Abstract
Drawing on narrative research conducted in the UK about self-injury and embodiment, this article presents a novel sociological analysis of self-injury, utilizing the concept of emotion work. By focusing explicitly on embodied methods of ‘doing’ emotion work, the article highlights the under-recognized importance of examining the practical, corporeal practices that can be involved in emotion work. I reflect on the sociological and theoretical significance of examining self-injury as embodied emotion work – both as an analytic concept and as a narrative resource.
Introduction
Self-harm is usually studied from a clinical perspective: however, sociological approaches have the potential to greatly improve understandings of the practice. Sociological work in this area is increasing, but there are important features of self-harm that remain under-explored. Different methods of self-harm are often overlooked – in this article I examine self-injury (cutting, burning or hitting the outside of the body) rather than self-poisoning. Further, emotional aspects of self-harm have been engaged with relatively superficially, despite this being a key clinical area, and there being a wealth of theoretical work on emotions that could be drawn on (Lupton, 1998b; Williams, 1998a; Williams, 2001). This article analyses accounts of self-injury as embodied emotion work. In doing this, I highlight hitherto under-examined aspects of both self-injury and emotion work.
Conceptualizing self-injury as embodied emotion work demonstrates a way in which self-injury can be understood sociologically. Accounts, and perhaps practices, of self-injury are shown to be closely tied to socio-cultural contexts. Sociological approaches should therefore be central to attempts to understand self-injury, challenging clinical psychological and psychiatric perspectives which tend to frame self-injury as a ‘problem’ located within the individual. Further, this analysis extends existing sociological theorization on emotion work. In contrast to research which has examined cognitive methods and feeling rules, I focus instead on the less frequently considered practical, corporeal ‘work’ that self-injury (and emotion work) entails. In my analysis, self-injury and emotion work are grounded in the bodies of people who self-injure, emphasizing the continued importance of incorporating bodies into sociological analyses of emotions, emotion work, and social life.
A general definition of self-harm in the UK is ‘self-poisoning or self-injury irrespective of the purpose of the act’ (NICE, 2004: 16). However, definitional problems have compromised academic and clinical understandings (Chandler et al., 2011). In particular, research has tended to gloss over the practical aspects of the behaviour, treating self-injury and self-poisoning as broadly similar. Although there are commonalities, I argue that in important ways the behaviours are distinct (see also Claes and Vandereycken, 2007). Self-injury involves the cutting, burning or hitting of the outside of the body. Self-poisoning usually involves (in the UK) ingesting higher than recommended doses of medication. Any damage caused by self-injury is immediately apparent, and can cause lasting (sometimes permanent) marks on the skin. In contrast, injuries caused by self-poisoning may not be obvious, and certainly tend not to leave lasting visible marks. The differential visibility of self-injury and self-poisoning may have important implications in how the behaviours are understood, an issue I discuss elsewhere (Chandler, 2010). Further, self-poisoning is far more likely to be treated clinically, making up around 80 per cent of self-harm presentations at Accident and Emergency departments (Hawton et al., 2004). Conversely, self-injury appears to be more common in (adolescent) community samples, but less likely to be treated clinically (Hawton et al., 2002). 1
Self-harm in general, and self-injury specifically, are said to be increasing in prevalence (Brophy, 2006). 2 However, measuring population levels of self-harm is notoriously difficult: it is often framed as a ‘secretive’ behaviour, and hospital presentations measure only a small proportion of all cases – a situation more acute with regard to self-injury (Hawton et al., 2002). The majority of studies that have sought to measure self-harm in community populations have focused on adolescent groups (Hawton et al., 2002; O’Connor et al., 2009). Hawton et al. (2002) found that around 13 per cent of adolescents reported ever self-harming and 6.9 per cent reported self-harm in the last 12 months, with around 65 per cent of this latter group reporting self-injury or self-cutting. Self-injury is increasingly understood to occur across a range of social contexts and groups (Adler and Adler, 2007). Despite this, statistics regarding the incidence of self-injury or self-harm among general adult populations in the UK do not currently exist.
