Hetero-sexting as mediated intimacy work: ‘Putting something on the line’

This article is concerned with women’s digitally mediated practices of creating and sending private sexual images to men, here referred to as ‘hetero-sexting’. Drawing on material from individual interviews with adult British women about their experiences of hetero-sexting, the article develops an understanding of women’s hetero-sexting practices as a form of female-conducted ‘mediated intimacy work’, constituted by a constant negotiation of female risk taking and male trustworthiness. In doing so, it shows how the women relied on and made active use of the sexting-related risk of digital image abuse as a means to establish and enhance trust and, as such, stress the significance of their hetero-sexting activities as performances of intimacy. Sexting-induced vulnerability was therefore both drawn on and dismissed within the very same accounts of hetero-sexting, as it was applied as a means to perform a new form of normative femininity, namely that of the agentic intimacy worker.


Introduction
I think sending like explicit pictures, you have to have like ultimate trust with the person that you're sending them to. [ . . . ] they're so like vulnerable and they're so intimate that it's like; it's a lot to think about. -Ruby (late teens) This article is focused on women's heterosexual 'sexting' practices, here understood as the digitally mediated practice of creating, sending and receiving self-made private sexual images. 1 More specifically, the article is concerned with women's accounts of their creating and sending of private sexual images to men, which I will refer to here as a form of 'mediated intimacy work' constituted by a constant negotiation of risk and trust. Drawing on material from individual interviews with adult British women about their experiences of sexting with men, I will illustrate how, with the risk that their images might be shared further in public without their consent being a prominent feature in women's experiences of heterosexual sexting practices, the concept of risk came to take on new and more complex meanings. In my interviewees' accounts, this sexting risk and associated notions of sexting-induced vulnerability were not strictly cast as something negative that they should mitigate and avoid -they were also cast as a resource for intimacy. 2 In providing an alternative, more positive account of the role of risk in women's heterosexual sexting practices -here forth referred to as 'hetero-sexting' 3 -this article expands on existing conceptions of the risk that sexting material might be shared publicly and non-consensually, as they occur in the existing literature on sexting and risk (Amundsen, 2019a;Renfrow and Rollo, 2014;Setty, 2018).
The field of research concerned with sexting is relatively novel, yet broad and quickly expanding. So far, a significant proportion of the existing research conducted in this area consists of qualitative empirical research and discourse analyses concerned with sexting practices as they take place in relation to notions of risk (Amundsen, 2019c;Döring, 2014;Setty, 2019). Arguably, this focus on sexting risk can be explained by sexting's common association with the non-consensual further online distribution of private sexual images in public. Here, I draw on Nicola Henry and Anastasia Powell's (2016: 405) concept of 'technology facilitated image-based abuse' in referring to this non-consensual activity as a form of 'digital image abuse'. I use this term broadly to refer to all forms of non-consensual further distribution of private sexual imagery online, regardless of whether the imagery is shared further as a result of hacking, by accident, as a joke, by someone with malicious intentions, or perhaps by someone who did not realise that the person depicted in the image is not OK with its further dissemination (Henry and Powell, 2016: 413).
There is limited data to demonstrate precisely how common digital image abuse is, but -as a recent report from the UK shows -those affected by this are 'disproportionally female' (Sharratt, 2019: 5). This gendered element of digital image abuse is also reflected in most of the existing research on sexting and risk, which is also marked by a predominant focus on the sexting practices of children or 'youth' -generally defined as young adults up until the age of 25 (Amundsen, 2019a: 480-481). The participants in the study on which I draw here were aged 18-38 and, as such, they were all legally defined adults. That being said, the literature concerned with child and youth sexting practices provides important insights regarding sexting and risk and, owing to this, it also offers a significant resource for making sense of the interview material presented in this article. Here, it is also worth noting that, while sexting is a complex practice that can occur in the context of all forms of human relations, this article will only address examples of sexting practices as they occur in the context of heterosexual relations of a romantic and/or sexual nature. This focus is determined by the fact that such heterosexual relations provide the main context in which the interviewees referenced here considered it too. Moreover, sexting is an activity that can be consensual, non-consensual or engaged with due to peer pressure (Bond, 2016). While stressing that the sexting activities discussed in this article were claimed by my interviewees to be entirely consensual, I also acknowledge that consent is an inherently contested concept (Popova, 2019). Moreover, the concept of consent might cover activities that, while being willingly engaged with, are still not chosen completely freely. For instance, some of the participants in the research project that this article stems from -but not quoted here -both stressed that it was their choice to engage in certain sexting activities and that they felt pressured into sexting due to societal or partner expectations (Amundsen, 2019b).

