(Un)twisted: talking back to media representations of eating disorders

Abstract

In 2014–2015, there were several news reports about a rise in the diagnoses of eating disorders (EDs), as attributed to the use of image-driven social media. Such coverage can be situated within a long history of concern in which those diagnosed with an ED are constructed as ‘especially vulnerable’ to the power of media images – a subjectivity which is pathologised and devalued precisely through its association with femininity. The most incisive objections to EDs being presented as a response to the ‘weight’ of media representation have come from Abigail Bray (2005) in her work on how anorexia is constructed as a reading as well as an eating disorder. In exploring the results of 17 semi-structured interviews with people who have experience of an ED discussing their encounters with media representations of EDs (material that is often co-opted into debates about the ‘toxic’ nature of media culture in this regard), this article uses empirical evidence to intervene in how ‘ED’ media subjectivities are often defined. In particular, it aims to explore the qualitative responses in the context of more ‘every day’ understandings of media engagement, thus working against the gendered pathologising which has persistently occurred.

Keywords:

• Eating disorder
• audiences
• media
• anorexia
• bulimia
• feminism

In 2014–2015, there were several news reports about a sharp rise in the diagnoses and treatment of eating disorders, as squarely attributed to the ‘rise of social media that has helped develop an obsession with image’ (Parry, 2014). Some of these reports focused on the general use of image-driven social media, such as Facebook and Instagram, in which the mediated circulation of youthful bodies is central. Others focused on how pro-eating disorder communities were finding new ways to incite competition and to ‘recruit’ members via social media, ‘exposing those susceptible to eating disorders to a new level of psychological pressure’ (Bingham, 2015).

Such coverage can be situated within a long history of concern which pivots on a perceived confluence between eating disorders (EDs), mediated images and highly ‘susceptible’ female audiences. To be sure, there is now a substantial body of empirical research which investigates the relationship between media influence and body image. But whilst some scholars suggest that clear and causal links have been established, others position the results as more uncertain and contradictory, suggesting that a consistent relation has not been established (see summary in Tiggemann, 2003). Either way, although it is problematic to entirely ignore the ubiquitous nature of media images and their potential implication within body/ eating distress, existing research does not substantiate the characterisation of the media reading practices of ‘eating disordered’ (ED) ‘subjects’ – especially those diagnosed with anorexia. In medical and popular discourse, the latter are routinely defined as occupying a ‘heightened state’ of susceptibility in relation to media culture (Thomsen, McCoy, & Williams, 2001), and as being ‘suggestively vulnerable’ (Levitt, 1997) to its influence. As this somewhat simplistic characterisation suggests, none of this work has emerged from Media and Cultural Studies’ approaches to audiences which challenged the hegemony of the media ‘effects’ model some time ago (see Barker & Petley, 1997). This shifted attention to how media audiences actively engage in the process of meaning making within broader operations of social power (Gillespie, 2005).

The most incisive objections to EDs being presented as a ‘fall out from experiencing too much representation’ (Probyn, 1988, p. 208) have come from Bray (2005) in her work on how anorexia is constructed as a reading as well as an eating disorder. Bray explains how a ‘dominant paradox has been mapped onto the anorexic subject’ which insists that ‘food refusal [is] … the direct result of the consumption of media representations …. An excessive media consumption is perceived to activate a pathological fear of corporeal consumption: over-reading produces under-eating’ (2005, p. 116). Bray examines how anorexia has become a ‘synecdoche for all [female] … pathological vulnerability’ to media forms – imagined reading practices which construct girls and women as infantilised and irrational and which have a long history (Ibid., p. 123). In doing so, Bray implicitly indicates how the construction of the ED ‘subject’, a position historically defined as feminine, is a matter of importance for feminist approaches to media audiences. Indeed, there is a whole history of empirical work in Feminist Media Studies (Ang, 1985; Radway, 1984; Ralph, 2011; Stacey, 1994), and more recently Girlhood Studies (Coleman, 2009; Holmes, 2016; Mitchell & Reid-Walsh, 2007), which has challenged the pernicious construction of female subjectivity as ‘excessively’ invested in, and effected by, mass mediated forms. Yet the discursive construction of ‘anorexic’ reading practices (which have in turn become the most visible representation of the relations between ED sufferers and ‘media’ consumption), arguably offer the most hyperbolic image of this devalued, feminine subjectivity. It is thus notable that such claims have not been subject to a similar empirical re-evaluation – an omission which this article seeks to address.

In exploring the results of 17 semi-structured interviews with people who have experience of an ED discussing their encounters with media representations of EDs (including anorexia, bulimia or EDNOS),¹ this article seeks to intervene in how their imagined media consumption practices are often defined. Rather than deliberately framing an interest in the interviewees’ experiences of media in terms of a negative and contaminating relationship between the textual and the material body (Bray, 2005, p. 118), the study invited participants to occupy a potentially more active position from which they could reflect on, evaluate and critique media representations. In order to do this, the article draws on feminist approaches to EDs, as well as feminist approaches to media audiences, using them to interrogate how alternative perspectives on ED experiences and media representations might be explored. This approach can be situated in relation to audience research which has offered a space for marginalised groups to respond to how they are represented in the mass media (e.g. Means, 2002). But it can also be placed in the context of critical feminist work on EDs which has drawn attention to the ways in which people diagnosed with an ED are often ‘insulted’ and ‘alienated’ by medical and popular discourses on their condition (Eckermann, 1997; Saukko, 2008), including those that invoke the media as the key source of blame (Boughtwood, 2005). In seeking to complicate the pathologising ways in which ED ‘audiences’ are often imagined to interact with media forms, my interest is in examining how ED representations are understood and evaluated via the subjective experience of the participants.

