Abstract
This systematic literature review aims to identify and critically examine the prevailing general trends of music education research that addresses issues of improvisation from 1985 to 2015. The study examined the main features of studies with impact that focus on musical improvisation and have been published in peer-reviewed music education journals. Data were organised on the basis of the following: 1) General publication features; 2) Topic; 3) Methodological approach; 4) Participant features; 5) Type of improvisation; 6) Definition of improvisation; 7) Findings; 8) Suggestions for practice. The study also takes a close look at the construction of the discourses through which improvisation has been framed in the field of music education, providing insights on how such discourses create particular pedagogical visions of improvisation. To this end, we have created a map of the different visions of improvisation pedagogy that the studied works point towards. These visions have been clustered in the following five categories: (i) from rupture of certainties to creative problematisation; (ii) return to the “natural” beginning—in search of humanness; (iii) improvisation as a learning tool; (iv) conserving and enlivening traditions; (v) improvisation as an impetus for creativity. The map proposed in this study is meant as a possible representation of the general trends that underpin music education research focusing on improvisation. This map can also be seen as a “tool” through which music educators can situate their practice and reflect on their particular ways of working with improvisation, possibly envisioning alternative ways forward.
Keywords
Introduction
That improvisation should be part of music education is now rather commonplace. We frequently hear that “[i]mprovisation is an important part of the young child’s life” (Brophy, 2001, p. 36) and therefore should be part of her/his education in music because it “is an essential feature of the art of making music” (Campbell, 2009, p. 140). At the same time, criticisms over the ways in which improvisation can and has been incorporated into education have sometimes led to rather aphoristic positions: “what we claim to be ‘teaching’ as improvisation in schools is not
While being aware of the contingency of our thinking, we aim to resist oversimplifications that create barriers to a critical approach of the educational relevance of improvisation. As Blum (1998) argues, improvisation has advanced through modernity as a “marked” term, that is, as a term always defined and construed the concept of “the composer” emerged in direct conjunction with a perceived opposition between “composition” and “improvisation.” It was in the decades around 1500 that new ideas began to be articulated, not only about musical authorship and the distinct professional identity of composers, but also about the difference between the composition as object, on the one hand, and improvisation as a practice, on the other. (p. 477)
As a result, improvisation has often been understood as the opposite of careful performance preparation (“on the spur of the moment”), unforeseeable, (“ex-improviso”), random (“fortuita”), an act deprived of reflection, an act that ignores any notion of adherence to rules (“sine meditatione”, “sans régle ni dessein”) (Blum, 1998). These conceptualisations, however, have advanced side by side with a perception of improvisation as a window towards unmediated freedom, as an act of transcending boundaries, imposed logics, and calculated modes of conduct (Blum, 1998; Kanellopoulos, 2013; Kramer, 2008; Landgraf, 2011; Piotrowska, 2012; Woodring-Goertzen, 1998).
It could be argued that this ambivalent perception of improvisation is a manifestation of the irreconcilable struggle that is the result of core modernist dualities: originality vs. stylistic meticulousness, immediacy vs. thorough planning of large forms, breaking away from habits and memory vs. creating perfection that endures in the form of complete musical works in accordance with the This largely modern sense of improvisation is built around a core antinomy: improvisation is recognized as a process that makes inroads towards musical, personal and sociopolitical freedom, and at the same time it is cast as a “pre-artistic”, fatally incomplete and largely marginal creative process. (p. 42)
The view of improvisation as an always-incomplete glimpse into unchartered freedom, and at the same time as a dangerous pathway to triviality and a threat to disciplined musical conduct might partly account for the—until recently—characteristic neglect of improvisation in musicology and philosophy of music (Bertinetto, 2013) and also for the rather defensive and resistant approach to improvisation that many music educators often adopt. We feel, nevertheless, that the ways in which this general condition has influenced music education requires a nuanced and systematic look at the ways in which music education practice and research have approached improvisation. This research is but a small step in this direction.
