I. Introduction

As a form of entertainment, song has been with us from early prehistoric times. As a form of language use, song has more recently become a tool for teaching English in the ESL/EFL (hereafter, EFL) classroom. The use of songs in the EFL lesson plan has been mainly for the variety they bring to the lesson and the enjoyment and motivation they bring to the EFL student. Language instructors have in general rated their use of song in the classroom as successful because of the stimulation and participation they so easily produce. Because the main use of song has been for entertainment, opponents dismiss its use in the EFL classroom as not being of instructional value. Little research has been done to determine how effective, if at all, the use of song is in teaching different aspects of English. This paper presents the results of experiments designed to determine the effectiveness of song in the learning of both lexical items and grammatical structures.

II. Review of the Literature

Research that has been done on the use of songs in teaching EFL is limited. Most of the material that is available concentrates on describing different ways of using songsin class. In this literature, the effectiveness of songs in teaching EFL is not seriously questioned, and their ability to capture the attention of the class seems to be the strongest justification for their use. Discussing why popular songs should be used, Dale Griffee (1992) states that "no one knows exactly why songs are powerful" (p. 4) and goes on to suggest that it is likely because they are non-threatening, are a satisfying art form, and are able to affect our emotions. However, he admits that although "all these reasons have been suggested, the answer remains a mystery" (p. 4).

Domoney and Harris (1993) support using pop music because it is a way of involving the class in "meaningful, enjoyable, and collaborative tasks" (p. 234). Nambir (1993) states, as the first reason for using songs, that "songs deal with the whole realm of human emotions and experiences" and appeal to all (p. 336). Lo and Li (1998) find the use of songs most valuable because they are motivating and "teachers and students alike find singing songs entertaining and relaxing" (p. 8). In giving a rationale for using songs in teaching EFL, the most often quoted reasons are those presented by Eken (1996, p. 46):

To present a topic, a language point, lexis, etc.

To practice a language point, lexis, etc.

To focus on common learner errors in a more indirect way.

To encourage extensive and intensive listening.

To stimulate discussion of attitudes and feelings.

To encourage creativity and use of imagination.

To provide a relaxed classroom atmosphere.

To bring variety and fun to teaching and learning.

These very same reasons could be given as support for using an interesting passage by an author or poet as realia. Treatment of the degree of effectiveness of the use of song in relation to more established forms of realia presentation is noticeably absent.

III. Method

The aim of this study is to answer the question: Are songs an effective means of learning vocabulary items and grammatical structures in comparison to an oral narrative presentation of the same material? This question is examined in terms of subject performance on a Korean-to-English translation test containing Korean equivalents of the English song lyric items being examined. For the investigation of this question, the following method was employed.

1. Subjects The subjects of this study were students enrolled at Chosun University in various English oral skills courses. The students were freshmen through seniors of mixed speaking proficiency, ranging from low-intermediate to advanced levels of proficiency, and were enrolled in six different courses. The students were taking these courses as either free-electives or major-electives. Since the courses were being taken because the students wished to, the motivation to learn was somewhat higher in comparison with the average required course.

2. Context of Investigation The present study was conducted during the regular class time and in the regular classrooms used by the subjects for their courses, with classes meeting for two or three 50-minute sessions per week. The courses were taught by native English speaking instructors, but in all cases, this investigation was conducted by the author. The various classrooms involved were comparable in their facilities and environment.

3. Design The six classes of subjects were divided into three test groups (two classes in each group), which ranged is size from 50 to 60 students each. One group served as the control group and was only tested, another group was exposed to only the lyrics in narrated form, and the third group was subjected to the song.

4. Materials The following materials were used to conduct this study:

a. A handout of the lyrics to the song "As Tears Go By," written by Mick Jagger and Kieth Richards (1964).

b. A cassette tape recording of the song "As Tears Go By" performed by The Rolling Stones (1965).

c. A handout of a random cloze test made from the lyrics of "As Tears Go By."

d. A Korea-to-English translation quiz consisting of three test items taken from the song lyrics. These test items were the Korean equivalents of:

Question 1: All I hear is the sound of rain falling on the ground.

Question 2: My riches can't buy everything.

Question 3: Smiling faces I can see but not for me.

