"Neo-dualism" is Perry's term for the type of property dualism defended by Chalmers. This is the idea that phenomenal experiences (or qualia) do not logically supervene on physical properties. (Below I'll distinguish between logical and natural supervenience.) Notice that neo-dualism about experience is entirely compatible with (indeed usually goes along with) functionalism about most or all of the rest of our mental lives.
1. The Zombie Argument.
Background: A "zombie" in philosophical usage is a being exactly like us in certain ways, but completely without qualia (i.e. without phenomenal consciousness, without the "what it's like" aspect of our mental lives). (Zombie arguments are also called, less vividly, "absent qualia" arguments.)
Exactly like us in which ways, exactly? That varies with different forms of the argument. To use the possibility of zombies as an argument against behaviorism, one needs only the assumption that the zombie behaves exactly like us. For an argument against functionalism, we need to assume that the zombie has the same functional states we do. For an argument against every form of (property) physicalism, a zombie must be a being exactly like us in every physical respect: an exact physical duplicate.
So the argument goes something like this.
1. I can conceive of a world physically identical with our world, but in which there are no qualia.
2. Therefore, there is a possible world physically exactly like ours, but without qualia.
3. Therefore, qualia do not logically supervene on physical properties.
4. Therefore, (property) physicalism is false.
The step from 2 to 3 uses the definition of logical supervenience: A-properties logically supervene on B-properties if and only if there are no two logically possible worlds with the same B-properties but different A-properties.
The step from 3 to 4 assumes that any version of physicalism implies that mental properties (including the property of having certain qualia) logically supervene on physical properties. (This is obviously true for behaviorism and the identity theory, and it is true for any version of functionalism that assumes that the realizers of functional properties must be physical.)
Note that step 3 is compatible with the view that qualia naturally supervene on physical properties. A-properties naturally supervene on B-properties if and only if there are no two naturally possible worlds (i.e. worlds compatible with the laws of nature in our own world) with the same B-properties but different A-properties.
2. The Knowledge Argument.
Background: the knowledge argument grows out of Thomas Nagel's argument in his famous paper, "What is it Like to Be a Bat?" The most-discussed current version of the argument, though, is due to Frank Jackson.
We imagine that Mary is a color scientist who knows all the physical facts about color perception. However, she has grown up in a black and white room and has never seen colors other than black and white.
1. Mary knows all the physical facts about color perception.
2. When Mary steps out of the room, she learns a fact she didn't already know (namely what it is like to experience red).
3. Therefore, what it's like to experience red is not a physical fact.
The argument here is extremely simple. If Mary knows all the physical facts, and yet learns something new about experience, then the new fact she learns about experience can't be a physical fact.
3. The Modal Argument.
The modal argument in its contemporary version stems from Saul Kripke (in his book Naming and Necessity and his article "Identity and Necessity"). Kim discusses some versions of the modal argument also, so this one should look pretty familiar. It is probably best thought of as an argument specifically against the identity theory. (Since Perry defends a version of the identity theory, it's relevant to his view.)
Consider a mental property M. According to the identity theory, there is a physical state P such that M = P.
1. It seems possible that M is not P
2. There is no way to explain away this apparent possibility (see below)
3. So it is possible that M is not P (from 1, 2)
4. If M = P, then it is necessary that M = P (necessity of identity principle)
5. So M ≠ P (from 3, 4)
The step from 3 and 4 to 5 rests on the fact that a proposition is necessary if and only if its negation is not possible.
The reason premise 2 is necessary is that there are some cases in which it seems possible that an identity could be false even though it isn't really possible. For example, Hesperus = Phosphorus, and (Kripke argues) this is necessarily true. (Both 'Hesperus' and 'Phosphorus' are names for the planet Venus, and there is no way Venus could have failed to be Venus.)
However, it may seem as though Hesperus might have failed to be Phosphorus. Hesperus is a name for the evening star, and Phosphorus, "bearer of light," a name for the morning star. It seems that we can imagine a situation in which these might have turned out not to be the same thing. But, Kripke argues, what we are imagining here is not really that Venus is not Venus. Rather, we are imagining a situation in which the star we see in the evening is not the same as the star we see in the morning.
For this explanation to work, it must be the case that the criteria for identifying at least one of Hesperus and Phosphorus could have picked out a different object instead. The criterion "brightest star in the evening" could have identified a different planet than Hesperus, i.e. Venus (if the orbits of the planets had been suitably different).
Kripke argues that it isn't plausible that something like this is at work in the case of M and P. In particular, our criterion for identifying M is just the way it feels; it seems that the criterion for identifying a phenomenal experience like M is just M itself. So there's no room to argue that this criterion could have picked out something different instead.
Perry calls his view "antecedent physicalism." The term "antecedent" doesn't actually describe a special version of physicalism -- Perry's view is just pllain old physicalism (in particular, a version of the identity theory). "Antecedent" is meant to capture the idea that Perry begins from a physicalist perspective, and asks whether the three arguments above are sufficient to show that he must give it up. He will argue that they don't.
Contents of the mind: experiences and thoughts. (Capacity for experience = sentience; capacity for thoughts = thought.)
We can have thoughts about experiences. In particular, we can have thoughts about sensations about pain. When we think about a sensation such as pain, we use a concept of pain. Perry suggests that the concept of pain includes both a phenomenological part and a psychological part. The psychological part consists of information about the typical causes and effects of the sensation (one might say about its functional role).
Theses that Perry takes to be part of common sense about the metaphysics of experience (epistemology comes in the next chapter):
1. Experiences and thoughts are inner causes and effects (p. 37).
2. Experiences have subjective characters (p. 38).
3. Some subjective characters are important because they are pleasant or
unpleasant (p. 40).
4. Some subjective characters are important because they carry crucial
information (p. 42). Perceptual experiences are a prime example.
5. Subjective characters are not external or historical aspects of inner states
(p. 43). (Perry's idea here is that although content may be "broad," subjective
characters aren't.)
Last update:
November 2, 2009.
Curtis Brown | Philosophy of Mind | Philosophy Department
| Trinity University
cbrown@trinity.edu