Doing Ethnography. Amanda Coffey
to conduct excellent ethnographic research without immersion in a field of study – for example, by applying ethnographic principles to research activities that might include a series of conversations and encounters with social actors. Here the field might be the site from which the social actors come, and through them be the site of exploration – for example, a series of ethnographic conversations with midwives in order to understand midwifery practice, or listening to and conversing with teachers in the staffroom of a school in order to make sense of teachers’ lives and selves in and out of school.
When a field of study is identified, pre-fieldwork practices or preliminary fieldwork can help to develop research questions and possible lines of inquiry. So, for example, undertaking some initial scoping research on the setting can often be possible and helpful. Collecting documents (see Rapley, 2018) in and of a setting can provide insights into the documentary realities of that setting, providing one way or indeed often multiple ways of ‘seeing’ and ‘reading’ a setting. Informal conversations and discussions, including exploring opportunities for access, might be possible, as might initial visits, some preliminary observation or the collection of visual images or sounds of the setting. Such activities actually mark the start of the research but can be usefully conceptualized as an initial or preliminary stage, enabling a more nuanced and focused research design to be developed, with the potential to scope out a range of lines of inquiry.
The field of study and research questions can also be developed out of our own interests and experiences, and completely legitimately so. A setting might well present itself in and through a personal experience. For example, there is a rich ethnographic tradition of teachers drawing on their experiences in undertaking school ethnographies (see Gordon et al., 2001, for an overview of ethnographic research in educational settings); equally, a number of sociologists have used their experiences of illness and health care settings to develop ethnographic analysis and insight on medical encounters and the body (see Beynon, 1987; Delamont, 1987; Horlick-Jones, 2011). The important thing here is that these personal experiences provide stimuli to engage ethnographically and reflexively with a field of study. They provide a way into and some important anchors for asking questions about the nature of experience and process. Indeed all ethnographic research builds on the resource of personal experience; by its very nature doing ethnography demands that the researcher is ‘there’ and, through personal engagement with the field, progressively focusing inquiry in order to develop rich descriptions and make sense of the field of study. So our personal interests and insights, to an extent at least, always guide our fieldwork practice, whether that be the stimuli for an initial interest in a topic or field of study, or through our personal reflections and experiences that enable fieldwork to become fine-tuned and detailed. The role of the ethnographic researcher in shaping and asking the questions, and in providing critical reflection is key. This does not assume a personalized, narcissistic or inward-facing research agenda. While the personal is intrinsic to the ethnographic project, it is important to acknowledge and question our own subjectivities and positionalities, drawing on but not being limited by our personal interests and questions.
Methods of data generation
There are a range of methods for generating data in ethnographic research. Hence, ethnographic research design includes making choices about data types and ways of generating those data. To ‘generate’ is a term used here with purpose and on purpose. The term ‘data collection’ implies that there are data to be collected; data which are already ‘out there’, ripe and ready to be harvested. Data collection can sound like rather a passive process, with a neutral researcher moving through a field as a data collector. This does a disservice to the processes and practices of ethnography, to the role of the ethnographer researcher and indeed to the complexities of social life. Data which might tell us something about the social world, and enable us to develop understandings about the life worlds of social actors demand a more active engagement on the part of the researcher. Data are not ‘there’, already in existence, simply to be collected up, organized and stored. Rather, ethnographic data are generated through various kinds of interaction with a social setting and/or social actors, crafted through our research practices. Indeed ethnographic data are all, in some way or another, co-productions between researchers, people in the field and the field itself. Data are made not caught.
The starting point for data generation for much ethnographic research is participant observation. This is where the researcher becomes a participant in the setting to a greater or lesser extent, in order to observe and record what is happening. Data in these circumstances, usually taking the form of fieldnotes, are made in situ during observation where possible, and expanded and developed after fieldwork. These data then are jottings and detailed notes, serving as a re-presentation of what has happened in the field. Participant observation is often held up as the gold standard for ethnographic data collection (see Atkinson and Coffey, 2002) and most clearly embodies key ethnographic research principles and the ethnographic spirit – being there, experiencing, watching, listening and feeling in order to understand and make sense. In ethnography there are different degrees to which participation in a setting is possible or desirable on the part of the researcher (see Gold, 1958, for a classical typology of researcher roles from participant through to observer; see also Chapter 5, this volume).
Participant observation can generate rich, layered data, but it can also demand a lengthy time commitment – sometimes stretching over several weeks, months, years or even decades (see Fowler and Hardesty, 1994; Okely and Callaway, 1992). This may influence the research design decisions that might or can be made. Time is an important factor to bear in mind. So too is the extent to which access to undertake participant observation can be successfully negotiated, as well as the ease or difficulty with which it might be possible to conduct observational fieldwork. In some settings, for example, being a participant observer in the setting may be dangerous, risky or impossible. It is worth noting here, however, that there is a long tradition of ethnographic participant observational fieldwork in settings that might have been initially considered difficult to access or dangerous or risky (see Nilan, 2002; Tewksbury, 2009).
Alongside, in addition to, or instead of participant observation, there are a range of other methods that usefully form part of the ethnographer’s data generation tool kit. Ethnographic interviewing – conversations with a purpose (Burgess, 1984; see Brinkmann and Kvale, 2018) – is a widely used method, both taking advantage of and being situated within the narrative turn within the social sciences (Czarniawska, 2004). Ethnographic interviewing builds on the anthropological approach to fieldwork – asking questions of social actors in the context of their everyday experiences, as part of participant observation. Interviewing, though, can also be utilized as an alternative to participant observation. Research can be ethnographic and be conducted through a series of extended interviews as conversations, and planned as such, recorded manually or digitally. Film also has a long history within ethnographic research, from the documentary style anthropological films of the early to mid-twentieth century through to a wider ranging visual revolution of the last few decades where still, moving and digital media have served to enhance and extend the repertoire of the ethnographic researcher for generating data for analysis. Ethnographic research design can include planning to generate researcher-produced still and moving images, supporting participants to create their own images, as well as gathering images of settings already created by participants or others (such visual data can include photographs and film, but also maps, pictures, digital media and other art forms). These approaches provide opportunities for differently generating data about social worlds. It is worthy of note here that digital technologies have also made it easier to capture the sounds of settings. Soundscapes can add a further sensory exploration of social life; noise and sound are very pertinent ways in which we experience and undertake our daily lives (Hall et al., 2008). The digital age has also brought with it new opportunities for data generation in ethnographic research. The digital landscape that encompasses websites, email communication, mobile phone technologies, geographical and mapping applications and social media means that there are a variety of ways in which social settings and social life are now digitally experienced, and through which researchers can engage with/in research settings. Ethnographic data can be gathered through participation in and engagement with these mediated technologies. This includes,