This thesis presents a linguistic ethnographic study of language ideologies and identities in a multilingual, university community in Kazakhstan: a university aspiring to put Kazakhstan’s ‘Trinity of Languages’ project, aimed at developing societal tri-lingualism in Kazakh, Russian and English, into practice. Data was collected at a Kazakhstani university from 2012 to 2013, combining participant-observation and fieldnotes, audio recordings and interviews. Drawing on the concept of heteroglossia (Bakhtin 1981), the research investigates how young people draw on ideologies of separate and flexible multilingualism (Blackledge and Creese 2010) and on the often contested indexicalities of Kazakh, Russian and English linguistic resources to negotiate identities as multilingual people in Kazakhstan, particularly in contexts of performance, and stance-taking. Consideration of these ideological and linguistic resources also sheds light on Kazakhstan’s wider ‘processes of ideological transformation’ (Smagulova 2008:195) and their real-life implications for multilingual people. Furthermore, the analysis highlights how participants construct stances towards translanguaging (Garcia 2009) and suggests that acts of contextualisation, which frame interactions as being more or less ‘on-stage’ or ‘off-stage’, shape the way that speakers draw on linguistic resources and their indexical meanings, and how these contexts can afford or constrain speaker agency in the negotiation of identities.

This thesis has two aims: developing resources for Arabic dialects and improving the speech recognition of Arabic dialects. Two important components are considered: Pronunciation Dictionary (PD) and Language Model (LM). Six parts are involved, which relate to building and evaluating dialects resources and improving the performance of systems for the speech recognition of dialects. Three resources are built and evaluated: one tool and two corpora. The methodology that was used for building the multi-dialect morphology analyser involves the proposal and evaluation of linguistic and statistic bases. We obtained an overall accuracy of 94%. The dialect text corpora have four sub-dialects, with more than 50 million tokens. The multi-dialect speech corpora have 32 speech hours, which were collected from 52 participants. The resultant speech corpora have more than 67,000 speech files. The main objective is improvement in the PDs and LMs of Arabic dialects. The use of incremental methodology made it possible to check orthography and phonology rules incrementally. We were able to distinguish the rules that positively affected the PDs. The Word Error Rate (WER) improved by an accuracy of 5.3% in MSA and 5% in Levantine. Three levels of morphemes were used to improve the LMs of dialects: stem, prefix+stem and stem+suffix. We checked the three forms using two different types of LMs. Eighteen experiments are carried out on MSA, Gulf dialect and Egyptian dialect, all of which yielded positive results, showing that WERs were reduced by 0.5% to 6.8%.

This thesis is a contribution to the field of learner corpus studies. It compares a number of different measures of accuracy and complexity in second language writing, applying those measures to a sample of 308 essays written by Kurdish university students majoring in English in three schools in Iraqi Kurdistan (at two years of study: third year and fourth year). It proposes an innovative method for measuring correctness, and integrates a number of different measures of accuracy into an Integrated Approach to Achievement. It first starts by applying the method of traditional error analysis to a sample of the data collected, and then as a result the research makes recommendations for measuring ‘correctness’ instead of concentrating on the analysis of errors. It then operationalizes those recommendations, proposing an innovative method of assessment of accuracy in L2 writing by assessing ‘correctness’ as a replacement for the measurement of error using standard methods of analysis (the T-unit and clause-based correctness analysis). As a third attempt it proposes a new method of analysis that takes various units into account and hence called the various-units-based correctness analysis. After that it brings together all the measures of accuracy in a novel and integrated assessment method called an Integrated Approach to Achievement (IAA). The thesis also uses various measures of syntactic complexity, including phrasal complexity. For measuring lexical complexity a recently developed program called the Lexical Complexity Analyzer (LCA) is used. The findings are important for both English writing pedagogy and assessment.

It is widely recognised that language learners require extensive vocabulary knowledge to cope with the demands of studying content fields in English. As well as being rich in general academic and technical terms, academic discourse has been shown to make frequent use of metaphor to express abstract concepts and to achieve rhetorical goals. While research has shown the benefits of raising learners' awareness of the underlying motivation of metaphorical expressions, these findings have yet to be applied to authentic classrooms over longer periods of study. This thesis examines the effects of raising Japanese learners' awareness of metaphorical expressions in a CLIL anthropology course. It examines the written work from two groups of learners: a control group whose language instruction focussed on academic and high frequency vocabulary and an experimental group who received instruction on course-specific metaphorical themes. Variation in metaphor production is compared for the two conditions and across learner abilities, and the interaction between the frequency, dispersion and salience of metaphors in classroom input and learner output is considered. The study then investigates the influences of word frequency, part of speech, phraseology and the L1 on learner metaphor production before concluding with recommendations for pedagogic practice and further study.

Module 3 presents the culmination of Module 1 and 2 research into learning environments modeled upon constructivist and self-determinist principles (authentic learning environments), the goal of which is to develop an understanding of factors that influence Japanese learners’ perceived values about learning environments and their propensity to engage in them. The study’s more specific goals are to ascertain the values learners assign to authentic learning environments (ALEs) and the reasons why they ascribe them; to ascertain the values these learners assign to instructor and peer relationships; to ascertain the relationships that exist between the values these learners assign to ALEs and the learners’propensity for engagement; and, to bring to light what potential such knowledge might hold for educators in Japan and beyond in the attempt to develop more functional curricula for learners. As the final installment of this modular dissertation, Module 3 will present the methodology used in the study, the results of the analyses of the collected data, a discussion of the findings and implications from those analyses, and recommendations for further research.

