Social Interaction

1. Introduction

Some people need support from another person to communicate with others due to their young age or physical or cognitive challenges. To some extent, one might claim that some people do not have the same communicative resources, or communicate differently, than the majority of the population. Assistance may be needed when they go to the doctors, or in school, or in other official places. In such situations, they can bring a companion; someone who will help them understand what is going on and help them express themselves about their needs and wishes. But as yet, the research on what those companions do is limited and fragmented. This is a particularly important issue for interactional research as it may help uncover strategies and practices that support people who require assistance of one kind or another to communicate and potentially ameliorate inequities and/or inequalities that they may face.

In order to shed further light on this important issue, this special issue brings together academic researchers from Europe and China/USA who have specialized in studying interaction involving people who need assistance and support to communicate in some settings and their companions. The articles collected here provide insight on, for example, companions' practices that help people who need assistance and support reach their goals and on how practices compare across different groups.

All the researchers in this special issue are particularly interested in the highly detailed aspects of interaction in the encounters and contexts under examination, and only work with videotaped naturally occurring communication in everyday and routine contexts. They concentrate on interaction involving companions to people who need assistance and support due to the following causes: dementia, visual impairment, congenital dysarthria, acquired aphasia, intellectual disability, epilepsy and multiple sclerosis. Others are receiving palliative care, or traditional Chinese medicine. In all articles, researchers use the advanced methods of multimodal conversation and interaction analysis established in Ethnomethodological Conversation Analysis (EMCA) for the analysis of videotaped interaction, supported by ethnographic notes and observations from fieldwork about the specific contexts in which the video data were collected. EMCA allows for very careful analysis of both spoken and bodily communication to identify the sequences of actions through which participants accomplish their social actions. The authors have different disciplinary backgrounds (e.g. anthropology, ethnography, linguistics, and speech and language therapy), are active in different European, Asian and American countries, and have a common research objective to investigate how communicative practices are supported by companions. Bringing together interactional research on companions that spans various geographical and cultural contexts has not been done before.

Note that the special issue's focus is not on the work of interpreters between different languages, on which there is already extensive research internationally. A companion almost always shares language with the person who has communication support needs, but for various reasons, e.g. due to developmental or acquired diseases, that person needs support in social interaction with others.

Companions and related participation roles have been investigated within a number of different contexts and with different analytical approaches. The bulk of research on the role of companions in interaction has been conducted in medical settings (Robson et al. 2016, Pino et al. 2021, Stommel & Stommel 2021, Yan & Yang 2024). This is likely to be a reflection of the fact that communication in these settings can contain complex and unfamiliar information, that poor communication can bring about serious threats to health, and that companions may be needed to help the patient put medical recommendations into action. At the time of Laidsaar-Powell et al.'s (2013) systematic review of triadic physician–patient–companion communication, most of the research available used quantitative methods. Studies provided data on how frequently companions were present in different settings and with different patient groups. Reviewed studies and similar ones published subsequently have provided quantitative data on how much time companions might claim for their own questions and responses and classified companion's contributions according to researcher defined categories. There has also been interest in abstracting from observational data a typology of different 'roles' for companions and also drawing up guidance for clinicians about how to facilitate helpful rather than unhelpful contributions from companions.

2. The Articles in the Special Issue

There are seven articles in the special issue. Four articles analyse companions' roles in healthcare, two in elderly care, and one in school. All seven articles are based on video recordings of naturally occurring interaction and one study (Cromnow et al., 2024/this issue) also involves audio recordings of phone calls. Six articles draw on data from Western countries, mainly from within the EU (Denmark, Finland, and Sweden), but also from the UK and Canada. Only one article draws on data from an Eastern country (Wei, 2024/this issue), with a pioneering study of triadic consultations in Traditional Chinese Medicine. One article explores companions in an institutional setting outside health or elderly care, namely classrooms in Danish schools (Toft & Due, 2024/this issue), providing new insights into how personal assistants to students with visual impairment navigate within the socio-material environment of the classroom.

