중국인 학습자의 한국어 어두 폐쇄음 발음에 대한 음향 음성학적 연구

Abstract

본 연구의 목적은 한국어를 공부하는 중국인 학습자의 한국어 어두 폐쇄음의 지각 및 산출의 오류 양상을 밝히고, 중국인 학습자가 발음한 한국어 어두 폐쇄음의 음향적 특성을 고찰하여 한국어 모어 화자와 비교 분석하는 것이다. 일반적으로 두 가지 계열의 폐쇄음을 갖는 여러 언어들과 달리 한국어는 평음, 경음, 격음의 세 계열 폐쇄음을 갖는다. 이러한 특이성 때문에 외국인 학습자에게 한국어 폐쇄음의 습득은 큰 장벽이 되고 있으며, 중국인 학습자 역시 한국어 폐쇄음의 습득에 어려움을 겪고 있다고 많은 연구에서 보고되었다. 그러나 기존의 연구들은 중국인 학습자가 한국어 폐쇄음 중 평음과 격음의 구분을 어려워한다는 것과 평음과 경음의 구분을 어려워한다는 것으로 혼재되어 나타나고 있다. 그래서 중국인 학습자가 평음, 경음, 격음을 각각 어떤 음소와 혼동하는지 명확히 알기 어렵고, 지각이나 산출 한 가지 측면만을 고찰한 연구가 대부분이어서 둘의 상관관계에 대해서도 알 수 없다. 이에 본 연구에서는 지각과 산출 양 측면에서 각 음소별 오류 양상을 구체적으로 밝혀 상관관계를 고찰하고, 중국인 학습자가 발음한 한국어 어두 폐쇄음의 음향적 특성을 한국어 모어 화자와 비교하여 그 차이를 분석해 보았다. 이를 위해 중국인 초급과 중급 학습자를 대상으로 청취 실험과 발화 녹음을 진행하였다. 먼저 청취 실험을 분석한 결과 중국인 학습자는 한국어 어두 폐쇄음 중 경음을 가장 잘 지각하고 평음과 격음은 상대적으로 지각을 어려워하는 것으로 나타났다. 평음은 격음과, 격음은 평음과 혼동하는 지각 오류가 많이 관찰되었고, 평음과 격음 모두 경음과 혼동하여 지각하는 비율은 매우 낮았다. 이러한 양상은 초급과 중급 학습자 모두에게서 동일하게 관찰되었다. 다음으로 발화 녹음을 분석하여 산출 양상을 살펴보았다. 초급과 중급 학습자 모두 경음의 산출률이 가장 낮았고, 학습 단계가 높아질수록 격음의 산출이 눈에 띄게 향상되었다. 평음은 경음에 비해서는 양호하나, 격음 대비 산출 오류가 많이 관찰되었다. 한편 학습 단계가 높을수록 평음과 격음 모두 지각률이 높아졌고, 그에 상응하게 산출 역시 향상되어 지각의 용이함이 산출의 정확도로 이어짐을 알 수 있었다. 그러나 경음은 학습 단계에 상관없이 높은 지각률을 보인 것과는 별도로 산출률이 세 음 중 가장 낮아 지각과 산출의 관계가 무조건 비례하지는 않음을 보여주었다. 이러한 현상은 [기식성]에 의해 유기음과 무기음 두 계열로 대립되는 중국어 폐쇄음 체계에서 그 원인을 찾아볼 수 있다. 중국인 학습자는 모국어의 영향으로 어두에 위치한 한국어 폐쇄음 중 기식이 있는 평음과 격음은 중국어 유기음에, 기식이 없는 경음은 무기음에 대응시킨다. 그 결과 무기음에 일대일 대응된 경음은 학습 초기 단계부터 지각이 용이하나, 유기음에 대응된 평음과 격음의 구분에는 혼동이 발생하는 것이다. 반면 산출 시에는 그들에게 친숙한 음성 자질인 [기식성]을 활용하여 경음에 비해 평음과 격음을 수월하게 산출할 수 있는 것이라 판단할 수 있다. 마지막으로 중국인 학습자가 발음한 한국어 어두 폐쇄음의 음향적 특성을 분석하였다. 이를 위해 산출 실험에서 사용한 발화 녹음 자료 중 어두에 폐쇄음이 위치한 특정 단어만을 분절하여 새로운 단어 파일을 생성한 후 파형, 스펙트로그램, 음높이 곡선 등을 참조하여 VOT와 F0를 측정하였다. 또한 중국인 학습자와 한국어 모어 화자 어두 폐쇄음의 음향적 특성을 비교하기 위해 한국어 모어 화자에도 동일한 절차로 실험을 진행하였다. 그 결과 한국어 모어 화자는 폐쇄음의 세 가지 발성 유형의 변별을 위해 VOT와 F0를 모두 사용하는 것으로 나타났으나, 중국인 학습자는 VOT만을 사용하였고 F0는 발성 유형의 변별에 크게 기여하지 않는 것으로 나타났다. 이러한 현상은 학습 단계가 낮을수록, 여성보다 남성 학습자에게서 더욱 뚜렷하게 관찰되었다. 이와 같은 실험 결과를 바탕으로 중국인 학습자를 위한 한국어 어두 폐쇄음의 교육 방안을 제안하였다. 발화 실험 분석 결과 중국인 학습자는 한국어 모어 화자와 달리 어두 폐쇄음을 발음할 때 F0를 거의 사용하지 않는 것으로 밝혀졌는데, 이에 착안하여 F0와 성조를 연결시켜 발화 학습에 활용하는 방법을 제시하였다. 또한 지각률이 낮게 나타난 평음과 격음은 지각 학습을, 산출률이 낮게 나타난 경음은 발화(조음) 학습에 집중할 필요가 있음을 밝혔다. 본 연구는 중국인 학습자의 한국어 어두 폐쇄음의 지각과 산출의 오류 양상을 구체적으로 밝히고 지각의 용이함이 산출의 정확도나 용이함으로 이어지는지 여부를 관찰하였다. 이를 바탕으로 발성 유형에 따라 선행되어야 할 학습을 제시하였고, 중국인 학습자가 발음한 한국어 어두 폐쇄음의 음향적 특성을 한국어 모어 화자와 비교하여 이를 활용한 폐쇄음의 실제 교육 방안을 제안하였다는 점에서 그 의의를 갖는다. |The purpose of this study lies in identifying correlation between perception and pronunciation of Korean word-initial stops by Chinese learners, contemplating the acoustic features of Korean word-initial stops pronounced by Chinese learners, and lastly comparing and analyzing the pronunciations of Korean word-initial stops by Chinese with those of Korean native speakers. Unlike many languages that have generally two types of stops, Korean has three-way contrasts: lax, tense and aspirated. Such unique feature often becomes a big barrier for foreigners to learn Korean stops, and there have been many studies demonstrating that Chinese learners also have difficulties in learning Korean stops. However, some studies state that Chinese learners find it difficult to distinguish lax from aspirated, while others conclude that Chinese learners have more difficulties in differentiating lax from tense. In addition, it is hard to find any study which examines pronunciation aspect or correlation between perception and production. Hence, this study focuses on analyzing the correlation between perception and outcome, as well as the patterns of pronunciation errors. First, in order to figure out the perception and production patterns of Chinese learners concerning Korean word-initial stops, we conducted listening test and speech recording for each learning stage. The result is that the highest perception was found in tense and recognition of lax and aspirated were lower. And more perception mistakes were observed in distinguishing lax from aspirated than from tense. As for aspirated, Chinese learners appeared to have more difficulty in distinguishing aspirated from lax rather than tense. When it comes to tense, both