Social Interaction

1. Introduction

Video ethnography is a powerful tool for the study of the everyday, situated human interactions of individuals with communicative differences. Over the last 30 years, ethnomethodology and conversation analysis (EMCA) methodologies have been used to document, precisely analyze, and thus clarify the nuanced social practices of individuals diagnosed with various communicative, cognitive (Antaki & Wilkinson, 2013; Wilkinson, 2019), and physical (Auer & Hörmeyer, 2017) disabilities. Recording and analyzing naturally-occurring interaction has been especially beneficial for the study of individuals with little to no speech production capacity. In the notable empirical and theoretical work of Charles Goodwin, Chil, a person with aphasia and limited linguistic production, is revealed to position himself as a competent interactant through elaborate, temporally-coordinated collaboration with others (C. Goodwin 2004). By shifting the analytical focus from speech to the organization of situated action, Goodwin’s work uncovers the cognitive life of Chil and elaborates upon Goffman’s model of participation frameworks (Goffman, 1981) beyond the traditional model of talk. Goodwin and Goodwin (2004) posit that privileging the stream of speech as an analytical focus for participation in interaction concomitantly denies full status of a participant who lacks fluent, complex speech ability.

By providing evidence and accurate grounds for observable social phenomena (Garfinkel, 1967; Sacks, 1992), EMCA methodologies can make “taken-for-granted” knowledge explicit by transforming tacit resources into topics that are elucidated in their own right (Watson, 2006). For the study of non-speaking populations, such as the non-speaking Autistic¹ population, the analysis of video data has much potential to surface facets of embodied interactional practices that may otherwise be missed, and to expand upon the rich accounts of Autistic sociality already existing in EMCA literature (e.g., Dickerson et al., 2007; Muskett et al., 2010; Sterponi & Shankey, 2015).

During data collection, the researcher plays a significant role in shaping an Autistic individual’s conduct, especially when the researcher enters a field site with a “roving camera” (Heath et al., 2007, p. 38) and becomes a part of the ongoing social and material ecology. In the situated activity of video recording interactions, the researcher can be argued to be in a participation framework with their participants as a bystander, and sometimes even as a ratified participant in the interaction (see also Edmonds, 2021/this issue, and Hofstetter, 2021/this issue). Technical decisions with the camera, interactions with participants on the site, and the orientation of the researcher’s body to that of the ongoing scene can all have consequences for the interactions that unfold. Although promising, the process of collecting video data of non-speaking Autistic individuals brings unique challenges. Firstly, Autistic individuals experience deep permeability with their environment (Conn, 2015; Sterponi & Chen, 2019). Data collection requires careful navigation of a constantly evolving social and material environment. Secondly, Autistic individuals can experience stigma for their bodily actions and are aware of their social rejection by others (Fein, 2018; Ochs, 2015). The researcher’s actions during the data collection process can amplify their felt difference from others. Lastly, they are a vulnerable population (Jaarsma & Welin, 2012), which requires that the researcher exercises sensitivity during the data collection process and after.

The technical work of data collection with non-speaking Autistic individuals has to be skillfully intertwined with careful navigation of emergent interactional circumstances. Furthermore, the researcher’s involvement in participation frameworks during the data collection process can engender unexpected participant roles. The quick and careful navigation of multiple participant roles can be complicated when non-researcher participant roles are invoked by the ongoing circumstances. Because Autistic subjects are vulnerable, the navigation of a researcher’s participant roles onsite can even be ethically blurred and contradictory, especially when these roles are summoned by participants within the ongoing scene.

