The sign of silence: Negotiating musical identities in an improvising ensemble

Group improvisation in music

This involves the creation of music as it is being played through the idiosyncratic contributions of two or more individuals, each interpreting and musically responding to the playing of the other(s). Any individual musical contribution is impromptu and is tailored to the sounds, rhythms and tonalities heard from accompanying musicians (Bastien & Hostager, 1988; Mazzola & Cherlin, 2009; Nettl & Russell, 2008; Seddon, 2005). Since all have input into the overall sound, the creativity in group improvisation is inherently social, rather than being attributable to or located within a single individual (Sawyer, 2003). As a spontaneous process of collaborative creativity effected through non-verbal social interaction, this musical practice is a unique psychological phenomenon that has not received the thorough investigation it merits (MacDonald, Wilson, & Miell, 2012).

Improvisation is a universal capacity since anyone who enjoys music can participate in group improvisation to some extent, irrespective of musical training (MacDonald, Kreutz, & Mitchell, 2012a; MacDonald & Wilson, in press; Pothoulaki, MacDonald, & Flowers, 2012). Yet study of this behaviour tends to be limited to examining the improvisation of individual performers judged to possess expertise, and has tended to focus on jazz improvisation. This music has, for example, been characterized by multiple tensions of ‘real-time’ interaction, demanding from the individual player a simultaneous focus on their own musical contribution and the contributions of those around them (Cook, 2004; Kenny, Gellrich, Parncutt, & McPherson, 2002). The importance of a shared awareness of known social practices in jazz has also been highlighted (Bastien & Hostager, 1988). Seddon (2005) proposes the concept of empathetic attunement facilitating communication between members of a jazz ensemble, according to which musicians reaching a point where each can apprehend the others’ musical perspectives through the use of six verbal or non-verbal processes in preparing a work for performance. It is further argued that the same broad processes are at play in both classical and jazz small ensembles, despite the varied degrees of constraint from scores and prearranged strategies (Seddon & Biasutti, 2009). A widespread metaphor of jazz as a ‘language’ is prevalent, with studies explaining group improvisation of music as analogous to conversation (Monson, 1996 ; Sawyer, 2005). Monson’s argument, that jazz musicians understand their practice in terms of a ‘vocabulary’ of musical components and a ‘grammar’ of how they should be put together, remains popular: to play the music you must learn to converse in the language. Yet improvisation has always been a musical component of most genres and is gaining increasing currency as a central mode of music-making, even being viewed as a ‘post-genre musical process’ (Bailey, 1993; Biasutti & Frezza, 2009; Lewis, 2007; MacDonald, Wilson, et al., 2012). As such, it is a much broader phenomenon of significant interest to social psychology, one which a rule-based language learning model of musical improvisation may be inadequate to explain.

Discursive psychology

In a similar way, discursive psychology has shown that transmission of information and generative grammar models cannot explain the full scope of verbal language use (Potter & Edwards, 1992). A central principle of discursive psychology is that talk is never just a neutral delivery of information or a statement of internal viewpoint. Any speech act inherently indicates:

By considering how talk is shaped by these considerations, researchers can enhance their understanding of communication (Potter, 2004; Potter & Wetherell, 1994). Analyzing verbal data for the structural features of discourse can illuminate what an individual’s expectations are when they engage in conversation, what aspects of a topic they treat as normal or subject to stigma, and what particular identities they support or resist for themselves (Wetherell, Taylor, & Yates, 2001). More significantly, by considering how all participants engage in a dialogue, we can see how those identity projects are shaped by perceived confirmation or challenge from others (Antaki & Widdicombe, 1998). The social construction of musical identities has been argued as central to music-making (MacDonald, Miell, Hargreaves, & Miell, 2002; MacDonald, Miell, & Wilson, 2005) and discursive psychology therefore has much to offer research on improvisation (Linson, Dobbyn, & Laney, 2012). For instance, when jazz musicians say improvising together is like language use, what function does that metaphor fulfil in maintaining a particular identity?

