Social Interaction

1. Overview

This special issue of Social Interaction: Video-based Studies of Human Sociality presents five compelling case studies reporting on the conduct of participant observation in video-based ethnography. In such research, video is used to record the moment by moment conduct of interaction by participants in their routine conduct of everyday events. That interaction is “co-operative”, in Goodwin’s terms (2018), in that it involves working together through participants’ continual monitoring of and accommodation to one another’s actions. The articles’ accounts make clear that the researcher, and the camera, are just as much participants in this social ecology of interaction as are any of the other participants.

In an essay introducing the special issue, titled “Researchers’ participation roles in video-based fieldwork”, Katila et al. (2021/this issue) raise the following questions about researcher participation during fieldwork:

1. What are the various roles that are available to researchers who are active participants in the video-recorded activity, and how can these roles be analyzed as unfolding in the moment-by-moment interaction?
2. What are the benefits and challenges of researchers participating in the interactions they study?
3. What ethical issues arise with the researchers’ different participation roles during data collection, especially when analyzing potentially vulnerable communities?

In the discussion that follows I will respond to those questions in two ways—first by some general comments on fieldworker participation in the social ecology of interaction, and second by reference to specific instances of participation as reflected upon and reported by the authors of the case study articles.

2. The Camera as Co-Operative Participant

In 1967 Paul Byers published an article titled “Cameras don’t take pictures.” In discussing still photography of human subjects he noted that people take the pictures, not the camera per se. Byers emphasized the quality of relationship between the photographer and the photographed, noting that acquaintance and affection between them influenced the overall orientation and visual framing of the photograph, the timing of the exposure, and the distance between the photographer and the photographed. The photograph was produced within social interaction and indexically portrayed its character. Byers was a skilled photographer—he was Margaret Mead’s cameraman and co-author in her study of interaction in a small academic conference (Mead and Byers, 1968). He also used cinema film to study timing and synchrony in social interaction.

Not only is the still photographer influenced by the photographer’s participation in interaction with the photographed subject but, as the case studies in this special issue demonstrate, the camera itself can influence what subjects do. This reactivity—the salience of the camera’s presence in scenes of everyday life—has been reflected upon more than has been the influence of the subjects on the photographer. (The camera or other recording device becomes an icon of being observed. To be photographed becomes very easily a matter of face threat, and in the interest of the subject’s face maintenance the photographer may give a warning—“say cheese.”).

One of my teachers, Edward T. Hall, made much use of still photography in his studies of “proxemics”—the cultural organization and meanings of interpersonal distance (e.g., Hall, 1966). He instructed us in how to use the still camera in fieldwork. His advice was to practice for hours at a time, walking around one’s living space looking through the camera’s viewfinder, and changing shutter speeds and aperture sizes while in motion. (In those days, camera settings were changed manually.) Hall said that we should become so familiar with the camera as an extension of ourselves that we could operate it smoothly without any hesitations or false starts so as not to draw attention to the mechanics of camera operation. (This was consistent with the then current notion of the participant observer’s role as “a fly on the wall” in order not to influence the conduct of the everyday interaction that was being observed. Much more on that later.) The point here is that Hall not only recognized that the observer’s presence can influence how the observed act but he also recognized that the camera could be more salient or less salient in the scene, hence its influence on participants was not constant but variable. When I began to use video with a handheld camera in fieldwork and taught students to do that, I reprised Hall’s instructions about the still camera. We practiced for hours with the camera, walking around living spaces and workspaces, looking through the camera’s eyepiece and adjusting focus and zoom distancing along the way (back then, aperture was automatically controlled but focus was done manually).

All this is to say that, while the camera doesn’t take pictures—people do—the camera can be considered as a participant in the local scene and in the relationships of mutual influence that are entailed in participation and in observation. The recording device’s salience in the scene changes from moment to moment, just as does the salience of the participant observer, and so it is useful to consider the camera as a character in its own right, working together with the observer.

All the articles in this special issue illustrate the changing salience of the video camera during the ongoing process of observation and recording. The cameras may be mounted on a tripod or handheld or be a body camera—in each case, sometimes the camera will be treated as more or less salient as recording takes place. In the article by Pehkonen et al., the camera suddenly became salient while recording blueberry picking in a forest when the camera was set down so the observer could swat a mosquito. Chen notes in her discussion of recording autistic research subjects that they are especially attentive to scrutiny by others, and also are especially attentive to the camera’s recording. She reports an instance in which two siblings could be considered to be performing for the handheld camera. She also reports another instance in which a young adult, who while being reprimanded by his teacher for snatching a cap from another young adult, looks pointedly at the handheld camera. It seems that as the teacher was drawing attention to his infraction, its documentation by the camera also became salient to the student as surveillance. Similarly, in the article by Goico on fieldwork focusing on deaf students in an elementary school classroom, tripod-mounted cameras became especially salient to those students at certain moments when they were engaged in infractions. Thus, the camera appears as relatively innocent and backgrounded at some moments while recording and then appears in other moments as a foregrounded symbol of pedagogical control.