Literature review
Broadly speaking, the existing qualitative research on sexting and risk can be divided into three main subsets. The first subset is concerned with risk discourses as they appear especially in educational campaigns and in media reports on sexting. These studies are critical towards the common framing of sexting as an inherently risky practice strictly associated with the social and inter-personal risks of bullying and/or reputational damage (Albury et al., 2017;Angelides, 2013;Dobson and Ringrose, 2016;Ringrose and Harvey, 2015: 205;Salter, 2016Salter, : 2726, especially as these might occur following a case of digital image abuse (Citron, 2014: 7-9). As such, this literature points out that these very campaigns, which aim to warn mainly youth about the dangers of sexting, might in fact cause as much harm as they set out to mitigate: in strictly focusing on how particularly youth can protect themselves from sexting related risk, these campaigns are critiqued for operating so as to cast young people's sexualities as something to be regulated and kept under control through extensive risk management (Hasinoff, 2015;Lee et al., 2013;Ringrose et al., 2013: 307).
The second subset of this literature is concerned with the ways in which discourses on sexting risk appear to focus more on the practices of women than, primarily, on those of men. A potential consequence of this gendered focus, is the fact that it is particularly young women who are called on to self-manage by way of risk mitigation (Amundsen, 2019c;Ringrose and Harvey, 2015). As follows, research has also explored how these gender unequal discourses on sexting risk shape, and form attitudes and responses to sexting in relation to gender (Lippman and Campbell, 2014). Much of this research has found that sexting practices are marked by a significant double standard with regard to gender, with women's practices being associated with the risk of their being branded as 'promiscuous' by peers and by an unknown public, with men being less likely to face similar social sanctions over their private sexual images (Ravn et al., 2019;Ringrose et al., 2013;Salter, 2016;Setty, 2019).
Finally, the third subset of the literature on sexting and the various risks associated with potential further non-consensual sharing of the sexting material, is particularly concerned with how people, and especially women, respond to risk discourses with regard to sexting, especially by engaging in risk mitigating practices (Setty, 2018). For example, this literature describes how risk mitigating activities can include careful considerations of which potentially identifying characteristics to depict in their private sexual images (Renfrow and Rollo, 2014). It also includes discussions of how participants claim to rely on trust in their sexting partner that they would not share the image further in public without their consent (Amundsen, 2019a). Here, I will further expand on the insights provided in this subset of sexting and risk literature as it is concerned with how people -and especially women -respond to and engage with the prevailing discourses regarding sexting and risk.
Drawing on material from individual interviews with adult British women about their experiences of hetero-sexting, I illustrate how sexting risk was not strictly cast by my interviewees as something negative that they should mitigate and avoid. Sexting risk was also understood as a resource for digitally mediated intimacy. In both expanding and problematising existing approaches to sexting risk as a primarily negative phenomenon, this article makes a unique contribution to the literature on sexting and risk by illustrating how sexting risk can operate as a source in the building of emotionally close heterosexual relations. Moreover, in introducing an innovative perspective on sexting risk as a resource for intimacy, this article will make a significant contribution to the field of mediated intimacy research (Attwood et al., 2017;Barker et al., 2018;Chambers, 2013) by showing how digital mediation informs how intimacy is both practised and perceived.
