Standard Southern British English Research Articles

Individual profiles amidst a multiethnolect: Acoustic hetereogeneity in London English

Multicultural London English (MLE; Kerswill and Torgersen 2008; Cheshire et al., 2011) arose in working class areas of London around 30 years ago through intensive, multiethnic social contact (Kerswill and Torgersen, 2021). The variety’s highly systematic phonological system has been seen as displacing traditional London vernacular varieties including Cockney, which have moved further East (Fox, 2015). MLE and Standard Southern British English are two strands among many, interwoven with features of the earlier London vernacular and features of non-MLE varieties (Sharma and Sankaran, 2011). Using preliminary data from a new project, Generations of London English, we present an acoustic analysis that shows both the stability and emerging maturity of MLE as well as strands of continuity from the past. The GOAT vowel shows a diverse range of variants indexing social class, ethnicity, age, and gender. By contrast, all Londoners participate in GOOSE-fronting, which is thus primarily an index of age, and women lead FOOT-fronting, a city-wide change showing less ethnic and social class sensitivity. A generalized use of labels such as MLE risks mischaracterizing the sociophonetic reality. We argue for analysis of individual phonetic profiles as unique intersections of historically layered features, diffusing differently through a heterogeneous population.

The search for linguistically coherent accents

Abstract Linguistic research refers to many related accents in Southeast England: Standard Southern British English (SSBE), Received Pronunciation (RP), Estuary English (EE), Cockney and Multicultural London English (MLE). However, there is inconsistency and imprecision in the demarcation of these accents based on linguistic and social factors. This paper delineates accents in Southeast England based on patterns of linguistic co-variation which we then relate to social predictors. We applied functional Principal Component Analysis to F1 and F2 measurements for diphthongs extracted from wordlist and passage productions for 193 young, south-eastern speakers. Principal Components were entered into a clustering analysis that identified patterns of linguistic co-occurrence. Three clusters emerge, broadly aligning with SSBE, MLE and EE for both linguistic and social factors. We illustrate the linguistic centre of gravity of the three diphthong systems for use as reference points in future research, and we discuss the need to make explicit how accents are defined.

Pausing and the ‘Othello Error’

The idea of detecting deception from speech is very attractive from a law enforcement perspective, yet research considering the possibility has yielded conflicting results, due to the practical difficulties in investigating the topic. Scientific research is yet to provide forensic linguistics with a reliable means of discerning lies from truths. The present study explores the relationship between truthfulness and pausing behaviour. Various aspects of the acoustics of pausing behaviour were investigated for Standard Southern British English in 30 mock police interviews from the DyViS database (Nolan et al. 2009). A novel distinction was made between prescribed and unprescribed lies, to delineate a potential source of differences in the unscripted content of speakers’ untruthful responses. Among pause duration measures, statistically significant differences were found across all three response types (truth, prescribed lie, unprescribed lie) for response latency, between truths and lies for initial filled pauses, and between unprescribed lies and the other response types for silent pauses. For pause frequency measures, only internal filled pauses showed a statistically significant difference: truths differed from both types of lies, but prescribed lies did not differ from unprescribed lies. Theories of cognitive effort and attempted control are drawn on in accounting for these findings.

The production of English monophthong vowels by Saudi L2 speakers

Abstract This study aims to describe the production of English monophthongs by nonnative Saudi speakers. It acoustically examined the production of English monophthongal vowels by Saudi second language (L2) speakers. Sixteen L2 participants produced twelve English monophthongs in carrier words. Comparable data were obtained in native Saudi Arabic and native Southern Standard British English to aid interpreting the L2 results. Formant frequencies and durations were measured for all vowels. Similar to Saudi Arabic, the English vowels produced by the Saudi L2 speakers occupied a smaller vowel space than the vowels produced by the native English speakers. There was overlap between many of the English vowel categories produced by the L2 speakers. However, the L2 speakers were found to employ similar durational characteristics to those employed by the native English speakers. L1 influence seems to play a major role in the L2 learners’ production of English vowels. The limited vowel space of the L2 speakers’ L1 appears to have limited their ability to produce the more acoustically peripheral English vowels.

