The intonation of Yes-No questions in Puerto Rican and other Caribbean varieties of Spanish has provoked great interest of the investigators, partly because of its unique circumflex contour that is different from the final rising contour common in other dialects (Sosa 1999; Armstrong 2010, 2012). Previous researches have also shown that in PR Spanish, the nuclear accent in echo-Yes-No questions that express surprise or ask for confirmation are represented with different tones than the information-seeking questions (Armstrong 2010). On the other hand, the Yes-No questions in Mandarin Chinese have been studied more for its syntactic variations. Two syntactic structures are believed to be alternative in the formation of Chinese information-seeking Yes-No questions: 1) the use of the sentence-final particle ma, known as the question marker and 2) the A-Not-A structure (Huang et.al, 2009). Nonetheless, little is known about syntactic variation across pragmatic contexts and the intonation of the questions (Lee 2000, 2005). The present study aims to investigate, above all, the intonational differences in Yes-No questions of four different pragmatic purposes: information-seeking, echo-surprise, confirmatory, and echo-repetition, in PR Spanish and Beijing Mandarin Chinese. It also considers any syntactic variation across the question types, especially in Chinese. Lastly, it considers the effect of different degrees of bilingualism of the participants on their intonation. For the study, an elicitation task with visual and audio guidance by means of a PowerPoint is used. The task has a Spanish section and a Chinese section. Each section consists of 20 contexts triggering Yes-No questions. Target items are divided into 4 blocks corresponding to the four contests. Spanish results show that as expected, most of the utterances were realized with falling intonation. At the same time, there are intonational differences among questions of different pragmatic contexts. Contradicting previous literature on PRS intonation (Armstrong 2010, Sosa 1999), the ´circumflex ´structure is preferred in information-seeking, confirmatory, and echo-repetition contexts, while echo-surprise context favors H*LL% final contour. In terms of bilingualism, the Spanish dominant speaker shows greater intonational variation across questions types. Some instances of rising intonation are attested probably due to influence of English or other varieties of Spanish. Chinese results show syntactic variations in the questions of different pragmatic contexts. The ma particle structure is favored in information-seeking and echo-repetition contexts, while A-not-A structure is preferred in confirmatory context. Yes-no question is scarcely found in echo-surprise context. In terms of intonation, there are effects of presence of particle and narrow focus. . / A Thesis submitted to the Department of Modern Languages and Linguistics in partial fulfillment of the Master of Arts. / Spring Semester 2017. / April 07, 2017. / Beijing Mandarin, Bilingualism, Intonation, Phonology-Syntax interface, Puerto Rican Spanish, Yes-no questions / Includes bibliographical references. / Carolina González, Professor Directing Thesis; Lara Reglero, Committee Member; Antje Muntendam, Committee Member.

In 1994, the South African democratic government declared 11 languages as official. Section (6) of the Constitution stipulates that all 11 official languages should enjoy equal status. Amongst other things, language policies, language institutions, and legislations were established to assist in implementing Section (6) of the Constitution. Provincial governments have been given the task to adopt a language policy that will ensure that the designated official languages are used, promoted, and developed equally. Mpumalanga has four designated official (provincial) languages: Afrikaans, English, isiNdebele, and SiSwati. This study aims to investigate the implementation of the South African language policy by a local municipality in Mpumalanga. The data presented in this thesis were gathered using semi-structured interviews with executive staff members as well as through questionnaires and office observations at the municipality. In addition, this thesis analysed physical municipal signage, Facebook posts by the municipality and members of the public, the official website of the municipality, and annual municipal reports. The findings indicate that there exist a variety of multilingual practices on the ground. However, these multilingual practices are mostly habitual and everyday; they are not due to the active implementation of a municipal language policy. Moreover, the findings of the study indicate that English is the preferred language at the municipality: English dominates in meetings, in written communication, on signage, official Facebook posts, and the website. The data also indicate that SiSwati has a strong presence at the municipality, but it is marginalised in other areas such as written and online communication. In addition, the data suggest that there is inequality in how the other provincial languages are used: there is only minimal presence of isiNdebele and Afrikaans. Moreover, Xitsonga, which is not a designated provincial language, has a strong presence in the municipality but no official status. This thesis argues that there is partial implementation of the language policy at the municipality. The challenges affecting the implementation includes the ideology around the use of English and the negative attitudes towards SiSwati that are present in the community. Lack of implementation is also a result of limited resources in the municipality. This thesis uses the data to formulate recommendations for the Mpumalanga provincial government. These recommendations can assist with the implementation of the language policy in government communication and the equal use of all the languages present in the municipality

