Search

Intelligibility refers to a targeted pronunciation level in English which enables non-native English speakers to produce and understand English speech uttered by both native and non-native English speakers (Abercrombie, 1949; Gimson, 2001; Cruttenden, 2014; Munro and Derwing, 2006; Levis, 2016). Instead of pursuing perfect mastery of English pronunciation, most researchers recommend intelligibility as an achievable and practical pronunciation goal (Gimson, 2001; Quirk, 1990; Jenkins, 2000;James, 2014). Although intelligibility is currently the focus of pronunciation studies and classroom instructions, it has not been applied in the Iraqi EFL classrooms and pronunciation research (see Al- Juwari, 1997; Ahmed, 2000; Mahud, 1998; Rashid, 2009; Khudhair, 2015; AlAbdely and Thai, 2016; Al-Owaidi, 2017). The theoretical assumption of the study is that an intelligibility level of universal validity for EFL learners is best achieved when speech performance in English is based on a native English speakers’ pronunciation model, namely Gimson’s (2001) Minimum General Intelligibility (MGI). Applying a mixed methods approach, the present study investigates the productive and perceptive intelligibility of Iraqi EFL learners in relation to foreign accent and accent familiarity. Productive intelligibility refers to learners’ English speech being understood by others. This is determined by a production intelligibility test. By contrast, perceptive intelligibility refers to the ability of learners to understand native and non-native English speech. This is determined by a perception intelligibility test. The study measures the above two aspects of intelligibility, identifies which aspects of foreign accent and accent familiarity most negatively affect intelligibility and determines the various strategies these learners use to overcome intelligibility failure. The overall quantitative analysis shows that Iraqi EFL learners are intelligible at the speech production and perception levels. However, these two aspects of intelligibility are negatively affected by the existence of segmental phonemes in English and Arabic that have no counterpart in the other language and by unfamiliarity with the speaker’s accent. The qualitative analysis identifies several segmental phonemic contrasts of a high functional load which are responsible for intelligibility failure and a list of strategies which the Iraqi EFL learners employ to overcome these failures. Based on the above findings and the nature of the pronunciation problems involved, the study suggests an intelligibility approach to the teaching of pronunciation for Iraqi EFL classrooms. The study concludes with a description of the research implications and applications that derive from the findings of the study.

This study aims to provide a better understanding of the similarities and differences of writing for two different journal contexts. The focus of the study is on the construction of Discussion Sections in research articles and in particular, an investigation of the similarities and differences in the use of rhetorical moves, citations and engagement in research articles in Indonesia-based Journals and International Journals publishing in the field of Applied Linguistics. Bringing together two established and complementary frameworks of genre analysis (Swales 1990, 2004) and the engagement system of Martin and White’s (2005) appraisal framework, the research provides a textual analysis of fifty Discussion Sections from published research articles taken from the two publication contexts. The overall aim of this study is to provide research-based evidence to assist in EAP instruction particularly in the teaching of English for Research Publication Purposes (ERPP), a growing sub-field within Applied Linguistics. Move analysis shows that Move 2: reporting results and Move 4: commenting on results are ‘the heart’ of Discussion Sections in both corpora. A higher number of Step options used in Move 4: commenting on results indicates that the main communicative focus of research article Discussion Sections in both Indonesian local and international publication contexts is about on commenting on the results. In terms of sequencing moves, this research shows that moves are not organised linearly. Citations analysis shows a dominant use of integral citations especially verb-controlling type in the Indonesian local corpus. It is suggested that this citation type requires less demand on synthesising various sources cited while Discussion Sections in the International corpus make greater use of non-integral citations which indicates a succinct synthesis of various sources. In terms of function, referring to literature is the most salient citation function in the Indonesian local corpus while attribution is the most dominant one found in the International corpus. To build more dialogue with readers especially when writers commenting on results, the research article Discussion Sections in the International corpus mobilise heteroglossic resource through the use of ‘contraction’ and ‘expansion’ features that create more space for refutation and build a writer-reader relationship. Conversely, Discussion Sections in the Indonesian local corpus deploy more ‘contraction’ resources that limit other external voices. To complement the textual analysis, Indonesian early-career academics as part of the community of the discipline and are working in the field of context were interviewed. Their perspectives suggest that it is important for results of genre analysis studies to be transformed into teaching materials and methods to enable novice writers to better understand research article writing conventions in the field of Applied Linguistics.

This thesis asks: what is the division of labour between the syntax and the semantics? The empirical focus is on the phenomenon of clausal embedding, whereby the grammar provides the resources to embed a clause within another clause, and the semantics provides the resources to represent an individual’s mental representations. The primary goal is to argue that that-clauses denote predicates of contentful entities – abstract objects, such as propositions, facts, and rumours. The major theoretical claim is that that-clauses function quite generally as modifiers in the compositional semantics, both when they compose with nominals and verbs. In order to cash this idea out, a strictly neo-Davidsonian approach to the syntax-semantics interface is outlined. In the syntax, arguments are severed from the verb; rather, they are incorporated as specifiers of functional heads. This is paralleled by a neo-Davidsonian semantics, where verbs denote predicates of eventualities, and thematic arguments are incorporated via metalanguage functions. Consequently all verbs, including attitude verbs, are argued to simply denote predicates of eventualities. Embedded clauses compose with attitude verbs as intersective modifiers – they specify the content of the verb’s eventuality argument.

Sperber and Wilson (1995:222) posit the term poetic effect for the peculiar effect of an utterance which achieves most of its relevance through a wide array of weak implicatures. Crucially, the input to pragmatic processing, which prompts the derivation of a poetic effect, is achieved via some stylistically pronounced linguistic feature: for example, a repeated lexical item, a peculiar syntactic form, a piece of alliteration, and so on. However, what has never been considered to any great depth from a relevancetheoretic perspective is how unusual elements of visuospatial form might also impact upon the reader’s basic understanding and wider interpretation of a given poetic text in ways that result in the derivation of specialised poetic effects. Therefore, the thesis posits a relevance-theoretic account of the cognitive-pragmatic effects of short linelength and line divisions, when employed and interpreted within complex forms of poetry. The account is split into two hypotheses relating to short line-length and line divisions respectively. Hypothesis 1 states that the use of short line-length leads to the majority of the text’s lexical material being perceived in a much slower, and therefore intense fashion, which consequently causes the lexical and encyclopaedic entries that such material links to within the mind to remain active for relatively longer periods of time. During such extended periods of lexical and encyclopaedic activation, literary readers are encouraged to inferentially process the text’s explicit-propositional content in relation to a range of further items of encyclopaedic-contextual material, which can give rise to arrays of additional contextual effects of a weakly implicit and therefore poetic nature. Hypothesis 2 states that line divisions are often intentionally utilised in poetic texts by writers in order to visuospatially separate integral syntactic units upon the page. This can encourage readers to pause and briefly consider, upon an anticipatoryhypothetical basis, the various possible pragmatic extensions of the text’s momentarily incomplete logical and propositional status, pre-line division as it were. The various pragmatic extensions may be formulated as arrays of weak explicatures, which for some readers may achieve poetic effects (in the specialised relevance-theoretic sense of the term). The process effectively constitutes the visuospatial equivalent of a deliberate ‘pause for effect’, which triggers a considerable degree of further inferential processing, and provides a distinct communicational ‘reward’ primarily at an explicit-propositional level.

This thesis describes the development of part-of-speech tagging resources for the Georgian language, consisting of i.) a new morphosyntactic language model for part-of-speech (POS) tagging purposes; ii.) tagging guidelines for tagging and post-editing; iii.) the KATAG tagset and iv.) the trained parameter files the probabilistic TreeTagger program needs to work on Georgian texts. A new morphosyntactic model of Georgian for part-of-speech tagging purposes is described in the thesis. The thesis also describes a tagset (KATAG) defined in accordance with a new morphosyntactic model of the language and a set of design principles and tagging guidelines. A stochastic methodology is used here to perform tagging in Georgian. Namely, the Treetagger - a probabilistic part-of-speech tagging program has been trained on Georgian texts. The justification for this choice is discussed. I use two tokenisation approaches in part-of-speech tagging. An accuracy of 92.41% using an enclitic tokenisation approach and accuracy of 87.13% was achieved using a non-enclitic tokenisation approach, corroborating my hypothesis that treating enclitic elements separately from the host words results in better tagging performance. To make the tagger program easily adaptable for a range of inputs (type, variety or genre of text), the performance of the probabilistic TreeTagger program was evaluated according to the obtained test set consisting of five different genres such as academic, informal, legal, fiction and news.

