Abstract: The advancement of Web 2.0 technologies has drastically extended the realm of self-expression, to the extent that personal and potentially controversial photographs are widely shared with public viewers. This study examined user-generated photographs of self-injury (SI) uploaded on a popular photo-sharing site Flickr.com, to explore how the photo uploaders represent their wounded bodies, whether there are any emergent discursive and visual conventions that (re)define "photographs of SI," and whether these emergent conventions affirm or resist dominant cultural discourses of SI. 516 photographs of SI uploaded by 146 Flickr members were analyzed using methods of visual content analysis and discourse analysis. The findings indicate that while dominant discourses largely determine the shaping of SI photographs, some uploaders subversively frame their wounds as a narrative of resilience, thereby transforming their wounds into an authentic source of self-expression.

Key words: self-injury; visual content analysis; discourse analysis; social media; Flickr

In recent years self-injury (SI), the deliberate injuring of one's body, has been reported to spread widely among developed countries, especially among youth and adolescents (NOCK & FAVAZZA, 2009; ROSS & HEATH, 2002). Cutting, burning, stabbing and piercing—SI is secretly, yet steadily gaining popularity among maladaptive persons as a viable coping strategy to relieve emotional pain and struggle (FAVAZZA, 1996). Along with the increasing prevalence of SI, Internet use by self-injurers has attracted considerable attention from health professionals, popular media, and the general public. A number of studies have reported the increasing use of the Internet by self-injurers, asserting that anonymity of online communication may provide a perceived low-risk venue to search for information, to find peers with similar interests, and to disclose their feelings (MURRAY & FOX, 2006; WHITLOCK, POWERS & ECKENRODE, 2006). [1]

While early online activities were based mainly on text-oriented interaction, recent advancement of user-friendly Web 2.0 technologies has provided new opportunities for self-injurers to produce and share multimedia content themed around SI. Content-sharing sites such as blogs, Facebook, Flickr, and YouTube increasingly serve as a creative outlet for self-injurers, where they display not only textual, but visual, audio and other multimedia materials. Through these venues, self-injurers began mediating their wounded bodies through photographs, videos, journals and other artistic pursuits, so as to share their experience with online peers. This however proposes a significant concern among health professionals regarding potential health risks for young adults, who have a higher SI rate and greater Internet usage than any other age-group (MITCHELL & YBARRA, 2007). In particular, graphic representation of SI often provokes anxiety among health professionals who tend to deem such materials "suggestive or triggering" (WHITLOCK, LADER & CONTERIO, 2007, p.1139) and carrying a considerable risk for maintaining and exacerbating self-destructive behavior (LEWIS, HEATH, SORNBERGER & ARBUTHNOTT, 2012). LEWIS and his colleagues (2011) conducted a content analysis of user-generated NSSI videos on YouTube and deemed the possible impact of graphic content on viewers as "worrisome" to health professionals (LEWIS, HEATH, ST. DENIS, & NOBLE, 2011, p.e552). [2]

Despite such concerns, however, very little is known about multimedia self-expressions featuring SI. Scholarly works on SI on the Internet tend to focus mainly on text-based expressions and interactions, while paying little attention to visual and multimedia elements that constitute crucial components of SI subculture online (STERNUDD, 2012). The sheer volume and burgeoning popularity of multimedia SI content suggest there is an urgent need for a better understanding of how SI is depicted online by user-generated visual content and how self-injurers frame their wounds via visual and textual modes of representation. [3]

This article reports empirical findings from a multimodal analysis of SI photographs on a photo-sharing site, Flickr, exploring 1. how the act of SI is framed and represented via user-produced photographs and concomitant textual descriptions, 2. whether these visual and textual narratives of SI contribute to the construction of alternative meanings and values of SI, and 3. whether these meanings are used to affirm or resist dominant cultural discourses of SI. In what follows, I first delineate unique characteristics of Flickr's social interface, conceptualizing Flickr photographs as a potential modality for self-expression and a stimulus for social interaction among Flickr members. This is then followed by an overview of research method that enables a multi-level examination of subject photographs, combining the method of visual content analysis (KRESS & VAN LEEUWEN, 1996) and discourse analysis (EDWARDS & POTTER, 1992; POTTER, 2003). After communicating main research findings, I discuss the ways in which the uploaders perform and enact the concept of SI via photographic mediation, and the emergent visual and verbal conventions that determine what constitutes "photographs of SI" on Flickr. [4]

With the rapid advancement of digital photographing technologies, our used-to-be overlooked quotidian world became picturesque. Photographic modes of self-expression have become an integral part of our everyday practice of identity construction, community building and cultural expression. Picture taking is no longer confined to life's most commemorative events such as weddings, graduations, and family trips; but such special scenes are now merged into countless moments of miniscule importance, namely, the everyday. As digital cameras and camera-equipped cell-phones become widely accessible to average users, we come to observe a myriad of photo-junkies holding their devices as they walk, pointing them randomly at streets, buildings, and passersby on almost every corner of the world. This emerging digital snapshot culture is also closely intertwined with online-based publication technologies. The advancement of user-friendly publishing and distributing platforms entices domestic photographs to rush out private albums into online repositories. As the channel between image capturing and display increasingly becomes leaky and permeable, digital representations of everyday life are now distributed widely via e-mails, photo-sharing websites and Social Network Sites (SNS) in anticipation of future sharing and redistribution. [5]

Among many similar sites, Flickr has become a prime destination for photo lovers who wish to share their work with millions of others. Founded in 2004, six billion photos were uploaded to the site by August 2011, and more than 3,000 images are added every minute (Flickr Blog). Flickr offers members the ability to upload, organize and share hundreds of pictures, as well as automatically post pictures taken by camera phones to blogs. Like SNSs, each user's page constitutes a node in a large decentralized network, through which members connect one another by posting comments on other's photos, subscribing to other's photo streams, and participating in public and private groups formed around shared interests. [6]

As a hotbed of domestic photography, Flickr displays millions of life's happy moments frozen in time—a smiling baby taking his first step, a happy bride and groom toasting flutes, and a starry-eyed student tossing her graduation hat. Yet at the same time, it also hosts thousands of photographs portraying thought-provoking representations of "the everyday" otherwise concealed from public eyes—like pictures portraying self-injured bodies. Until the rise of interactive and user-friendly publication platforms like Flickr, images of SI functioned like asteroids randomly scattering around the cyber-universe, only occasionally fostering a sense of underground camaraderie. To some extent, Flickr's vast social network and user-generated filtering system normalized formerly tabooed themes like SI. Unlike topic-centered online communities where the moderators often judge suitability of posting content prior to publication, most of the time Flickr does not filter the material in advance; the ethical and aesthetic judgments on images usually happen after the fact on the user end, by members' voluntary self-censoring and through interactive channels including flags, ratings, comments, and public discussions on groups.¹⁾ As a result, Flickr displays numerous varieties of photos next to each other, with little regard to choice of subject, quality of photographic technique, or identity of photo uploaders. Absorbed into a vast stream of user-generated content, pictures of SI become a drop in the ocean, harmoniously coexisting with other less controversial images. [7]

