Earticle

This paper aims at investigating the cross-linguistic generality and intra-language specificity in processing morphosyntactic and semantic on-line processing in Korean often called head-final language using ERPs(Event-related Poentials). Before the current experiment, sub-classification for errors was firstly conducted into sentence-final-error(syntactic/semantic-final-error) and sentence-mid-error (syntactic/ semantic-mid-error) according as the same type of errors can be arisen in any loci of complement or predicate in Korean. Our experimental results shows that, as expected, P600(positive peak at about 600 ms after the stimulus onset) and N400(negative peak at about 400 ms after the stimulus onset) were observed in each error type respectively. However, specifically the centered activation area in sentence-final-error is moved into more posterior site from anterior position which is different from the topographic distribution occurring at sentence-mid -error. This finding suggests that the error-occurring position can be main factor having effect on error processing area as well as error types indexed with ERP characteristics such as latency or amplitude of the current component.

이 연구는 최근 급증하고 있는 베트남인 한국어 학습자들을 좀 더 효율적으로 교육할 수 있는 방안을 탐구 하기 위해 베트남인 한국어 학습자들의 쓰기 오류를 분석하여 음소 체계와 어휘 의미 측면에서의 오류 양상을 살펴보았다. 이를 위해 베트남인 한국어 학습자 78명의 쓰기 과제를 분석해서 자음, 모음, 어휘 의미 측면에서 의 오류를 살펴보고 유형별 교육 방안을 제안하였다. 이 연구의 결과 음소와 관련해서는 경구개파찰음, 평음과 경음, 유음의 받침 표기, 이중모음 ‘ㅟ’ 등의 오류가 많이 나타났으며, 어휘 의미 관련해서 모국어인 베트남 어의 다양한 어휘 의미를 한국어 사용 상황에 맞는 어휘와 구별하지 못해 오류를 많이 범하였다. 베트남인 한 국어 학습자들의 효율적 한국어 교육을 위해서는 오류 양상과 원인을 꾸준히 수집하고 분석해나가야 한다. 또 한 교육현장에서는 발음교육과 어휘 표기 교육, 작문 교육과 어휘 사용 교육 등을 통합적으로 해나가는 방안에 대해 더욱 더 고민해나가야 할 것이다.

This study examined the writing errors of Vietnamese Korean learners in terms of phonemic system and vocabular meaning to explore ways to educate Vietnamese Korean learners more efficiently. To this end, the writing tasks of 78 Vietnamese Korean learners were analyzed, errors in terms of consonant, vowel, and vocabular meaning were examined, and an educational plan for each type was proposed. As a result of this study, many errors such as affricate, plain and hard sound, the use of 'liquid' in a consonant, and double vowel 'ㅟ' appeared in relation to phoneme, and many errors were made in relation to vocabular meaning because various vocabular meanings of the native Vietnamese language could not be distinguished from those that fit the Korean language use situation. For effective Korean language education for Vietnamese Korean learners, it is necessary to continuously collect and analyze the pattern and cause of errors. In addition, it is necessary to think more about ways to integrate pronunciation education, vocabulary notation education, writing education, and vocabulary use education in the educational field.

영어권에서는 통계적 기반 의미 오류 검사기에 대한 연구가 활발하게 진행되었으나 한국어에서는 자주 나타나는 오류를 중심으로 오류 단어를 검사할 문맥이나 공기정보를 갖는 규칙 기반의 연구가 이루어져왔다. 본 논문에서는 표제어나 사전 기반 범주화 정보로 표현되고 있는 기존 문법 검사기의 성능을 개선하는 방안으로 한국어 명사 어휘 의미망인 KorLex를 활용하는 방법을 연구한다. 특히, 용언은 목적어나 주어에 사용되는 명사의 표제어가 의미 제약 조건으로 사용되는 예가 많다. 본 논문에서는 용언의 의미 제약 조건을 명사 표제어 단위에서 명사 의미 그룹(class) 단위로 확대하여 문법 검사기의 의미 오류 검사 규칙의 의미 오류 제약 조건을 일반화 하는 방안을 연구한다.

