Young Women’s Mental Illness and (In-)visible Social Media Practices of Control and Emotional Recognition

Abstract

What “counts” as a mental illness–related image matters. Most research attention has focused on distressing or recognizable mental illness–related visual practices, yet this offers partial insight into youth mental health. Using visibility and practice theories, I share an in-depth case study exploring the social media practices of four young women, aged 14–17 years, engaged with an Australian adolescent psychiatric service. They describe how being visible to others on social media potentially produces anxiety and burdens them to respond to others’ questions or unhelpful support. In response, they engage in practices of control to manage the vulnerability of mental illness and burdensome sociality. Their mental illness–related media practices are often invisible; they rework mental illness through ambiguous, supportive or humorous practices or, through imagined intimacy, engage with images that feel relatable to them even if the images do not depict recognizable mental illness content or employ recognizable hashtags or titles. These insights complicate “what counts” as mental illness–related content or practices on social media and challenge researchers and practitioners to consider the sociotechnical contexts that shape young people’s mental health.

What “counts” as a mental illness–related image matters. Most research attention has focused on distressing or recognizable mental illness–related visual practices. However, this offers partial insight to how young people “do” mental health online. Feuston and Piper (2018) argue that research exploring mental illness–related posts on social media often adopts a representational approach that “offers a narrow, captured view,” one that directs researchers toward an algorithmic “coded gaze” that attempts to classify, quantify, and ultimately predict user practices (p. 2). For example, this gaze may analyze mental health-related content collected by a shared hashtag on Instagram or track mentions of psychiatric terminology on Twitter. This gaze potentially misinterprets dynamic and diverse digital practices as well as changing platform cultures. Equally, it directs researchers and practitioners toward particular youth practices that engage with mental health or illness (such as self-injury or suicidal content) at the expense of more quotidian media practices and what might influence these. Although this attention is warranted as a means of supporting young people through mental ill-health, researchers and practitioners who rely on limited insights may ignore what might be more important or relevant to a young person and their mental health.

Moving beyond this approach, this project explores the social media practices of four young women, aged 14–17 years, engaged with a metropolitan Australian adolescent psychiatric service (APS), through an in-depth case study employing image collection, interview, and social media analysis methods. Using theories of visibility and media practices, I consider how visibility on social media produces anxiety, both lived and potential, and how this influences their media practices. Here, images and practices are not easily “seen” or “visible” as mental illness–related; practices are “invisible,” and afford a feeling of emotional recognition, affective community or intimate co-presence that I call imagined intimacy. I focus on visual presence and visibility, not because media approaches are dominated by the visual, but as it attends to what it means to “be seen” as unwell and a young woman. Visibility here is as much about the potential for being seen and its flow on potential affects (such as distress, inadequacy, anxiety, fatigue, etc.) as it is about actual or lived experiences of being seen.

Visualizing Mental Illness

Research to understand young people’s experiences of mental ill-health and illness typically relies on psychological and biomedical frameworks that often erase the heterogeneity of young people’s social, emotional, and material experiences. Even at clinical level, mental illness is an obtuse concept that often relies on shared knowledge, local practitioner experience, and individual and community beliefs, attitudes and values (Ringer & Holen, 2016). In this article, I use the term “mental illness” to engage with the personal ways the young women in this study chose to identify; this is not to erase the complex, and often inequitable or discriminatory histories and realities of psychiatry and psychology worldwide. While honoring the participants’ preferred term of mental illness, I acknowledge that it also reflects how young people’s anxiety may be psychopathologized or subject to clinical therapy in unhelpful or abusive ways.

Social media provide young people with the opportunity to make “visible” their experiences of mental illness. As Fullagar et al. (2017) outline, social media hold an informal pedagogical potential as they circulate ideas about mental health risk and illness that are entangled with diagnostic cultures, clinical expertise, and activist movements. Tembeck (2016) highlights “selfies of ill health” as a vernacular autopathographic genre where illness selfies visibly and publicly perform what would otherwise be an invisible illness. Similarly, depression is made visible through depression hashtags that link content, likes, comments, and shares, adding value to distinct images as they become a searchable group (McCosker, 2016). Of the literature that explores social media visibility and mental illness, most attends to self-injury-related practices afforded by platforms and media, suggesting mediated self-injury images are a nuanced and contextual way of “seeing” mental illness (Seko & Lewis, 2018; Sternudd, 2012). This work challenges assumptions that self-injury is inevitable as a result of mediated social contagion or “being triggered.” Rather, the possibility of multiple and even contradictory feelings is attached to these practices without precluding triggers or distress, including allowing individuals to “make . . . the inner experience visible” (Sternudd, 2012, pp. 14, 26). These practices center the emotional or affective experience of living with mental ill-health.

