To see the other types of publications on this topic, follow the link: Language transfer (Language learning) Germany.
Dissertations / Theses on the topic 'Language transfer (Language learning) Germany'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Language transfer (Language learning) Germany.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Samperio, Sanchez Nahum. "General learning strategies : identification, transfer to language learning and effect on language achievement." Thesis, University of Southampton, 2016. https://eprints.soton.ac.uk/412008/.
Full textAbstract:
Each learner has a set repertoire of general learning strategies that he or she uses despite the learning context. The purpose of this study is to identify the general learning strategies that beginner learners of English have in their repertoire, the transfer of such strategies to language learning and the predictive value they have in language achievement. It is also intended to discover the effect that the teaching of not frequently used general learning strategies have on learners’ language achievement. Additionally, to identify possible differences in strategy types and frequency of strategy use in low and high strategy users as well as high and low achievers of beginner English language learners. This study followed a mixed-methods research methodology by collecting numerical data by means of a 51-item general strategies questionnaire (Martinez- Guerrero 2004) applied in two administrations. The sample consists of 118 beginner English language learners in a language center at a northern Mexican University. Data were analyzed with the SPSS and Excel software. The qualitative data was collected through twenty individual semistructured interviews; furthermore, three one-hour-forty minute strategy instruction sessions were included as the treatment. Quantitative results show that learners have a more frequent use of Achievement Motivation, Cognitive and Concentration strategies; and less frequent use of Study, Study Organization, and Interaction in Class strategies. Qualitative findings indicate that learners use Study and Study organization and Concentration strategies largely in both general learning and language learning. Qualitative data complement and extend the quantitative data gathered in the questionnaire. No significant differences were found on the type of strategies that learners use in general learning contexts and language learning, which suggests that learners transfer their learning strategies from their general strategy repertoire to language learning as the first tools to deal with language learning tasks. A positive correlation was found between learning strategy use and language achievement test scores. Achievement test scores were primarily predicted by the use of Achievement Motivation and Interaction in Class strategies, and to a lesser extent by affective and study strategies. Strategy instruction sessions had no significant increase in the adoption and use of strategies. Furthermore, high and low achievers and strategy users seem to use the same type of strategies; the frequency of strategy use and how they use the strategy represented the difference between types of learners. Finally, a number of language learning strategies emerge from qualitative data that learners use in language learning. Pedagogical implications of the findings of this study provide a potential framework to help not only teachers but also institutions in identifying and teaching new and specific learning strategies.
2
Zhang, Yuan Ph D. Massachusetts Institute of Technology. "Transfer learning for low-resource natural language analysis." Thesis, Massachusetts Institute of Technology, 2017. http://hdl.handle.net/1721.1/108847.
Full textAbstract:
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2017.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 131-142).
Expressive machine learning models such as deep neural networks are highly effective when they can be trained with large amounts of in-domain labeled training data. While such annotations may not be readily available for the target task, it is often possible to find labeled data for another related task. The goal of this thesis is to develop novel transfer learning techniques that can effectively leverage annotations in source tasks to improve performance of the target low-resource task. In particular, we focus on two transfer learning scenarios: (1) transfer across languages and (2) transfer across tasks or domains in the same language. In multilingual transfer, we tackle challenges from two perspectives. First, we show that linguistic prior knowledge can be utilized to guide syntactic parsing with little human intervention, by using a hierarchical low-rank tensor method. In both unsupervised and semi-supervised transfer scenarios, this method consistently outperforms state-of-the-art multilingual transfer parsers and the traditional tensor model across more than ten languages. Second, we study lexical-level multilingual transfer in low-resource settings. We demonstrate that only a few (e.g., ten) word translation pairs suffice for an accurate transfer for part-of-speech (POS) tagging. Averaged across six languages, our approach achieves a 37.5% improvement over the monolingual top-performing method when using a comparable amount of supervision. In the second monolingual transfer scenario, we propose an aspect-augmented adversarial network that allows aspect transfer over the same domain. We use this method to transfer across different aspects in the same pathology reports, where traditional domain adaptation approaches commonly fail. Experimental results demonstrate that our approach outperforms different baselines and model variants, yielding a 24% gain on this pathology dataset.
by Yuan Zhang.
Ph. D.
3
Jin, Di Ph D. Massachusetts Institute of Technology. "Transfer learning and robustness for natural language processing." Thesis, Massachusetts Institute of Technology, 2020. https://hdl.handle.net/1721.1/129004.
Full textAbstract:
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2020Cataloged from student-submitted PDF of thesis.
Includes bibliographical references (pages 189-217).
Teaching machines to understand human language is one of the most elusive and long-standing challenges in Natural Language Processing (NLP). Driven by the fast development of deep learning, state-of-the-art NLP models have already achieved human-level performance in various large benchmark datasets, such as SQuAD, SNLI, and RACE. However, when these strong models are deployed to real-world applications, they often show poor generalization capability in two situations: 1. There is only a limited amount of data available for model training; 2. Deployed models may degrade significantly in performance on noisy test data or natural/artificial adversaries. In short, performance degradation on low-resource tasks/datasets and unseen data with distribution shifts imposes great challenges to the reliability of NLP models and prevent them from being massively applied in the wild. This dissertation aims to address these two issues.
Towards the first one, we resort to transfer learning to leverage knowledge acquired from related data in order to improve performance on a target low-resource task/dataset. Specifically, we propose different transfer learning methods for three natural language understanding tasks: multi-choice question answering, dialogue state tracking, and sequence labeling, and one natural language generation task: machine translation. These methods are based on four basic transfer learning modalities: multi-task learning, sequential transfer learning, domain adaptation, and cross-lingual transfer. We show experimental results to validate that transferring knowledge from related domains, tasks, and languages can improve the target task/dataset significantly. For the second issue, we propose methods to evaluate the robustness of NLP models on text classification and entailment tasks.
On one hand, we reveal that although these models can achieve a high accuracy of over 90%, they still easily crash over paraphrases of original samples by changing only around 10% words to their synonyms. On the other hand, by creating a new challenge set using four adversarial strategies, we find even the best models for the aspect-based sentiment analysis task cannot reliably identify the target aspect and recognize its sentiment accordingly. On the contrary, they are easily confused by distractor aspects. Overall, these findings raise great concerns of robustness of NLP models, which should be enhanced to ensure their long-run stable service.
by Di Jin.
Ph. D.
Ph.D. Massachusetts Institute of Technology, Department of Mechanical Engineering
4
Utgof, Darja. "The Perception of Lexical Similarities Between L2 English and L3 Swedish." Thesis, Linköping University, Department of Culture and Communication, 2008. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-15874.
Full textAbstract:
The present study investigates lexical similarity perceptions by students of Swedish as a foreign language (L3) with a good yet non-native proficiency in English (L2). The general theoretical framework is provided by studies in transfer of learning and its specific instance, transfer in language acquisition.
It is accepted as true that all previous linguistic knowledge is facilitative in developing proficiency in a new language. However, a frequently reported phenomenon is that students see similarities between two systems in a different way than linguists and theoreticians of education do. As a consequence, the full facilitative potential of transfer remains unused.
The present research seeks to shed light on the similarity perceptions with the focus on the comprehension of a written text. In order to elucidate students’ views, a form involving similarity judgements and multiple choice questions for formally similar items has been designed, drawing on real language use as provided by corpora. 123 forms have been distributed in 6 groups of international students, 4 of them studying Swedish at Level I and 2 studying at Level II.
The test items in the form vary in the degree of formal, semantic and functional similarity from very close cognates, to similar words belonging to different word classes, to items exhibiting category membership and/or being in subordinate/superordinate relation to each other, to deceptive cognates. The author proposes expected similarity ratings and compares them to the results obtained. The objective measure of formal similarity is provided by a string matching algorithm, Levenshtein distance.
The similarity judgements point at the fact that intermediate similarity values can be considered problematic. Similarity ratings between somewhat similar items are usually lower than could be expected. Besides, difference in grammatical meaning lowers similarity values significantly even if lexical meaning nearly coincides. Thus, the obtained results indicate that in order to utilize similarities to facilitate language learning, more attention should be paid to underlying similarities.
5
Casula, Camilla. "Transfer Learning for Multilingual Offensive Language Detection with BERT." Thesis, Uppsala universitet, Institutionen för lingvistik och filologi, 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-412450.
Full textAbstract:
The popularity of social media platforms has led to an increase in user-generated content being posted on the Internet. Users, masked behind what they perceive as anonymity, can express offensive and hateful thoughts on these platforms, creating a need to detect and filter abusive content. Since the amount of data available on the Internet is impossible to analyze manually, automatic tools are the most effective choice for detecting offensive and abusive messages. Academic research on the detection of offensive language on social media has been on the rise in recent years, with more and more shared tasks being organized on the topic. State-of-the-art deep-learning models such as BERT have achieved promising results on offensive language detection in English. However, multilingual offensive language detection systems, which focus on several languages at once, have remained underexplored until recently. In this thesis, we investigate whether transfer learning can be useful for improving the performance of a classifier for detecting offensive speech in Danish, Greek, Arabic, Turkish, German, and Italian. More specifically, we first experiment with using machine-translated data as input to a classifier. This allows us to evaluate whether machine translated data can help classification. We then experiment with fine-tuning multiple pre-trained BERT models at once. This parallel fine-tuning process, named multi-channel BERT (Sohn and Lee, 2019), allows us to exploit cross-lingual information with the goal of understanding its impact on the detection of offensive language. Both the use of machine translated data and the exploitation of cross-lingual information could help the task of detecting offensive language in cases in which there is little or no annotated data available, for example for low-resource languages. We find that using machine translated data, either exclusively or mixed with gold data, to train a classifier on the task can often improve its performance. Furthermore, we find that fine-tuning multiple BERT models in parallel can positively impact classification, although it can lead to robustness issues for some languages.
