Earticle

A parasitic gap, an illicit gap, is acceptable when it is dependent on a real gap and the parasitic gap construction containing two gaps is difficult to understand due to the structural complexity. Moreover, these constructions are very rare in Korean learners' input because they are not learned in L2 classes. This study aims to investigate Korean learners' syntactic knowledge and processing of parasitic gaps inside subject islands using an offline scalar judgement task and an online self-paced reading task. The offline task showed the sensitivity of the licensing constraint. The online task indicated that Korean learners computed the internal representation of the parasitic gap by applying the parsing principle and the strategy during reading the sentence, which was similar to native speakers' processing. Therefore, these results suggest that there exists the logical problem in the acquisition of parasitic gap constructions despite the poverty of the stimulus.

Jun, Jong Sup. 2011. Children's Syntactic Knowledge of English Derivational Morphology: Evidence from Early Suffixes. Korean Journal of Linguistics, 36-4, 1053-1073. Previous studies on the acquisition of English derivational morphology suggest that children acquire the syntactic knowledge of derivational morphemes after Grade 4 in elementary school; that is, children do not learn the parts of speech (POS) change between a root and its affixed form (e.g. drive  driver) until they become nine years old. On the contrary, the present study explores preschool children's knowledge of English derivational morphology, and shows that children have some syntactic knowledge at age 3-5. The evidence comes from English-speaking children's use of so called early suffixes, i.e. -er/or, -ie, and -y. We extracted preschool children's derivatives from the HSLLD corpus (Dickinson & Tabors, 2001) in the CHILDES database (MacWhinney & Snow, 1985). We then cross-classified the words in terms of children's age, the semantic transparency, the POS change, and the suffix type. The log-linear regression analysis shows that preschool children use derivational suffixes following the generalization that is sensitive to the POS of the derived words. This finding supports the view that children have the implicit syntactic knowledge of derivational morphology during the preschool years. (Hankuk University of Foreign Studies)

인공지능, 딥러닝, 머신러닝 등이 괄목할만한 발전을 이루면서 2016년경부터 100여개의 언어를 비롯하여 가장 보편적으로 사용되어 온 Google Translate (구글기계번역기)는 자연언어처리(NLP) 분야와 외국어 학습 등 언어활용 분야에 독보적인 역할을 하고 있다. 본 논문은 구글기계번역기, Google Translate에 있어서, 영어모국어화자가 가진 통사문법적-범주적 지식의 중요성과 그 영향력에 대해 살펴보고자 한다. Jackendoff (1999)는 맹인을 위한 독서기계 (Reading Machine)등을 구축하려면 통사구조적 지식과 문법적 분해력(parsing)이 매우 중요하고, 적어도 현재의 컴 퓨터는 엄청난 발전을 이루기는 하였으나, 인간의 두뇌를 따라갈 수 없다는 결론을 내렸다. Jackendoff가 논의했던 몇 가지 어휘항목과 통사구조적 중의성을 활용하여, Google Translate 기계발음번역기를 통해 그의 주장을 확인하는 실험을 실시하고, 그 결과를 분석하는 것이 이 논문의 목표이다. 이 연구는 Jackendoff의 주장처럼 L1 화자가 내재화 한 통사문법적, 범주-구조적 지식은 NLP, 혹은 “독서기계”등의 구축에서 중요하며, 이는 Chomsky (1986, 2005)등에 서 논의된 내재적언어 (I-language)의 핵심이라는 점을 시사한다.

This paper highlights the influence and the importance of the syntactic-grammatical knowledge on “the reading machine”, appeared in Jackendoff (1999). Due to the lack of the detailed testing and implementation in his research, this paper tests an extensive data array using a component of Google Translate, currently available freely and most widely on the internet. Although outdated, Jackendoff’s paper, “Why can’t Computers use English?”, argues that syntactic-grammatical knowledge plays a key role in the outputs of computers and computer-based reading machines. The current research has implemented some testings of his thought-provoking examples, in order to find out whether Google Translate can handle the same problems after two decades or so. As a result, it is argued that in the field of NLP, I-language in the sense of Chomsky (1986, 1995 etc) is real and the syntactic, grammatical, and categorial knowledge is essential in the faculty of language. Therefore, it is reassured in this paper that when it comes to human language, even the most advanced “machine” is still no match for human faculty of language, the syntactic-grammatical knowledge.

