Linguistic Knowledge Research Articles

CLIP-Llama: A New Approach for Scene Text Recognition with a Pre-Trained Vision-Language Model and a Pre-Trained Language Model

This study focuses on Scene Text Recognition (STR), which plays a crucial role in various applications of artificial intelligence such as image retrieval, office automation, and intelligent transportation systems. Currently, pre-trained vision-language models have become the foundation for various downstream tasks. CLIP exhibits robustness in recognizing both regular (horizontal) and irregular (rotated, curved, blurred, or occluded) text in natural images. As research in scene text recognition requires substantial linguistic knowledge, we introduce the pre-trained vision-language model CLIP and the pre-trained language model Llama. Our approach builds upon CLIP’s image and text encoders, featuring two encoder–decoder branches: one visual branch and one cross-modal branch. The visual branch provides initial predictions based on image features, while the cross-modal branch refines these predictions by addressing the differences between image features and textual semantics. We incorporate the large language model Llama2-7B in the cross-modal branch to assist in correcting erroneous predictions generated by the decoder. To fully leverage the potential of both branches, we employ a dual prediction and refinement decoding scheme during inference, resulting in improved accuracy. Experimental results demonstrate that CLIP-Llama achieves state-of-the-art performance on 11 STR benchmark tests, showcasing its robust capabilities. We firmly believe that CLIP-Llama lays a solid and straightforward foundation for future research in scene text recognition based on vision-language models.

Can Linguistic Knowledge Improve Multimodal Alignment in Vision-Language Pretraining?

The field of multimedia research has witnessed significant interest in leveraging multimodal pretrained neural network models to perceive and represent the physical world. Among these models, vision-language pretraining (VLP) has emerged as a captivating topic. Currently, the prevalent approach in VLP involves supervising the training process with paired image-text data. However, limited efforts have been dedicated to exploring the extraction of essential linguistic knowledge, such as semantics and syntax, during VLP and understanding its impact on multimodal alignment. In response, our study aims to shed light on the influence of comprehensive linguistic knowledge encompassing semantic expression and syntactic structure on multimodal alignment. To achieve this, we introduce SNARE , a large-scale multimodal alignment probing benchmark designed specifically for the detection of vital linguistic components, including lexical, semantic, and syntax knowledge. SNARE offers four distinct tasks: Semantic Structure, Negation Logic, Attribute Ownership, and Relationship Composition. Leveraging SNARE , we conduct holistic analyses of six advanced VLP models (BLIP, CLIP, Flava, X-VLM, BLIP2, and GPT-4), along with human performance, revealing key characteristics of the VLP model: (i) Insensitivity to complex syntax structures, relying primarily on content words for sentence comprehension. (ii) Limited comprehension of sentence combinations and negations. (iii) Challenges in determining actions or spatial relations within visual information, as well as difficulties in verifying the correctness of ternary relationships. Based on these findings, we propose the following strategies to enhance multimodal alignment in VLP: (1) Utilize a large generative language model as the language backbone in VLP to facilitate the understanding of complex sentences. (2) Establish high-quality datasets that emphasize content words and employ simple syntax, such as short-distance semantic composition, to improve multimodal alignment. (3) Incorporate more fine-grained visual knowledge, such as spatial relationships, into pretraining objectives. 1

Enhancing L2 Writing through Scaffolding Concept: A Study With Reference To Rural Students in a Government College of Thanjavur, Tamil Nadu

This paper attempts to highlight the importance of Scaffolding concept in L2 writing. As writing requires complex psycholinguistic skills it is necessary study writing to come out with new insights. A number of studies have highlighted the importance of learner scaffolding. This paper tries to answer the following research question. How does scaffolding concept in L2 writing, enhance the writing ability of the target group of learners?It touches upon a number factors that affect L2 writing. Besides, the paper makes use of scaffolding frame work with regard to four major components namely, content knowledge, context knowledge, process knowledge and linguistic knowledge and finally it reveals the indispensable role of scaffolding framework in L2 writing context.