Sociological work on self-harm (and self-injury) is increasing. However, definitions are not always clear, and, reflecting existing clinical biases (Chandler et al., 2011), research has focused on certain groups: women (Harris, 2000; Hodgson, 2004; Inckle, 2007), young people (Scourfield et al., 2011) and clinical patients (Hadfield et al., 2009; Redley, 2003). An exception to this is the work of Adler and Adler (2007) who have examined the sociological characteristics of self-injury from a deviance and social learning perspective, using a more diverse, community-based sample. Despite these important contributions self-injury remains an under-theorized and little understood behaviour. Sociological approaches can offer a great deal in expanding and challenging existing understandings of self-injury, and self-harm (see Cresswell, 2005 for an analysis of service user challenges to understandings of self-harm). In particular, work on self-injury and self-harm has generally overlooked the embodied and practical aspects of the behaviours (with the notable exception of Inckle, 2007). Self-injury and self-harm are inherently embodied, and my own work demonstrates the importance of an embodied approach in understanding explanations for self-injury.
Emotions and Emotion Work
Emotions are increasingly central to attempts to explain self-injury. A review of the clinical literature found that in studies which collected self-reported reasons for self-injury, the most commonly cited was ‘affect regulation’ (Klonsky, 2007). Affect regulation is a clinical term incorporating descriptions such as: ‘stopping negative feelings’, ‘relieving anxiety/fear’, ‘tension release’, and ‘stress management’ (Klonsky, 2007: 230–1). Clinical survey research which asked about feelings before and after self-injury has endorsed the hypothesis that self-injury is carried out in order to increase positive feelings, or decrease negative feelings (Klonsky, 2007).
Clinical research on the emotional aspects of self-injury is limited in a number of ways. The importance of the socio-cultural context of emotions is rarely, if ever, acknowledged. Emotions tend to be treated as universal and relatively unproblematic categories, with data generally collected through the use of highly structured clinical questionnaires. Although in some cases familial contexts are addressed (Wedig and Nock, 2007), little attention is given to the wider socio-cultural contexts of individuals who self-injure and how this might account for the emotional problems being reported. More usually, emotional problems are located within the individual who self-injures: they ‘struggle to regulate’ their emotions; they have difficulty naming emotional states; they are ‘unable’ to continue with normal tasks while experiencing strong emotions (e.g. Gratz, 2007). Clinical accounts therefore obscure socio-cultural contexts that might contribute to or mediate these ‘problems’.
Qualitative studies also report that self-injury is closely tied to emotions and emotional regulation (Horne and Csipke, 2009). Alexander and Clare (2004) found that self-injury was described as a method of ‘releasing’ painful emotions. Crouch and Wright’s (2004: 192) study of understandings about self-harm in an adolescent in-patient psychiatric unit found that patients frequently reported that self-harm caused a ‘release’ which helped individuals to ‘calm down’. Similarly, Adler and Adler (2007) reported that some participants referred to their self-injury as being a way of ‘expressing’ emotions. Solomon and Farand (1996: 115) found some participants suggested self-injury was a way of ‘transforming’ emotional distress into a more manageable physical pain. Some of these accounts also address the social contexts in which self-injury can take place, often focusing on problems with families and relationships.
Both lay and clinical accounts share a concern with the management or regulation of emotional experience. Many of these accounts are limited in important ways. Clinical work treats emotions in a reductionist manner: emotions are treated as individual experiences rather than inter-relational ‘complexes’ (Burkitt, 1997; Williams, 1998a), avoiding any engagement with the wider socio-cultural contexts of emotional life. Both types of accounts tend to gloss over the ways in which people who self-injure explain their use of self-injury in regulating their emotions; descriptions of what self-injury ‘does’ tend to be taken at face value. Finally, existing accounts of self-injury and emotions rarely address the embodied nature of emotions.
The focus on emotional management and regulation in existing literature points to the potential usefulness of Hochschild’s concept of emotion work. Despite numerous challenges (Bolton and Boyd, 2003; Wouters, 1989), emotion work continues to be both influential and valuable in sociological theorization (Freund, 1990; Frith and Kitzinger, 1998; Taylor, 2010; Theodosius, 2006, 2008). Hochschild’s (1979) original formulation suggests there are three techniques through which people carry out emotion work: cognitive, bodily and expressive. Hochschild was clear that cognitive, bodily and expressive emotion work were not practically separable. However, much subsequent work has tended to focus on analyses of feeling rules, rather than the techniques used by people to ‘do’ emotion work. Where techniques are examined these are generally cognitive or expressive types of emotion work (Bolton and Boyd, 2003; Theodosius, 2006). Bodily methods are less often addressed (Knights and Thanem, 2005) – though in some work, the embodied nature of emotions is occasionally noted (e.g. Theodosius, 2008). Indeed, although sociological work on emotions increasingly acknowledges embodiment, much existing writing on emotion work engages very little with embodied methods that individuals might use to ‘do’ emotion work. This mirrors a similar trend in writing on self-harm, where the bodily practices of ‘doing’ self-harm, and how these are accounted for and understood, are frequently under examined.
