- ARC Centre of Excellence for the Dynamics of Language, 2015, Department MemberThe University of Sheffield, Human Communication Sciences, Alumnus, and 4 moreadd
- Linguistics, Language Acquisition, Psycholinguistics, Second Language Acquisition, Sociophonetics, Applied Linguistics, and 17 moreSpeech perception, Phonetics, Phonology, Acoustic Phonetics, Speech Sciences, Developmental Psychology, Speech Communication, Accents, British Dialects and Accents, Sociolinguistics, Language Learning, Statistical Learning, Speech Acoustics, Auditory Perception, Vowels, Speech Production, and Dialects of Englishedit
- I am a speech and language researcher in the Linguistics Department at the University of Potsdam. I was recently awar... moreI am a speech and language researcher in the Linguistics Department at the University of Potsdam. I was recently awarded a grant from the Australian government to carry out a new project at the MARCS Institute (Western Sydney University) on "Dynamic properties of Australian vowels in production and perception" that examines the extent to which different properties in the speech signal can serve as distinct cues for (1) phoneme and speaker identity and (2) learning new words in non-native dialects. The first output is a paper (to be published in the Journal of the Acoustical Society of America) reporting on new Australian English data collected by Jaydene Elvin (MARCS Institute) as part of her PhD thesis. I hold a PhD from the University of Sheffield and Bachelors and Masters degrees from University College London (UCL). I am working on several other research projects that examine variability in speech production and perception which make use of acoustic analysis and behavioural and neuroimaging methods. My research involves colleagues from the MARCS Institute (Western Sydney University), Cardiff Metropolitan University and Australian National University.edit
Many cross-language and L2 speech perception studies have been conducted on English sounds and a limited number of speakers or synthetic tokens have been used for auditory stimuli. The Spanish listeners of the present study were presented... more
Many cross-language and L2 speech perception studies have been conducted on English sounds and a limited number of speakers or synthetic tokens have been used for auditory stimuli. The Spanish listeners of the present study were presented with natural tokens of Dutch vowels produced by males and females selected from the corpus reported in Adank et al. [(2004). J. Acoust. Soc. Am. 116, 1729–1738]. The results show that single category assimilations are common and that certain Dutch vowels frequently assimilate to Spanish diphthongs. Predictions are made for Spanish learners’ initial stage in the acquisition of the Dutch vowel system.
Research Interests:
Recent vowel corpora show that there are often clear acoustic differences between vowels produced in different phonetic contexts. We expand on a recent corpus of Northern Standard Dutch (NSD) vowels by including a variety of consonantal... more
Recent vowel corpora show that there are often clear acoustic differences between vowels produced in different phonetic contexts. We expand on a recent corpus of Northern Standard Dutch (NSD) vowels by including a variety of consonantal contexts. Our results show that there are very clear contextual effects on the spectral and temporal properties of NSD vowels. The most striking effect is the apparent 'fronting' of vowels in alveolar contexts, which has not previously been reported for Dutch. Classification with a supervised learning algorithm reveals some substantial differences between our acoustic measurements and those reported in earlier studies on NSD vowels.
Research Interests:
Peruvian Spanish (PS) and Iberian Spanish (IS) learners were tested on their ability to categorically discriminate and identify Dutch vowels. It was predicted that the acoustic differences between the vowel productions of the two... more
Peruvian Spanish (PS) and Iberian Spanish (IS) learners were tested on their ability to categorically discriminate and identify Dutch vowels. It was predicted that the acoustic differences between the vowel productions of the two dialects, which compare differently to Dutch vowels, would manifest in differential L2 perception for listeners of these two dialects. The results show that although PS learners had higher general L2 proficiency, IS learners were more accurate at discriminating all five contrasts and at identifying six of the L2 Dutch vowels. These findings confirm that acoustic differences in native vowel production lead to differential L2 vowel perception.
