Over the last 12 months I have been travelling around the UK interviewing classical guitar-makers. It’s been a really interesting experience and although I have not yet begun to analyse the data, I thought it was worth writing up a few notes on the process up to this point, partially in support of a seminar I am running at our doctoral study school this month about the ‘art of listening’ (cf. Les Back, 2007) and hope that this account offers students some insight into the work of interviewing.
So far, I have interviewed 30 luthiers 1 My interviews with luthiers are unstructured, with most conversations lasting around 2 hours. On a few occasions, I was invited to stay for lunch and kept the recorder running while we talked. I also kept the recorder running when people were showing me around their workshop. I think all of this constitutes the ‘interview’, which was
“based on a clear plan that you constantly keep in mind, but are characterized by a minimum of control over the people’s responses. The idea is to get people to open up and let them express themselves in their own terms, and at their own pace.”Bernard (2006, 211)
Unstructured interviews are not unplanned nor informal interviews. Mine were planned quite carefully, from the initial selection of interviewees, the timing and logistics of the interview and the preparatory list of things I wanted to cover. There was a formality that defined the start and end of each interview, and between those two points, a desire to create a relaxed, natural conversation that remained on topic.
Selection of interviewees
I selected to interview most participants based on their response to a survey that I conducted between August and October 2018. Next, I will say a bit about the survey as background to the interviews:
Over eight months in 2018, I compiled a list of classical guitar-makers in the UK, living and deceased. I began by contacting members of a popular Internet forum for classical guitar, stating that I was interested in whether it was possible to construct a genealogy of UK classical guitar-makers, identifying the relationships between teachers and students and therefore the passing on of craft knowledge. Members of the forum offered names and references to consult.2 I created a spreadsheet of publicly available information so as to organise what I was finding, including any details about the luthier’s education. Many luthiers have biographical details on their website so as to offer potential customers an insight into the length of their experience, their values and approach to lutherie, how and from whom they learned their craft. Within a week of searching, I had a spreadsheet of 81 luthiers, both deceased and alive; within two months, 100 names. I made the list publicly readable, so that forum members might be prompted to offer new entries.
At the time of issuing the survey in August 2018, there were 130 names on my spreadsheet and I was aware of 102 luthiers who constituted the total living population. Since then, I have learned of a further three working classical guitar-makers and there are certainly more amateur luthiers who have attended night classes or short courses and made one or two instruments.3 Nevertheless, it has surprised some people I have subsequently interviewed that there are over 100, the vast majority of whom are still active.
The survey was piloted with 3 luthiers, whose responses were included in the final results. After piloting the survey in July, I first posted information about it on the Internet forum, inviting UK luthiers to complete it. I wanted to see what the response was from this method before sending it directly to the list of people I had compiled. Only four people completed it over the course of a month (all within a few days of me posting to the forum). At the beginning of September, I then sent an email with information about the research and a link to the online survey to all 101 luthiers I had contact details for and sent a reminder two weeks later. I closed the survey at the end of October 2018.
In total, 61 individuals responded to the survey and 49 people offered follow-up interviews. So far, I have interviewed 21 survey respondents plus a further 9 people who either subsequently contacted me or who I contacted directly. In selecting people for interview I wanted to achieve good coverage of experience, including both younger and older luthiers; those who said it was their main occupation and those for whom it is not; individuals who were self-taught and those who went to college; people who had taught on short courses, in college and privately one-to-one, and those with and without a formal qualification in musical instrument making. Furthermore, I also sought to interview people who had achieved significant reputations for their work, and occasionally I interviewed people because it was convenient to do so e.g. if they lived near someone else I planned to interview. I have two more interviews planned in September and a further three I would like to do over the following year, if possible.
I interviewed 30 people across the UK, from Edinburgh to St. Ives, over the course of 12 months. I was awarded £2541 funding from my university to cover the costs of this (it also paid for trips to the British Library and London Metropolitan University archives for document research) and the entire budget has now been spent. Another implicit cost is the time it took me to do the interviews. Like other academics on the national contract, I am allocated 222hrs/year for scholarship and research purposes, which amounts to about a day a week, and some of that time was used for conducting interviews. I also used annual leave and weekends because it was more convenient to me and/or the interviewee to do so. My point here is that the time spent on research and on annual leave and on other work became more fluid than it might otherwise have been in order to visit people in a time-effective manner, grouping interviews whenever I could in different parts of the country. More commonly, I would take a day to visit two people who lived in the same region of the country, interviewing one in the morning and the other in the afternoon. I took three trips lasting several days to Northumberland and Scotland, East and West Sussex, and Devon and Cornwall, usually incorporating bank holidays and weekends.4
I conducted one interview via Skype because the interviewee’s plans changed at the last minute and he said it would be more convenient to do it this way. The interview went well, but it was the shortest interview I conducted (49 mins) because online communication doesn’t allow for a relationship to be developed in the same way as being there in person. You can’t drink coffee together, talk about the lutherie books they have on their shelves, visit their workshops, look at their instruments, and so on. In one sense, online interviews could be seen as efficient in terms of time and money, but in my case the quality of the data was less rich.
