Florida State University Libraries

Transcription

1 Florida State University Libraries Electronic Theses, Treatises and Dissertations The Graduate School 2007 The Effect of Song Translation vs. Non- Native Original Language Performance in Japanese on Emotional Response of Japanese Participants Makiko Chiashi Follow this and additional works at the FSU Digital Library. For more information, please contact

2 THE FLORIDA STATE UNIVERSITY COLLEGE OF MUSIC THE EFFECT OF SONG TRANSLATION VS.NON-NATIVE ORIGINAL LANGUAGE PERFORMANCE IN JAPANESE ON EMOTIONAL RESPONSE OF JAPANESE PARTICIPANTS BY MAKIKO CHIASHI A thesis is submitted to the College of Music in partial fulfillment of the requirement for the degree of Master of Music Degree Awarded: Spring Semester, 2007 Copyright 2007 Makiko Chiashi All Rights Reserved

3 The members of the committee approved the thesis of Makiko Chiashi defended on December 5, Jayne M. Standley Professor Directing Thesis Clifford Madsen Committee Member Dianne Gregory Committee Member The Office of Graduate Studies has verified and approved the above named committee members. ii

4 ACKNOWLEDGEMENTS I would like to show my gratitude for all the people who have brought so much happiness and excitement to my life. Especially, I would like to thank. Dr. Standley for your wisdom and guidance through out the course of study at Florida State University. Thank you for making the process of writing the thesis enjoyable for me! Dr. Madsen, Dr. Darrow and Gregory for all the great lessons passed on me. Mike Silverman for polishing my writing. Betty Brown for helping me with the statistics. Sabina Barton for agreeing to make recordings of the song. You are an amazing. George Recker for all the lessons and chatting I had with you. I am starting what I always had in the back of my head now. And I love it! Matt Miller for taking time to read my writing. I owe you a lot of gyoza! And my family and friends all over the world (you know who you are!) for all the support and laughs! iii

5 TABLE OF CONTENTS List of Tables v Abstract vi CHAPTER 1 - INTRODUCTION 1 CHAPTER 2 - REVIEW OF LITERATURE 3 Song Preference 3 Emotion and Music 4 Assessing Response to Music 6 Melody and Lyrics 7 Song Translation 8 Multicultural Music Education 9 Diversity and Multicultural Aspects in the Music Therapy Curriculum 11 Lyric Performance by Non-Native Speakers 13 Purpose Statement 14 Hypothesis 14 CHAPTER 3 - METHOD AND PROCEDURE 15 Participants 15 Design 16 Music 17 Procedure 17 CHAPTER 4 - RESULTS 19 CHAPTER 5 - DISCUSSION 26 Limitation of the study 28 Implication for Music Therapy 29 Suggestion for Future Research 29 Conclusion 29 APPENDIX A - Human Subjects Committee Approval 31 APPENDIX B - Informed Consent Form 33 APPENDIX C - Questionnaire 35 APPENDIX D - Raw Data 39 REFERENCES 42 BIORAPHICAL SKETCH 48 iv

6 LIST OF TABLES 1. Demography of the Participants Question 1. Soothed by the Song Question 2. Disturbed by the Song Question 3. Strength of Emotional Response Question 4. Listen When Homesick Question 5. Hevner s Adjective Scale Descriptions of Rating for English and Japanese Version T-test for the Japanese Version by Gender T-test for the English Version by Gender Correlation between All the Ratings T-test for Hevner s Adjective Groups Comments 24 v

7 ABSTRACT The purpose of this study was to examine whether music therapists should sing in a nontranslated, inexperienced foreign language to match the song preference of a client or patient from a different cultural background or whether they should sing the translated English version of the same song. The participants (N =34) were Japanese people who were student at a large state university in the southeast. The participants listened to two song versions; one in original Japanese version and the other in English translated version of a Japanese traditional song sung by a music therapy student whose native language was English. A questionnaire was used to measure the participants emotional responses to the songs. The participants answered by using a seven point Likert-scale about how much they were soothed or disturbed by the songs, how strong their emotional response was to the song, how much they would like to listen to the song when feeling homesick, and their description of the song by choosing an Hevner s adjective group. A General Linear Model repeated measure test was applied to the research data to determine whether there was any differences between the participants responses to two versions of the songs, and whether the subjects length of stay in the United States had any effect. The statistical analysis revealed that there were no significant differences in any questions but one. However, there were several correlations between the questions. This study supported the hypothesis that the participants emotional response would not be affected by song translation or non-native original language performance. However, when performing a song with an unfamiliar language, it was recommended to consult with a native speaker for the language to ensure high quality of performance. Further research regarding music therapy with multi-cultural clients or patients is strongly suggested to better treat the growing population in the United States. vi