Research questions
This systematic literature review aims to identify and critically examine prevailing general trends of music education research that addresses issues of improvisation. 1 As an indicator that an article has had some impact in our field, we have used the 10-citation rule. Furthermore, this study takes a close look at the construction of discourses through which improvisation has been framed in the field of music education, providing insights on how such discourses create particular pedagogical visions of improvisation. In this sense, it comes close to Mantie’s (2013) critical examination of discourses constructed through “popular music pedagogy” scholarly studies.
This aim has led to the formulation of the following research questions:
(1) What are the main features of studies that address issues of musical improvisation and have been published in peer-reviewed music education journals?
(2) What visions of improvisation pedagogy emerge through the approaches to improvisation that these studies take?
The contribution of our study to knowledge advancement may be seen as twofold. First, we aim at identifying general features of music education studies that address issues of improvisation. This has been the result of an extensive content analysis and the descriptive statistics it yielded. In this sense, this study complements review studies such as those of Running (2008), Henry (1996), Rohwer (1997), and more recently Chandler (2018), who have focused on creativity, composition, creativity assessment, and improvisation in elementary general music respectively. Secondly, and on a more interpretative level, we aim at understanding how the notion of improvisation, its role and value for musical practice, and its educative potential have been construed through these studies. To this end, we will propose a conceptual map that dynamically represents (a) the different
Research design
In this study, our ambition has been to go beyond summarising research findings in the area of improvisation pedagogy. This research can be seen as an
Method
Our study focuses on improvisation studies published in music education scholarly journals between 1985 and 2015. We have included studies from the mid-1980s onwards since it was during that time that music education research began to exercise an increasing and considerable influence on the content and rationale of music education curricula on an international scale. From the mid-1980s onwards, scholarly research journals began to give voice to research developments that reflected the lessons learned from the radical initiatives that had been growing since the 1960s (Finney, 2011; Paynter & Salaman, 2008). It was during that time that music education steadily advanced towards acknowledging the need for a sustained and critical dialogue between (a) psychologically informed research traditions, (b) radical teaching initiatives stemming from the creative music in education movement, and (c) everyday multilevel actual teaching concerns (Grashel & LeBlanc, 1998; Roulston, 2006; Swanwick 2008; Welch et al., 2004; Yarbrough, 1984, 1996). 2 These advancements gave rise to the publication of a variety of music education research journals in the 1990s and the 2000s; moreover, numerous music education research methods textbooks began to appear internationally, acknowledging the need both for more diverse methodologies and for studying a greater variety of music education practices (Colwell, 1992; Kemp, 1988, 1992; Phelps, 1980; Phillips, 2008). 3
Sample selection
The sample of our study consists of papers published in leading music education journals. The journals were drawn from the Finnish Publication Forum (JUFO).
4
Eighteen music education journals were identified by this system, out of which twelve are ranked by JUFO as level 1, five as level 2, and one as level 3, the highest level of the ranking. Online search engines such as Jstor, Sage, ProQuest, Cambridge, Taylor&Francis, and Informit as well as the journals’ own web pages were used. In those cases where online access was not available, searches were performed manually. When possible, multiple sources were used in order to crosscheck findings. The headword used was
Our decision to study works with some impact on the field of music education led us to use the 10-citations rule, meaning that articles with less than 10 citations at the time of conducting this study (academic year 2015–2016) were excluded from our analysis. To discover the citation count for each article, we used Google Scholar. 6 Citation analysis has previously been used in journal content analysis as a tool for identifying journal prestige in music education research (Hamann & Lucas, 1998), and influential studies and prominent trends of music education research (Diaz & Silveira, 2014; Rutkowski, Thompson, & Huang, 2011; C. P. Schmidt & Zdzinski, 1993). Although not unproblematic (Bornmann & Daniel, 2008; van Raan, 2004; Woolgar, 1991), citation count is considered to be a fairly reliable indicator of research impact (Bornmann, Mutz, Neuhaus, & Daniel, 2008): “Citation-based bibliometric analysis provides indicators of international impact, influence” (van Raan, 2004, p. 27).