Incorporating automatic speech recognition methods into the transcription of police-suspect interviews: factors affecting automatic performance

IntroductionIn England and Wales, transcripts of police-suspect interviews are often admitted as evidence in courts of law. Orthographic transcription is a time-consuming process and is usually carried out by untrained transcribers, resulting in records that contain summaries of large sections of the interview and paraphrased speech. The omission or inaccurate representation of important speech content could have serious consequences in a court of law. It is therefore clear that investigation into better solutions for police-interview transcription is required. This paper explores the possibility of incorporating automatic speech recognition (ASR) methods into the transcription process, with the goal of producing verbatim transcripts without sacrificing police time and money. We consider the potential viability of automatic transcripts as a “first” draft that would be manually corrected by police transcribers. The study additionally investigates the effects of audio quality, regional accent, and the ASR system used, as well as the types and magnitude of errors produced and their implications in the context of police-suspect interview transcripts.MethodsSpeech data was extracted from two forensically-relevant corpora, with speakers of two accents of British English: Standard Southern British English and West Yorkshire English (a non-standard regional variety). Both a high quality and degraded version of each file was transcribed using three commercially available ASR systems: Amazon, Google, and Rev.ResultsSystem performance varied depending on the ASR system and the audio quality, and while regional accent was not found to significantly predict word error rate, the distribution of errors varied substantially across the accents, with more potentially damaging errors produced for speakers of West Yorkshire English.DiscussionThe low word error rates and easily identifiable errors produced by Amazon suggest that the incorporation of ASR into the transcription of police-suspect interviews could be viable, though more work is required to investigate the effects of other contextual factors, such as multiple speakers and different types of background noise.

The perception–production link varies with stages of L2 development and vowel properties

Abstract This study investigates whether the perception–production link in phonological acquisition varies with stages of L2 development. It also examines whether the perception–production link for L2 vowels varies according to vowel properties, or whether the L2 vowels match or mismatch with L1 vowels. Korean learners of English in the UK were divided into more experienced and less experienced groups based on age of arrival and length of residence. The learners completed English vowel production, English vowel identification, and English–Korean vowel mapping tasks with English words (e.g., beat, bot). The intelligibility of their production and the accuracy of their vowel identification were assessed. Results show that only the more experienced learners’ perception and production were significantly correlated. Among L2 Standard Southern British English vowels similar to L1 Korean vowels (/i, ʌ, u/) only /i/ showed a significant correlation between perception and production, while among L2 vowels dissimilar to L1 vowels (/eɪ, ɘʊ, ɑ, ɒ/) only /ɑ/ showed a negative correlation, indicating that the correlation varied with vowel properties. The study contributes to the L2 phonological acquisition literature by exploring the perception–production link in terms of different stages of L2 development and L2 vowel properties.

Investigating the Effects of Speaker Variability on Arabic children’s Acquisition of English Vowels

This study investigated whether speaker variability in phonetic training benefits vowel learnability by Arabic learners of English. Perception training using High-Variability stimuli in laboratory studies has been shown to improve both the perception and production of Second Language sounds in adults and children and has become the dominant methodology for investigating issues in Second Language acquisition. Less consideration is given to production training, in which Second Language learners focus on the role of the articulators in producing second language sounds. This study aimed to assess the role of speaker variability by comparing the effect of using High-Variability and Low-Variability stimuli for production training in a classroom setting. Forty-six Arabic children aged 9-12 years were trained on 18 Standard Southern British English vowels in five training sessions over two weeks and were tested before and after training on their vowel production and category discrimination. The results indicate that Low-Variability stimuli may be more beneficial for children, however, High-Variability stimuli may alter some phonetic cues. Furthermore, the results suggest that production training may be used to improve the perception and production of Second Language sounds, but also to inform the design of Second Language pronunciation learning programmes and theories of Second Language acquisition.

Speaker-specificity in speech production: The contribution of source and filter

This study examines the extent to which speaker-specific information is encoded in different features of vocal output and the relationships between those features. A range of acoustic features, grouped as source (laryngeal voice quality measures and fundamental frequency) and filter features (formants and Mel-frequency cepstral coefficients; MFCCs), were extracted from the vocalic portion of the hesitation marker um for 90 male speakers of Standard Southern British English. Little overall correlation between the sets of features was observed, suggesting no strong interdependence between source and filter in our data. Although filter features were consistently better at discriminating between same- and different-speaker pairs compared with source features, combining source and filter has the potential of producing the lowest error rates and the strongest speaker discrimination scores. Taken together, results show that source and filter provide complementary speaker-specific information. However, the extent of the improvements in speaker discrimination performance when combining source and filter varied across speakers. We explore potential explanations for this finding and discuss the implications for source-filter theory, and for applied fields such as speaker recognition and forensic speech science.