8 CHAPTER I INTRODUCTION There has been a rapid increase in the number of the people who are from other countries in the United States. Immigration is the number one cause of population growth. It has been reported that 2.25 million people come to the United States as both legal and illegal immigrants annually. Moreover, women among the immigrant population give birth to 750,000 babies a year. As immigrants start their lives in the United States, they bring different cultures, languages and aspects to American society. As minority cultures become visible in the United States, nearly all aspects of society, such as business, education, and the way of living are influenced. For example, businesses may be run by multi-lingual staffs (Castelluccio, 1996), or the products may have instructions in many languages other than English. We see advertisements and entertainment that are specifically targeted for minority populations especially, in particular, for the Hispanic audiences (Segal & Sosa, 1983). Moreover, there are a greater number of education programs for students who are non-native English speakers. Now more than ever, the minority population in the United States is a focus of attention. According to United States Census Bureau, about 20 % of the United States population does not speak English at home (2003). Languages other than English are dominantly spoken at homes in Arizona, 26.4%; California, 41.0%; Florida, 24.2%, New Mexico 36.1 %; New York, 27.5%; and Texas, 32.5%. In the major cities and urban centers in the United States, there is a high percentage of people who speak English less than very well : Chicago, 17.1 %; Dallas, 26.3%; Houston, 24.0 %; Los Angels, 31.1%; and New York City, 24.6% (2003). Due to this increase in diversity, music therapists treat minority groups as clients or patients. The more diverse the country becomes, the more likely minority clients or patients are to increase. The probability that music therapists in the United States will have clients or patients from minority cultural backgrounds in sessions is high and it is expected to continue to increase (Toppozada, 1995; Moreno, 1988; Darrow & Molloy, 1998). 1

9 In the American Music Therapy Competencies issued by the American Music Therapy Association, there is only one competency criterion that mentions culture; develop and maintain a repertoire of music for age, culture, and stylistic differences. However, AMTA does not specifically require schools that offer music therapy programs to incorporate in their curricula the necessity for students to learn to play music from different cultures such as non-western culture. It is emphasized in the music therapy literature that it is important to use live, age appropriate, and preferred music for the clients to have the best outcomes from the session (Krout, 2000; Walworth, 2003). It is expected that there will be numerous instances where a therapist may not be familiar with songs and styles a client prefers particularly if that individual is from minority population. Also, such preferences may be in a language with which the therapist is unfamiliar. In such instances, what is the best approach for a music therapist? This study investigates possible approaches for a music therapist to best serve with clients or patients from other cultures. 2

10 CHAPTER II REVIEW OF LITERATURE Song Preference Music therapy literature often suggests the use of music familiar to the client for therapy sessions to obtain potential therapeutic outcomes (Robb, 1999; Walworth, 2003; Lord & Garner, 1993; Staum, 1993). As an example of therapeutic interventions, rap or hip/hop therapy is used in urban areas in the United States for the adolescents who are at risk, since the contents of rap/ hip hop music are the type of music containing themes most similar to their lives (DeCarolo, ; Tyson, 2002). Music preference can be influenced by age, socio-economic status, gender, and cultural background (Herrington & Capella, 1994). Mashkin and Volgy (1975) found that among college students who preferred country music most had a more traditional viewpoint of roles for sexes than did those who preferred folk music. Russell (1987) measured familiarity and likeability of college students of popular music on the sales chart. Results showed that the longer the songs stayed on the sales chart, the more familiar they were to the students. However, there was no proof to support the notion that the students preferred the songs that remained on the carts for longer periods of time. Therefore, the causation between familiarity and likeability was not clear from this research. The result of the research by Hargreaves, Messerschimidt and Rubert (1986) contradicts previous research results. Fifty-four undergraduate students listened to familiar and unfamiliar excerpts from two music genres, classical and pop. They were to rank the quality and likeability of each excerpt by using a Likert-scale measurement; 1 being dislike, 7 being like and 1 being poor quality, 7 being good quality. No matter how much musical training the subjects had, they ranked familiar music higher in both likeability and quality. For music to exist, there must be a performer. Even though it may be a popular and well-liked piece of music, listeners preference toward that music might differ depending on who performs or the quality of performance. Over 200 middle-school and college students 3

11 participated in research aimed at determining if the race of a performer impacts the musical preference of the listener (McCrary, 1993). The participants listened to a variety of music sung by both white and black singers. The races of the singers were not told to the participants. It appeared that while the participants who were Black had strong preferences for the songs that they thought were sung by Black singers, the participants who were White had broad-mindedness for their preferences. Even though the song was familiar to a patient or client, if the therapist does not execute the song well it might not sound familiar to him/her. Moreover, it may lose effectiveness for therapeutic outcomes because of poor quality of performance. Baker (2001) suggested that recorded music may sound the most familiar to patients because it was almost impossible to recreate the quality and characteristics of the original recording, especially in clinical settings. Relaxation or reduction of anxiety is often addressed as an objective in medical music therapy sessions. In such cases, patients preferred music is used most often. Walworth (2003) compared the effectiveness in reducing anxiety by 1) patients preferred genre or artist, 2) a particular song of patients preference, and 3) no music. Listening to music lowered patients anxiety more than did no music, but the difference in effectiveness between two treatments was not significant. Furthermore, listening to a preferred relaxing song recording selected by college students or no music promoted more relaxation when compared with listening to hard rock music. Even among students who preferred to listen to hard rock music on a daily basis, they were not more relaxed than before they listened to hard rock music (Burns et al., 2002).Therefore, this research suggests that a preferred song of a client or patient is not always the most effective. Songs chosen for music therapy sessions must be appropriate for the nature of the particular objective. Emotion and Music People often experience a change or surge of emotion while listening to music or even, unexpectedly, afterwards. The emotion can be positive or negative depending on the listener s 4

