Towards a holistic model of functions of music listening across cultures: A culturally decentred qualitative approach

Towards a holistic model of functions of music listening across cultures: A culturally decentred qualitative approach Psychology of Music 40(2) 179 200 The Author(s) 2010 Reprints and permission: sagepub. co.uk/journalspermission.nav DOI: 10.1177/0305735610381885 pom.sagepub.com Diana Boer and Ronald Fischer Victoria University of Wellington, New Zealand Abstract The present article explores the functions of music listening from a cross-cultural perspective. We present a model of functions of music listening based on a multicultural qualitative approach. Our model covers personal, social and cultural musical experiences. Seven main functions of music listening were identified: music in the background, memories through music, music as diversion, emotions and self-regulation through music, music as reflection of self and social bonding through music. Our model was confirmed in an independent sample using a cross-method validation. Quantitative analyses of the qualitative data explored the salience of functions of music listening across four sub-samples: Asian and Latin-American sub-samples being more collectivistic and non- Anglophone Western and Anglophone Western sub-samples being more individualistic. Across all sub-samples the self-regulation function was the most important personal use of music, bonding was the most important social use of music and the expression of cultural identity was the most salient cultural function of music regardless of listeners cultural background. Apart from these similarities which point towards universalities, we also revealed cross-cultural differences pointing towards culture-specific uses of music. Limitations in the methodology of this exploratory cross-cultural approach and future directions in cross-cultural psychology of music are discussed. Keywords cross-cultural comparison, cultural functions, functions of music, individual functions, music listening, qualitative method, social functions Music has been identified as a strong human universal (Blacking, 1974; Cross, 2001; Falck & Rice, 1982; Merriam, 1964). In a comprehensive review of comparative music perception and cognition, Carterette and Kendall (1999) provide convincing evidence for the universality of sensory, perceptual, and cognitive processes involved in musical activities independent of social or cultural settings. Nevertheless, they present equally strong support for the view that these Corresponding author: Diana Boer, Centre for Applied Cross-Cultural Research, School of Psychology, Victoria University of Wellington, PO Box 600, Wellington, New Zealand. [email: diana.boer@gmx.net]

180 Psychology of Music 40(2) processes are conditioned by social and cultural forces (Carterette & Kendall, 1999: 727). While research has been conducted regarding the basic perceptual and cognitive processes in musical activities, the universalities and cultural specificities of broader psychological functions of music remain elusive. Does music serve similar psychological functions across cultures? This intriguing question has not been systematically investigated within music psychology. Previous research has demonstrated that demographic aspects, such as age and gender, are important factors that contribute to an individual s listening behaviour. For instance, differences between the musical preferences of male and female listeners have been reported (Colley, 2008; O Neill, 1997), just as gender roles have been shown to influence musical behaviour (Maidlow, 1999), while age or developmental stage of the listeners affect the type of music that is preferred (Hargreaves, Comber, & Colley, 1995). Nonetheless, cultural background has not attracted much research interest heretofore; hence, the current study focuses in particular on culture in an attempt to address this gap in contemporary research. With globalization and rising cultural mobility and diversity, determining and considering cultural differences and universalities becomes increasingly essential. However, since Merriam s (1964) seminal work and the paradigm shift in ethnomusicology from culture-general to culture-specific approaches in the 1980s (Falck & Rice, 1982; Nercessian, 2002), to the best of our knowledge, no systematic study has been conducted to examine the functions of music across more than one culture. The present article provides an exploratory cross-cultural examination of this topic with a specific focus on the functions of music listening. The aim is to develop a holistic model of functions of music listening by taking cultural perspectives into account. Musical behaviour has been explored in various disciplines, including ethnomusicology, sociology and psychology. As North and Hargreaves (2008) indicate, the level of analysis can vary from intra-individual to ideological. Similarly, each person can experience music as an individual (when remembering events of adolescence while listening to Nirvana or Led Zeppelin), social (when visiting a festival or club with friends) or cultural phenomenon (feeling proud when Stephan Raab participated in the European song contest Eurovision in 2000 as the German contestant). In order to gain a thorough understanding of these different types of musical experiences, these three levels are explored by drawing upon functions discussed in the psychological, sociological and ethnomusicological literature. Further, we introduce previous cross-cultural approaches to general psychological functions of music and methodological issues related to those approaches. The functions of music listening: three levels of musical experience Hargreaves and North (1999) conclude that the psychological functions of music embrace three broad domains, specifically cognitive, emotional and social functions. Functions of music listening in the cognitive domain are investigated in studies on the perception, recognition and memory of musical elements and characteristics, such as pitch, melody, rhythm, structure or complexity (e.g., Addessi & Caterina, 2005; Deliège & Sloboda, 1997; Ockelford, 2004); the immanent ability of music to trigger autobiographical memory (Cady, Harris, & Knappenberger, 2008; Schulkind, Hennis, & Rubin, 1999); and music s influence on cognitive performance (Furnham & Stephenson, 2007; Lesiuk, 2005; Schellenberg, Nakata, Hunter, & Tamoto, 2007). The emotional domain receives attention in studies about emotions perceived in music, emotions and physical reactions induced through music and emotions expressed through music (Juslin, 2005; Juslin & Laukka, 2004; Nawrot, 2003). In everyday life, music can be utilized in mood regulation (Husain, Thompson, & Schellenberg, 2002; Saarikallio & Erkkilä, 2007), to

