Tanya Kant
University of Sussex, Brighton, UK

[Abstract]

‘The film you quote. The songs you have on repeat. The activities you love. Now there’s a new class of social apps that let you express who you are through all the things you do’ (Facebook, 2014).

Introduction: Performing the Self in the ‘Like’ Economy

Facebook’s apps network – an ‘ecosystem’ designed to ‘deeply integrate’ (Zuckerberg, 2011) commercial lifestyle, gaming, entertainment and shopping applications into Facebook – is a key component of Facebook Inc.’s current operational and economic infrastructure. As Mark Zuckerberg repeatedly emphasised during the 2014 F8 conference keynote, Facebook apps play an essential role in Facebook Inc.’s newest expansion strategy – that is, to become a ‘cross-platform platform’ (Zuckerberg et al., 2014) that connects not just friends, family and acquaintances but the millions of platforms, websites, ‘stacks’ and services that currently constitute the web. To date, the external products and services that apps integrate into Facebook include (but are not limited to) games, movie, book and music services, ticket purchasing programs, bookmarking software, photo editors, health and fitness trackers, events managers, comic strip creators, lifestyle forums, interactive cook books and stress relief programs – there are over seven million apps currently available on Facebook (Stasticbrain, 2014), which generate an income of around $3 billion for Facebook and its third party associates (Zuckerberg et al., 2014).

As such, Zuckerberg’s ambition to become the ‘platform of platforms’ extends beyond mere rhetoric – Facebook already underpins a new form of socio-economic connectivity that constitutes what Gerlitz and Helmond call the new ‘”like” economy’ (2013), that is, a cross-platform, socio-technical ‘infrastructure that allows the exchange of data, traffic, affects, connections, and of course money’ (2013: 1353). Though Gerlitz and Helmond cite Facebook’s ‘like’ and ‘share’ buttons as the primary nodes of mediation in this developing ‘social web’, apps also play a crucial role in connecting this complex socio-technical infrastructure. As Fagerjord notes, apps – once known as applications – no longer simply function as ‘software designed to do a specific task’ (2012: 2), they are now also utilised as profitable, ‘sticky’ and frictionless bridges, used to carry massive amounts of data produced and consumed by users between ‘integrated, monopolistic outlets’ (2012: 4) such as Facebook.

Facebook apps, however, were not simply designed to situate Facebook as the ‘cross-platform platform’ – they were also apparently created to facilitate and enhance new modes of user self-expression. Rolled out in 2011 along with Timeline, Facebook’s newest profile manifestation, apps were pitched to users as ‘the perfect way to express who you are’ (Zuckerberg, 2011). As the introductory quote of this paper emphatically suggests, apps are discursively framed by Facebook as tools that let users articulate – and more importantly publicise – ‘who they are’ by accommodating any user lifestyle preference, ‘interest token’ (Liu, 2008), socio-cultural taste, consumer choice or even affect; ‘your runs, your bike rides, your cooking, your eating, your sleeping, your happiness, your fashion, anything you want’ (Zuckerberg, 2011, my emphasis).

Facebook’s predictably lofty promises to enhance user self-expression should of course be critically scrutinised – there are a number of economic, techno-architectural and socio-cultural implications that call Facebook’s claims in to question, as this paper will explore. However, the capacity of social networks sites (SNSs) such as Facebook to accommodate and maintain some form of selfhood is well-established – critiques of how individuals construct, negotiate and perform online identity have existed for many decades. As scholars such as J Sage Elwell (2014) Rob Cover (2012) and Theresa Sauter (2013) note, as the web becomes a pervasive and ubiquitous presence in individuals’ everyday lives, the distinction between ‘online’ and ‘offline’ self-performance is becoming increasingly unhelpful; ‘day-to-day processes of self-formation are becoming more and more implicated with new digital tools’ (Sauter, 2013: 2). Furthermore, these digital tools are increasingly underpinned by what Tarleton Gillespie calls ‘algorithmic logic’ – that is, a socio-technical rationale enacted, governed and maintained by both human and computational actors (2014). This paper considers Facebook apps as one such automated ‘tool’– and potential algorithmic actor – in the writing of the self in everyday life.

