Dominique Sportiche
UCLA
Wh-movement of embedded subjects in French triggers the surprising appearance of a second wh-word looking item, a special qui:
(1) Qui tu crois qui dort
who you think who is asleep
who do you think is asleep
Since Kayne 1976 motivating the idea that this special qui is a form of the tensed C que, a substantial body of work has worried about how to formulate this exceptional que to qui rule and its link to the ECP (much like that deletion in English in comparable circumstances). The central motivation for the existence of this rule is based on Kayne's detailed analysis of the French relative clause complementizer system.
I will propose a reanalysis of the C system of French relative clause which:
(i) suggests that this second qui is a wh-element
(ii) decouples the que/qui rule from subject extraction
(iii) links the que/qui phenomenon to the fact that French has a double paradigm of pronouns, one of weak (or clitic) forms and one of strong forms.
These conclusions create a challenge regarding the analysis of (1) which now indeed displays two wh-elements.
After documenting a number of new properties of the construction in (1) (both in French and in Dutch) strongly reminiscent of properties found in partial wh-movement constructions (in German or Hindi etc..), I will suggest that (1) is not a case of embedded subject wh-extraction but rather involves extraction of the subject of a pseudo relative, which will help derive some of these new properties.
Reception to follow in 1413 Marie Mount Hall.