the Paradox of Language Acquisition at every turn. We should be delighted if some aspects of this problem can be
reduced to the child doing some relativelysimple statistical analysis of the input; but on the other hand, we should not
assume that this approach scales up to a solution of the entire problem.
4.8 How Universal Grammar can be related to genetics
In order for Universal Grammar to be an innate cognitive specialization, it must be transmitted genetically, just like
anything else innate. But what does it mean for it to be transmitted genetically?
It is certain that the genes cannot directly code a set of functional principles. All they can do is code the synthesis of
proteins under particular environmental circumstances, which in turn guide certain unknown parameters of brain
growth. Edelman (1992), Elman et al. (1996), and Deacon (1997) provide fascinating discussions of how brain
architecture develops and differentiates. But I think it fair to say that the manner in which this process is guided by
genetics or anything else is pretty much a mystery at the moment—and this is only at the level of turning genetic
instructions into neural architecture. On top of this lies the mystery pointed out in Chapter 2—how neural
instantiation supports functional organization, especially at a level as complex as language. So at the moment I think
there is really no hope of understanding in any detail the wonderfully indirect mechanisms for genetic transmission of
Universal Grammar. Infifty years, perhaps....
Jeffrey Elman et al. (1996) and, following them, Terrence Deacon (1997) mount a series of important arguments
against a detailed innate language-learning capacity, based on this proble mof genetic trans mission. It is worth
addressing these arguments in some detail, as they provide a concise distillation of many arguments in opposition to
Universal Grammar over the years.
First, they argue that the only way to control behavior of an organis mis through adjust ment of synaptic weights. But
genes cannot adjust synaptic weights—they can code only general guidelines to brain growth. At the same time,
however, these authors explicitlysay theyare nativists, and theyare ready to grant complexinborn instincts to animals.
It is only UG that theyobjectto. But then, we might ask, howare animal instincts coded on thegenes? There is just as
much mystery here. As Chomsky says (1965: 206),“Every known species has highly specialized cognitive capacities”;
and all the evidence of the past thirty-five years has amplified this statement substantially.
I suggest that the proper questions to ask on this score are the following:
- How cananysort of animal behavior—spatial orientation and navigation, bird songs and nest-building,
primate call systems, sexual selection, child-rearing,