12 mood, style of music, or any memories associated with the music (Abeles & Chung, 1999). According to Gabrielsson, music can be felt to express emotion as well as to arouse emotion (2001, p. 425). What triggers any changes in listeners emotion through music has been of interest to a number of researchers. Sloboda and Juslin (2001) noted there were two elements that affect listeners emotion; one was musical qualities (character, structure, and styles etc.) and the other was associated memories or experiences from the past. Musical elements such as pitch and tempo (Hevner, 1937) and pitch and rhythm (Schellenberg, Krysciak, & Campbell, 2000) were examined to determine if they are responsible for eliciting the emotional change in listeners. Hevner s results showed positive emotions such as happiness and excitement were associated with music that had faster tempos in moderate pitches and dissonant harmonies. Conversely, negative emotions such as sadness were emitted by music with slow tempos in low pitches. In the research by Shellenberg and colleagues, the participants rated music with frequent pitch and rhythmic changes as happy. When they rated sad music, frequent pitch change was still the element that made the music sound sad, but rhythm change was shown not to have any affect. Other researchers focused on characteristics of the listeners as the elements responsible for emotional response to music. Musicians versus non-musicians had similar emotional responses to music regardless of their musical training (Bigand, Vieillard, Madurell, Marozeau, & Dacquet, 2005; Kallinen, 2005; Madsen, Brynes, Capperella-Sheldon, & Brittin, 1993). In the study by Kallinen and colleagues, emotions such as joy and sadness were the easiest emotions elicited by musical excerpts. Iwaki, Tanaka, and Hori (2003) and Iwanaga and Moroki (1999) examined effect of listeners preferred music on their emotions. Preferred music helped people to fall asleep faster if they were used to listening to the music before sleeping (Iwaki, Tanaka, & Hori, 2003). This may be a result of classical conditioning whereby listening to music is associated with falling asleep for those who are accustomed to listening to music before bed. Iwanaga and Moroki (1999) investigated how people s emotional responses (relaxation, vigor and tension) as well as physiological responses (heart rate, respiration and blood pressure) were affected by preferred music. While preferred music was effective in lowering the tension of the 5

13 participants, exhilarating music aroused feelings of vigor and tension in the participants. In terms of reducing the level of tension of the subjects, what mattered most was the type of music, not the preference. Listeners cultural backgrounds have also been taken into consideration with regard to difference in emotional responses in research studies. Participants who had western cultural backgrounds were exposed to music from Hindustani, North India (Backwill & Thmopson, 1999). Even though they were totally unfamiliar with the tonality and the instrumentation from this culture, they were able to respond to emotional content in accordance with the intention of the songs. A similar result was found in the study by Gregory and Varner (1996). The participants from both western culture and Indian culture listened to a variety of excerpts from western classical, Indian classical and New Age music. Using a modified version of Hevner s adjective checklist, the listeners chose adjectives to match the emotion they felt from listening to the excerpts. There was a subtle disagreement in choice of adjectives between western and Indian; however, overall differences in their cultural background did not affect the choice of adjectives. Assessing Response to Music The tools used in measuring people s emotional response to music can be divided into three categories; self-report, expressive behavior and psychological measurement (Sloboda & Juslin, 2001). Self-report is the most frequent measure used in research. Among measures of self-report, the most common tool to appear in music literature is an equal interval scale, such as a Likert-scale. The scale has a straight line with numbers that are placed in equal intervals. The participants are asked to choose a number on the line to show how much they like or dislike, agree or disagree, and so on. This scale is used widely when measuring the intensity of the subjects attitude toward music (Wapnick, 1976). Hevner s adjective list (1935, 1936) is one of the most widely used tools to examine response to music in terms of mood. It consists of eight groups of adjectives, totaling 67 words. Adjectives in each group share closely related meanings (Radocy & Boyle, 2003, p. 328). It has 6

14 been used for studies that examined the mood of music (Gregory & Varey, 1996), listeners responses to music (Hevner, 1937), and listeners perceived mood states (Goins, 1998). The check list has been modified by Hevner herself (Fransworth, 1969) and other researchers (Gregory & Varey, 1999) so that it can fit with particular study designs. Melody and Lyrics Music can be found everywhere in daily life. There are few TV shows without music, people unconsciously hum a tune when they are happy, and stores play music that matches the images of their brands (Herrington & Capella, 1994). Often music is tied to a certain product in a commercial or to a slogan since it can work as a conditioned or unconditioned stimulus in a psychological phenomenon of classical conditioning (Murray & Murray, 1996). When listening to music with lyrics such as pop music, art songs, and catchy phrases on TV, there are two elements to listen to; melody and lyrics. In the brain, speech and singing are generated in the right and left hemisphere laterally (Callan et al., 2006). A study by Brown, Martinez and Parsons (2006) showed that the brain processes a melody part of a tune in the right hemisphere and the lyrical part of a tune in the left hemisphere. They are processed in the brain side by side, but yet as two different kinds of information. When changing one s mood with music, what influenced the mood was not the lyrics of music, but the melody (Sousou, 1997). In another study that examined if the melody and lyrics of a song shared an equal role in changing one s mood, it was found that melody was more powerful in terms of changing one s mood (Ali & Peynircioglu, 2006). It also appeared that lyrics played a more important role in changing one s mood when music was perceived to have negative emotions (sad or angry). On the other hand, melody was more influential when the emotion that was related to the music was positive (happy or calm). There has been public concern that heavy metal and rap music might reinforce problematic behavior among youth because of the aggressive or irrational lyrical contents. However, there is no concern about contemporary classical music using unusual beats and 7

