Speech Technology and Argentinean Welsh

Total Page:16

File Type:pdf, Size:1020Kb

Speech Technology and Argentinean Welsh Speech technology and Argentinean Welsh Elise Bell The University of Texas at El Paso [email protected] Abstract under-resourced languages, including Welsh and other Celtic languages, making creation of speech This paper argues for increased efforts to technologies for these languages more challeng- source Welsh language data from the pop- ing. As we undertake that challenge, it is vital that ulation of Welsh speakers in Argentina. we consider the source of the data on which our The dialect of Argentinean Welsh is under- technology is based. As less-resourced language resourced even in comparison to other speech technology becomes more broadly accessi- Celtic languages, which are already con- ble, speakers who deviate from the norms explic- sidered less-resourced languages (LRLs). itly or implicitly assumed by the technology will Argentinean Welsh has been shown to begin to come in contact with it. Depending on diverge from other dialects of Welsh the variety inherent in the data underlying the sys- in the realization of acoustic contrasts tem, those marginalized speakers may or may not such as voice-onset time and vowel dura- be able to successfully take advantage of speech tion. These differences potentially obscure technology. The aim of this paper is to highlight phonemic contrasts in the language, cre- the particular areas of speech technology develop- ating homophony absent in other dialects. ment that may create obstacles or pose problems The inclusion of Argentinean Welsh data for users, and to propose the addition of a partic- in training sets for future Welsh speech ular source of acoustic data that lies outside the technology development will increase the norm for Welsh language technology. The main applicability of such technology to other speaker group of concern here is speakers of Welsh speaker communities whose Welsh speech in Argentina, but the arguments that follow apply may not align with that currently in use for to second language (L2) or non-fluent speakers of model training, including second-language Welsh as well. and non-fluent speakers. Compared to dialects of Welsh spoken in Wales, Argentinean (or Patagonian) Welsh is extremely 1 Introduction under-resourced and under-researched. Documen- The development of speech language technology tation efforts amount to a handful of citations such as automatic speech recognition (ASR) de- (Jones, 1984; Jones, 1998; Sleeper, 2015; Bell, pends on the availability and accessibility of large- 2017), and to my knowledge, only one speech cor- scale language data sets, both spoken and written. pus. Little is known about how the dialect of Welsh The information in these data sets is used to create spoken in Argentina differs from other dialects of statistical generalizations that form the basis for Welsh, although there are several reasons to expect speech technologies including speech recognition, dialectal variation. The effects of bilingualism on text-to-speech systems, and grammatical parsing. speech production are well documented (Flege et Large resources of this type are less available for al., 1997; Flege et al., 2003; Escudero, 2009), and all adult speakers of Argentinean Welsh are bilin- c 2019 The authors. This article is licensed under a Creative Commons 4.0 licence, no derivative works, attribution, CC- gual with Spanish (if not trilingual with English or BY-ND. another language). Dialect differences may also Proceedings of the Celtic Language Technology Workshop 2019 Dublin, 19–23 Aug., 2019 | p. 16 arise from the effects of second language (L2) ac- using the Paldaruo app3 and website to crowd- quisition of Welsh. Differences in speech produc- source the collection of Welsh utterances (Prys tion due to these effects may include the merg- and Jones, 2018). Utterances were elicited with a ing of phonemic categories, or the use of differ- set of target words and sentences designed to col- ent acoustic cues in contrast production. Because lect a representative phoneme set. The project is these effects are fairly inextricably tied up with currently available through the Mozilla Common- the effect of Spanish bilingualism on Argentinean Voice project4 where users can contribute and eval- Welsh in general, they will not be treated sepa- uate recordings, and where a portion of the vetted rately here. This paper presents a brief overview data is available for download. of the state of Welsh language speech technology There are several other text and speech corpora and resources, followed by a short discussion of available for the Welsh language. The Language the history and modern context of the Welsh lan- Technologies Unit has created multiple text cor- guage in Argentina. Subsequently, I present evi- pora, including one of social media posts5 as well dence that experimentally observed differences be- as a million-word corpus consisting of various reg- tween Welsh dialects support the inclusion of Ar- isters of Welsh writing.6 Researchers at Bangor gentinean Welsh data in future speech technology University’s ESRC Centre for Research on Bilin- development efforts. gualism in Theory & Practice have also produced two publicly available corpora of Welsh bilingual 1.1 Welsh speech technology speech.7 One of these, the Patagonia corpus, is to my knowledge the only publicly available col- Speech recognition and speech synthesis technolo- lection of Argentinean Welsh speech. While such gies rely on (relatively) large amounts of acoustic conversational corpora are invaluable for the study data, which must be transcribed orthographically of syntactic and morphological phenomena (Carter (in the case of a grapheme-based speech