Multiple and conjoined wh-questions in Najdi Arabic

Abstract

This thesis investigates the syntax of Conjoined Question Phrase Construction (such as when and where do you….?) in Najdi Arabic (NA). NA allows such questions under certain constraints, which are sensitive to the categorial status of the fronted conjoined wh-phrases, i.e. argumental wh-phrases vs. adjunct wh-phrases. With the first constraint, when the subject and the object are wh-phrases, the subject wh-phrase is what should be fronted, whereas the object wh-phrase remains in situ (as in who saw what?). The second constraint concerns cases when the multiple wh-phrases are an argumental wh-phrase and an adjunct wh-phrase. The argumental wh-phrase should not be a subject here but rather an object wh-phrase, which in turn should remain in situ, while the adjunct wh-phrase is fronted (as in when did John see whom?). The third constraint pertains to instances with fronted conjoined adjunct wh-phrases. Here the two wh-phrases should be fronted, separated by the coordinating conjunction wa ‘and’ (not any other coordination conjunction as in when and where did you see the man?). Using the main theoretical assumptions of the Minimalist Program (henceforth MP) (Chomsky, 1995 and subsequent work), Phase Theory (Chomsky, 1999, 2000, 2001 and subsequesnt work), Rizzi’s (1997) split CP-system and Nunes` (2001, 2004) Sideward Movement, this thesis provides a unified analysis of the three constraints mentioned above. As for the first constraint, the study argues that the head of CP is endowed with an EPP feature. C° attracts the first wh-phrase it c-commands to its Spec. This implies that argumental wh-phrases do not need to move to the left periphery driven by their own requirements. This accounts for why the second object wh-phrase remains in situ, which I assume moves at LF. As for the second constraint, the study argues that adjunct wh-phrases move to the left periphery because they have a strong [Q] feature. When the subject is a wh-phrase, it is the closest wh-word to the head of the CP and it will be attracted by the EPP feature on C °, something that leads to the ungrammaticality of the question. Even if C°’s need to have its Spec filled would be fulfilled by the subject wh-phrase, the question remains ungrammatical because the adjunct wh-phrase (having a strong [Q] feature) is banned (by the subject wh-phrase) from moving to CP. This accounts for why questions with an adjunct wh-phrase and an object wh-phrase (which has a iv low position in the derivation) are grammatical. As for the third constraint, the study argues that this is a consequence of two factors: the strong featural content of adjunct wh-phrases and CP not being recursive. In order to solve the apparent tension (of moving two wh-phrases into one structural position), NA devises what I label as pseudo-coordination. The two wh-phrases are conjoined under one XP, i.e. &P. The implementation of this combination and insertion is executed through sideward movement (à la Nunes 2001, 2004). This thesis also investigates questions starting with a wh-word leiʃ “why” followed directly by another wh-word. The discussion reveals that the wh-word leiʃ “why” in such context is not a wh-word but a discourse particle which reflects the speaker`s surprise at the question at hand.