest Only.
*Philosophical Issues*, Ed. E. Villeneuva. Atascadero, Calif.:
Ridgeview.
Hardin, C., L. (1988). *Color for Philosophers: Unweaving the Rainbow.*
Indianapolis, MA: Hackett.
Harman, G. (1997). Explaining Objective Color in Terms of Subjective
Reactions. In Alex Byrne and David R. Hilbert (Eds.), *Readings on
Color* (Vol. 1: The Philosophy of Color, pp. 247-261). Cambridge,
MA: MIT Press
Hilbert, D., R. (1987). *Color and Color Perception.* A Study in
Anthropocentric Realism. Stanford: Center for the Study of Language
and Information
Hilbert, D., R., Kalderon M., E. (2000). Color and the Inverted
Spectrum. In Steven Davis (Ed.): *Color Perception: Philosophical,
Psychological, Artistic and Computational Perspectives* (pp.
187-214). New York: Oxford University Press
Hunt, R., G., W. (1982). A model of colour vision for predicting colour
appearance. *Color Research and Application, 7,* (2), 95-112.
Jackson, F. (1977). *Perception.* Cambridge: Cambridge University Press.
Jackson, F., & Pargetter, R. (1997). An Objectivist's Guide to
Subjectivism About Color, In Alex Byrne and David R. Hilbert (Eds.),
*Readings on Color* (Vol. 1: The Philosophy of Color, pp. 67-79),
Cambridge, MA: MIT Press.
Jakab, Z. (1999). Overlooking the resources of functionalism?,
*Behavioral and Brain Sciences, 22,* (6), 957
Jakab, Z. (2000). Ineffability of qualia: a straightforward naturalistic
explanation. *Consciousness and Cognition, 9,* (3), 329-351
Jakab, Z. (2001). Color experience: empirical evidence against
representational externalism. Ph.D. thesis, Carleton University,
Ottawa. Available at: www.carleton.ca/iis/TechReports
Jakab, Z. (2002). Intrinsic colors - and what it is like to see them. In
Mausfeld R., and Heyer, D. (Eds.): *Colour Vision: From Light to
Object.* Oxford: Oxford University Press, forthcoming.
Johnston, M. (1997). How to Speak of the Colors. In Alex Byrne and David
R. Hilbert (Eds.), *Readings on Color* (Vol. 1: The Philosophy of
Color, pp. 137-176). Cambridge, MA: MIT Press
Julesz, B. (1971). *Foundations of Cyclopean Perception.* Chicago:
University of Chicago Press.
Kirk, R. (1994). *Raw Feeling.* Oxford: Clarendon Press.
Kosslyn, S., M. (1980). *Image and Mind.* Cambridge, MA: Harvard
University Press.
Kosslyn, S., M. (1990). Mental Imagery. In D. N. Osherson, S. M.
Kosslyn, & J. M. Hollerbach (Eds.), *An Invitation to Cognitive
Science* (Vol. 2: Visual Cognition and Action, pp. 73-97).
Cambridge, MA: MIT Press
Kosslyn, S., M. (1994). *Image and Brain. The Resolution of the Imagery
Debate.* Cambridge, MA: MIT Press
Langsam, H. (2000). Why colours do look like dispositions. The
*Philosophical Quarterly, 50,* 68-75.
Marr, D. (1982) *Vision. A computational investigation into the human
representation and processing of visual information.* New York:
Freeman
Matthen, M. (1999). The Disunity of Color. *Philosophical Review, 108,*
47-84.
McGilvray, J., A. (1994). Constant colors in the head. *Synthese, 100,*
197-239.
McGinn, C. (1996). Another look at color. *Journal of Philosophy, 93,*
537-553.
McLaughin, B., P. (2002a) The place of color in nature. In Mausfeld R.,
and Heyer, D. (Eds.): *Colour Vision: From Light to Object.* Oxford:
Oxford University Press, forthcoming.
McLaughlin, B., P. (2002b). Color, consciousness, and color
consciousness. In Quentin Smith (Ed.): *New Essays on
Consciousness.* Oxford: Oxford University Press, forthcoming.
Nisbett, R., & Wilson, T. (1977). Telling more than we can know.
*Psychological Review, 84,* 231-259.
Raffman, D. (1995). On the Persistence of Phenomenology. In Thomas
Metzinger (Ed.): *Conscious Experience* (pp. 293-308). Schöningh:
Imprint Academic.
Rey, G. (1995). Towards a Projectivist Account of Conscious Experience.
In Thomas Metzinger (Ed.) *Conscious Experience* (pp. 123- 142).
Schöningh, Imprint Academic.
Ross, P. W. (2001). The Location Problem for Color Subjectivism.
*Consciousness and Cognition, 10,* 42-58.
Shoemaker, S. (1994). Phenomenal Character. *Noûs, 28,* (1), 21- 38.
Smith, M. (1993). Color, Transparency, Mind-Independence. In John
Haldane and Crispin Wright (Eds.): *Reality, Representation and
Projection.* Oxford: Oxford University Press
Stillings, N., A., Weisler, S., E., Chase, C., H., Feinstein, M., H.,
Garfield, J., L., & Rissland E., L. (1995). *Cognitive Science: An
Introduction.* Cambridge, MA: MIT Press
Thompson, E. (1995). *Colour Vision: A Study in Cognitive Science and
the Philosophy of Perception.* London and New York: Routledge.
Thompson, E. (2000). Comparative Color Vision: Quality Space and Visual
Ecology. In Steven Davis (Ed.): *Color Perception: Philosophical,
Psychological, Artistic and Computational Perspectives* (pp.
163-186). Oxford University Press, New York
Tye, M. (1995). *Ten Problems of Consciousness.* Cambridge MA: MIT
Press
Tye, M. (2000). *Consciousness, Color, and Content.* Cambridge MA:
MIT Press
Werner, J., S., & Wooten, B., R. (1979). Opponent chromatic mechanisms:
Relation to photopigments and hue naming. *Journal of the Optical
Society of America, 69,* 422-434.
--
--------------------------------------------------------------------------
Patrick Wilken
Postdoctoral Fellow in Biology, Caltech
"A person with a watch knows what time it is, a person with two is never sure."
-anon.