In this article I take the work of Frith and Kitzinger (1998) as a starting point in assessing the narrative functions of talk about self-injury as emotion work. Frith and Kitzinger noted the potential pitfalls of using emotion work only as an analytic category (see also Taylor, 2010). Their analysis of the use of emotion work in the narratives of women describing unwanted sexual encounters demonstrated the importance of examining emotion work as a participant resource, as well as an analytic category. Frith and Kitzinger suggested that women used narratives about emotion work to frame themselves as active and responsible, rather than as ‘victims’. Their analysis challenged more traditional approaches to examining emotion work, which, they argued, take interview accounts as ‘evidence’ that emotion work is occurring. Frith and Kitzinger demonstrated that analysing the narrative functions of interview talk can highlight links between explanatory accounts and wider socio-cultural concerns: in this case highlighting women’s resistance to dominant ‘victim’ narratives. Similarly, my analysis examines the ways in which emotion work was used in the narratives of people who self-injure, rather than focusing solely on the presence of such talk.
Theorizing Embodied Accounts
Rose (2003: 54) has suggested that emotional or mental health complaints are attributed more and more to malfunctions in the physical brain. Following this, individuals are more likely to ‘define key aspects of one’s individuality in bodily terms … and to try to reform, cure or improve oneself by acting on that body’ (2003: 54), a view Rose terms somatic individuality. Arguments analogous to somatic individuality have been used to account for the proliferation of physical treatments for mental illness, in particular pharmaceuticals (Lyon, 1996), though this argument is not without its critics (Abraham, 2010). Alongside this is the understanding that these (contested) changes in the framing of particularly ‘emotional’ problems have had wider effects on the ways in which people understand, account for and interpret their emotional lives (Stepnisky, 2007). Applying Rose’s concept of somatic individuality to accounts of self-injury as embodied emotion work may offer some indication as to why self-injury is accounted for in certain ways: in particular, the use of biomedical terminology to explain the efficacy of the practice. I suggest that accounts of self-injury might be interpreted as non-pharmaceutical methods of ‘working on’ the self, via the management of emotions through the body.
In a different, though equally embodied vein, Williams (1998b) has theorized the ‘performance’ of health in late modernity. He argues that the competing desires for control and release, along with the inherent uncontrollability of bodies, are reflected in ‘performances’ of health. Reflecting contradictory understandings of what it means to be healthy, health is about both control and release. These arguments are related to understandings about what it means to be emotionally healthy. Lupton (1998a: 47–8) explored lay understandings of emotions, finding that emotional control was valued, but simultaneously the need for emotional release and expression was acknowledged. Motifs of control and release are also prominent in accounts of self-injury: thus, these narratives could be seen to reflect contradictory socio-cultural understandings of (emotional) health, simultaneously expressing the need/desire for both release and control.
Williams, Lupton and Rose share a concern with the embodied nature of social life and of individual attempts to account for practices and understandings related to health and emotions. As such, their work is useful in expanding existing sociological understanding of self-injury. In the remainder of this article I demonstrate that self-injury can be understood as a form of embodied emotion work. I suggest ways in which the concept of emotion work can be expanded, both through a focus on embodiment, and by incorporating theoretical perspectives on embodiment, rationality and modernity.
Methodology
Sampling and Recruitment
The data reported here are drawn from narrative, life story research that explored the ‘lived experience’ of self-injury (Chandler, 2010; McCormack, 2004; Riessman, 1993). I explicitly excluded self-poisoning, though some participants had experience of both types of self-harm. In contrast to much previous work on self-injury, this project was oriented towards the socio-cultural contexts in which self-injury took place, focusing on the life stories of 12 people who had self-injured. Sampling was purposive, and I aimed to recruit equal numbers of men and women who had a range of experiences with self-injury and with formal services. Participants were recruited from non-clinical community sites in a Scottish city, to increase the chances of including people who had not engaged with formal support services. Adverts were placed in community centres and on an online community website. In addition, some participants were recruited via word of mouth and snowballing.
Data Collection
Each person was interviewed on two occasions between April 2007 and May 2008. The first interview explored the participant’s life story, guided in most cases by a life-grid (Wilson et al., 2007). The life-grid is a research tool that allows participants to build up a ‘picture’ of their life story by filling in sections of a grid. The grid featured columns for each year of the participant’s life, and six rows covering the themes: education; employment; where and who I lived with; family and relationships; personal interests; health and ill-health. Participants used the grid in different ways, some sticking rigidly to the structure offered by the grid, while others used the topics as starting points for discussion. The second interview addressed self-injury more directly, with discussion exploring themes arising from the research more generally, along with issues identified during the participant’s first interview. The research was intended to be collaborative, 3 thus participants were encouraged to bring their own themes to the second interview.
