Earticle

This study explores Chinese learners’ use of English passive voice through a comparative analysis of two learner corpora(ICLE_Ch and LOCNESS) and a reference corpus(BNC 2014). By analyzing the frequency of English passive constructions, syntactic errors, semantic prosodies, and over-passivization, this study reveals significant differences between Chinese learners and native English speakers. The findings show that Chinese learners use the passive voice less, associate it with more negative connotations, and make more syntactic errors, especially in agreement and passive participle formation. This study also found that Chinese learners tend to overuse the passive voice, especially with inanimate subjects. These differences are attributed to differences in Chinese and English syntax and semantico-pragmatics. This study infers the pedagogical implications that Chinese students should be encouraged to use more passive constructions in their English writings when necessary and to pay attention to the rules of tense and subject-verb agreement and passive participle formation. They should also be taught to avoid excessive passivization and the tendency to link English passives to negative semantic prosody.

Most task-based research examined the effects of task complexity on various aspects of L2 production, and this study aims to perform an in-depth analysis of task complexity effects on error patterns observed in L2 writing in terms of morphological, lexical, syntactic, and mechanical errors. 39 learners of English carried out one written task with three levels of task complexity, operationalized as the number of elements. Results of a series of repeated-measures ANOVAs revealed that greater task complexity led to significant increases in the amount of all four error types and the numbers of morphological, lexical, and local syntactic errors per T-unit. It was also found that learners are more likely to make errors in terms of plurality, word choice, and article use when they need to mention more elements when carrying out complex tasks.

Ideally, Korean-Chinese translation is to be done by translators who can use both Korean and Chinese as their mother tongue. In reality, however, most translators speak Korean as their native language and they received special training of translation. Thus it is an approved essential step to get proof reading by native Chinese speakers. Despite the step, it is virtually impossible to avoid mistranslation. Therefore, the present study purposed to analyze errors in Korean-Chinese translation and their types to give help to Korean-Chinese translators. The subject of this study was the Chinese home page of Seoul City. The researcher collected errors in the site from February to March in 2003, and examined errors by classifying them into vocabulary errors, syntactic errors, problems found from comparing the source language to the target language and cautions in writing translated texts.

The purpose of this study is to investigate different types of errors identified from the essays written by 68 Asian EFL learners at a Korean university. The learner corpus containing these essays consisted of 2498 errors; based on the schemes suggested by previous researchers, those errors were analyzed according to error classification, i.e., sentence fragments, run-ons, comma splices, word order, verb tense, verb form, subject verb agreement, word form, determiners, prepositions, and word choice. Of all the error types committed by college-level learners, word choice had the highest error frequency, accounting for around 33%. Word choice issues in EFL students’ writing could be due to a traditional way of English learning such as rote memorization of vocabulary and purely receptive reading skills. Thus, instructor suggestions on practical learning methods concerning word choice in writing can be certainly helpful and teachers would also need to pedagogically utilize EFL written word choice strategies in a more effective way.

This paper examines syntactic errors in 50 Chinese college EFL learners' writing. The syntactic errors include the following errors of ‘run-on sentence’, ‘subject ellipsis’, ‘word order’, ‘conjunction’, ‘relative clause’, and ‘subject-verb agreement’. Their error rates are 32.20%, 28.00%, 17.79%, 11.01%, 7.62%, and 3.38%, respectively. We argue that syntactic errors are the gravest ones in writing because they give a bad impression to the rater. We propose that all these errors are caused by interlingual differences between English and Chinese, mainly derived from different language typology with the former as a hypotactic and the latter as a paratactic language. Accordingly, Chinese EFL teachers are suggested to explicitly teach their students English grammar via directing their attention to these interlingual syntactic differences.

With regard to the effects of educational system pathology for improving language learning, through syntactic and grammatical errors and contrastive linguistic analysis theory, the researcher has examined the errors of persian speakers majoring in Arabic writing. The participants of this study includes the students of Isfahan and Kashan universities from 2007 to 2008 (2007- 2008) amongst those 56 students at their 7th semester are chosen. The instrument used in this study is a piece of writing (composition). Moreover, random sampling has been used. That is, randomly two universities of Isfahan province have been selected, and amongst the students of these universities, the 7th semester students have been chosen. Based on the results of this study, it has been proved that interlingual and intralingual interference are two of the most important sources responsible for the Syntactic and Grammatical errors in students writing.