Xri is a Khoekhoe language spoken among the Griekwa people of the Northern Cape, South Africa, and was thought until recently to be extinct. Fieldwork conducted in 2018 and 2019 documented Xri spoken by 27 semi-speakers and rememberers, the last speakers of this dying language. The absence of satisfactory studies of the phonetic and phonological effects of language obsolescence and death in African languages, particularly the endangered Indigenous Click Languages, necessitates further investigation. I describe the phonetic and phonological effects on Xri of language contact with Afrikaans over 170 years and critique previous studies of Xri. Innovative data collection techniques used to obtain the data are detailed. Xri phonemes not found in Afrikaans are more likely to undergo change, and the production of key classes of phonemes—such as nasal vowels—are characteristic of informants with high spoken Xri proficiency. The distinctions between click types are unstable in the speech of most informants but there is minimal loss of click realisation, and click accompaniments are resistant to change. A metric developed for measuring speaker competency is also demonstrated. Informants' spoken Xri competency is measured based on their syntactic, morphological, phonetic, and tonological performance (50%), as well as overall lexicon size (50%). Informants are divided into three group case studies by competency score, which are shown to correspond to the degree of change in their realisation of Xri phonetic features. Click sounds have persisted in the speech of even informants with low Xri proficiency, and the findings support the hypothesised salience of clicks as a phonological class. The accompaniments of click phonemes, however, displayed greater resilience to change than the click phones themselves. The contextual biographical data obtained support the linguistic assessment of the estimated date of Xri moribundity by 1960. The metric developed to measure speaker competency has—with further testing—the potential to contribute to future research in critically endangered language research. The data collection methods used for this study are also recommended for future research in situations of language death.

Originating in the Konkan coast of Maharashtra, India, a large group of people immigrated to South Africa. The migrations took place in the period of colonialism in the Cape and Natal, and sometimes extended into the later Union and Apartheid periods. These people have settled into living in Cape Town and could essentially be the largest Kokni group outside of India. Generations later, the language “Kokni” is still spoken (maintained) among the Kokni people. However, this is true for the elders, as they still maintain strong ties to India. The Kokni youngsters however have shifted away from the Kokni language as a result of the schooling system in which English and Afrikaans are dominant. Despite this, some youth still maintain various lexical items from the Kokni language in their everyday conversations in which English and Afrikaans are dominant. The study at hand set out to determine which Kokni lexicon (vocabulary) categories had survived and are still maintained among the post-shift generation of the Kokni youth, who now have English as their main language and Afrikaans as their second language. Additional emphasis was on determining whether gender, birth order, or grandparents in the home affects retention of the Kokni language lexicon (vocabulary). In order to do so, 40 Capetonians of Kokni descent, of both sexes between 18 to 35 years, born and raised in Cape Town, were recorded taking part in sociolinguistic interviews. The mixed-method approach was used to gather the background demographics and lexicon of the youth. Afterward, the data was organised and analysed using Guttman scaling; known as implicational scaling in Linguistics (Guttman, 1944; Babbie, 2011; Mesthrie, Chevalier & McLachlan, 2015). The data shows evidence of particular lexical categories being maintained more than others. Kinship terminology, typical food dishes, counting and every day vocabulary are among these aforementioned lexical categories. This confirms that the Kokni youth have shifted away from the language, toward English and Afrikaans, yet maintained some Kokni lexical categories.