In recent years, research in Linguistic Landscape Studies (LLS) has demonstrated increasing interest in people and practice around signs, investigating signs through the lens of everyday activities such as walking and tattooing (see e.g. Lou 2016, 2017; Peck and Stroud 2015; Stroud and Jegels 2014). While the field already provides an interactional framework for analysing here-and-now action around signs - geosemiotics (Scollon and Scollon 2003) - there are empirical steps still to go to capture and interrogate the social interactions in which the everyday formation and experience of Linguistic Landscape (LL) are embedded. This thesis focuses on struggles around signs in the Bahnhofsviertel, Frankfurt. Based on 6 months of ethnographic fieldwork in the Bahnhofsviertel, it provides a panorama of processes of social change in the district and their manifestation in public signage, including demographic shifts and tensions and hostility regarding perceived gentrification. It then focuses in on a particular aspect of the scene - struggles around stickers. Putting up stickers in public space is a civil offence in Frankfurt and, as part of a broader urban development plan for the Bahnhofsviertel, local authorities have been cleaning away stickers from public fixtures in the district since 2014. The thesis generates an interactional perspective on LL focused on these struggles around stickers. Observations of stickerers, their metacommentary, and official documentation on signage are drawn together in an analysis guided by Interactional Sociolinguistic theory and Goffmanian interactional theory. The analysis focuses on two main dimensions of struggles around stickers: the motivations behind stickering and sticker removal; and sticker emplacing as an interactional accomplishment. With regard to the first dimension, the stickerers and the local authorities are found to have two separate kinds of main concern: the stickerers want to communicate social, political and commercial messages and bring them to the foreground of people's attention, while the local authorities want to generate a particular environmental style and congenial background to people's everyday tasks. The incompatibilities between the two parties and their projects lie in their differing visions of how people move through space and cognise the material environment. With regard to the second dimension, sticker emplacing is found to be guided by two main interactional concerns. On the one hand, stickerers choose a time and place for their messages, and monitor their stickers once in place, in anticipation of the moves of cleaners and political/commercial rivals. On the other hand, in the moment of emplacing, stickerers orient to different moral accountabilities, including the possible activation of 'remedial interchange' (Goffman 1971) by CCTV operators and other observers. The data and analysis have implications for the concepts of 'transgression' and 'emplacement' in LLS. Approaching the emplacing of stickers as interactional activity points to various contingencies influencing how, where, and by whom signs get put up in public space that a post hoc interpretation of a sign's location would miss. Moreover, it is shown that transgression is not a clear-cut distinction to be applied uniformly to signs like stickers, rather transgressiveness is decided in interaction, in the ritual work performed by social actors.

Describing the numerous factors that constrain and promote particular aspects of linguistic behaviour in interaction is very difficult. The recent adoption of more advanced quantitative methods has enhanced this modelling, leading to a greater understanding of linguistic patterns. At the same time, the increase in availability of digital recordings and storage capacity for such recordings is leading to increasingly large corpora of complex linguistic data for such investigations. The Sounds of the City corpus is one such example and is the corpus we model throughout this thesis. The corpus is an electronic real-time corpus of Glaswegian vernacular, which consists of a searchable, multi layered database of 58 hours of recordings from 136 speakers, recorded between 1970 and 2010 with orthographic transcripts and automatically phonemically segmented waveforms, amenable to automatic acoustic analyses of durational and resonance characteristics of speech. Vowel formant measurements provide a numeric representation of a spoken vowel and are a commonly used metric to measure linguistic variation and change, with each vowel having multiple formant measures, which correspond to the resonances of the vocal tract. The first three vowel formants are important perceptual cues for the successful recognition of vowel qualities. Current quantitative modelling methods consider each formant separately, inferring characteristics on each formant measurement assuming independence between each formant. This assumption for most vowels seems misplaced, as formant measures are often correlated with one another. In this thesis, we extend upon current modelling techniques applied to sociolinguistic corpora by introducing a Bayesian hierarchical model which models the first three formant measures for each vowel simultaneously, taking into consideration the correlation present between such measures. We also implement reparameterisation methods to alleviate issues caused by highly correlated samples, which is often observed in MCMC output for models applied to datasets with nested structures, a common feature in sociolinguistic corpora. These models not only account for the complex nested structure of the data and uncover the underlying dynamics of language just like classical mixed effects models, but now additionally account for the correlation between formants, providing a more accurate representation of factors driving linguistic variation and change. The output from the Bayesian hierarchical model is visualised as a graphical model. Graphical models provide a visual representation of the conditional dependence between variables, making them an attractive inference tool. We combine the hierarchical model and jointly infer the relationship between vowel formant measurements using the precision estimates from the hierarchical model as input to a Bayesian Gaussian graphical model. The resulting graph utilises a chain graph like structure which visually informs the user which factors have a significant effect on vowel variation, corresponding to each formant, and also the relationship present between the first three formants. This novel inference tool helps to aid the understanding of complex model output much like the ones fitted to the Sounds of the City corpus, though can easily be applied to numerous modelling problems.

It has been more than ten years since the official 2005 publication of the Lexical Priming Thoery (LPT), when various applications and expansions have been made to advance this theory in many dimensions. The majority of this kind of study has been done in English, while little research has been done in Mandarin Chinese—a genetically different language from English, with a notable exception of Hoey and Shao (2015), who addressed issue by undertaking a preliminary study of several notions as defined in LPT. In order to fill in this gap, this thesis explores the fundamental concepts noted in LPT, namely collocation, colligation and semantic association, to provide more evidence of LPT's ability to explain the linguistic features of Chinese and, based on that evidence, to determine whether LPT can be further advanced or expanded to address the characteristics of Chinese. Furthermore, we attempt to expand LPT's dimensions in the field by comparing the lexical and syntactical similarities in and differences between English and Chinese. By examining the evidence presented by the given nodes (world and ?? shi4jie4) and the clusters containing the nodes, we aim to assess the potential of applying LPT in cross-linguistic studies of English and Chinese. A holistic analysis of the nodes and phraseology units containing the nodes show that the three concepts as defined in LPT are also applicable to Chinese. With respect of collocation, we have found that collocation of Mandarin Chinese operates on both single characters and nestings. With respect of colligation, we have found that there is an interaction between the node items' inter- and intra-colligation. The colligation is primed in an interaction with collocation and semantic association. The parameters interchange with and are dependent on each other. The interaction between collocation, colligation and semantic association operates on ?? shi4jie4 implies that the relation between grammar and lexis of Chinese may need to be re-considered as well. First, the "borrowed" grammatical class of "word" needs to be re-defined, especially in a cross-linguistic study between English and Chinese. The English definition of word are not tied in with Chinese due to the latter's distinctive morphological features. We propose, with evidence found for ?? shi4jie4 that nesting as defined in LPT could be applied as a basic corresponding pattern and departure point for cross-linguistic study between English and Chinese. Second, we propose a new approach of describing Chinese grammar after testing the validity of LPT for Mandarin Chinese that the grammatical category a Chinese character and nesting is assigned shall be an outcome of the character or nesting's grammatical primings. The grammar of Chinese is an accumulation and interweaving of the primings of a character or a nesting's collocation, colligation and semantic association. Components of nestings are primed to combine in certain way (order), to occur at certain position within a nominal group, and to serve a certain function in a sentence. By comparing the differences and similarities exhibited by cross-linguistic equivalence (world and ?? shi4jie4), we discovered that there is no absolute "similarity" or "differences" regarding the lexical and syntactical priming features between the two languages. The divergence lies in the congruence. We could not restrict the cross-linguistic study to one unit, rather, we ought to analyse the features from the perspective of co-textual interaction. A more abstract unit, for example, the semantic association needs to be considered.

