back printable version

Problems in definitions of composition and improvisation in jazz and jazz education

My Ph.D research concerned definitions of jazz. It was an exploratory study that examined the ways in which jazz is defined differently in education than in what I called the ‘real world’ outside it. In the work, I went on to investigate the nature of this distinction, and also to explore the more general issue of the changes that occur to musical styles when they enter educational contexts. The research examined the literature of both jazz and jazz education, and included long interviews with jazz musicians who were also educators. This paper begins with more specific observations about composition and improvisation in jazz and jazz in education, and then goes on to consider some more general findings.

A. Composition and Improvisation

In his seminal article about improvisation, ‘Thoughts on Improvisation: a Comparative Approach’ (1974), Nettl questions the Western distinction between improvisation and composition and argues instead for the concepts of ‘fast’ and ‘slow composition’. Taking a cross-cultural approach, he suggests a continuum between two kinds of musician. One, like Schubert or musicians from the Middle Eastern and Indian improvised traditions, writes or invents at great speed, in a way which he describes as basically improvisatory, whether on paper or by ear. The other, like Beethoven or Bruckner, continually reworks and achieves a different kind of coherence. In both cases he identifies the ‘model’ as crucial. The model is the essence of the work, the modal configuration, the given pitch or rhythm context or the theoretical terms used to describe musical elements. For Nettl, the relationship between the finished text and its models should be the focus of any analysis. Whether music is improvised or composed, the focus of musicological endeavour should be the examination of the density with which given models are used, the degree of audibility of such models and the closeness of the finished product to the model. Seen in this way, the identification of what the models are in the first place is immaterial.

A central finding of my work was that classical music was an important reference point for both writers and interviewees in their discussions of jazz. Many definitions of jazz referred to classical music or compared it to classical music in some way. However, such comparisons occurred more frequently in education, where definitions concerning the status, value and musical processes of jazz all tended to be defined through direct or indirect reference to the status, value and musical processes of classical music. Gabbard (1995) also notes this phenomenon and suggests it may be due to a concern intrinsic to education to legitimate teaching and research in whatever is being taught.

Evidence of the tendency to link jazz with classical music was to be found, for example, in that:

The works of Duke Ellington were described as like Bach, Mozart, and tended to valued as having unities or coherences of various kinds associated with classical compositions.

Standards used to define such canons tended to be those of classical music and were focused, for example, on valuing motive, harmony and counterpoint as found in jazz performances. In discussion of jazz in education, for example, counterpoint was a feature described as desirable, even though it is a feature not normally defined as valuable by real world musicians of the time. Owens’ (1974) examination of motivic development in Charlie Parker’s improvising and Stewart’s Schenkerian analysis of the solos of Clifford Brown are other examples of the use of analytical procedures grounded in classical musicology.

At the same time, the literature indicates that specialised jazz musicological techniques are beginning to develop. For example, Gunther Schuller’s idea that the various bands of Miles Davis’ long career carry his ‘fingerprint’ defines a different kind of coherence on his work. Spring and Bash both use the idea of the ‘formula’ as a way of defining melodic material found in jazz improvisation in ways that are more flexible than Reti’s ‘motive’. There was also evidence in the literature of new ways of defining improvisation, group interaction, self-expression and the role of the groove in jazz. Jazz studies is gradually coming of age.

In the view of Krin Gabbard, academics and educators in jazz are still concerned primarily with defining canons rather than with discussing how they come to be constructed, or subverting them. Marcia Citron discusses the processes by which canons develop through the definition of what she calls sets of parameters, which articulate the standards by which music is seen as valuable. In Citron’s terms, the parameters used by educators to assign value to jazz by creating such canons tend to be those associated with the canons of classical music. Specifically, my findings indicate that the parameters used to discuss jazz often refer more to compositional aspects of the music than to those concerning improvisation.

In short, the educational playing-field is sloped in direction of classical music. One result of this is that in education improvisation is often defined as in some way ‘harder’ for classical musicians. The work of jazz educators, it seems, is therefore focused not simply on teaching jazz but also on what might be termed unlearning classical music. Educators teach jazz by overcoming the barriers set up by classical music in education to the learning of jazz style. Jazz educators interviewed, for example, consistently discussed working on the ‘classical rhythm habits’ of their students and their students’ problems with working by ear and with improvising.

B. Other findings

Other related but more general findings concerning real world jazz and jazz in education and are summarised below:

The substyles of jazz

Bebop and fusion were found to be more blurred in the real world, while in education they were defined in more clearcut ways.

Bebop was found to be more central to jazz in education than outside it.

Fusion was found to be less central and also educationally ‘harder’ than other styles, and was therefore put later in curriculum sequences. This was in direct contradiction to accounts of students taught, and indeed of the experiences of the interviewees themselves, both of whom said they had come to jazz initially through real world experience of bands such as Weather Report and Miles Davis’ Bitches Brew work in the late 1960s.

more recent or ‘creative’ repertoire and improvisation vocabulary was found to feature less in discussion of jazz in education, which tended to focus on past jazz and on recreation of earlier styles.


