114
7 Materialism: The Brain As Computer
and gave rise to the discipline of cognitive science, which is the culmination
of physicalism. According to cognitive science, thinking and problem-solving
are computations, and all mental activity consists of manipulating symbols in
the brain.
There is no hard evidence to support that idea, and it is plainly false: Indeed,
brains operate almost exclusively with Gestalt thoughts and images, whereas
computers are forever constrained to operate serially. The hallmark of mental
processes is that they are global and unifying, whereas any process that moves
forward step by step, in linear fashion, is quintessentially mechanical.
It is an interesting fact that everything computers do can be done just
as well (but more slowly) on the calculating machine invented by Charles
Babbage in the 1820s, which works with cogs and pinions. If you believe in
the computational theory of mind, you must likewise believe that a gadget of
cogs and gears can do everything a brain does.
The computational theory of mind has a clever and controversial way of
explaining how computers are able to carry out intelligent tasks which (when
we perform them) require the understanding of meaning in language. All
language, either natural or artificial, has two aspects: First is its syntax, which
consists of the rules of grammar and word order which make it possible to
communicate with language at all. Second is its semantics, which is the way
words and sentences correspond to things in the world, and therefore have
meaning. Computers operate mechanically on symbols, and are therefore able
to handle the syntax of language, but computers have no access to meaning.
It has been suggested that this difficulty can be overcome by tying the syntax
of computer languages to their semantics. In other words, word meanings can
be coded into the syntax—so semantics runs parallel to the syntax, although
there is no contact between the two. A computer pays attention only to the
shapes of the symbols (that is, the syntax), but it is claimed that it may operate
in a manner which respects semantic constraints. In a nutshell, the shapes of
symbols and the ways they connect with each other can be made to mimic
the way the objects which the symbols represent combine with each other. So
according to this claim, the way computers process symbols can be made to
mirror the way objects relate to the world. Thus, though they have no semantic
power, computers can be made to operate as if they understood meaning.
Cognitive scientists claim that it is due to this parallelism that intelligence
and meaning are mechanically possible. They also claim that this very mechanism explains how meaning is understood by people. According to this view,
meanings (and perhaps sensations too) don’t exist as such: They exist only
virtually—hitching a free ride on the computations carried out by the brain.
The claim is that brains operate as if sentience and meaning existed; when we
7 Materialism: The Brain As Computer
and gave rise to the discipline of cognitive science, which is the culmination
of physicalism. According to cognitive science, thinking and problem-solving
are computations, and all mental activity consists of manipulating symbols in
the brain.
There is no hard evidence to support that idea, and it is plainly false: Indeed,
brains operate almost exclusively with Gestalt thoughts and images, whereas
computers are forever constrained to operate serially. The hallmark of mental
processes is that they are global and unifying, whereas any process that moves
forward step by step, in linear fashion, is quintessentially mechanical.
It is an interesting fact that everything computers do can be done just
as well (but more slowly) on the calculating machine invented by Charles
Babbage in the 1820s, which works with cogs and pinions. If you believe in
the computational theory of mind, you must likewise believe that a gadget of
cogs and gears can do everything a brain does.
The computational theory of mind has a clever and controversial way of
explaining how computers are able to carry out intelligent tasks which (when
we perform them) require the understanding of meaning in language. All
language, either natural or artificial, has two aspects: First is its syntax, which
consists of the rules of grammar and word order which make it possible to
communicate with language at all. Second is its semantics, which is the way
words and sentences correspond to things in the world, and therefore have
meaning. Computers operate mechanically on symbols, and are therefore able
to handle the syntax of language, but computers have no access to meaning.
It has been suggested that this difficulty can be overcome by tying the syntax
of computer languages to their semantics. In other words, word meanings can
be coded into the syntax—so semantics runs parallel to the syntax, although
there is no contact between the two. A computer pays attention only to the
shapes of the symbols (that is, the syntax), but it is claimed that it may operate
in a manner which respects semantic constraints. In a nutshell, the shapes of
symbols and the ways they connect with each other can be made to mimic
the way the objects which the symbols represent combine with each other. So
according to this claim, the way computers process symbols can be made to
mirror the way objects relate to the world. Thus, though they have no semantic
power, computers can be made to operate as if they understood meaning.
Cognitive scientists claim that it is due to this parallelism that intelligence
and meaning are mechanically possible. They also claim that this very mechanism explains how meaning is understood by people. According to this view,
meanings (and perhaps sensations too) don’t exist as such: They exist only
virtually—hitching a free ride on the computations carried out by the brain.
The claim is that brains operate as if sentience and meaning existed; when we