Why should our mind-reading abilities be involved in the explanation of phenomenal consciousness?*

Diana I. Pérez

CONICET- Universidad de Buenos Aires

Abstract

In this paper I consider recent discussions within the representationalist theories of phenomenal consciousness, in particular, the discussions between first order representationalism (FOR) and higher order representationalism (HOR). I aim to show that either there is only a terminological dispute between them or, if the discussion is not simply terminological, then HOR is based on a misunderstanding of the phenomena that a theory of phenomenal consciousness should explain. First, I argue that we can defend first order representationalism from Carruthers' attacks and ignore higher order thoughts in our account of phenomenal consciousness. Then I offer a diagnostic of Carruthers' misunderstanding. In the last section I consider further reasons to include mindreading abilities in an explanation of phenomenal consciousness.

KEY WORDS: Representationalism; High-order theories of consciousness; Mind-reading abilities

Resumen

En este trabajo, considero las discusiones recientes dentro de las teorías representacionalistas de la conciencia fenoménica, en particular las discusiones entre el representacionalismo de primer orden (FOR) y el representacionalismo de orden superior (HOR). Mi objetivo es mostrar que, o bien hay sólo una diferencia terminológica entre ellos, o bien si la discusión no es simplemente terminológica, entonces HOR está basada en una mala comprensión del fenómeno que una teoría de la conciencia fenoménica debe explicar. En primer lugar, argumento que se puede defender al FOR de los ataques de Carruthers e ignorar los pensamientos de orden superior en nuestra explicación de la conciencia fenoménica. Luego ofrezco un diagnóstico de la incomprensión de Carruthers. En la última sección, considero razones adicionales para incluir a las habilidades mentalistas en una explicación de la conciencia fenoménica.

PALABRAS CLAVE: Representacionalismo; Teorías de la conciencia de orden superior; Habilidades mentalistas

What are we talking about when we engage in philosophical discussions about phenomenal consciousness? Are we discussing that peculiar feature which distinguishes humans from all other species in the world? Or, are we talking about a primitive kind of consciousness, the kind of consciousness we share with other species (for example, the great apes) and pre-linguistic children? I do not know whether there is a neutral way to answer this question. Probably not. But, I believe that what makes us unique in the animal kingdom are our cognitive and linguistic abilities, and not the subjective character of our experience. I think I am in line with the philosophical tradition: the renewed interest in consciousness in philosophy over the last 30 years began with Nagel's famous question about the subjective character of the bat experience ("What is it like to be a bat?" Nagel 1974). I will base my paper on this conception of phenomenal consciousness, and, from this conception, I will consider recent discussions within the representationalist theories of phenomenal consciousness, in particular, the discussions between first order representationalism (FOR) and higher order representationalism (HOR). I aim to show that either there is only a terminological dispute between them or, if the discussion is not simply terminological, then HOR is based on a misunderstanding of the phenomena that a theory of phenomenal consciousness should explain.
The explicit difference between HOR (in Carruthers' version) and FOR is the need to include a special relationship with our mindreading abilities to give a philosophical account of phenomenal consciousness. In my view, we can defend first order representationalism from Carruthers' attacks and ignore higher order thoughts in our account of phenomenal consciousness. The paper is organized as follows. The first section presents both theories. The second section considers the main arguments against first order representationalism given by Carruthers, and tries to provide a first order reply. The third section offers a diagnostic of Carruthers' misunderstanding. The last section considers further reasons to include mindreading abilities in an explanation of phenomenal consciousness.

I. Varieties of Representationalism

If we consider the current discussions about phenomenal consciousness, we will find abundant literature on the controversy between representationalism or intentionalism and anti-representationalism. Representationalism about phenomenal consciousness holds that the phenomenal character of experience is one and the same as (or, in its weaker form, supervenes on) representational content that meets further conditions. Anti-representationalism denies this claim, holding that there exist (in reality, or just in some not too distant possible worlds) cases with the same phenomenal content but with different representational content (inverted spectrum cases) or cases with the same representational content but with different phenomenal content (inverted Earth cases). In this paper I will focus on the representationalist theories of phenomenal consciousness.

