|Title:||The phonetics of Modern Greek Rhythm and Its Phonological Implications||Add Dissertation|
|Author:||Amalia Arvaniti||Update Dissertation|
|Email:||click here to access email|
|Institution:||University of Cambridge, Department of Linguistics|
|Linguistic Subfield(s):||Phonetics; Phonology;|
|Abstract:||Speech rhythm has been the subject both of phonetic and of phonological studies. However, neither exclusively phonetic nor exclusively phonological studies have achieved an adequate description of speech rhythm. The phonetic distinction into stress- and syllable-timed languages has not been supported by the results of acoustic studies. Despite the lack of empirical evidence for the two rhythmic categoris, metrical phonology has adopted the distinction and incorportated it into metrical representation. Moreover, the principles of metrical phonology, which was originally intended as a description of English rhythm, have been adopted uncritically for the rhythmic description of widely different languages, such as French, Polish and Greek.
In contrast, the present dissertation adopts an interdisciplinary approach to the study of rhythm. The metrical structure of Greek is investigated by means of perceptual experiments and acoustic analyses of natural speech. The main topics under study are the acoustic correlates of primary and rhythmic stress and the perceptual and acoustic nature of the two stresses in a host-and-clitic group with antepenultimate stress. The experimental results form the basis on which a phonological representation of Greek rhythm is elaborated.
The empirical data show that the acoustic correlates of primary stress in Greek are amplitude, duration and fundamental frequency. Concerning the stresses in a host-and- clitic group the following picture emerges: the enclitic (added) stress is indirectly shown to be more prominent than the original lexical stress of the host (secondary stress); the secondary stress, which has often been equated to rhythmic stress, is perceptually and acoustically distinct from rhythmic stress and identical to a lexical stress which is not the designated terminal element (i.e. the head) of the phonological phrase it belongs to. No convincing acoustic evidence for rhythmic stress was found.
The above results suggest that the rhythmic structure of Greek is based chiefly on stress and is flatter and more flexible than has previously been assumed, as lapses and clashes are tolerated to a considerable extent. The rhythmic features of Greek suggest first that stress is the main contributor to rhythm, irrespective of the rhythmic category a language is thought to belong to; second, that binary rhythmic patterns are not universal as phonologists have often assumed. Thus, it is proposed that the stress-/syllable-timing classification be replaced by an abstract phonological representation, which reflects the contribution of stress to speech rhythm. This representation should be adapted to the particular rhythmic features of each language. It is proposed that Greek rhythm is best represented by n-ry branching trees which have only 5 levels: the syllable, the phonological word, the phonological phrase, the intonational phrase and the phonological utterance.