Dr Kirsty McDougall

My research interests in phonetics broadly fall into two main areas, speaker characteristics and the phonetics of varieties of English. I use experimental methods and approaches involving spontaneous speech, and am concerned with improving our understanding of the roles of individual, social and historical factors in theories of speech production and perception.

My research to date has primarily focussed on speaker characteristics and forensic phonetics. My Ph.D., entitled ‘The Role of Formant Dynamics in Determining Speaker Identity’, investigated ways of using formant frequency dynamics to characterise a speaker and developed techniques using polynomial equations. In the DyViS project, ‘Dynamic Variability in Speech: A Forensic Phonetic Study of British English’ (UK ESRC RES-000-23-1248), with Francis Nolan, Gea de Jong and Toby Hudson, we created the first large-scale forensically oriented speech database for English (100 male speakers of SSBE in several speaking styles), whose methodology has since inspired the development of a number of new forensic phonetic databases. As well as enabling me to expand my research on formant dynamics, the DyViS database facilitated our studies of fundamental frequency distribution in SSBE and sound change as a source of speaker-distinguishing information, and has since been used extensively by many other forensic phonetic researchers.

Also with a forensic focus, I have an ongoing programme of research investigating the potential of disfluency features to distinguish speakers. In collaboration with Martin Duckworth, I have developed and implemented TOFFA (Taxonomy of Fluency Features for Forensic Analysis), a methodology for the analysis of fluency behaviour in forensic casework. We are extending this work across different speaking styles, larger populations, more varieties of English, and other languages.

Another area of my interest within forensic phonetics is the use of earwitnesses to identify a voice in legal cases where no recording of the voice is available. I am consulted by the police as an expert witness regarding the construction of ‘voice parades’ for such cases, and have been researching the use of the statistical technique multi-dimensional scaling to improve the fairness of the procedure for selecting the foil voices for a voice parade. My research projects in earwitness evidence include the VoiceSim project, ‘Voice Similarity and the Effect of the Telephone: A Study of the Implications for Earwitness Evidence’ (UK ESRC RES-000-22-2582), my British Academy Postdoctoral Fellowship project, 'A Phonetic Theory of Voice Similarity', and most recently the IVIP project ‘Improving Voice Identification Procedures’ (UK ESRC ES/S015965/1), for which I am Principal Investigator. IVIP is an interdisciplinary project bringing together colleagues in linguistics, psychology, criminology and law. The project is investigating whether higher accuracy in earwitness identification can be achieved by improving voice parade methodology, and exploring ways to improve the interaction of the criminal justice system with the use of earwitness evidence.

Beyond yet still related to forensic phonetics, I have an ongoing interest in the phonetic realisation of varieties of English, motivated by questions relating to speaker characteristics, language variation and change, and sociophonetics. As a native speaker of Australian English, I am particularly interested in this variety and have published and presented a number of acoustic and sociophonetic studies of variation in the production of its consonants. I am currently engaged in research into plosive realisation in mainstream and Aboriginal varieties of Australian English with Debbie Loakes.

I welcome enquiries from potential M.Phil. and Ph.D. students.

 

PhD Supervisees

 

 

Name Thesis title Date Completed Linda Gerlach Automatic assessment of voice similarity and its implications for forensic applications. 2025 Daniel Lee     Chloe Patman     Alice Paver     Tallulah Buckley