Reducing Consciousness by Making it Hot
A Review of Peter Carruthers' Phenomenal Consciousness
Robert W. Lurz
Department of Philosophy
Brooklyn College, CUNY
2900 Bedford Avenue
Brooklyn, NY 11210
U.S.A.
rlurz@brooklyn.cuny.edu
Copyright (c) Robert W. Lurz 2002
PSYCHE, 8(05), September 2002
Previously held: http://psyche.cs.monash.edu.au/v8/psyche-8-05-lurz.html
KEYWORDS: Phenomenal consciousness, worldly-subjectivity, experiential-subjectivity, higher-order and first-order theories.
REVIEW OF: Peter Carruthers. (2000). Phenomenal Consciousness: A Naturalistic Theory. Cambridge: Cambridge University Press. $60 hbk.
347 pp. ISBN: 0521781736.
ABSTRACT: Our conscious experiences are said to possess a unique property called phenomenal consciousness. Why these and only these
states of us have this property has proved to be an exceedingly
difficult question for philosophers and scientists to answer. In fact,
some have claimed that this question constitutes the hard problem of
the mind-body problem, one which cannot be solved by the standard
methods of contemporary science. In his most recent book, Phenomenal
Consciousness, Peter Carruthers offers a bold, original and
scientifically acceptable solution to this hard problem: the
dispositional higher-order thought (HOT) theory. I describe the main
line of argument in Phenomenal Consciousness for Carruthers'
dispositional HOT theory and present three places where the argument
seems most vulnerable. I end the review with a very positive endorsement
of Phenomenal Consciousness, recommending it as compulsory reading for
anyone interested in the contemporary philosophical and scientific
debate over the nature of phenomenal consciousness.
1. Introduction
Our conscious experiences are said to possess a unique property called
phenomenal consciousness. Why these and only these states of us have
this property has proved to be an exceedingly difficult question for
philosophers and scientists to answer. In fact, some have claimed that
this question constitutes the hard problem of the mind-body problem,
one which cannot be solved by the standard methods of contemporary
science. Peter Carruthers strongly disagrees. In his most recent book,
Phenomenal Consciousness, Carruthers offers a bold, original and
scientifically acceptable solution to this hard problem. Carruthers'
book provides a clear overview and analysis of all the important
philosophical theories of phenomenal consciousness on the market, along
with many lucid accounts of relevant, up-to-date findings on
consciousness from neuroscience and psychology. I highly recommend
Phenomenal Consciousness to anyone interested in the current
philosophical and scientific debate over the nature of phenomenal
consciousness.
Phenomenal Consciousness is quite large in scope, covering the full
range of contemporary philosophical and scientific views on
consciousness -- from Thomas Nagel's 1974 paper, "What is it like to be
a bat?", to recent work by Lawrence Weiskrantz on blindsight. I do not
intend, therefore, to give a full account of all that takes place in the
book. Rather, I would like to give a rough overview of (what I take to
be) the book's main line of argument, leaving my few criticisms for the
very end.
2. Overview of Phenomenal Consciousness
In Phenomenal Consciousness, Carruthers puts forward and defends a
dispositional higher-order thought (HOT) theory of phenomenal
consciousness. Roughly, the theory states that "phenomenal consciousness
consists in a certain sort of intentional content ('analog', or
fine-grained), held in a special-purpose short-term memory store in such
a way to be available to higher-order thoughts about the occurrence and
nature of those contents; and that in virtue of such availability (given
the truth of some or other form of 'consumer semantics') all of those
contents are at the same time higher-order ones, acquiring a dimension
of seeming or subjectivity" (p. xiii).
Carruthers takes his theory to provide a "reductive explanation" of
phenomenal consciousness, by which he means that it "describes a way of
linking together cognitive structures and contents (in terms which do
not themselves presuppose phenomenal consciousness), any instantiation
of which is supposed to be metaphysically sufficient for phenomenal
consciousness to occur" (p. 257). There are no possible worlds,
according to Carruthers, in which the cognitive structures and contents
described in his account are appropriately instantiated in a creature's
mind and that creature fails to undergo phenomenally conscious states.
Carruthers also takes his dispositional HOT account to describe "the
actual natural constitution of phenomenal consciousness, in this
world" (p. 259). In our world, any creature that fails to appropriately
instantiate these cognitive structure and contents does not, according
to Carruthers, have phenomenally conscious states.
Carruthers has two main objectives in his book: (1) to defend his
account against alleged counterarguments designed to show that the
cognitive structures and contents described in his account are either
metaphysically insufficient for phenomenal consciousness (arguments for
qualia) or are not constitutive of phenomenal consciousness in this
world (arguments for phenomenal consciousness in animals and infants),
and (2) to show that his account is superior to any of the major rival
naturalistic accounts of phenomenal consciousness currently on the
market. Before an overview is given of these objectives, a few words
need to be said about the intended explanandum of Carruthers' account.
2.1. On The Definition of "Phenomenal Consciousness"
Phenomenal consciousness, according to Carruthers, "is the property
which mental states possess when it is like something to have them"
(p. 13). To illustrate, imagine that you are looking at a red tomato and
a green lime under good lighting conditions. There is something that it
is like for you to see the red tomato, and what it is like is different
from what it is like for you to see the green lime. There is a
subjective difference between your seeing the tomato and your seeing the
lime, and this difference is a difference in phenomenal consciousness.
