Abstract
The purpose of this study was to review flow research in music contexts from 1975 until the first quarter of 2019. Specifically, frequencies/percentages were calculated for (a) output in five-year periods; (b) publication type; and (c) methodologies employed, including measurement instruments used. Content analyses were also conducted on topics covered. Using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology, a total of 3,341 records were examined, with 95 studies eventually included in the analysis. Findings indicated a steady increase in research output over the last 25 years. Studies overwhelmingly sampled participants from Western populations, especially the United States. The majority of quantitative studies used self-report questionnaires, of which those developed by Jackson and colleagues were most prevalent. Among the qualitative approaches, interviews and observations were the most common means of data collection. Topics covered in the studies reviewed include the psychophysiological aspects of flow, transmission and group experience of flow, the association of flow with a range of positive outcomes, factors that contribute to flow experiences, and flow experiences of young children. Implications for future research were proffered in light of the findings.
Sometimes a person reports having an experience of extreme joy, a feeling of ecstasy for no apparent good reason: a bar of haunting music may trigger it, or a wonderful view, or even less—just a spontaneous sense of well-being. (Csikszentmihalyi, 1990, p. 49)
Extreme joy. Ecstasy. Indeed, music seems particularly germane to humans experiencing flow—the phenomenological state of deep enjoyment and absorption in an activity (Csikszentmihalyi, 1990). Csikszentmihalyi’s writings on flow frequently incorporate discussions of music. He points out, for example, that composing is an enjoyable and rewarding activity (Csikszentmihalyi, 1975) and those who play musical instruments tend to be in flow states more often (Csikszentmihalyi, 1997). For Csikszentmihalyi, close listening to music not only prevents one from being bored and anxious, but also facilitates flow experience, and so important is music to him that he laments that whenever there are budget cuts in schools, music is often one of the first subjects to be removed from the curriculum (Csikszentmihalyi, 1990).
Flow research in music contexts has been gathering momentum in recent years (e.g., Clementson, 2018; Cohen & Bodner, 2019). Chirico, Serino, Cipresso, Gaggioli, and Riva (2015) reviewed 10 quantitative flow studies concerning music and found that dispositional flow was the most frequently examined. Kim (2016) examined seven studies on flow in music contexts and showed how being in flow enhanced musicians’ ability to perform. Keay (2018) analyzed the content of 51 studies on flow and music and classified them into three categories: music performance, music teaching, and music consumption. Some other recent explorations in flow research include the physiobiological aspects of flow (e.g., de Manzano, Theorell, Harmat, & Ullén, 2010), the association between flow and emotion (e.g., Marin & Bhattcharaya, 2013), and group flow (e.g., Bishop, 2018; Sawyer, 2016). However, as far as can be determined, there is no published study that has sought to comprehensively document, review, and present the overall landscape of flow research in music contexts from the initial formulation of flow theory. Such a study is important as it can offer insights and point towards new directions.
Accordingly, the purpose of this study was to review flow research in music contexts from 1975 until the first quarter of 2019. The year 1975 was chosen because, according to Csikszentmihalyi (1990), the “first book that described the flow experience directly was Beyond Boredom and Anxiety (Csikszentmihalyi, 1975)” (p. 242). Specifically, frequencies and percentages were calculated for (a) output in 5-year periods; (b) publication type (e.g., journal articles, book chapters); and (c) methodologies employed (e.g., quantitative, qualitative) and methods including measurement instruments used. Content analyses of the topics covered were also conducted. The specific research questions were:
What was the change in the number of studies from 1975 until the first quarter of 2019?
Which methodologies were used?
What topics did the studies cover?
What are the implications of the findings for future research?
Method
Methodological guidelines from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA; Moher, Liberati, Tetzlaff, Altman, & the PRISMA Group, 2009) were used to generate data for this review. Relevant literature published between 1975 and the first quarter of 2019 were searched on four databases, namely, PsycINFO, PubMed, Web of Science, and Google Scholar. The search string used were “music” AND “flow.” The two authors were the reviewers for the screening of the records. Any discrepancies between the two authors were to be discussed and moderated by a third reviewer. Backward citation search was also employed to identify any records that may have been left out during the database search.