6
Panzeri-Alvarez, Christina. "Metacognition and language transfer for an English language development transitional program." CSUSB ScholarWorks, 1998. https://scholarworks.lib.csusb.edu/etd-project/1780.
Full text
7
Tse, Siu-ching, and 謝兆政. "Cross linguistic influence in polyglots: encoding of the future by L3 learners of Swedish." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2012. http://hub.hku.hk/bib/B4842187X.
Full textAbstract:
The current study aims to investigate the source(s)of cross linguistic influence(CLI)on the production of future encoding strategies by L1 Cantonese learners of L3 Swedish who speak L2 English. In the literature of third language acquisition (TLA) research, the language status of native and non-native languages as well as genetic and (psycho)typological language distance are identified to be important to TLA processes but the current knowledge is insufficient to inform which factor(s) is/are more influential. Given the close genetic distance between English and Swedish and the status of English as a second language, it is hypothesized that CLI on L3 Swedish comes from L2 English rather than L1 Cantonese. Any confirmation or rejection to this hypothesis serves to inform the relationship of language status and language distance to TLA. To test this hypothesis, linguistic background questionnaire and a picture elicitation task are designed to record the production of future ideas in the three languages. Through qualitative and quantitative analyses, mixed sources of CLI from Cantonese and English are identified. An equidistance representation of non-native languages is also identified in which non-native English and Swedish respectively show similar degree of cross linguistic matching in relation to native Cantonese regardless which of them is the principal source of CLI. The hypothesis of differentiation of linguistic representation in the minds of polyglots is therefore proposed and further verification and investigation is required.published_or_final_version
Linguistics
Master
Master of Arts
8
Meftah, Sara. "Neural Transfer Learning for Domain Adaptation in Natural Language Processing." Thesis, université Paris-Saclay, 2021. http://www.theses.fr/2021UPASG021.
Full textAbstract:
Les méthodes d’apprentissage automatique qui reposent sur les Réseaux de Neurones (RNs) ont démontré des performances de prédiction qui s'approchent de plus en plus de la performance humaine dans plusieurs applications du Traitement Automatique de la Langue (TAL) qui bénéficient de la capacité des différentes architectures des RNs à généraliser à partir des régularités apprises à partir d'exemples d'apprentissage. Toutefois, ces modèles sont limités par leur dépendance aux données annotées. En effet, pour être performants, ces modèles neuronaux ont besoin de corpus annotés de taille importante. Par conséquent, uniquement les langues bien dotées peuvent bénéficier directement de l'avancée apportée par les RNs, comme par exemple les formes formelles des langues. Dans le cadre de cette thèse, nous proposons des méthodes d'apprentissage par transfert neuronal pour la construction d'outils de TAL pour les langues peu dotées en exploitant leurs similarités avec des langues bien dotées. Précisément, nous expérimentons nos approches pour le transfert à partir du domaine source des textes formels vers le domaine cible des textes informels (langue utilisée dans les réseaux sociaux). Tout au long de cette thèse nous proposons différentes contributions. Tout d'abord, nous proposons deux approches pour le transfert des connaissances encodées dans les représentations neuronales d'un modèle source, pré-entraîné sur les données annotées du domaine source, vers un modèle cible, adapté par la suite sur quelques exemples annotés du domaine cible. La première méthode transfère des représentations contextuelles pré-entraînées sur le domaine source. Tandis que la deuxième méthode utilise des poids pré-entraînés pour initialiser les paramètres du modèle cible. Ensuite, nous effectuons une série d'analyses pour repérer les limites des méthodes proposées ci-dessus. Nous constatons que, même si l'approche d'apprentissage par transfert proposée améliore les résultats du domaine cible, un transfert négatif « dissimulé » peut atténuer le gain final apporté par l'apprentissage par transfert. De plus, une analyse interprétative du modèle pré-entraîné, montre que les neurones pré-entraînés peuvent être biaisés par ce qu'ils ont appris du domaine source, et donc peuvent avoir des difficultés à apprendre des « patterns » spécifiques au domaine cible. Issu de notre analyse, nous proposons un nouveau schéma d'adaptation qui augmente le modèle cible avec des neurones normalisés, pondérés et initialisés aléatoirement qui permettent une meilleure adaptation au domaine cible tout en conservant les connaissances apprises du domaine source. Enfin, nous proposons une approche d’apprentissage par transfert qui permet de profiter des similarités entre différentes tâches, en plus des connaissances pré-apprises du domaine sourceRecent approaches based on end-to-end deep neural networks have revolutionised Natural Language Processing (NLP), achieving remarkable results in several tasks and languages. Nevertheless, these approaches are limited with their "gluttony" in terms of annotated data, since they rely on a supervised training paradigm, i.e. training from scratch on large amounts of annotated data. Therefore, there is a wide gap between NLP technologies capabilities for high-resource languages compared to the long tail of low-resourced languages. Moreover, NLP researchers have focused much of their effort on training NLP models on the news domain, due to the availability of training data. However, many research works have highlighted that models trained on news fail to work efficiently on out-of-domain data, due to their lack of robustness against domain shifts. This thesis presents a study of transfer learning approaches, through which we propose different methods to take benefit from the pre-learned knowledge on the high-resourced domain to enhance the performance of neural NLP models in low-resourced settings. Precisely, we apply our approaches to transfer from the news domain to the social media domain. Indeed, despite the importance of its valuable content for a variety of applications (e.g. public security, health monitoring, or trends highlight), this domain is still poor in terms of annotated data. We present different contributions. First, we propose two methods to transfer the knowledge encoded in the neural representations of a source model pretrained on large labelled datasets from the source domain to the target model, further adapted by a fine-tuning on few annotated examples from the target domain. The first transfers contextualised supervisedly pretrained representations, while the second method transfers pretrained weights, used to initialise the target model's parameters. Second, we perform a series of analysis to spot the limits of the above-mentioned proposed methods. We find that even if the proposed transfer learning approach enhances the performance on social media domain, a hidden negative transfer may mitigate the final gain brought by transfer learning. In addition, an interpretive analysis of the pretrained model, show that pretrained neurons may be biased by what they have learned from the source domain, thus struggle with learning uncommon target-specific patterns. Third, stemming from our analysis, we propose a new adaptation scheme which augments the target model with normalised, weighted and randomly initialised neurons that beget a better adaptation while maintaining the valuable source knowledge. Finally, we propose a model, that in addition to the pre-learned knowledge from the high-resource source-domain, takes advantage of various supervised NLP tasks
9
Mozafari, Marzieh. "Hate speech and offensive language detection using transfer learning approaches." Electronic Thesis or Diss., Institut polytechnique de Paris, 2021. http://www.theses.fr/2021IPPAS007.
