[논문]Teaching Pronunciation Using Sound Visualization Technology to EFL Learners

Min, Su-Jung; Pak, Hubert H.

문제 정의

Training techniques for the production of L2 contrasts have been developed and examined in several experimental studies (Chapelle, 1998; Chun & Plass, 1996; Doughty, 1991; Leather, 1983; Leon & Martin, 1972; Swain & Lapkins, 1995). These studies conducted experiments to show that certain aspects of L2 speech production can be trained, and that production training might also lead to improved perception. However, the development of training techniques for L2 pronunciation is in its infancy, and there is much to be explored in assessing whether various methods of pronunciation training are effective in enabling subjects to accurately produce L2 contrasts (see Pennington & Richards, 1986, for discussion).
This paper will show that computer-based training using sound visualization technique can be effectively implemented to Korean EFL learners to produce and perceive English pitch and durational contrast for words in sentences as well as for words in isolation.

가설 설정

In summary, the questions addressed by this study were: (1) Is production training that provides f0 contours as visual feedback effective in assisting the overall acquisition of English pitch and duration contrasts? Does the overall improvement of the trained group from the pre- to the post-tests differ from the improvement of the control group? (2) Having participated in the training that provided isolated words, phrases, and sentences, does the trained group improve its test scores in both word-inisolation and word-in-sentence contexts? (3) Does this production training have effects on both production and perception? If production learning occurs, is it accompanied by perceptual learning?
In addition to our main research question, we address three secondary questions: (2) Can training with synthetic speech stimuli help Korean listeners learn to ignore duration differences between vowel pairs and pay attention to vowel quality instead? (3) Will trainees generalize newly acquired knowledge about vowels to new speech tokens and new speakers? And (4) Will new knowledge about vowels acquired from training be retained after the training has stopped?
In addition to our main research question, we address three secondary questions: (2) Can training with synthetic speech stimuli help Korean listeners learn to ignore duration differences between vowel pairs and pay attention to vowel quality instead? (3) Will trainees generalize newly acquired knowledge about vowels to new speech tokens and new speakers? And (4) Will new knowledge about vowels acquired from training be retained after the training has stopped?
In summary, the questions addressed by this study were: (1) Is production training that provides f0 contours as visual feedback effective in assisting the overall acquisition of English pitch and duration contrasts? Does the overall improvement of the trained group from the pre- to the post-tests differ from the improvement of the control group? (2) Having participated in the training that provided isolated words, phrases, and sentences, does the trained group improve its test scores in both word-inisolation and word-in-sentence contexts? (3) Does this production training have effects on both production and perception? If production learning occurs, is it accompanied by perceptual learning?
In summary, the questions addressed by this study were: (1) Is production training that provides f0 contours as visual feedback effective in assisting the overall acquisition of English pitch and duration contrasts? Does the overall improvement of the trained group from the pre- to the post-tests differ from the improvement of the control group? (2) Having participated in the training that provided isolated words, phrases, and sentences, does the trained group improve its test scores in both word-inisolation and word-in-sentence contexts? (3) Does this production training have effects on both production and perception? If production learning occurs, is it accompanied by perceptual learning?
The first question of interest was whether the trained group improved its overall test scores from the pre- to the post-tests significantly more than the control group. The trained group of subjects showed a larger improvement in their overall test scores (including production and perception tests) from the pre-test (52.
A number of studies have shown that perception and production of words or segments in isolation is different from that of words in sentence contexts (Kalikow &Swets, 1972; Krashen & Terrell, 1983; Leather, 1983; Munro & Derwing, 1995). The second question in this study, therefore, was whether production training that focuses on words embedded in phrases or sentences has an effect on trained subjects' performance in isolated word and sentence contexts.