Research Interests:
Vowel inventories vary across languages in terms of the phonological vowel categories within them and the phonetic properties of individual vowels. The same also holds across different accents of a language. The four studies in this... more
Vowel inventories vary across languages in terms of the phonological vowel categories within them and the phonetic properties of individual vowels. The same also holds across different accents of a language. The four studies in this project address the role of listeners’ native accents in the cross-language acoustic and perceptual similarity of vowels. Study I explores the acoustic similarity of Northern Standard Dutch (NSD) vowels to the vowels in two accents of British English, namely Standard Southern British English (SSBE) and Sheffield English (SE), and demonstrates that some NSD vowels are acoustically most similar to different SSBE and SE vowels and that other NSD vowels differ in the degree of acoustic similarity to SSBE and SE vowels. Study II examines how SSBE and SE listeners use spectral properties to identify English monophthongs and finds that SSBE and SE listeners differ on some monophthongs, broadly in line with the spectral differences between naturally produced SSBE and SE vowels. Study III investigates SSBE and SE listeners’ discrimination accuracy of five NSD vowel contrasts, which cause British English learners of Dutch perceptual problems, and shows that SE listeners are generally less accurate than SSBE listeners. Study IV tests SSBE and SE listeners’ perceptual similarity of NSD vowels to English vowels and reveals that SSBE and SE listeners differ on some NSD vowels. The present findings demonstrate the influence of listeners’ differential linguistic experience on speech perception and underscore the importance of accounting for listeners’ particular native accents in cross-language studies.
Research Interests:
This review examines research on speech perception by both native and non-native listeners. The development of speech perception in infancy is first considered and a theoretical model that accounts for this is introduced. A brief overview... more
This review examines research on speech perception by both native and non-native listeners. The development of speech perception in infancy is first considered and a theoretical model that accounts for this is introduced. A brief overview then follows of several research areas under the umbrella of non-native speech perception, namely cross-dialect, cross-language and second-language speech perception. It is shown that non-native and native speech perception is critically shaped by the specific ways in which speakers use acoustic cues in speech production.
Research Interests:
This paper examines to what extent acoustic similarity between native and non-native vowels predicts non-native vowel perception and whether this process is influenced by listeners’ native and other non-native dialects. Listeners with... more
This paper examines to what extent acoustic similarity between native and non-native vowels predicts non-native vowel perception and whether this process is influenced by listeners’ native and other non-native dialects. Listeners with Northern and Southern British English dialects completed a perceptual assimilation task in which they categorized tokens of 15 Dutch vowels in terms of English vowel categories. While the cross-language acoustic similarity of Dutch vowels to English vowels largely predicted Southern listeners’ perceptual assimilation patterns, this was not the case for Northern listeners, whose assimilation patterns resembled those of Southern listeners for all but three Dutch vowels. The cross-language acoustic similarity of Dutch vowels to Northern English vowels was re-examined by incorporating Southern English tokens, which resulted in considerable improvements in the predicting power of cross-language acoustic similarity. This suggests that Northern listeners’ assimilation of Dutch vowels to English vowels was influenced by knowledge of both native Northern and non-native Southern English vowel categories. The implications of these findings for theories of non-native speech perception are discussed.
Research Interests:
Evidence of distributional learning, a statistical learning mechanism centered on relative frequency of exposure to different tokens, has mainly come from short-term learning and therefore does not ostensibly address the development of... more
Evidence of distributional learning, a statistical learning mechanism centered on relative frequency of exposure to different tokens, has mainly come from short-term learning and therefore does not ostensibly address the development of important learning processes. The present longitudinal study examines both short- and long-term effects of distributional learning of phonetic categories on non-native sound discrimination over a 12-month period. Two groups of listeners were exposed to a two-minute distribution of auditory stimuli in which the most frequently presented tokens either approximated or exaggerated the natural production of the speech sounds, whereas a control group listened to a piece of classical music for the same length of time. Discrimination by listeners in the two distribution groups improved immediately after the short exposure, replicating previous results. Crucially, this improvement was maintained after six and 12 months, demonstrating that distributional learning has long-lasting effects.
