Return to index of reviews

Sweet Dreams:
Philosophical Obstacles to a Science of Consciousness
by Daniel C. Dennett
The MIT Press, April 1, 2005
178 pages, hardcover
Review by Jim Walker

In this little book, Dennett attempts to clarify his thinking about his hypothesis of consciousness to his mentors, critics and the lay pubic. The book comprises various essays and lectures given from 1999 to the present. Dennett explains that nothing supernatural or mysterious occurs in the brain and that, in principle, a sophisticated software/hardware machine or countless stupid homunculi (each of which could consist of a machine-itself) could account for the phenomenon we call consciousness.

If you haven't read Dennett's Consciousness Explained or his other works on his brain theories, you may not understand a lot of his explanations. In Consciousness Explained he presented his "Multiple Drafts" model and in this book he provides other metaphors such as his "Fame in the Brain" to "Fantasy Echo" models (further extending his "Multiple Drafts" model).

Dennett's demystifying of consciousness has apparently upset a lot of people, especially philosophers and religionists. He tries to show the problems with the "Zombic Hunch" idea, a popular misconception held by many philosophers that tries to separate consciousness from brain function. According to Dennett, the Zombi illusion compares with the illusion of the sun going around the earth, in that we now know not to trust our visual illusions. The same with the Zombi Hunch; sure it seems valid, but once you understand the problems, the illusion becomes benign.

As a scientist, Dennett upholds the idea of third-person science, and that a scientist could gain knowledge about consciousness strictly from this third-person stance, something which he calls "heterophenomenology." Although I can understand that scientists might gain enough knowledge about consciousness to eventually construct a conscious entity (a hardware/software brain, perhaps), through third-person means, but that does not necessarily mean that scientists can understand the experience of consciousness. Apparently Dennett disagrees. He thinks that as long as enough data gets communicated by the conscious entity to the scientist, then the scientist should have the ability to understand conscious states including emotions and feelings! In an extraordinary statement Dennett asks, "are we really so sure that what it is like to see red or blue can't be conveyed to one who has never seen colors in a few million or billion words?" This brings up a recollection from George Miller's observation that a human being can only hold 6 ideas in the mind at one time (plus or minus 2). Maybe a super-brained evolved alien could hold a million words in its head, but I can't even remember what I thought five minutes ago! I just don't see how humans could ever understand experienced states without actually experiencing them. An heterophenomenological scientific understanding of consciousness can't possibly include everything (see the Bronowski quote below), but it could build a consciousness in the same way that a scientist who has limited chess knowledge might construct a chess computer program to beat anyone.

There is no absolute knowledge. And those who claim it, whether they are scientists or dogmatists, open the door to tragedy. All information is imperfect.
--Jacob Bronowski (The Ascent of Man "Knowledge or Certainty")

Although I find Dennett's ideas fascinating and enlightening, I still do not understand a lot about what he says about qualia. Yes, I understand the problems of the Zombic Hunch, and I agree with him (I think). Yes, I understand that nothing mysterious or supernatural occurs during consciousness. I understand his Multiple Drafts and Fame in the Brain model (I think). But I do not understand his resolution of the qualia problem (or does he have one?). Now perhaps Dennett doesn't consider qualia a problem, but I see no satisfactory explaining for the mechanism of feelings, sensations, emotions (qualia). Dennett admits the difficulty to deny qualia and that he has worked on the task for years without scant progress. But what does he mean by this? Does he really deny qualia and that the task for him involves convincing others? Or does he accept qualia as a real phenomena and he hasn't made scant progress in explaining it? I suspect he means that qualia exists only as a user illusion and that it emerges out of the Multiple Drafts model, but I don't have certainty he means this.

As a non-scientist perhaps my ideas represent a weak naïve view but in my model, the mind consists of at least two broad categories: Hardware/software (which involve detecting, thinking, and reacting) and Qualia (which involves feelings and emotions). The hardware/software part does not require qualia. For example, computer robots (true zombies) can perform simple thinking, detecting and reacting (software algorithms, camera eyes, touch sensors, chemical detectors, etc.). Qualia, on the other side of the token, does not require thinking (at least for a short period). For example, by practicing no-thought through mediation or biofeedback methods, one can experience sensations without thought at all. According to Victor S. Johnston (read the review on his book here), feelings give meaning to the thinking part of our brain and serve as the foundation of all semantic networks. I might even go further to speculate that qualia serves as the most important criteria for consciousness. If you had to choose only one, would you pick thinking over feelings (which would make you a robot) or feelings over thinking (which would make you dumb and stupid, but at least you would feel)?

No doubt that feelings involve electrical/chemical reactions in the brain and that they don't require mysterious explanations. Anyone who has taken Lysergic Acid Diethylamide or Salvia divinorum knows that even a minute amount of chemicals can dramatically alter the way the brain feels. Neurotransmitters, hormones-- chemicals, no doubt contributes to feelings in the brain, but how?

I think the genius of Dennett comes from his explaining the hardware/software part of the brain. His Multiple Drafts model explains a lot toward how the mind sees, thinks and calculates which, no doubt, one could use to eventually build a robot that sees, thinks and calculates like a person. But even robots today can see, think, and calculate. To say that feeling will just emerge, as long as the hardware/software gets built to the right complexity doesn't satisfy as an explanation. I still see no explanation from Dennett or anyone else on how these electro-chemicals (or anything else) in the brain can create feelings.

It may not answer many of the difficult questions, but Dennett's book gives us the latest ideas, problems and thinking of consciousness. It also serves as a good reference book for future brain theorists to explore.

I look forward to his future writings about one of life's greatest, yet answered, questions.

