Since sociology has been historically concerned with the question of how social order is possible, its main focus has been on exploring (the imposition of) conventions, rules, norms, every day routines, etc. As a result, the study of creative processes has been mostly left aside. This has changed within the last years, but sociology is still lacking profound studies of creative processes, such as composing, improvising, playing, inventing, etc.

My contribution is aimed at filling this theory driven gap. For this purpose, I conducted an empirical research project to explore improvising processes in music from the perspective of the sociology of action and interaction. Two genres were chosen to enable an explicitly contrastive analysis so that a higher level of generalization could be achieved: free jazz and flamenco.

A qualitative, ethnographic data collection strategy was applied. Data were collected in three steps: first, audio and visual recordings were made of the performances of three free jazz trios and two flamenco ensembles (a duo and a trio), all of them professional musicians, in a recording studio. One at a time, the groups were invited to the recording studio of the Institute for Composition and Electroacoustic of the University for Music and Visual Arts in Vienna to perform improvisations, so that participant observation could be made while they performed. In a further stage, the ensembles were confronted – immediately after playing – with the audio-visual material in which they could see and hear their performances, the audio- visual data functioning as a trigger for the musicians to make free comments, observations, and evaluations about their improvisations. Also, as part of this second step, a non-structured interview with the group was carried out. Afterwards, the musicians were asked to name the best improvisation piece of the session. In a third step, the researcher accompanied each of the musicians separately to the technical room and heard with him or her the selected piece of music, carrying out an individual in-depth interview, within which the artist told her what she/he thought and did during each sequence of the piece. This way, the subjective experience of each musician and his/her ways of acting regarding the action-interaction process of improvisation could be examined, all of this being again video-audio recorded. As a result, the same selected piece of music had been examined and narrated from the point of view of each of the musicians, which enabled to reconstruct the process of improvisation at the level of the individuals and at the level of the interaction within the group.
The method employed for the analysis of the observation protocols, and the individual and group interviews was grounded theory, a qualitative, flexible and systematic strategy that enables the formulation of a theory based on the analysed data and that does not force empirical data into theoretical corsets.

As a result, an action and interaction model emerged, that aims to explicate improvisation processes from the double perspective of action and interaction theory. It consists of four dimensions: 1) the (malleability of the) musical material, 2) the (more or less intense) interaction between the players, 3) the attitude of the actors (from “open” to “closed”), and 4) the (more or less emergent) music. These dimensions are closely intertwined with one another and each of them can vary from a minimum to a maximum.

Improvising is modelling musical material. By musical material, a key category within my model, I mean not only perceptible sounds, scores or instruments, but also the knowledge that musicians carry with them and that is incorporated as sedimented experience. Depending on the musical genre, the material is more or less malleable, i.e., can be formed while performing. The more malleable the material is, the more intense is the interaction between the musicians that respond to new or unanticipated forms of it. In an “open attitude mode”, the artists experience a high receptivity towards the musical materials offered by the others within the performance and respond according to them: imitating, continuing musical ideas, etc. A high degree of the first three dimensions leads to a high degree of the emergent quality of the music performed. The more emergent the music is, the more it results from the playing process. At the same time, a high emergent music influences the other dimensions back.

The advantage of the proposed model consists in the fact that it does not postulate a single criterion (i.e. primary vs. secondary musical parameters) to decide whether improvisation is or is not taking place, and conceives of it as a continuum from minimum to maximum. It takes into consideration the mutual influences of four central, empirically driven dimensions of improvising processes. Including such “variables” that have been normally left aside by improvisation models, like musical material, or attitude, the creative practice of improvising can be comprehended in a more holistic, broaden way. The four dimensions discussed, as well as their mutual influences, which could be complemented and completed in further studies, show that, for musical improvising, the knowledge of the artists (their material), their attitude towards contingency, their interaction, and their music as output of their action are mutually influential factors that determine, on the whole, creativity in performance.