15 irregular rhythms that might affect the mental state of the young people. The research that attempted to discover the causation between listening to heavy metal music or rap music and adolescents maladaptive behavior have not been successful (Gardstorm, 1999; Scheel & Westefeld, 1999). The results of Gardstorm s study suggested lyrics of those music genres might influence their feeling, but they do not trigger problematic behaviors. Rap music has been one of the most controversial genres of music since it emerged in the mainstream in early the 1980 s because of provocative content of the lyrics (Fenster, 1995). Often rap music contains lyrics about violence, drugs, and sex (Krohn & Suazo, 1995) that generate public concern for the listeners. However, Armstrong (1993) found that there were noticeable similarities between the lyrics of rap and country music. Both genres share negative topics such as violence, arrest, assault, poverty, and masculinity. Even though rap and country music are similar in context, only rap music seems to receive negative attention from society. When rap lyrics were presented as country or folk songs or songs that were sung by white artists, the perceptions toward the lyrics were more positive than when they were presented as rap music or music sung by Black artists (Fried, 1996 & 1999). Therefore, when a piece of music is not presented in the original complete form that includes melody, lyrics, artist and performance, people s perception toward the song may vary. Song Translation In translation, the translation is everything (Yoshimoto, 2001, p. 51) Translation is more than a process of exchanging a word with another word that has the same meaning in a different language (Sakai, 2006). If it is a bad translation, it can lose the flow, tone and even message of the original text. A well-known Japanese novelist who has had many of her works translated into other languages noted, a professional is truly amazing-they will take my words and spin a magic in their own language that replaces them with the appropriate wordings and expressions in the reader s language (Yoshimoto, 2001, p. 51). The more people want to exchange information, and explore other s culture and art, the more translation will be in demand 8

16 (Hirano, 2001). Both Yoshimoto and Hirano are famous novelists in Japan. For Japanese novels to be shared abroad, translation is necessary. However, there are other forms of art that do not require translation to transcend borders such as fashion and music. Often music goes beyond its own country s borders without getting any help from translators. Needless to say, instrumental music is universal. However, even for the music with lyrics, translation is not always necessary. The recent phenomenon of global rap/hip hop music is a notable example (Pennycook, 2003). Rap/hip hop was born in the 1970 s in the Bronx among black people as a form to express their life and culture (Wikipedia, 2006). Ever since the birth of rap/hip hop culture in the Bronx, it has expanded its popularity to white populations within the United States (Sullivan, 2003), and even to Asia and Europe (Bennett, 1999; Pennycook, 2003). If rap/hip hop is a unique form of cultural art among Black people in the United States, how has it become enjoyed and appreciated among other racial groups? Sullivan (2003) explains that the gap between races that enjoy rap music has narrowed. She examined the attitude of Black and white adolescents toward rap music in the United States. One of the answers she obtained from a white adolescent about why he listened to rap music was because of the good beat. While Black adolescents tend to listen to the messages contained within the lyrics that may relate to themselves, white adolescents tend to listen to the beat of the music. Thus, one form of music can be appreciated by different racial groups since they may be listening to different elements of the music. Multicultural Music Education Elliot (1990) says, culture is not something that people have, it is something that people do. In the United States, there are people from various backgrounds and ethnicities. They live, talk, and eat in diverse ways which make the country a multicultural society. A particular culture may seem different or difficult to understand for the people from another culture. The same concept can apply to music. A certain kind of music from one culture can sound strange or different to the people from another culture. Music itself is often considered as a form of culture 9

17 (Elliot, 1990; Moreno, 1988). Therefore, there are as many different cultures as there are different kinds of music in this country. A group of people categorize music that they are accustomed to listening to as our music and, the other unfamiliar music as their music. This leads people to the tendency to think our way is the right way (Schmid, 1994) or our music is the right music. It is often emphasized how important it is to incorporate multicultural aspects into music education in the hope of students having a better understanding of other cultures. Music educators are aware that music programs must adapt to changes in the multicultural population of the classroom, which is representative of American society. However, music educators should not use music from different cultures in the classroom without any knowledge or explanation about the cultures to students (Campbell, 2002). To make the musical experience in a classroom meaningful, teachers can start exploring a variety of musical cultures in schools and the community. What makes multicultural music education successful is teachers willingness to educate themselves with unfamiliar territory of music from other cultures (Kelly & Van Weelden, 2004). When teaching music from another culture, it is important not to stray from the original form or authenticity. For example, using authentic instruments, singing in the native language, and having guest artists who specialize in the culture can be an effective lesson for the students (Goodkin, 1994). Edward (1998) conducted an experiment to find out what kind of approach is effective when teaching fourth graders Native American music. The students were divided into groups. Each group was assigned to a specific study station; one with large-group with authentic instruments, one with guest artists, one with authentic instruments, one with non-authentic instruments, and the other with regular curriculum. The result taken from students written responses showed that groups that had authentic instruments gave students effective and positive learning experiences. Moreover, the student group who had the guest artists responded that their learning experience could not have been taught by anyone else. According to Elliot (1990), music is a four-dimensional concept, meaning music involves a doer, a doing, something done, and a context in which the doing is done. These four dimensions are formed through culture in which music is born. This four-dimensional concept 10