Exploration of Business English Teaching Mode in Applied Undergraduate Colleges Based on OBE Concepts

In the context of globalization, the demand for business English professionals is increasing, placing the application-oriented undergraduate institutions under the critical responsibility of cultivating business English graduates with practical abilities and professional competencies. However, the traditional teaching model tends to prioritize the impartation of linguistic knowledge, neglecting the cultivation of actual application skills, which fails to meet market demands. The Outcome-Based Education (OBE) philosophy, which is guided by learning outcomes and emphasizes the development of students' abilities, offers new insights for educational reform. This article reviews the current state of research on business English teaching models in application-oriented undergraduate institutions, highlighting the shortcomings of traditional models. Based on the OBE concept, this article explores the reconstruction of training objectives, the optimization of curriculum systems, the reform of teaching methods, and the improvement of evaluation systems as four core elements, and proposes implementation strategies involving strengthening top-level design, faculty development, school-enterprise cooperation, and information technology construction. These measures not only enhance the quality of education but also boost students' competitiveness in the workplace, providing valuable references for the reform of business English teaching models in application-oriented undergraduate institutions.

Ethical practice in participant-centred linguistic research

Abstract This article outlines ethical principles for ‘participant-centred linguistic research’ (PCLR), a term we coin to incorporate a range of linguistic research approaches that place importance on the involvement of participants. Linguistics, as a field, has strengthened its focus on participant-centred and socially situated research, recognising the value of better understanding our participants’ practices and linguistic knowledge. However, this also brings ethical challenges for our research practice. Drawing on three differing UK-based case studies from the authors’ own work, the article explores complex issues that can arise during PCLR and establishes four key principles that cut across our varied experiences. Firstly, we address participant consent and confidentiality, establishing the principles: 1. Informed consent and ethics protocols are dialogic processes and 2. Expectations around confidentiality and anonymity can shift during a project. Secondly, we address our research relationships with participants, our key principles being: 3. The researcher-participant relationship is complex and variable and 4. Close attention must be paid to power dynamics within the research setting. Ultimately, we argue that the human interactions and relationships involved in PCLR mean research may inevitably be somewhat unpredictable; researchers therefore need an understanding of the ethical parameters of their practice to navigate these complexities.

Электронный корпус татарского языка на базе модели лингвистических графов знаний

The article presents a new version of the electronic corpus of the Tatar language, updated based on a linguistic knowledge graph model for Turkic languages. This new version of the corpus allows for information description across multiple linguistic levels: morphonological, syntactic, and semantic, through the use of knowledge graphs to represent linguistic data. This approach enhances corpus functionality, enabling searches that incorporate syntactic and semantic information. A distinctive feature of the electronic corpus implementation is that the model employed aligns closely with the structural and functional characteristics of Turkic languages and serves as a foundation for developing various software products for semantic text processing in Turkic languages. In particular, these products include the linguistic portal "Turkic Morphme" and the new version of the Tatar language electronic corpus, "Tugan Tel.".

Incorporating Cultural Sensitivity Into ESL Curricula To Foster Pragmalinguistic Competence

This study explores strategies for incorporating cultural sensitivity into English as a Second Language (ESL) curricula to enhance learners’ pragmalinguistic competence, the ability to use language pragmatically and appropriately within social contexts. By examining approaches such as culturally relevant content, cross-cultural pragmatics training, and practical language activities, this study analyzes how integrating cultural sensitivity into language instruction fosters learners’ ability to navigate linguistic nuances. Findings suggest that a culturally sensitive curriculum not only enriches linguistic knowledge but also significantly enhances learners’ competence in communicating appropriately across diverse cultural contexts.

L1 as a tool for providing corrective feedback

This article delves into the controversial practice of utilizing a student’s first language (L1) as a teaching resource in second language (L2) learning environments. Initially, strategies such as code-switching/code-mixing and translanguaging were considered signs of poor linguistic ability. There was a strong push towards using only the target language in foreign language education, aiming to limit the first language’s interference and foster a deeper immersion in the new language. However, later research has shown the benefits of incorporating the first language in bilingual education and language learning processes. It’s argued that a student’s knowledge in their native language can actually support their comprehension of a second language, suggesting that transferring certain linguistic or conceptual knowledge from L1 to L2 can be advantageous. This perspective encourages the strategic use of this knowledge transfer in teaching methods. Moreover, the text points to positive results from various studies on the positive impact of L1 usage in L2 classrooms. These insights pave the way for further exploration into the application of the first language in adult English as a Second Language (ESL)/English as a Foreign Language (EFL) education, particularly regarding providing corrective feedback.