Research Interests:
This study compares the duration and first two formants (F1 and F2) of 11 nominal monophthongs and five nominal diphthongs in Standard Southern British English (SSBE) and a Northern English dialect [Sheffield English dialect]. F1 and F2... more
This study compares the duration and first two formants (F1 and F2) of 11 nominal monophthongs and five nominal diphthongs in Standard Southern British English (SSBE) and a Northern English dialect [Sheffield English dialect]. F1 and F2 trajectories were fitted with parametric curves using the discrete cosine transform (DCT) and the zeroth DCT coefficient represented formant trajectory means and the first DCT coefficient represented the magnitude and direction of formant trajectory change to characterize vowel inherent spectral change (VISC). Cross-dialectal comparisons involving these measures revealed significant differences for the phonologically back monophthongs /ɒ, ɔː, ʊ, uː/ and also /ɜː/ and the diphthongs /eɪ, əʊ, aɪ, ɔɪ/. Most cross-dialectal differences are in zeroth DCT coefficients, suggesting formant trajectory means tend to characterize such differences, while first DCT coefficient differences were more numerous for diphthongs. With respect to VISC, the most striking differences are that /uː/ is considerably more diphthongized in the Northern dialect and that the F2 trajectory of /əʊ/ proceeds in opposite directions in the two dialects. Cross-dialectal differences were found to be largely unaffected by the consonantal context in which the vowels were produced. The implications of the results are discussed in relation to VISC, consonantal context effects and speech perception.
Research Interests:
Theories and methods modelling vowel quality in terms of vowel inherent spectral change (VISC) have been developed and tested overwhelmingly on North American English (AE) dialects, which raises the question of their generalisability in... more
Theories and methods modelling vowel quality in terms of vowel inherent spectral change (VISC) have been developed and tested overwhelmingly on North American English (AE) dialects, which raises the question of their generalisability in non-AE dialects and other languages. The present paper examines VISC as an aspect of vowel quality in Standard Southern British English (SSBE) and Northern Standard Dutch (NSD). Despite markedly different VISC patterns, SSBE vowels are analysable along the same lines as in AE. While the same mostly holds for NSD, VISC is found to be more important for determining SSBE vowel quality, especially for SSBE nominal diphthongs. Additionally, a pair of NSD diphthongs presents a challenge for current theories and methods as they are acoustically similar. In line with studies on AE, theorising vowel quality in terms of VISC aids descriptions of vowels and removes the need to treat nominal monophthongs and diphthongs in different ways.
Research Interests:
Aims and Objectives: This study investigates the effects of individual bilingualism and long-term language contact on monophthongal vowel productions in English and Welsh. Design: To this end, we recorded the Welsh and English vowel... more
Aims and Objectives:
This study investigates the effects of individual bilingualism and long-term language contact on monophthongal vowel productions in English and Welsh.
Design:
To this end, we recorded the Welsh and English vowel productions of two sets of Welsh-English bilinguals differing in home language use, as well as the English vowel productions of English monolinguals.
Data and analysis:
The data were analysed acoustically, with a focus on spectral and temporal properties. Comparisons were then made within each language and cross-linguistically.
Findings:
The results of a cross-linguistic acoustic comparison revealed a high degree of convergence in the monophthong systems of Welsh and English, but also some language-specific categories. Interestingly, at the individual level we found no effect of linguistic experience on vowel production: the two sets of bilinguals and the English monolinguals did not differ in their realisation of English vowels, and the two sets of bilinguals did not differ in their realisation of Welsh vowels.
Originality:
This is one of few studies to examine the effect of linguistic background on variation in Welsh and English bilingual speech, and the first to compare the speech of Welsh-English bilinguals and English monolinguals. More specifically, it investigates the extent to which a speaker’s home language can affect phonetic variation in a close-knit community of speakers and in a situation characterised by long-term language contact.
