Tracking representationalism:

Lycan, Dretske, and Tye

David Bourget and Angela Mendelovici

February 26, 2013

(penultimate draft)

Abstract

This paper overviews the current status of debates on tracking rep-

resentationalism, the view that phenomenal consciousness is a matter
of tracking features of one’s environment in a certain way. We overview
the main arguments for the view and the main objections and chal-
lenges it faces. We close with a discussion of alternative versions of
representationalism that might overcome the shortcomings of tracking
representationalism.

1 Introduction1
There is something that it is like to be you. Perhaps you are currently
having a visual experience of something blue, tasting something sweet, or
feeling a twinge of pain in your foot. These experiences partly characterize
the phenomenal, qualitative, or subjective aspect of your mental life. This
phenomenon of there being something it is like for you to be in certain states
is phenomenal consciousness.
According to physicalism, consciousness is a physical phenomenon. Phys-
icalism offers an attractively simple ontology. The problem is that any pu-
tative description of conscious states in physical terms seems to leave out
1
This paper is thoroughly co-authored. It will appear in Andrew Bailey (ed.), Philos-
ophy of Mind: The Key Thinkers. Continuum (forthcoming)

1
their most crucial feature, their phenomenal character. It seems utterly in-
explicable why being in a certain physical state should be like anything at
all. As Colin McGinn (1989) puts it, it’s hard to see how the technicolor of
consciousness can arise from the physical and functional features of soggy
grey matter.
This chapter explores a relatively new strategy for understanding con-
sciousness in physical terms, tracking representationalism. It’s a two-step
strategy. The first step accounts for phenomenal consciousness in terms of
intentionality, the aboutness of mental states. The second step accounts for
intentionality in terms of a physical tracking relation to the environment.
We focus on the views of three of the most influential proponents of
tracking representationalism: William Lycan, Fred Dretske, and Michael
Tye. Section 2 describes tracking representationalism. Section 3 explores
some of its motivations. Section 4 overviews some of the challenges it faces.
Section 5 briefly discusses alternative versions of representationalism.

2 Tracking representationalism
Tracking representationalism is a theory of phenomenal consciousness, the
“what it’s like” of being in certain states. Mental states that exhibit phe-
nomenal consciousness are also known as phenomenal states or experiences.
Phenomenal states are said to have phenomenal properties.
Tracking representationalism aims to understand consciousness in terms
of another mental phenomenon: intentionality. Intentionality is the about-
ness or directedness of mental states. For example, you can think about
your mother, believe that you live on Earth, or desire that it rains. These
states exhibit a kind of directedness or aboutness. What a mental state is
about is its intentional content (or just its content).
Mental states that exhibit intentionality can be said to have intentional
properties, for example, the property of representing redness, or the property
of representing that you live on Earth. It is useful to distinguish between
pure and impure intentional properties. A pure intentional property is a
property of representing a certain content. An impure intentional property

2
is a property of representing a certain content in a certain manner. For
example, the property of representing redness is a pure intentional property,
while the property of representing redness in imagination is an impure in-
tentional property.2 We will say more about impure intentional properties
shortly.
It is fairly uncontroversial that thoughts, beliefs, and desires exhibit
intentionality. Many perceptual states seem to exhibit intentionality as well.
For example, visual experiences seem to present us with shapes, colors,
and other features of our environments. It is natural to describe them as
having contents involving shapes, colors, and other such features. Tracking
representationalism claims that all conscious states exhibit intentionality,
and that intentionality is the main ingredient in phenomenal consciousness.
More precisely, tracking representationalism combines three doctrines.
The first is representationalism:

Representationalism Every phenomenal property is identical to some

(pure or impure) intentional property.3

Intuitively, representationalism is the view that phenomenal consciousness

is just a special kind of intentionality. A mental state’s representational
nature exhausts its phenomenal nature.
Representationalism comes in two main varieties: pure representation-
alism and impure representationalism. Pure representationalism states that
phenomenal properties are identical to pure intentional properties. Impure
representationalism states that phenomenal properties are identical to im-
pure intentional properties.
2
We borrow this way of defining intentional properties from Chalmers (2004).
3
Lycan, Dretske, and Tye offer slightly different definitions of representationalism. For
Lycan, representationalism is the view that “the mind has no special properties that are
not exhausted by its representational properties, along with or in combination with the
functional organization of its components.” (Lycan, 1996, p. 11) Dretske defines represen-
tationalism as the view that “[a]ll mental facts are representational facts.” (Dretske, 1995,
p. xiii) Tye takes representationalism to be the view that “phenomenal character is one and
the same as representational content that meets certain further conditions.” (2000, p. 45)
We believe our working definition captures the core thesis defended by Lycan, Dretske,
and Tye.

3
 According to pure representationalism, phenomenal consciousness is a
matter of intentional content alone. What it is like to be in a mental state is
determined solely by what that mental state represents. The challenge fac-
ing pure representationalism is that it seems that states that represent the
same contents can nonetheless differ in phenomenal character. For example,
the following four states arguably involve the same intentional property of
representing redness, but involve different corresponding phenomenal prop-
erties:

(Perc-red) Perceptually experiencing unique (pure) red

(Think-red) Thinking about unique red

(Nonconc-red) Nonconsciously representing red (e.g. in early visual pro-

cessing)

(Belief-red) Having a standing belief about unique red

(Think-red) is an example of an occurrent conceptual state. It is occurrent

in that it is entertained, undergone, or active. It is conceptual in that it
involves concepts. Other occurrent conceptual states include beliefs and
desires that you are currently entertaining. (Belief-red) is an example of a
standing conceptual state, a state that involves concepts but is not occurrent.
(Nonconc-red) is an example of an occurrent state that may or may not be
conceptual and that we are in no sense aware of.
These cases involve distinct phenomenal characters. (Belief-red) and
(Nonconc-red) do not have phenomenal characters. While it is controversial
whether (Think-red) has phenomenal character, it is uncontroversial that it
has a different phenomenal character than (Perc-red). But if the same inten-
tional properties do not always give rise to the same phenomenal properties,
then phenomenal properties cannot be identified with intentional properties
and pure representationalism is false. For such reasons, most representation-
alists endorse impure representationalism, on which the manner in which a
mental state represents its content can make a difference to its phenomenal
character. Lycan, Dretske, and Tye are all impure representationalists.

