The psychological functions of music listening

Why do people listen to music? Over the past several decades, scholars have proposed numerous functions that listening to music might fulfill. However, different theoretical approaches, different methods, and different samples have left a heterogeneous picture regarding the number and nature of musical functions. Moreover, there remains no agreement about the underlying dimensions of these functions. Part one of the paper reviews the research contributions that have explicitly referred to musical functions. It is concluded that a comprehensive investigation addressing the basic dimensions underlying the plethora of functions of music listening is warranted. Part two of the paper presents an empirical investigation of hundreds of functions that could be extracted from the reviewed contributions. These functions were distilled to 129 non-redundant functions that were then rated by 834 respondents. Principal component analysis suggested three distinct underlying dimensions: People listen to music to regulate arousal and mood, to achieve self-awareness, and as an expression of social relatedness. The first and second dimensions were judged to be much more important than the third—a result that contrasts with the idea that music has evolved primarily as a means for social cohesion and communication. The implications of these results are discussed in light of theories on the origin and the functionality of music listening and also for the application of musical stimuli in all areas of psychology and for research in music cognition.


INTRODUCTION
Music listening is one of the most enigmatic of human behaviors. Most common behaviors have a recognizable utility that can be plausibly traced to the practical motives of survival and procreation. Moreover, in the array of seemingly odd behaviors, few behaviors match music for commandeering so much time, energy, and money. Music listening is one of the most popular leisure activities. Music is a ubiquitous companion to people's everyday lives.
The enthusiasm for music is not a recent development. Recognizably musical activities appear to have been present in every known culture on earth, with ancient roots extending back 250,000 years or more (see Zatorre and Peretz, 2001). The ubiquity and antiquity of music has inspired considerable speculation regarding its origin and function.
Throughout history, scholars of various stripes have pondered the nature of music. Philosophers, psychologists, anthropologists, musicologists, and neuroscientists have proposed a number of theories concerning the origin and purpose of music and some have pursued scientific approaches to investigating them (e.g., Fitch, 2006;Peretz, 2006;Levitin, 2007;Schäfer and Sedlmeier, 2010).
The origin of music is shrouded in prehistory. There is little physical evidence-like stone carvings or fossilized footprints-that might provide clues to music's past. Necessarily, hypotheses concerning the original functions of music will remain speculative. Nevertheless, there are a number of plausible and interesting conjectures that offer useful starting-points for investigating the functions of music.
A promising approach to the question of music's origins focuses on how music is used-that is, it's various functions. In fact, many scholars have endeavored to enumerate various musical functions (see below). The assumption is that the function(s) that music is presumed to have served in the past would be echoed in at least one of the functions that music serves today. Of course, how music is used today need have no relationship with music's function(s) in the remote past. Nevertheless, evidence from modern listeners might provide useful clues pertinent to theorizing about origins.
In proposing various musical functions, not all scholars have related these functions to music's presumed evolutionary roots. For many scholars, the motivation has been simply to identify the multiple ways in which music is used in everyday lives (e.g., Chamorro-Premuzic and Furnham, 2007;Boer, 2009;Lonsdale and North, 2011;Packer and Ballantyne, 2011). Empirical studies of musical functions have been very heterogeneous. Some studies were motivated by questions related to development. Many related to social identity. Others were motivated by cognitive psychology, aesthetics, cultural psychology, or personality psychology. In addition, studies differed according to the target population. While some studies attempted to assemble representative samples of listeners, others explicitly focused on specific populations such as adolescents. Most studies rely on convenient samples of students. Consequently, the existing literature is something of a hodgepodge.
The aim of the present study is to use the extant literature as a point of departure for a fresh re-appraisal of possible musical functions. In Part 1 of our study, we summarize the results of an extensive literature survey concerning the possible functions of music. Specifically, we identified and skimmed hundreds of publications that explicitly suggest various functions, uses, or benefits for music. We provide separate overviews for the empirical literatures and the theoretical literatures. This survey resulted in just over 500 proposed musical functions. We do not refer to each of the identified publications but concentrate on the ones that have identified either more than one single function of music listening or a single unique function that is not captured in any other publication. In Part 2, we present the results of an empirical study whose purpose was to distill-using principal components analysis (PCA)-the many proposed functions of music listening. To anticipate our results, we will see that PCA suggests three main dimensions that can account for much of the shared variance in the proposed musical functions.