Full textAbstract:
Une des promesses des plateformes de réseaux sociaux (comme Twitter et Facebook) est de fournir un endroit sûr pour que les utilisateurs puissent partager leurs opinions et des informations. Cependant, l’augmentation des comportements abusifs, comme le harcèlement en ligne ou la présence de discours de haine, est bien réelle. Dans cette thèse, nous nous concentrons sur le discours de haine, l'un des phénomènes les plus préoccupants concernant les réseaux sociaux.Compte tenu de sa forte progression et de ses graves effets négatifs, les institutions, les plateformes de réseaux sociaux et les chercheurs ont tenté de réagir le plus rapidement possible. Les progrès récents des algorithmes de traitement automatique du langage naturel (NLP) et d'apprentissage automatique (ML) peuvent être adaptés pour développer des méthodes automatiques de détection des discours de haine dans ce domaine.Le but de cette thèse est d'étudier le problème du discours de haine et de la détection des propos injurieux dans les réseaux sociaux. Nous proposons différentes approches dans lesquelles nous adaptons des modèles avancés d'apprentissage par transfert (TL) et des techniques de NLP pour détecter automatiquement les discours de haine et les contenus injurieux, de manière monolingue et multilingue.La première contribution concerne uniquement la langue anglaise. Tout d'abord, nous analysons le contenu textuel généré par les utilisateurs en introduisant un nouveau cadre capable de catégoriser le contenu en termes de similarité basée sur différentes caractéristiques. En outre, en utilisant l'API Perspective de Google, nous mesurons et analysons la « toxicité » du contenu. Ensuite, nous proposons une approche TL pour l'identification des discours de haine en utilisant une combinaison du modèle non supervisé pré-entraîné BERT (Bidirectional Encoder Representations from Transformers) et de nouvelles stratégies supervisées de réglage fin. Enfin, nous étudions l'effet du biais involontaire dans notre modèle pré-entraîné BERT et proposons un nouveau mécanisme de généralisation dans les données d'entraînement en repondérant les échantillons puis en changeant les stratégies de réglage fin en termes de fonction de perte pour atténuer le biais racial propagé par le modèle. Pour évaluer les modèles proposés, nous utilisons deux datasets publics provenant de Twitter.Dans la deuxième contribution, nous considérons un cadre multilingue où nous nous concentrons sur les langues à faibles ressources dans lesquelles il n'y a pas ou peu de données annotées disponibles. Tout d'abord, nous présentons le premier corpus de langage injurieux en persan, composé de 6 000 messages de micro-blogs provenant de Twitter, afin d'étudier la détection du langage injurieux. Après avoir annoté le corpus, nous réalisons étudions les performances des modèles de langages pré-entraînés monolingues et multilingues basés sur des transformeurs (par exemple, ParsBERT, mBERT, XLM-R) dans la tâche en aval. De plus, nous proposons un modèle d'ensemble pour améliorer la performance de notre modèle. Enfin, nous étendons notre étude à un problème d'apprentissage multilingue de type " few-shot ", où nous disposons de quelques données annotées dans la langue cible, et nous adaptons une approche basée sur le méta-apprentissage pour traiter l'identification des discours de haine et du langage injurieux dans les langues à faibles ressourcesThe great promise of social media platforms (e.g., Twitter and Facebook) is to provide a safe place for users to communicate their opinions and share information. However, concerns are growing that they enable abusive behaviors, e.g., threatening or harassing other users, cyberbullying, hate speech, racial and sexual discrimination, as well. In this thesis, we focus on hate speech as one of the most concerning phenomenon in online social media.Given the high progression of online hate speech and its severe negative effects, institutions, social media platforms, and researchers have been trying to react as quickly as possible. The recent advancements in Natural Language Processing (NLP) and Machine Learning (ML) algorithms can be adapted to develop automatic methods for hate speech detection in this area.The aim of this thesis is to investigate the problem of hate speech and offensive language detection in social media, where we define hate speech as any communication criticizing a person or a group based on some characteristics, e.g., gender, sexual orientation, nationality, religion, race. We propose different approaches in which we adapt advanced Transfer Learning (TL) models and NLP techniques to detect hate speech and offensive content automatically, in a monolingual and multilingual fashion.In the first contribution, we only focus on English language. Firstly, we analyze user-generated textual content to gain a brief insight into the type of content by introducing a new framework being able to categorize contents in terms of topical similarity based on different features. Furthermore, using the Perspective API from Google, we measure and analyze the toxicity of the content. Secondly, we propose a TL approach for identification of hate speech by employing a combination of the unsupervised pre-trained model BERT (Bidirectional Encoder Representations from Transformers) and new supervised fine-tuning strategies. Finally, we investigate the effect of unintended bias in our pre-trained BERT based model and propose a new generalization mechanism in training data by reweighting samples and then changing the fine-tuning strategies in terms of the loss function to mitigate the racial bias propagated through the model. To evaluate the proposed models, we use two publicly available datasets from Twitter.In the second contribution, we consider a multilingual setting where we focus on low-resource languages in which there is no or few labeled data available. First, we present the first corpus of Persian offensive language consisting of 6k micro blog posts from Twitter to deal with offensive language detection in Persian as a low-resource language in this domain. After annotating the corpus, we perform extensive experiments to investigate the performance of transformer-based monolingual and multilingual pre-trained language models (e.g., ParsBERT, mBERT, XLM-R) in the downstream task. Furthermore, we propose an ensemble model to boost the performance of our model. Then, we expand our study into a cross-lingual few-shot learning problem, where we have a few labeled data in target language, and adapt a meta-learning based approach to address identification of hate speech and offensive language in low-resource languages
10
Mau, Pui-sze Priscilla. "Cross-language transfer of phonological awareness in Chinese-English bilinguals." Click to view the E-thesis via HKUTO, 2006. http://sunzi.lib.hku.hk/hkuto/record/B36889301.
Full text
11
Yuan, Boping. "Directionality of difficulty in second language acquisition of Chinese and English." Thesis, University of Edinburgh, 1993. http://hdl.handle.net/1842/17576.
Full textAbstract:
This thesis is concerned with the investigation of directionality of difficulty in second language acquisition (SLA) by Chinese-speaking learners learning English as a foreign language (EFL) and by English-speaking learners learning Chinese as a foreign language (CFL) . Chinese allows both subject PRO in finite clauses and object pro. However, subject PRO in finite clauses and object pro give rise to ungrammaticality in English. Unlike Chinese, in which topics can be base-generated, English does not allow a base-generated topic. Chinese and English are also different in that while English reflexives can only take a local subject in finite clauses as their antecedent (thus a short-distance reflexive), the Chinese reflexive ziji can take the matrix subject as well as the embedded subject as its antecedent (thus a long-distance reflexive) . With respect to these differences between the two languages, our focus is on whether it is more difficult for CFL learners to acquire subject PRO, object pro, base-generated topics and the long-distance reflexive in the acquisition of Chinese than for EFL learners to unlearn subject PRO, object pro, base-generated topics and the long-distance reflexive in the acquisition of English. The results of our study suggest that there is no single direction of difficulty in the SLA of Chinese and English. In terms of object pro, the direction of difficulty is from Chinese to English. However, in acquiring and unlearning the subject PRO, neither CFL learners nor EFL learners seem to have much difficulty. As for base-generated topics, it is found that the acquisition of this feature by CFL learners is more difficult than the unlearning of this feature by EFL learners. The results concerning the acquisition of the Chinese long-distance reflexive ziji by CFL learners suggest that a lack of long-distance binding for ziji is fossilized in these learners' interlanguage (IL) grammars of Chinese. Based on the findings in this study, we argue that the directionality of difficulty in SLA can only be studied with respect to individual language features and that the mere existence of relevant positive evidence in the input is not a guarantee that there will be a change in the learner's IL grammar. There are many factors involved in deciding the direction of difficulty in SLA. These factors include the availability of informative evidence to the learner, the possibility that the learner makes use of the evidence available for the restructuring of his IL grammar of the target language, the learners' ability to process the relevant data in the input, and the interaction between the structure in the learners' L1 and the inherent developmental stage of the target language.
12
Conradie, Simone. "Verb movement parameters in Afrikaans : investigating the Full Transfer Full Access hypothesis." Thesis, McGill University, 2005. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=85899.
Full textAbstract:
This thesis sets out to test the Full Transfer Full Access hypothesis (FTFA), which claims (i) that second language (L2) learners start out with the parameter settings instantiated in their first language (L1) grammars ('full transfer') and (ii) that they can subsequently reset parameters to the target L2 settings where these differ from the L1 settings, provided the required (triggering) positive evidence is available in the L2 input ('full access').Three studies on the L2 acquisition of two verb movement parameters, the V2 parameter and the Split-IP parameter (SIP), are reported. The first study investigates 'full access', testing whether English-speaking learners of Afrikaans, who started acquiring the L2 in early childhood and are thus child L2 learners, can reset the two parameters. The second study investigates 'full transfer' and 'full access' by testing whether English-speaking and German-speaking learners start out with different settings of the two parameters and whether the English-speaking learners can reset the parameters. All participants in this study are adult L2 learners, which facilitates a comparison of child L2 acquisition (first study) with adult L2 acquisition. The third study investigates whether Afrikaans-speaking learners of French can acquire knowledge of the ungrammaticality of certain construction types that are allowed in their L1 but not in the L2 (although the languages share the same parameter setting), despite the fact that there seems to be no positive evidence to this effect in the L2 input. It is argued that, taken together, the studies provide evidence in support of the FTFA.
The original contribution of this thesis lies in (i) investigating both verb movement parameters (instead of only one), (ii) providing a thorough discussion of the relevant syntactic properties of Afrikaans, (iii) investigating the L2 acquisition of Afrikaans, and (iv) addressing the question of how learners go about acquiring a parameter setting ([+SIP]) in cases where both the L1 and the L2 share the parameter setting but the L1 exhibits a superset of the properties exhibited by the L2.
13
Kwan, Chung-hin. "An investigation of English errors of Hong Kong secondary 1 and secondary 5 students and their relationship with mother tongue Cantonese transfer." Hong Kong : University of Hong Kong, 1997. http://sunzi.lib.hku.hk/hkuto/record.jsp?B18939326.
Full text
14
Popa, Simona. "Language transfer in second language acquisition. Some effects of L1 instruction (Romanian) on L2/L3 learning (Catalan/Spanish)." Doctoral thesis, Universitat de Lleida, 2016. http://hdl.handle.net/10803/380548.
Full textAbstract:
In migration contexts, the diversity of languages in contact triggers the processes of second language (L2) acquisition and language transfer; as well as drawing attention to the importance of mother tongue (L1) maintenance.
