|
|
|
Think of three robots, all of which behave exactly like successful well-adapted animals except that we know there is a computer in each of them running a program which is responsible for their behaviour.
Animal A has a mackintosh computer, and program p running on it.
Animal B has a PC computer for a brain, but it has the same program running on it.
This is like animal A having WORD written for the mackintosh running on it and animal B having the same program WORD running on it (only this time WORD written to run on the different hardware, the PC as distinct fro m the Mac).
Because it's the same program, when a tiger approaches, the same subroutines - let's say the same calculations - will get performed, although because the hardware is different, what is going on physically in the two animals will be different.
Identity theory says because what is going on physically is different, so will be the experience of the two animals (identity theory identifies the phenomenal experience with the physical goings-on). So the identity theorist says that A and B would have different experiences.
Machine functionalism says au contraire that the phenomenal experience of the animal is to be identified with the calculation being performed. And because the program is the same in the two animals, the same calculation will be being performed in both cases. So machine functionalism says that the phenomenal experience of the two animals will be the same, even though there are different physical processes going on. Both animals will have the same experience - because the same calculation is being carried out. It doesn't matter that the two physical processes are different.
Animal C has different hardware again - say it's a Sun cluster or something. But is also has a completely different program running on it, not a version of WORD. It's a word-processor, but a completely different one. The end point however is the same - Animal C shows behaviour that is as successfully adapted to its environment as is the behaviour of A or B.
But in the case of animal C, when the tiger approaches, a different set of subroutines will be triggered. It's a different program, so although it will do the same job, it does it in a different way.
Teleological functionalism says: as far as the phenomenal experience of C is concerned, this difference in calculation being carried out doesn't matter. Confronted with the tiger, it will have the same experience as A or B. Teleological functionalism is able to say this because it identifies the phenomenal experience not with physical process (identity theory) nor with calculation being carried out (machine functionalism) but with the function that that bit of the program is performing. If all three programs have sections in them designed to get the body ready for flight say, all three animals will have the same phenomenal experience (the feeling of terror, say) when the relevant section of their program is running. In A and B that section will involve the same calculations, (even though the physical goings on in the two computer-brains will be different). In the case of animal C however not only will the physical goings-on be different when the tiger appears: the calculations performed by its program will be different too. But the function performed (getting the animal ready to flee) will be the same in C as it is in A and B. And according to the teleological functionalist, this will mean that the phenomenological experience (feeling of terror) will be the same also.
VP
Updated 29:04:03
 |  |
Revised 12:03:03
