Logotipo MLAG

Find a word in this Page


Index » Members » Sofia Miguens » Anti-representationalism


D. Dennett’s brand of anti-representationalism:
a key to philosophical issues of cognitive science

«There are two major traditions in modern theorizing about the mind, one that we will call representationalist, and one that we’ll call eliminativist. Representationalists hold that postulating representational states is essential to the theory of cognition (…) Eliminativists, by contrast, think that psychological theories can dispense with such semantic notions as representation. According to eliminativists the appropriate vocabulary for psychological theorizing is neurological or perhaps behavioral» Fodor & Pylyshyn 1988, Conexionism and cognitive architecture, Cognition, 38.

1. Cognitive science, mind and method

Cognitive science research bears on our understanding of fundamental questions about the mind, such as the nature of representation, consciousness and action. Drawing on such research D. Dennett’s work of the last thirty five years offers proposals about how to think about each of these issues. His views are admittedly controversial – he tends to be accused of insensitivity to the ‘real’, first-person, problems of the mind , but in fact he provides a comprehensive, cognitive science grounded, account of the nature of subjectivity.
For Dennett intentionality, and not consciousness, is the most basic issue about the mind , and the notion of interpreter is essential to deal with it (Dennett 1987). His theory of consciousness involves a critique of qualia and the Cartesian Theater, but also positive views, namely a funcionalist model aimed at explaining the status of the centredness, seriality and continuity of mental life and a belief-like, second-order, account of self-awareness (Dennett 1991). It is from views on intentionality and consciousness – which will be explored in this article – that a practical philosophy arises, concerning what must be in place in terms of cognition for the mentalistic concepts of ‘person’ and ‘action’ to apply to cognitive systems (Dennett 1983, Dennett 2003).
Anti-representationalism is the core of Dennett’s view of the mind. Anti-representationalism is the idea according to which for there to be mind there need be no such thing as real inner representations in those physical systems we take to be rational agents. For Dennett this means that there’s no such thing as representations short of a stance, a strategy towards certain physical systems. Many people consider these formulation already unintelligible, or question begging . Critics in general took it as interpretivism (i.e. mind would depend on a merely instrumental attribution of mentality by an interpreter). In fact, in order to fully understand Dennett’s brand of anti-representationalism we must see it as a more sophisticated view, including (i) the idea that mentalistic notions aplly to agents globally considered and not to inner states, (ii) the fact that they apply depends on recognition of patterns of behaviour and also on agents inner design for recognition (which means questions of representation and design should come together in the theory of mind ), (iii) a position about the metaphysics of cognition, which involves taking seriously the difference between vehicles of representation (e.g. neural events) and what is represented. The whole view may arguably push Dennett into denying metaphysical realism (Haugeland 1997: 267-304) but it is certainly not instrumentalist.
Before addressing Dennett’s general positions I want to point out something that may help a reader navigate through his work. Given his dislike of tempting subjects such as qualia and zombies, and also his dislike of overlinguistified and overargued approaches to questions of mind, Dennett sometimes seems to be very far from protocols of argument in philosophy, not so much a philosophers’ philosopher as a philosopher for those outside of the discipline who are interested in philosophical issues. He casts a suspicious look on some ongoing debates in the philosophy of mind and often goes straight to cognitive science work . This is on the one hand a symptom of what Dennett calls his technophile inclinations - in his own words, if not for being born into a arts and humanities academic family he would probably have become an engineer (Self-Portrait )-, but we may also see it as a quinean, naturalized epistemology, imperative put to practice in terms of method, disrupting supposed boundaries between what should be scientific and what should be philosophic in the theory of mind. In consequence, Dennett’s work often takes the form of an effort to relate cognitive science with phenomenological descriptions of mind (taking phenomenology to mean a neutral description of mental goings-on, at the personal level, the level we access our own minds ). Bringing together this phenomenological take a cognitive system may have of itself with a third-person approach involves specific problems: cognitive science is involved in what Dennett calls (Dennett 1991) the dismantling of the witness protection program or a close-up of the observer. There’s no witness, no observer inside a cognitive system which we take to be mental (ourselves included, of course) and still there’s mind.
Yet, although taking cognitive science seriously involves trying to find out whether self-knowledge can be disrupted from a third-person point of view, this may in fact go together with the impossibility of being an eliminativist. And that is exactly the case with Dennett. Anyway if we consider that self-knowledge stops - it goes only as far as the mind goes, or rather, not even that far, and definitely does not reach cognitive architecture or physical workings - Dennett’s decision not to separate philosophical theory of mind from third-person approaches to cognition doesn’t seem such a bad one.