The use of this tool allows us to create a representative picture of prevailing trends in music education research that addresses issues of improvisation, leaning on studies that can be seen as having a strong impact in our field. However, the use of the 10-citations rule induces a limitation: as citation frequency increases gradually over time (Hancock, 2015), post-2011 papers had less than 10 citations, and had to be excluded from our analysis. This selection process decreased the number of studies included from 185 (published in 17 research journals) to 77 (in 11 journals). Table 1
7
shows the music education journals we looked at based on JUFO; it also shows
List of music education journals (
Frequency and relative frequency of articles published in 1985–2015 with the headword improvis in the title or abstract. **Frequency and relative frequency of articles with = > 10 citations (Google Scholar in 2015) and included in the study sample.
Analysis procedures
The analysis began with reading each of the 77 articles a minimum of three times. Data were organised on the basis of a rubric used to record each article. This rubric included the following: 1) General publication features, 2) Topic, 3) Methodological approach, 4) Participant features, 5) Type of improvisation, 6) Definition of improvisation, 7) Findings, 8) Suggestions for practice.
The methodological approach used in each study (no. 3 in the list above) was further categorised as follows: quantitative, qualitative, mixed methods, practice-driven descriptive essays, philosophical, or literature review. Type of improvisation (no. 5 in the list above) included instrumental improvisation, vocal improvisation, solo improvisation formats, group improvisation, and improvisation genre. The latter was further categorised as: western art music; popular; jazz/blues; world musics; children’s songs/singing games; tonal, non-genre-specific; “free” 8 music; not specified. For studies that focused on more than one genre, a mark was placed in all relevant categories.
In studies with empirical data, participant features (no. 4 in the list above) were categorised as follows: 9 1) level of education, 2) gender, 3) ethnicity, 4) marginality, 10 and 5) music involvement. In addition, we recorded the country where the data were collected. In order to refrain from making assumptions, only specific information regarding gender, ethnicity, and marginality was used. If no details were given, data were classified as “not specified”, aligning with Ebie (2002). If the information aligned with more than one category, a mark was placed in all relevant categories. Level of education was categorised in the following way: birth to kindergarten (ages 0–6), primary (ages 6–12), secondary (intermediate, high school, ages 12–18), tertiary (college/university, 18–), and professional (teachers, musicians).
Participants’ music involvement was coded as systematic (instrumental tutoring of more than 1 year, music teachers, or further education in music) or casual (general teachers, non-music majors, no or less than 1 year of experience in learning a musical instrument). For studies that had participants with a variety of music involvement and/or main instruments a mark was placed in each relevant category. The categories used in our rubric were decided on the basis of a brief review of content analysis studies (Ebie, 2002; Kratus, 1995; Rutkowski et al., 2011; Silveira & Diaz, 2014; Tirovolas, & Levitin, 2011; Yarbrough, 1984). The first stage of the analysis resulted in condensed descriptions of each of the 77 articles. To answer the first research question, descriptive statistics were elicited on the basis of the rubric presented above.
The second stage aimed at identifying the visions of improvisation pedagogy that emerged (research question 2). We first created a list of possible approaches to improvisation inspired by interdisciplinary literature on improvisation (including historical and cultural musicology, ethnomusicology, theatre studies, literary theory, music education, and music therapy). The list served as an abductive hypothesis, enabling the researchers to “enter the field with the deepest and broadest theoretical base possible and develop their theoretical repertoires throughout the research process” (Timmermans & Tavory, 2012, p. 180; Agar, 1999). Thus, our study borrows the logic of instrumental case studies, where abductive reasoning can be employed using already developed “instruments and preconceived coding schemes” (Stake, 1994b, p. 243) in the process of analysis.