The song and its lyrics were selected because that song is slow and the words are enounced. In addition, because of its date of release, it is a song that is not familiar to most college-aged Koreans. In the cloze test, approximately one half of the words were omitted. These included hear, riches, smiling faces, and see from the lyrics used in the quiz questions above. In the three test items only the translation to English of the underlined parts were of concern. These items were selected because riches was expected to be an unfamiliar lexical item, the string smiling faces I can see was considered to be an unfamiliar grammatical structure, i.e., an inverted clause and all appeared in what was considered an unfamiliar structural environment and at the same time carried a meaning other than its most commonly used one.

5. Procedure Each of the three groups the control group, the lyrics group, and the song group were given an unannounced pretest near the beginning of the semester. This pretest was the translation quiz containing the three items described above. The students were informed that the quiz was part of a research project and would in no way affect their course grade. Sufficient time (approximately 15 minutes) was given for the students to complete the quiz. For the control group, the quiz was followed by regular class work. For the lyrics group, the quiz was followed by a discussion on the song lyrics. The lyrics were recited to the students a total of three times, and each line was repeated an additional three times by the instructor in the course of the discussion of the meaning of the lyrics. The song group heard the song three times as they were doing the cloze test. They also heard the lyrics repeated an additional three times by the instructor during the ensuing discussion on interpreting the lyrics.

One week later, each group was administered a posttest. This was the same quiz that was administered as the pretest. As before, the quiz was completely unannounced. The students had not been instructed to prepare for a future quiz nor were they advised to study the material presented and discussed earlier. This first posttest was followed four weeks later by a second posttest, which was identical in every respect to the pretest and the first posttest.

6. Data Analysis Because the quizzes were administered unannounced, there were a considerable number of students who were absent for at least one of the quiz administrations. If a subject was not present for one of the administrations, their test results for all quizzes were not used. Accordingly, test results were used from 30 subjects in the control group, 22 students in the lyrics group, and 32 students in the song group. Each of the three tests was scored with a possible score of one point for each question and a total possible score of three points for each test. Scores awarded were 0.0, 0.5, or 1.0 points, depending on the degree of accuracy of the underlined portions of the test items in 4d above, not on the accuracy of the test item as a whole. The raw score averages for each test by group and the percentage increase from one test administration to the next are as in Table 1:

Table 1. Average Group Test Scores and Percentage Change

Group

Pretest

Posttest 1

[% Increase]

Posttest 2

[% Increase]

Control

0.567

0.733

[29.3%]

0.750

[02.3%]

Lyrics 1

0.891

1.674

[87.9%]

1.652

[1.3%]

Song 1

0.539

2.171

[302.8%]

2.609

[20.2%]

The results of the tests were compared, for the three test items collectively and for each item individual

ly, using the t-test to determine if there was a significant difference resulting from the different treatments that the three groups received.

IV. Results and Discussion

1. Comparison of Test Scores Within Each Group First, the t-test was administered to compare the test scores of each test for each group. This showed that for the control group there was no significant difference between the pretest and posttest 1 scores or between the pretest and posttest 2 scores. However, there was a significant difference (p < .05) between the pretest and posttest 2 scores. Scores increased for each consecutive test that the control group was administered, as they also did for the other groups, with only one exception. The results showed that the average increase in scores from one test to the next was relatively small. The learning that did take place between test administrations can only be attributed to student interest in the test questions, as the subjects expressed no familiarity with the song or its lyrics when questioned after the last quiz.

The lyrics group showed a significant difference between the pretest and posttest 1 and between the pretest and posttest 2. No significant difference was recorded for the comparison of posttest 1 and posttest 2, indicating that any learning that took place from test-taking between these two tests was offset by forgetting of material learned over the four-week period between the two administrations. For the song group, significant differences were recorded between each test compared. This indicates that the amount of retention of material even after four weeks was significantly high. This contrast with the lyrics group can be accounted for by the song-stuck-in-my-head phenomenon (Murphey, 1992a,b) whereby the music of a song facilitates the repeating of the song's lyrics during a time of relative quite following the audition of a song.

2. Comparison of Test Scores Between Groups Comparisons were made between groups for total test scores for each quiz and for individual test questions on each quiz. Results of the t-test for the control group and the lyrics group showed no significant difference on individual question or total test scores on the pretest, but showed a significant difference for all but question 3 on posttest 1 and for all scores on posttest 2. Similarly, a comparison of the control and song groups showed no significant differences for any pretest scores but showed a significant difference for all posttest 1 and posttest 2 scores. This simply indicates that learning took place in both the lyric and song groups.