As scholars such as Sauter (2013) note, a plethora of potential ‘technologies of the self’ — to use Foucault’s (1994) much-mobilised term — are embedded in the operational structure of Facebook: status updates, videos, photos, sharing links, location tagging and ‘likes’ (amongst other functions) all have the potential to facilitate facets of user self-expression. As I shall further explore, Facebook’s cross-platform apps can also be considered to be one such ‘technology of the self’, since they carry a unique set of parameters as tools for self-writing through their automated operational protocols and capacity to connect across platforms. For example, many Facebook apps have the ability to post automatically published status updates (herein referred to as ‘autoposts’) to an individual’s Facebook friends network, often without their knowledge or immediate consent (at the time of posting). Autoposts by apps can take on a number of forms, but most refer to an in-app action or achievement by a user and are written on behalf of a user in first or third person; for example, ‘xxx xxxx is listening to Serious Time by Mungo’s Hi Fi on on Spotify’, ‘Batman & The Flash: Hero Run – I’ve just scored 22, 323 points!’ or ‘I’ve just run 5.99 miles on MapMyRun’ (see Figure. 1).

Figure 1. Examples of autoposting from Facebook (2014).

The ability of apps to algorithmically write in the user’s stead, as well as connect across platforms, gives rise to a host of fresh critical implications in regards to app use as a way ‘to express who you are’: what does it mean to use apps as tools for self-expression? If, as Cover proposes, ‘the establishment and maintenance of a profile is not a representation or biography but performative acts, which constitute the self and stabilise it over time’ (2008: 181), then what does it mean to a have an app algorithmically perform an act of selfhood as part of users’ Facebook profiles? [1] What does it mean for an app to automatically yet unintentionally give away our ‘guilty pleasures’ to our online audiences – our ‘trashy’ song preference or our pornographic pleasures? What kind of self can be constituted under ‘the logic of the algorithm’, as well as through the social connectivity that currently pervades the web?

Drawing on the accounts of sixteen Facebook users, this paper explores the performative implications of apps as tools for self-expression. These vignettes are primarily structured around participant experiences of largely unwanted autoposts by apps and also examine participant strategies of ‘coping’, resisting or exploiting the connection between their apps and Facebook accounts. Finally, the paper explores the impact that autoposts have on how they view other people in their networks. As I will further explore, participants recounted a number of complex, tense and often unwilling encounters with autoposting apps on Facebook – including apps disclosing ‘guilty pleasures’ such as trashy songs or sexually suggestive content to participants’ Facebook friends, Spotify ‘adding an event’ to a participant’s ‘past’ and the framing of other people’s game app posts and invites as ‘chavvy’.

Participants’ accounts suggest that in many instances, autoposts by apps work to intervene and at times disrupt the carefully staged identity performances that users commonly enact on Facebook (Van Dijck, 2013). By considering critical notions such as ‘context collapse’ (boyd and Marwick, 2011), ‘taste performance’ (Liu, 2008) and ‘grammars of action’ (Agre, 1994), I will argue that apps function not just as tools for self-expression, but as unwanted actors in the writing and performing of selfhoods on Facebook. If utterances of selfhood on Facebook are performative, then apps hold the constitutional capacity to restrict and even regulate the self in ways that transcend the boundaries of Facebook.

