Methods 

The autoethnographic methods presented here are audio journaling and reflective (written) journaling. We then turn to ethnographic methods: participant-observation, informal conversations and interviews. These are followed by some reflections on verbal accounts and music-making.

Audio journaling

Journaling, in a general sense, is a common tool in (auto)ethnographic research for "recording personal thoughts, daily experiences, and evolving insights” (Hiemstra, 2001, p. 20). Here, audio journaling was used as the main form of journaling, which is a form of verbal protocol where reflections related to a particular phenomenon are captured by 'thinking aloud' in combination with audio recordings. One of the reasons for working with audio journaling is that it enables you to capture lived experiences in a more direct way than its written counterpart. While written journals offer rich reflection, they may not always capture spontaneous expression and reactions in the moment, seen that writing may involve the person to deliberate his or her thoughts carefully, and choose to refine them several times before completion (Sawhney, Graver & Breitkopf, 2018). A potential metaphor could be to see audio journals as captured improvisations in comparison to a written text, which behaves more like – and is sometimes referred to – a composition.

The reflections that I recorded were often related to ongoing work, such as preparing scores by the piano , with the advantage that I could capture my thoughts while still sitting by the instrument, thus maintaining the momentum of the work process. It could also be in conjuction with musical sessions; for instance, as a way of recording my initial reflections when walking away from a rehearsal space. Reflections that were recorded outside of the musical space were mainly recorded while walking, as this generally seemed to be conducive to seeing the project from new perspectives. As demonstrated by a landmark study conducted at Stanford university, walking can significantly boost creative ideation in real time and shortly after, including divergent thinking (Oppezzo & Schwartz, 2014).

It should be mentioned that these reflections were recorded in English instead of my mother tongue (Swedish), which allowed me to use an application for automatic transcription. This posed some initial challenges, as reflecting in a second language can impede the natural, spontaneous expression that comes with speaking one's native language. However, my assessment was that the added value of capturing these reflections in such an accessible way compensated for this risk. Also, it removed the need to translate the reflections later, something that would involve an extra element of interpretation.

Reflective (written) journaling

Reflective journals were used as a written complement to their verbal counterpart, especially when the journaling wasn't directly connected to work with preparing material. However, as the project unfolded I gradually abandoned the written journals in favor of audio recordings.

Participant-observation 

The musical sessions with participating musicians were studing using participant-observation, a method where researchers study the practices of communitites through participating in mutual activitites. Williamon et al. (2021, p. 88-89) points out that participant-observation can be particularly useful in practice-based music research, when the emphasis is on process; e.g. activities related to composing, rehearsing, learning, and performing. An advantage with this method is that it gives access to naturally occuring behaviours associated with musical practices and introspections underlying the music making (p.89).  As demonstrated by Gibson (2006), performing and improvising together with research participants – in combination with subsequent interviews – can be an effective way of achieving contextually detailed conversations that would be difficult to attain through interviews only. It can also serve as a way of shedding light on the ineffable experiences – experiences that cannot be conveyed using words only – that are intrinsic to music performance (Yip, 2020). 

Observations related to the musical situations focused on departure points and ways of engaging with and adapting the musical material. Sessions were documented through audio recordings in combination with field notes and/or audio journaling done by me soon after the sessions. Following the focus of the project, the field notes that were written down soon afterwards included information about the musical processes; what repertoire we tried, what approaches we took to musical media and the musical language of the compositions, our reactions to these approaches, and what the main insights and revelatory moments were. Similar information could also be recorded using audio journals soon after a session – e.g. when walking away from a rehearsal space – when the experiences were still fresh in memory. This was common when there were moments that stood out; important insights, revelations, and musical turning points. Such salient moments were often revisited later with the aid of the recordings, which could include transcribing dialogue related to important events.

