Thesis (M.A.T.) -- School for International Training, 2006. / Advisor -- Bonnie Mennell Includes bibliographical references (leaves 64-65).

This thesis is concerned with the description of individual experiences of (architectural) space in a social milieu. Architecture, while considered to be primarily concerned with space as its medium, has a very impoverished (or occasionally, very contorted) verbal language in which to discuss space. The author, as a beginner teacher, noted this in attempts to explore spatial experience with students of architecture, and resolved with their help to generate an appropriate verbal vehicle. The main body of the thesis relates this attempt and accounts for its failure. The Thesis, thus, follows three intertwined streams. 1) A scientific investigation into means for the description of human experience of (architectural) space, using methods developed from Kelly's Personal Construct Theory Repertory Grids. 2) A partially developed spatial analytic language, my personal response to 1) above, which is to be seen as the start of a new research programme that may last many years (the future of which is outlined). 3) An account of a personal learning experience both from, around and through each of these. These streams are organised into three parts. Part 1: Background Studies - into work in associated areas and fields, with an assessment of their relevance to the undertaking presented here. Part 2: The Experiments - attempting (and failing) to create a language, and the transition from verbal to visual, with critical arguments and observations. Part 3: A New Beginning - learning from the failure of Part 2, and the argument for and commencement of a new research programme.

Humans possess a rich semantic knowledge of words and concepts which captures the perceivable physical properties of their real-world referents and their relations. Encoding this knowledge or some of its aspects is the goal of computational models of semantic representation and has been the subject of considerable research in cognitive science, natural language processing, and related areas. Existing models have placed emphasis on different aspects of meaning, depending ultimately on the task at hand. Typically, such models have been used in tasks addressing the simulation of behavioural phenomena, e.g., lexical priming or categorisation, as well as in natural language applications, such as information retrieval, document classification, or semantic role labelling. A major strand of research popular across disciplines focuses on models which induce semantic representations from text corpora. These models are based on the hypothesis that the meaning of words is established by their distributional relation to other words (Harris, 1954). Despite their widespread use, distributional models of word meaning have been criticised as ‘disembodied’ in that they are not grounded in perception and action (Perfetti, 1998; Barsalou, 1999; Glenberg and Kaschak, 2002). This lack of grounding contrasts with many experimental studies suggesting that meaning is acquired not only from exposure to the linguistic environment but also from our interaction with the physical world (Landau et al., 1998; Bornstein et al., 2004). This criticism has led to the emergence of new models aiming at inducing perceptually grounded semantic representations. Essentially, existing approaches learn meaning representations from multiple views corresponding to different modalities, i.e. linguistic and perceptual input. To approximate the perceptual modality, previous work has relied largely on semantic attributes collected from humans (e.g., is round, is sour), or on automatically extracted image features. Semantic attributes have a long-standing tradition in cognitive science and are thought to represent salient psychological aspects of word meaning including multisensory information. However, their elicitation from human subjects limits the scope of computational models to a small number of concepts for which attributes are available. In this thesis, we present an approach which draws inspiration from the successful application of attribute classifiers in image classification, and represent images and the concepts depicted by them by automatically predicted visual attributes. To this end, we create a dataset comprising nearly 700K images and a taxonomy of 636 visual attributes and use it to train attribute classifiers. We show that their predictions can act as a substitute for human-produced attributes without any critical information loss. In line with the attribute-based approximation of the visual modality, we represent the linguistic modality by textual attributes which we obtain with an off-the-shelf distributional model. Having first established this core contribution of a novel modelling framework for grounded meaning representations based on semantic attributes, we show that these can be integrated into existing approaches to perceptually grounded representations. We then introduce a model which is formulated as a stacked autoencoder (a variant of multilayer neural networks), which learns higher-level meaning representations by mapping words and images, represented by attributes, into a common embedding space. In contrast to most previous approaches to multimodal learning using different variants of deep networks and data sources, our model is defined at a finer level of granularity—it computes representations for individual words and is unique in its use of attributes as a means of representing the textual and visual modalities. We evaluate the effectiveness of the representations learnt by our model by assessing its ability to account for human behaviour on three semantic tasks, namely word similarity, concept categorisation, and typicality of category members. With respect to the word similarity task, we focus on the model’s ability to capture similarity in both the meaning and appearance of the words’ referents. Since existing benchmark datasets on word similarity do not distinguish between these two dimensions and often contain abstract words, we create a new dataset in a large-scale experiment where participants are asked to give two ratings per word pair expressing their semantic and visual similarity, respectively. Experimental results show that our model learns meaningful representations which are more accurate than models based on individual modalities or different modality integration mechanisms. The presented model is furthermore able to predict textual attributes for new concepts given their visual attribute predictions only, which we demonstrate by comparing model output with human generated attributes. Finally, we show the model’s effectiveness in an image-based task on visual category learning, in which images are used as a stand-in for real-world objects.

