David Chalmers:
THE CONSCIOUS MIND (Oxford University Press, 1996)


Home | The whole bibliography | My book on Consciousness

(Copyright © 2000 Piero Scaruffi | Legal restrictions - Termini d'uso )


The Australian philosopher David Chalmers presents a formidable theory of consciousness. Basically, Chalmers believes that consciousness is due to "protoconscious" properties that must be ubiquitous in matter and that "psychophysical" laws, not of the "reductionist" kind that Physics employs, will account for how conscious experience arises out of those properties. There is, instead, nothing mysterious about our cognitive faculties, such as learning and remembering: they can be explained by the physical sciences the same way they explained physical phenomena. Chalmers therefore changes the scope of the mind-body problem, by enlarging the "body" to include the brain and its cognitive processes, and by restricting "mind" to conscious experience. Cognition migrates to the body. Consciousness, on the other hand, is truly a different substance, or, better, a different set of properties, and just cannot be explained by the "natural" laws of the physical sciences. The study of consciousness requires a different set of laws, just because consciousness is due to a different set of properties.

The book is organized like a mathematical treatise, with definitions first, a few corollaries and finally the main argument.

Chalmers contends that the mind is more than just conscious experience, and by this he probably means that there is more in the brain than just consciousness ("mind" is an ambigous term, and some probably use it interchangeably with "consciousness", in which case Chalmers' statement would be a contradiction in terms). For Chalmers, "mind" is any state of the brain that causes behavior. For example, I may drink because I am thirsty, I may move my hands because I want to grab an object, I may buy a plane ticket because I believe the fare will go up. These "mental" states may or may not be conscious. Chalmers therefore distinguishes between the conscious experience, that he calls the "phenomenal properties of the mind", and the mental states that cause behavior, that he calls the "psychological properties of the mind". Phenomenal states deal with the first-person aspect of the mind, whereas psychological states deal with the third-person aspect of the mind.

Psychological properties have, by his definition, a "causal" role in determining behavior. Whether a psychological state is also a phenomenal (conscious) state does not matter from the point of view of behavior. What conscious states do is not clear, but we know that they exist because we "feel" them.

Mental properties can therefore be divided into psychological properties and phenomenal properties. These two sets can be studied separately. It turns out that psychological properties (such as learning and remembering) have been and are studied by a multitude of disciplines, and in a fashion not too different from physical properties of matter (given their "causal" nature), whereas phenomenal properties constitute the "hard" problem. A psychological property causes some behavior, no less than most material properties. A phenomenal property is a fuzzier object altogether.

Chalmers also distinguishes awareness and consciousness: awareness is the "psychological" aspect of consciousness. Whenever we are aware, we also have access to information about the object we are aware of. Awareness is that access. It is a psychological state that has a "causal" nature. "Consciousness" is a term more appropriately reserved for the phenomenal aspect of consciousness (for the emotion, for the feeling).

Chalmers is, de facto, separating the study of cognition from the study of consciousness. Cognition is a psychological fact, consciousness is a phenomenal fact. Psychological facts, by virtue of their causal (or functional) nature, can be explained by the physical sciences. It is not clear, instead, what science is necessary to explain consciousness. To start with, Chalmers focuses on the notion of supervenience.

Chalmers goes to a great extent to clarify the theory of supervenience, but mostly ends up proving how flaky that theory is. A set B of properties supervenes on a set A of properties if any two systems that are identical by properties A are also identical by properties B. For example, biological properties supervene on physical properties: any two identical physical systems are also identical biological systems.

"Local" supervenience is a stricter variant of supervenience: two identical paintings are not worth the same amount, because one is the original and the other one is a replica, therefore financial properties do not supervene locally on physical properties. Whenever the context matters (in this case, who painted the painting), local supervenience fails. Local supervenience implies global supervenience, but not viceversa. (One could object that an "exact" copy of a painting "is" identical to the original for all purposes. Chalmers' idea of a replica presupposes that it is not an exact, atom by atom, copy, it is a similar painting that an art expert can tell from the original).