The purpose of this study was to investigate: 1) the relationship between L2 vocabulary/ L2 syntactic knowledge and L2 reading comprehension and 2) the relative role of vocabulary and syntactic knowledge in L2 reading comprehension. L2 reading comprehension was examined as literal, inferential, and evaluative comprehension as well as general reading comprehension. For this study, 47 Korean software high school EFL learners were administered a vocabulary knowledge test, a syntactic knowledge test, and a reading comprehension test. Results revealed that vocabulary as well as syntactic knowledge significantly correlated to general reading comprehension; however, vocabulary knowledge was correlated to literal and inferential comprehension, and syntactic knowledge to inferential and evaluative comprehension. For vocabulary knowledge, all of the frequency bands except 2K and 4K word families of the Vocabulary Size Test (VST) (Nation & Beglar, 2007) significantly correlated to general reading comprehension and inferential comprehension; however, low-frequency vocabulary showed the connection with literal comprehension, and high-frequency vocabulary with evaluative comprehension. Additionally, only syntactic knowledge played a crucial role in general reading comprehension and inferential comprehension, after controlling for other variables. Vocabulary knowledge was not found to be a significant predictor of reading comprehension. The findings imply the need for teaching syntactic knowledge as well as a broad range of vocabulary, in a Korean EFL reading comprehension class.

Jun, Jong Sup. 2011. Children's Syntactic Knowledge of English Derivational Morphology: Evidence from Early Suffixes. Korean Journal of Linguistics, 36-4, 1053-1073. Previous studies on the acquisition of English derivational morphology suggest that children acquire the syntactic knowledge of derivational morphemes after Grade 4 in elementary school; that is, children do not learn the parts of speech (POS) change between a root and its affixed form (e.g. drive ? driver) until they become nine years old. On the contrary, the present study explores preschool children's knowledge of English derivational morphology, and shows that children have some syntactic knowledge at age 3-5. The evidence comes from English-speaking children's use of so called early suffixes, i.e. -er/or, -ie, and -y. We extracted preschool children's derivatives from the HSLLD corpus (Dickinson & Tabors, 2001) in the CHILDES database (MacWhinney & Snow, 1985). We then cross-classified the words in terms of children's age, the semantic transparency, the POS change, and the suffix type. The log-linear regression analysis shows that preschool children use derivational suffixes following the generalization that is sensitive to the POS of the derived words. This finding supports the view that children have the implicit syntactic knowledge of derivational morphology during the preschool years. (Hankuk University of Foreign Studies)

Two different approaches have been taken in an attempt to explain DPs in early child language: Some argue that children's internal grammar is not different from that of adults and others insist that any determiner in early child language lacks functional categories or the features in adult language. This study investigates these assumptions with respect to one of DPs, the articles. In particular, it examines an article usage that has a bearing on this question, articles in formulae, in the data of twenty-one English-acquiring children below MLU 3.0. The main finding of the study is that the majority of the children's utterances containing articles in the data cannot be considered formulaic. Although some previous research has found early article use to be limited to rote-learned sequences, this study concludes that articles in early English child language comprise an adult-like functional category. (Gyeongin National University of Education)

The purpose of the study was to explore the relative contribution of both vocabulary knowledge and grammatical knowledge as predictors of reading comprehension performance. To accomplish this, a total of 337 college students voluntarily participated in the study, all of whom were then taking an English reading course at two different universities. The measurement instruments used for the study were based on three sections: vocabulary, grammar, and reading comprehension. Each section consisted of two different tasks. For all the tasks except the recall task, the scoring procedure proved easy and expedient, with 1 point being assigned for correct items, and 0 point for incorrect ones. To obtain valid measures for the recall task, the learners were scored using Johnson’s (1970) unweighted pausal unit analysis. For the data analysis, structural equation modeling (SEM) was conducted to evaluate the role of syntactic knowledge and vocabulary knowledge where it predicts the L2 reading ability of the learners. The SEM analysis indicated that vocabulary knowledge and grammatical knowledge variables made contributions to the prediction of reading comprehension performance, with vocabulary knowledge outperforming grammar in its predictive power. Implications and limitations of the study are also discussed in this study.