The present study examines the processes of L2 acquisition (Catalan and Spanish), L1 (Romanian) maintenance, and L1transfer, in the case of 130 immigrant Romanian students, as well as the effect of attendance at L1 classes and length of residence on the three languages analysed. Accordingly, three parallel language competence tests were applied in seven public schools of Compulsory Secondary Education in Catalonia. Generally, the results indicate that the language transfer from the L1 to the L2s occurs and a longer length of residence facilitates the learning of Catalan and Spanish, but, at the same time, hinders the level of competence in L1. Also, attendance at Romanian classes seems to influence the maintenance of the mother tongue and the acquisition of the second languages.En contextos de migració, la diversitat de llengües en contacte esdevé processos d’adquisició de segones llengües (L2) i de transferència lingüística; a més de revifar el debat sobre la importància del manteniment de la llengua materna (L1). En el següent treball s’exploren els processos d’adquisició de l’L2 (català i castellà), del manteniment de l’L1 (romanès) i de la transferència lingüística de l’L1, de 130 estudiants immigrants d’origen romanès; així com l’efecte d’assistir a classes d’L1 i el temps d’estada, en les tres llengües estudiades. Per a aquest propòsit, s’han aplicat tres proves paral•leles de competència lingüística en set instituts d’Educació Secundària Obligatòria de Catalunya. A nivell general, els resultats indiquen que la transferència lingüística de l’L1 a les L2s sorgeix i que un major temps d’estada afavoreix l’aprenentatge del català i del castellà però, al mateix temps, va en detriment del nivell del coneixement adquirit en la seva L1. Així mateix, l’assistència a classes de romanès sembla influir en el manteniment de la seva llengua materna i en l’aprenentatge de segones llengües.
En contextos de migración, la diversidad de lenguas en contacto desencadena procesos de adquisición de segundas lenguas (L2) y de transferencia lingüística; además de reavivar el debate sobre la importancia del mantenimiento de la lengua materna (L1). En el siguiente trabajo se exploran los procesos de adquisición de L2 (catalán y castellano), del mantenimiento de la L1 (rumano) y de la transferencia lingüística de la L1, de 130 estudiantes inmigrantes de origen rumano, así como el efecto de asistir a clases de L1 y el tiempo de estancia, en las tres lenguas estudiadas. Para ello, se han aplicado tres pruebas paralelas de competencia lingüística en siete institutos de Educación Secundaria Obligatoria de Cataluña. A nivel general, los resultados indican que se da la influencia de la L1 en las L2 y que un mayor tiempo de estancia favorece el aprendizaje del catalán y del castellano, pero, a su vez, va en detrimento del nivel de conocimiento adquirido en su L1. Asimismo, la asistencia a clases de rumano parece influir en el mantenimiento de su lengua materna y en el aprendizaje de segundas lenguas.
15
Kung, Shui Man Jessica. "Language maintenance or language shift ? : a study of South Asian ethnic minorities' Chinese language learning in Hong Kong." HKBU Institutional Repository, 2012. http://repository.hkbu.edu.hk/etd_ra/1353.
Full text
16
Mau, Pui-sze Priscilla, and 繆佩詩. "Cross-language transfer of phonological awareness in Chinese-English bilinguals." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2006. http://hub.hku.hk/bib/B36889301.
Full text
17
Halvardsson, Gustaf, and Johanna Peterson. "Interpretation of Swedish Sign Language using Convolutional Neural Networks and Transfer Learning." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-277859.
Full textAbstract:
The automatic interpretation of signs of a sign language involves image recognition. An appropriate approach for this task is to use Deep Learning, and in particular, Convolutional Neural Networks. This method typically needs large amounts of data to be able to perform well. Transfer learning could be a feasible approach to achieve high accuracy despite using a small data set. The hypothesis of this thesis is to test if transfer learning works well to interpret the hand alphabet of the Swedish Sign Language. The goal of the project is to implement a model that can interpret signs, as well as to build a user-friendly web application for this purpose. The final testing accuracy of the model is 85%. Since this accuracy is comparable to those received in other studies, the project’s hypothesis is shown to be supported. The final network is based on the pre-trained model InceptionV3 with five frozen layers, and the optimization algorithm mini-batch gradient descent with a batch size of 32, and a step-size factor of 1.2. Transfer learning is used, however, not to the extent that the network became too specialized in the pre-trained model and its data. The network has shown to be unbiased for diverse testing data sets. Suggestions for future work include integrating dynamic signing data to interpret words and sentences, evaluating the method on another sign language’s hand alphabet, and integrate dynamic interpretation in the web application for several letters or words to be interpreted after each other. In the long run, this research could benefit deaf people who have access to technology and enhance good health, quality education, decent work, and reduced inequalities.Automatisk tolkning av tecken i ett teckenspråk involverar bildigenkänning. Ett ändamålsenligt tillvägagångsätt för denna uppgift är att använda djupinlärning, och mer specifikt, Convolutional Neural Networks. Denna metod behöver generellt stora mängder data för att prestera väl. Därför kan transfer learning vara en rimlig metod för att nå en hög precision trots liten mängd data. Avhandlingens hypotes är att utvärdera om transfer learning fungerar för att tolka det svenska teckenspråkets handalfabet. Målet med projektet är att implementera en modell som kan tolka tecken, samt att bygga en användarvänlig webapplikation för detta syfte. Modellen lyckas klassificera 85% av testinstanserna korrekt. Då denna precision är jämförbar med de från andra studier, tyder det på att projektets hypotes är korrekt. Det slutgiltiga nätverket baseras på den förtränade modellen InceptionV3 med fem frysta lager, samt optimiseringsalgoritmen mini-batch gradient descent med en batchstorlek på 32 och en stegfaktor på 1,2. Transfer learning användes, men däremot inte till den nivå så att nätverket blev för specialiserat på den förtränade modellen och dess data. Nätverket har visat sig vara ickepartiskt för det mångfaldiga testningsdatasetet. Förslag på framtida arbeten inkluderar att integrera dynamisk teckendata för att kunna tolka ord och meningar, evaluera metoden på andra teckenspråkshandalfabet, samt att integrera dynamisk tolkning i webapplikationen så flera bokstäver eller ord kan tolkas efter varandra. I det långa loppet kan denna studie gagna döva personer som har tillgång till teknik, och därmed öka chanserna för god hälsa, kvalitetsundervisning, anständigt arbete och minskade ojämlikheter.
18
Tipa, Thep-Ackrapong Steffensen Margaret S. "Fossilization a case study of practical and theoretical parameters /." Normal, Ill. Illinois State University, 1990. http://wwwlib.umi.com/cr/ilstu/fullcit?p9101128.
Full textAbstract:
Thesis (D.A.)--Illinois State University, 1990.Title from title page screen, viewed November 15, 2005. Dissertation Committee: Margaret Steffensen (chair), Irene Brosnahan, Bruce Hawkins, Maurice Scharton, Sandra Metts. Includes bibliographical references (leaves 211-218) and abstract. Also available in print.
19
Mostert, Annamarie. "A language programme evaluation : English as language of learning and teaching / A. Mostert." Thesis, North-West University, 2008. http://hdl.handle.net/10394/2032.
Full text
20
Levesque, Guy-Luc. "Lexico-Semantic Influence in Interlingual Transfer." PDXScholar, 1994. https://pdxscholar.library.pdx.edu/open_access_etds/4771.
Full textAbstract:
The present study replicates research by Tomoko Takahashi (1984) on lexico-semantic patterns used by students in an acquisition poor environment. The purpose of the current study was to determine how an acquisition rich environment affects learners' use of four lexico-semantic patterns: congruence occurs when the Ll definition of a lexical item forms a one-to-one correspondence with the L2 lexical item; convergence occurs when the Ll lexical item has broader applications than the L2 lexical item; divergence occurs when the L2 lexical item has broader applications than the Ll lexical item; and semantic gap occurs when the Ll lexical item has no appropriate corresponding L2 lexical item (Takahashi, 1984). The instrument, a lexico-semantics test, is the same instrument used in Takahashi's study. It was designed to measure which patterns are most frequently used by Japanese EFL students learning English. The results, unlike Takahashi's, suggest that beginning and advanced ESL students use the four patterns equally well. No significant difference was found between the two groups. These results are contrary to what had been expected. However, they show that the proposed hierarchical order of difficulty of congruence, convergence, divergence and semantic gap is the same in both studies. The results also indicate that the acquisition· rich environment seems to dramatically improve beginners' performance of the four patterns. Since the instrument was designed for EFL students (an acquisition poor environment) it may not have fully challenged the advanced ESL students (an acquisition rich environment) while challenging the beginning students. This may have been due to the fact that the students in the present study received a great deal of input from the acquisition rich environment, which could account for their increased ability to restructure hypotheses about L2 vocabulary items. In conclusion, more studies are needed to determine the complete role of the four lexico-semantic patterns in vocabulary acquisition. An expanded follow up study that fully tests the advanced and beginning ESL learners' ability could determine whether both groups progress along a language continuum with respect to the use of the four lexico-semantic patterns. Furthermore, although the patterns may serve, in a limited capacity, as indicators of a learner's difficulties in vocabulary acquisition, a wider body of research is needed before they can be applied in a language learning environment.
21
Lategano, Antonio. "Image-based programming language recognition." Master's thesis, Alma Mater Studiorum - Università di Bologna, 2020. http://amslaurea.unibo.it/22208/.
Full textAbstract:
Nel presente lavoro di tesi viene affrontato per la prima volta il problema della classificazione dei linguaggi di programmazione mediante approcci image-based. Abbiamo utilizzato alcune Convolutional Neural Network pre-addestrate su task di classificazione di immagini, adattandole alla classificazione di immagini contenenti porzioni di codice sorgente scritto in 149 diversi linguaggi di programmazione. I nostri risultati hanno dimostrato che tali modelli riescono ad apprendere, con buone prestazioni, le feature lessicali presenti nel testo. Aggiungendo del rumore, tramite modifica dei caratteri presenti nelle immagini, siamo riusciti a comprendere quali fossero i caratteri che meglio permettevano ai modelli di discriminare tra una una classe e l’altra. Il risultato, confermato tramite l’utilizzo di tecniche di visualizzazione come la Class Activation Mapping, è che la rete riesce ad apprendere delle feature lessicali di basso livello concentrandosi in particolare sui simboli tipici di ogni linguaggio di programmazione (come punteggiatura e parentesi), piuttosto che sui caratteri alfanumerici.