My original aims were broad and exploratory; however, throughout the course of the research my analytic focus was drawn towards talk about bodies, emotions and communication, and these became the three main themes in my analysis. Each of these is present in academic and lay accounts of self-injury, though they had been little discussed in existing sociological work. That I was not explicitly oriented towards studying emotional or embodied aspects of self-injury from the start of my research is potentially problematic. Indeed, because of my initially broad focus I did not routinely or systematically collect data on emotions, or on embodied, non-verbal aspects of the interviews. Participants nevertheless included talk about both emotions and bodies in their accounts of their self-injury. Therefore, I was able to focus my later analysis on narratives of and about emotional and embodied understandings of self-injury.
Data Analysis
Analysis was a mixture of thematic and narrative techniques. Thematic content coding was conducted on the interview transcripts, while in my analytic writing I attempted to remain sensitive to the broader narrative contexts of the interviews. The transcribed interviews were coded using NVivo into the three broad themes of bodies, emotions and communication. This was followed by sub-coding, which was more inductive. Throughout analysis, I engaged with the context in which the accounts were given, acknowledging the place of each piece of coded text within participants’ broader narratives. Thus, although inevitably imperfect, I attempted to avoid ‘fracturing’ participants’ responses (Riessman, 1993). Participants’ interview talk was not viewed as a way of accessing their ‘inner thoughts’ or tapping into what they ‘really did’ when they self-injured (Atkinson, 1997). Rather, interviewees’ talk was examined as an account of their self-injury, specific to the context of their interviews with me. Thus, my own position and impact as collaborator and co-producer of knowledge was acknowledged throughout my analysis (Stanley and Wise, 1993).
Findings: Self-injury and Emotion Work
The final sample was relatively diverse, especially in contrast to the homogeneous nature of samples used in much of the existing literature. 4 Participants were aged between 21 and 37 years old, five were male, seven female. Although the majority were studying for or had gained higher educational qualifications, the socio-economic backgrounds and current employment status of participants were more mixed. Importantly, participants reported a wide range of experiences with self-injury. Some had injured themselves regularly and extensively over many years, whilst others had injured themselves on only a few occasions. All participants said they had cut themselves, many also reported burning (using cigarettes, hot knives or metal) and hitting (using fists or hammers) themselves. Participants talked about similarly various experiences with psychiatric and medical services: three participants were heavily involved in psychiatric treatment, including in-patient hospital stays; two participants had never sought formal help for their self-injury. Other participants reported varying involvement with services including receiving counselling, psychiatric medication and support from their general practitioner, visits to Accident and Emergency departments, and use of voluntary support agencies (e.g. The Samaritans).
During the interviews participants frequently explored the reasons they had self-injured, discussing the ways in which they felt self-injury had ‘worked’ for them and also describing self-injury as a form of ‘work’. Often, these accounts were explicitly embodied: referring both to feelings or emotions, as well as the practical, corporeal actions involved in self-injury. Participants’ descriptions of the ‘work’ that they did during self-injury, and their accounts of what was happening to their bodies when they self-injured parallel a number of theoretical approaches to embodiment: motifs of ‘control’ and ‘release’; and neurochemical understandings of bodies. In most cases, participants’ accounts of the emotional aspects of their self-injury referred to some form of ‘working on’ their emotions through their body. In the following sections I introduce two particular ways in which participants described their self-injury as ‘working on’ their emotions. First, accounts which utilized the motif of control and release; and, second, explanations that described self-injury as eliciting or creating emotions.
Control and Release
Release, relief and control were used by many participants when describing their self-injury. Release and relief were generally used interchangeably. For some participants, the release that self-injury effected was said to allow them to regain ‘control’ over their emotions, selves and in some cases lives more generally. For other participants, self-injury was described as a way of controlling otherwise uncontrollable feelings. I illustrate these issues through an examination of the narratives of two participants: Anna and Belinda. 5
Anna, aged 33, had first injured herself at the age of 14, when she broke her wrist with a hammer. She continued to hit herself and break bones until her early twenties. After a few years with no self-injury, she began to cut herself at the age of 27. Anna was engaged with psychiatric services, she had received several psychiatric diagnoses and had been prescribed psychotropic medicine. When I spoke with her, Anna reported cutting herself regularly and increasingly extensively. She said she was worried that she might kill herself, as during her self-injury she sometimes felt suicidal (see Solomon and Farand, 1996, for similar findings on the potentially blurred line between self-injury and suicide).