Video data collection of Autistic interaction surfaces questions that are central to video-based fieldwork: What are the technical and interactional challenges of navigating multiple participant roles of the researcher during data collection? What are ethical issues that arise with emergent participant roles during data collection? This paper examines the interplay of participant roles in collecting interactional video data of non-speaking Autistic individuals. In the first analysis, I examine two siblings in an intimate multisensorial interaction. I focus on the complexity of handling a video camera while navigating simultaneously-occurring participant roles, empirically demonstrating the Autistic child’s awareness of the larger participation frameworks at play. The second analysis builds upon the first by examining the ethical dimensions of engaging in these complex interactional dynamics within data collection, as well as the ethical implications of the Autistic individual’s awareness of the researcher’s participation frameworks. I examine a tense interaction between an Autistic adult and his teacher. Both participants are misaligned in their agendas, which leads to an accumulation of mutual tension between them. I unearth possible participant orientations to the camera and discuss implications for video recording such difficult and unanticipated situations. Lastly, I show how the summoning of the researcher’s involvement in the ongoing scene puts the researcher in blurred and contradictory positionalities. I argue that in developing the professional vision (Goodwin, 1994) of the EMCA scholar (Katila & Raudaskoski, 2020) during data collection, the creation of participation frameworks moment-by-moment is ethical in nature.

2 Methods, Data and Settings

In this paper, I am the primary researcher and author for data and analysis. The social interactions examined in this paper come from two different video-ethnography corpora that I collected in Singapore. The researcher (me) and the participants were all native Singaporeans. The languages spoken onsite were English and Singapore Colloquial English (Leimgruber, 2013). As the primary researcher, I gained access through ethical approval and permission from the university² , written consent from my participants’ guardians, and additionally, from the Autism institution where necessary.

The first 4h corpus documents Alex, a 10-year-old, non-speaking boy diagnosed with Autism, and his interactions with his parents and sister (14-year old) as they engaged in different activities at home. The family is English-speaking and Singaporean-Chinese. At the time of data collection, it was my second time visiting the family in the context of data collection. Prior to this visit, I had known them for about two years because the mother of Alex and Bridget, Aunty, was my own mother’s friend. I select video data of a 30s interaction between Alex and his sister Bridget, where they sit on the floor beside me and participate in a multisensorial interaction with their hands, bodies, and feet.

The second corpus documents social interactions that transpire over a period of two weeks in an Autism institution. This facility enrolls youth and adults diagnosed with moderate to severe Autism between the ages 19-55 and offers various skills training and activities for the students for about six hours a day, Monday through Friday. Data (30h) were gathered from two classes of twelve non-speaking students between the ages 20 and 30 engaging in activities such as taking walks, going for outings, gardening, crafting, and swimming. The analysis centers on an episode in which a participant displays what seems to be observable as distress behavior.

Video data was transcribed according to the transcript conventions by Jefferson (2004). Multimodal transcription follows the style of Selting (2010) and additional screenshots and images follow C. Goodwin (2018). During data analysis, I include myself—the researcher—as a co-participant in the ongoing interactions. However, as I transcribe and analyze the data, I draw upon my recollection of being present at the site, ethnographic data on my participants’ backgrounds, and literature on Autism, which allow me to attend to myself as a participant within the events that unfold.

3 Researcher participant roles and participant orientations

This section begins from the notion that the participant-observer’s and participants’ reflexive orientation toward each other constitute a participation framework (see also Goico, 2021/this issue). Participants, including the researcher, produce courses of actions which position themselves to perceive in ways that are relevant to the activities in progress (C. Goodwin, 2007). In this section, I focus on a 30s interactional sequence that focuses on Alex in interaction with his sister, Bridget. I examine the interactional work undertaken by me partaking in simultaneously-occurring participation frameworks, and how my ongoing participation has an effect on the technical work of the camera. I then examine the Autistic child’s orientation to my participation frameworks and discuss implications for unearthing the communicative competencies of Autistic individuals.

I had been video recording for about an hour prior to this data extract, and the recording was taken during the first of two days filming the family. I had paused filming because my video camera had a low battery signal. Aunty, who inquired on my camera status, began the activity of searching for extra batteries for me. As we engaged in conversation, Bridget and Alex were seated on the floor by my right, and Bridget started to inch towards Alex. I noticed her movement and anticipated the potential forthcoming of an interaction between the siblings. I therefore turned on the camera in my hand and began the data collection process.