Things are not as clear as this illustration might suggest, however. For
Carruthers makes a rather subtle distinction between two senses of the
phrase 'what it is like', only one of which is to be identified with
phenomenal consciousness. According to Carruthers, there is a conceptual
and a real distinction between "what the world (or that state of the
organism's own body) is like for an organism, and what the organism's
experience of the world (or of its own body) is like for the organism"
(p. 128). The former sense of subjectivity Carruthers calls
"worldly-subjectivity"; the latter sense he calls
"mental-state-subjectivity" or "experiential subjectivity". It is this
latter sense of subjectivity that Carruthers identifies as phenomenal
consciousness (p. 129, n. 7).
Upon first encountering this rather subtle but important distinction
between worldly and experiential subjectivity, the reader may struggle
to make sense of it -- at least, this reader did. And I found that the
distinction was given its clearest expression in Chapter 6, where
Carruthers puts forward an argument -- drawing on cases from everyday
life and scientific research -- for the existence of non-conscious
experiences. The upshot of the chapter -- at least, with respect to its
bearing on the worldly-subjectivity/experiential-subjectivity
distinction -- is that non-conscious experiences have
worldly-subjectivity without having experiential-subjectivity; whereas,
conscious experiences have both types of subjectivity. To illustrate the
conscious/non-conscious distinction and its bearing on the
worldly-subjectivity/experiential subjectivity distinction, let me
appeal to two rather well-known cases which Carruthers uses to show the
existence of non-conscious experiences: the absent-minded driver case
and the blindsight case.
The Absent-Minded Driver Case: An absent-minded driver is paying little
or no attention to what he is seeing on the road, and yet he keeps his
car on the road and successfully navigates it around other vehicles.
Imagining himself as such a driver, Carruthers maintains that:
I surely must have been seeing, or I should have crashed the car.
Indeed, my passenger sitting next to me may correctly report that I
saw the vehicle double-parked at the side of the road, since I
deftly turned the wheel to avoid it. Yet I was not conscious of
seeing it, either at the time or later in memory. My percept of that
vehicle was not a conscious one. (p. 149)
The Blindsight Case: In cases of blindsight, subjects who have areas of
their striate cortex damaged claim to see nothing in certain blind areas
(scotomas) of their visual fields but retain the remarkable ability
(manifested on forced-choice test trials) to visually discriminate
objects and features within their scotomas. Blindsight subjects can,
Carruthers reports, "accurately trace the movement of a light across a
screen in the blind portion of their visual field, entirely by
guess-work, professing that they are aware of nothing; and they can also
discriminate colours without conscious awareness" (p. 155).
In both these cases, it seems, the subjects are (in some sense) visually
aware of items in their environment -- the absent-minded driver is aware
of the double-parked vehicle, and the blindsight subject is aware of the
movement of the light across the screen. The world in both these cases,
according to Carruthers, is "subjectively presented" (via the
stimulation of the subjects' relevant sensory organs) to the subjects as
being a certain way. In this respect, Carruthers maintains, their visual
experiences have worldly- subjectivity (p. 128). However, as Carruthers
points out, these subjects are not consciously aware of these items in
their environment: their respective perceptual experiences occur below
the level of consciousness. Consequently, it is quite intuitive to say,
as Carruthers does, that there is nothing that it is like for these
subjects to undergo these non-conscious perceptual experiences (p. 176).
Their perceptual experiences, then, lack what Carruthers calls
experiential subjectivity or phenomenal consciousness.
With the worldly-subjectivity/experiential-subjectivity distinction
sketched, let us return to your visual experiences of the red tomato and
the green lime. Your visual experiences, we shall assume, are conscious.
You are not suffering from absent-mindedness or blindsightedness, and
you are focusing your attention on the tomato and the lime before you.
Now, your visual experience of the tomato represents the tomato as being
a certain way (namely, as being red and round) and, as a result, has
worldly-subjectivity. But, there is also something that it is like for
you to undergo this visual experience, and, as a result, it also has
experiential-subjectivity (or phenomenal consciousness). The same can be
said, mutatis mutandis, for your visual experience of the lime.
However, the experiential subjectivity of your visual experience of the
red tomato is distinct from that of your visual experience of the green
lime: what it is like for you to see the red tomato is different from
what it is like for you to see the green lime. An important part of
Carruthers' project in Phenomenal Consciousness is to give a
naturalistic account of this type of distinction by appealing only to
cognitive structures and intentional contents of mental states that do
not presuppose the existence of phenomenal consciousness. And, as
indicated above, Carruthers attempts to do this by showing that the
intentional contents of your conscious experiences are different. The
difference in the experiential subjectivity between your consciously
seeing the red tomato and your consciously seeing the green lime resides
entirely in the difference between the contents of your experiences. And
this difference in the intentional contents of your experiences resides,
in part, in the fact that your experiences -- since they are conscious
-- are available to your higher-order thought (HOT) module in your
brain.
2.2. Some Important Details of Carruthers' Account
Some important details of Carruthers' theory need to be added at this
point; and we can do this, in part, by looking at the following diagram
which Carruthers provides.
Figure 1.
Dispositional HOT Theory
According to Carruthers' theory, conscious experiences -- such as your
conscious visual experience of the tomato -- are percepts (or
analog/fine-grained representations) that represent first-order features
(redness, roundness, spatial relations, etc.) of objects in the
environment. And conscious experiences (percepts) are made conscious in
virtue of their being made available to the subject's higher-order
thought (HOT) forming module via a special short-term memory store C.
