The Interaction of cues to phrasing and cues to lexical stress in Greek

Abstract

Research on the alignment of F0 movements has shown that they align with the segmental string. This is the “segmental anchoring” hypothesis. In addition, research on the duration of segments at the edges of prosodic domains has shown that duration is a robust cue for the demarcation of prosodic levels, however, the exact operation of those cues under different conditions is still a matter of research. This paper addresses the question of how the alignment of F0 of prenuclear pitch accents and the duration of segments in Prosodic Words of Greek might interact, when produced under different prosodic boundary and lexical stress placements. We used Prosodic Words of the short “article + noun + clitic”, bearing an L*+H prenuclear pitch accent. From the results for the duration of the segments of the prosodic word we did not find support for mechanisms that have been found to operate on other languages, such as English and Dutch. The mechanisms examined were those of preboundary lengthening, polysyllabic shortening, accentual lengthening and articulatory strengthening. We found that the proclitic is shorter than the enclitic, which indicated a shortening of the proclitic. We also showed that the distance of the stress and pitch accent from the prosodic boundary causes a change on the duration of the segments of the noun of the prosodic word. That is, segments belonging to a word with stress further away from the boundary are longer than the ones with stress closer to the boundary. Regarding the alignment of F0 movements, we found that the Low tone aligns with the onset of the accented syllable, but is influenced by the existence of a word boundary. Surprisingly, we found that the H tone is not influenced by the existence of a prosodic boundary and aligns after the onset of the first postaccentual vowel, regardless of whether the clitic is a proclitic or an enclitic. Position of stress seemed to influence the alignment of the F0 movements as well, since when the pitch accent lay further away from the prosodic boundary there was more time for the rise from L to H to take place.