Supplementary visits to the original articles were made in order to crosscheck and crystallise our interpretation of the approaches that prevailed in each study. The emerging characterisations of the approaches were compared against and parallel to each other, ensuring comprehensiveness and accuracy of interpretation. As Timmermans and Tavory (2012) have argued, “abduction reflects the process of creatively inferencing and double-checking these inferences with more data” (p. 168), with the aim of looking for plausible “new concepts” that might meaningfully account for new data. This process of analysis (see Figure 1) led to the identification of 11 approaches to improvisation. Exploring the pedagogical implications of these 11 approaches, and the ways in which they were related to each other in the data, in pairs or groups, led to the proposition of five overarching themes that describe the visions of improvisation pedagogy in these studies.

Selection of research material and process of analysis.
Results – Research Question 1: What are the main features of studies that address issues of musical improvisation and have been published in peer-reviewed music education journals?
Τopics studied
The scope of the studies proved to be broad, employing a number of theoretical and methodological perspectives informed by a variety of disciplines. In order to present an overall view of the studied topics, we compared and grouped all relevant information, ending up with nine headings that include subtopics addressed (see Table 2). Each article was placed under one particular heading. Topics related to musical development were the most frequent (31.2%), steadily attracting music education researchers (for an overview of changes in topics studied across time see Figure 2).
List of topics studied and subtopics addressed, frequency and relative frequency of topics studied in the data sample (

Relative frequency of topics studied in the data sample (
More articles dealing with practical teaching methods seem to have been published at the beginning of the period under study, while studies dealing with how teachers feel about teaching have been more frequent in recent times. A focus on issues of meaning-making in improvisation and musical responsiveness has also been a rather recent development. Also, four studies, three by a prominent scholar in music education (Bresler, 2005, 2006, 2009) and one by an important theorist of qualitative research (Stake, 1994a), draw on music improvisation as a metaphor and model for understanding the creative fluidity that inheres in the process of carrying out qualitative research.
Methodological approaches
Quantitative methods had the highest representation (36.4%), followed by qualitative approaches (24.7%) (Figure 3). Prior to the year 2000 only three studies (3.9%) employed qualitative methods. However, after 2000 employment of qualitative methods began to rise significantly. Qualitative research approaches include ethnography, grounded theory, action research, case study, naturalistic inquiry, and narrative inquiry. Practice-driven descriptive essays (19.5%) reached a peak prior to the year 2000, gradually decreasing afterwards, possibly due to the rise in qualitative studies as well as to an increasing interest in the pursuit of philosophical approaches to improvisation (11.7%). Studies with empirical data (

Relative frequency of methodological approaches in the study sample (
Features of the empirical studies
Data generation techniques
Solo improvisation tasks were the most popular data generation technique, measuring individual effort with or without an accompaniment. This was used in 40.4% (
Participants
School students (primary and secondary) were the focus of 44.2% (

Frequency and relative frequency of participants’ level of education in empirical studies (
A total of 61.5% of the studies focused on participants with systematic music involvement (Table 3), with a prevalence of wind instruments (31.3%). Instruments were not specified in 53.1% of studies with empirical data, particularly in studies with music teachers or tertiary music education students. Only four studies (Burnard, 2000a, 2000b, 2002; Mang, 2005) provided a detailed description of the participants’ ethnic background. Most studies (53.8%) were conducted in North America (Figure 5), 30.8% took place in Europe, while one study employed distribution of questionnaires in North and South America, as well as in Australia (that of Madura Ward-Steinman, 2007).
Frequency* and relative frequency of level of music involvement in empirical studies (
If the information aligned with more than one category, a mark was placed in all categories.

Frequency and relative frequency of country where data were collected in empirical studies (n=52).
Type of improvisation
Instrumental improvisation was the focus in 40.4% (
Most studies (38.5%) focused on tonal but non-genre-specific music (Figure 6); 30.8% of the studies focused on jazz and blues improvisation genres. This was followed by “free” music (19.2%). An explicit focus on western art music, world musics, and popular musics was particularly rare.