2. A few remarks about the origins of Intentional Systems Theory

Since Dennett’s anti-representationalism results from what was going on in philosophy in the 50s and 60s, I will take a step back to reformulate the question about Dennett’s anti-representationalism as a question about what makes for an antirepresentationalist such as Dennett. In epistemological and ontological questions, Dennett’s work was inspired by ideas of Ryle, Wittgenstein, Quine and Putnam. Ryle showed him that the alternative between materialist and idealist monism depended on an incorrect approach to the problem of mind , and Wittgenstein the groundlessness of any presumption of epistemic authority or direct access to the structure of the world in mental descriptions . Wittgenstein and Ryle together provided an example of how description of mind at the personal level should go, and of how talk of mind should be taken, if not as authoritative introspective description of thinking substances. From Quine, Dennett took the naturalist outlook and the regard for natural science, and also (as Fodor put it) nihilism about meaning – the idea that ‘there are no semantic engines’. Putnam’s functionalism, a sophisticated kind of materialism, allowed him to articulate his opposition to psychophysical identity theory and gave him the idea of the mental as incorrigibility of personal level acess. His thinking about action was influenced by Ch. Taylor’s philosophical critique of psychological behaviourism, as an utopian search of a pure data language to explain behaviour, as well as by E. Anscombe’s analysis of intentional action in Intention.
The starting point for the theory of mind is thus ‘phenomenology’, and that’s what Wittgenstein (in the Philosophical Investigations) and Ryle (in The Concept of Mind) do: a report of mental happennings, whether these be feelings of pain or mathematical thoughts, without any presumption of explanation, or inquiry about physical basis of mind. Strangely enough, these philosophers, on whose work Dennett molds his own, are often taken as behaviorists. Yet, behaviourism, Skinner type of behaviourism, was exactly what stood opposed to the methodological revolution in psychology which prompted Dennett‘s philosophy, and which we can loosely call cognitivism, in the sense of informational-computational approach to the mind. Dennett has always opposed what he called ‘peripherist behaviorism’ and defended the indispensable role of intentional and teleological descriptions of behaviour - initially he called this position centralism (Dennett 1969), later Intentional Stance (Dennett 1987) . Psychological and philosophical behaviorism have always been rather different anyway, and it is philosophical behaviorists Dennett takes as models. But there’s one point where Dennett profoundly disagrees with a rylean-wittgensteinian approach to the mind: the ideia that a self-sufficient conceptual analysis could be a proper philosophical method, which easily pairs up with contempt for natural science . What’s important about Wittgenstein and Ryle is the fact that they try to conceive the status of the personal level – the level at which we describe our minds from within, and where pulleys and levers questions, i.e. mechanical considerations, are not legitimate. Such personal/subpersonal distinction is vital, when dealing with mind and cognition, to set apart what is descriptive and what is explanatory. Still, even if personal level approach should consist of phenomenological data-gathering for the theory of mind, this should be brought together with scientific research about cognition. Luckily, to help him step past the lack of interest in the philosophical bearing of science, he had brought from America the influence of Quine, and the imperative, for a proper quinean, not to overlook science . This led Dennett to a hybrid philosophy / cognitive science style of approach, which sets him equally apart both from Ryle and Wittgestein conceptual analysis as from much contemporary philosophy of mind.
Putnam’s functionalism came as a final touch. In his manifests from the 60s , Putnam declared the mind-body problem to be not an empirical problem but a logical, linguistic one, with nothing to do with a supposedly unique character of human subjective experience, since it would arise for any system characterized by an assimetry of access to its own physical and functional (logical) states. So from Putnam, Dennett took not only functionalism as a critique of identity theory but also a characterization of mind as incorrigible self-access to functional states .
The result is Intentional Systems Theory (IST henceforth): a moderate realism about the nature of representation (Dennett 1998, 95), going together with a teleofunctionalist view of content and the refusal to consider psychological explanation as explanation properly speaking. As for consciousness, the theory includes a deflationary view of (so called) qualia, and a second-order, belief-like, account of awareness, based on a functionalist model – the Multiple Drafts Model (Dennett 1991). Dennett’s gradualist position about ‘persons’ and ‘actions’ (Dennett 1983, Dennett 2003) comes of these.

