The influence of subliminal crosstalk in dementia narratives

Introduction

In this article, we present conversation analysis (CA) focusing on aspects of a single extended interaction between researchers and a person attending a dementia day care centre. The interaction was recorded as part of a recently completed ethnographic study that aimed to compile a film record of people with dementia engaged in various activities. The filmed material was intended to model ways of promoting communication and reducing social exclusion for practitioner–students in Dementia Studies (Capstick, Chatwin, & Ludwin, 2015). One element of the project involved capturing informal conversations between members of the research team and people attending the day centre in order to elicit their personal narratives. However, a significant and re-occurring problem in obtaining data usable for educational purposes was the consistently high level of background noise in the environment where our interactions took place.

From a technical perspective, capturing audio and video data in crowded or noisy fieldwork environments is never easy. There can be innumerable variables, ranging from theoretical concerns about researcher gaze, down to practical issues such as where to position equipment so that recordings are as clear as possible. This is obviously a greater concern if the data are to be used – as was ours – for linguistic and micro-interactional analysis. Unlike some day centres that cater for people with dementia, this one had a lively and vibrant atmosphere – particularly in the large communal area where we conducted much of our fieldwork. Ironically, although stimulating for the people who used the day centre, this routinely chaotic environment caused us considerable difficulty when trying to capture one-to-one interactions that were clear of intrusive background noise – particularly loud voices from other nearby conversations. Initially, we considered many of our disappointingly ‘contaminated’ recordings to be unusable. However, during the process of attempting to clean up and extract serviceable audio, it became apparent that elements that we initially regarded merely as annoying extraneous noise, occasionally turned out to have a peculiar resonance with our primary data.

In this article, we therefore discuss hidden micro-interactional relationships that were uncovered by analysing the interplay between our primary data source and ‘noise’ interactions occurring nearby. Elements of this corpus have been analysed elsewhere in the context of socio-cultural resistance in people with dementia (see Capstick & Chatwin, 2016). Here, however, we focus specifically on a micro-interactional analysis that actively incorporates external noise elements, as a form of interlocutor, to reveal elements of subliminal co-construction that would not otherwise be evident. What we are basically concerned with then, from an interactional perspective, are variants of multi-party interaction; interactions involving more than two interlocutors (Glenn, 1989; Pallotti, 2001; Shriberg, Stolcke, & Baron, 2001). A novel feature of the data we present here is that one of the multi-party elements in this field of audio production is an unintended consequence of the interactional environment (i.e. the lounge area in a noisy day centre). This element will be termed ‘crosstalk’. Crosstalk is usually to be avoided and simply refers to signals from different sources that unintentionally interact with one another to cause unwanted effects (see, for example, Kim, Kim, &Yoo, 2007; Wrigley, Brown, Wan, & Renals, 2005). In this analysis, rather than trying to avoid the effects of crosstalk, we actively seek them out.

Methods

Our methodological approach is based on the conventions of CA. CA is largely concerned with the analysis of the verbal communicative practices that people routinely use when they interact with one another. Utilising video and audio recordings of naturally occurring interaction, a highly detailed method of transcription is designed to capture the minutiae of speech, including aspects of non-verbal behaviour (see Box 1). In its pure form, CA provides an analytical tool that can expose the underlying structural ‘rules’ governing the ways activities are composed and organised (Drew, Chatwin, & Collins, 2001). At a basic level, CA is particularly suited to revealing the relative temporal and sequential relationships between interlocutors as they interact with one another – a feature that is key to revealing the phenomena we describe here.

CA has been used in a wide variety of settings, with medical and health-related areas being particularly well represented. It has, for example, been applied to primary care interactions (Heath, 1995; Heritage & Stivers, 1999; Mangione-Smith, Stivers, Elliott, McDonald, & Heritage, 2003) health visiting (Heritage & Sefi, 1995); counselling, (Perakyla, 1995); mental health and specialist neurological consultations (Plug & Reuber, 2007) and complementary and alternative medicine consultations in a variety of therapeutic modalities (Chatwin, 2008, 2009; Ruusuvuori, 2005a, 2005b; Ruusuvuori & Lindfors, 2008). With specific focus on the micro-interactional characteristics of dementia-related interaction, Watson (1999) described aspects of trouble, indicating behaviour and repair strategies in conversations between people with dementia and family members. More recently, work by Blackburn and Reuber (2015), and Jones et al. (2016), has outlined the use of socio-linguistic cues to aid interactional profiling and diagnostic processes.

Our analysis is also grounded in research that has sought to understand the wider social implications of communication and interaction in dementia care settings. Research utilising broader discourse-based approaches, for example, is fairly well represented, even though the field in general currently receives significantly less attention than many other areas of health and social care. Relevant studies include early work by Cohen-Mansfield and Werner (1997) who examined, what they termed at the time, as the management of verbally disruptive behaviours in nursing home residents; Åkerström (2002) who described the way in which talk about aggressive patients was formulated in a particular way among workers in a nursing home, and Ward, Vass, Aggaewal, Garfield, and Cybyk (2008), who utilised video and ethnographic observation to capture the interactional dynamics of the care home.