Implications:
The findings demonstrate pervasive phonetic convergence in a language contact situation with a historical substrate. They also indicate that a homogeneous peer group with shared values can override the effects of individual linguistic experience.
This study investigates the effects of individual bilingualism and long-term language contact on monophthongal vowel productions in English and Welsh.
Design:
To this end, we recorded the Welsh and English vowel productions of two sets of Welsh-English bilinguals differing in home language use, as well as the English vowel productions of English monolinguals.
Data and analysis:
The data were analysed acoustically, with a focus on spectral and temporal properties. Comparisons were then made within each language and cross-linguistically.
Findings:
The results of a cross-linguistic acoustic comparison revealed a high degree of convergence in the monophthong systems of Welsh and English, but also some language-specific categories. Interestingly, at the individual level we found no effect of linguistic experience on vowel production: the two sets of bilinguals and the English monolinguals did not differ in their realisation of English vowels, and the two sets of bilinguals did not differ in their realisation of Welsh vowels.
Originality:
This is one of few studies to examine the effect of linguistic background on variation in Welsh and English bilingual speech, and the first to compare the speech of Welsh-English bilinguals and English monolinguals. More specifically, it investigates the extent to which a speaker’s home language can affect phonetic variation in a close-knit community of speakers and in a situation characterised by long-term language contact.
Implications:
The findings demonstrate pervasive phonetic convergence in a language contact situation with a historical substrate. They also indicate that a homogeneous peer group with shared values can override the effects of individual linguistic experience.
Research Interests: Phonology, Welsh, Welsh linguistics, Language Acquisition, Phonetics, and 12 moreDialects of English, Sociolinguistics, Language Variation and Change, Speech Production, Sociophonetics, Bilingual Education, Speech Communication, Acoustic Phonetics, Adolescent, Bilingualism, Language contact, and British Dialects and Accents
Acoustic studies of several languages indicate that second-formant (F2) slopes in high vowels have opposing directions (independent of consonantal context): front [iː]-like vowels are produced with a rising F2 slope, whereas back... more
Acoustic studies of several languages indicate that second-formant (F2) slopes in high vowels have opposing directions (independent of consonantal context): front [iː]-like vowels are produced with a rising F2 slope, whereas back [uː]-like vowels are produced with a falling F2 slope. The present study first reports acoustic measurements that confirm this pattern for the English variety of Standard Southern British English (SSBE), where /uː/ has shifted from the back to the front area of the vowel space and is now realized with higher midpoint F2 values than several decades ago. Subsequently, we test whether the direction of F2 slope also serves as a reliable cue to the /iː/-/uː/ contrast in perception. The findings show that F2 slope direction is used as a cue (additional to midpoint formant values) to distinguish /iː/ from /uː/ by both young and older Standard Southern British English listeners: an otherwise ambiguous token is identified as /iː/ if it has a rising F2 slope and as /uː/ if it has a falling F2 slope. Furthermore, our results indicate that listeners generalize their reliance on F2 slope to other contrasts, namely /ɛ/-/ɒ/ and /æ/-/ɒ/, even though F2 slope is not employed to differentiate these vowels in production. This suggests that in Standard Southern British English, a rising F2 seems to be perceptually associated with an abstract feature such as [+front], whereas a falling F2 with an abstract feature such as [-front].
Research Interests:
This study presents the first acoustic description of the vowel space of a Papuan language – Nambo, spoken in southern Papua New Guinea – based on duration and first and second formant measurements from 19 adult male and female speakers... more
This study presents the first acoustic description of the vowel space of a Papuan language – Nambo, spoken in southern Papua New Guinea – based on duration and first and second formant measurements from 19 adult male and female speakers across three age groups (young, middle-aged, senior). Phonemically, Nambo has six full vowels /i, e, æ, ɑ, o, u/ and a reduced vowel tentatively labeled /ə/. Unlike the full vowels, the quality of /ə/ showed great variation: seniors’ and young females’ realizations tended to be more open and retracted than those by young males, while middle-aged speakers’ productions fell between these two variants.
