Tonal alignment, scaling and slope in Italian question and statement tunes

Unlike in languages such as English and Standard Italian, the Italian spoken in Naples shares with other Southern varieties the use of a very similar rising-falling (LHL) tune for both yes/no questions and narrow focus statements (D'Imperio, 2001; Grice et al., to appear). However, it has been claimed that the temporal alignment of the accent peak is later in the pitch accent of yes/no questions (L*+H) than in that of statements (L+H*) and that this alignment difference is employed by native speakers to perceptually identify the two tunes (D’Imperio and House, 1997). This study acoustically tested the hypothesis that all three tonal targets of the rise-fall are timed and scaled differently in questions and statements. Moreover, slope differences for both rise and fall were also tested by employing logistic regression modeling. Two speakers of Neapolitan Italian produced utterances whose target words differed in question/statement modality, syllable structure and segmental environment. The results show that all three targets within the rise-fall are timed later in questions than in statements. By contrast, no systematic difference was found for the slope of the rise nor for the slope of the fall. The exact contribution of F0 height to signaling the contrast could not be determined, though. In fact, while one speaker marked the difference by producing higher peaks for statements, the other did not produce any difference.