At an organisational level, recent work by Kontos, Miller, Mitchell, and Cott (2011) highlighted the need to understand interrelations between care home workers and their supervisors, and several studies also employing observational and ethnographic approaches have focused on aspects of daily living in care homes. These have included levels of engagement, activity and communication (see: Bowie & Mountain, 1993; Gilloran, McGlew, McKee, Roberts, & Wight, 1993; Hallbierg & Norberg, 1990; Nolan, Keady, & Grant, 1995; Schreiner, Yamamoto, & Shiotani, 2005). Studies have also addressed the dynamics developing between carers and care home residents, and have shown a disparity between the socially oriented interests of residents and the task-based agenda of the care staff (Bender & Cheston, 1997; Iwasiw & Olson, 1995). Similarly, the content and context of caring encounters have been explored in studies by those such as Coupland, Coupland, Giles, and Kenwood (1998) and Lancy (1985).

The data

In the session that forms the basis of this article, we examine extracts from a fieldwork recording of an interaction between a member of the day centre (Don) and two researchers. When the interaction took place, the day centre lounge was particularly noisy, making clear sound recording extremely difficult. Multiple interactions were occurring between staff and residents in the room and, of most relevance to this article, among a group of three care staff in an adjoining kitchen area. In each of our extracts, we have used a simplified version of the notation system developed for CA (Sacks & Jefferson, 1995). An important feature of this system is that it allows the various simultaneous strands of conversation and ambient noise to be presented in a way that reflects their relative temporal placement; the square brackets ([) in the text indicate points at which different layers of talk and/or noise overlap (a complete list of the symbols used is given in Box 1). In addition to Don, the participants are the three members of day centre staff (CS1, CS2 and CS3) and one researcher (Res). A second member of the research team was present, filming the encounter, but was not directly involved in the interaction.

The work presented here formed part of a broader study that was funded by NIHR-SSCR and received ethics approval No. 12-IEC08-0034 in September 2012 from the NHS Social Care Research Ethics Committee. Don was the first of 11 people who we ultimately worked with using principles of Participatory Video (Milne, Mitchell, & De Lange, 2012) to create digital stories about things that were of importance and value to them. Information about the broader study can be found in Capstick et al. (2015) and on the NIHR-SSCR website http://www.sscr.nihr.ac.uk/PDF/Insights/IN23.pdf.

In the analysis that follows, we focus on three discrete sequences of interaction. However, for clarity, we provide three data extracts for each sequence: the first being ‘clean’ in that it contains only details pertaining to our primary data – the interaction between Don and the researcher. The second includes only the ‘noise’ data (primarily the external talk generated by the carers in the nearby kitchen area). The third version of each sequence includes both interactions superimposed upon one another and establishes the temporal relationships between them.

Don's narrative

Like many people with dementia Don has difficulties with short-term recall, and his speech was sometimes repetitive; his long-term memory is good and is keen to talk about his wartime experiences, even though not all of his memories of that time in his life are happy ones. Like the three men with dementia whose wartime narratives are discussed by Capstick and Clegg (2013), Don talk in a frank and revealing way about his time on fuel supply with the RAF, with an emphasis on the subjective domain and personal lived experience. Don volunteered to join the RAF at the age of 17, and worked in a variety of war zones, including the Middle East, throughout the duration of the Second World War. In our first extract, he is describing his demobilisation:

Extract 1.1 (Don’s narrative)

If we switch to the parallel extract involving only the care staff:

Extract 1.2 (Care staff narrative)

In extract 4, Don resumes his wartime narrative. In the section transcribed here, he is describing how the RAF learned to intercept and shoot down flying bombs that were sent over from Germany – ‘kites' being slang for British fighters.

Extract 2.1 (Don's narrative)

Our final extract (below) offers another example of overt thematic resonance:

Extract 3.1 (Don's narrative)

Discussion and conclusion

In this article, we have focused on a single, discrete, interlocutory phenomenon occurring in a relatively niche field of communication research. We have argued elsewhere (Capstick & Chatwin, 2016) that neither the dominant biomedical discourse on dementia, nor its psycho-social alternative offer a sufficient account of the complexity of communication by, and between, people with dementia and those who care for them in formal group care environments. In particular, such models tell us little about the resources drawn on by people with dementia, as agentic social actors, to make sense of the situations in which they find themselves, or to compensate for the communicative difficulties that usually accompany this condition. We would suggest that the material we have outlined here opens up further questions about how external influences might also impact on the content and structure of the narratives produced, not only by people with dementia, but also by practitioners working in such environments.

At a pragmatic level, we would argue that incorporating the kind of peripheral crosstalk we outline here is directly relevant to the micro-analysis of interactions in dementia settings, or indeed, any similar therapeutic arena where external cognitive influences (i.e. noise) might have an impact on vocal and narrative production. What we would emphasise is that our data illustrate that this impact may not simply be disruptive. In the case of Don, for example, the crosstalk between his narrative and that of the care staff serves to generate a layer of interaction (and meaning) which, whilst it may be largely subliminal, can still be directly observed. Whilst our initial interpretation was that Don struggled to be heard and had to compete with the staff for attention, for example, an alternative possibility now presents itself. That is, that Don’s story had already permeated and interpenetrated talk within the care environment to such an extent that it was shaping the staff’s communication with each other. There is a strong possibility, also, that the care staff – being used to Tom’s story, and anticipating what he would be telling the researchers, perhaps also feeling excluded from the film-making process – are establishing their own right to possess a narrative and to have their voices heard. The study of ‘crosstalk’ can therefore help to provide a deeper understanding of how micro-interactions may be co-constructed in care environments. We would suggest that a more detailed study of narrative production in different dementia environments, and with groups of people occupying different positions in their internal hierarchies, could provide further insight into such phenomena, and also into organisational cultures.