Earticle

This study explores the issues of using present indicative in place of shall in the English translation of Korean statutes. Due to the Plain Language legislative attempt over the past few decades, shall is now rapidly being replaced by the simple present and other performative forms in many common law countries. Against this backdrop, this study compiled three sub-comparable corpora consisting of translated Korean Statute Corpus (TKSC) and non-translated British Statute Corpus (BSC) and American Statute Corpus (ASC), and examined the frequencies, types, and patterns of present indicative in each corpus. In particular, the present study separately calculated the frequencies of present indicative forms that convey prescriptive force using CLAWS 7 tagged text files. Analysis of the three comparable corpora shows many discrepancies in the use of present indicative between translated corpus and non-translated corpora. While the simple present is restrictively used in definition provisions in TKSC, many present simple verb forms are found in ASC and BSC as a performative marker to indicate constitutive rules, application provisions, deeming provisions, and other declaratives.

Lexical bundles play an important role in supporting fluency, coherence, and rhetorical clarity in academic writing. Their use is particularly relevant in argumentative essays, where writers must present claims, organize ideas, and maintain cohesive argumentation. Although previous research has examined lexical bundles in learner writing, limited work has compared learner-produced bundles with those found in AI-generated texts, despite the increasing use of artificial intelligence in educational settings. This study investigates the lexical bundles used in argumentative essays written by Indonesian EFL learners and compares them with bundles appearing in AI-generated texts. The objective is to identify the most frequently used bundles in each corpus, examine their structural and functional patterns, and describe the differences observed. A corpus-based comparative method was employed, drawing on two datasets: a learner corpus of university student essays and an AI-generated corpus produced by DeepAI and SuperAI. Lexical bundles were identified and categorized by structure (NP-based, VP-based, PPbased, others) and function (research-oriented, text-oriented, and participant-oriented). Frequency and distribution measures were also calculated. The findings show that the two corpora differ in their most frequent bundles, structural compositions, and functional distributions, with each corpus displaying distinct patterns of lexical bundle use in argumentative writing.

Nominal tautological constructions (NTCs), as a special kind of syntactic and semantic structure, have aroused considerable interests among linguists from different perspectives. By analyzing the data from the Corpus of Contemporary American English (COCA) and Center for Chinese Linguistics PKU (CCL), this paper conducts a contrastive study of NTCs in English and Chinese, aiming to find out the similarities and differences on the one hand, and the cognitive interpretation on the other hand. The analysis of data from the COCA and CCL corpora shows that NTCs in English and Chinese exhibit different distributional features, providing a basis for further analysis. It is found that similarities of English and Chinese NTCs are often embodied in similar basic repetitive structures and similar semantic representation such as ‘acceptance, tolerance, or distinction’. The differences between English and Chinese NTCs lie in either the language-specific or culture-specific interpretation of some types of NTCs in each language. Finally, this paper tries to expound the cognitive interpretation of NTCs from the perspective of categorization and decategorization theory.

This study investigates the productivity and frequency of the near-synonymous derivational suffixes -ish and -like in the British National Corpus. In particular, it examines the bases to which they attach and the meanings imparted in each instance. Some of the findings correspond to work conducted with other corpora or in the theoretical realm, but the study also found a definite pattern of -like attaching to adjectives to form adverbial expressions, in much the same was that -ly does, though conveying a unique meaning. The study also examines the use of adverbial modifiers of words formed with the two suffixes and examines how these serve to modulate the semantic contributions of the affixed words.

This paper aims to compare generic features of a specific genre in Korean and English language. For this purpose, written advertisements from magazines in two languages were compiled and small corpora of equivalent genres were analyzed from a perspective of comparative genre analysis. The focus of analysis was given to the ‘lexical’, ‘structural’, and ‘other choices’, by exploring the discourse values within a specific context. The case is to raise foreign language learners’ awareness of language form within a meaningful context, represented in a relevant genre. In this study, the comparison is narrowed down to written advertisements through cross-linguistic comparison. It shows that comparative corpora of familiar genres can be used to help learners realize the ways FL is explored to achieve particular communicative goals. Furthermore, it can lead naturally to the discussion of cultural and social aspects reflected in the language.