22
Becker, Thomas. "Contrastive analysis for teaching Koine Greek case syntax to Russian-speaking students." Theological Research Exchange Network (TREN), 2007. http://www.tren.com/search.cfm?p023-0209.
Full text
23
Kayama, Yuhko. "Second language acquisition of Japanese relative clauses." Thesis, McGill University, 2005. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=100636.
Full textAbstract:
This thesis investigates the second language acquisition of Japanese relative clauses (RCs) by native speakers of English or Korean. Researchers have argued that Universal Grammar (UG) can be accessed in adult second language (L2) acquisition. The Full Transfer Full Access (FTFA) Hypothesis (Schwartz and Sprouse 1994, 1996) claims that after initial transfer from L1, learners reset their parameter values and are able to acquire L2 properties that are different from their L1. Japanese and English relativization: while Japanese relativization does not involve movement in the syntax (Kuno 1973, Saito 1985, etc.), English relativization involves obligatory operator movement and is subject to movement constraints like Subjacency. Because of Subjacency, certain RC structures are prohibited in English. In Japanese, on the other hand, a zero pronominal, pro, is base-generated in the gap position of an RC and coindexed with the relative head noun, and thus long distance relativization is possible. I argue that the presence or absence of pro is also a parametric difference between Japanese and English; pro is present in Japanese, but not in English. In order for English speakers to acquire Japanese RCs, they need to reset the values of these two parameters.Experimental studies were conducted with high-intermediate and advanced learners of Japanese (15 English speakers and 18 Korean speakers). Tests examined the learners' knowledge of the two grammatical properties in Japanese---namely, the lack of wh-movement and the presence of pro. Korean is different from English but similar to Japanese with respect to the parameters in question. Following FTFA, it is hypothesized that English-speaking learners initially transfer their L1 values, and that eventually they are able to switch parameters to the L2 values by accessing UG. The results of several tasks (including interpretation tasks and judgment tasks) confirm this hypothesis; while Korean speakers generally performed well irrespective of proficiency levels, English-speaking intermediate learners transferred their L1 values, failing to accept grammatical Japanese sentences that are not possible in English. English-speaking advanced learners, on the other hand, performed better than intermediate learners, and exhibited evidence that they had acquired the two properties of Japanese, supporting FTFA.
24
Luk, Yuen-chau. "The role of phonological awareness in second language reading." Click to view the E-thesis via HKUTO, 2006. http://sunzi.lib.hku.hk/hkuto/record/B36890613.
Full text
25
Gürel, Ayşe. "Linguistic characteristics of second language acquisition and first language attrition : Turkish overt versus null pronouns." Thesis, McGill University, 2002. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=38201.
Full textAbstract:
This thesis investigates the binding of overt and null subject pronouns in second language (L2) acquisition and first language (L1) attrition of Turkish. The aim is to provide a comparative investigation of language transfer effects in the ultimate state of the L2 and L1 grammar. More specifically, it examines transfer effects from English L1 and English L2 into the grammars of Turkish L2 and Turkish L1, respectively.In this thesis, I propose that the Subset Condition (Berwick, 1985; Manzini & Wexler, 1987) can account for transfer phenomena observed in both L2 acquisition and L1 attrition. I argue that the subset relation that holds between the L1 and the L2 can be a predictor for the extent and duration of cross-linguistic transfer in L2 acquisition and L1 attrition. In other words, whether or not a particular property will resist L2 acquisition and undergo L1 attrition can be determined by looking at the subset relationship between the L1 and the L2 with respect to that property.
The prediction is that in configurations where the 'influencing language' (L1 in L2 acquisition and L2 in L1 attrition) is the superset of the 'affected language' (L2 in L2 acquisition and L1 in L1 attrition), L1 transfer effect will persist in L2 acquisition and we will see more signs of L2 transfer into the L1 grammar, resulting in more attrition effects.
Pronominal binding is chosen to investigate such cross-linguistic transfer effects. English and Turkish differ with respect to governing domains and types of pronominals present in two languages. Turkish, being a pro-drop language, allows null subject pronouns in main and embedded clauses. It also has a special type of anaphoric pronominal, kendisi, for which English has no corresponding form.
Two experiments were conducted to test L2 acquisition and L1 attrition of binding properties of Turkish overt and null subject pronouns under the influence of English. Participants included native English-speakers living in Turkey (end-state L2 Turkish speakers) and native Turkish-speakers living in North America (end-state L2 English speakers). Overall, results obtained from the two studies reveal cross-linguistic transfer effects in the manner predicted. In particular, properties of English overt pronouns (e.g., him/her) are transferred onto the overt Turkish pronoun o in L2 acquisition and in attrition, whereas properties of the Turkish null pronoun and the anaphoric pronominal kendisi are unaffected by English.
26
Berdugo, Oviedo Gloria. "Sentence processing strategies by bilinguals." Thesis, McGill University, 1990. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=60031.
Full textAbstract:
Sentence processing strategies by 10 bilingual speakers (Spanish: L1 and English: L2) were compared to 10 monolingual speakers of Spanish and 10 of English. Word-by-word reading times for ambiguous and non-ambiguous sentences were recorded. A paraphrasing task was used to determine the attachment preferences for a prepositional phrase placed after the verb. Multivariate analyses of the data showed that bilingual speakers combined strategies from both languages when processing ambiguous sentences in L2 that are non-ambiguous in L1. Furthermore, overall results showed that the three groups were sensitive to the attachment of the prepositional phrase. The results are discussed in terms of their implications for theories of sentence processing and of language transfer.
27
Chan, Carol Suk Oi. "A cross-sectional study of syntactic errors in English composition by ESL students in Hong Kong : aspects of negative transfer." HKBU Institutional Repository, 2002. http://repository.hkbu.edu.hk/etd_ra/501.
Full text
28
Inagaki, Shunji. "Transfer and learnability in second language argument structure : motion verbs with locationaldirectional PPs in L2 English and Japanese." Thesis, McGill University, 2002. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=38492.
Full textAbstract:
This thesis investigates how the outcomes of the acquisition of second language (L2) argument structure will vary depending on the nature of the learner's first language (L1). The focus is on motion verbs appearing with a prepositional/postpositional phrase that expresses the final endpoint of the motion (goal PP). In English, manner-of-motion verbs (e.g., walk ) and directed motion verbs (e.g., go) can appear with a goal PP as in John walked (went ) to school. In contrast, Japanese allows only directed motion verbs to occur with a goal PP. Thus, Japanese motion verbs with goal PPs form a subset of their English counterparts. I propose an analysis of these crosslinguistic differences in terms of different incorporation patterns in lexical-syntax (Hale & Keyser, 1993). L1 transfer and learnability considerations (White, 1991b), then, lead me to hypothesize that Japanese-speaking learners of English will be able to acquire the L2 representation on the basis of positive evidence, but that English-speaking learners of Japanese will have difficulty acquiring the L2 representation due to the lack of positive data motivating the restructuring of the L1 representation to the L2. A series of experiments tested these hypotheses using grammaticality judgment and picture-matching tasks. Results in general supported this prediction, suggesting that whether the L1 constitutes a subset of the L2 or vice versa indeed affects the outcomes of L2 argument structure. The results indicate full involvement of L1 and UG in L2 acquisition, thus supporting the Full-Transfer/Full-Access model of L2 acquisition (Schwartz & Sprouse, 1994).
29
Keung, Yuen-ching. "Transfer of cognitive skills in learning to read Chinese (L1) and English (L2) among HK elementary students." Click to view E-thesis via HKUTO, 2006. http://sunzi.lib.hku.hk/hkuto/record/B37090720.
Full text
30
Ananth, Priya. "Acquisition of tense and aspect in Toki 'when' clauses in Japanese as a second/foreign language." Columbus, Ohio : Ohio State University, 2007. http://rave.ohiolink.edu/etdc/view?acc%5Fnum=osu1187208767.
Full text
31
Graber, Elizabeth. "Evaluating Speech-Language Pathologists’ Transfer of Continuing Education Learning Outcomes to Their Practice Settings." Cleveland State University / OhioLINK, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=csu1593547874361376.
Full text
32
Lam, Ka-ki Fiona. "The verb use in Cantonese monolingual and Cantonese-English bilingual children." Click to view the E-thesis via HKU Scholors Hub, 2005. http://lookup.lib.hku.hk/lookup/bib/B38279277.
Full textAbstract:
Thesis (B.Sc)--University of Hong Kong, 2005."A dissertation submitted in partial fulfilment of the requirements for the Bachelor of Science (Speech and Hearing Sciences), The University of Hong Kong, June 30, 2005." Also available in print.
33
Wijayanto, Agus. "Interlanguage pragmatics of refusal strategies by Javanese EFL learners." Thesis, University of Aberdeen, 2011. http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=183672.