Non-conscious experiences -- such as those of the absent-minded driver
and blindsighted subject -- are not made available to the subject's HOT
forming module by way of being held in C; rather, they are held in a
different short-term memory store N, from which they are made available
to the subject's motor-output system, either directly or by way of the
subject's action-schemas module.
On Carruthers' account, both conscious and non-conscious percepts may
represent the very same first-order features of external objects. A
blindsighted subject's non-conscious visual experience of a red tomato,
for example, may represent the tomato as being red and round, just as
your conscious visual experience of the tomato represents the tomato as
being red and round. This commonality in the first-order contents of
these experiences explains, according to Carruthers, the fact that what
the world is like for the blindsight subject (whilst she is aware of
the tomato in her blind region) is (or may be) what the world is like
for you (whilst you are aware of the tomato). Furthermore, since your
conscious visual experience of a green lime represents a different
first-order property (namely, greenness) from that represented (we are
supposing) in the blindsighted subject's non-conscious visual experience
of the red tomato, what the world is like for you (whilst you are
aware of the lime) is not what the world is like for the blindsighted
subject (whilst she is aware of the tomato in her blind region).
First-order contents of percepts, on Carruthers' account, are taken to
explain the worldly-subjectivity of experiences.
But, as we saw above, there is a difference in the experiential
subjectivity between conscious and non-conscious experiences, and a
difference in the experiential subjectivity among conscious
experiences of different types (e.g., the difference between consciously
seeing a red tomato and consciously seeing a green lime), which needs to
be explained. Carruthers' explanation of these differences runs roughly
as follows. The fact that a conscious experience is available to a
subject's HOT module -- that is, available to the subject to think about
it -- adds a higher-order dimension to the experience's intentional
content. Conscious experiences, on Carruthers' account, not only
represent first-order properties of objects in the environment (as
non-conscious experiences do), but they also represent higher-order
properties -- properties of the experience itself. Carruthers explains:
Where before these were first-order analog representations of the
environment (and body), following the attachment of a HOT system
these events take on an enriched dual content. Each experience of
the world-body becomes at the same time a representation that just
such an experience is taking place; each experience with the content
red(a), say, is at the same time an event with the content seems
red(a) or experience of red(a).<1> (p. 242)
The difference in the experiential subjectivity between conscious and
non-conscious experiences, according to Carruthers, resides entirely
in the difference in the intentional contents of these types of
experiences. Non-conscious experiences have only first-order contents;
whereas, conscious experiences have first-order and higher-order
contents. And the difference in experiential subjectivity among
conscious experiences of different types resides entirely, according
to Carruthers, in the different higher-order contents of these
experiences:
The difference between the feel of a phenomenally conscious
experience with the analog content red(a), on the one hand, and
the feel of a phenomenally conscious experience with the analog
content green(a), on the other, is that the former also has the
higher-order content seems red(a) whereas the latter also has the
higher-order content seems green(a). So each phenomenally
conscious experience has its distinctive form of subjectivity by
virtue of acquiring a higher-order analog content which precisely
mirrors, and represents as subjective, its first-order content.
(p. 243)
Finally, it should be noted that Carruthers takes the intentional
contents of percepts to be individuated narrowly -- in abstraction from
the objects and properties in a creature's environment -- and amenable
to a naturalistic explanation in terms of the narrow functional roles
that these percepts play within the mind of the creature that possesses
them -- what Carruthers calls "consumer semantics" (pp. 95-112). The
narrow functional role of conscious percepts -- specifically, their
availability to a creature's HOT module via a special short-term memory
store C -- metaphysically determines their unique, narrow, higher-order
intentional contents, which, in turn, metaphysically determines their
possession of phenomenal consciousness. On Carruthers account,
phenomenal consciousness boils down to a property that percepts possess
in virtue of a unique, narrow functional role that they play within the
mind of the creature that possesses them.
2.3. Two Main Tasks in Phenomenal Consciousness
Carruthers sets himself two main tasks in his book, as mentioned above.
The first is to defend his account against alleged counterarguments, and
the second is to show that his account is superior to any other major
rival account of phenomenal consciousness on the market. With regard to
the first task, Carruthers addresses (among others) four traditional
counterarguments against higher-order/functionalists accounts of
phenomenal consciousness: the absent-qualia argument, the
inverted-qualia argument, the inverted-earth argument, and the
argument for phenomenal consciousness in animals and infants.
The first two arguments purport to show that it is metaphysically
possible -- by way of its being conceivable -- that there be a creature
A who is functionally isomorphic to a creature B, who realizes the
functional architecture in Carruthers' account (as illustrated in Figure
1 above), and yet A's percepts either lack phenomenal consciousness
altogether (the absent-qualia argument) or the phenomenal character of
A's percepts are inverted relative to B's (the inverted-qualia
argument). The third counterargument (the inverted-earth argument)
purports to show that it is metaphysically possible -- again, by way of
its being conceivable -- that there be a creature A who is functionally
heteromorphic to a creature B, who realizes the functional architecture
in Carruthers' account, and yet A's percepts have the same phenomenal
character relative to B's. Were either of these three alleged
metaphysical possibilities genuine, they would show that the functional
architecture described in Carruthers' account fails to metaphysically
fix the nature and existence of phenomenal consciousness.
Carruthers responds to the absent- and inverted-qualia arguments by
showing that at most these arguments prove the conceptual possibility
of such cases of absent and inverted qualia; but they do not prove the
metaphysical possibility of such cases, which is what is needed to
disprove his account. In his response to these two arguments, Carruthers
relies upon a sharp distinction between the property of phenomenal
consciousness (the explanandum of his account) and the concept of
phenomenal consciousness (which he takes to be a recognitional concept).