영어권에서는 통계적 기반 의미 오류 검사기에 대한 연구가 활발하게 진행되었으나 한국어에서는 자주 나타나는 오류를 중심으로 오류 단어를 검사할 문맥이나 공기정보를 갖는 규칙 기반의 연구가 이루어져왔다. 본 논문에서는 표제어나 사전 기반 범주화 정보로 표현되고 있는 기존 문법 검사기의 성능을 개선하는 방안으로 한국어 명사 어휘 의미망인 KorLex를 활용하는 방법을 연구한다. 특히, 용언은 목적어나 주어에 사용되는 명사의 표제어가 의미 제약 조건으로 사용되는 예가 많다. 본 논문에서는 용언의 의미 제약 조건을 명사 표제어 단위에서 명사 의미 그룹(class) 단위로 확대하여 문법 검사기의 의미 오류 검사 규칙의 의미 오류 제약 조건을 일반화 하는 방안을 연구한다.

With video compression standards such as MPEG-4, a transmission error happens in a video-packet basis, rather than in a macroblock basis. In this context, we propose a semantic error prioritization method that determines the size of a video packet based on the importance of its contents. A video packet length is made to be short for an important area such as a facial area in order to reduce the possibility of error accumulation. To facilitate the semantic error prioritization, an efficient hardware algorithm for face tracking is proposed. The increase of hardware complexity is minimal because a motion estimation engine is efficiently re-used for face tracking. Experimental results demonstrate that the facial area is well protected with the proposed scheme.

이 논문은 2018년 ‘탈BL’ 논쟁 이후 대중화에 성공한 드라마 「시맨틱 에러」 (2022)에 주목함으로써, BL 드라마 팬덤의 정동(affect)을 페미니즘 문화 비평의 흐름 속에서 탐구한다. 드라마 「시맨틱 에러」의 정동은 온라인 공간을 바탕으로 만들어졌으며, 대상에 과도하게 몰입하는 현상을 의미하는 ‘과몰입’ 문화와 밀접하게 연동되어 있다. 드라마 「시맨틱 에러」는 원작 소설에 있던 ‘19금’ 요소를 삭제함으로써 ‘12세 이용가’ 콘텐츠로 탈바꿈하였는데, 이러한 개작 과정은 ‘BL의 대중화’를 둘러싼 상반된 분석을 야기할 수 있는 지점이라는 점에서 각별히 짚어볼 만하다. 가령 ‘안전한 판타지로 잘 각색되었다’는 평가는 이 작품을 BL 드라마화의 성공 사례로 보게 한다. 그런데 「시맨틱 에러」를 둘러싼 ‘과몰입’ 현상에 주목해보면, 이 드라마를 둘러싼 정동이 ‘안전한 판타지’라는 규정 바깥에서 활성화되고 있음이 확인된다. 이에 본 연구에서는 드라마 「시맨틱 에러」 팬덤의 문화 실천을 검토하기 위해, ‘과몰입–어펙트(exmersion-affect)’라는 관점을 제안한다. 본 논문은 드라마 「시맨틱 에러」를 둘러싼 정동이 서사를 일방적으로 전달하는 구조(창작자→독자/수용자/감상자)를 통해 발생하는 것이 아님을 주장한다. 이를 관계의 형성 과정을 이분법적으로 보는 시각에서 벗어나 ‘내부–작용(Intra-action)’으로 검토함으로써, 서사를 감상하는 방식이 ‘얽힘·연루(entanglement)’의 과정으로 이뤄지고 있음을 고찰한다. 즉 이 연구는 드라마 「시맨틱 에러」가 ‘탈BL 담론’과 ‘안전한 판타지’ 사이의 긴장 관계에 놓여 있던 텍스트라는 점에 착안하여, ‘과몰입–어펙트’라는 조어를 통해 오늘날 드라마 감상 문화의 특성을 새롭게 설명함으로써 정동 연구의 틀을 다각화하는 것이다.

