These results and their interpretation are still controversial. However, again, the dispute need not be couched in the
oversimplified terms of a“grammar gene”thatdeterminesa“grammar box.”First, only some aspectsof SLIlanguage
are impaired, suggesting multiple loci of genetic control. This comports with the view I have been urging throughout
this chapter, that Universal Grammar should not be viewed as a monolithic unit. Second, I take it as given that a
particular stretch of genetic material normally controls the development of many different and apparently unrelated
aspects of the body. So we should not demand a“smoking gun,”a genetic defect that affects all and only language.
Like all the evidence presented here, the evidence fro mSLI on its own is only suggestive,not conclusive. But taken as
a whole, I think the body of evidence does begin to show an overwhelming pattern.
4.9.4 Language creation
What Ifind the most striking evidence for a prespecified skeleton for language are the situations in which children
create a language where there was none before. There are three cases.
- Deaf children whose parents do not know a signed language have been observed to improvise a gestural
communication system, sometimes called“home sign”(Goldin-Meadow and Mylander 1990). Though their
parents make use of the sign system as well, it is clearly being initiated by the children: at any stage of the
system's development, the children use a greater variety of signs with a greater complexity of combination
than the adults. To the extent the systems have been studied, they display certain rudiments of grammatical
structure: consistent word order and incipient morphological marking. Where can the consistent structuring
of these systems come from, if not from the child's f-expectations of what linguistic communication is
supposed to be like?
•“Pidgin”languages have developed many times in the course of history, when speakers of several mutually
incomprehensiblelanguages have been thrown together. Pidgins typically borrow vocabulary from the parent
languages, often in phonologically degraded form. They lack stable word order, and their grammatical
organization has a rudimentary“Me Tarzan, you Jane”flavor, lacking inflection and subordination.
The interesting thing is what happens next. Derek Bickerton (1981) documents in detail that children of a pidgin-
speaking communitydo notgrow up speaking the pidgin, but rather use the pidginas raw material for a grammatically
much richer system called a“creole.”In particular, he traces the transition fro mthe Hawaiian pidgin of i mported
workerstotheHawaiiancreoleoftheir children;speakers ofbothofthesewerestillaliveat thetimeofhisfieldworkin