‘Suggestively vulnerable’? media bodies/material bodies

In critiquing the dominance of medical and psychiatric models of EDs which prioritise biological and individual factors (and which define EDs as a mental illness), critical feminist perspectives are concerned with culture – namely the relationship between EDs and the construction of gender and sexual difference (Bordo, 1993; Orbach, 1986; Warin, 2010). In resisting a pathologising division between those who have/ do not have EDs, feminist work in the field has often sought to position eating/ body distress on a continuum with the discourses and practices associated with normative femininities (Malson, 2009). The earlier feminist work on EDs often foregrounded the role played by media discourses in this regard (Bordo, 1993; Orbach, 1986), but later feminist interventions have been more critical of the ways in which EDs are reduced to a problem of ‘body image’, a simplification which neglects the broader range of cultural forces which contribute to gender inequities and are expressed through eating and body distress (Malson, 2009, p. 136). It is thus somewhat contradictory that the feminist models have been critiqued for failing to explain why not all women develop EDs by reference to the role of the media – why so many ‘individuals do not develop the [problem] … even though they have been exposed to the same cultural environment’ (Brumberg, 1988, p. 38). Yet it is from this perspective that the idea of ED sufferers having a ‘vulnerability’ in relation to media content sneaks in and gains legitimacy, especially in work emerging from medicine and the social sciences.

In their brief overview of research on ‘Eating Disorders and the Role of the Media’ in the Journal of the Canadian Academy of Child and Adolescent Psychiatry, Wendy Spettigue and Katherine A. Henderson outline how:

… females who have already internalized the thin beauty ideal and/or who already have high levels of body dissatisfaction are the most vulnerable [to an ED] … [Further], women with anorexia engage in heavy media use and describe their consumption of fashion magazines as an ‘addiction’, with many saying that the greatest media dependency occurred after their eating disorders had taken control of their lives. (2004, pp. 3, 4)

The language here is pathologising and draws on a discourse of substance abuse. The quotation also directs us back to Bray’s suggestion that such representations pathologise female subjectivity more widely. After all, the reference to girls/ women who have ‘internalized the thin beauty ideal’ or who experience ‘high levels of body dissatisfaction’ could be seen as approximating the subjectivities of normative femininity (Bordo, 1993; Orbach, 1986). Furthermore, such characterisations are either presumptive, or they are yoked to a limited number of early studies which set out to explicitly ‘measure’ the relationship between media effects and EDs in highly reductive ways (e.g. Murray, Touyz, & Beumont, 1996). To be sure, work on body image has increasingly recognised how a range of differential factors within the category of ‘woman’ may moderate or inflect how images of the body are negotiated (including age, ethnicity and existing eating habits) (e.g. Kilpela, Becker, Wesley, & Stewart, 2015). In addition, feminist work on the body – including that focused on girls – has sought to theorise the body beyond a binary of representation/ materiality (Budgeon, 2003; Coleman, 2009; Frost, 2005). Rather than perceiving the female body as a passive slate awaiting the inscription of oppressive media discourses, it asks us to consider how bodies ‘are experienced through … images [my emphasis]’ (Coleman, 2009, p. 70), and to understand the female body as a process (or event) which unfolds with images (Ibid., p. 1; Budgeon, 2003). But while related, such work has not directly addressed the imagined media subjectivities of the ‘ED subject’, whose relations with media culture remain marooned within a sea of problematic assumptions.

There is clearly a distinction between examining the role of the media in the aetiology of EDs and thinking about responses to media representations of EDs. But such representations are in fact often co-opted into debates about the ‘toxic’ role of the media in the production of EDs. The American psychoanalyst Hilde Bruch (1988) popularised the idea that media representations of EDs act ‘pedagogically in so far as [they] … instruct women on how to reproduce the symptoms’ of such problems (Bray, 2005, p. 120), and this perspective has permeated subsequent work in the field (Levitt, 1997; Thomsen et al., 2001). Steven R. Thomsen et al. for example describe how ED stories may ‘back-fire when directed toward at risk women’ (2001, p. 61), and how once an ED has developed, readers may ‘twist these warning messages to suit their needs and purposes [my emphasis]’ (Ibid.). This term ‘twist’ – which I have also used in the title of this article – is revealing in suggesting the highly problematic ways in which the relations between ED sufferers and media culture are imagined. Not only does this term imply a type of warped or deviant reading of media ‘messages’, but it also places the source (and responsibility for) the ‘pathology’ onto the individual themselves. After all, with regard to media representations of anorexia at least, a number of scholars have noted how such portrayals are deeply contradictory in their bid to both pathologise and glamourise self-starvation (Holmes, 2015; Warin, 2010; Way, 1995; Whitehead & Kurz, 2008) – a contradiction that is indicative of the ways in which the very thin (or even emaciated) female body is on a continuum with desirable ideals of normative femininity. To acknowledge such instabilities, and thus to ‘untwist’ the claims to a distorted view, is to reveal the extent to which it might be cultural constructions of femininity that are ‘warped’, rather than the minds of the ED sufferer themselves.

Explorations of how the mass media might influence the development of an ED have far outnumbered studies of how EDs are portrayed in the media (Shepherd & Seale, 2010, p. 487), and existing work has included news stories (Ibid., Strother, Lemberg, Chariese Stanford, & Turberville, 2014), radio phone-ins (Brooks, 2009), women’s magazines (Whitehead & Kurz, 2008) and celebrity coverage (Holmes, 2015; Saukko, 2008). Research on the online construction of EDs has burgeoned in the last 15 years, largely with respect to ‘pro-ana’ or ‘pro-mia’ discourse, which is now increasingly articulated via social media networks (e.g. Boero & Pascoe, 2012; Yeshua-Katz & Martins, 2013). With its focus on both user and professionally-generated content, there is a debate about whether social media should be conceptualised as part of ‘the mass media’ (van Dijck, 2013), and in the context of this research, I left it up to the participants to decide what constituted ‘media representations’ of EDs. Particularly interesting here has been the anxiety generated by the figure of the ED subject as a media produser (rather than ‘excessive’ media consumer), involved in the production, manipulation and circulation of ‘thinspirational’ images online, as well as the broader meanings of what ‘EDs’ are (see Ferreday, 2003). But the media coverage of such cultures has done little to challenge the wider association between young femininity, EDs and media susceptibility. Not only have they helped to make visible a series of heightened anxieties about girls’ ‘unsupervised use of digital technologies’ (Martinez, 2007), but the relationship between the mediated and the material body is still frequently framed in terms of ‘risk’.