This study examines the affective dynamics of BL drama fandom through a feminist cultural criticism framework, focusing on the narrative formation of Semantic Error (2022), a Korean drama that gained mainstream popularity in the wake of the 2018 “Tal-BL” (Leaving BL) discourse in Korea. The affect surrounding Semantic Error emerged primarily in online spaces and was closely tied to practices of intense emotional immersion. While the drama’s success has often been attributed to its status as a “safe fantasy, ” this study situates such interpretations within the broader historical context shaped by the Tal-BL discourse. The adaptation of Semantic Error?which removed the original novel’s explicit sexual content and was reproduced as a ‘suitable-for-age-12’ drama?prompted conflicting explanations regarding its widespread appeal. Yet when attention is directed to the phenomenon of “ over-immersion affect, ” it becomes evident that the drama’s affective force operates beyond the boundaries of the “safe fantasy.” To analyze the cultural practices of Semantic Error’s fandom, this study proposes the concept of exmersion-affect, which highlights how affect is continually regenerated rather than simply delivered in a one-way narrative structure? from creator to reader/viewer?. Grounded in the new materialist feminist idea of entanglement, we argue that affect emerges through non-linear, ongoing relational processes rather than through a linear model of narrative transmission. By positioning Semantic Error within the tension between the Tal-BL discourse and the appeal of safe fantasy, this study diversifies the analytical frameworks of BL drama affect studies and suggests a non-linear, entangled understanding of text?audience relations through the lens of exmersion-affect.

최근 BL의 급부상으로 인하여 OTT 플랫폼을 기반으로 한 드라마화가 이어지고있다. 범용성을 추구하는 지상파 드라마와 달리 OTT 드라마는 다양한 취향을 반영하는 전략으로 확고한 소수의 향유층을 만족시키면서 정체성을 확보해왔다. 문제는 소수 취향의 개인적인 장르인 BL이 가장 대중적인 장르 중 하나인 드라마로 전환될 때 은밀함과 대중성이라는 두 요소 사이의 간극을 어떻게 극복하느냐이다. 웹소설 원작의 <시맨틱 에러>는 드라마에서 수용 가능한 BL 요소의 수위 조절과 OTT 플랫폼의 향유 특성에 따른 BL 취향의 극대화 사이에서 이율배반적인 균형점을 찾음으로써 시장성을 확보했다. 기존 BL 드라마와 차별화한 <시맨틱 에러>의 드라마화 전환 전략은 크게 세 가지로 볼 수 있다. 1) 로맨스 컨벤션의 전면화, 2) BL 정체성의 유연한 유지, 3) 영상과 음악의 전략적 활용이다. 첫째, BL 요소를 연성화하고 로맨스 컨벤션을 전면화함으로써 BL 장르에 익숙하지 않은 새로운 향유자의 유입을 용이하게 하였다. 둘째, 웹소설을 거쳐 웹툰으로 재현한 이미지와 배우들의 싱크로율을 높이고 대사를 그대로 활용함으로써 원작과의 연결성을 유지했다. 웹소설과 달리 드라마는 성애 묘사와 같은 BL 요소를 구현하는 데 한계가 있지만 적절한 수준에서 기존 향유자들의 상상 이미지를 강화해 BL의 정체성을 유지하고 즐길거리를 제공했다. 셋째, 주요 캐릭터의 감정을 대변하는 색채와 OST의 효과적 개입은 서사와 어우러져 드라마 자체의 소구 요소로서 적극 활용되었다.

The recent rise of BL has led to a wave of dramas based on OTT platforms. Unlike terrestrial dramas that strive for universality, OTT dramas have secured their identity by satisfying an entrenched minority audience with strategies that reflect diverse tastes. The question is how to overcome the gap between the two elements of secrecy and publicity when BL, a personal genre with a minority taste, is transformed into drama, one of the most popular genres. <Semantic Error>, based on a web novel, secured its marketability by finding a balance between controlling the level of acceptable BL elements in the drama and maximizing BL tastes according to the enjoyment characteristics of OTT platforms. Differentiated from existing BL dramas, <Semantic Error>'s dramatization transformation strategy can be seen in three main ways. They are: 1) the totalization of romance conventions, 2) the flexible maintenance of the original content identity, and 3) the strategic use of video and music. First, by softening BL elements and front-loading romance conventions, it facilitated the influx of new audiences unfamiliar with the BL genre. Second, the drama maintained its connection to the original by increasing the synchronization rate of the images and actors reproduced in the webtoon through the web novel and utilizing the same dialogues. Unlike web novels, dramas have limitations in implementing BL elements such as depictions of sexuality, but at the right level, they reinforced the imaginary images of existing users to maintain the identity of BL and provide entertainment. Third, the colors that represent the emotions of the main characters and the effective intervention of the OST were actively utilized as appeal elements of the drama itself in combination with the narrative.

