Rickert on embedded computation (was re: science of consciousness.)

Jerry Hull ZZZghull at stny.lrun.com
Tue Apr 28 20:43:33 EST 1998


On 28 Apr 1998 15:52:50 -0500, rickert at cs.niu.edu (Neil Rickert)
wrote:

>No, I disagree.  In fact this was the sort of thing that the
>disagreement between Bill Modlin and me was about.  We can say that
>something is a computation without having to map it into the action
>of a formal Turing machine.  From my perspective, the Turing theory
>is that of an idealized mathematical model of computation.  It is not
>a constraint on any actual computation, that it is required to
>conform to the idealized model.  We generally don't expect our
>idealized models to exactly correspond to reality.  Rather, the
>expectation is that the model fits well enough to be useful for
>theoretical analysis.

I disagree, in turn.  It is only insofar as a PC or whatever conforms
to the Turing model that it can be said to execute algorithms.  That
is what it *means* to be Turing equivalent:  the TM defines the
effective procedure.  When your PC departs from the TM spec in this
respect, it is *broken*.  TM *is* an idealization when it comes to
infinite tape, perhaps.  But insofar as performing syntactic
operations on symbols is concerned, as soon as your CPU ceases to be
TM fastidious, you'll need a trip to the shop.  That it fit the model
is critical for its practical utility; forget theoretical analysis.

--
Jer
"If you meet the buddha, kill the buddha."
Lin-chi (dead)



More information about the Neur-sci mailing list