Yet the visualization of mental illness on social media may not be direct, explicit, or easily interpreted. What becomes “counted” as a visual representation of mental illness may suggest more about how researchers understand mental illness than how people engage in visibility practices online. Analyzing self-harm posts on Tumblr, Seko and Lewis (2018) describe a shift from “direct depictions of self-injured bodies to re-appropriations of popular media content that figuratively represent emotional struggles” (p. 180). Most self-injury tagged posts in their study do not explicitly represent self-injury practices but rather convey emotion through images and content from popular culture that rarely relate to self-injury or even mental health. The platform affordances and vernacular cultures of Tumblr further shape how users express and recirculate self-injury content; reblogging images and memes re-appropriates visual content unrelated to self-injury, and expresses the multiplicity of emotional experiences of self-injury rather than an explicit aesthetic of one’s own cutting. This reflects research investigating other Tumblr communities where users recirculate “often ambiguous, fragmented and multimodal” (Gonzalez-Polledo & Tarr, 2014, p. 5) content on and between blogs to engage with emotional rather than narrative dimensions of experience (Cho, 2015; Hanckel et al., 2019; Kanai, 2017a). Studying mental health on Instagram, Feuston and Piper (2018) use multiple digital ethnographic and interview methods to understand how Instagram users share mental health and illness experiences. They recognize that what “counts” as a mental illness–related image or practice is already coded, classified, or defined through hashtag or keyword searches where certain images or content will be classified as relevant (e.g., tags, such as #depression but not #journey). As per my introduction, this may misinterpret individuals’ and communities’ practices.

Visibility and Affective Media Practices

Visibility refers to “the ability to be seen” and also “symbolic or representational visibility, which is about the legitimacy of some phenomena, people, groups or objects to be noticed, perceived or articulated” (Tiidenberg, 2018, p. 13). Brighenti (2010) identifies visibility as a relation between aesthetics (the visual) and power (politics) where being seen and seeing others is not necessarily symmetrical (or reciprocal) and involves the imagination or perception of being visible. This does not necessarily presuppose visibility as positive or productive, and its inverse, invisibility as undesirable; we can imagine the value of not being visible in a broader public (e.g., the pleasure or safety of being in a “hidden” counterculture) or the risks, anxiety, or violence that emerge from being visible in an unwelcoming public sphere.

On social media, mediated visibility of oneself necessarily extends beyond selfies and bodies (Tiidenberg & Whelan, 2017), and encompasses symbolic representation as well as how a person presents themselves “to be seen as someone” (Tiidenberg, 2018, p. 64). Berriman and Thomson (2015) suggest that “young people are driven by a dual emotional imperative: seeking to navigate between the potential emotional pleasures derived through praise and recognition, whilst simultaneously attempting to avoid the anxiety and distress of being exposed to criticism and derision” (p. 13). Making oneself visible is a practice in wanting to “be seen as” someone in a particular light, such as a friendly person, a capable person who copes well with challenges, or a person blending “into the crowd.”

For young women in particular, visibility encompasses more than simply the presence of their bodies on social media. Dobson (2015, p. 162) argues, drawing on Harris (2004), that young women in the Global North are culturally urged to make not only their bodies visible, but also their inner selves, “to make themselves known,” while at the same time are regulated and derided for being “attention-seeking” (Dobson, 2015; Senft, 2008). To Burns (2015, p. 1730), selfie practices in popular discourse are framed as feminine and devalued as trivial, narcissistic, or excessive, reflecting wider social discourse that “enforces [women’s] compliance with social norms and governs the participation of women within the online public sphere.” Online participation requires young women to carefully balance participation within expected social norms or become ridiculed, excluded, or harassed.