Previous studies have used a genre approach and move-based analysis as a tool for unpacking discourse organisation of text. Likewise, corpus-based approaches, as a bottom-up method, have been applied to investigate vocabulary patterns, their distribution and typical linguistic characteristics within the text. These two approaches are considered complementary, offering insights into the discourse patterns of a genre. Few studies, however, have attempted to combine these two methodological perspectives and then develop a pedagogic intervention based on the description. This study integrated move-based and phraseological approaches with the specific objectives of scrutinising the macro- and micro-structure of applied linguistic research articles written in English and indexed in the Thai Citation Index (TCI) database, along with investigating Thai novice researchers’ and graduate students’ perceptions of article writing for publication and responses to a workshop. The research comprised two constitutive strands: one descriptive, and one interventionist. The descriptive phase began with developing a move-based coding protocol. I then examined the four conventional sections of 50 research articles (RAs) to identify the ways in which the discourse of these texts was organised. A corpus-based approach, with qualitative support, was then applied to extract pedagogically interesting n-grams: functional n- grams and content-based n-grams, from a dataset of 110 RA texts (the original group of 50 plus an additional 60). Concerning the interventionist strand, I carried out workshops applying the knowledge obtained from the descriptive phase to raise Thai novice scholars’ and graduate students’ awareness vis-à-vis article writing. Data regarding perceptions of this process were elicited through semi-structured interviews with six faculty members and two graduate students.

In order to be an academic researcher, one needs to be recognised for having expertise in a certain field or discipline. This process of positioning oneself in a field and being recognised for it can be a confounding one that arguably most researchers have to go through. What are the discursive and social practices that researchers engage in so as to be recognised as an academic in a certain field or discipline? The dilemmas or struggles that researchers faced in the process of positioning and establishing themselves as academics are examined from a discursive perspective. The data came from twenty-seven qualitative interviews with academics, ranging from early-career researchers to Professors Emeriti, who work in universities in the UK. As a qualitative case study of researchers in applied linguistics and related fields, it is inspired by ethnographic studies about identity construction in its exploration of how researchers construct identities for themselves or position themselves as applied linguists and their struggles by drawing upon. Informed by Bakhtinian notions of polyphony and positioning theory, it argued that researchers self-position and resisted being positioned through dialogic utterances and voicing. This thesis conceptualised academic struggles as enacted through discursive acts between interview participants (interviewer and respondent) and made sense of by drawing upon tacit, shared or sometimes unshared knowledge about academia. It proposed a model to show how social practices (such as academic struggles) are mediated through language (utterances and discursive acts) and perpetuated in a cyclical and continuous process. The thesis analyzed discursive acts and pragmatic resources such as voicing and humour to demonstrate how struggles are enacted through a negotiation of understanding and positioning. Through the process of understanding and accounting for these struggles in the interview, interlocutors evoked their beliefs, assumptions and ideas about which aspects of academia are more valued or less valued by fellow researchers, institutions and other stakeholders in higher education. These beliefs were often tacit knowledge and hence understood as discourses about academia. The thesis contributed insights into researchers' and institution's valuation beliefs of what constitute 'good' research outcomes, preferred kinds of impact, valued publishing practices and conducive research environments. The thesis argued that the struggles faced by xi researchers often emerged from incongruences between what individual researchers aspire toward and what institutions value.

Building good relationships at work is crucial for individual wellbeing and workplace satisfaction. Yet, managing these relationships is far from easy and concerns over relationship management and getting along well are integral to people’s daily lives (Knobloch, 2010). Research indicates that good relationships are particularly scarce in intercultural teams (e.g. Mannix & Neale, 2005; Stahl, Maznevski, Voigt, & Jonsen, 2010), yet little research has been undertaken to investigate just how it is that team members and colleagues get along and how they relate around working together. To address this gap, this thesis draws on 25 hours of interactional data from the meetings of an intercultural team of MBA students recorded over 8 months. Analysis of the transcripts is supplemented by observations and interviews conducted with team members at the start and end of their teamwork. The study investigates one talk activity in depth, troubles talk, that is demonstrated to have played a central role in relating in the team. It explores how rapport management (Spencer-Oatey, 2008) is done in troubles talk across different domains and provides a thick description of troubles talk itself. It also explores the functions troubles talk seems to fulfil for relating in the team including building common ground, a shared perspective, shared norms, empathy, solidarity and trust, team member identity and group mood, in addition to supporting team member coping. The findings reveal that these functions are realised through a number of different strategies that can be used in troubles talk, including: (reciprocal) self-disclosures, troubles humour, swearing, commiserating and developing shared narratives. Troubles talk as such appears as a kind of super-strategy (cf. Brown & Levinson, 1987) in which many other strategies for relating can be embedded, and which seem less permissible in other types of talk. The thesis thus advances our understanding of relational strategies and practices around relating especially in workplace contexts and of troubles talk, a seemingly ubiquitous everyday talk phenomenon. It concludes by proposing some theoretical developments around relating and rapport management and offering recommendations for future research.

Cross-border collaboration has become the rule for a growing number of enterprises in post-modern societies. In this context, constructing trust is directly related to establishing and maintaining strategic partnerships and alliances. This PhD project investigates the development and performance of trust between Greek and Turkish business partners in a cross-border collaboration setting. Trust is understood as a dynamic construct that operates mainly in the interactional order. It is perceived as a situated and negotiated accomplishment, something partners do in interaction either in institutional settings or in their everyday personal lives (Candlin and Crichton, 2013; Pelsmaekers, Jacobs and Rollo, 2014). The project adopts an ethnographic approach and seeks to capture the ways in which trustworthiness is understood, warranted and performed by participants. The study attends to the plea for more qualitative studies on trust and contributes to literature concerned with how trust develops and evolves over time in cross-border collaboration settings. The research data include 56 hours of semi-structured ethnographic interviews with business partners, and audio and video recordings of natural interaction, including formal meetings, dinners, visits and everyday talk. The data are analysed from an interactional, sociolinguistic perspective. Trust development in the data is intrinsically linked to perceptions of trustworthiness. The findings suggest that trustworthiness perceptions depend on the perceived expertise and morality of partners, which are then affected by social relationships, shared group membership and the interactional histories that these generate. Shared group membership emerges as particularly significant in the context of a specific case in the data. The analysis shows that by identifying with each other on the basis of shared attributes, in this case regional identity, partners can develop affinities and close personal relationships that facilitate the development of trust. The thesis offers insights also with regard to how trust is enacted in workplace discourse. The findings suggest that negotiation of business decisions during meeting talk constitutes a good arena for the study of trust, as it provides business partners with the opportunity to claim trustworthy personas for themselves through relevant positioning. The work adds to the literature on organisational trust from a discourse analytical perspective and its findings are relevant both to workplace discourse scholars and practitioners interested in workplace relationships.

Problem solving in the workplace is a high stakes activity that has important implications for organisations and individuals. Although problem-solving has attracted a lot of attention, there is little research on the interactional aspects of problem-solving processes. There are also few attempts to make interactional research relevant to business studies on organisational problems. In this thesis, I adopt a discursive approach and understand problems as locally produced in and through interaction. I draw on one case study and discuss data collected in one multinational company. Through interviews and interactional data, the thesis investigates how organisational ‘problems’ are constructed (and occasionally solved). The discussion focuses on the discursive resources employees draw upon in formulating problems and their professional roles and responsibilities negotiated in problem solving meetings. Special attention is paid to the meeting event and the HQ-subsidiary context in relation to the ways in which issues are negotiated and ratified as ‘problems’. The findings show that in interviews, language and culture, as abstract concepts, constitute key resources for the construction of organisational problems. In participants’ talk about language and culture, the individuals’ ideologies are enacted, and the abstract concepts become critical means for positioning (them)selves and others in the organisational setting. In problem-solving meetings, I focus on interactional activities which emerge in the timeframe of the meeting event. The patterns of these activities are not linear but shaped by the participants’ interactional and institutional positioning through which their professional roles are enacted. I argue that the processes of constructing organisational problems are contingent on the ability of individuals to access and challenge dominant institutional and professional discourses and ideologies. Based on the analysis, I propose a model that visualises interactional moves that constitute problem-solving activities in the meeting event. I conclude the discussion of the data by also proposing a model on the HQ-subsidiary relationships as emergent in the problem-solving interaction.