A range of positions was found in this area. Jazz was defined as African-American, as American and also as articulating a range of other identities too, and all these positions occurred in and outside education. It was hard to discern clear cut differences in this area.

However all ethnicity was generally much less explicit in education as a defining feature. Musical elements associated with African American music-making tended to be downplayed in discussion of jazz and labels which associated specific ethnic identities and their associated performance practices were found to be used much less often in educational discussions.

Group Interaction

Qualities of social interaction were defined consistently as important, and included: listening, blending, supporting each other, complementing each other, a sense of trust. The groove was considered a vital place where such musical qualities were demonstrated by bands or individual band members in their music-making. However such qualities of group interaction were mentioned less often in education and were also valued much less.

A tension was also revealed between a need to control interaction of learning in jazz and need to allow freedom of interaction in improvising. It seems that, as educational music-making becomes more teacher-led, these interactive qualities are articulated less.

The jazz journey towards openness, growth and self-knowledge

Real world jazz musicians were also defined as on a life-long journey towards musical self-knowledge and openness through self expression and a process of musical growth.

This journey was much more emphasised in discussion of real world jazz, though it appeared in some interviewee discussion of learners too.

In the real world, data indicated that the learner had greater control over the content of music-making and has more opportunities for self-expression and group interaction.

Three main findings followed on:

A general discomfort was found with any kind of categorisation of jazz in the real world, while in education there was an opposite tendency to categorise by constructing boundaries between styles.

The range of definitions of jazz was generally much narrower and more rigid in education. As educational knowledge, jazz was generally more rigidly defined and canonical.

There was an emphasis on valuing musicological features associated with compositions, which included motive, harmony, counterpoint and piano touch. In this way, it seems the status of jazz was felt to be increased through the valuing of the same features as those valued of classical music. Composition tended to be valued over improvising, and improvisations were valued as if they were compositions.

The possibility was also considered of a feedback relationship between jazz in education and jazz in the real world. This first occurred to me while reading the accounts of more recent jazz by Stuart Nicholson and Steven Elworth. The young lions generation of the 1980s including Wynton Marsalis et al. successfully pioneered a jazz education and indeed a view of jazz history and jazz performance which focuses primarily on the celebration of the African American jazz greats of the past. While this approach has its validity and was in some ways long overdue, it has been seen as excluding other forms and certainly falls into the category of constructing clear boundaries between jazz and other styles. As such, it can be seen as going against tendencies to blur boundaries found to be characteristic of real world jazz. Can it be a coincidence that this approach to boundary construction in jazz has come from those musicians who were the first to go through formal jazz education in the 1960s and 70s?

Jazz and improvised musics seem to be more suited to the kinds of curriculum organisation which facilitate open stylistic boundaries. Yet in education, we need clear definitions of success and failure in assessments and clear course titles and curriculum structures. Put at its most challenging and also admittedly at its most theoretical, the problematic nature of definitions of real world jazz is antithetical to the need to codify and transmit it in a structured way as educational knowledge. The main implication of this is the need for educators to design new forms of educational knowledge which embody and facilitate open and flexible definitions of musical styles, while still imparting to learners a coherent set of skills and understandings.

Charles Beale
Proffessor of Jazz Piano, Royal College of Music
Senior Jazz Consultant, Associate Board of Royal Schools of Music
June 2001

Brief Bibliography

Bash, L. (1981) The Historical and Pedagogical Implications of Motivic Formulae as supplied by the Charlie Parker recordings of "Cherokee", in Jazz Research Papers 1981, Manhattan Kansas: NAJE Publications.

Citron, Marcia. J. (1993) Gender and the Musical Canon, Cambridge: Cambridge University Press.

Elworth, Steven B. (1995) Jazz in Crisis 1948-58, in Gabbard, 1995: 57-75.

Gabbard, Krin (1993) The Jazz Canon and it’s Consequences, in Annual Review of Jazz Studies 5, 1993, Rutgers Institute of Jazz Studies, Newark NJ.

Gabbard, Krin (ed.) (1995) Jazz among the Discourses, Durham and London: Duke University Press.

Nettl, Bruno (1974) ‘Thoughts on Improvisation: A Comparative Approach’, in Musical Quarterly Jan. 1974, 60, 1.

Nicholson, S. (1990) Jazz: the modern Resurgence, London: Simon and Schuster.

Owens, Thomas (1974) Charlie Parker: Techniques of Improvisation, unpublished Ph. D. Thesis, University of California, Los Angeles.

Schuller, Gunther (1986a) Musings: The Musical Worlds of Gunther Schuller, Oxford and New York: Oxford University Press.

Spring, Howard (1990) The Use of Formulas in the Improvisations of Charlie Christian, in JazzForschung Jazz Research 22, Graz Austria: Akademische Druck - u. Verlagsanstalt.

Stewart, Milton Lee (1973) Structural Development in the Jazz Improvisational Technique of Clifford Brown, unpublished Ph.D., University of Michigan