자음군단순화와 관련된 몇 문제

최은영 어문연구학회 2013 어문연구 Vol.75 No.-

한국어의 어말자음군은 어말이나 자음 앞에서 두 자음이 모두 실현될 수 없다는 특징을 가지고 있다. 이는 종성에 하나의 자음만 실현되는 음절구조제약에 따른 것으로 이를 자음군단순화라 한다. 이러한 자음군단순화의 실현 양상을 체계적이고 간결하게 규칙화하는 것은 어렵다. 그러므로 본고는 자음군단순화의 규칙을 찾기에 앞서 방언에서 실현되는 자음군단순화의 현상의 특징을 먼저 검토하고자 한다. 한국어의 어말자음군은 크게 ㄹ이 선행하는 자음군과 ㄹ이 아닌 다른 자음이 선행하는 자음군으로 나눌 수 있다. 본고는 ㄹ외의 자음이 선행하는 비ㄹ계 자음군을 자료로 하여 방언에서 어떻게 실현되고 있는지를 검토하였다. 방언의 자료는 제주도를 제외한 8개 지역을 대상으로 하되 한국정신문화연구원의『한국방언자료집Ⅰ~Ⅷ』(1987~1995)에 준하여 논의하였다. 본고는 비ㄹ계자음군단순화와 관련하여 자음군의 방언별 실현 양상을 검토하고 그에 관련하여 특이하게 실현되는 양상을 검토하였다. 방언에서는 기존의 연구와는 다르게 실현되는 양상들이 있다. 크게 후행음절과 관련된 것과 자음군단순화가 두 자음이 아닌 제3의 자음으로 실현되는 경우가 있었다. 후행음절과 관련된 것으로는 두 가지 특이한 실현 양상을 검토하였다. 첫째, ‘ㄳ, ㅄ’ 자음군은 자음이 후행하든 모음이 후행하든지 간에 경음으로 실현되어야 한다. 그러나 경음이 아닌 평음으로 실현되는 양상을 확인하였다. 둘째, ‘ㄶ’ 자음군은 자음이 후행하면 격음으로 실현된다. 그러나 평음으로 실현되는 양상이 있었다. 그리고 어말자음군 다음에 후행하는 음절의 자질을 검토하여 어떠한 양상을 포착할 수 있는지를 확인하였다. 마지막으로 모음이 후행하는 경우 연음이 되어 자음군이 모두 실현되어야 하지만 하나의 자음이 탈락하여 재구조화가 실현되기도 한다. 이러한 재구조화가 방언별로 어느 정도 실현되고 있는지를 검토하였다. 이와 같이 자음군단순화와 관련된 몇 가지의 문제점들을 통하여 자음군단순화의 실현에 대한 특질을 규명하는 데 도움을 주고자 한다. The Korean word final consonant cluster has the feature that both of the two consonants in the coda, Korean word final consonant clusters, shouldn't be pronounced in front of the coda or the consonants in in the initial position. This is why a restriction of only one consonant in the word final consonant cluster having sound by the syllable structure constraints. This rule is called the consonant cluster simplification. It is very hard to make the aspect of the consonant cluster simplification a rule systematically and simply. Therefore I intend to examine the aspects of the consonant cluster simplification which can be seen in dialects in this paper before I try to find the rule for the consonant cluster simplification. The Korean Word final consonant cluster is largely divided into the two groups. One is the consonant cluster starting with ‘ㄹ’. The other is the consonant cluster not starting with ‘ㄹ’. I have examined how the aspects of Korean Word final consonant clusters, especially not starting with ‘ㄹ’ in the word final consonant clusters, appear in dialects, Most of the date dealt with in this paper are from『The Collection of Korean Dialects』(1987~1995), by The Academy of Korean Studies, which are collections of dialect data from 9 districts, namely the whole country. But the date of Je-ju were not studied in this paper. I have found two aspects in those dialects, which the former studies had passed by. One is related with the following syllable, and the other is the case which the consonant cluster simplification appears not with one of the final consonants but with a third consonant. Especially, I've examined the two extraordinary aspects related with the following syllable. First, the consonant cluster ‘ㄳ, ㅄ’ should be sounded into a tense sounds whether they are in front of the consonants or the vowels. But I have found the extraordinary aspects that they are sounded into lax consonants not tense sounds. Second, the consonant cluster ‘ㄶ’ should be sounded into aspirated sounds in front of the consonants. But I have also found the extraordinary aspects that they are sounded with lax consonants. Through Examining the features of the syllables behind the Word final consonant cluster, I have able to catch how the Word final consonant cluster are sounded. Finally, in front of the vowels, the latter one in the Word final consonant cluster should be linked, and then both in the Word final consonant cluster should be pronounced. But I have also found the extraordinary aspect that one of the consonant in the word final consonants is dropped out by the reconstruction. So I have tried to examine how the reconstruction appears in each district except Je-ju. I hope that my research help look into the features of the consonant cluster simplification minutely.