REVIEW OF THE RESEARCH ON THE FUNCTIONS OF MUSIC
Discussions and speculations regarding the functions of music listening can be found in both theoretical literature concerning music as well as in empirical studies of music. Below, we offer a review of both literatures. The contents of the reviews are summarized in Tables A1, A2. Table A1 provides an overview of theoretical proposals regarding musical function, whereas Table A2 provides an overview of empirical studies regarding musical function. Together, the two tables provide a broad inventory of potential functions for music.

THEORETICAL APPROACHES
Many scholars have discussed potential functions of music exclusively from a theoretical point of view. The most prominent of these approaches or theories are the ones that make explicit evolutionary claims. However, there are also other, non-evolutionary approaches such as experimental aesthetics or the uses-andgratifications approach. Functions of music were derived deductively from these approaches and theories. In addition, in the literature, one commonly finds lists or collections of functions that music can have. Most of these lists are the result of literature searches; in other cases authors provide no clear explanation for how they came up with the functions they list. Given the aim of assembling a comprehensive list, all works are included in our summary.

Functions of music as they derive from specific approaches or theories
Evolutionary approaches. Evolutionary discussions of music can already be found in the writings of Darwin. Darwin discussed some possibilities but felt there was no satisfactory solution to music's origins (Darwin, 1871(Darwin, , 1872. His intellectual heirs have been less cautious. Miller (2000), for instance, has argued that music making is a reasonable index of biological fitness, and so a manifestation of sexual selection-analogous to the peacock's tail. Anyone who can afford the biological luxury of making music must be strong and healthy. Thus, music would offer an honest social signal of physiological fitness.
Another line of theorizing refers to music as a means of social and emotional communication. For example, Panksepp and Bernatzky (2002, p. 139) argued that in social creatures like ourselves, whose ancestors lived in arboreal environments where sound was one of the most effective ways to coordinate cohesive group activities, reinforce social bonds, resolve animosities, and to establish stable hierarchies of submission and dominance, there could have been a premium on being able to communicate shades of emotional meaning by the melodic character (prosody) of emitted sounds.
A similar idea is that music contributes to social cohesion and thereby increases the effectiveness of group action. Work and war songs, lullabies, and national anthems have bound together families, groups, or whole nations. Relatedly, music may provide a means to reduce social stress and temper aggression in others. The idea that music may function as a social cement has many proponents (see Huron, 2001;Mithen, 2006;Bicknell, 2007).
A novel evolutionary theory is offered by Falk (2004a,b) who has proposed that music arose from humming or singing intended to maintain infant-mother attachment. Falk's "puttingdown-the-baby hypothesis" suggests that mothers would have profited from putting down their infants in order to make their hands free for other activities. Humming or singing consequently arose as a consoling signal indicating caretaker proximity in the absence of physical touch.
Another interesting conjecture relates music to human anxiety related to death, and the consequent quest for meaning. Dissanayake (2009), for example, has argued that humans have used music to help cope with awareness of life's transitoriness. In a manner similar to religious beliefs about the hereafter or a higher transcendental purpose, music can help assuage human anxiety concerning mortality (see, e.g., Newberg et al., 2001). Neurophysiological studies regarding music-induced chills can be interpreted as congruent with this conjecture. For example, music-induced chills produce reduced activity in brain structures associated with anxiety (Blood and Zatorre, 2001).
Related ideas stress the role music plays in feelings of transcendence. For example, (Frith, 1996, p. 275) has noted that: "We all hear the music we like as something special, as something that defies the mundane, takes us "out of ourselves," puts us somewhere else." Thus, music may provide a means of escape. The experience of flow states (Nakamura and Csikszentmihalyi, 2009), peaks (Maslow, 1968), and chills (Panksepp, 1995), which are often evoked by music listening, might similarly be interpreted as forms of transcendence or escapism (see also Fachner, 2008).
More generally, Schubert (2009) has argued that the fundamental function of music is its potential to produce pleasure in the listener (and in the performer, as well). All other functions may be considered subordinate to music's pleasureproducing capacity. Relatedly, music might have emerged as a safe form of time-passing-analogous to the sleeping behaviors found among many predators. As humans became more effective hunters, music might have emerged merely as an entertaining and innocuous way to pass time during waking hours (see Huron, 2001).
The above theories each stress a single account of music's origins. In addition, there are mixed theories that posit a constellation of several concurrent functions. Anthropological accounts of music often refer to multiple social and cultural benefits arising from music. Merriam (1964) provides a seminal example. In his book, The anthropology of music, Merriam proposed 10 social functions music can serve (e.g., emotional expression, communication, and symbolic representation). Merriam's work has had a lasting influence among music scholars, but also led many scholars to focus exclusively on the social functions of music. Following in the tradition of Merriam, Dissanayake (2006) proposed six social functions of ritual music (such as display of resources, control, and channeling of individual aggression, and the facilitation of courtship).
Non-evolutionary approaches. Many scholars have steered clear of evolutionary speculation about music, and have instead focused on the ways in which people use music in their everyday lives today. A prominent approach is the "uses-and-gratifications" approach (e.g., Arnett, 1995). This approach focuses on the needs and concerns of the listeners and tries to explain how people actively select and use media such as music to serve these needs and concerns. Arnett (1995) provides a list of potential uses of music such as entertainment, identity formation, sensation seeking, or culture identification.
Another line of research is "experimental aesthetics" whose proponents investigate the subjective experience of beauty (both artificial or natural), and the ensuing experience of pleasure. For example, in discussing the "recent work in experimental aesthetics," Bullough (1921) distinguished several types of listeners and pointed to the fact that music can be used to activate associations, memories, experiences, moods, and emotions.
By way of summary, many musical functions have been proposed in the research literature. Evolutionary speculations have tended to focus on single-source causes such as music as an indicator of biological fitness, music as a means for social and emotional communication, music as social glue, music as a way of facilitating caretaker mobility, music as a means of tempering anxiety about mortality, music as escapism or transcendental meaning, music as a source of pleasure, and music as a means for passing time. Other accounts have posited multiple concurrent functions such as the plethora of social and cultural functions of music found in anthropological writings about music. Non-evolutionary approaches are evident in the usesand-gratifications approach-which revealed a large number of functions that can be summarized as cognitive, emotional, social, and physiological functions-and the experimental aesthetics approach, whose proposed functions can similarly be summarized as cognitive and emotional functions.