3. The Observer as Co-Operative Participant

The camera’s relationships with participants in interaction are dynamic because the participants’ relationships with one another are dynamic as well. Nothing sits still in the ongoing course of real time performance of concerted social action. Goffman (1964) defined the social situation as an organization of focused attention during which participants constantly monitor one another’s actions, and since his time, a succeeding generation of close analysis of audiovisual recordings of interaction confirms his insight. What listeners are doing while they are listening influences what speakers are doing while they are speaking. Moreover, as concerted activity shifts from moment to moment, so do patterns of attention and engagement among participants. These relationships of mutual attention and influence have been labeled differently (as “participation frameworks” by Goffman (1981), as “postural positions” and “contexts” by Scheflen (1973), as “F formation” by Kendon (1990), as “participation structures” by Erickson and Shultz (1977/1997), and as changes in “contextualization” by Gumperz (1982). Shifts in relationships among participants have been called changes in “alignment” and in “footing” by Goffman and many others (Goffman 1981). Yet all these labels and notions point to the same basic phenomenon: in the ongoing course of interaction among participants there is continual potential for rearrangement of configurations of mutual attention and co-operative action, from the current present moment to the next moment. While certain participation frameworks may be sustained for a time, they can potentially shift at each new moment.

From this social ecological perspective on the conduct of face-to-face interaction, it’s apparent that the video camera, together with the participant observer as the camera’s operator, are just as much participants in the interactional ecosystem as are any of the other participants. It follows that as participation frameworks and activity systems change within interactional occasions, so the participation (and salience) of the camera and the ethnographer will change. Not to accommodate to these shifts in the organization and conduct of interaction—as all the other participants are doing—would be disruptive. Indeed, it would be asocial. The appropriate kinds of participation by the ethnographer and camera—where to look, what to say and to whom, when and how to move with others and with one’s self—are all matters of adaptation in the moment.

In this special issue, all the articles report shifts in participation framework and footing among interactional participants, including the participant observer during the course of fieldwork. Hofstetter, in her article describing fieldwork during board gaming and rock climbing, reports that at certain times she became a complete participant in a board game as it was being recorded. She also reports dramatic and sudden shifts from partial participation into complete participation while rock climbing, during moments in which she was connected by ropes to a fellow climber and it became necessary for her to take action to prevent her fellow climber’s fall. Pehkonen et al. report moments while recording search dog training in which it was necessary for the observer not to move while hand-holding a camera, so as not to distract the dogs from their searching task. Edmonds reports shifts in the course of her work in an animal shelter while recording with a body camera. Participating fully in cleaning animal cages because volunteers for that work were scarce, when a fight broke out among groups of chimpanzees Edmonds was not able to record that, or to record shelter staff reaction to it. Rather, she needed to continue in cleaning the cages. Her labor was necessary then, just as Hofstetter’s effort as a participant was momentarily needed to ensure the safety of her rock-climbing partner.

4. Participation as Practical Conduct

As Garfinkel emphasized in his seminal work, Studies in Ethnomethodology (1967), participants in everyday interaction are knowledgeable agents who take practical action, i.e. action that is opportunistically fitted to the particular circumstances of the moment. As practical social actors we can have expectations about what is likely to happen in an event, but we can never be exactly sure of what will happen next. Thus, we are not so much following socialized cultural rules (as “cultural dopes” (Garfinkel, 1964, p. 244) but are employing judgment capacities in what Bourdieu calls an overall “sense of the game” (Bourdieu and Wacquant 1992, pp. 120-121). We are making sense as we go along. The experience of timing in the ongoing flow of interaction is what in Greek is called kairos—the time of opportunity, of particular turning points that call for particular adaptive moves. In contrast to the continuous character of clock time, what in Greek is called chronos, the time of opportunity, kairos, is discontinuous. Practical social actors know how to handle this discontinuity, to “seize the moment” and make use of whatever comes to hand at that point in time. (An apt metaphor for the practical social actor is the bricoleur, the French handyman who re-purposes objects and techniques priorly designed for a certain function to perform a function that is necessary in a current moment. Examples of bricolage include wrapping medical tape around the drainpipe of a kitchen sink in order to stop a water leak, or, if lacking a wood chisel, using a screwdriver as a substitute.) In the ongoing flow of social interaction participants can never do nothing—they are always doing something. Sometimes what is appropriate is culturally pre-scripted action, but sometimes what is appropriate and adaptive is interactional bricolage. (For elaboration, see Erickson, 2004, pp. 134-174.)