3. Representation: sententialism, eliminativism or interpretation

But what is anti-representationalism, if not straight Churchland-style eliminativism? In IST’s case it results from (i) a quinean option for interpretation in the theory of mind (ultimatelly explored through the idea of real patterns recognized from the Intentional Stance), (ii) darwinism (it is design that accounts for both patterns and recognition, and Darwinian principles of evolution by natural selection should account for the origin and nature of design). This eventually leads to a denial of metaphysical realism .
While a self-professed physicalism , IST’s first concern is in fact the status of mentalistic language in the theory of cognition. It was this option that eventually led to the defense of a real patterns (as opposed to Language of Thought intentional realism) view of the nature of representation. Whenever the behaviour of some physical system is being understood or predicted mentalistically, the intentional stance is being adopted (according to Dennett, there are other stances we may take towards a physical system, namely the physical stance – the physicist’s approach, and the design stance – the biologist, or computer programmer, focus on function) . By taking the intentional stance a cognitive system makes sense of the behaviour of another without any idea of inner workings. How is it possible that minds read minds this way? It is possible because patterns of behaviour, relations between the system and those aspects of environment which matter to it, are being recognized by another cognitive system. Before seeing this interpretivism as untenably behavioristic and anti-naturalist we should consider that we needn’t even be talking about humans here – what Dennett is saying is that we cannot avail ourselves with the notion of representation without considering relations between agents and their environment. What we, in the theory, call representations depends on interpretation in that it depends on recognizing, from the intentional stance, the embeddeness of behaviour of agents in the environment. It is important here that there are basic representations, which are action-oriented and about what is, for specific agents, relevant in the environment . Contrary to what instrumentalist readings assume, Dennett thinks that an interpreter (until now, a mind-recognizer) does not arbitrarily create intentional interpretations of other cognitive systems, rather representation talk is made legitimate by recognition of environment-embedded patterns of behaviour. This should be accounted for as a matter of design-for-representation-using and and representation-recognizing and inevitably it leads from the question of representation to that of the nature of design. Thus, Dennett has to come clear with his view of design: is it real or not? To sum up, he thinks design is real in that it results from an algorithmic, mindless process of evolution by natural selection (Dennett 1995), but depends on interpretation in that functions – and functions of inner design-for-representation included – depend on recognition to be taken as specific functions .
So although representation relations depend on recognition, they do subsist and more must be said about their content. Dennett does this in terms of a telefunctionalist account: content is not within, where vehicles of representation occur, it is a function of function, externally determined. But what is external? If action-oriented representations are the basic type of representations, the world is represented differently by different cognitive systems. For instance, different animal perceptive systems evolved to respond to the enviroment’s affordances, so different cognitive systems will be different world-makers (taking well known examples from the literature, there will be worlds in which location-of-nectar is especially salient, worlds where that’s the case with possible-nutrition-which-flies-quickly-through-visual-field, or toxic-surface-water-to-be-avoided, etc). This Dennett calls notional worlds, a concept that does not apply to vehicles of representation within a cognitive system, nor to the world directly, but rather to what is represented .
What do these views about nature and content of representations amount to when it comes to giving an account of psychological explanation? Basically, they keep Dennett from subscribing to anything Fodor’s hard-core representationalist realism buys him: causal thinking about the intentional level of reality, psychological explanation as nomological, and so proper explanation - the whole constellation of special sciences, ceteris paribus laws, independence of psychological generalizations from physical implementation. But for Dennett, Fodor’s sentences-in-the-head view conflates intentionality with design: being an engineering hypothesis Language of Thought commits the rylean sin of taking a conceptual anwer to be a causal answer. The representational- computational theory of mind takes notions (language, representation) which aplly to the whole cognitive system as if they could do explanatory work at the subpersonal level. For Dennett this amounts to letting a ghostly central observer sneak in in the theory. Plus, Dennett sees no reason to accept the common design (Dennett 1978: 90) of cognitive systems Language of Thought implies – for him subpersonal cognitive psychology is just a theory of implementation of that which, intentionally interpreted, are representations and computations.
The next move is to clarify the relations between folk psychology, (scientific) subpersonal cognitive psychology and IST as a philosophical view of cognition (Dennett 1987, 43). And the way he sees it, there is no reduction of intentional characterizations to physical descriptions – but also no explanatory power of mentalistic notions.
