ABSTRACT

The aim of this article is to explore certain aspects of interrogative sentences in Italian—in particular, the “wh-in-situ” constructions, where at least two distinct wh-phrases co-occur in a clausal structure and only one appears to be overtly dislocated to the left periphery. Descriptively, Universal Grammar (UG) allows three distinct types of strategies for multiple wh-questions: languages like Bulgarian and Polish, where all wh-elements move to the front (1a); languages like Japanese and Chinese, where all wh-elements stay in situ (1b); languages like Italian and English, which combine these two opposite strategies and move only one wh-element to the front while leaving the other in situ (1c). 1

[CP wh-phrase C wh-phrase C [TP…t…t…]] (Bulgarian, Polish,…)

[CP [TP wh-phrase…wh-phrase]] (Chinese, Japanese,…)

[CP wh-phrase C [TP…t… wh-phrase]] (English, Italian,…)

The literature in the field is vast. Among others, see the seminal works by Richards (1993) and Bošković (1999); see also Boeckx and Grohmann (2003) and Cheng and Corver (2006) for papers on this issue and related ones, and Bayer (2005) for an updated critical review of the wh-in-situ literature. In this article, I will concentrate on the third class of languages only and limit my observations to Italian syntax. I will propose that this language is actually included in the first class, (1a)—that is to say, all wh-phrases move to the left periphery, although the final linear order is altered by a syntactic mechanism to be illustrated here. It is generally assumed that UG provides the possibility of checking the wh-features of multiple wh-phrases in two ways, thus splitting all languages into three categories: the wh-features of the phrase within the CP field are locally checked by a head endowed with wh-features in the left periphery; those of the phrase within the TP field are instead checked in situ. 2 The three-way partition comes from the fact that there are languages where the two strategies are mixed. The central proposal of this article is that the analysis given for Italian is not correct. Both wh-phrases undergo movement, and wh-feature checking is performed in the same portion of the left periphery. There is no wh-in-situ in this language; rather, the “wh-in-situ effect” (WISE) is due to a movement operation that rearranges the phrases, yielding the observed linear order.