Voice input is often required in many new application environments such as telephone-based information retrieval, car navigation systems, and user-friendly interfaces, but the low success rate of speech recognition makes it difficult to extend its application to new fields. Popular approaches to increase the accuracy of the recognition rate have been researched by post-processing of the recognition results, but previous approaches were mainly lexical-oriented ones in post error correction. We suggest a new semantic-oriented approach to correct both semantic level and lexical errors, which is also more accurate for especially domain-specific speech error correction. Through extensive experiments using a speech-driven in-vehicle telematics information application, we demonstrate the superior performance of our approach and some advantages over previous lexical-oriented approaches.

배경 및 목적: 초등 3-4학년은 읽기유창성이 완성되는 시기로 읽기의 최종목표인 읽기이해에 도달하기 전에 읽기유창성이 선행되어야 한다. 따라서 언어문제를 동반한 언어학습장애 아동들의 읽기유창성 능력을 살펴보고 오류 특성을 구체적으로 알아보았다. 방법: 초등학교 3-4학년에 재학 중인 언어학습장애 아동 17명, 학년과 지역을 일치시킨 일반아동 17명, 총 34명으로 하였다. 읽기유창성은 이야기글 2개로 살펴보았으며 분당 정확하게 읽은 음절수, 정확률, 속도로 살펴보았고 오류는 의미/음운 오류로 나누어 오류수, 오류 유형을 살펴보았다. 결과: 언어학습장애 아동의 읽기유창성을 살펴본 결과 분당 정확하게 읽은 음절수, 정확률, 속도가 일반아동보다 유의미하게 낮았다. 또한 전체, 의미/음운 오류수 모두 유의미하게 높았다. 오류 유형을 살펴보자면 의미오류는 대치가 가장 많이 나타났으며 일반아동은 자기수정, 반복, 생략, 첨가 순으로 언어학습장애 아동은 반복, 생략, 첨가, 자기수정 순으로 나타났다. 음운오류는 반복, 대치, 자기수정이 공통적으로 나타났고 일반아동은 첨가, 생략 순으로 언어학습장애 아동은 생략, 첨가 순으로 나타났다. 논의 및 결론: 본 연구에서는 초등 3-4학년 언어학습장애 아동의 읽기유창성과 의미/음운 오류 특성에 대하여 확인하였다. 언어와 읽기에 어려움을 보이는 언어학습장애 아동은 글을 자동적으로 읽는 데 어려움을 보였으며 의미/음운 오류의 특성이 두드러지는 것으로 나타나 평가와 중재에서 이를 고려하여 읽기유창성을 확인해야 할 것이다.

Objectives: The time period between grades 3 and 4 is the period when reading fluency (RF) is achieved. RF should be achieved before reaching the final goal of reading, and reading comprehension. Therefore we investigated the RF of children with language learning disabilities (LLD) with language problems. And investigated the characteristics of reading error in detail. Methods: Seventeen children LLD in the 3rd and 4th grades of elementary school, 17 typically developing (TD) children who matched in grade level and were from the same area, for a total of 34 children. RF was examined through reading two stories, and the number of syllables correctly read per minute, accuracy, and speed were examined. The errors were classified into semantic and phonological errors, and the number of errors and error types were examined. Results: It was found that the number of syllables correctly read per minute, accuracy, and speed of reading by children with LLD was significantly lower than that of TD children. Also, the total of both semantic and phonological errors was significantly higher. Substitution was the most frequent type of semantic error. The most common types of semantic errors exhibited by TD children were self-correction, repetition, omission, and addition respectively; whereas the errors exhibited by children with LLD appeared in the order of repetition, omission, addition, and self-correction. Phonological errors were common in the order of repetition, substitution, and self-correction. For TD children addition was more common than omission, and for children with LLD omission was more common than addition. Conclusion: In this study, we examined the RF characteristics of 3-4 graders with LLD and identified semantic and phonological error characteristics. Children with LLD who have difficulties with language and reading have difficulty reading text automatically. In particular, the characteristic of semantic and phonological errors was prominent. Therefore, it is necessary to consider RF in the course of assessment and intervention.