4
 Impure representationalist views can be further divided into two main
types. According to one-manner impure representationalism, every phe-
nomenal property is identical to an impure intentional property of the form
representing C in manner M, where M is the same manner of representation
for all phenomenal states. M demarcates phenomenal from non-phenomenal
states, so we will refer to it as the demarcating manner of representation.
In effect, the one-manner view’s appeal to manners of representation serves
to preclude certain states, such as (Think-red), (Nonconc-red), and (Belief-
red), from having phenomenal properties. Within the class of states ex-
hibiting the demarcating manner, a mental state’s phenomenal properties
are determined by its intentional properties. Dretske and Tye both endorse
one-manner representationalism.
The second type of impure representationalism ascribes a larger role to
manners of representation. According to Lycan (1996), each sensory modal-
ity has a corresponding manner of representation. Visual experiences rep-
resent visually, auditory experiences represent aurally, and so on. Each of
these manners of representation factors into the resulting phenomenal char-
acter of the experience. On his view, content only determines phenomenal
character within a sensory modality. This is intra-modal representational-
ism.4
The second component of the tracking representationalism defended by
Lycan, Dretske, and Tye is reductionism about manners of representation:

Reductionism about manners of representation Manners of represen-

tation are physical or functional properties.

Dretske, Tye, and Lycan take the relevant manners of representation to be

reducible to physicalistically acceptable entities, such as functional roles or
evolutionary histories.
Both Dretske (1995) and Tye (1995, 2000) take the demarcating manner
to have two components, one that precludes occurrent and standing con-
ceptual states, like (Think-red) and (Belief-red), from having phenomenal
4
Weaker views are possible. For example, weak representationalism merely asserts that
phenomenal states are essentially representational.

5
character, and one that precludes states that we are in no sense aware of,
like (Nonconc-red), from having phenomenal character.
Dretske (1995) and Tye (1995, 2000) claim that phenomenal states have
a certain “poisedness” to impact on central cognition. They both take
poisedness to be part of the demarcating manner, though they understand
it slightly differently.5 According to Dretske and Tye, (Nonconc-red) and
other states that we are in no sense aware of do not supply information to
conceptual systems, and so they do not have phenomenal properties.
Dretske (1995) and Tye (1995, 2000) also take representing nonconcep-
tually to be part of the demarcating manner of representation. For Dretske,
nonconceptual contents are those that are represented by innate (or “sys-
temic” (Dretske, 1995, p. 12)) representations. Tye does not offer a precise
account of nonconceptual representation, but he makes some remarks that
suggest that it may be a matter of whether the representation allows us to
pick out instances of the same property on different occasions (Tye, 2000,
pp. 62-3). Since (Think-red) and (Belief-red) have purely conceptual con-
tents, they do not represent in the demarcating manner, and so they do not
have phenomenal properties.6
Lycan (1996) is less committal on how to characterize the relevant man-
ners of representation posited by intra-modal representationalism. He sug-
gests that manners are constituted by functional roles, but he does not
specify which roles.
The third component of tracking representationalism is the tracking the-
ory of intentionality:

The tracking theory of intentionality Intentionality is (or derives from)

a tracking relation.

According to the tracking theory, intentionality is a matter of detecting,

5
Tye understands it as being a matter of being poised to affect beliefs, desires, and other
conceptual states, while Dretske understands it as being a matter having the function of
supplying information to conceptual systems.
6
These accounts of the demarcating manner of representation compete with several
accounts of the neural correlate of consciousness found in the scientific literature. Much
of the scientific discussion of the problem of consciousness assumes something like the
tracking representationalist view (c.f. Crick 1994, Baars et al 1998, Edelman 1989).

6
carrying information about, or otherwise correlating with features of the en-
vironment. For example, the concept chair might represent chairs because
it detects, carries information about, or correlates with chairs.7
Lycan, Dretske, and Tye each has his own favorite refinement of the
tracking theory. On Dretske’s teleological view, a representational state
x represents the property F iff x has the function of indicating F (1995,
p. 2). The relevant notion of function is a teleological one; it is a matter of
something’s “job” or “purpose.” In cases of phenomenal states, the relevant
functions derive from their evolutionary history. Tye’s view is that a state
S of creature c represents that P just in case, “if optimal conditions were to
obtain, S would be tokened in c if and only if P were the case; moreover, in
these circumstances, S would be tokened in c because P is the case.”8 (2000,
p. 136) Lycan (1996) does not settle on a specific version of the tracking
theory, but endorses a broadly evolutionary approach.9
Recall that tracking representation aims to account for consciousness in
physical terms. Together with a physicalistically-kosher account of manners
of representation, the tracking theory is key to accomplishing this goal. On
tracking representationalism, consciousness is a matter of intentionality and
manners of representation. Intentionality is a matter of tracking, which is a
physical relation, and the manners of representation are functional roles or
evolutionary histories.

3 Motivations for tracking representationalism

In this section, we consider motivations for representationalism in general
(subsections 3.1-3.3) and motivations specifically for tracking representa-
tionalism (subsection 3.4).
7
Tracking theories need not claim that all intentional states get their content directly
from tracking. For instance, a common view is that all atomic representations get their
contents from tracking, while composite representations get their contents compositionally
from atomic states.
8
See Tye (2000, Chapter 6) for Tye’s account of optimal conditions.
9
The teleological approach to intentionality is defended by Millikan (1984), Neander
(1995), and Papineau (1993).

7
3.1 Sensory qualities
Lycan’s main consideration in favour of representationalism is that it pro-
vides a neat theory of the qualities we are aware of in perception (1987;
1996). When we introspect on our perceptual experiences, we notice various
qualities. For example, when you introspect upon a perceptual experience of
a blue circle, you notice qualities like blueness and roundness. One central
question about the nature of perception can be put as follows: What are
these qualities and where do they fit into our overall theory? Lycan puts
the problem in terms of phenomenal objects, where phenomenal objects are
the bearers of sensory qualities: What are phenomenal objects?
One view of perception, naive realism, takes phenomenal objects to be
external world objects. On this view, the qualities we are aware of in per-
ception are real properties of external objects. The circle you experience
exists in the external world, and blueness and roundness are properties of
this circle. Unfortunately, this view has trouble accounting for the qualities
we are aware of in hallucination. In hallucination, we are also aware of sen-
sory qualities, but there needn’t be an external object with such qualities.
If you are hallucinating a blue circle, there needn’t be anything blue and
circular before you. So it looks like phenomenal objects can’t be external
world objects after all.
Another view of perception, the sense data view, takes phenomenal ob-
jects to be mental objects, or sense data, and sensory qualities to be prop-
erties of sense data. When you see a blue circle, you have a blue and round
sense datum in your mind. Blueness and roundness are properties of the
sense datum. This view faces many challenges, but one is particularly rel-
evant for present purposes: It involves a commitment to apparently irre-
ducibly mental particulars, and so it seems incompatible with a physicalist
theory of the mind. This leads Lycan (1987) and many other theorists to
reject the view.
A third alternative, the adverbialist theory of perception, denies the
assumption that there are phenomenal objects. Sensory qualities are not
properties of something. They are not really properties at all. Instead,
adverbialism claims that sensory qualities are ways of perceiving. When