"Logical" supervenience (loosely, "possibility") is also a stricter variant of supervenience: some systems could exist in another world (are "logically" possible), but do not exist in our world (are "naturally" impossible). Elephants with wings are logically possible, but not naturally possible. Systems that are naturally possible are also logically possible, but not viceversa. For example, any situation that violates the laws of nature is logically possible but not naturally possible. Natural supervenience occurs when two sets of properties are systematically and precisely correlated in the natural world. Logical supervenience implies natural supervenience, but not viceversa. In other words, there may be worlds in which two properties are not related the way they are in our world, and therefore two naturally supervenient systems may not be logically supervenient. (Although it is not clear what is logically possible that is naturally impossible: it all depends by one's definition of "our" world and by one's scientific knowledge, as ancient Greeks would have certainly considered Einstein's spacetime warping a natural impossibility).

Chalmers then argues that most facts supervene logically on the physical facts (if they are identical physical systems they are identical, period). There are few exceptions and consciousness is one of them. Consciousness is not logically supervenient on the physical.

This is by far the weakest part of the book. Once one "sees" that there is truly only one kind of supervenience, the magician's trick is revealed. What Chalmers is saying is, quite simply, that the physical sciences can explain everything except consciousness, and he uses his several variants of supervenience to prove it mathematically. Truth is that, at the end, we have to take his word for it.

So Chalmers conclude that consciousness "cannot" be explain by physical sciences (more appropriately, cannot be explained "reductively"). The first 132 pages, basically, read like a (more or less) rigorous proof that consciousness cannot be explained by existing science.

Unlike other philosophers, Chalmers does not conclude that consciousness cannot be explained, tout court, but only that it cannot be explained the way the physical sciences explain everything else: by reducing the system to ever smaller parts. Chalmers leaves the door open for a "nonreductive" explanation of consciousness. Chalmers is claiming that Materialism is false, that's all. His proof is weak at best. Following the same reasoning, one could prove that Biology is false. Chalmers' argument basically goes back to the "zombie" question: if a physical copy of you is built by some futuristic machine, would that copy of you experience the same feelings you experience? Chalmers argues that physical identity is not enough, but, honestly, his 132-page proof does not amount to much more than an act of faith.

Whatever the merits of the proof, his claim is that Materialism is doomed. Chalmers does not rule out "monism",t he theory that there is only one substance; he only rules out that the one substance of this world is matter as we know it with the properties we currently know. So even his claim that Materialism is false strongly depends on the definition of "matter". Materialists would certainly still call "matter" the one substance that includes the known properties of matter, in which case Chalmers' theory would be simply a new materialist theory of mind.

Then Chalmers proceeds to present his own theory of consciousness, that he calls "naturalistic dualism" (but might as well have called "naturalistic monism"). It is a variant of what is known as "property dualism": there are no two substances (mental and physical), there is only one substance, but that substance has two separate sets of properties, one physical and one mental. Conscious experience is due to the mental properties. The physical sciences have studied only the physical properties. The physical sciences study macroscopic properties like "temperature" that are due to microscopic properties such as the physical properties of particles. Chalmers advocates a science that studies the "protophenomenal properties" of microscopic matter that can yield the macroscopic phenomenon of consciousness.

His parallel with electromagnetism is powerful. Electromagnetism could not be explained by "reducing" electromagnetic phenomena to the known properties of matter: it was explained when scientists introduced a whole new set of properties (and related laws), the properties of microscopic matter that yield the macroscopic phenomenon of electromagnetism.

Similarly, consciousness cannot be explained by the physical laws of the known properties but requires a new set of "psychophysical" laws that deal with "protophenomenal properties". Consciousness supervenes naturally on the physical: the "psychophysical" laws will explain this supervenience, they will explain how conscious experiences depend on physical processes.

Chalmers emphasizes that this applies only to consciousness. Cognition is governed by the known laws of the physical sciences.

Chalmers then turns to the relationship between cognition and consciousness. Phenomenal (conscious) experience is not an abstract phenomenon: it is directly related to our psychological experience. Consciousness interacts with cognition and that interactaction gets expressed via what Chalmers calls "phenomenal judgements" ("I am afrai", "I see", "I am suffering"). These are acts that belong to our psychological life (to cognition) but that are about our phenomenal life (consciousness).