I.1. There are different types of representationalist theories depending on the "further conditions" that a given representational content should meet in order to be identified with phenomenal content. Tye (1995, 2000) and Dretske (1995) propose a theory of phenomenal character according to which the further constraints are: (1) that the representation should be poised for use by the cognitive system, and (2) that the representation has nonconceptual content, that is, that the representational content should be connected with the agent's beliefs and judgments about her experience, but it has to be independent of the possession of the concepts that a third person (or the individual) should use in order to characterize the experience and independent of the concepts which constitute the individual's beliefs about her own experience. Tye (1995) called this theory the "PANIC theory" of phenomenal character and summed up his view:

Phenomenal content [...] is content that is appropriately poised for use by the cognitive system, content that is abstract and nonconceptual. I call this the PANIC theory of phenomenal character: phenomenal character is one and the same as Poised Abstract Nonconceptual Intentional Content (Tye 1995, p. 137).

Tye assumes that "the cognitive system" refers to the normal adult human being's cognitive system, which includes a mindreading system.¹ But in Tye's explanation of phenomenal consciousness, this mindreading system has no special role. Phenomenal content depends upon its being poised for use by our decision-making mechanisms, or by our fixation beliefs mechanism, but phenomenal content does not necessarily have to be the object of higher order thoughts.
Summing up Tye's view, phenomenal character is identified with nonconceptual representations which are: (1) the outputs of sensory modules; (2) the inputs to the central system/conceptual modules; (3) items not directly tied to action (the relation with action is mediated by conceptual representations); (4) items whose existence and content is independent of the possession of a public language, and also independent of the possession of concepts, that is, of stored schemas in memory; (5) intentional, because they can misrepresent things, or even represent something that is nonexistent, and they are fine-grained;² and (6) items whose semantics is externalist (defined by causal covariation under optimal conditions).
Phenomenal content does not occur within the sensory modules. On the contrary, Tye holds that experiences and feelings are identical to the outputs of sensory modules, which are, at the same time, inputs to the cognitive system. (Tye, 1995, p. 137) He sums up his proposal pointing out four differences between beliefs and experiences, all of which are dependent upon Tye's characterization of nonconceptual (or nondoxastic) representation. (1) There is a difference in content: experiences represent determinate features; beliefs represent determinable ones (Tye, 1995, p. 139). (2) The vehicles of representation are different; experiences represent as maps or matrices do; beliefs represent as sentences do. (3) Experiences do not make use of memory-stored representations, and beliefs must draw on such representations. (4) Experiences are poised for use by the cognitive system; beliefs are part of the cognitive system. (Tye 1995, p. 142).
Although Tye's first examples of phenomenal consciousness are perceptual experiences, he explicitly extends his representational theory to many other experiences, including afterimages, bodily sensations, background feelings, emotions and moods. Let us consider Tye's theory of pain. According to Tye, pains are sensory experiences which track (under normal conditions) certain sorts of disturbances in the body, for example bodily damage. Pains are, therefore, representations of different kinds of disturbances in the body.³ The response of the brain to our bodily damage is as mechanical as the response to external stimuli and, as in the case of vision and other perceptual experiences, there are no concepts involved in this process. Pain, according to Tye, is not identical to the awareness of pain: while the first is a nonconceptual representation of a certain disturbance in the body, the second one is a cognitive state, which brings experience under concepts (Tye, 1995, p. 115). The parallelism between external sensory representations and internal ones is complete.

I.2. On the other hand, quite recently, Peter Carruthers developed a new version of higher order representationalism, which gives our mindreading abilities a special role in explaining phenomenal consciousness.
He says:

My own view about phenomenal consciousness is a form of representationalist, or intentionalist, one. I think that phenomenal consciousness consists in a certain sort of intentional content ("analog" or fine-grained) that is held in special-purpose functionally individuated memory store in such a way as to be available to a faculty of higher order thought (HOT) (Carruthers 2005, p.8).