At most, the conceivability of such cases described in these two
counterarguments, Carruthers maintains, shows that our recognitional
concept of phenomenal consciousness cannot be functionally defined; they
do not show that the property that this concept picks out cannot be a
functional property.<2>
With regard to the inverted-earth argument, Carruthers maintains that
such cases -- cases in which an earthling is given inverted-color lenses
and brought to a place where the colors of things are inverted relative
to their counterparts on earth -- merely shows that the wide
causal/functional roles of percepts do not metaphysically fix their
phenomenal character; they do not show that the narrow
causal/functional roles do not. And it is the latter metaphysical
possibility that must be shown, Carruthers maintains, in order to
disprove his account.
Finally, the fourth counterargument, to which Carruthers considers and
responds, purports to show that the higher-order architecture described
in his account is not constitutive of phenomenal consciousness in our
world. According to this counterargument, it is intuitively plausible
that many animals and young children possess phenomenally conscious
experiences. That is, it is intuitively plausible that there is
something that it is like for (say) a cat or a two-year-old to see a
scurrying mouse or a red ball, and that what it is like for each is
different from what it is like for each to see a stationary mouse or a
green block, or to hear a trumpet blast, or to feel an itch, and so on.
However, it is also plausible -- and there is some rather suggestive
scientific evidence (see Clements and Perner (1994) and Povinelli
(1996)) that corroborates this -- that these creatures do not have the
ability to form higher-order thoughts about their own perceptual states
and, hence, do not have minds that contain the HOT module which, on
Carruthers' account, is essential for phenomenal consciousness.
Carruthers' responds to this counterargument by attempting to show that
the intuitive belief behind it is groundless and easily explained
away.<3> The intuition that animals and young children possess phenomenally conscious experiences, according to Carruthers, is an illusion, resulting from the way we attribute mental states to these creatures. Using the attribution of an experience to a cat as a case in point, Carruthers explains the mechanism behind this illusion as follows:
When we ascribe an experience to the cat we quite naturally (almost
habitually) try to form a first-person representation of its
content, trying to imagine what that experience might be like 'from
the inside'... But when we do this what we do, of course, is imagine a
conscious experience -- what we do, in effect, is represent one of
our own experiences, which will then bring with it its distinctive
subjective phenomenology. So we are subject to a kind of cognitive
illusion -- an illusion which arises because we cannot consciously
imagine a mental state which is unconscious and lacking any
phenomenology. When we imagine the mental states of non-human
animals we are necessarily led to imagine states which are
phenomenological; and this leads us to assert (falsely, if the
arguments given here are correct) that if non-human animals have any
mental states at all (as they surely do), then their mental states
must be phenomenally conscious ones. (p. 198)
Now, there are theories of phenomenal consciousness on the market --
specifically, first-order (FO) theories -- that are consistent with and
offer an explanation of the intuition that animals and young children
have phenomenally conscious percepts. Carruthers, however, attempts to
show that such theories suffer serious problems which his own
dispositional HOT account does not. And this brings us to the second
main task in Carruthers' book: showing that his account is superior to
any other major rival account of phenomenal consciousness on the market.
Carruthers describes and objects to no less than six different
contemporary accounts of phenomenal consciousness: three different kinds
of FO accounts (which he attributes to Fred Dretske, Michael Tye, and
Robert Kirk, respectively) and three different kinds of higher-order
(HO) accounts (which he attributes to David Armstrong, David Rosenthal,
and Dan Dennett, respectively). Much is to be learned from Carruthers'
penetrating analysis and criticisms of these rival accounts, but for the
sake of brevity, I shall focus on his analysis and criticism of FO
accounts.
Roughly, FO accounts attempt to explain the nature of phenomenal
consciousness in terms of an experience's first-order intentional
content and first-order functional role. The main goal of FO accounts,
according to Carruthers, is to
Characterise all of the phenomenal -- 'felt' -- properties of experience
in terms of the representational contents of experience (widely
individuated). On this view, the difference between an experience of red
and an experience of green will be explained as a difference in the
properties represented -- reflective properties of surfaces, say -- in
each. (p. 114)
On FO accounts, the phenomenal difference (for instance) between your
conscious visual experience of a red tomato and your conscious visual
experience of a green lime resides entirely in the different first-order
contents of these experiences. Your visual experience of the tomato
represents an external object as having the properties of being round
and red; whereas, your visual experience of the green lime represents an
external object as being round and green. Although FO theorists disagree
over what fixes an experience's first-order representational content,
they agree that there is no need to attribute (as Carruthers does)
higher-order representational contents to experiences -- contents about
the experiences themselves -- in order to account for the phenomenal
difference among different experiences.
FO theorists acknowledge, however, that there are some mental states
which possess intentional contents, such as beliefs and thoughts, but do
not possess phenomenal properties. According to FO theorists, the
difference between these mental states and sensory states, which do
possess phenomenal properties, resides in the distinct first-order
functional role that sensory states play in the mind. According to FO
theorists, sensory states, unlike beliefs and thoughts, are (to use
Carruthers' words) "the output[s] of the various (peripheral)
perceptual systems (sight, hearing, touch, etc.), which [are] presented
as input[s] to the various (central) cognitive systems charged with
fixing [first-order] beliefs, generating plans, and controlling
movements" (p. 115). Carruthers captures the essentials of the
functional architecture of the FO account with the following diagram:
Figure 2.