A t-test comparison of the lyric and song groups showed that there was a significant difference between the lyric group test scores and the higher song group scores in all areas of posttest 2 except question 1. This indicates that song is indeed more conducive to learning both lexical items such as riches and structural patterns such as inverted clauses. The lack of significant difference for test item 1 may be attributed to the fact that it contained both new lexical and new structural content too much new material to be learned from just a few exposures to the string of words.

3. Comparison of Song Group Question Scores by Test For the song group, a comparison of posttest 1 and 2 scores with pretest scores expectedly showed a significant difference for all three questions. However, a significant difference between posttest 1 and posttest 2 also appeared for test item 3. This indicates that song is particularly conducive to structural pattern learning.

V. Repeat Experiment

In Section III.5, it is stated that the song group was administered a cloze-type listening activity as part of the exposure to the song. The lyrics group was not administered this cloze activity. To remove this variable from the experiment and to attempt to repeat the results of the original experiment, a second experiment was conducted during the spring semester of 2003. The materials and procedures were the same for the lyrics and song groups in this second experiment (hereafter, lyrics2 group and song2 group, respectively), except that the same cloze-type test was administered to the lyrics2 group, with the lyrics read by the instructor three times, as was administered to the song2 group. The song2 group consisted of 31 students in two different classes, and the lyrics2 group was comprised of 23 students in three classes.

The results of data analysis revealed that lyrics2 and song2 groups performed comparable to the original lyrics and song groups. That is, the percentage increases in posttest 1 average scores over pretest scores and posttest 2 scores over posttest 1 scores were higher for the song2 group than for the lyrics2 group (Table 2). This was also true for the original song and lyrics groups.

Table 2. Average Group Test Scores and Percentage Change

Group

Pretest

Posttest 1

[% Increase]

Posttest 2

[% Increase]

Control

0.567

0.733

[29.3%]

0.750

[02.3%]

Lyrics 1

0.891

1.674

[87.9%]

1.652

[1.3%]

Lyrics 2

0.652

2.022

[210.1%]

2.152

[06.5%]

Song 1

0.539

2.171

[302.8%]

2.609

[20.2%]

Song 2

0.435

1.516

[248.5%]

1.726

[12.2%]

Test score averages for each group are charted in Fig. 1, and the percentage of change in test score averages is illustrated in Fig. 2 below.

VI. Conclusions

This study, though admittedly very small in scope, leads to the conclusion that a song-based language-learning activity can be a very effective English teaching technique. It not only equaled but surpassed the similar activity lacking song. New vocabulary and new grammatical structures are both more easily learned when presented in songs than in narrated form. This is due to the-song-stuck-in-my-head phenomenon (T. Murphey, 1990, and A. Cash, 2001). Though more extensive study is necessary to verify these preliminary findings, the indications are that incorporating songs into EFL lesson planning is not only an effective, but a highly effective, teaching technique.

Fig. 1. Average Test Scores by Group



Fig. 2. Percentage Change in Average Test Scores by Group



References:

Cash, A. (2001). Can a tune get stuck in your brain? Healing Music E-zine, 1(4). Retrieved on February 25, 2003, from

http://www.healingmusicenterprises.com/oct2001.html

Domoney, L., & Harris, S. (1993). Justified and ancient: Pop music in EFL classrooms. ELT Journal, 47(3) 234-241.

Eken, D.K. (1996). Ideas for using songs in the English language classroom. English Teaching Forum, 34(1) 46-47.

Griffee, D.T. (1992). Songs in action. Herfordshire, England: Prentice Hall International.

Lo, R.S.M., & Li, H.C.F. (1998). Songs enhance learner involvement. English Teaching Forum, 36(3) 8-11.

Murphey, T. (1992a). Music and song. Oxford: Oxford University Press.

Murphey, T. (1992b). The discourse of pop songs. TESOL Quarterly, 26(4), 770-774.

Murphey, T. (1990). The song-stuck-in-my-head phenomenon: A melodic din in the head. System, 18(1), 53-64.

Nambiar, S.A. (1993). Pop songs in language teaching. In J.W. Oller, Jr. (Ed.), Methods that work (pp. 335-338). Boston, MA: Heinle and Heinle.