Die vorliegende Arbeit mit dem Titel „Ebíra English in Nigerian Supersystems: Inventory and Variation“ befasst sich mit einer kleinen Varietät des Nigerianischen Englisch, die für eine Untersuchung aus zwei Gründen besonders geeignet erscheint: Einerseits bin ich selbst Mitglied dieser Volksgruppe, was mir einen besonderen Zugang zu guten, aktuellen und vor allem natürlichen Sprachdaten ermöglicht. Diese sind für eine soziophonetische Untersuchung mit den Konzepten und modernen Methoden der Variationslinguistik von besonderer Bedeutung. Andererseits ist die vorliegende Arbeit keine weitere Studie über die großen Systeme des nigerianischen Englisch oder über die beiden größten und bereits relativ gut untersuchten Systeme des Yoruba-Englisch im Südwesten des Landes oder des Hausa-Englisch im Norden, sondern über eine relative kleine Gruppe dazwischen, die historisch zunächst von den Yoruba und später immer mehr von den Hausa-Sprechern beeinflusst wurde und nach wie vor beeinflusst ist. Diese empirische soziophonetische Studie stellt zwei Forschungsfragen: FF1) Welches Vokalinventar besitzt Ebíra Englisch? Diese Frage ergibt sich aus den widersprüchlichen Ergebnissen vorheriger Untersuchungen (zu Nigerianischen, Yoruba- bzw. Hausa-Englisch) und soll hier erstmals in einer Analyse von digitalen Aufnahmen von 28 jüngeren und älteren Männern und Frauen (16 bzw. 12) aus den Jahren 2014-2016 untersucht werden. Diese Aufnahmen wurden im Rahmen von soziolinguistischen Interviews gemacht, die die bekannten Sprachstile (nach Labov) umfassen: Wortliste, Lesepassage (die bewährte Kurzgeschichte The Boy who Cried Wolf mit jeweils 90 vorkommenden englischen Vokalen) und Konversation. Diese Frage ist auch vor dem Hintergrund des Einflusses der beiden nahen Hauptvarietäten Yoruba- und Hausa-Englisch interessant (FF1b). Auf der Grundlage von fast 15.000 extrahierten Vokalen erfolgte jeweils nach der sorgfältigen Aussortierung unbrauchbarer oder unvollständiger Daten eine quantitative Untersuchung mit Hilfe des Analyseinstruments PRAAT, mit dem sich die Vokalqualität in Form von Formanten messen und darstellen lässt. Die Untersuchung umfasste die bekannten Monophthongkontraste (nach Wells` lexical sets) FLEECE & KIT, FOOT & GOOSE (+ USE ), LOT & THOUGHT & STRUT , TRAP & BATH & lettER , sowie NURSE , und die relativen Diphthonge FACE , GOAT und CURE. FF2) Welche sprachlichen und sozialen Variablen können die Variation dieses Ebíra Englisch Vokalsystems erklären? Neben den bekannten sozialen Variablen Alter (bzw. Altersgruppe), Geschlecht, Mehrsprachigkeit und Bildung wurden v.a. die sprachlichen Variablen Vokaldauer, phonetische Umgebung der Vokale und Sprachstil untersucht. Interessanterweise war für eine so detaillierte Analyse der Variation die zunächst recht groß wirkende Anzahl der extrahierten Vokale nicht in jedem Fall groß genug oder nicht gut genug verteilt.

The variety of middle-class speakers in St. John’s conforms to some degree to mainland Canadian-English pronunciation norms, but in complex and distinctive ways (Clarke, 1985, 1991, 2010; D’Arcy, 2005; Hollett, 2006). One as yet unresolved question is whether speakers of this variety participate in the Canadian Shift (cf. Clarke, 2012; Chambers, 2012), a chain shift of the lax front vowels that has been confirmed for many different regions of Canada (e.g. Roeder and Gardner, 2013, for Thunder Bay and Toronto, Sadlier-Brown and Tamminga, 2008, for Halifax and Vancouver). While acoustic phonetic analyses of St. John’s English are rare, some claims have been made that urban St. John’s speakers do not participate in the shift, based on two or six speakers (Labov, Ash & Boberg, 2006; Boberg 2010). Other researchers with larger data sets suggest that younger St. John’s speakers participate in mainland Canadians innovations to different degrees than mainlanders (e.g. Hollett, 2006). The Canadian Shift has not been uniformly defined, but agreement exists that with the low-back merger in place, BATH/TRAP retracts and consequently DRESS lowers. Clarke et al. (1995), unlike Labov et al. (2006), assert that KIT is subsequently lowered. Boberg (2005, 2010), however, emphasizes retraction of KIT and DRESS and suggests unrelated parallel shifts instead. In this PhD thesis, I demonstrate the presence of the Canadian Shift in St. John’s, NL, conforming to Clarke et al.’s (1995) original proposal. In my stratified randomly-sampled data (approx. 