Children with early language delay form a heterogeneous group. Although a significant number will catch up and develop language in the normal range, some will continue to have difficulties with language. Predicting the outcome for these children represents a challenging task for clinicians. It has been suggested that the assessment of sociocognitive skills contributes distinctively to the prediction of persistence of language and communication difficulties and the nature of these difficulties. In the absence of standardized assessments in Saudi Arabia for children with early language delay, this study aimed to take a first step to filling this gap by developing a battery of early sociocognitive and language measures. The battery consisted of six measures assessing sociocognitive and language skills using direct and indirect methods, some existing and some newly developed or adapted for this project. Sociocognitive measures were the Early Sociocognitive Battery (ESB; Chiat & Roy, 2006b), together with a new Motor Imitation test (MI) and Sociocognitive Questionnaire (SCogQ); language measures included the Sentence Repetition test (Wallan, Chiat, & Roy, 2011), a new Arabic research adaptation of the Language Use Inventory (O’Neill, 2009), and a preschool adapted version of the Arabic Picture Vocabulary Test (Shaalan, 2010). Since this project was performed in a very different language culture and included a wider range of sociocognitive and language measures than most previous studies, a second aim was to investigate relations between the different sociocognitive and language skills. The battery was administered to 161 Saudi children between the ages of 2;0-3;5 years, divided into three six-months age groups and almost equally divided into boys and girls. Addressing the first aim of this study, results showed that all the measures with the exception of the SCogQ were reliable, valid, and age sensitive. These findings suggest that the measures are fit for purpose and have the potential to identify children with early language delay. Parental concern matched children’s performance on direct and indirect measures of language for the majority of children. Turning to the second aim of the study, regressional analyses using the three language assessments as outcome measures showed that the ESB and MI were important predictors of pragmatic language and receptive vocabulary when other measures had been taken into account. It is concluded that the substantial set of data that this study has produced on the wide-ranging battery of assessments can serve as a reference for clinical comparison and as a foundation for standardization with a fully representative sample of young Saudi children. These measures not only enable the formal identification of a delay in Saudi preschoolers but are also informative about strengths and difficulties and can guide intervention. The results add to current understanding of the role sociocognitive skills play in language development, and provide the foundation for longitudinal research investigating relations to longer term outcomes.

Archwilia’r traethawd hwn rai o’r problemau a wynebir wrth hyrwyddo iaith fechan, gan ddefnyddio’r Gymraeg yn brif enghraifft. Amlinella’r bennod gyntaf effaith negyddol unffurfrwydd monolithig y wladwriaeth-genedl ar amlieithrwydd, gan ei wrthwynebu ar y sail y gall medru sawl iaith wella hyblygrwydd meddwl a pherthnasau rhwng gwahanol grwpiau. Yn yr ail bennod cyflwynir damcaniaeth cymdeithaseg iaith a chynllunio ieithyddol ac amlinellir yr hyn y gallai disgyblaeth marchnata ei wneud i newid agweddau ac ymddygiad o blaid dwyieithrwydd. Ym mhennod 3 cyflwynir canlyniadau ymchwil a wnaed ar 219 o breswylwyr Dyffryn Teifi, ardal a chanddi ddwysedd siaradwyr Cymraeg uchel. Archwilir sawl agwedd ar yr iaith, e.e. defnydd iaith, dyheadau siaradwyr at y dyfodol, mesurau cynllunio ieithyddol a phroblemau a photensial grðp ail iaith cynyddol yr ardal. Ym mhennod 4 cymherir agweddau 324 o unigolion tuag at yr iaith, traean ohonynt yn ddisgyblion chweched dosbarth mewn ysgolion Cymraeg, traean arall yn ddisgyblion uniaith Saesneg a thraean yn ddysgwyr. Cyflwynir eu hatebion i restr gyffredin o gwestiynau parthed defnyddioldeb y Gymraeg, hunaniaeth genedlaethol a mesurau cynllunio ieithyddol. Pwysleisir pa mor bwysig yw siaradwyr ail iaith a’r rhai di-Gymraeg wrth geisio goresgyn y rhwystrau i sefyllfa ieithyddol wedi ei ‘normaleiddio’. Ym mhennod olaf y traethawd dadansoddir rhwystr arall i ddwyieithrwydd, sef y methiant anferth i gynhyrchu siaradwyr newydd, rhugl o oedolion. Cynigir mai methodoleg hen ffasiwn sydd ar fai am hyn a chynigir cyrchddulliau ‘ymennydd gyfeillgar’ i gymryd eu lle. Archwilir un o’r rhain, Suggestopedia, yn fanwl, a chesglir ei fod o leiaf mor effeithiol â chyrchddulliau eraill ac, o bosib, yn fwy effeithiol o dipyn. Cloir y traethawd drwy bwysleisio mai dim ond trwy gynllunio’n strategol wrth ystyried anghenion siaradwyr newydd o oedolion y goresgynnir y rhwystrau ar lwybr dwyieithrwydd

This thesis examines the bilingualism of nine families of Pakistani origin living in England, to gain an understanding of their experience of living through a period of rapid intergenerational language change. The families are all linked to the medical profession and the second generation are educationally successful. The study aims to gain insights into the social and linguistic processes of language shift. It also aims to contribute to understanding of the position of successful ethnic minority young people. Qualitative research methods are used, including interviews and participant observation and analysis of data is interpretive. A corpus of language use is analysed, focusing on intergenerational interaction and code-switching. Discussion of educational issues draws on policy documents. The fIrst three chapters constitute the preliminary section of the thesis. They present the research group, the social, linguistic and educational contexts, a theoretical framework and a critical description of the methodological evolution of the study. Chapters four to nine are data-based and focus on different aspects of the families' bilingualism. Accounts of early language nurture are analysed; bilingual skills and repertoires are investigated and the relationship between linguistic and cultural continuity is explored. A study of in-family bilingual interaction is presented and language education policy and its relevance to family concerns is considered. The study draws coherence from recurring themes, and insights into the experience of transitional bilingualism. Concern for educational success emerges strongly. The different signifIcance of using the minority languages inside the family and outside is revealed and the importance of long-term developments and cultural continuity is clear. Language education policies do not always reflect the priorities of the families. The linguistic situation of the families is very unstable but individuals may well take up very different intellectual positions in relation to their linguistic future.