Languages are culturally transmitted through a repeated cycle of learning and communicative interaction. These two aspects of cultural transmission impose (at least) three interacting pressures that can shape the evolution of linguistic structure: a pressure for learnability, a pressure for expressivity, and a pressure for coordination amongst users in a linguistic community. This thesis considers how these sometimes competing pressures impact linguistic complexity across cultural time. Using artificial language and iterated learning experimental paradigms, I investigate the conditions under which complexity in morphological and syntactic systems emerges, spreads, and reduces. These experiments illustrate the interaction of transmission, learning and use in hitherto understudied domains - morphosyntax and word order. In a first study (Chapter 2), I report the first iterated learning experiments to investigate the evolution of complexity in compositional structure at the word and sentence level. I demonstrate that a complex meaning space paired with pressures for learnability and communication can result in compositional hierarchical constituent structure, including fixed combinatorial rules of word formation and word order. This structure grants a productive and productively interpretable language and only requires learners to acquire a finite lexicon and a finite set of combinatorial rules (i.e., a grammar). In Chapter 3, I address the unique effect of communicative interaction on linguistic complexity, by removing language learning completely. Speakers use their native language to express novel meanings either in isolation or during communicative interaction. I demonstrate that even in this case, communicative interaction leads to more efficient and overall simpler linguistic systems. These first two studies provide support for the claim that morphological and syntactic complexity are shaped by an overarching drive towards simplicity (or learnability) in language learning and communication. Chapter 4 reports a series of experiments assessing the possibility that the simplicity bias found in the first two studies operates at a different strength depending on the linguistic level. Studies in natural language learning and in pidgin/creole genesis suggest that while morphological variation seems to be highly susceptible to regularisation, variation in other syntactic features, like word order, appears more likely to be reproduced. I test this experimentally by comparing regularisation of unconditioned variation across morphology and word order in the context of artificial language learning. I show that language users in fact regularise unconditioned variation in a similar way across linguistic levels, suggesting that the simplicity bias may be driven by a single, non-level-specific mechanism. Taken together, the experimental evidence presented in this thesis supports the hypothesis that the cultural and cognitive pressures acting on language users during learning and communicative interaction - for learnability, expressivity and coordination - are at least partially responsible for the evolution of linguistic complexity. Specifically, they are responsible for the emergence of linguistic complexity which maximises learnability and communicative efficiency, and for the reduction of complexity which does not. More generally, the approach taken in this thesis promotes a view of complexity in linguistic systems as an evolving variable determined by the biases of language learners and users as languages are culturally transmitted.

In this thesis, I examine morphological, syntactic and semantic finiteness properties of a selected set of converbal, adverbial clauses in Jejuan (Koreanic, South Korea), following a Canonical Typology approach. This typological framework relies on the construction of a so-called Canonical Ideal that is the logical convergence of all criteria defining a certain concept. Thus on the basis of finiteness criteria, I build an ideal of a canonically finite clause, which the individual properties of a particular Jejuan adverbial clause are compared to. In this way, I situate Jejuan adverbial clauses in the typological space of Canonical Finiteness. This ensures cross-linguistic comparability through a rigorous application of this concept to a particular language. Drawing on elicited and spontaneous language data, this study shows that the finiteness properties of Jejuan adverbial clauses are not uniform, and bundle into larger patterns only to a limited extent: one can identify a class of 'canonically non-finite' clauses, yet most clause types do not group into larger classes of finiteness properties. On the one hand, no adverbial clause ever confirms with the canonical ideal in the entirety of its properties, meaning that they are in fact non-finite at least in some respect. On the other hand, a particular clause type may be more canonically finite on the morphological level, yet less so on the syntactic or semantic level, and vice versa. As a conclusion, the findings support current tendencies in the theoretical literature which suggest that neither a binary, nor a gradual theorisation on finiteness provide satisfactory accounts: in fact, the Jejuan results ask for an examination under a multi-dimensional angle which allows for various mismatches between different linguistic domains. Given this, I argue that the Canonical Typology model is a welcome framework that can capture the diversity of crosslinguistic finiteness manifestations in a rigorous, yet multi-faceted manner, enabling the comparison of different languages in a principled way.

In 2007 the Myanmar government made a decision to standardise the country's two sign languages, Yangon Sign Language and Mandalay Sign Language. The project was initiated without community consultation. While this paternalistic approach to sign language planning and policy is widespread, there is a paucity of academic research that explores deaf people's responses to policy making. This study presents an ethnographic account of language planning and policy (LPP) in the Yangon deaf community, giving visibility and voice to deaf people. LPP is examined at different levels, demonstrating the complex and dynamic interactions between language policy in education, unofficial community language policy and top-down attempts at standardisation. Experiences of language use in school are shown to shape unofficial LPP in the community, influencing language ideologies and linguistic practices, as well as wider beliefs regarding language, equality and citizenship. The study also highlights the agency of the community, demonstrating how participants negotiate, and subvert, official LPP by constructing their own unofficial policy towards the standardised language in accordance with their ideologies, interests and agendas. Throughout the thesis, attention is drawn to the need for LPP research to go 'beyond language' and adopt an interdisciplinary approach in order to understand more completely the implications and outcomes of LPP. The findings also contribute to ongoing scholarly debate regarding the interplay between LPP and social justice. It is suggested that a more critical approach is required, one that questions the assumed moral imperative of interventions such as mother-tongue education and language rights.

Gorwaa is a South Cushitic language of Tanzania whose nouns are particularly complex. Based on detailed documentation of the language (most of which is openly accessible for consultation in an online archive (Harvey 2017)), this dissertation provides a first description of Gorwaa grammar, with a particular focus on the noun. Additionally, a grammatical analysis of Gorwaa nouns is developed using the Distributed Morphology architecture and Minimalist syntax. This offers a different perspective from the typically functional analyses available for South Cushitic languages thus far. Following a general sketch of Gorwaa grammar, as well as a brief introduction into the theoretical framework, each subsequent chapter of the dissertation focuses on one subpart of the noun and its morphosyntactic characteristics. Composed of several identifiable subparts (e.g. the stem, the suffix, and the linker), each of which in turn presents a rich array of variants, the Gorwaa noun is an ideal entry point for inquiry into Gorwaa as a system, as adequate explanation of nouns in this language touches on all the major modalities of grammar (phonology, morphosyntax, semantics, and pragmatics). Syntactically, the stem is formed of a root, whose characteristics (phonetic, semantic, and categorial) are determined by the larger syntactic structure in which it is found. Distinguished by two broad groups of morphosyntactic characteristics (those which are regular and those which are listed), the suffix is formed of (maximally) three syntactic heads: Cl (classifier), # (quantifier), and n ('little n'). In order to bear a number value (Sg or Pl), nouns must be classified and quantified. Nouns unvalued for number ('general' number) are neither classified nor quantified. The little n head is the site of the paradigm, itself established as a grammatical formative realized as a specific suffix through instructions post-Spellout. Grammatical gender is a diacritic feature, also realized post-Spellout, making Agree a necessarily post-Spellout operation (cf. Bobaljik 2008). The linker is agreement morphology on the syntactic head D. Cases of mismatch between the form taken by the linker and the gender value of n represents the interpretable (semantic) features of the referent of the noun (itself the external argument of n), intervening in agreement relations between D and n. This mechanism is extended to account for adjectival number agreement on nouns of general number.