Not only did the relation between representation and design lead Dennett to Darwinism but also to what he calls, in Kinds of Minds (Dennett 1996) an evolutionary point of view on ontology and metaphysics. From an evolutionary perspective there’s a plurality of minds. But can pluralism can be defended together with physicalism, understood as priority given to the physical stance, and a constitutive reference to physics, as basic science, when the question is what fundamentally exists? Again, the question of reduction comes up. There are several ways to be an anti-reductionist about the mind (i.e. to deny that the intentional may be reduced to entities and laws situated below within the framework of a hierarquical conception of nature), the main ones represented in the philosophy of mind by Fodor and Davidson. To understand Dennett’s anti-reductionism we must focus not on the situation where a human instrumentally interprets another as mental, but rather on the status of what is represented, and on the concept of an interpreter.
What is the ‘interpreter’? Answering this question implies clearly stating the metaphysical implications of IST. J.Haugeland does that in an illuminating way, bringing IST together with an idea which is actually one of its main targets: J. Searle’s concept of intrinsic intentionality - the ideia that one can legitimately distinguish real believers from which are merely predictable from the intentional stance . Haugeland suggests that we use searlean intrinsic intentionality and Dennett’s interpreter to clarify one another. Both Dennett and Searle claim that intentionality is normative, and Searle proposes the notion of aspectual shape (of anything mental) to say what is meant by normativity . Now, whereas Searle simply characterizes aspectual shape, and cannot account for it, Dennett, with his idea of interpreter, can. Let’s formulate the problem as ‘how can physical configurations of brains resulting from evolution by natural selection be intrinsically normative’?, and take a game of chess as an example of taking something as something. How does a player see that a horse threatens a pawn? No causal history accounts for the perception of horse-as-horse or pawn-as-pawn: only a stance, a constitutive commitment, from which depends the very reality of chess phenomena, does. These are not identifiable with physical pieces of ivory (or patterns on a screen, or marks on a paper); rather they are constitued by standards . Whose standards? The subject’s – and what Dennett is saying abut representation is that the constitutive role of interpretation stands to mentality as the subject stands to the reality of chess phenomena.
Haugeland sees the intentional stance view of representation as kantian constitution: objects of perception, action or thought, are understandable as what they are only in terms of a commitment of a subjectivity-qua-unity (i.e. the interpreter). It is this interpreter we need to understand in order to have a metaphysics of representation. The interpreter is neither a place in the brain, nor a self, it’s more like a taking-as function. This normativity is a feature of subjectivity which for instance a neurobiological theory of consciousness cannot account for, nor should, because it is not by itself a theory of representation, but of the vehicles of representation. The chess analogy is useful to dispel any suspicion about the mysterious character of normativity – what is there is a stance, without which there’s neither subjectivity nor objecthood: that’s the two-sided nature of mind as representation, necessary for searlean aspectual shape which distinguishes anything mental as mental. Through this we can understand searlean intrinsic intentionality as subjectivity to itself, or self-mental-interpretation, and also see the other side of the problem of representation as what B. Cantwell Smith (Cantwell Smith 1996) calls the ‘origin of objects’. Cantwell Smith points out that physics does not buy us objects - individuated, continuous through time, discrete and identifiable represented objects – but only particularity and locality. Physicalist theories of mind should take into consideration, instead of assuming existence identity or persistence across time of intentional objects. Being a represented object is not a local property of a space-time region: objects (individuation in what is represented) are there only for certains entities. Object is not a causal notion, but an historical concept, and that’s the main reason why representational notions should not be taken as physically effective.
Haugeland and Cantwell Smith help shaping Dennett’s anti-representationalism and also his view of psychological explanation: anti-representationalism is not a view about human agents’ capacity for interpreting one another by means of strategically attributing mental states, but a view about the nature and status of representation in the world according to which it is only at the level of a subject / object interface that ontologies (Dennett’s notional worlds) are defined (even if it may still be claimed that the world is metaphysically one). So, in spite of accusations of behaviourism and instrumentalism, IST in fact involves the irreducibility of subjectivity – not of qualia but of this essential role of an interpreter for representation. This also means there are at least two different unities a theory of mind must account for: the interpreter, the unity involved in intentionality, and another unity, a represented unity in a mind, for itself - the self, which, from a cognitive point of view, Dennett sees as virtual unification in a multiple agents cognitive system. The first is a metaphysical question about cognition; the latter is a question about the individual subject’s subpersonal mechanisms of cognition, to be answered in terms of brain . This is crucial for the next step in the theory of mind: a theory of consciousness, partly built on this view of representation, but also involving the proposal of a cognitive model, and a set of epistemological considerations about qualia. The fact that Dennett defends a representationalist theory of consciousness (i.e. consciousness is a form of representation, principles for understanding intentionality and consciousness are the same) makes it necessary, in order to understand his anti-representationalism, to deal with intentionality and consciousness together.