18 can be applied to music therapy. Music therapy is a multi-dimensional therapy; there is a therapist, music that is played by a therapist, treatment that is received by clients or patients, and the therapeutic outcome from the session. If one of the four dimensions in music is lacking, the power of music may not be as effective as it should be. Diversity and Multicultural Aspects in the Music Therapy Curriculum As the United States becomes more diverse, the ways of business, advertisement, education and politics have adapted to target minority populations in this country. For instance, when calling numbers, there is almost always an option for Spanish, we see more commercials with minority actors or actresses on TV, and there are classes for students whose native language is not English in public schools. Also, the president makes speeches in Spanish in hopes of gaining popularity with the minority population. Universities offer courses focused on interacting with people from different countries in academic areas such as business, politics, international affairs, and social work. Johnson (1987) talked about a training program for counseling psychologists to learn how to work with clients from different backgrounds and ethnicities. He said being aware of cultural difference was the key to success for the psychologist when dealing with clients from various backgrounds. The research that was conducted by Chin (1996) indicates that among 534 higher education institutions that were approved by the National Association of Schools of Music, half of these schools did not have any multicultural music courses listed in the bulletins during the academic year. Chin believes that music professionals should take the lead in highly needed education in multiculturalism. Rarely are multicultural music classes specially designed for music therapy students. Often, they are expected to take such courses from general music classes (Chase, 2003). Chase emphasized the need for multicultural music courses specifically designed for music therapy students so that we as health care professionals can be effective when dealing with clients or patients from other cultures. Toppozada (1995) conducted a nation-wide survey among professional music therapists 11

19 with regard to their knowledge of, and attitude toward, relevant multicultural issues to determine if there should be training opportunities for music therapists or students to expand multicultural aspects in the profession. The results showed that about 87 % of the respondents considered culture in making music choices for the session, and about 78% agreed that music therapy students should have training in multiculturalism. Moreover, the majority of the respondents thought clients cultural background was relevant. Even though many of the respondents answered that they would choose songs for the session taking into consideration clients culture, it was unknown how they incorporated cultural aspects into actual sessions from this survey. A similar research study was conducted by Darrow and Molloy (1998) to further Toppozada s findings. The response to questionnaires indicated that many professional music therapists were aware of the fact that they needed to be sensitive to cultural backgrounds of clients or patients. However, there was doubt as to whether there was enough preparation to deal with cultural differences among patients or clients in their school work. Many subjects mentioned how they learned and deepened their multicultural knowledge through their experience. Requiring a class that teaches music therapy students multicultural music in the music therapy curriculum was advocated by Moreno (1998). She acknowledged the fact that there was a growing minority population in the United States and said it was inevitable that music therapists would have clients from other cultural backgrounds. Knowing a simple phrase in a foreign language would improve communication with people from different countries. Also knowing and using music from their culture will build initial rapport with clients or patients. Multicultural music can be used not only for people from a specific country, but also as unfamiliar music to clients or patient to add diversity to the session. By using music from another culture in a session with the clients or patients from that culture, Moreno expects that it can maximize the potential outcome of the session through sharing culture and values. 12

20 Lyric Performance by Non-Natives Speakers As the world becomes better connected using advanced technology in exchanging information through internet and media, listening to music from other countries has become easier. However, the western style of music that includes classical, jazz and country has long been a common music style in Japanese culture prior to the generation of internet. Watanabe (1982) noted the western style of music was widely accepted in Japan because it suits the modernized life style. Japanese traditional music using instruments such as the koto and shakuhachi does not seem to take root in Japanese life nowadays. Jazz music was introduced to Japanese society in the early 1900 s in the dance halls and through the radio, and American country music made its emergence in the mid 1900 s (Thompson, 1992). As much as Japanese people enjoy listening to the original versions, there are many Japanese singers who try to make their musical career by singing western style music in the original English language. It is common for Japanese people to go to the United States to pursue a career as singers to learn from the origins of the western style of music. When they come back to Japan to perform, they may sound authentic to the ears of Japanese people who are not proficient in English. However, their singing may not reach the standard of native English speakers. Thompson (1992) expresses his concern about well-known Japanese country singers who try to emulate American country songs in their repertoire. The secret of Nelson s popularity, however, is not so much his voice as his phrasing, the way he sells the song. All of this is lost when Whisky River becomes Whisky Liver, or when the singer confides that he just can t wait to get An Za Wo Agen (p. 35). He noted that the message of the song was carried by the lyrics. If the song is sung by a singer who has less than perfect pronunciation of the original language, the message may not make as strong an impact as when sung by a native. 13

21 Purpose Statement Research shows that when teaching or experiencing music from different countries or cultures, it is more effective to have authentic instruments and guest performers who specialize in the particular music from the culture (Goodkin, 1994; Edward, 1998). In music therapy settings, when the client or patient is not from the United States or whose musical preference is not typical of American culture, it may be unrealistic to have a guest musician or to prepare authentic instruments just for that session. In that case, a music therapist is expected to use the best approach possible. Using recorded music may be one possibility. However, live music is often the most desired form of music to generate maximum potential outcome from the session. Translation of music from another culture or another language may be an alternative. The purpose of the current study is to examine whether music therapists should sing in a non-translated, inexperienced foreign language to match the song preference of a client or patient from a different cultural background or whether they should sing the translated English version of the same song. Hypothesis The participants emotional response will not be affected by song translation or non-native original language performance. 14

