Concept-Formation, Complexity and Social Domains: Investigating Emotion(S) in a Prison Setting

Introduction

1.1 In relation to its more general parameters the paper discusses two main methodological issues, against a background of qualitative data on emotion regulation in prisons. The first concerns the usefulness of the domains model of social reality for sorting and organising qualitative data. In this regard the model facilitates an examination of segments of social reality, which are often overlooked. Thus, it helpfully yields a contribution to the agency-structure problem in social analysis. Second, the paper focuses on theoretical and methodological issues around theory generation in relation to the collection and analysis of qualitative data. In particular the paper highlights the way in which the use of a conceptual scaffold as an initial, but flexible orienting device, encourages theoretical elaboration and conceptual development. Using a study of in-cell television the concept of 'techno-therapeutic control' was generated. This was achieved by linking different social domains, and offering an explanatory tool to account for prisoners' social behavior and particularly, their expression of emotion. This strongly anchored empirical concept emerged from the research data, but in a manner quite unlike that of the grounded theory approach to theory generation.

1.2 The following blends a discussion of findings with important methodological issues. It first establishes how the social domains model facilitates research into emotions, power and control in the prison setting by offering a multi-dimensional model of social reality- as entangled social domains articulated through social relations of power. The model enables researchers to examine different modes of emotional expression, from the point of view of the intersecting domains of psychobiography, situated activity, social setting (prison), and the wider context of the criminal justice system. Second, it explores how qualitative research on power and control can be accessed and how this helps to throw light on links between emotions and social reality.

1.3 These twin emphases prepare the ground for the heart of the paper, which focuses squarely on the processes of coding and analysis of data used in the study of in-cell television. The creation of core concepts (rather than themes) led to theory development more generally. Using the adjoining methodology of adaptive theory, it permitted inputs from a diverse range of general theories and research findings across different areas (prison, emotions and mass media).

Emotions, Power & Control in Prison Life

2.1 The domains model assists in constructing narratives about prisons, emotions and mass media which are crucial in developing concepts from the primary data. As Houston and Mullan-Jensen (2011: 270) point out, the model offers researchers 'smaller domains or units of analysis' rather than abstract general concepts such as structure and agency. This helps to orientate research data, and at the same time, encourages 'psychological depth and sociological width' (Houston 2010). In the context of prison life, the combined influences of the four principal domains (psychobiography, situated activity, social settings and contextual resources), serve to focus the analysis of qualitative data onto different aspects of the emotion/control coupling -and their effects on prisoner's experiences with television in their cells.

Psychobiography

2.2 This domain captures the effects of 'critical experiences' throughout people's lives (2006:274) and provides an insight into how social agents deal with and manage their everyday life. Layder stresses that because social agents 'exist both 'inside' and 'outside' society' (Layder 2006:274), emotions can be 'disruptive' to social relations and action and can thus powerfully impact on the more remote aspects of social life. While different individuals have different capacities to control their social situations (2004:11) emotions are not always under control. Individuals also have different levels of commitment to social situations and their broader social enterprises. At the heart of this is a person's 'ontological security'^[2].

2.3 Such factors influence a person's capacity for personal control -to alter and adapt to their circumstances. This is evident for example in prison settings, where some prisoners are much more able to cope with incarceration than others (Liebling 1999; Cohen and Taylor 1972). The techniques and adjustments made by some prisoners, for example those sentenced to life in prison, are varied and are not fixed, as they can evolve and change over time (Jones and Schmid 2000; Sapsford 1978). Imprisonment puts pressure on one's basic security and this is reflected in how prisoners deal with boredom and frustration (Knight 2015), and manage, particularly long sentence with large amounts of unstructured time (Cohen and Taylor 1972). Mass media, including television offer ways in which prisoners remain connected to their sense of self and carve out autonomous identities (Jekwes 2002). Silverstone (1996) has argued that television provides important routine orientated material and offers audiences comfort in the predictability of television broadcasts. Television is therefore reliable, trustworthy and punctual, and as a consequence, contributes to achieving an individual's 'ontological security' (Silverstone 1996). In sum, psychobiography underlines the importance of personal efficacy and actual and felt ability to control one's immediate circumstances.