본 연구는 초등저학년 읽기부진 아동의 구문능력 특성을 구문지식과 구문인식능력으로 나누어 살펴보고자 하였으며, 읽기능력과 구문능력이 관련이 있는지를 살펴보았다. 초등학교 1-3학년의 읽기부진아동 36명, 일반아동 36명으로 총 72명이 참여하였다. 읽기부진아동은 한국어읽기검사 결과 낱말해독과 읽기이해검사의 합산점수인 읽기지수 I에서 표준점수가 90점 이하인 아동을 대상으로 하였다. 두 집단 간 구문과제 유형에 따른 차이를 분석하였으며, 읽기능력과 구문능력과의 관계를 보기 위해 피어슨 상관분석을 실시하였다. 연구 결과, 읽기부진아동집단은 일반아동집단보다 구문과제 모두에서 낮은 수행력을 보이는 것으로 나타났으며, 두 집단 모두 구문인식과제에서 더 어려움을 보이는 것으로 나타났다. 하지만 상호작용효과가 나타나지 않아 읽기부진아동이 구문과제 유형에 따라 수행력에 영향을 받지는 않는 것으로 나타났다. 읽기부진아동과 일반아동 모두에서 읽기능력과 구문능력이 상관관계가 있는 것으로 나타났다. 본 연구 결과, 초등저학년 읽기부진아동은 또래 일반아동과 비교하였을 때 낮은 구문능력을 보이며, 이러한 낮은 구문능력이 읽기어려움에 영향을 준다고 해석할 수 있다. 이후 읽기부진집단을 낱말해독부진아동, 읽기이해부진아동의 개별적인 집단으로 하여 구문특성을 살펴보는 연구가 필요할 것이다.

The purpose of this study was to examine the syntactic knowledge and syntactic awareness for early school-aged children with reading difficulty and whether syntactic ability and reading skills were correlated with each other. The participants were 36 children with reading difficulty and 36 their grade-matched typically developing children who are grades 1 to 3. The results showed that children with reading difficulty performed significantly lower in both syntactic knowledge and syntactic awareness task compared to typically developing children. As well there was not an interaction effect between group and syntactic tasks. There were all significant correlations between two groups, except between decoding and syntactic knowledge for children with reading difficulty. The results of this study suggest that poor syntactic ability may impede the reading development for children with reading difficulty. Further studies should be carried out to explore syntactic skills for children with decoding problems and children with specific reading comprehension problems separately.

本研究探讨了朝鲜族儿童的汉语语素意识、词汇知识和句法知识对阅读理解中的影响。研究结果表明，汉语语素意识和句法知识都能显著预测汉语阅读理解的准确性，汉语语素意识比句法知识具有更强的预测作用，并且在不同的年级预测效果稳定，而词汇知识的预测效果则不稳定。本研究为理解汉语语素意识、词汇知识、句法知识与汉语阅读理解之间的关系提供了新的证据。

This study explored the functions of morphological awareness, vocabulary knowledge and syntactic knowledge for Korean-Chinese children in their Chinese reading comprehension. The results showed that both morphological awareness and syntactic knowledge can significantly predict the accuracy of Chinese reading comprehension, and that morphological awareness has a stronger predictive effect than syntactic knowledge, and its effect is stable at different grades, but vocabulary knowledge does not. This study provides new evidence in Chinese reading to understand the relationship between morphological awareness, vocabulary knowledge, syntactic knowledge and Chinese reading comprehension.

The purpose of the present study was to examine the role of semantic and syntactic knowledge in the first language acquisition of Korean classifiers. The elicited classifiers production test(EPT) was conducted to 105 children aged from 2 to 7. EPT consisted of 16 classifiers and two items for each classifier. 32 items were divided into 2 major semantic features: animacy and inanimacy. The semantic features of inanimacy were subcategorized into 3 features such as neutral, shape and function. The results revealed that; 1) children produced the correct structure of classification from the very early age with correct word order of the noun phrase showing early fundamental syntactic knowledge; 2) The earliest response pattern was to respond to all nouns in the same way using a neutral classifier showing no apparent semantic basis for their choice; 3) Children didn't show any preference for animate, shape, or function classifiers.