Jazz musicians and discourses of improvising

Our studies of jazz musicians’ talk as talk, rather than as a window onto what happens in music, have described how musicians say they manage to improvise together; explored how the context of this talk is relevant; and considered whether musicians’ talk about improvising together behaves as discourse. We progressed from the belief that, if music is created in a group, it is worth asking a group about it, and initially conducted focus group interviews with professional jazz instrumentalists living in central Scotland about their musical practice (MacDonald & Wilson, 2005; Wilson & MacDonald, 2005). While no definitions of jazz were unanimously agreed in the discussions, structural features such as maintaining swing feel and a balance of collective and individual practices were used to distinguish jazz improvising from other types of music. Participants also discussed what comprised the life of a jazz musician, positioning themselves in relation to these emergent identities. Eclectic professional practice created a demanding lifestyle; in this context, the group aspect of jazz playing is integral to related musical identities. These individuals were a limited population, all male and acquainted; their discussion of jazz practice highlighted the importance of social context to identity as a musician, but also the ways in which positive and negative aspects of this identity were foregrounded to invite particular interpretations of these identities. A second study involved individual interviews with male and female jazz musicians in London not known to the authors, aiming to tap into personal accounts to set against the ‘public’ discourse arising at focus groups of musicians (MacDonald & Wilson, 2006). This latter study mapped patterns of identity work around gender that operated to support power structures around gender, and therefore demonstrated these musicians’ discourse around their professional practice as necessarily hegemonic.

In all these data, participants discussed views and experience of playing music and improvising. The content of what they said about improvising corroborated many features other authors have observed. For instance, this passage from a bass player ostensibly reports that jazz is perceived as a language to be learned by those who play it:

I mean, there’s people who do improvised music but people would say they’re not jazzers. I’m not going to name anybody, but there’s a few people on the scene who (.) they will say it’s not (.) it’s not jazz because they’re not using the language, it hasn’t come from the tradition. So there is a, not a snobbery, but there is a, you know, there is a certain expectation of certain rules to be followed or certain traditions to have at least been studied or absorbed. (quoted in MacDonald & Wilson, 2006, p. 65)

However, the way in which this is said indicates this claim has implications for how she is presenting herself to the interviewer and the recorder. The respondent attends to the context of the research interview by not asking questions (treating this as the interviewer’s role), and stressing that she will not ‘name names’ (line 2). She also constructs two categories of people – people who do improvised music and people who say this is not jazz – without expressly aligning herself with either. Nevertheless, she recognizes a possibility that people who claim ‘jazzer’ authority might be construed as snobbish (line 4) and seeks to repair this by stressing the minimal, and hence reasonable, nature of their concern (‘certain’, ‘at least’). The elements of discourse are therefore all observable and construct a more complex understanding of the relationships between players than a straightforward exchange or interplay of musical contributions.

A consistent pattern of description brought into play to position a speaker or other people as one type of musician constitutes a repertoire: ‘a lexicon or register of terms and metaphors drawn upon to characterize and evaluate actions and events’ (Potter & Wetherell, 1994, p. 138). We identified two distinct repertoires used by jazz musicians when talking to account for what they play (Wilson & MacDonald, 2005). According to ‘mastery’, improvisation is understood and controlled through mastery of skills and knowledge; according to ‘mystery’, improvisation is instinctive and uncontrolled, arising from unfathomable inspiration. These repertoires have distinct implications for identity: mastery emphasizes achievement and authority, but suggests that anyone can improvise given time, while mystery implies innate exclusivity of improvisers, but accords them less authority in their practice.

Both repertoires are used to maintain exclusivity of a socially negotiated in-group. Within the focus groups, the repertoire of mastery was always endorsed and accepted:

while the repertoire of mystery was always treated with irony where it was deployed:

From this perspective, jazz musicians’ talk of ‘learning the language’ of jazz operates to bolster in-group identity in a fluid and competitive environment. When they talk about improvisation, jazz musicians use discourse to invite or counter particular versions of themselves and their improvising. However, the studies outlined earlier did not clarify whether this discourse simply represented a narrow professional milieu rather than improvising per se. For instance, a similar use of mastery and mystery repertoires has recently been identified in the talk of conservatoire students about playing classical music (Juuti & Littleton, 2010), and both jazz and classical musicians may access these repertoires in genre-specific ways (Dobson, 2010). Furthermore, although these studies sought to understand how musical communication might operate in group improvisation – through transmission or construction – the data rarely strayed into first person accounts of playing experience, and all participants played with different people. There was therefore a need to explore further how diverse types of musicians talked about a shared experience of improvising and how they maintain and negotiate their musical identities through related discourses.