제안 방법

Subjects were trained individually in the language laboratory of the Kongju National University. In each of the 10 training sessions, the subjects opened each model audio file, listened to the model (isolated word, word in phrase, or word in sentence), and watched the pitch pattern of the model in real time. Then they produced the utterance on another empty window, and the program overlaid the model pitch contours onto the subjects' pitch contours.
Production and perception tests were conducted before and after training. Twentyone words including pitch and durational contrasts were chosen for production tests, for example, /i/-/I/, as in beat vs.
For the perception tests of words-in-isolation, subjects first clicked a play button on a computer screen to listen to each word spoken in isolation. Subjects then chose by clicking one of the nine pitch patterns (HL, LH, HLL, LHL, LHH, HLLL, LHLL, LHHL, LHHH; H = high, L = low) of prosody graphs that matched the word they had just heard. This method required the ability to identify not only the pitch but also the duration of the words.
The activities using this Sound Visualization Technique leads students to predict, identify, sort, match, discover rules, and exchange information. The activities are intended to help students to become more aware of their own English pronunciation, stress and intonation, and ultimately speech rhythms, and those of native English speakers to analyze them to produce various features of pronunciation in relevant context, both visually and audibly.
The trained subjects' attention was directed to both pitch and duration of the utterances they practiced producing with the Sound Visualization Technique. The focus of the training was for subjects to match the f0 contours of words, phrases, and sentences to those of the models. Minimal word pairs were only implicitly used in phrases and sentences in 7 out of 10 sessions.
, scores on the nine words that appeared in training were excluded from the overall production scores reported earlier). The production scores on the 12 novel words by the trained versus control groups were calculated for both the pre- and the post-tests. The results were comparable to those on all production words.
The structure of the experiment was as follows: one group of native Korean speakers was trained to produce English pitch and durational contrasts in isolated words and in words embedded in phrases or sentences. This trained group took a pretest (before training) and a post-test (after training) in both production and perception.
Then they produced the utterance on another empty window, and the program overlaid the model pitch contours onto the subjects' pitch contours. The subjects repeated this procedure, producing the same material until their pitch contours matched the model contours. The subjects were able to listen to model tokens whenever they wanted.
First, the subjects who participated in the production training improved their overall test scores significantly more than the control subjects did. The trained subjects' attention was directed to both pitch and duration of the utterances they practiced producing with the Sound Visualization Technique. The focus of the training was for subjects to match the f0 contours of words, phrases, and sentences to those of the models.
These pairs and triplets of words were presented in isolation or embedded in phrases or sentences. The training consisted of 10 training sessions. Sessions 1-3 presented a total of 47 words in isolation.
Two of the subjects (one in the training and one in the control groups) had studied French in high school for less than 4 years, but the others had no experience of studying any other foreign languages. The training group participated in a pre-test, training, and a post-test, and the control group took the pre- and the post-tests at the same time as the training group. The control group, in addition, was given the opportunity to participate in the training after the post-test although their training data were not part of the present experiment.
In each of the 10 training sessions, the subjects opened each model audio file, listened to the model (isolated word, word in phrase, or word in sentence), and watched the pitch pattern of the model in real time. Then they produced the utterance on another empty window, and the program overlaid the model pitch contours onto the subjects' pitch contours. The subjects repeated this procedure, producing the same material until their pitch contours matched the model contours.
The structure of the experiment was as follows: one group of native Korean speakers was trained to produce English pitch and durational contrasts in isolated words and in words embedded in phrases or sentences. This trained group took a pretest (before training) and a post-test (after training) in both production and perception. A control group also took the same pre- and post-tests, but did not participate in training.
Thirty English words were produced both in isolation (word-in-isolation context) and in carrier sentences (word-in-sentence context). Thus, there were a total of 60 utterances (30 words2 speakers) in each of the four tests (the word-in-isolation and the word-in-sentence tests in both the pre- and the post-tests), randomly ordered and divided into six blocks. For the perception tests of words-in-isolation, subjects first clicked a play button on a computer screen to listen to each word spoken in isolation.
It may not be clear to L2 learners whether a pitch rise or fall is phonetically relevant. To remedy such problems, a handout was made for each session, in which explanation was given on how to pronounce the utterances using prosody graphs. A prosody graph (Fig.

대상 데이터

Perception tests also consisted of a pre-test and a post-test. Each test contained 30 novel words that did not appear in training. The perceptual stimuli were recorded by two female native speakers of English.
Eight native speakers of Korean (5 males and 3 females) who were taking a secondyear English Conversation course in the department of English Education at the Kongju National University (50 min a day, five times a week) participated. The participants were randomly assigned to a training and a control group (n = 4 each).
Session 4 presented words in 15 short phrases, and Sessions 5-10 presented words in a total of 57 sentences of different length. The speech materials were recorded by four native speakers of English (2 males and 2 females), as models for participants' production training. The materials were all recorded on DAT cassettes, and then digitized at 20 kHz on a PC.
The training materials included 33 pairs and 2 triplets of English words contrasting in pitch and duration. These pairs and triplets of words were presented in isolation or embedded in phrases or sentences.

데이터처리

The percent correct scores of the word-in-isolation and the word-in-sentence tests in pre- and post-tests were calculated. These perceptual scores and the production intelligibility scores (i.e., percent correct scores of the native English judges' identifications of the subjects' production) were examined in a mixed design analysis of variance (ANOVA). The factors were Group (trained vs.

성능/효과

In conclusion, the findings in the present experiment suggest that the production training method developed and tested in this study was effective in enabling native Korean speakers to produce and perceive English pitch and durational contrasts. This exploratory experiment, however, motivates further studies to address various theoretical and pedagogical issues on the production and perception of L2 contrasts.
The results obtained in this study support positive responses to all four of the research questions posed. The most important outcome was that the trainees’ perception improved significantly on all of the phonetic contrasts on which they were trained, whereas a control group showed no such improvement. This study yielded three major findings.
The training resulted in the increased ability to perceive the pitch and durational patterns of the words, and the increased intelligibility of the subjects' production as evaluated by native English judges. The trained group improved on the practiced words but also shown a significant gain in their testing on novel words.

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증

Teaching Pronunciation Using Sound Visualization Technology to EFL Learners 원문보기

Abstract ▼ AI-Helper

주제어

AI 본문요약
AI-Helper

문제 정의

가설 설정

제안 방법

대상 데이터

데이터처리

성능/효과

이 논문을 인용한 문헌

저자의 다른 논문 :

관련 콘텐츠

원문 보기

원문 URL 링크

이 논문과 함께 이용한 콘텐츠

AI-Helper ※ AI-Helper는 오픈소스 모델을 사용합니다.

선택된 텍스트

연합인증

Teaching Pronunciation Using Sound Visualization Technology to EFL Learners 원문보기

Abstract ▼ AI-Helper

주제어

AI 본문요약 엑셀 다운로드 AI-Helper

문제 정의

가설 설정

제안 방법

대상 데이터

데이터처리

성능/효과

이 논문을 인용한 문헌

저자의 다른 논문 :

Min, Sujung (6)

관련 콘텐츠

원문 보기

원문 URL 링크

이 논문과 함께 이용한 콘텐츠

AI-Helper ※ AI-Helper는 오픈소스 모델을 사용합니다.

선택된 텍스트

AI 본문요약
AI-Helper