The notion of participation as practical action has implications for how a participant observer comes to understand the practical conduct of what is being observed. A minimally participating observer—whether watching firsthand or in replaying an audiovisual recording—is presented with limitations on what can be known that way—not having to take practical action yourself. Description based on careful but distanced watching can be behaviorally precise but still lack hermeneutic validity — ethnographic adequacy. Watching from the sidelines, one misses the phenomenological feel of what it is actually like to be playing the game. For the minimally participating ethnographer there is a temptation to portray social action descriptively as gravitationless—weightless, effortless. Such description is sometimes interpretively appropriate, but if what is going on in a given encounter is struggle then it is misleading to portray that social action as effortless—lacking in suspense and uncertainty. Rather, to get an insider’s full sense of the game it is necessary to join in the game. (In my own career, my fieldwork conduct has changed from minimally participating observation to much more participatory engagement in action research. On the implications of this not only for hermeneutic validity in research-based understanding but for qualities in relationships that develop with informants as collaborators in inquiry and action, see Erickson, 2006.)

In discussing fieldwork during rock climbing, Hostetter observes that the kinesthetic experience of engaging with ropes and the rock face provides the ethnographer with insight into rock-climbing experience that video recording and analysis, by itself, could not provide. This is fieldwork learning by doing, literally at rope’s end. Goico, by being placed momentarily by students in positions of authority, comes to understand the finely nuanced tugs between teacher supervision and student noncompliance that are continually at play as underlife and face management within the everyday conduct of school classroom life. The embodied experience of sweeping out animal cages provided Edmonds with a member’s sense of animal care workers’ experience and practical conduct.

5. Ethics in the Participatory Conduct of Fieldwork

Ethics in seeking permission from informants and conducting ethnographic fieldwork with them involve matters that go beyond the conventional notions of research ethics that provide the basis for the guidelines and mandates of Institutional Review Boards (IRBs). It is important to recognize that IRB guidelines were initially developed to address ethical issues in medical research. The basic principle was “do no harm.” In medical research, a prime risk of harm to research subjects occurs at the point of data collection. The subject holds still (as a “patient” rather than as an “agent”) while some intervention is done—a surgical procedure, a medicine administered. Then the consequences of the intervention are examined. Interventions are considered as entailing different amounts of risk. Even a low risk data collection procedure, such as a blood draw, has slight risk—of infection or discomfort.

In video-based ethnographic fieldwork, the risks of harm are different from those in medical research and the points in time at which harm might occur are also different. Research subjects in an ethnographic study are agents, not patients, and in the ongoing course of routine social interaction they, and the co-operative action in which they are engaged, never hold still. Data collection does not proceed through intervention by the researcher—indeed, in the classical image of participant observation, intervention (i.e., interference in the normal conduct of everyday life) is to be avoided. When done with respect and genuinely informed consent, fieldwork doesn’t actually hurt people. (In teaching about this, I say that, unlike an x-ray photograph as an information collection technique, videotaping does not in itself cause harm to those whose everyday practices are being recorded.) What is potentially hurtful comes later in the research process—at the point of research reporting. And there the potential harm is not physical but social—the possibility of face threat or institutional/community sanction as people’s everyday conduct is revealed through description and perhaps also through the circulation of video recording. Who will see the video records and read detailed narrative vignettes or transcriptions? Under what circumstances? This is what can make research subjects anxious while fieldwork is taking place and after it’s been completed.

What is “consent” and “being informed”? The IRB informed consent procedures, based on experimental or quasi-experimental design in intervention research, presume that the purposes and procedures of research will not change during the course of data collection. It is thus ethically responsible to inform research subjects of the study’s purposes and risks before their participation as subjects begins, and to seek consent at the outset in circumstances that do not involve explicit or implicit coercion. However, the situation is more complicated in ethnography. As ethnographic fieldwork proceeds across time the focus of inquiry often evolves—neither the research participants, nor the research questions, nor the observer stay entirely constant in what they are doing. Thus, it is ethically appropriate to inform subjects recursively during fieldwork as the focus of inquiry develops, and if that focus has changed substantially, to seek amended consent from informants, not necessarily in writing, but at least orally, as informed “assent.”