A lot of my travelling was by car because it was easier for me to predict how long it would take for me to get to a person’s house or workshop and because people don’t always live within easy reach of a train station. Trains are a more expensive way to travel too. Typically, I would hire a car at £25/day + fuel and using my phone as a SatNav, I could very accurately predict when I would arrive at each destination. I hired cars because it is university policy to do so when travelling more than 100 miles round trip, which was always the case. Also, I do not use my own car for work, so am not insured to drive it for ‘business purposes’. It may seem rather banal to talk about car hire here, but it had an unanticipated effect on me as a researcher for two reasons: First, it complicated the idea of being ‘at work’. When I was travelling for interviews at weekends and on leave I was doing so, insured, for a ‘business purpose’ (i.e. research) but outside of conventional work time. Second, I found myself wanting to explain to interviewees about the car hire because, as they typically walked me to the vehicle as I was leaving, I felt ‘inauthentic’ about getting into a new and pristine car. I’ve never had a new car (our actual car, bought last year, is 10 years old) and I felt that the hired car might suggest something about me which I do not identify with.
Finally, I want to say that although it has been tiring at times driving around the country, it’s also been a pleasure to meet and talk with everyone and to see so much of the UK, driving to places I’ve never been and experience the beauty of the countryside where luthiers often choose to live.
Conducting the interviews
After the initial welcome and offer of tea/coffee, I would begin each interview by ensuring that the person had completed the consent form and understood the nature of the research and by doing this, we might talk a bit about my background and how I came to do the research. I checked to see whether they had opted for anonymity and everyone, without exception, said they wished to be referred to by their actual name if quoted in subsequent publications. At times during interviews, they might indicate that something they said was off the record (or ‘I maybe shouldn’t say this…’) and I will use my judgement to ensure that no-one is quoted in a way that is likely to cause any negative repercussions. If in doubt, I will check with them again.
I had an interview schedule, which I glanced at occasionally to ensure that I touched on each of the same questions and themes in each interview. I told the interviewee I had prepared these questions but that I preferred not to work through them systematically in the form of a Q&A. Instead, I asked the person to “tell your story” and I said that I would prompt them when necessary to ensure that my questions were answered in the course of the conversation. Interviewees were happy with this. As we sat around their kitchen table or in their workshop, it made sense to take this approach and, I hope, allowed them to relax. Several people thanked me for the opportunity the interview gave them to reflect on their lives and work.
As I said above, the interviews would usually take around 2 hours (a couple of interviews with retired luthiers, lasted up to 5 hours). I turned on the audio recorder as soon as I could because interviewees would start saying things of value to the research almost as soon as I’d entered their home or workshop. I also learned to keep the recorder running until I was almost out of the door, for the same reason. Early on, when I thought an interview was over and had turned the recorder off, we then launched into further discussion which I had to quickly recall and write down while sat in the car outside after I’d finally left. Although at first I was worried that interviewees might feel uncomfortable with a recorder running, one person said they were pleased I was using it because a previous interview they had given to someone else was not accurately reported and he blamed it on them not having recorded the interview.
For the first couple of interviews, I used a dedicated digital recorder but soon switched to using my phone, which worked flawlessly. It had plenty of storage space; the microphone was very sensitive and automatically filtered out background noise; the battery lasted for hours (and could be charged while in the car between interviews), and it was unintrusive to have a phone on the table or in my hand as we walked around the workshop. I used the VoiceRecord7 app for iPhone which also allowed me to easily and wirelessly export the audio from my encrypted phone, to my encrypted laptop and then backed up to secure cloud storage provided by my university, which I did shortly after each interview.
I also took photographs of each interviewee and took along a good quality camera in addition to the camera on my phone. On a couple of occasions I forgot to take photos so I subsequently got into the habit of taking my camera out of my bag and putting it on the table to indicate to the interviewee that I’d like to take photos but also to remind me to use it. I used the dedicated camera to take portraits, usually in the workshop, to act as a visual reminder of the meeting and possibly for publication. Occasionally I would ask people if I could take a photo for Instagram, where there are a lot of luthiers sharing photos of their work. This was a way for me to share what I was doing with a small and interested group of ‘followers’ – mainly luthiers; to indicate the ‘journey’ of the research, and to humanise it with images of participants. Guitar-makers in the UK, as I have seen and been told, are quite isolated with many people working alone at home, and social media appears to be a way that is drawing some people together to share their work. The consent form made it clear to me how people wanted to be identified and, as I said above, every single interviewee chose to be identified by their real name and to allow me to take photos. I made it clear to interviewees that the purpose of my research was not to evaluate their work or promote the businesses of specific individuals.