Situated Activity

2.4 This domain highlights the direct and interactive relationships individuals have with each other. It is the 'delivery system of social behaviour' (Layder 2004:21) where arrivals and departures are marked out and punctuate encounters in social life (ibid:44). It is also a 'gathering point' for power which becomes drawn into the activity between and among social agents (ibid:50). Encounters are not emotionless and can potentially unsettle and thwart basic security. For some individuals, 'avoidance' of situated activity or withdrawal from social encounters can resolve these tensions. Several sociological studies of prisoners show that some withdraw from prison culture in an attempt to avoid psychological and physical harm (Jewkes 2002; Sapsford 1978; Cohen and Taylor 1972; Irwin and Cressey 1962). Yet this is not altogether feasible or completely beneficial.

2.5 Layder describes situated activity as having a 'compelling enticement' (2006:279) since interaction provides opportunities to gain reward through inclusion, approval and identification as well as 'validation and support' of emotional needs (2004:25). In the context of the prison this withdrawal can be viewed negatively especially by those in control of the prison. Knight (2014) highlights how these withdrawal techniques contribute to prisoners' emotional management in which mediated materials provide valuable resources to counter adverse impacts of withdrawal, such as loneliness and isolation. This domain therefore acts as a filter between the psychobiographical and the more impersonal or structural domains. Knight's (2014) exploration of cell sharing in prison demonstrates how an individual's input into situated activity can have important consequences on the ways in which prisoners live together and interact (see Lull (1990) on 'family viewing'). In sum, the domain of situated activity points to the importance of situational and interpersonal aspects of emotion, power and control.

Social Settings

2.6 Settings are the immediate environments of situated activities and take the form of 'local aggregations of reproduced social relations, positions and practices' (2006:280). Formal settings like schools or prisons, are highly structured in terms of routines and hierarchies; features characteristic of 'total institutions' (Goffman 1991: 15). Interaction is largely defined by formal status and position; for example a prisoner will wear a different set of clothes from a prison officer, distinguishable by a 'uniform'. By contrast in informal settings – such as friendship groups or family networks- roles are less distinct and observable and encouragement to adopt certain roles is not formally orchestrated.

2.7 Intricate forms of control and power can be traced across the formal and informal settings of the prison (see Goffman (1991) in his study of total institutions). Prisoner studies outline the ways in which social groups define and shape their experiences of prison organisation. For example, Crewe's work on the prisoner society (2009) and drug dealing cultures in prison (2006) offers insightful perspectives on forms of compliance and resistance. In addition, the norms and values of the social setting are reflected in prison policy. Liebling et al's (1999) analysis of the incentive and privilege scheme and also safer custody policies (2002) reveals much about the ways in which prison policy is both experienced and enacted by staff. Television research has also demonstrated how gender roles and family dynamics impact on the orchestration of domestic routines (Gray 1992, Lull 1990, Morley 1986). In sum, as a domain the social setting points to the significance of prison as a site of organizational power and control and shapes the subjective worlds (emotion) and objective (constraints) dimensions of both inmates and staff. With television in situ, these features offer complexity to contemporary understandings of the prison.

Contextual Resources

2.8 This domain highlights the historical accumulation of economic, political and cultural resources and how their distribution influences (but does not fully determine) their availability and accessibility to groups based on gender, age, class, ethnicity and status. The domain thus shapes the ways in which situated activity and the inner-self is experienced, and managed. Contextual resources refer to the range of cultural resources accumulated over time, such as values, mores, knowledge, artefacts and representations in the media. We draw on these to inform us about ourselves and others. Like settings, contextual resources are not, entirely objective ; they are 'brought to life' and mobilized through activity. For example, historically the aims of imprisonment such as punishment and rehabilitation, have seemingly attained a relatively independent status as policy objectives. However the living reality of imprisonment brings these aims into the heart of everyday life of prison. Media, including television, are an important source of contextual resources, and the ways in which social agents access and subsequently interpret and employ them are an important feature of audience behaviour. The degree to which audiences can access these resources influences other domains. Examples include the informal resources prisoners draw upon to do their prison time and the power of prisoner folklore (well documented by Cohen and Taylor 1972). The types or modes of doing prison time are also shaped by such cultural resources as masculinity^[3] (Jewkes 2002) and adaptations to prison life (Jones and Schmid 2000) and modes of coping (Liebling 1999). We go on now to show how using these domains can lead to concept formation from the analysis of qualitative data.