4. Consciousness as an imprecise Cartesian Theater

Dennett’s theory of consciousness is bound to frustrate fans of phenomenal consciousness, qualia and zombies: it is a deflationary theory, in which conciousness is explained in terms of representation, self-access, and incorrigibility of self-reports (in fact, qualia and zombies don’t even make sense within a representational theory of consciouness). Dennett’s theory of consciousness deals not with the so-called hard problem of consciousness, but with a set of problems (kinds of access, unification of a stream of consciousness from multiple drafts at the subpersonal level, self-representation and virtual center, self-reports, the difference language makes in a mind, etc). The type of solutions Dennett comes up with make it obvious that what he is offering is at least partly a cognitive model. This earned him the accusation of offering no arguments and no answer to the metaphysical problem of phenomenal consciouness. For Dennett, though, this is how a theory of consciousness should be, if it does not go after bogey-problems.
Dennett’s theory of consciousness is developped mainly through the Multiples Drafts Model (put forward in the 90s, in Consciousness Explained, as a successor of the Brainstorms model). Funcionalist models (which are, in spite of the proclaimed opposition to the Cartesian Theater , the starting point for a belief-like theory of inner awareness) are the constructive part of Dennett’s theory of consciouness - through them Dennett aims at filling the (supposed) abyss between physiology and phenomenology. Plus, they allow him to deal with empirical issues of cognitive science, such as the status of mental images and dreams, production of language, blindsight, temporal anomalies (such as color phi-phenomenon and Libet’s problems of backward referral in time and delay of consciousness of intention) (Dennett 1992). The destructive, and better known, part of his theory of consciousness is an argument for quining qualia – decisive for rejecting the usual ways of vindicating phenomenal consciousness: zombies, knowledge argument, what-it’s-like.
How does one go from a cognitive model to the elimination of qualia? Unlike what happens when dealing with intentionality (the agent’s unity is simply assumed to consider behaviour and adaption to the environment), dealing with consciousness involves zooming into cognitive systems, characterized at subpersonal level(s) as (i) hardware-level parallel distributed processing, (ii) functional level competing agents. The assumption of a unity or center as given goes, and the unified represention of the system by the system becomes part of what is to be explained . Dennett’s critique of qualia is simply a matter of exploring the epistemological consequences of this absence of a real center where everything comes together for an observer -the Cartesian Theater metaphor for the mind. Still, even if the close-up of the observer reveals the absence of a real center, our own mind appears to each of us as an (imprecise) Cartesian Theater, and that is what the Multiple Drafts Model (MDM) aims at explaining. This means formulating subpersonal level hypothesis about what makes conscious experience unified, serial and continuous and also about what makes for our sense of control of intention. According to the MDM, unification depends on a user’s ilusion of a virtual serial machine implemented on parallel distributed processing hardware (the brain of the body the self will call its own). Virtual is obviously not unreal – the mental life of the system is very much real for the system itself - and Dennett is not saying consciouness is an illusion (the illusion would be to think of the unified self as a separate substance, an out of the physical world pure experiencing) but trying to account in terms of cognitive architecture for the way we appear to ourselves. Self-representation is another necessary condition for the sense of being owner and author of phenomenology, and for the unification and centralization of control and intention. The need for inner communication pushes for a represented unity (in language using humans, eventually an ‘I’) . This centralized virtual arena for unified control of behavior is necessary for what phenomenologically is feeling a self from within. Still, unified intention and initiation are also user’s illusions in the sense that the ‘control-component’ (as the Brainstorms model has it) is replaceable with autonomous agents which can be ‘dismissed’ (i.e. they are bottom-up explainable). In such conditions, disunity and break-down lurk – the various pathologies of identity which have always interested those interested in mind (from schizophrenia and paranoia to multiple personality disorder and split brain) reflect this absence of a real center.