First-Order Representationalism
The sensory states that are contained in a subject's E box, according to
FO theorists, are phenomenally conscious states in virtue of their
availability to the subject's first-order belief-forming and
practical-reasoning modules. And the different sensory states that
reside in a subject's E box differ in phenomenal character, according to
FO theorists, in virtue of their distinct first-order intentional
contents.
The chief problem for FO accounts, according to Carruthers, is their
inability to give an adequate account of the conscious/non-conscious
distinction for experiences. Recall that, for Carruthers, phenomenal
consciousness is possessed by, and only by, conscious experiences. FO
theorists, Carruthers argues, are faced with an intractable trilemma
(pp. 168-179). They can either (i) accept that all and only conscious
experiences are phenomenally conscious but deny that there is any
genuine conscious/non-conscious distinction for experiences to be
explained; (ii) accept that all and only conscious experiences are
phenomenally conscious, accept that there is a genuine
conscious/non-conscious distinction for experiences to be explained, and
attempt to explain this distinction in terms of a first-order functional
architecture; or (iii) deny that all and only conscious experiences are
phenomenally conscious, accept that there is a genuine
conscious/non-conscious distinction for experiences to be explained, and
attempt to explain this distinction in terms of a higher-order
functional architecture.<4>
Neither of these three options, Carruthers argues, is plausible. Option
(i) runs afoul of certain facts --- gleaned from everyday cases and
scientific studies -- which strongly suggests the existence of
non-conscious experiences (pp. 147-168); option (ii) fails to explain
why there should be anything that it is like for a subject to possess a
conscious (as opposed to a non-conscious) experience (pp. 170-174); and
option (iii) is committed to the highly counterintuitive view that
non-conscious experiences are phenomenally conscious (pp. 175-179). The
only plausible way out of this trilemma, Carruthers suggests, is to
abandon FO accounts entirely and accept a HO account of phenomenal
consciousness, such as his own dispositional HOT account.
3. Problems for Phenomenal Consciousness
Notwithstanding its many virtues, the main chain of reasoning in
Phenomenal Consciousness, I believe, has a few weak links. In this
section, I wish to point out three of these links.
3.1. Taking the Trilemma by a Horn
Insofar as I understand Carruthers' trilemma against FO accounts, I do
not see how the FO theorist could escape from it by passing through its
horns. Rather, it seems to me that to escape from the trilemma, the FO
theorist must somehow blunt one of its horns; and of the three horns
presented, option (ii) seems the most intuitively plausible. So, it
seems to me that the FO theorist should, if he can, take up and defend
option (ii). But, contrary to what Carruthers maintains, I believe that
the FO theorist can do this without impaling himself on option (ii)'s
point.
Carruthers maintains that any FO theorist who takes up option (ii) fails
to explain why conscious and only conscious experiences are phenomenally
conscious. It is not entirely clear to me, however, why Carruthers
thinks that this is so. But, on one plausible reading, he appears to
reason as follows. Any FO theorist who takes up and defends option (ii)
must give an account of the conscious/non-conscious distinction for
experiences which does not appeal to higher-order mental states, such as
the state of being aware of one's own experiences. On such an account,
then, conscious experiences need not be experiences that their
possessors are (or are able to be) aware of. But, this, then, seems to
make it mysterious why conscious experiences should be any more
phenomenally conscious than the non-conscious experiences of blindsight
subjects and absent-minded drivers.<5> For blindsight subjects and absent-minded drivers, recall, are not (and are unable to be) aware of their respective experiences, and it is precisely this fact about these individuals which makes it intuitively plausible to say that their respective experiences are not phenomenally conscious (p. 176). Phenomenally conscious experiences -- experiences that there is something-it-is-like to have -- are, intuitively, experiences that their possessors are (or are able to be) aware of (p. 14). Therefore, without an appeal to higher-order mental states -- specifically, the state of being aware of one's own experiences -- the FO theorist who takes up option (ii) seems to be unable to explain why conscious experiences, and only conscious experiences, should be phenomenally conscious.
If this is Carruthers' reasoning on pages 170-174, then I find it
compelling only if there is but one reading of the expression 'S is
aware of his experience E' which makes the claim 'phenomenally conscious
experiences are experiences that their possessors are (or are able to
be) aware of' true, and that that reading is a higher-order reading. On
a higher-order reading, to say that S is aware of E is (roughly) to say
that S is aware of E as such -- that is, that S is aware that he has
an experience E. But, is this the only plausible reading of 'S is
aware of E' which makes the claim 'phenomenally conscious experiences
are experiences that their possessors are (or are able to be) aware of'
true? I'm not sure that it is. In fact, there appears to be a
non-higher-order reading of this expression which is just as likely to
make this claim true as the higher-order reading. On this
non-higher-order reading, to say that S is aware of E is simply to say
that S is aware of what E represents.<6> And to be aware of what
something, x, represents is not the same as being higher-order aware
that x represents such-and-such.
To illustrate this difference, consider an analogous distinction in
cases of trompe l'oeil paintings. When one is fooled by a trompe
l'oeil painting -- say, a painting of an outdoor scene -- one is aware
of the painting but not as such. One is not aware that there is a
painting of an outdoor scene present. For if one were aware of the
painting in this way, one would not be fooled by it. Rather, when one is
fooled by the trompe l'oeil painting, one is aware of the painting
simply in the sense of being aware of what it represents. In such
cases, one is aware of what is represented in the painting -- an
outdoor scene, for example -- without being aware that a painting
represents it.
