Re: Chomsky vs. Skinner on Language

From: Stevan Harnad (
Date: Sat Mar 02 1996 - 19:44:44 GMT

> From: "Liz Lee" <>
> Date: Tue, 27 Feb 1996 09:43:21 GMT
> The
> reason Behaviorism doesn't have these answeres is that it does not
> consider 1. that there are innate structures involved in behaviour 2.
> that there is no reason to theorise or explain what is going on inside
> the brain when we learn or carry out any other type of behaviour.

1 is tight, but 2 is garbled: Yes, behaviourism has nothing to say about
innate capacities (and tries to explain grammatical capacity as shaped
by reinforcement, whereas it is actually inborn and could not be shaped
by reinforcement because the child never makes the errors that would
have needed to be corrected by feedback). But behaviourists, in
rejecting not only the unobservable evidence of introspection, but
unobservables in general, rueld out the possibility of theorising about
the unobservable structures and processes that CAUSE behaviour.

In the first instance, those structures and processes are theoretical,
and must be discovered the way any theory is discovered: By looking at
the observable data (behaviour) and trying to come up with causal
explanations: hypotheses, models.

The brain's "behaviour" IS observable, but it is still just data. By
looking at brain data you don't "see" the sturctures and processes that
cause behaviour; you still need theoretical models, which can then be
tested by behavioural data and perhaps even cross-tested against brain

> The field of Cognitive Science has grown in response to the question -
> What is it that allows us not to use the grammatical errors described
> by Chomsky? In fact what is going on inside our heads every time we
> make a decision, answer a question or remember the past? Until we can
> explain how we are able to do what we do and all this without even
> "thinking" about it, there are more questions than answeres.

Yes, Chomsky introduced the notion of "competence" -- in
contradistinction to "performance" (behaviour): We perform (behave)
grammatically because we have a certain CAPACITY, and it is that
capacity that must be explained. It turns out that that capacity is
based on a very complex set of rules, UG, and that set of rules cannot
have been learned by trial and error shaping of grammatical behaviour.
These rules are unconscious and inborn. We did not learn them, yet we
"know" them, though we are not conscious of them: we cannot introspect

This opens the door to many other unconscious rules and processes --
some innate, some learned -- of which we are not conscious, and that are
not directly observable as either behaviour or brain activity, yet they
are the real causes of our behaviour, the bases of our behavioural
capacity. THEY are what cognitive theory must discover and test.

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:15 GMT