Exploring the Importance of Traditional Schools in Maintaining the Heritage Language: A Study of Tshivenḓa from Tshimbupfhe, South Africa

This study aimed to explore the critical role of traditional schools—specifically murundu and vhusha—in preserving the Tshivenḓa language and cultural heritage in Tshimbupfhe, in the Limpopo Province, South Africa. Traditional schools play a significant role in preserving languages and cultural traditions that might otherwise be at risk of erosion due to globalisation, modernisation, or cultural assimilation. A deeper comprehension of their role allows individuals to appreciate more fully how these schools contribute to the preservation of cultural diversity and heritage. The study sampled 31 participants. This study adopted a qualitative approach and relied on face-to-face semi-structured interviews to collect and analyse data. The theoretical framework used to underpin the study was the Afrocentricity theory. The findings of the study illustrated that traditional schools serve as pivotal institutions for the transmission of linguistic and cultural knowledge, integrating indigenous values, practices, and language into their curricula. The study provides insights into how traditional schools facilitate the transmission of cultural values, practices, and languages to younger generations. By exploring the methods and content of traditional education, the research highlights how these institutions contribute to cultural continuity and identity. Secondly, the study contributes to the documentation and preservation of indigenous educational practices and pedagogies. It offers a comprehensive account of how traditional schools operate, including their teaching methods and cultural rituals, thereby preserving valuable knowledge for future generations. Furthermore, this study enhances the understanding of traditional schools’ significance and supports the preservation of language and cultural heritage.

The construction of linguistic identities in talk about food among Tibetan heritage language learners

ABSTRACT This ethnographic article examines the interactional processes through which Tibetan-Canadian heritage language learners agentively construct their linguistic identities in everyday family interactions. Because food is centrally involved in family routines, as well as broader cultural practices and the articulation of individual tastes, talk about food provides a site for analyzing negotiations of individual agency amid expressions of shared identity. Drawing from previous language socialisation scholarship that approaches family meal times as a resource for constructing identity, this paper analyzes children’s conversations at meal-times, during food preparation, and about food preferences. Through interactional analysis of twelve months of longitudinal video ethnography in two Tibetan-Canadian families, I found that, in these activities centered on food, children used multilingual and multimodal resources to negotiate authority over their cultural and linguistic knowledge. Specifically, children agentively displayed their knowledge of Tibetan language forms, elaborated on their capacities to participate in cultural activities, and instructed parents about their individual preferences. By analyzing these interactional displays of knowledge through the concepts of agency and co-operative action, I argue that children construct their identities as speakers of Tibetan, despite the dominance of English in their language repertoires.

Robust visual question answering via polarity enhancement and contrast

The Visual Question Answering (VQA) task is an important research direction in the field of artificial intelligence, which requires a model that can simultaneously understand visual images and natural language questions, and answer questions related to images. Recent studies have shown that many Visual Question Answering models rely on statistically regular correlations between questions and answers, which in turn weakens the correlation between visual content and textual information. In this work, we propose an unbiased Visual Question Answering method to solve language priors from the perspective of strengthening the contrast between the correct answer and the positive and negative predictions. We design a new model consisting of two modules with different roles. We input the image and the question corresponding to it into the Answer Visual Attention Modules to generate positive prediction output, and then use a Dual Channels Joint Module to generate negative prediction output with great linguistic prior knowledge. Finally, we input the positive and negative predictions together with the correct answer to our newly designed loss function for training. Our method achieves high performance (61.24%) on the VQA-CP v2 dataset. In addition, most existing debiasing methods improve performance on VQA-CP v2 dataset at the cost of reducing performance on VQA v2 dataset, while our method not only does not reduce the accuracy on VQA v2 dataset. Instead, it improves performance on both datasets mentioned above.

The Effects of Varying Practice Modes on L2 Learners’ Vocabulary Retention Through Songs: Listening, Singing, and Oral Reading

Singing or listening to songs is one of the most common L2 activities outside the classroom, and many language teachers integrate songs in their teaching; however, the effect of varying song practice modes on L2 learners’ retention of linguistic knowledge is under-researched. Eighty-seven EFL learners of university students were recruited and divided into three subgroups. To avoid different song different effect, the study adopted a counterbalanced designed, in which each subgroup practiced three songs under three modes: listening + singing, listening + reading lyrics orally, and listening only. Three aspects of vocabulary knowledge (aural and written meaning recognition, written form production) were measured in the pre-test, one-week delayed, and eight-week delayed post-tests. HLM was performed to analyze the data. Results showed that the main effects of the four variables (times, practice modes, vocabulary dimensions, songs) had significant effects on L2 learners’ vocabulary retention, but the interaction effects between times and the three other variables were less salient. Overall, students scored higher for singing mode and for written meaning recognition; however, different songs and times seem to have played an important but complicated role on vocabulary retention. Applications of integrating songs in teaching L2 are discussed.