In wider terms, the research on mass media representations of EDs reflects a long-standing concern about the adequacy of media reporting, and the implications this has for public understandings of EDs (Shepherd & Seale, 2010). Such concern is suggested by the ways in which organisations such as Beat (in the UK) and Mindframe (in Australia)² have developed guidelines for media reporting on EDs, and Beat also act in a consultative capacity to advise on such representations. I do not, however, provide a fuller summary of such interventions here, nor attend to the debates about how media constructions of EDs might have developed, changed or ‘improved’ (see Shepherd & Seale, 2010). As acknowledged in the introduction, my interest is in examining how representations were understood and evaluated via the experiences and words of the participants.

Method

The national eating disorder charity Beat offer a service which helps to recruit participants for research studies (Beat, 2015). Participants were primarily recruited via an advert on Beat’s research page, as well as the circulation of the study via their social media profiles. In addition to being 18 or over, the criteria for participating in the study was simply that respondents had experience of anorexia, bulimia or EDNOS,³ and considered that talking about media representations of EDs would not cause significant distress. Whilst girls and women still constitute the majority of ED sufferers (albeit it in a context in which male EDs are seen to be on the rise) (Strother et al., 2014), I did not limit the recruitment criteria along gender lines. This was because the aim was to explore – and potentially critique – the construction of a pervasively pathologised and feminised subject position, rather than simply ‘answer’ such claims with empirical responses from female participants. Ultimately, the study recruited 16 females and one male: eight of these had experience of anorexia and one of a-typical anorexia⁴ ; six had experience of bulimia, and two had experience of EDNOS. The ages of the participants ranged from 19 to 45, with the majority (13) bunching between the ages of 22–32. Twelve defined their ethnicity as white British, one as white British Jewish, one as white Welsh, one as white Scottish, one as white European and one as German/Pakistani. The fact that the study predominantly recruited white respondents unfortunately reflects the ways in which media constructions of EDs still tend to carry a white bias – a partiality discussed by participants within the study itself. The sample also replicates the prevalence of white voices in qualitative research into EDs (Cortney, Gleaves, Cepeda-Benito, Carmen Fernandez, & Rodriguez-Ruiz, 2005), a bias that continues despite growing evidence that EDs are not confined to Caucasian groups. . In this regard, although this article is justified in referring to the participant voices as silenced when it comes to negotiations with media culture, it remains important to recognise that some voices are always more silenced than others.

In seeking to disrupt the discourses through which the relationship between media texts and EDs are often defined, a decision was also made not to limit the recruitment of participants to those with experience of anorexia. Anorexia is constructed as the most hyperbolic condensation of the ‘contaminating relationship’ (Bray, 2005, p. 117) between media/ material bodies. But such discourses are also used to encompass other types of eating disorder such as bulimia and EDNOS, both of which – however problematically – have been associated with a ‘desire for thinness, body dissatisfaction, and fear of weight gain’ (Burns, 2004, p. 269). To focus exclusively on the voices of those who have been diagnosed with anorexia would also replicate and reinforce the ways in which anorexia represents the more ‘spectacular’ and visible ED, and the ways in which people with different eating problems are subject to various forms of marginalisation within medical, scholarly and popular discourse (Burns, 2004; Squire, 2003).

Some participants considered themselves to be recovered whilst others did not, and the criteria for participation was deliberately inclusive in this respect. The voices of those with an ED are often considered to be ‘outside of the true’ (Saukko, 2008, p. 61) – as complexly bound up with highly gendered discourses of pathology, deceit and a ‘warped’ view of reality. As this flatters the notion of a ‘skewed’ or ‘twisted’ relationship with media culture, I felt it important to include the voices of those who are still perceived to be ‘in’ their ED. It is also within this same spirit that I do not indicate the status of each participant (in terms of a present or past relationship with an ED), treating them all as simply people with experience of eating/ body distress whose responses to media representations are worth respecting and listening to.

Participants took part in a one-to-one interview that lasted on average about one hour. The interviewer provided a brief description of the study at the start, and disclosed that she herself had experience of anorexia, which lasted for 20 years. The decision to self-disclose was taken for various reasons. First, it was prompted by what felt congruent and authentic to me as a researcher: I could not have comfortably undertaken the interviews pretending that I had no first-hand experience of an ED, and thus experience of what it was like to be positioned by media representations of EDs. Second, my decision to self-disclose was prompted by the long-running debate on the dynamics of the interviewer-interviewee relationship and the power dynamics in which the interview takes place, a debate that has been central to feminist work with women across disciplines (Tang, 2002). In critiquing a positivist epistemology, feminist qualitative research has invested in models which seeks to reduce the gap between ‘expert’ and participant (Oakley, 1981), and has thus often included discussions of personal experience and reflexivity (Tang, 2002). But the possibility of a ‘non-hierarchical’ relationship (Oakley, 1981), as based on gender congruence and perceptions of shared experience, has long been subject to critique. As such, feminist work has explored the range of social attributes – such as class, race, age and educational capital – that can shape the balance of power within the interview encounter (Tang, 2002).

Although the disclosure of my experience cannot be seen as circumventing any potential differentials of power, it was nevertheless offered as a form of reciprocity (see Ribbens, 1989, p. 594), and I indicated my willingness to answer any questions about my own ED history. In this regard, I was keen that participants did not feel like a ‘goldfish in a bowl’, interesting because of their (so-called) ‘aberrant’ or ‘exotic’ media reading practices. Indeed, whilst it would be problematic to characterise the interviews as simply ‘friendly free-flowing discussion[s]’ (Tracy, 2013, p. 56) – and I fully recognise that this article is also a chance for me to speak back to how ‘I’ was represented – they emerged, at times, as focused conversations (Spitzack, 1988, p. 55) in which we shared experiences of eating/body distress. Although I acknowledge that my personal disclosure could have had various implications that were not openly articulated by the respondents, and that interview data is always produced within situated and complex relations of power, my admission was repeatedly received in positive terms, as expressed during the interview itself, or via subsequent email contact.