recog- et al., 2010; Webb-Davies, 2016), the acoustic data nition system) or phonetically (in the case of a they contain is not always of high enough quality, phoneme-based system). The collection, analy- nor is the corpus large enough, to stand alone as the sis, and processing of this data requires resources sole source for development of speech technology. including people-hours, funding, and often, par- A brief history of Welsh in Argentina is presented ticipation of community members in data crowd- below, followed by a discussion of the benefits that sourcing efforts (Prys and Jones, 2018). Currently, Argentinean Welsh data may have on future devel- available Welsh speech technology is fairly lim- opment of Welsh speech technology. ited (compared to larger-resourced languages like English). Much of what is available has been 1.2 Argentinean Welsh produced by the Welsh Language Technologies The presence of Welsh in Argentina is due to 1 Unit, based at Bangor University. Tools pro- the mid-19th century efforts of a group of Welsh duced by the Language Technologies Unit range speakers, led by Michael D. Jones, who sought to from front-end resources accessible to the public establish a Welsh colony away from the influence (a vocabulary website plugin (Jones et al., 2016), of the English language and British government a Welsh language spelling and grammar checker (Williams, 1975). In 1865, following an agree- (Prys et al., 2016)) to back-end tools such as a ment with the government of Argentina, a Welsh part-of-speech tagger (Prys and Jones, 2015) that colony was established in the Patagonia region of are open source and accessible to researchers out- of the country. Today, descendants of the origi- side of the unit itself. The unit has also devel- nal 200 colonists (and of the several thousand who oped speech recognition and synthetic speech tech- followed in subsequent years) still maintain the nologies that are of particular relevance to this pa- Welsh language and culture in Argentina. Mod- 2 per. These include the development of Macsen, a ern Welsh speakers are clustered in two areas of Welsh-language personal digital assistant based on 3https://apps.apple.com/bs/app/paldaruo/ data collected by the Languages Technologies Unit id840185808 4https://voice.mozilla.org 1www.bangor.ac.uk/canolfanbedwyr/ 5http://techiaith.cymru/data/corpora/ technolegau_iaith.php.en twitter 2http://techiaith.cymru/2016/05/ 6http://corpws.cymru/ceg/ introducing-macsen 7http://bangortalk.org.uk/ Proceedings of the Celtic Language Technology Workshop 2019 Dublin, 19–23 Aug., 2019 | p. 17 Chubut Province, in Dyffryn Camwy on the At- developed. Specific aspects of Argentinean Welsh lantic coast, and Cwm Hyfryd to the west. variation, which may be due to synchronic effects Although inter-generational transmission of the from first language Spanish, the effect of lifelong language waned during the 20th century, revital- Spanish bilingualism, or diachronic dialect diver- ization efforts were spurred in 1965, the centennial gence, are discussed below. of the original colony’s establishment. The cen- Today, all adult speakers of Welsh are at least tennial celebration renewed interest in Welsh cul- bilingual, either with English (in Wales) or with ture and language, and by the 1990s several lan- Spanish (in Argentina). This situation compli- guage initiatives were established which still ex- cates what might otherwise be a straightforward ist today. These include Welsh language medium dialect comparison between differing varieties of primary schools, annual Eisteddfodau (traditional Welsh. Cross-linguistic influence from competing poetry and song competitions), and an ongoing languages Spanish and English is entangled with teacher exchange program with Wales through the other linguistic pressures, including effects of first 8 Welsh Language Project. language (L1) on second language
Recommended publications
  • Salient Features of the Welsh Accent That Are Chosen to Be Portrayed in Film
    MA Language and Communication Research Cardiff University Salient Features of the Welsh Accent that are Chosen to be Portrayed in Film Andrew Booth C1456511 Supervisor: Dr Mercedes Durham Word Count: 16,448 September 2015 Abstract The accent portrayed by an actor in films has many different implications to the audience. For authenticity, the filmmakers need their accent to be as close to genuine speech as possible. The Welsh-English accent in film is portrayed in many different ways; the aim of this study is to investigate which features are viewed as salient to filmmakers when portraying a Welsh accent. This dissertation focuses the portrayal of salient features of the Welsh-English accent in the film Pride (2014). Pride was chosen because it can compare Welsh to non-Welsh actors who portray a Welsh-English accent. The research is carried out on the film using both auditory and acoustic analysis. Tokens from the film were coded in terms of their realisations for analysis and comparison to previous literature. This research produced a number of key findings: first, the Welsh actors supported previous research on patterns of realisation for Welsh-English. Second, the non-Welsh actors recognised and produced the key features of a Welsh-English accent. Finally, the features presented are salient when representing a Welsh accent in film. In summary, theories such as accommodation, language transference, hypercorrection, fudging and transition were used to explain variation of accents. This research argues for a multi- methodological approach to analysing different features of a Welsh-English accent in film. ACKNOWLEDGEMENTS I would like to thank my supervisor, Dr Mercedes Durham for her instrumental advice and insight over the past few months.