8
you see something blue, you perceive bluely. “Bluely” is an adverb that
modifies the verb “perceive,” which is why the view is called “adverbialism.”
Likewise, when you see a blue circle, you perceive bluely and roundly. But
problems arise in the case of complex experiences. What happens when you
perceive a blue circle and a red square? Do you perceive bluely, roundly,
redly, and squarely? The problem is that on this adverbialist treatment this
would be indistinguishable from perceiving a blue square and a red circle.
This problem is known as the many properties problem and was first put
forth by Frank Jackson (1977).10
Lycan (1987, 1996) argues that representationalism offers the most plau-
sible account of phenomenal objects. It allows us to maintain that sensory
qualities are properties of phenomenal objects, but deny that these phenom-
enal objects are external world objects or sense data. Instead, the represen-
tationalist can say that phenomenal objects are intentional objects, and
sensory qualities are represented properties of these intentional objects.11
What is an intentional object? On the most general characterization, in-
tentional objects are represented objects, or objects that intentional states
are about. For example, a fear that the economy is failing has the economy
as its intentional object, and a belief that Santa Claus is jolly has Santa
Claus as its intentional object. On Lycan’s view, then, when you see a blue
circle, your experience involves an intentional object that is blue and round.
In other words, your experience involves a represented object that is repre-
sented as having the properties of blueness and roundness. As the example
of believing that Santa Claus is jolly illustrates, intentional objects need not
exist. Thus, by taking phenomenal objects to be intentional objects, the
representationalist can allow that sometimes phenomenal objects do not ex-
ist. In a case of hallucination, you are aware of an intentional object having
sensory qualities, but this object happens not to exist.
While this argument for representationalism is best developed in Lycan
(1987), echoes of this reasoning are also found in Dretske (1995; 2003), Tye
10
The problem is further developed by Lycan (1987), who argues that the only available
adverbialist solution leads to intentionalism.
11
The view that the objects of perception are intentional objects can be traced back to
Anscombe (1965) and Hintikka (1969).

9
(1995), and Harman (1990). As we will see, the argument from transparency
bears some similarities to this argument as well.

3.2 The transparency of experience

Perhaps the best known argument for representationalism is the argument
from transparency. Gilbert Harman (1990) introduces this argument and
Tye (1991; 1995; 2000) further develops it.12
The argument from transparency is based on introspective observations
about what we do and do not notice in experience. In a nutshell, the intro-
spective observations are these:

1. When we pay attention to our experiences, we don’t notice qualitative

properties attributed to our experiences themselves.

2. When we pay attention to our experiences, the qualitative properties

that we notice are attributed to the objects of our experiences.

To borrow Harman’s (1990) famous example, when Eloise sees a tree, she
notices the greenness of the leaves, the brownness of the trunk, and the
overall shape of the tree. All the qualities that she notices seem to her to be
properties of external objects, not of her experience. Our experiences seem
to be characterized by properties of the sorts that external objects have, not
properties that sense data or other mental entities have. Moreover, intro-
spection reveals exactly the same properties whether or not corresponding
objects exist. All this suggests that the nature of experience is exhausted
by represented properties of represented objects.
Though advocates of transparency focus on visual experiences, the trans-
parency observation is supposed to hold for all experiences. When we in-
trospect on our auditory experiences, for example, we notice properties of
sounds, such as their loudness and pitch, but we don’t notice any qualitative
12
The transparency of experience has its historical roots in G. E. Moore (1903), but,
as Kind (2003) convincingly argues, the version of the transparency thesis Moore had in
mind was considerably weaker than the one Harman and Tye endorse.

10
features of the our auditory experiences themselves.13,14
The claim that experience is transparent has been challenged on various
grounds. Kind (2003) and Loar (2003) suggest that, though we are not
normally aware of qualitative features of our experiences, with some effort,
we can become aware of such features. But perhaps the most common
kind of objection comes in the form of examples of experiences in which
we do seem to be aware of qualitative features of experience, for example,
experiences involved in blurry vision. Since these potential counterexamples
to transparency are also potential counterexamples to representationalism,
we will discuss them in section 4.

3.3 Co-variation between intentional content and phenome-

nal character
The intentional properties of experience seem to covary with their phenom-
enal properties. Phenomenal greenness seems to go with represented green-
ness, phenomenal loudness with represented loudness, and so on. Put other-
wise, if two experiences differ in their intentional properties, then they also
differ in their phenomenal properties, and vice versa. Representationalism
can explain this: Phenomenal properties just are a species of intentional
properties.
Tye offers a version of this motivation as part of his argument from
transparency (2000, p. 48). Suppose your experience of a ripe tomato has
a particular intentional property. It also has a particular phenomenal prop-
erty. If we change the intentional property, say by changing the represented
color of the tomato, then your phenomenal properties change too. Tye
claims that this holds for all experiences. If two experiences differ in their
intentional contents, then they also differ in their phenomenal characters.
13
Harman’s (1990) argument from transparency stops there. Tye (2000)’s version of the
argument from transparency, however, combines the transparency observations with an
additional observation, which we will consider in the next subsection.
14
It is important to note that Harman only denies that we are aware of any qualita-
tive features of our experiences. This is consistent with the possibility of introspecting
non-qualitative features of experiences, for example, their representational or temporal
features. Some authors take the transparency thesis to rule out such observations. But
this makes the transparency thesis stronger than required by Harman’s argument.