Language and Knowledge of the World

Abstract: This article addresses language and knowledge of the world within the communicative process, as it is understood that knowledge of the world plays a preponderant role in effective communication between language users. The main objective of this article was to demonstrate the relevance of knowledge of the world in the communicative process, showing that for the construction of a textual meaning or understanding of the text, several processes occur and the mastery of linguistic knowledge is not a sufficient condition for calculating the sense. This study's starting question is the following: does mastery of linguistic knowledge in itself constitute a sufficient and/or exclusive condition for the construction of textual meaning? To answer this question, bibliographical research was adopted as the study methodology, and the type and method of approach were qualitative and inductive respectively, taking into account that the inductive method starts from the examination of particular data, sufficiently verified and inferred the results of the part examined to the entire population (Marconi & Lakatos, 2003). To carry out this study, participant observation and interviews were also chosen as data collection techniques. Thus, this study showed that for effective communication, that is, construction of textual meaning within the communicative process, a large part has to do with knowledge of the world manifested through shared knowledge about sociocultural conventions and/or pragmatic aspects of language, contradicting the perceptions according to which linguistic knowledge, which is limited only to the domain of the grammatical rules of the language and the lexicon, means that the individual would be able to understand the text.

Linguistic Acceptability

Abstract: This article is about linguistic acceptability, and in this article, linguistic acceptability is treated as a synonym of textual acceptability. This article has as fundamental objective to understand the factors that contribute to the establishment of linguistic acceptability, and to continue this study, the starting question was the following: What are the factors that contribute to the establishment of linguistic acceptability? Therefore, to answer this question, the line of bibliographical research was adopted as a methodology in order to understand the different perspectives linked to this theme in order to understand and comprehend the determining conditions that contribute to textual acceptability in a given communicative event. In this way, it is understood that this article can substantially contribute to raising awareness among potential speakers of a language about the essential criteria that guarantee textual acceptability, but also about its relevance in the language teaching and learning process. Thus, in view of the above, it is concluded that textual acceptability is characterized by the ability of the receiver to attribute a certain meaning to the linguistic sequences produced by his interlocutor, there are several criteria that must be mobilized, relevant for example for linguistic knowledge, sociolinguistic knowledge and pragmatic knowledge, that is, the learner of a given language must simultaneously master linguistic conventions and sociocultural conventions on the use of language within the speaking community

Acceleration in multilingual children: the case of French

ABSTRACT The acquisition of the null-subject property has received much attention. Languages differ as to whether they allow (pronominal) subjects to be unexpressed. French is generally assumed to be a language not allowing null-subjects, while languages like Italian and European Portuguese are treated as null-subject languages. The present study locates these language types within the null-argument hierarchy. This hierarchy “emerges” through the interaction of (non-language specific) cognitive strategies and linguistic features. In such a hierarchy, Italian and European Portuguese are less marked than French. The study of the longitudinal data of three Italian/Portuguese-French bilingual children between the age of 1;0 to 5;0 shows that bilingual children can reach the French target grammar more quickly than monolingual French-speaking children, even if they have a “weak” language. The children are accelerated with respect to (pronominal) subject realizations and target-like inflection of finite verbs. The findings indicate that bilingual children are not statistical learners, but they can rely on prior knowledge attained in their respective other, radically different, and less marked language. Consequently, educational programs with a focus on multilingualism (like pedagogical translanguaging) which aim to activate prior linguistic knowledge can be effective to compensate for disadvantages related to critical periods.

Körtvélyes község nyelvi tájképének és a lakosság helynévismeretének kapcsolata

The Relationship Between the Linguistic Landscape of the Village of Körtvélyes (Hrušov) and the Knowledge of Local Place-Names by the Population My thesis is in line with my previous research, which focuses on the linguistic landscape and the name semiotic landscape, bilingualism and place-name research and its utility. The present paper is also based on this theme, and aims to present one aspect of the research conducted in a selected village in the Rožňava district. In the literature review, I will therefore discuss the above-mentioned concepts, explaining them in order to make the results more interpretable. I will then describe the location of the research, Hrušov, and the methods used in the research. The aspect on the basis of which I will examine this research in this paper is the relationship between linguistic landscape, bilingualism and knowledge of place-names. Thus, in presenting the results, I will describe the possible relationship between the name semiotic landscape, bilingualism, and knowledge and use of place names. n the conclusion, I would like to point out the importance of the name semiotic landscape and the importance of preserving Hungarian place-names.

