Earticle

This study compares the productivity of the Korean nominal derivational suffixes –jeok, –seong, and –hwa by applying Baayen’s (2009) measures of morphological productivity. Using the New Yonsei Korean Corpus 4 and 5, this study measures realized productivity, expanding productivity, and potential productivity based on type frequency, hapax legomena, and token frequency. It further examines high-frequency words and hapax legomena to interpret how productivity is realized in actual word formation. The results show that –jeok exhibits the highest realized productivity, whereas –seong and –hwa display relatively higher potential productivity. The findings also show that the productivity of the three suffixes is realized in different lexical and semantic patterns. This study suggests that morphological productivity should be interpreted not only through quantitative measures but also in relation to actual lexical patterns and the semantic functions of derivational suffixes.

This study aims to examine how one of the core conceptual terms in Confucianism, ren (仁), is translated and semantically represented in James Legge's English translation of the Analects. To this end, a total of 59 chapters in which ren appears were extracted, and the translated vocabulary was classified and frequency analyzed, along with an in-depth semantic analysis of key passages. The analysis shows that Legge translated ren, which appears 109 times in the Analects, predominantly into the ‘virtue’ family of terms (94 times, approx. 86%), while also rendering it variously as ‘perfect virtue’, ‘benevolence’, ‘principle’, and ‘good’, depending on the context. In particular, Legge understood ren as a comprehensive virtue, but when it manifested in concrete actions such as filial piety and fraternal submission (孝悌) and loving others (愛人), he used the ‘bene-’ family of terms. When de (德) and ren co-occurred in the same sentence, he translated ren as ‘principle’ or ‘perfect virtue’ to clarify the meaning of the sentences. This translational approach demonstrates that James Legge, as a Western missionary, undertook his translation of the Analects in order to facilitate the understanding of his readers.

This study investigates the contrastive linguistic features of the Korean suffix '-jeok' and its functional equivalent 'de' in Chinese and '-teki' in Japanese, focusing on their morphological and lexical realization within the framework of Third Language Acquisition. Based on lexical data extracted from the Yonsei Korean textbooks, the research classifies five types of contrastive patterns: monosyllables, word order variations, negative expressions, idiomatic expressions, and semantic-oriented reconstructions. The findings reveal that while approximately 75.9% of the '-jeok' lexicon shows full correspondence across the three languages, the remaining lexical items exhibit notable asymmetrical correspondences. Specifically, Korean and Japanese share high morphological similarity due to their common Sino-character derivation, whereas Chinese demonstrates additional semantic flexibility and favors context-oriented expressions. These results provide significant pedagogical implications for Korean language education, suggesting that learners' prior language experiences (L1 and L2) may act as cognitive filters that trigger both positive and negative transfers. Furthermore, the study highlights the need for a contrastive instructional approach that distinguishes the grammatical and semantic functions of '-jeok' to minimize cognitive interference for multilingual learners.

This study quantitatively analyzes the usage patterns of the Japanese response marker soudesune and its plain-form equivalents soudane and soune in Japanese animation, comparing the results with Son (2021), which examined real-world talk show data. Using an online multimedia corpus comprising 69 animation titles (approximately 427 hours), 314 tokens were analyzed in terms of semantic function (agreement vs. non-agreement), speaker gender, and accompanying nonverbal behavior. The main findings are as follows. First, the distribution of the two functions differed significantly across the three expressions, and 97% of the soune tokens were produced by female speakers, suggesting that soune functions as a gendered role language. Second, speakers uttering non-agreement soudesune were shown on screen at a notably high rate (90%), indicating the producers' intent to visualize the speaker's deliberation. Third, in sharp contrast to the real-world data, in which agreement tokens were accompanied by nodding in 91% of cases, nodding was observed in only one instance; instead, head-lowering, head-raising, and hand-to-face gestures conveyed speakers' attitudes. These results suggest that the primary factor determining the meaning of soudesune is the preceding utterance, while nodding serves as a supplementary cue.

This paper examines the selection conditions of antonymic adjective pairs in the modern Chinese interrogative construction “yǒu(有)+duō(多)+adjective?”. Korean grammar books on modern Chinese generally state that this construction tends to combine with adjectives carrying active or positive meanings. However, they do not sufficiently explain what “active” or “positive” means in this context. To address this issue, this study investigates the actual distribution of antonymic adjective pairs in “yǒu(有)+duō(多)+adjective?” interrogatives, based on data from the BCC corpus. It then analyzes the principles of adjective selection from the perspectives of markedness, scale structure, and discourse context. The corpus analysis shows that adjectives such as cháng (長), kuān(寬), gāo(高), dà(大), zhòng(重), and shēn(深) occur far more frequently than their antonymic counterparts. This tendency does not mean that these adjectives are positive in terms of value judgment. Rather, they function as unmarked terms that can neutrally ask about the value of an entire scale. In contrast, marked terms such as duǎn(短), zhǎi(窄), ǎi(矮)·dī (低), xiǎo(小), qīng(輕), and qiǎn(淺) are selected mainly when a specific direction of the scale is highlighted in discourse. However, cases such as rè(熱)/lěng(冷), pàng(胖)/shòu(瘦), and cū(粗)/xì(細) show that adjective selection cannot be explained by markedness alone. The type of scale structure, the typical properties of the subject, and discourse salience also play important roles. Therefore, the selection of antonymic adjective pairs in this construction is based primarily on the quantitative openness of unmarked terms, but is also shaped by scale structure and discourse context.