beginning and intermediate level learners showed high perception of 80% or more on average, but the opposite was observed in production, implying that high perception does not automatically lead to accuracy or easiness of production. Such errors in perception and production result from the system of Chinese stops. Chinese stops are divided into two: aspirated and unaspirated depending on [aspirated]. Chinese learners sensitively react to [aspirated] out of phonetic properties. Therefore, Chinese learners match both lax(pronounced with weak breathing) and aspirated (produced with strong breathing) to aspirated, while they match tense to unaspirated. As a result, distinguishing lax from aspirated often leads to perception errors. And the perception of tense was relatively higher since Chinese learners match tense to unaspirated one-on-one. When it comes to production, however, the pronunciation of lax and aspirated appear to be easier since lax and aspirated adopt [aspirated] which is familiar phonetic property to Chinese learners. Next, I performed speech recording, in order to examine the acoustic features of word-initial stops produced by Chinese learners and to compare them with those pronounced by Korean native speakers. Out of recorded speech, I collected only specific words with word-initial stops and created a new word file, to measure and analyze VOT and FO of segmented initial consonants by utilizing waveform, spectrogram, and pitch curve. According to the test, Korean native speakers turned out to use both VOT and FO in order to distinguish three types of stops. Tense was differentiated from the other two via VOT, while lax was distinguished from the others by FO. However, Chinese learners could differentiate three types of stops only through VOT, while FO didn't seem to contribute much to distinction of phonation types. Such result was more apparent in Chinese learners at lower learning level and in female than male learners. Based on the result, I suggest education plans depending on phonation types so that Chinese learners learn Korean word-initial stops. Since FO is identified as a major discriminating parameter, I reason that it is important to help Chinese actively utilize FO. In the context, I suggest education plans which connects Chinese intonations with FO of Korean stops. Furthermore, since the easiness of perception and production is different depending on phonation types, I suggest that perception study should precede for the learning of lax and aspirated, while production study should be a preceding condition for the learning of tense. The significance of this study lies in that it looks into the correlation between perception and production of Korean word-initial stops, and the patterns of perception and production mistakes made by Chinese learners, thereby examining whether easier perception leads to the higher accuracy or easiness of production. It is also meaningful that the study suggests education plans depending on phonation types, and it compares the acoustic features of Korean word-initial stops pronounced by Chinese learners with those produced by Korean native speakers, thereby identifying difference between the two, and suggesting the ways for Chinese learners to learn Korean stops properly.; The purpose of this study lies in identifying correlation between perception and pronunciation of Korean word-initial stops by Chinese learners, contemplating the acoustic features of Korean word-initial stops pronounced by Chinese learners, and lastly comparing and analyzing the pronunciations of Korean word-initial stops by Chinese with those of Korean native speakers. Unlike many languages that have generally two types of stops, Korean has three-way contrasts: lax, tense and aspirated. Such unique feature often becomes a big barrier for foreigners to learn Korean stops, and there have been many studies demonstrating that Chinese learners also have difficulties in learning Korean stops. However, some studies state that Chinese learners find it difficult to distinguish lax from aspirated, while others conclude that Chinese learners have more difficulties