The initial search was conducted on April 12, 2019 and yielded 3,341 records (PsycINFO: 423; PubMed: 341; Web of Science: 1,537; Google Scholar: 1,040). After the removal of 555 duplicates, 2,786 records were screened by their titles and abstracts. Of the 2,786 records, 2,637 were eliminated as they were not relevant (e.g., not related to music, different kinds of flow, such as blood flow and air flow). The authors then retrieved the full texts of the remaining 149 records and screened them critically for their eligibility according to the following inclusion criteria: (1) The full text was available and written in English, (2) music was the primary domain examined, and (3) the understanding of the term “flow” followed that of Csikszentmihalyi (e.g., Csikszentmihalyi 1975, 1990). Dissertations and theses were excluded. Five articles were added from the citations of the full-text articles; additionally, two articles were suggested by colleagues in the field. Finally, a total of 95 records were included in this review. Subsequently, the writings were content analyzed for publication types, demographics, methodologies, and topics covered (Andrés, 2009; Hart, 1998; Petticrew & Roberts, 2006). To ascertain accuracy and consistency, the findings were cross-checked between the two authors (e.g., Morse, Barrett, Mayan, Olson, & Spiers, 2002). Figure 1 documents the details of the process undertaken to generate the data for this present review.
The PRISMA flowchart.
Results
Figure 2 presents the research output in five-year periods. As far as can be determined from the methodological procedures outlined above, the first published study in which music was the principal domain examined was Custodero’s (1998) observation of flow experiences among young children engaged in musical learning. Following Custodero’s work, we found a steady increase in flow research output, from two studies by the end of 1999 to 95 by the first quarter of 2019, indicating a sustained rising interest over the last two and a half decades (see Figure 2).
Flow in music contexts: Research output in five-year periods.
Of the 95 records, 85 (89.5%) were in the form of journal articles, while the remaining were book chapters (n = 5) and conference papers (n = 5). Of the journal articles, 55 (64.7%) were published in journals focused on music, music psychology, music education, and the arts, while the remaining 30 (35.3%) appeared in journals that were not focused on music and the arts but in areas such as medical science, psychology, vocational behavior, motivation, emotion, and personality. Journals that published the most research articles on flow in music contexts were the Psychology of Music (n = 14), Frontiers in Psychology (n = 8), Musicae Scientiae (n = 6), and Music Education Research (n = 5). These four journals accounted for over one third (38.8%) of all published journal articles on flow in music contexts.
Among the empirical studies, U.S. participants were the most frequently sampled (n = 30). This was followed by the UK (n = 12), Australia (n = 10), Portugal (n = 5), Canada, Italy, and Singapore (n = 4 each), Brazil, Germany, Spain, and Sweden (n = 3 each), Austria, Belgium, Finland, Ireland, Israel, and Slovenia (n = 2 each), and Greece, The Netherlands, and Romania (n = 1 each). Consistent with psychological research in other domains, most researchers sampled participants from “WEIRD” (Western, Educated, Industrialized, Rich, and Democratic) countries (Henrich, Heine, & Norenzayan, 2010; King & McInerney, 2014).
Analysis of methodologies
Of the 95 records, 43 (45.2%) were quantitative, 26 (27.4%) employed qualitative approaches, 13 (13.7%) used mixed methods, and the remaining 13 (13.7%) were non-empirical (e.g., theoretical/conceptual articles and literature reviews). Empirical research, therefore, comprised the majority (86.3%) of research on flow in music contexts. Tables 1 to 4 present summaries of articles grouped according to the various methodologies. Within each table, the studies are presented chronologically, from the earliest to the most recent. For each study, we provided information on the participant(s), topic(s), study design, methodology, and key findings.
Quantitative Studies (n = 43).
Qualitative Studies (n = 26).
Mixed Methods Studies (n = 13).
Non-Empirical (n = 13).
Among studies that used quantitative measures, the vast majority (52/56 studies = 92.9%) used self-report questionnaires. The flow scales designed by Jackson and colleagues (Jackson & Eklund, 2002, 2004; Jackson & Marsh, 1996; Jackson, Eklund, & Martin, 2010; Jackson, Martin, & Eklund, 2008; Martin & Jackson, 2008) were the most frequently used (32/56 = 57.1%; e.g., Cohen & Bodner, 2018; de Manzano et al., 2010; Fritz & Avsec, 2007; Valenzuela & Codina, 2014). These scales quantify flow experience either as a disposition (Dispositional Flow Scale) or as a state (Flow State Scale); they are also available in long (36 items) and short versions (9 items). This was followed by the Experience Sampling Method (Csikszentmihalyi & Csikszentmihalyi, 1988; Csikszentmihalyi & Larson, 1987; Csikszentmihalyi, Larson, & Prescott, 1977), where participants had to complete a self-report questionnaire whenever they received a signal from a pager they had with them (4/56 = 7.1%; e.g., Clementson, 2018; O’Neill, 1999). In four out of the 56 studies (7.1%), the authors developed their own questionnaires (e.g., Bloom & Skutnick-Henley, 2005; Diaz, 2011). Four studies used questionnaires adapted from other studies; two of these (Cunha & Carvalho, 2012, 2013) adapted Custodero’s (1997, 2005) Flow Indicators in Musical Activities (FIMA) to assess flow experience during music learning activities. Two studies (3.6%; Ruth, Spangardt, & Schramm, 2017; Stupacher, Witte, & Wood, 2017) used the Flow Short Scale (Engeser & Rheinberg, 2008; Rheinberg et al., 2003).
