i Abstract English and Japanese Questions Hisashi Morita The Centre for Comparative Philology and General Linguistics Doctor of Philosophy The University of Oxford Hilary Term, 2002 This thesis aims to show that there are certain universal phenomena observed among interrogative constructions of natural languages and to present consistent and simple analysis of the phenomena by examining the syntax and semantics of these constructions. The main claims are the following. First, I will argue that both overt (or SS) and covert (or LF) movement of wh-phrases are constrained by the same sets of syntactic rules. In-situ wh-phrases, which are thought to move covertly, have not been seen as subject to syntactic constraints such as the subjacency condition. On the basis of this, Huang (1981/82), among others, claims that covert movement is not subject to such syntactic constraints. I will show that this is not the case by presenting evidence for a pied-piping mechanism, which enables in-situ wh-phrases to escape violations of the subjacency condition. In this way, I will provide support for the first claim that overt and covert movement are equally conditioned. The second claim is that, contra Pesetsky’ s (2000) claim, the behaviour of in-situ wh-phrases does not differ from language to language. I will argue that non-D-linked in-situ wh-phrases adopt covert movement whereas D-linked in-situ wh-phrases can optionally undergo Agree. The third major claim in this thesis is that there is always one overt wh-movement per interrogative clause. This is the case even in the case of multiple-wh questions. In order to provide sufficient support for this claim, it is necessary to provide an explanation for why every wh-phrase must move in Bulgarian and other similar languages with multiple wh-questions. Following Bošković (1998), I will argue that all overt movement except wh-movement is caused by phonologically relevant focus features in the language. I will also show that the WH feature of a wh-phrase is one type of contrastive-focus phenomenon and that this phenomenon requires two distinct features: semantically relevant and phonologically relevant features. It follows that multiple overt movement of wh-phrases is necessary for languages which require movement of a phonologically focused phrase (e.g. Bulgarian) whereas a single overt movement of a wh-phrase is observed in languages without movement of focused phrases (e.g. German).