This mixed-methods study examined students’, parents’, and teachers’ perceptions of the factors sustaining private English education in South Korea. Surveys involved 150 students, 36 parents, and 73 teachers, and semi-structured interviews involved four parents, two teachers, and one English education policy administrator. Survey data were analyzed using descriptive statistics, Welch’s ANOVA, and Games–Howell post hoc tests, while qualitative data were analyzed thematically. All three groups generally viewed private English education as helpful for English learning and achievement, with improved school grades as a major reason for participation. Agreement that distrust of school education motivates participation was relatively low. Group differences were significant in 11 of 13 common survey items, but not for perceived effects on English grades and future opportunities. The persistence of private English education was linked to school grades, school assessment practices, college admissions competition, parental expectations, and fear of falling behind. Reducing dependence on private English education requires aligning school instruction with assessment, strengthening differentiated and remedial learning, and ensuring equitable English learning opportunities regardless of families’ economic circumstances.

This study conducts a critical discourse analysis (CDA) of China-related economic headlines published in 2025 by the Korean newspaper JoongAng Daily, using the transitivity system of Systemic Functional Linguistics (SFL). It examines how China’s economy is represented in Korean media discourse and how ideological meanings are embedded in news headlines. The analysis shows that material processes (64.6%) are the most frequent, followed by relational (18.3%) and mental processes (8.5%). The predominance of material processes indicates that China’s economy is mainly represented through action-oriented processes. The findings suggest that JoongAng Daily frames China–South Korea economic relations in terms of competition and potential threat, thereby foregrounding China’s rapid economic rise and expanding influence.

Repatriated international Chinese language teachers often face multifaceted challenges, including reverse culture shock and career transition difficulties. Drawing on Conservation of Resources (COR) theory, the Broaden-and-Build framework, and reverse culture shock, this study investigates how psychological capital (PsyCap) influences career development, and examines the mediating roles of social support and cultural re-adaptation. Using structural equation modeling on data from 369 repatriated teachers, results indicate that PsyCap significantly and directly predicts career development (β=0.447, p<0.001; total effect=0.884), functioning as a primary driver rather than merely a mediator. Social support partially mediates the relationship between PsyCap and career development (indirect effect = 0.335, 95% CI [0.235, 0.434]). We identify a distinct pathway—from psychological advantage to network resources to professional achievement—that resonates strongly with the logic of guanxi (relationship-building) and collective efficacy characteristic of collectivist societies. Notably, cultural re-adaptation showed no significant direct effect on career advancement (β=0.117, p=0.136). From a COR perspective, early-stage repatriation is a resource-draining defensive response focused on homeostasis, not a gain cycle; thus, it yields no immediate promotional dividends. Findings suggest that intervention strategies should pivot from generalized cultural adaptation training toward cultivating PsyCap and institutionalizing professional support networks.

This study examines how Korean research on AI-assisted English writing is represented at the level of publication abstracts by applying text-mining procedures to a bounded corpus of English-language abstracts indexed in RISS between 2015 and 2025. To this end, we analyzed a corpus of 180 English-language abstracts retrieved from peer-reviewed journal articles and graduate theses indexed in RISS. A sequential text-mining workflow was employed, combining word-frequency, TF–IDF, n-gram, co-occurrence, LDA, and CONCOR clustering to identify macro-level trends. The results show that ChatGPT, feedback, and learning constitute central semantic hubs, indicating an increased abstract-level emphasis on LLM-supported feedback and process-oriented writing support alongside continuing attention to automated evaluation and linguistic accuracy. In addition, CONCOR clustering identified four higher-order thematic domains: (1) language generation and translation support, (2) feedback structures emphasizing grammar and self-regulated revision, (3) AI-integrated instructional design in digital environments, and (4) collaborative learning and evaluationoriented empirical research. Across these domains, the abstracts recurrently foreground feedback, linguistic accuracy, learning processes, instructional support, and teacher mediation. Taken together, these findings provide a basis for cautious interpretive reflections on AI-assisted writing research and suggest directions for future corpus expansion, full-text analysis, and mixed-method validation.

Generative AI may support autonomous language learning, but its feedback can extend beyond the learner’s intended revision target. This study examines the characteristics and limitations of Gemini-generated feedback on Japanese dialogues written by early-intermediate university students in Korea and considers its potential for autonomous learning. The data consisted of 99 dialogues produced by 32 students. After an exploratory examination of prompt conditions intended to reduce input burden and excessive rewriting, one prompt was selected and applied to all dialogues. The results showed that, even when instructed to revise only grammatical and lexical errors while preserving the original text, Gemini also modified expressions related to conversational naturalness and pragmatic appropriateness. It also produced missed errors and inappropriate revisions. These findings suggest that Gemini can provide feedback beyond linguistic accuracy, but learners need to evaluate whether revisions preserve their original meaning and communicative intentions.