A multi-methodological approach to the instrumental phonetic and perceptual study of rhotic sounds in varieties of English.

Rhotics ('r'-sounds) have intrigued phoneticians perhaps more than any other class of consonants due to their phonetic complexity, diversity of articulation and susceptibility to rapid and extensive changes over time. This research will focus on the phonetics and phonology of rhotic sounds using a range of phonetic instruments and experimental methods, each of which addresses a different aspect of these sounds. Some look at how they are made in speakers' vocal tracts, some at their acoustic structures during transmission, and others at how they are perceived and processed by the brain. By combining methods that complement each other, this research hopes to paint a more complete picture of the different r-sounds in English and their similarities and differences than has been available until now.
Rhotic sounds are an important group of sounds in the world's languages for several reasons. They are important for phonetic and phonological theory because, while it is widely felt that they form a distinct class of sounds, it is far from clear what they all have in common, i.e. what defines the class of rhotic sounds. The class is generally held to include such phonetically different sounds as trilled 'r' produced with two or three beats of the tongue-tip against the upper gum, the French 'r-grasseye' produced with the back of the tongue, and the modern urban British English 'w'-like sound associated in the media with celebrities such as Jonathan Ross. The class of rhotics is thus seen to cut across the established classifications of traditional phonetic theory which is based on how sounds are made rather than how they are heard. This research will try to ascertain the relationship between the production and perception of these sounds, and whether the defining feature of the class of rhotics might be in the way they are perceived rather than in the way they are produced.
Rhotic sounds are important for developmental and clinical reasons because, no matter what language is being learned, children often find them the most difficult sounds to master in their acquisition of pronunciation systems and development of pronunciation skills. People with speech disorders often find that 'r' is one of the sounds most affected by their medical condition. The kinds of sounds that children and speakers with speech and/or hearing disorders produce in place of the 'correct' r-sound can provide insight into how r-sounds are represented in their speech production systems. The results of this research may prove useful for planning speech therapy intervention with populations affected by developmental and/or acquired speech and/or hearing disorders.
Rhotic sounds are important sociolinguistically because different social groups within the same language often have their own distinct form of r-sound which can function as a group identity-marker. For example, in French and German a back /r/ (articulated with uvular constriction) is considered socially prestigious, while a tapped or trilled /r/ is stigmatised. In English English, both uvular and trilled /r/ are non-standard and may elicit negative perceptions of the speaker. In Scottish English, by contrast, tapped/trilled /r/ is the socially acceptable norm. However, we do not know what the articulatory and acoustic thresholds are that have to be crossed before listeners perceive rhoticity, or how much variation there might be in these thresholds in different varieties of English. These questions are important for our understanding of the social significance of rhotics and have implications for forensic work in terms of speaker profiling and the evaluation of earwitness evidence where the rhoticity or non-rhoticity of a suspect's speech is at issue.

Outside of the academic community the outcomes associated with Group 1's pilot study have the potential to impact on the knowledge base and professional practices in Speech and Language Therapy (SLT) and Clinical Audiology (CA). Janet Addison, the SLT Manager for NHS Leeds Community Health Care, and Chris Monaghan-Doyle the Principal Clinical Scientist (Audiology) for Leeds NHS Trust, have both expressed an interest in their services being involved as beneficiaries. Health professionals from these services will be invited to the meetings and workshops to learn more about what instrumental and experimental phonetic research can offer them, and for them to let the academic phoneticians know what their clinical interests and priorities are. Dr Nick Thyer, who lectures in Audiology at the University of Leeds, will be a member of the proposed Research Network. Other members of the proposed Research Network (underlined) have already been engaged in clinical and developmental research concerning the phonetics of rhotics, e.g. Heselwood & Howard's (2002) study of rhotic production by speech-impaired speakers, Coleman, Knight & Ros's (2010) study of rhotic production by hearing-impaired speakers, and Knight, Dalcher & Jones' (2007) study of the acquisition of rhotic sounds by children in their first language. If we can adapt the battery of methods as outlined above in the Academic Beneficiaries section so it can be applied to acquisitional and developmental data, and to clinical data involving speech- and/or hearing-impaired populations, we will have achieved something of lasting benefit to Speech and Language Therapy and Clinical Audiology as well as to the academic study of speech learning and speech and hearing pathologies.

The outcomes associated with Group 2's pilot study have the potential to impact on the knowledge base and professional practice of forensic phonetics by bringing greater phonetic precision to an understanding of how rhotics vary across accents and varieties of English so that speaker profiling can be more accurate, and by informing evaluation of the reliability of earwitness evidence where rhoticity is an issue in speaker identification. Two members of the proposed Network, Dr Paul Foulkes and Dr Dominic Watt, are heavily involved in forensic phonetics both in teaching on the MSc in Forensic Speech Science at the University of York, and as consultants on phonetics and regional varieties of English to JP French Associates, a leading UK commercial organisation in forensic linguistics. Professor Peter French, its Director, is keen to benefit from this work and to attend, or send members of his team to attend, events organised as part of the Network. It should throw light on the reliability of ear-witness evidence involving rhotic accents (see e.g. French & Harrison 2006).

References
Coleman, Knight & Ros (2010) Acoustic characteristics of /r/ produced by people with hearing impairments.
British Association of Academic Phoneticians Colloquium, March.29-31.
French & Harrison (2006) Investigative and evidential applications of forensic speech science. In Heaton-
Armstrong, Shepherd, Gudjonsson & Wolchover (eds.) Witness Testimony: Psychological, Investigative and
Evidential Perspectives. Oxford University Press.
Heselwood & Howard (2002) The realisation of English liquids in impaired speech: A perceptual and
instrumental study. In Windsor, Kelly & Hewlett (eds) Investigations in Clinical Linguistics & Phonetics.
Mahwah: Lawrence Erlbaum. Pp.225-241.
Knight, Dalcher & Jones (2007) A real-time study of rhotic acquisition in British English. Proceedings of the
XVIth International Congress of Phonetic Sciences.