This study describes the phonological structure of the Tegem language, a little-known Niger-Kordofanian language spoken by around 2000 people in Sudan. The research follows the basic linguistic theory in identifying the segments, investigating their phonotactic pattern, and identifying their functional role in meaning distinction. The study is based on lexical items collected from two Tegem language speakers via wordlists elicitation sessions. That provides the core basis for a detailed foundational description of the phonetic and phonological features of consonants, consonant sequences, vowels, syllables, and tones in Tegem. The description includes a brief account of relevant morphophonemic phenomena such as the voicing assimilation, consonant labialization and noun class sound alternations. Tegem consonants and vowels phonemes are categorized into two and three categories respectively. The consonants comprises of five obstruent and nine sonorant phonemes out of 20 phones. The vowels include four front, two central, and four back vowels phonemes out of 12 phones. Both the consonants and vowels are very common to occur in a phonologically (and morphologically) complex clusters. There are phonotactic constrains on such sequences conditioned by the environment where they occur. The study explored those sequences as bisegmental structures of adjacent segment sequences. The suprasegmental analysis found six closed and seven open syllables in Tegem where the monosyllabic lexemes of CVC and CVV are the most salient among its 13 syllable types. The research also recognized a pattern of backness (±back) vowel harmony in the disyllabic nouns and adjectives. The syllable is determined as the bearing unit of the lexical tone in Tegem, i.e. change in the syllable tone is contrastive. The lexical tones include two level tones: high (H) and low (L), and four contour tones: falling (F), rising (R), falling-rising (FR), and rising-falling (RF). The amount of the linguistic data in this study and its description form a solid foundation for further investigation of this poorly documented language.

Includes bibliographical references.

This research has been designed to investigate regional differences in a variety of South African English known as General South African English, between the White communities of Durban and Cape Town respectively. The research is socio-phonetic in its focus and therefore three variables, the PRICE, NURSE and KIT vowels of Wells’ (1982) lexical sets, were selected for acoustic analysis, a selection which was guided partly by the researcher’s own intuitions and partly by observed correlations in the impressionistic literature between certain realisations of these vowels and the Cape and Kwa-Zulu Natal as regions.

Includes abstract. / Includes bibliographical references (leaves 98-104). / The aim of this study is to investigate socio-cultural change in two prestigious school environments. It considers how female isiXhosa mother tongue speakers, who attend prestigious English schools, are undergoing changes in identity, which are mirrored in the accent of the prestigious variety of English they speak.

Includes abstract. / Includes bibliographical references. / While the language attitudes and reported behaviours of migrants have long been of interest to linguists, educationalists and sociologists, increased levels of global mobility and technological activity are changing the nature of migration. This mini-thesis considers competing paradigms of mobility including diaspora, transnationalism and super-diversity and emerges at the recognition that the shape of migration has changed considerably over the last 20 years, especially in the South African context. This new migration, characterised in this paper as a shift from diaspora to transnationalism, might have significant consequences for the way migrants conceptualise host countries and countries of origin. This study sought to investigate the language attitudes and behaviours of a family of recent Sri Lankan migrants to South Africa. The aim of the study was to describe their attitudes and reported language behaviours, and having done so, to consider whether, in theory, any of these language attitudes or behaviours might be related to longer-term language attitudes and behaviours such shift, maintenance or loss.

Includes abstract. / Includes bibliographical references. / This research contributes to the wider fields of sociophonetics and the social dialectology of English in South Africa. The study looks at three vowel sets; GOOSE, BATH and KIT taken from Wells (1982). The study was designed to identify and attempt to explain potential differences in pronunciation amongst speakers in an English-speaking community living in Cape Town and classified as 'Coloured' during apartheid. The community in question has used English as their first language for several generations and has enjoyed some of the economic advantages attached to this while at the same time being the victims (historically) of discrimination and marginalization. The study looks at the speech of twenty speakers. Using the methods of variationist sociolinguistics, it aims to investigate what correlations can be drawn between these speakers. It examines whether the speech of the informants can be correlated along lines of social class, education, personal background and occupation. In addition, the study looks (albeit briefly) at issues of language usage and social identity with regard to these twenty speakers.