According to Carruthers, the "further condition" a representation should fulfill to be phenomenally conscious is to be disposed to impact the mindreading system, that is, it should have some potential impact on the subject's theory of mind. In other words, it is because we have a mindreading faculty capable of higher order thoughts about our own experiences that they become phenomenally conscious. Carruthers defends a version of consumer-semantics, which gives two different contents to the percept - a conscious one ("analog seeming red") and an unconscious one ("analog red"). He labels his theory "dual content theory", or "dispositional higher order thought theory" of phenomenal consciousness (in opposition to Tye's and Dretske's "first-order representationalism").
The key to Carruthers' proposal is dual content but he does not give sufficient details about it. He characterizes this content as "analog content" (following Dretske 1981), but he does not propose a clear distinction between analog and digital content. He simply says, "Intentional contents are analog when they have a finer grain than any concepts that the subject can possess and recall. Intentional contents are nonconceptual when their existence is independent of any concepts that the subject possesses" (Carruthers 2005, footnote 5, p. 102). The first characterization is weaker than the second, and Carruthers prefers it because it is compatible with the idea that certain analog contents could be partly conceptual. This is a crucial difference between FOR and HOR, because FOR holds that phenomenal content is entirely nonconceptual, while HOR holds that phenomenal experiences are partly conceptual and partly nonconceptual because they have two contents. The first order one is nonconceptual and the second order one should be conceptual because of its second order nature. (Carruthers' examples of these contents are: "analog-experienced-green" or "analog-experienced-smooth", Carruthers 2005, p. 68).
Carruthers also distinguishes his position from other higher order theories: (a) inner-sense theories, which characterize consciousness as a kind of perception of a first order state, and (b) actualists' higher order thought theories according to which, in order to be conscious, a given state should be the object of an actual higher order thought, and not just be disposed to produce a higher order thought, as Carruthers claims. These other forms of representationalism go beyond this paper. Summing up Carruthers' theory, he holds that a phenomenally conscious content is that which is potentially/dispositionally (not actually) the content of cognitive (not perceptual) higher order representations, in particular, they have to be the potential content of a system that can distinguish between the world and our experiences about the world, that is, a theory of mind system, capable of separating appearance from reality.

II. Arguments against FOR

There are many arguments against representationalism, for example, the above inverted-type arguments that I will not consider here. I will focus on the arguments against first order representationalism, not against representationalism in general. Carruthers proposed three main arguments to reject the first order version of representationalism.
The first argument runs as follows. First order representationalism does not distinguish between (a) what the world is like for an organism and (b) what the experience of the world is like for this organism. Tye mixes both but they are (at least conceptually, according to Carruthers) different: (a) is a property of the world, (b) is a property of the organism. Carruthers says:

We [...] need to distinguish between two different sorts of subjectivity - between worldly subjectivity and mental state subjectivity between phenomenal properties of the world and phenomenal properties of one's experience of the world (Carruthers 2005, p. 43).

The "dual content" that Carruthers proposes is designed precisely to distinguish between these two kinds of subjectivities - worldly subjectivity, identified with first order content, and mental state subjectivity, identified with the higher order content. Given that first order representationalism only considers one kind of content, the distinction cannot be made.
In my view, Carruthers demands too much to a theory of phenomenal consciousness. This theory has to explain what our sensations, feelings and emotions are, independently of the way in which we conceptualize them. Let me explain. One of the main arguments for representationalism (an argument that Tye mentions once and again: Tye 1995, p. 30, Tye 2000, Chapter 3) is based on the transparency of experience. According to this idea, when we are attending to how things look to us, we are not aware of any inner object or thing or any qualities of our experience; we are just aware of the external objects and their qualities. Hence, we cannot make the distinction Carruthers holds in this first argument between the experience of the world and the experience of our mental states, while we are attending to our experience. Of course, we can make this distinction when we think about our own experiences, but we cannot make this distinction while we are undergoing the experience. This is precisely why Cartesian arguments like the dreaming argument or the malicious genius argument are so powerful. There is nothing in the experience itself that allows us to distinguish between the qualities of the world experienced and the qualities of the experience of the world: in the experience there is only one kind of quality. All the rest depends upon our cognitive/conceptual system (in particular, our theory of mind system) which allows us to differentiate between appearance and reality.