Flickr's relative tolerance for content seems to encourage uploaders of SI photographs to share their lived experience within the context of image sharing. The social architecture of Flickr constitutes the site of encounter between image and viewer, where the boundaries between the realm of experience (the practices of self-injuring) and that of representation (the photographic enterprise by which the images are conveyed to the viewer) become increasingly permeable and ambiguous. Once uploaded to Flickr, photographic representation of self-inflicted wounds takes a new materiality that provides content uploaders with a mediated venue for self-expression, as well as a stimulus for social interaction with other Flickrists. In this light, Flickr's ability to turn photo publication into interactive dialogue points to the importance of examining how Flickr photographs represent SI, and to what extent these visual narratives reflect or refract dominant medical discourse that deems SI as a symptom of mental disorder (NOCK & FAVAZZA, 2009). Apart from pathologized image of "self-injurer," uploaders of SI photographs may become able to reinterpret their photographs as cultural objects to be shared, which in turn may engender a new form of solidarity and sociality beyond dominant, often hegemonic discourses. [8]

A picture tells a thousand words. This cliché implies the capacity of image to communicate myriads of meanings, even those unable to be verbalized and otherwise left unheard. But this cliché also suggests that the image is often overcrowded with information, which makes reading it tremendously difficult practice. For instance, Roland BARTHES (1977) famously insists that a photograph invites too many interpretations and thus always needs accompanying texts to fix meaning. In the absence of verbal texts, the photograph becomes "a message without a code" (p.17), a precariously "floating chain of signified" (p.39). The notion that image demands linguistic "anchorage" (p.37) has long made visual representation subservient to text. [9]

A counterargument to such Barthean tendency arose from the field of social semiotics. Gunter KRESS and Theo VAN LEEUWEN (1996), the leading figures in this field, assert that the image is "an independently organized and structured message, connected with the verbal text, but in no way dependent on it—and similarly the other way around" (p.18). Visual and verbal modes of representation are both grounded upon the same sociocultural codes, but each takes different forms in realizing meaning. In this regard, KRESS and VAN LEEUWEN emphasize the importance to account for "visual grammar," the culturally produced regularity that controls visual modes of representation. [10]

Inspired by the social semiotic approach, the present study delves into multimodality of Flickr photographs. As a photo-sharing social media, Flickr enables a photograph on its own to serve as a powerful vehicle of self-expression and social interaction. Yet, user-generated texts (titles and descriptions) attached to the images can also reinforce meanings embedded in the visual representation. Given that Flickr (and the Internet in general) is an inherently multimodal medium that harmonizes different channels of representation, it is necessary to explore visual and textual modes as a complementary dyad that together constitutes certain representational conventions vis-à-vis existing discourses around SI. Instead of considering visual meaning in isolation, this research thus examines content SI photographs and their composition, along with user-generated texts that imply their context of publication. [11]

The unit of analysis in this study was "photographs of SI," defined as photographic images capturing intentionally injured and/or scarred bodies or body parts. This definition modifies and extends STERNUDD's (2012) definition of SI pictures,²⁾ in order to be attentive to different methods of SI employed by each uploader, specific conditions of wounds captured in the photographs, and the visual and compositional techniques used to depict the wounds. [12]

The sampling process began with 1,389 images retrieved on October 6, 2011, by searching the keyword "self-injury" with the search engine built into Flickr interface. This research obtained ethics approval from York University Research Ethics Board under the condition that the materials examined belong to the public domain. To adhere to this ethical guideline, the whole process of data collection was done without signing into a Flickr account, so as to retrieve only photos publicly accessible to any Internet users. Of 1,389 images, 96 were identified as non-photographic images (i.e. illustrations, screenshots) and thus excluded in the analysis. The remaining 1,293 photographs were manually observed to sift "photographs of SI" out of the entire dataset. Images excluded from this study were varied in subjects; pictures capturing accidentally made wounds and fake wounds using make-ups; a series of artistic pictures portraying a crying woman in white garment with a crown of thorns; snapshots of razor blades, scissors, and knives; a shot of injured guinea pig. A number of uploaders made a tribute to To Write Love On Her Arms (TWLOHA), a grass-roots SI-awareness movement dedicated to supporting people struggling with depression, self-injury and suicide. Some pictures were dedicated to Self-Injury Awareness Day,³⁾ such as images of an orange ribbon that symbolizes the awareness movement. These photographs, albeit featuring SI, were excluded from the dataset because they did not directly capture injured bodies. [13]

The sifting process identified 516 photographs of SI posted by 146 uploaders.⁴⁾ First, a visual content analysis was conducted according to the following valuables: 1. parts of the body injured (i.e. forearm, thigh, wrist), 2. method of SI (i.e. cutting, burning, carving, stabbing), and 3. conditions of injuries (i.e. bleeding wounds, scars, welts). As for methods of SI, when the skin looked injured with a sharp object, leaving thin cuts in a straight line or lines, it was coded as "cutting." "Carving" was used when words or symbols were carved on the skin. When a wound was seen inflicted with a pointed instrument, such as a pin, needle, or injector, it was categorized as "stabbed" wound. In the presence of more than one method of SI (i.e. a word carved into old scars), all methods were recorded. The conditions of wounds were determined according to the degree of recovery. Besides a fresh bleeding wound, a mark or scab on the skin was categorized as a "scar" when it looked healed up, while a red or pink ridge or swelling on the skin was deemed a "welt" in the process of healing. In case the picture captured more than one condition of injuries (i.e. bleeding cuts on old scars), multiple codes were assigned. [14]

In addition to content, the use of different visual techniques was examined in order to identify the "visual grammar" of the photographs of SI. According to KRESS and VAN LEEUWEN (1996), the visual grammar of photography consists of numerous variables, including the choices of subjects, image processing devices (camera, lens, filter), and photographic techniques (framing, focus, angle, lightening, color saturation, etc.). Among several variables, this study explored the following three variables to determine compositional techniques of the subject photographs: 1, distance and focus, 2. method of self-portrait (position of camera), and 3. point of view (direction of camera gaze). [15]

In regard to photographic techniques, KRESS and VAN LEEUWEN deem distance and focus as one of the most powerful elements that symbolizes "social distance" (p.124). A close-up shot suggests intimate and personal relationship, depicting the subjects as belonging to the same group as the photographer. On the contrary, an image that depicts the subjects in a distance conveys them as strangers, signifying mental distance the photographer feels with the subjects. While the idea of social distance has originally stemmed from the analysis of portraiture, KRESS and VAN LEEUWEN argue that it could also be applicable to the visual representation of objects and environment: for instance, when an object is shown at a close distance in advertisement, the viewer feels as if he/she is using the tool, reading the book, or eating the food. Following their guideline, distance and focus were determined by looking at whether the picture focused solely on wounds or showed other parts of body such as the face of the uploader. A close-up shot was defined as a picture capturing wounded body parts only, while head, face, torso and other body parts were cropped out of these pictures. Very close shot was deemed as "extreme close-up" in which injured skin took up the entire frame, while the background got obscured and indefinable. Photos taken from a distance and which showed more full human figures were categorized as self-portraiture. For the self-portraits, whether the represented person looked straight into the camera was also coded, in order to determine whether the photo establishes "imaginary relation" (KRESS & VAN LEEUWEN, 1996, p.118) with the viewer.⁵⁾ [16]

