Rhetorically, this book is a masterpiece. Without a doubt, Dennett is a master of the persuasive argument, and he pulls out all the stops here: see the gradual shifts in word choice ("the familiar Cartesian Theater" at the beginning morphs by the end into "our old nemesis, the Cartesian Theater"), very carefully chosen metaphorical models in order to pull in all kinds of technological magic, the way he performs a mea culpa without actually admitting guilt near the end of the book, and prevalent rhetorical questions and blanket declarations to mask over the places where things are most awkward. It's really well done. If it weren't for the theoretical issues (really, one major theoretical issue), this would easily be a 5 star book. It is both rigorous yet accessible, and it is a wonderfully enjoyable exploration of consciousness and its limitations. The book has a lot to teach people about their own self-experience, even (and especially!) those who aren't prone to agree with him (such as me). To borrow a term Dennett uses to describe a fellow philosopher's work, this book is an "instructive failure". It's definitely worth your time to read.

I read this book, however, with a particular question in mind: when I see something in my mind's eye, where does that image reside? (Note that I have a vivid mind's eye — some people don't even have a mind's eye at all, and don't even know they're missing one! See this Discover article for more.) This book was referred to me because it was apparently a solid materialist argument and explanation. Unfortunately, Dennett simply doesn't have an answer for that question. His rhetoric obscures that fact very well (although he cops to side-stepping questions of "the nature of things" in Appendix A), but ultimately he takes a functionalist approach, and that handily does away with the very evidence which I was looking for him to explain.

To see how this happens, it is first necessary to realize that we can't actually prove anyone else is conscious in the same sense we are. As was popularized in The Matrix, we could be dreaming and not realizing it, and everyone else could be projections of our own consciousness. Like in A Beautiful Mind (SPOILER ALERT), we could be populating our experience with non-existent people. No test proves anything more than the fact that it is a very clever illusion or bit of calculation going on — but, ultimately, our own consciousness is uniquely experienced by us. Everyone else could be a zombie (in the philosophical sense), and we have no way to know differently. This is not the point Dennett starts with, but it's implicit within Dennett's functionalist approach. The move Dennett makes is to then project that back on to us, and say that we, too, have no consciousness in any real sense. How could you prove differently to anyone else? What objectively accessible function can you perform as a conscious entity which these potentially unconscious entities could not?

The problem is that the experience of having consciousness — not just computational complexity or a narrative world, but some "audience in a Cartesian Theater" — is not simply a "doctrine" (that word choice is a great example of Dennett's excellent and subtle rhetorical moves), but stems from evidence. It's not functional-objective evidence, but it is evidence none the less. That evidence is the experience of having consciousness. As much as you chip away at what the consciousness is and does (and Dennett does that excellently), there is always still an "I" which is having the experience. That's the whole point of Descartes, and Dennett simply does not feel the need to account for it, instead presenting the epistemological limitations of his "heterophenomonology" as ontological statements. But a map is not the territory, and a scientific model is not the thing itself. Simply because heterophenomonology cannot take with subjective evidence seriously does not mean that subjective evidence should not be taken seriously. This point is very poetically (and popularly) driven home in Nature, Man and Woman, and Dennett is aware of it, but simply dismisses it as not relevant. If you read through my progress in this book, you'll see lots of places where those dismissals are explicitly identified, because they drove me nuts. (Here's an example.)

To a certain extent, though, he is right: phenomonology is not relevant within the context of science, where heterophenomenology is sufficient. As a text in cognitive science, Dennett's approach is really insightful and makes many challenging points. In the world of philosophy, though, the phenomonology needs to be accounted for. It's not philosophically sufficient to assume an air of "anthropological" condescension towards at least our own phenomonology — that's discounting a vital piece of evidence without basis, and if it cannot be accounted for, then consciousness (at least our own consciousness) is insufficiently explained. Dennett claims that consciousness is just an illusion — but who is perceiving the illusion? If it is a deception, who is being deceived? That claim is actually begging the question, and Dennett responds by retreating to functionalism, and saying it is our memories that are being deceived. But I do not simply remember being conscious — I am conscious of being conscious! But that sentence doesn't even make sense in the functionalist conception. But Dennett seems to be paraphrasing Groucho Marx: "Who are you going to believe? Me or your own mind's eye?"

It is notable at this point, though, to mention that the kind of "anthropological" condescension Dennett advocates has actually come under an extreme kind of fire as a Eurocentric and fundamentally racist pattern of thought which results not in real knowledge, but more in generating a hybrid of European presumptions and European interpretative "data". For more on this point, see Provincializing Europe: Postcolonial Thought and Historical Difference. I bring this up because so much of the argumentation feels like 19th century thinking, not 21st century thinking, and the model Dennett proposes reads more like an artifact of philosophy's own artifices and methodologies than an account of reality. Appeals to the specialness of language (strictly conceived), to Newtonian physics, and conflations between non-materialism and dualism really reinforce that sense.

On top of all of this, Dennett builds heavily upon ideas of evolutionary psychology and memes, which are roundly destroyed as non-scientific materialistic doctrines in The Spiritual Brain: A Neuroscientist's Case for the Existence of the Soul. There is also some significant ethological ("animal-psychological") claims about the communication and cognition of animals which are simply empirically wrong given research done by people like Marc Bekoff. And the idea that we could have immortality by "copying our brain's software" is not very comforting to this particular copy (a point made well in The Prestige). There is no serious engagement with psi or spirituality, as though they are self-evidently irrelevant to discussions of consciousness.

With all these complaints, however, the book still gets 4 stars. It's definitely worth a read, and it is really an eye-opening revelation of just how complicated the apparently simple idea of "the conscious mind" really is. Highly suggested.