Wackernagel in the Language of the Rigveda. A Reassessment

Abstract In this paper I discuss several models of Wackernagel cliticization in the language of the Rigveda. Following a short overview of Rigvedic syntax I argue for a discourse functional slot in the left periphery. I show that Hale (1987) and (1996) can be falsified and that Krisch (1990) and Hock (1996) suffer from major drawbacks. In my own account I argue for a prosodic approach, in which WL2 clitics are hosted by the first phonological word in a sentence and WL1 clitics by the first phonological phrase. This phrase corresponds to the left periphery in the syntactic structure, which is obligatory in the language of the Rigveda. Using WL1 clitics as a diagnostic tool for determining syntactic structure, I show that Wh-subjects belong to the left periphery. Concerning the syntactic status of infinitive phrases, I demonstrate that rationale clauses are CPs, whereas all other types of embedded infinitive phrases are not.