Nominal copular sentences can be distinguished in canonical (1) subject+copula+predicative expression and inverse predicative expression-copula-subject (2) (Moro 1997): (1) [TP [SubjDP. The picture [PP of the wall]]i is [SmallClause_ti [DPthe cause [PP of the riot]]]]. (2) [TP [PredDP. The cause [PP of the riot]]j is [SmallClause [SubjDPthe picture [PP of the wall]]_tj. ]] In a Self-Paced Reading experiment combined with a sentence comprehension task on Italian nominal copular sentences (40 native Italian participants) we tested the sub-extraction of a wh- PP, either from the referential DP («of which wall ...» ), namely the subject, or from the predicative DP («of which riot»), the predicative nominal, in both canonical (1) and inverse copular sentences (2). Italian shows overt number agreement between the copula and the grammatical subject independently of the position of the subject. We expect sub-extraction to be acceptable from the predicate (according to L-marking; cf. Chomsky 1986, Cinque 1990) but not from the subject. Theoretically only sub-extraction from the canonical predicate position should meet the requirement of L-marking, but these contrasts were never tested before in a Self-Paced Reading. Participants show faster reading times in various regions and are more accurate in answering the comprehension question when the extraction takes place from a predicative DP («of which riot») in both canonical (1) and inverse (2) sentences, confirming that the structural asymmetry between referential subject and predicative DP has a central role in both the processing and the comprehension of nominal copular sentence as predicted by Moro (1997, 2006). Keywords: copular sentences, syntax, Wh extraction
