This paper describes a corpus survey of [Verb + EXAMINATION] to explore the formal specifications of support verb constructions (SVCs). Previous studies described the properties of SVCs by focusing on the apparent semantic and syntactic paucity of support verbs (Brugman, 2001; Newman, 1996; Wierzbicka, 1982) and the process of argument transfer triggered by the complements (Grimshaw & Mester, 1998; Grimshaw, 1990). However, these verb-centered approaches face the issue of ambiguity between light and heavy senses in context (e.g., make_light a diagnosis, make_heavy a certificate), and coverage of low frequency support verbs (e.g., sustain an injury). To address these issues, we point out the need for formal specifications of SVCs targeted on their compliments. Our corpus analysis of [Verb + EXAMINATION], in which the deverbal noun examination possesses its own argument structure, revealed the types of verbs preferred in SVCs and the grammatical properties of examination (e.g., the occurrence of an of-phrase).

This study developed the preliminary list of lexical bundles (e.g., on the other hand, the fact that) for argumentative essay writing and explored its potential applications to English for General Academic Purposes (EGAP) practice to prepare undergraduate students for their future use in academic written English genre (e.g., research papers). The list, called the Essay Bundles List (EBL), was created by extracting frequently used lexical bundles from opinion- and source-based argumentative essays by L1 English speakers. Corpora consulted include the International Corpus Network of Asian Learners of English (ICNALE), Louvain Corpus of Native English Essays (LOCNESS), PERSUADE2.0, Michigan Corpus of Upper-Level Student Papers (MICUSP), and British Academic Written English (BAWE). A total of 3,768 bundles were compared with the list of academic written English (Academic Formulas List: AFL) to confirm EBL applicability. The results showed that the EBL covers approximately 80% of the AFL, indicating its potential as an EGAP wordlist. Correspondence analysis of the top 21 frequent bundles in opinion- and source-based essays and the AFL revealed that the opinion-based bundles (e.g., I believe that) can be made suitable for academic written English with the use of inanimate subjects (e.g., it is true that), while source-based discourse bundles (e.g., in order to) imply their direct applicability. The EBL was refined to 127 bundles according to their difficulty levels on the Common European Framework of Reference (CEFR) scale (A2, B1, B2) in proficiency order. This study suggested that, basic referential bundles (e.g., the fact that) and objective stance bundles (e.g., this means that) are appropriate for A2 and B1 students. Discourse bundles (e.g., to begin with, on the one hand) should remain a focus throughout the progression from B1 to B2. Advanced referential bundles, such as the existence of, are most suitable for instruction at the B2 level.

Corpora have been utilized for purposes of language pedagogy. One of the approaches, data-driven learning (DDL), uses corpora or corpus-based materials in language classrooms. DDL is an inductive language learning method in which learners explore authentic language data and discover linguistic patterns on their own. While empirical studies on DDL have been increasing (Boulton & Cobb, 2017), there are only a few experiments on “data-driven construction learning” (Gilquin, 2021). The present paper reports the results of an experiment aimed at testing the effectiveness of data-driven construction learning and evaluating learners’ attitudes towards DDL. In the experiment, two groups of Japanese learners of English learned the way construction (e.g., Goldberg, 1995; Luzondo Oyón, 2013) with one group learning through DDL and the other through a traditional form-focused instruction. DDL in this study includes an explicit explanation of the target construction, following the tenet of applied construction grammar (Gilquin & De Knop, 2016). The effectiveness of the two methods was measured and compared by means of pre- and post-tests (sentence production and translation tasks). Additionally, the learners’ attitudes towards DDL were investigated through a post-questionnaire. The improvement of both sentence production and translation tasks in the post-tests demonstrated that both DDL and the traditional instruction were effective. Also, the participants’ attitude towards DDL was found to be positive. However, learners who received the traditional instruction outperformed those who received DDL. Therefore, the present study concludes that even though DDL was effective, other teaching methods could be more beneficial for learners, depending on the difficulty of a target construction and learners’ proficiency. This paper also argues that learners can benefit from DDL in various ways, such as developing general cognitive skills, and hence it is suggested that incorporating DDL into a classroom activity or employing it as an out-of-class activity would be advantageous.