Jiang Linghui. 2014. A Comparative Study on Chinese ‘有(you)’ and Korean ‘있다(issta)’ with Korean-Chinese Parallel Corpus. Journal of Linguistic Studies 19(3), 255-280. This paper attempts to make a comparative study on Chinese ‘有(you)’ and Korean ‘있다(issta)’ through Korean-Chinese parallel corpus by implementing the usage analysis of sentences. In Chinese, ‘有(you)’ has many meanings, and the most frequently used is ‘to possess' or 'to own’ and ‘to exist’. Generally, Chinese-studying students think Chinese ‘有(you)’ is the same as Korean ‘있다(issta)’. It is because in many cases Chinese ‘有(you)’ is directly interpreted or translated to Korean ‘있다(issta)’ as a one-to-one correspondence. This paper reorganized the meanings of Chinese ‘有(you)’ through the existing dictionaries and grammar books, observed and analyzed the meanings by the sentences which were extracted from Korean-Chinese parallel corpus, and contrasted Chinese ‘有(you)’ with Korean ‘있다(issta)’ based on these sentences. (Yonsei University)

A good foreign language curriculum consolidates the three important principles: sequence, continuity and integration. Since English textbooks are the incarnation of the curriculum used in foreign language classrooms, a comparative investigation of the principles in English textbook corpus enlightens differences in English education of South Korea (SK) and North Korea (NK). To this end, the current study analyzes how graded sets of vocabulary are distributed to discover the sequence in the textbooks, and what the repetition rate is by looking into Standardized Type Token Ratio (STTR) to find out the continuity in the textbooks, and what keyword list is in both positive and negative keyword list of the NK corpus against the SK corpus to elicit integration property of the curriculum. The findings are: (1) NK textbooks are structural in sequence, and no such structural sequence is found in SK textbooks. (2) SK textbooks recycle vocabulary more frequently than NK textbooks, and this leads to more rote memorization activities in NK textbooks. (3) NK textbooks are more integrated with other academic subjects such as science, geography and ideology.

The purpose of this paper is to investigate and compare the use of three time adjectives new/old/young in high school textbooks of Korea and Japan. The main findings of the study are as follows. First, while the order(high to low) of frequency in the use of time adjectives such as new:135/231: old: 125/188 and young 73/100 is consistent between Korean and Japanese textbooks, the frequency of the three adjectives used is higher in Japanese textbooks than in Korean textbooks. Second, the proportion of time adjective+noun expressions out of all common combined nouns in both textbooks is as low as 30%, hence it is reasonable to state that there is more commonality than difference in combined nouns of adjective+noun expressions of Korean and Japanese textbooks. Third, the proportion of common time adjective+noun expressions between COCA(top 20) and Korean/Japanese textbooks is just approximately 55% and 52%, respectively. These results strongly suggest that we should consult native English speakers’ corpus when we write and publish English textbooks.

This study investigated the structural and functional differences between formulaic sequences in College Scholastic Ability Test (CSAT) and SAT reading assessment texts. The results of the RANGE program showed that SAT used more diverse and difficult words compared to CSAT. The frequency analysis revealed that CSAT used more formulaic sequences than the SAT counterpart. This suggests that CSAT used more repeated expressions, whereas SAT used diverse vocabulary items. The structural analysis showed that noun phrases were the most dominant in CSAT, whereas prepositional phrases were pervasive in SAT. The functional analysis showed that both corpora relied heavily on referential expressions. The results indicate that referential bundles are dominantly used in institutional writing (Biber & Barbieri, 2007). In accordance with the previous studies, the results suggest that high frequency formulaic sequences can be different according to the register. In CSAT, connectives and discourse organizers were prevalent. This may be due to the characteristics of question types in CSAT. The results may indicate that formulaic sequences in texts are partly influenced by the characteristics of a register.