The objectives of this study are mainly two-fold. First, it is to analyze the cohesion characteristics, vocabulary, and grammar knowledge of the writing tasks, which English learners produced using dictogloss protocols. The second aim is to identify the relative contributions of lexical and grammatical knowledge to the writing products yielded by the participants. For this purpose, 45 English learners at a university completed four different tests and then completed two types of dictoglass tasks with two different topics over four weeks. The collected data were analyzed using a web-based analysis software called Coh-Metrix. The results of the analysis indicated that participants in the group with higher vocabulary scores yielded higher mean scores in all six dependent variables within three categories of fluency, lexical complexity, and grammatical complexity. Of six dependent variables, three measures of Total Word Count, Latent Semantic Analysis, and Average Sentence Length showed statistically significant differences in group comparisons. No statistically significant results, however, was found regarding the category of lexical complexity. With relative contributions of lexical and grammatical knowledge, both vocabulary tests were analyzed as significant factors affecting learners’ task performance. However, only one grammar test - error identification test - was analyzed as a variable that affected task performance. (203 words)

목적 : 형태인식이란 형태소의 구조를 의식적으로 인지하고 그 구조를 파악하거나 조작하는 능력을의미한다. 본 연구는 연결어미 및 파생어의 두 형태인식 유형에 따라 학년집단 간에 어떠한 차이를보이는지 살펴보고자 하였다. 또한 학년집단에 따라 형태인식 유형 각각을 예측하는 변인이 무엇인지밝힘으로써 두 유형의 형태인식이 서로 다른 방식으로 처리되는 과정인지 탐색하고자 하였다. 방법 : 초등학교 1학년에서 4학년 131명을 대상으로 저학년과 중학년 집단으로 나누어 형태인식검사와 어휘, 구문이해, 단어읽기 및 읽기이해 검사를 실시하였다. 그 후 학년과 형태인식 유형에 대해 교차요인설계를 실시하였고, 각 학년집단 별로 상관분석 및 형태인식 검사를 종속변수로 기타변인을 독립변수로설정한 단계별 회귀분석을 실시하였다. 결과 : 연구 결과 학년집단과 형태인식 유형에 따른 주효과와상호작용 효과가 모두 나타나 학년이 올라갈수록 연결어미와 파생어 형태인식이 증가하는 것을 알 수있었다. 단계별 회귀분석 결과 저학년 집단과 중학년 집단 각각에서 연결어미와 파생어 형태인식이서로 가장 큰 설명력을 갖는 것으로 나타났다. 그러나 저학년과 중학년에서 연결어미 형태인식을 가장잘 예측하는 기타변인은 구문이해로, 파생어 형태인식을 예측하는 기타변인은 읽기이해로 서로 다르게나타났다. 결론 : 본 연구결과는 연결어미와 파생어 형태인식이 서로를 가장 잘 설명하는 변인으로서언어의 의식적 조작이라는 메타언어 속성을 공유하지만, 반면 굴절과 파생이라는 서로 다른 실현과정을거치는 것으로 나타났다. 후속연구는 언어 및 읽기장애를 지닌 아동집단을 대상으로 깊이 있게 탐색되어야 할 것이다.

Purpose : Morphological awareness refers the ability to consciously recognize the morphemic structure of words and to reflect on and manipulate the structure. In this study, the morphological awareness task included inflectional forms (connective endings) and derivational forms (derivational affixes). The purpose of this study was to investigate whether two types of morphological awareness, inflectional and derivational, skills are dissociable in its processing. Methods : The participants were 131 first through fourth grade children divided into two grade groups. The two groups performed on tasks of morphological awareness, vocabulary, syntactic comprehension, word reading, and reading comprehension. Stepwise multiple regression was conducted to determine differences between two groups and to find out which variables differently explained the two types of morphological awareness scores in each grade group. Results : First, a statistically significant difference was found between the grade groups and between the two types of morphological awareness tasks through grades. In other words, the ability of manipulating morphemic structures increased over grades. Second, two types of morphological awareness accounted for the unique variances each other which means those two types share common attributes of metalinguistic ability. The results also showed, in both grade groups, syntactic comprehension accounted for the inflectional forms of morphological awareness task and the derivational forms of morphological awareness was predicted by reading comprehension. Conclusions : The results of the study support the hypothesis that inflection and derivation rules should be considered to be different. Future research needs to identify in which ways these processing occurs in various disabled groups.