Visibility online is managed through different practices to negotiate exposure and how people are seen by others. Drawing on Marwick and boyd’s (2011) “context collapse,” Litt and Hargittai (2016) articulate how social media users negotiate how media flattens multiple audiences into one through savvy attention to one’s imagined audience by “thinking about a general abstract audience or by thinking about a more targeted audience” when posting (p. 9). Miller (2016) identifies this as scalable sociality where users choose different platforms or channels to connect with different groups or content. For Abidin (2016), this is “visibility labour,” the often tacit and analogue work of managing social media visibility whereby Influencers and their followers “self-posture and curate their self-presentations so as to be noticeable and positively prominent” to various stakeholders through practices such as tagging or re-gramming content on Instagram (p. 90). The logic of this labor may value the amplification of Influencer content, but does not assume that all visibility labor necessarily aims to increase the visibility of content. For example, the indiscriminate labor of “spammer” followers in Abidin’s research were perceived as too visible and their efforts were counter-productive to being noticed favorably by Influencers. Visibility in this way is a series of affective practices that focus “on the emotional as it appears in social life and tries to follow what participants do” (Wetherell, 2012, p. 4). An affective practice “typically pulls together or orders in relation to each other patterns of body/brain activity, patterns of meaning-making, feelings, perceptions, cognition and memories, interactional potentialities and routines, forms of accountability, appraisals and evaluations, subject positions and histories of relationships” (p. 236).

Practices direct attention to media platforms, their affordances, and associated platform vernacular (Gibbs et al., 2015), mutually constituting spaces and incorporating affect and feelings. Here, affordances align to Nagy and Neff’s (2015) proposition of “imagined affordances” to demarcate the affordances that “emerge between users’ perception, attitudes and expectations; between the materiality and functionality of technologies; and between the intentions and perceptions of designers” (p. 1). Similarly, vernacular affordances—the shared and individual ways people understand affordances—influence users’ practices on interconnected, multiple levels of scale, including affordances at the level of infrastructure, app, device, feature, and interface (McVeigh-Schultz & Baym, 2015). This limits how individuals negotiate control or visibility on social media; practices are constrained by platform functions, interfaces, algorithms, and infrastructure.

Methods

The case study design (Hendry, 2018) was informed by feminist media, digital ethnographic, and visual methodologies toward everyday media practices (Pink et al., 2016; Rose, 2014; van Zoonen, 1994). Case study approaches investigate phenomenon “within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident” (Yin, 2003, p. 13). The narrow, in-depth scope of this study contrasts to public health studies with larger samples to instead more closely follow the media practices of a small number of participants. While the study may be critiqued for its limited size or capacity for generalization, case studies do not claim to be representative but rather “generalizable to theoretical propositions” (p. 10).

I recruited participants from an outpatient program of a public APS in Melbourne, Australia. The APS compromises inpatient and outpatient psychiatric services, and outpatient programs for young people aged 12–18 years. Reasons for admission vary, but APS uses clinical and psychosocial measures to assess mental state, risk (e.g., harm to self or others), and family and social context.

During May to September 2015, I invited young people (regardless of gender) aged 14–17 years to participate by posting flyers around APS, attending program sessions to introduce myself, and asking APS practitioners to share flyers. Potential participants contacted me via email, text message, or a website. Parents or carers provided formal consent for young people to participate and participants completed assent forms to build their understanding of the project. Participants were paid for each stage of the project (AUD30); this mirrors emerging best practice for mental health services as progressively more peer workers and advocates are, rightly, being compensated for their contributions (Seymour, 2013).

I collected data between June 2015 and March 2016, using multiple, staged methods to create case studies of each participants’ practices. In the first stage, I asked participants to email five to eight images related to “#identity” (this prompt was determined during a pilot study involving practitioners and young advocates) prior to our first interview. Participants were encouraged to interpret the task in their own way. I used the images as a “probe” (Boehner et al., 2012) during the first interviews to prompt reflection and “make things visible” rather than as a representation of participants’ identities (Rose, 2014, p. 27). Later, semi-structured and unstructured interviews were held at local library meeting rooms, offices at participants’ schools, or closed rooms at APS. When participants allowed, I also collected data from their public social media accounts (downloading images, taking screenshots, copying text or URLs, or taking written notes).

The final corpus of data included the #identity images from each participant; interview audio-recordings and transcripts; digital copies of images shared in interviews (e.g., screenshots or downloaded); interview notes (e.g., description of an image a participant was showing me); emails from participants; screenshots and hyperlinks from social media accounts; and notes from each data collection stage (as reflexive analytical memos; Snyder, 2012). I checked the accuracy of transcripts and de-identified all data (e.g., surnames, school names, friends’ faces in Instagram photos). I used a paper-based method for analysis, printing data to physically and visually “see” the links between groups of data. Employing Attride-Stirling’s (2001) thematic network analysis for both visual and textual contents allowed me to explore the data at different levels and “thematic networks [that] aim to facilitate the structuring and depiction of themes” (p. 387).

