Building candidate sets

The main aim in formulating Universal Grammar is to identify the processes that play a role in constructing linguistics objects and to uncover the conditions that these processes are subject to. The linguist�s task has, however, not ended after these processes and the constraints on them have been discovered, since this still leaves the massive amount of variation between languages to be explained. The existence of this variation indicates that Universal Grammar actually defines a large set of alternative structures from which the individual languages make a selection. Further, it seems that even within a single language it is possible to choose between alternative structures to express certain meanings, where the choice does not depend on strictly syntactic considerations (e.g. information structure). The question how the actual structures within a certain language are selected plays an important role in the current linguistic debate. The minimalists program suggests that language-particular grammars are constructed such that only a limited number (perhaps only one) of the potential structures survive by placing additional restrictions (normally encoded in the lexicon) on the application of the syntactic processes. Optimality Theory suggests that the linguistic processes apply uniformly in all languages, so that all potential structures are potentially available in any language, and that the language selects a certain structure as the optimal one on the basis of considerations that need not be of a strictly syntactic nature. Since the two assumptions have wide ramifications for the overall structure of the grammar, comparing and evaluating proposals that take either of these assumptions, or a combination thereof, as their point of departure has a high priority.