This research points out the problems with some semantic interpretation shown in the first volume of the Great Chinese-Korean Dictionary by selecting items of three expressions. The author divided this study into three parts. First, error analysis of semantic interpretation of 'Il-Ssang-Yang-Ho'(一雙兩好), second, 'Sam-Moo'(三武), and third, 'Bul-Seong-In-Sa'(不省人事) and 'In-Sa-Bul-Seong'(人事不省). This study is briefly summarized as follows. First, this research examines the errors in the semantic interpretation of 'Il-Ssang-Yang-Ho' and suggests some modification by reviewing a variety of literature such as 'Yoo-Rim-Oe-Sa'(儒林外史), 'Cheong-Pa-Jab-Ji'(清波雜志) 'Go-San-Ang-Ji'(高山仰止), 'Seo-Ho-Yi-Jip'(西湖二集) and 'Gi-Mae-Hwa-Gui-Ryo-Suh-Gak'(寄梅花鬼鬧西閣). The findings of the study suggest that the meanings of 'Il-Ssang-Yang-Ho' can be interpreted as 'matching well each other', 'both look good together' and 'both are good together', but the meaning like 'the couple lives in conjugal harmony' is inappropriate and thus should be deleted. In addition, the reasons of the erroneous semantic interpretation of the item, 'Il-Ssang-Yang-Ho' shown in the Great Chinese-Korean Dictionary are revealed by comparing and contrasting the items presented in various dictionaries including 'Dae-Han-Hwa Dictionary'(大漢和辞典), 'Joong-Mun Great Dictionary'(中文大辭典), and 'Han-Uh Great Dictionary'(漢語大詞典). Second, through the examination of historical facts, this study finds out that 'Sam-Moo' (三武) includes 'Tae-moo-je, Moo-je, and Moo-jong' but does not include 'Do-moo-je.' Furthermore, this study discusses the inappropriateness of the expression 'oppression' shown in the item, 'Sam-Moo' by comparing it with other great dictionaries. In particular, this study argues that one of the item expressions, 'annihilating buddhists' does not represent objective historical views and thus introduces Ryu-ang's argument that it is appropriate to interpret the expression as 'prohibiting buddhists.' Third, this study finds out that the meanings between the two words, 'Bul-Seong-In-Sa' and 'In-Sa-Bul-Seong' in reverse order are the same and thus it discusses the problems with the lack and wrong usage of examples of the item in the dictionary as well as the discrepancy between a word and its word class in semantic interpretation. Moreover, it points out the lack of clarity of the interpretation, 'not knowing the proper manner as a human being' and thus suggests that the meaning should be modified to 'being ignorant of the world as a human being'.

“Jianyuju”(‘N1 + V1 + N2 + V2’) is a predicate type that arouses people’s constant discussion. In Modern Chinese “Jianyuju” means a sentence with its object plays as subject of the second verb simultaneously. The first verb in a “Jianyuju” has meaning of incurring and results in appearance of the second verb. This paper intends to have a semantic feature analysis and grammatical error pattern of “Jianyuju(兼語句)” in Modern Chinese. In respect of syntax, ‘V1’ of this construct is a combination of object and subject, in respect of semantic ‘N1’ is the conductor of ‘V1’, ‘N2’ is the receiver of ‘V1’ and conductor of ‘V2’, and there is a controlling relation between ‘N1’ and ‘N2’. This paper focus on [-致使] semantic feature and compare with “Shi Jianyuju(使兼語句)” and “Ling Jianyuju(令兼語句)”. After analyzing the errors of students, the essay classifies the errors into four types: lack of elements, redundant adding of elements, errors of elements, errors of the structure. Since the main problems concentrating on the usage of the first type of “Jianyuju” structure, the essay analyzes the errors of each type and gives some possible reasons for them.