    [Show full text]
  • Old and Middle Welsh David Willis ([email protected]) Department of Linguistics, University of Cambridge
    Old and Middle Welsh David Willis ([email protected]) Department of Linguistics, University of Cambridge 1 INTRODUCTION The Welsh language emerged from the increasing dialect differentiation of the ancestral Brythonic language (also known as British or Brittonic) in the wake of the withdrawal of the Roman administration from Britain and the subsequent migration of Germanic speakers to Britain from the fifth century. Conventionally, Welsh is treated as a separate language from the mid sixth century. By this time, Brythonic speakers, who once occupied the whole of Britain apart from the north of Scotland, had been driven out of most of what is now England. Some Brythonic-speakers had migrated to Brittany from the late fifth century. Others had been pushed westwards and northwards into Wales, western and southwestern England, Cumbria and other parts of northern England and southern Scotland. With the defeat of the Romano-British forces at Dyrham in 577, the Britons in Wales were cut off by land from those in the west and southwest of England. Linguistically more important, final unstressed syllables were lost (apocope) in all varieties of Brythonic at about this time, a change intimately connected to the loss of morphological case. These changes are traditionally seen as having had such a drastic effect on the structure of the language as to mark a watershed in the development of Brythonic. From this period on, linguists refer to the Brythonic varieties spoken in Wales as Welsh; those in the west and southwest of England as Cornish; and those in Brittany as Breton. A fourth Brythonic language, Cumbric, emerged in the north of England, but died out, without leaving written records, in perhaps the eleventh century.
    [Show full text]
  • Why Learning to Read Is Easier in Welsh Than in English: Orthographic Transparency Effects Evinced with Frequency-Matched Tests
    Applied Psycholinguistics 22 (2001), 571–599 Printed in the United States of America Why learning to read is easier in Welsh than in English: Orthographic transparency effects evinced with frequency-matched tests NICK C. ELLIS and A. MARI HOOPER University of Wales, Bangor ADDRESS FOR CORRESPONDENCE Nick C. Ellis, School of Psychology, University of Wales, Bangor, Gwynedd LL57 2DG, United Kingdom. E-mail: [email protected] ABSTRACT This study compared the rate of literacy acquisition in orthographically transparent Welsh and ortho- graphically opaque English using reading tests that were equated for frequency of written exposure. Year 2 English-educated monolingual children were compared with Welsh-educated bilingual chil- dren, matched for reading instruction, background, locale, and math ability. Welsh children were able to read aloud accurately significantly more of their language (61% of tokens, 1821 types) than were English children (52% tokens, 716 types), allowing them to read aloud beyond their compre- hension levels (168 vs. 116%, respectively). Various observations suggested that Welsh readers were more reliant on an alphabetic decoding strategy: word length determined 70% of reading latency in Welsh but only 22% in English, and Welsh reading errors tended to be nonword mispro- nunciations, whereas English children made more real word substitutions and null attempts. These findings demonstrate that the orthographic transparency of a language can have a profound effect on the rate of acquisition and style of reading adopted by its speakers. Systems that are noisy and inconsistent are harder to sort out than systems that are reliable and categorical. There is now a large body of research demonstrating that greater ambiguity in the mappings between the forms and functions of a particular language causes less successful learning because of a larger degree of competition among the cues in the learning set (Bates & MacWhinney, 1987; MacWhinney, 1987).