I commence this article with an introduction to the methods for data collection and theory used as means of analysis. Next, I draw on my interview to material show how my interviewees cast female-to-male sexting practices as a form of 'mediated intimacy work' that enable women to both reflect, establish and enhance notions of intimacy in their heterosexual relations. In doing so, I demonstrate how these interviewees saw hetero-sexting risk not simply as an obstacle to this kind of intimacy work, but also as a resource for intimacy. Drawing on this finding I argue, by way of conclusion, that my interviewees' accounts of hetero-sexting not only work to illustrate women's complex understandings of risk in mediated intimacy work generally and hetero-sexting specifically, but also the unequal distribution of power in the heterosexual intimate domain according to gender. Here, I also claim that, in blurring the distinguishing lines between public and private and in extending the temporal dimensions of risk and vulnerability, digital mediation enhances the existing elements of risk and vulnerability in intimacy work. As such, it also operates so as to further entrench existing gender inequalities in the heterosexual intimate domain.

Methods and means of analysis
This article is grounded in a broader exploratory project on adult women's experiences of using new media and digital technology in their romantic and/or sexual relations. The project was conducted with a particular focus on women's accounts and experiences of sexting and the quotes included in this article are drawn from the individual semi-structured interviews that I conducted for this study. Prior to each interview, I informed interviewees that -while I was interested in exploring the influence of new media and digital technology on women's romantic and/or sexual relations more broadly -I was particularly concerned to learn more about their experiences of creating, sending and receiving private sexual images, through the use of new media and digital technology.
Before I commenced on the participant recruitment for these interviews, the project was granted ethical approval by the Department of Sociology of the University of Cambridge's Ethics Committee. I then recruited for a broad range of interview participants, with the only requirements being that participants self-defined as 'woman', were aged 18 and over and were based in Cambridgeshire, United Kingdom. Purposive and snowball sampling were used to connect with interview participants and, between June 2016 and February 2017, 44 individual interviews were conducted. Each interview was recorded and transcribed verbatim, before I did a thematic analysis of every transcript with coding software Atlas.ti (Braun and Clarke, 2006). All participants in this research project were self-selecting -they were not offered anything in return for their participation, beyond drinks and snacks during interviews. Prior to every interview each participant signed an informed consent form.
The women interviewed for this study were aged 18-38. Out of the 44 interviewees, 22 were single and 22 were in relationships. Thirty-eight of the women had no children, with six having one or more children. The vast majority of the sample was highly educated: 42 of the 44 women had one or more higher education degrees and/or were currently undertaking one. Forty-three of the women were employed or currently undertaking a higher education degree, with one of the women identifying as a 'stay-at-home mother'. One interviewee identified as Mixed British/Asian, two as East Asian, two as Mixed British/African, four as South Asian, and 35 as White British or White Other. In terms of sexuality, two of the interviewees argued that they were 'not sure' about their sexualities, with one identifying as lesbian, 1 as 'queer', 10 as bisexual, and 30 as heterosexual. Even though this means that several of the interviewees were, had been and/or desired to be involved in same-sex relations, they almost exclusively addressed sexting as a heterosexual activity. When I asked them why they thought this was the case, the majority claimed not to be sure and/or made it clear that this was not something that they had given much thought. Overall, the participants contributing to this study were predominantly White and highly educated, and it is likely that this has informed the findings presented here. Thus, in the final section of this article, I suggest some ways to address these limitations in future research. To protect the confidentiality of every interviewee, each has been assigned a pseudonym consisting of a randomly allocated standard British name.