The purpose of this study is to find out educational method of Equality Particles, {mankeum/cheorum/kachi} for Korean learners. In Korean linguistics, these morphemes can not be explained clear. Moreover in Korean language education, only some limited meaning of {mankeum/cheorum/kachi} is introduced at the level 2, and then their actual usage is showed many errors. To identify the semantic similarities and differences of {mankeum/cheorum/kachi}, it is used Sejong 21'th century's Project corpus and Korean learner's error corpus. As a result, {mankeum} is used as a bound noun and a particle, {cheorum} as a particle and {kachi} can be used as a particle and a adverb. In this case investigations, {mankeum}'s meaning is extended, "the difference between a certain amount of> different levels of degree> degree of similarity of> equivalence" In contrast, {cheorum} has the meaning of 'degree of similarity' and 'equivalence' between two comparative objects. Lastly, {kachi} is derived from {katta} as an adjective and has the meaning 'equivalence' like {cheorum}. (SoonChunhyang University)

The focus of this paper will be to investigate the properties of the Korean reflexive pronouns {caki, sasin} and to analyze the errors made using these reflexive pronouns through the corpus of Korean learners. These two words {caki} and {sasin} are the traditional reflexive pronouns and are different from the Chinese words {自己}. The typical properties are like this. First, in general there are no differences in terms of usage between these two words {caki, sasin}. Second, only {sasin} can be used when wanting to emphasizing the usage. Third, in the reflexive usage, {caki} has many constraints. For example, one constraint that {caki} has is that you cannot use it when referring to first and second person. In contrast, {sasin} has no constraints. Also {sasin} can be used as the antecedent (honorable person) and in terms of frequency is also used more often than {caki}. However Korean learners use {caki} more than{sasin} because of the Chinese word {自己} as well as overgeneralization.

Korean reflexive pronoun, {caki}, {sasin}, generic usage, intensifying usage, inference, overgeneralization

음성인식기를 거친 결과는 오류를 포함할 수 있으며 이를 다른 자연어처리 응용에 이용하기 위해서는 오류의 검출과 수정과정이 필수적이다. 음성인식 오류 후처리는 그 성격상 문자인식 후처리와는 다른 접근 방법을 필요로 하며, 본 인구에서는 잡음환경을 제외한 특정 도메인에 국한된 음성발화 상황에 초점을 맞추고자 한다. 후처리 방법에 있어서는 통계적 접근과 패턴매칭에 의한 접근 방법이 있으며, 본 연구에서는 특정 도메인에서 사용되는 어휘의 의미정보를 포함하는 패턴을 자동으로 생성시켜 이에 의한 오류 검출 및 수정 방안을 제안한다. 본 실험에 사용된 도메인은 차량정보센터용 음성정보 제공 시나리오이며 상용 음성인식기를 후처리를 위한 개발 툴로 사용하였다.

학생들은 문제해결과정에서 체계적이고 일관적인 오류를 범한다. 오류 분석은 학습의 어려움을 해결하고, 학생들의 오류를 수정하기 위한 대표적인 전략이다. 언어네트워크는 개념 간 연결을 분석하여, 인지구조를 분석하는데 주로 활용된다. 연구의 목적은 언어네트워크를 활용하여 학생들의 오류 유형을 분석하는 것이다. 이를 위해, 고등학생 18명을 대상으로 분석하였다. 검사지는 발성사고법을 통해 세포분열에 대한 문제해결과정을 분석하기 쉽도록 구성되었다. 수집된 자료는 언어네트워크로 구조화되었다. 과학적인 개념을 가진 학생들과 오류를 가진 학생들의 언어네트워크를 비교하여 오류 유형을 분석하였다. 분석한 결과 오류 유형은 4가지로 분류되었다. 오류의 원인은 개념 간 연결에서 개념의 결여나 연결의 결여로 나타나고 있었다. 또한 하위 개념의 이해는 상위 개념을 이해하기 위한 결정적인 역할을 하고 있었다. 이를 통해 개념과 연결의 결여에 의한 오류의 원인이 학생들에게 적절한 피드백에 대한 시사점을 제공하고 있다. 추후연구에서는 오류를 수정하기 위한 효과적인 교수학습 전략을 제시되길 기대한다.

Students make systematic and consistent errors to solve the problems. The error analysis is the representative educational method used to solve learning difficulties and correct students’ errors. Semantic Network Analysis(SNA) has been used for analyzing the relationships among concepts and configuring the network of cognitive structure. The purpose of this study is to analyze students’ error types using SNA. To this end, 18 high school students were investigated. The questionnaire was developed to make the student observation process of problem solving for cell division questions, based in think aloud. The collected answers from the students were visualized SNA. Comparing the semantic networks of students in terms of scientific concepts to students in making the error corresponding to the types of errors, the results of error types were driven. The results were identified four representative types of errors. There was found lack of concepts and connections among concepts. And a subordinate concept was the critical factor to understand a superordinate concept. Therefore, it is important to verify the sources of errors originated from deficiency of the concepts and connections to give an accurate feedback to students. For the future study, it is expected that more useful instruction strategy to correct errors should be established.