Arabic is ruled by standard norms firmly fixed in the Koran. It is significant that the standard Arabic was derived from the Hejaz (the western part of Saudi Arabia) dialects. So the standard Arabic does not depend on the Koreish dialect only, but also on the other dialects of western Arabia. Nowadays many Arabs and non-Arabs make many errors in their use of Arabic. This is because of Arabic's very complicated grammatical rules which make it hard to find anyone who can write and speak Arabic without errors. On the contrary, we could find them make errors in the syntactical, morph-syntactical, and lexical errors are evident. Therefore this research aims to study the varieties of grammatical errors in order to discover why speakers make those mistakes errors in Arabic, and how they could avoid them. Particular emphasis is placed on the range of permissibility, i.e. whether these errors may be accepted and considered as part of standard Arabic.

This study is a study on the syntactic errors that frequently appeared in writing problems of second-year Korean language students at the National University of Thailand. It is a study to analyze errors related to investigations in which Thai Korean majors made the most mistakes in writing classes at actual schools and to suggest alternatives to them. To this end, in the second-year writing class, which the researchers were in charge of, a writing task that students would be familiar with was assigned and analyzed. As a result of analyzing the writing tasks of Thai second-year Korean students, the researchers, as expected, found the most errors in the use of the survey. In addition, unlike Thai, which is an isolated language, the characteristics of Korean, which is a deadlock language, have shown various types of grammatical errors related to investigation and mother use. As the first study, this study first conducted an analysis of investigation errors. In terms of syntax, the confrontation, omission, and excessive addition of investigations were the most prominent, and various errors were directly identified in connection endings, tubular endings, tense, sentence word order, and response of sentence components. In terms of vocabulary semantics, there were many inscriptions due to the incorrect use of investigations, and other errors were found such as the use of context-nonconforming vocabulary and unclear expressions that were sometimes difficult to grasp the context. It is thought that one of the main reasons for the occurrence of errors in writing of Korean students majoring in Korean is the difference and difficulty according to the characteristics of Thai and other Korean. As an alternative to this, first, I think in the writing class, classes by level are required, not by grade Second, it is necessary to continuously learn the difference between Korean colloquial and octopus. Third, it was suggested that it is effective to teach the continuity of frequently appearing vocabulary in relation to the survey.

본 연구에서는 일련의 연구에서 수집된 영작문 오류 유형의 정제된 자료를 토대로 연관 규칙을 생성하고, 학습을 통해서 효용성이 검증된 연관 규칙을 활용해서 영작문 데이터의 형태 통사 오류를 자동으로 탐지한다. 영작문 데이터에서 형태 통사 오류를 찾아내는 작업은 많은 시간과 자원이 소요되는 작업이므로 자동화가 필수적이다. 기존의 연구들이 통계적 모델을 활용한 어휘적 오류에 치중하거나 언어 이론적 틀에 근거한 통사 처리에 집중하는 반면에, 본 연구는 데이터 마이닝을 통해서 정제된 데이터에서 연관 규칙을 생성하고 이를 검증한 후 형태 통사 오류를 감지한다. 이전 연구들에서는 이론적 틀에 맞추어진 규칙 생성이나 언어 모델 생성을 위한 대량의 코퍼스 데이터와 같은 다량의 지식 베이스 생성이 필수적인데, 본 연구는 적은 양의 정제된 데이터를 활용한다. 영작문 오류 유형의 형태 통사 연관 규칙을 생성하기 위해서 Apriori 알고리즘을 활용하였다. 알고리즘을 통해서 생성된 연관 규칙 중 잘못된 규칙이 생성될 가능성이 있으므로, 상관성 검정, 코사인 유사도와 같은 규칙 효용성의 통계적 검증을 활용해서 타당한 규칙만을 학습하였다. 이를 통해서 축적된 연관 규칙들을 영작문 오류를 자동으로 탐지하는 실험에 활용하였다.

In many cases, a dialogue system can't extract the correct analysis information of a user's spoken utterance, because of its own ungrammatical components. Therefore, in order to perform a correct before it performs the syntactic processing. In this paper, we use a real dialogue corpus and classify these ungrammatical errors as 4 categories : phonological, syntactic, semantic errors that consist of speech reparis and inversions, and propose an algorithm to detect and correct the errors. In short, this paper proposes a method to detect and correct the speech repairs and inversions that are classified as the phonological and syntactic errors to implement a robust dialogue system. And, through the test of real dialogue data, this paper shows an efficiency of the proposed algorithm.