22 CHAPTER III METHOD AND PROCEDURE Participants There were thirty-four Japanese people who volunteered to participate in this study. They were contacted via or phone by the researcher prior to the experiment. The majority were students who came from Japan to study at a large university in the southeast. The rest were Japanese people who worked or lived in the area. Among them, eighteen were female and sixteen were male. The degrees they sought varied from bachelor to doctorate. Their length of stay in the United States ranged from two months to twenty-four years. The average length of stay was 6.27 years. Table 1. Demography of the Participants Gender Length of stay in the US (yrs.) 1 F 1 2 F F 8 4 M M F 8 7 F 3 8 F M M M M 9 13 M 2 14 M 14 15

23 Table 1. (Continued) 15 M 3 16 M F F F F 9 21 M F F 8 24 F 9 25 F 6 26 M 4 27 F 5 28 F 5 29 M F M 5 32 F M M 15 Design The study used subjects as their own controls to respond to two music selections, one in original language sung by a non-native speaker of that language, and the same song translated into English and sung by the same person. The order of song presentation was counter balanced to control for order effect. 16

24 Music A well known Japanese song titled furusato (home town) was used. This song was composed in 1914 by Tatsuyuki Takano (lyrics) and Teiichi Okano (melody) to be sung in elementary level schools in Japan (Wikipedia, 2006). Since then, it has been a favorite of Japanese people over the generations. Currently, the tune widely appears in music text books for elementary school and middle school levels (Nakano, 2004). This song was even featured in the winter Olympics held in Nagano, Japan as a representative of Japanese traditional tunes. This particular song was chosen for the study due to the content of the lyrics that were considered suitable for living situations and moods of Japanese people who were in the United States. Two different versions of the song were used; one in original Japanese version, and one in an English translated version by Greg Irwin (Sakai, 2003). They were sung for recording for this study by a graduate music therapy student who was a native English speaker. The singer had no prior experience of speaking or singing in Japanese. The songs were taught to music therapy student by the researcher who is Japanese. The recording device used was the Olympus digital voice recorder DS-2. The songs were played through ipod nano with SONY headphones MDR-201 for the experiments. Procedure Each of the participants was tested individually in a quiet place such as the library, lounges on campus or their homes at their convenience. The experiment took an average of 10 minutes to complete. The participants were given the following before starting the experiment; a consent form, two blue questionnaires for evaluating the Japanese version, and a yellow questionnaire for evaluating the English version of the song. These evaluations included questions regarding their responses to each version of the song measured by a 7 point Likert-scale, a Hevner s adjective check list (Fransworth, 1969) and a sheet for any comments they had about the songs or the experiment. After they read the consent form, they listened to the 17

25 songs using the ipod nano and headphones with their comfortable volume without looking at a questionnaire. The participants were not informed who the singer of the recording was. After each listening, they were asked to fill out a questionnaire. They were allowed to take as long time as they needed to complete it. Half of the subjects listened to the Japanese version first and the English version next and the other half listened in reverse order to control for order effect. 18

26 CHAPTER IV RESULTS A General Linear Model repeated measures test was applied to determine 1) whether there was any difference between the participants responses to two versions of the songs, and 2) whether there was any difference in the scores for each question depending on the participants length of stay in the United States. Additionally, a t-test was applied to compare the difference between the scores for each question between the genders of the participants. Finally, another t-test was used to determine if there was a difference in the participants description of the Japanese and English versions of the song using Hevner s adjective lists. Additionally, the comments written by the participants after the experiment were examined. The results from a General Linear Model repeated measures test are shown in the following tables for each question. Table 2. Question 1. Soothed by the Song df Mean Square F Sig. Within Groups (Songs) Between Groups (Length of stay) Interaction (Songs x Length of stay) Error (Within) Error (Between)

27 Table 3. Question 2. Disturbed by the Song df Mean Square F Sig. Within Groups (Songs) * Between Groups (Length of stay) Interaction (Songs x Length of stay) Error (Within) Error (Between) * Statistically significant. Note. p< Table 4. Question 3. Strength of Emotional Response df Mean Square F Sig. Within Groups (Songs) Between Groups (Length of stay) Interaction (Songs x Length of stay) Error (Within) Error (Between) Table 5. Question 4. Listen When Homesick df Mean Square F Sig. Within Groups (Songs) Between Groups (Length of stay) Interaction (Songs x Length of stay) Error (Within) Error (Between)

28 Table 6. Question 5. Hevner s Adjective Scale df Mean Square F Sig. Within Groups (Songs) Between Groups (Length of stay) Interaction (Songs x Length of stay) Error (Within) Error (Between) The descriptive statistics of the ratings for questions on the English and Japanese versions are shown in Table 7. Table 7. Description of Rating for English and Japanese Versions English Japanese M SD M SD Q1: Soothingness Q2: Disturbance Q3: Emotional Response Q4: Homesick Q5: Hevner's Adj The statistical analysis revealed that there were no statistically significant differences between the two song versions for all the questions but one. There was a statistically significant difference between the ratings for two song versions for question 2 (Table 3). The Japanese version sung by the English speaking music therapy students was significantly more disturbing to Japanese persons. There were no statistically significant differences in the ratings for each question depending on the subjects length of stay in the United States (Table 2-5). Finally, there were no statistically significant interactions between songs and length of stay. Table 7 shows means for the ratings of the two versions of the songs. A t-test was used to compare the means of scores on the Japanese version and English 21