We report below on a novel study addressing these gaps. In particular we sought to answer:

Method

Participants for this study were recruited from The Glasgow Improvisers Orchestra (GIO), an ensemble rehearsing and performing together regularly, in which both authors played (www.glasgowimprovisersorchestra.com). This ensemble is dedicated to free improvisation, a growing movement eschewing genre parameters in the interests of maximizing innovation in music. Its line-up includes more than 20 players from a wide range of musical backgrounds including jazz, classical, pop and avant-garde, allowing us to look beyond jazz practice. Furthermore, these interviewees were able to reflect on a body of musical experience shared with each other.

Individual in-depth interviews (n = 10, two female and eight male) about improvising and playing in the ensemble were conducted by GW; his identity as a fellow ensemble member is likely to have been an influence on the data and we reflect on this in the analysis. Since it was intended to conduct in-depth analysis of rich data, a sample of 10 musicians diverse in gender and musical backgrounds was sought. Eleven ensemble members, including all female ensemble members, were approached by the authors on this basis and invited to take part, with only one declining. Participants were assured of confidentiality and it was made clear that participation was voluntary at all times; interviews were tape recorded with consent and lasted between 45 and 120 minutes. They took place at participants’ convenience in either their own homes, a university room, or in cafes. Interviewees were asked to describe their own experience of instances of group improvising within the ensemble.

The authors both coded transcripts independently for themes and categories of conversational action following the analytic practices of discursive psychology (Potter, 2004). Thus each researcher noted under their own headings recurring features that they observed in the content of interview talk, as well as any striking patterns in how accounts were presented, endorsed or challenged, and particularly considered structural features of discourse analogous to those identified in their previous studies. We discussed and refined these headings and patterns to arrive at definitions that both could recognize, then re-examined the transcripts to identify divergent instances and test whether the explanations developed could account for all these and arrive at a consistent interpretation of the data. Finally, reflexive features of the interview process and context were considered in relation to this interpretation, in particular the relationship between interviewer and interviewees as fellow ensemble members.

Results

Discussion

Our findings indicate that, when musicians talk about improvising, they construct a version of their musical practice shaped by musical identity work. These constructions have significant implications for researchers studying creativity in general and improvisation in particular (Clarke, 2012). For example, the importance of instrumental virtuosity may be contested when exploring issues around aesthetics and quality. Viewing musical events as ambiguous and open to multiple interpretations by both performers and listeners challenges the importance of a canon of skills as a prerequisite for successful improvisation. The different ways in which a disparate group of free improvisers construct their musical practice indicates the importance of recognizing some of jazz musicians’ talk about improvising as an expression of their professional milieu rather than a universal or authoritative account of improvising in its broadest sense.

Musicians in this ensemble were able to construct versions of themselves as improvisers whose practice was the more valid because it lacked mastery or informed control, something we have never observed jazz musicians to utter without such a claim being undercut with irony. Yet ensemble members could also endorse the importance of instrumental mastery and theoretical knowledge. What seems significant is that privileging one or other repertoire in their talk required a consequent acknowledgment that alternative versions of improvising were tenable, to avoid potentially excluding colleagues from identity as improvisers. Thus the social nature of group improvisation, when it takes place among musicians of varied stripes, may require all concerned to maintain a flexible approach to what should constitute improvisation. This counters the version of improvisation expressed, for instance, in ethnomusicological studies of jazz musicians (Berliner, 2002; Monson, 1996), where assimilation of a core set of skills, knowledge and procedures is prescribed for improvisation to take place. This may be the basis on which a great many successful jazz musicians see themselves operating; but the findings here underline the importance for psychologists of not treating this as necessarily the way that all people or groups will improvise. Gibson, in studying jazz musicians’ accounts of improvisation, finds a similar emphasis placed on the acquisition of necessary embodied skills, but locates their centrality within the practices of a specific art world (in Becker’s sense) around jazz (Becker, 1984; Gibson, 2006). Our findings suggest a great many more modus operandi in improvisation that could be studied and mapped; or indeed, that models of improvisation must be anticipated as being subject to change or obsolescence just as ways of making music have been transformed across centuries and cultures.