영동방언의 자음군단순화와 음운규칙 간의 관계

박환옥 ( Park Hwanoak ) 세명대학교 지역문화연구소 2020 지역문화연구 Vol.19 No.1

이 글에서는 현대국어에 존재하는 어간말 자음군에 비추어 충청북도 영동군에서 실현되는 자음군단순화의 실현양상과 자음군단순화와 음운규칙 간의 관계를 알아보는 것을 목적으로 한다. 영동지역의 경우, 체언 2개, 용언 12개의 어칸말 자음군이 확인된다. 이 방언의 체언 어간말 자음군은 ‘ㄱㅆ’, ‘ㅂㅆ’만 존재한다. 용언에 나타나는 어간말 자음군은 ‘ㄵ, ㄶ, ㄺ, ㄻ, ㄼ, ㄿ, ㅀ, ㄹㅡㅇ, ㅂㅆ’이다. 영동방언의 자음군단순화와 음운규칙 긴의 관계를 살펴보면, 경음화는 대부분의 자음군에서 발생한다. 유음화를 보이는 어형들은 ‘ㅀ, ㄹㅡㅇ’을 말음으로 가지는 자음군이고, 연구개음화를 보이는 어형들은 ‘ㄵ, ㄻ, ㅂㅆ’을 말음으로 가지는 자음군들로, 자음군단순화가 먼저 일어난 후에 연구개음화가 일어날 수 있다. 유기음화는 ‘ㄶ, ㅀ’을 말음으로 가진 어형들의 경우 ‘뚫-’을 제외하고 유기음화에 의해 ‘ㅎ’와 자음이 축약을 일으켜 단순화할 수 있는 자음이 소멸되어 자음군단순화가 적용되지 않는다. 비음화는 어간말에 개재자음이 삭제된 이후에 비음화가 일어날 수 있다. This paper aims to describe the realization patterns and the relations of phonological rules on stem-final consonant clusters in Yeongdong dialect. In Yeongdong dialect, there are 2 nouns and 12 predicates(verbs and adjectives) which have the stem-final consonant clusters. This paper shows that this dialect has two noun stem-final consonant clusters, which are ‘ㄱㅆ’, ‘ㅂㅆ’ and the stem-final consonant clustersof verbs and adjectives are ‘ㄵ, ㄶ, ㄺ, ㄻ, ㄼ, ㄿ, ㅀ, ㄹㅡㅇ, ㅂㅆ’. In case of the relations between the consonant cluster simplification and phonological rules, tensification are realized in most of the consonant clusters. Liquidization are shown in the consonant clusters which have ‘ㅀ, ㄹㅡㅇ’ as the stem-final consonant clusters. The consonant clusters which have ‘ㄵ, ㄻ, ㅂㅆ’ as the stem-final consonant clusters show velarization, and it is realized after the consonant cluster simplification. As for Aspration, stems except ‘뚫-’ that have stem-final cluster such as ‘ㄶ, ㅀ’ don’t show any consonant cluster simplification due to the phonological reduction of ‘ㅎ’ with the following consonant. Nasalization are shown after deletion of the Intervenient Consonant.

Acquisition of English Onset Clusters by Chinese and Korean EFL Learners

Hao, Chunhui,Seok-Keun Kang 국제언어인문학회 2016 인문언어 Vol.18 No.-

This study accounts for the acquisition of English onset /s stop r/, /s stop/ and /stop r/ clusters by Chinese and Korean learners. Participants in this study include 10 low level and 10 high level Mandarin and Korean EFL learners each. The material includes 27 words of target clusters. The results show that Mandarin and Korean speakers tended to make more modification in the more marked tri-consonant clusters than the less marked bi-consonant clusters. In addition, high level subjects performed noticeably better than low level subjects in both groups. Korean and Mandarin speakers’ simplification strategies were different: Mandarin speakers inserted /u/ or /ə/ into the clusters, whereas Korean speakers preferred to replace one of the consonants (often /r/) with another consonant in tri-consonant clusters and /stop r/ clusters. Both Korean speakers and lower English proficiency Mandarin speakers mistakenly aspirated the voiceless stops /p, t, k/ in the onset /s stop/ clusters.