Boer and Fischer 181 cope with a crisis (Behne, 1997; Lehmann, 1994) or as a stress reliever (North, Hargreaves, & O Neill, 2000; Tarrant, North, & Hargreaves, 2000). Finally, music impacts the social domain by providing guidelines for the construction and expression of identity (Tarrant et al., 2001; Tekman & Hortaçsu, 2002) and by supporting interpersonal relationships (Bakagiannis & Tarrant, 2006; Hargreaves & North, 1999; Rentfrow & Gosling, 2006). Much of the psychological literature, however, assumes that music listening is an individual and solitary activity. This assumption may, in fact, be linked to the cultural background of the many music psychologists who hail from Western societies, which typically tend to be more individualistic (Hofstede, 2001). Thus, the literature focuses predominantly on cognitive and emotional functions of music at the level of the individual and neglects collective aspects of musical experiences. This imbalance is increasingly criticised by various psychological scholars (Juslin, 2005; Juslin & Laukka, 2004; Lehmann, 1994; MacDonald, Hargreaves, & Miell, 2001; North & Hargreaves, 2008; Rentfrow & Gosling, 2006). Meanwhile, sociologists and ethnomusicologists also remind us that the collective aspects of music make up a fundamental feature contributing to social and cultural settings (DeNora, 2000; Frith, 1987; Merriam, 1964; Mitchell, 1996). Music can thus serve as a symbolic representation of social and cultural values and identities (Merriam, 1964), while also having the ability to communicate personal values and identities to others (Frith, 1987). The transmission of cultural identities, norms and values in music can support conformity and validate societies, and thereby contribute to cultural continuity and integration (Merriam, 1964). Individuals can use music to create a self-definition and find a place within society (Frith, 1987). Frith (1996) subscribes to a broad and flexible definition of identity when he claims that experiencing popular music is an experience both of individuality and collective identity. DeNora (2000) argues that music is a resource for producing social life. As an everyday activity, music furnishes the social space with material-cultural resources for feeling, being and doing. This is part of how the habitat for social life its support system is produced and sustained (DeNora, 2000: 129). Clearly, music serves functions at the individual, social and cultural levels, each of which are underpinned by psychological processes. However, most psychological research lacks a comprehensive analysis of these three different levels of musical experience (North & Hargreaves, 2008). The under-representation of social components in music psychological research has been mentioned previously, yet the neglect of culture in music psychology is even more profound (North & Hargreaves, 2008; Walker, 2004). A substantial number of models have been developed proposing psychological functions of music listening (Behne, 1997; Lehmann, 1994; North et al., 2000; Tarrant et al., 2000; Sloboda, 2005), and additional published articles list psychological functions of music listening based on previous literature (Hargreaves & North, 1999; Schäfer & Sedlmeier, 2009). However, cultural aspects of musical experience are hardly examined in these sources. The current article advances research by simultaneously examining individual, social and cultural experiences with music. Examining all three levels aims to gain a holistic picture of the psychological functions of music beyond the conventional focus on individual experience. We specifically focus on cultural experiences in musical functioning, addressing an initial gap in music psychological research. The second gap entails a lack of research on non-western audiences in psychological research. Only a limited number of psychological studies investigate musical experiences in non-western samples (Tekman & Hortaçsu, 2002) or compare limited numbers of samples from Western and non-western settings (Balkwill & Thompson, 1999; Eerola, Himberg, Louhivuori, & Toiviainen, 2006; Gregory & Varney, 1996). The current article aims to provide a more holistic and inclusive perspective

182 Psychology of Music 40(2) of the functions of music by taking culture into account in the examined musical experience, as well as in the sampling approach. Cross-cultural psychology and its methodologies provide fertile grounds from which an inquiry on musical functioning can arise, given its concern with the systematic study of behaviour and experience as it occurs in different cultures, and is influenced by culture and results in changes in existing cultures (Triandis, 1995). Hence, cross-cultural psychological studies on music are introduced, alongside methodological issues that must be addressed when dealing with samples from more than one cultural group. Cross-cultural psychology of music A small number of studies has taken a cross-cultural psychological perspective while investigating the functions of music listening. For instance, Saarikallio (2008) examined the use of music in mood regulation. Her study of Finnish and Kenyan adolescents revealed that the nature of mood regulation by means of music is similar across the two samples coming from rather disparate cultures. Meanwhile, Saarikallio (2008) found certain cross-cultural differences in how this regulation was realized. Specifically, Kenyan adolescents verbalized three additional processes of mood regulation, namely the use of music to foster concentration, gain energy, and express happiness. Similarly, Gregory and Varney (1996) conducted a study on affective responses to music by comparing European and Asian listeners. They found that European and Asian listeners responded quite differently to Western and Asian music. Gregory and Varney (1996) suggest that the affective responses to music seem to be determined by the cultural background of the listener. In their study on Pakistanis, Rana and North (2007) uncovered striking similarities in the role of music in everyday life when compared to a British study (North, Hargreaves, & Hargreaves, 2004). Finally, Schäfer, Sedlmeier, and Tipandjan (2008) examined 17 functions of music and their links to music preferences in both a German and an Indian sample. Their results revealed that in India, music fulfils the same functions in everyday life and to the same extent as in Germany. However, the link between these functions and music preferences was not as strong as in the German sample. These cross-cultural studies suggest that general universalities, as well as certain cultural specificities, exist in the functions of music listening. It bears repeating, however, that these studies predominantly focus on the individual functions of music; social aspects are given much less attention. Furthermore, in all four studies, instruments measuring functions of music were developed based on findings from Western samples. Those instruments were then applied in non-western samples, aiming to replicate or generalize findings from Western samples. The pitfalls of such an approach have been intently discussed within cross-cultural psychological literature (Berry, 1989; Segall, Lonner, & Berry, 1998). Certain methodological issues, introduced in the following section, need to be addressed when conducting cross-cultural research. Methodological issues in cross-cultural research Models constructed in Western settings are often applied in non-western settings, even though it remains questionable whether the examined construct is exhaustively captured in non-western settings (Berry, 1989; Segall et al., 1998). Construct bias is an important methodological issue in such an imposed approach. Domain under-representation is one facet of construct bias, which appears when important aspects of the domain that a theoretical variable is assumed to account for are not represented in the measurement instrument (Fontaine, 2005, p. 803).

Boer and Fischer 183 Berry (1989) and Segall et al. (1998) proposed an approach that incorporates various cultural perspectives of a particular phenomenon and then draws conclusions about psychological similarities across cultures. In order to develop a framework of functions of music listening that is applicable across cultures, this study thus aims to consider multiple cultural perspectives. Van de Vijver and Leung (1997) promoted a decentred approach to ensure a cultural balance and to avoid construct bias in cross-cultural research. Using a decentred approach, multiple cultural perspectives on the phenomenon in question are gathered. The input from culturally diverse individuals reduces the likelihood of domain under-representation. In decentred approaches, systematic sampling is desired (Van de Vijver & Leung, 1997). Participants from more than two cultures are sampled, which vary systematically in cultural dimensions, such as values, beliefs, or self-construal, in line with the respective research question (Van de Vijver & Leung, 1997). A dimension of cultural variability that seems particularly relevant for our purposes is individualism-collectivism. Within the last two decades of cross-cultural research, individualism-collectivism has proven to be the most prominent cultural dimension for the psychological investigation of culture (Brewer & Chen, 2007; Hofstede, 2001; Oyserman, Coon, & Kemmelmeier, 2002). Individualistic cultures can be characterized as valuing personal autonomy; the individual strives towards personal goals and is perceived as an independent, self-reliant being (Triandis, 1995). Countries that are regarded as most individualistic are situated in North America, Central and North Europe and Australasia (Hofstede, 2001). Collectivistic cultures, on the other hand, value social embeddedness; the individual s priority is to pursue group goals over individual goals and the collectivist self is interdependent and inseparable from the collective, such as family, friends or other collectives (Oyserman et al., 2002; Triandis, 1995). Countries deemed as the most collectivistic are situated in Latin and South America and many parts of Asia (Hofstede, 2001). The sampling of this research focused on the individualism-collectivism dimension of cultural variability for two reasons. First, as noted above, previous psychological research has predominantly focused on individualistic perspectives of musical experiences and has neglected collective aspects, an imbalance which most likely stems from the individualistic background of so many researchers in this field. Second, the individualism-collectivism dimension posits the possibility that individuals from different cultural ends of this dimension experience and use music differently. This has not been examined, yet. Hence, we sampled participants from various collectivistic (East and South East Asian, South American) as well as more individualistic (European, North American, and Australasian) societies in our studies. Detecting culture specifics, universals or systematic cultural variations along cultural dimensions is a significant challenge that a single study can certainly not hope to achieve independently. In fact, a series of studies is required, with our approach providing one exploratory piece of a much larger picture. Given the lack of research on cultural variations in musical perception and reception, our study is particularly timely (Huron, 2008). In this article, we present a multicultural Study 1a, in which qualitative data was used to construct a model of functions of music listening that seeks to be applicable across cultures. However, prior to an explicit examination of cultural similarities and variations in the model developed, the model is hoped to be validated in an independent sample in Study 1b, using a distinct methodology and sample composition. This attempts to address possible methodological constraints of the proposed model. Once cross-method evidence of the validity of our proposed model has been established, cultural variation along the individualism-collectivism dimension in Study 2, in which we reanalyse the data of Study 1a, will be explored. Thus, alongside the empirical construction

