Update 21/9/98
Recordings
We have recorded the Newcastle data, and the data
from Dublin will be recorded next week.
Labelling
Kimberley Farrar is working on the Newcastle data. At the moment, she is
labelling the reading passage, and is looking at the relationship
between phonological category membership and F0 peak alignment in IP
initial pitch accents. Orla Lowry has labelled the Belfast reading
passage and a large section of the Belfast free conversation.
Bronwen Evans who has been working for the project as a volunteer
over the summer has started working on an undergraduate dissertation
based on the data from GSB which she has labelled. She is planning
to carry out a comparison of phonological labels across the five
speaking styles. Margit Aufterbeck from the University of Heidelberg
has joined us as a volunteer for three weeks. She is labelling data
from Cambridge.
Publications
We have submitted a paper to Journal of Phonetics:
Grabe, E., Nolan, F., and Farrar, K.J. (submitted). Pitch accent realisation in four varieties of British English.
Our findings show that different varieties of British English can share intonological specifications, but differ in the way these specifications are realised in F0: when voiced material is scarce, pitch accent are compressed in General Southern British English and Newcastle English, but truncated in Belfast and Leeds English. Secondly, the data show the the reverse is also possible. Different varieties can be characterised by the same phonetic realisation effect, but apply this effect to rather different pitch accents.
A further paper will be published in the proceedings of ICSLP '89 (Sydney, Australia): Grabe, E., Nolan, F., and Farrar, K.J. (1998). IViE - A Comparative Transcription System for Intonational Variation in English.To appear in Proceedings of ICSLP1998, Sydney, Australia.
In this paper, we offer an alternative to ToBI, the de facto standard for the transcription of English intonation. IViE differs from ToBI in that (1) it offers an explicit separation of phonetic and phonological levels of intonation, and (2) it unpacks the phonological transcription of intonation into a number of levels, and thereby facilitates prosodic labelling. Labellers begin by determining the location of rhythmically prominent syllables and boundaries. Then they decide which of the rhythmically prominent syllables are associated with pitch movement. Finally, they draw up a phonological account of the data.