The project was reviewed by the hospital Human Research Ethics Committee (HREC) that services APS from May 2014 to April 2015. As per the National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council [NHMRC], 2014) guiding Australian ethics review, the project was classified as high risk as the project explored mental illness and participants were under 18 years. Beyond institutional requirements, ethical concerns persist even as I write this article. I use pseudonyms chosen by the participants to maintain confidentiality. Following Markham’s (2012) discussion of fabrication as a practice to protect privacy, I have fabricated or modified quotes and media content. This approach also reiterates that research articles are always “partial accounts and snapshot versions of truth—our best effort to encapsulate for particular audiences the studied experience of everyday life” (p. 341). Each participant will have aged out of the service this year and this temporal distance disallows APS clinicians to identify the young women in this article, adding another layer of confidentiality.

Negotiating Mental Illness and Visibility

Introducing Claudia, Padawan, Daisy, and Beatrix

Akin to the demographics of young people engaged with APS and similar services in Australia, my four participants were young women of White Anglo ethnicity, mostly middle-class, and living with at least one parent. Most of the four women had experienced at least one inpatient APS hospital admission. All had significant absence from school, and three had attended an APS school program to support their return to school, employment, or an alternative education setting. Collectively, their list of mental ill-health experiences or diagnoses include childhood trauma, anxiety, depression, sexual abuse, emerging personality disorder, self-injury, suicidality, disordered eating, and bipolar disorder. ¹

Claudia (17 years) had performed in youth theater and created and shared funny videos and images on Instagram: performing in costumes with her puppy or acting out impromptu skits. She preferred not to use other typical youthful platforms, such as Facebook or Snapchat, as she was cautious about her visibility to others, even though she chose the predominantly visual Instagram as her one social media account. Padawan (14 years) was a highly visible and frequent social media user. She maintained intimacy with friends through gestures of care on other platforms, sending hilarious memes and ugly selfies on Snapchat to cheer up her cousin. She stressed how important it was to find queer communities on Tumblr but how it was, at the same time, confusing to her. Padawan was critical of “middle-class mothers” who do not “get” social media and are offended by mental illness or sexual diversity content. Daisy (14 years) was the most publicly visible on social media—Pinterest in particular—in supporting other young people struggling with mental ill-health. She had experienced bullying at school that extended into Instagram. On Pinterest, she collected and pinned encouraging memes and inspirational images to promote mental health awareness and often posted messages to users she noticed were struggling. Similarly, Beatrix (16 years) avidly collected, curated, posted, and reblogged “emotionally authentic” content on her (now 8 years old) public Tumblr blog. She collected so many images—her mobile phone storage was at capacity—that she deleted apps to create more space. She maintained a highly controlled and curated presence on Facebook to avoid friends’ questions or offers of support.

Next, I discuss how each young woman negotiated her own visibility and that of their mental illness experiences. First, I examine how each young woman expressed feeling anxious about the visibility of mental illness and being visible to others. I then outline their strategic practices to control visibility and finally their “invisible” media practices to feel a sense of emotional recognition.

The Anxiety and Burden of Visibility

Although each young woman engaged in different social media practices on different platforms, they all stressed that being “visible” online—exposing their faces or bodies or sharing emotional distress, confusion or pain—potentially produced anxiety, stress, and worry. They were concerned about how they were visible to others and who was visible to them.

For example, Claudia felt pressured as a young woman to present “well” to others while experiencing “severe anxiety.” She emphasized that “if you’re honest [about your mental illness] you get backlash,” therefore, it was “safer not to post.” Likewise, Daisy described a similar tension: selfies posted because she “feels good” received either encouragement from others or being told to “get over yourself . . . you’re so stuck up.” It was difficult for her to balance the “double standard” that emerges when young women are judged about their online presentation, but Daisy reinforced that “as long as you’re happy it’s all good in the end.”

Visibility on social media sometimes meant responding to others’ questions and offers of support. Although this was not necessarily critique or harassment, they often rejected these connections. Padawan emphasized that social media “can be just too much,” as “everyone’s on it, and everyone’s like talk to me, talk to me, you know, or look at me and all that.” This undesired and unintentional attention required balancing between managing her visibility and her perception of others’ wanting to talk with her. She “just feel[s] so pressured” by her friends as they expect her to talk to them “constantly” on Facebook Messenger even though she sees them at school. On Messenger, her Friends can see a visible, green circle next to her name on the platform, even though she is more likely to be looking at something funny in another browser tab. Being visible miscommunicated that she was “available” to her friends. Similarly, Beatrix worried about replying to an encouraging message from an unknown Tumblr user in case they knew her outside of Tumblr. Because the message was anonymous (and she did not have any known friends on the platform), she felt she did not need to answer it, “That’s another thing with my anxiety, if someone messages me and it’s anonymous, I’m like ‘Oh my god, what if it’s somebody I know?!’ [laughs].”