This thesis explores the structure of Blanga (iso 639-3 code blp), a previously undescribed and undocumented Austronesian language of the Solomon Islands. It opens with a sociolinguistic and ethnographic introduction to the language, their speakers and the linguistic ecology of the region, followed by a presentation of the main phonological features and an analysis of some major phonological processes, including strategies for hiatus avoidance, which constitute the base of several surface phenomena, and stress assignment, the theoretical importance of which is also revealed by surface processes. A detailed description of major morphological and morphosyntactic aspects is then made available, which continues with a discussion of Aktionsart, types of predicates and their subcategorisation frames, in an attempt to define thematic roles and identify macroroles, according to the RRG principles of lexical decomposition. Sentence-level coordination and subordination are subsequently looked at. The final chapters focus on the relations established between a predicate and its (direct) arguments and their implications for the general theory. Blanga does not use formal means of encoding semantic roles but the speakers are able to identify them based on their intuitions of verb semantics, on pragmatic and cultural knowledge, and on discourse context. There is also considerable variation in the language with respect to the order of constituents in a clause, which, in conjunction with morphosyntactic markers and prosody is used to encode topic and focus. Blanga, therefore, employs completely different means of encoding the two primary sets of predicate-argument relations. Because of the lack of a voice distinction and of other constructions in the language, evidence for the necessity of employing a third set, that of grammatical relations, is limited to verb agreement, Equi-NP coreference, coreference in chained clauses and, partially, causativisation, which comes to confirm their construction-specific character.

This thesis investigates the cognitive underpinnings of Scalar Implicature phenomenon. Here I present a series of experiments in three domains of research for scalars: (i) scalar diversity phenomenon, (ii) implicature priming and (iii) the time course of access to pragmatic enrichments. I adopt a broadly Gricean theoretical approach with local pragmatic enrichment to the design of the studies and argue that this approach can shed light on the phenomena. The results of the experiments also lend support to the theoretical perspective taken. This thesis introduces a new perspective to interpret scalar diversity phenomenon. Given the observation that different scalar terms give rise to unembedded scalar implicatures at different rates, experiments presented in Chapter 2 and 3 suggest that one source of scalar diversity is the strength of association between a scalar term and its upper-bounding local enrichment. It provides indirect evidence that local enrichment impacts on the interpretation of unembedded scalars. More direct evidence of an effect of local enrichment in unembedded scalars is found in implicature priming. Experiments presented in Chapter 4 find unembedded and embedded scalar enrichments could prime each other, indicating local pragmatic enrichment as a shared mechanism involved in both. In addition, this thesis presents research on the time course of access to local pragmatic enrichment of 'some', which reveals no delay in pragmatic enrichment vis a vis semantic interpretation. Overall, this thesis argues for an integrated Gricean system that allows for scalar phenomena to be explained by two mechanisms, a global inference mechanism and a local enrichment mechanism.

This thesis investigates the role of cooperation for pragmatic inferences. The notion of cooperation that is proposed as relevant for discussing the relationship between cooperation and communication is that of joint action. Different theories of communication are reviewed together with the different roles that they assign to cooperation in the context of communication. The study of communication in non-cooperative contexts is used as a way to inform the role of cooperation in communication. Different predictions are derived from Grice’s (1989) account and Sperber and Wilson’s accounts (Sperber & Wilson, 1986/1995; Sperber et al. 2010) regarding what happens to implicatures in non-cooperative contexts. A series of experimental studies investigates communication in non-cooperative contexts and tests the prediction derived from Grice’s account that hearers will not derive implicatures from the utterances of uncooperative speakers. Overall, the results of these studies are not in support of Grice’s prediction. They instead support the view that because of a dissociation between comprehension and epistemic acceptance of communicated content (Sperber et al., 2010; Mazzarella, 2015a) uncooperative contexts do not affect the inference of implicatures but only the acceptance of their content. Lastly, this thesis touches on the topic of the source of relevance for an utterance, which is treated as a theory neutral notion corresponding to what different theories formalise as the Question Under Discussion (Roberts, 1996/2012) or the shared purpose of interlocutors (Grice, 1989). The results of an experimental study on this topic suggest that the exhaustivity of an utterance as an answer to the possible QUDs in a context affects the choice of which QUDs the utterance will be taken to be addressing. Ultimately, this thesis provides initial experimental evidence on how cooperation (or lack thereof) affects pragmatic inferences and puts forward a novel experimental approach to this line of research.

The manifold connections between language and economic affairs have been approached intermittently from a variety of angles. Lack of a sustained focus, however, means less attention has been paid to financial and economic concepts in a nonspecialised public sphere, how their meanings travel through use, or how and to what extent they shape an audience's understanding of economic affairs. The current thesis investigates how an online news audience draws attention to particular economic and financial terms, why some terms attract more attention, and which contextual features influence potential uptake of the terms. The data consists of media reporting during selected moments of the euro crisis (2010-2011), and subsequent online user-comments. Under investigation are comments produced in The Guardian/The Observer Online, and, to a lesser extent, The (Sunday) Times, around critical moments when Greece, Ireland, and Portugal applied for financial aid (23 April 2010, 21 November 2010, and 6 April 2011, respectively). The thesis employs an analytical lens of metapragmatics, incorporating insights from linguistic pragmatics and principles of Computer-Mediated Discourse Analysis. The focus lies on how, and to what extent, news user-commenters engage with specific terminology of the reporting through their own metalanguage surrounding particular financial and economic terms. Drawing upon Agha's (2007) framework of language use in society, as well as processes of mediatization in online contexts, the thesis argues that three main categories of financial/economic terms may be established based on the amount and type of metalanguage they attract. Besides (i) everyday, common terms and (ii) terms that are highly contentious, (iii) a substantial portion of terms allows commenters to both strategically bring to the fore existing connotations surrounding them, as well as introducing connotations specific to the momentary interaction. The semantic ambiguity of theoretical economic/financial concepts in a public setting plays a key role in the processes witnessed in the third category. The thesis closes with a discussion of the findings' implications for ideology-in-language research, as the data reveal varying degrees of audience receptivity to particular lexis. It also reiterates the need for 'a turn towards the audience' in discourse-analytical investigations.

This thesis discusses a number of open questions and explores various understudied and unstudied aspects of Modern South Arabian (MSA) linguistics and MSA studies at large. Namely, it contains an extensive literature review which offers a summary of the most significant works in the field, a discussion about the internal sub-grouping and the internal cohesion of Modern South Arabian within Semitic, a grammatical sketch of the Jibbali/Shehret variety spoken on al-Ḥallāniyya island in the Kuria Muria Archipelago: this description focuses on the differences between this and the better-studied mainland varieties. The syntax section, however, takes into account not only the insular variety, but also central and eastern Jibbali/Shehret, in order to present a clearer picture of the syntactic features of this language. In the chapter that follows, a discussion about the lexical interferences of Arabic and Austronesian in Modern South Arabian is presented. Furthermore, the thesis contains three appendices: the first one is a description of the Dhofar inscriptions, re-labelled South-eastern Arabian inscriptions in view of new findings in the Yemeni governorate of Mahra and the examination of epigraphic evidence from Soqotra. The second appendix contains a number of texts in Kuria Muria Jibbali/Shehret, ranging from a selection of Miranda Morris's recordings from the 1980s to recordings proceeding from personal fieldwork made in 2017, with interlinear morpheme glossing. Finally, the third appendix contains a glossary of the above-mentioned Jibbali/Shehret variety. It is argued that Modern South Arabian studies are overall still in an incipient phase, and a research agenda is set up and proposed.

Two eye tracking experiments investigated whether italicising contrastively focused words facilitates processing. In speech, pitch accents can mark focus, signalling that there exist alternatives to the marked element which are relevant to interpreting the expression. In writing, typographic emphasis (italics, bold, etc.) can be used in a similar manner, particularly to mark contrastive focus (asterisks delimit italicisation here): ‘We expected Mark to bring *Mary* to dinner. However, when he arrived, he was with *Ellen*.’ Previous studies have shown that processing is facilitated by congruent marking of information structure with pitch accents, and impeded by incongruent marking. This study sought similar effects for typographic emphasis. Eye movements of participants were tracked as they read short texts (dialogues in Experiment 1, narratives in Experiment 2) in which a contrastively focused target word was italicised, or not. Experiment 2 also manipulated contrast on the target word. To the author’s knowledge, this is the first eye tracking study that investigates the interaction of these factors within continuous text. Incongruence, where a non-contrastive word was typographically emphasised, or a contrastive word was not emphasised, was predicted to cause processing difficulty manifesting as re-reading. Results did not suggest that incongruence had a processing cost, although some effects were found, including (amongst more subtle effects) longer overall fixations on target words, and higher likelihoods of them being fixated at all. These results are discussed from several perspectives, including a ‘zero-impact’ account, which holds there to be no linguistic effect of typographic emphasis, and a more probable ‘effects elsewhere’ account, which suggests the type of re-reading examined here may not have been the correct place to look. The possibility that the visual contrast inherent in typographic emphasis may simply ‘catch the eye’ rather than have a linguistic effect is also investigated, and assessed as being unlikely.