184 Psychology of Music 40(2) of a model of the functions of music listening, culture specifics, universals or systematic cultural variations are sought to be identified in the salience of the identified functions of music. For this purpose, the occurrences of identified themes in the responses of the participants were compared across four cultural sub-samples, varying along the individualism-collectivism dimension. Individualism-collectivism was not directly measured; instead, we relied on past research to classify participants according to their cultural backgrounds (Hofstede, 2001; Inglehart & Baker, 2000; Schwartz, 2006). Study 1 Study 1 is a qualitative study that aims to develop an empirically-based holistic model of the functions of music listening. First, in Study 1a we analysed responses to open-ended questions about the individual, social and cultural experiences with music in young people s lives from a variety of cultures. The answers were coded across all groups to reduce ethnocentric bias in interpreting the themes. Then, a second sample was used to test whether or not the proposed model of the functions of music listening could be validated. For this purpose, an independent qualitative study was designed to further assess individuals experiences with music. Study 1b addresses four of the limiting factors of Study 1a: 1) participants were a convenience sample drawn from the general public (as opposed to targeted music lovers and the possible self-selection of participants in Study 1a); 2) participants were sought to encompass a broader age range (as opposed to the young participants in Study 1a); 3) a paper-and-pencil format was used (rather than the online survey in Study 1a); and 4) Study 1b used a different set of questions about musical experiences, focusing on expected emotions and physical reactions to music (rather than on a broader scope of musical experience in Study 1a). Study 1b was designed with the central premise that the reoccurrence of all functions of music would validate the existence of those functions, independent of method and sample. Method of Study 1a Data collection. Samples were targeted based on two criteria, high musical commitment and cultural variation in the individualism-collectivism dimension based on participants cultural background. The targeted population entailed music fans and young people, as this population is composed of committed music listeners (LeBlanc, Sims, Siivola, & Obert, 1996), who would be most likely to provide rich input with respect to content. In order to obtain the targeted samples, we took advantage of the internet (Gosling, Vazire, Srivastava, & John, 2004; Reips, 2000; Skitka & Sargis, 2005). The link to the online survey was posted on discussion boards of musicrelated websites. 1 These websites targeted a range of popular music styles in each country, such as hip hop, pop, rock, metal and music in general. The websites originated in seven countries, which were selected based on a systematic sampling approach (Van de Vijver & Leung, 1997), as they cover three individualistic (New Zealand, the US, and Germany) and four collectivistic (Hong Kong, the Philippines, Brazil, and Singapore) societies (Hofstede, 2001). The survey was conducted in English in Singapore, Hong Kong, the Philippines, Brazil, New Zealand, and the

Boer and Fischer 185 US. The proficiency in English is fairly high in the sampled Asian countries, as English is an official language of these countries. Furthermore, English was the language of the Asian websites involved in the study; therefore, a language barrier was deemed unlikely for Asian participants. In the Brazilian sample, on the other hand, lower English proficiency may have been a limiting factor. Issues that may arise when using qualitative data in a second language are discussed in more detail below. The survey in Germany was conducted in German, which was also the language of the German websites. German responses were translated by the first author and corroborated in a committee approach by three bilingual colleagues (one academic staff member and two postgraduate students). We applied the multiple site entry technique by posting the survey in five webpages per host country. This technique was suggested by Reips (2000) as a means of reducing self-selection bias in internet sampling. However, with this target population, representativeness of our results can by no means be claimed. Study 1b aims to countervail some of the methodological issues related to age, musical commitment, and self-selection faced in the samples of Study 1a. Participants. The sample consisted of 222 participants in Study 1a. 2 Among the participants, the average age was 23 years (SD = 9.05; range = 13 69), and 58% of the participants were female. Regarding their cultural backgrounds, the participants can be categorized into four broad cultural clusters, based upon their stated home country (that is, the country in which they grew up) and ethnicity, namely Anglophone Western, non-anglophone Western, Asian, and South-American (Table 1). 3 The first two clusters represent mostly individualistic cultures, and the latter two represent mostly collectivistic cultures (Hofstede, 2001). Similar clusters have been empirically identified in large cross-national studies (Inglehart & Baker, 2000; Schwartz, 2006). Using participants native country as a proxy for cultural background has been deemed an acceptable technique for exploratory studies (Matsumoto & Yoo, 2006). 87% of the participants resided in the seven targeted countries (30 in Brazil, 5 in Hong Kong, 82 in Germany, 32 in New Zealand, 7 in the Philippines, 23 in Singapore, and 13 in the US), and the remaining participants were born in the targeted countries, but resided elsewhere, or they came from other Asian, South-American, Anglophone and non-anglophone Western societies. Those participants whose country of birth differed from their country of residence (17%) were categorized based on their primary ethnic or cultural affiliation. The categorization based on birth country was at variance from the cultural cluster of the country of residence for only 5% of the participants. Excluding these individuals did not affect any of the findings of the current study. Questionnaire. The online questionnaire consisted of two parts. The first section dealt with musical preferences, musical activity and demographic details. The second part asked Table 1. Sample description by cultural cluster (Study 1a, 2) Sub-sample Male Female All N Age: M (SD) N Age: M (SD) N Age: M (SD) Anglophone Western 26 25.38 (10.59) 22 26.45 (14.35) 49 26.20 (12.41) Non-Anglophone Western 17 24.47 (5.41) 77 16.84 (3.93) 95 18.38 (5.33) Asian 29 26.48 (9.92) 15 21.67 (4.51) 44 24.84 (8.72) South-American 17 27.35 (9.39) 17 24.76 (4.59) 34 26.06 (7.40) Whole sample 89 25.94 (9.25) 131 20.04 (7.96) 222 22.56 (9.05)