Making the affective experiences of mental illness visible to others—online or off—was difficult or unhelpful. Struggling with emotional distress or anxiety was often a personal, overwhelming experience and others’ attempts to be supportive were inappropriate, burdensome, or exhausting. Even though they could not recall a time when they were challenged or bullied as a result of others’ stigma toward mental illness, they avoided directly making their mental health struggles visible. Beatrix shared that experiencing a panic attack or anxiety was extremely unique, “I don’t know anybody else who has the same level of anxiety and panic that I do. I try to keep it to myself a bit . . . I will just sit there and feel like this big ball of just . . . anxiety, panic, just a big ball of panic.” Claudia expressed that she would not share images about mental illness as it would feel like “following the crowd” and therefore less authentic, and also that she was afraid of the “huge amount of mixed responses”, she would receive because of the stigma related to mental illness. She was comfortable sharing a photo of her broken wrist on Instagram but not sharing when she had a terrible day as she feared “others as to what they might see or what might be revealed.” For her, any potentially supportive comments were unwelcome, “that’s not the kind of support that I want. I don’t feel like I need, I don’t know, 50 people commenting on a photo being like “Here for you any time!,” it’s just so impersonal!”

Each young woman stressed that they preferred to be in control of managing their own emotional distress. They preferred support that did not make them feel like a burden or demanding of others. Visibility of their mental illness on social media increased the potential for others to assume they needed support or care. Shunning others’ care allowed them to avoid negotiating other people’s support which was more taxing than caring for themselves. At the time, they did not place responsibility for their care to their close family or friends, and did not seek help from their peers. Beatrix stressed that young people should also embrace this responsibility as “they need to be taught responsibility about taking care of themselves and avoiding the things, at least for a while, that are going to trigger them.” Likewise, Daisy understood recovery as “stuff you need to do for yourself.” Talking to people had sometimes been helpful but sometimes they misinterpreted her needs in unhelpful ways. She distinguished between feeling understood in comparison to feeling misread:

. . . but parents, they can also turn things into a bigger thing . . . if you said “I’m feeling suicidal,” they feel like “Okay I’m checking you every five minutes, you’re not allowed to go anywhere on your own,” and it’s like just because I feel it doesn’t mean I’m going to do it!

Managing and Controlling Visibility

In response to the potential anxiety and stress of social media visibility, each participant engaged in practices to curtail the visibility of their mental illness and manage how others might connect with them. Practices of control were “visibility labour” (Abidin, 2016) to mitigate imagined or experienced anxiety. These practices included choosing particular devices or platforms; tagging and untagging, as well as deleting content; and managing the visibility of their faces, bodies, and names on platforms.

Choosing Platforms and Devices

Strategically choosing a platform, given its affordances and vernacular, is critical for scalable sociality (Miller, 2016) and for each young woman to manage her visibility. Beatrix distinguished between her emotional “real self” on her pseudonymous Tumblr and her highly curated Facebook profile. Similarly, Padawan described the visibility and pressure to connect afforded by Facebook contrasts with Tumblr, as “there’s a lot more freedom in Tumblr . . . it’s social media but it’s kind of just for me . . . I don’t worry about people waiting for me to message them.” On Tumblr, she could “hide from people [that know her], but still be on social media.” Snapchat also offered her freedom to select who sees her funny selfies and memes, without sharing them with every follower, unlike Instagram where her images could potentially reach her 500 followers and others scrolling past her feed.

Choosing different devices, often those shared among family members or passed around at school, required them to negotiate not only what was shared by platforms or apps but also what digital traces remain on devices. Illustrating this, Daisy preferred to use Pinterest on her mother’s iPad because of its large screen. Yet although “technically it is my mum’s,” she sometimes pretends she can’t find it when her mother asks for it. She carefully deletes search histories or downloaded images she believes her mother would consider “inappropriate” or “harmful.” This protected her mother and avoided Daisy feeling like a burden. Daisy could not recall a time when this occurred, but stressed she “just knew” it would upset her mother although they had not explicitly discussed it. The devices Daisy uses shape her practices, not in regards to “access” as having or not having a personal device, as Vickery (2015) challenges, but access as it influences how she maintains her privacy and mother’s trust even as the device is only temporarily “hers.”