15세기 ㅅ계 합용병서의 음가에 대한 재검토

김정인 한말연구학회 2023 한말연구 Vol.64 No.50

. In the early stages of the creation of Hunminjeongeum, the ㅅ-clusters within morphemes can be considered as consonant clusters. However, in the late 15th century, examples of ㅅ appearing before the ㅅ-clusters ('봇ᄭᅡ',etc) suggest that the preceding ㅅ was redundant, indicating that the ㅅ-clusters in this context was a fortis [s’]. Ultimately, within morphemes, the ㅅ-clusters transitioned from consonant clusters to [s’]. In the 16th century, there is an increase in the usage of ㅅ+ㅅ-clusters which seemed not renovated form, and this increase, may be attributed to non-phonological motivation influencing the pronunciation shift of the ㅅ-clusters. Among the various instances of the s-clusters of anlaut, those changed from the forms pronounced as lenis [s] are considered fortis [s’]. In contrast, those consistently appearing as the ㅅ-clusters since the early stages of Hunminjeongeum creation are considered consonant clusters. Treating the same character in the same period as having two different underlying phonemes is cumbersome. Furthermore, assuming that these characters originally represented fortis, it is implausible that the Hunminjeongeum creator assigned the characters which included ㅅ that had not even been realized. However, if we were to consider the pronunciation of the ㅅ-clusters of anlaut from the beginning as consonant clusters, it is difficult to establish a motivation for the change. Therefore, we suggest that the change of the pronunciation of the ㅅ-clusters of anlaut can also be attributed to the influence of non-phonological motivation.

한국어 음절말 자음군 발음 오류 양상과 특성 : 베트남인 한국어 학습자를 중심으로

안성민 이중언어학회 2023 이중언어학 Vol.91 No.-

This study provides basic data for improving Korean pronunciation education for Vietnamese learners of Korean by examining the patterns and characteristics of their pronunciation errors in syllable-final consonant clusters. The error rate by consonant-cluster was ‘ㄾ’ (65.9%), ‘ㄿ’ (65.5%), ‘ㄼ’ (64.9%), ‘ㄺ’ (54.0%), ‘ㄻ’ (47.7%), ‘ㄳ’ (42.2%), ‘ㄵ’ (24.6%), ‘ㅄ’ (5.6%); there were many pronunciation errors in the consonant cluster including ‘ㄹ’. More errors occurred when a syllable-final consonant cluster preceded a consonant than a vowel, and there were differences in the error patterns. Pronunciation error rates when the trailing element was a consonant did not differ between levels 2 and 3, while a significant difference appeared only in level 4. By grammatical category, the pronunciation error rates of nominal was lower than that of verbal, and in terms of vocabulary difficulty, beginner-level vocabulary was lower than intermediate-level vocabulary. Vocabularies for selecting the first consonant were lower than the vocabularies for selecting the second consonant by location of the selected consonant. In level 2, there were many error types in the pronunciation of the second consonant, but in level 3, there was more attempt to pronounce the first consonant even with a pronunciation error. Pronunciation error rates involving trailing vowels showed significant differences among all levels, regardless of grammatical category or vocabulary difficulty. According to the position of the selected consonant, the pronunciation error rate was high for the vocabulary that selected the second consonant. Level 2 showed an error of omitting a consonant even when accompanied by a trailing vowel, while level 3 showed failed attempts to pronounce both consonants. In level 4, as the success rate for articulation of the preceding consonant increased, the pronunciation error rate decreased. 본 연구는 베트남어 한국어 학습자의 음절말 자음군 발음 오류의 양상과 특성을 살펴봄으로써 한국어 발음 교육 개선을 위한 기초 자료를 제공한다. 자음군별 오류율은 'ㄾ'(65.9%), 'ㄿ'(65.5%), 'ㄼ'(64.9%), 'ㄺ'(54.0%), 'ㄺ'(54.0%), 'ㄻ ' (47.7%), 'ㄳ' (42.2%), 'ㄵ' (24.6%), 'ㅄ' (5.6%); 'ㄹ'자음을 포함한 자음군에서 발음 오류가 많았다. 음절말 자음군에 후행하는 요소가 모음일 때보다 자음일 때 더 많은 오류가 발생하였고, 오류 양상에서도 차이가 있었다. 후행 요소가 자음일 때의 발음 오류율은 2급과 3급에서 차이가 없었고, 4급에서만 유의미한 차이가 나타났다. 어휘 난이도는 초급 어휘의 발음 오류율이 중급 어휘의 발음 오류율보다 낮았다. 선택된 자음의 위치에 따른 발음 오류율은 첫 번째 자음 선택 어휘가 두 번째 자음 선택 어휘보다 낮았다. 2급에서는 두 번째 자음의 발음 오류 유형이 많았으나, 3급에서는 발음 오류가 있어더라도 첫 번째 자음을 발음하려는 경향이 있었다. 후행하는 요소가 모음일 때의 발음 오류율은 문법 범주나 어휘 난이도와 상관없이 모든 수준에서 유의미한 차이를 보였다. 선택한 자음의 위치에 따라 두 번째 자음을 선택한 어휘의 발음 오류율이 높았다. 2급의 경우 모음이 후행할 때도 자음을 생략하는 오류를 보였고, 3급에서는 두 자음을 모두 발음하지 못하는 오류를 보였다. 4급에서는 선행 자음의 조음 성공률이 높아지면서 발음 오류율도 낮아졌다.