11
In other words, intentional properties supervene on phenomenal properties.
Tye’s overall argument for representationalism is an inference to the
best explanation: The best explanation of the observed relationship between
phenomenal properties and intentional properties and the transparency ob-
servation is that phenomenal properties are just a species of intentional
properties.
Frank Jackson (2004, p. 109) offers the reverse observation in favor of
representationalism: If two experiences differ in their phenomenal proper-
ties, then they also differ in their intentional properties. For example, a
visual experience of a red pen has certain phenomenal properties, and it
has certain intentional properties. Suppose we change your experience so
that it has different phenomenal properties, say, by switching the red pen
with a blue pen. Now your experience also has different intentional prop-
erties. Jackson takes similar observations to hold across all experiences. If
this is right, then phenomenal properties supervene on intentional proper-
ties. But then, Jackson argues, intentional properties suffice for phenomenal
properties.15

3.4 Providing a physicalist theory of consciousness

As we mentioned at the outset, phenomenal consciousness seems to resist
physicalist treatment. By providing a two step reduction of consciousness to
tracking, tracking representationalism offers an attractive physicalist theory
of consciousness. The first step reduces consciousness to intentionality and
manners of representation. This steps relies on the arguments discussed in
subsections 3.1-3.3. The second step reduces intentionality to a species of
tracking relation, and manners of representations to broadly physical prop-
erties. The reduction of intentionality to tracking is somewhat plausible
since many of the currently popular theories of intentionality are tracking
theories. Combining the two steps, then, offers us a reduction of phenomenal
consciousness to broadly physical properties. The technicolour of experience
is just a matter of tracking relations between internal states and properties
15
Byrne (2001) also develops an argument along these lines.

12
in the world, when such relations have the right additional physical prop-
erties to be conscious. 3.5 The mapping problem The mapping problem of
consciousness is the problem of specifying a model that can predict which
phenomenal states accompany which physical states. The mapping problem
is independent of the ontological problem of determining the metaphysical
nature of phenomenal consciouness, i.e. whether it is physical or irreducibly
mental. For example, knowing that conscious states are physical would not
by itself tell us which physical states are identical to which phenomenal
states. Conversely, knowing which physical states go with which phenom-
enal states would not tell us whether the correlated states are identical or
related in some other way instead. While philosophers have traditionally
focused on the ontological problem, solving the mapping problem is crucial
for answering various practical questions concerning consciousness, e.g. Can
comatose patients feel pain? Could electronic circuits ever feel anything? Is
there anything it is like to be an ant?
Tracking representationalism underpins all existing theories of conscious-
ness that could provide a solution to the mapping problem, including those
that have been advanced by neuroscientists and psychologists (e.g. Crick
1994, Baars et al 1998, Edelman 1989). These possible solutions share three
components corresponding to the three components of the tracking repre-
sentationalist view. First, phenomenal properties are associated 1-to-1 with
intentional contents (this component is secured by the one-manner repre-
sentationalist view). The second component is a biological or functional
characterization of the property that distinguishes physical states that are
associated with phenomenal states from other physical states (e.g., Crick’s
40hz hypothesis, or Dretske and Tye’s accounts of the demarcating man-
ners). The last component is the identification of the intentional contents of
experiences with what their neural vehicles track in the outside world (this
is secured by the tracking theory of intentionality). Taken together, these
components specify a 1-to-1 correspondence between phenomenal properties
and physical states or physical state types, thereby purporting to solve the
mapping problem. All theories of consciousness that come anywhere close
to specifying such a correspondence consist in the three above components;

13
they are all forms of tracking representationalism.
Alternative approaches are not merely hard to justify, they are hard to
imagine or even formulate. It is very hard to see how else we could specify
a general relation between phenomenal states and physical states than by
identifying the intentional contents of the former with the tracked contents
of the latter. In our view, this may be the best available motivation for
tracking representationalism.16

4 Objections
This section considers objections to tracking representationalism. Some are
directed specifically at tracking versions of representationalism (subsection
4.1). Others are directed at representationalism in general (subsection 4.2).
We close with objections to impure representationalism (subsection 4.3).

4.1 Objections to tracking representationalism

4.1.1 4.1.1 Inverted Earth

Ned Block (1990; 1996) argues that the combination of representationalism

with a tracking theory of intentionality leads to implausible consequences.
Imagine a planet just like the Earth except for two small differences: First,
all colors have been inverted. On Inverted Earth, ripe bananas are blue,
blueberries are yellow, ripe tomatoes are green, and grass is red. Second,
color names are also inverted. For example, inhabitants of Inverted Earth
use the word “yellow” for the color that we call “blue.” One night, you are
transported to Inverted Earth without your knowledge. Before you wake
up, a spectrum inverting device is inserted into your optic nerve (again,
without your knowledge). As a result, you notice nothing unusual about the
colors of things when you wake up. It is highly plausible that you could in
principle never figure out that you are now on a planet with inverted colors.
16
However, we are skeptical that tracking representationalism succeeds at solving the
mapping problem. It is not clear to us that tracking under a physically- or functionally-
specified manner correlates well with phenomenal consciousness. Some of the objections
in section 4 speak to this point.

14
In particular, you might go on living the rest of your life there without ever
noticing anything unusual about the colors of things. This seems to be a
perfectly consistent scenario.
This apparent possibility is inconsistent with some forms of tracking
representationalism. On Earth, you had a brain state s that tracked the color
red. After you’ve spent some time on Inverted Earth, this state will come
to track the color green. As a result, tracking representationalism seems
to predict that after sufficient time has passed you would experience green
when you are in s on Inverted Earth. For example, you should eventually
come to experience green when in the presence of ripe tomatoes. But it
seems clear that you would not, given the spectrum inverted device that has
been implanted in you.
This argument does not work against all versions of tracking representa-
tionalism. In particular, it does not bear on views on which what s represents
is the same on Earth and Inverted Earth. This is the case on teleological the-
ories. Take for example Dretske’s view that what an experience represents
is a matter of what it has the biological function of indicating. Your internal
states do not acquire new biological functions on Inverted Earth (those were
fixed in the course of evolution on Earth), so this kind of tracking view is
compatible with the scenario as described.

4.1.2 Swamp-person

While teleological views easily escape the objection from Inverted Earth,
they face another powerful objection: the Swamp-person objection.17
Though it is highly unlikely, it seems perfectly possible that a molecule
for molecule duplicate of you could suddenly come into existence in a swamp
as a result of a lightning strike. Moreover, it seems plausible that this
Swamp-person would have conscious experiences. A Swamp-person could
see the swamp, feel the mud, and have all the other experiences that you
would have if you suddenly found yourself in a swamp.
The problem for teleological versions of tracking representationalism is
that Swamp-person’s mental states have no evolutionarily-determined bio-
17
This objection was inspired by a thought experiment from Davidson (1987).