Functions of music as they derive from literature research
As noted, many publications posit musical functions without providing a clear connection to any theory. Most of these works are just collections of functions of music from the literature. Not least, there are also accounts of such collections where it remained unclear how the author(s) came up with the functions contained. Some of these works refer to only one single function of music-most often because this functional aspect was investigated not with the focus on music but with a focus on other psychological phenomena. Yet other works list extensive collections of purported musical functions.
Works that refer to only one single functional aspect of music include possible therapeutic functions for music in clinical settings (Cook, 1986;Frohne-Hagemann and Pleß-Adamczyk, 2005), the use of music for symbolic exclusion in political terms (Bryson, 1996), the syntactic, semantic, and mediatizing use of film music (Maas, 1993), and the use of music to manage physiological arousal (Bartlett, 1996).
The vast majority of publications identify several possible musical functions, most of which-as stated above-are clearly focused on social aspects. Several comprehensive collections have been assembled, such as those by Baacke (1984), Gregory (1997), Ruud (1997), Roberts and Christenson (2001), Engh (2006), and Laiho (2004. Most of these studies identified a very large number of potential functions of music.
By way of summary, there exists a long tradition of theorizing about the potential functions of music. Although some of these theories have been deduced from a prior theoretical framework, none was the result of empirical testing or exploratory datagathering. In the ensuing section, we turn to consider empiricallyoriented research regarding the number and nature of potential musical functions.

EMPIRICAL INVESTIGATIONS
A number of studies have approached the functions of music from an empirical perspective. Two main approaches might be distinguished. In the first approach, the research aim is to uncover or document actual musical functioning. That is, the research aims to observe or identify one or more ways in which music is used in daily life. In the second approach, the research goal is to infer the structure or pattern underlying the use of music. That is, the research aims to uncover potential basic or fundamental dimensions implied by the multiple functions of music. This is mostly done using PCA or factor analyses or cluster analyses that reduce a large number of functions to only a few basic dimensions. In some cases, the analyses are run exploratively whereas in other cases, they are run in a confirmatory way, that is-with a predefined number of dimensions. The empirical studies can be categorized according to several criteria (see Table A2). However, when discussing some of the most important works here, we will separate studies where respondents were asked for the functions of music in open surveys from studies where the authors provided their own collections of functions, based on either literature research or face validity.