This study investigates developmental patterns in the written English of Japanese EFL learners from the 7th to 12th grade by analyzing a corpus of approximately 30,000 words. The corpus was annotated with 37 types of error tags to examine error patterns and changes across different parts of speech and error type. The analysis revealed some findings. While overall error rates generally decreased with school year progression, there was a temporary increase in the ninth grade, third year of junior high school. This increase suggests that students at this stage attempt to use more complex linguistic features or newly learned grammatical structures. Regarding syntactic development, the analysis showed a clear development from simple to complex sentences, with senior high school students increasingly using more sophisticated structures involving conjunctions. These results provide insights into the characteristics of Japanese learners’ linguistic development across different parts of speech and demonstrate qualitative changes in language use as students’ progress through school years.

The main aim of this paper is to investigate whether the abstract noun advantage can be followed by an appositive that-clause and whether the auxiliary verb can is essential within the clause to explain advantage. A descriptive study using both quantitative and qualitative approaches, incorporating corpus analysis and informant surveys, reveals two key findings. First, constructions in which advantage is followed by an appositive that-clause are observed in the spoken British English as well as in written forms of both British and American English. Second, the auxiliary verb can in such clauses is not a mandatory component for the noun advantage but is context-dependent. Its occurrence is strongly influenced by the meanings of “ability” or “circumstantial ability” conveyed in the appositive clause. These findings offer valuable insights for revising the treatment of advantage in reference books and dictionaries.

This study employs a quantitative method, latent Dirichlet allocation topic model, to examine the distinctive thematic and lexical characteristics of poems by Alfred, Lord Tennyson, and his brother, Charles Tennyson, focusing primarily on Poems by Two Brothers (1827), the first published collection by Alfred, Lord Tennyson. It has been said that there is some ambiguity surrounding the poems’ authorship within the collection, originally published without annotations indicating each poem’s author. This research leverages topic model to uncover patterns in the diction and thematic inclinations of the two brothers. The results of the latent Dirichlet allocation analysis indicate that each poet gravitated toward certain specific topics as dominant themes in his works. Topic 17, which is associated with romantic sentiments and primarily physical descriptions of women, is prevalent in the poems written by Charles Tennyson, while Topic 1, which concerns themes of masculinity, enthusiasm, and battle, is prominent in the poems written by Alfred Tennyson. These findings highlight the distinctive differences between the two brothers in word usage. It is noteworthy that this study represents a novel application of topic model in examining characteristic diction of the two poets, offering internal evidence of the distinct word usages within the Tennysons’ collaborative collection. This exploration underscores that topic model is effective in distinguishing thematic tendencies as well as the characteristic diction of the two authors, Alfred and Charles Tennyson.

In this paper, we introduce the concrete interface, various on-board corpora, search functions, implemented statistical processing, and expected applications of Parallel Link (Ver. 2.0), an analysis tool for Japanese-English and English-Japanese parallel corpora that has finally been completed. In particular, four new parallel corpora, BSD, Coursera, Hiragana Times, and JENAAD, have been added to the database, bringing the total number of parallel corpora to 13 and increasing its volume by around 1.3 times. An overview of these added corpora, as well as those to be added in the future, will be presented in detail. The new functions added in Ver. 2.0, such as the context display function, the distribution graph function for each corpus of collocations, the voice function, and the Google Translate link function, will also be introduced.

This study explores the development and implementation of data-driven learning (DDL) tools for middle and high school students to overcome barriers in English language education. DDL, proposed by Johns (1991), emphasizes the use of corpus tools for pedagogical purposes, enabling learners to discover language patterns through active analysis of linguistic data. Despite its potential, DDL has not been widely adopted in primary and secondary education due to challenges such as the complexity of authentic texts and the need for specialized training. To address these issues, we developed hDDL2, a user-friendly DDL support software tailored for younger learners. The hDDL2 tool incorporates an educational corpus adjusted for vocabulary and grammar levels appropriate for junior and high school students. hDDL2 effectively supports inductive, student-centered learning, aligning with modern educational goals. The tool’s features, including easy-to-use search functions and grammar quizzes, facilitate the discovery and retention of language rules. This paper discusses the progressive development of hDDL2, and its application to beginner-level classroom settings. The results suggest that hDDL2 can bridge the gap between theoretical DDL approaches and practical classroom implementation, promoting more engaging language learning experiences for beginning-level students.