“Just Enough” of Oneself

The visibility of participants’ bodies and faces, as well as their “real” name differed by platform, and allowed them to “negotiate platform expectations and segment audiences” (van der Nagel, 2017, p. 327). For those on Facebook, each woman used their real name and shared at least one photograph with their face. On Instagram, only Beatrix cropped her face out of images, whereas Claudia, Daisy, and Padawan each included their faces and first real names, unlike Tumblr and Pinterest where Padawan, Beatrix, and Daisy use pseudonyms on their more emotionally vulnerable and public accounts. Beatrix’s Tumblr username was a French translation of a word she loves, whereas Daisy describes choosing a contradicting phrase for her Pinterest username to reflect the ambiguity of recovery. Choosing a platform username “signal[s] information about them, what they say, and to who they intend to say it,” highlighting different naming strategies claim agency and control (van der Nagel, 2017, p. 313).

No participant shared an image of her face or body for the #identity task or during interviews, even though data were collected during the height of popular selfie interest in Australia. Yet although each shares her face on at least one platform, they do not regularly share and upload selfies. Sharing “too many” facial photographs was regarded as seeking “too much” attention. This was not determined by an explicit number of images, but a felt sense of exceeding the perceived social norms of a platform. For example, Claudia’s Instagram feed features no selfies but her face in photos and videos, often performing silly faces with her dog. Like Claudia, Daisy’s face features in photos taken by other people or in photos with friends. While Padawan’s Instagram profile is public and lists her full name, her profile shares 10 photographs to her 350 followers. Her face is in only two images, with one fitting a typical selfie definition as she sits taking the photograph in front of a mirror. Beatrix’s only photo of her face is on Facebook even though she uses the platform predominantly for group messaging. Beatrix’s legs appear in one Instagram image next to her book and sunglasses at the beach; she stresses that she shares pictures of places rather than herself or things.

Beatrix shares “just enough” on Facebook including brief status updates or mundane photographs to avoid questions she imagines she would receive if she was absent online (e.g. “what is wrong with you?”). She posted seven photos to Facebook, selecting images that balanced looking good “but not too good” such as her school formal photos. Limiting participation on Facebook (especially where participation is visible to others) also allows Beatrix to share a “heavily edited version of me” that hides who is able to really know her. Even though she only shares images of her face on Facebook (and not Instagram or Tumblr), she shares “just enough” of herself without being too personal or visible to others. Likewise, Daisy and Claudia also shared emotionally and socially neutral images of themselves that were good enough. These images, like photographs of their face, balanced being seen as wanting too much attention from others while also not communicating being unwell to others. “Just enough” participation allows them to connect at a distance with friends and followers without contravening perceived platform norms.

Tagging and Deleting

While tagging and collating content on social media increases the potential for it to circulate within and between platforms and search engines (McCosker, 2016), tagging practices for these young women carefully directed the attention of specific people to content. Padawan recounted that she does not tag or share the content she finds for her Instagram followers or Facebook Friends. Instead, she selects images to tag for “my cousin who lives two hours away . . . I don’t have her on Tumblr, because we have our own Tumblrs, and follow each other on Instagram.” This affords intimate and visible connection to cheer up her cousin when she is feeling low while not including all of her Instagram followers.

After experiencing bullying at school and via Instagram, Daisy’s school co-ordinator encouraged her to delete Instagram to avoid harassment. Instead she blocked and deleted followers to manage her vulnerability. She expressed gleefully that her bullies had not blocked her, and that she is able to follow their public Instagram profiles and see what they were doing. After changing schools, her Instagram feed began to feature more friends. She amplified the sociality of her Instagram images through playful captions and hashtags to reinforce her relationships, such as troupe, “my lovelies” crew and #lovethem. For Beatrix, untagging and unfollowing content on Facebook, alongside presenting “just enough” of herself, allows her to closely control the assumptions her friends might make about her health based on what they see. Yet even as she stressed it was important to edit and update her Facebook page, she does not delete posts on Tumblr—even her first posts reblogging One Direction photos—as “I’m somehow more satisfied with it [her blog] after I stop editing.”

“Invisible” Visual Practices

In the previous section, I highlighted practices that contained how visible each young woman, and their mental illnesses, were to others. I now turn to how they did make visible their experiences, albeit through ultimately “invisible” practices.