The second argument is the following. First-order representationalism has problems explaining some non-conscious experiences like the absentminded driver or the sleepwalker. In these cases we have first order representations of the environment that are poised for the guidance of behavior, in fact, they are actually  controlling behavior, although the subject is not conscious of them and reports that he has not seen the streets, the traffic light, other people, and so on. First order representationalism has a dilemma: either these experiences are not phenomenally conscious even if they meet the constraints proposed by FOR, (and so, the definition should be reformulated), or these experiences are phenomenally conscious but inaccessible to the subject. Tye chooses the second horn of the dilemma.
Let us consider Tye's explanation of these kinds of cases. Tye reminds us of the distinction between phenomenal consciousness and awareness: the subject is aware of those things to which he has access, but being aware is one thing and being phenomenally conscious is another thing. Being aware depends upon the possession of concepts, and the application of these concepts to experience. Tye says, "Consciousness of one's own mental states is higher order consciousness and in this case concepts are required" (Tye 1995, p. 115). Tye proposes this basic distinction. One thing is the pain experience, and another thing is pain awareness. The second is a higher order conscious experience for the subject because of the many conceptual elements in it, which are absent in the first.
Carruthers accuses Tye of generating an infinite regress because Tye would presumably need either an infinite set of "phenomenal concepts", one for each hue of color, for example, or he would be damned to lose the fineness of grain of experience. However, Carruthers, in my view, misinterprets Tye: he does not require the application of a special kind of concept, but just ordinary concepts. I am aware of having a stabbing stomachache, but I am not aware of experiencing the specific red₂₉. If awareness is a conceptual state, as Tye holds, it necessarily lacks the fineness of grain that our nonconceptual experience has. Carruthers seems to hold that there are some representations with both analog and conceptual content at the same time. Yet, as I said before, it is necessary to explain how a given content could be, at the same time, conceptual and nonconceptual.
The second step is my own answer to these cases from a first order perspective. In my view, the correct answer to these cases is that we have to distinguish two kinds of behavior: rationally controlled behavior and automatic behavior. Moreover, only the first one is involved in Tye's characterization of "phenomenal consciousness": the representations that produce only automatic motor answers are not poised to the cognitive system of decision-making and belief fixation. In fact, we are usually unaware of the automatic actions we perform and do not remember having acted (as in the sleepwalker case). The kinds of cases considered in the argument against FOR include automatic behaviors like driving or sleepwalking, which are neither rationally controlled by the subject nor usually remembered, and there would seem to be no problem holding that those cases are not phenomenally conscious.

The last argument is more complex. Carruthers offers six desiderata for a successful reductive theory of phenomenal consciousness:
1) Asuccessful reductive theory of consciousness should explain why phenomenally conscious states have a subjective aspect to them.
2) Asuccessful reductive theory of consciousness should explain why there should seem to be such a pervasive explanatory gap between all the physical, functional and intentional facts, on the one hand, and the facts of phenomenal consciousness, on the other. There is a growing consensus that this can be explained by appealing to "phenomenal concepts" or "recognitional concepts", which have no conceptual connections to physical, causal-role or intentionalcontent concepts (Loar 1990, Tye 1999). So, the theory should allow for pure recognitional concepts of experience.
3) A successful reductive theory of consciousness should explain why people believe that the properties of their phenomenal experience are intrinsic, being non-relationally individuated. (This is the qualia intuition.)
4) successful reductive theory of consciousness should explain why their possessors consider phenomenally conscious experience ineffable.
5) should explain why their possessors consider it private.
6) should explain why their possessors consider it infallible, not just privileged known.

