Chapter 3: Modularity

From: Harnad, Stevan (
Date: Sat May 10 1997 - 21:55:26 BST

Chapter 3: Modularity

This chapter says a lot of fashionable things about independent modules
in the brain, but I'm afraid it does not have much substance.

You all remember our discussion of tasks that were "cognitively
penetrable" (in which your knowledge changes your performance)
for example, the Monte Hall Problem; and tasks that are "cognitively
impenetrable", like the Mueller-Lyer Illusion.

Cognitively IMpenetrable systems in the brain are thought to be
independent of other systems; they do what they do on their own,
and what you think or know or learn does not make any difference
to them.

The senses (vision, hearing, touch, smell) certainly seem to be
independent modules. (Synesthesias, in which, say, sounds evoke
colours in your mind, or vice versa, are just couplings; they
don't actually play a role in perception besides just being there
for some synesthetes.)

Besides the independent senses, there is at least one other important
module, and it's one you learned more in Chapter 7: Grammar is
an independent module; you "know" Universal Grammar, in the sense that
you can use it from a very early age, but you know it only "implicitly,"
or "procedurally" (in the sense of Chapter 10. You know HOW to produce
and recognise the grammatically well-formed sentence of your language,
but you don't know how you do it; you don't know the rules.

This Chapter touches upon several of the things you learn from later
chapters: The "Poverty of the Stimulus" refers to the fact that the
child "learns" to speak without ever violating Universal Grammar (UG).
For a better explanation see:

The concept of "double dissociation" is introduced here. As you will
see in Chapter 10, this is an important effect in neuropsychology,
one that allows you to strengthen the case for the reality of theories
about brain function: If there are two functions (say, to be trivial,
seeing and hearing) and there are two brain injuries (say, occipital
and temporal lobe) so that if there is injury in one you get one
deficit and not the other, and vice versa, then that suggests that our
idea of the two independent functions (modules?) has a stronger basis
in the evidence. This is an example of the way neuropsychology can help
us explain the mind: It gives us ways of testing some theories, to see
whether we are separating functions correctly.

This is also the Chapter in which we encounter the Frame Problem. For
that, please see the other discussions in this year's skywriting.

Some of the child's "theory of mind" ideas are first mentioned
in this chapter, but you will only understand them after you
have read the later chapters on "metarepresentations," "speaker meaning"
(vs. "sentence meaning"), pragmatics and autism.

One thing you might try to do when you are revising is to
link to:

If you use the simple or advanced search engines, the Web
can become your personal Encyclopedia, bringing in, for
example, courses at other Universities around the world
that are describing, explaining, discussing the same topics
we are: There's nothing like 2-3 different angles on the
same topic to make everything make sense. And then there's
still always Skywriting to use if you still aren't sure.

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:52 GMT