Anna’s discussions of her self-injury suggested that at times it might be experienced as ‘out of control’; however, her descriptions more often indicated that self-injury allowed her to ‘release something’ and thus regain control:
… when the situation seems to spiral and I’m whooo losing it. Em and it was like right, regain control, this is what I’m gonna do, I’m gonna cut myself, well it wasnae as calculated as that … but cut myself … and I cut myself, my right arm I cut myself and it just wasnae, … it wasnae deep it was just ken what I mean it was just, crappy cuts … this is gonna sound so bad em, … and so I covered it up … and I was like ah, no … ken, it’s not happening, so I got my blade and I cut my other arm and … it … was, literally like I could feel it and hear it sortae like tearing open, […] that was it, that was the one, it was like, it’s worked this time that, fine, d’you know what I mean? […] … and it’s, like, releasing something … and then when that whatever it is is released then your sortae regaining control … s’what it’s all about, it’s all about, control.
6
Anna’s narrative here is embodied, and demonstrates the ‘work’ that some participants said they had to put into self-injury for it to be experienced as effective. Anna suggests that she increasingly felt she had to inflict deeper wounds in order to effect this ‘release’. Anna is also clear here that her self-injury is a reaction to a ‘situation’. She related a number of different episodes of self-injury, each of which was in response to situations or people that she found difficult to deal with. In particular, these appeared to be circumstances where she felt out of control:
If I’m no in control of a situation, or if I’m no in control of what’s happenin’ … that’s when I self-harm, that’s when it’s like I have to, I have to regain this or I have to … like … take control or control something, […] sometimes I start to panic aboot things, and the only way I can stop panicking about it and think rationally about it is … cut myself … it just like, I dunno it makes me just stop I suppose and then, it’s like right ok, deal wi it. So I think it’s like getting control or gaining control.
Here Anna’s narrative refers to explicitly emotional issues. She contrasts the ‘panic’ she sometimes feels with ‘thinking rationally’ which she said she was able to do following self-injury. This parallels theoretical work by Williams (1998a) on the late modern desire for control, and the tensions between rationality and emotions (see also Barbalet, 2001). Anna’s description of her self-injury draws on broader socio-cultural motifs which privilege rationality and calmness. Her self-injury is described as a way in which she achieves calmness and rationality, even in the face of situations in which she feels panicky and out of control. Thus, self-injury – more usually framed as irrational – becomes a rational, perhaps logical response to a difficult situation. This argument is similar to that made by Harris (2000) in her analysis of letters written by women who had cut themselves. Harris’s respondents associated the ‘relief’ afforded by self-injury with ‘relief’ at still being alive. In contrast, participants in my research described relief or release as being tied to control – over situations, emotions or selves.
Belinda was aged 21, and had been injuring herself (mainly by cutting) since she was about 16. Belinda described her self-injury as changing in nature over time. Initially, she felt she had self-injured ‘impulsively’ at least in part to seek attention. Later, she said she self-injured in a more ‘controlled’ way, relating this directly to her attempts to manage her thoughts and feelings:
… that’s just how I feel I guess, there’s so many thoughts and ideas and, and feelings, and everything just sort of, I don’t know it’s just everywhere (laughs) and it just has to come out – well, it doesn’t really have to come out. I just want it to sort of stop and slow so I can sort it, into something that I can, … understand and deal with, and em, yeah, and I guess sometimes when I can’t do that and I just, can’t do it and it’s just really frustrating, that’s when I, need to feel something, […] something that I know is real and it’s there and it’s concrete […] my arm hurts it’s bleeding, it’s a feeling, instead of just confusion and … not understanding things […] and trying to make something of, whatever’s in my head, to make it into something understandable and manageable …
Belinda associates the practical, corporeal results of self-injury – a bleeding, painful arm – with being better able to ‘manage’ the confusion in her head. Thus we have some elaboration here of the idea that self-injury might ‘transform’ emotional pain into physical pain. For Belinda, self-injury is described as a way of distracting, if not stopping, confusing thoughts and feelings by providing an alternative focus. Importantly for Belinda, the bleeding, painful arm is framed and perhaps experienced as something that is more ‘real’. This might offer some explanation as to why accounts which emphasize the transformative power of self-injury are so readily repeated and accepted. Such accounts may reflect broader socio-cultural privileging of physical over emotional suffering (Bendelow, 2009: 80). Belinda suggests it is the very disembodied, amorphous nature of emotions that makes them less preferable to ‘physical’ pain which is described as more real, and, perhaps, more authentic. In addition, Belinda’s description of self-injury as ‘working’ to stop feelings of confusion could be seen to represent a non-pharmaceutical, non-clinical method of ‘working on’ emotional problems through bodily means, lending support to Rose’s (2003) concept of somatic individuality.