29 version by the participants gender. The descriptive statistics and results of a t-test for the participants gender and the ratings on the questions for the Japanese version in shown in Table 8, and for the English version in shown in Table 9 (Female, N=18; Male, N=16). Table 8. T-test for the Japanese Version by Gender Mean SD t df Sig. Female Male Female Male Q1: Soothingness Q2: Disturbance Q3: Emotional Response Q4: Homesick Q5: Hevner's Adj Table 9. T-test for the English version by Gender Mean SD t df Sig. Female Male Female Male Q1: Soothingness Q2: Disturbance Q3: Emotional Response Q4: Homesick Q5: Hevner's Adj There were no significant differences by the participants gender in rating the questions for either the Japanese or English versions. The correlations between scores on all of the questions from both versions were examined and shown in the Table

30 Table 10. Correlation between All the ratings E Q1 E Q2 E Q3 E Q4 E Q5 J Q1 J Q2 J Q3 J Q4 J Q5 E Q ** 0.46** 0.53** E Q ** -0.48** ** E Q3-0.49** ** * E Q * 0 E Q J Q J Q J Q * 0.02 J Q J Q5 1 ** Correlation is significant at the 0.01 level (two-tailed). * Correlation is significant at the 0.05 level (two-tailed). Several statistically significant correlations between questions were revealed. For the English version, the participants who rated high for question 1 ( how much were you soothed by the song? ) rated question 2 ( how much were you disturbed by the song? ) lower. Also, there was a negative correlation between questions 2 and 3 ( how strong was your emotional response to the song? ); and the questions 2 and 4 ( If you were experiencing being homesick, how much would you like to listen to the song? ) for the English version. The participants who rated high for the soothingness and the strong emotion to the song of the English version rated high for how much they wanted to listen to it when they were experiencing homesickness. There was only one statistically significant correlation between the questions for the Japanese version (questions 3 and question 4). The participants who had stronger emotional response to the song rated high for how much they wanted to listen to the song when they were experiencing homesickness. There were some significant correlations between the questions for the English version and the Japanese version. There was a positive correlation between both versions for question 2 indicating that the participants who were disturbed by the English version were also disturbed by the Japanese version. Similar correlations were found between question 2 and question 3 for the both versions. Finally, a t-test was used to determine whether there was a difference in the participants 23

31 choice of Hevner s adjectives group for the two versions of the song. There were no statistically significant differences between the subjects choice of adjective groups for the Japanese and the English versions of the song. Descriptive statistics and results from the t-test are shown in Table 11. Table 11. T-test for Hevner s Adjective Groups Mean SD t df Sig. J E J E Adj. Groups All participants were given an opportunity to make written comments about the experiment afterwards. 52 % of the participants made comments in either English or Japanese. All these comment were specifically about the songs. Table 12. Comments 38.9 % Negative comments about Japanese version (the accent / wrong pronunciation of Japanese) 38.9% Prefer listening to a singer who understands what he/she is singing about 16.7% Positive comments about English version (sound natural to the singer) 11.1% Prefer Japanese version (easy to listen to) Among those who made comments, 38.9 % noted the accents and wrong pronunciations in the Japanese version were bothersome while listening. The same percent of subjects noted that they would not want to listen to the singer who did not understand the words of the song. Over 16 % reported the English version spoke to them more than the Japanese version because the English version sounded natural for the singer. One of the participants commented the 24

32 English version sounded more soothing because it was clear the singer was singing with understanding of the emotions behind the song. However, 11.1 % of them noted they liked the Japanese version more since they could not listen to the whole music (melody and context) in the English version because of the effort to understand the words. 25

33 CHAPTER V DISCUSSION The purpose of this study was to examine whether music therapists should sing in a non-translated, inexperienced foreign language to match the song preference of a client or patient from a different cultural background or whether they should sing the translated English version of the same song. The statistical analysis of the research results indicated that there were no significant differences in the subjects emotional responses during either the Japanese or the English translated versions of a song sung by a music therapy student who was a native English speaker. However, one of the questions had a statistically significant difference in the ratings for two song versions. Neither the participants gender nor the length of stay in the United States had any effect on how they responded to any of the questions about the song versions. The results from a statistical analysis revealed that the participants were significantly more disturbed by the Japanese version than the English version of the song. The mean of rating for question 2 how much were you disturbed by the song? for the Japanese version was less than 3, but it was significantly higher than the mean rating for the English version. This finding can lead to the assumption that the accents and wrongly pronounced words in the Japanese version were more disturbing to the participants than listening to a familiar Japanese song in English. However, the mean ratings for both versions were less than 3 in a 7 point Liker-scale. Therefore, neither of the versions had much negative effect on the participants responses. This can be supported also by the fact that the means for question 3 how strong was your emotional response to the song? were the same for both versions. There were correlations between the questions and other variables. On the English version, the subjects who rated high on question 1, how much were you soothed by the song? scored low on question 2, how much were you disturbed by the song? and vice versa. These negative correlations were to be expected as the questions were contrary to each other. Similar correlations were observed between questions 2 and 3 how strong was your emotional response to the song?, and the questions 2 and 4, if you were experiencing being home sick, how much 26