Full textAbstract:
The study investigated similarities and differences between refusal strategies conducted by British native speakers of English (NSE) and Javanese learners of English (JLE). The data were elicited, using discourse completion tasks (DCT), from 20 NSE and 50 JLE. Comparative data concerning refusal strategies in Javanese were elicited from 35 native speakers of Javanese (NJ) to provide a baseline for investigating the extent to which differences between JLE and NSE could be explained by the influence of L1 pragmatics. The refusal strategies were classified based on modified refusal taxonomy by Beebe et al. (1990) and were analysed into sociopragmatic and pragmalinguistic strategies. Z test and Chi Square (χ2) were applied to test the statistical significance of differences between JLE and NSE usage. The study found that all three groups employed broadly similar sequential orders, frequencies of occurrences, and contents of both semantic formulae and adjuncts. Some differences were found, however, in which the strategies of the two Javanese groups (JLE and NJ) were more alike than either was to NSE. These findings suggest that distinctive JLE usages (i.e. different from NSE) are either due to the influence of L1 (negative pragmatic transfer) or simply deviation (idiosyncratic usage). The former occurred mainly in the utilization of politeness strategies by the Javanese groups. The salient elements of Javanese cultural values and their relation to the expression of politeness are discussed in some detail, and are shown to be reflected in the English of Javanese learners. The latter (deviations) appeared to arise from a conflict between JLE speakers’ notions of “correct” grammar and word meanings, on the one hand, and the pragmalinguistic demands of the interaction, on the other hand.
34
Brown, Cynthia A. 1967. "Acquisition of segmental structure : consequences for speech perception and second language acquisition." Thesis, McGill University, 1997. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=34704.
Full textAbstract:
Through an investigation of the acquisition of feature geometric representations in first and second language acquisition, this dissertation demonstrates how the Feature Geometry theory contained in Universal Grammar actively guides and constrains the acquisition of segmental representations by children. In addition, it demonstrates how the mature feature geometry in a speaker's mental grammar restricts the range of nonnative phonemic contrasts that he or she will be sensitive to in the input and, hence, able to acquire as an L2 learner.Three related areas of research are explored and integrated in this work: first, a theoretical study explores the feature-geometric representation of sonorant and non-sonorant laterals, based on their behavior in a variety of phonological processes cross-linguistically, and suggests that [lateral] is not a phonological feature, but rather that laterality is a phonetic property that derives from a specific feature-geometric representation; second, an experimental study investigates the acquisition of phonemic contrasts by English children and demonstrates that segmental representations are acquired in a uniform order that is consistent with properties of Feature Geometry; finally, a series of experimental studies examines the perception and acquisition of the English /l-r/, /b-v/, /p-f/, /f-v/ and /s-theta/ contrasts by native speakers of Japanese, Mandarin Chinese and Korean.
The findings from each of these studies are synthesized to obtain a comprehensive picture of how segmental representations are acquired and how this L1 knowledge impinges on the acquisition of L2 phonemes: it is argued that the monotonic acquisition of feature-geometric structure by young children restricts their sensitivity to particular non-native contrasts, and the continued operation of this existing feature geometry in adult speech perception constrains which non-native contrasts adult learners will be sensitive to in the L2 input and, therefore, capable of acquiring; the circumstances in which the native grammar facilitates perception of non-native contrasts and in which acquisition is possible are also discussed.
35
Alasfour, Aisha Saud. "Grammatical Errors by Arabic ESL Students| An Investigation of L1 Transfer through Error Analysis." Thesis, Portland State University, 2018. http://pqdtopen.proquest.com/#viewpdf?dispub=10826886.
Full textAbstract:
This study investigated the effect of first language (L1) transfer on Arabic ESL learners’ acquisition of the relative clauses, the passive voice and the definite article. I used Contrastive Analysis (CA) and Error Analysis (EA) to analyze 50 papers written by Arabic ESL students at the ACTFL Advanced Mid proficiency level. The analysis was paired with interviews with five advanced students to help determine whether L1 transfer was, in fact, influencing students’ errors predicted by CA.
Students in this study made L1 errors along with other errors. Although no statistical difference was found between the frequency of transfer and other (non-transfer) errors, L1 transfer errors were still common for many learners in this data. The frequency of the relative clause L1 transfer errors was slightly higher than other errors. However, passive voice L1 errors were as frequent as other errors whereas definite article L1 errors were slightly less frequent than other errors. The analysis of the interviews suggested that L1 still played a crucial role in influencing learners errors.
The analysis also suggested that the frequency of transfer errors in the papers used in this study might have been influenced by CA-informed instruction students received and students’ language level. Specifically, learners reported that both factors helped them reduce the frequency of L1 transfer errors in their writing.
The teaching implications of this study include familiarizing language instructors with possible sources of errors for Arabic ESL learners. Language instructors should try to identify sources of errors by conducting their own analyses or consulting existing literature on CA paired with EA. Finally, I recommend adopting a CA-informed instruction to help students reduce and overcome errors that are influenced by their L1.
36
Child, Michael W. "Cross-Linguistic Influence in L3 Portuguese Acquisition: Language Learning Perceptions and the Knowledge and Transfer of Mood Distinctions by Three Groups of English-Spanish Bilinguals." Diss., The University of Arizona, 2014. http://hdl.handle.net/10150/333340.
Full textAbstract:
Interest in Portuguese has steadily increased over the last decade in universities across both North and South America (Carvalho 2002, 2011), principally among Spanish speakers. Generally speaking, Portuguese for Spanish-speakers courses have been designed around the theory that Spanish-speaking students will benefit from cross-linguistic influence (CLI, or transfer) due to the typological similarity that exists between Portuguese and Spanish (see Júdice, 2000). Related to this, the Typological Primacy Model, or TPM (Rothman, 2011), states that CLI in L3 acquisition principally comes from the language that is perceived to be typologically similar to the target language (psycho-typology, see Kellerman, 1983), resulting in both positive and negative transfer. Although there is a high degree of typological similarity between Spanish and Portuguese, it is unknown whether or not this linguistic proximity is equally salient to all learners and whether or not learners view this linguistic proximity as an advantage or a disadvantage when learning Portuguese. Furthermore, some studies have suggested that the context in which one's Spanish is acquired may play a role in the different types of CLI evident among different Spanish-speaking learners of Portuguese (e.g., Carvalho & da Silva, 2006; Johnson, 2004; Koike & Gualda, 2008). Consequently, Carvalho (2002, 2011) has called for more empirical evidence to shed light on the nature of CLI between Spanish and Portuguese. This dissertation, consisting of three main studies, seeks to answer this call by examining the effects of language background on L3 Portuguese acquisition among three groups of Spanish-speaking bilinguals: L1 Spanish (L1S) bilinguals, L2 Spanish (L2S) bilinguals, and heritage speakers of Spanish (HS bilinguals). Results from both quantitative and qualitative analyses of questionnaire data from the first study suggest that although all participants view Spanish as the principal source of CLI in L3 Portuguese acquisition, L2S bilinguals and HS bilinguals perceive the role of Spanish as significantly more facilitative when learning Portuguese than do L1S bilinguals. The second and third studies used a sentence completion task and a preference/grammaticality judgment task (see Ayoun, 2000) to measure bilingual students' knowledge of mood distinctions in Spanish in obligatory and non-obligatory contexts, respectively, and how they transfer that knowledge to Portuguese. Results indicate that the L2S group scored significantly lower on both measures of mood distinctions in obligatory contexts in Spanish, but transferred over more of their knowledge to Portuguese than either the L1S or HS groups. Similarly, results suggest that L2S bilinguals do not understand the variable nature of mood distinctions in non-obligatory environments, but show almost identical strategies of mood selection in both Spanish and Portuguese. In contrast, L1S and HS bilinguals display knowledge of the variable nature of mood distinctions in Spanish in these contexts but show marked differences in mood selection between the Spanish and Portuguese tasks. The results of these studies contribute to L3 acquisition literature by emphasizing the complexity involved in determining the role of the background languages in CLI and by highlighting the importance of the context of acquisition in CLI. In addition, the results provide more empirical evidence regarding the differences between how different groups of Spanish-speaking bilinguals transfer their knowledge when acquiring L3 Portuguese.
37
Bäck, Jesper. "Domain similarity metrics for predicting transfer learning performance." Thesis, Linköpings universitet, Interaktiva och kognitiva system, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-153747.
Full textAbstract:
The lack of training data is a common problem in machine learning. One solution to thisproblem is to use transfer learning to remove or reduce the requirement of training data.Selecting datasets for transfer learning can be difficult however. As a possible solution, thisstudy proposes the domain similarity metrics document vector distance (DVD) and termfrequency-inverse document frequency (TF-IDF) distance. DVD and TF-IDF could aid inselecting datasets for good transfer learning when there is no data from the target domain.The simple metric, shared vocabulary, is used as a baseline to check whether DVD or TF-IDF can indicate a better choice for a fine-tuning dataset. SQuAD is a popular questionanswering dataset which has been proven useful for pre-training models for transfer learn-ing. The results were therefore measured by pre-training a model on the SQuAD datasetand fine-tuning on a selection of different datasets. The proposed metrics were used tomeasure the similarity between the datasets to see whether there was a correlation betweentransfer learning effect and similarity. The results found a clear relation between a smalldistance according to the DVD metric and good transfer learning. This could prove usefulfor a target domain without training data, a model could be trained on a big dataset andfine-tuned on a small dataset that is very similar to the target domain. It was also foundthat even small amount of training data from the target domain can be used to fine-tune amodel pre-trained on another domain of data, achieving better performance compared toonly training on data from the target domain.