The purpose of this study is to investigate on usage for 'Kajang, Deo, Deol' of degree adverbs and extract a list of co occurrent words with 'Kajang, Deo, Deol by the part of speech presented in Korean University textbooks for Korean learners. For this study, the corpus was analyzed through classifying 59 Korean university textbooks to extract co occurrent words named predicates by the part of speech at different level of textbooks. The result shows that 'Kajang' and ‘Deo’ has a higher frequency of cooccurrence with gradable antonym adjectives in each level. In contrast, ‘Deol’ has rare cooccurrence with adjectives. In the aspect of usage with verbs of ‘Kajang’ and ‘Deol’ is highly collocated with ‘like’ verb in every level. Especially, ‘Kajang’ is collocated with some psychological verbs. ‘Deo’ is highly collocated with some kind of action verbs intermediate and advanced level. In the aspect of usage with adverbs, ‘Kajang’ is collocated with ‘mani’ ‘Deo’ is collocated with ‘jom’ in basic level and ‘jeomjeom, jogeum, holshin’ in intermediate and advanced. In the aspect of usage with nouns ‘Kajang’, ‘Deol’ is collocated with ‘static noun’ but ‘Deol’ has rare frequencey of ‘noun’. Noteworthy results are that ‘Deo’ and ‘Deol’ has some marked meaning when collocated with special action verbs. Deo has a marked meaning of ‘ strengthening’ when combined with action verbs such as ‘eat, give, wait, go’ and 'Deol' have a marked meaning of 'incompletion'.when combined with action verbs such as ‘wake up, do, leak, write. In hence, I will also present some possibility of application for Korean Language Education for Korean learners based on Corpus analysis. Eventually I expect that the results of this study will be applied to teaching methods of the degree adverbs 'Kajang', 'deo', 'deol' for Korean learners, and construction of Korean textbooks.

The purpose of this study is to compare the usage of the coordinator ‘and’ between the International Corpus of English(USA) and a Korean learner corpus of English. For this study, 120 sentences were randomly selected, both from the ICE and the Korean learner corpus that was collected from basic English writing courses in university. The authors analyzed the syntactic and semantic aspects of the coordinator to find out the similarities and differences between native English speakers and Korean learners of English. In this study, we examine the eight types of connotation by Quirk et al.(1985) and the six types of coordinated constituents, viz. NP, PP, AP, AdP, VP and S to find out if Korean learners use it in a proper way. This study argues that they are not fully aware of linguistic principles such as iconicity and grammatical knowledge of English such as post-nominal modifier, nominalization and collocation. Based on the results, it suggests that Korean learners should learn and be exposed to the coordinator in various register.

The study aims to build a Maritime Law English Corpus related with oil spill, to comparatively analyze key semantic domains, and to explore similarities and differences among three subcorpora by using a network analysis. Key semantic domains are extracted from three different maritime law texts such as international public maritime law, international private maritime law, and agreements of maritime accidents through Wmatrix keyness analysis to compare each of three corpora with BNC Sampler Informative. We focus on key semantic domains shared with three subcorpora by using a Netminer 4.0 program. Compensation agreements for maritime accidents related with oil spill will be made with regard to international public maritime law and international private maritime law. Comparative analysis of maritime law texts will make a great contribution to corpus linguistics and will also provide fodder for identifying the differences and similarities between the similar laws through quantitative analysis.

The aim of this paper is to compare the frequency in Korean degree adverbs between a Korean corpus and Korean Language Teaching textbooks (i.e., Seoul National University Press, Yonsei University Press, Gyeongsang National University Press, Sogang University, and Kyung Hee University). The results indicate that the frequency of Korean degree adverbs in a Korean corpus is almost the same as the frequency occurring in Korean Language Teaching textbooks. This analysis also shows what type of degree adverbs are most frequently used in a Korean corpus, both in spoken and written form, and which degree adverbs should be included in Korean Language Teaching textbooks for beginners in the future. Considering the high frequency of some degree adverbs, such as “maeu,” “deo,” and “gajang” in real-life situations, this study suggests teaching these adverbs at the beginning level. Regarding teaching degree adverbs, this study will contribute to publishing Korean language teaching textbooks with more authentic data.