Whilst I did not know in advance how participants would respond, I wanted to ask questions which did not pivot on nor assume a harmful relationship between the textual and the material body (Bray, 2005, p. 118). In so doing, I sought to create a space in which participants could evaluate media representations of EDs, whilst I also wanted to introduce alternative themes to those which dominate medical/popular discourse in this field. As such, the interview schedule covered key five areas: (1) how conscious participants were of media representations of EDs; (2) to what extent they evaluated them; (3) whether they recollected the first time they came into contact with a media story about EDs; (4) whether they felt that there was anything not represented in media representations of EDs and (5) how they would represent their experience of having an ED. At the same time, the emphasis on seeking to provide a forum for alternative perspectives needs to acknowledge that there is no ‘outside’ space from which to speak back to media representations of EDs. Just as the framing of this article has had to enter into dialogue with existing medical, psychiatric and popular constructions of the relations between ED sufferers and the media, so do the responses of the participants: such discourses constitute part of the reality of what ‘EDs’ are and how they come to mean. As such, this process of negotiation or discursive exchange is evident in the participant responses themselves.

The interviews were recorded on a digital voice recorder and transcribed, and the participant data was anonymised at the point of writing.⁵ Informed by wider feminist qualitative research on EDs (e.g. Boughtwood & Halse, 2010; Burns, 2004), I then conducted a thematic discourse analysis of the transcripts which involved three different stages. Upon a first read through the transcripts I deduced that the responses clustered around three broad themes. These were (1) the extent to which participants saw themselves as being ‘triggered’ by media representations of EDs; (2) their most striking memories of media representations of EDs (3) and their evaluations of their perceived ‘authenticity’. I then conducted an in-depth qualitative analysis of these themes (Smith, Harre, & Van Langenhare, 1995), looking for recurring patterns and commonalities. Lastly, I critically reflected on how the respondents positioned themselves in relation to normative discourses on the relationship between media and EDs, and the extent to which this conformed to, subverted, or negotiated with, the gendered discourses which conflate ‘eating’ and ‘reading’ disorders.

But as this process suggests, it is also important to acknowledge that qualitative research into EDs (or indeed any kind of qualitative research) does not unproblematically give the participants ‘voice’ in speaking back to dominant discourses (Malson, Finn, Treasure, Clarke, & Anderson, 2004; Saukko, 2008). As Saukko observes in her qualitative ED research, there is a potential tension between the desire to ‘listen carefully and faithfully’ to the women’s experiences of eating and body distress, and the impetus to critically assess the discourses … from which their voices are made …’ (2008, p. 77). In this regard, questions of ‘voice’ are pressing not only in the solicitation and construction of the data (in which regard I have reflected on my own positionality), but also with respect to its analysis and interpretation.

In terms of approaching a conceptual framework for the data, I knew that I wanted to take from Media and Cultural Studies’ work on audiences a rejection of a media ‘effects’ theory paradigm. I thus approached the data with an understanding of audience-text relations as always negotiated within complex, contingent and contextually specific ways (e.g. Gillespie, 2005). But I did not begin the research with a preconceived idea about which aspects of the work on media audiences would be relevant – a focus which ultimately became multi-faceted as it responded to the range of themes within the interview data itself. As such, this article uses a range of different conceptual lenses informed by feminism (moving across issues of trigger warnings, emotion and affect, to negotiations of the ‘real’), in a bid to provide alternative perspectives which may push at the parameters of the debate, and begin to disturb the ways in which ED ‘reading practices’ are often defined.

Trigger warning: this media could seriously damage your health

The fact that those diagnosed with an ED are encouraged to understand their problem as in part a response to media culture, is suggested by the ways in which this was frequently the first topic to come up, despite the absence of questions or prompts on this theme. Some participants expressed critical and sometimes angry views on the prevalence of ‘media blaming discourses’ about EDs (Boughtwood, 2005). Acknowledging the extent to which constructions of young femininity are inextricably linked to the idea of the ‘anorexic’ as a ‘vulnerable’ media reader, the male participant in the study was quick to criticise the ‘very mono-causal coverage of anorexia … which plays into the idea of anorexics as just bony girls who read too many copies of Cosmo’ (P12). Some participants explained how their decision to take part in the study was prompted by a desire to prove that they were not ‘in fact media sponges’ (P11), whilst others rejected the causal narrative of media effects with such assertions as ‘I find it incredibly insulting and patronising that such a … [perspective] is out there’ (P9).

But that is very far from saying that this causal relationship, specifically with regard to media images of EDs, was not endorsed at other points in the interviews. This theme primarily emerged in relation to the idea of media content as an ED trigger – a term introduced and used by the participants themselves. Originating in the domain of psychiatry and in relation to a focus on trauma, the concept of a ‘trigger’ is understood as something that initiates and/or aggravates a form of behaviour or response (Smith, 2014). As an adjunct to this, the term ‘trigger warning’ has a long history in relation to media culture, as used to alert viewers to the potentially upsetting or offensive nature of media content. But the term is now also frequently used in online discussions of sensitive issues and topics. This can include problems associated with addiction (drug and alcohol abuse, EDs or self-harm) as well as discussions/ representations of sexual abuse or suicide, or hate speech directed at marginalised groups (Kyrola, 2015). Within this context, trigger warning discourse has prompted heated debate (particularly in the US) about the politics of such interventions within academic, social media and/ or activist spaces (Smith, 2014), with proponents seeing them as a form of ethical ‘protection for vulnerable groups’, and detractors attacking them as extensions of a ‘neoliberal culture of overprotection, excessive self-involvement, and celebration of victimhood …’ (Kyrola, 2015, p. 131).

As has been discussed within both popular and scholarly arenas, there are often gendered dimensions to such debates. This is not only in terms of feminism’s role in helping to originate trigger warnings (in relation to making the personal ‘political’ and challenging the relegation of pain to the private sphere (Carter, 2015; Smith, 2014)), but also with regard to how the ‘triggered’ subject position is constructed as ‘feminised’ in highly negative ways. Indeed, it is this latter perspective which is expressed by the participants above: such discourses are described as ‘insulting’ and ‘patronising’, whilst the comments also allude to the gendered nature of such constructions (as in ‘just bony girls who read too many copies of Cosmo’).