The narratives of Belinda and Anna demonstrate two ways in which motifs of control and release were utilized by participants in accounting for the ways in which self-injury ‘worked’ to manage feelings or emotions: Anna gained control by ‘releasing’ something during self-injury; Belinda was able to manage her emotions in a controlled way, using the ‘concrete’ effects of self-injury to ‘distract’ herself. Simon Williams (1998b: 443) has claimed that late modernity is characterized by competing and contradictory ‘imperatives’ of control and release, rationality and desire. However, while Williams frames control as denial and release as pleasure, narratives about self-injury complicate this already difficult distinction. Anna’s account frames the ‘pleasure’ experienced through self-injury as being associated with the (bodily, self, social) control that the release was understood to effect. Both Belinda and Anna highlight control as a positive aspect of their practice of self-injury. I would suggest that Williams’s analysis is overly broad, failing to account for the different levels of control individuals experience in their day-to-day lives. For individuals experiencing little or no control over their self, body or life, any moment of snatched control might be experienced as pleasurable.
Eliciting Emotions
While the accounts discussed above refer to self-injury as releasing or controlling emotions, other participants described using self-injury to work on their emotions by creating or eliciting feelings through self-injury. These narratives were strikingly similar to some of Hochschild’s (1979) earlier descriptions of emotion work. However, I suggest that accounts of self-injury offer an explicit method of ‘doing’ emotion work – more transparent than some of the descriptions of emotion work related by Hochschild. Significantly, these accounts frequently drew on biomedical or technical language to explain how self-injury was understood to function.
Francis suggested that his self-injury had partly been oriented towards eliciting emotions where previously he felt ‘numb’. Francis was 25 and had injured himself on three or four occasions between the ages of 21 and 23. Francis’s ‘emotional problem’ was that he did not feel upset enough about events that had happened in his past, or did not feel able to express or display feelings he felt he ‘should’ be feeling:
I wasn’t pretending that I wasn’t upset but I would just, I wasn’t letting people know I was upset, if you see what I mean. I think that em, that got to me, after a while, that I felt I was incapable of feeling anything, you know incapable of emotion and …. Em …. I didn’t like that, I wanted to be able to feel I wanted to, you know, live or experience stuff or … and so, self-harming was, you know a way of, feeling, pain, you know feeling pain ’cos it was something.
Francis’s discussion here reflects some of the difficulty faced when attempting to articulate emotions or feelings. There is a blurring between what is ‘really’ felt, what is displayed, and what is communicated. Francis’s account indicates that he felt he ‘should’ be feeling upset, but he struggled to identify whether he was ‘really’ upset, or whether he was having difficulty expressing his upset. Self-injury is framed as offering a way of feeling something less elusive, less opaque and – as Belinda mentioned above – more concrete: the pain, blood, and visible nature of self-injury are described as being more tangible than emotions. As with Belinda’s description of self-injury as transformative, Francis’s account privileges physical suffering over emotional suffering. In both cases physical pain is framed as more visible, and therefore, more ‘real’, or authentic. Francis and Belinda provide these accounts by drawing on ‘acceptable’ understandings about the relative desirability and importance of ‘emotional’ versus ‘physical’ pain.
Justin was 29 and had self-injured between the ages of around 16 and 24. Justin had a relatively positive narrative about his early self-injury, between the ages of 16 and 18. He described his early self-injury as working on feelings of social anxiety; a situational response to events where he was struggling to feel ‘appropriately’:
I definitely remember kind of, you know, getting sort of [a] rush, from it, you know if you were feeling a bit down and you kind of just, you know, saw the blood and then you’d be like, you know that would kind of, you know, give you a kind of rush, em, and then you could kind of, you know, … I mean I was always quite shy as a child and stuff, em, … and, I guess like if it was in sort of social situations where there was a whole bunch of people downstairs, you could, you know, cut yourself, get a rush, and then, you could kind of, […] you know almost like a sort of, almost like drinking […] basically, like it would give you a kind of you know, Dutch courage or something.