Images were also coded in accord to three methods of taking self-portraits: 1. photographing one's body with the camera in an outstretched arm, 2. photographing a reflection in the mirror, and 3. capturing oneself from a distance (with an aid of self-timer or remote control). This variable aims to illuminate the strategy of self-presentation adopted by the uploaders, allowing examination of the degree to which they distance and objectify their own body. Closely interrelated to the method of self-portraits was the point of view, a variable that was used in order to examine from which direction the photograph was taken. When photographing one's own body, the photographer has three choices in perspective; capturing his/her body from a first-person perspective, turning the camera onto him/herself and photographing as others would do (from a second-person perspective), or shooting one's reflection in a mirror. Body images taken from a first person's point of view reproduce how the body owner sees him/herself, how his/her body parts appear to the eye of the photographer. On the contrary, pictures shot from a distance through a second person's point of view show the photographer's body in a way he/she could not have seen by him/herself. The outcome may surprise the photographer, as he/she otherwise has no way to obtain such a view without technological help. A self-portrait in a mirror resembles the pictures taken from the second person's perspective, but it gives a slightly different outcome as a mirror image swaps right and left. The photographs were coded by whether they were taken from a first-person perspective, a second person perspective, or in a mirror, so as to identify the uploader's choice in direction of gaze. Like the method of self-portrait, this visual cue was expected to illuminate cognitive and perceptual preferences of the photographers. [17]

In order to assure the reliability of the coding procedure, this study employed the method of single-coder reliability check (i.e. KOUPER, 2010) and coded a small subsample again after sufficient time and compared the results. A randomly chosen subsample of pictures (N=77, about 15% of the sample) was coded twice with an interval of two weeks between coding sessions. The second coding session generated identical results, which suggests the coding categories were reliable enough in differentiating among the categories. [18]

After analyzing content and composition of SI photographs, the study identified three distinct types of photographs. A discourse analysis of texts attached to these images was then conducted to explore the context of publication. The coding was informed by the principles and practices of discourse analysis developed in social psychology that allows an exploration of how people utilize language to perform particular social actions and construct subjective meanings (EDWARDS & POTTER, 1992; POTTER, 2003). Following social constructivist epistemology, this analytic position sees discourse as a form of dynamic social action through which people construct and perform distinct versions of reality (POTTER, 2003). The present study benefits from this perspective in examining how uploaders frame their pictures in reference to particular discourses, and to specific contexts in which they constitute their own interpretation of SI. It also allows a researcher to question the possibility of new and emergent frame of reference they utilize in presentation of the self with respect to the issue of SI. [19]

The analysis was carried out in a step-by-step process. First, the textual information attached to the images was carefully read to identify recurring themes of explaining the photographs and the act of SI. During these readings, I noted extracts relevant to the research questions (i.e. uploaders' subjective interpretation of the photos, particular references to dominant medical and cultural assumptions about SI) and re-read these extracts to identify particular discursive tendency shared among content uploaders. Second, I organized the extracts according to how the uploaders position themselves and others vis-à-vis dominant discourses of SI. Identifying discursive positioning (DAVIS & HARRÉ, 1990; WETHERELL, 1998) taken by speaker illuminated how prevailed assumptions of SI might have influenced what and how the uploaders talked about their pictures. In so doing, I paid special attention to how the accounts were structured, what meanings they communicate, and how the uploaders frame themselves. The gender of uploaders was specified when it was identifiable, being inferred from their Flickr profiles, their self-portraits, and interactions with the other users. While all user-generated texts examined were publicly accessible, all extracts presented below were paraphrased and information that can identify individual users was omitted to ensure the privacy of the uploaders. [20]

Most images sampled for this research were raw, candid digital snaps assumingly taken by lay photographers. Some images have been edited to add texture or layers, to adjust brightness, exposure, contrast or saturation, to inscribe messages onto the photographs, or to assemble multiple images into one (i.e. a pair of sequential photos arranged side by side). Among them, most obvious manipulation was converting colored photos to black and white or sepia. Of 516 images, colored pictures (N=443) occupied the majority of the sample, but there were also 66 black and white pictures and 7 sepia-toned pictures. Some uploaders displayed different versions of the same photo in one set or collection (i.e. different angles, different cropping one color and one black and white etc.). [21]

Regarding the injured body parts depicted in the photographs, approximately 60% of images portrayed SI performed on arms (N=305), among which forearms were the most common location for SI (50% of all images, N=258, see Table 1). Thighs and legs were the second most common location, which together constituted 19% of all images. However, wrist was not so common location for SI among the subject images, with only 8.7% of pictures portrayed wounds on wrists. Even in the photographs portraying multiple scars on arms, wrists tended to be left clean and uninjured.

In terms of methods of SI adopted in the photographs, cutting overtly dominated the other methods. As Table 2 suggests, in 89% of all photographs (459 images out of a total of 516) cutting was utilized as a method of self-harming, leaving line-shaped wounds in various conditions. Though actual tools used for SI were hard to determine from wounds, some pictures captured tools along with injured bodies, including razor blades, art knifes and scissors. Most cuts ran across the arm or thigh, instead of running parallel to blood vessels. Stabbing was the second most popular among the subject images and the majority of them were assumingly done by safety pins or injections. Many singes on skin appeared to be cigarette burns, since they were circular with pinkish color on the outside of the ring. Additionally, 19 pictures captured a word or symbol carved into skin. Most messages appeared short, consisting of one or two words including "hate," "fat," "help me," or "love U." Symbols tended to be simple, like heart, star or triangle, but one picture depicted a complex chemical formula carved on thigh. Whereas most of the time a single method was adopted, some self-injurers combined different methods, for instance, carving and cutting, cutting and burning that left wounds in different shapes and conditions.