Carruthers' strategy is to show that his theory can explain these features, while FOR cannot. Except for (1) which is basically the same argument as the first one I mentioned above, all the other desiderata are concerned with the way in which we conceptualize our experience, that is, the way in which people think or know their experience, not with the experience itself. At this point Carruthers changes the subject. I will not analyze his explanation of these points here. I merely point out that his explanation includes phenomenal concepts, purely recognitional concepts, the kind of concepts that the private language argument wanted to refute (Wittgenstein 1953). These special kinds of concepts play a central role in his explanation. I will return to this topic below. But, I will first show why these desiderata are not adequate for a theory of phenomenal consciousness: this will be the topic of the next section.

III. In defense of FOR or what a theory of phenomenal consciousness should supposedly explain.

Carruthers says:

I propose to argue that phenomenal consciousness will emerge, of metaphysical necessity, in any system where perceptual information is made available to higher order representations in analog form, and where the system is able of recognizing its own perceptual states, as well as the states of the world perceived. For by postulating that this is so, we can explain why phenomenal feelings should be so widely thought to possess the properties of qualia - that is, of being non-relationally defined, intrinsic, private, ineffable, and knowable with complete certainty by the subject (Carruthers, 2005, pp. 45-6).

It is clear that Carruthers holds that a theory of phenomenal consciousness should explain not only this phenomenon, but also why we conceive of this phenomenon in the way we do. This is exactly the point that I criticize. It does not seem adequate to expect that much for this theory. Surely, why we think about phenomenal consciousness as we do depends more on the way we actually think and the concepts that we have, particularly psychological concepts, than on the peculiarities of the phenomenon itself. (Granted, for the sake of argument, we think about it as Carruthers says we do and this way of thinking about phenomenal consciousness is not in itself implausible, as Dennett 1988 argues). In fact, the way in which Carruthers' theory explains the desiderata that first order representationalism cannot explain depends on the fact that we can distinguish between appearance and reality, and that we have purely recognitional concepts, in short, that we have concepts like "seeming red" as opposed to "being red", undoubtedly a difference belonging to our conceptual cognitive system. As I said above, in the very experience of the world there is no place for the appearance/reality distinction. When we philosophically explain phenomenal consciousness, we want to explain the peculiar subjective character that Descartes experienced while sitting in his chair meditating, independently of whether that flow of experience turned out to be veridical or the product of a malicious genius. Therefore, a theory of experience (or phenomenal consciousness) does not have to explain this distinction, which depends on something further, that is, our conceptual capacity to consider our experiences the experiences of a world that exists, independently of us, about which we can make mistakes. True, to make this distinction, we need the categories of the theory of the mind and, surely, only human beings can enjoy a theory like this. The point is whether the explanation of our conception of experience as we conceive of it, given the ordinary psychological concepts that we possess, should form part of the theory of qualia. Let me consider three reasons why it should.
The first reason is to believe that a philosophical theory of qualia need not be concerned with qualia itself, but that it should provide a conceptual elucidation of our concept of quale, and therefore should explain why we think about phenomenal consciousness as we actually do. Nevertheless, this does not seem to be Carruthers' reason because he affirms that he is not giving an analysis of a concept, but a substantive theory about a certain phenomenon: consciousness (Carruthers 2005, p. 47). That is to say, Carruthers is in line with the orthodox view that contemporary philosophers of mind hold. They say that a naturalistic vision of philosophy should provide a good explanation of a certain phenomenon (for example, conscious experience), using data from the different scientific disciplines while considering conceptual questions. But, he does not believe (and I agree) that the philosophical activity could be exhausted by conceptual analysis.
The second reason would be that representationalism aims to be a reductive theory of qualia, and therefore brute identities postulated by first order representationalism would not suffice to offer a genuine reductive explanation (Carruthers 2005, p. 103). In fact, according to the functional model of reduction, the first step for a reductive explanation is the conceptual analysis of the phenomenon to be explained (Chalmers 1996, Jackson 1998, Kim 1998). However, here we find ourselves in the following paradoxical situation: the concepts that Carruthers adds to complete the partial reduction offered by first order representationalism are phenomenal concepts, which cannot be functionalized! Even if there is no consensus about the way in which we have to understand this special kind of concept, these concepts are conceptually irreducible, non-analyzable in terms of other concepts. Carruthers realizes that what is added does not have the characteristics required by the functional model of reduction and he subtly changes the subject. He says that his theory does not offer an analysis of these concepts, but "allows us to explain its existence" (Carruthers 2005, p. 103), something that first order representationalism would not be able to do. For the sake of argument, let us suppose that we need not include these concepts to give a correct account of qualia. It is far from clear why first order representationalism would not be able to explain its existence. In fact, what we need is an explanation of the existence of a peculiar type of concept, particularly special kinds of mental concepts, that is, the concepts by which we think about our mental life (in this case about our experience). The theory of the mind system is that which produces these kinds of concepts. It is not true that a first order representationalist cannot give an account of these concepts. The fact is that he considers that he does not need them to state a theory of qualia. If the first order representationalist had been asked to explain what human beings thought about their own experiences, he would have given an answer concerning the conceptualization of the experience, not the experience itself, and he would probably have included phenomenal concepts in his explanation (see Tye, forthcoming).
There is a third reason, which I will analyze in the final section because it constitutes, in my opinion, the most important argument against the position that I am trying to defend.