This research project was very much informed by my particular interest in the intersection of gender and power, especially with regard to the roles that these play in practices of digitally mediated intimacy. Following John Thompson (1995: 13), I understand power as 'the ability to intervene in the course of events and to affect their outcome'. The interviews were shaped by an open interview guide, centred on questions about the interviewees' perceptions and experiences of digitally mediated intimacy and sexting practices. I approached interviewing more as a practice of listening than as one of asking a set of given questions, meaning that each question asked was crafted in response to what the interviewee first told me (Roulston, 2010: 139). Owing to this, each interview ended up being a deeply reflective practice, and no interview was similar to another (Roulston, 2010: 4). That being said, some themes or issues occurred frequently in every interview and, when that happened, I picked up on these themes to probe further. Two such themes were the issues of hetero-sexting as mediated intimacy work and of sexting-induced vulnerability as a resource for women to draw on in order to enhance the significance of their hetero-sexting practices as the performance of intimacy. In what follows, I analyse these themes by drawing on the concepts of 'emotion work' (Hochschild, 1983: 7) and 'male emotional domination' (Illouz, 2012: 104-105), which I explain in the next paragraph. In doing so, my aim is not to generate insights that can be generalised to tell us something about the experiences of all women as a whole -my sample is too small and narrow for empirical generalisation. Rather, I follow Shani Orgad (2019: 17-18) in treating my interview material as 'focused points of human experience that can teach something about a more general problem' (Sennett and Cobb, 1972: 45) -namely that of women as intimacy workers in a heterosexual intimate domain that is increasingly digitally mediated. Thus -while acknowledging that 'gender' is a socially defined concept constructed at the intersection of a range of 'axis of differentiation' (Brah and Phoenix, 2013: 76) -the analysis conducted in this article will focus on articulations of female gender as they are socially defined (Butler, 1990) and discussed during interviews in relation to notions of intimacy, risk, vulnerability and trust.
In developing an understanding of my interviewees' conception of women's heterosexting practices as a form of mediated intimacy work, I draw on the concept of emotion work (Hochschild, 1983: 7). This concept entails acts that are carried out in a private context and that involve 'the management of feeling to create a publicly observable facial and bodily display' so as to enable the further development of particular feelings in others (Hochschild, 1983: 7). Emotion work is therefore thought to be of use value, meaning that it is a practice through which something can be achieved (Hochschild, 1983: 7). In terms of my interview material, the resource that the interviewees were the most eager to obtain in conducting their emotion work was a sense of intimacy, which is why I refer to women's sexting practices as a form of mediated intimacy work. In coining the term 'mediated intimacy work', my aim is also to highlight how the use of new media and digital technology for such intimate purposes distinguishes it from other forms of intimacy work. As I will discuss, digital mediation shapes and forms practices, as well as perceptions, of intimacy. When using the term 'intimacy', I follow Lynn Jamieson (1998: 1-2) in understanding it as 'a very specific sort of knowing, loving and 'being close to' another person' -an emotional state that is achieved through the mutual sharing of one's innermost thoughts and feelings with another. This understanding of intimacy is hence based on disclosure -on the sharing of personal information that otherwise would be kept private (Attwood et al., 2017: 249). Following this perception of intimacy, intimacy is hence perceived as an emotional state that is grounded in a willingness to take risk by rendering oneself vulnerable to someone else. Understood as such, intimacy is also based on one's willingness to place trust in this other person that they will not act on their vulnerability in any way that is contrary to one's intimate intentions. To make sense of this reliance on trust in intimacy, I draw on Denise Rousseau et al'.s (1998: 395) understanding of trust as 'a psychological state comprising the intention to accept vulnerability'.
The emphasis on vulnerability and trust in intimacy becomes particularly prominent in relation to the interview material presented in this article, especially because my interviewees described the mediated intimacy work conducted through hetero-sexting primarily as the doing of women. Crucially, this means that -in order to establish and enhance intimacy -it is primarily women who render themselves vulnerable and hence reliant on the trustworthiness of their male partner. To make sense of this unequal distribution of mediated intimacy work in hetero-sexting, I draw on Illouz's (2012: 104-105) understanding of male emotional domination in heterosexual relations. According to Illouz (2012: 104-105), this concept captures that which occurs when one side or partner is capable of controlling the emotional interaction in a relationship by remaining more detached than the other, hence demonstrating their superior ability to make free choices in the given sphere and also to constrain the free choices of the other. Following Illouz (2012: 104-105), women's inclination to conduct mediated intimacy work and men's relative lack of desire to engage in similar practices might thus be seen as an example of the unequal distribution of power between women and men in the intimate domain. In this article, I draw on the example of women's practices of hetero-sexting to show how the risks generated by sexting as a form of mediated intimacy work affect women more strongly than men and, consequentially, render women more emotionally dependent on their male partners. However, I am also going to problematise any understanding of sexting risk and female vulnerability as a possibility with inherently negative consequences.