Each interview felt like a unique and valuable experience. I was worried that the recording app might malfunction (it never did) and would occasionally check it was recording throughout the interview, drawing attention to the fact that we were being recorded. I took notes during the interview, too, mainly to aid my navigation of the recordings when I come to analyse them. Consequently, I’m able to look at my notes and be reminded about some of the things we discussed and the order in which they were discussed.
Although the interviews were what researchers call ‘unstructured’, rather than ‘semi-structured’ or ‘structured’, they were purposeful and productive. The interviews were naturalistic but didn’t happen naturally. Interviewees understood why I had visited them and that this is my work – they didn’t want to waste my time or theirs. By asking them to “tell your story”, my aim was to “get people onto a topic of interest and get out of their way.” (Bernard, 2006, 216) In effect, I had defined the focus of the interview and the interviewee was determining the content. Once settled and relaxed, I tried to let the interviewee take the lead. (Bernard, 2006, 217) This didn’t mean that I simply sat back and listened but my role in the interview was to listen carefully, probe and steer the conversation just enough to stimulate the information I was looking for without making the interview about me. Sometimes it was enough to let people pause and think about what they wanted to say next, at other times I would summarise where we had got to and they would pick up again from there. There are a variety of ways to keep a conversation going and in the right direction (see Bernard, 2006, 217-222) As we get older and interact with more people in different social situations, we gain experience in keeping conversations going, or changing the subject, showing enthusiasm and interest in what other people have to say, and digging deeper when someone has said something interesting. Such experience and ‘social skills’ are drawn on and developed over the course of multiple interviews.
One regret is that had I the time, I would have listened to and even transcribed each of the interviews before doing the next one, but it just wasn’t possible alongside everything else. That said, it is still possible to learn from the experience of each interview and the responsibility I have as a researcher to my interviewees quickly became apparent. People were not only giving me their time, but were openly talking about their lives and keenly supporting the research, rather than simply aiding it. I was often congratulated on the work I was doing and thanked for making lutherie the subject of my research. People were usually keen to know what would happen after the interview and it became clear to me that I was doing the research on behalf of others as well as trying to satisfy my own questions. 5 This sense of responsibility to the research and its participants has been hugely motivating and again blurs the boundaries of professional and personal identity because I feel I have a personal obligation to the people I met, whose homes and workshops I visited, and to carry through with what I said I was going to do.
There was in some cases a very good rapport with the luthiers I interviewed and this may have been aided by a number of ‘response effects’ – the “measurable differences in interview data that are predictable from characteristics of informants, interviewers, and environments.” (Bernard, 2006, 239) Such characteristics could be: That I am a 45 year old male of similar class background to many of the people I interviewed; that I can demonstrate both a personal and professional interest in and experience of classical guitar-making; that I understand the craft process, the history of the tradition and its language; that I could often tell the interviewer something they didn’t know about their own craft tradition; that the interviews took place in the familiar surroundings of their home or workshop; that I dressed appropriately (clean and casual!) and had prepared adequately, and that I was open and honest with people about my motivations and hopes for the research. Had I been, for example, a 21 year old female from a higher social class, with no workshop experience, perhaps a research assistant recruited for the interviews, rather than the main researcher, then I may have received different responses – no less true and honest – but more limited in their depth and breadth.
Having said that, the interviews are not direct access to the ‘truth’ – nothing is – and I acknowledge that people could have been telling me what I wanted to hear, not dishonestly but in an effort to make the experience seem worth my while. (see e.g. Hammersley (2003) for a round up of views on this). We are all susceptible to telling the truth how we want to see it or how we wish it were, to put ourselves in the best light, and tend to remember or prioritise our achievements rather than our failures. I am aware that when asking people about their college experience 40 years ago, or the feedback they got from a famous guitar player, there is the opportunity for error or exaggeration, and it is part of my work to make a judgement on that both during the interview process (in my response and my personal notes) and in the analytical process following the interview.
A complement to my research so far would be to observe people at work rather than ask them about their work. This is something I am doing at Newark College, having spent around 50 hours observing students and tutors in the guitar workshop over 12 visits so far, amounting to 10,000+ words of notes and 170 photographs. In an auto-ethnographic sense, I have also observed and take notes about my own guitar-making and in a limited way, I have seen my teacher, Roy Courtnall, at work, too. On many occasions, I have thought that it would enrich the study to spend time observing a single maker for a prolonged period but the logistics of this, even after someone was willing to put up with the intrusion, are complex, not least needing a month to observe the making of just one instrument.
You do the best you can under the circumstances.