Deaccenting in Spontaneous Speech in Barcelona Spanish

Abstract

Recent literature on Spanish intonation assumes that deaccenting occurs when a lexical item fails to cue stress via an F0 rise or some other pitch movement through its stressed syllable. Inspired by the findings and suggestions for future research by Face (2003), the present study fills in research gaps by examining seven potential influences on deaccenting, working with spontaneous speech, and addressing the understudied Barcelona dialect of Spanish. The analysis of 160-170 minutes of spontaneous speech data collected at the Universitat Autònoma de Barcelona reveals that the odds of deaccenting increase in words that are high frequency in Spanish, have fewer syllables, are verbs or adverbs, are uttered multiple times within a recent timeframe, or are in initial or medial positions of the phonological phrase. Finally, high frequency verbs and adverbs, as well as adverbs, nouns, and verbs with fewer syllables are all especially prone to deaccenting.

Intonation in Spanish Classroom-style Didactic Speech

“…Such words are seen as semantically light and thus articulated with less F0 movement by instructors. This explanation fits into the scheme of Rao (2009), who finds all three of the reasons given here to be relevant to deaccenting in spontaneous speech. 9 The deaccented items in normal speech are on average longer than those in didactic speech and less common across Spanish, but they do represent information that is either repeated or obvious based on the passage at hand.…”

“…9 The deaccented items in normal speech are on average longer than those in didactic speech and less common across Spanish, but they do represent information that is either repeated or obvious based on the passage at hand. Therefore, examining didactic speech, faster read speech, and the spontaneous speech from Rao (2009), leads to the emergence of a type of continuum of permissible words to deaccent based on speech style that may be based on hearer involved. In more animated read speech directed at L2 learners, shorter, common, or given words are subject to deaccenting, while faster read speech addressing native speakers allows for pitch reduction of longer words, provided that they are repeated.…”

“…In more animated read speech directed at L2 learners, shorter, common, or given words are subject to deaccenting, while faster read speech addressing native speakers allows for pitch reduction of longer words, provided that they are repeated. In spontaneous interviews, Rao (2009) finds all of these to be common, along with phrase position and grammatical category. We do find that the slower rate of didactic speech results in phrases with fewer stress words (on average 1-2 per phrase) than phrases in normal speech (3-4 words per phrase).…”

“…Regarding F0 movement, each stressed word in both didactic and normal speech was measured for rise and fall time in milliseconds (ms), peak height, alignment with or displacement from the stressed syllable, and secondary stress. An F0 excursion had to measure at least 7 hertz (Hz) in order to be considered a peak (based on O"Rourke, 2006; Rao, 2009;Willis, 2003). Once all measurements were taken, a series of paired t-tests statistically evaluated the differences between didactic and normal speech styles with regard to the phonetic variables in question.…”

The Necessity of Both Naturally-Occurring and Elicited Data in Spanish Intonational Phonology

“…This variation in accenting in naturally-occurring speech opens the door to research on deaccenting, and research on Spanish deaccenting has begun to take place (e.g. Ortega-Llebaria 2006, OrtegaLlebaria, Prieto & Vanrell 2007, Rao 2006, 2008, 2009. Continued work in this area will lead to a better understanding of what conditions deaccenting and how meaning is affected by the accented/deaccented distinction.…”