15
logical functions, since Swamp-person has no evolutionary history. But then
they do not have the biological functions relevant to determining intentional
content. This means that they cannot represent anything by the lights of
teleological accounts of intentionality. Given the kind of representational-
ist view under consideration, this in turn implies that Swamp-person would
have no experiences, which seems implausible. Inverted Earth and Swamp-
person together constitute a dilemma for tracking representationalism.18

4.1.3 Causal efficacy

Tracking representationalism also faces issues regarding the causal role of

experiences. This objection applies to all views on which phenomenal prop-
erties are wide properties, that is, properties that involve factors outside
of one’s body at the time of their instantiation. Tracking representational-
ism implies that phenomenal properties are wide properties because it takes
them to involve relational properties involving external objects and/or his-
torical events.
Suppose you are looking for a tomato. You open your refrigerator, lean
forward, and see a tomato on a shelf. As a result, you extend your arm, grab
the tomato, and close the door. It seems plausible that your arm movement
is caused by the combination of mental states you are in. Simplifying a little,
it seems that the movement is caused by (i) your desire to take a tomato out
of the refrigerator, (ii) your visual experience of a tomato in the refrigerator,
and (iii) your belief that you can pick up the tomato thus experienced by
making the relevant arm movement. But if the property instantiated in the
visual experience in question is a wide property involving such factors as your
evolutionary history or a link between your internal brain states and features
of your environment, it cannot play an immediate causal role in bringing
about your arm movement. Wide properties cannot have immediate effects
on one’s movements. In principle, it should be possible to fully explain your
arm movement in terms of what went on in your head and the rest of your
18
Tye (2000) attempts to offer an account of the intentionality of phenomenal states
that avoids the two horns of the dilemma constituted by these objections. In our opinion,
the resulting account seems implausibly ad hoc.

16
body at the time. If phenomenal properties are a kind of tracking property,
they are not immediately efficacious in bringing about behavior.19
A typical reply to this objection is to grant the point but try to find
some other explanatory role for wide experiences to play. For example,
Dretske (1995, p. 196) grants that on his view phenomenal properties cannot
be immediate causes of bodily movements. However, he goes on to argue
that they can play another kind of explanatory role. For example, the
fact that a certain phenomenal state has the function of tracking a certain
external property might seem to explain why its tokens tend to occur in the
presence of instances of the property. Phenomenal states might well have
an explanatory role on Dretske’s and similar views. However, this does not
erase the fact that tracking intentionalism is incompatible with the seemingly
obvious fact that instantiations of phenomenal properties have immediate
behavioral effects. The objection was not that tracking intentionalism makes
phenomenal properties explanatorily redundant, but that it is incompatible
with their playing an immediate role in bringing about bodily movements.

4.1.4 The mismatch problem

According to Tye’s, Dretske’s, and Lycan’s versions of tracking representa-

tionalism, the intentional content of a mental state is what it tracks and
the phenomenal character of a mental state is its intentional content. In
short, the phenomenal character of a mental state is what it tracks.20 The
problem is that the phenomenal characters of some of our experiences are
not plausibly identified with the features in our environment that we track.
Consider an experience of red. This state plausibly tracks something like
the disposition to reflect electromagnetic radiation with a wavelength of ap-
proximately 650 nm. The problem is that this does not seem to be plausibly
19
See Horgan (1991).
20
Tye, Dretske, and Lycan endorse an identity version of representationalism, on which
phenomenal features are identical to intentional features and phenomenal character is
identical to intentional content. It is also possible to hold a mere determination version
of representationalism, on which phenomenal features are not identical to intentional fea-
tures, but instead are determined by them in some way. The objection in this section
applies to this weaker type of view as well. See fn. 22.

17
identifiable with the experience’s phenomenal character, what it’s like to
have the experience. The problem is that phenomenal redness and surface
reflectance properties seem utterly dissimilar. Many other qualities we are
aware of in our experiences seem to have no matching properties in our en-
vironment: the painfulness of pain, the sound of a harp, and the feeling of
heat are not plausibly identified with physical properties of things in our
bodies or environment.21,22
Tracking representationalists might reply to this worry by appealing to
the phenomenal concepts strategy. The apparent distinctness between phe-
nomenal redness and surface reflectance properties is an illusion created by
the fact that we have two different ways of representing the same property.
Somehow, the different operating principles of the two ways of representing
are incompatible, making it hard for us to see phenomenal redness is the
same thing as some reflectance property. Whether this sort of reply can
succeed is a hotly debated topic, but we remain skeptical.23

4.1.5 Phenomenal externalism

Many of the above objections to the tracking component of tracking repre-

sentationalism arise from its commitment to phenomenal externalism, the
view that the phenomenal character of a mental state is at least partly de-
termined by factors outside the subject’s body. According to phenomenal
21
Since the tracking representationalist wants to be a physicalist, she should not re-
spond to these worries by claiming that there are non-physical properties corresponding
to redness, painfulness, etc.
22
In footnote 20, we noted that there are possible versions of representationalism on
which phenomenal character is not identical to, but is instead merely determined by, in-
tentional content. These views escape the present objection in its current form, since
they do not claim that phenomenal characters are identical to tracked contents. However,
the apparent dissimilarity between the phenomenal character of experiences of red and
what they track poses problems for this view as well. The problem is that there seems to
be nothing special about the tracked content that is responsible for its having the phe-
nomenal character of redness, rather than the phenomenal character of greenness or the
phenomenal character of pain. In other words, there appears to be no clear connection
between the nature of certain surface reflectance properties and the nature of the phe-
nomenal characters that they are supposed to determine, making it baffling how exactly
the determination is supposed to take place.
23
See Balog (1999, 2012), Chalmers (2004, 2007), Stoljar (2005), and Alter and Walter
(2007)

18
externalism, two intrinsic duplicates (duplicates from the skin in) might dif-
fer in their phenomenal properties. The Swamp-person and Inverted Earth
objections target specific versions of phenomenal externalism on which phe-
nomenal properties are partly determined by historical or non-historical en-
vironmental factors, respectively, while the causal efficacy objection and the
mismatch problem are problematic for all versions.24
Phenomenal externalism itself might seem outlandish independently of
any downstream consequences. Dretske, Lycan, and Tye, however, have
each offered positive arguments for phenomenal externalism (Dretske 1996;
Lycan, 2001; Tye, forthcoming).

