The Classification of Languages and Dialects into Rhythm Classes and the Production of Non-native English Prosody: A Comparative Study

Abstract

The world's languages are generally classified into two categories on the basis of their rhythmic organization. The most well-known classification, proposed by Pike (1945), is the stress-timed language classification, in which stresses tend to recur at regular time intervals. English is one example of such a language category. Syllable-timed languages, where all syllables - stressed or unstressed - are produced at equal duration, are exemplified by French. Later, a third group was equally proposed: mora-timed languages, exemplified by Japanese and Tamil. A mora is a unit of timing. Each mora takes the same length of time to say (Ladefoged, 1975). However, the main problem in categorizing these types of languages was trying to find experimental evidence or acoustic correlates of language rhythm in the speech signal. Since the late 1960s, linguists have been trying to provide empirical support for the isochrony/regularity theory with varying degrees of success. The most that can be said for its legitimization is that a tendency toward isochrony is best felt at a perceptual level. In this research, it was hypothesized that a perceptual study could generate more convincing results relevant to rhythm typology. This approach has been adopted by using discrimination experiments based on spectral inversion in order to decide the rhythmic status of many languages as well as many Arabic and English dialects. Based upon the results of this study, it made sense for listeners to use rhythm as an essential cue to classify languages and dialects according to their rhythmic properties. Listeners were also able to perceive rhythmic differences between languages and dialects belonging to the same rhythm class. This implies the possibility of subclasses of broad rhythmic categories (Ghazali et al., 2002; Hamdi et al., 2005; Ben Abda, 2004). When non-native speakers of English produced read and spontaneous speech, they tended to use some of their first language prosodic features in the production of English suprasegmentals, which resulted in differences between native and non-native English in terms of rhythm and stress.