/l/ and /r/ production by Korean and Japanese speakers of English : What factors are influential for the production\ulcorner

박시균,Park See-Gyoon The Korean Society Of Phonetic Sciences And Speech 1999 말소리 Vol.37 No.-

본 논문은 한국인과 일본인이 영어 유음 /l/과 /r/을 발음하는데 있어서 어떤 요소들이 가장 큰 영향을 미치는가를 알아보는데 중점을 두고 있다. Park & Ingram (1995)에서는 한국인과 일본인의 영어 발음인지(perception) 실험을 통해 동 문제에 대해 분석을 해보았는데 본 논문에서는 같은 주제를 발음 실험을 통해 그 해답을 찾고 있다. 실험의 결과는 언어 의존적인 요소인 모국어 음운 체계의 영향뿐만 아니라 언어 독립적인 요소인 화자(speaker), 유음이 나타나는 위치(position), 유음의 유형(type) 그리고 영어 유음을 발음하는 각 개인 등의 요소들이 모두 한국인과 일본인의 영어 유음 /l/과 /r/ 발음에 영향을 미치고 있음을 보여 주고 있다. 이는 인지 실험에서도 드러난 바로 인지와 발음은 서로 밀접한 관계를 맺고 기제가 일어나고 있음을 보여 준다.

아랍인의 한국어 단순 모음 인지 연구

박시균(Park, See-Gyoon),김지영(Kim, Ji-Young) 한국언어학회 2016 언어 Vol.41 No.3

It is well known that L2 experience and L1 interference are the most important factors in L2 sound acquisition based upon the models of Flege (Speech Learning Model: SLM; Flege, 1987a, 1987b; Bohn & Flege, 1992) and Best (Perceptual Assimilation Model: PAM; Best, 1993, 1994, 1995; Best & Tyler, 2007). In our perceptual experiments on Korean monophthongs by Arabic learners of Korean, we examined influence of the two important factors of L2 experience and L1 interference posed by SLM and PAM. The results showed that the prediction based upon SLM and PAM was not completely correct but acceptable only in part. Why the two models were not totally acceptable should be investigated continuously. Our experiments results revealed, however, that some phonetic features such as lip rounding and the front-back position of the tongue, might have acted as the third important factor which could influence on the two existing factors.

일본어의 길이 단위 ‘모라’의 특수성에 관한 연구

박시균(See-Gyoon Park),하타나카츠유미(Tsuyumi Hatanaka) 사단법인 한국언어학회 2014 언어학 Vol.0 No.70

It is well accepted amongst Japanese speakers and many Japanese scholars that a timing unit ‘mora’ is language-specific in Japanese. We, however, are not certain that the mora is language-specific, in particular in terms of length, in Japanese because, as Beckman (1982) challenged, the mora was non-identical in length depending upon different kinds of morae. To examine our skepticism on the mora length, we designed a perception experiment. In the experiment, 4 sets of the motto-moto like ‘sokuon’ opposition and 2 sets of kanai-kannai like ‘hatsuon’ opposition were read by two native Japanese speakers and the difference of length of the geminate and single consonants of each pair was splitted from 1 (single consonant) to 10 (geminate consonant) and this splitted tokens were given to a native Korean group (101 subjects) and a native Japanese group (270 subjects). Our experiment result showed that unlike our prediction the Japanese group’s perceptual border was nearer to single consonants than that of the Korean group. This means that the ‘mora’ as a timing unit is not in an absolute position. However, the Japanese group’s result showed more perceptual consistency than that of the Korean group. This result seems to say that while Koreans rely solely upon the length factor of the consonant in question for distinguishing long and short tokens of the Japanese words, Japanese people use other factors such as the lengths and fo of the vowels before and after the consonant in question and the pitch-accent of the words in addition to the length factor of the consonant in question.

딥러닝 기반 실시간 객체 추적을 이용한 3D 프린터 출력 시간 측정

임현아,문비,오요셉 대한산업공학회 2022 대한산업공학회 추계학술대회논문집 Vol.2022 No.11

외국어 화자에게 어려운 한국어 자음 분석 및 교육 방안 모색 : 영어 화자를 중심으로

박시균(See-Gyoon Park),권병로(Pyong-Ro Kwon) 국어문학회 2008 국어문학 Vol.45 No.-

This article studies difficult Korean consonants, in particular, stops, liquid, and geminate consonants to English native people who are learning Korean. In section Ⅱ, the authors scrutinize the discrepancies in perception and production of Korean stops, liquid and geminate sounds between English speakers learning English and native Koreans by means of the analyses of the experiments’ results. The results of the experiments claim that orthography, morphemes, phonological rules as well as sounds themselves are all important and that teachers teaching Korean sounds to foreigners should consider all the relevant factors in teaching Korean sounds. Section Ⅲ is an important section because it gives us several tips of how to teach Korean sounds and phonological processes to English speaking people learning Korean.