This account parallels Hochschild’s (1979: 561) example of someone at a party reflecting: ‘I made myself have a good time’. In contrast to the account presented by Hochschild, Justin is more explicit about the method he used to improve his mood and his ability to subsequently deal with social situations. Further, Justin suggested that self-injury ‘worked’ because self-cutting caused an ‘adrenaline rush’:
I think the first time it was associated with kind of a rush and, and a buzz […] it was more of a, sort of the, … the blood and the, … sort of you know adrenaline […] rush, em, … so, … yeah, with, I wouldn’t really say that it was, you know done for the, the pain.
Such accounts of self-injury could be seen to lend support to Rose’s (2003) concept of somatic individuality. Justin, and other participants in my research, accounted for the success of their self-injury in terms of its neurochemical effects. Both Rease and Mark said that self-injury caused endorphins to be ‘released’ or to ‘flow’, which created a feeling of well-being. This latter explanation reflects an oft repeated biomedical understanding of how self-injury functions, though not one borne out by clinical research as yet (Chandler et al., 2011).
Justin emphasized the difference in sensation between the ‘buzz’ and any suggestion that he cut himself to feel ‘pain’. In his description of this, Justin tied the ‘work’ he did on his emotions explicitly to his body: both in terms of the practical, corporeal actions involved (self-cutting), and in the way in which he accounted for the effects of these actions (neurochemical). Mark made a similar distinction, also using neurochemical terminology to explain physical sensations:
I mean both of them you’ve got endorphins flowing if you’ve, if you’ve got up the nerve to pick up a knife and cut yourself or you’re being thrown out of a club by a bouncer you’re, the adrenaline is flowing and all the rest of it […] so the physical sensation I think is probably pretty much the same.
In this discussion Mark attempted to account for the lack of pain he felt during self-cutting, comparing it to an incident where a friend had been injured but had not noticed. Like Justin, his account draws on technical, biomedical terms to explain and describe physical sensations. Other participants suggested that self-injury did not hurt because the brain ‘released’ pain-relieving ‘chemicals’. These accounts drew on neurochemical understandings of the body, and also on metaphors of releases or flows, similar to those identified by Lupton (1998b) in her work on lay understandings of the emotional body.
These narratives highlight the difficulty participants had in accounting for their self-injury as both emotional and physical, whilst also treating emotions and physicality as separate. In particular, this issue was expressed when participants discussed the role of pain in their self-injury. These difficulties both reflect and challenge the enduring power of dualist understandings of bodies and emotions, lending support to Williams and Bendelow’s work on pain and the limitations of dualism (1998).
The practice of self-injury itself, and indeed embodied emotion work in general, also support Rose’s concept of somatic individuality. Both refer to explicit attempts to improve the (emotional) self by acting on the body. If emotional problems are increasingly understood to originate in the body, whose hormones or chemicals are somehow malfunctioning (and often treated physically with drugs), then acting towards that body through self-injury, or other methods of embodied emotion work, might be a more understandable course of action. Thus, accounts of self-injury might be more likely to draw on narratives which implicate the (neurochemical) body in attempts to improve emotional well-being.
Conclusion
The ideas presented here are based on in-depth, qualitative research with a small number of people who have self-injured. In this article I have explored in detail one particular way of accounting for self-injury – that it is an embodied method of managing emotions. There are, however, many different interpretations and understandings of self-injury that would benefit from sociological attention, some of which I explore elsewhere (Chandler, 2010). Further sociological work is needed to more fully examine the suggestions and conclusions I have made here. In particular, future work should attend to potential contrasts between accounts of self-injury among different social groups (e.g. men and women, younger and older people).
The analysis of self-injury offered here demonstrates the usefulness of explicitly engaging with embodied methods of doing emotion work. The accounts I have discussed highlight the importance of examining the corporeal ‘work’ that can be involved in embodied methods of doing emotion work, such as self-injury. This ‘work’ was central to participants’ accounts of their self-injury, demonstrating that even when not directly attached to economic activity, emotion work can involve (physical) labour. My analysis of self-injury also serves to emphasize the relative lack of attention given to other embodied methods of doing emotion work, despite Hochschild’s early (and in some ways pioneering) emphasis on the embodied nature of both emotions and emotion work. Even where emotion work is considered solely as an analytic category (Frith and Kitzinger, 1998), attending to embodied aspects of this ‘work’ allows a more explicit examination of the different methods individuals might use to manage their emotions. Future research should address embodied methods, rather than taking for granted that people simply ‘do’ emotion work. Doing so would allow a more practical, corporeal orientation of our understanding of social life (Crossley, 2007).