One general theme of the articles concerns companions' role for enabling or disabling the accompanied person's agency in talk and decisions about themselves, e.g., by intervening on their behalf or taking their side in relation to an institutional representative (e.g., a doctor). All articles contribute to increasing the knowledge on how companions' actions may differ between different institutional contexts, different cultural contexts, and the nature of the disability or condition of the accompanied person. Another general theme of the articles concerns companion's roles for including or excluding the accompanied person from participating in an institutional activity (dancing, problem solving with a Virtual Assistant, end of life talk, and learning in school, and a public communication service). The articles provide a more nuanced and detailed perspective on the role of companions in comparison with previous research that uses less context sensitive methodologies.

3. Presentations of the Articles

Ruusuvuori, Antaki and Chinn (2024/this issue) compared whether companions of patients with intellectual disabilities (in UK annual health checks) intervene differently from those accompanying patients with multiple sclerosis or epilepsy (in Finnish neurological consultations). Companions can understand both types of patients as having difficulties with either accessing experiences or memories or putting them into words, as a part of their condition or diagnosis, and the article focuses on whether companions orient to this difference when they choose to intervene on behalf of the patient. The analysis showed that companions to a patient with intellectual disabilities tended to strongly respect the patient's agency and entitlement to speak to their own experiences in three different ways: a) by allowing the patient time to volunteer the answer to the question themselves, b) by glossing inadequate answers as being a temporary failure to remember, and c) by constructing a no-problem answer. However, companions to patients who were examined for or had a diagnosis of epilepsy or multiple sclerosis, were found to take a more varied, proactive and interventionist approach. When showing entitlement to the knowledge concerning the patient's situation, they rather used plain assertions than the sensitive turn designs mentioned above, or self-selected when the patient had a primary right to answer, although also showing some orientation to the patient's primary role as the respondent by gazing at the patient or by negotiating with them on the correct answer. Consequently, when the patients have a lifelong intellectual disability, companions may choose to act more sensitively in how they speak on their behalf, hence potentially orienting to a larger social vulnerability compared to an illness that has developed or has been diagnosed later in patients' life.

Pino, Land and Hoey (2024/this issue) analysed conversations at a large UK hospice involving end of life (EOL) patients and explored the implications of companions' actions for explicit and focused talk about the progression of the disease and the patient's mortality. The study departures from previous claims that companions can merely avoid or refuse to engage with EOL issues, and proposes instead a more rounded and nuanced understanding of companion's role in enabling or hampering discussions of mortality. The analysis looked closely at local contexts where a patient alluded to disease progression or EOL, and the following response by the person (e.g., a spouse) accompanying the patient to the consultation with the psychotherapist. The analyses demonstrated that although companions in one instance did not elaborate on a patient's allusion EOL, they usually did not avoid EOL issues, but were rather pursuing other interactional projects, and as a consequence did not actively promote explicit and focused discussion about EOL matters. Those interactional projects were in line with the sequence of action underway and also relevant within a wider activity in progress. Instead of a contextually relevant elaboration on a patient's allusion to EOL, companions were found to introduce a related but distinct matter or introducing a positive consideration, which were relevant responses within the local context of interaction. These findings support the consideration that the presence of companions can add complexities to healthcare interactions, since their actions contribute to shape conversational environments that were not favourable to immediately elaborating on patients' allusions. But these findings also show that the responsibility for achieving a sustained focus on EOL matters in the interaction depends on all participants' actions, including the health care practitioner. The authors suggest that participants should handle talk about disease progression and EOL as delicate matters and therefore not raise them explicitly or unilaterally, but rather develop them interactionally in a stepwise manner.

