The aim of the ABFAB project is to investigate attitudes to regional accents in Britain today, and the effects that accent bias may or may not have on access to the professions among speakers of different varieties of English in the UK. Combining methods from linguistics, psychology, and economics, the project will focus on the role of accent and other speech factors in people's ability to judge competence in hiring contexts, both members of the public and recruiters in law firms in London and Leeds.
Every person’s vocal tract is unique. Information about its shape is encoded in the speech signal and provides cues for speaker identification. Understanding this encoding is essential for forensic speaker comparison across different speech recordings, but currently there is no direct evidence of how vocal anatomy affects the speech signal. This project addresses the issue with three research questions: which anatomical features contribute most to systematic variation in speech; what is the range of variation of these features within the population; and to what extent are anatomically-derived differences in speech detectable by humans and automatic systems? These questions will be addressed with a novel combination of acoustic modelling and magnetic resonance image analysis. Results will impact forensic casework by highlighting discriminatory features, measuring their impact on listeners, and providing population data. Furthermore, understanding the links between anatomy and acoustics has impact for speech and language therapy and speech synthesis applications.
This project investigates linguistic variation and change in the dialects spoken in three urban centres of the North East of England: Newcastle upon Tyne, Sunderland, and Middlesbrough. As varieties of ‘North East English’, these dialects share many linguistic properties, but the people who speak them have distinctive local identities and their language differs grammatically in ways that have yet to be systematically quantified. Our analysis focuses on grammatical, lexical and discourse-pragmatic features in recordings of people from these three locales, which were made for the ESRC-funded project ‘The Use and Utility of Localised Speech Forms in Determining Identity’ (TUULS, Llamas et al. 2016-19). Using quantitative sociolinguistic methods, we examine the extent of dialectal variation in the area and track the spread of linguistic changes over time and across geographical space. By considering the socio-historical context of each community, we also explore why linguistic differences arose within a relatively small region.
Theoretical accounts of the strategies used by children to learn the structures of words and grammatical features of languages differ considerably, but our knowledge of what is possible is limited by the existing focus on a relatively small number of languages associated with industrialised nations. Here, we investigate grammatical features and structures that may be expressed in a variety of different ways. Examples of grammatical features include number (eg the distinction between singular and plural), or gender (eg distinguishing masculine and feminine in languages like French), features expressed within the shape of the word and associated items. Grammatical structure may be manifested in agreement across the separate words of a noun phrase. This project investigates the acquisition of inflectional morphology, ie grammatical features and structures as reflected in the word forms and associated agreement, in Gújjolaay Eegimaa, a language of the Atlantic family of the Niger Congo phylum spoken in Southern Senegal. This language has a gender system of the type traditionally known as a noun class system. Noun class systems with complex gender agreement are characteristic of the Niger-Congo languages.
This project explores the properties of the mass/count distinction across languages focusing specifically on languages that display the peculiar property of "plural mass nouns" (There were waterS on the floor). We employ experimental as well as theoretical techniques in order to establish the relevant syntactic and semantic properties of these languages and more generally understand the ways in which languages can vary in the ways they encode the distinction between objects and substances.
Research suggests that Deaf children and Deaf young people have much higher rates of emotional and behavioural problems than hearing children and young people, yet assessment tools in English are not accessible for Deaf children and their families. This project translated one such screening tool, the Strengths and Difficulties Questionnaire (SDQ), into British Sign Language (BSL) which can be completed by parents or teachers or by the child and scored by a clinician.
There are two main strands to this project: one sociophonetic, the other forensic. On the sociophonetic side, our interest is in the relationship between mobility and social class as a predictor of accent variation in urban north-east England. The emphasis here is on group-level linguistic differences. What does it mean to say that speech forms are localised versus supralocal? What makes north-eastern accents distinct, yet similar? On the forensic side, we take a Bayesian reasoning approach, focussing on individual speaker differences in long-term spectral features that we hypothesise to be essentially accent-independent. The goal is to test the viability of combining the database of recordings we collect for the TUULS project with an existing database, DyViS, which represents a highly dissimilar variety (viz., Southern Standard British English). Should interspeaker differences within the two respective accent groups be greater than differences between them, we can justify combining TUULS and DyViS into a single, much larger reference population. Significant benefits would be brought with respect to cost and time savings in speaker comparison work if we can demonstrate that the collection of case-specific reference data is generally unnecessary.
The aim of the project is to compare the performance of different methods for forensic voice (or speaker) comparison – from linguistics and phonetics, acoustics, and automatic speaker recognition (ASR) – on the same set of recordings. We will explore the performance of the methods to assess their relative strengths, the consistency of their results and error patterns, and thus the potential for different methods to be integrated into a single framework. The ultimate aim is to improve methods in forensic voice comparison, taking a major step towards the development of a methodology that is more transparent, validated, and replicable. This outcome will benefit academics and forensic practitioners, the public, judicial systems, and investigative/security agencies.
Our team specialises in the study of phonological development. Our research focuses on infants and toddlers, including observational studies of production and experimental studies of word recognition and word learning. We study the relationship between babble and early words, and the shaping of a child's knowledge of the sound system of the language they are learning. We also study how the speech of the adults around them affects infants' learning.