중국, 일본, 베트남 학습자의 한국어 음절 말 자음군 발음 연구

이화진(Lee, Hwajin),하호빈(Ha, Hobin) 중앙어문학회 2023 語文論集 Vol.96 No.-

This study aims to analyze the Korean syllable final consonant cluster which is spoken by advanced learners from China, Japan and Vietnam. Most learners tend to be confused about Korean syllable final consonant cluster - LC cluster > NC cluster > GC cluster > OC cluster - due to the complexity of consonant cluster simplification. In terms of substantives, a member of word final consonant cluster tends to be deleted in a wrong way for the followed vowel. The deletion of cluster member in this case is based on the convention of native Korean speakers reconstructs the word final syllable with a morphological boundary. In terms of predicates, a member of stem final consonant cluster tends to be deleted in a wrong way for the followed consonant especially in the occlusive sequences with LC cluster. The difference between both patterns correlates with the intuition about the morphological transition of an inflectional language such as Korean or Japanese.

중국, 일본, 베트남 학습자의 한국어 음절 말자음군 발음 연구

이화진,하호빈 중앙어문학회 2023 語文論集 Vol.96 No.-

본고는 한국어 고급 학습자들의 음절 말 자음군 발화 양상과 거기에서 나타나는 특징을 종합적으로 비교 고찰하는 것을 목적으로 한다. 이를 위해 중국, 일본, 베트남 학습자들을 대상으로 한국어 음절 말 자음군 발화 실험을 진행하였다. 실험 결과, 음절 말 자음군의 유형별 발화 오류율은 언어권에 상관없이 ‘유음계 > 비음계 > ㅎ계 > 폐쇄음계’의 순으로 나타났다. 그리고 체언의 어말에서는 모음이 결합할 때, 용언의 어간 말에서는 자음이 결합할 때 오류율이 높게 나타났다. 특히 용언 어간의 ‘ㄺ’와 ‘ㄿ’에 장애음이 연쇄하는 경우 [말다], [을다]와 같은 오류가 많이 나타났고, 용언 어간의 ‘ㄼ’에 비음이 연쇄하는 경우 [넘네]와 같은 오류가 많이 나타났다. 마지막으로 비음계 음절 말 자음군의 경우, 용언 어간 ‘앉-’에서는 오류가 거의 나타나지 않았고, 체언 ‘삶’에서만 오류가 많이 나타났다. 언어권별로 보면, 세부적으로 약간의 차이는 있으나 전체적으로 비슷한 오류 양상이 나타났다. This study aims to analyze the Korean syllable final consonant cluster which is spoken by advanced learners from China, Japan and Vietnam. Most learners tend to be confused about Korean syllable final consonant cluster - LC cluster> NC cluster > GC cluster > OC cluster - due to the complexity of consonant cluster simplification. In terms of substantives, a member of word final consonant cluster tends to be deleted in a wrong way for the followed vowel. The deletionof cluster member in this case is based on the convention of native Korean speakers reconstructs the word final syllable with a morphological boundary. In terms of predicates, a member of stem final consonant cluster tends to be deletedin a wrong way for the followed consonant especially in the occlusive sequences with LC cluster. The difference between both patterns correlates with the intuition about the morphological transition of an inflectional language such as Korean or Japanese.

