Corrective questions exploit focalization to correct and replace a preceding question much like corrective statements exploit focalization to correct and replace a preceding statement. This paper examines Italian corrective questions in depth, presenting new experimental data supporting their grammatical status. It also shows that Italian corrective questions constitute root interrogative clauses, lack a terminal rise, and qualify as canonical questions when assessed with respect to Farkas’ criteria (Farkas 2022). Corrective questions are theoretically significant because they simultaneously contain a wh-phrase and a corrective focus in a root clause. Their grammaticality challenges the widely held hypothesis that wh-phrases and corrective foci must both be interpreted in a left-peripheral FocP projection that can only hosts one or the other. This hypothesis incorrectly predicts corrective questions to be impossible. I will argue, instead, that corrective foci can be interpreted in-situ, and that focalization à la Rooth (1992) and contrast à la Neeleman and Vermeulen (2012) naturally extend to corrective questions and straightforwardly account for their interpretation and corrective import. The paper also examines focus fronting, discussing the object / non-object divide that emerges from the experimental data and showing that non-object corrective foci may front even when a wh-phrase is present.