Techno-therapeutic Control: A Bridging & Synthesizing Concept

3.1 In relation to the environment of the prison, each social domain pinpoints an area in which a particular aspect of prisoners' emotions and felt sense of control, are significantly impacted. Of course, from the point of view of the prisoners' lived experiences, the sometimes radically different influences of the domains blend into each other to create a composite effect in daily life. In this sense the mutual effects of agency and structure in the prison setting are synthetically bridged in the intersection of domain influences. The use of the domains model as a point of departure and as a means of sorting and organising data provided indicative guidelines for the development of bridging and synthesizing 'techno-therapeutic control'.

3.2 Thus techno-therapeutic control became focally important in bringing together the elements of this study – in-cell television use, prisoners' emotions, power, control, and the routine of daily prison life. The remaining parts of this paper document how the concept of techno-therapeutic control was developed as an on-going part of the study. This involved a dialogue between 'internal' data and analysis with wider 'extant' theories and evidence. In sum television in prison is a therapeutic tool used to help manage and control prisoners' emotions and behaviour.

1. Data Analysis: Developing A Core Concept

4.1 The domains model provides a conceptual 'scaffold' (Layder 2005) which can be adapted as research evolves thus allowing theory development to be solidly embedded in data. However the adaptive researcher uses the scaffold as a series of access points not only for empirical data but also to 'extant theory'- evidence that sits outside the primary data collected by the researcher. This enhances visibility of the empirical data and its wider context by facilitating dialogue between data and external evidence resulting in a broader discussion. The next part of this paper demonstrates how theoretical concepts were developed during the analytical phases of the research using the adaptive approach.

Phase 1- Segmenting the Data: Pre-Coding

4.2 From this process the data from the interview transcripts are fragmented and placed into groups. Here 'topics of relevance' can be identified. Similar to open coding, pre-coding (Layder 2005:54 & 77) involves 'categorizing the data' (Strauss and Corbin 2007:61). Boeije (2010:96) points out that codes designate units of information that can be catalogued, thus allowing the researcher to develop early 'concepts' (or themes). Here readings of the prisoner transcripts were categorized according to three broad categories which would eventually provide 'core and satellite codes' (ibid:56). These core codes included talking about television, looking after yourself and tensions and dilemmas. For each core code a series of sub-codes were identified. These three core codes with associated sub-categories were not fixed, rather they were regarded as 'provisional codes to be subsequently firmed-up and 'validated' by on-going data collection and analysis and which, on this basis, may eventually be adopted' (Layder 2005:55). So in this first phase they provided an elementary framework to begin to identify relationships between the codes and sub-codes.

4.3 The Figures 1, 2 and 3 highlight the three initial core codes, with their sub-codes. In forming these sub-codes, the researcher took a pragmatic view of what these sections of data do. For instance the talking about television core code includes data segments about how prisoners use their television and how it shapes their everyday lives. For example,

My partner obviously I've been with from school and I've got three kids and basically my partner's kind of like me, we're both Pisces and she likes wildlife a lot, documentaries, but she's more into soaps than I am. So of a night time I will talk to my partner on the phone and she will say what are you watching tonight and I will say I'm watching this and she will watch it at the same time. In a way I feel like I'm with her if you get what I'm saying. I know that sounds a bit strange but telepathically I feel like I'm with her (Joshua- prisoner)

4.4 For Joshua television is used as a mechanism to connect with his partner to bring about feelings of intimacy. This captures the kinds of relationships prisoners have with television, including their attitudes to television viewing within the prison context. For example Ryan talks about how television can help him manage and plan his sentence,

You sail through with telly. There is always a new series. Like 'Prison Break' that is 6 months gone, a 6 month break and then it is on again. (Ryan- prisoner)

Talking about Television Core Code

Looking after Yourself Core Code

4.5 This core concept captures how prisoners adapt to prison life and how they use television and other activities to care for themselves. In particular the ways in which they seemed to be talking about television (and even avoiding it) served as mechanisms to enhance their well-being and to cope with the deprivations imprisonment imposes on them. For example Alan talks directly about the ability to cope in prison

It [TV] helps some things, some people can't do it, some people can. There are the strong minded and the weak. Some are copers and some are strong minded. You are still in jail, if they take the telly it doesn't matter. If you are sat in your cell with no TV it is the same as the day before, exactly the same as yesterday. Days, weeks and months go by, they are taken. It does for me sometimes, it depends what's on. (Alan- prisoner)

4.6 Sunny for example observes how his relationship with television is amplified as a result of imprisonment,

I am not addicted to TV, but sometimes I don't feel like I'm in prison when I'm watching. (Sunny- prisoner)