An Articulatory Study on Consonant Cluster Simplification in L1 Korean and L2 English

Gwanhi Yun 한국영어학회 2023 영어학 Vol.23 No.-

The present study aims to investigate to what extent and how consonant cluster simplification occurs in L1 Korean and L2 English coda consonant clusters. To this end, four types of coda clusters, including /lp/ and /lk/ were examined with Korean speakers in two ultrasound imaging experiments. Experiment 1 showed that the variability patterns in the production of coda clusters emerged, including three types of variants for /lk/ (e.g., “partial velar,” “full velar,” and “zero velar” and two types of variants for /lp/ sequences (e.g., /l/-deletion and /p/-deletion). Their realization patterns varied depending on cluster type and individual speakers. “Partial velar” variants, particularly, are indicative of gestural hiding and gestural overlap inducing articulated but inaudible consonant clusters. Experiment 2 demonstrated that no-deletion variants were more dominant than deletion forms, along with hyperarticulation variants for English CC clusters (43% vs. 37% vs 20%). These findings provide support for gradient realizations of /CCC/ clusters in L2 English such as /ptk, ktp, lpt, lkt/, depending on cluster type and L2 speakers. Compared to the production of L1 Korean CCC, “no-deletion” variants were elicited more frequently, showing the exemplars of gestural hiding and gestural overlap.

논문(論文) : ≪설문해자(說文解字)≫ 독약(讀若)에 반영된 복자음 성모(聲母)

권혁준 ( Hyok Jun Kwon ) 중국어문연구회 2015 中國語文論叢 Vol.0 No.70

This study is designed to reveal the types of “Initial Consonant Clusters”(“clusters” hereafter) as reflected in “the Shuowen Duruo[讀若] Glosses”. As the time span of Old Chinese[OC] covers presumably more than 1,000 years, we will have to assume that there had been a number of phonological changes within the period. This seems to be the case with the cluster system, which would reduce in its types and numbers over time. We have some evidence against varieties of clusters existing in the pre-Qin time, which, however, eventually turned into simple consonants as they developed into Middle Chinese. Therefore, the process of cluster-simplification deserves our discussion. There are some materials that show what types of clusters were retained at the late stage of OC, such as the Shuowen Duruo Glosses, Paranomastic glosses[聲訓], yiwen[異文], tongjia[通假] as well as the annotations of the Eastern Han Scholars. This study takes the Shuowen Duruo Glosses as a main material. We start our discussion with reviewing the two antecedent studies on the consonantal system as revealed in the Shuowen, and take Zheng-Zhang Shangfang’s OC Phonology as the point of departure. We find his transcriptions on some characters problematic and attempt to revise them by providing some evidence. Thus we establish certain types of clusters which our predecessors failed to reconstruct or falsely reconstructed if they did at all. We also find that there are a number of types of clusters which had disappeared in the Duruo Glosses. That, however, does not necessarily mean that all of these missing clusters had turned into simple consonants as of the Eastern Han because they are not reflected in the Glosses. In conclusion, we have Cl, C·r, s(-)C and others reflected in the Glosses. This study may provide only a sketch description of the cluster system of the Eastern Han, but will make contribution to tracing back the phonological system of the time.

A Perception-based Syntagmatic Account of Consonant Cluster Simplification in Korean

Mira Oh(오미라) 한국언어학회 2018 언어 Vol.43 No.1

This study aims to provide a perceptual account for consonant cluster simplification (CCS) in Korean verb stems. I investigate what motivates the differences between the invariable and variable types of CCS, and I argue that the disparity between the two types in Korean results from the amount of contrast in terms of the distinctive feature specifications between consonants within a cluster. I make five claims by investigating both types of CCS in Korean verb stems. First, CCS in Korean is motivated by the requirement that every segment be perceptually salient. Second, the two types of CCS are determined based on the perceptibility difference among the competing consonants within the cluster. Third, the perceptual account of CCS better explains“the effect of the following consonant”than syllable-based analysis. Fourth, the constraint ranking that utilizes perceptually motivated markedness and faithfulness constraints within the Optimality Theory framework can comprehensively account for the two types of CCS. Fifth, the proposed constraint ranking also explains how to produce truncata that are frequently used in fast speech or casual speech for the purpose of economy. In this study, I show that in contrast to previous studies of CCS in Korean in syllabic terms without reference to context where the consonant cluster appears, syntagmatic context needs to be taken into consideration to comprehensively understand CCS in Korean. This study has a general theoretical implication that phonetic perceptibility can directly influence patterns in a phonological process.