Among the four studies that did not use self-report questionnaires, authors sought to observe and quantify flow (e.g., Custodero, 1998; Gloor, Oster, & Fischbach, 2013). Custodero (1998) developed the Flow Indicators in Musical Activities (FIMA) measure to code visible signs of flow experience in young children during music learning activities, while Gloor, Oster, and Fischbach (2013) observed a jazz band and measured participants’ flow using the oscillations and changes in the level of body movement with sociometric badges.
Interview was the most common qualitative approach and was used in 21/39 (53.8%) of all studies including qualitative components. These interviews were often semi-structured and sought to elicit responses on how the participants felt when engaged in various musical activities (e.g., Bailey & Davidson, 2002; Garces-Bacsal, Cohen, & Tan, 2011). Sample interview questions included, “How do you feel when you are singing?” (Bailey & Davidson, 2002, p. 252), and “How did you feel when you started the programme? How do you feel now?” (Ascenso, Perkins, Atkins, Fancourt, & Williamon, 2018, p. 14).
The second most typical qualitative approach involved observations of participants (17/39 = 43.6%). These observations were often video-recorded (e.g., Custodero, 2005; St. John, 2006; Waite & Diaz; 2012). One common instrument used to code video data was Custodero’s FIMA (1997, 2005) (e.g., Cunha & Carvalho, 2011). Another common source of data collection was fieldnotes (e.g., Clementson, 2018; St. John, 2006). For instance, Clementson (2018) recorded fieldnotes while observing rehearsals to capture the actions of the participants, such as interactions between students and comments made. These fieldnotes were then analyzed together with the rehearsal transcripts.
Eight out of 39 studies including qualitative components (20.5%; e.g., Bernard, 2009; Bloom & Skunick-Henley, 2005; Lamont, 2012) employed neither interviews nor observations. These studies used approaches such as qualitative surveys (e.g., Bloom & Skunick-Henley, 2005) and narrative writing describing and reflecting on experiences (e.g., Bernard, 2009; Lamont, 2012). The Music in Flow Survey (Bloom & Skutnick-Henley, 2005; Kirchner, Bloom, & Skutnick-Henley, 2008) included a “write-in” component that asked participants to describe the recent flow experience they had while performing a musical activity. The survey also included the elements that needed to be present in order to be considered a flow experience, such as “attention was focused on the music and not on task-irrelevant thoughts” (Bloom & Skutnick-Henley, 2005, p. 25). Narrative writing approaches included free descriptions and reflections (e.g., “Describe in your own words the strongest, most intense experience of music that you have ever had”; Lamont, 2012, p. 579).
Topics covered
In addition to the methodologies employed, Tables 1 to 4 also present lists of topics covered by each of the 95 records. Salient topics that emerged among studies that employed quantitative methods included music performance (e.g., Iusca, 2015; Marin & Bhattacharya, 2013; Wrigley & Emmerson, 2013), music practice (e.g., Butkovic, Ullén, & Mosing, 2015; de Araujo & de Andraude, 2013; Miksza & Tan, 2015), performance anxiety (e.g., Khalsa & Cope, 2006), and motivation (e.g., Heller, Bullerjahn, & von Georgi, 2015; O’Neill, 1999). Among the qualitative studies, music education, especially those that involve younger participants, was a common topic (e.g., Addessi & Pachet, 2005; Chen-Hafteck & Schraer-Joiner, 2011; Cunha & Carvalho, 2011; Custodero, 1998). Other topics examined using qualitative approaches included group/collective flow (e.g., Hytönen-Ng, 2013; Sawyer, 2006) and creativity (e.g., Hill, Hill, & Walsh, 2018; Lage-Gómez & Cremades-Andreu, 2019; Morrow, 2013).
