Sociolinguistics Symposium 19
Freie Universität Berlin | August 21-24, 2012
Programme: accepted abstracts
Abstract ID: 306
Part of General Poster Session (Other abstracts in this session)
Prosodic phrasing perception in Rio de Janeiro dialect: spontaneous and reading speech
Authors: Serra, Carolina Ribeiro
Submitted by: Serra, Carolina Ribeiro (Universidade Federal do Rio de Janeiro, Brazil)
The main goals of this research, which focuses on prosodic phrasing, in Brazilian Portuguese, are: (i) to find a correlation between the prosodic constituents boundaries -- as described by the Prosodic Hierarchy Theory (Nespor & Vogel 1986) -- and the perception and production of spontaneous and reading speech breaks; (ii) to describe the phonetic/phonological characteristics and the syntactic ranking of perceived and non perceived edges; and (iii) to capture the most relevant cues to the perception of prosodic boundaries for each speech style (Blaauw 1994).
Our corpus includes 10 extracts of spontaneous (informal interviews) and reading speech (orthographic transcription of spontaneous speech). In the perception test, 11 referees heard the oral sample, without punctuation, and marked the perceived breaks. The five speakers and the 11 referees are graduate students from Rio de Janeiro, between 22 and 38 years old.
The results point out to the fact that prosodic breaks are mainly perceived at the intonational phrase boundary, regardless the speech style: spontaneous speech (91%) and reading speech (99%). In the last one, 64% of the intonational phrase boundaries, described by the Prosodic Hierarchy Theory, were perceived as breaks, but, in the first one, only 37% were perceived. The leading cue to a break perception is a pause (92%) and this fact explains, by itself, 87% of the perceptive results. The most usual nuclear contour in both styles is H+L* L% (the neutral declarative contour in Portuguese), but its frequency at perceived breaks draws a distinction between reading speech and spontaneous speech (67% and 30%, respectively). In spontaneous speech, contours like L+H* H% and L*+H H% are also produced (34%). In general, descendant nuclei in reading speech are predominant, as well as the edge tone L; in spontaneous speech, both the descendant and ascendant nuclei distribution and low or high boundaries are similar. According to the statistic test, the L edge was globally significant for perception.
The results show that perceived intonational phrases present a higher number of syllable and prosodic words (PW) than those which were not perceived, for all speakers and styles. According to the statistical analysis, size affects the prosodic phrasing and the break perception.
The statistical results point out to the fact that breaks are mainly perceived at the matrix phrase limit (reading speech: 59%; spontaneous speech: 61%), showing the endurance of the matrix phrase edge/intonational phrase boundary mapping. Spontaneous speech displays a more frequent variation, as regards the relation of predictive, perceived and produced phrasing.
We conclude that reading and spontaneous speech share the same prosodic grammar, performed by the same type of phonetic/phonological cues, no matter being more consistent in reading speech and more diffuse in spontaneous speech, which adds a higher difficulty in spontaneous speech, than in reading speech, in the systematic perception of prosodic boundaries.
BLAAUW, E. 1994. The contribution of prosodic boundary markers to the perceptual difference between read and spontaneous speech. Speech Communication, 14, p.359-375.
NESPOR, M. & VOGEL, I. 1986. Prosodic phonology. Dordrecht: Foris.