186 Psychology of Music 40(2) open-ended questions regarding the meaning and function of music in people s lives. These questions addressed three levels, namely the personal meaning of music, the social meaning of music and the cultural meaning of music. Since asking multiple questions enhances the validity and reliability of capturing the phenomenon empirically, three questions sought to capture the personal significance of music to the participants. The first question targeted the meaning of music in each individual s life (What does music mean to you? Please write your thoughts about the role music plays in your life.). The second question related to the influence individuals perceived music to have in their lives (How does music influence your life?). The third question targeted situational experiences with music (Think about one specific situation when you were listening to music in the last three days. Please describe what you thought, felt and did in that situation.). Two social contexts were used to gain an understanding of the social significance of music, entailing being with friends (What role does music play when you are hanging out with your friends?) and the meaning of music for the family (What is the meaning of music for your family members?). The cultural significance of music was sought by asking two further questions (What is the meaning of music in your home country? and What is the meaning of music in your cultural community?). Participants completed survey versions that randomly contained three of the seven openended questions in an attempt to shorten the survey. Analytical strategy. The content of the responses to the open-ended questions was analysed using thematic analysis. Thematic analysis is a qualitative method that identifies, analyses and reports pattern within data (see Braun & Clarke, 2006) and was performed to identify themes and sub-themes of the functions of music. One central advantage thematic analysis has over, for instance, content analysis or grounded theory, is that it allows for large data sets to be handled in a flexible manner. The participants cultural background did not explicitly enter the analytical approach for two main reasons. First, the aim of Study 1a was to develop one model that is applicable to various cultures, not one model for each cultural group. Thus, the analyses included the data set as a whole, as it is built upon culturally rich data input. A second reason for not explicitly considering the cultural background was to reduce ethnocentrism in coding the answers. In qualitative research, objectivity is not (or cannot be) pursued (Greenfield, 2000); still, an attempt was made to minimize bias in the cultural interpretations during the analyses of the responses by analysing the data without reference to the cultural or demographic origins of the respondents. Such an approach rules out that responses are overinterpreted or misinterpreted with regard to their cultural content or origin. The cultural background of participants will be examined in greater detail in the quantitative analyses of the data set presented in Study 2. In the following section, the various identified functions of music listening are discussed. Extracts from respondents answers will be presented in English with corrected spelling and with reference to the respondent s demographic background (M = male, F = female). Results and discussion of Study 1a Seven main themes were identified: music in the background, memories through music, music as diversion, emotion in music and self-regulation through music, music as reflection of self and social bonding through music. Each of the seven main functions of music consists of a varying number of sub-functions, comprising the rich content of each function. 1. Music in the background. Music is often used as a background element while doing something else; it can also fill gaps and help pass the time. The use of music as a background element

Boer and Fischer 187 while being engaged in other activities was mentioned by a Singaporean (M, 36 years), who considered that music is something meant to be played (softly) in the background while [ ] doing something else. Previous research has referred to the background function as a facet of individual use, for example, as a diffuse listening style, where the focus is not on the music but on other activities (Behne, 1997), or the use of music for distraction from routine tasks (Sloboda, 2005). Our results also revealed the social element of this function, which was, for instance, brought up by a participant from Malaysia (M, 34 years), who uses background music to keep the ambience warm and cosy when spending time with friends. In such social settings, music can thus serve the function of creating the right atmosphere. When music serves a background function, it is not the centre of attention. 2. Memories through music. Particular songs can trigger memories and can connect the listener to his/her past. Music can bring back memories of events, life stages, relationships and emotions or memories of loved ones, including friends and relatives. A 15-year-old female participant from England stated: I also have lots of memories linked to music, so listening to a particular song can remind me of someone. The function of reminiscence has been studied previously as a facet of individual use of music (Behne, 1997; DeNora, 2000; Frith, 1987; Schulkind et al., 1999; Sloboda, 2005; Sloboda, O Neill, & Ivaldi, 2001). However, this particular function of music can also operate when music is enjoyed in the company of friends or family members. In this case, joint listening to a particular song can trigger memories of shared moments. A 32-year-old female from New Zealand said that we think of our memories when listening to music with friends. Therefore, music is a means of reminiscing alone or with others. 3. Music as diversion. Music is also a medium of entertainment, as it is used for feeling good and enjoying oneself. Music is mainly to have fun, as a 15-year-old female from Germany stated. In addition, music is essential for dancing. As a Brazilian (F, 28 years) mentioned, along with listening to the music, there is the dancing that helps on the feeling good part. According to Merriam (1964), music serves as a form of entertainment in all cultures. He also posits that the entertainment function is accompanied, for instance, by physical responses, such as dancing or aesthetic enjoyment, which the data of our study supports. The diversion function incorporates the simple enjoyment of music without connection to prior mood and emotions (Saarikallio & Erkkilä, 2007; Sloboda, 2005; Sloboda et al., 2001). It also includes social aspects, such as dancing with friends (DeNora, 2000; Merriam, 1964). 