Intonational change studies in the literature only amount to contact-induced change. A seek for universals may only proceed through extensive example drilling and comparative research, since triggers of distinct outcomes usually show...
moreIntonational change studies in the literature only amount to contact-induced change. A seek for universals may only proceed through extensive example drilling and comparative research, since triggers of distinct outcomes usually show phonetic (thus, language-specific) differences. The most natural candidate to be a universal, non language-specific, factor influencing change is pragmatics, and answers may come from intonational phonology, which analyses the correspondence between phonetic form and pragmatic function. However, it is not easy to find minimal pairs only differing in pragmatics (sentence-type) undergoing different evolutions in time. To complicate the picture, iconicity (Givón 1990), namely a tendency to one-to-one form-function correspondence, is highly disrespected in intonational phonology, especially in situations of bilingual or multiple repertory and systems show, on one side, different contours associated to the same pragmatic sentence type (intonational allophony, or redundancy, often a product of bilingualism) and on the other side, different pragmatic functions represented by the same contour (polysemy, ambiguity). As to the distribution of allophonic variation in multiple repertories, Chambers and Trudgill (1998) distinguish two kinds of systems: whilst mixing systems have 2 well distinct phonetic forms in the repertory, each one associated to a language of origin, fudging systems show intermediate phonetic patterns with characteristics of both original patterns. Accommodation Theory (Giles et al. 1991) is based on the claim that long-term convergence of patterns is the result of extensive short-term accommodation of speakers, and Trudgill (2014) suggested that long-lasting change is gradual and happens " below the level of conscious awareness ". In a study on the evolution of the intonation of two groups of bilingual speakers in Majorca, Romera and Elordieta (2013) observe that the contour used for statements and that of questions evolved in two different ways, in that the two groups' original patterns would phonetically converge to an intermediate form and their shape gradually approach one another only in the case of questions. Dealing with different allophonic distributions, the authors did not ascribe the different behaviour to the pragmatics of questions vs statements but to the difference in saliency of the contours' shape, that is, a (contingent) characteristic of the phonetic form. Our work starts from the analysis of the phonological system of the variety of Italian spoken by two age groups from Pescara (eastern Italy). Studying the intonational contours associated with 35 sentence types we found, by the elder speakers, the same allophonic variety of pitch accents associated to 2 different sentence types, contrastive statements and information-seeking yes-no questions, consisting (figure 1) of 2 main pitch accents, A) a " pretonic peak " fall, transferred from the local language, Pescarese, with a high peak aligned at the end of the pretonic syllable, and B) a " tonic peak " late fall with a peak fully inside the tonic wowel, influenced from Rome Italian, together with a minority of intermediate patterns, being either overlapping patterns (a " double " pretonic+tonic peak (C1) pattern and its " extended peak " variant (C2), with the two peaks joined by a plateau) or a midway pattern (a single " midway peak " (D) usually aligned between the onset and the nucleus of the tonic syllable. When coming to the youngsters, however, the evolution of this common distribution proves to be sensitive to sentence type: when associated to contrastive focus, the main, well distinguished, pitch accents A) and B) keep prevailing in number, in a predominatly mixing system, and the number of intermediate patterns, being especially overlapping patterns, mildly increases, whereas in information-seeking questions the midway fudged type D) outnumbers A) and B). Dealing with the same range of phonetic shapes undergoing change across the two sentence types, and having different outcomes, this difference in change type can only depend on sentence-type (and not on the phonetic shape itself) and is compatible with the hypothesis that questions, being more interactive, listener-oriented sentence types, may trigger change " below the level of consciousness " and therefore are more prone, as a long-term effect, to gradual phonetic convergence of contours to midway fudged forms, more than contrastive statements, which are a more speaker-oriented, type, favoring maintainance.