음성실험을 통한 한국어 학습자에 대한 어중 경폐쇄음 교육 방안 모색

박시균(Park See-Gyoon),김지영(Kim Ji-Young) 우리말학회 2019 우리말연구 Vol.57 No.-

본 연구는 한국어 학습자들을 위한 효율적이고 실제적인 경음 지도 방안을 모색하고자 중국인, 몽골인, 베트남인 한국어 학습자들을 대상으로 ‘아빠(VC*V)’, ‘압바(VCCV)’, ‘압빠(VCC*V)’ 유형의 단어를 읽게 한 후 한국인들이 어떻게 인지하는지 살펴보았다. 또한 한국인과 한국어 학습자의 음성자료를 비교하여 어떠한 유사점과 차이점을 보이는지 살펴보았다. 인지실험 결과 한국인들은 ‘아빠(VC*V)’ 유형에서는 경음으로 뿐만 아니라 평음으로도 인지하였지만 ‘압바(VCCV)’, ‘압빠(VCC*V)’ 유형에서는 4개의 토큰을 제외하고 모두 경음으로 인지한 것을 알 수 있었다. 산출실험 결과, ‘아빠(VC*V)’ 유형의 경우에는 한국인의 인지에서 오류를 보인 피험자의 음성자료와 한국인의 음성자료의 차이점을 비교해 보았다. 그 결과 한국어 학습자들의 선행모음의 길이와 VOT는 길게 나타났으나 폐쇄구간의 길이는 현저히 짧게 나타난 것을 알 수 있었다. ‘압바(VCCV)’, ‘압빠(VCC*V)’ 유형의 경우에는 한국인이 경음으로 인지한 음성자료와 한국인의 음성자료의 유사점을 비교해 보았다. 이 두 유형으로 조음을 하면 한국어 학습자들과 한국인의 폐쇄구간의 길이와 선행모음의 길이가 유사해지는 것을 알 수 있었다. 즉 ‘아빠(VC*V)’ 유형으로 조음할 때보다 ‘압바(VCCV)’나 ‘압빠(VCC*V)’ 유형으로 조음할 때 경음으로 조음할 확률이 상대적으로 크게 높아진다고 할 수 있다. 따라서 어중 초성에 위치한 경음을 제대로 조음하지 못하는 한국어 학습자들에게 선행자음이 있는 ‘압바(VCCV)’나 ‘압빠(VCC*V)’ 유형을 이용하여 지도한다면 경음의 긴장성을 파악하는 데 도움을 줄 수 있으며 결과적으로 경음의 정확한 산출을 이끌어 낼 수 있을 것이다. This study dealt with Korean people’s perception of the tokens of 1) ap*a(VC*V), 2) appa(vccv) and 3) app*a(VCC*V) types produced by Chinese, Mongolian and Vietnamese learners of Korean in order to find effective and actual ways of teaching Korean fortis stops to foreign learners. Also the authors carried out a production test followed by the comparison of speech data produced by foreign learners and native Koreans to elicit similarities and differences of the data of the two groups. The perception test showed that Korean listeners perceived ‘ap*a(VC*V)’ type tokens as not only ‘fortis’ but ‘lenis’ and ‘aspirated’ ones. However they mostly perceived ‘appa(vccv)’ and ‘app*a(VCC*V)’ type tokens as ‘fortis’ ones only with 4 exceptions. The ‘ap*a(VC*V)’ type misperceived tokens of foreign learners were compared with Koreans’ equivalent tokens. The tokens of foreign learners showed longer lengths of preceding vowels and VOT but remarkably shorter length of closure section than those of Koreans. For ‘appa(vccv)’ and ‘app*a(VCC*V)’ type tokens Koreans and foreign learners appeared to have similar lengths of closure section and preceding vowels. This result seems to show that foreign learners’ production of ‘appa(vccv)’ and ‘app*a(VCC*V)’ type tokens can lead to higher rate of accuracy of Korean ‘fortis’ stops than that of ‘ap*a(VC*V)’ type tokens.

일본인과 호주인 한국어 학습자의 한국어 겹자음 길이 비교 연구

박시균(See-Gyoon Park),하타나카츠유미(Tsuyumi Hatanaka) 사단법인 한국언어학회 2012 언어학 Vol.0 No.64

This paper aimed 1) to compare the length of Korean geminate vs single consonant sounds produced by Korean, Japanese and Australian speakers and 2) to analyze Korean liquid geminate vs single sound pronunciations produced by Japanese learners who do not have geminate liquids in their L1. The Japanese Beginner Group, whose L1 has a length distinction between geminate and single consonants as in Korean. produced length ratios closer to those of Koreans than the Australian Intermediate Group did. This showed that L1 influence was stronger than the experience of L2 learning. Even though Japanese learners (both the Beginner Group and the Experienced Group) could well distinguish Korean geminate and single obstruents and nasals in their production. their phonetic realization of liquid geminate and single sounds was neither clear nor similar to that of Koreans. Clear phonetic, not phonological, differences in quality between similar sounds in L1 and L2 seems to be an obstacle for L2 learners which cannot be overcome in a short time. The geminate structure of the sounds seems to give L2 learners more difficulties when they produce them.