Rather, in what follows, I refer to my interviewees' accounts of hetero-sexting to show how risk has come to take on a complex and multi-faceted meaning in the heterosexual intimate domain as both a vehicle and an obstacle for female power.

Sexting risk as a reflection of intimacy
During my conversations with the interviewees about their practices of hetero-sexting with male relations, one of the most prominent themes of conversation was the concept of risk, closely associated with an understanding of vulnerability. All the interviewees were conscious of and very clear to me about the fact that women's creating and sending of private sexual images to men exposed them to the risk that their images might be shared further in public without their consent. Relatedly, we also discussed the sense of vulnerability that this experience of sexting risk generated in them. When we discussed the concept of vulnerability, it appeared to me that my participants operated with a highly complex understanding of this concept. While acknowledging that their sexting-induced vulnerability put them at risk of extensive digital image abuse, it did not necessarily render them more reluctant to engage in risky practices of hetero-sexting. Rather, many of the women, like Hannah (mid-20s) cast the risk in hetero-sexting as what actually made this activity exciting, stating that 'the fact that um you're trusting somebody else with that image of you in compromising positions, [ . . . ] I guess the risk factor is kind of what makes it exciting'. Instead of casting vulnerability as simply a detriment to intimacy, many of my interviewees also presented it as what makes hetero-sexting worthwhile, primarily because it is a means through which intimacy can be performed. This point can, for instance, be illustrated by Imogen's (mid-20s) quote below, in which she explains that the level of sexting-induced vulnerability and risk that she was willing to take in hetero-sexting could be seen as an illustration of the level of intimacy already established in the given relation: I send sexual images differently depending on who I'm sending them to. So like, in long term committed relationships I wouldn't have a problem sending them over text or WhatsApp where they can be like saved and stored. And also, I wouldn't have a problem with [ . . . ] it being easy to identify me in them. Like, I wouldn't mind having my face in them for example. But if it was kind of a flirty thing with someone I wasn't that serious with, I would probably still send them, but I would probably send them on Snapchat and I probably would not have my face in them. Just cos then even if someone were to screen-shot them, I could just be like 'well, there is nothing you can do with that because you can't identify it as me'. Kind of thing. So I think that's my experience, I am kind of conscious of being careful about the way I use it and like I would only send like permanent pictures to people that I was like in a kind of committed relationship with that I trusted to not abuse that even if we broke up or whatever. Um so I think my experience has been completely positive in the sense that [ . . . ] I have never felt worried about sharing sexual images.
As we can see from Imogen's quote above, the risky content of her private sexual image was thought to operate as a representation of how close she felt to the intended receiver. The closer she felt to the person for whom the private sexual image was meant, the more significant was her experience of trust, and the more risk she was willing to take in terms of what to show in her private sexual image. Risk, that is, was seen as an opportunity to enact the level of trust and intimacy already established in the relation. Understood as such, this quote by Imogen is also a good example of how my interviewees saw sexting-induced vulnerability and risk as a reflection of the level of intimacy in the given relation. Intimacy, that is, determined how much risk the women were willing to take by rendering themselves vulnerable through exposure. This quote is also an illustration of how my participants' experience of intimacy also tended to be reflected in their choice of mediating technology. In 'long term committed relationships', Imogen made use of digital messaging services like WhatsApp, which allow users to save and store private sexual images -something that increases the risk of them being shared further in the future. However, in relations of a less serious nature -meaning that they are marked by a lower level of trust -Imogen was more likely to make use of messaging services like Snapchat, where she can send private sexual images that disappear after a limited amount of time and hence only are available to the recipient for a brief moment. (That being said, as Imogen also points out, the recipient is still able to take a screenshot of the photo, meaning that they might nevertheless be able to save and store a copy of it).