Jieun Lee and Hyoeun Choi (2022). A study on the quality of patent neural machine translation: A comparison of omission and syntactic errors in the Korean-English translations by patent-specialized Patent Translate and WIPO Translate. This paper aims to evaluate the quality of patent translations produced by the two patent-specialized machine translation engines, EPO’s Patent Translate and WIPO’s WIPO Translate. For manual evaluation, four experienced patent translators or patent translation service managers evaluated the quality of 106 English sentences from the translations of 30 Korean patent abstracts by the two MT engines. In the automatic evaluation, Patent Translate slightly outperformed WIPO Translate, whereas in the manual evaluation WIPO Translate outperformed Patent Translate. According to the error annotations provided by the evaluators, WIPO Translate produced more omission errors than Patent Translate but handled the complex syntax of the source text better, while Patent Translate produced more syntactic errors than WIPO Translate. The results indicate that in automatic evaluation, MT outputs with fewer omissions were rated higher, while in manual evaluation, comprehensible and accurate syntactic structures appeared to determine the overall quality evaluation. (Ewha Womans University, Korea)

본 연구의 목적은 브로카 실어증, 베르니케 실어증, 명칭실어증 간의 문장이해능력 차이를 살펴봄과 동시에 자연회복 기간내 환자군과 자연회복 기간이 지난 환자군들간의 문장이해능력 차이를 살펴보고자 하였다. 연구에 참여한 대상자는 총 36명(실어증자 18명, 정상인 18명)이었다. 실어증 환자의 문장이해 특성을 살펴보기 위해서 사용한 실험과제는 음운적 오류, 통사적 오류, 의미적 오류가 있는 문장과 없는 60문장을 판단하게 하는 과제를 사용하였다. 실어증 유형에 따른 음운, 의미, 통사별 문장이해능력을 살펴본 결과, 세 실어증 모두 통사>음운>의미 순으로 오류가 많았으며, 자연회복 기간 내 그룹과 자연회복이 지난 그룹간의 문장이해능력을 살펴본 결과 유의미한 차이가 없었다. 실어증 유형에 따른 문장이해능력은 실어증 모두 통사>음운>의미 순으로 오류가 많았으며, 실어증 발병후 자연회복 기간에 따른 문장이해능력에는 유의미한 차이가 없었다.

This study aims to investigate the difference in sentence comprehension characteristics according to the types of aphasia and the difference in the length of the period after onset of aphasia. 36 subjects participated in this study: 18 aphasic subjects and 18 normal subjects. The task for the examination was to jurge if visually presented sentences are correct or wrong. The sentences were prepared with regard to the individual realms (phonology, semantics and syntactics) of language, according to the characteristics of patients with aphasia. The comparison of sentence comprehension by the patients with aphasia according to the realms of language showed that for all the 3 types of aphasia, the highest error rates were in the realm of syntactics, followed by phonology and semantics. The comparison of sentence comprehension according to the length of period after the onset the aphasia, between the group in natural recovery period and the group after natural recovery period showed no significant difference. This study show that the patients with aphasia were common in that they had difficulty in the realm of syntactics.

본　연구는 한국 청각장애학생의 조사오용에 미치는 통사론적 영향과 음운론적 영향을 검토하기위해 일본 청각장애학생의 조사오용과 비교하는 것을 목적으로 했다. 대상자는 한국 농학교 고등부에 재학　중인 학생 44명과 일본 농학교 고등부에 재학 중인 학생 51명으로 총 95명이었다. 그 결과 통사적으로는 한국과 일본 모두 명사와 동사의 관계를 나타내는 격조사의 오용이 명사와 명사의 관계를 나타내는 격조사의 오용보다 유의하게 많았다. 또한, 한국 청각장애학생의 조사오용의 평균산출률은 한국어 조사의 특징인 음운론적 영향에 의한 조사오용을 빼기 전에도 후에도 일본 청각장애학생의 조사오용의 평균산출률보다 유의하게 높았다. 이것으로 한국의 청각장애학생들의 조사오용에 미치는 음운론적 영향은 크지 않을 가능성을 시사했다. 그러나 음운론적 영향의 조건의 차이가 있는 조사「와/과」에 한해서는 음운론적 영향에 의한 조사오용의 평균산출률이 통사론적 영향에 의한 조사오용의 평균산출률보다 높은 경향을 보였다.

The present study examined syntactic and phonological influences on particle errors in Korean students with hearing impairments by comparing them to particle errors in Japanese students with hearing impairments. The participants were 44 high school students in deaf schools in Korea and 51 high school students in deaf schools in Japan for a total of 95 students. The results were as follows. Both in Korean and Japanese students, particle errors due to syntactic influence were significantly higher for case-markers that express the relationship between a noun and a verb as compared to errors in case-markers that express the relationship between two nouns. In addition, the average occurrence ratio of particle errors in Korean students with hearing impairments was significantly higher than the average occurrence ratio of particle errors in Japanese students with hearing impairments, even after exclusion of particle errors due to phonological influence, which are typical of the Korean language. This suggests that phonological influence on particle errors is limited in Korean students with hearing impairments. For the particle “wa/kwa” however (the phonological formation of which is irregular in comparison to other particles influenced by phonological factors), the occurrence of particle errors due to phonological influence tended to be higher than particle errors due to syntactic influence.