Throughout most of the filming of this data, I recall being physically positioned to face Aunty while simultaneously holding the camera by my side towards Bridget and Alex (Figure 1). My bodily configuration thus afforded me the opportunity to become a participant in two simultaneously occurring participation frameworks (C. Goodwin and Goodwin, 2004; M.H. Goodwin, 2006; C. Goodwin, 2007): the activity of attuning to Bridget and Alex, and the other, the activity of conversing with their mother, Aunty.

Figure 1³

At the start of the recording, I spend some time adjusting the camera to Bridget’s and Alex’s interaction while simultaneously conversing with Aunty. In the following extract, I focus on an audio-recording of my conversation with Aunty, as well as the shifting camera angles from which Bridget and Alex were filmed. I demonstrate how the interactional work I undertake has an effect on my technical work with the camera. Screenshots from the video are converted to line drawings and edited to visually demonstrate the participants’ positions within the frames.

Extract 2

Extract 2 begins when the camera is turned on. Aunty uses Singlish, a colloquial variety of Singapore English (Wee, 2014; Wong, 2014) that is an unmarked code for informal interaction, expressing familiarity and solidarity between interlocutors (Goddard, 1994). Singlish is rich with pragmatic particles such as ah, leh and lah (Wong, 2014), which are used to convey certain attitudes or assumptions about the reactions of others (Gupta, 1994). In Line 1, Aunty’s utterance “nothing leh”, is an announcement that there were no batteries in the closet, with the discourse marker “leh” (Botha, 2019) indicating an apology and that she had tried her best. Aunty’s use of “leh” is a marker for informal interaction, and I reciprocate in a similar manner. I respond to Aunty in Singlish, stating “it’s okay lah” (Line 2). “It’s okay” lets Aunty know that I would not require a battery, and “lah” discourse particle indicates reassurance (Wee, 2004). In engaging with Aunty, I keep the interaction friendly, using the appropriate kinship term “Aunty” so as to sustain our participation framework.

The interactional work I undertake is coupled with technical challenges in camera placement. The video camera has a 3-inch digital screen that allows me to view what the camera is capturing. At the start of the recording, I hold the camera close to my face to look at the digital screen, thus capturing the interaction from a slight top-down perspective. As I produce my verbal utterance in Line 2, the video camera veers to the far right as seen in screenshot 2. Alex falls off the frame, and the camera shot is slanted. I continue to engage in conversation with Aunty and proceed to express my gratitude to her (Line 3). My video camera remains slanted to the right and I recall noticing the slant through quick glances at the digital screen. It is only in Line 4, when Aunty produces her turn, that I am able to adjust the focus of the video camera and include both Alex and Bridget in the shot frame (screenshot 3). In Line 5, I then adjust the video camera even further: I zoom into the participants’ interaction such that I capture as much of the body as possible within the camera frame (screenshot 4).

In the extract above, I have shown how I had to manage dual roles, one as attending to Aunty, and the other as attending to the children’s interaction. The interactional work undertaken also creates camera technical challenges experienced by me in navigating both participation frameworks. In the next segment, I analyze two extracts, but I turn the analytical focus to Alex and examine his attunement to the camera moment-by-moment. Bridget and Alex’s interaction involves the affective co-engagement of its participants in a constitution of intimacy (M. H. Goodwin, 2017). Such interactions are challenging to capture not only because they require careful attunement of the camera to the interaction, but also because participants have to feel comfortable enough to display such intimacy to an observer.

The phenomenon of “doing being observed” in video ethnography data has long been documented and theorized. In Harvey Sacks’ Lectures in Conversation (published posthumously in 1992), Sacks discusses an excerpt from the beginning of a group therapy session, where participants attend to the microphone in a theatrical, play-like fashion. In a similar vein, the researcher becomes an audience member to be entertained by the siblings once she begins the recording process. The following extract describes the same interaction as the one analyzed above, but the analysis instead focuses on Alex.

