Several Occitan dialects spoken in Western Piedmont exhibit no 1sg subject clitic form (a situation widely attested in Northern Italian dialects), although interrogative clauses with a 1sg subject feature an enclitic particle ke, identical to the complementiser. Many attempts have been made to interpret this ke as a reflex of Lat. E(G)O ‘I’ or originating from the reanalysis of the analogical -k displayed by verbs such as dik ‘I say’. Rather, I claim that ke is what it seems, namely the complementiser, and entertain the hypothesis that in these varieties the complementiser ended up satisfying EPP-like features.
To support this analysis, I examine data from a wider area including Provençal, Gascon and Ibero-Romance dialects in which the complementiser is used as a proclitic particle in assertive clauses (Etxepare 2010 a.o.). On the theoretical side, I submit the hypothesis that the peculiar behaviour of the complementiser ke in that area has to do with the checking of a ‘speaker’ feature in the left periphery of the clause.
©2014 by Walter de Gruyter Berlin/Boston