As for the conditions of wounds, most looked superficial and nonfatal, damaging only the epidermis without causing the significant loss of tissue. Conditions of wounds were largely split into two patterns: pictures that portrait fresh, bleeding cuts or wounds made by deliberate self-harm and ones that capture old scars that are already healed, or welts in the process of healing. 62.2% of the images (N=321) captured scars completely healed up, while 181 photographs (35.1% of all images) depicted freshly made, open wounds and blood. There were 93 pictures (18% of all images) capturing healing welts in the process of healing, filled with pinkish granulation tissue. Some pictures presented stitched (N=18) and bandaged wounds (N=5), which indicated medical attention (either official or unofficial) was paid to the injury. In the majority of pictures which depicted bleeding wounds, multiple conditions of wounds existed in one skin, like freshly opened wounds made upon old scars or healing welts. Co-existence of several wounds in different healing stages suggests that SI was repeatedly performed with a certain time period before the body was healed up. Some pictures captured deep scars left by recurrent self-harming, which made the skin look like a washing board. When uploading pictures capturing the same SI event, the uploaders oftentimes organized them in the chronological sequence by which images were produced. For instance, an uploader posted 24 pictures of an injured thigh taken in a same day, organized the series starting with a picture depicting the moment when a hypodermic needle was injected into her skin. Next to this picture were a couple of photos depicting blood dripping down her thigh, and then close-up shots capturing the stabs made by the injectors. However, some uploaders intentionally disrupted the chronological order and displayed photos in a non-linear manner. For these uploaders, as PINK (2001) suggests, they may have had other narrative elements more meaningful than original shooting order, thus flexibly arranged their images in a way that their subjective experience could be better reflected.

In terms of distance and focus, 385 pictures were close-up shots depicting injured body parts only. There were also 40 extreme close-shots of injured skin, in which skin was filling up the whole picture, thus obscuring which body part the skin belongs to. While close-up and extreme close-up pictures occupied 82.4% of the entire sample, there were 91 photographs (17.6% of all images) that resembled traditional self-portrait type pictures, capturing faces of the photographers along with their injured bodies. In 20 of these 91 self-portraits, the faces of the persons were unrecognizable, as they faced down or covered their face by hands, arms, books or other objects. In some of these "faceless" pictures people sunk their heads into their hands as if groaning in wordless pain. Twenty-seven self-portraits captured photographers looking directly into the camera, while in 44 pictures the represented persons averted their gaze. This suggests that about one third of these self-portraits (27 out of 91) portrayed self-injurers as establishing "imaginary contact" with the viewer, "demanding" some affective interaction. These portrayed facial expressions varied from smile to sadness to indifference; some glance up at the camera with a pouty face, while others stare at the viewer with tearful eyes. Although it is not of direct relevance to the present study, a female-dominated tendency was conspicuous among the sampled self-portraits. Of the 91 self-portraits, 88 depicted female self-injurers with their bodies. It seems also worth mentioning here that all self-portraits depicted old scars or healing wounds, rather than fresh, bleeding wounds. [25]

As for the methods of self-portrait, about 73% of the images (N=377) indicated that the photographer held the camera on one hand when they took the shot. There were only seven self-portraits in mirrors despite the popularity of this method among Flickrists.⁹⁾ In a related vein, 61.6% of the pictures (N=318) were taken from the first person's point of view, looking down the cuts on thighs or looking up the arm facing inward. On the contrary, in 151 pictures the body was portrayed from the second person's perspective, depicting, for instance, two arms stretching toward the lens or scratches inscribed on the lower back.

Analyses of the condition of wounds (Table 3) and visual techniques (Table 4) illuminate that there are several types of SI photographs coexisting on Flickr, which can be roughly divided into three groups. The first group consists of documentary-type photographs that aim predominantly at bearing witness to SI. These are mainly the close-up shots of fresh wounds that tend to exclusively depict injury itself and the body parts where SI was actually done. The focus is on the immediate aftermath of SI inscribed on one's body—wounds accompanied by blood bear vivid witness to self-harming—and act as a visual record of the action. These photographs tend to be taken from a first person perspective with the camera in an outstretched hand, presenting the photographer's body in the way that he/she has seen him/herself. [27]

In contrast, pictures in the second group appear as a mode of reminiscence, representing visual evidence of the past struggle with SI. Rather than documenting an ongoing event, the photographs belonging to this group capture old scars that were already healed up, or welts in the process of healing. In these pictures scars are presented as an index of the past event, reminding the self-injurer the moment of SI that has happened a long time ago. The pain embalmed in the documentary type photographs lessens its affective impact in these images, recalling instead the past suffering endured. Some uploaders gave additional meaning to their scars by writing a message or a symbol on their scars in support of SI awareness movement. Like the first group, most of these images were captured from the first person point of view with the camera at arm's length, but some were taken from a distance, from the viewer's perspective. [28]

The final group is composed of self-portraits that depict injured body as a whole. Whilst the other two groups consist of fragmentary documentations of body parts displayed out of context, pictures in this group visually implied the entire identity of the uploaders. In these pictures, wounds were portrayed as important yet were one of many identity markers associated with the represented person. In showing their wounds, the represented persons often acted out emotions such as anguish, agony, sadness and happiness with facial and bodily expressions. The gaze and gesture displayed by the photographed persons play a pivotal role in performative self-expression. A happy face may appear more relevant to the viewer if the gaze is directed rather than averted toward the camera, as this suggests that the photographed person invites the viewer into a relation of social affinity with him/her. A writhing gesture of agony may invoke sympathy, as this may appear the portrayed person asks the viewer's help. These types of pictures thus do not simply carry visual evidence of SI, but represent deliberative performance by the uploaders by means of photographic self-expression. [29]

Following the visual content analysis, accompanying texts to three types of pictures (documentary, reminiscence, and self-portrait) were examined to identify the context of production and publication. While content and compositional analyses illuminate the visual intent in shaping of SI photographs, user-generated texts attached to the images add a supplemental layer of meaning to the images. [30]

Of 516 pictures 47 had neither title nor description, but the rest of pictures (N= 469) had at least either title or description. In some cases, uploaders did not create a title for their images, but simply kept file numbers automatically given by their digital camera (i.e. DCP1277 or IMG0356) or inserted the date the photo was taken as titles. These titles and descriptions seemed designed for personal categorization purposes, rather than giving additional information to the viewer. But other uploaders utilized titles and descriptions to explain what the image entails, for what purpose it was taken, and what it means to them. In these cases, textual information interacts synergistically with image, adding extra layers of meaning otherwise unavailable in the picture. [31]

Pictures that depicted fresh, bleeding wounds tended to carry short titles and descriptions. Of 181 pictures that portrayed the immediate aftermath of SI, about two-third of them only had short titles composed of less than three words, such as "self-injury," "blood16," and "new cuts." Longer titles and descriptions were equally succinct, simply describing what is going on in the photograph in a straightforward manner, i.e. "self-injury done by razorblade on forearm." While some uploaders attached a paragraph-length text to their photographs, the verbal communication tended to remain minimal in this group of photographs. Although accompanying texts attached to documentary type photographs do not offer much information to help viewers understand the context of production and sharing, they nonetheless present a kind of murmuring to oneself that can certainly provide a nuanced understanding of the photos. For instance, the following extract implies the uploader's motivation for SI.