목적: 학령기 아동들이 문장을 능숙하게 사용하는 것뿐만 아니라 정확한 문장을 사용하는 것이 중요하다. 문장 자체에 오류가 있으면 청자나 독자가 내용을 이해하는 데 어려움을 겪을 수 있다. 성인기까지 문장 오류들이 나타나므로 효과적인 의사소통을 위하여 문장지도의 조기 개입이 필요하다. 이를 위하여 초등학교 저학년아동의 구문 산출 특성을 분석하였다. 방법: 본 연구에서는 구어와 문어 능력이 정상범위에 속하는 초등학교2학년(20명)에게 Mayer(1969)의 ‘Frog, where are you?’를 보여준 후 자발적으로 이야기를 말하거나 쓰게 하였다. 전사한 자료에서 구문 산출성(어절 수, 절 수, AS-unit 수, AS-unit 당 어절 수, AS-unit 당 절 수)과구문 연결성(종결어미, 연결어미, 전성어미, 접속부사), 구문 오류율(연결어미 오류, 접속부사 오류, 호응 오류, 순서 오류, 생략 오류, 중복 오류) 등을 분석하였다. 말하기와 글쓰기 간 구문 산출성ㆍ복잡성, 연결성, 오류율의 차이를 비교하였다. 결과: 초등학교 저학년 아동은 글쓰기보다 말할 때 더 많은 어절과 절을 산출하였다. 그러나 말하기와 글쓰기 간 구문 길이와 구문 복잡성의 차이는 없었다. 구문을 연결할 때 연결어미를가장 많이 사용하였다. 종결어미는 글쓰기에서 유의하게 더 많이 사용하였고, 전성어미는 말하기에서 더 많이사용하였다. 문장 구조의 오류율 측면에서, 말하기와 글쓰기 모두에서 주어 생략 오류율이 가장 높았고 목적어 생략 오류율, 연결어미 오류율이 높았다. 결론: 주어 생략이 빈번한 우리말에서 구문 산출성과 연결성, 오류를 분석하기 위하여 AS-unit을 활용할 수 있었다. 학령기 저학년 아동의 구문 지도 시 특정 오류에 대한 정보를 제공한 것에 의의가 있다.

Purpose: It is important that school-age children use correct sentences as well as have proficient use of sentences. If there are errors in sentences, a listener or a reader may have difficulty understanding the content. It is necessary to teach syntactic structure to school-age children for effective communication, otherwise they may produce sentences with errors until adulthood. Thus, the syntactic characteristics of elementary school students were analyzed. Methods: Participants in Grade 2 (N=20) had normal spoken and written language competences in standardized tests. Participants voluntarily spoke and wrote stories after viewing the picture book, 'Frog, Where Are You?' (Mayer, 1969). Syntactic productivity (e.g., total number of words), complexity (e.g., clauses per AS-unit), syntactic connectivity (e.g., conjunctive ending of predicate, conjunction), and grammatical errors (e.g., conjunction, omission) were analyzed. A paired-sample t-test was used to compare the difference between speaking and writing of the syntactic productivity, complexity, connectivity, and the ratio of errors. Results: The students produced more words and phrases when speaking compared with writing. There was no difference in syntactic complexity between speaking and writing. The conjunctive endings of predicates was used most when connecting clauses. The use of final endings of predicates was statistically significant in writing more so than in speaking, whereas changing endings was used more in speaking. In the ratio of grammatical errors, subject omission errors were the most frequent and object omission errors and conjunctive ending errors were frequent in speaking and writing. Conclusions: AS-units can be used to analyze syntactic productivity, complexity, connectivity, and grammatical errors in Korean with omitted subject frequently. It is meaningful to provide information on specific errors in sentence instruction.

This study investigates the potential and limitations of generative artificial intelligence (AI) in Korean language education by evaluating how well current models identify and correct syntactic errors made by learners. Multilingual models (ChatGPT, Claude) and Korean-specific models (EXAONE, QwenKO) were tested using a learner corpus. The results revealed notable performance differences. EXAONE showed high accuracy in error detection, highlighting the strengths of Korean-focused models, but its performance in error correction was limited and inconsistent across error types. Claude and GPT demonstrated more balanced results, particularly in rewriting corrected sentences, while QwenKO showed generally low accuracy. Overall, Korean-specific models excelled in identifying errors, whereas larger multilingual models performed better in sentence-level correction. All models struggled with errors requiring contextual or pragmatic understanding, suggesting that current AI systems still lack discourse-level competence. This study provides empirical insight into the capabilities of generative AI for syntactic error processing and offers pedagogical considerations for its informed application in Korean language instruction.