The concept of the ‘trigger’ now pervades both popular and medical constructions of EDs. As such, given that the ED sufferer is already positioned as vulnerable in relation to media encounters, the feminised subject position implicit within the trigger warning debates is doubly invoked here. The participants in the study did sometimes talk about media representations of EDs as triggering, although only one participant in the sample, a 45 year old woman who began to suffer from bulimia in the 1980s, saw such material as playing a definitive role in the development of her eating problems. In describing her encounters with the fictional novel and TV movie The Best Little Girl in the World (1981) which tells the story of 17 year old Casey Powell’s descent into anorexia and bulimia, she recounts her experience thus:

At the time I got all my ideas [on vomiting] from this book … and movie and I’d take notes … I would keep in my closet every article [about the movie] I could find in any magazine or newspaper. It was my secret box, my ammunition, my toolbox. (P2)

The participant talked openly about witnessing her mother’s struggle with bulimia as well as her father’s binge eating, both of which she understood as contextual risk factors in her development of an ED, but clearly felt that the media representations were the most definitive cause. Yet it is notable that even in this passage her media encounter is articulated in ways which problematise associations of passivity or vulnerability. The word ‘ammunition’ (which, as with ‘trigger’, draws on a discourse of weaponry) indicates an active preparation for combat or battle. In fact, she imagines here a kind of firing out and not simply a ‘taking in’ (as the ‘sponge’ metaphor would have us believe).

In comparison, other participants made more fleeting references to what they saw as triggering media encounters, or described practices which were implicitly framed as protective measures in this regard. So one commented how ‘I have a filter on my YouTube that means nothing [like that] will show up’ (P9), whilst another explained: ‘If I see a really thin woman I consciously scroll away from it’ (P4). What was more striking was how assumptions about the interpretative frameworks of the participants were also used by others – from medical professionals to well-meaning peers – to effectively prescribe or censor access to media content. So one respondent recalled how ‘I was told by like a doctor or someone from very early on [in treatment for anorexia] like “don’t read any of that stuff”’ (P9), whilst another explained:

One of my friends emailed me and said ‘so I’ve seen this really great film called God Help the Girl [2014] which I think you’d love, but I’m not sure I should let you see it until you’ve recovered’ [my emphasis]. (P12)

Pervasive assumptions about ‘suggestible’ subjectivities clearly led people to issue ED sufferers with their own trigger warnings which, rather than offering the choice that supporters of trigger warnings espouse, sail close to the argument that they ‘pathologi[ze] … the reading practices of already marginalized groups, and ignor[e] their capacity to read and view critically’ (Kyrola, 2015, p. 140).

Indeed, whilst feminism has been central to the origination of trigger warnings, feminists have also raised concerns about their potential ‘mis-use’, as well as the ways in which they are themselves imbricated within the production (and not simply the description) of audiences’ ‘triggering’ relations with texts (Kyrola, 2015). Whilst triggers are often articulated in highly personal ways, we can also consider how they become collectively ‘seen, recognized and experienced’ as triggering through ‘their circulation and repetitive framing’ (Ibid., p. 137). In this regard, it was clear from the interviews that a persistent ‘voice’ in this framing was medical discourse. This was most starkly articulated by a 19 year old participant with experience of anorexia, and two admissions as an in-patient:

Treatment always tells you to avoid triggers … so I felt for a long time that all these [media] things should be triggering me because I was part of this illness that got triggered by things … I don’t think that that is necessarily the case for everyone, but once you are told that those things set you off … well, I mean, do they? [emphasis in original]. (P11)

That is absolutely not to suggest that what the participants’ describe as triggering experiences are not genuine and intensely felt (Ibid.). But it is to bring into focus how discursive constructions of EDs, including those offered by the media and medical discourse, are also constitutive of people’s realities, experiences and self-constructions (Malson et al., 2004), reinforcing a critique of EDs as an ‘innate’ and individualised pathology. Recognition of this dialogue is entirely absent in previous accounts of how the testimonies of ED ‘subjects’ ‘prove’ the causal and ‘toxic’ power of media culture (Murray et al., 1996; Thomsen et al., 2001) – an omission which abstracts such responses from broader relations of social power.

‘Too close to home’: affect and memory

The most negative discourse on media effects has often pertained to behavioural effects (Barker & Petley, 1997), such as those that I have sought to challenge in the framing of this article. But we can also clearly talk about media texts as having emotional effects, including, for example, shock, sorrow, excitement or disgust (Ibid., p. 34). What is now termed ‘affect’ has arguably been central to feminist theorising (Kyrola, 2014, p. 16), including feminist work on audiences (Bird, 2003; Stacey, 1994), for many years. Furthermore, the increasing turn to affect in feminist research on the body has often been positioned as a corrective to the apparently deterministic prioritisation of representation (Burns, 2009, Kyrola, 2014, pp. 15, 16). Work on EDs has focused on questions of affect and the ‘body-as-experienced’ (Burns, 2009, p. 126, Ferreday, 2003; Malson, 2009; Warin, 2010), in part as a counter to the over-emphasis on the cultural imagery of the body which is prioritised in medical, media (and some feminist) constructions of EDs. This is particularly so with regard to bulimia – which is not ‘visible’ in the same way as anorexia – and some feminist scholars have called for a greater acknowledgement of the bulimic body as a lived experience or process of embodiment which cannot simply be read as a ‘text’ (Burns, 2009; Squire, 2003). I return to these debates, and the specific representation of bulimia, in the next section of the article.

Although the previous discussion of triggering can be seen as intimately linked to questions of affect (Carter, 2015; Kyrola, 2015), I focus here on the ways in which participants often spoke about their encounters with media representations in what could be seen as affective and emotional ways, especially when invoking frameworks of memory. All audience research is essentially memory research (Stacey, 1994), in so far as participants in audience studies are often engaged in recalling media encounters, experiences, reactions and investments. The archive of media memories stretched back considerably further for some participants – such as those whom (like me), developed an ED in the 1980s or 1990s. In conceptualising memory as the active production of meaning (Kuhn, 2002, p. 253), approaches to media and memory within Film, TV, Media and Cultural Studies have occupied a constructivist position in which the ‘reality of the [remembered] incident is less significant than the way in which the incident is remembered by the participant from their present-day viewpoint’ (Ralph, 2011). Also important for my purposes here are the ways in which memories have been used to contest wider metanarratives (Radstone & Hodgkin, 2003, p. 10): one of the reasons that I asked participants if they recalled their first encounter with a media representation of an ED was to explore if discourses other than a pernicious behavioural trigger might be revealed.