This study investigates the linguistic outcomes of war-induced dialect contact between speakers of two dialects of Iraqi Arabic (IA): the qiltu dialect spoken in Hīt (HIA) and the gilit dialect spoken by migrants from the gilit areas to Hīt. It differs from previous contact-based studies on Arabic dialects in that it investigates dialect change in the speech of the local residents rather than in the speech of the migrants. The HIA dialect has been noted for the loss of certain local linguistic features at the expense of gilit features with a wider areal distribution (Al-Ani, 1978; Khan, 1997). The thesis provides a quantitative sociophonetic description of the speech of 36 male and female HIA speakers and of their accommodation to the gilit dialect by examining the impact of four independent variables (age, gender, Social Network Integration (SNI), and speakers’ attitudes) on the use of four phonological variables (two consonantal and two vocalic). At the consonantal level, variation in the use of the uvular stop (q) and the velar stop (k) are examined. At the vocalic level, the variation in the insertion of the epenthetic vowel [i] within onset consonant clusters in word initial positions and the lowering of the short high vowel /i/ into [a] in the context of pharyngeal consonants is examined. This is the first quantitative sociophonetic study of its kind to be conducted on HIA to examine whether dialect levelling is taking place and to compare the findings to those of other sociolinguistic studies on IA qiltu dialects (e.g. Mosuli Iraqi Arabic) spoken in communities that have undergone similar sociodemographic changes as Hīt. Over the course of the last three decades, Hīt has seen high levels of internal migration of people from gilit-speaking cities, namely the capital Baghdad. Previous research has suggested that as a consequence of spatial mobility and dialect contact, language features with a wider socio-spatial currency become more widely used at the expense of more locally specific features (Britain, 2009). This mobility disrupts large-scale close-knit localised networks, which have over time retained highly systematic and intricate sets of socially built linguistic standards (Milroy, 2002). This study shows that speakers’ age, gender and SNI have an important impact on the acquisition of gilit features, with SNI being the most important independent variable. HIA speakers with more open friendship networks i.e. high scorers on the SNI scale were the leaders in adopting gilit features. There is sound change in progress and gilit features spread faster among the younger generation. In all age groups, male speakers produced more gilit features than female speakers did. The results suggest that the spread of the gilit features at the expense of the HIA local features is determined by social, linguistic and social-psychological factors. It was shown that speakers’ attitudes play an important role in their linguistic behaviour. Socially salient phonological variables such as (k) are levelled only minimally and the vocalic features are levelled more and adopted more frequently than the consonantal features. Epenthesis of the vowel [i] is restricted to monosyllabic words. Vowel lowering is not limited to the context of velar and pharyngeal consonants. The outcome of dialect contact is levelling, which involves the replacement of local linguistic forms with other linguistic features with a wider geographical spread (Cheshire et al., 1999:1) and motivated by individuals having open networks to people outside their community.

The interpretation of the Chinese simple reflexive ziji can be ambiguous between local vs. long-distance interpretation. Chinese features three verb types in relation to the interpretation of ziji: the introverted/self-oriented verb (VT1) only allow a local interpretation; the extroverted/other-oriented verb (VT2) only allow a long-distance interpretation; the ambiguous/context-dependent verb (VT3) allow both interpretations depending on the discourse-context. Hence, the current work focuses on how factors such as verb-semantic and discourse-context information influence the interpretation of ziji by a corpus study and a self-paced reading study. The corpus study examines the distribution of the three verb types, indicating that compared with VT1 and VT3, VT2 is used less with ziji. Because only VT2 provides unambiguous evidence for a long-distance interpretation of ziji, the variations of the three verb types in the input of Chinese will result in a protracted acquisition of the long-distance interpretation of ziji. Also, the role of verb-semantic orientation and discourse prominence affecting the interpretation of ziji is supported based on the corpus data. The self-paced reading study investigates how verb-semantic and discourse-context information used as retrieval cues guide the interpretation and real-time processing of ziji by native Chinese speakers and English-speaking learners of Chinese. The findings are as follows. English-speaking learners of Chinese are able to acquire the long-distance interpretation of ziji, even if the long-distance interpretation is ruled out by their L1 (English). With Chinese proficiency increasing, they allow less long-distance interpretation of ziji with VT1. In addition, although native Chinese speakers and English-speaking learners of Chinese are sensitive to both cues, they do not weigh the two cues in an equal way. In particular, native Chinese speakers rely more on the verb-semantic cue to interpret ziji, however, the discourse-context cue can over-rule the verb-semantic cue. Whereas English-speaking learners of Chinese rely more on the discourse-context cue (less on the verb-semantic) to interpret ziji. Also, with Chinese proficiency increasing, they become more reliance on the verb-semantic cue, however, their reliance on the discourse-context cue is not decreased. Moreover, English-speaking learners of Chinese are generally slower than native Chinese speakers during real-time processing of ziji. English-speaking learners of Chinese process more when they encounter the verb before ziji, while native Chinese speakers take longer time to process ziji and onwards. Furthermore, English-speaking learners of Chinese are more susceptible than native Chinese speakers to the retrieval interference when there is a conflict between the two cues. In conclusion, L2 acquisition of the long-distance interpretation of ziji by English-speaking learners of Chinese supports a probabilistic approach to L2 parameter (re)setting. Also, the interpretation and real-time processing of ziji by native Chinese speakers and English-speaking learners of Chinese supports a cue-based approach to language processing and comprehension.

The existence of two types of second language knowledge, namely acquired and learned, implicit and explicit (or more accurately, modular and extra-modular second-language knowledge) has long been accepted by different theories within the field of second language acquisition. However, this consensus is hampered in part by a lack of empirical evidence of how each type of second language might manifest itself in the production and comprehension of the second-language learner. Adopting the Modular Cognitive Framework (Truscott & Sharwood Smith, 2004a; 2004b), this thesis aims to establish a detailed account of the nature of development of the two types of knowledge; how different input experiences stimulate the development of different types of L2 knowledge. The thesis also aims to explore how L1 cross-linguistic influence would fade with more frequent encounters with L2. The linguistic constructions tested are unaccusative and unergative verbs. The basis of choosing the linguistic constructions is the L2 English classroom practice of oversimplifying the notion of intransitivity and avoiding explicit teaching of syntactic and semantic aspects of the unaccusative and its argument. The study incorporates two empirical tasks measuring L2 comprehension and production: a reaction-time reading comprehension task, comparing Najdi Arabic L2 English learners' response time to sentences with unaccusative verbs compared to response time to sentences with unergative verbs and an error-correction and rule-verbalisation task comparing Najdi Arabic learners' ability to correct ungrammatical sentences with unaccusative verbs compared to correcting ungrammatical sentences with unergative verbs. Findings from the current experimental study support what the Competing Systems Hypothesis (Rothman, 2008) seems to suggest: while L2 production could be guided by both modular as well as extra-modular L2 knowledge, L2 comprehension could only be guided by modular L2 knowledge.