Phonetic and phonological sound changes in an agent-based model

The focus of the study is an agent-based model (ABM) concerned with simulating phonological stability and change using real speech data from a population of speakers. At the core of the model was a flexible and agent-specific association between acoustic exemplars and phonological categories. This was achieved by means of two general-purpose unsupervised machine learning algorithms: the first grouped exemplars into acoustic clusters, the second identified sets of clusters which largely contain exemplars of the same word classes. The model was tested first on data from Standard Southern British English, where a shift of /u/ to the front of the vowel space was expected not to cause any phonological re-categorisation. The simulation indeed showed phonological stability despite the phonetic change. Using the same settings, the ABM was then applied to data from New Zealand English in which a merger of /eə, ɪə/ diphthongs has taken place in the last 50 years. Compatibly with that, the simulation showed a shift from /eə/ towards /ɪə/ along with the neutralisation of the diphthong contrast. The results are discussed with respect to the conditions necessary to model a merger, taking into account both findings on marginal phonological relationships and principles from exemplar theory.

Investigating the Efficacy of Using Online Resources for Production Training in Learning Non-Native Vowel Contrasts

The current study investigates the benefits of virtual phonetic training on second language (L2) vowel production and examines whether production training improves perception. While the benefits of production instructions were investigated with instructions that were delivered by either a virtual or a face-to face teacher, a few studies investigated the benefits of online resources for second language (L2) learning with tutor audio instructions and feedback. Fifty-five Arabic speakers (21 served as a control group) took part in 3 one-to-one training sessions and were trained on producing seven Standard Southern British English (SSBE) vowels (/ ɪ, ɛ, ʌ, ɑː, ɔː, ɒ, ʊ/) that have been found to be difficult for Arabic learners of English. Improvements were assessed by their vowel identification, category discrimination accuracy and their vowel intelligibility. The results showed some improvements in vowel intelligibility and that vowel learning paved the way for improvement in perception, relative to the control group. The results indicate that using web-based tools along with audio instructions are beneficial for L2 learners and confirmed previous findings about the relationship between improvement in speech perception and production.

Native Danish listeners’ evaluation of English accents

This paper examines native Danish listeners' attitudes towards five native English accents and some of the factors which are likely to influence listeners' evaluations. Forty-seven native Danish listeners participated in a verbal guise test in which they rated samples from five English varieties on status/competence (the power dimension), solidarity and voice quality dimensions, and had to guess the speaker's origin: Standard Southern British English (SSBE), General American (GA), Australian English (AUS), Scottish English (SCO), and Southern US English (SUS). Additionally, the listeners stated their accent preference and responded to questions regarding their English media consumption. The standard varieties SSBE and GA were rated highly on the power dimension, but downgraded on solidarity. The varieties AUS, SCO, and SUS were rated more positively on solidarity, but downgraded on power. SUS was correctly identified most frequently, followed by SSBE and GA, while the listeners had problems identifying SCO and AUS correctly. Accent preference and English media consumption were related to some of the individual traits, and the listeners' ability to identify GA correctly was related to their English media consumption. Overall, the present results suggest that future studies should examine the influence of media consumption on attitudes towards accents in greater detail.

T‐tapping in Standard Southern British English: An ‘elite’ sociolinguistic variant?

AbstractSocial class is one of the key axes of sociolinguistic variation, but the speech of those at the top of the class spectrum—the elite—is rarely studied. While T‐glottalling has spread widely across British English accents, a competing variant—T‐tapping—has attracted little scholarly attention in the United Kingdom. This article presents a study of elite speech by examining sociolinguistic variation in T‐tapping among adolescent speakers of Standard Southern British English. Data were collected from interviews with teenagers aged 16–19 at two schools in Hampshire, UK. T‐tapping is led by those who previously attended private school and is used more by boys than girls in formal speech. The findings suggest that T‐tapping may be used to index a combination of authority and informality, which is invoked by elite speakers to assert themselves from a position of privilege while maintaining an image of openness and approachability.