Extract 3

As seen in Extract 3, just as the video recording begins, Bridget sticks her tongue out at Alex. Her invitation to Alex seeks playful collaboration from him, and he imitates her tongue-protrusion gesture (Metzoff & Moore, 1977) by also protruding his tongue (Figure 3.1). In Line 1 when Aunty exclaims “nothing leh”, Alex shifts his gaze from Bridget towards the camera (Figure 3.2), showing his awareness of being observed. Alex then shifts his gaze again towards his mother (Figure 3.3). Alex’s shift in visual attention to the camera and then to Aunty demonstrates an awareness of the ongoing participation frameworks, the first of him being observed and the second of Aunty’s conversation with me.

Alex’s shifts in eye gaze are significant when considering his Autism diagnosis. Despite gaze aversion in face-to-face interaction being common in Autistic interaction, Autistic individuals have been shown to be deliberate in eye gaze shifts for a variety of reasons, including to reduce stress (Jaswal & Akhtar, 2019), and even as an interactional resource (Dickerson et al., 2005; Korkiakangas & Rae, 2014; Dindar et al., 2017). Alex’s shifts in eye gaze in this data indicate his understanding of the complex interactional dynamics at play.

As a researcher, my own onsite attunement to the interaction between the siblings can be argued to sanction and even encourage social interaction. It is also clear that Alex is cognizant that he is receiving the attention of the camera. Furthermore, my involvement with Aunty in one participation framework and the siblings in another allows for the siblings’ interaction to unfold independently without engagement with Aunty. In the moments immediately after Extract 2 and Extract 3, the following interaction occurs between the siblings. They orient their faces to the camera as the sequence unfolds in a performative act of doing-being-observed.

Figure 4

Figure 4.1

Figure 4.2

Gif Figure 4.1-4.2

Figure 4.3

Figure 4.4

Gif Figure 4.3-4.4

In Figure 4.1, Bridget inches even closer to Alex, puffs out her right cheek, and extends it to him. By so-doing, Bridget orients her face towards the camera. Alex imitates Bridget by puffing his right check out (Figure 4.1). Both Alex and Bridget lean their bodies toward each other, and Alex de-puffs his right cheek, puffing up his left (Figure 4.2). As Alex does so, he looks directly at the camera. Alex and his sister lean towards each other in Figure 4.3, bringing their puffed cheeks towards one another. When their cheeks touch, Alex glances at the camera (Figure 4.3). As they move apart again, Alex and his sister close their eyes and release the air from their cheeks (Figure 4.4). Alex’s and Bridget’s interaction is affectionate and intimate, and their faces and bodies are oriented to the camera as they share the moment.

As I have demonstrated in this section, the presence of a camera can occasion an expectation for an interaction to unfold. Alex’s acute awareness of the camera is evident through his explicit shifts in eye gaze throughout the video recording. I have also shown how the camera could have played a significant role in engendering particularly affectionate, play-like actions from the siblings, as performed for the camera. I have discussed the multiple participant roles I play within different participation frameworks and how I attune to each of my participants differently, cooperating in their respective activities-in-progress. By surfacing my multiple participant roles during the data collection process, I have demonstrated how my own embodied actions are inextricably intertwined with the local contextures I document. Alex’s ability to participate in such affectionate interaction, and his acute awareness of the participation frameworks at hand, defy clinical characteristics of Autistic individuals as having an inability to emotionally reciprocate (DSM-5, 2013).

4 Ethical dilemmas in blurred participant roles

I have shown in the previous example how the act of recording itself comprises a participant framework. The researcher’s work requires navigation of onsite participation frameworks while managing camerawork that is attuned to the Autistic individual. The complexity of the researcher’s work can sometimes be made even more challenging when unanticipated, difficult events occur during data collection. For example, how should the researcher position herself during situations of tension and distress (see also Wootton, 2012) that may emerge during data collection? If, as seen in the above section, Autistic individuals are aware of the intricate participation frameworks that exist during data collection, what are the implications of the researcher’s involvement in an ongoing scene?