This paper analyzes the error types and characteristics of Korean learners through the error analysis of the auxiliary verb 'Hui(会)' in Mandarin, and provides educational suggestions according to the situation of Korean learners. First, through the analysis of previous studies, the traditional and newly added usages of the auxiliary verb 'Hui(会)' are examined, and the semantic functions of 'Hui(会)' that should be dealt with in educational grammar are identified as 'ability', 'belief', 'pattern', 'promise'. Next, as a result of analyzing the error sentences of Korean learners, the error rate appeared in the order of omission error, substitution error, and abuse error, and the characteristics of each error type could also be confirmed. Finally, as a result of examining the causes of errors in the use of 'Hui(会)' by Korean learners, it was found that it was mainly due to educational vacuum, misleading, and interference in their mother tongue. Also, focusing on the major problems revealed through error analysis of Korean learners, descriptions of semantic function and usage, and solutions to be noted in the teaching process were presented.

배경 및 목적: 본 연구는 초등 저학년 발달성 난독 아동을 대상으로 의미지식을 고려한 문단글 읽기유창성 특성을 알아보고 오류분석을 하고자 하였다. 방법: 초등 저학년 발달성 난독 아동 30명을 수용 어휘력 검사 결과에 따라 어휘정상 난독 집단과 어휘부진 난독 집단으로 나누었으며, 두 집단은 학년, 지능, 해독수준이 일치하였고, KOLRA의 해독 수행력 또한 일치하였다. 본 연구에서는 문단글 읽기유창성의 정확도와 속도를 비교하고 의미지식을 고려하여 오단서 분석(MA)과 언어패턴 분석(LPA)으로 오류빈도를 비교하였다. 결과: 어휘부진 난독 집단은 오단서 분석(MA)에서 대치 및 수정 오류가 유의하게 높은 빈도로 나타났다. 또한 언어패턴 분석(LPA)에서는 자소-음소 불일치형 오류에서 유의하게 높은 오류빈도를 보였다. 논의 및 결론: 자동성 이론에 따르면, 유창한 읽기를 위해서는 의미적 단위로 글을 이해하며 읽어내야 한다. 해독 수준이 동일한 난독 집단 내에서도 어휘부진 집단의 경우 의미지식의 제한으로 문단글읽기유창성에서 더 많은 오류를 보였다. 한국의 초등 1-2학년 교과서 내 문단글에는 문법형태소가 결합된 자소-음소 불일치형 어절이 다수 포함되어 있다. 따라서 음절 단위뿐만 아니라 3중 단어 구성론을 고려하여 읽기 유창성 오류를 살펴보아야 한다.

Objectives: The purpose of this study was to investigate the characteristics of text reading fluency error analysis of Korean 1st and 2nd graders with developmental dyslexia considering semantic knowledge. Methods: Thirty developmental dyslexic children were divided into two groups of 15 children, one group with poor vocabulary and the other not. Both groups had similar word decoding abilities as determined by Korean Language based Reading Assessment (Pae, Kim, Yoon, & Jang, 2015). The children had normal intelligence and difficulties in reading. The task was the text taken from Korean Language based Reading Assessment. Considering semantic knowledge, the performance of text reading fluency and error types, the results of miscue analysis (MA) and Linguistic pattern analysis (LPA) were compared. Results: The group with poor vocabulary exhibited a lower frequency of errors in substitution and correction by miscue analysis. They also showed a lower error frequency in phoneme-grapheme non-correspondent errors by linguistic pattern analysis. Conclusion: According to the automaticity theory, we need to read texts with understanding on a semantic basis. Even with the same level of decoding, there were more errors in text reading fluency due to limited semantic knowledge in the group with poor vocabulary. Paragraphs in textbooks in the 1st and 2nd grades in Korea contain several phoneme-grapheme non-correspondent ‘eojeols’ that combine grammatical morpheme graphemes. Therefore, in addition to reading errors of grapho-phonology in syllable units, reading errors considering triple word-form bases should be examined.