Regional variation in British English voice quality

AbstractThis study considers regional variation of voice quality in two varieties of British English – Southern Standard British English and West Yorkshire English. A comparison of voice quality profiles for three closely related but not identical northern varieties within West Yorkshire is also considered. Our findings do not contradict the small subset of previous research which explored regional and/or social variation in voice quality in British English insofar as regionality may play a small role in a speaker’s voice quality profile. However, factors such as social standing and identity could perhaps be even more relevant. Even when considering homogeneous groups of speakers, it is not the case that there is a cohesive voice quality profile that can be attached to every speaker within the group. The reason for this, we argue, is the speaker-specificity inherent in voice quality.

5–9-year-olds categorisation of regional accents: The role of exposure to variation

This study investigates primary school children’s ability to categorise speakers based on regional accent variables found in the UK. Thirty-three children (5;7–9;9) in York took part in a task in which they were asked to group speakers according to whether they sounded as though they came from the same place. Results showed that children were much more accurate at grouping together different speakers according to a Yorkshire/Standard Southern British English accent distinction, rather than a Yorkshire/Scottish or a Yorkshire/North East accent distinction. The distinctiveness and the familiarity of the Standard Southern British English accent features can account for this finding. Furthermore, the children’s general exposure to variation was found to predict their accuracy in the task; children with more exposure to regional variation performed better when grouping speakers based on their pronunciation of the same phoneme embedded within the same word. Building on the results of a previous study with preschool children, these results are discussed in line with an exemplar model of indexical learning.

Examining the success of three kinds of cross-language similarity in predicting English listeners’ discrimination of Danish vowels

Several methods have been used in second language speech research to predict nonnative listeners’ discrimination of foreign language speech sounds. The present study examined how well three of these methods predict the discrimination of the four tightly spaced Danish unrounded front vowels by native listeners of Standard Southern British English (SSBE), whose four front vowels are evenly spaced along the close-open dimension. Experiment 1 examined the ecphoric similarity of the Danish and English vowels in a perceptual assimilation task in which L1 SSBE listeners used native labels and goodness-of-fit ratings in identifying the Danish vowels. Experiment 2 examined the perceptual similarity of the English and Danish vowels in a graded discrimination task in which the same listeners rated the degree of dissimilarity of Danish-English vowel pairs. The third prediction of discriminability was derived from an acoustic comparison of the Danish and English vowels in the F1/F2 space. The same participants as in Experiments 1 and 2 then discriminated three Danish front vowel contrasts. None of the three methods was found to fully successfully predict discrimination of the Danish vowels by the SSBE listeners. The presentation will discuss reasons for the lack of predictive power of the three widely used methods.

Pronunciation Error on English Diphthongs Made by EFL Students

This research addresses pronunciation errors of English diphthongs made by EFL students. The data were collected from pronunciation recordings of 9 post-graduate students who were studying linguistics in the second year. Three of the respondents were male and 6 female students. Studies show that different backgrounds and cultures affect sounds and sound styles as they are related to pronunciation or language styles. The purpose of this research was to analyze and describe the pronunciation errors of English diphthongs made by EFL students. Diphthong is divided into two types, namely GA (General American) and SSBE (Standard Southern British English) or commonly referred to as British accent. To collect the data, the researchers recorded students’ pronunciations using a smartphone. The data collected were then analyzed by employing qualitative and quantitative descriptive methods. The results show that 4 students used SSBE diphthong accent and 5 used GA diphthong accents. In this analysis, there were also 4 students who made pronunciation errors in pronouncing [eɪ], [ɛə], [ʊə] and [aɪ] diphthongs, 3 students had problems about [aʊ] and 2 students mispronounced [oʊ] diphthong.

