Focus and wh in Jamaican Creole: Movement and exhaustiveness 1 Stephanie Durrleman & Ur Shlonsky Université de Genève 1 Introduction Wh & ex-situ focus constructions in Creoles have been analysed in terms of movement to the left periphery (Saramaccan: Aboh (2006), Jamaican: Durrleman (2008), Papiamentu: Kouwenberg & Lefebvre (2007)) and as clefts (Haitian: Lumsden & Lefebvre (1990) Manfredi (1993), Jamaican: Patrick (2007), Veenstra (2008)). Our goal is to reassert the cartographically-rooted left-peripheral account, as in Rizzi (1997; 2004a) and related work. We consider data from Jamaican Creole (JC) but we believe the analysis extends to other Creoles. Section 2 shows that JC focus and wh-questions involve internal rather than external merge. Section 3 explains why a cleft analysis cannot account for these properties. Section 4 summarizes and sharpens the arguments in favour of the left-peripheral approach, originally developed in Durrleman (2008), and discusses the driving force for movement and its interpretive correlates. Section 5 suggests that the formal difference between exhaustive and contrastive focus is at the root of some of their distributional properties. Section 6 is the conclusion. 2 Focus/wh in JC: External merge or internal merge in the left periphery? Several considerations militate in favour of the thesis that the focused constituent in JC is internally and not externally merged. 2 The grammaticality of (1) is naturally explained if the focused constituent containing the anaphor imself reconstructs into a position c-commanded by the subject. Reconstruction is a property of movement. (1) A fi imself Jan did tiif da mango de A Prep himself John Past steal Det mango there ‘It’s FOR HIMSELF that John stole that mango’ 1 Many thanks to our Jamaican Creole informants J. Farquarson, C. Forrester, M. Forbes, T. Tame, E. Miller and in particular, T. De Lisser. Thanks also to H. Devonish for the Guyanese Creole data, to M. Finney for the Krio data, to D. Cserzo, G. Puskas and E. Varga for the Hungarian data and to E. Biloa for the Tuki sentences. Comments by L. Haegeman and discussions with V. Bianchi and G. Bocci are gratefully acknowledged. Usual disclaimers apply. 2 See e.g., e.g. Koopman (1984; 2000), Manfredi (1993) and Aboh (2006).