Surveys about the functions music can have
A number of studies have attempted to chronicle the broad range of musical functions. Most of these studies employed surveys in which people were asked to identify the ways in which they make use of music in their lives. In some studies, expert interviews were conducted in order to identify possible functions. Table A2 provides a summary of all the pertinent studies including their collections of functions and-where applicable-their derived underlying dimensions. We will restrict our ensuing remarks to the largest and most comprehensive studies. Chamorro-Premuzic and Furnham (2007) identified 15 functions of music among students and subsequently ran focus groups from which they distilled three distinct dimensions: emotional use, rational use, and background use. Some of the largest surveys have been carried out by Boer (2009). She interviewed more than a thousand young people in different countries and assembled a comprehensive collection of musical functions. Using factor analysis, she found 10 underlying dimensions: emotion, friends, family, venting, background, dancing, focus, values, politic, and culture. (Lonsdale and North, 2011, Study 1) pursued a usesand-gratifications approach. They identified 30 musical uses that could be reduced to six distinct dimensions. In a related study employing a larger sample, the same authors came up with eight distinct dimensions: identity, positive and negative mood management, reminiscing, diversion, arousal, surveillance, and social interaction (Lonsdale and North, 2011, Study 4). When interviewing older participants, Hays and Minichiello (2005) qualitatively identified six dimensions: linking, life events, sharing and connecting, wellbeing, therapeutic benefits, escapism, and spirituality.
The various surveys and interview studies clearly diverge with regard to the number of different musical functions. Similarly, the various cluster and factor analyses often end up producing different numbers of distinct dimensions. Nevertheless, the results are often quite similar. On a very broad level, there are four categories that appear consistently: social functions, emotional functions, cognitive or self-related functions, and physiological or arousalrelated functions (see also Hargreaves and North, 1999;Sedlmeier, 2009, 2010).

Empirical studies using predefined collections of functions of music
Apart from the open-ended surveys and interview methods, a number of studies investigating musical functions begin with researcher-defined collections or even categories/dimensions. Some of these predefined collections or categories/dimensions were simply borrowed from the existing published research, whereas others were derived from specific theoretical perspectives.
Empirical studies on functions of music emerging from specific theoretical approaches. Some of the above mentioned theoretical approaches to the functionality of music have been investigated in empirical studies. Boehnke and Münch (2003) developed a model of the relationship of adolescents' development, music, and media use. They proposed seven functions of music that relate to the developmental issues of young people (such as peer group integration, physical maturation, or identity development). In two studies with a large number of participants, Lonsdale and North (2011) applied the model of media gratification (from McQuail et al., 1972) and used a collection of 30 functions of music they assembled from literature research and interviews. In both studies, they ran factor analyses-reducing the number of functions to six dimensions and eight dimensions, respectively. Lehmann (1994) developed a situations-functions-preference model and proposed that music preferences emerge from the successful use of music to serve specific functions for the listener, depending on the current situation. Lehmann identified 68 ways in which people use music, from which he was able to reduce them to 15 music reception strategies (Rezeptionsweisen) such as compensation/escapism, relaxation, and identification. Misenhelter and Kaiser (2008) adopted Merriam's (1964) anthropological approach and attempted to identify the functions of music in the context of music education. They surveyed teachers and students and found six basic functions that were quite similar to the ones proposed by Merriam (1964). Wells and Hakanen (1997) adopted Zillmann's (1988a,b) mood management theory and identified four types of users regarding the emotional functions of music: mainstream, music lover, indifferent, and heavy rockers.