Second language perception of English vowels by Portuguese learners: The effect of stimulus type

The present study investigated L2 English vowel perception and the effect of stimulus type on the identification of vowel segments that present difficulties for Portuguese learners. It also examined the effect of subject factors such as age of acquisition, length of formal instruction, language use and vocabulary size, on the L2 learners’ perceptual performance. Twenty-nine adult Portuguese learners of English were tested on six English vowels (/iː ɪ ɛ æ ɜː ʌ/) with two tasks, differing in stimulus type, i.e., in the lexical status of trials (real words and pseudo words) in which the target vowels were auditorily presented. The testing stimuli consisted of 72 trials with real CVC words and 72 trials with pseudo CVC words, naturally produced by two speakers of standard southern British English (SSBE). The L2 vocabulary size of the participants was measured with two receptive vocabulary size tests and the language background data, viz. age of learning, length of formal instruction and L2 use was collected with a questionnaire. Results confirmed the Portuguese learners’ difficulties in accurately categorizing the target vowels, particularly when identifying the vowel target sounds embedded in pseudo words, which suggests that L2 phonological categories may be established after lexical forms. Furthermore, a significant correlation was found between L2 language use and accurate perception of four of the target vowels, which indicates that the more frequently learners use the target language, the more accurate is their L2 English vowel perception.

Identification of English vowels by non-native listeners: Effects of listeners’ experience of the target dialect and talkers’ language background

This study investigates how second language (L2) listeners’ perception is affected by two factors: the listeners’ experience with the target dialect – North American English (NAE) vs. Standard Southern British English (SSBE) – and talkers’ language background: native vs. non-native talkers; i.e. interlanguage speech intelligibility benefit (ISIB) talker effects. Two groups of native-Korean-speaking listeners with different target English dialects – L1-Korean listeners of English as a second language (ESL) in the USA and L1-Korean ESL listeners in the UK – were tested on the identification of 12 English vowels spoken by native and non-native (L1-Korean) talkers of NAE and SSBE. The results show that the L2 listeners’ experience with the target dialect had a significant impact on the accuracy of their identification of the L2 vowels. However, no ISIB-talker effects were observed for the L1-Korean listener groups regardless of the listeners’ differences in experience with the two varieties of English. The study adds to the L2 sound acquisition literature and the ISIB literature by looking into L2 learners’ identification of L2 vowels, taking into account the learners’ differences in experience with two standard varieties of English (NAE and SSBE) and the interaction between the learners’ experience with the two varieties and ISIB-talker effects. It also sheds some light on the issue of adult L2 learners’ ability to learn the vowels of a new target variety.

Assessing the effects of accent-mismatched reference population databases on the performance of an automatic speaker recognition system

Automatic Speaker Recognition (ASR) systems are designed to provide the user with statistics relating to the similarity of two or more speech samples and to the typicality of those shared features in the wider population. When an ASR system is used as part of a forensic investigation, the user must decide what counts as the appropriate ‘wider population’ and select a reference database accordingly. While it has generally been held that the voices populating the reference database should be similar in accent to that of the samples under consideration, the degree to which the accents should correspond has until now not been investigated empirically. We report in this article on a study in which the composition of the reference database was systematically varied in terms of accent, using corpora of samples of Standard Southern British English and of three subvarieties spoken in North-East England (Newcastle, Sunderland, Middlesbrough).

Exploring the relationship between voice similarity estimates by listeners and by an automatic speaker recognition system incorporating phonetic features

The present study investigates relationships between voice similarity ratings made by human listeners and comparison scores produced by an automatic speaker recognition system that includes phonetic, perceptually-relevant features in its modelling. The study analyses human voice similarity ratings of pairs of speech samples from unrelated speakers from an accent-controlled database (DyViS, Standard Southern British English) and the comparison scores from an i-vector-based automatic speaker recognition system using ‘auto-phonetic’ (automatically extracted phonetic) features. The voice similarity ratings were obtained from 106 listeners who each rated the voice similarity of pairings of ten speakers on a Likert scale via an online test.Correlation analysis and Multidimensional Scaling showed a positive relationship between listeners’ judgements and the automatic comparison scores. A separate analysis of the subsets of listener responses from English and German native speaker groups showed that a positive relationship was present for both groups, but that the correlation was higher for the English listener group.This work has key implications for forensic phonetics through highlighting the potential to automate part of the process of selecting foil voices in voice parade construction for which the collection and processing of human judgements is currently needed. Further, establishing that it is possible to use automatic voice comparisons using phonetic features to select similar-sounding voices has important applications in ‘voice casting’ (finding voices that are similar to a given voice) and ‘voice banking’ (saving one's voice for future synthesis in case of an operation or degenerative disease).