In this short description accompanied with a picture of bloody arm, the uploader alludes to the voluntary and repetitive nature of her SI and asserts that it is "the warm blood" that entices her to continue the act. While this extract appears too blunt to provide a clear reason for what the act really means to the uploader, it well describes the physical excitement the uploader feels when blood oozing out of the wound, and implies the motivation behind the uploading of this photo to visually reproduce the pleasure. Given wide prevalence of blood fetish among body modification subculture (VALE & JUNO, 1989), it is assumed that the uploader may resonate herself with such subcultural, pro-SI type discourse. Nevertheless, this description appears somewhat introverted and self-contemplative in its goals, rather than intending to offer additional information to the viewers regarding context of publication. Similarly, the following description attached to a photograph of numerous bleeding cuts on leg shows little intent to offer additional information, but rather presents subjective feeling toward SI.

In Extract 2, the uploader positions herself in line with dominant medical assumption and deems SI as something shameful, driven by the "lack of self-control." It is not an enjoyable, deliberative practice, but an impulsive and pathological act to fulfill her self-destructive urge. Though this short description does not allow us to fully grasp the intentionality of the uploader, we can assume that this picture was uploaded partially for the purpose of self-discipline, to prevent the uploader from doing the same thing again in the future. The self-reproachful attitude taken by this uploader marks a sharp contrast to the subject position taken in Extract 1. Despite some of the striking resemblances in content and composition, the two extracts above indicate the contrasting semantic intentions behind these photographs. [34]

Compared to the documentary type photographs, texts associated with pictures that depicted old scars tended to be longer and provide more detailed explanation about the context of the photographs. In some cases, the uploaders inserted poems or lyrics from popular songs as descriptions that serve as background music to the silent photographs. In other cases, they provided personal narratives behind the images, recalling the past memory of SI. The extract below exemplifies such accompanying texts in which the uploaders reminisce about the past struggle.

"I suffered from depression for a few years and I used self-injury as a way of relieving the pain I felt inside ... I stopped cutting about two years ago and I hope I can keep up. Most of my scars are already gone" (Extract 3). [35]

In this description attached to a close-up shot of a scarred arm, the uploader recalls her past, when she was harming herself to relieve inner distress. Her scars look white and faded, symbolizing the past long gone but never forgotten. The interpretative position taken by the uploader is in alignment with medical discourse, as she frames SI as a pathological act caused by mental ill-health (long-term depression) and implies that she feels proud of herself for having quitted and stayed away from this old habit. In the meantime, however, the uploader thinks her scars are worth photographing and preserving, before all the scars are gone. The last line, "Most of my scars are already gone," appears highly indicative in this sense, as it implies the uploader's conscious decision to save the scars from oblivion—at least by means of photographic embalming. Here visual and verbal channels synergistically work to inform the viewer of the ambivalent position of the uploader: on the one hand the old wounds repulsively remind her of a painful struggle she has been through, yet on the other they will be missed if they disappear completely. [36]

Texts attached to self-portraits represent intriguing subject positions taken by the uploaders. In contrast to the first and second groups that focus solely on damaged skin, these self-portraits show wounds as part of the whole body, and accompanying descriptions tend to reflect positive interpretation of SI and the injured body. In engaging with positive self-construction, the uploaders often make a reference to dominant medical and cultural assumptions and then position themselves in opposition to these dominant discourses. The following description well exemplifies this rhetorical strategy:

"I am not as lovely as I once was. I am not as shameful as I used to be. Self-harm is not a sign of weakness, it's a sign of strength and survival. I call them 'war injuries' because that's what they are" (Extract 4). [37]

This description was accompanied with a self-portrait of a smoking woman with multiple scars on her forearm, taken from side view. In this extract the uploader posits a contradictory understanding of her body; she first alludes to normative beauty ideal that embraces clean, flawless skin and declares that her self-injured body is not as "lovely" as unharmed one, indicating that it is perhaps repulsive and disgusting even to her eyes. Yet, she insists that she no longer feels shame at the wounds, for they symbolize not her weakness but "strength and survival." She even confers the title of "war injuries" on her scars, emphasizing survivability and endurance represented by the old wounds. The shift from dominant discourses that conceive scars as "a sign of weakness" to alternative definition of them as "war injuries" illuminates the complicated process of self-acceptance the uploader has gone through. In positioning herself as a survivor who has gone through physical and emotional difficulties, the uploader gives a new meaning to her scars and reframes her body as something she owns, controls, and is proud of. [38]

While Extract 4 explains what the images meant to the uploaders, it sounds more self-contained than interactive, primarily declaring the uploader's resiliency and self-pride to be a SI survivor. However, some uploaders go one step further and display their self-portraits for an altruistic purpose, to empower similar others struggling with SI. The following extract was attached to a self-portrait in a mirror, in which the photographer wears no-sleeve shirts and shows off numerous cuts on her upper arms.

"I am a photographer and a recovering self-harmer. A tough journey, but there's hope. I know I am beautiful and there is hope. Recovery is possible. You are beautiful" (Extract 5). [39]

In this description, the uploader first positions herself as "a photographer" as well as "a recovering self-harmer." By showing two identities in parallel, she simultaneously establishes her authenticity to tell personal narratives of SI and authority to display her self-portrait as a work of art. On the one hand, the old scars inscribed on her skin eloquently tell the "tough journey" and symbolize the struggle endured. On the other hand, the self-declaration of identity as a photographer allows the uploader to posit a counterargument to conventional beauty ideals and aestheticize her and fellow self-injurers' bodies. This identification makes wounds appear no longer pathological or repulsive, but as subjects worth capturing via photographic beautification. Perceiving her self-injured body to be beautiful not only affirms the embodied experience of the uploader but also affirms those who share similar SI experiences. Because the photographer found herself "beautiful" through the "tough journey" to quit SI, similar others who are struggling with SI are equally "beautiful," and their entire self ("You") is also celebrated. In this vein, this photograph delivers a powerful message of hope and empowerment to self-injurers on Flickr, which actively transforms body hatred and shame into love and self-pride. [40]

Analysis of content and composition of SI photographs reveals that dominant cultural discourses have infiltrated the Flickr images in a number of ways. One of the most conspicuous impacts of cultural discourse can be seen in the injured body parts captured in the photographs. About 60% of the photographs depicted wounds inflicted on arms, especially forearms, which were deemed in previous study as the most common part for self-harming in popular media portraits (WHITLOCK, PURINGTON & GERSHKOVICH, 2009). In some sense, popularity of arms cannot be unexpected, since arms are one of the most easily reachable body regions. But besides accessibility, arms seem to symbolize the non-suicidal and repetitive nature of the act, since limb injuries in general have lower mortality rate and faster recovery speed than head, neck and trunk injuries. Worthy of attention here is the relative scarcity of wounds on wrists, which have long been defined by medical professionals as a sign of suicidal intent (i.e. GREYDANUS, 2011). Relative unpopularity of wrists and other fatal and sensitive areas like necks, nipples or genitals thus suggests non-suicidal tendency among the uploaders, who carefully chose body parts to keep SI nonfatal and long lasting. [41]