This thesis examines language variation and change in a context of minority language revitalisation. In particular, I concentrate on young fluent speakers of Scottish Gaelic, a minority language of Scotland that is currently undergoing revitalisation. Data from three groups of speakers are presented: older speakers in the Isle of Lewis, a Gaelic heartland area in north-west Scotland; adolescent Gaelic-speakers in Lewis learning the language in immersion schooling; and adolescent Gaelic-speakers in immersion schooling in Glasgow, an urban centre where Gaelic has not traditionally been spoken as a widespread community language. The sociolinguistic analysis examines potential language changes, explores patterns of linguistic variation, and uncovers the role that Gaelic plays in identity formation for each of the participants. In order to gain an insight into the role of Gaelic in different speakers’ lives, I report on ethnographic studies carried out in Lewis and in Glasgow. The phonetic analysis then explores patterns of variation in the production of laterals, vowels, and tone and intonation. The results indicate large differences between the speech of older and adolescent speakers in Lewis, while differences between young speakers in Lewis and Glasgow suggest that Glasgow Gaelic is developing as a phonetically and socially distinct variety of the language. For example, older speakers in Lewis speak Gaelic as a partial tone language, unlike young people in Lewis and in Glasgow. Differences are also present between young people in Lewis and in Glasgow, such as in the acoustics of the vowel [ʉ], the production of the lateral system, and intonation patterns. The developments detailed in this thesis are the result of a complex interaction between the internal sound structure of Gaelic, language contact with varieties of English, identity construction, and differing conceptions of the self. All of these factors are conditioned by the status of Gaelic as a minority endangered and revitalised language. In exploring these avenues, I advance an account of language variation and change and apply it to a context of minority language revitalisation.

Statistics show that the world wide web is dominated by a few widely spoken languages. However, in quieter corners of the web, clusters of minority language speakers can be found interacting and sharing content. This study is the first to compare three such clusters of Irish language social media users. Social network analysis of the most active public sites of interaction through Irish – the Irish language blogosphere, the Irish language Twittersphere and a popular Irish language Facebook group – reveals unique networks of individuals communicating through Irish in unique and innovative ways. Firstly, it describes the members and their activity, and the size and structure of the networks they share. Then through focused discourse analysis of the core prolific users in each network it describes how the language has been adapted to computer-mediated communication. This study found that the largest networks of Irish speakers comprised between 150-300 regular participants each. Most members were adults, male, and lived in towns and cities outside of the language’s traditional heartland. Moreover, each group shared one common trait: though scattered geographically, through regular online interaction between core members they behave like communities. They were found to have shared histories, norms and customs, and self-awareness that their groups were unique. Furthermore, core users had adapted the language in new and innovative ways through their online discourse. This study is the first comprehensive audit of who is using the Irish language socially on the web, where they are forming networks online, and how they are adapting the language to online discourse. It makes a unique contribution in re-imagining what constitutes an Irish language community in the context of the Network Society. In the process, it contributes to the growing body of sociolinguistic research into globalisation and local identity on the web.

This study is concerned with the linguistic analysis of a cognitive training programme for offenders which was run at Prison X in 1996. Several Cognitive Skills classes run by prison officers and attended by groups of five to eight prisoners were videotaped and analysed to investigate the discourse practices used in these sessions. I also explored the written discourse of the Cognitive Skills Handbook used by the offenders as a reference-text for running the classes. In my research, I have borrowed insights from Critical Discourse Analysis (CDA), particularly Fairclough's three-dimensional model of discourse, as it forms a framework for studying language in its relation to power and ideology. I have attempted to show through this case study that the discursive practices investigated are ideological in that they produce and reproduce unequal power relations in the way they represent and classify offenders. Following the Hallidayan tradition, I have taken a systemic functional approach as my point of departure for the analysis and interpretation of texts.

This research focuses on improving stock tweet sentiment classification accuracy with the addition of the linguistic features of stock tweets. Stock prediction based on social media data has been popular in recent years, but none of the previous studies have provided a comprehensive understanding of the linguistic features of stock tweets. Hence, applying a simple statistical model to classifying the sentiment of stock tweets has reached a bottleneck. Thus, after analysing the linguistic features of stock tweets, this research used these features to train four machine learning classifiers. Each of them showed an improvement, and the best one achieved a 9.7% improvement compared to the baseline model. The main contributions of this research are fivefold: (a) it provides an in-­depth linguistic analysis of stock tweets; (b) it gives a clear and comprehensive definition of stock tweets; (c) it provides a simple but effective way to automatically identify stock tweets; (d) it provides a simple but effective method of generating a localised sentiment keyword list; and (e) it demonstrates a significant improvement of stock tweet sentiment classification accuracy.