한국인 영어학습자의 발음 오류 원인 분석과 교육 방법 모색

박시균(See-Gyoon Park) 사단법인 한국언어학회 2004 언어학 Vol.0 No.40

교수법을 활용한 한국어 장애음 교육 : 중국인 여성 결혼 이민자를 중심으로

박시균(Park, See-Gyoon) 한국언어문학회 2011 한국언어문학 Vol.77 No.-

This paper studies two listening tests of Chinese novice learners on Korean obstruents and a Korean listening evaluation for the production data of Korean obstruents by the Chinese learners. Based on the result of the 12 Chinese subjects at the first experiment, they were divided into 3 groups to go through different teaching methods for pronunciation education. Group A went with the cognitive approach, Group B with the audio-lingual method and Group C with the task-based approach, respectively. Each group was set with different time scales by differential stages and curriculums, aligned by each method’s characteristic. After the experimental lessons, the second test which was the same as the first test at the procedure was run to find out the efficiency of the given teaching methods. In terms of statistics, the experimental results between groups did not bring out a meaningful result; however the improvement at the in-group results was meaningful. In particular a far more efficiency in Groups B and C with the audio-lingual method and the task-based approach than Group A based on the cognitive approach, was observed. The audio-lingual method and the task-based approach teaching method showed that they are the efficient methodologies for teaching Korean sounds to the beginner level of foreign learners.

한국어, 영어, 일본어의 음성, 음운 차이에 대한 연구

박시균(See-Gyoon PARK) 국어문학회 2015 국어문학 Vol.58 No.-

L2 음운 습득 분야에서 L1의 전이는 연구의 중심에 서 있었다. 대조분석가설(CAH)을 주창한 Lado(1957) 이후 ‘L1의 전이’는 L2음 습득 모델들에서도 주요한 요소로 자리 잡았다. 두 개의 모델이 이 분야의 학자들에게 가장 선호되고 영향을 끼쳐 왔는데 첫 번째 모델은 Flege(1987a, 1987b)가 주창한 ‘음성학습 모델(SLM)’이고 두 번째 모델은 Best(1993, 1995)가 주창한 ‘지각동화모델(PAM)’이다. 이 두 모델은 모두 ‘L1의 전이’와 ‘L2 경험’을 L2음 습득의 중요 요소로 보고 있다. 본고에서 필자는 ‘L1의 전이’와 ‘L2 경험’과 관련하여 네 개의 다른 경우를 상정하여 필자의 실험들로 이 두 모델을 검증하였다. 네 개의 경우 중 두 개는 이 두 개 모델의 예측이 맞는 것으로 나타났지만 나머지 두 개는 예측이 결과와 대비해서 성공적이지 못했다. 예측이 성공적이지 못한 두 개의 경우에는 ‘제3의 요소’가 필요하였다. 그리고 이 제3의 요소는 ‘유표성’과 직접적으로 관련이 있었다. ‘제3의 요소’가 L2 학습자들이 L2 음을 습득하는데 ‘긍정적’으로 작용하는 경우 이는 목표로 하는 L2음이 ‘무표적’인 방향에 서 있었고 반대로 ‘제3의 요소’가 ‘부정적’으로 작용하는 경우 L2음은 ‘유표적’인 방향에 서 있었다. ‘L1의 전이’와 ‘L2 경험’이 L2 음 습득에서 중요한 요소들이지만 ‘제3의 요소’도 이 두 요소와 동등하게 중요하다. L1 transfer has been a main theme in the area of L2 sound acquisition. Since Lado(1957) advocated the Contrastive Analysis Hypothesis(CAH) where ‘L1 transfer’ was a main factor, L1 transfer has been a main factor in the following L2 sound acquisition models. Two models have been favorite and influential to the scholars in this area. One was Speech Learning Model(SLM) asserted by Flege(1987a, 1987b). The other was Perceptual Assimilation Model(PAM) claimed by Best(1993, 1995). Both models put the emphasis on ‘L1 transfer’ and ‘L2 experiences’ as main factors of L2 sound acquisition. In this paper, I verified these two models with my own experiments for four different cases in association with L1 transfer and L2 experiences. Out of four cases two were explained by the two models’ predictions but the other two were not successful with their predictions. The third factors, whether those are ‘positive’ or ‘negative’, were needed to solve the problems of the two models’ insufficient predictions. Also the third factors were directly related to the ‘markedness’. If the factors are ‘positive’ those are related to ‘unmarked’ direction of the targeted L2 sounds and if ‘negative’, those are related to ‘marked’ direction of them. Although ‘L1 transfer’ and ‘L2 experiences’ are important factors in L2 sound acquisition, the third factors are equally important to the other factors. The detailed relationship amongst theses three factors should be resolved in future researches.