    [Show full text]
  • New Approaches to Brittonic Historical Linguistics Abstracts
    New Approaches to Brittonic Historical Linguistics Abstracts Gwen Awbery Aberystwyth University / University of Wales Trinity St David Historical dialectology and Welsh churchyards Memorial inscriptions on gravestones are an important source of evidence for dialect variation in Welsh in the past. They exist in truly enormous numbers, are found throughout the country, commemorating people from all walks of life, and the tradition of using Welsh in this context goes back to the mid-eighteenth century. Most revealing are the poems which form part of the inscription. Some have rhyme schemes which work only with very specific features of dialect; others are slightly garbled versions of poems by well-known writers, where the influence of dialect features can explain the changes made to the original. Since the location and date of each inscription is known, it appears possible to build up a picture of where and when these dialect features were in use. The situation is not totally straightforward, however, and there are inevitably problems which arise in dealing with this material, ranging from the need for extensive and time- consuming fieldwork, to the difficulties of reading inscriptions on worn and damaged stones, and the disconcerting tendency of some poems to show up in unexpected locations. Bernhard Bauer Maynooth University Close encounters of the linguistic kind: the Celtic glossing tradition The project Languages in Exchange: Ireland and her Neighbours (LeXiN) aims for a better understanding of the linguistic contacts between British Celtic and Irish in the early medieval period. The focal point of the investigation forms the “Celtic glossing tradition”, especially the vernacular glosses on the computistic works of Bede and on the Latin grammar of Priscian.
    [Show full text]
  • Welsh Mutation and Strict Modularity © Õìth August Óþõé Florian Breit
    §Z f§u± WELSHMUTATIONANDSTRICTMODULARITY WELSHMUTATION & STRICTMODULARITY Florian Breit esis submitted in partial fullment of the requirements for the degree of Doctor of Philosophy in Linguistics Department of Linguistics University College London Õìth August óþÕÉ Florian Breit: Welsh Mutation and Strict Modularity © Õìth August óþÕÉ Florian Breit £§h£Z «¶£u§ê«§: John Harris «¶f«oZ§í «¶£u§ê«§: Andrew Nevins is document was typeset using the typographical look-and-feel classicthesis Minion Pro (version ¦.ó) for LYX by André Miede. e type faces used are by Robert Slimbach and Euler by Hermann Zapf. For the greater glory of God, and in memory of Uncle Peter, who taught me the importance of love, the meaning of sorrow, and the struggles of happiness. Declaration I, Florian Breit, conrm that the work presented in this thesis is my own. Where information has been derived from other sources, I conrm that this has been indicated in the thesis. London, Õìth August óþÕÉ ............................. Florian Breit vii Abstract e generative view of the language faculty is premised on the modularity of mind. In its most restrictive form, strict modularity, this predicts that phonological processes cannot be sensitive to syntactic information and vice versa. Initial Consonant Mutation in languages such as Welsh appear to falsify the strict modularity hypothesis in relation to the morphosyn- tax–phonology interface. is is because mutation involves the triggering of a phonologically regular process in a morphosyntactically determined environment. Consider the Welsh data in (Õ): (Õ) /kaT/ ‘cat’ (a) [i XaT] ‘her cat’ (b) [i kaT] ‘his cat’ e plosive–fricative alternation in (Õ) cannot be purely phonological in nature, because the phonological environment in (Õa) and (Õb) is identical.
    [Show full text]
  • Phonetics of Southern Welsh Stress
    Phonetics of Southern Welsh Stress Zirui Liu Abstract This study is concerned with the phonetic manifestations of primary and secondary stress in Southern Welsh. The study found effects of primary stress on duration and F2 in the language. Specifically, vowels with primary stressed had longer duration and horizontally expanded vowel space when compared to unstressed vowels. Although there was significant difference in F0 between primary stressed vowels and their unstressed counterparts, F0 trajectory suggested no effect directly associated with stress. No significant effects of secondary stress were found in the present study. Due to the position shift of primary stress from the ultima to the penultimate (around the eleventh century) (Williams & Ball, 2001), vowels that occupy the ultima syllables were also included in the investigation. No clear stress related phonetic prominence was found for ultima vowels, except for word-final lengthening. Keywords: stress, prosody, Welsh, Celtic 1 Introduction Lexical stress is a suprasegmental linguistic phenomenon, which can provide rhythmic structure in a fixed-stress language such as in Turkish Kabardian and Polish (Gordon & Applebaum, 2010; Hayes, 2009). It can also encode lexical information in a language with phonemic stress, as in English and Spanish (Hayes, 2009). Earlier denotations on word level stress emphasised the strength of articulation during production of the stress bearing syllable, more specifically on ‘force’ and ‘loudness’ (Jones, 1949; Bloomfield, 1933; Bloch & Trager, 1942; Fry, 1958; Kingdon, 1958; Morton & Jassem, 1965). This paper aims to examine which phonetic measures correlate with word stress in Welsh spoken in the southern region of Wales (primary and secondary). Not much research on stress has been done in the Celtic language family.
