Walking and talking

[Brian MacWhinney]

John, Clare, and Info-CHILDES,
  First let me reply to John Limber that he is right to question my
knee-jerk invocation of modularity for this issue.  As Phil Dale also noted
in a note to me, it is really the resource competition issue that is central
here and not the modularity issue.  Let me try to rephrase what I meant
here.  What I meant was that, if the infant or the developmental process can
control the skills of walking and talking to some degree separately, then
they must be at least partially modular.  They don't have to be totally
encapsulated and all of that, just weakly modular.  Then, if there is
resource competition and if one of these developmental processes gets all of
the developmental, attentional, and processing resources, the other would
presumably suffer.  This is what Hull and Hull were thinking when they
brought up the amusing example of talking interfering with bladder control.
(It seems as if the young child is somehow suffering stage fright.)
  Now, it is certainly also true that, as Phil Dale points out, a positive
relation between walking and talking (as opposed to no relation) would be
consistent with any theory, modular or not.  However, if there is a positive
relation that is promoted by mutual positive interactions, then you have to
have a theory of how those positive relations work.  For example, Tina
Bennett pointed out to me that one could expect a positive relation between
walking and talking if you view both as expressions of a growing
independence on the part of the child and the attendant separation from the
mother.  Various writers have made this point over the years and it seems
interesting and possibly testable.
  I'm not at all suggesting that the data support any of these positions,
just noting that there are lots of interesting possibilities.  I might also
suggest that the fact that no negative relation of the type noted by Hull
and Hull has been detected in recent work might well be due to our tendency
to rely on group data which might well mush together some kids that have a
negative relation with others that show a positive relation.
   I am sure that other readers of this list are better able to address
Clare's interesting question about cognitive-motor relations than I.
However, I might at least point to the prevalence of motor items on infant
tests such as the Bailey, along with the fact that the items that are most
predictive of later IQ are not the motor items, but rather the language
items.

--Brian MacWhinney