4.2 General objections to representationalism

4.2.1 Perceptual distortion

The objections in the preceding section are targeted specifically at tracking

representationalism. Other objections apply to representationalism indepen-
dently of any commitment to the tracking theory. Most of these objections
involve cases of perceptual distortion. These are cases in which what we ex-
perience is distorted compared to the way we take the world to be, but we do
not tend to regard our experiences as misrepresenting. All objections of this
sort aim to show that representationalism is false because some experiences
in the same sensory modality can differ in phenomenal character without
differing in content. If true, this would imply that there is something more
to experience than representing a certain content in the kinds of manners
described by Lycan, Dretske, and Tye.
Blurry vision is the most widely discussed example of perceptual distor-
tion. When you see blurry, you do not tend to think of your experience as
presenting you with some blurry or fuzzy object. This makes it hard to see
what intentional contents might characterize blurry vision. This line of argu-
24
Jackson (2004) combines representationalism with an internalist version of something
like the tracking theory. On Jackson’s view, intentionality is a matter of our relations
to the environment, but intrinsic duplicates are related to the environment in the same
ways. This view avoids the Inverted Earth and Swamp-person objections, but it is unclear
whether it avoids the causal efficacy objection and the mismatch problem.

19
ment against representationalism originates from Boghossian and Velleman
(1989).
Dretske’s (2003) view is that blurry experiences in fact do represent
fuzzy objects. Tye (2003) suggests that what characterizes blurry experi-
ences is that their contents leave the contours of objects indeterminate in a
certain way. A more recent proposal by Allen (forthcoming) is that blurry
experiences represent their objects as having multiple contours.
It is difficult to adjudicate the question of what blurry experiences rep-
resent qua blurry experiences. However, we suggest that blurry experiences
can easily be seen to differ in representational content compared to non-
blurry experiences, because blurry experiences clearly leave out some infor-
mation about the world. Blurry experiences always involve a deficiency in
detail.25
Peacocke (1983) presents a set of objections from perceptual distortion
against the view that the phenomenal character of an experience is deter-
mined by its content. The best known of these objections involves two
experiences in which one sees two identical trees from different distances.
Tree 1 and Tree 2 do not look different in size, and so the experiences of
Tree 1 and Tree 2 do not differ in content. But these experiences differ in
phenomenal character corresponding to the different apparent sizes of the
trees. So it does not seem that the perspectival aspects of the phenomenal
character of visual experiences are captured by their contents.
Lycan (1996), Harman (1990), and Tye (1996) offer accounts of the
perspectival aspect of vision. According to Lycan, we need two layers of
content to explain perspective: a layer that represents objects in objective
three-dimensional space, and a layer that represents colored shapes in a two-
dimensional space. In the case of the two trees, the two experiences are alike
with respect to the first layer of content but their second layers of content
involve different sized shapes. Harman and Tye suggest that the difference
in content between the two tree experiences is one in situation-dependent
properties such as being large from here.
25
See Bourget (Ms.).

20
4.2.2 Allegedly contentless phenomenal states

Pains, moods, and emotions are challenging for representationalism because

it is not at all clear what they represent. Even if they have intentional
properties, it is not clear that their phenomenal properties can be identified
with any of their intentional properties.
Tye (2008) suggests that pains represent bodily damage at a bodily loca-
tion, and that they represent the damage as bad. Different pains represent
different types of bodily damage. For example, a stabbing pain represents
sudden damage at a well-defined location, whereas an ache represents inter-
nal damage at a vaguely defined location. These contents determine pain’s
phenomenal character.
Tye’s view attributes two different types of content to pains: bodily
damage, and badness. Different kinds of worries arise for each content.
While bodily damage is plausibly tracked, it is not clear that it captures the
phenomenal features of pain. The second component, the representation of
badness, does seem to capture an aspect of pain’s phenomenal character.
However, it is not clear that the tracking representationalist can appeal to
this content, since it is not clear that this is a content that we can track.
These worries are specific instances of the mismatch problem discussed in
subsection 4.1.4.26
According to Tye, emotions represent objects as (1) having evaluative
features (e.g. as being dangerous, invasive, or foul), and (2) causing or being
accompanied by a bodily disturbance (e.g. a racing heart, or perspiration).27
As in the case of pain, questions arise as to whether these contents can both
be tracked and capture emotions’ phenomenal characters.
Moods seem to escape this kind of treatment, since they do not seem
to represent external items at all. For example, an experience of sudden
elation, free-floating anxiety, or pervasive sadness do not seem to qualify
any object. There are several options open to the representationalist. Tye
26
Bain (2003), Klein (2007), Seager and Bourget (2007), Bourget (2010), and Mende-
lovici (2010) offer alternative representationalist theories of pain.
27
Seager (2002), Montague (2009), and Mendelovici (forthcoming) also defend represen-
tationalist accounts of emotions.

21
has suggested that moods represent departures from the “range of physical
states constituting functional equilibrium” (1995, p. 129). More recently,
Tye (2008) has suggested that at least some moods represent the world in
general as having affective properties.28 Mendelovici (forthcoming) suggests
that moods represent the same properties as their corresponding emotions,
but that they do not represent any objects as having these properties. Moods
represent mere properties. Kind (forthcoming) objects that all these views
fail to capture the phenomenal character of moods.

4.3 Objections to reductionism about manners of represen-

tation
Objections have also been raised against the reductive view of manners
of representation that is part of tracking representationalism. The spe-
cific proposals put forward by Dretske and Tye have been questioned on
a number of grounds. On their view, what distinguishes phenomenal from
non-phenomenal representation (beyond differences in content) is that the
former, but not the latter, are nonconceptual and relevantly poised to influ-
ence cognition. One simple objection to this view is that some subconscious
states influence cognition and have nonconceptual content without being
phenomenally conscious.29 The case of blindsight also seems to show that
being poised and nonconceptual is not sufficient for a representation to be
phenomenally conscious (Block, 1995).
The appeal to nonconceptual content might also be questioned on other
grounds. According to Dretske and Tye, only nonconceptual contents con-
tribute to phenomenal character. This precludes thoughts from having phe-
nomenal characters, since they are conceptual states. This might seem im-
plausible to some (see Horgan and Tienson (2002), and Pitt (2004)). Relat-
edly, it is quite plausible that conceptual contents are sometimes involved
in perceptual states and are responsible for certain phenomenal features of
those states. For example, there is a phenomenal difference between seeing
28
Tye describes such states as emotions, but many would classify them as moods.
29
See Seager and Bourget (2007).