    [Show full text]
  • Syllable Structure and Vowel-Zero Alternations in Moroccan Arabic and Berber Mohamed Lahrouchi
    Syllable structure and vowel-zero alternations in Moroccan Arabic and Berber Mohamed Lahrouchi To cite this version: Mohamed Lahrouchi. Syllable structure and vowel-zero alternations in Moroccan Arabic and Berber. The Routledge Handbook of African Linguistics, 2018. halshs-01723842 HAL Id: halshs-01723842 https://halshs.archives-ouvertes.fr/halshs-01723842 Submitted on 15 Jun 2021 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Syllable structure and vowel/zero alternations in Moroccan Arabic and Berber Mohamed Lahrouchi (CNRS & University of Paris 8) The Modern Arabic dialects of the Maghreb have three phonemic vowels /i, a, u/. These dialects are generally said to have lost the short vowels of Classical Arabic (henceforth, CA) and to have developed a short central vowel [ə], used to break up consonant clusters.1 A rapid comparison between CA and Moroccan Arabic (henceforth, MA) actually reveals that the latter lost length contrast in the vowels. In a number of items shared by these languages, there is a regular change whereby the long vowels of CA correspond to short vowels in MA, whereas short vowels in CA disappear in MA, resulting in consonant clusters often simplified by means of vowel epenthesis.
    [Show full text]
  • Brythonic “Second Lenition” Revisited: Theoretical and Typological Considerations on the Rise of Brythonic Voiceless Spirants
    Brythonic “Second Lenition” Revisited: Theoretical and typological considerations on the rise of Brythonic voiceless spirants Pavel Iosad Moscow State University [email protected] Colloquium Societatis Celto-Slavicae II Moscow, RAS Institute of Linguistics & MSU 14–16 September 2005 Abstract I consider some of the theories that have been proposed to account for the development of Proto-Celtic unvoiced stops in Brythonic. I show that the analysis involving “direct” spirantisation of doubled consonants, with no degemination, such as the one proposed in (Isaac, 2004), does not only make some wrong predictions but also rests on assumptions that appear untenable within the findings of modern phonological theory. I review the arguments put forward by, for example, Thomas (1990), and show how they must be amended to fully account for all of the problematic data. 1 The background PC unvoiced in stops in Brythonic: /p t k/ → /b d g/ intervocalically, /f T x/ after liquids and in gemination (cf. calch ‘chalk’ ← L calcem, cath ‘cat’, L cattus). Mutations going back to sandhi Welsh has nasal mutation after fy, PB *men ‘my’ but not after eu, PB *es¯on ‘their’; only voiced stops undergo nasal mutation after numerals such as naw, PB *nawan ‘nine’; spirant mutation after ei, PB *esyas ‘her’, a(c), PB *ak ‘and’, chwe, PB sweks ‘six’, tri, PB *tr¯ıs. Breton and Cornish have no systematic counterpart to nasal mutation; almost all of the above items, including o ‘their’, show evidence of having at least at some point triggered the counterpart to W spirant mutation. 1.1 The two theories • (Jackson, 1953; Martinet, 1955; Isaac, 2004) suggest that spirants evolved out of fortes geminates due to a “sharpening of expiratory force” (Jackson, 1960) or the prolongation of the continuant articulation present in aspirated consonants and the resulting loss of constriction (Isaac, 2004).