4. Emotion in music. Many participants stated that music has the capacity to convey emotions and to trigger emotions or emotional and physical reactions. Particular songs are therefore specifically chosen for their emotional content in order to express a particular emotional state of the participants. One participant (F, 30 years) from Germany summarized the emotional function of music as follows, Music can transport and express emotions. And what is more powerful than a song or just a tune which is able to bring tears in your eyes? As previously mentioned, the emotional effects of music have attracted the attention of many psychologists. Most prominent among them, Juslin and colleagues (Juslin & Laukka, 2004; Juslin & Sloboda, 2001) examined how emotions in music are perceived and how music triggers emotions in the listener. In light of these processes, listeners use music to express their emotions to their environment (Baacke, 2002; DeNora, 2000; Frith, 1987) as well as to regulate their emotions and mood (Behne, 1997; Saarikallio & Erkkilä, 2007). A participant from New Zealand (F, 20 years) noted the interconnectedness of the induction, perception and affection of emotions and mood in and through music in stating, Music influences me by changing, affecting or reinforcing whatever mood I m in.

188 Psychology of Music 40(2) 5. Self-regulation through music. The listener actively utilizes music to alter or improve his or her current state of mind (beyond mood). This function encompasses a diverse set of psychological domains that can be altered by music. Music can help to relax and relieve stress (Behne, 1997; Lehmann, 1994; Sloboda, 2005; Sloboda et al., 2001) and to enhance creativity and intellectual focus. Listening to music can reduce loneliness, while offering a means of escape from this world (Baacke, 2002; Lehmann, 1994; Tarrant et al., 2000). For instance, a 37-yearold female from New Zealand labels music as an escape and a comfort from the difficulties of life. Certain music can assist in venting frustration and aggression (Behne, 1997; Lehmann, 1994; Tarrant et al., 2000). Music thus acts as a form of therapy; it can ease sorrow and negative moods in times of crisis (Juslin, 2005; Lehmann, 1994; Saarikallio & Erkkilä, 2007; Sloboda, 2005). 6. Music as a reflection of the self. Music can express three different kinds of identity, individual, social and cultural identity. First of all, music can be an outlet for personal self-expression. It allows for the expression of a person s individuality and lifestyle. For a 31-year-old female in New Zealand, music is a way for me to express my individuality liking a particular sound or band was a way of saying who and what you stood for. Furthermore, music expresses and influences values and attitudes; it can act as inspiration, and can offer guidance (Baacke, 2002; DeNora, 2000; Frith, 1987; Hargreaves & North, 1999; MacDonald, Miell, & Wilson, 2005; Merriam, 1964). Indeed, music has positively influenced the personal development of many of the participants, as one 33-year-old German woman emphasized, And music is a huge enrichment of your whole life. Music can make you wise I don t know where I would be today without Beethoven. Second, music indicates social identity by signifying group membership, for instance, belonging to a particular social group (like alternative or rave) or the current cool group in school (Frith, 1996; Tarrant et al., 2001). Third, music can reflect cultural identification, commenting on its history and the current conditions for the members of this cultural group. In this regard, a 29-year-old Romanian woman states that the folk, traditional music is more related to national identity, Romanian customs and traditions, and history. Our results emphasize that not only individual identity (Baacke, 2002), but also social and cultural identities, which have been discussed in sociological and ethnomusicological literature (DeNora, 2000; Frith, 1996; Merriam, 1964) are salient in the participants utilization of music. 7. Social bond through music. The last function of music to be discussed entails social bonding with friends or family. Music can provide an opportunity for a collective activity, such as discussing and listening to music or going to concerts together. These shared musical activities can influence relationships with friends and family members and can even create a special bond within a family or a circle of friends. A 17-year-old male participant from the US declared that music gives us a topic. We discuss it It helps us bond in a way that nothing else can. Social psychological investigation of the social bonding function started to flourish recently (Bakagiannis & Tarrant, 2006; Boer, 2009; Hargreaves & North, 1997; North & Hargreaves, 2008; Rentfrow & Gosling, 2006). However, most studies have focused on providing empirical evidence for the bonding character of shared musical preferences, while the underlying processes of the collective experiences of music listening or concert visits remain unclear. To summarize, in this multicultural qualitative study, we identified seven functions of music listening to create a holistic model. Needless to say, these seven functions of music are not genuinely new discoveries, as they have been described and analysed in previous psychological, sociological and anthropological literature. However, to the best of our knowledge, they have not previously been compiled into one holistic model of the functions of music. The novel