Wei (2024/this issue) analysed the role of companions in a health care context that differs from the previous studies, namely Traditional Chinese Medicine (TCM) consultations, focusing specifically on local contexts of conversation when patients resist the doctors' treatment recommendations or lifestyle advice. In China, there is a cultural norm to be accompanied by someone when going to the doctors, and TCM encounters are usually quite informal and open for bystanders to join in. Most TCM consultations are routine, so patients do not always bring specific medical concerns, which also was the case in the current study. TCM doctors usually alternate between recommending medical treatment and giving lifestyle advice, because of the importance of enforcing certain lifestyle rules to achieve effective TCM treatments. It is common for patients to resist the doctor's advice. Unlike many Western medical contexts, TCM consultations may therefore provide a more interpersonal approach where companions are provided opportunities to engage actively in the interaction, but research on how this is done and also on the role of companions is scarce. Therefore, this study aimed to provide a more nuanced understanding of companion roles in TCM encounters. The analysis focused specifically on instances where the companion was invited by doctors after a patient's display of resistance. Results showed that efforts from the doctor to enlist the companion in facilitating the medical or lifestyle agenda was not always effective. In a majority of the analysed instances, the companions chose to side with the patients' resistance and declined to offer the requested help. The companion's own involvement in implementing the patient's treatment or change of lifestyle appeared to influence the companion's choice. When asked to be a more active agent of the doctor's agenda at home, and hence attributed more responsibility for success of the treatment, the companion was less likely to accept the doctor's request for help. The study demonstrated that companions seem to navigate two distinct sets of concerns: relational and medical. Most companions chose to side with the patient rather than facilitating the medical agenda, which in these data constituted being a good companion.

Cromnow, Samuelsson, Danielsson, and Plejert (2024/this issue) investigated when and how professional companions intervened in conversations when clients with communication disorders (congenital dysarthria or acquired aphasia) were using a Swedish communication facilitation service to participate in phone calls. The client and the professional companion were physically co-present and could see each other, while the conversation partner was audibly present on the phone via a loudspeaker. The analysis demonstrated that the clients commonly used eye gaze to invite the companion to intervene. When accompanying the client with congenital dysarthria, the companion routinely added prosodic or syntactic completions or addressed uncertainties regarding the client's turn-completion. Interventions were found to be missing, however, after sequences with overlapping talk. When accompanying the client with acquired aphasia, the companion intervened when there was a potential lack of response on behalf of the client or the conversation partner, when the client used resources visually unavailable to the conversation partner (e.g., writing or embodied moves), or in some unclear cases where the companion probably intervened on the basis of previous knowledge about the client. Interventions were found to be missing, however, in relation to more complex problematic contexts such as when the client's answer disaligned with the conversation partner's question. These results are relevant for the design of communication facilitator training and the development of the communication facilitation service. When facilitating talk involving clients with dysarthria, companions can focus on the details of speech and the organisation of turn taking, while facing a much more varied range of challenges when facilitating clients with aphasia: involving the progression of the talk, the turn taking, the use of embodied communicative resources, and the context sensitive understandability of all participants. In phone calls involving clients with aphasia, the companion takes on a more comprehensive responsibility for the conversation, a difference the authors argue is not a reflection of the guidelines for communication facilitators professional role, but of the local affordances of accomplishing meaningful conversations in interaction.

Kosurko and Webb (2024/this issue) explored the role of the companion in supporting the engagement of people living with dementia (PLDW) in dance activities that take place in long-term residential care settings in Canada, Finland, and the UK. The analysis demonstrates that the role of the companion and the nature of companionship in these data emerge from the social action of engaging PLWD in dancing, and that a wide range of participants are locally constructed as companions - family, volunteers, staff, and also PLWD - rather than just facilitating staff. Touch was found to be a prominent resource to facilitate engagement in dancing. Touch was used by staff facilitators to engage joint attention in moments when residents display disengagement with the group dance activity. Facilitating staff members also used touch to establish connectedness between two residents, and also to facilitate understanding between two residents (a couple) on how to engage in a dance. Facilitating staff members also used shepherding touch on the dance floor when brokering shared understanding between another staff member and a resident PLDW. The analysis also showed sequences where the participants were taking turns at being the companion, when three different participants responded to a PLWD's disengagement from the ongoing dance activity, first two staff members and finally another PLWD. In all these cases, the action sequence began with a participant's disengagement and then different participants used touch in the subsequent companionable action towards a willing (re)engagement. One of the main contributions of the study is that the role of companion is found to emerge locally through the participants' engagement in embodied social action, and that it can be shared by multiple participants, including the PLWD themselves.