    [Show full text]
  • New Perspectives in Celtic Studies
    New Perspectives in Celtic Studies New Perspectives in Celtic Studies Edited by Aleksander Bednarski, Paweł Tomasz Czerniak and Maciej Czerniakowski New Perspectives in Celtic Studies Edited by Aleksander Bednarski, Paweł Tomasz Czerniak and Maciej Czerniakowski This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Aleksander Bednarski, Paweł Tomasz Czerniak, Maciej Czerniakowski and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7076-5 ISBN (13): 978-1-4438-7076-4 TABLE OF CONTENTS Preface ....................................................................................................... vii Part I: Language Chapter One ................................................................................................. 3 Prolegomena to a Study of Welsh Vocalism Sabine Asmus and Cormac Anderson Chapter Two .............................................................................................. 17 Diphthongs in the North of Wales Paweł Tomasz Czerniak Chapter Three ............................................................................................ 31 Translating or Mistranslating
    [Show full text]
  • Adult New Speakers of Welsh: Accent, Pronunciation and Language Experience in South Wales
    languages Article Adult New Speakers of Welsh: Accent, Pronunciation and Language Experience in South Wales Meinir Williams * and Sarah Cooper College of Arts, Humanities and Business, Bangor University, Bangor LL57 2DG, UK; [email protected] * Correspondence: [email protected] Abstract: This study examines the experiences of adult new speakers of Welsh in Wales, UK with learning pronunciation in Welsh. Questionnaire data were collected from 115 adult L2 speakers with English as an L1 located in South Wales. We investigated self-reported perceptions of accent and pronunciation as well as exploring which speech sounds were reported to be challenging for the participants. We also asked participants how traditional native speakers responded to them in the community. Perceptions of own accent and pronunciation were not rated highly for the participants. We found that speaker origin affected responses to perceptions of accent and pronunciation, as well as speaker learning level. In terms of speech sounds that are challenging, the results show that vowel length as well as the consonants absent in the L1 (English) were the most common issues reported. A range of responses from traditional native speakers were reported, including speaking more slowly, switching to English, correcting pronunciation or not responding at all. It is suggested that these results indicate that adult new speakers of Welsh face challenges with accent and pronunciation, and we discuss the implications of this for language teaching and for integration into the community. Citation: Williams, Meinir, and Keywords: second language acquisition; speech production; accent; pronunciation; new speakers; Sarah Cooper. 2021. Adult New minority language bilingualism Speakers of Welsh: Accent, Pronunciation and Language Experience in South Wales.
    [Show full text]
  • Pre-Aspiration in Welsh English: a Case Study of Aberystwyth a Thesis
    PRE-ASPIRATION IN WELSH ENGLISH: A CASE STUDY OF ABERYSTWYTH A THESIS SUBMITTED TO THE UNIVERSITY OF MANCHESTER FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF HUMANITIES 2015 MICHAELA HEJNÁ SCHOOL OF ARTS, LANGUAGES AND CULTURES List of Contents List of Contents ........................................................................................................................ 2 List of Figures .......................................................................................................................... 7 List of Tables ......................................................................................................................... 11 Abbreviations of laryngeal aspects ........................................................................................ 13 Abbreviations of oral aspects ................................................................................................. 14 Abbreviations of journals ....................................................................................................... 14 Abstract .................................................................................................................................. 15 Lay Abstract ........................................................................................................................... 16 Declaration ............................................................................................................................. 17 Copyright ..............................................................................................................................
    [Show full text]
  • Intonation in Anglesey Welsh and Uses the Patterning to Discuss an Appropriate Underlying Phonological Representation of the Intonation
    INTONATION IN ANGLESEY WELSH Sarah E. Cooper School of Linguistics and English Language Bangor University PhD 2015 Abstract This thesis investigates the intonation system of Anglesey Welsh, an under researched variety with regards to its suprasegmental aspects. The main aim of this thesis is to provide a fine-grained intonational investigation into the realisation of the structural components of the intonation system. The secondary aim considers a functional hypothesis proposed by Haan (2002), that there may be a trade-off between the amount of lexico- syntactic marking used to cue interrogativity (e.g. inversion, wh-words) and the amount of intonational marking (e.g. higher and later accent peaks). With regards to the phonetic realisation of the structural elements, two features are investigated: the vertical scaling (or height) and the temporal alignment to the segmental string. The materials were manipulated to test the effect that a) grammatical function and b) temporal constraints had on the scaling and alignment of the structural components. The results show that the scaling (height) was significantly affected by the grammatical function of the test sentences, with the structural components in questions being realised higher than statements. With regards to alignment, there was evidence of significant adjustments in the location of some structural components as a function of the number of syllables available in the sentence. The final part of this thesis discusses issues surrounding a phonological analysis of the intonation system using the scaling and alignment data. It is proposed that Anglesey Welsh questions and statements are best analysed as having a rising L+H* in both prenuclear and nuclear position, followed by a low boundary tone L%.
    [Show full text]