Boer and Fischer 189 contribution of the proposed model involves the incorporation of a holistic range of functions and the more balanced coverage of social elements, as compared to previous models. Before possible cultural variations in the proposed functions of music are explores, another sample is needed to validate the proposed functions of music listening. Thus, Study 1b examines the cross-method validity of the seven functions of music proposed. Method of Study 1b Participants. The participants included 74 persons of the general population, who were randomly approached in the central business district of Wellington, New Zealand and invited to participate in the study. The average age of the participants was 31 (SD = 12.08); 49% were female. More than half of the participants were born in New Zealand (63%); 20.5% were born in other Anglophone countries, such as Australia, the UK or the US; and the remaining 16.5% were born in other countries, such as Greece, Kuwait or Spain. Questionnaire and analytical strategy. The participants were given two tasks. First, they were to describe the emotions aroused by the thought of listening to music; and second, they were to note their specific physical reactions to listing to music. These tasks focus on the anticipated reactions and expected effects of listening to music. Despite the different frame of the questions in Study 1b, all seven functions of music were expected to be present if they are genuinely relevant functions of music. The responses were categorized according to the seven main functions of music listening by the first author. In order to establish interrater reliability, 50% of the responses were coded by a second coder. Reliability was assessed using Cohen s Kappa measuring agreement. Landis and Koch (1977) suggest that K = 0.40 to 0.59 indicate moderate interrater reliability, 0.60 to 0.79 substantial, and above 0.80 outstanding agreement. The average interrater agreement for the coding of Study 1b was acceptable (K = 0.65) and ranged from moderate agreement for the self-regulation function (K = 0.41) to outstanding agreement for the background, memory and social bonding functions (K = 1.0). The percentage of participants who mentioned each respective function of music in their responses to the two questions (averaged) is reported below. Results and discussion of Study 1b All seven main functions of music made their appearance in the responses of the participants in Study 1b. Participants mentioned the self-regulation function recurrently (58% of all participants), as music helps to relax, to relieve stress, to energise, and offers a form of escapism (as one 39-year-old female from New Zealand reports, being able to separate myself from the hustle and bustle of everyday life ). The diversion function of music reappeared in 39% of the responses in terms of enjoyment and the desire to dance. Participants responses clearly emphasized a range of feelings that the emotional function of music epitomizes (37% of the responses), such as happiness, sadness, excitement, loss, as listed by a 22-year-old male participant from Fiji. Furthermore, a number of physical reactions were described, such as an increased or decreased heart rate, finger tapping, or smiling. The function of music as an expression of the self was sporadically mentioned in 10% of the responses as a source of curiosity, inspiration and guidance in life. The reminiscence function was present in 9% of the responses as a memory that music can trigger. 3% of the responses reported a feeling of connectedness and friendship when listening to music, representative of the social bonding function. A limited number of

190 Psychology of Music 40(2) participants (2% of responses) indicated that they listen to music in the background while engaging in other activities. In an independent sample in Study 1b, all seven functions of music listening reoccurred. The reoccurrence of functions was supported by a substantial interrater agreement, indicating cross-method evidence of the validity of these functions. With these results, the universalities or cross-cultural variations in the prevalence of the seven functions of music listening can now be explored. Study 2 The aim of Study 2 was to explicitly examine the cultural similarities and variations in the developed model of the functions of music listening. For this purpose, the qualitative data set of Study 1a was reanalysed using quantitative methods. As mentioned above, previous psychological research on music listening has focused on the level of the individual by investigating the cognitive and emotional functions of music. The source of this focus on the individual can be based upon the research questions that were previously asked or on the incorporation of mainly individualistic samples. Our two research questions thus target these two facets. First, what are the most prevalent functions of music for the individual, social and cultural experience of music? We posit that the questions at each level of musical experience elicit the respective functions of music; for the individual experience of music, functions with personal focus, such as emotional and self-regulation functions, will be most prevalent; for the social experience with music, the social bonding function will be most prevalent; and as for cultural experiences with music, the function including cultural identity (reflection of the self) will be most prevalent in the responses. Our second research question targets the cultural background of participants: Are there differences in the occurrence of functions of music listening in the responses from individuals with different cultural background? As we have previously argued, collective musical elements may be more prevalent among individuals from collectivist cultures, given that they ostensibly prioritize group goals over personal ones. Similarly, individual elements, such as emotional and reminiscence functions, might prove to be more prevalent among individuals from individualistic cultures. Our analyses aim to explore systematically whether music functions 1) appear universally (across all four cultural regions), or 2) differ systematically along the cultural individualism-collectivism dimension, or 3) differ across the four geographical regions (culturespecific). We are the first to investigate such differences; therefore, the analyses are exploratory. Method Participants and questionnaire. In Study 2, the qualitative responses of Study 1a were reanalysed using quantitative techniques. Hence, participants, data collection procedures and questions have already been described above. According to Boyatzis (1998), thematic analysis is a process that can be applied on a continuum of qualitative and quantitative methods. The two poles of this method continuum are entirely qualitative description of the data (as presented in Study 1a) on one pole and purely quantitative methods of statistical analysis (as presented in Study 2) on the other. Analytical strategy. Moving beyond the verbal description of the identified functions, we quantified the functions of music by means of nominal coding (presence absence coding) of the