Hall, Albert, and Peel (2024/this issue) investigated dyadic everyday interaction between an elderly person with dementia (PWD) and a virtual assistant (VA, named Alexa) in a domestic setting in order to outline what 'virtual companionship' might involve as an interactional achievement. The PWD lived in an apartment complex in the UK specialised in dementia care and was video recorded when interacting with Alexa. The analysis focused on the actions achieved by the PWD's verbal utterances within sensor range of Alexa, and how change of footings may implicitly index the user's orientation to Alexa as a co-participant and possible companion in the sense of a friend. The analysis demonstrates that the PWD discriminated between talk that made sure to involve and establish Alexa as a recipient and talk that avoided involving Alexa, hence demonstrating an understanding of and orientation to Alexa as a possible participant (or companion) in terms of a problem-solving agent. Talk that involved Alexa were initiated with a four-turn summons sequence, involving Alexa's 'wake word'. Subsequent commands were made in the shape of directives, while more complex requests could use conventional markers of contingency, such as modal verbs and 'please'. In activities, however, where Alexa was of no use to her, the PWD engaged in 'talking to herself', e.g., by way of doing 'online commentary', making sure not to use the wake word. The PWD could also engage in talk 'about' Alexa, e.g., when complaining about the VA for not being helpful, hence treating the VA as a blameable other, using anthropomorphic references and gendered pronouns. In these sequences, the PWD also made sure not to activate Alexa by avoiding the wake word. Despite that these orientations to Alexa are not analysed as treating the VA as a human or as a person, they are still argued to emulate the interactional practices of 'companionship' in terms of 'friendship'. Similar to interaction between humans, the footings towards Alexa shifted dynamically between constructing the VA as more or less companion-like and they were shown to be locally sensitive to different social actions.

Toft and Due (2024/this issue) analysed the role of companions in a different institutional context from the other studies in the special issue, namely classroom interaction in a Danish regular chemistry class that involved a visually impaired student (VIS). The companions were personal learning support assistants that facilitated VIS's interactions with teachers and other students, who were often fellow students or teachers without formal training in supporting VISs. The analysis focused on how a VIS, an assistant, and a teacher worked together toward solving a chemistry assignment, where the assistant helped the VIS with reading and writing in tables with his electronic braille note-taker. Results showed that in these problem-solving activities, the assistant mainly enacted the role of being a fellow "learner" or "student" in dyadic interaction with the teacher, a role that temporarily sidelined the VIS from interacting with the teacher. These roles demonstrably differed from the institutionally situated identity as a VIS's assistant, which Toft and Due discuss as potentially problematic from a learning point of view. When the companion acted as a "student", the companion was also positioned as the primary recipient of the teacher's feedback, and the VIS was positioned as a bystander. This rather specific participation framework was orchestrated by specific socio-material circumstances that presented particular possibilities and constraints. The visual technology used in the chemistry class, such as the computer screen that displayed the tables, was unavailable for the VIS and this became consequential for the accomplishment of the learning activity, and worked to structure the unfolding interaction. Toft and Due argue that learning materials should be made available for visually impaired students so that they can access them more independently, and also that problem solving interaction should be organised in ways that allow VIS to engage in the problem-solving process, rather than being introduced to products of it after the fact.

4. Discussion

The researchers in the studies presented here explore different ways that companion roles and 'companionship' are constituted. Companions are mostly non-professional or professional persons. Non-professional companions comprise relatives or friends who accompany a person in need of help to a doctor or therapist, but also classmates that are temporarily assigned as personal learning support assistants, and they all share the background of having no education or training for the role. Professional companions comprise service providers in the Swedish communication service and facilitating staff at a UK long-term residential care home for PLWD. The Virtual Assistant Alexa deviates from this pattern, being a non-human device that the PLWD occasionally treats interactionally as a possible 'participant' or 'companion' in the sense of an everyday friend or a practical resource. In most of the articles, the companion (or companionship) is associated with a specific participant with a more or less fixed supportive identity. In Kosurko and Webb's article, however, the companion's role is much more dynamic and shifting between different participants. Other PLWD occasionally also take on the role as companion to a fellow resident, mainly accomplished with embodied resources due to the specific activity context of dancing. This also occur in the article by Hall et al, where the PLWD dynamically shifts between positioning the Virtual Assistant Alexa as a friend and problem solver, as someone to complain about, or as a non-recipient.