Frequency and relative frequency of music genre in empirical studies (n=52).
Results – Research Question 2: What visions of improvisation pedagogy emerge through the approaches to improvisation that these studies take?
Visions of improvisation pedagogy in music education research
Our analysis yielded a set of five visions of improvisation pedagogy, which manifest themselves through eleven ways of approaching improvisation and improvising. In this paper we argue that particular visions of improvisation pedagogy lead to concrete pedagogical actions that take place in

A map of prevailing visions of improvisation pedagogy as they emerge through the approaches to improvisation that music education research studies address.
Below, we present the five visions of improvisation pedagogy and the approaches to improvisation in a non-hierarchical order.
Vision I: From rupture of certainties to creative problematisation
This vision of improvisation pedagogy sees improvisation as a means for cultivating a more
The roots of this vision can be traced back to the experimental music practices of the post-war era (Kutschke, 1999; Nyman, 1999; Reynolds, 1965). Envisioning improvisation in education as a means of creative becoming is closely connected to approaching improvisation as an
An emphasis on rupture entails an approach to improvisation as a mode of music making that poses and problematises issues of how we live together, addressing issues of personal freedom and socio-musical inequalities:
Vision II: Return to the “natural” beginning—in search of humanness
This vision rests on a more psychologically-oriented stance, paying particular attention to a student’s personality and its moulding. It is shaped by an understanding of improvisation as
This vision adopts a broader view of improvisation as a
Vision III: Improvisation as a learning tool
This vision approaches improvisation as a
Studies that adopt this vision value improvisation as a
Adherence to this vision leads to pedagogical work that uses improvisation as a means for deepening and expanding learned skills (Addison, 1988), focusing on technical and psychological skills that are integral to music-making (Addison, 1988; Beegle, 2010), leading to musical development (Harrison & Pound, 1996) as well as contributing to an enhanced appreciation of music (Parisi, 2004). Furthermore, this vision has significantly contributed to the development of a body of research that uses improvisation as a tool for assessing aspects of musicianship, or for determining the level of musical or skill development (e.g., Beegle, 2010; Guilbault, 2004; Paananen, 2006); this has also contributed to a body of literature that relates to the content and the structure of aural skills curricula (Azzara & Grunow, 2003; Spiegelberg; 2008). It must be noted, however, that concerns have been raised as to whether music educators’ employment of improvisation as a learning strategy does justice to the complexities of improvisation practice (see, e.g., Hickey, 2009).
Vision IV: Conserving and enlivening traditions
As a result of the intersections between ethnomusicology, jazz studies and music education (Berliner, 1994; Elliott, 1995; Nettl, 2012; Sudnow, 1993), a growing body of music education studies seem to acknowledge the various roles that improvisation plays in a variety of musical traditions. Thus, they approach improvisation as a
Vision V: Improvisation as an impetus for creativity
This vision values improvisation for its contribution to the generation of ideas, and as a
Discussion
A call for broadening the scope of research
In this article we have explored some general features of studies with impact that address music improvisation and were published in peer-reviewed music education journals between 1985 and 2015. Our study shows that research that addresses improvisation in secondary school and community music contexts, as well as studies that focus on participants with varied musical backgrounds and experiences are still far from becoming a widely acknowledged and discussed subfield. Important inroads might also need to be paved by future studies on improvisation in world musics (
Furthermore, future research might need to pursue more closely intermedia improvisation practices in education, as well to develop “practice
Our study demonstrates that the role of improvisation in inclusive practices, and its potential contribution to social cohesion through empowering students who can be described as socially, economically, or culturally marginalised, has not achieved the prominence we feel it deserves. To argue for more research in that direction does not of course imply that improvisation should be seen merely as a remedy to issues of community building. There is a need for critical approaches to improvisation and its relation to notions of power, and to how improvisation creates its own (hidden or explicit) hierarchies. To that we should add the value of researching improvisation as a mode of creative practice in the face of contemporary educational contexts, which have imposed dramatic changes in the role of creativity in education (Kalin, 2018; Kanellopoulos, 2015).
