(Bio)linguistic variation: have/ be alternations in the present perfect

The discussion in this article is based on a set of data which reflect microvariation in a closely connected set of languages (Romance varieties), as well as variation between more distant languages (Romance and Albanian). In Section 11.2, we present the basic evidence. In Section 11.3, we review the notions of person split, transitivity, and voice in terms of which we analyze the data in Section 11.4. Having discussed person-split systems in some detail elsewhere (Manzini and Savoia 2007), we concentrate on auxiliary selection according to transitivity/voice, arguing in particular that object-to-subject movement is insufficient to characterize it. The same conclusion has been reachedmore than once in the literature—by early work such as Burzio (1986), where be selection in Italian is governed by a disjunctive rule, only half of which refers to movement—and by recent work such as Reinhart and Siloni (2005), where the unaccusative treatment of voice (reflexive, etc.) is explicitly argued against. Contra Burzio (1986) we maintain that a unified account of auxiliary selection is possible; and contra Reinhart and Siloni (2005) we argue that Case cannot be responsible for it. In Section 11.5, we introduce some varieties where the various parameters cross-cut, and in particular selection according to transitivity/voice depends on the person split. In the introductory and concluding section (Sections 11.1, 10.6) we relate our empirical aims and results to general issues, insisting on variation, the lexicon, and the LF interface once again.