배경 및 목적: 본 연구는 초등 저학년과 고학년에서 읽기이해를 예측하는 요인을 살펴보았다. 첫째, 저학년과 고학년 간에 낱말해독과 언어능력의 읽기이해 설명력이 차이가 있는지를 살펴보았다. 둘째, 언어능력을 어휘지식과 구문지식으로 나누어 각 언어영역이 학년집단에 따라 다르게 읽기이해를 설명하는지를 살펴보았다. 방법: 초등학교 1-3학년 108명과 4-6학년 68명으로 총 176명의 아동이 참여하였다. 모든 아동은 비구어성 지능지수가 70점 이상인 아동이다. 종속변인으로 읽기이해, 읽기이해를 예측하는 독립변인으로 낱말해독, 수용어휘, 문장따라말하기, 음운인식, 빠른이름대기를 검사하고, 위계적 회귀분석을 실시하였다. 결과: 첫째, 저학년 아동은 낱말해독과 수용어휘, 빠른이름대기가 읽기이해를 설명하는 변인으로 나타났다. 반면에 고학년에서는 문장따라말하기, 음운인식, 빠른이름대기능력이 읽기이해를 설명하는 요인으로 나타났다. 둘째, 학년에 따라 읽기이해를 설명하는 언어영역은 다른 것으로 나타났다. 저학년에서는 의미지식인 수용어휘가 읽기이해를 설명하는 변인이었으며, 고학년에서는 구문지식인 문장따라말하기가 읽기이해를 설명하는 것으로 나타났다. 논의 및 결론: 저학년에서는 낱말해독능력이 읽기이해능력에 중요한 변인이 되는 반면에 학년이 올라갈수록 낱말해독능력보다는 읽은 내용의 의미를 파악할 수 있는 언어능력이 읽기이해에 더 영향을 주는 것으로 나타났다. 또한 저학년에서는 개별적인 어휘에 대한 지식이 읽기이해에 더 영향을 주는 반면에, 고학년에서는 문장 내에서, 그리고 문장과 문장 간의 구문구조를 파악하는 능력이 읽기이해에 더 영향을 주는 요인으로 밝혀졌다. 낱말해독 정확도뿐 아니라 읽기속도가 읽기이해에 영향을 주는지를 살펴보는 후속연구가 필요하다.

Objectives: Little is known about the prediction of reading comprehension for early and later elementary grades in Korea. The purpose of this study is to explore the relative contribution of word decoding and language skills to reading comprehension across grades. Specifically, this study examined whether vocabulary and syntactic knowledge might differentially affect reading comprehension across grades. Methods: One hundred six children in grades 1 to 3 and 68 children in grades 4 to 6 participated in this study. All of the children scored 70 or higher on the nonverbal intelligence test. All participants were tested to evaluate reading comprehension, word decoding, receptive vocabulary, syntactic skills, phonological awareness and rapid naming. Hierarchical multiple regressions were conducted to determine which variables of word decoding and language would contribute to reading comprehension. Results: In the early elementary grades, word decoding, receptive vocabulary, and rapid naming accounted for unique variances in reading comprehension. In the later elementary grades, however, syntactic skills, phonological awareness, and rapid naming were main the differentiating factors in reading comprehension levels. Conclusion: The results of this study were consistent with previous studies which have indicated that word decoding is a stronger predictor of reading comprehension in early elementary grades while language skills are a more important predictor of reading comprehension in late elementary grades. Further studies should be carried out to explore the use of word decoding speed as well as word decoding accuracy to measure their respective contribution to reading comprehension in transparent orthography.