In this section, I demonstrate how the multiple participant roles of the researcher during data collection undoubtedly highlight pertinent ethical issues in conducting EMCA video-based fieldwork on Autistic individuals. I hope that this section makes a contribution in uncovering some of the ethical dimensions involved in data collection.

In a recent “ethical turn” in anthropology, ethics and morality have been foregrounded as “co-constructions of the observer and the observed” (Fassin, 2014, p. 432), and themes such as vulnerability and suffering have been explicitly invoked and discussed in a range of activities (e.g., Throop, 2010; Zigon, 2011). The researcher, through the concept of an ethnographic epoché, is invited to actively extend the limits of data interpretability so as to include her own subjectivity and to attend to her self-experience (Throop, 2012). The approach taken in EMCA seeks to elucidate the tacit knowledge-base of participants, and how their local contingencies affect their behaviors and interactions (Watson, 2006). In the following analysis, I first analyze, as accurately as possible through EMCA, the unfolding scene and my participant roles within it. In the spirit of ethnographic epoché, I then reflect upon the blurred positionality of these participant roles and the ethical dilemmas that are unearthed as a result.

The extract of focus in this section involves mounting tension between teacher and student, and eventually a display of heightened emotion, which I will call a “distress display”. Although my purpose for conducting fieldwork was to study Autistic interaction more generally, during data collection I encountered instances of distress displays. I understood that these displays could happen, but I did not anticipate the unpredictability, frequency, or intensity of the displays when I conducted fieldwork in the institution daily. Whereas the invocation of ‘distress’ in this episode suggests that the person enacting the display is in distress, I use the term to posit that these are observable displays of seeming distress, not if the individual is necessarily experiencing distress itself. In the interaction examined below, taken about one and a half weeks into the data collection process, two classes of 12 students and 4 teachers are out for a morning walk around the Autism institution, an activity that is a part of their morning routine at least once a week. The walk itself takes place on a long public pathway in between residential houses, with rows of plants and trees running down both sides. The group has been walking for about half-an-hour that morning, and students and teachers alike are traversing at varying distances from one another (Figure 5).

Figure 5

Prior to the data extract, Sam, a 26-year-old non-speaking male Autistic student at the institution, had just grabbed the hat of another classmate. The student who lost his hat appeared unbothered and continued walking ahead. Sam’s teacher, a female in her 20s, puts pressure on Sam to return the hat (Figure 5), issuing verbal instructions and using various forms of control touch (Cekaite, 2015). Upon examining the larger corpus, it seems Sam often takes others’ personal items and is then asked to return them by his teachers. In this episode, Sam’s teacher issues verbal instructions to return his classmate’s hat, but she is met by continued resistance from Sam. Because the teachers and students are outdoors in a public area, the teachers’ top priority is to ensure that the activity of walking proceeds smoothly, and that risk toward the students or the public is minimized. When Sam sprints off, he presents a risk to progressing the activity, and his teacher runs after him. She catches up to him and grabs his shirt to hold him back. Figure 6 depicts the moment-to-moment sequence of events that follow.

Figure 6

In Figure 6, the teacher has just caught up with Sam and grabs his shirt (Figure 6.1). Once the grabbing occurs, Sam immediately turns around and attempts to grab his teacher’s hat (Figure 6.2). The teacher spreads her feet into a wider standing stance, pushing Sam away from her. Sam then attempts to grab his teacher’s glasses (Figure 6.3). His teacher leans back and pushes him away from her but continues to have a strong grasp on both his shoulders. From Figure 6.3, it is clear that the teacher is attempting to prevent the grabbing of her hat, and possibly, even to protect herself as Sam attempts to grab her items. By maintaining a hold on Sam’s shoulders, she is also continuing with a form of control touch to prevent him from running away as he did earlier. Both Sam and his teacher push their forearms against one another: Sam’s teacher keeps her hands on Sam’s shoulders, her own elbows locked at the joint so as to keep Sam at a certain distance from her. The tussle between Sam and his teacher leads to the interaction in Extract 7.