Tensions and Dilemmas Core code

4.7 Across the narratives it was evident that managing themselves within the prison context is complex and fraught with tensions and dilemmas. In particular their access to television was not straightforward and maintaining ownership of a television in prison requires compliance and also a financial commitment from the prisoner. For example,

TV is a privilege, they can take them away from you if you get a nicking. (Ryan- prisoner)

We're paying for it. There have been times where, alright then, there's been times where I've been in my cell for a month but like I said then I've gone to the seg^[4], that one pound is still coming off my money while I'm down the seg and I ain't got a telly. So it all differs. If they were giving it us without us paying for it then I suppose they could say, yes, it's a privilege, but at the moment I don't think it is. I think it's something we're paying for and therefore we should have a right for it. (Joshua- prisoner)

The narratives highlight how sometimes this is difficult to maintain, particularly in an environment which brings together challenging and vulnerable individuals.

4.8 These concepts were also initially adopted for the staff transcripts and data fragments from the staff interviews were also attributed to the codes. It was possible to identify from staff transcripts that they were sensitive to tensions and dilemmas of television in prison, as well as how television is a mechanism for looking after yourself. However there was other material in the staff transcripts that could not be coded using these labels. This is because some of their discussions were heavily focused on policy and procedures, such as prisoners' access to television, their daily work, and use of policy. At this stage a new core code was defined called staff. For example Tim, a prison officer, talks about the role that television has in maintaining order and control,

I think the main difference for staff with prisoners having their own in cell televisions, they're a lot easier to put away behind the doors. They appear to be happier behind the doors because they're occupied. And because they're occupied, they're far less likely to be on the cell bells requesting telephone calls or for any other reason that they may want to be coming out their cells. Funnily enough, if there's something important on the television that they want to watch, they generally don't come out to association. If there's football on television, they'll not come out on association. So the management of prisoners, particularly on the evening, is far, far easier. (Tim- prison officer)

4.9 As fieldwork progressed and more interviews were completed with prisoners, these initial categories were sufficient, as 'newly collected data from comparison cases do not change the outcomes so far' (Boeije 2010:114). A second wave of analysis began to refine and review these initial categories. At this stage the framework remains temporary until a number of checks are completed.

Phase Two of Segmenting: Axial Coding

This second phase refined and extended (or abandoned) the categories as stipulated in the initial segmenting phase. The development of the final categories and codes emerged as a result of intensive reading of the transcripts (prisoners and staff) and re-organisation or 'axial coding' (Strauss and Corbin 1990). The open coding now becomes restricted as,

…the researcher moves from a situation of total openness in coding to a form of closure dictated largely by the empirical data themselves. (Layder 2005:54)

4.10 In order to close off the coding process it can be helpful to also consider some of the wider literature in the fields of inquiry. At this point some of the key literature on prisons and audiences of television was revisited and new topics were explored. This helped to achieve closure and to realise the connections between codes and sub-codes. Thus a 'dialogue' between the emerging (primary) data and extant theory accelerated and intensified as analysis progressed (Layder 2005:55). All parts of the transcripts received a code and some parts received more than one code and thus appeared across different core codes in some instances. One example includes an exert from Leon's interview,

Well there is without TV, especially at night locked behind your door. Boredom is poisonous, it is mental poison. You can easily get distressed and suicidal in here. TV keeps you occupied. (Leon- prisoner)

4.11 In this instance this section of the data was attributed four codes and filed into the following code categories- emotion, well-being and therapy, television activity, prison life. These were further filed into sub-codes- boredom, bang-up, self-preservation (using television) and suicide. In doing this, the researcher is then able to understand and observe close relationships between the codes. In this example Leon's description of boredom emphasizes how the context of the prison specifically in relation to being locked inside his cell amplifies negative and self-destructive feelings. Television offers some respite for these occurrences. Coding in this way is beneficial to spotting trends across the sample. It was then possible to see where specific codes were coded or multiple coded. Leon was not the only person interviewed to relate boredom and the prison experience in these ways, and this was a reoccurring theme.