38
Mogofe, Romulus Asaph. "Integrating language literacy skills in teaching physical sciences in Riba Cross District, South Africa." Thesis, University of Limpopo, 2016. http://hdl.handle.net/10386/1590.
Full textAbstract:
Thesis (M. A. (Education)) -- University of Limpopo, 2016Poor performance, by English Language Learners (ELLs), in Science examinations remains a thorny issue in countries where English is not the home language. Research around the world and the Department of Basic Education in South Africa have long recommended the integration of Language Literacy skills in the teaching of Physical Sciences in order to solve this issue. Despite that, learners’ poor performance in Physical Sciences examinations has been found to be positively related to low language literacy skills. The questions are: Do Physical Sciences teachers integrate language literacy skills in teaching the subject?; If yes, to what extent is the integration of language literacy skills practiced in Physical Sciences classroom? In an attempt to answer the above questions, a quantitative survey was carried out in Riba Cross District of Sekhukhune Region of Limpopo Province in South Africa. 211 learners and five teachers from selected nine schools took part in the study and questionnaires were used to collect data. Data were analysed using descriptive and inferential statistics and the Statistical Package for the Social Sciences (SPSS) version 22 was used. The results indicate that Language Literacy skills are integrated into the teaching of Physical Sciences in Riba Cross District, despite concerns raised by the teachers. The areas of concern include letting learners to argue using evidences and writing reports. Furthermore, schools with large classes have challenges in integrating Language Literacy Skills in the teaching of Physical Sciences. Therefore, further studies are recommended which should integrate both qualitative and quantitative approaches in school contexts.
39
Valenzuela, Elena. "L2 ultimate attainment and the syntax-discourse interface : the acquisition of topic constructions in non-native Spanish and English." Thesis, McGill University, 2005. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=102224.
Full textAbstract:
This thesis investigates the syntax-discourse interface in adult, end state second language (L2) acquisition. Specifically, it examines topic constructions in Spanish and English, namely Clitic Left Dislocation (CLLD) and Contrastive Left Dislocation (CLD), which exhibit both syntactic and discourse level properties. In both cases, topics occur at the left periphery of clauses and reintroduce a subset of a known set previously mentioned in discourse. Sensitivity to specificity is available in Spanish but not in English. The interpretation of the topicalized element as either generic or specific depends on the presence or absence of the clitic. Data from a bidirectional study are reported in order to investigate the issue of L1 transfer as well as the question of whether acquiring a new property is easier than losing a property.Three current theories of acquisition are examined, namely, the Full Transfer/Full Access model (FT/FA), Failed Functional Features Hypothesis/Representational Deficit Hypothesis (FFFH/RDH), and Optionality theory. The theories are considered with respect to their predictions about L2 ultimate attainment and the syntax-discourse interface.
Participants were 15 end state speakers of L2 Spanish/L1 English and 15 end state speakers of L2 English/L1 Spanish were tested. Three tasks were administered for each study (Sentence Completion Task, oral Sentence Selection Task, and oral Acceptability Task) targeting topic constructions and associated interpretive properties in the respective target languages. Results for the L2 Spanish study indicate that while the syntactic properties of CLLD were acquired, the interpretive properties of specificity were fossilized (i.e. clitic was overgeneralized). Results for the L2 English study also showed that the syntactic properties were acquired but the specificity distinction of the L1 was transferred into the L2. Results demonstrate that problems with specificity occur in both directions as predicted by the theory of Optionality. In both cases, the optionality found at the interpretive level was L1-based. Namely, non-target L1 forms co-existed with the L2 forms.
40
Visaggi, Salvatore. "Multimodal Side-Tuning for Code Snippets Programming Language Recognition." Master's thesis, Alma Mater Studiorum - Università di Bologna, 2021. http://amslaurea.unibo.it/22993/.
Full textAbstract:
Identificare in modo automatico il linguaggio di programmazione di una porzione di codice sorgente è uno dei temi che ancora oggi presenta diverse difficoltà. Il numero di linguaggi di programmazione, la quantità di codice pubblicato e reso open source, e il numero di sviluppatori che producono e pubblicano nuovo codice sorgente è in continuo aumento.
Le motivazioni che richiedono la necessità di disporre di strumenti in grado di riconoscere il tipo di linguaggio per snippet di codice sorgente sono svariate. Ad esempio, tali strumenti trovano applicazione in ambiati quali: la ricerca di codice sorgente; la ricerca di possibili vulnerabilità nel codice; la syntax highlighting; o semplicemente per comprendere il contenuto di progetti software.
Nasce così l'esigenza di disporre di dataset di snippet di codice allineati in modo adeguato con il linguaggio di programmazione.
StackOverflow, una piattaforma di condivisione di conoscenza tra sviluppatori, offre la possibilità di avere accesso a centinaia di migliaia di snippet di codice sorgente scritti nei linguaggi più usati dagli sviluppatori, rendendolo il luogo ideale da cui estrarre snippet per la risoluzione del task proposto.
Nel lavoro svolto si è dedicata molta attenzione a tale problematica, iterando sull'approccio scelto al fine di ottenere una metodologia che ha permesso l'estrazione di un dataset adeguato.
Al fine di risolvere il task dell'identificazione del linguaggio per gli snippet estratti da StackOverflow, nel lavoro svolto si fa uso di un approccio multimodale (considerando rappresentazioni testuali e di immagini degli snippet), prendendo in esame la tecnica innovativa di side-tuning (basata sull'adattamento incrementale di una rete neurale pre-addestrata).
I risultati ottenuti sono confrontabili con lo stato dell'arte e in alcuni casi migliori, in considerazione della difficoltà del task affrontato nel caso di snippet di codice sorgente che presentano poche linee di codice.
41
McEleney, Sarah Nicole. "La influencia del conocimiento de otras lenguas en la identificación de palabras en español L2 un estudio de los modelos bia y bia+ /." Oxford, Ohio : Miami University, 2009. http://rave.ohiolink.edu/etdc/view?acc%5Fnum=miami1248105154.
Full text
42
Kwan, Chung-hin, and 關仲顯. "An investigation of English errors of Hong Kong secondary 1 and secondary 5 students and their relationship with mother tongueCantonese transfer." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 1997. http://hub.hku.hk/bib/B3195943X.
Full text
43
Matsumura, Tomomi. "The Use of Evidentials in Hearsay Contexts in Japanese and English." PDXScholar, 2017. https://pdxscholar.library.pdx.edu/open_access_etds/4153.
Full textAbstract:
Evidentials are one of the language codes that convey the speaker's beliefs in terms of the degree of reliability of information and how the speaker obtained information (Chafe, 1986; Ishida, 2006; McCready and Ogata, 2007). Evidentials play an important role in communication since they also function to show the speaker's attitude toward the interlocutors by making a sentence softer (Trent 1998). In his theory of territory of information, Kamio (1990, 1997, 2002) proposed that pragmatic rules of evidentials are different in Japanese than in English. Ishida (2006) studied these differences and argued that learners of Japanese (JF learners) face difficulties when conveying information with evidentials in Japanese.
This study aimed to see how learners of Japanese as a foreign language (hereafter JF learners) use evidentials differently from Japanese native speakers by replicating Ishida's (2006) study. Discourse data, consisting of utterances produced by Japanese L1 speakers (J-speakers) in Japanese and JF learners in Japanese and English, was produced based on twelve situations where participants were instructed to convey hearsay information verbally to the specific third party. Collected productions were compared to see differences in the frequencies of evidentials, in general, and in terms of the source of information, the timing of conveyance, and the addressees. The study showed that JF learners use evidentials less frequently both in Japanese and English than J-speakers. This less frequent use of evidentials by JF learners in both languages indicates a transfer effect from English to Japanese. Additionally, a close look at each evidential use revealed other factors that might impact the speaker's choice of evidentials. This study revealed that JF learners tend to prefer to use evidentials such as soo da and to itta, while they tend to avoid using lexical evidentials such as rashii, mitai da, and yoo da which require more practice. These findings have implications for future research and for a classroom pedagogical approaches to the use of evidentials.
44
Kanwal, Jasmeen Kaur. "Word length and the principle of least effort : language as an evolving, efficient code for information transfer." Thesis, University of Edinburgh, 2018. http://hdl.handle.net/1842/33051.