A female participant who developed eating problems in her early teens and was later diagnosed with atypical anorexia, recounted her memories of reading the American actress Portia de Rossi’s memoir Unbearable Lightness, which details her struggle with both anorexia and bulimia. She explained how: ‘I tried to read it. But it like related to me so much I thought no, [I’m] not gonna read it … There was too much identification – it was too emotional’ (P4). Similarly, the storyline in which the character of Hannah Ashworth (Emma Rigby) developed anorexia/ bulimia in the British teenage soap Hollyoaks in 2007 was recollected by no less than five respondents in such terms as:

I found it really upsetting and like I wanted to cry whenever I was watching … not being able to watch it when other people were in the room because it was like too close to home …. (P8)

Or:

It wasn’t nice to watch someone go through the same sort of symptoms … The vomiting and the not taking care of herself. When [Hannah’s friend] Melissa died of anorexia I remember being really upset and telling my mum about it. (P14)

Concern has been raised about the conceptually imprecise nature of the term audience ‘identification’ (Barker, 2005), and as Martin Barker observes, it has also been co-opted by effects discourse, in so far as it might be seen to imply an immersion in which ‘audiences los[e] … self-awareness’ and thus become more malleable and ‘suggestible’ (Ibid., p. 357). The fact that female audiences, at a discursive level at least, have often been constructed as lacking an ‘appropriate’ distance from the screen or text (or as over-identifying with what they see in terms of an ‘excessive’ emotional involvement) (Doane,1987; Modleski, 1982) is thus no coincidence here – further consolidating discourses on female audiences as uncritical and ‘passive’.

But the responses from the participants intervene in complex ways in such constructions. On one level, what comes through in each of the responses above is indeed the suggestion of an unbearable closeness with the character or narrative in the text – in which it was painful or shocking to see aspects of their own experience externalised. In the second response, this is related specifically to the context of viewing – the familial home – at a time when the participant’s bulimia was undisclosed. But all three responses also articulated their memories as less encounters in which they ‘lost’ themselves in the text (or as experiences which produced or perpetuated destructive behaviours), than as a kind of ‘reality check’ in relation to how they gauged and acknowledged their own distress at the time. This is explicitly suggested in the third response which hints that the storyline led to an expression of grief by the participant in which the dangers of anorexia were both confronted and articulated. This highlights how the idea that media representations of EDs might have positive pedagogical implications has been entirely ignored in the deterministic emphasis on their harmful and triggering effects. Furthermore, what is striking here is that opposed to the divisive discourse on media images of EDs making female ‘ill’, these accounts foreground recognition, connection and empathy between women: from reader to book, or viewer to screen. This is less in terms of the medical or psychiatric emphasis on girls/ women lacking an ‘appropriate’ distance from the representation (or wanting to narcissistically ‘become’ it), than it is in audiences finding momentary glimpses of recognition in a media landscape that has so often pathologised or rejected them.

‘Really authentic and raw’: judging representations of EDs

Even whilst describing affective investments in particular representations, participants would simultaneously talk reflexively about their construction: so the Hannah story in Hollyoaks was also seen by one of the same respondents above as unrealistic (and wanting) because ‘it was very like, “oh we’re finished with this storyline now – let’s do something else …”’ (P8). This was especially so in terms of what were seen as the potentially constraining influence of generic conventions, or the status of the text as a commercial product. Such knowledge and reflexivity seems far from surprising in the context of Media and Cultural Studies work on audiences (Ang, 1985; Barker & Brooks, 1998), and it only gains significance in relation to the medical and popular discourses which position ED sufferers as media ‘dupes’ – a construction which the evaluative nature of the participant responses functions to critique.

For example, this emphasis on reflexivity and the status of the ED story as a media product often emerged in relation to my question about what, if anything, was absent or marginalised within media representations of EDs. In this regard, nine participants foregrounded the more mundane (but difficult) experience of the recovery process which ‘just doesn’t have popular appeal in the same way’ (P9), isn’t ‘shocking enough to sell papers’ (P10), or ‘just doesn’t make a great media narrative’ (P3). But in the formation and articulation of these judgements, there were clear differences between the participants: those who did not fit the category of (‘typical’) female anorexia felt that they had a less expansive array of media constructions to judge and appraise – a fact that cautions against the homogenisation of both media representations of EDs, as well as the participant responses in this study. Such respondents talked openly about what they saw as the continued predominance of anorexia in the media (especially coverage that featured images of the emaciated female body) and the impact of this on their own self-conceptions. This was particularly so for the participants with experience of bulimia.

As discussed, bulimia is seen as being less clearly etched on the contours of the body – a factor that has encouraged its neglect within media, medical and feminist discourses, each of which have tended to read EDs through the inscribed, visible surface of the body (Burns, 2004; Malson, 2009; Squire, 2003). The participants with experience of bulimia shifted between insisting that their problem was indeed harder to represent or visualise (‘there is no symbolic image, aside from perhaps a toilet’) (P5), to a concern that to do so would be more irresponsible, thus acknowledging the possibility of the trigger discourse. Relating back to the status of the ED representation as an economic product, others felt that bulimia didn’t have the right kind of commercial shock value, and as one explained:

[Bulimia] is seen as being more visceral, the vomit and the excrement – the ways in which people shed their calories. It’s shocking to see an anorexic body, whereas people would just be disgusted to see the contents of someone’s stomach. (P3)