Consenting and informing, as continuous processes developing over time, also involve recognition of the risks to subjects that can result from the behaviorally specific descriptive reporting that characterizes ethnography. In my own work, in long-term (often year-long) video-based ethnography and action research with early grades classroom teachers, I have negotiated consent agreements that reduce the risk of face threat to the teachers by specifying conditions for written descriptive reporting and for viewing of video footage by audiences. At the outset of a study, while negotiating entry to a field setting, the consent form I have developed for teachers says that article or book length ethnographic report drafts will be shared with the teacher for review prior to publication. If the teacher disagrees with the characterization of her teaching practice, she does not have the right to censor the report, but does have the right to write a rejoinder commentary of disagreement with the researcher’s interpretations which would be published along with the report. Also concerning video footage, if certain unanticipated events take place fortuitously that might be embarrassing to the teacher, or to students and their parents—e.g. a child’s emotional “meltdown” while being videotaped, or a lesson that goes entirely off the intended track—the consent form for the teacher (and for parents) states that they have 48 hours to decide whether or not to request that the video footage showing that untoward event be erased. Such a request from a teacher or from a parent ensures erasure. Also, the consent form for teachers and for parents guarantees that no one in the local setting can ask to see footage from the classroom (including the school’s principal and teaching colleagues, as well as parents) without the teachers’ consent. (The rationale here is that risk of harm from institutional or community sanction for being portrayed in narrative description or on video is greatest in the local setting. Research subjects who are videotaped can be anonymized when the footage is shown to general scholarly audiences, but that is not possible when the video is shown locally.)

Over the past forty years, in studies involving long-term observation and literally hundreds of hours of video recording, no teacher, student, or parent has requested that I erase a single instance of video footage, nor have school administrators or parents asked to see footage without the consent of the teacher. The same holds for written descriptive reporting. Teachers have never asked to publish a written disclaimer, although as a result of their review of manuscript drafts I have sometimes revised my reporting to accommodate their suggestions for changes in interpretation.

It follows that when it is culturally appropriate (and admittedly, that is not always the case), explicit assurances about the researcher’s responsibility for preventing face threat in reporting provide a foundation for the development and maintenance of trust with research subjects, and this is especially important in circumstances in which there are asymmetries of power and authority in relations among interactional participants. (This includes many situations of family and community life, as well as situations in formal organizations such as schools and other workplaces.) When trust is threatened during the course of fieldwork, anxiety rises about the camera’s recording and even about the observer’s gaze and note-taking. The researcher’s activity comes to be seen as panoptical surveillance whose consequences could be harmful. Thus, the fieldworker’s engagement with research subjects needs to go beyond the limiting principle of “Do no harm.” It involves living up to the positive injunction “Do unto others as you would have them do unto you.”

It’s apparent from the previous discussion that ethical aspects in fieldwork relations, as well as in research reporting, are not simply fixed at the outset of a study. They are continually under negotiation and adjustment, on a daily and even momentary basis, within the kairos of immediately contingent circumstances. Here, too, the insight from ethnomethodology about practical reason and practical conduct is pertinent. Fieldworkers, in their relations with the people they are studying, are not simply following rules. As practical social actors they are making sense, reflexively, in concert with those with whom they are engaged in social interaction during the course of doing fieldwork. (For further discussion, see Guillamin and Gillam (2004), Paoletti (2014), and Shaw et al. (2019)).

In the articles in this special issue, each fieldwork situation presents the fieldworker with particular ethical dilemmas and responsibilities during the conduct of observation and participation. And it should be noted that the potential for research subjects’ vulnerability (as in “vulnerable subjects”) is not a fixed quantity. Vulnerability can emerge for a research subject in one moment and shift in the next. Hofstetter was ethically obliged to respond to the immediate needs of her rock-climbing partner during “belaying” partnership in climbing. (In contrast, during Hofstetter’s board game participation the stakes for life chances with her partners were not high, but they were very high in her partnership during rock climbing.) Pehkonen and colleagues, as they videotaped search dog training, would have been irresponsible to move with their cameras so as to maintain within the visual frame of the recording the fast-paced action of the dog training. They were recording in circumstances in which such movement with the camera could have confused the dogs and thus interfered with the trainer’s work.