The accompanied participants in the studies need help and support for different reasons. In two of the studies, companions are found to orient differently to the supported person, depending on the nature of this reason. Ruusuvuori et al. show that companions may choose to intervene more sensitively in how they speak on behalf of patients with a lifelong intellectual disability, as compared to patients with multiple sclerosis or epilepsy that develops later in life. Also, Cromnow et al. show that the companion takes on a more comprehensive responsibility for the conversation in phone calls involving clients with acquired aphasia, as compared to in calls with clients with congenital dysarthria, in orientation to the specific affordances of each persons' diagnosis.

All but two studies analyse the role of companions within a triadic participation framework. In the healthcare studies, data are drawn from different types of consultation talk between a patient and a doctor or therapist, and the companion mediates or intervenes between them. In the school study, the interaction setting is multiparty, but the local participation framework is still analysed as a triangulation involving the student, the teacher and the mediating personal assistant. In the two studies of elderly care, however, the participation frameworks are much less clearly triadic, or not triadic at all. In Kosurko and Webb, the dance activity involves many different PLWD and facilitating staff members, and the frameworks are more shifting and dynamic. In Hall et al., the framework is analysed as dyadic between one PLWD and a Virtual Assistant and therefore does not involve a human conversational partner. Hall et al. therefore looks more into the nature of 'companionship' as an interactional achievement than into the companion as a mediator, broker or facilitator between two other persons. Both articles about elderly care are also drawn from more domestic or homelike settings, with recordings in residential homes, without the consultative dimension that dominates the healthcare studies. When considering all seven studies, companionship seems to appear more clearly in triadic participation frameworks, and less clear in domestic multiparty frameworks when the person with communication vulnerabilities is not a typical service recipient.

In three of the studies, the companion's actions are found to be beneficial for the client, patient, or resident, e.g., increasing their possibilities to be understood, to resist unwanted recommendations by the doctor, or to get (re)involved in a social activity (dancing). In three of the studies, however, the companions' actions are found to be more complex. Pino et al. show that actions by companions to EOL patients do not contribute to establish explicit talk about EOL matters, but also (which is the key point) that they do not actively evade EOL talk. Cromnow et al. show that while companions mostly facilitate the phone call (e.g., with repairs on a detailed level), the companion does not intervene in relation to some more complex interactional phenomena (e.g. overlapping talk or mismatches between initiatives and their response in terms of action type). And finally, Toft and Due show that the student with visual impairment, although receiving support regarding the outcome of problem solving, is excluded from (learning from) the actual problem-solving process that is achieved within dyadic talk between the assistant and the teacher. All authors discuss how the results of their studies may inform the institutional conversational practices, the training of professional companions, or recommendations and advice to non-professional companions.

An observation based on the articles in this special issue and also other EMCA studies of companions in interaction, is the possibility of a gradient of legitimation or entitlement of the companion to intervene in the interaction. On the one end of that gradient scale, there are factors that are more fixed or predetermined and related to social roles and the socio-material circumstances. Such forms of companionship are more institutionally determined and scripted, but are of course still established and negotiated in orientation to the contingencies of interaction (see for example Cromnow et al. and Toft & Due). On the other end of that gradient scale, there are more context sensitive and dynamic factors that do not relate to social roles. Such forms of companionship are mainly unscripted and negotiated within the local interaction (see for example Kosurko & Webb, Ruusuvuori et al., and Pino et al.). On the basis of the articles in the special issue, it can be argued that a companions' entitlement is sensitive to the capabilities of the accompanied person. A companion who intervenes in talk involving less capable participants with often lifelong conditions or disabilities does that with less entitlement and greater sensibility, while companions intervening in talk involving a participant that are more fully capable with conditions or disabilities often acquired later in life intervenes with greater entitlement and less sensibility.

This special issue has specialized in using multimodal EMCA methods and principles when collecting and analysing data. So what does EMCA add to the existing wider field of research on companions? As the articles in this special issue demonstrate, companionship is carefully calibrated within the contingencies of embodied interaction and spatio-social environments and multimodal EMCA can get inside the interactional practices that achieve that. For example, the identity work of participants is often ideology or value based, requiring them to do or refrain from doing certain things. CA provides tools for examining the roles participants are actually orienting to in the emerging interaction (which may depart from expected behaviors) and how roles are negotiated and emerge from local contingencies, in ways that cannot be demonstrated with quantitative approaches or other qualitative approaches. The detailed findings of EMCA research can not only help companions, and the persons they support, to work more efficiently with professionals such as doctors, welfare officials, and teachers, but can also help professionals to use the competencies and knowledge of companions in their work.