Examining the ways in which embodied emotion work is used as a participant resource (Frith and Kitzinger, 1998) by people who have self-injured affords an important, further layer of analysis: offering additional affirmation of the importance of embodiment in emotional accounts, as well as providing a more nuanced analysis of emotion work itself. Participants drew on narratives incorporating embodied emotion work to justify and account for their use of self-injury to manage their emotions. They did this in two inter-related ways: framing self-injury as rational; and drawing on biomedical language to lend authority to their accounts.
Emotion work was used as a way of presenting self-injury as a rational – if not always successful – response to problematic emotions and situations. Participants frequently set up self-injury as a method of gaining control over otherwise uncontrollable or intangible feelings. In this way, their accounts reflected dominant, dualist understandings regarding the irrational nature of emotions (Jackson and Scott, 1997; Williams, 1998a). These accounts challenge psychiatric interpretations of self-harm, which more usually frame the practice as irrational and impulsive (Redley, 2010). Thus, participants portrayed themselves as ultimately rational, capable actors who were addressing – albeit in an unconventional manner – unwanted, undesirable emotions.
In many accounts, the embodied nature of self-injury was framed as central to explaining the efficacy of the practice in improving emotional states. In these cases, participants often drew on technical, biomedical language. Using such language could be seen as an attempt to lend weight to accounts of self-injury which, as noted above, is more usually framed as irrational and impulsive. By framing self-injury as a practice which acts on the body in ‘concrete’ ways – affecting endorphins or adrenaline flows – such accounts appropriate the power of dominant biomedical narratives about bodies. These narratives reflect the increasing encroachment of biomedical explanations into accounts of emotions and emotional life (Rose, 2003; Stepnisky, 2007).
Biomedical accounts of self-injury, and those which emphasize rationality over emotions, each incorporate enduring and powerful dualist modes of thought (Crossley, 2001; Williams and Bendelow, 1998). Explanations of self-injury that privilege rationality over excess emotionality suggest that these states are opposed and mutually exclusive. Similarly, where narratives about self-injury are embodied through the use of biomedical language, biomedicine and physicality are also privileged over emotions. Thus, although self-injury is increasingly associated with ‘emotional problems’ in clinical literature, individual accounts of the practice complicate this understanding. These accounts do implicate ‘emotional problems’, but self-injury is offered as a rational, biomedically justified response, grounded in practical actions (cutting, burning, hitting). Perhaps significantly, self-injury is framed as generally more successful, or preferable, to other potential responses: talking, or ‘cognitive’ emotional work. In this way, self-injury might be seen to broadly mirror biological approaches to understanding ‘mental’ illness, approaches which some have warned might lead to the ‘silencing’ of ‘entire realms of life’ (Stepnisky, 2007: 203). Simultaneously, focusing on self-injury in terms of individual ‘emotional problems’ could well ‘silence’ attempts to examine wider social processes that might contribute towards, or shape potential responses to, such problems.
It is possible, however, that each of these findings at least partly resulted from the nature and contexts of the research interviews. As noted above, many of my participants were university educated, and this coupled with my position as an educated person, researching self-injury academically, may have encouraged certain types of accounts. Emphasizing rationality, and drawing on technical, biomedical language to lend authority to accounts of self-injury may have been specific to the context of this research and the individuals involved with it. Again, this highlights the need for further research on self-harm which examines more diverse groups of people, especially those living in different socio-economic contexts.
Accounts of self-injury as embodied emotion work further demonstrate both the problematic nature and enduring power of dualist approaches to mind/body, rationality/emotions (Crossley, 2001; Williams and Bendelow, 1998). Although narratives about the practice of self-injury do endeavour to split off mind and body, emotions and rationality, the very embodied nature of self-injury complicates and frustrates these attempts. In particular, accounts that indicate ‘work’ being done on the emotions, via the body, force an acknowledgement of the interconnected nature of mind and body. Thus, sociological analyses of self-injury as embodied emotion work offer a more nuanced, practical and potentially more useful perspective than clinical accounts which instead tend to reproduce dualist interpretations of emotions, bodies, rationality and mind.
Footnotes
Acknowledgements
Thanks to the research participants for sharing their time, insights and experiences with me. Thanks to the ESRC for funding the research on which this article is based. The SHI Study Group at the University of Edinburgh, along with Angus Bancroft, Gethin Rees, and Steve Platt, provided essential commentary on various drafts of the article. Finally, thanks to the anonymous reviewers, and the Editors, for their helpful suggestions and comments.