ABSTRACT
There are two variables built into the model proposed in Chapter 7 that allow us to account for cross-linguistic variation. The first of these is intrinsic to Optimality Theory: while it is assumed that constraints are universal, the ranking of the constraints may be language-particular. This puts us in a position to derive a certain amount of crosslinguistic variation from a re-ranking of constraints. The second variable included in the present model concerns the feature specification of the pronouns: languages have different pronominal paradigms. In interaction with the various constraints determining the distribution of pronouns (e.g. economy constraints), this leads to different output conditions as well.