Meanwhile, the popularity of arms as the location for SI also indicates a complex, contradictory psychology behind the act. On the one hand, unlike body parts like face, hands, ears and neck that are usually exposed to other's gaze, arms can easily be covered with shirtsleeves, wristbands or bracelets if the body owner wishes to hide the wounds. Yet on the other hand, the fact that arms are more publicly visible than belly, breasts, and back may give visual privilege to the wounds on arms compared to wounds on more concealable body parts. As such, the arm-injurers are to some extent capable of allowing others to constantly check their injuries and putting the wounds on display on their own discretion. While for some this capacity to control visibility of wounds can appear "manipulative" or an "attention-seeking" gesture as often deemed by popular culture discourse (FAVAZZA, 1996), the ambivalent visibility of arms seems to give the self-injurers a sense of control over visibility of their body, at least choosing when and to whom, they show or hide their wounds. Visible yet concealable, arms may provide self-injurers with a venue for performative self-control. [42]

In a similar vein, the methods of SI employed by the uploaders also reflect existing cultural assumption. The overwhelming percentage (about 90%) of the photographs examined in this study depicted cuts and scars on body parts, indicating popularity of cutting among the self-injurers. This echoes the trends across media portraits of self-injury, in which cutting is depicted as the most dominant method of SI (WHITLOCK et al. 2009). Many pictures portrayed several shallow cuts, rather scratch-like injuries made by a sharp instrument. Most cuts looked superficial and do not penetrate the skin, though some of them had considerable length and depth. Most the time, cuts were multiple and uniform in shape, often oriented in the same direction or in a criss-cross pattern. While the prevalence of shallow wounds reconfirms a non-suicidal tendency among the uploaders, it also reinforces the "delicate cutters" stereotype proposed by health professionals in the 1960s (BRICKMAN, 2004) that equalizes "self-injury" with nonfatal, repetitive cutting. Also intriguing is that some uploaders carved messages onto their skin. In his analysis of SI photographs, STERNUDD (2010) focuses on cutting that takes a form of words and concludes that the majority of messages expresses anger or self-disgust. Although messages of anger (i.e. "hate") or self-disgust (i.e. "fat") were found in the present study, there were also less hurtful messages like "love U" and symbols carved onto one's skin, including aesthetically sophisticated ones. Such symbols, shapes and certain designs, express personal meaning significant to the uploader, as well as their aesthetic impulse. Compared to linear-shaped cuts or scratches, these wounds resemble the outcomes of scarification, a form of artistic body modification that applies artistic design onto the body through carving. In this light, the uploaders who chose to inscribe words and symbols have more communicative and perhaps aesthetic intent than the other self-harmers. Moving their act from the realm of self-destruction to the domain of artwork, these uploaders may enact artistic practice to use their skin as a medium for aesthetic self-expression. [43]

Regarding conditions of wounds, the analysis reveals what individual uploaders valued amongst different meanings they have given to their wounds. There is a polarizing trend among the uploaders whether they put more weight on the past or on the present. About 35% of photographs documented gaping wounds dripping blood, while more than 80% of the pictures showing either scars or healing wounds made in the past. The qualitative analysis of texts attached to the photographs gives an explanation to this split tendency among the uploaders: one group foregrounds documentary function of photography and uses it to embalm the moment when the violence was inflicted on the body. For these uploaders, pictures of the freshly injured body serve both as records of life events and as memorabilia for catharsis worth remembering. On the contrary, the uploaders who found the past wounds more worthy to photograph than the immediate aftermath of SI utilized photography as a medium to tell a story about a struggle endured. Unlike bloody wounds that bear vivid witness to self-destruction, the old scars symbolize the healing process and attest to the painful experience the uploaders went through. Of course the healing process can be equally painful as the act of SI, but scars depicted in these pictures testify that the uploaders came through the struggle and succeed in discontinuing SI (at least at the time the photos were taken). [44]

Along with the content analysis, visual techniques adopted by the uploaders imply several visual conventions emerging among the Flickr photographs of SI. The examination of distance and focus suggests an overt dominance of close-shots that exclusively depicted injured body parts and extreme close-ups of wounded skin. The dominance of close-up shots resonates with the result of the study by STERNUDD (2010), which found close-up pictures depicting exclusively injured skin to be statistically the "most frequent representation of cutting" (p.232). In these photos, the wounded bodies were denaturalized and reconstructed as fragments removed from normal contexts and scale, then displayed as independent phenomena without any contextualized features. In the extreme close-ups, the body becomes even more decontextualized, losing all identity markers usually inscribed to the skin, such as gender, age, race or cultural backgrounds. The effect of this intentional fragmentation appears two-fold: on the one hand, it deprives the self-injured body of its context, depth and other attributes necessary for meaning making. The body becomes objectified, taken out from the context, thus loses its power to act on its own. Yet on the other, due to the very absence of context, depth, and cultural connotation, the damaged skin obtains a new ontological meaning apart from individual self-injurers, symbolizing ubiquity of pain and suffering beyond the boundary of individual bodies. In other words, deprived from the original body, the wounds embalmed onto photography no longer remain silent; instead, such a photo groans, screams and perhaps eloquently speaks on behalf of people suffering from physical pain, reminding the viewers their own experience of wounding. Here the injured skin replaces face of self-injurer, assimilating injured skin to the concept of the self. [45]

In addition to distance and focus, the use of two other visual cues—method of self-portrait and direction of camera gaze—also indicate visual story-telling strategies prevalent among the uploaders. More than 70% of the pictures were taken by holding the camera out at arms length and over 60% of pictures were taken from the first person perspective. These framing styles contribute to the reproduction of what the uploader has seen and experienced, allowing the uploaders to accurately preserve what their eyes have captured, while at the same time giving the viewer an impression that he/she is the owner of this damaged skin. The pseudo-subjective experience mediated via these pictures is doubtlessly an embodied one, since the images invoke the body itself, bodily experience, and materiality of skin damaged, and allow for the future reenactment of the scene of SI. Taken from the viewpoint of the self-harmers, the photographic embodiment of pain demands the viewer to share the physical pain, pent-up feeling and emotional struggle embalmed on the surface of photography. That is, the wounds-only, first person viewpoint close-ups do not give the viewer enough distance to objectify the body as a distinct "other." Rather, by causing perceptual confusion between the body of uploader and that of observer, they externalize otherwise isolated and internalized pain and make it a sharable experience via online network. What seems critical here is that this function of the photograph goes beyond the "inexpressibility" of physical pain theorized by Elaine SCARRY (1985). The incommunicable nature of pain, which, according to SCARRY, not only resists language but even actively destroys it, has long given power to external authority to explain the sufferers' feeling on their behalf. But the user-generated pictures of SI on Flickr eloquently speak on their own, for the sake of the uploaders. Through the irreversible flow of self-mediation the damaged skin is converted into an object of artistic consumption and acquires unique affectivity and sharablity may overcome the original unsharability of physical pain. As a visual reminder of body's "woundabilty," or the pain as a ubiquitous, humanly shared experience, the pictures of SI may construct a potential to bridge between the viewer and the uploader. [46]