Extract 7

In Line 52, Figure 7.1, Sam and his teacher are in a face-to-face spatial configuration (Kendon, 1990) with the teacher’s hands on Sam’s shoulders and Sam’s forearms resting on top of his teacher’s forearms. Both are staring into each other’s eyes as the teacher softly pleas with Sam to stop (Line 52). Both face each other for a brief moment before Line 53, where Sam screams “Haiyah:::”, elongating the vowels in his vocalization together with fluctuations in prosodic contour. Sam’s body simultaneously flings outwards toward the direction of the camera (Line 54, Figure 7.2).

The researcher, being attuned to the ongoing interaction, is in a tricky position when the above episode unfolds. During data collection it can sometimes be impossible to make quick decisions about recording: I did not realize at the time that a distress display was about to happen. It was only afterwards, during data analysis, that I was able to unpack the unfolding events. Reviewing this data, perhaps a first pertinent ethical question would be whether or not I should have filmed this segment at all. The increasing mutual tension between Sam and his teacher clearly cumulated to the distress display. However, given that the camera can occasion doing-being-observed and that the Autistic individual is sensitive to the participation frameworks at play, the presence of a camera during this sequence could have been involved in the unfolding circumstances. In this sequence, Sam in Line 54 flings his body towards the camera before dropping onto the floor, suggesting that there is a possibility Sam was orienting to the camera. In the moment of data collection though, I was not able to predict this outcome of the interaction.

Conversely, documentation of these difficult moments is valuable in revealing 1) how Autistic emotionality is embedded within social interactional contexts, and 2) how the otherwise tacit codes held by the participants play out in everyday circumstances. For example, Sam’s outburst could at first glance be categorized as a ‘meltdown’: an intense emotional response to overwhelming circumstances, characterized by a complete loss of control (Lipsky, 2011). Sam’s possible awareness of and orientation to the camera during his distress display suggests that he may have had more control over his behavior than what is otherwise suggested by a meltdown.

Through the above analysis, it is clear that Sam and his teacher are in misalignment in terms of their agendas. The teacher’s priority lies in teaching Sam not to take others’ possessions and to return items to their rightful owners. Sam’s priority, on the other hand, lies in holding onto the hat, and he continues to struggle against complying with his teacher’s agenda. The extract reveals the ways in which certain Autistic behaviors are responded to. The teacher meets Sam’s mounting resistance with repeated verbal instruction and various forms of control touch. In addition, her insistence that Sam return others’ belongings reveals how teachers may hold normative moral codes, such as returning others’ properties, even if prioritizing such codes may lead to distress. It is unclear whose moral codes the teacher is acting upon: Is the hat-returning her own personal prerogative, the Autism institution’s rules, or perhaps even influenced by the larger medical model of Autism that presents larger institutional forces beyond this specific institution? Any of these factors, or a combination of factors, could have played a role in the teacher’s insistence. This data extract is but a starting point toward unpacking the multilayered socio-cultural and socio-political influences underlying practices that govern Autistic behavior.

Another issue that the above extract raises involves the teacher’s use of control touch on Sam. There are practical constraints present at the site, namely the ratio of teachers to students, which is 1:3, the small size of the teacher in comparison to the adult male student, and the Autism institution’s regard for keeping order amongst students and moving forward with the day’s schedule. Through the sequential unfolding of the interactional sequence presented in this section, it becomes clear that the teacher resorts to control touch as an attempt to more easily maintain order in this public setting, and also to protect herself. The practical circumstances that shape this sequence of events bring to light the daily challenges in day-to-day life within this institution.