5. Future Directions

Although featuring an article on companions in the classroom, this special issue illustrates a lack of knowledge within the wider field of companion research on the role of companions in public service encounters outside healthcare and elderly care. Education, social work, the judiciary (police and courts), the correctional service, and the employment agency are just a few examples of institutional contexts featuring triadic encounters involving an accompanying relative or friend, which are yet unexplored but appear as perspicuous settings (Garfinkel, 2002, p. 181) for identifying companions' roles in institutional settings with other agendas and goals than in health and elderly care.

Another interesting but yet fairly unexplored context in research on companions are ordinary conversation and domestic settings. Kosurko and Webb as well as Hall et al. investigated domestic and everyday activities within an institutional dementia care setting and found that the role of 'companion' was more dynamic and shifting there compared to consultations in healthcare. It would be interesting to develop this line of research to also look at 'companions' or 'companionship' in ordinary conversation involving participants in need of communication support. Are triadic conversational frameworks involving a 'companion' endemic to institutional settings or do they also occur in ordinary conversation, and in that case, what type of actions are companions contributing with there and how are they achieved?

In this special issue, Wei analysed companions in a non-Western healthcare setting, and described some cultural differences with regard to the wider role of Chinese Traditional Medicine in the Chinese society, but also with regard to the companions' entitlement to intervene in consultations with doctors, as well as entitlement to take oppositional stances towards the doctor's recommendations. There are yet very few similar studies of companions' roles in non-Western institutional (or ordinary) settings, and such research could contribute to developing institutional practices in those settings, but also provide comparative grounds for identifying aspects of Western institutional practices that may become more visible in comparative work.

One of the studies in the special issue (Hall, Albert, and Peel) explored a single PLWD's dyadic orientation to a voice activated Artificial Intelligence (AI) model as a 'companion' in everyday life. Current AI-based tools do not seem to be developed for supporting conversation, but have been used by some speech and language therapists to help people with Motor Neurone Disease to slow down the negative impact of the increasing dysarthria on the patients' possibilities to participate in social interaction (Cave 2024). The analysis showed, however, that the tool may need further development before working in a conversational environment (Cave, 2024), and hence also as a 'companion'. The technological advancements in the AI field are moving fast, however, as well as research on conversational AI (e.g., Stokoe et al. 2024), and there may come new or more developed types of AI-based tools that could be used as 'companions' in the near future, within triadic or other types of participation frameworks.

We also suggest widen the scope of types of 'companions' to explore with EMCA methodologies. One yet unexplored 'companion' is 'voice hallucinations' (auditory verbal hallucinations) experienced by people living with Psychosis, and the way they are treated and oriented to in conversational talk. For a Conversation Analysis study using conversational data obtained within an intervention study, see Deamer et al. (2020). Voice hallucinations are obviously only possible to hear by one of the participants, but they can still be oriented to as 'conversational partners' and hence play a role in the talk in similar ways as a third party 'companion'. Qualitative research on voice hallucinations has been dominated by interview studies and intervention studies, and using EMCA on naturalistic data could further the knowledge on how voice hallucinations impact the everyday life quality of people living with Psychosis, but also inform professional practice in different institutional settings.

Another type of 'companion' that would be interesting to explore using EMCA methodologies are pets and animals. Pets may be oriented to as companions in a home setting when a patient with communicative vulnerabilities is visited by a healthcare professional, and animals may be oriented to as companions in institutional contexts. For example, interview and observation studies have shown that horses (Malcolm et al., 2018) and guinea pigs (O'Haire et al., 2014) have a positive effect on the social abilities of children with autism. These studies are not using EMCA methodologies, however, so the details of the way the animals support the children's social abilities are yet to be explored.

Investigating how companions contribute to and influence users' opportunities to participate in meetings with professionals in institutional conversations is of great importance to increase participation for people who communicate differently, or who face challenges with their communication. The knowledge that is produced in the articles of this special issue can also support companions by drawing attention to and strengthening them in their everyday and recurring work, regardless of whether they are carers or professionals, something that has great indirect significance for the quality of life of people with communicative challenges.