본 연구는 모어가 다른 일본어학습자들의 접속조사 ｢から｣와 ｢ので｣의 습득과정에 나타나는 중간언어의 양상에 대해 두 가지 접속조사의 사용기준을 전항에 접속하는 항목과의 접속형태, 오용경향, 문체라는 측면에서 분석한 것이다. 본 연구의 분석을 통해 밝혀진 일본어학습자의 접속조사 ｢から･ので｣의 습득경향을 정리하면 다음과 같다. (1) 일본어학습자의 ｢から･ので｣는 ｢から｣를 중심으로 사용되며 학습레벨이 올라갈수록 ｢ので｣의 사용이 늘어난다. 여기에는 학습자들의 덩어리 표현이라는 중간언어, 오용의 적극적인 회피전략 등이 영향을 미치고 있다. (2) ｢から･ので｣의 오용에 나타나는 특징으로서 의미용법 상의 오용이 거의 없고, 접속하는 형태에 의한 오용이 대부분을 차지한다는 점, ｢から･ので｣의 습득에 있어 회피가능한 대체문법항목이 존재하고, 두 가지 항목 사이의 치환이 가능하기 때문에 타 문법항목에 비해 상대적으로 오용이 적게 나타난다는 점을 들 수 있다. (3) ｢から･ので｣가 사용된 문의 문체적 측면에서는 선행연구와는 달리 ｢から→정중체｣, ｢ので→보통체｣의 형태로 사용되는 것으로 나타났다. 이는 학습자에게 있어 접속사 ｢ですから｣가 덩어리 표현으로 정형화되어 있어 문중에서도 그대로 사용되기 때문이라 생각할 수 있으며, 이것이 오용의 회피수단으로서도 활용될 가능성이 있다. (4) ｢から･ので｣에 접속하는 형식을 분석한 결과, ｢から｣의 경우 ル형식과 접속하고 있는 경향이 강하며, 상대적으로 イ형식과의 접속이 적고, ｢ので｣의 경우 ル형식과 접속하는 수가 가장 많았으나 그 수는 タ형식과의 접속과 큰 차이가 없었다. 이러한 결과를 통해 ｢から｣의 경우, 학습자에게 ｢だから｣라고 하는 덩어리가 가장 강하게 존재할 가능성이 있으며, ｢ので｣의 경우는 ｢なので｣, ｢たので｣라는 덩어리가 사용에 영향을 미치고 있음을 짐작할 수 있다. (5) 위와 같은 사용경향은 언어유형별로 크게 차이를 보이지 않았다. 이는 이러한 학습자들의 유니트 형성 등의 중간언어 전략이 모어와 상관없이 비교적 보편적으로 형성되고 사용됨을 나타내는 결과라 할 수 있다.

I analyzed this study from an item located about an intermediate language to appear in the acquisition process of “Kara” and “Node” of the Japanese learner who varied in mother tongue forward, a tendency of the misuse, three sides of the writing style. The results of the analysis are as follows. (1) The use of “Kara” and “Node” of the Japanese learner is carried out around “Kara”, and use of “Node” increases as a learning level rises. It is thought that the strategies that are going to evade an intermediate language, misuse lump expression of the learner positively influence it here. (2) Most of the legal misuse for the meaning is not seen, and an avoidable substitute grammar item exists in a point that misuse by the form to be connected occupies most of, the acquisition of “Kara” and “Node”, and characteristics to emerge for misuse of “Kara” and “Node” include the point where misuse is hard to relatively in comparison with other grammar items appear because the substitution between two items is possible. (3) In the side of the writing style of the sentence that “Kara” and “Node” was used for, use in the form of “Node → plain style” was seen “Kara → formal style” unlike a precedent study. “Desukara” of the conjunction hardens for a learner, and this is made a fixed form as expression, and this is because it is just used in the sentence, and the possibility that I am used as evasion means of the misuse is thought about. (4) As a result of having analyzed a form to connect with “kara” and “Node”, in the case of “Kara,” form and the tendency that I connected were strong, and there was relatively little connection with the “I-form”. On the other hand, in the case of “Node,” there were the most cases to connect with “Ru-form and Ta-form”. It is thought that the lump by the learner is formed to a learner from such a result in each item. (5) A wide margin is not seen according to a language type, and, as for the tendency to these use, as for the intermediate language strategies such as the unit formation of the learner, it is thought that the formation is used regardless of mother tongue relatively and universally.

本研究は、母語の異なる日本語学習者の接続助詞｢から｣と｢ので｣の習得過程に現われる中間言語の様相について、前接する項目との接続、誤用の傾向、文体という3つの側面から分析したものである。分析の結果は次の通りである。 (1)日本語学習者の｢から･ので｣の使用は、｢から｣を中心に行われ、学習レベルが上がるにつれ ｢ので｣の使用が増える。ここには学習者の固まり表現という中間言語、誤用を積極的に回避しようとする戦略などが影響していると考えられる。 (2)｢から･ので｣の誤用に現われる特徴として、意味用法的な誤用はほとんど見られず、接続する形による誤用が大半を占めるという点、｢から･ので｣の習得において回避可能な代替文法項目が存在し、二つの項目の間の置き換えが可能であるために他の文法項目に比べ、比較的に誤用が現われ難い点が挙げられる。 (3)｢から･ので｣が使われた文の文体的側面においては、先行研究と異なり、｢から→丁寧体｣、｢ので→普通体｣の形での使用が見られた。これは学習者にとって接続詞の｢ですから｣が固まり表現として定形化され、文中においてもそれがそのまま使われるためであるとともに、誤用の回避手段としても用いられる可能性などが考えられる。 (4)｢から･ので｣に接続する形式を分析した結果、｢から｣の場合はル形式と接続している傾向が強く、比較的にイ形式との接続が少なかった。一方で｢ので｣の場合は、ル形式とタ形式に接続する場合が最も多かった。このような結果から、それぞれの項目に学習者に学習者による固まりが形成されていることが考えられる。 (5)以上のような使用傾向は、言語類型別に大差が見られず、学習者のユニット形成などの中間言語戦略は母語と関係なく比較的かつ普遍的に形成･使用されていると考えられる。