4.12 In the case of this study six core codes were finalised. These categories remained independent of each other at this stage and were not hierarchical; relationships between the fragments of data at this stage were not assumed. However within the codes themselves it was recognized there was significant synthesis across the sub-codes in order for them to be grouped together. For example the staff core code mainly refers to the work they do and the policy they administer. For example Tim, a prison officer talked about the way in which policy was organised when televisions were first introduced to prison cells,

…enhanced^[5] prisoners were the only ones with televisions. Enhanced status appeared to have a lot more status, if you like, prisoners were more keen to get on to enhanced because there was a perception that there was more to offer on enhanced than standard or basic. (Tim- prison officer)

4.13 The three categories from the pre-coding phase continued to be used, and three additional categories were created. In particular, the emphasis of prisoners using television to look after themselves whilst in prison was prevalent in the prisoner narratives. Moreover, staff were also sensitive to the care-giving qualities that television delivers to people, especially whilst they are locked in their cells. On closer reading it was evident that their descriptions were related to the ways in which they wished to control and actively manage their emotions. Hence, emotion was elevated to a core code. For the staff category, although many of the codes developed for the prisoners were appropriate, another category was needed to reflect reference to policy and prison work such as behaviour management. Prison staff were also instrumental in providing an insight into the introduction of in-cell television, and this also needed to be accounted for. Pre-coding identified these differences in perspectives between prisoners and staff. The final additional core code, study participation was included for respondents' reference to their views and attitudes to taking part in the study.

4.14 Descriptive and also interpretive or conceptual labels were identified for codes (Miles and Huberman 1994). For a descriptive label the researcher was able to code data fragments where expressions of emotions or discussions of emotions were apparent- for example all respondents talked about boredom and thus these instances were coded,

I find that if I'm left alone, that's when I get bored my mind goes, it is why I am in here, it pisses me off it gets me thinking and that is not a good thing to do (Stuart- prisoner)

4.15 Conceptual labels, however, draw directly from wider concepts. For example, 'emotional labour' – the work officers do that demands emotional work is highlighted by Paul,

So interaction between prisoners and between staff has reduced, in my eyes…because they'd rather sit and watch TV… it is quite important that you get interaction in prison if you really want to tackle re-offending because a prison officer's job should be that you'll sit talking and sit and say, this is not the way to go, there is a different way of life…I think it's reduced it down to virtually nil, if I'm honest. (Paul- prison officer)

4.16 This was added as a code to describe where respondents talked about the emotional adjustments they make in their interactions with others. This was especially relevant to staff where they talked about the work they do and was informed directly by Hoschchild's term for managing emotion (1983^[6]):

But I think in general the majority of the population respect the fact that they have got televisions. And the other thing you have got to take into account, when you penalise one you have to penalise another because they are double cells. If you have got one with a behavioural problem and he smashes his television you penalise the other person by taking that television away. So we wouldn't normally do that, we would normally place them on report and go down the discipline issue about damaging the cell… generally there is not a big issue with damage to the property, because they know once they start damaging it the chances are they are not going to get another one… (Paul- prison officer)

4.17 Paul's work with prisoners is rationalised by procedural details, in order to mitigate the amount of emotional labour required of the officers. They too actively manage prisoners' emotions.

4.18 Another example of a conceptual code derives from Lull's (1990) typology of 'social uses of television' which provides an existing framework to explore the social uses of television in the prison context and how they differ to the domestic environment. This question was important for understanding the peculiarity of viewing in the prison setting. It was apparent that the modes of engagement with television both mirror and subvert the social uses defined by Lull's research of television viewing in the home. For example,

I just lay down on my bed and watch it. Sometimes I will be sitting up but basically I relax and lay down and watch it. (Ned- prisoner)

4.19 For Ned television can act as a relaxant in line with Lull's notion of affiliation and avoidance. Other social uses relate to the environmental features of television viewing. In the case of Alan this provides him with some background noise

I get bored sometimes, writing a letter keeps me occupied for a bit. I get bored in afternoons, we leave the TV on, but not watching it… (Alan- prisoner)

As such the fragments of the data were coded in line with the conceptual codes defined by Lull.

4.20These conceptual codes were inserted into the Talking About Television core code.

4.21 It was also apparent from the prisoner interviews that respondents did not refer exclusively to the role that television had in their lives but to other activities and types of interaction. To account for these differences, using TV and not using TV sub-codes were included, to distinguish between these differing activities. For instance, when prisoners and even staff talked about techniques for self-preservation or forms of therapy, these codes provided a distinguishing marker to file the fragments of data accordingly. At this stage coding was dealt with on a case-by-case basis and memos were created via an 'integrative' process and this allowed a consideration of relationships across codes as well as between cases (respondents) (Boeije 2010:124).