Not only centredness and seriality but also continuity of phenomenology has to be explained: part of Dennett’s exploration of consciousness concerns the role of neglect and filling in for what may not be there in experience (from memory, to dreams, to perception) and his basic thesis is that if there’s no observer there’s no need for filling in. More specifically, awareness of x cannot exist without a belief and it is by principle impossible for the subject to decide, in what concerns his awreness of x, between appearance and reality (this may be called first-person verificationism) . The blind spot, in vision, is emblematic of the kind of non-problematic neglect (we might call it normal anosognosia) Dennett is after and thinks is significant. In where there are no agents anxious for information, there’s no problem if no information arrives. Dennett generalizes this to our feeeling of our own phenomenology .
Having replaced the observer with specialized agents and identified awareness with availability, in memory, of knowledge-that is essential to understand what is probably the main test-case for MDM and first-person verificationism: the supposed stalinesque and orwellian revisions Dennett tests on colour phi-phenomenon and several other examples (Dennett 1991). It was already apparent in the Brainstorms model that memory was vital for Dennett’s conception of consciousness. Personal level access of the system to itself, which is ultimately what Dennett calls consciousness, depends on different types of subpersonal accesses. Namely expression, i.e. publication of any mental content, even to oneself (e.g. thinking linguistically), involves subpersonal transit between memory and mechanisms for reportability. If content of awareness is everything registered in memory, it may decay or be interfered with before it is recovered for public access. If ‘public’ should be identified with personal and includes the linguistic acess the system has to itself, this determines the epistemic prospects of introspective authority, since what we call semantic intentions are based on the connection between control and reportability. Given the absence of difference between processing and status of contents of which we’re aware, consciousness depends on probes – it is self-monitorization which make consciousness as inner-awareness possible. This is developed through an higher-order view of consciousness (Dennett 1991:303). A system is conscious if and only if higher-order mental states, mental states which are about other mental states of the system, occur. And these mental states are, according to Dennett, beliefs, not perceptions .
It is based on these ideas about how mental life appears to conscious beings such as ourselves that Dennet discusses and quines (i.e. eliminates) qualia. For him, the term names intrinsic, ineffable, private, uncorrigibly known characteristics of our mental lives. With qualia so defined, the claim that there are no qualia, amounts to saying that there could not be, in a subject's mental experience, something that is immediately known and ineffable. There’s no real seem, no pure non-conceptual appearing (Dennett 1991, 369-411) - there simply cannot be any differences in conscious experience that subjects are not explicitly aware of .
What follows from this? Dennett never says cognitive systems such as ourselves (in which there are no qualia) are not conscious - in fact, the very object of a theory of consciousness are those systems, among intentional systems, which are conscious. Dennett theory of consciousness applies exclusively to humans – other animals, or even babies, simply are not subject to the Cartesian Theater illusion: those are minds less centralized that adult, linguistic, human minds, and not self-aware . So even if Dennett does away with any aprioristic distinction between natural and artificial and intends to have us look at cognitive performances of actual, biologically based, minds as situated, with other (biological or not) possible kinds of minds, in the same design space (Dennett 1995), he also proposes criteria for human minds as a specific kind of minds. Cognitive architectures for communication and language make for thresholds: there are characteristics the minds of linguistic creatures possess because they are linguistic. Allowing namely for mental acts such as endorsing and affirming one’s own beliefs and making voluntary decisions, language simply makes human minds much more powerful than those of any other animals – for Dennett although this is not a difference in kind, only a difference in degree, it is still big enough to make all the difference in moral terms .
7. Conclusion