Sexting risk to establish intimacy
Like Imogen, several of the women participating in this study also explained how they acted upon their own vulnerability in order to foreground the significance of their hetero-sexting practices as mediated intimacy work. As such, the risk of further nonconsensual sharing also provided them with the opportunity to 'stretch the conventional meanings' of their own vulnerability (Ringrose, 2011: 101). That is, the women explained how they operationalised vulnerability as a means to stress the significance of hetero-sexting as mediated intimacy work, hence pushing against the regulating abilities of sexting risk. Indeed, as illustrated in the following quote, also by Imogen (mid-20s), the majority of my interviewees appeared to associate sexting-induced vulnerability not just with the existing level of intimacy in the given relation, but also with its establishment: The more risk Imogen was willing to take in terms of what to disclose, the more serious and intimate she considered her relationship to be. Evidently, this was made clear to her partner as well: Imogen explained to him that, in making herself vulnerable to public identification upon the event of digital image abuse, she was also showing him how close she felt to him. Furthermore, in doing so, she was placing a moral obligation on her partner to protect the private sexual images, thus emphasising her expectation that this notion of intimacy and trust would continue to develop into the future.

Sexting risk to enhance intimacy
Imogen's understanding that a willingness to take risk is a way to express and establish intimacy was reflected in the majority of my interviews. Many of the women in this study described the risk inherent in their practices of hetero-sexting as something that they actively employed as a means to highlight the significance of their images being created as mediated intimacy work. These images were hence created not just so as to establish intimacy or to reflect the level of intimacy already existing in the relation, but also so as to further enhance it through their expression of trust in their partner. For instance, in answering my question of why it is that people create private sexual images to share with their romantic and/or sexual partners, Fiona (mid-30s), explained it as such: To Fiona, the appeal of creating and sending private sexual images to a male partner was directly associated with the employment of risk as a way of establishing trust and enhancing intimacy. This view was echoed in many of my other interviews, where the women claimed that by sending more explicit and/or identifying material -for example by showing their face or tattoos -they were also contributing to the further development of intimacy in their heterosexual relations. In these narratives, hetero-sexting as mediated intimacy work was directly linked to women's active use of their own at-risk status as a means to express trust and, in doing so, spur on the development of intimacy. 4 As argued by Jade (early 20s): suddenly have a responsibility now that they didn't have before, which is to earn the trust that you've given them or to [ . . . ] respect it [or] something.
As we can see from Jade's quote above, in creating private sexual images she was in fact drawing on her at-risk status as a means to both enact and establish trust and, with that, trigger the further development of intimacy. Like Imogen (mid-20s) and Fiona (mid-30s), Jade thought of this practice as a way to place a moral responsibility on the receiver of the private sexual image. In sharing this image with him, she also acted so as to bestow on him some responsibility to protect it from further distribution, and thus also to protect her from the harms associated with digital image abuse (Citron, 2014: 7-9). Indeed, the image was created with the expectation that the future receiver would feel obliged to 'earn that trust' by not acting on Jade's vulnerability, and by treating Jade and the private sexual image with the respect she felt that they deserved. As such, the private sexual images were thought to work as a means to bring the creator/sender and the receiver closer together, through a shared awareness of the risk involved for the woman in creating and sending the image and the trust bestowed on the man in her sharing it with him. This quote is thus a good illustration of how the risk inherent in hetero-sexting for women was seen as a resource to be drawn on so as to create a stronger bond between the women and their male partners.

Mediated intimacy work as enabled by risk-trust negotiations
Many of my interviewees operated with a similar understanding of sexting related vulnerability as a means to express and enhance the level of perceived intimacy between them and their partners. The level of risk that they were willing to take in creating and sending private sexual images -hence rendering themselves vulnerable -was generally used as a means to indicate how much trust they had in their male partner, and hence how close they felt to him. The more vulnerable they were willing to make themselves in their private sexual images by, for instance, including identifying characteristics like their face, the more significance they granted these images as mediated intimacy work. In other words, by rendering themselves identifiable and therefore even more vulnerable to abuse upon the event that their images were shared non-consensually and in public, they also aimed to express, establish and/or enhance intimacy. In fact, this risk-trust negotiation had become so integral to their understandings of sexting as mediated intimacy work that private sexual images that failed to express a level of vulnerability generally received no uptake as intimacy, precisely because they did not simultaneously express a level of trust in the receiver. This point is clearly demonstrated by the women's attitudes to the private sexual images that they sometimes received from their male sexual and/or romantic partners.