4.22 It was through this axial coding process that the researcher began to observe the value of the typologies offered by Lull (1990) and Sykes (1999). It was evident that Sykes' famous pains were actually limited in highlighting the complexity of the felt experiences and hence an important reflection of how the prison experience has been described. In realising this limitation the reading of the transcript meant the research could draw out a range of different emotions such as happiness

Like last night there was a big game on, Chelsea, but it wasn't in the TV Times, one of the lads told me, I was really pleased…it was a nice surprise, I was very happy about that (Bill- prisoner)

4.23 Another example is frustration,

Sometimes I do the opposite to him. I can't stand Jean Claude Van Dam and Steven Segal or crime and The Bill. I like Panorama, Dispatches, News on ITV, but not regional. The Bill is most frustrating, it is police orientated , a warped perspective of what police do. (Will- prisoner)

4.24 At this stage some of the interpretive codes developed from extant theory such as Lull (1990) and Sykes (1999) were insufficient typologies to work with. However the use of Sykes' typology began to pose important questions about the value of these generic pains of incarceration. First it was observed that these pains conflated the felt experiences and thus it was more appropriate to code data fragments as emotional responses such as loneliness, sadness, or fear rather than using Sykes' typology to inflict pain by denying: liberty, goods and services, heterosexual relationships, security and autonomy. This then replaced the conceptual framework offered by Sykes to describe deprivations of liberty and threats to security. However, experimenting with both descriptive and conceptual codes during the axial coding phase, results in important and early indications of theory formation. Moreover 'dialogue' between established typologies and data pin-points the salience of the active silencing of emotion in broader discussions of emotion and criminal justice (De Haan and Loader 2002). Furthermore some approaches to research

…provide rich information about the subject's lived experience…but then fail to contextualize it fully from a sociological standpoint that looks closely at the interrelationship between agency and structure, the micro and macro and the subjective and objective dimensions of social life. (Houston and Mullan-Jensen 2011:270)

2. Reassembling the data: Concept-indicator links

5.1 Layder's (2005:80) view of social reality as four entangled domains highlights the multi-dimensional nature of social life. This framework offers opportunities to reassemble data to make sense of the relationships between the inter-related domains which, in effect, 'bridge ' or 'mediate ' between behavioural and structural aspects of social reality (ibid:80). Several techniques were employed to develop the final concepts. The volume of data stored in the categories was vast and of course not all of the coded fragments could appear in outputs of the research. Instead the research selected material that enabled 'linking verbatim extracts with the identified themes, helps to ensure the research is trustworthy' (Houston and Mullan-Jensen 2011: 269). In order to make these kinds of decisions, a number of inquiries were conducted using Nvivo to discover which fragments or codes^[7] were related to other codes and the kinds of patterns and anomalies that were present. These included using visuals, or 'code stripes' to see which codes appeared together or close to each other to establish links. Searching and counting were also used to identify salience and intensity of issues disclosed in the interviews (Seale and Silverman 1997). For example, boredom was mentioned by all of the respondents. Moreover, when boredom was mentioned, this fragment appeared close to descriptions of the prison regime such as pains of incarceration, especially bang-up (locked in cells) and discussions about how prisoners looked after themselves, such as well-being or the fear of deterioration. Leon's description of boredom was included earlier and can also demonstrate how these sub-code are used

Boredom is poisonous, it is mental poison. You can easily get distressed and suicidal in here. TV keeps you occupied. Even just changing the channels using the remote, it keeps you focused. (Leon- prisoner)

5.2 It was also noticeable that when respondents talked about boredom, television was useful in filling or killing time. Hence bridges between these features of prison life could be identified. At this point data from the television use diaries was drawn in to indicate the relevance between these codes (Layder 2005:85). For instance data from the diaries was able to confirm how much time people were using television and when this was likely to take place during any given day. As a result, the salience of boredom could be contextualized within these daily practices and in some cases highlighted the urgency of wanting to kill time and make it go quickly.

