Implicit in section 423.2.2's polemic against avant-garde textual theory, it might seem to a hasty reader, is an invitation to the practitioner of translation or of advanced (e.g. university-level) native language pedagogy or of non-native language pedagogy to accept instead the whole package of contemporary linguistic science. In order to emphasize that is not what is intended, we are devoting an entire section to the idea that contemporary technical linguistics is also irrelevant to the work of such practitioners of the language arts.
By technical linguistics we mean careful contemporary research as a whole, from the latest developments in phonology and morphology to formal syntax and semantics, as distinct from the more accessible forms of early generative work that had come to the fore in the sixties and made such an impact on neighbouring disciplines like philosophy and literature that their image of linguistics to this day is based on early generative grammar.
This is not the place to show that technical linguistics in general, and in particular its formal syntactic and semantic component which could seem relevant to workers in the translation field, must be seen as a success story in descriptive and explanatory progress compared to earlier work. Let us take it for granted that technical linguistics does represent a great advance. We shall even assume, with some idealization, that the principles discovered by technical work today provide the underpinnings of the effects described in the classical generative grammars of the sixties, and that thus there is a standard sort of scientific continuity in the enterprise, with earlier effects being derivable from later theories via descriptively motivated oversimplifications.
Our question in the present section is whether technical generative grammar as it is practiced today is of any relevance to the language arts in general and translation in particular. We will answer this question in the negative, and diagnose this as a problem, to be addressed in sections 423.2.4 and 423.2.5. While our remarks are applicable to all careful forms of generative syntax and semantics taken seriously in the leading universities today, we will take Chomsky's theory of Government and Binding (presented in Chomsky (1981) and associated commentarial writings) as representing the field as a whole (See the unit under 412.17 for some details of this theory). We do this for the sake of brevity and because a bigger exercise would yield similar results.
The problem is best seen with respect to a concrete case. Consider the category of questions. The GB (Government and Binding) form of technical syntax does not formally recognize the existence of questions, although GB writings do make a considerable amount of pretheoretical use of the term 'Question'. What GB does offer, is an account of certain pretheoretically delimited domains of data in terms of the serious theoretical concepts that belong to various modules or subtheories, such as the proper government module built around the Empty Category Principle [which says that nonpronominal empty categories must be properly governed] and the binding module built around the binding principles which include "principle C" [the principle that an R-expression must be A-free within the domain of the closest operator, if any, that binds the R-expression]. These modules, coupled with the thematic module imposing a one-to-one pairing on thematic roles and arguments and the bijection module imposing a similar pairing on operators and variables (thus forcing every operator, include every "question word" [a pretheoretical notion], to move to a non-A-position at LF if it has not done so by S-structure), end up constituting a network of options and impossibilities in which it is possible to retrieve and make sense of all the information about "questions" that would normally be provided by a sixties-style descriptive transformational grammar of the kind that translators and other consumers of applied linguistics tend to feel comfortable with. |