Facebook Apps as Tools for Self Performance

Since the popular emergence of computer mediated communication (CMC) technologies, notions of what it means to express selfhood or identity online have been explored, celebrated and berated by a huge number of scholars (Sauter, 2013: 3). As media theorists such as Frances Dyson have noted, the early internet was popularly perceived as ‘a medium that dissolves race, class and gender, producing instead “liquid” identities’ (2005: 86). Despite these hopes, the contemporary web that most users experience is not populated by ‘liquid’ identities or multiple subjectivities – as Marwick (2013), van Zoonen (2013), Karppi (2012) and van Dijck (2013) note, SNS sites are predominantly populated by very much identifiable individuals who are expected, as Mark Zuckerberg famously claimed, ‘to have one identity’ (cited in Van Dijck, 2013). As Marwick notes, though ‘the move to commercial social software such as social network sites, blogs, and media-sharing services has brought with it an impetus to adhere to a fixed, single identity’ (2013: 368), this commercial impetus does not fit how most individuals experience identity formation and construction in everyday life. Similarly, Van Dijck states:

The mantra of people having one authentic or ‘true’ identity… betrays a fundamental misjudgement of people’s everyday behaviour. Ever since Goffman, it is commonly accepted that people put on their daily lives as staged performances where they deliberately use the differentiation between private and public discursive acts to shape their identity. (2013: 212)

Despite these assertions that identity construction consists of a ‘staged performance’ rather than a single or ‘authentic’ identity, Facebook insists that their tools – such as apps – can assist users in ‘expressing who they *are’, implying that users do not ‘perform’ their identity on Facebook; they are *the selves they portray on Facebook. Participants’ testimonies suggest that this is far from the case, as I will further explore, yet Facebook’s discursive construction of identity suggests that a users’ single, verifiable self can indeed should be expressed via their Facebook use.

The commercial benefits for the harvesting of users’ profile information are sizeable. As Karppi (2012) notes, this data can be efficiently translated into monetary value through targeted advertising, behavioural marketing and other personalisation strategies. However, such information does not just generate profit for Facebook and its associates, it also holds expressive and affective value for users; as Liu recognises SNS profiles constitute for users ‘an expressive arena for taste performance’ (2007: 252). Yet though Facebook and its advocates celebrate the input of this information as beneficial to user self-expression, scholars such as Marwick, Cover, Karppi, John Killoran (2002) and Laron Janier (2010) note that Facebook’s attempts to determine users’ interests can be critiqued as standardising and restrictive. For example Marwick states that:

Facebook profiles cannot be altered and thus all adhere to the look and feel of the site. As a result, user customization is restricted primarily to filling out predefined fields, such as favourite books, music, television and films. (2013: 14)

Such criticisms highlight the homogenising influence of commercial interests on profile pages that work to restrict the modes of self-expression available to users of SNSs. These critics suggest that Facebook’s operational architecture may indeed provide a means for users to express elements of their selfhood, but under a technological regime that restricts the construction of identity to normative, regulatory and commercially viable frameworks. Furthermore, Software Studies theorists such as Philip Agre and Kitchin and Dodge have proposed that these computational frameworks could have performative effects for the users that ‘act’ through them. Agre’s ‘grammar of action’ model proposes that the implementation of computational architectures such as Facebook can ‘constitute that reorganization of… existing activity’ (1994: 11) suggesting that self-expression can be rearticulated, and therefore regulated, to fit the computational grammar embedded into technological frameworks, networks and interfaces. Through such grammar, as Karppi states, ‘impressions of the self are built according to the platform’ (2012: 293).

What role then do apps play in these frameworks? Apps have been scrutinised by scholars from a plethora of disciplines for their effect on factors such mental health (Richardson, 2013), familial well-being (Wilson and Yochim, 2013) and identities (Walker, 2014), as well as for their role in ‘gamifying’ socio-cultural practices such as physical self-improvement techniques (Williamson, 2015) and learning pathways (Fallon, 2013). Though the content of various apps has been critiqued through various theoretical frameworks for some time, the form and function of the app in and of itself – as an object or a relation (Clough, 2014) – is a newly-developing consideration, recently explored by scholars such as Manovich (2014), Matviyenko (2014), and Mellamphy and Biswas Mellamphy (2014). This paper seeks to contribute to this developing field of app studies by considering both the content and relations that cross-platform apps present to users, specifically in relation to identity performance, as well as through the lived experiences of Facebook apps users. It is Gerlitz and Helmond’s emphasis on social-technical relations mobilised by Facebook that makes the ‘like economy’ a valuable framework for considering that connective capabilities of Facebook apps in relation to self-expression. As the vignettes highlight, it is often the cross-platforms connections created by apps (between Spotify and Facebook, Instagram and Facebook, Slideshare and Facebook, etc) as well as the autoposted content disclosed by these apps, that problematises the role of apps in self-articulation and identity performance.