10,000 vowels, 34 interviewees, stratified as to age, gender, socioeconomic status, and “local-ness”), results from Euclidean distance measures, correlation coefficients, and linear, as well as logistic, mixed-effects regression show that (1) young St. John’s speakers clearly participate in the shift; and that (2) age has the strongest and a linear effect. Continuous modeling of age yields even more significant results for participation in a classic chain shift (6% decrease in lowering per added year). My findings also confirm that the change seems to have entered the system via formal styles (cf. Clarke, 1991, 2010, for TRAP in St. John’s). Traditionally, the linguistic homogeneity on a phonetic level of the Canadian middle class has been explained by Canada’s settlement and migration patterns of the North American Loyalists from Ontario to the west (cf. Chambers, 2009). Newfoundland’s settlement is distinct, in that the British and the Irish were the only two relevant sources. If settlement were the only crucial reason for a shared pronunciation of Canada’s middle class from Vancouver to St. John’s, the Canadian Shift should be absent in the latter region. I suggest three reasons for middle-class St. John’s’ participation in the Canadian Shift: 1) Newfoundland’s 300-year-old rural-urban divide as a result of its isolation, through which British/Irish features are attributed to rural und lower social class speakers; 2) the development of the oil industry since the 1990’s, through which social networks changed according to the perception of social distance/closeness; and 3) the importance of the linguistic marketplace, which is high in St. John’s due to 1) and 2).:List of Tables viii List of Figures x 0 Prologue – Variationist Sociolinguistics 1 1 Introduction 27 2 English-speaking Canada and its Vowel Shifts 31 3 Newfoundland and its Englishes 77 4 Data and Methodology 107 5 Analysis and Discussion 243 6 Conclusion 363 Bibliography 375 Appendices 409 A Interview Questionnaire 409 B Normality Tests per Speaker and Age Group 423 C Vowel Plot of Median Formant Values 433 D Results for the Assumptions of T-tests 435 E Results from Decision Trees and Optimal Binning 439 F Results from Regression Analyses 449 G Résumé 457 H Deutsche Zusammenfassung der Dissertation 461 I Eidestattliche Erklärung zur Eigenständigkeit 469

Word stress is crucial for language perception and production as it enables lexical access and speech segmentation. Phonetically, word stress is cued through higher pitch, duration, and intensity. These cues are weighted differently in production and perception by language learners from different backgrounds. Slavic English learners, especially with a fixed-stress West Slavic language background like Czech but also with a variable-stress South Slavic background like Bulgarian face difficulties in word stress acquisition. Their stress placement errors cannot be fully explained by direct transfer and highlight a more complex relation between the target and the native language. This thesis thus aims to explore the perception and production of English word stress cues by speakers of Slavonic Englishes based on evidence from an EEG study on Slavic and German English word stress cue perception and a phonological and phonetic study on Czech English (CzE) features. These findings are used for the conceptualization of a perception and production training tool. First, the concepts of cue and feature are reviewed in an effort to bridge psycho- and sociolinguistics and create credible language stimuli. Then, the Event-Related Potentials study demonstrates that Slavic and German English speakers successfully process first- and second-syllable stress changes cued through pitch, duration, and intensity. Still, speakers of Slavic languages (with a rather syllable-timed rhythm) perceived first and second syllable stress similarly whereas speakers of German (with a rather stress-timed rhythm) perceived the second syllable deviation as more salient. A study on the role of word stress in the CzE feature inventory shows that syllable-timed rhythm also affects production, as speakers lack vowel reduction and add additional stress to longer words. Compared to Bulgarian English (BgE), CzE shows differences in stress features but an overall lack of consistency in their use to differentiate primary from secondary and unstressed syllables. These findings are integrated in the conceptualization of a High-Variability Phonetic Training (HVPT) system featuring segmentals and suprasegmentals. Overall, this thesis finds group differences in foreign language perception and production of word stress and applies them in the development of customized language learning solutions.:List of Figures ix List of Tables xiii List of Abbreviations xv Acknowledgements xvii 1 Introduction 1 1.1 Issue: Perception and production of English word stress cues 1 1.2 Language selection 5 1.3 Association 8 1.4 Research questions and outline 10 2 Background 13 2.1 Word stress 13 2.1.1 Assignment 17 2.1.2 Shifting 23 2.1.3 Rhythm 25 2.2 Word stress perception and production 26 2.2.1 Word stress cues and features 27 2.2.2 Sensitivity to word stress 30 2.2.3 Word stress acquisition 31 2.3 Methods for studying word stress perception and production 35 2.3.1 Behavioral methods 35 2.3.2 Psychophysiological methods 36 2.3.3 Phonological and phonetic methods 47 3 From Cues to Features 49 3.1 Introduction 50 3.2 The concepts of cue and feature in Psycho- and Sociolinguistics 51 