The private sexual images that my interviewees claimed to receive from their male partners as part of their hetero-sexting activities most often consisted of depictions of male genitals, a genre of private sexual images commonly referred to as 'dick pics'. Generally speaking, the women claimed that the negotiations of risk and trust that are so integral to their own private sexual images were lacking from their male partner's private sexual images, primarily because private sexual images only depicting a penis involved no taking of risk by showing any identifying characteristics. As such, their images did not communicate a level of trust in the intended receiver through reduced anonymisation and enhanced vulnerability. The depictions of men's dick pics hence also failed to take on the meaning of mediated intimacy work. This view was echoed in several interviews, but perhaps most clearly in my interview with Whitney (late 20s): I'm not massively fussed about receiving them [men's private sexual images] [ . . . ]. Uh I would actually much rather see their face and like have kind of friendly or funny pictures, cos I feel like that's actually more intimate than sending an anonymised nude.
According to Whitney, an image of a man's face would have more meaning to her than an 'anonymised nude' because the face, unlike an unidentifiable 'nude', could express some level of an intimate connection between sender and receiver. The enjoyment that Whitney derived from receiving a man's sexual self-representation did not originate in the nudity itself or in the sexual element of the picture, but from its operating as an expression of an emotional connection between sender and receiver, from its constituting an act of mediated intimacy work. Importantly, Whitney's quote is illustrative of what I take to be the women's general perception of men's private sexual images, namely that there was a disconnect between what the women wanted from these images -an expression of intimacy through the negotiation of risk and trust -and what they usually received: a purely sexual and anonymous display of a man's penis. While the depictions of men's private sexual images most often were highly sexually expressive, they did not communicate much emotion. Rather, the missing depiction of a risk-trust negotiation in their private sexual images operated so as to express a level of emotional detachment, thus rendering the performative element of these images inherently different to those created by the women as mediated intimacy work.

Discussion and conclusion
In this article, I have shown how my interviewees operated with a complex understanding of sexting risk, both as a negatively loaded possibility to avoid and as a resource on which to draw in order to enhance the significance of their hetero-sexting as mediated intimacy work. This perception of sexting-induced vulnerability and risk as an opportunity for intimacy is resembling the way in which risk is framed in some economic parlance, where greater risk is associated with the possibility of greater financial profit (Lupton, 2013: 9). That is to say, while the materialisation of risk remains cast as a negative, potential outcome, the very engagement in risky activities is perceived as a positive activity due to it being understood as potentially profitable. Introducing a more positively loaded conceptualisation of hetero-sexting risk, this article has made an important contribution to the existing research on how people respond to risk discourses with regard to sexting, wherein sexting risk is generally cast as a negative possibility that those who engage in it seek to avoid (Amundsen, 2019a;Renfrow and Rollo, 2014;Setty, 2018). In bringing in this new conception of sexting risk, the findings presented in this article are also in line with those presented in other articles on online sexual cultures by, for example, Marijke Naezer (2018) and Matt Hart (2017). Their research -which explores young people's engagement in 'online sexual activities' (Naezer, 2018) and the posting of 'naked self-photographs' to Tumblr blogs (Hart, 2017) -show how research participants operate with understandings of risk that are infinitely more complex and multi-faceted than are perceptions of risk simply as something to avoid.