22
the duck-rabbit as a duck versus seeing it as a rabbit.30
It is noteworthy that this approach to demarcating properties is open to
general objections to physicalism and functionalism. For example, Chalmers’
(1996) zombie argument applies. It is on the face of it conceivable that one
has brain states that track features in one’s environment while being poised
and nonconceptual, but one has no phenomenal experiences. If conceiv-
ability is a guide to possibility, this suggests that such a scenario is possi-
ble, which is incompatible with the claim that consciousness is a matter of
poised, nonconceptual tracking. Jackson’s (1982) knowledge argument also
applies without modification. See chapter XXXX for more on these debates.
That the usual objections to functionalism and physicalism apply suggests
that tracking representationalism might not make the ontological problem
of consciousness easier after all. 5.

5 Other kinds of representationalism

Most of the objections discussed in previous sections are not effective against
representationalism on its own. They target representationalism in combi-
nation with the tracking theory of intentionality and/or reductionism about
manners of representation. This has led several authors to reject the track-
ing theory and/or the reductionist view of manners of representation while
retaining the representationalist component.31
One alternative to tracking representationalism is mere representation-
alism. Chalmers (2004) and Pautz (2010a,b) endorse representationalism
without endorsing a physicalist view of intentionality or manners of repre-
sentation. Mere representationalism does not aim to provide a physicalist
theory of consciousness or a solution to the mapping problem, but it might
still seem to improve our understanding of consciousness. It at least provides
us with a better understanding of the internal structure of consciousness,
30
These objections are discussed in Mendelovici (2010). More objections to Dretske’s
and Tye’s accounts of demarcating properties can be found in Seager (1999; 2003), Kriegel
(2002), Byrne (2001, 2003), and Seager and Bourget (2007).
31
Crane (2003), Chalmers (2004; 2006), Seager and Bourget (2007), Pautz (2009; 2010),
Mendelovici (2010), Bourget (2010).

23
which might help formulate hypotheses regarding its relation to physical
structures. Mere representationalism also retains the benefits of reductive
representationalism as far as the problem of hallucination discussed earlier
goes.
Another alternative to tracking representationalism, the pure represen-
tationalist view briefly discussed in section 2, challenges the assumption that
manners of representation are needed to explain consciousness. On this view,
consciousness is simply intentionality: the difference between conscious and
nonconscious representations is simply a difference in content. One differ-
ence in content that might be relevant is that nonconscious intentional states
have less determinate contents than phenomenal experiences. Other differ-
ences could be relevant. Bourget (2010) and Mendelovici (2010) defend pure
representationalist theories of consciousness.

6 Conclusion
Tracking representationalism is a relatively new theory of phenomenal con-
sciousness. As our cursory overview shows, it has generated much discussion
over the past few years. Whether or not it ultimately succeeds, it has chal-
lenged and reshaped the contemporary understanding of the relationship
between consciousness and intentionality.

Bibliography
Keith Allen (forthcoming). Blur. Philosophical Studies.
G. E. M. Anscombe (1965). The intentionality of sensation: A grammatical
feature. In Ronald J. Butler (ed.), Analytic Philosophy. Blackwell.
Torin Alter and Sven Walter (eds.) (2007). Phenomenal Concepts and
Phenomenal Knowledge: New Essays on Consciousness and Physicalism.
Oxford University Press.
J. L. Austin (1962). Sense and Sensibilia. Oxford University Press.
Bernard J. Baars, J. B. Newman and John G. Taylor (1998). Neuronal
mechanisms of consciousness: A relational global workspace approach.

24
 In Stuart R. Hameroff, Alfred W. Kaszniak and A. C. Scott (eds.), To-
ward a Science of Consciousness II. MIT Press.
Katalin Balog (2012). In Defense of the Phenomenal Concept Strategy.
Philosophy and Phenomenological Research 84 (1):1-23.
Katalin Balog (1999). Conceivability, possibility, and the mind-body prob-
lem. Philosophical Review 108 (4):497-528.
David Bain (2003). Intentionalism and pain. Philosophical Quarterly 53
(213):502-523.
Ned Block (1996). Mental paint and mental latex. Philosophical Issues
7 :19-49.
Ned Block (1995). On a confusion about a function of consciousness. Brain
and Behavioral Sciences 18 :227-247.
Ned Block (1990). Inverted earth. Philosophical Perspectives 4 :53-79.
Paul A. Boghossian and J. David Velleman (1989). Color as a secondary
quality. Mind 98 (January):81-103.
David Bourget (Ms.), A General Reply to the Arguments from Blur, Double
Vision, Perspective, and Other Kinds of Perceptual Distortion Against
Representationalism.
David Bourget (2010a). Consciousness is Underived Intentionality. Noûs 44
(1):32-58.
David Bourget (2010b). The representational theory of consciousness. Dis-
sertation, Australian National University.
Alex Byrne (2003). Consciousness and nonconceptual content. Philosophical
Studies 113 (3):261-274.
Alex Byrne (2001). Don’t PANIC: Tye’s intentionalist theory of conscious-
ness. A Field Guide to the Philosophy of Mind.
Alex Byrne (2001). Intentionalism defended. Philosophical Review 110
(2):199-240.
Alex Byrne and Michael Tye (2006). Qualia ain’t in the head. Noûs 40
(2):241-255.
David J. Chalmers (2006). Perception and the fall from Eden. In Tamar
S. Gendler and John Hawthorne (eds.), Perceptual Experience. Oxford
University Press.

25
David J. Chalmers (2004). The representational character of experience. In
Brian Leiter (ed.), The Future for Philosophy. Oxford University Press.
David J. Chalmers (2004). Phenomenal concepts and the knowledge ar-
gument. In Peter Ludlow, Yujin Nagasawa and Daniel Stoljar (eds.),
There’s Something About Mary: Essays on Phenomenal Consciousness
and Frank Jackson’s Knowledge Argument. MIT Press.
David J. Chalmers (2007). Phenomenal concepts and the explanatory gap.
In Torin Alter and Sven Walter (eds.), Phenomenal Concepts and Phe-
nomenal Knowledge: New Essays on Consciousness and Physicalism.
Oxford University Press.
Tim Crane (2003). The intentional structure of consciousness. In Quentin
Smith and Aleksandar Jokic (eds.), Consciousness: New Philosophical
Perspectives. Oxford University Press. Crick, Francis (1994). The As-
tonishing Hypothesis: The Scientific Search for the Soul. Scribners.
Brian Cutter and Michael Tye (2011). Tracking representationalism and the
painfulness of pain. Philosophical Issues 21 (1):90-109.
Donald Davidson (1987). Knowing one’s own mind. Proceedings and Ad-
dresses of the American Philosophical Association 60 (3):441-458.
Fred Dretske (2003). Experience as representation. Philosophical Issues 13
(1):67-82.
Fred Dretske (1996). Phenomenal externalism, or if meanings ain’t in the
head, where are qualia? Philosophical Issues 7 :143-158.
Fred Dretske (1995). Naturalizing the Mind. MIT Press.
Fred Dretske (1993). Conscious experience. Mind 102 (406):263-283.
Gerald M. Edelman (1989). The Remembered Present: A Biological Theory
of Consciousness. Basic Books.
Katalin Farkas (2008). Phenomenal intentionality without compromise. The
Monist 91 (2):273-93.
Gilbert Harman (1990). The intrinsic quality of experience. Philosophical
Perspectives 4 :31-52.
Jaakko Hintikka (1969). The logic of perception. In Jaakko Hintikka (ed.),
Models for Modalities. Reidel.