34 would you like to listen to the song? The more the participants were disturbed by the song, the weaker their emotional responses to the song and the less likely they wanted to listen to the song when experiencing homesickness. Again, these correlations were expected due to the content of the questions. When the subjects had high ratings for question 1, they had stronger emotional responses to the song and they were more likely to want to listen to the song while being homesick. Thus, the correlations between the questions on the English version were very consistent with the content of the questions. For the Japanese version, there was only one significant correlation found between the questions. The subjects who had higher ratings on the emotional response tended to rate high on question 4 which asked how much they would like to listen to the song when experiencing homesickness. Other correlations between the questions that were seen on the English version were not found on the Japanese version. This may have been because the participants attention was not fully focused on listening to the Japanese version as a whole. Even though the singer had no experience in singing or speaking in Japanese, she produced a Japanese version of the song with high accuracy. However, it did not sound authentic enough to the Japanese people as evidenced by their comments. Over 38 % of the participants who left a comment said they could not help noticing the accents. Therefore, instead of listening to the all elements in the song simultaneously, their attention may have been shifted from focusing on listening to the entire song to attending to the words that sounded wrong or odd. Contrary to the researcher s expectation, there were no statistical differences between the adjective groups the participants chose to describe each song version. In fact, most of them chose the same adjective group for both songs. The researcher anticipated a different result because many of the subjects giggled while listening to the Japanese version. It was thought that the wrong pronunciations of words made by the singer caused this. However, it appeared it did not become a strong enough reason to interfere with the subjects emotional description of the song. Many of them noted the accents in the Japanese version were noticeable and a little annoying. Although results showed it was not a large enough factor to disturb participants 27

35 responses, it left them with negative depictions of the song. Participants comments were divided into two groups; ones who preferred the Japanese version and the ones who did not prefer the Japanese version. The first group of subjects mentioned they preferred the Japanese version more because it did not require them to carefully listen as opposed to having to listen to the English version carefully so that they would understand every word sung. However, the other group of subjects commented that they preferred the English version since they did not want to listen to a singer who did not understand the context of the song. During the Japanese version of the song, some participants noted I did not feel any emotion from the song and I would like a Japanese person to sing a Japanese song. Limitation of the Study There were some factors that may have possibly limited the results of the study. The first factor may have been the song choice. The song used in the study, furusato, was a very popular traditional Japanese song. In fact, all of the participants knew the song. However, it was not the type of song the participants would listen to on a daily basis when they reminisced about their home. Also, the song had a very simple melody with a medium tempo, making it easier to perform for a singer who was an English native speaker. The Japanese version of the song was performed well as evidenced by the participants asking the researcher if the singer spoke Japanese. If it had been a pop song with a fast tempo and complex melody lines, it may have been more difficult to sing in Japanese and might have sounded different from the original version, thus influencing results. Another factor was how the questions were asked in the questionnaire. As depicted in the results, there were no significant differences in the participants emotional responses to either of the songs except for one question. However, the written comments made by the participants were not always consistent with these findings. Although neither of the songs soothed or disturbed them very much, there was a clear difference in their impressions toward the two songs according to their written comments. Unfortunately, this study was not able to objectively and quantitatively measure this difference. 28

36 Implications for Music Therapy The purpose of study was to determine a possible approach for the growing minority client populations in music therapy settings. The research results suggest that the listeners emotional response would not be different if the music therapist sang in an inexperienced original language of a song or in English that was translated. This result allows music therapists options when dealing with a client or patient from another culture. Although the translated version of the particular song used in the study was available, this might not be always the case. Therefore, if the English translation was not available for a song, it is recommended that a music therapist sing an original version of the song. There is a possible benefit of singing the song in its original version: by being willing to use the language, it might help build rapport with a client or patient as suggested by Moreno (1998). However, both practice and consultation with a native speaker for mistakes in pronunciation are strongly suggested before performing the song in a music therapy session. Suggestions for Future Research During the study, only the Japanese language was tested. Japanese is a language that can be easily transliterated phonetically so that a person with no background in learning Japanese can say a word with high accuracy (Tsuge, 2004). However, languages such as Chinese and French would be very difficult to write phonetically since there are tones and sounds that do not exist in English. Thus, it is conjectured that the quality of performance in inexperienced Chinese or French would be poor. Therefore, if one of these languages were used in similar research, it may have a different result from the present study. Future research could examine non-traditional music from other cultures. Conclusion The hypothesis that the participants emotional response will not be affected by song 29

37 translation or non-native original language performance was supported by the research results. It is crucial for music therapists to be open-minded and flexible to accommodate music therapy session plans for minority populations. Under the current situations where most program do not provide special courses to teach future music therapists multi-cultural music, music therapists must learn about other cultures and expand their repertoire on their own. From this study, a possible approach for dealing with a client or patient from another culture was revealed. However, further studies regarding music therapy with multi-cultural clients or patients are much needed and encouraged. 30

38 APPENDIX A Human Subject Committee Approval 31

39 32