The syntax of Dutch embedded fragment answers: on the PF-theory of islands and the wh/sluicing correlation

Abstract

This paper presents new evidence in favour of Merchant’s (2004, 2008) PF-theory of islands, which states that island-sensitivity is due to the presence of PF-uninterpretable traces at PF. This new evidence is provided by two types of Dutch embedded fragment answers: whereas one type is island-sensitive, the other one is not. The former differs from the latter in that it involves an extra movement step, leaving an extra trace. Moreover, this paper argues that the wh/sluicing correlation (van Craenenbroeck and Liptak 2006, 2009) makes the correct predictions regarding the (non-)embeddability of fragment answers in Dutch and English. The wh/sluicing correlation states that there is a correlation between the type of wh-movement a language exhibits and the types of clausal ellipsis attested in that language. I show that it follows straightforwardly that, unlike in Dutch, embedded fragment answers are not attested in English.