Phonetics/phonology research cluster workshop on speech rhythm

Add to your list(s) Download to your calendar using vCal

• V. Dellwo, B. Post, C. Graham, A. Li, E. Schmidt
• Friday 25 March 2011, 09:30-13:00
• Wine Room, King’s College.

If you have a question about this talk, please contact Marco Aldo Piccolino Boniforti.

9:30–11:00	Student presentations – Calbert Graham, Aike Li, Elaine Schmidt (University of Cambridge)
11:00–11:30	Coffee break
11:30–12:30	Speech rhythm perception – Prof Volker Dellwo (Universität Zürich)
12:30–13:00	A cross-linguistic study of rhythmic development – Dr Brechtje Post (University of Cambridge)

Speech rhythm perception (V. Dellwo)

Speech rhythm is predominantly a perceptual phenomenon but yet the mechanisms by which we perceive the rhythm of a speech signal are still widely unknown. In the present talk I will report on a study that compared which durational information is more salient in terms of the auditory impression of the rhythm of a sentence: (a) the durations of consonantal and vocalic intervals or (b) the durations of syllables. For a listening experiment, durational cues of seven German sentences produced by ten German speakers (five male, five female) were extracted under two different conditions. In condition I, consonantal intervals were turned into /s/-sounds and vocalic intervals into /a/-sounds (cv delexicalization). In condition II, entire syllables were turned into /sa/ syllables with a generic durational /s/:/a/ ratio of 5.5 : 4.5 (sy-delexicalization). In a rating task 21 native listeners of German judged on a seven point scale for 140 delexicalized sentences (7 sentences x 10 speakers x 2 delexicalization conditions) how well they matched the auditory rhythm of their originals. Results revealed that sy-delexicalized stimuli are typically rated significantly higher than the cv-delexicalization. It was found that when syllabic cues, in particular the number of syllables in a sentence, were not obtainable from the cv-delexicalization, listener ratings were poor. The impact of the results on how listeners may perceive rhythmic differences between languages using cv-delexicalized speech are discussed.

A cross-linguistic study of rhythmic development (B. Post)

The characteristic rhythm of a language – traditionally referred to in terms of stress- and syllable-timing – has been claimed to emerge from various phonological properties like syllable structure, vowel reduction, stress placement, and phrasing. If so, learning to produce an appropriate rhythmic pattern as a child depends not only on sufficient motor control, but also on the acquisition of those properties. This implies that, while rhythmic development may start early, it potentially encompasses the entire period of phonological development until approximately age 9.

In this paper, we investigate to what extent the acquisition of rhythm is indeed language-dependent, and emerges gradually as language-specific systemic properties are acquired. We comparing English (stress-timed), Spanish (syllable-timed), and Catalan (mixed/intermediate rhythm class) 2-, 4-, and 6-year-olds, to test the hypothesis that rhythmic differences emerge in parallel with the acquisition of phonology. Hence, cross-linguistic rhythmic differences should already be apparent at age 2, but become stronger with age, reflecting phonological and prosodic differences. Also, rhythmic patterns remain untarget-like even at age 6, depending on the complexity of the systemic properties of the target language.

The results confirm that there are ambient effects of segmental and syllabic properties of the target language already at age 2, and that rhythm development is language-dependent. The results also show that incomplete acquisition of phonological systemic properties rather than lack of motor control plays a role in the non-target-like timings observed in younger children. However, syllabic and segmental properties cannot fully explain the findings. Factors such as stress placement and phrasing, which are some of the properties which are not yet fully acquired at 6, also play a role, supporting a more refined, multi-systemic model of rhythmic development.

This study is part of ‘APriL’, an international collaborative project with Lluïsa Astruc (Open University), Elinor Payne (University of Oxford), and Pilar Prieto and Maria del Mar Vanrell (both Universitat Pompeu Fabra), funded by the Generalitat de Catalunya (Ref. 2007 PBR 29 and Ref. 2009 PBR 00018 ; PI Prieto) and the British Academy (Ref. SG-51777; PI Post).

This talk is part of the Phonetics & Phonology Research Cluster series.

This talk is included in these lists:

• Phonetics & Phonology Research Cluster
• Wine Room, King’s College

Note that ex-directory lists are not shown.