Relationships between linguistic knowledge, linguistic awareness, and argumentative writing among upper elementary bilingual students

Relationships between linguistic knowledge, linguistic awareness, and argumentative writing among upper elementary bilingual students

Problems of Foreign Language Teaching for Adult Veterinary Medicine Specialists: Reasons and Their Solution

Learning a foreign language by adults is one of the most critical aspects of lifelong learning in Ukraine. Many adults have worked in certain spheres of the Ukrainian economy; they must improve their qualifications once in a certain period. This should include specialized subjects and learning a foreign language. However, the organization of this process requires much effort and the unique skills of a language teacher since teaching a foreign language to adults can be fairly challenging, facing many problems. The article aims to determine the main problems and their solution in teaching foreign languages to adults. Methodology. We focus more on learning a foreign language in continuous education, using the example of specialists in the veterinary field of Ukraine. To conduct research, we interviewed 14 people aged 45 and over and 18 people aged 35-40 who studied at the Stepan Gzhytskyi National University of Veterinary Medicine and Biotechnologies of Lviv. In addition, English classes were held with eight persons from among those interviewed. The survey was conducted as an interview, where participants were asked to answer presented questions. Results. After interviewing the respondents of our study and analyzing the learning process, we can single out the most common problems when adults learn English: psychological, phonetic, lexical, and grammatical. The research results have demonstrated that teaching a foreign language to adult specialists requires the teacher's deep linguistic and psychological knowledge. Considering the problems raised during teaching, two groups of issues are solved by frequent training, explaining materials, memorizing rules, and vocabulary. Conclusion is psychological problems require a creative approach from a teacher to solve them.

Foreign language talker identification does not generalize to new talkers.

Listeners identify talkers less accurately in a foreign language than in their native language, but it remains unclear whether this language-familiarity effect arises because listeners (1) simply lack experience identifying foreign-language talkers or (2) gain access to additional talker-specific information during concurrent linguistic processing of talkers' speech. Here, we tested whether sustained practice identifying talkers of an unfamiliar, foreign language could lead to generalizable improvement in learning to identify new talkers speaking that language, even if listeners remained unable to understand the talkers' speech. English-speaking adults with no prior experience with Mandarin practiced learning to identify Mandarin-speaking talkers over four consecutive days and were tested on their ability to generalize their Mandarin talker-identification abilities to new Mandarin-speaking talkers on the fourth day. In a "same-voices" training condition, listeners learned to identify the same talkers for the first 3days and new talkers on the fourth day; in a "different-voices" condition, listeners learned to identify a different set of voices on each day including the fourth day. Listeners in the same-voices condition showed daily improvement in talker identification across the first 3days but returned to baseline when trying to learn new talkers on the fourth day, whereas listeners in the different-voices condition showed no improvement across the 4days. After 4days, neither group demonstrated generalized improvement in learning new Mandarin-speaking talkers versus their baseline performance. These results suggest that, in the absence of specific linguistic knowledge, listeners are unable to develop generalizable foreign-language talker-identification abilities.

Issues and Countermeasures in Chinese FLTRP Primary School Textbooks under the New Curriculum Standard

The New Curriculum Standard (NCS) is the norm and quality requirement of the Chinese national educational curriculum as well as the basis of teaching management, curriculum evaluation, and textbook compilation. In textbook design, the selection of textbook materials should meet the objectives and requirements of the NCS. Since appropriate teaching materials are essential to curriculum standards and effective teaching, we must follow the consistency between textbook materials and curriculum standards to achieve an ideal degree of conformity with the textbook system and teaching objectives and promote the sustainable development of the English curriculum reform. However, the current primary school English textbooks fall short of the NCS, which leads to difficulties in implementing teaching tasks and achieving teaching goals. Given these difficulties, we should analyze the causes from three dimensions: linguistic knowledge, cultural awareness, and learning strategies. In addition, a high recurrence of phonetic knowledge, systematic grammar presentation, diversified learning styles, level grading, and in-depth involvement of cultural factors are keys to the perfect conformity of the NCS and English textbooks.