Language use and Chomsky's methodological claims

In New Horizons, Chomsky points out his suspicion that the property of language called "displacement property" (i.e. when syntactic units -- phrases, lexical items -- are interpreted in a different syntactic position in the expression) is forced by legibility conditions; that is, that this property exists due to interpretive requirements imposed on language by our systems of thought (since they have these properties, as indicated by studies of language use). These requirements of legibility conditions, by their turn, are imposed on language by the architecture of the mind/brain (Chomsky, 2000:12-13). This is to say that the "displacement property" is an interface property. And that properties of the architecture of the mind impose constrains over the architecture of language and relational systems, linking them pro usability benefits. Again, each expression is taken to be created by the computational mechanism but it needs to be presented in a "readable form" to the interface mechanisms (where expressions are to be interpreted at the semantic interface and phonetically realizable at the phonetic interface). 

Concerning language use, Chomsky is skeptical about our technical capacities to approach the mechanisms and processes that play key roles in this type of behavior. He does not deny that topic is crucial, but he directs the complexity of the topic to our current understanding of the biology of the mind (restrictions imposed by the brain) and its representational capacities. 

Accordingly, the type of questions about semantics he considers to be fruitful involve semantic (representational) features (2000:16). Given the lexical item 'book', humans can refer to the 'material factor of the lexical item' or to 'its abstract component' (e.g. the person -- or agent, as I prefer -- may intend to talk about a physical object, such as when saying "My book is over there", or about an abstract entity, when saying "My book is finally finished"; or both.). The point is that humans have cognitive resources for conceptualizing in these terms, and not in others, as far as we can understand and formulate the contents processed. That is one way -- fruitful, he would say -- of addressing such representational/biological resources of the mind/brain. And this is a challenging enterprise, according to the linguist's view. 

For Chomsky, the way we humans talk about books, for example, reflects the way we humans think. Thus, it is better for us to formulate semantic questions of a precise type. Instead of trying to describe "what thing" the expression refers to or pick out in the world or in some "belief space," we should turn ourselves to the question of how are the semantic features used when we think and talk using lexical items, for instance (see 2000:17). 

Since many factors seem to enter into play when we use language, we can start by approaching lower-level mechanisms and then target more complex ones. This is Chomsky's methodological advice. "Generative grammar seeks to discover the mechanisms that are used, thus contributing to the study of how they are used in the creative fashion of normal life" (2000:17). 

The study of the mechanisms that are connected to, or enter into play in, the generative processing was the central concern of Cartesian inquires and still is central for the understanding of language and mind. We can conclude this line of thought by pointing out two questions to think about when pursuing a research problem of this type. First, we can ask ourselves Is it treatable by a (current) naturalistic investigation?, if not, Is this related to any component that is treatable by a (current) naturalistic investigation?

For Chomsky, some relevant questions are simply not yet treatable, or are meaningless or are pseudo-questions of the type of how do things work?

Reference

CHOMSKY, N. 2000. New Horizons in the Study of Language and Mind, with a Forward by Neil Smith. Cambridge, England: Cambridge University Press.