in differentiating lax from tense. In addition, it is hard to find any study which examines pronunciation aspect or correlation between perception and production. Hence, this study focuses on analyzing the correlation between perception and outcome, as well as the patterns of pronunciation errors. First, in order to figure out the perception and production patterns of Chinese learners concerning Korean word-initial stops, we conducted listening test and speech recording for each learning stage. The result is that the highest perception was found in tense and recognition of lax and aspirated were lower. And more perception mistakes were observed in distinguishing lax from aspirated than from tense. As for aspirated, Chinese learners appeared to have more difficulty in distinguishing aspirated from lax rather than tense. When it comes to tense, both beginning and intermediate level learners showed high perception of 80% or more on average, but the opposite was observed in production, implying that high perception does not automatically lead to accuracy or easiness of production. Such errors in perception and production result from the system of Chinese stops. Chinese stops are divided into two: aspirated and unaspirated depending on [aspirated]. Chinese learners sensitively react to [aspirated] out of phonetic properties. Therefore, Chinese learners match both lax(pronounced with weak breathing) and aspirated (produced with strong breathing) to aspirated, while they match tense to unaspirated. As a result, distinguishing lax from aspirated often leads to perception errors. And the perception of tense was relatively higher since Chinese learners match tense to unaspirated one-on-one. When it comes to production, however, the pronunciation of lax and aspirated appear to be easier since lax and aspirated adopt [aspirated] which is familiar phonetic property to Chinese learners. Next, I performed speech recording, in order to examine the acoustic features of word-initial stops produced by Chinese learners and to compare them with those pronounced by Korean native speakers. Out of recorded speech, I collected only specific words with word-initial stops and created a new word file, to measure and analyze VOT and FO of segmented initial consonants by utilizing waveform, spectrogram, and pitch curve. According to the test, Korean native speakers turned out to use both VOT and FO in order to distinguish three types of stops. Tense was differentiated from the other two via VOT, while lax was distinguished from the others by FO. However, Chinese learners could differentiate three types of stops only through VOT, while FO didn't seem to contribute much to distinction of phonation types. Such result was more apparent in Chinese learners at lower learning level and in female than male learners. Based on the result, I suggest education plans depending on phonation types so that Chinese learners learn Korean word-initial stops. Since FO is identified as a major discriminating parameter, I reason that it is important to help Chinese actively utilize FO. In the context, I suggest education plans which connects Chinese intonations with FO of Korean stops. Furthermore, since the easiness of perception and production is different depending on phonation types, I suggest that perception study should precede for the learning of lax and aspirated, while production study should be a preceding condition for the learning of tense. The significance of this study lies in that it looks into the correlation between perception and production of Korean word-initial stops, and the patterns of perception and production mistakes made by Chinese learners, thereby examining whether easier perception leads to the higher accuracy or easiness of production. It is also meaningful that the study suggests education plans depending on phonation types, and it compares the acoustic features of Korean word-initial stops pronounced by Chinese learners with those produced by Korean native speakers, thereby identifying difference between the two, and suggesting the ways for Chinese learners to learn Korean stops properly.