The view that cognition is computation needs little introduction. Propelled by the writings of innumerable thinkers (e.g., [19]; [2]; [16]; [36]; [37]; [28]; [18]; [20]; [23]; [14]; [2e]; [35]; [17] -- and this touches but the tip of a mammoth iceberg of relevant writing), this view has reached every corner of, and indeed energizes the bulk of, contemporary Artificial Intelligence (AI) and Cognitive Science (Cog Sci). The view has also touched nearly every major college and university in the world; even the popular media have, on a global scale, preached the computational conception of mind. Of course, this conception is as protean as it is pandemic; the cognition-is-computation slogan competes for equal time with a number of others. For example, for starters we have
There are differences, and in some cases significant differences, between these sorts of locutions. (We discuss some of these differences below.) But surely there is a great and undeniable (though confessedly vague) commonality in the works -- a commonality captured, for example, by Haugeland:
What are minds? What is thinking? What sets people apart, in all the known universe? Such questions have tantalized philosophers for millennia, but... scant progress could be claimed... until recently. For the current generation has seen a sudden and brilliant flowering in the philosophy/science of the mind; by now not only psychology but also a host of related disciplines are in the throes of a great intellectual revolution. And the epitome of the entire drama is Artificial Intelligence, the exciting new effort to make computers think. The fundamental goal of this research is not merely to mimic intelligence or produce some clever fake. Not at all. AI wants only the genuine article: machines with minds, in the full and literal sense. This is not science fiction, but real science, based on a theoretical conception as deep as it is daring: namely, we are, at root, computers ourselves ([18], p. 2).
This conveys the core spirit of ``Strong" AI, which wavers not a bit in the face of questions about whether sensors and effectors are necessary, or where in the Chomsky Hierarchy from finite state automata to Turing Machines people fall. Nonetheless, it will facilitate matters if we have a rather more focussed version of the doctrine on the table. Accordingly, we will say, following [16] closely, that the cognition-is-computation view -- sometimes called `Computationalism,' sometimes `Strong AI,' etc. -- amounts to the following: People (or minds, or brains) are computers. Computers, in turn, are essentially Turing Machines (or other equivalent automata). Hence, the boundaries of computability define the boundaries of cognition. It's easy enough to render this view wholly declarative, and hence quite unmistakable, using the predicate calculus, once the relevant predicates are defined by Mx iff x is a Turing Machine and Px iff x is a person. For simplicity, let's say:
Proposition 1.
.
Later on we'll propose a more
``fine-grained" version of
Computationalism, one which makes explicit reference to both cognition
and computation.
But before moving toward this version,
one remark. There have of course been prior attacks on
Computationalism (e.g., John Searle's [35] infamous
Chinese Room Argument). To our knowledge, however, none of these
attacks, with the possible exception of [26] and Chapter IX
of [2e], have attempted to show that the falsity of this
doctrine follows from the very foundations upon which it's
built. In the case of Searle's Chinese Room Argument, everything
hinges on an ingenious but undeniably fanciful thought-experiment. The
Argument From Irreversibility is based, for the most part, on provable
facets (e.g., reversibility) of computation which appear
by the lights of both common-sense and elementary physics to be at odds
with what we can come to know about cognition.