Empirical studies on functions of music emerging from literature research.
A number of studies have made use of predefined musical functions borrowed from the existing research literature. The significance of these functions and/or their potential underlying structure has then been empirically investigated using different samples. As mentioned, not all of those studies tried to assemble an exhaustive collection of musical functions in order to produce a comprehensive picture of the functions of music; but many studies were focused on specific aspects such as the emotional, cognitive, or social functions of music. Schäfer and Sedlmeier (2009) collected 17 functions of music from the literature and found functions related to the management of mood and arousal as well as self-related functions to be the ones that people highly ascribe to their favorite music. Tarrant et al. (2000) used a collection of 10 functions of music from the literature and factor analyzed them resulting in three distinct dimensions of music use: self-related, emotional, and social. Sun and Lull (1986) collected 18 functions of music videos and were able to reduce them to four dimensions: social learning, passing time, escapism/mood, and social interaction. Melton and Galician (1987) identified 15 functions of radio music and music videos; and Greasley and Lamont (2011) collected 15 functions of music, as well. Ter Bogt et al. (2011) collected 19 functions of music from the literature and used confirmatory factor analysis to group them into five dimensions. In a clinical study with adolescents, Walker Kennedy (2010) found 47 functions of music that could be reduced to five dimensions.
By way of summary, extant empirical studies have used either an open approach-trying to capture the variety of musical functions in the course of surveys or questionnaire studies-or predefined collections of functions as they resulted from specific theoretical approaches or from literature research. These different approaches have led to quite heterogeneous collections of possible musical functions-from only few functions posited by a specific hypothesis, to long lists arising from open surveys. Moreover, although the many attempts to distill the functions of music to fewer dimensions have produced some points of agreement, the overall picture remains unclear.

THE STRUCTURE AMONG THE FUNCTIONS OF MUSIC
With each successive study of musical functions, the aggregate list of potential uses has grown longer. Questionnaire studies, in particular, have led to the proliferation of possible ways in which music may be relevant in people's lives. Even if one sidesteps the question of possible evolutionary origins, the multitude of hundreds of proposed functions raises the question of whether these might not be distilled to a smaller set of basic dimensions.
As noted earlier, previous research appears to converge on four dimensions: social functions (such as the expression of one's identity or personality), emotional functions (such as the induction of positive feelings), cognitive or self-related functions (such as escapism), and arousal-related functions (such as calming down or passing time). These four dimensions might well account for the basic ways in which people use music in their daily lives.
Notice that cluster analysis and PCA/factor analysis presume that the research begins with a range of variables that ultimately capture all of the factors or dimensions pertaining to the phenomenon under consideration. The omission of even a single variable can theoretically lead to incomplete results if that variable proves to share little variance in common with the other variables. For example, in studying the factors that contribute to a person's height, the failure to include a variable related to developmental nutrition will led to deceptive results; one might wrongly conclude that only genetic factors are important. The validity of these analyses depends, in part, on including a sufficient range of variables so that all of the pertinent factors or dimensions are likely to emerge.
Accordingly, we propose to address the question of musical functions anew, starting with the most comprehensive list yet of potential music-related functions. In addition, we will aim to recruit a sample of participants covering all age groups, a wide range of socio-economic backgrounds, and pursue our analysis without biasing the materials to any specific theory.

FUNDAMENTAL FUNCTIONS OF MUSIC-A COMPREHENSIVE EMPIRICAL STUDY
The large number of functions of music that research has identified during the last decades has raised the question of a potential underlying structure: Are there functions that are more fundamental and are there others that can be subsumed under the fundamental ones? And if so, how many fundamental functions are there? As we have outlined above, many scientists have been in search of basic distinct dimensions among the functions of music. They have used statistical methods that help uncover such dimensions among a large number of variables: factor analyses or cluster analyses.
However, as we have also seen, the approaches and methods have been as different as the various functions suggested. For instance, some scholars have focused exclusively on the social functions of music while others have been interested in only the emotional ones; some used only adolescent participants while others consulted only older people. Thus, these researchers arrived at different categorizations according to their particular approach. To date, there is still no conclusive categorization of the functions of music into distinct dimensions, which makes psychological studies that rely on the use of music and its effects on cognition, emotion, and behavior still difficult (see also Stefanija, 2007). Although there exist some theoretically driven claims about what fundamental dimensions there might be (Tarrant et al., 2000;Laiho, 2004;Schubert, 2009;Lonsdale and North, 2011), there has been no large-scale empirical study that analyzed the number and nature of distinct dimensions using the broad range of all potential musical functions-known so far-all at once.
We sought to remedy this deficiency by assembling an exhaustive list of the functions of music that have been identified in past research and putting them together in one questionnaire study. Based on the research reviewed in the first part of this study, we identified more than 500 items concerned with musical use or function. Specifically, we assembled an aggregate list of all the questions and statements encountered in the reviewed research that were either theoretically derived or used in empirical studies. Of course, many of the items are similar, analogous, or true duplicates. After eliminating or combining redundant items, we settled on a list of 129 distinct items. All of the items were phrased as statements in the form "I listen to music because . . . " The complete list of items is given in Table A3, together with their German versions as used in our study.