Research Interests: Languages and Linguistics, Phonology, Phonetics, Sociolinguistics, Speech Sciences, and 12 moreSociophonetics, Linguistics, Speech Communication, Fieldwork in linguistics, Papua New Guinea, Acoustic Phonetics, Papuan linguistics, Speech, Experimental Phonetics, Papuan Languages, Vowels, and Field linguistics
This study provides a thorough acoustic analysis of the 18 Australian English monophthongs and diphthongs produced in a variety of phonetic contexts by young adult speakers from Western Sydney. The 18 vowels are well separated by duration... more
This study provides a thorough acoustic analysis of the 18 Australian English monophthongs and diphthongs produced in a variety of phonetic contexts by young adult speakers from Western Sydney. The 18 vowels are well separated by duration and mean and dynamic formant trajectory information. Vowel durations and formant trajectories were affected by the consonantal context in which they were produced, particularly those produced in the /hVd/ context. Finally, the results indicate that capturing aspects of vowel inherent spectral change may be useful in future cross-dialectal and cross-linguistic studies. Copyright 2016 Acoustical Society of America
Research Interests: Acoustics, Languages and Linguistics, Phonology, Phonetics, Dialectology, and 16 moreDialects of English, Sociolinguistics, English language, English, Australia, Speech Acoustics, Sociophonetics, Linguistics, Speech Communication, Acoustic Phonetics, Speech, Accents, Experimental Phonetics, Vowels, Australian English, and English accents
Speech scientists have long noted that the qualities of naturally-produced vowels do not remain constant over their durations – regardless of being nominally "monophthongs" or "diphthongs". Recent acoustic corpora show that there are... more
Speech scientists have long noted that the qualities of naturally-produced vowels do not remain constant over their durations – regardless of being nominally "monophthongs" or "diphthongs". Recent acoustic corpora show that there are consistent patterns of first (F1) and second (F2) formant frequency change across different vowel categories. The three Australian English (AusE) close front vowels /iː, ɪ, ɪə/ provide a striking example: while their midpoint or mean F1 and F2 frequencies are virtually identical, their spectral change patterns distinctly differ. The present study utilizes a pre-attentive discrimination paradigm with electroencephalography to assess AusE listeners' perceptual sensitivity to close front vowels with different F1 × F2 trajectory lengths (TLs) and directions (TDs). When TLs are modest, there is an asymmetry in perceptual sensitivity: closing vowels, e.g., /iː/ whose trajectory terminates high in the F1 × F2 vowel space, are perceptually prominent, whereas centering vowels, e.g., /ɪ, ɪə/ whose trajectories end more centrally, are not. However, when TLs are exaggerated, the asymmetry in the perceptual sensitivity to the two TDs is substantially reduced. The results indicate that, despite the distinct patterns of spectral change of AusE /iː, ɪ, ɪə/ in production, its perceptual relevance is not uniform, but rather vowel-category dependent.
Research Interests: Cognitive Science, Acoustics, Languages and Linguistics, Phonology, Brain Imaging, and 12 morePhonetics, Dialectology, Dialects of English, Auditory Perception, Speech perception, Speech Communication, Acoustic Phonetics, Cognitive Neuroscience, EEG Signal Processing, Change detection, Vowels, and Australian English
Australian English (AusE) /iː/, /ɪ/ and /ɪə/ exhibit almost identical average first (F1) and second (F2) formant frequencies and differ in duration and vowel inherent spectral change (VISC). The cues of duration, F1 × F2 trajectory... more
Australian English (AusE) /iː/, /ɪ/ and /ɪə/ exhibit almost identical average first (F1) and second (F2) formant frequencies and differ in duration and vowel inherent spectral change (VISC). The cues of duration, F1 × F2 trajectory direction (TD) and length (TL) were assessed in listeners’ categorization of /iː/ and /ɪə/ compared to /ɪ/. Duration was important for distinguishing both /iː/ and /ɪə/ from /ɪ/. TD and TL were important for categorizing /iː/ versus /ɪ/, whereas only TL was important for /ɪə/ versus /ɪ/. Finally, listeners’ use of duration and VISC was not mutually affected for either vowel compared to /ɪ/.