This implies that further research is needed to map out which aspects of our knowledge of improvisational processes represent particularities of professional or institutional milieus, and how we might revise our understandings on an ongoing basis by exploring alternative arenas of musical practice. For instance, Davidson and Good (2002) observed the influence of interpersonal dynamics on how classical musicians perform together, and a more recent comparison of the accounts of jazz musicians and classical string players has drawn attention to how nuances in their discourse reflect structural features of their working practice and professional context (Dobson, 2010). Piano students in a musical conservatoire have been observed to valorize a repertoire of mastery in talking about their studies, and it seems reasonable to consider that possession of exclusive technique should be a key feature of professional identity work among those dedicating years of their lives to acquiring this status through music (Juuti & Littleton, 2010). Greater understanding of how professional identity work is distinguishable from music identity work could be achieved with wider study of improvisation.

The study we report involves musicians known to each other and the researchers, and with musical experiences in common, and has yielded rich insight into the complexity of interaction and sense-making when improvising. This is in part achieved through recognizing the specific context of musicians being interviewed by a fellow ensemble member and using this to inform interpretation of the data. If the social processes of improvised music are potentially much more complex (and therefore much more interesting for psychological research) than previous genre-specific studies have suggested, there is therefore considerable scope for enriching and extending these findings by undertaking similar investigation into improvisation wherever it is found around the world.

The ways in which participants talked about improvising together were consistent with the understanding of communication in discursive psychology (Potter & Edwards, 1992). Individuals were seen to ascribe meanings or intentions to how they and others took part in improvising, but those meanings were negotiable and constructed in response to conversational context. However, the findings also demonstrate how the transmission models of communication that are frequently used to explain the non-verbal interaction of group improvising can be fruitfully superseded by a model of conversation as discourse. It is a tenet of the turn to language that we cannot access the internal individual worlds of musical behaviour, only verbal versions of it. These are inevitably subject to social construction and may be inadequate to account for such an embodied and transient experience. However, the findings demonstrate that these musicians recognized a particular non-verbal event within a shared musical improvisation as something irresolvably ambiguous. While there is a significant amount of evidence to indicate the effects of specific structural parameters upon musical communication, this finding underlines that music has an essentially ambiguous quality, and that elements such as preference and structure interact in a number of sophisticated ways to produce meaning for the listener (Knox, Beveridge, Mitchell, & MacDonald, 2011) – including someone listening to their fellow performers. Moreover, when recounting that particular event, the improvisers treated its multiple potential meanings as consequential in both preferred and dis-preferred ways for the identities of themselves and fellow players, and oriented to them accordingly.

This is important as, no matter what specific emotion a performer wants to convey in any given performance on behalf of themselves or a composer, as listeners we filter everything we hear though our own listening histories, experiences and preferences. What is heard is therefore interpreted in innumerable idiosyncratic and context-specific ways (Cross, 2005; MacDonald, Kreutz, & Mitchell, 2012b). Our findings suggest strongly that this is no less the case among musicians improvising together. In the course of performing this music, an individual improviser may make a contribution to the music with a particular intent; but both musical content and intent will be shaped by the identities which that individual can construct for themselves and others within the social and temporal context of performance. Yet that musical contribution may be interpreted, and responded to, in idiosyncratic and unpredictable ways by the other improvisers, depending in turn on the meanings they assign to it in the context of what has gone before, and the identities they recognize for themselves and others in the group.