Obscuring Mental Illness

Each young woman stressed that as their recovery progressed, there was no value in “obviously distressing” mental illness images and instead they preferred to engage in more complex imaging practices, without hashtagging, collating, or captioning images with mental illness–related text. Akin to Seko and Lewis’ (2018) and Feuston and Piper’s (2018) research, their mental illness images and practices were mostly untraceable, indirect, or unrecognizable, even as these practices were highly emotional or personal. By curating “invisible” images on platforms that allowed anonymity or ambiguity, participants felt relief from the pressures and frustrations associated with being visible. In this way, their mental illness experiences and practices were invisible and hiding in plain sight.

Daisy articulated the significant work of her mental illness “journey” to manage her own chronic anxiety, history of self-injuring behavior, and ongoing school disengagement. Her Pinterest username and bio opaquely referenced her recovery work, referring to aggressively fighting anxiety and the vulnerability required for healing, as “some people think I’m really sweet a lot . . . so it’s like my inner, tough stuff.” Of the four, Daisy circulated the most explicitly “distressing” content, such as self-injury, suicidality, and depression on several of her well organized Pinterest boards, but she used hashtags and board titles that did not relate to mental illness, such as Psychology, Story-writing, and Wounds and Gore. Her choice of board names and tags blurs how these images might be understood. Daisy was not hiding this content but rather genuinely engaged with the pinned images through those board themes.

Each young woman talked about the importance of images that feel relatable for them. Often these images were mundane or expressed everyday stressors using screenshots from TV shows or artistic images of scenery or fashion. For Beatrix, although most of her pink, fashion-related or text-based images on Tumblr were reblogged, she asserted that “just because I didn’t create, doesn’t mean that it’s not mine . . . I picked these specific things . . . together it makes a bigger image” to express an affective sense of her mental illness. Collectively, this bigger image offered a sense of emotional recognition and curated “just who I am I guess, how I feel.”

Vulnerability Through Humor

Humor, whether watching funny clips, scrolling through comedic content or producing humorous images and videos, provided relief from distress, overwhelm, and judgment. These practices were invisible: tracked by platforms and algorithms, but often not shared by participants or made visible to their friends or followers. Humor was a socially appropriate means of sharing distress and narratives of pain as “humor creates value from unpalatable feelings which can then be circulated . . . and is key to defusing the perception that one may be overly impacted by the feelings of frustration, weakness or shame unbefitting a strong, neoliberal subject” (Kanai, 2017b, p. 6). For Claudia, humor facilitated self-esteem and socially acceptable attention from others. She wrote stand-up comedy and finds it useful as long as “it’s not too close and it’s not too personal, or else you just get angry and bitter.” She performs skits on her Instagram—none related to mental illness—yet posted a video on her private YouTube channel performing a funny song about catastrophising thoughts and anxiety at her school’s Mental Health Week. Claudia reflected that feelings that were “too much” were safe in comedy; her emotions were not muted but rather amplified for comedic value. Humorous transgression avoided critique of being seen as self-centered, while remaining agreeably accessible and visible to others’ attention and potential interaction.

Humor also afforded social connection. Baym (1995) identifies the role of humor to create solidarity where a group identity emerges from sharing humorous content or conversation, one that is premised on the assumption of a shared recognition of the layered meanings circulating humor, including humor sharing self-deprecation (Ask & Abidin, 2018). Beatrix shared a comic about the “real life horror” of social anxiety as one of her #identity images. Laughing, she explains it is “definitely something I can relate to . . . it’s done in a way that’s humorous . . . I can look at it and go I feel like this, it makes me sad, it gets me!” Padawan and Beatrix each stressed the value of humor to feel emotionally recognized, even though much of what they engaged with was unrelated to mental illness (e.g., videos of cute animals).

Imagined Intimacy

Each young woman expressed the value of social distance from others to feel emotionally understood and recognized; here, escaping demands of sociality afforded an imagined intimacy. Both Padawan and Daisy stressed that particular platforms, Tumblr and Pinterest, respectively, allowed social distance (as an imagined affordance; Nagy & Neff, 2015) from connecting with others directly, and this offered emotional freedom. Padawan articulated that on Tumblr, “no one can tell when you’re on online . . . there’s a lot more freedom with Tumblr . . . it’s not like a social . . . it’s social media, but it’s kind of just for me, and . . . online friends.”