© 2018 Acoustical Society of America
© 2018 Acoustical Society of America
Research Interests:
The present study investigated the extent to which variation along formant trajectory dimensions-considered both separately and simultaneously-manifests in the English diphthongs CHOICE, FACE, MOUTH, GOAT and PRICE. The sources of... more
The present study investigated the extent to which variation along formant trajectory dimensions-considered both separately and simultaneously-manifests in the English diphthongs CHOICE, FACE, MOUTH, GOAT and PRICE. The sources of variation were phonemic category, flanking consonants, speaker's gender and speaker's regional background. Formant trajectory dimensions were overall and time-varying first (F1) and second (F2) formant trajectories as parameterised by their means and time-varying slopes and curvature using the discrete cosine transform. Phonemic category was robustly predicted by variation in the combination of F1 and F2 means and slopes, whereas flanking consonants were not. Gender was strongly predicted by variation in F1 and F2 means, but not in F1 and F2 time-varying aspects. Regional background emerged in the variation of both F1 and F2 means, slopes and curvature to roughly similar extents. Given the acoustic multidimensionality of vowels, variation is most appropriately viewed with a multivariate approach.
Research Interests:
Second-language (L2) speech perception research has largely focused on learning to detect phonetic variation which contributes to phonemic differences. However, phonetic variation may emerge for other reasons, e.g., different speakers. In... more
Second-language (L2) speech perception research has largely focused on learning to detect phonetic variation which contributes to phonemic differences. However, phonetic variation may emerge for other reasons, e.g., different speakers. In native language acquisition, speech perception is flexible enough to permit speaker-related phonetic variation while at the same time preserving phonemic identity. In L2 speech learning, it remains unclear how speaker-related phonetic variation is learned or processed. To test how monolinguals and L2 learners handle speaker-related phonetic variation in detecting phonemic differences, the present study investigated the discrimination of novel words spoken in a familiar and unfamiliar accent. Compared to monolinguals, near-native L2 learners made slightly more errors overall and were quicker to attribute large phonetic differences to phonemic differences rather than speaker-related differences. This highlights the challenging task of learning naturally permissible variation in L2 phonemic categories.
Research Interests:
Generalised Additive Models (GAMs) provide a flexible way for modelling non-linear relationships, which is a welcome addition to research on vowel acoustics. If formant trajectory modelling is conducted solely within the confines of how... more
Generalised Additive Models (GAMs) provide a flexible way for modelling non-linear relationships, which is a welcome addition to research on vowel acoustics. If formant trajectory modelling is conducted solely within the confines of how GAMs are implemented in statistical packages, such as mgcv (Wood, 2017), there are some limitations with respect to how the acoustic (temporal and spectral) properties of vowels are to be modelled. This is because, for the time-varying properties of formant trajectories, mcgv disallows relationships between predictors and disallows relationships between dependent variables, meaning crossed and multivariate data structures, common in experimental designs, cannot be easily modelled. The present piece outlines a way in which the utility of GAMs for characterising formant trajectories of vowels might be (1) extended to such data structures and also (2) how theories of vowel perception can guide GAM modelling to capture perceptually relevant properties of formant trajectories. It is proposed that the use of GAMs can act as a “stepping stone” to analyses which can more closely reflect researchers’ data structures and research aims. An example of how GAMs might be used in this way is provided. It is concluded that further work is required to refine how the weaknesses of GAMs for capturing important data structure information can be compensated for, so that the strengths of GAMs for formant trajectory curve fitting can be fully capitalised on for research on vowel acoustics.