5.3 Following Miles and Huberman (1994) a matrix format was employed to observe relationships between an extensive range of codes and also between cases. This was useful for testing ideas and assumptions. Each matrix was stored and labelled to help build a picture of relationships or 'links'. For example, it was noted that prisoners' use of television is often used to create 'sociability' and was particularly salient when they talked about watching local news and soap operas. Sunny, a prisoner explained why local news was important,

It is very important. I am disconnected and without TV I can only get information by family and friends. You've got fresh news on TV, otherwise you are guessing about what is happening outside. Live TV is very important. For example the opening of the new shopping centre in X. It told me what it looks like, it gave me something fresh, fresh thinking and then I am still connected. I have a big family and friends and brought up with big groups and it helps me stay connected to them and it tells me this world is ok and one day you will go out and so it is important. (Sunny- prisoner)

5.4 In order to make these claims the matrix approach assists in spotting the linkages between codes. In this instance it was possible to observe overlap and closeness between the prison life code along with relationships with TV texts and staying connected. In doing this it was then possible to test how this aspect of sociability conformed to what other researchers have said about 'sociability' (Moores 2004). Within this same matrix it was also apparent most prisoner respondents were being highly selective in their viewing. Here a significant number expressed a desire to learn from television and pointed towards the social learning category of the social uses of TV category. It became apparent that prisoners' desire to learn from television was closely associated or positioned in their discussion near or next to fragments in the looking after yourself category, in particular wellbeing, self-preservation and fear of psychological deterioration. This did not feature in staff cases, so instead codes were created to capture how prisoners are cared for - safer custody, purposeful activity and decency.

3. Final Concept Development- Elaboration

6.1 The discovery of patterns and relationships across this data allowed what Layder (2005:116) calls 'theory elaboration'. These 'conceptual frameworks' can draw upon extant theory and emergent data. In this study the most significant concept was 'control' because it directly linked across the core codes and thus was 'bridging' between subjective and objective features of social reality (ibid:92). In recognising 'control' as a bridging and binding feature across the data it was possible to identify the nuanced nature of control:

• personal or control of the self
• situated control within and across interaction in social settings
• mediated control within and across interaction with television

6.2 Concepts can thus be developed not simply from the primary data a researcher collects, but can occur in conjunction with external concepts. For example the concept of personal control, was elaborated by Layder's (2004) thesis on emotion and Rose's (1999) work on 'self-regulation'. This dialogue with literature enables researchers to develop conceptual arguments that can be positioned with wider conceptual discussions and contrasts with grounded theory. Grounded theory restricts researchers to interrogating their (primary) data in order to draw out conclusions. The adaptive approach departs from this uni-lateral style of analysis and encourages visibility of concept-data indicators both within the data as well in relation to wider conceptual frameworks.

6.3 In progressing this further the fragments of data stored under the emotion category were re-listed and aligned to the concept of personal control. These fragments had relationships or 'linkages' to other concepts such as looking after yourself and prison life: tensions and dilemmas. This first stage of 'primary elaboration' (Layder 2005: 117) is to begin developing a conceptual map, which is 'the centre of a potential web from which others extend outwards'. All the categories (code headings) were listed under this new category, and this was repeated for all three forms of control. By identifying relationships (connections between fragments as well as disconnections) the original codes were sometimes amalgamated and given new interpretive names. Figure 4 below highlights this:

Figure 4.

6.4 A process of 'secondary elaboration' (ibid:121) demonstrates that substantive categories are relevant; thus 'clusters' can be created to capture a number of related categories. For a cluster to remain in the final concept group there had to be a direct relationship to the over-arching principle of control. Where this was not immediately apparent they were stored elsewhere, but not abandoned completely. In conjunction with this process, 'tertiary elaboration' assisted in drawing more explicitly from general or extant theory (ibid:122). As Figure 4 shows, relationships between a number of core and sub-codes had to be observed in order to finalise the concept of techno-therapeutic control. This three step process looks like this:

1. Linkages between codes- clusters

6.5 A review of the fragments of data within these categories was necessary in order to ratify these relationships; to understand explicitly what aspects of prison life, for example, motivated them to look after themselves. Adjoining this were responses from staff, who also viewed television in similar ways. Bill, a prisoner, highlights how the experiences of deprivations such as loneliness can be alleviated by television;

It is about mind control. TV pacifies you, like for the young and the hot heads that come in here. TV becomes your friend. If you have got time to spare TV is there. (Bill- prisoner)

6.6 However in positioning these deprivations Bill explains how other things can be missed more acutely,

When I think about what I miss, TV isn't one of them. There are things like beer and women but when you ask people you never hear them say they miss TV. (Bill- prisoner)

6.7 For Paul, a prison officer, talks about how prisoners' access to television is based on eligibility of access

And I think quite rightly if we are going to deprive people of access to media, we allow them newspapers, we allow them articles in the library, and for me its just another extension of that. Its part and parcel of the 21st century… I do think its something that prisoners have a right to expect a certain standard of care. (Paul- prison officer)

As a result of reviewing this kind of data it is possible to begin to elaborate on the perspectives of the respondents.