Findings also suggest that some uploaders took one step further and revealed their identity via self-portrait type images. Although the controversial and triggering nature of SI and negative stereotypes attached to the act might have made many uploaders hesitant to publicize their faces, in 71 pictures the uploaders came out from under the cloak of online anonymity and showed their faces alongside their wounds. While the actual reason for their decision is not knowable to us, the accompanying texts give a glimpse to what motivated these uploaders to disclose their identity. The discursive strategy used by these "face-full" uploaders was to frame their wounds as a proof of struggle endured and position themselves as "survivors" who have overcome the hardships. This positioning allows the uploaders to reframe SI against dominant cultural discourses and reshape it as a mode of enactment of one's resiliency. In so doing, some beautified their wounds and sequentially themselves, so as to empower other self-injurers who have gone through similar struggle. The performativity of damaged body maximized in these self-portraits, to the extent that the wounds symbolize a shift from shame to pride, body hatred to love, and offers the uploaders a sense of confidence and agency. While this aestheticizing tendency was occasionally seen in pictures without revealing photographer's face, it was most conspicuous among the face-full uploaders. [47]

However, this conceptual shift should not be taken to imply that uploaders have gained complete freedom to choose subjective positions. Instead, dominant medical and cultural discourses are still pretty much at work in the mind of uploaders, who fully understand the socially controversial nature of their act. As such, their subjective positions become highly ambivalent and somehow contradictory; on the one hand, they see their wounds repulsive and shameful in line with conventional beauty ideals. Yet on the other, the wounds, especially old scars inscribed on the body, bear witness to embodied experience of survival and resiliency, which appears to be authentic and empowering to other self-injurers. To take an identity as a "self-injurer" on Flickr, thus requires the uploaders to perform two opposing positions; one that resonates with medical and cultural discourses that pathologize the wounded body, and the other that enhances an alternative view to beautify the same body as an authentic vehicle of the "true self." [48]

Despite the findings and implications, the present research has a few shortcomings worth addressing. One noticeable limitation is that the present study examined user-generated content without directly approaching the content uploaders, thus was unable to reveal why the self-injurers photograph and share their wounded bodies online, and what they gained from such experiences. Future research could employ more direct approach to investigate what is behind the content through interviewing the uploaders. Future research should also explore how viewers receive and react to photographs of SI. Given that Flickr allows viewers to post comment to pictures, analyzing comments attached to SI photographs might provide further insight into meanings viewers connect to the images, and the extent to which dominant discourses influence the mode of reception. [49]

SI photographs on Flickr represent a complex process of identity performance, in which the self-injured body becomes a site where different discourses intersect. While Internet communities of self-injurers have long been criticized for normalizing SI and/or promoting pro-SI attitudes (i.e. WHITLOCK et al., 2006), the current study has illuminated more nuanced subjective positions taken by the uploaders. In expressing themselves via visual and verbal channels, the uploaders formulate and share their own realities and identity narratives against stigmatizing discourses. In so doing, however, they tend to oscillate between dominant discourses and alternative discourse against them, framing the act in resonance with dominant cultural assumptions while giving alternative interpretations to their injured bodies. In this oscillation, SI and its outcome are often described as two separate entities: whereas SI is framed pathetic and shameful, or at least something not to be proud of, scars on skin are also seen a positive memorabilia that symbolizes survivability and resiliency of the uploader. This ambivalent attitude allows for a new interpretation of scars as authentic and to some extent socially acceptable, without normalizing SI itself. For some uploaders, the medium of photography provides a vehicle for presenting authentic first-person accounts of self-harming experience, and at the same time to constitute injured body as subject of aesthetic gaze. For these Flickrists, to take a photograph of self-inflicted wounds is to acknowledge their own vulnerability and "healablity," so as to shift their injured body from the domain of mental illness to the realm of advocacy and recovery. Uploading wounds on Flickr is, in this light, appears to be an ambivalent practice that renders the experience of self-expression as a paradoxical mix of personal catharsis and social performance. [50]

1) This does not mean that Flickr is a sensory-free heaven for all users. It occasionally makes a compromise with local politics and legislation, thereby limiting user access to specific content. When Flickr introduced a user-generated filtering system (SafeSearch system) in 2007, users with Yahoo! accounts registered in Germany, Singapore, Hong Kong, India, and South Korea were prevented from turning SafeSearch off, due mainly to local age-verification laws and penalties in those countries. This decision has provoked vigorous protests particularly in Germany, where a number of Flickrists called for protest against new restrictions by posting anti-censorship messages on their photostreams and formulating Flickr groups for this matter. As a result, Flickr relaxed censorship for German users, granting them access to "moderate" content (SafeSerach Moderate). Yet, users in Germany and Romania still cannot see "restricted" content due to local terms of use (Flickr FAQ/Content Filters). <back>

2) In his study of SI photographs, STERNUDD (2012) defined his sample as "photographic self-portraits of SI that commonly depict close-ups of fresh cuts or scarred body parts" (p.422). This definition appears too narrow for the scope of this research, since it limits the method of SI to cutting and focuses only on close-up photographs. <back>

3) According to LifeSIGNS (Self-Injury Guidance and Network Support), a UK-based grassroots SI support organization, Self-Injury Awareness Day (SIAD) on March 01 is a day for global self-injury awareness, education and advocacy. In order to raise awareness about self-injury, LifeSIGNS encourages people to wear an orange wristband or hold a key ring, disseminate fact sheets available on their websites, and mention SIAD on social networks such as Twitter, Facebook and Tumblr. <back>

4) All photographs analyzed in this study were publicly accessible at the time of data retrieval (October 2011). Since then, some uploaders have deleted or changed the privacy settings of their work. Some of them are no longer active at the time of writing this article (December 2012). That a sense of privacy presented by the uploaders has been constantly changing appears suggestive when we study use-generated materials online. <back>

5) In regard to photographic portraiture, KRESS and VAN LEEUWEN (1996) further emphasize the power of gaze that shapes our way of seeing the people inside picture frame. They point out that photographs in which the represented person looks directly at the viewer's eyes can establish a social relation with the viewer even on an imaginary level. When the photographed person looks straight into the camera, he/she "demands" the viewer to return the gaze and "enter into some kind of imaginary relation with him or her" (p.118). Emotions depicted or stimulated by a photograph also contribute to the humanization of the subject. Portrayed persons may smile, frown, or scornfully stare at the viewer in a photograph, dragging the viewers into a quasi-social, affective mode of interaction. Such portraitures make a fundamental contrast to portraits in which a represented person glances aside or hides his/her eyes to avoid eye contact. The viewers will see people in these pictures passive and submissive, who simply "offer" information without demanding anything back from the viewer. <back>

6) Total percentages will not equal 100% as multiple codes can be assigned to individual photograph. <back>

7) Total percentages will not equal 100% as multiple codes can be assigned to individual photograph. <back>

8) Total percentages will not equal 100% as multiple codes can be assigned to individual photograph. <back>

9) This traditional method of taking a self-portrait has gained popularity among Flickrists. For instance, there is a popular Flickr group called MirrorProject@Flickr, which is a Flickr extension of the popular website MirrorProject that has gathered self-portraits taken in mirrors or other reflective surfaces since 1999. As of December 2012, MirrorProject@Flickr has over 6,500 members and 25,700 pictures in its group stream. <back>

Brickman, Barbara J. (2004). "Delicate" cutters: Gendered self-mutilation and attractive flesh in medical discourse. Body & Society, 10(4), 87-111.