My role in these sessions was as a full participant, effectively giving me the dual role as a participant and a researcher. The latter was generally more emphasized afterwards, as I revisited the documented material or reflected on the sessions in retrospect. This was made clear to the participating musicians; although I took part in sessions as a performer – often leading the creative work – the sessions were a part of a research project, led by me, where the approaches and musical processes were studied. As Gunnarsson (2019) points out:

Even when studying one's own professional practice, as is sometimes the case in research on practical knowledge, one participates on different terms from those one is studying. Simply by assuming the role of researcher, one distances oneself from the professional role one normally assumes. (p. 236)

Thus, not only did my actions affect the processes from the perspective of music performance, the fact that these processes were also under scrutiny, and that there was a research agenda present, might also have a certain impact on musical choices and ways of reflecting over the processes.

Informal conversations

Besides typically being a part of participant-observations, the use of informal conversations can function as a qualitative research method in its own right, with the advantage that one can achieve greater ease of communication and produce more naturalistic data (Swain & King, 2022). Three types of informal conversations were used: Conversations that took place during the music sessions, conversations in conjuction with music sessions – i.e. soon before or after – and conversations with respondents. 

Semi-structured interviews

Interviews could take on three main forms; as listening sessions, public discussions, and interviews with key respondents. These were were all semi-structured, i.e. based on a combination of prepared and spontaneous questions. All of the sessions were documented with audio recordings.

Listening sessions

Given the centrality of listening, many interviews were organized as listening sessions, typically following a performance or a recording session. These were organized as one-to-one sessions, or in small groups (i.e. focus group interview). The advantage of working with listening sessions as a setting for mutual reflection is that it involves a natural setting from a musician’s perspective, one that is frequently a part of performing, recording, and reflecting on the results of those activities afterwards, i.e., not necessarily organized for the purpose of a research study. However, the mutual listening and reflection can also be used to enhance a interview situation; as demonstrated by Zadig, Lyberg-Åhlander & Folkestad (2016), recordings of music-making situations can be effective material to use as "triggers” in an interview situation – then referred to as stimulated recall – where participants can experience recorded events from an outside perspective, while still having insights from the inside regarding motivations and intentions (Sherin & Han, 2004). The stimulated recall method has its roots in the psychology of memory training and language learning, where the aim is to stimulate the recall of an interview or discussion by reproducing a situation – e.g. by audio or video recording (Bilodeau, Fox & Blick, 1963). Bloom describes the basic idea as follows: "that a subject may be enabled to relive an original situation with vividness and accuracy if he is presented with a large number of the cues or stimuli which occurred during the original situation." (Bloom, 1953, p. 161).

The one-to-one sessions were done online, due to the geographical distances, and – to some extent – due to the ongoing pandemic at the beginning of the project.  The group-based sessions were organized as physical meetings that took place in a room equipped with a sound system. For these sessions, I prepared questions related to evaluating our performances and different approaches to the music that we had worked with, i.e. what did you think of that approach with…? However, these were merely there as a guide; the ambition was to attain a free-flowing conversation based on the music and aspects we picked up while listening. As we listened, the participants were also encouraged to ’think aloud’ when reacting to something they heard.

Panel discussions

Panel discussions were held on five occasions, following concerts in the last year of the project. These were conceived as semi-structured focus group discussions, i.e. a type of interview where a group of people discusses a given topic or issue in-depth. These were all moderated; the first two by external moderators, the remaining three by me. Furthermore, two of these sessions had guest musicians (see musical outcome/concerts). All of the discussions were based on interview guides that I had prepared, e.g. questions related to our approaches to working with the pieces, how we perceived approaches and the processes when working with the music. There was also room for questions from the audience.

The challenge with being a moderator and participant simultaneously, as was the case in some of the sessions, resembles that of being an participant-observer in that you can easily be restricted by your own perspective when performing; while an external moderator might have a certain critical distance, the performer perspective can make you ’blind’ to some aspects of the music-making process. On the other hand, it offers an opportunity to go in-depth using insights from the inside; this way, direct experiences from moments of making music together, and previous knowledge of how the material was approached can lead to more detailed follow-up questions and exhaustive reflections on a particular aspect.