This thesis considers how language development takes place over time by a group of 124 secondary school students of English. A series of five studies were conducted for this purpose using the tools and methods from corpus linguistics and written discourse analysis. Specifically, the thesis presents a detailed analysis of (1) how a set of function words (that, to and of) were used by these students over a 24-month period, and (2) how narrating practices concerning the structure of selected individual texts changed over time. The two distinct strands of investigation, both of which based on an inductive methodology, highlight, on the one hand, the extent to which there are common as well as unique aspects of language use observed across time and space (Francis et al., 1996, 1998) and, on the other, the role of human agency and meaning making practices in using linguistic resources over time and in shaping and constructing texts within and across individuals. Taken together, the overall inductive methodology and an emphasis on treating all instances of the conventionally labelled ‘learner language’ as equally valid features of natural human language use, show clear advantages over alternative approaches based on a deficit model.

The prevailing approach to speech recognition is Hidden Markov Modelling, which yields good performance. However, it ignores phonetics, which has the potential for going beyond the acoustic variance to provide a more abstract underlying representation. The novel approach pursued in this thesis is motivated by phonetic and phonological considerations. It is based on the notion of pseudo-articulatory representations, which are abstract and idealized accounts of articulatory activity. The original work presented here demonstrates the recovery of syllable structure information from pseudo-articulatory representations directly without resorting to statistical models of phone sequences. The work is also original in its use of syllable structures to recover phonemes. This thesis presents the three-stage syllable based, pseudo-articulatory approach in detail. Though it still has problems, this research leads to a more plausible style of automatic speech recognition and will contribute to modelling and understanding speech behaviour. Additionally, it also permits a 'multithreaded' approach combining information from different processes.

Research suggests that discourses around Muslim girls position them as having ‘wasted potential’ (Archer, 2002) and being oppressed by parental expectations around marriage (Basit, 1995/1996). In contrast, when talking about their aspirations Muslim girls themselves draw on discourses around personal choice within the bounds of parental expectations (Archer, 2002; Ahmad, 2001). This study explores the discourses used by teachers and Muslim girls, how Muslim girls are positioned within these discourses, and the implications for Muslim girls’ experiences in school. Foucauldian Discourse Analysis (as described by Willig, 2008) is used to analyse semi-structured interview data from five teachers and focus group data from five Muslim girls in year 9. Discussions around race in relation to the practice of Educational Psychologists are rare, and so the usefulness of a discursive approach to the practice of Educational Psychologists in relation to race is also discussed.

This thesis is a corpus-based study of two spoken academic corpora in English as the (foreign) medium of instruction (EMI) context. The first corpus is compiled of transcripts of academic lectures by non-native speakers (NNS) from an EMI medical college in Saudi Arabia. To compare the data, a second corpus is compiled of similar transcripts by native speakers (NS) taken from the British Academic Spoken English (BASE) corpus. The first part of the research qualitatively and quantitatively investigates the use of English discourse markers (DMs) on two levels: Structural (e.g. okay, so, because) and Interactional (e.g. okay?, I mean, any questions?). Structural DMs are found to function frequently as Topic Initiators, Topic Developers, Summarizers, and Closers, and occur more frequently in NS lectures' discourse. Interactional DMs, which function as Confirmation Checks, Rephrasers and Elicitors, are found to occur more frequently in the NNS lectures. This thesis demonstrates that the uses of DMs by the NS and NNS lecturers are affected by discourse context, pedagogic goals, personal lecturing styles, interaction with students and the need to create a conducive learning environment. The second part explores the use of Arabic discourse markers (ADMs) in the NNS lecture discourse on similar Structural and Interactional levels. Interactional ADMs (e.g. ya3ni {means}, mufhoom? {understood}) have a higher overall frequency than Structural ADMs (fa {so}, laanu {because}). The third part of this thesis explores the pedagogical functions of English-Arabic code-switching (CS) in the NNS lectures. When the purpose of CS is to make meaning clearer and convey knowledge more efficiently, it is not a language barrier but an effective communicative strategy. The data shows that CS is used mainly in seven roles in the NNS lecture discourse: (1) solidarity, (2) reiteration, (3) elaboration, (4) topic, (5) elicitation, (6) checking comprehension and (7) classroom management.

This thesis has developed out of a classroom practitioner’s desire to find out what works when trying to improve both the quality of the writing and the attainment of GCSE Business Studies students. It investigates the body of literature on writing approaches and in the context of the current educational framework in UK secondary schools considers how suggested writing interventions can be used. It describes the use of an exploratory randomised controlled trial conducted with 14-15 year-old students in English state schools who were studying business start-ups as part of a course in ‘Business Studies’. Students participating in the trial were randomly assigned to an intervention or control group within each class in each participating school. The intervention uses a ‘Story Grammar’ strategy to improve students’ reasoning by increasing the frequency and complexity of their use of ‘connectives’ such as ‘when’, ‘if’ and ‘because’. The analysis reports positive effects of the intervention on students’ understanding as judged by the use of a standard examination style mark scheme, and the number and complexity of connectives used by students in their extended writing. The design of the experiment and its practicability as a model for investigating effects of classroom interventions are discussed.