STRESS CLASH RESOLUTION IN ENGLISH AS L1 AND L2

Résumé

This paper aims to analyze which strategy speakers of English as L1 and Brazilian speakers of English as L2 use to solve stress clash, a phenomenon in which two syllables bearing primary stress are adjacent in different words forming a phonological phrase such as [thirteen men]. The representation of stress clash, as well as the operation that allows its undoing, is one of the justifications whereby Liberman & Prince (1977) propose the metrical grid. The clash depends on information about the metrical level in which it occurs. The simple phonetic adjacency is not enough to characterize a clash. This paper is based on the work of Pike (1945), Selkirk (1984), Major (1985) to show distinctions between language types of rhythm (stress/syllable-timed ones) as well as a non-categorical polarization for rhythm in languages (Barbosa, 2000, 2002) and Silva Jr, (2013), Fragozo (2017) when comparing native and Brazilian speakers of English for the choice of stress clash solving strategies. For the Methods, we have done acoustic analysis of what we extracted and normalized the vowel duration values and applied to a statistical analysis using ANOVA test to check the degree of variance between Brazilians and native speakers of English. Our results show that native and Brazilian speakers of English use distinct strategies to solve stress clash: the former using stress retraction, which undoes primary stress clashes by moving the first stress of the clash to the left and the latter using silent demibeat addition, inserting a short pause in between the clash environment.