METHOD
Participants were asked to rate how strongly they agreed with each item-statement on a scale from 0 (not at all) to 6 (fully agree). When responding to items, participants were instructed to think of any style of music and of any situation in which they would listen to music. In order to obtain a sample that was heterogeneous with regard to age and socioeconomic background, we distributed flyers promoting the Internet link to our study in a local electronics superstore. Recruitment of participants was further pursued via some mailing lists of German universities, students from comprehensive schools, and members of a local choir. As an incentive, respondents got the chance to win a tablet computer. A total of 834 people completed the survey. Respondents ranged from 8 to 85 years of age (M = 26, SD = 10.4, 57% female).
Notice that in carrying out such a survey, we are assuming that participants have relatively accurate introspective access to their own motivations in pursuing particular musical behaviors, and that they are able to accurately recall the appropriate experiences. Of course, there exists considerable empirical research casting doubt on the accuracy of motivational introspection in self-report tasks (e.g., Wilson, 2002;Hirstein, 2005;Fine, 2006). These caveats notwithstanding, in light of the limited options for gathering pertinent empirical data, we nevertheless chose to pursue a survey-based approach.

RESULTS
Principal component analysis revealed three distinct dimensions behind the 129 items (accounting for about 40% of the variance), based on the scree plot. This solution was consistent over age groups and genders. The first dimension (eigenvalue: 15.2%) includes statements about self-related thoughts (e.g., music helps me think about myself), emotions and sentiments (e.g., music conveys feelings), absorption (e.g., music distracts my mind from the outside world), escapism (e.g., music makes me forget about reality), coping (e.g., music makes me believe I'm better able to cope with my worries), solace (e.g., music gives comfort to me when I'm sad), and meaning (e.g., music adds meaning to my life). It appears that this dimension expresses a very private relationship with music listening. Music helps people think about who they are, who they would like to be, and how to cut their own path. We suggest labeling this dimension selfawareness. The second dimension (eigenvalue: 13.7%) includes statements about social bonding and affiliation (e.g., music helps me show that I belong to a given social group; music makes me feel connected to my friends; music tells me how other people think). People can use music to feel close to their friends, to express their identity and values to others, and to gather information about their social environment. We suggest labeling this dimension social relatedness. The third dimension (eigenvalue: 10.2%) includes statements about the use of music as background entertainment and diversion (e.g., music is a great pastime; music can take my mind off things) and as a means to get into a positive mood and regulate one's physiological arousal (e.g., music can make me cheerful; music helps me relax; music makes me more alert). We suggest labeling this dimension arousal and mood regulation. All factor loadings are reported in Table A3.
In order to analyze the relative significance of the three derived dimensions for the listeners, we averaged the ratings for all items contained in each dimension (see Figure 1). Arousal and mood regulation proved to be the most important dimension of music listening closely followed by self-awareness. These two dimensions appear to represent the two most potent reasons offered by people to explain why they listen to music, whereas social relatedness seems to be a relatively less important reason (ranging below the scale mean). This pattern was consistent across genders, socioeconomic backgrounds, and age groups. All differences between the three dimensions are significant (all ps < 0.001). The reliability indices (Cronbach's α) for the three dimensions are α = 0.97 for the first, α = 0.96 for the second, and α = 0.92 for the third dimension.