Situating this data within the larger literature on Autism surfaces more complex issues. When considering the hypersensitivity of Autistic individuals to touch (Baron-Cohen et al., 2009; Fitzgerald, 2013; Sapey-Triomphe et al., 2019), the teacher’s use of corporeal control may have contributed to the student’s distress. Such data begins to unpack the larger implications behind how Autism is defined institutionally and how certain behaviors are responded to in educational practices. Autistic scholarship such as Nolan and McBride (2015) discuss how in the medical institution, the Autistic sensory experience is constructed as a disablement. The rhetoric of deficit-driven medical models of Autism pathologizes the “lived body” of the Autistic, which may have an influence on educational practices involving control touch. Examining the management of Autistic bodies in interaction supports establishing a Goffmanian basis for Autism diagnosis based on “local interaction order” rather than the traditional Foucauldian-style accounts behind the clinical label of Autism (Maynard & Turowetz, 2017). Distress episodes such as the one analyzed above unveil socio-interactional contingencies to Autistic emotionality. These analyses can play a crucial role in redefining Autism as a starting point for the reconceptualization of educational practices that situate Autistic sensibilities within co-participation.

There is much value in capturing these difficult interactions on video. However, the filming of such interactions is delicate and necessitates caution on the part of the researcher, who has the potential to further perpetuate Autistic disablement. Further ethical issues ensue after, when I, as the researcher, am summoned to be involved in the unfolding scene (see also Edmonds, 2021/this issue, for participants’ expectations on the researcher’s contributions to the ongoing activity).

Extract 8

After Sam flings his body toward the camera, he then turns around and sinks his body to the ground as in Extract 8, Line 55. In Figure 8.1, Sam’s teacher loosens her grip on Sam’s forearms. In Figure 8.2, she turns around. Behind her and in front of her is a stretch of path, and all the other teachers had already walked far ahead, leaving Sam and his teacher behind. The teacher therefore turns to me and summons me into achieving her mission of hat-returning (Line 56). My utterance “Yah can” in Line 57 latches onto the teacher’s and I comply with the request by turning off the camera and taking the hat. In my actions it is clear that I am responding to the needs of the teacher in these unforeseen circumstances. However, my alignment with the teacher presents some ethically conflicting consequences. Even though Sam has just let out a distress display, and is now lying on the floor, the teacher continues to prioritize the returning of the hat. Prior to line 56 and 57, Sam and his teacher have misaligning agendas that are sustained throughout the interaction and that eventually lead to Sam’s distress display. My alignment with the teacher means that I enter into a participation framework with her that involves rectifying Sam’s violation of the normal ethical code of respecting the property of others. By complying with her, I inadvertently demonstrate my stance (M. H. Goodwin, 2008) against Sam’s goals and partake in fulfilling the teacher’s objective. Of significance is also the fact that I, presenting as a neurotypical adult, align with the neurotypical adult teacher without centering the Autistic participant in the interaction. When as researchers we are “not ourselves members of historically marginalized populations" (Edmonds, 2021/this issue), our alignment with those in power can perpetuate vulnerability. The ethical bind of the researcher is precisely in her blurred positionality as a researcher as she is co-opted into the activities of the institution.

In the above sequence, I analytically explicate the participants’ tacit knowledge-base about proper adherence to the ongoing interaction, including my own as a researcher in the scene. In light of my participation in the scene, I inevitably partake in the moral code set forth by the teacher. This data raises important questions. Is Sam responsible for his actions as a competent knowing actor? Is the teacher of a person with vulnerabilities responsible? What is my responsibility as a researcher in this scene? As an EMCA researcher, my role as a bystander shifts when I become a ratified participant. By returning the hat, I am complicit in my participants’ moral code, whether it is shaped by the teacher’s own practices, the rules of the Autism institution, or the larger medical institution within which Autism is defined.

5 Discussion

The ethical role of the participant observer in bringing light to human suffering and vulnerability has been present in EMCA and anthropology studies of medical, legal, and other institutions. For the EMCA researcher, studying the Autistic population surfaces questions that are core to the nature of video-based fieldwork. In this article, I have shown how the EMCA researcher can play an active role in negotiating and shaping the very interactions being investigated. I begin by demonstrating how during data collection, the researcher has to undertake complex interactional work by managing multiple participant roles in different simultaneously-occurring participation frameworks, while orienting the camera to ongoing interactions. I show how an Autistic child is aware of these multiple participation frameworks and performs doing-being-observed. Participants’ awareness of the camera have ethical implications in other more difficult recording circumstances. I unveil the ethical nature of participation frameworks as they are created moment-by-moment within the data collection process. Ethical dilemmas emerge when decisions to continue filming or not have to be made, or when the researcher enacts one participant role over another when she participates in the ongoing scene.