Among some of the more recent developments in flow research, researchers have found relationships between flow and a range of psychophysiological measures. De Manzano et al. (2010) examined 21 pianists from Sweden and found positive associations between flow and activity of the zygomaticus major facial muscles, minimum blood pressure, heart rate variability, and respiratory depth. Inverse relationships were also found between flow and heart period, maximum blood pressure, and pulse pressure. In a study involving two conductors, six opera singers, and six dancers, Thomson and Jaque (2011) found decreased levels of regulatory capacity and cardiac autonomic balance when performers were experiencing flow. The association between flow and emotion has also been examined. For example, Marin and Bhattcharaya’s (2013) study involving 76 piano performance students found that daily practice and trait emotional intelligence were predictors of flow.
Researchers have found that flow can be transmitted from one person to another and also experienced in groups (i.e., group flow). Bakker (2005) aimed to test two hypotheses: (1) that job resources (e.g., social support and coaching by one’s supervisor) can positively influence music teachers’ challenge–skill balance, which then contributes to their likelihood of experiencing flow; and (2) that through emotional contagion, there can be a crossover of flow from music teachers to their students. Participants were 178 music teachers and 605 students from the Netherlands. Findings indicated empirical support for the two hypotheses. Accordingly, the author stresses the importance of having sufficient job resources and that these resources are not only important for teachers, but also for the well-being of their students. Drawing from music and theatre, Sawyer (2006) presents examples of group creativity, unpacks group flow and posits “When a group is performing at its peak” (p. 158) as its definition. He also explains how ensemble musicians who are making music together in a state of group flow function optimally as a team, gel as one musical entity, and work collectively towards shared musical goals.
Flow has also been found to be associated with a range of positive outcomes. MacDonald, Byrne, and Carlton (2006) examined the relationships between flow, creativity, and the quality of group compositions by 45 first-year university students from Scotland. Findings indicated a positive relationship between flow and creativity. Fritz and Avsec (2007) investigated how flow experiences when engaged in a range of musical activities (e.g., solo and orchestral performances, rehearsals) were related to the subjective well-being of 84 Slovenian music academy students. Findings indicated that several dimensions of flow (e.g., clear goals, challenge–skill balance, and concentration) were positively associated with subjective well-being. In cross-cultural research involving band students from the United States and Singapore, significant positive associations were found between flow, grit, and commitment to band (Miksza, Tan, & Dye, 2016; Tan & Miksza, 2018).
A number of researchers have found factors that contribute to flow experiences. O’Neill (1999) examined the motivational and social factors of 60 music students from the UK. Findings indicated that high achieving students and non-specialist students experienced flow more than moderate achieving students. Bailey and Davidson’s (2002) phenomenological study sought to determine the psychological and physiological benefits, if any, of active group singing for seven members of a Canadian choir of homeless males. The themes that emerged included numerous clinical and engagement benefits that aligned with flow theory. Diaz (2011) examined how a short task designed to induce mindfulness meditation affected American university students’ flow, aesthetic response, and perceived attention while listening to music. Findings indicated that the mindfulness task increased participants’ ability to focus on listening without being distracted, thus altering their listening experiences. Participants also reported differences between their aesthetic responses and flow, as each explained a different type of positive and heightened affective experience.
Finally, young children have been found to experience flow while engaged in a range of musical activities. Custodero (2002, 2005) examined young children’s flow experiences in different music learning environments and found that development, environment, and the temperament of children influence musical engagement. She also suggests that music educators acknowledge music making as an absorbing and rewarding activity before considering flow as a framework to evaluate children’s musical engagement. Addessi and Pachet (2005) observed young children (3–5 years old) in Italy engage with an interactive musical system called “the Continuator.” This system has the ability to produce music as though a human were playing the keyboard. Csikszentmihalyi’s flow theory was then used as an interpretive lens through which to examine how the system attracted and sustained the interests of the participants.
Discussion
The purpose of this study was to review flow research in music contexts from 1975 until the first quarter of 2019. Findings indicate that the first published work on flow in music context was by Custodero (1998), that is, 23 years after the publication of Beyond Boredom and Anxiety (Csikszentmihalyi, 1975)—the first book to describe flow directly. Thereafter, we found a steady increase in research output over the last 25 years. There are several possible reasons that may explain our findings. Empirical approaches to examine flow became more established only during the late 1980s (Moneta, 2012). For example, while the experience sampling method was introduced in 1977, the first studies using it to explore flow were not published until 1987 (Csikszentmihalyi & Larson, 1987; Csikszentmihalyi et al., 1977). This may explain why flow research in music contexts only began to gain momentum from the beginning of the 21st century. Furthermore, the increase in research output might have been influenced in part by a similar upward trend in the larger field of positive psychology during the same period (Kim, Doiron, Warren, & Donaldson, 2018), as well as the relatively more developed research field on flow in sport contexts (Swann, Keegan, Piggott, & Crust, 2012; Swann, Piggott, Schweickle, & Vella, 2018).