Bulimia is constructed here as almost too ‘transgressive’ to be represented – a theme that has been taken up in the limited amount of feminist work in the field (Burns, 2004; Squire, 2003). In drawing on work on abjection (Kristeva, 1982) and ‘pollution’ (Douglas, 1966) for example, Sarah Squire discusses how bulimia is seen to threaten the ‘tightly controlled boundaries’ of both bodily and social orders. Not only is it signified by the ‘movement of bodily matter … threatening the established boundaries of body and self’ (2003, p. 21), but the cultural disgust surrounding bulimia – and its relative invisibility – is inextricably imbricated within wider cultural norms of femininity (Burns, 2004; Squire, 2003, p. 21). Indeed, several participants, but especially those with experience of bulimia, explicitly discussed the ‘binarized hierarchy’ (Burns, 2004, p. 272) within which anorexia and bulimia are often caught. Within this dualism, anorexia is often constructed as signifying self-control, persistence, obedience, transcendence and purity, as opposed to bulimia’s association with chaos, moral weakness, sexual promiscuity and ‘mess’ (Burns, 2004). Feminist work on both medical and media constructions of EDs has explored how, even whilst anorexia is itself pathologised, such an opposition functions to prefer and reject particular ideas about desirable and normative femininities based on restraint, fragility and purity (Burns, 2004; Squire, 2003; Whitehead & Kurz, 2008). As one participant with experience of bulimia commented: ‘It’s like I’m less of a woman you know? I feel like I should be able to restrict and deny, but I can’t [my emphasis]’ (P17). This suggests the power of such constructions to mark out acceptable female norms – even whilst the ‘binarized hierarchy’ between anorexia and bulimia was recognised and evaluated. But such an apparent transgression of dominant cultural discourses on femininity could also be positively embraced by others: in answer to my question about how she would represent bulimia, one participant (who was undertaking a PhD in performance art) excitedly explained how it would be deliberately confrontational and include ‘big vats of fake vomit’ (P2).

The discussion of soap opera in the previous section offered responses which might be read in terms of Ien Ang’s concept of ‘emotional realism’ – an investment in the portrayal of recognisable subjective experiences. Although the participants critiqued aspects of the narratives as unrealistic (and over-determined by generic or commercial concerns), they still found truth of feeling in the Hollyoaks narrative (Ang, 1985). With regard to a wider appraisal of the extent to which media representations of EDs were thought to be ‘realistic’ or ‘authentic’, most participants, as the discussion above implies, also drew more widely on their own experience of an ED, even whilst they also looked outside of this to comment on types of omission or bias.⁶ The ways in which media audiences judge the reality – or in relation to fiction, verisimilitude – of media texts has been discussed within Communication Studies (see Busselle & Greenberg, 2000), as well as work on media audiences (Ang, 1985; Barker, 2005). Although the emphasis on personal real life experience has been considered within such frameworks (Busselle & Greenberg, 2000), work on the construction of reality/ realism has more often argued that claims and judgements in this regard are made in relation to other media texts. Richard Armstrong refers to this as a ‘a sort of sliding scale’ in which judgements of realism and reality are ‘defined more or less by the [texts] … which came before it and went after it’ (Armstrong, 2005, p. x).

Both of these perspectives were evident in the interviews. For example, blogs were often foregrounded as more real in direct comparison to women’s/girl’s magazines (discussed below). In fact, what is interesting here is how texts or genres which have historically been labelled as women’s/girl’s culture – such as teen and women’s magazines and celebrity coverage – were recurrently positioned at the ‘bad’ end of the continuum where the truth value of media representations were concerned. The equation between mass culture and the feminine clearly has a long history, as the framing of this article has explored. But what is crucial here is that the metaphor of consumption, and thus the analogy between eating and reading/ viewing in which the subject ingests ‘trash’, ‘junk’ or ‘rubbish’ (Radway, 1984), is central to such gendered constructions, including the allegorical constructions of EDs as ‘reading disorders’ (Bray, 2005, p. 117). As a result, and in something of a double ideological manoeuvre, largely female bodies are seen to be ‘eaten away’ (Ibid.), not simply by media culture, but by those popular forms most associated with girls and women.

In reflecting this pervasive devaluation, many of the participants implicitly distanced themselves from this fare, variously positioning such representations as being both irresponsible and ‘unreal’. In contrast, the representations valued as more realistic (and useable) were personal blogs about EDs, which appeared in eight of the respondent accounts. Often defined by the participants as ‘self-published media’ and considerably ‘better’ than the media forms that preceded them, one respondent talked about the posts by Clare Greaves on the Huffington Post, which detail her experience of having an ED, as well as her experience of self-harm:

[They offer] a really authentic and raw kind of brutal experience …. Her articles vary from week to week. Sometimes she is doing really well, and other weeks really struggling. She talks about the thoughts she’s having … A few weeks ago she ended up just taking herself to an eating disorder unit as she was worried that she would just do something. All of these things I think are quite subversive, and quite dangerous, but it’s the first time I’ve really come across someone being able to speak like that …. (P11)

This response, among others, offers a complex insight into why such media forms might be claimed as realistic, welcome or significant. In sanctioned or dominant media narratives of anorexia, particularly those aimed at the female audience, it has historically been the voices of those who have recovered, or at least semi-recovered, who are encouraged to contribute to cultural and media discourse on anorexia as a ‘problem’. Within this framework, subjects should exhibit a certain level of physical and temporal distance from their ‘anorexic’ self, and certainly renounce its destructive implications. This in turn exemplifies the ways in which the anorexic voice is positioned ‘outside of the true’ (Saukko, 2008, p. 6). Yet it is perhaps precisely the ambiguity of Greave’s status in this respect that the respondent finds ‘subversive and dangerous’, or ‘the first time’ she has ‘really come across someone being able to speak like that …’.

Whilst self-published media might well be positioned here as a valuable chance to ‘speak back’ to more traditional media representations of EDs, it did not consistently emerge as such across the interviews. In articulating perspectives not radically dissimilar to the press discourse on the topic (e.g. Bingham, 2015), seven participants expressed concern about the relationship between EDs and social media in which young girls were imagined as both the key produsers and recipients of such interactions. But unlike the press discourse, which tends to foreground people being ‘indoctrinated’ by pro ED material, the participant talk focused on what was presented as pro-recovery material in all its forms.⁷ So one respondent described the genre of ‘My eating disorder journey’ on YouTube, which often focuses on experiences of anorexia:

I think that the very sleek production of some of these videos is misleading. [They tend to follow a structure which goes] …. ‘my life was great! How wonderful was my life!’ and then ‘How terrible was my life! [with an ED]!’ and then ‘How wonderful is my life again!’ Then it’s ‘here’s a picture of porridge, and salad, and oh, I really don’t look like I’m better at all’. Those kind of videos cement in so many ways the shame of people whose ED doesn’t fit this very neat, frail ‘protect-me’ type thing. (P9)

Unlike the discussions of the blogs above, the participant frames the videos as (re)narrowing and standardising a ‘polished’ representational framework for anorexia that is elsewhere associated with the mass media. It also worth noting here that these stories, which conform to the medical narrative on EDs, are often produced by white Western girls from predominantly affluent backgrounds – further foregrounding their partiality, and links to ‘old’ media representations of EDs (Holmes, 2016).