Boer and Fischer 191 responses. Each response received a code regarding each identified function. If the function was present in the response, the coded value was 1; if the function was absent in the response, the coded value was 0. These binary codes allowed for structured statistical analyses of the responses, including frequency analyses across cultural clusters (non-anglophone Western, Anglophone Western, Asian, and South-American; see Table 1). The research questions will be explored using Chi-square tests. If systematic differences between individualistic and collectivistic sub-samples in the frequency analyses are uncovered, then claims about cultural variations along the individualismcollectivism dimension can be made (Van de Vijver & Leung, 1997). The general prevalence of the functions of music listening and cultural differences in the occurrence of these functions at the three investigated levels were separately examined for the individual (3 questions), in social settings (2 questions) and at the cultural level (2 questions). For the following analyses we merged the responses to the three questions at the individual level, the responses to the two questions at the social level, and the responses to the two questions at the cultural level in order to increase reliability. The data was coded by the first author. In order to establish interrater reliability, all responses were independently coded by a second researcher. The codings of the second coder overlapped by 92% with the first author s codings. Interrater agreement was again assessed using Cohen s Kappa analogue to the procedure in Study 1b. The average interrater agreement of main themes was acceptable with K = 0.62, ranging from K = 0.49 (diversion, emotions) to K = 0.81 (memories). According to Landis and Koch (1977), our coding of main themes shows a substantial reliability, ranging from moderate to outstanding agreement. Results and discussion The prevalence of the functions of music listening at three levels The first question was to ask what functions of music listening would be most salient at each level of questioning. The occurences of all functions at the individual, social and cultural level of musical experiences are presented in Table 2. The self-regulation function was listed with the greatest frequency by participants at the individual level. The second most salient function at the individual level was the emotional function, followed by identity expression. At the social level of musical experience, the social bonding function of music was, as expected, the most salient function in the overall sample. The second most salient function at the social level was the diversion function, followed by the background function. With respect to the cultural meaning of music, the most frequently mentioned functions of music were the expression of (cultural) identity and music as a means of diversion. These findings support our expectations that the questions at each level of musical experience elicit the functions of music listening with the respective focus. Some functions are particularly meaningful at certain levels, such as the expression of cultural identity at the cultural level and social bonding at the social level. However, the results also reveal an intriguing pattern; all seven functions were identified (although to varying degrees) at all three levels. Furthermore, for instance, the diversion function occurs with a similar frequency in all three levels. These findings suggest a complex interplay between level of meaning and the holistic collection of the functions of music. The final question revolves around whether or not cultural variations exist in these patterns.

192 Psychology of Music 40(2) Table 2. Prevalence of seven functions of music listening at three levels of musical experience (percentage of occurrence in responses; N=222) Function of music Individual experience (%) Social experience (%) Cultural experience (%) Background 12 17 6 Memories 9 6 1 Diversion 22 18 21 Emotion 27 7 8 Self-regulation 36 6 8 Identity 22 6 32 Social bond 6 34 5 The prevalence of the functions of music listening across cultural groups Beginning with the individual-level question, the explicit social function of music (social bonding) was not more salient in the collectivist sub-samples (Table 3). Two statistically meaningful cultural variations arose in the appearance of music as a trigger of memories and emotions. A greater number of non-anglophone Western participants than participants from other cultural clusters indicated that music listening evokes memories. A similar variation occurred across cultural groups in the emotional function of music; significantly more non-anglophone Western participants than Anglophone Western, South-American and Asian participants mentioned experiencing the emotional functions of music. This finding may point towards a culture-specific musical functioning for non-anglophone Westerners, which appears independent of the individualistic dimensions, as Anglophone Westerners did not show these patterns. For both detected cross-cultural variations, however, it could be argued that the variations may be due to the rather distinct demographic composition of the non-anglophone Western sub-sample. This sub-sample was significantly younger than the other sub-samples (F(3, 218) = 14.19, p < 0.001; Table 1). Furthermore, this sub-sample was predominantly composed of adolescent girls, who may be particularly committed to using music in their emotional and reminiscent moods. In order to overcome the incomparability of the sub-samples, we matched the sub-samples by dropping non-anglophone Western participants below the age of 16. This Table 3. Cross-cultural comparison of functions of music listening at the individual level (percentage of occurrence in responses; N = 222) Individualistic samples Collectivistic samples Function of music Anglophone Western (%) N = 49 Non-Anglophone Western (%) N = 95 Asian (%) N = 44 South American (%) N = 34 Chi-square df = 3 Background 8 12 19 8 5.16 Memories 4 16 6 4 12.00** Diversion 32 21 15 24 6.35 Emotion 21 41 12 20 25.09*** Self-regulation 27 43 33 33 6.62 Identity 20 21 29 20 2.44 Social bond 12 5 1 6 7.21 Note. * p <.05, ** p <.01, *** p <.001; Significance (2-sided)