The challenge in the study of cognition is whether it allows us to think of the world as ultimately mindless and Dennett’s philosophy is an answer to that challenge. It is, thus, especially significant that his anti-representationalism does not in any way amount to an elimination of subjectivity from our conception of the world. In fact, it is remarkable that he has been accused of being something like a third-person absolutist, insensitive to ‘real’, first-person problems of mind, when in fact the central thesis of IST concerns the indispensability of mentalism. It is one thing to keep us from an appeal to intuition and to something unique about our kind of minds, and to argue that frontiers between mental and non-mental may be fuzzier, and the transition more gradual, than what we would intuitively believe, another thing to think we could ever do away with semantic notions when dealing with cognition.


Brook, A. and Ross, D. 2002. Daniel Dennett. Cambridge: Cambridge University Press.
Cantwell Smith, B. 1996. On the Origin of Objects. Cambridge MA: MIT Press
Dahlbom, B, (ed). 1993. Dennett and His Critics. Oxford: Blackwell
Dennett, D. 1969. Content and Consciousness. London: Routledge and Kegan Paul.
Dennett, D. 1978. Brainstorms. Cambridge MA: MIT Press
Dennett, D. 1983. Elbow Room - Varieties of Free Will Worth Wanting. Cambridge MA: MIT Press.
Dennett, D. 1987. The Intentional Stance. Cambridge MA: MIT Press.
Dennett, D. 1991. Consciousness Explained. Boston: Little Brown.
Dennett, D. 1995. Darwin’s Dangerous Idea. New York: Touchstone.
Dennett, D. 1996. Kinds of Minds. New York: Basic Books.
Dennett, D. 1998. Brainchildren. London: Penguin.
Dennett, D. 2003. Freedom Evolves. London: Penguin.
Fodor, J and Lepore, E 1993. Is Intentional Ascription Intrinsically Normative, in Dennett and His Critics, edited by B. Dahlbom, 137-161. Cambridge MA: Blackwell.
Haugeland, J, 1997. Having Thought. Cambridge MA: Harvard University Press.
Miguens, S., 2002. Uma Teoria Fisicalista do Conteúdo e da Consciência – D. Dennett e os debates da filosofia da mente. Porto: Campo das Letras.
Nagel, T., 1995. Other Minds. Oxford: Oxford University Press.
Ross, D., Brook, A. and Thompson, D., 2000. Dennett’s Philosophy. Cambridge MA: MIT Press.
Searle, John, 1993. The Rediscovery of the Mind. Cambridge MA: MIT Press.