Analogously, being aware of what one is experiencing (when one is
experiencing an F) is not the same as being aware that one is
experiencing an F (when one is experiencing an F). The latter is a
higher-order mental state -- one that represents a lower-order mental
state (one's experience of an F) as such -- but the former is not.
Nevertheless, in being in the former mental state -- in being aware of
what one is experiencing (when one is experiencing an F) -- one is,
in one sense, aware of one's experience. One is aware of one's
experience in just the same way as one is aware of the trompe l'oeil
painting when one is aware of what it represents. Normal sighted
subjects are aware of many of their conscious visual experiences in this
non-higher-order sense of 'aware of'. When I consciously see a red
tomato, for instance, I am aware of what (N.B.: not that) I am
seeing, as evidenced by the fact that I can sincerely say, describe, or
remember, what (N.B.: not that) I am (was) seeing.<7> And in the non-higher-order sense of 'aware of', I am thereby aware of seeing a red tomato.
However, not all subjects who have visual experiences of items or
features are (or are able to be) aware of what they are experiencing
when they are experiencing these items or features. Blindsight subjects,
recall, are able to experience some features within their blind regions,
as evidence by their performance on forced-choice test trials. But,
these subjects are not aware of what (N.B.: not that) their visual
experiences represent, as evidenced by their inability to sincerely say,
describe, or remember what (N.B.: not that) they are (were)
experiencing.<8> These subjects, in a word, are unaware of what they
are in fact visually aware of. In the non-higher-order sense of 'aware
of,' then, blindsight subjects are not aware of seeing those features within their blind regions. A similar description can be given, mutatis mutandis, for the non-conscious visual experiences of absent-minded drivers.<9>
I see no reason why the FO theorist cannot use this non-higher-order
reading of 'S is aware of E' as a way of explaining why, on his FO
account, all and only conscious experiences are phenomenally conscious.
The non-conscious visual experiences of blindsight subjects and
absent-minded drivers are not phenomenally conscious, according to the
FO theorist, because the possessors of these experiences are wholly
unaware of what these experiences represent; whereas, the conscious
visual experiences of normal subjects are phenomenally conscious,
according to the FO theorist, because the possessors of these experience
are (or are able to be) aware of what these experiences represent. Put
bluntly: there is something that it is like to have an experience E if
one is (or is able to be) aware of what E represents, but there is
nothing it is like to have an experience E if one is not (or is unable
to be) aware of what E represents.
The mystery for FO accounts of why conscious experiences should be any
more phenomenally conscious than the non-conscious experiences of
blindsight subjects and absent-minded drivers seems to disappear once
this non-higher-order reading of 'S is aware of E' is made explicit and
pressed into service. It is true, of course, that on such an account,
conscious experiences need not be experiences that their possessors are
(or are able to be) aware of, in the higher-order sense of 'aware of'.
But, this leaves it mysterious why conscious experiences should be any
more phenomenally conscious than the non-conscious experiences of
blindsight subjects and absent-minded drivers only if one assumes that
a subject must be (or must be capable of being) higher-order aware of E
in order for E to be phenomenally conscious. But, such an assumption
merely begs the question against a FO account of phenomenal
consciousness. And since there appears to be a legitimate
non-higher-order reading of 'S is aware of E' which seems to explain why
all and only conscious experiences are phenomenally consciousness,
Carruthers must show, and not merely assume, that on such a reading, FO
accounts still fail to explain why all and only conscious experiences
are phenomenally conscious. On the face of it, the FO theorist seems
equipped to blunt the second horn of Carruthers' trilemma.<10>
3.2. Explaining Away an Intuitive Belief
I'm not sure about other people, but I believe that many animals have
conscious perceptual experiences and beliefs about items and facts in
their environments. My reason for holding this belief is not that I have
succeeded in imagining what it is like for animals to perceive or have
beliefs about items and facts in their environments - for I have not.
Rather, my reason is simply that a lot of animal behavior can be
predicted and explained quite well in terms of the concepts and
generalizations of our folk psychology, and these concepts and
generalizations are about conscious mental states. Our folk theory is
a theory of how conscious mental states interact and produce behavior.
It is not a theory about how unconscious mental states interact and
produce behavior. Unconscious mental states fall under their own laws,
and it is the project of scientific psychology to discover them.
Therefore, since our folk theory does a fairly good job in explaining
and predicting quite a lot of animal behavior, there is some grounds to
believe that animals have the kinds of mental states that our folk
theory's concepts and generalizations are about - that is, conscious
mental states.
Carruthers claims that such a belief is, in fact, groundless, since
there is no real need to attribute conscious mental states to animals in
order to explain their behavior. He writes:
Everything that the cat does can be explained perfectly well by
attributing beliefs, desires, and perceptions to it. There is no
explanatory necessity to attribute conscious beliefs, desires, or
perceptions. All we really have reason to suppose, in fact, is that
the cat perceives the smell of the cheese. We have no independent
grounds for thinking that its percepts will be phenomenally
conscious ones. Certainly such grounds are not provided by the need
to explain the cat's behaviour. For this purpose the concept of
perception, simpliciter, will do perfectly well. (p. 199)
But, to what theory does Carruthers think that these concepts of belief,
desire, and perception belong? If they belong to our ordinary folk
theory, then they are concepts of conscious beliefs, desires, and
perceptions. On the other hand, if they belong to some other theory -
say, a scientific theory about unconscious beliefs, desires, and
perceptions - then clearly the burden is on Carruthers to demonstrate
that everything that animals do can be explained perfectly well with
this theory. But as far as I know, no such demonstration has been made.