GENERAL DISCUSSION
Since the earliest writing on the psychology of music, researchers have been concerned with the many ways in which people use music in their lives. In the first part of this paper, we reviewed literature spanning psychological, musicological, biological, and anthropological perspectives on musical function. The picture that emerged from our review was somewhat confusing. Surveying the literature from the past 50 years, we identified more than 500 purported functions for music. From this list, we assembled a somewhat catholic list of 129 non-redundant musical functions. We then tested the verisimilitude of these posited functions by collecting survey responses from a comparatively large sample. PCA revealed just three distinct dimensions: People listen to music to achieve self-awareness, social relatedness, and arousal and mood regulation. We propose calling these the Big Three of music listening.
In part one of our study we noted that several empirical studies suggest grouping musical functions according to four dimensions: cognitive, emotional, social/cultural, and physiological/arousal-related functions. This raises the question of how our three-dimensional result might be reconciled with the earlier work. We propose that there is a rather straightforward interpretation that allows the four-dimensional perspective to be understood within our three-dimensional result. Cognitive functions are captured by the first dimension (self-awareness); social/cultural functions are captured by the second dimensions (social relatedness); physiological/arousal-related functions are captured by the third dimension (arousal and mood regulation); and emotional functions are captured by the first and third dimensions (self-awareness + arousal and mood regulation). Notably-as can be seen with the items in Table A3-there is a dissociation of emotion-related and mood-related functions. Emotions clearly appear in the first dimension (e.g., music conveys feelings; music can lighten my mood; music helps me better understand my thoughts and emotions), indicating that they might play an important role in achieving self-awareness, probably in terms of identity formation and self-perception, respectively. However, the regulation of moods clearly appears in the third dimension (e.g., music makes me cheerful; music can enhance my mood; I'm less bored when I listen to music), suggesting that moods are not central issues pertaining to identity. Along with the maintenance of a pleasant level of physiological arousal, the maintenance of pleasant moods is an effect of music that might rather be utilized as a "background" strategy, that is, not requiring a deep or aware involvement in the music. The regulation of emotions, on the other side, could be a much more conscious strategy requiring deliberate attention and devotion to the music. Music psychology so far has not made a clear distinction between music-related moods and emotions; and the several conceptions of music-related affect remain contentious (see Hunter and Schellenberg, 2010). Our results appear to call for a clearer distinction between moods and emotions in music psychology research.
As noted earlier, a presumed evolutionary origin for music need not be reflected in modern responses to music. Nevertheless, it is plausible that continuities exist between modern responses and possible archaic functions. Hence, the functions apparent in our study may echo possible evolutionary functions. The three functional dimensions found in our study are compatible with nearly all of the ideas about the potential evolutionary origin of music mentioned in the introduction. The idea that music had evolved as a means for establishing and regulating social cohesion and communication is consistent with the second dimension. The idea of music satisfying the basic human concerns of anxiety avoidance and quest for meaning is consistent with the first dimension. And the notion that the basic function of music could have been to produce dissociation and pleasure in the listener is consistent with the third dimension.
In light of claims that music evolved primarily as a means for promoting social cohesion and communication-a position favored by many scholars-the results appear noteworthy. Seemingly, people today hardly listen to music for social reasons, but instead use it principally to relieve boredom, maintain a pleasant mood, and create a comfortable private space. Such a private mode of music listening might simply reflect a Western emphasis on individuality: self-acknowledgement and well-being appear to be more highly valued than social relationships and relatedness (see also Roberts and Foehr, 2008;Heye and Lamont, 2010).
The results of the present study may be of interest to psychologists who make use of music as a tool or stimulus in their research. The way people usually listen to music outside the laboratory will surely influence how they respond to musical stimuli in psychological experiments. For those researchers who make use of music in psychological studies, some attention should be paid to how music is used in everyday life. The three dimensions uncovered in this study can provide a parsimonious means to identify the value a person sets on each of three different types of music use. It is also conceivable that individual patterns of music use are related to personality traits, a conjecture which may warrant future research.
With regard to music cognition, the present results are especially relevant to studies about aesthetic preferences, style or genre preferences, and musical choice. Recent research suggests that musical functions play an important role in the formation and development of music preferences (e.g., Schäfer and Sedlmeier, 2009;Rentfrow et al., 2011). It will be one of the future tasks of music cognition research to investigate the dependence of music preference and music choice on the functional use of music in people's lives.
By way of summary, in a self-report study, we found that people appear to listen to music for three major reasons, two of which are substantially more important than the third: music offers a valued companion, helps provide a comfortable level of activation and a positive mood, whereas its social importance may have been overvalued.