Since any musical contributions and responses require subjective social constructions of the process taking place and the performers involved, the improvised music necessarily constitutes an emergent, non-verbal discourse. This is a version of group improvisation that much better represents the excitement, complexity and diversity of improvised music than the idea of people who have learnt simply an objectively right way to pass about commonly understood musical components. It also stands in some contradiction to frequent assumptions that improvisers reach shared musical meanings in playing together (e.g., Borgo, 2004). Previous studies of jazz ensembles have examined creativity and communication in the rendition of works that exist to some extent in advance, even if only as a sequence of chord changes and an assumed style (e.g., Seddon & Biasutti, 2009). The musicians in our study, however, were concerned with creating, rather than rendering, a piece of music entirely in performance, with no prearranged parameters. Furthermore in a large ensemble, visual cues are less readily transmitted, particularly where no leadership role is assigned. In this context, the meaning of musical contributions must truly be constructed in a social context.

While Sawyer has argued (Sawyer, 2005, p. 49) that the unpredictability of improvised jazz means that ‘a musician cannot know the meaning of her own playing until the other musicians have responded’, our findings suggest that, in free improvisation, she may not even know then, but can only construct meanings for that music in contingent ways. This suggests an important qualification of Seddon’s (2005) empathetic attunement: group improvisation depends on the social construction of other players’ musical perspectives, rather than necessarily being able to ‘see’ them as intrinsic characteristics. What we observed in the accounts of these musicians shows significant divergence in the interpretation of musical contributions and objectives between individuals, despite the ensemble having a strong awareness of, and enthusiasm for, a collective identity.

The focus on choosing not to play as a particular musical event and behaviour has allowed this study to reach novel conclusions about what takes place in group improvisation, and highlights an important direction for future research in this area. Research pursuing cognitive modelling of improvisation focuses on physical processes at the expense of social processes (Berkowitz, 2010; Brattico & Tervaniemi, 2010; Johnson-Laird, 2002; Limb & Braun, 2008; Pressing & Sloboda, 1988). Yet the qualitative investigation of musicians’ talk about improvising tends not to require musicians to focus on specific musical events, or compare analysis of talk with analysis of their actual playing. This is perhaps surprising. However, musicians may often find difficulty in, or resist, verbalizing an exact account of what takes place in the extraordinarily rapid, embodied and non-verbal processes of improvising (MacDonald & Wilson, 2006). Bastien and Hostager (1988) provided an early exception to this rule, having asked musicians to talk them through video recordings of their playing, but treated musicians’ accounts as reportage rather than discourse. Discursive psychologists have repeatedly called for qualitative research to broaden its scope beyond interview transcripts (Potter & Hepburn, 2005). Although we have gained insight on improvisation through focusing on a specific instance that emerged as recurrent in our data, it might be more beneficial for future research to look beyond generic interviews with musicians and seek their descriptions of live or recorded musical acts to a greater extent; or, to undertake discourse analysis of recorded music itself.

There is a research imperative to enhance our understanding of improvisation as a unique form of collaborative creativity crucial to the landscape of contemporary music, and as spontaneous musical communication that is universally accessible regardless of technical proficiency. We conclude that the creative, communicative and social process of musical improvising in groups can best be understood when the entirety of improvisational practice is taken into consideration and its various contexts (musical, physical, cultural and interpersonal) acknowledged. While improvisation is a practice that cuts across all musical boundaries, the study of its wider musical applications is in its infancy. There is considerable need for further psychological research to enrich our capacity to understand improvisation by looking beyond particular genres, and beyond particular professional milieus or the musical activity of ‘specialists’. However, there is a particular danger in trying to formulate research questions about artistic or creative behaviour that we proceed from a working definition of what the processes under investigation should be, whereas the very nature of artistic endeavour is to innovate and develop new and unique approaches, to redefine what constitutes creative practice. Qualitative research methods are the most appropriate approach to understanding what people do when they engage in the arts for this reason, since they proceed from individual understanding of the object of research interest, rather than testing an established definition. The continued investigation of group improvising through discursive psychology in particular can enable research into the arts to keep pace with the protean rate of change in music as in other art forms, rather than seeking to privilege particular versions of improvising as they become institutionalized.