The findings of this study have several implications for future research. To begin with, flow research in music contexts may benefit from a greater diversity of participants. The vast majority of research studies covered in this paper examined participants from “WEIRD” populations. Singapore (e.g., Tan & Miksza, 2018) was the only Asian country covered. Citing a study of top psychological journals which found that “96% of subjects were from Western industrialized countries—which house just 12% of the world’s population,” Henrich et al. (2010) argued that in order to “understand human psychology, behavioral scientists must stop doing most of their experiments on Westerners” (p. 29). Research on populations beyond those in WEIRD countries and studies involving cross-cultural comparisons (e.g., Tan & Miksza, 2018) may contribute to a more culturally-responsive and nuanced understanding of flow in music contexts. In writing about flow and music, Csikszentmihalyi (1990) noted how in “every known culture, the ordering of sound in ways that please the ear has been used extensively to improve the quality of life” (p. 108). Future research that examines flow experiences of participants making music in non-Western musical settings, such as the Gamelan ensemble, may offer additional insights and advance research on flow in music contexts.
This study also offers some fruitful methodological implications. Among investigations using quantitative approaches, the vast majority (92.9%) used self-report questionnaires. The relative lack of studies that seek to quantify flow through observation may be due to practical constraints, as it is logistically more challenging to observe large numbers of participants. Further theorizing to determine the observable, measurable characteristics of flow more precisely—assuming this is possible—may suggest additional ways of collecting data for comparison with data gathered from self-report questionnaires. Among investigations using qualitative approaches, interviews and observations were the most common means of data collection (79.5%). Less common approaches, such as qualitative surveys (e.g., Bloom & Skunick-Henley, 2005) and narrative writing (e.g., Bernard, 2009; Lamont, 2012) may be employed by more researchers to expand the range of possibilities for qualitative research on flow. As these involve writing rather than speaking (interviews) or doing (observations), new insights may be gleaned provided participants find writing easier than talking, which may not necessarily be the case. Another advantage of collecting data in written form is to reduce the potential for interviewer effects influencing participants’ responses (Kreuter, 2008; Meho, 2006).
Our content analysis of topics covered in the studies reviewed has implications for flow researchers. In particular, researchers may build on the work of recent explorations in flow research, such as group flow (e.g., Bishop, 2018; Sawyer, 2006). The fact that music-making and music education often happen in groups (Tan & Miksza, 2019) may have contributed to interest in how flow experienced individually can be transferred to others (Bakker, 2005), and how flow is experienced communally (Sawyer, 2006). As Csikszentmihalyi (1990) noted, “Singing in a choir and playing in an amateur string ensemble are two of the most exhilarating ways to experience the blending of one’s skills with those of others” (p. 112). Further research on flow beyond that which is experienced by individuals may provide opportunities to obtain new insights for music and music education. For example, while self-report questionnaires on individual flow have been used in music contexts, researchers could similarly adapt group or collective flow questionnaires (e.g., Salanova, Rodríguez-Sánchez, Schaufeli, & Cifre, 2014) to examine group flow in musical contexts quantitatively. Work on the psychophysiological aspects of flow (e.g., de Manzano et al., 2010; Jaque, Karamanukyan, & Thomson, 2015) could also be explored further. In particular, interdisciplinary work involving the domains of music and medical science may be helpful when exploring how flow induced by music and/or musical activities could be valuable for therapeutic and rehabilitative purposes.
One major limitation of this study was that it covered only records retrieved using PRISMA guidelines. As one of the inclusion criteria was that articles and other sources should be written in English, we omitted research published in other languages. Future studies could examine flow research in music contexts published in other languages for the purposes of comparison. Furthermore, in limiting our search to “flow” as conceptualized by Csikszentmihalyi, we might have left out research on experiences that could have been considered to be flow but were not termed as such. While Csikszentmihalyi has made significant contributions to coining the term “flow,” theorized it extensively and supported the theory with empirical support, the experience itself may well go further than that described in the literature, academic and professional, that is currently available. It may be helpful for researchers to examine similar experiences that are termed in different ways, such as the Chinese notion of wuwei (Slingerland, 2003; Tan, 2016).
Notwithstanding these limitations, this review has provided a bird’s eye view of the research literature and made some recommendations for the future. Sustained research endeavors on flow in music contexts will doubtless contribute to our ever-growing understanding of how Csikszentmihalyi’s (1990, p. 49) “bar of haunting music may trigger” the “experience of extreme joy” and “feeling of ecstasy.” And this makes life worth living.