It seems to me that issues of research ethics were especially foregrounded in the other three articles. In each of those cases there were circumstances of considerable power asymmetry in the research site. Chen’s documentation of an autistic student’s angry resistance while being reprimanded by his teacher/caretaker involved Chen’s aligning in the moment with the teacher’s exercise of authority—a change from her previous alignment with the student. This led her to wonder about the ethical appropriateness of continuing to record as the unexpected incident developed. (Since she had not anticipated this as a potential problem during the process of negotiating research permission—with a “video erasure” provision—she was left worrying about this after the recording was done.) Goico also found herself repeatedly engaged (and then repeatedly disengaged) as an authority figure in ongoing relations with students and their teacher in the classroom. This is an inevitable consequence of pervading power asymmetry in teacher-student relationships in classrooms, and of power asymmetry in student-student relationships as well. Which lines of solidarity will the fieldworker establish—those with the teacher or those with students?

Finally, Edmonds’ ethical choices in her fieldwork at the animal sanctuary seem to me to be even more poignant and complicated than in those that arose in the other fieldwork cases. She reports power and prestige asymmetry at the animal sanctuary in the relationships between supervisors, volunteers, and full-time animal caretakers. There was a distinct division of labor, a racial/social class hierarchy, and a language hierarchy in the setting. Administrators, researchers, and student volunteers from France were speaking a cosmopolitan version of that language while Cameroonian animal care workers were speaking a pidgin version of English (and presumably also a locally creolized version of French.)

The Francophones did “think work” and the pidgin speaking care workers did difficult physical labor in inherently unpredictable circumstances. (Just as in the Finnish forest described by Pehkonen et al., the appearance of mosquitos was likely but not exactly predictable in the moment, so in the Cameroonian wildlife sanctuary as described by Edmonds, outbreaks of conflict among animals and recalcitrance in their relations with caretakers and volunteers were expectable but not precisely predictable.) When a fight broke out among chimpanzees outside the animal shelter building, for Edmonds to have stopped her cage cleaning in order to be able to record managers down the hall as they discussed what to do about the fight, would have been to abandon her responsibility in assisting the care workers. She notes that there was a serious shortage of volunteers to help the full-time care workers, i.e. the care workers had at that time a special kind of “vulnerability” in their work life. Which alignment would Edmonds adopt—that of solidarity with the managers and animal researchers or that of solidarity with the care workers? The labor of the latter was “manual” and thus lower in prestige than that of the former, just as the speech of the care workers was lower in prestige than that of the managers. Yet it was the care workers who had direct responsibility for and practical knowledge of the animals under their care. Edmonds couldn’t have it both ways. Her ethical choices, repeatedly, were to align with the care workers.

6. Conclusion

As noted at the outset of this commentary the social interaction that is documented, analyzed, and reported in video-based ethnography is in Charles Goodwin’s terms co-operative—a matter of participants working together—their actions in relations of continual mutual influence in real time, by which they complete one another’s actions in concert with one another (Goodwin, 2018). From that point of view the fieldworker is just as much a participant in the social ecology of the present and next moment as is any other participant. In other words, Goodwin’s conception of co-operation is foundational for understanding not only how everyday social interaction is able to take place at all, but how fieldworkers are inevitably participating in such interaction as they are studying it.

Another foundational notion comes from Garfinkel’s (1967) understanding of the participant in social interaction as a knowledgeable social actor who assesses the immediate circumstances that are present in the kairos of the moment and who takes practical action that adapts opportunistically to those particular circumstances. Moreover, as Garfinkel and Lynch later observed in their call for social studies of the everyday work of scientific researchers (see Lynch, 1993)—the researcher is also a practical actor in the world, taking action opportunistically. Thus, the researcher operates on the same epistemic plane as any other social actor, with the same limits on information and the same need for improvisatory skills in taking immediate action that any other participant has. This applies not only to research in the physical sciences but to social research as well. The ethnographer has neither more nor less knowledge and capacity for action than do the research subjects who are being studied.

Also foundational for this discussion have been Goffman’s observations on footing (1981). Since interaction is not exactly predictable from moment to moment, relations of alignment and mutual influence among participants are continually changing, and thus the fieldworker’s participation changes as well.

In sum, when we consider together the case studies presented in this special issue and we take together the insights of Goodwin, Garfinkel, and Goffman about the nature of interaction and its practical conduct in real time, we are returned to the even more foundational insight of the ancient philosopher Heraclitus. Recall his claim that all things are continually in flux. It follows that, as video-based ethnographers, in the moment-by-moment conduct of fieldwork, we and our cameras can never step in the same river twice. If, as Heraclitean-informed fieldworkers, we’re being realistic, we will not presume that constancy in participatory role is even possible during fieldwork’s course. Or that non-participation in the fieldwork setting is possible either—let alone desirable.