From Tones to tunes: Effects of the f0 prenuclear region in the perception of Neapolitan statements and questions
Résumé
no abstract
According to some theories of intonational meaning (Pierrehumbert and Hirschberg, 1990), tune meaning is the result of the separate contribution of each of its sub-parts, i.e. pitch accents and edge tones. However, most research has mostly focussed on the nuclear configuration (composed of the nuclear accent, the phrase accent and the boundary tone), since it has been less or more implicitly regarded as the semantic bulk of intonation contours, while the contribution of the prenuclear region to tune meaning is still understudied. In Neapolitan Italian, the nuclear accent carries information about intonation modality. Specifically, differences in early (e.g., L+H*) vs. late (e.g., L*+H) alignment are used to differentiate yes/no questions intonation and narrow focus statements, both in production and perception. Furthermore, when a prenuclear (L)H* accent precedes the nuclear one, the two intonation modalities also differ in the shape and slope of f0 region following the prenuclear accent. In statements, the f0 rapidly falls from the H prenuclear peak to the region immediately after the end of the prosodic word, whereas in questions the f0 fall after the prenuclear peak is shallower and it has a convex shape. This difference has been previously accounted for by an edge tone marking the end of the Accentual Phrase (AP), and differently specified in statements (LAP) and questions (HAP) (Petrone and D'Imperio, 2008a). In this paper, we explore the hypothesis that such a difference in tonal specification would help Neapolitan listeners to recover the contrast between questions and statements even when information about the alignment of the nuclear accent is not available. To verify this, we carried out two pilot experiments, in which auditory speech stimuli were gated at different locations of the sentence. In Experiment I, listeners were asked to identify natural stimuli either as questions or statements. In Experiment II, they had to rate gated stimuli, manipulated as for tonal scaling of the AP, on five semantic scales. Our results showed that the prenuclear contour carries enough information in order to distinguish the two intonation modalities and that AP scaling manipulation significantly affects listeners' judgments. This challenges the idea that only the nuclear configuration is relevant for the questions-statements distinction, thus implying that tune meaning is the result of the interaction between the prenuclear and nuclear f0 contours.