IV. Two final objections

The third reason that Carruthers could offer in his defense is that our phenomenal experience is the way it is in part because we think about it in the way we actually do. That is to say, if a person has the exact same perceptual faculties but does not have higher order thoughts about percepts like ours, then he does not have the same type of phenomenal experience that we have.
I have two concerns with this idea. First, if we accept this objection, it seems that it might be just a mere verbal disagreement between the first order representationalist and the higher order one; that is, there is no substantive difference between them that goes beyond the way they decide to use such words as "qualia". In my opinion, under a stringent reading of first order representationalism, phenomenal consciousness is the output of the perceptual system that turns out to be a potential input to a cognitive system like that of the human being, a system with a theory of the mind. Thus, under this reading, first order representationalists would also be able to maintain that without conceptualizing it in the way we actually do, we would not have the experience that we have. (I do not claim that this idea is in Tye's or Dretske's theories. Even more so, as I claimed before, it would seem that by "cognitive system" they understand something less committed to the anthropocentric idea that this system should be something similar to the normal adult human being's cognitive system. This is why their explanation of phenomenal consciousness can be extended to prelinguistic beings like babies, or nonlinguistic beings such as, say, bats). In other words, if the reason to include our mindreading abilities in the explanation of phenomenal consciousness is that there is an essential or constitutive relationship between the way in which we conceptualize our experience and the experience itself, then, we would probably not be able to distinguish between first order representationalism and a higher order one. (Therefore, this is not a good argument for higher order representationalism). For both theories, it is a contingent fact about the human cognitive architecture that our mind reading abilities are involved in phenomenal consciousness but for both it is not conceptually necessary that it should be so.
I think that the answer to the question about the mere verbal disagreement can be yes or no, depending on the way in which we understand the idea of the second order content of experience. If we interpret it as I have tried to describe it, as thoughts whose objects are the outputs of perceptual modules - as genuine higher order thoughts - for example, beliefs or judgments about my own experience, then there is no real disagreement between Carruthers and Tye. They both propose the same architecture although there could be a difference regarding which part of that architecture we should identify as the qualitative content for the subject. But, on reading some other papers we can see that Carruthers has doubts about whether genuine thoughts are involved in phenomenal consciousness, and therefore to consider them as states possessing genuine conceptual content, or whether the higher order states which are involved in phenomenal consciousness are perceptual. Carruthers holds a kind of higher order perception theory, and consequently describes the contents identified with phenomenal experience as nonconceptual content (or not "completely" conceptual content) that is, as an analog content such as "analog seeming red". Under this reading of Carruthers' proposal, there is a real disagreement between Carruthers and Tye, but in my opinion, the burden of the proof is on Carruthers' side, because he owes us an explanation of the nature of this strange analog and, at the same time, conceptual content.
My second argument concerns the most basic question: What are qualia? It is true that in describing qualia we necessarily appeal to the appearance/reality distinction, to the subjective/objective distinction. Children acquire these concepts at around the age of three. Some psychologists even claim it is earlier: the 18-month-old toddler who pretends that the banana is a telephone already distinguishes between appearance and reality. The child knows that the banana is not a telephone but pretends it is one. When we have a conscious experience of red, for example, it is not true that we have a conscious experience that something appears to be red (as the argument of the transparency of the experience showed). Clearly, we are able to reconceptualize the experience of red in this way, but this implies mastering the appearance/reality distinction, besides mastering the concept of red. Since Wittgenstein, we all know that concepts like "seeming red" are parasitic on the concept "red". In fact, during most of our conscious life, our everyday life is such that we are conscious of our experience of the world, not of the presence of mere appearances before us. This is the sick case, the anomalous case, that in which we do not see reality as it is, but only as we conjecture it to be. Thus, it does not seem to be constitutive of our experience of the world that we conceptualize it as being a mere experience of the world, like a far away and independent state of the world, which could be exactly as it is without the existence of the world.
Let us return to the initial question: What is phenomenal consciousness? The standard answer (and I recommend Chalmers 1996 in his paradigmatic book) is that it is "the most familiar thing in the world" (Chalmers 1996, p. 3), the one that accompanies us constantly, not the sick sensation that the world could not have been exactly what we are in fact perceiving. Therefore, a philosophical explanation of qualia does not need to involve the possession of the appearance/reality distinction by the subject or that we may possibly think that our experience of the world is just an illusion. It is the explanation of what is most familiar to us and, therefore, of the case where we do not even consider the distinction between appearance and reality. This is the phenomenon that a theory of qualia should explain, the phenomenon that is most familiar and constant, the one that brings us closer to more primitive kinds of minds, the one that we can enjoy before we are "contaminated" by philosophy.