The present study explores the representations of ‘Arabs’ and ‘Americans’ in American and in Arab news media discourse between 2001 and 2011 through an evaluation of the metaphors used to describe each group. The study chiefly argues that metaphors in Arab and American news media discourse are used for rhetorical and ideological motives. The main research aims of this study are to find out what metaphors used to represent Arabs and Americans occur in the New York Times (the NYT) on the one hand, and in the Al Jazeera English and the Arab News (the ANM) on the other hand, whether there are similarities or differences between metaphors in each news media source, and to explain the purposes underlying the choice of these metaphors. The study is based on the experientialist cognitive linguistic Conceptual Metaphor Theory developed by Lakoff and Johnson (1980) and the rhetorically, cognitively and corpus oriented approach to metaphor Critical Metaphor Analysis developed by Charteris-Black (2004). The main corpus approach that this study uses is Metaphor Pattern Analysis proposed by Stefanowitsch (2006) for the collection of metaphor data. The corpus used consists of news articles about Arabs and Americans extracted from the online editions of the NYT, and the ANM between 2001 and 2011. Findings on the metaphors used to describe Americans and Arabs in the NYT demonstrate that the use of metaphors describing Americans is noticeably but not always motivated by the ideological purpose of representing them as a unique and exceptional nation thus supporting the Myth of American Exceptionalism. It is additionally argued that the use of metaphors referring to Arabs can be explained partly by the rhetorical ideological goal of associating them with danger, violence and threat. In ANM, findings revealed that a number of metaphors describing Americans have underlying rhetorical and ideological motives similar to those in the NYT. The image of Americans as a unique nation is reiterated in the ANM. As regards the use of metaphors occurring in collocation with Arabs in the ANM, the findings demonstrated that they are partially motivated by an underlying belief that associates them with powerlessness, passivity and incapability. Furthermore, and despite the importance of rhetorical ideological purposes, the findings also show that metaphors used do not always necessarily convey positive evaluations of the self or negative evaluations of the other. In addition to its contribution to our understanding of the dominant representations of Americans and Arabs in two inherently different news media sources, the thesis has proposed adjustments to Critical Metaphor Analysis as relevant to this project, and has also contributed to research in critical discourse analysis.

The English language proficiency and academic literacy needs of increasing numbers of incoming students from diverse language and educational backgrounds constitute one of the major challenges faced by English Medium of Instruction (EMI) universities, both in English speaking 'inner circle' countries (Kachru, 1985) and elsewhere, as a consequence of the globalisation of higher education. Unprecedented pressure on universities to address the manifold academic language needs of diverse cohorts of students has resulted in a move to fashion academic English language development mechanisms, known as English for academic purposes (EAP) programmes in the UK. A growing body of research dedicated to academic English provision has thus emerged. Despite the burgeoning interest in this area of enquiry, however, research in the UK, unlike other native English-speaking contexts (e.g. Australia), has primarily focused on pre-entry EAP programmes - referred to in this context as pre-sessional courses - rather than post-entry EAP provision, normally referred to as in-sessional courses. This study therefore examined the characteristics and efficacy of current in-sessional programmes on offer across British universities and the principles underlying the design and delivery of such programmes. It also evaluated the effectiveness of the provision and those factors determining its efficacy. To this end, a mixed-method approach consisting of a large-scale online survey and a series of follow-up interviews was employed, in which three primary stakeholder groups participated: students, subject specialists and in-sessional staff. The findings suggest that, despite some variation, the provision tends to constitute primarily non-credit-bearing English for general academic purposes courses, predominantly targeted at non-native speakers of English, and offered centrally via different units across institutions. In terms of effectiveness, the results showed that the provision is perceived as less effective in terms of adequacy, relevance, and needs analysis respectively. In addition, new evaluation sub-criteria emerged, other than those reflected in best practice principles literature on EAP, based on stakeholders' perceptions of the provision. These findings contribute to the existing literature on in-sessional EAP provision by providing a more comprehensive understanding of what current in-sessional provision across British universities looks like, and identifying those factors which determine its efficacy, based on both best practice principles in EAP and stakeholders' perceptions.

This study investigates in-service teacher education and training (INSET) practices in Qatar, and employs action research (AR) as an investigative methodology. It is an attempt to improve the researcher's practice as a teacher trainer by exploring the local INSET setting and investigating an interventionist training programme. It had its origins in personal observation that local INSET is struggling to provide effective learning opportunities to teachers who generally engage in the minimal required INSET. This study was conducted over a five-year period in three stages, with three different cohorts of in-service teachers of English language teaching (ELT). The three stages were needs analysis, implementation and reflection. The needs analysis stage aimed at gathering context information to inform the INSET design and delivery in the succeeding stages. The implementation stage consisted of three AR cycles. In the first cycle, the three-month INSET delivery involved 12 sessions, comprising 48 hours of face-to-face training. The second cycle involved around 25 INSET events for over one and a half months. In the third cycle, each of the three sessions and a demo lesson was delivered every other two weeks to allow for training input implementation and online collective discussion by trainees. The quantitative and qualitative data were collected by means of multiple methods. Quantitative data were statistically analysed while qualitative data were thematically analysed. The findings of the study provided insights on the teachers' attitudes towards INSET, main challenges, perceived success factors, preferred training themes, engagement, and trainer role. This study contributes to understandings of INSET in the Qatari context. It was conducted at a time of major reforms where INSET is seen as one of the main tools to address policy makers' concerns about the educational system. Albeit, minimal academic research has been conducted on the effectiveness of the teacher training programmes offered to TESOL practitioners in Qatar. The findings provide ways to improve INSET, shed light on how the role of a trainer can contribute to its success, explain why teachers in Qatar may not be motivated to take part in INSET, and propose a framework for INSET design and delivery in the Qatari and other ELT contexts.

This thesis deals with the construction of the European Union (EU) as negotiated among contributors to the English Wikipedia between 2001 and 2015. It focuses on the Talk Page (TP) which accompanies the Wikipedia article on the EU and provides a space for Wikipedia contributors to discuss controversial issues regarding the article. The EU has received considerable attention in Critical Discourse Studies (CDS), addressing e.g. questions regarding language policy and discourses surrounding topics connected to the EU (e.g. Muntigl, Weiss, & Wodak, 2000; Unger, Krzyżanowski, & Wodak, 2014; Wodak, 2007a). However, private individuals’ attempts to make sense of the EU when facing the task of defining it have hardly been touched upon. In this context, Wikipedia constitutes an ideal repository of data as it has recorded debates on the institution since 2001. Taking a corpus-assisted approach (cf. Baker, 2006), I examine how contributors from various backgrounds have grappled with their understanding of the EU. Additionally, this study explores aspects of Wikipedia since this collaboratively created encyclopaedia has received little research attention. Taking the EU on Wikipedia as a starting point, this thesis presents a foray into how Wikipedia can be approached from a CDS perspective. That is, on the one hand, it identifies central aspects of this website’s structure and addresses policies that guide Wikipedia operations and thus shape Wikipedia data. On the other hand, it examines the site’s societal impact/relevance and evaluates to what extent it can function as a transnational public sphere. Findings suggest that a substantial part of discussions amongst Wikipedians addresses the classification of the EU along the continuum between confederation and unified country, depending on different views concerning member states’ sovereignty. Wikipedia’s policies and the nature of the debates further suggest that the TP can, to some extent, serve as a transnational public sphere.

This thesis critically analyses the intriguing behaviour of plural inanimate nouns triggering partial agreement in subject–verb–object (SVO) word order in Modern Standard Arabic (MSA). The analysis focuses on three significant morphosyntactic features of agreement: animacy, gender and number. I follow integrated theoretical assumptions of current Minimalism and Distributed Morphology (DM). I operate on the consonantal root-and-pattern property of MSA to argue for a DM approach to feature analysis. I argue that assignment of gender values takes place during syntax before PF whereas the exponence of agreement features takes place post-syntactically at phonological form. I argue for a decompositional analysis for the structure of determiner phrase (DP) in which agreement features are not located at the same syntactic head within the nominal. Following Ritter (1993) and Zabbal (2002), number heads its own syntactic projection, and thus number phrase (NumP) is the syntactic locus for inflectional number values of the nominal. I argue for two locations for number values: NumP, which hosts the distributive reading, and n, which hosts the collective reading interpretation. I also argue for a decompositional analysis of nP in which there are two ns; the closest to the root hosts the interpretable gender that is assigned based on the semantic properties of the nominal, whereas the higher n hosts the uninterpretable gender assigned arbitrarily to the noun. I argue that this structure provides a satisfactory analysis for the agreement behaviour of mixed-agreement nouns. With respect to agreement, I argue for an Agree-based approach in which the probe enters into an Agree relationship with a goal in its c-commanding domain. Following Doron (2000), I assume that the presence of an Extended Projection Principle feature on T is optional and conditioned by the goal’s capability to move to [Spec, TP].