Davis, Bronwyn & Harré, Rom (1990). Positioning: The discursive production of selves. Journal for the Theory of Social Behaviour, 20(1), 43-63.

Favazza, Armando R. (1996). Bodies under siege: Self-mutilation and body modification in culture and psychiatry. Baltimore: The Johns Hopkins University Press.

Kress, Gunter R. & van Leeuwen, Theo (1996). Reading images: The grammar of visual design. London: Routledge.

Lewis, Stephen P.; Heath, Nancy L.; St. Denis, Jill M. & Noble, Rick (2011). The scope of nonsuicidal self-injury on YouTube. Pediatrics, 127(3), e552-557.

Lewis, Stephen P.; Heath, Nancy L.; Sornberger, Michael J. & Arbuthnott, Alexis E. (2012). Helpful or harmful? An examination of viewers' responses to nonsuicidal self-injury videos on YouTube. The Journal of Adolescent Health, 51(4), 380-385.

Mitchell, Kimberly J. & Ybarra, Michele L. (2007). Online behavior of youth who engage in self-harm provides clues for preventive intervention. Preventive Medicine, 45(5), 392-396.

Murray, Craig D. & Fox, Jezz (2006). Do Internet self-harm discussion groups alleviate or exacerbate self-harming behaviour? Australian e-Journal for the Advancement of Mental Health, 5(3), 1-9.

Nock, Matthew K. & Favazza, Armando R. (2009). Nonsuicidal self-injury: Definition and Classification. In Matthew K. Nock (Ed.), Understanding nonsuicidal self-injury: origins, assessment and treatment (pp.9-18). Washington DC: American Psychological Association.

Pink, Sarah (2001). Doing visual ethnography: Images, media and representation in research. London: Sage.

Potter, Jonathan (2003). Discourse analysis and discursive psychology. In Paul Marc Camic, Jean E. Rhodes & Lucy Yardley (Eds.), Qualitative research in psychology: Expanding perspectives in methodology and design (pp.73-94). Washington DC: American Psychological Association.

Ross, Shana & Heath, Nancy (2002). A study of the frequency of self-mutilation in a community sample of adolescents. Journal of Youth and Adolescents, 31(1), 67-78.

Scarry, Elaine (1985). The body in pain: The making and unmaking of the world. New York: Oxford University Press.

Sternudd, Hans (2010). The discourse of cutting: A study of visual representation of self-injury on the Internet. In Jane Fernandez (Ed.), Making sense of pain: Critical and interdisciplinary perspective (pp.237-248). Oxford: Inter-Disciplinary Press.

Sternudd, Hans (2012). Photographs of self-injury: Production and reception in a group of self-injurers. Journal of Youth Studies, 15(4), 421-436.

Vale, V. & Juno, Andrea (1989). Modern primitives: Tattoo, piercing, scarification—An investigation of contemporary adornment & ritual. Hong Kong: Re/search Publications.

Wetherell, Margaret (1998). Positioning and interpretative repertoires: Conversation analysis and post-structuralism in dialogue. Discourse and Society, 9, 384-412.

Whitlock, Janis L.; Lader, Wendy & Conterio, Karen (2007). The Internet and self-injury: What psychotherapists should know. Journal of Clinical Psychology, 63(11), 1135-1143.

Whitlock, Janis L.; Powers, Jane L. & Eckenrode, John (2006). The virtual cutting edge: The Internet and adolescent self-injury. Developmental Psychology, 42(3), 407-417.

Whitlock, Janis L.; Purington, Amanda & Gershkovich, Marina (2009). Media, the Internet, and nonsuicidal self-injury. In Matthew K. Nock (Ed.), Understanding nonsuicidal self-injury: Origins, assessment and treatment (pp.139-155). Washington DC: American Psychological Association.

Yukari SEKO completed her doctorate in communication and culture at York University, with a focus on new media studies and qualitative Internet research. Aspiring a career spanning communication studies and health research, Yukari is currently doing her postdoctoral fellowship in the Social Aetiology of Mental Illness (SAMI) Training program at the Centre for Addiction and Mental Health (CAMH). Her research interests include the therapeutic use of social media, eHealth, youth and adolescent mental health, Internet research ethics, and online-mediated intervention.

Postdoctoral Fellow
Social Aetiology of Mental Illness (SAMI) Training Program
Centre for Addiction and Mental Health (CAMH)
455 Spadina Ave. Suite 300 
Toronto, ON M5S 2G8 
Canada

Seko, Yukari (2013). Picturesque Wounds: A Multimodal Analysis of Self-Injury Photographs on Flickr [50 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(2), Art. 22,
http://nbn-resolving.de/urn:nbn:de:0114-fqs1302229.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License

Injured body parts	Number of occurrences	% of photos examined (N=516)⁶⁾
Arm: forearm(s)	258	50.0
Arm: upper arm(s)	28	5.4
Arm: whole	19	3.7
Leg: thigh(s)	78	15.1
Leg: leg(s)	20	3.9
Wrist	45	8.7
Hand	8	1.6
Belly	6	1.2
Head	3	0.6
Knee	3	0.6
Breast	2	0.4
Finger	2	0.4
Hip	2	0.4
Lip	2	0.4
Back	1	0.2
Neck	1	0.2
Unidentifiable	47	9.1

Method	Number of occurrences	% of photos examined (N=516)⁷⁾
Cutting	459	89.0
Stabbing	32	6.2
Burning	21	4.1
Carving	19	3.7
Scab picking	5	1.0
Hair pulling	3	0.6
Bruising	1	0.2
Scratching	1	0.2
Unidentifiable	3	0.6

Condition of wounds	Number of occurrences	% of photos examined (N=516)⁸⁾
Scar healed up	321	62.2
Bleeding fresh wound	181	35.1
Welt, healing wound	93	18.0
Stitched wound	18	3.5
Bruise	7	1.4
Bandaged wound	5	1.0
Scab	3	0.6
Unidentifiable	3	0.6

	Number of occurrences	% of photos examined (N=516)
Distance and focus
Close-up: close-up to body part(s)	385	74.6
Close-up: extreme close-up to skin	40	7.8
Self-portrait: person looks straight into camera	27	5.2
Self-portrait: person glances aside/closes eyes	44	8.5
Self-portrait: face unrecognizable	20	3.9
Method of self-portrait
Holding the camera in an outstretched hand	377	73.1
Setting the camera in distance	132	25.6
Reflection in a mirror	7	1.4
Point of view
First person perspective	318	61.6
Second person perspective	151	29.3
Reflection in a mirror	7	1.4
Unidentifiable	40	7.8