The purpose of this study was to investigate how African heritage language learners in Grade 3 constructed their language identity in a culturally diverse urban classroom. The data reflects the Grade 3 learners’ beliefs and feelings toward being educated in an English-medium school instead of attending a school where they were taught in their home language. A theoretical framework, based on Bronfenbrenner’s ecological system informed my study. The framework provides a holistic picture of how the Grade 3 learners constructed their language identity, as it did not only focus on the learners as such. The ecological system examined the unique aspects of the learners within their microsystem, mesosystem, exosystem, macrosystem and chronosystem. In this exploration of multiple cases studies from a social constructivism perspective, the researcher collaborated with two Grade 3 teachers, seven parents and six Grade 3 African heritage language learners, to create a mosaic by using semi-structured interviews, documents, observational field notes, narrative reflections, photographs, drawings and a reflective journal. Participants’ perceptions of attending an English-medium school revealed how the Grade 3 learners constructed their language identity through assimilation. The findings revealed that the Grade 3 participants did not favour their home language or heritage culture as much as they favoured the Western culture. The following factors influenced the construction of the Grade 3 learners’ language identity: their environments, the school setting, their teachers’ attitude toward teaching in a culturally diverse classroom, their parents’ attitude toward their attending an English-medium school and lastly their friends. / Dissertation (MEd)--University of Pretoria, 2017. / Early Childhood Education / MEd / Unrestricted

This thesis investigates whether studying through English has an effect on the written texts secondary school students produce in their first language, Spanish. Research in bilingual education has tended to focus on students‟ language proficiency and academic achievement as opposed to investigating differences in discourse norms. However, an increased awareness of the role of discourse in language use and the culturally-specific nature of discourse, along with a growing concern over the dominance of the English language in Europe and elsewhere, have widened the range of questions identified as requiring investigation in bilingual education. Popular understanding suggests that English speakers make different choices from Spanish speakers in particular rhetorical situations. Although research into these contrasts can be problematic, there is support for the existence of cultural preference in the selection of options, and specific areas of similarity and difference between Spanish and English discourse have been suggested. This study then looks at one group of secondary three (3º E.S.O.) Spanish students studying through English on a Content and Language Integrated Learning (CLIL) programme in Spain, and compares two sets of argumentative texts written in Spanish by the CLIL students with equivalent texts written by non-CLIL students in the same school. Forty-eight texts are examined in total, and the comparison draws on previous research into discourse differences between Spanish and English texts from contrastive rhetoric, systemic functional linguistics (SFL), and other fields, using tools from SFL for the textual analysis. Areas of analysis include use of clause complexes, multiple Theme and thematic progression, and also genre structure and text organisation strategies for argumentative writing. The main contrasts are found to be in length of t-units, use of simplexes versus complexes, use of multiple Theme, and some issues of text structure. These differences largely correspond to contrasts found in studies comparing written Spanish and written English text, with the CLIL students‟ texts showing features more commonly associated with English writing. Additional data from analysis of the geography textbooks used by the CLIL and non-CLIL students, questionnaires administered to these students, and interviews with their teachers are also used to explore the possible CLIL effect on the students‟ written text. The study discusses how discourse conventions associated with English text in contrast with Spanish text may have influenced the Spanish writing of the CLIL students, considering possibilities including the effect of direct and indirect teaching, and the more general impact of the CLIL programme. Also explored is the question of whether this possible influence of English on Spanish language use is a matter for concern or not, with increased work on language and discourse awareness suggested as a potential response. Methodological issues raised through the study relating to aspects of text analysis and of data collection are also addressed.

This work refers to the topic of bilingual mathematics education. It is focused on monitoring the communicational influences of teacher's and pupil's language, categorization of the disturbing effects and possibilities, how the teacher should prepare for them and how to solve them. A significant role plays the reflection of the teacher, as the results were done on the basis of action research during teaching in England, Germany and the Czech Republic. The teaching was realized in a second language as well for the teacher as for the pupils. The teacher was at the same time experimenter as well as author of this work. The theoretical part results from dealing with scientific literature about bilingualism, bilingual education, methodology of CLIL (Content and Language Integrated Learning), relations between first language, language of mathematics and the foreign language. In the practical part an analysis of lessons performed in mathematics with the subject of probability is carried out. The teaching was focused on playing a game with dice and deciding whether the game is fair or not. The lessons were audio taped and transcriptions of the spoken language were made. An analysis of the language, teacher's mistakes, communicational interference, proportional relations on communication of the pupils and of the...