자연어 처리에서 언어에 대한 지식은 전자사전과 문법규칙으로 구성되어 서로 상보적 관계에 있고, 각 어휘에 대한 품사 및 기타 자질-값에 의해 매개된다. 이러한 언어지식을 전통적인 방법에서는 국어자료의 분석에 경험이 많은 언어전문가의 직관에 다분히 의존하여 정의하였고, 말뭉치를 이용한 자동 획득 기법에서는 태그세트를 먼저 설정하고, 이 태그를 원시 말뭉치에 부착하여 태깅된 말뭉치로부터 자동으로 통계적 분석을 통하여 획득한다. 그런데 두가지 접근방법이 가지고 있는 공통적인 문제점은 품사나 자질-값의 정의 및 할당기준, 선악의 평가기준, 튜닝에 대한 적극적 대처 등이 마련되어 있지 않다는 점이다. 이 연구에서는 이러한 문제점의 발생원인을 말뭉치 분석 과정에서 살펴보고, 품사 및 자질-값의 설정과 할당기준을 마련하는 방법론 및 이를 적극적으로 지원하는 도구를 설계한다.

이 글은 활용 개념과 수형도를 근간으로 기술되어 온 한국어 문법 모델에 대한 대안으로 표지 개념과 모빌 구조를 제시하고 이를 바탕으로 개발된 한국어 구문분석기의 특성을 소개하고자 한다. 먼저, 조사와 어미를 독자적인 토사 단위인 표지로 처리함으로써 국부 구조의 통사 범주와 문법적 기능을 명확하고 일관되게 구분할 수 있으며, 모빌 구조는 한국어의(상대적) 자유 어순 현상을 효과적으로 기술할 수 있다. 이에 의거한 문법 모형은 언어학적 지식과 구문분석 엔진 사이의 독립성을 향상시킴으로써 향후 구문분석기의 성능 개선을 보다 용이하게 한다. 이 글에서 소개하는 구문분석기는 언어학자에 의해 구축된 지식을 이용한다는 점에서 지식기반적이라고 할 수 있는데 여기에는 동사의 하위범주화 정보, 첨어 유형정보, 의미정보가 핵심적인 언어 지식으로 이용된다. 모빌 구조에 의한 구문분석은 국부 구조를 단순화함으로써 구문적 중의성을 최소화하며, 의미정보는 주어진 술어의 논항적 자격을 검증하는 기준으로 작용하여 구문적 중의성을 감소시키고 정확한 분석을 가능하게 한다.

This study aims to investigate phonological factors that affect the ordering of binomials in English, based on the Korean EFL learners’ knowledge of binomials. Forty undergraduate students were asked to perform an acceptability judgment task to test their awareness of almost irreversible 40 English binomials. The result showed that they made the most errors in the ordering of Adj+Adj binomials, followed by the types of V+V, Prep+Prep, N+N, and Adv+ Adv. Their lack of knowledge of binomials was proven to be unstatable, when analyzing the scores of the test. It was also found that Korean English learners had insufficient knowledge of phonological constraints that play a key role in the binomial ordering, in which the first word has higher vowels, more front vowels, shorter vowels, fewer initial and more final consonants, less obstruent initial and more obstruent final consonants, fewer syllables, and rhythmic alternation.

Yoon, Hyeonji, Lee, Sang-Ki, & Seong, Sanghui. 2017. Effects of Syntactic Priming on the Development of Grammatical Knowledge by English L2 Learners: A Research Synthesis. Korean Journal of Linguistics, 42-3, 483-510. This study aimed to make a comprehensive overview of the effects of syntactic priming on acquiring L2 grammatical structures. A total of 20 primary studies published in peer-reviewed journals (and 27 independent experiments from them) were retrieved through the systematic and comprehensive literature search. The meta-analysis of the primary studies revealed a medium-sized effect of syntactic priming (d=0.54). That is, assuming that the chance of each competing structure to occur is equal, the average weighted odds ratios of the target structure X to the alternate structure Y increased by 22.75%. Meanwhile, the effects of syntactic priming were found more robust with a greater effect size (d=1.09) when the primed condition was contrasted with the unprimed condition, showing that the chance for the target structure to occur increased by 37.85% under the primed condition. Other detailed findings and their implications are discussed. (Kwangnam Middle School, Korea National University of Education, & Jeonjuseo Middle School)