There are also challenges with doing these discussions in an open setting that should be addressed; for one thing, the participants might feel more inhibited in the presence of an audience, in comparison to a closed setting. Also, the responses in these situations might be articulated differently than if the discussions were held solely with a group of performers, based on a desire to communicate with the people in the room. An advantage with doing these discussions in connection to concerts, however, is the involvement of the audience; besides the added value of research communication, the presence of an audience during the discussions gives a potential for follow-up questions, and the possibility of addressing aspects that were not considered by the moderator or the participants.

Interviews with key respondents

Interviews were also conducted with key respondents in the field – more specifically improvisers with prior experience of combining Western classical music with improvisation – with the purpose of gaining more knowledge on how other musicians approached the studied phenomenon. These interview situations can be considered as collegial interviews; in my role as an "insider" in the field, I had prior knowledge that helped me both formulate relevant questions and follow up on these when deepening was desired. However, as Shutzberg (2019) points out, insiderness can never be taken for granted. For instance, my role as a researcher could entail a distance to those I interviewed, and thus compromise the collegial closeness. One way to counteract this was for me to approach the interviews as conversations, where I as interviewer could involve myself and include my own experiences and affects, rather than slavishly following the interview guide. This dialogical approach, sometimes referred to as reflexive dyadic interviewing, can be seen as a consequence of the collegial relationship (Chew-Graham et al., 2002), and is according to Shutzberg (2019) something that occurs more often the more involved the interviewer is in the subject. As Gubrium & Holstein (1997) point out, it is not primarily a conscious strategy to get the respondent to open up, but rather a desire to have a conversation where both parties – in a non-hierarchical way – can share thoughts and experiences, where the interviewer's task is largely about being responsive to the joint creation of meaning and the emotional dynamics that arise. Open-ended questions were prepared by me, such as Can you tell me about your previous experiences with improvising over the repertoire by X? and How did you prepare the material?, with occasional follow-up questions.

Reflections on verbal accounts and music-making

When considering interviews as a way of gaining insights into the practices of improvising musicians, we should be cautious about (over)interpreting the verbal statements in which musicians – and other practitioners – describe their considerations and reasoning after the fact, e.g. the thoughts that takes place during the musical situation. In a discussion of his concept of reflection-in-action, Donald Schön describes jazz musicians' reflections as follows: ”[W]e need not suppose that they reflect-in-action in the medium of words. More likely, they reflect through a ’feel for the music’ which is not unlike the pitcher's ’feel for the ball.’ (Schön, 1983, p. 56).

There is good reason to believe that what happens in the moment of musical creation does not involve words or reflection in a verbal sense. At the same time, Vaisey (2009) describes how interviewing by its very nature puts us in contact with a discursive consciousness, a type of consciousness that does a good job of providing us with rational explanations and reasoning that are not necessarily linked to the real motives behind our actions. There is thus a risk that, when we have to articulate a course of events afterwards, we describe the actions as rationally guided rather than something that arose as a result of unconscious processes or intuition. In the domain of psychology, this phenomenon is known as post hoc rationalization, as a way of rationally justifying our actions after the event (post hoc).

Also, in addition to looking at the time aspect, it may be relevant to look at the context in which knowledge is situated, the practice in which meaning-making takes place. As Alvesson & Sköldberg (2017) point out, qualitative interviews do not only give us access to the participants' thoughts and motivations linked to a particular practice, they produce situated accounts, situated to the interview and adapted to the particular logic of the interview practice. Joosse & Marshall (2020) refer to this phenomenon as ex situ -rationalization, meaning that we rationalize based on the norms and meaning-making associated with the new context, based on the current practice.  This is an important point, and a reason for why many of the conversations – formal or informal – were done in connection to musical situations; settings that are more in situ.