Bruno Marchal writes:

> It gets cumbersome to qualify everything with "given the appearance of > a physical world". As I have said before, I am not entirely convinced > that comp is true,

Nor am I. (Remind that no machine can, from its first person point of view, be "entirely convinced" that comp is true. Comp is an axiom for a theory, and the beauty of it is that comp can explain why it has to be a guess. The "yes doctor" has to be an act of faith. It is (meta?)-theological.

> precisely because because such ideas as a conscious computation > supervening on any physical process

This does not follow at all. We have already have some discussion about this and since then I have a more clear-cut argument. Unfortunately the argument is based on some result in mathematical logic concerning the distinction between real numbers and integers. We can come back on this in another thread. For a logician there is a case that "real number" are a simplification of the notion of natural number. An identical polynomial equation can be turing universal when the variables are conceived to belong to the integers, but is never turing universal when the variables belong to the reals. Well, a case can be made that the vacuum

> or on no physical process may be considered absurd.

This would be fair enough in case the idea that consciousness supervenes on physical processes was not absurd in the first place. In all your post you do assume comp. For comp to be false you have to assume actual physical infinities and give a reason why consciousness supervenes on that. But in some reasoning it seems clear to me you talk life if comp is true, when referring to the functional role of neurotransmitters, the fact that slight change in the brain are permitted, etc.

Standard computationalism says that mental processes supervene on physical processes, and moreover that these physical processes with their attendant mental processes may be emulated by a digital computer. The problems with this theory are:

1. The implementation problem: everything can implement a computation if you look at it the right way. Normally this is of no consequence - mapping the vibration of atoms in a rock to a word processing program would be at least as difficult as building a conventional computer and writing the software for it - but if computations can be conscious, then the conscious computations are hiding all around us. 2. The Maudlin/Marchal argument showing that even if you specify that a computer must handle counterfactuals in order to avoid the "trivial" conclusion (1), you end up concluding that physical processes are irrelevant to consciousness. You (BM) think that (1) is absurd but (2) is OK; Maudlin thinks that (1) and (2) are both absurd, and that therefore computationalism is a flawed theory. You would like to keep computationalism but drop the computers, i.e. the supervenience thesis. I am not certain which I would rather drop: computationalism or the idea of disembodied consciousness.
> It is quite possible, for example, that there is something special > about the structure of the brain which leads to consciousness, and a > digital computer will not be able to copy this, even if it copies 3rd > person observable behaviour. Against that idea is the question of why > we didn't evolve to be zombies, but maybe we would have if nature had > electronic circuits to play with.

You are saying that zombie are possible if comp is false. I can agree. Actually I believe that comp entails the existence of a notion of local zombie (which can make you believe that they are conscious during a time).

> If I had to guess between comp and not-comp I don't think I could do > better than flipping a coin.

Comp is my working hypothesis, and I tend to consider that arguing for or against comp is a bit of a waste of time (especially given that comp is undecidable for machines). Still I am astonished that you would flip a coin on that matter. Comp is just the statement that "I" am turing emulable at some level. Even if we have to emulate the quantum state of the entire galaxy (some "generalized brain") comp remain true unless we have both:
- my brain = the "complete physical" description of the galaxy
- the "object" galaxy is not turing emulable.

It is possible to drop computationalism and keep a form of functionalism, without introducing magical effects or even any new physics. For example, it is possible that consciousness is a property of actual neuronal activity, and although you might be able to emulate this activity using a digital computer, even to the point where you can build a substitute brain that behaves just like the biological equivalent, it won't be conscious. In order to build a proper replacement brain you would have to copy the actual physical structure of the neurons, not just emulate them. I am not sure that I believe this theory but it is theoretically possible.
Stathis Papaioannou
Be one of the first to try Windows Live Mail.
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at

Reply via email to