And this is so, in part, because there simply isn't at present a fully
developed scientific theory of unconscious beliefs, desires, and
perceptions with which to run such a demonstration. But, until someone
produces such a scientific theory and demonstrates that it explains
everything that animals do, Carruthers, it seems, has given little
reason to think that my belief in animal conscious mentality is
groundless.
I am also not convinced that Carruthers succeeds in explaining away my
belief that many animals have conscious mental states. According to
Carruthers, recall, I hold this belief because whenever I attribute such
mental states to animals, I undergo -- "almost habitually" -- the
following sort of process of imagination:
- I imagine "from the inside" what it is like for the animal to possess the mental state which I attribute to it;
- in doing (i), I inevitably imagine what it is for me to possess this mental state; and
- in doing (ii), I inevitably imagine a conscious mental state.
As a result of doing (iii), I am naturally led, Carruthers maintains, to
"assert that if non-human animals have mental states at all (as they
surely do), then their mental states must be phenomenally conscious" (p.
198).
What I find puzzling about Carruthers' explanation here is that it fails
to describe what typically goes on in my mind when I attribute mental
states to animals. Typically, I do not find myself trying to imagine
"from the inside" what it is like for an animal to have the mental
states that I attribute to it. When I say that the cat sees the mouse,
or that the dog smells the buried truffle, I do not usually find myself
trying to imagine what it is like for these animals to have their
respective perceptual experiences. And in many cases, I know full well
that I could not possibly imagine what it is like for an animal to have
the mental states that I attribute to it. I am quite sure, for instance,
that I could not imagine what it is like for a dog to smell a buried
truffle, and so I do not even try to imagine this. Nevertheless, I
certainly believe that dogs do smell buried truffles and that their
olfactory experiences in such cases are conscious. Carruthers'
explanation, then, would appear to be unable to account for why I should
find it so natural to attribute conscious mental states to animals when
I do not (and, in some cases, can not) imagine "from the inside" what it
is like for the animal to have the mental states that I attribute to it.
Carruthers offers the following reply to this puzzling feature of his
explanation:
[I]t is plausible that what happens in such cases is that we
generalize from those experiences which we can imagine. Since any
experience which we can imagine must have the properties of feel and
what-it-is-likeness, we naturally assume that any experience
whatsoever (even those which we cannot imagine) must be similarly
phenomenally conscious. (p. 199)
But, if I do engage in such acts of generalization, I do not seem to be
aware of them. I do not recall consciously making such a generalization.
And it is certainly not the sort of explanation that I would give for
why I believe that many animals have conscious mental states.
Furthermore, it should be noted that I am not prone to such acts of
generalizations in other domains. For instance, I am not prone to think
that all closed-sided figures have fewer than eight sides simply because
all closed-sided figures that I can imagine have fewer than eight sides,
or that all music of Beethoven's is but four notes long simply because
all music of Beethoven's that I can imagine is but four notes long, and
so on. So, why think it plausible to suppose that I (and others) engage
in this sort of generalization when it comes to the mental states of
animals? It seems to me that a better explanation for why I (and perhaps
others) believe that many animals have conscious mental states is that a
lot of animal behavior can be predicted and explained quite well in
terms of the concepts and generalizations of our folk psychology, and
these concepts and generalizations are about conscious mental states.
At least, this is the sort of explanation that I would offer for my
belief.
3.3. The Transparency Theses
Carruthers' dispositional HOT theory of phenomenal consciousness seems
to run afoul of a particular version of the transparency thesis. One
version of this thesis (endorsed by Harman (1990), Tye (1995), and
Dretske (1995)) states that there are no introspectible,
non-representational features of sensory experiences: when we introspect
a sensory experience, we are not aware of any feature of our experience
over and above those features which our experience represents. Call this
the representational transparency thesis. There is, however, another,
stronger version of the transparency thesis, which we can call the
first-order transparency thesis. According to this thesis, when we
introspect a sensory experience, we are not aware of any feature of our
experience over and above those first-order features which our
experience represents -- where first-order features are those features
possessed by external, non-mental objects.<11> So, for instance, when I see a red tomato and am introspectively aware of seeing a red tomato, I am not aware of any feature of my visual experience over and above the first-order properties it represents the tomato as having -- such as, the particular color, shape, location, texture, etc. of the tomato. I am not, in other words, introspectively aware of having a visual experience as of a visual experience of a red tomato; I am simply introspectively
aware of having a visual experience as of a red tomato.
Carruthers' dispositional HOT account appears to be perfectly consistent
with the representational transparency thesis. After all, part of his
project in Phenomenal Consciousness is to explain away those putative
introspectible, non-representational features of experiences (p. 94).
But, Carruthers' account appears to be inconsistent with the first-order
transparency thesis. For on his account, every conscious experience is
not only a first-order representation of features of external objects
but a "higher-order experience" which represents "that just such an
experience is taking place; each experience with the content red(a),
say, is at the same time an event with the content seems red(a) or
experience of red(a)" (pp. 242-3). On Carruthers' account, then, it
would appear that when I am introspectively aware of seeing a red
tomato, I am not only aware of having a visual experience as of a red
tomato, I am also aware of having a visual experience as of a visual
experience as of a red tomato.