This PhD thesis develops a critical account of discursive practices of bordering in the EUropean migration regime. By articulating recent advances from the fields of Critical Migration and Border Studies and Discourse Studies, it develops a theoretical and methodological framework that enables grasping discursive borders in their heterogeneity. On a broader level, it is interested in re-approaching post-structuralist and materialist strands of theory and analysis by going back to their beginnings in structural Marxism and psychoanalysis. EUrope’s discursive borders are scrutinised through the lens of different contexts that allow emphasising the entangled nature of policy, academic, and activist discourse. First, the present research scrutinises a set of practices of discursive bordering with a relatively high stability over time. Adopting a post-colonial, macro-historical perspective, it shows how EUrope’s colonial history infuses the conceptual apparatus of the EU’s contemporary migration policy. This serves as a foundation for the following chapters, that examine practices of discursive bordering from a micro-enunciative and a situated perspective. While the second analysis focuses on the construction and supraversion of the labour / refugee divide in German discourses on EUropean migration, the third shows how discursive borders are turned into a political stake in a migrant protest. This allows conceiving of categorisation and differentiation as discursive practices that are scattered in time and space, and characterised by resonances, contradictions, and subversions instead of following a common rationality or having a central point of reference.

The purpose of this study was to develop the Talk Skills pedagogic intervention, implemented in the Korean adult L2 learning context, which aims to raise awareness of effective L2 talk and teach oral communicative strategies that help students to achieve it. The study is underpinned by theories that foreground the importance of language use in L2 classrooms, focusing, most importantly, on the relationship between interaction and second language acquisition, and sociocultural theory for language learning. Review of the literature showed that students had the best opportunities for language learning when classroom talk embodies characteristics such as students giving opinions, offering reasons, sharing information, respectfully challenging each other, attempting to reach agreement, negotiating meaning, noticing and building upon gaps in their language and promoting language learning through scaffolding and emergent language. This type of talk is termed here exploratory talk for language learning. However, research into the Korean context showed that Korean L2 learners encounter problems with classroom group oral interaction that inhibit the production of this kind of talk and that may lead to unfulfilled potential for learning. This led to the hypothesis that adult Korean L2 learners could benefit from lessons that raise awareness of this kind of talk and learn strategies to help achieve it. Drawing on previous attempts at metacognitive awareness raising of effective classroom talk, as well as literature on oral communicative strategy training, the Talk Skills intervention was developed using a design-based research (DBR) methodology. The scope of the project was limited to exploring the soundness and local viability of the intervention, using lesson transcript data, student interview feedback, my own field notes and expert appraisal from my course tutors to refine the intervention across two iterations. Initial impact of the project was also explored by analysing feedback from a small number of teachers who have used elements of the intervention in their adult English language courses. Taken as a whole, this thesis argues that Korean adult L2 learners can benefit from metacognitive awareness raising of exploratory talk for language learning and the learning of oral communicative strategies to help achieve this kind of talk. The thesis further argues that this aim can successfully be achieved using a design-based research methodology to both develop the Talk Skills intervention as a pedagogic tool, and further offer specific insight into instructional techniques, student engagement and teacher’s interactional roles that aid the success of its implementation. Finally, this thesis argues that as DBR is an underutilized methodology in the field of L2 research, the Talk Skills project offers a useful example of DBR for practitioner researchers wishing to embark on intervention design and development.

In 2010, Article 124 of the new Ecuadorian Higher Education Law stipulated that university students need to master a foreign language as a requirement for graduation. Subsequent regulations specified that this requirement had to be a B1 level, based on the Common European Framework of Reference (CEFR). To identify the responses, challenges and tensions that universities experienced in complying with the requirements of Article 124, an explanatory sequential mixed methods study was conducted. This comprised a survey that was administered to language centre Directors, EFL teachers and EFL students in 14 universities located in 10 different cities in the country, in-depth interviews conducted with teachers, and focus group interviews with students from 3 universities. Results of this research showed that responses of participating universities focused mainly on their physical and information and communications technology (ICT) infrastructure development, and less on pedagogy and management due to dependence on the use of overseas textbooks and the adoption of the Common European Framework language indicators as the proposed EFL outcomes for ELT programmes. Challenges faced by university stakeholders included lack of budget for infrastructure and ICT improvement, effective ICT integration, identification of a target EFL language level, and a lack of status of English in the participating universities. Out of these responses and challenges, tensions emerged related to teaching qualifications and access to appropriate institutional resources for teachers. Using Mahboob and Tilakaratna’s (2012) Principles-Based Approach for English Language Teaching Policies and Practices as a lens through which to analyse EFL language policy, the study found that the principles of collaboration, alignment and transparency were not sufficiently realised and there was a lack of evidence and empowerment among Ecuadorian universities. Thus, the study proposes a more contextualised and consensual approach to formulating EFL language policy, in which English can be integrated into institutional processes that promote globalisation and the internationalisation of universities.

The last two decades have seen major interventions in naturalisation legislation across much of the ‘industrialised’ world, with successive UK administrations introducing and refining citizenship tests, language proficiency requirements, and ceremonial performances in an explicit bid to elevate the acquisition of citizen-status from a ‘right’ to a ‘privilege’ (Home Office, 2013a). In this thesis, I draw on theorisations of the border as a geopolitical, a biopolitical, and a ‘neuropolitical’ (Isin, 2004) phenomenon, and thus conceptualise these newly-instituted ‘citizenisation’ (Fortier, 2017) processes as a political project, as part of an assemblage of bordering practices through which to heighten and manage the anxieties of the neurotic subject within the securitised state, and to realise the nation state in domopolitical terms as a home (Walters, 2004). While discourse researchers have certainly provided much important empirical insight into the strategies employed in contemporary citizenship documentation, and in the broader discursive contexts in which these citizenship regimes are situated (e.g. Wodak et al, 2009; Löwenheim and Gazit, 2009), I argue that the citizenship regime cannot be understood as wholly constituted through the linguistic practices of élite policy and media actors, but should instead be conceptualised more broadly as discursive practice: as a complex set of linguistic, material, and symbolic practices. From this premise, I situate my research within a particular local context, drawing together a data set which takes into account policy documents and testing materials, observations of testing and ceremonial rituals, and interviews with twenty actors involved in the production of the citizenship regime, including citizenship officials and recent and prospective citizens. Articulating themes across the data set, I look to the ways in which both the state and the Good citizen-subject are performed and secured within this discursive landscape, and find important moments of disruption and contestation to this dominant discursive formation.

Can the learning of another language influence the way we interpret and describe emotional situations? Following previous research on emotions and crosslinguistic influence (Jarvis & Pavlenko 2010; Pavlenko 2002b; 2008d; 2014; Pavlenko & Driagina 2007), this thesis examines whether the learning of emotion words in English as a foreign language (L2) can influence the way Arabic speaking learners interpret emotions when there are no translation equivalents for a given emotional concept in their first language (L1). By examining English language learners from different foreign language learning contexts in Kuwait, i.e. immersion and non-immersion, this thesis examines whether the learning of another language might affect their lexical choices when describing the same emotional situation in their L1 Arabic. It also examines whether or not possible differences in their identification and expression of emotion in the two different languages can be attributed to crosslinguistic influence. This study focuses on the emotion words excitement/excited/exciting and frustration/frustrated/frustrating, as these emotions depicted by these English emotion words differ in the way they are encoded and conceptualized both in meaning and in emotional weight in Arabic. The study adopts multiple methodologies such as narrative elicitation via film recall as well as one-on-one interviews to supplement the data. The data revealed evidence of L2 influence on the types of emotion words used in the L1, as well as an L2 influence on the L1 descriptions of emotional states in the immersion learners' data. There was also an influence of the L1 on the use of L2 emotion words and descriptions of emotional states in the nonimmersion learners' data. The results suggest that foreign language immersion contexts can facilitate the internalization of L2 conceptual emotion categories in nonequivalent and partially equivalent L2 emotion words.