This study aims to explore stylistic differences between translated and non-translated news articles and their implications. While the stylistics of literary texts has long been a favorite subject in translation studies, that of non-literary texts has drawn little attention from researchers, despite the significance in translator education. Against this backdrop, this study attempts to highlight the stylistic aspects of non-literary texts based on the checklist categories of stylistics suggested by Leech and Short (2007): lexical categories, figures of speech, grammatical categories, and context and cohesion, with a focus on the last two. To this end, techniques of corpus stylistics, including keyword analysis, are employed to examine a parallel corpus of English news articles and their Korean translations, and a comparable corpus of Korean news articles compiled by the author. The findings of this study reveal stylistic differences between translated and non-translated news articles in such stylistic features as average sentence length, demonstratives, conjunctions, and direct and indirect quotations.

본 연구는 DeepSeek가 한국인 중국어 학습자의 작문 오류를 수정하는 데 있어 어떠한 양상과 특징을 보이는지 분석하고, 이를 인간 교사와 비교한 연구이다. 이를 위해 글로벌 중국 어 중간언어 코퍼스에서 한국인 학습자의 오류 문장을 무작위로 추출하여 DeepSeek-R1 모 델과 인간 교사의 수정 결과를 분석하였다. 주요 연구 결과는 다음과 같다. DeepSeek는 재현율(Recall, 95.86%) 면에서 인간 교사 (84.83%)보다 우수하여 학습자의 오류를 거의 놓치지 않고 찾아내는 높은 민감도를 보였다. 그러나 정확도(Precision)는 49.47%로 인간 교사(85.42%)에 비해 현저히 낮았는데, 이는 오류가 없는 부분까지 수정하는 과도 교정 현상이 빈번했기 때문이다. DeepSeek는 오타 수정 및 어휘 오류 식별에 강점이 있으며, 문장을 더 문어체적이고 격식 있게 다듬는 경향이 있다. 반면, 인간 교사는 최소 수정 원칙을 준수하며 학습자의 본래 의도와 어조를 최대한 유지하는 신중함을 보였다. DeepSeek는 대규모 데이터를 빠르게 처리하고 수정하는 데 유용하며, 인간 교사는 문맥 에 맞는 자연스러운 표현과 교육적 관점에서의 정확한 피드백을 제공하는 데 탁월하다. 결론 적으로 인공지능과 인간 교사의 협력은 중국어 교육 현장에서 수정의 효율성과 정확성을 동 시에 높일 수 있는 실질적인 방안이 될 수 있다.

This study analyzes the patterns and characteristics of DeepSeek in correcting writing errors made by Korean learners of Chinese and provides a comparative analysis with human teachers. For this purpose, error-containing sentences from Korean learners were randomly extracted from the Global Chinese Interlanguage Corpus, and the correction results of the DeepSeek-R1 model were compared against those of human teachers. The key research findings are as follows. DeepSeek demonstrated superior performance in terms of recall (95.86%) compared to human teachers (84.83%), showing high sensitivity in identifying almost all learner errors. However, its precision was significantly lower at 49.47%, compared to 85.42% for human teachers, due to frequent over-correction—modifying parts of the sentence that were originally error-free. DeepSeek showed strengths in correcting typos and identifying lexical errors, with a tendency to refine sentences into a more formal and written style. In contrast, human teachers adhered to the principle of minimum revision, exercising caution to preserve the learner's original intent and tone as much as possible. While DeepSeek is highly effective for rapidly processing and correcting large-scale data, human teachers excel at providing natural expressions suited to the context and accurate feedback from an educational perspective. In conclusion, the collaboration between AI and human teachers can serve as a practical approach to simultaneously enhance the efficiency and accuracy of error correction in Chinese language education.