Daisy explained that unlike Instagram, there were “secret rules of Pinterest” and “you can pin what you want, and pin when you want . . . it’s okay if you pin a lot or pin a little.” Daisy experiences Pinterest as a “silent community” where individual users all repinned content without conversation. She explained that “when you’re going through that stuff you don’t want to help, to get out of it. You want to know that other people understand what you’re going through too and you’re not just really alone.” Daisy’s most unequivocal board, titled Mental Health Help, expressed “understanding, like if you’re going through a hard time, here’s some supportive stuff to help you get through it.” She explained that the board was “good because there are other people out there who are going through tough times, and if you’re helping someone like that, it’s like yesss!” She perceived the visibility of others re-pinning her images as evidence of them as being like her and through this she felt a sense of connection that did not rely on confirming if other users were like her.

Similarly, Beatrix imagined a sense of intimacy and emotional recognition on Tumblr. Her Tumblr blog was an “art gallery” where she did not interact directly with other users but stressed that she felt understood on Tumblr. Likewise, Cho (2015) illustrates the “experience of Tumblr” as “walking through a million different constantly shifting galleries” that circulate affect through “aesthetics, intimation, sensibility, and movement” (p. 46). Engagement with visual content, much like wandering through an art gallery, is unstructured yet forms an affective sense of belonging; for Beatrix, belonging emerged through her posts, other users, and the platform itself.

Discussion

Visibility as afforded by social media is burdensome for each young woman; they negotiate the experienced and potential effects of visibility and how others see them. Engaging in affective practices, they mediate their visibility on social media within the sociotechnical affordances of different platforms. These affordances shape possibilities for visibility, and this must be taken into account when deciding “what counts” for mental health research. As Tiidenberg and Whelan (2017) propose, faces and bodies may not be the only “selfies” users share as “self-representation [becomes] a set of practices, conventions, and norms of content production and consumption” (p. 143). An image of a pink dress printed with a nostalgic phrase or scrolling through hilarious memes may be just as much a mental illness–related practice as posting an image of self-injury cuts. Visibility here comes to reference social visibility as these young women attend to the social dimensions of privacy and participation, rather than their visibility to social media companies or how platforms ambiguously shape their online experiences (see Gerrard, 2018).

The effects of mental illness stigma, self-expectations to be responsible for their own care, and managing feeling burdensome contribute to the anxiety of being visible. Yet the assumption that increased visibility is related to increased anxiety may not neatly carry into social media spaces (Lasén, 2013). Being visible, engaging with emotionally relatable images, and feeling understood on platforms like Tumblr and Pinterest ameliorated anxiety and distress. “Invisible” practices, where mental illness images did not directly depict explicit mental illness themes, conveyed the feeling of living with mental ill-health. This imagined intimacy challenges and reworks Turkle’s (2011) assertion of social media users being “alone together.” Emotional recognition with content affords being together alone, “a sense of comforting togetherness” (Kanai, 2017a, p. 9) and produces an imagined intimacy or feeling of others—presumably other people who create and circulate images and videos—that these young women perceive as like them. This offers a type of support that allows participants to feel in control and centers their feelings and the affective experience of mental illness.

Each young woman’s affective media practices highlight the inadequacy of current paradigms that inform youth mental health interventions that focus on moderating or banning problematic content, while at the same time, encouraging young people to share their “authentic” stories. Understanding the relationship between visibility practices and stigma requires careful contextual analysis. Exposure to stigma always concerns the visible (Goffman, 1991), but the outcome of visibility is never determined prior to recognition (or misrecognition).

The visibility of mental illness experiences on social media allows us to address the underlying challenges young people face (boyd et al., 2009), but if mental illness–related practices and images are obscured or ambiguous, we must first engage with young people’s broader media practices and cultures. If these young women strategically and tactically negotiate their visibility online, albeit in different ways, how might this challenge health promotion efforts that seek to build awareness or promote help-seeking behavior on moderated platforms that require their digital visibility as a condition of participation? How might the increasing visibility of mental illness and anxiety circulating on platforms, and connection with others be something to avoid rather than seek out? How do expectations related to resilience or experiences of care shape these young women’s reluctance to embrace support from others? Their media practices challenge how researchers and practitioners understand mental illness online. “Invisible” mental illness–related practices may not be visible to researchers without the participation of communities of young people living with mental ill-health. How we understand “what counts” as a mental illness–related image or practice must engage with these young people. To take this challenge seriously, future research may benefit from shifting away from starting with problematic media content to explore, with young people, their practices, and experiences. Studying their feelings, attitudes toward, and experiences of mediated connection, care, and support, besides the content they engage with, may better inform mental health initiatives that more productively respond to their lived experiences.