In other interviews, it was the phenomenon of pro-recovery material on the image-based social network site Instagram that was a topic of significance, attracting such comments as ‘I can’t tell you how many people have an Instagram account based around their recovery. That really worries me’ (P11); ‘users almost value their recovery by how many followers they have’ (P16); ‘I would see it as people colluding … posting pictures of how they are struggling’ (P7) or ‘teenage girls are like … addicted to it, posting all day’ (P17). These constructions in turn enable respondents to occupy a flattering position of power – a stance so regularly denied them in constructions of their own media subjectivities – in which they can comment on (and see through) what they see as the ‘faux’ nature of such recovery cultures and identities. Such a knowledgeable position is always occupied, in symbolic terms, at the expense of someone else, and it is again the devalued subjectivity of the female adolescent that is at stake here. This not only returns us to where this article began, but it also offers a clear reminder of the complex and pervasive ways in which popular discourse on EDs helps to constitute people’s perceptions or experience of eating/body distress, even whilst such constructions may be evaluated, rejected or critiqued.

Conclusion: (un)twisted

The central aim of this article has been to give space to the participants’ complex negotiations with media discourses on EDs in ways which challenge the simplistic perception that experiences of eating/body distress are a reaction to the ‘weight’ of media images. In doing so, the article has aimed to widen the discursive repertoire of how such relationships are commonly conceived, including what are in many ways quite routine and predictable ways of engaging with media forms (at least from the perspective of Audience Studies). This can be perceived as a strategy of ‘ordinarisation’ in so far as the media responses offered by the participants encourage us to situate them within more everyday understandings of media engagement that are very far from ‘twisted’, ‘aberrant’ or extraordinary.

But although this may be indicative of the ways in which ‘local processes of knowledge production’ have the potential to critique and challenge ‘more widespread and institutionalised discourses’ (Day & Keys, 2008, p. 4), this does not necessarily equate to a transfer of social and cultural power, or a clear material disruption of the discourses at stake. As this article has explored, the bid to give ‘voice’ in this regard cannot unfold in a discursive space untainted by dominant conceptions and assumptions. Despite the shift to seeing eating problems as multifactorial in origin, constructions of EDs as media induced retain a pervasive currency. As the interview responses make clear, such constructions have significant implications for the participants’ identities and self-conceptions on a number of different levels, even whilst they are negotiated and evaluated. Furthermore, in terms of the wider political implications of this topic, the potential ideological work performed by such discourses may span far beyond the construction of a ‘vulnerable’ or ‘suggestible’ media subject positon. Indeed, the construction of EDs as a form of ‘reading disorder’ (Bray, 2005) is a pathologising label and as such, it should be placed on a continuum with the wider pathologisation of women’s eating and body distress, and the attendant connotations of both perceptual or mental ‘distortion’. Pathologising labels can be ‘used to strip people of their civil liberties’ (Day & Keys, 2008, p. 8) – a fact worth remembering in the context of ED treatment which is frequently involuntary and punitive, and in which the majority patient demographic remains teenage girls (see Gremillion, 2003; HCSIC, 2014).

But this article also has implications for feminist work beyond the specific terrain of EDs. The imagined media reading practices which this article has contested form part of a larger cultural history in which female audiences have been positioned as ‘infantilised’, and in which media culture has been assumed as ‘detrimental to their minds and bodies’ (Bray, 2005, p. 120). The construction of the feminised ED ‘subject’ has endured as one of the most hyperbolic representations of this subjectivity and as such, it should be a focus of concern for media work on audiences, especially that originating from a feminist point of view.

Although I have written this article at time when, in the context of the web 2.0 landscape, the very use value of the term ‘audience’ is under attack from a number of media scholars (Nightingale, 2011, p. 2), it attests to the ways in which public and popular discourses on ‘audiencing’ continue to matter. As Virginia Nightingale suggests, ‘being an audience is essential to cultural participation, the ways in which it effects how we understand ourselves (our identity) and our power to control the world around us …’ (2011, p. 1). The right to ‘participate’ in this regard, without gendered forms of pathologisation, marginalisation or juvenilisation, is very much worth fighting for.

Notes on contributor

Su Holmes is Reader in Television at the University of East Anglia. She is the author of British TV and Film Culture in the 1950s (Intellect, 2005), Entertaining TV: The BBC and Popular Programme Culture in the 1950s (MUP, 2008) and The Quiz Show (EUP, 2008). She is also the co-editor of Understanding Reality TV (Routledge, 2004), Framing Celebrity (Routledge, 2006), Stardom and Celebrity: A Reader (Sage, 2007), and In the Limelight and Under the Microscope: Forms and Functions of Female Celebrity (Continuum, 2011). She has recently been working on feminist approaches to eating disorders and has articles published on this topic in Participations, International Journal of Cultural Studies and Feminist Media Studies.

Disclosure statement

No potential conflict of interest was reported by the author.

Acknowledgements

With thanks to the 17 participants: it was a great privilege to listen to your stories and opinions. With thanks to beat for assisting with the recruitment process.

Notes

1. Eating Disorder Not Otherwise Specified. Those diagnosed with EDNOS may display many of the symptoms of anorexia and/or bulimia, but are seen as not fitting the full criteria for either. It is estimated that 40–60% of people who seek treatment for an ED fall into this category (http://nedc.com.au/ednos).

2. https://www.b-eat.co.uk/assets/000/000/072/BeatMediaGuidelines_original.pdf.

http://www.mindframe-media.info/for-media/reporting-suicide/Downloads/?a = 5140.

3. EDs linked to obesity were not considered as they are often associated with different cultural representations and assumptions.

4. In medical terms, someone diagnosed with atypical anorexia fits much of the criteria for anorexia, aside from the very low BMI.

5. I have replaced names with a participant number (i.e. P1) which reflects the order in which the interviews were transcribed.

6. For example, the predominance of white females in representations of EDs was noted by most of the participants.

7. This probably reflects that fact that many – but not all – of the participants had been through some process of recovery, and had engaged with or searched for recovery material online.