3.2.1 Cue 51 3.2.2 Feature 55 3.2.3 From features to cues, from cues to features 59 3.3 The need for non-native experimental stimuli in Psycholinguistics 59 3.3.1 The validity of stimuli representing language cues and features in experimental designs 61 3.3.2 Mirroring features of non-native English 61 3.4 Considerations in the design of non-native experimental stimuli in Psycholinguistics 63 3.4.1 Methodological challenges 63 3.4.2 Using cues to represent features 64 3.5 Implications for TESOL 66 3.6 Conclusion 67 4 ERP Evidence for Slavic and German Word Stress Cue Sensitivity in English 69 4.1 Introduction 70 4.2 Materials and methods 78 4.2.1 Stimuli and paradigm 78 4.2.2 Participants 81 4.2.3 Experimental procedure 82 4.3 Results 84 4.4 Discussion 89 4.4.1 Slavic and German word stress sensitivity in English 89 4.4.2 Applications in language technologies and learning 97 4.5 Conclusion 98 5 Czech English Word Stress Features 101 5.1 Introduction 102 5.2 Methodology 105 5.2.1 Participants 105 5.2.2 Interview schedule 105 5.2.3 Transcription and data analysis 106 5.3 Results and Discussion 109 5.3.1 Czech English features 109 5.3.2 Word stress features 120 5.4 Applications 125 5.5 Conclusion 132 6 Training English Word Stress Perception and Production with Technology 135 6.1 Introduction 136 6.2 Background 137 6.2.1 Teaching word stress 137 6.2.2 Technology in word stress teaching 139 6.3 HVPT tool concept 141 6.4 Discussion 144 6.5 Conclusion 145 7 General Conclusion 147 7.1 Summary and conclusions 147 7.2 Contributions to the field 150 7.3 Limitations and outlook 153 References 157 A Appendix 189 A.1 Data availability 189 A.2 German summary 189 A.3 Supplementary materials 192 A.4 Eigenständigkeitserklärung 200

The variety of middle-class speakers in St. John’s conforms to some degree to mainland Canadian-English pronunciation norms, but in complex and distinctive ways (Clarke, 1985, 1991, 2010; D’Arcy, 2005; Hollett, 2006). One as yet unresolved question is whether speakers of this variety participate in the Canadian Shift (cf. Clarke, 2012; Chambers, 2012), a chain shift of the lax front vowels that has been confirmed for many different regions of Canada (e.g. Roeder and Gardner, 2013, for Thunder Bay and Toronto, Sadlier-Brown and Tamminga, 2008, for Halifax and Vancouver). While acoustic phonetic analyses of St. John’s English are rare, some claims have been made that urban St. John’s speakers do not participate in the shift, based on two or six speakers (Labov, Ash & Boberg, 2006; Boberg 2010). Other researchers with larger data sets suggest that younger St. John’s speakers participate in mainland Canadians innovations to different degrees than mainlanders (e.g. Hollett, 2006). The Canadian Shift has not been uniformly defined, but agreement exists that with the low-back merger in place, BATH/TRAP retracts and consequently DRESS lowers. Clarke et al. (1995), unlike Labov et al. (2006), assert that KIT is subsequently lowered. Boberg (2005, 2010), however, emphasizes retraction of KIT and DRESS and suggests unrelated parallel shifts instead. In this PhD thesis, I demonstrate the presence of the Canadian Shift in St. John’s, NL, conforming to Clarke et al.’s (1995) original proposal. In my stratified randomly-sampled data (approx. 10,000 vowels, 34 interviewees, stratified as to age, gender, socioeconomic status, and “local-ness”), results from Euclidean distance measures, correlation coefficients, and linear, as well as logistic, mixed-effects regression show that (1) young St. John’s speakers clearly participate in the shift; and that (2) age has the strongest and a linear effect. Continuous modeling of age yields even more significant results for participation in a classic chain shift (6% decrease in lowering per added year). My findings also confirm that the change seems to have entered the system via formal styles (cf. Clarke, 1991, 2010, for TRAP in St. John’s). Traditionally, the linguistic homogeneity on a phonetic level of the Canadian middle class has been explained by Canada’s settlement and migration patterns of the North American Loyalists from Ontario to the west (cf. Chambers, 2009). Newfoundland’s settlement is distinct, in that the British and the Irish were the only two relevant sources. If settlement were the only crucial reason for a shared pronunciation of Canada’s middle class from Vancouver to St. John’s, the Canadian Shift should be absent in the latter region. I suggest three reasons for middle-class St. John’s’ participation in the Canadian Shift: 1) Newfoundland’s 300-year-old rural-urban divide as a result of its isolation, through which British/Irish features are attributed to rural und lower social class speakers; 2) the development of the oil industry since the 1990’s, through which social networks changed according to the perception of social distance/closeness; and 3) the importance of the linguistic marketplace, which is high in St. John’s due to 1) and 2).