This diploma thesis, titled "Pythagorean Theorem in CLIL lessons", deals with one of the possible strategies in bilingual education - CLIL method which links teaching of a non- linguistic subject and a foreign language, mathematics and English, in this case. For the realization of CLIL method I have chosen the topic of the introduction to Pythagorean Theorem. The aim of this thesis was to check whether the chosen topic is suitable for CLIL lessons, then the influence of a foreign language on the active involvement of pupils in the lesson and on understanding of the subject matter. At the beginning, I focus on general integration of the content subject and a foreign language and then I describe the preparation of the experiment. The core of this thesis is the project of the CLIL lesson. The project was realized in the experimental groups, the gained data have been processed on the base of observation and analysis of the final individual work and the questionnaire. The results show the potential of the CLIL method as a new possibility for bilingual teaching of content subjects.

Thesis (M.Ed. (Language Education)) --University of Limpopo, 2011 / This dissertation investigates the challenges that history learners in rural black schools are faced with during examinations with the use of English as a language of learning and teaching. The aim of the research is to investigate whether English, which is used as a language of learning and teaching in schools, is a barrier to the learning and teaching of history. It is believed that the use of English has implications on the results that learners obtain during summative assessment. The literature revealed that any language of learning and teaching, which is not the mother tongue of learners, has an impact on the process of learning. Assessment, which determines whether the learning outcomes have been attained or not, is affected by factors such as the type of assessment, for example, formative or summative, the purpose of assessment and the rules and regulations that go along with assessment. In order to gain a greater insight and comprehensive understanding of the research problem, that is, the challenges that are faced by learners during summative assessment, both the qualitative and quantitative approaches were used. A case study design was adopted. The research techniques employed were interviews, document analyses and observations. The findings indicated that the environment in which the learners are learning second language has impact on the learners’ language proficiency. They further indicate that assessment in itself poses challenge to learners, in addition to the challenges caused by the language barriers. It was also indicated that the educators and the learners prefer to use English as the language of learning and teaching, despite all the challenges they face when they use English. The important findings were discussed with the aim of arriving at conclusions, shortcomings recommendations.

This study intended to investigate the language learning strategies used by learners of English as a foreign language, aiming to find the amount of strategies and the domain differences of the strategies used / to reveal the link between strategy use and success levels / and to find out the difference in strategy use between genders and its influence on their achievement in English. 257 (153 male, 104 female) students from Atilim University English Preparatory School participated in the study. At the time of the study all the participants were in the same proficiency level, and were distributed to different classes of the same level. The data were gathered through strategy inventory for language learning (SILL) of Oxford (1990), which was translated to Turkish by Cesur and Fer (2007). The instrument, based on Oxford&rsquo / s (1990) classification of the language learning strategies, is composed of 50 items in six subscales. The participants responded to the inventory before the end of the level they were in. The data were analyzed through SPSS (15.0) to find the relationship of language learning strategies, gender and achievement in learning the target language. To reveal the interconnections between these factors, independent t-tests and an ANOVA test, along with post hoc procedures were performed on the gathered data. The findings of the study revealed that use of language learning strategies are positively effective in success in English, that females were significantly more successful than males in terms of achievement tests, and that they used more language learning strategies in learning English. Depending on the statistical results, it is discovered that there is a significant connection between gender, language learning strategies and achievement in English.

This investigation reports on a study that explores the views of students, teachers and parents about the Communicative Language Teaching (CLT) approach in learning English as a second language in Bangladesh. This study focuses on the improvement of English language outcomes in Bangladesh. Though compulsory for fifteen years of schooling, public examination results indicate that students perform poorly in English. This research is conducted at the secondary schools in Bangladesh where English is compulsory because of its global nature as the second or foreign language. Mainstream students learning English using the Communicative Language Teaching (CLT) approach are facing many challenges. The study employs a mixed methods approach which includes qualitative interviews, semi-structured focus group interviews and observations; and quantitative data involved achievement tests to find gaps between oral and written attainment, in order to determine the effectiveness of CLT developing language skills to communicate in and outside the classrooms. ESL learners in Bangladesh have been using CLT for nearly two decades, but the attainments are not satisfactory particularly in oral language (listening and speaking) compared to written language (reading and writing). Four schools (two high and two low achievement) were selected from two divisional cities according to the public examination results. Five students, all English subject teachers and five parents from each school were invited randomly to participate. Findings indicated a confirmation of the gap between oral and written language achievements and highlighted that CLT is not working effectively to develop communicative competence to the learners. The participants identified several factors causing this. Among these were large class sizes (number of students), an extensive curriculum, insufficient class time (duration), an inappropriate examination system, excessive teacher workload, lack of parent awareness of CLT, and negative relationships between home and school. All of these factors impact on the effectiveness of CLT in Bangladesh. All participants agreed that CLT as an approach is better than other approaches used in Bangladesh to develop English language learning, but the varied interpretation and implementation (practice) makes it less effective. Therefore, they suggested some issues for local and national level policy makers that could enhance the CLT practice in Bangladesh.