26
Terence E. Horgan (1991). Actions, reasons, and the explanatory role of
content. In Brian P. McLaughlin (ed.), Dretske and His Critics. Black-
well.
Terence E. Horgan and John L. Tienson (2002). The intentionality of
phenomenology and the phenomenology of intentionality. In David J.
Chalmers (ed.), Philosophy of Mind: Classical and Contemporary Read-
ings. Oxford University Press.
Frank Jackson (2004). Representation and experience. In Hugh Clapin,
Phillip Staines and Peter Slezak (eds.), Representation in Mind. Elsevier.
Frank Jackson (1977). Perception: A Representative Theory. Cambridge
University Press.
Frank Jackson (2004). Representation and experience. In Hugh Clapin,
Phillip Staines and Peter Slezak (eds.), Representation in Mind. Elsevier.
Amy Kind (2003). What’s so transparent about transparency? Philosophical
Studies 115 (3):225-244.
Colin Klein (2007). An Imperative Theory of Pain. Journal of Philosophy
104 (10):517-532.
R. Kraut (1982). Sensory states and sensory objects. Noûs 16 (May):277-93.
Uriah Kriegel (2011). The Sources of Intentionality. Oxford University
Press.
Uriah Kriegel (2002). PANIC theory and the prospects for a representational
theory of phenomenal consciousness. Philosophical Psychology 15 (1):55-
64.
David Lewis (1983). Individuation by acquaintance and by stipulation.
Philosophical Review 92 (1):3-32.
Brian Loar (2003). Transparent experience and the availability of qualia. In
Quentin Smith and Aleksandar Jokic (eds.), Consciousness: New Philo-
sophical Perspectives. Oxford University Press.
William G. Lycan (2001). The case for phenomenal externalism. Philosoph-
ical Perspectives 15 (s15):17-35.
William G. Lycan (2001). A simple argument for a higher-order represen-
tation theory of consciousness. Analysis 61 (269):3-4.
William G. Lycan (1996). Consciousness and Experience. MIT Press.

27
William G. Lycan (1996). Layered perceptual representation. Philosophical
Issues 7 :81-100.
William G. Lycan (1987). Phenomenal objects: A backhanded defense.
Philosophical Perspectives 3 :513-26.
William G. Lycan (1987). Consciousness. MIT Press.
Colin McGinn (1989). Can we solve the mind-body problem? Mind 98
(July):349-66.
Angela Mendelovici (forthcoming). Reliable Misrepresentation and Tracking
Theories of Mental Representation. Philosophical Studies.
Angela Mendelovici (2010). Mental Representation and Closely Conflated
Topics. Dissertation, Princeton University.
Ruth G. Millikan (1984). Language, Thought and Other Biological Cate-
gories. MIT Press.
Michelle Montague (2009). The logic, intentionality, and phenomenology of
emotion. Philosophical Studies 145 (2):171-192.
G. E. Moore (1903). The refutation of idealism. Mind 12 (48):433-453.
Karen Neander (1995). Misrepresenting and malfunctioning. Philosophical
Studies 79 (2):109-41.
David Papineau (1993). Philosophical Naturalism. Blackwell.
Adam Pautz (2010a). An argument for the intentional view of visual expe-
rience. In Bence Nanay (ed.), Perceiving the World. Oxford University
Press.
Adam Pautz (2010b). Why explain visual experience in terms of content?
In Bence Nanay (ed.), Perceiving the World. Oxford University Press.
Adam Pautz (2009). What are the contents of experiences? Philosophical
Quarterly 59 (236):483-507.
David Pitt (2004). The phenomenology of cognition, or, what is it like to
think that P? Philosophy and Phenomenological Research 69 (1):1-36.
William E. Seager (2003). Tye on consciousness: Time to panic? Philosoph-
ical Studies 113 (3):237-247.
William E. Seager (2002). Emotional introspection. Consciousness and
Cognition 11 (4):666-687.

28
William E. Seager (1999). Theories of Consciousness: An Introduction and
Assessment. Routledge.
William E. Seager and David Bourget (2007). Representationalism about
consciousness. In Max Velmans and Susan Schneider (eds.), The Black-
well Companion to Consciousness. Blackwell.
Wilfrid S. Sellars (1975). The adverbial theory of the objects of sensation.
Metaphilosophy 6 (April):144-160.
Daniel Stoljar (2007). Consequences of intentionalism. Erkenntnis (Special
Issue) 66 (1-2):247-70.
Daniel Stoljar (2005). Physicalism and phenomenal concepts. Mind and
Language 20 (2):296-302.
Michael Tye (forthcoming). Phenomenal externalism, lolita, and the planet
xenon. In Terence E. Horgan and David Sosa (eds.), Collection on the
Philosophy of Jaegwon Kim. MIT Press.
Michael Tye (2009). Consciousness Revisited: Materialism Without Phe-
nomenal Concepts. MIT Press.
Michael Tye (2008). The experience of emotion: An intentionalist theory.
Revue Internationale de Philosophie 62 :25-50.
Michael Tye (2003). Blurry images, double vision, and other oddities: New
problems for representationalism? In Quentin Smith and Aleksandar
Jokic (eds.), Consciousness: New Philosophical Perspectives. Oxford
University Press.
Michael Tye (2000). Consciousness, Color, and Content. MIT Press.
Michael Tye (1996). Perceptual experience is a many-layered thing. Philo-
sophical Issues 7 :117-126.
Michael Tye (1995). Ten Problems of Consciousness: A Representational
Theory of the Phenomenal Mind. MIT Press.

29