The association of intimacy with female vulnerability that appeared within my interviewees' accounts is, in itself, nothing new. For example, Sarah Bracke (2016: 67) cites Robin James (2015) in arguing that normative femininity, as traditionally understood, was associated with the embodiment of vulnerability. As such, women were encouraged to act in a manner that foregrounded their timidity and fragility. However, most academic literature on the contemporary construction of normative femininity under conditions marked by the prevalence of a distinctly postfeminist sensibility, argue that this association of femininity with vulnerability is now rejected and replaced by the figure of the active, freely choosing, and empowered female individual (Baker, 2008;Gill, 2007;Stringer, 2014). What sets my interview material apart from these other conceptions of vulnerability is therefore that, here, agency is promoted not through the rejection of vulnerability, but through its being embraced. My interviewees claimed to actively use their own vulnerability by placing themselves at risk as a means to enact intimacy and, as such, their own power and agency within the heterosexual intimate domain. Vulnerability was hence both drawn on and dismissed within the very same accounts, as it was applied as a means to perform a new form of normative femininity, namely that of the agentic intimacy worker.
The understanding of women as intimacy workers presented here has much in common with Rosalind Gill's (2009: 351-352) concept of 'intimate entrepreneurship', an interpretative repertoire that casts relationships as a form of labour and the maintenance and establishment of relationships as goals that demand 'research, planning and strategy'. The intimacy worker, as she was presented to me during interviews, appeared as a feminine subject engaged in the continuing task of reflecting, establishing and enhancing intimacy through the engagement in risky practices of hetero-sexting. Indeed, heterosexting as mediated intimacy work was presented as a project for women to plan, regulate and manage in a highly individualised manner, much in line with the neoliberal ideals articulated through postfeminism (Elias et al., 2017: 39;Rose, 1996: 157). As such, hetero-sexting also provides a further illustration of Illouz's (2012: 104-105) claim regarding male emotional domination by showing how mediated intimacy work is something that it mainly is women who take on in the heterosexual intimate domain, not men. Essentially, this also means that it is especially women who render themselves vulnerable by exposing themselves to sexting risk and reliant on the trustworthiness of their male partners. While digital mediation creates new opportunities to conduct intimacy work -like hetero-sexting -it also works to further entrench existing inequalities in the heterosexual intimate domain.
Arguably, the digital mediation of hetero-sexting as intimacy work rendered its gendered dimensions -particularly with regard to the distribution of risk and trust -even more explicit, both because digital affordances has made the violation of sextinginduced trust easier, and because the technology used for intimate disclosure might be considered inherently unsafe and unpredictable. For example, the affordances of much new media and digital technologies complicate the boundary between public and private -thus introducing to acts of mediated intimacy work a heightened risk that private and intimate utterances like sexting material appear in public. Moreover, as digitally mediated communication occurs across space and time, digital mediation also works to stretch notions of risk and vulnerability in experiences of intimacy into the unforeseeable future, hence expanding their temporal dimensions and amplifying their role in intimacy. For those about to embark on further research on the topics of sexting, intimacy and digital mediation, the insights presented here constitute an excellent starting point, precisely because it shows how the affordances, nature and material aspects of digital mediation shape practices and perceptions of intimacy. Future research could, for example, look more closely at these mediating technologies in order to examine more closely how technologies inform intimacy and, relatedly, interpersonal gender relations. It would also be interesting to see future research that addresses the limitations of this study, for instance by exploring how sexting risk is experienced in romantic and/or sexual relations between women, by women with lower education levels, and/or by women who are not predominantly White. There are clear indications that understandings and experiences of sexting risk might vary between different sub-categories of women, and this requires further exploration. For example, the hacking of comedian Leslie Jones' website, followed by the distribution of racist messages alongside her stolen private sexual images, suggest that acts like intentional digital image abuse are not just informed by harmful ideas about gender, but also about 'race' and ethnicity (Lawson, 2018;Levin, 2016). Going forward, it would also be very interesting to see similar research on hetero-sexting conducted with a focus on the experiences of adult self-identifying men. The findings from such research would not only contextualise the findings presented here. It would also enable a testing of the assumptions presented in this article about the ways that gender and power inform practices of mediated intimacy work in heterosexual relations.

Author note
I confirm that this work is original and has not been published elsewhere, nor is it currently under consideration for publication elsewhere.