philip.rose@anu.edu.au
Username: philip.rose@anu.edu.auParticipant details
Research InterestsMy research is in two main areas: The Phonetics and Phonology of Tone, and Forensic Speaker Recognition. Since 1982 I have researched how the linguistic category of tone in varieties of Chinese, Thai and Tibetan varies cross-linguistically at the acoustic-phonetic and physiological levels. Examples of this research are:
The first demonstration that tonal F0 is produced in different ways (i.e. with different degrees of pulmonic involvement) in different tone languages. The development and refinement of the first normalisation techniques for the linguistic-phonetic comparison of tonal acoustics. The first demonstration of the existence of an exotic aryepiglottic phonation type in a Chinese dialect. The first investigation of tonal aerodynamics. The first description and analysis of Tibetan Spelling Chant (the chant Tibetan Lamas use to spell words in the Buddhist Canon). Currently I am working on a two-volume book with Dr Zhu Xiaonong which describes acoustically the tones and tone sandhi in the Chinese Wu Dialects in the whole province of Zhejiang. These dialects show the most complicated tonal behaviour in the world.
Since 1994 I have been researching forensic speaker recognition, and doing FSR case-work. My book 'Forensic Speaker Identification' addressed the correct evaluation of forensic speech-science evidence within the proper Bayesian framework, and my book 'The Technical Comparison of Forensic Voice Samples' is the legal reference for FSR. I am currently researching the application of multivariate likelihood ratios in FSR; how to bring FSR in line with evidence evaluation in DNA profiling; and how to combine automatic and traditonal methods of FSR within the Bayesian framework (this last is a four year ARC-funded project).
|