2. Elaboration

6.8 Several theories were therefore useful to expand upon the descriptions offered by staff to 'throw light on the data' (Layder 2005:127) and also produce 'hybrid ideas' (ibid:126). Here literature was drawn upon for elaboration. In particular the work of Rose (1999), Layder (2004), Garland (1991) and Silverstone (1996) were important for providing an elaborative framework with which to inform the primary data. At this point it is possible to begin to appreciate how people and organisations manage emotion and provide evidence to support what they do with them.

3. Concept formation- Control: Techno-therapeutic control

6.9 This process was repeated until all of the final concepts were developed by the 'conjoint use of empirical research and aspects of general theory' (Layder 2005:129). These concepts provided smaller frameworks to begin full elaboration and they appear as key themes in the research report. The concept of 'techno-therapeutic control' is one of a number of concepts developed in this project on in-cell television in prison. What this analysis demonstrates is that forms of control were adopted by prisoners and staff, which were distinctive in the prison context. The term 'techno' relates to the role of technology (television and other digital media) have in delivering and helping to achieve therapeutic care. All stakeholders were complicit in this exercise and television acts as an important mechanism for supporting modes of control which help prisoners to manage their emotions, especially those that magnify pain and deprivation. Yet in order to understand the context of control, elaboration on the social setting and contextual resources was necessary to provide a robust account of psychobiographies and situated activity (Layder 2004). In this respect the landscape of the prison with it's formal and informal norms and values can be positioned in debates about power and control.

Conclusion

7.1 This article has offered a blend of empirical information and a discussion of methodological issues in relation to social domain theory and the adaptive approach. In developing and forming theoretical concepts this strategy helps researchers to reach aspects of complex social and psychological phenomena- such as emotion, which are often concealed. The approach allows researchers to integrate emergent primary data with extant theories and engage in early critical analysis. Together domain analysis and the adaptive approach encourage researchers to look for and discover explanatory insights into the significance and role of emotions in a stratified but integrated way. By confronting the complexity of the issues involved concept formation moves beyond a descriptive account to an approach attuned to producing explanatory concepts around emotion and control in prison life – like 'techno-therapeutic control' – The salience of this approach to analysis has resulted in the generation of evidence revealing the intricacy of the relationship between prisoners' emotions and how prison institutions manage them. This methodological approach highlights how the social and psychological prison experience is subtle, complex and nuanced.

---

Notes

¹ The ethnographic strategy was influenced by Moores' (1993) research . In this context it seeks to capture varying perspectives of the television audience and observe the culture of television viewing – see also Silverstone et al 1991, Gray 1992, Jewkes 2002& Bird 2003.

² This deviates from Laing's discussion on the contrasts between 'ontologically secure' and 'ontologically insecure' people in his text The Divided Self (1990).

³ The nature of this study meant that only male prisoners participated. Gersch's (2004) study of media use in prison did include a small sample of women. A study which employs feminist perspectives is necessary to delineate gendered features of prisoner audiences.

⁴'seg' is shorthand for segregation. This is an area in the prison where prisoners who are difficult or are been punished for bad behaviour are placed and removed from the mainstream prison population.

⁵ The Incentive and Earned Privilege system is a mechanism for rewarding prisoners for good behaviour. Privileges can also be removed for poor compliance. There are currently four tiers to this enhanced, standard, entry and basic (HMPS 2011 PSI 11/2011)

⁶ The selection of the emotions listed here was not straight-forward. Opinion is divided as to what constitutes an emotion. Scherer (2005) provided the most accessible model of emotion, taking into account the 'subjective feeling component' (2005:698) as well as the more widely accepted components such as cognition, neurophysical, motivational and motor expression of emotion. Overall emotion is 'expression, bodily symptoms and arousal, and subjective experience' (ibid:698). For the purposes of this study a focus on the subjective and expression as well as self reports on bodily symptoms and arousal, such as crying and laughter are included in this category, as affective responses. Scherer is also one of the few theorists to acknowledge boredom as an emotion (see Barbalet 1999). It was evident that boredom was described in these different ways across the narratives.

⁷ Codes are referred to as Nodes in Nvivo. We shall refer to them as codes here.

---

References