Full textAbstract:
In 1935 the linguist George Kingsley Zipf made a now classic observation about the relationship between a word's length and its frequency: the more frequent a word is, the shorter it tends to be. He claimed that this 'Law of Abbreviation' is a universal structural property of language. The Law of Abbreviation has since been documented in a wide range of human languages, and extended to animal communication systems and even computer programming languages. Zipf hypothesised that this universal design feature arises as a result of individuals optimising form-meaning mappings under competing pressures to communicate accurately but also efficiently - his famous Principle of Least Effort. In this thesis, I present a novel set of studies which provide direct experimental evidence for this explanatory hypothesis. Using a miniature artificial language learning paradigm, I show in Chapter 2 that language users optimise form-meaning mappings in line with the Law of Abbreviation only when pressures for accuracy and efficiency both operate during a communicative task. These results are robust across different methods of data collection: one version of the experiment was run in the lab, and another was run online, using a novel method I developed which allows participants to partake in dyadic interaction through a web-based interface. In Chapter 3, I address the growing body of work suggesting that a word's predictability in context may be an even stronger determiner of its length than its frequency alone. For instance, Piantadosi et al. (2011) show that shorter words have a lower average surprisal (i.e., tend to appear in more predictive contexts) than longer words, in synchronic corpora across many languages. We hypothesise that the same communicative pressures posited by the Principle of Least Effort, when acting on speakers in situations where context manipulates the information content of words, can give rise to these lexical distributions. Adapting the methodology developed in Chapter 2, I show that participants use shorter words in more predictive contexts only when subject to the competing pressures for accurate and efficient communication. In a second experiment, I show that participants are more likely to use shorter words for meanings with a lower average surprisal. These results suggest that communicative pressures acting on individuals during language use can lead to the re-mapping of a lexicon to align with 'Uniform Information Density', the principle that information content ought to be evenly spread across an utterance, such that shorter linguistic units carry less information than longer ones. Over generations, linguistic behaviour such as that observed in the experiments reported here may bring entire lexicons into alignment with the Law of Abbreviation and Uniform Information Density. For this to happen, a diachronic process which leads to permanent lexical change is necessary. However, crucial evidence for this process - decreasing word length as a result of increasing frequency over time - has never before been systematically documented in natural language. In Chapter 4, I conduct the first large-scale diachronic corpus study investigating the relationship between word length and frequency over time, using the Google Books Ngrams corpus and three different word lists covering both English and French. Focusing on words which have both long and short variants (e.g., info/information), I show that the frequency of a word lemma may influence the rate at which the shorter variant gains in popularity. This suggests that the lexicon as a whole may indeed be gradually evolving towards greater efficiency. Taken together, the behavioural and corpus-based evidence presented in this thesis supports the hypothesis that communicative pressures acting on language-users are at least partially responsible for the frequency-length and surprisal-length relationships found universally across lexicons. More generally, the approach taken in this thesis promotes a view of language as, among other things, an evolving, efficient code for information transfer.
45
Randolph, Gerda Ann Packard. "Building written language: A program for second language literacy in English." CSUSB ScholarWorks, 2000. https://scholarworks.lib.csusb.edu/etd-project/1866.
Full text
46
姜源貞 and Yuen-ching Keung. "Transfer of cognitive skills in learning to read Chinese (L1) and English (L2) among HK elementary students." Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2006. http://hub.hku.hk/bib/B37090720.
Full text
47
Andersson, Kristin. "Explicit and implicit comparison of English and Swedish in English course books for year 7 and year 11 in Sweden." Thesis, Karlstads universitet, Institutionen för språk, litteratur och interkultur, 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-47873.
Full textAbstract:
This paper compares and contrasts how linguistic transfer has been incorporated in English course books for year 7 and year 11 in Swedish schools. This study has focused on finding explicit and implicit reference to transfer and interference in English course books for Swedish students by identifying references from the authors where they have pointed to similarities and differences between Swedish and English. The results of the study showed that comparisons between the languages were used in all books, but explicit references were more common than implicit references. Of the four different linguistic levels (pronunciation, grammar, vocabulary, pragmatics) examined in this study, grammatical comparisons were the most frequent. However, the authors of the books for year 7 used more comparisons between the languages than the authors of the year 11 books. One conclusion that can be drawn from the present study is that it is up to the authors of the course books to decide to what extent they should compare and contrast the two languages, since the national curricula for the subject English give no indications of how the subject of transfer should be dealt with in language acquisition.Den här uppsatsen jämför hur lingvistisk överföring behandlas engelska kursböcker för årskurs 7 och årskurs 11 i svenska skolor. Studien har fokuserat på att undersöka i vilken mån författarna uttryckligen eller underförstått refererat till överföring och transfer genom att identifiera var kursböckerna innehåller information om likheter och skillnader mellan svenska och engelska. Resultaten från denna studie visar att sådana jämförelser mellan språken fanns i alla böcker, men att uttrycklig jämförelse var det vanligast sättet för författarna att visa på likheter och skillnader. Av de fyra olika lingvistiska nivåer som användes i denna studie (uttal, grammatik, ordförråd, pragmatik), förekom jämförelse av grammatik mest. Författarna till böckerna för årskurs 7 använde sig mer av jämförelser mellan språken än författarna till böckerna för årskurs 11. En slutsats som man kan dra av denna studie är att det är upp till författarna själva att bestämma i vilken mån de ska jämföra de två språken, eftersom läroplanen för ämnet engelska inte ger någon information om hur överföring ska hanteras i språkinlärning.
48
Ramponi, Alan. "Knowledge Extraction from Biomedical Literature with Symbolic and Deep Transfer Learning Methods." Doctoral thesis, Università degli studi di Trento, 2021. http://hdl.handle.net/11572/310787.
Full textAbstract:
The available body of biomedical literature is increasing at a high pace, exceeding the ability of researchers to promptly leverage this knowledge-rich amount of information. Although the outstanding progress in natural language processing (NLP) we observed in the past few years, current technological advances in the field mainly concern newswire and web texts, and do not directly translate in good performance on highly specialized domains such as biomedicine due to linguistic variations along surface, syntax and semantic levels. Given the advances in NLP and the challenges the biomedical domain exhibits, and the explosive growth of biomedical knowledge being currently published, in this thesis we contribute to the biomedical NLP field by providing efficient means for extracting semantic relational information from biomedical literature texts. To this end, we made the following contributions towards the real-world adoption of knowledge extraction methods to support biomedicine: (i) we propose a symbolic high-precision biomedical relation extraction approach to reduce the time-consuming manual curation efforts of extracted relational evidence (Chapter 3), (ii) we conduct a thorough cross-domain study to quantify the drop in performance of deep learning methods for biomedical edge detection shedding lights on the importance of linguistic varieties in biomedicine (Chapter 4), and (iii) we propose a fast and accurate end-to-end solution for biomedical event extraction, leveraging sequential transfer learning and multi-task learning, making it a viable approach for real-world large-scale scenarios (Chapter 5). We then outline the conclusions by highlighting challenges and providing future research directions in the field.
49
Han, Young Joo. "Feedback and Transfer in Second Language Writing: A Qualitative Study of ESL Students' Experiences." The Ohio State University, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=osu1498779154665613.
Full text
50
Kruesemann, Heike. "Language learning motivation and the discursive representations of German, the Germans, and Germany in UK school settings and the press." Thesis, University of Reading, 2018. http://centaur.reading.ac.uk/77934/.
Full textAbstract:
Language learning in UK secondary schools is in sharp decline. Of the three most commonly taught languages, French, Spanish, and German, German uptake is dropping at the fastest rate. Societal attitudes towards languages and the target language communities of speakers are commonly blamed for the decline, however, few previous studies have investigated this area via empirical evidence. The current study explores the relationship between motivation for German learning in adolescent language learners in England, and the representations of German, the Germans, and Germany in private and public UK discourses. Through a mixed methods, cross-sectional research design, private grassroots discourses in a school setting and public discourses in the national press are examined to gain an insight into how beliefs and attitudes around German are constructed and conceptualised, and results are related to the factors that underlie learner attitudes, motivation for German learning and language uptake decisions. The study’s theoretical framework draws on a range of key concepts from second language (L2)-specific and mainstream psychological motivation theories, such as Gardner’s socio-educational model, as well as on aspects of cognitive-situated and self-based models. Main participants were 506 13 to 16-year-old German learners from a range of four secondary schools at the time they were asked to decide whether to continue or drop German. Further respondents included four German teachers and four head teachers from the participating schools. For the school settings strand, learner data were collected via focus groups and a questionnaire combining items which generated quantitative (such as motivation mean scores) as well as qualitative (such as metaphor) data; teacher and head teacher data were collected via interviews. For the public discourse strand, a specialised corpus of 40,000 UK national newspaper articles around German, the Germans, and Germany was compiled, and explored using discourse analysis techniques. Four research questions investigated the motivational dimensions relating to learners’ choices to continue or discontinue with the study of German, how German is represented in discourses of key players in school settings, how German is represented in newspaper discourses in the press, and the relationship between public and private discourses around German, the Germans, and Germany in the UK. Contrary to what is commonly thought about motivation for language learning, results suggest that adolescent learners in England are motivated to continue German not by instrumental rationales, but rather by their enjoyment of the classroom learning situation, and by a sense of personal relevance. The growing elitification in language learning in the UK manifests in the study’s data, in that a higher socio-economic background is associated with continuing German, more conducive attitudes, a higher sense of personal relevance, and a view of language learning as a worthwhile process which requires effort and persistence. In press discourses, German is mainly represented in terms of politics and war, Germans mainly in terms of war, and Germany mainly in relation to other countries and football. The wider discourses in the press are reproduced in a reciprocal relationship with private discourses in the school setting. The study contributes to knowledge by presenting new insights into the motivation of adolescent German learners in the UK, and by validating elements of pre-existing motivation models, such as self-determination and self-worth theory. Furthermore, through its novel design of bringing together private and public discourse domains, it provides empirical evidence for previously unsubstantiated claims of links between the representation of target language speakers and communities in the mass media, and language learner motivation at English secondary schools.