Moreover, it seems to us that future music education research might need to develop stronger links with the burgeoning field of improvisation studies, with experiments with improvisation and radical problematisations that come from the fields of critical musicology (e.g., Stefanou, Ragkou, Peki, Pazarloglou, & Papoutsi, 2016; Székely, 2008), historical musicology (e.g., Wegman, 1996), and philosophy of music (e.g., Goehr, 2016). It is noteworthy that, with one exception (MacGlone & MacDonald, 2017), none of the important edited volumes that focus on improvisation and were published after 2015 contain a single essay on its educative dimensions (Born, Lewis, & Straw 2017; Caines & Heble, 2015; Lewis & Piekut, 2016a, 2016b; Siddall & Waterman, 2016). 11
A tool for further reflection
In response to our second research question, this paper has also proposed a map of different visions of improvisation pedagogy that the investigated studies point towards. The proposed five visions of improvisation pedagogy, with the 11 different approaches to improvisation towards which they point, illustrate the plurality that exists in how improvisation has been understood in the literature reviewed in this study. One important conclusion that can be drawn is that music education studies have moved beyond the mysticism that used to surround past approaches to improvisation, a mysticism that denied any sort of role for improvisation in the process of education (see Watson, 2010).
Our data show a strong preference for

Frequency and relative frequency of approaches to improvisation adopted in the study sample (n=77).

Frequency of approaches to improvisation in studied topics.
Our data confirm that, far from being a marginal and peripheral mode of musical practice in music education, improvisation has become a way of addressing, highlighting, and cultivating qualities that are of core importance to music and its role in human lives. It can therefore be said that music education studies that address issues of musical improvisation have indeed tried to inquire into improvisation’s links to core aspects of what it means to be musically educated, and the sometimes irreconcilable struggle between conflicting forces that this process induces: how to enable students to delve into extant modes of musical practice without impeding their spontaneity; how to enable critical reasoning while fostering community building; how to develop modes of study that are close to students’ natural learning processes while advancing technical mastery; how to allow for innovative thinking while preserving long-cherished traditions “authentically”. Thus, music education’s apprehension of improvisation seems to have gone beyond the freedom vs. triviality polarity mentioned at the start of this article.
The map proposed in this study is meant as a possible representation of general trends that underpin music education research that addresses improvisation. In addition, we suggest that this map may also function as a way of conceptualising the tensions that arise in different music education situations where improvisation plays a part. Thus, it can be used as a framework for situating our particular ways of working with improvisation in our everyday teaching practice. In this sense, the visions of improvisation pedagogy proposed in this paper might work as a map that assists our reflection on the
In this sense, in her/his everyday engagement with improvisation, every music teacher “produces” a new version of the map. However, as van Manen (1991) aptly states, “[a]s I reflect pedagogically on my daily living with children I discover my pedagogical nature, its present limits and possibilities” (p. 532). Thus, every version of the map may be subject to change, as one reflects upon and experiments with different approaches to the question of what role improvisation should play in our everyday teaching practice. Different answers to the question of the educational value of improvisation produce different visions of improvisation pedagogies, thus creating distinctive “pedagogical moments” of improvisation. Our map can be seen as a tool through which music educators can situate their practice and reflect upon it, possibly envisioning alternative ways forward. As such, it is an example of how theory might inform practice.
Footnotes
Acknowledgements
Sincere thanks are due to the community of researchers at the Music Education doctoral seminar of the Sibelius Academy, University of the Arts Helsinki; also to Danae Stefanou, Eleftheria Tseliou, Heidi Westerlund and Christopher TenWolde for their constructive comments at various stages of this project. We would also like to thank the anonymous reviewers of RSME for their sharp comments and critical observations.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is part of the ArtsEqual project (project no. 293199) funded by the Academy of Finland’s Strategic Research Council and its Equality in Society program.
Notes
Author biographies
![]()
