464
PARTISAN REVIEW
g ramma r o f a language specifi es one such cognitive structure. Others
include physical science, commo n sense, and the principles tha t enab le
a person
to
recognize faces. Although peopl e act as th ey do a t leas t
partly because of the cognitive structures they possess, Chomsky
beli eves th ere is no real prospect fo r a theory tha t wou ld account fo r
thi s, whi ch he thinks will probabl y remain a mys tery fo rever. But
psychologists can try
to
characterize various cognitive structures, just
as linguists have characterized the grammars of various languages, and
they can study the ways in which con ceptua l structures are acquired.
Chomsky argues tha t different conceptu a l structures are acq uired
in different ways. Specific aspects of a g iven concep tua l structure may
be inn ate and unl earn ed. For exampl e, much of the visua l sys tem is
appa rentl y "w ired in " ahead of time. Spatial perception is no t lea rned
but inn a te. We mi ght expec t tha t our recogniti on o f faces simil arl y
depends upon an innate schema of faces. And Chomsky expl a in s a t
leng th why there is reason to suppose th at lang uage lea rnin g invo lves
fillin g in a n inna te gramma ti ca l schema tism . Language learning is no t
simpl y the application of genera l intelli gence
to
the lea rning of
language but invo lves a specifi c facult y of language.
In
Chomsky'S view, th ere is no interes ting genera l theo ry of
learning. Different cogniti ve structures are lea rned or acquired
III
different ways. So-called learning th eory in psychology has been
entirely concerned with the uninteres tin g margina l cases o f learning
which do not depend on specifi c schema ta but must proceed by tri a l
and erro r.
Chomsky goes on to sugges t tha t the development of science
depends on a specific science faculty, an innate schema fo r scientific
theori es. This makes science poss ibl e but it also means there are
impo rtant limits to science, sin ce th ere is no evoluti o nary reason for
the science faculty to be suited to the di scovery and understanding o f
every aspect of rea lity. We may, for exampl e, never be a bl e to develop a
signifi cant science tha t would account fo r human behav ior.
We are specifi call y "des igned" for some tas ks, such as lang uage
learning. Other tas ks are beyond o ur cognitive ca pacity-perhaps the
tas k of developing a theory that wou ld account for actua l human
behav ior is an exampl e. Pro bl ems a t the border of cognitive capacity,
like those involved in chess, "wi ll provide opportunity fo r intrig uing
intell ectual pl ay.. . . H ere we wou ld expect
to
find tha t the sli ght
differences between individua ls are magnified to striking divergence o f
a ttitude." Chomsky adds,
As crea ti ve minds approach the limits of cognitive capacity, no t onl y
will th e act of crea tion be limited to a talented few, but even th e