Kabata, Kaori & Lee, Jeonghwa. 2010. A Corpus-based Comparative Study of Completives in Japanese and Korean. Korean Journal of Linguistics, 35-3, 837-862. Data from written and spoken corpus was analyzed to examine the synchronic usage patterns of completives in Japanese and Korean. Taking a panchronic approach, the present study aimed to determine how the similarities and differences of the synchronic behaviors of completives in the two languages might be attributable to the similarities or differences of their grammaticalization paths. Do the completives with similar lexical sources exhibit similar usages patterns between the two languages, or do they demonstrate some differences, and if they do, what factors might underlie such differences? The results indicated that the completive auxiliaries of the two languages exhibit a striking parallelism, but that they are at different stages of grammaticalization. Some discrepancies were also found between the usage patterns and semantic distributions discussed in the previous literature. (University of Alberta, The Cyber University of Korea, University of California, Berkeley)

Kim, Haeyeon. 2013. A Comparative Study of the Meanings and Uses of Weight-Denoting Words Kapyepta 'Light' vs. Mwukepta 'Heavy' in a Korean Written Corpus. Korean Journal of Linguistics, 38-1, 115-140. This research attempts to characterize two opposite adjectives kapyepta 'light' vs. mwukepta 'heavy' through an examination of a Korean written corpus. This study first examines the frequency of these two adjectives in the database, showing that kapyepta is more frequently used than mwukepta in the database. Second, it examines the frequency of grammatical functions of kapyepta vs. mwukepta, showing that about half of the kapyepta tokens take the adverbial form, although they are used both attributively and predicatively. Third, this study explores co-occurrence relations and collocation patterns of the two adjectives, showing that most of the kapyepta tokens are used to modify or predicate nouns which denote abstract notions, actions/activities, and attitudes, while mwukepta co-occurs mostly with nouns of concrete objects, actions/activities, and abstract entities. Fourth, this work explores semantic properties of kapyepta vs. mwukepta based on the examination of semantic categories of co-occurring nouns. This study also shows that idiomatic expressions can be explained in terms of conceptual metaphor theory. Finally, this research suggests that a corpus-based analysis is a useful tool in characterizing uses and meanings of many closely related words in Korean. (Chung-Ang University)

This study compares English and Korean’s exploitations of the heart in association with emotion and thought. While the heart as a container of emotion has generally been acknowledged across languages, the conceptualization of the heart associated with thought is regarded as language-specific and not holding in English. Examining the corpus data of English and Korean constructions, this research finds that this is an overly simplified conclusion and that each language uses the heart in describing both emotion and thought, to a greater degree in the former and to a lesser degree in the latter. The two languages differ in preferred conceptualizations of the heart associated with thought: Korean conceptualizes the heart as a container of thought and English conceives of the heart as a thinking site. The two languages are common in that they utilize the heart in describing intuitive thoughts and secrets, though in different patterns and to different degrees:Among thoughts contained in Korean speakers’ hearts, hidden thoughts were the most prominent and intuitive ideas were found the most among thoughts in English speakers’ hearts. The Confucian culture of valuing smooth human relationships by means of not revealing oneself and the Western culture of stressing rational reasoning and empirical evidence in knowledge seem to be reflected in Korean hidden thought constructions and English intuitive thought constructions, respectively.

This study examines the translation of Korean temporal connective endings into Chinese through the lens of cognitive linguistics. Drawing on Han Kang’s novel Yeosu and its Chinese translation, it introduces ‘Syntactic Translation Entropy’ as a quantitative tool to assess cognitive processing in translation. Analysis of 300 literal translations reveals that high-entropy endings involve greater translational diversity and necessitates dynamic contextual inference, whereas low-entropy endings exhibit stable form–meaning correspondences that provide reliable reference points for pedagogy. An analysis of 300 literal translations shows that high-entropy endings display greater translational variability and require dynamic contextual inference, whereas low-entropy endings demonstrate stable form–meaning correspondences that provide a data-driven foundation for the subsequent discussion of cognitive translation pathways. In addition, the study applies the concept of syntactic distortion proposed by Schaeffer et.al (2016) and 谭业升 (2020), together with 王寅 (2005)’s broad definition of grammaticalization, to 158 instances of grammaticalized translation. The findings show that shifts in word order, grammatical categories, and relations between subordination and coordination are pervasive, and that translators employ conceptual restructuring to preserve semantic accuracy and coherence. Overall, the research provides methodological innovation and the oretical insights. It offers practical guidance for translation practice and new perspectives for cross-linguistic studies.