EMCA methodologies have much potential to contribute to a deeper understanding of Autism in documenting, accurately analyzing, and clarifying the social interactions that unfold in sometimes conflictual circumstances. In my analyses, I have shown how it is not always possible to anticipate potential ethical issues as they unfold in the moment. Ethical considerations of the researcher therefore extend beyond data collection and lie precisely in how the researcher analyzes and interprets the data. A notable example of ethical considerations beyond data collection lies in another piece of data, which involves Sam having a meltdown in a corridor. Because Sam had two teachers with him, there was no need for me to provide help. However, halfway through the recording, Sam stops screaming, turns to the camera and says what might have been, “get out.” I did not realize at the time that Sam might have addressed the camera or me until I had reached the transcription phase. In this case, I eventually decided to remove the video from my corpus. Beyond data collection, the process of careful EMCA transcription is itself theoretical (Ochs, 1979) and an ethical endeavor. In the future, clear boundaries with participants about researcher-involvement should be set at the very start of data collection. Ideally, a member of the target population should participate in the design of protocols for ethical data collection practices, and perhaps even in the data collection process itself (Stack & McDonald, 2014; Tanabe, 2018).

In developing a professional vision (Goodwin, 1994; Katila & Raudaskoski, 2020) for the researcher of Autistic interaction, this paper draws attention to the ethical implications of the researcher’s moment-to-moment participation in the data collection process. Issues examined in this work should be discussed in relation to other important ethical discussions on other aspects of research with vulnerable populations, such as assuming competence to consent (Danby & Farrell, 2004) through accessible consent-taking (Cameron & Murphy, 2006; Speer & Stokoe, 2014), anonymization of data (Mondada, 2014), and avoiding ableist language in discussing vulnerable communities such as the Autism community (Bottema-Beutel et al., 2020).

Leaving oneself open to the lived experience of one’s participants means making oneself vulnerable to transformation through research practice (Liberman, 1999; Paoletti, 2014; Throop, 2018). Through the process of collecting, analyzing and interpreting data, the EMCA researcher is continually developing her learning much like other occupations: She becomes a professional within a community of practice, positioning herself and expressing different practice-linked identities in countless existential spaces (Raia, 2018). For the EMCA researcher, ethical issues at the data collection phase continue into transcription and analysis, where the data collection experience is relived and replayed through to publication and beyond. Given the involvement of the researcher in the local contextures of the data and the ethical dilemmas they may face, there may be a form of emotional labor (Hochschild, 1979) involved in video ethnography research (Shaw, 2019). I invite more discussion on how academic communities of practice can play an active role in preparing and supporting scholars wanting to partake in work of this nature through ethical dialogue, and the prioritization of researcher wellbeing.

Acknowledgements

I wish to thank all participants in this study for their time and generosity during the data collection process. I thank the guest editors and reviewers for their invaluable comments on my drafts, especially Julia Katila for her continuous support through the publication process. I thank also Federica Raia, Melissa Ho and Victor Zhuang for constructive comments that helped later revisions of this paper. Lastly, I wish to express my utmost gratitude to K.K. Luke for guidance on this paper, and for seeing me through my first rounds of data collection as a young scholar.

^{1 In this paper, I follow the preferences of the Autistic community for “identity-first language” (e.g., Autistic people) rather than “person-first language” (e.g., people with Autism) (Sinclair, 2013).↩}

^{2 Data for this study received official Human Subjects Approval Information by University of California, Berkeley, and Nanyang Technological University.↩}

^{3 Artwork by Kevin Di Pasupil↩}