Position of Polish Clitics: An HPSG Approach

The aim of this paper is to provide an HPSG analysis of linear positions of Polish pronominal clitics. The distribution of Polish clitics, though rather free, is more constrained than that of other NPs. Polish clitics either occur preverbally or must closely follow the verb. We treat all Polish pronominal clitics as syntactic items and propose an LP constraint to account for their distribution. This constraint uniformly accounts both for preverbal and postverbal positions of clitics. Moreover, it captures the adjacency contrast between preverbal and postverbal positions of clitics. We briefly compare our approach with two possible alternatives: an analysis based on the lexical weight, cf. (Abeillé and Godard 1998), and topological fields of (Kathol 1995). The former might turn out to be too general if lite items in Polish are not restricted to clitics only. The latter, on the other hand, seems too restrictive for a ‘free’ word order language like Polish. 1 Polish Pronominal Clitics as Syntactic Items Polish pronouns may occur in two forms: weak and strong. The two forms can be morphologically distinct, cf. (1), but most weak pronominals in Polish double as strong forms, e.g., nas ‘us’ or im ‘them’, cf. (Rappaport 1988), (Spencer 1991, ch.9). In order to avoid confusion, we exemplify the behaviour of clitics only with morphologically distinct forms throughout the paper. (1) a. Jan John go himweak / jego himstrong zobaczył. saw ‘John saw him.’ b. Che,tnie willingly mu himweak / jemu himstrong pomagamy. help-we ‘We help him willingly.’ Weak pronouns are clitics, i.e., they are prosodically deficient items. Unlike strong forms, they do not bear heavy stress and cannot be used in isolation. (2) a. Jan John zobaczył saw JEGO himstrong / *GO, himweak aand nie not Piotra. Peter 1I wish to thank Anne Abeillé, Adam Przepiórkowski, Marek Świdziński for discussion and suggestions, and two anonymous reviewers for useful comments. I benefited also from remarks from the audience at the CLIN98 Meeting, especially from Frank Richter. Needless to say, all remaining errors and omissions are my own.