I say that Carruthers' account appears to be inconsistent with the
first-order transparency thesis because he is not entirely clear on the
nature of the analog mode of presentation seems red(a) contained in
the dual content of conscious experiences. On most occasions, Carruthers
takes this analog mode of presentation to be equivalent to the mode of
presentation experience of red(a), which is a higher-order mode of
presentation -- presenting to the subject a property of the subject's
experience (namely, the property of being of red(a)). On this
reading of seems red(a), Carruthers' account is clearly inconsistent
with the first-order transparency thesis. But, sometimes Carruthers
appears to take seems red(a) as presenting to the subject a
first-order property of the distal object of the experience. In this
way, seems red(a) is on par with the first-order analog mode of
presentation red(a), in that both present to the subject a property of
the distal object of the experience. Carruthers appears to treat seems
red(a) in this first-order way when he attempts to explain "the
so-called 'transparency' of perceptual experience" in terms of the
machinery of his HOT account:
This is now easily and satisfyingly explained. The reason why you do
not discover any additional properties of your experience when you
concentrate your attention on it -- in addition, that is, to the
properties of the world (or body) represented -- is that there are
no such properties. All that happens when you focus your attention
on your experience of the ripe red tomato is that you attend to a
state with an analog content representing redness(a), which also
represents seeming redness(a). And either way, to focus your
attention on this state is to focus on the redness represented. (p.
248-9)
But, if all I am introspectively aware of, when I am introspectively
aware of seeing a red tomato (for example), are the properties of the
world (in this case, the properties of the tomato) represented in my
visual experience, then seems red(a) and red(a) must present to me
properties of the world (the tomato), not properties of my visual
experience of the world (the tomato). Hence, in the above paragraph,
seems red(a) and red(a) appear to be taken as first-order modes of
presentation of the world (the tomato).
Taking seems red(a) in this first-order way certainly allows
Carruthers' account to be consistent with the first-order transparency
thesis, but it does so only to saddle it with other problems. First, if
seems red(a) and red(a) both present the tomato as having a certain
property, the question arises whether these properties are the same or
different. If they are the same, then one would like to know how seems
red(a) and red(a) are different? Carruthers, I would think, would
want to individuate these different modes of presentation in terms of
their different functional roles. But, the problem here is to say how
their functional roles would differ on his HOT account. Since seems
red(a) is not -- we are supposing here -- a higher-order mode of
presentation, it would seem unlikely that it should be functionally
individuated in terms of its availability to a creature's HOT module. On
the other hand, if these modes of presentation pick out different
properties, then one would like to know what these two different
properties are and how they are related. We have some rough idea of what
the property of the tomato is which is picked out by red(a), but what
is the property of the tomato which is picked out by seems red(a)? And
how are these two properties related? Does one supervene on the other,
or are they merely correlated?
Finally, and perhaps more importantly, if seems red(a) is a
first-order mode of presentation, then why think that one needs to
appeal to a HOT module in a creature's mind in order to account for its
first-order representational content? And if one need not appeal to a
HOT module in a creature's mind in order to account for the content of
seems red(a), and if the content of this mode of presentation is (in
part) what constitutes the phenomenal character of a conscious
experience (as Carruthers maintains), then why think that one needs to
appeal to a HOT module in a creature's mind in order to explain the
nature of phenomenal consciousness? It makes some sense to appeal to a
HOT module if seems red(a) is taken to be equivalent to the
higher-order mode of presentation experience of red(a). But, if
seems red(a) is taken to be a first-order mode of presentation, there
would appear to be little motivation for accepting a HOT theory of
phenomenal consciousness.
On the face of it, Carruthers' HOT account seems to face a dilemma. If
seems red(a) is a higher-order mode of presentation, then his account
runs afoul of the first-order transparency thesis; but if seems red(a)
is a first-order mode of presentation, then there appears to be little
motivation to accept a HOT account of phenomenal consciousness.
4. Conclusion
I suspect that Carruthers has replies to my three criticisms in section
3. And I hope that he does. For his project in Phenomenal
Consciousness -- to solve the hard problem of the mind-body problem
-- is truly a noble one, and one that I would like to see accomplished
along functionalist lines. What Phenomenal Consciousness unarguably
accomplishes, however, is a clear presentation of the strongest and most
sophisticated attempt to date to solve this hard problem by means of a
HOT theory. For this reason alone, Phenomenal Consciousness, is
compulsory reading for anyone interested in the contemporary
philosophical and scientific debate over the nature of phenomenal
consciousness.
References
Armstrong, D. (1997). What is consciousness? In N. Block, O. Flanagan,
and G. Guzeldere (Eds.), The nature of consciousness (pp.
721-728). Cambridge, MA: MIT Press.
Carruthers, P. (1989). Brute experience. Journal of philosophy, 86,
258-269.
Carruthers, P. (2000). Phenomenal consciousness. Cambridge: Cambridge
University Press.
Clements, W. and Perner, J. (1994). Implicit understanding of belief.
Cognitive development, 9, 377-394.
Dretske, F. (1995). Naturalizing the mind. Cambridge, MA. MIT Press.
Harman, G. (1990). The intrinsic qualities of experience. In N. Block,
O. Flanagan, and G. Guzeldere (Eds.), The nature of consciousness
(pp. 663-675). Cambridge, MA: MIT Press.
Lurz, R. (2001). Begging the question: A reply to Lycan. Analysis,
61.4, 313-318.
Lurz, R. (forthcoming). Neither HOT or COLD: An alternative account of
consciousness. Consciousness and Cognition.
Povinelli, D. (1996). Chimpanzee theory of mind? In P. Carruthers and P.
Smith (Eds.), Theories of theories of mind (pp. 293-329).
Cambridge: Cambridge University Press.
Tye, M. (1995). Ten problems of consciousness. Cambridge, MA. MIT
Press.