Notes

* Previous versions of this paper were presented at the Simposio Lenguaje, mente y conocimiento. 40 años de Crítica, Instituto de Investigaciones Filosóficas, UNAM, Mexico, April 9 - 13, 2007 and the XVII Interuniversitary Workshop on Philosophy and CognitiveScience: P. Carruthers, Palma de Mallorca, May 24-26, 2007. I thank P. Carruthers, L. Skidelsky, E. Txapartegi, and the audiences at the two conferences. Financial support has been provided by ANPCyT (PICT 2005/33150) and CONICET (PIP no. 5987).

1 I do not discuss here whether these mindreading abilities depend upon the possession of a module or whether they belong to the central system. I will call it a "mindreading system", a neutral characterization.

2 Tye 1995, 4.3. Tye 2000, 3.3.

3 See Tye 1995, p. 113.

References

1. Carruthers, P. (2005), Consciousness. Essays from higher order perspective. Oxford, OUP.         [ Links ]

2. Chalmers, D. (1996), The Conscious Mind. Oxford, OUP.         [ Links ]

3. Dretske, F. (1995), Naturalizing the Mind. MIT Press.         [ Links ]

4. Nagel, T. (1974), "What Is it Like to Be a Bat?", Philosophical Review, pp. 435-450.         [ Links ]

5. Tye, M. (1995), Ten Problems of Consciousness. Cambridge, Massachusetts, MIT Press.         [ Links ]

6. Tye, M. (2000), Consciousness, color and content. Cambridge, MA, MIT Press.         [ Links ]

7. Tye, M. (forthcoming), "A Theory of Phenomenal Concepts" http://www.utexas.edu/cola/depts/philosophy/faculty/tye/Theory.pdf        [ Links ]

8. Wittgenstein, L. (1953), Philosophical Investigations, Oxford, Basil BlackwellLinks ] Helvetica, sans-serif">.