Chalmers realizes a paradox: phenomenal judgements, that are about consciousness, belong to cognitive life, therefore can be explained reductively, but he just proved that consciousness cannot be explained reductively. The way out of the paradox is to assume that consciousness is not relevant, that we can explain phenomenal judgements even if/when we cannot explain the conscious experience they are about, i.e. the explanation does not depend on that conscious experience, i.e. that feeling or emotion is irrelevant.

Chalmers cautions that this conclusion does not necessarily imply that consciousness (as in "free will") is irrelevant for behavior, but it surely does smell that way. If we can explain behavior about consciousness without explaining consciousness, it is hard to believe that behavior requires consciousness.

Chalmers takes these facts literally: our statements about consciousness are part of our cognitive life, and therefore can be explained quite naturally, just like any other behavior. I speak about my feelings the same way I raise a hand. There is a physical process that explains why I do both. It also happens that we "are" conscious, not just that we talk about it, and that part cannot be explained (yet). If we had a detailed understanding of the brain, we could predict when someone would utter the words "I feel pain". So Chalmers believes that our talk about consciousness will be explained just like any other cognitive process, just like any other bodily process. This is not the same as explaining the conscious feelings themselves, and it leaves open the option that feelings are but an accessory, an evolutionary accident, a by-product of our cognitive life with no direct relevance to our actions.

This conjecture represents a quantum leap for philosophy of mind. Since Descartes, the dilemma has been how do body and mind communicate. Chalmers realizes that "body" extends to the brain, and brain is responsible for many phenomena that we consider mind and that are no more mysterious than the movement of a hand. Therefore, within the Cartesian dichotomy, "body" must be enlarged to encompass brain processes and "mind" must be restricted to conscious experience. Otherwise, most of the mystery is not a mystery at all: the way "mind" remembers or learns is no more mysterious than the way a muscle gets stronger or weaker. What is mysterious is that "remembering" and "learning" are sometimes associated with conscious experience. That is the real puzzle: how does a brain process of remembering (that is ultimately an electrochemical process of neurons triggering each other) communicate with our conscious life of feelings and emotions that seems to be located in a completely different dimension? The paradox to be explained is not that body and mind communicate but that cognition and consciousness communicate.

Chalmers also offers an explanation of phenomenal judgement based on the theory of information. After all, his definition of "cognition" is pretty much that of "information processing": cognition is the processing of information, from the moment it is acquired by the senses to the moment it is turned into bodily movement. Information is what "pattern" is from the inside. Consciousness is information about the pattern of the self. Information becomes therefore the link between the physical and the conscious. Since information is ubiquitous, he also gets entangled in the question whether everything has feelings, and, of course, if experience is ultimately due to information, there is no reason why anything would not be associated with "experience". Just like every other physical property we know is widespread in the universe, there is no reason why "experience" (defined as the macroscopic effect of "protophenomenal properties") should no be widespread. Objects may well have a degree of consciousness. Chalmers' "natural dualism" is therefore a close relative of "panpsychism".

Furthermore, if information leads to experience, there must be a lot more experience than we "feel" because the brain processes a lot more information than we are aware of. But then parts of the brain may have experience that does not travel to the "I". The "I" is not necessarily all the is experienced by the brain. The "I" may simply be a chunk of coherent information out of the many that arise all the time in the brain.

The book includes two chapters on popular subjects, just for the heck of it: one on Artificial Intelligence and one on Quantum Mechanics. The latter is another reason to buy the book.

Chalmers' arguments are adorned with lots of subtleties for philosophers, but Chalmers is certainly aware that those philosophical subtleties tend to annoy readers from other disciplines (and tend to age badly).

The bottom line is that Chalmers believes consciousness can be explained by studying nonphysical properties of matter, and that the mind-body problem must be recast as a cognition-consciousness problem.

The book offers such a broad survey of theories of mind and contemporary neuroscience that it's hard to point out any significant omission. However, it would not hurt to include a treatment of the nascent discipline of self-organizing/ complex systems. An almost contemporary book is Stuart Kauffman's "Reinventing the Sacred", in which one of the most famous purveyors of self-organizing systems explains why he thinks that consciousness cannot be reduced to the physical. There is obviously another school of thought (grounded in mathematical equations) that is reaching conclusions similar to Chalmers' conclusions, but it is not given any relevance in this book. It is not even clear where it would fit in Chalmers' classification of theories.


Permission is granted to download/print out/redistribute this file provided it is unaltered, including credits.