A few quotes from the book:

What we now know is that each of us is an assemblage of trillions of cells, of thousands of different sorts. Most of the cells that compose your body are descendants of the egg and sperm whose union started you (there are also millions of hitchhikers from thousands of different lineages stowed away in your body), and to put it vividly and bluntly, not a single one of the cells that compose you knows who you are, or cares.

For we now know that the "miracles" of life-- metabolism, growth, self-repair, self-defense, and of course, reproduction-- are all accomplished by dazzlingly intricate, but nonmiraculous, means.

The sheer existence of computers has provided an existence proof of undeniable influence: there are mechanisms-- brute, unmysterious mechanisms operating according to routinely well-understood physical principles-- that have many of the competences heretofore assigned only to minds.

The critics have found that it's hard to say, exactly: qualia, feelings, emotions, the what-it's-likeness (Nagel) or the ontological subjectivity (Searle) of consciousness.

It still seems as if the Earth stands still and the Sun and Moon go around it, but we have learned that it is wise to disregard this potent appearance as a mere appearance. It still seems as if there's a difference between an thing at absolute rest and a thing that is merely not accelerating within an inertial frame, but we have learned not to trust this feeling.

When the brain takes the suggestion, the brain is forming a belief or expectation, not painting a picture for itself to look at.

I don't maintain, of course, that human consciousness doesn't exist; I maintain that it is not what people often think it is.

Banished along with the Emperor are what might be called the Imperial Properties: the two most mysterious varieties being the Qualia Enjoyed by the Emperor and the Imperial Edicts of Conscious Will.

Once we appreciate all the nonmysterious ways in which the brain can create benign "user-illusions," we can begin to imagine how the brain creates consciousness.

Yes, it is indeed difficult to deny that there are qualia. I've been working on the task for years, with scant progress!

Everybody's favorite example of qualia are "subjective colors," such as the luscious yellow you enjoy when you look at a ripe lemon or the breathtaking shade of warm pink you see in the western sky during a glorious sunset.

...are we really so sure that what it is like to see red or blue can't be conveyed to one who has never seen colors in a few million or billion words?

As long as your homunculi are more stupid and ignorant than the intelligent agent the compose, the nesting of homunculi within homunculi can be finite, bottoming out, eventually, with agents so unimpressive that they can be replaced by machines.

Your body is made up of some trillions of cells, each one utterly ignorant of all the things you know. If we are to explain the conscious Subject, on way or another the transition from clueless cells to knowing organizations of cells must be made without any magic ingredients.

There is not such thing as first-person science, so if you want to have a science of consciousness, it will have to be a third-person science of consciousness.

If all that matters is the computation, we can ignore the brain's wiring diagram, and its chemistry, and just worry about the "software" that runs on it. In short-- and we now arrive at the provocative version that has caused so much misunderstanding-- in principle you could replace your wet, organic brain with a bunch of silicon chips and wires and go right on thinking (and being conscious, and so forth).

Fame in the brain provides, perhaps, a useful way of thinking about the "political" access that some contents may have to the reins of power in the ongoing struggle to control the body, but it has nothing to say about the brute, lower-order what-it-is-like-ness of phenomenal consciousness.

Now perhaps dogs have similar reflective episodes in their inner lives; if they do, then surely they are just as conscious as we are, in every sense. But I hypothesize-- this is, the empirical going-out-on-a-limb part of my view-- that they do not.

Let me sum up. I have ventured (1) the empirical hypothesis that our capacity to relive or rekindle contentful events is the most important feature of consciousness-- indeed, as close to a defining feature of consciousness as we will ever find; and (2) the empirical hypothesis that this echoic capacity is due in large part to habits of self-stimulation that we pick up from human culture, that the Joycean machine in our brains is a virtual machine made of memes.


Series Foreword
1 | The Zombic Hunch: Extinction of an Intuition?
1 The Naturalistic Turn
2 The Reactionaries
3 An Embarrassment of Zombies
4 Broad Functionalism and Minimalism
5 The Future of an Illusion
2 | A Third-Person Approach to Consciousness
1 Scientists from Mars
2 Folk Theories and Philosophy
3 Heterophemomenology Revisited
4 David Chalmers as Heterophenomenological Subject
5 The Second-Person Point of View
3 | Explaining the "Magic" of Consciousness
1 The Thankless Task of Explaining Magic
2 Dismantling the Audience
3 The Tuned Deck
4 | Are Qualia What Make Life Worth Living?
1 The Quale, An Elusive Quarry
2 Change Blindness and the Question about Qualia
3 Sweet Dreams and the Nightmare of Mr. Clapgras
5 | What RoboMary Knows
1 Mary and the Blue Banana
2 "Surely" She'll Be Surprised
3 You Had to Be There!
4 RoboMary
5 Locked RoboMary
6 | Are We Explaining Consciousness Yet?
1 Clawing Our Way toward Consensus
2 Competition for Clout
3 Is There Also a Hard Problem?
4 But What about "Qualia"?
5 Conclusion
7 | A Fantasy Echo Theory of Consciousness
1 Fleeting Fame
2 Instant Replay
8 | Consciousness: How Much Is That in Real Money?

To obtain this book, click below:
Sweet Dreams: Philosophical Obstacles to a Science of Consciousness

Other books by Daniel C. Dennett:

Freedom Evolves (2003)

Brainchildren (1996)

Kinds of Minds (1996)

Darwin's Dangerous Idea (1995)

Consciousness Explained (1991)

The Intentional Stance (1987)

Elbow Room (1984)

The Mind's I with Douglas Hofstadter (1981)

Brainstorms (1978)

Content and Consciousness (1969)

Return to freethinkers home page