Though Gerlitz and Helmond do not explicitly explore apps as tools for identity performance, they argue that the ‘like’ and ‘share’ buttons used to connect platforms and services currently afford users a very limited ‘horizon of possibility’ (2013:1353) that only allows for the expression of positive – rather than negative or critical – sentiment. They speculate that apps might expand this horizon of possibility, as apps facilitate a discursive and expressive framework that extends beyond merely being able to ‘like’ or ‘recommend’ something. They note that:

… when creating an app, developers are prompted to define verbs that are shown as user actions and to specify the object on which these actions can be performed. Instead of being confined to ‘like’ external web content, users can now ‘read’, ‘watch’, ‘discuss’ or perform other actions. (2013: 1353)

Do apps then hold a way out of the restrictive framework that Facebook imposes on its users? The existence of over seven million apps certainly suggest that apps allow users to ‘express who they are’ by accommodating forms of expressive action that go beyond simply ‘liking’. However, Gerlitz and Helmond highlight that ‘automatic posts’ by apps – that is, the autoposts that take the focus of this paper – may problematise the expressive potentials of apps:

These new apps come with the controversial feature of frictionless sharing and automatically post performed activities to the ticker once users have signed up. (2013: 1353)

The ability of apps to autopost on a user’s behalf on friends’ Newsfeeds or Tickers [2] certainly has sparked controversy amongst Facebook users. Facebook is now downplaying the significance of autoposts in users’ activity, admitting that ‘people often feel surprised or confused by stories that are shared without taking an explicit action’ (Facebook blog, 2014). These sentiments were certainly echoed by participants. Though many participants enjoyed using apps, experiences involving autoposting were overwhelmingly negative, as I will shortly explore. Furthermore, if the ‘performed activities’ of apps are considered in the Butlerian sense – that is as a peformative act that discursively ‘enacts and produces’ the subject which it names (1993: 13) – then autoposts take on a far more profound significance than just being confusing. Apps’ performative capacity to algorithmically write utterances of selfhood on behalf* *of users actually works to negate their expressive potential, by (re)writing selfhoods that conform to the socio-technical grammar of the ubiquitously connected ‘like’ economy.

Method: Investing in the ‘Like’ Economy

In total, sixteen participants were interviewed as part of this research project, all of whom took part in semi-structured, face-to-face interviews designed to explore participant engagement with apps and services on Facebook. [3] Participants were recruited through the ‘Plugged-in Profiles’ research page – a Facebook page established and maintained as part of the project. The page was accompanied by a research survey, which some participants filled out prior to their interviews. All participants were assigned pseudonyms in order to protect their identities.

Establishing a Facebook page involved self-reflexively joining the ‘like’ economy that this project seeks in part to critique. The page gained eighty nine ‘likes’ (that, is users who have subscribed to the page) and calls for participants were seen by over one thousand Facebook users. However, as Gerlitz and Helmond note, the nature of the ‘like’ economy is grounded on social connectivity, meaning that many of my subscribers (and subsequent interviewees) were largely recruited from friend networks close to my own pre-existing network on Facebook [4]. As a result, my participants share some socio-cultural similarities in terms of their demographics – for example are all 24—30 years old and based in the UK. It is thus important to note that the accounts of my participants do not reflect the plethora of possible identities or demographics on Facebook. As is clear from the following vignettes, my participants accounts are embedded within socio-cultural normative tastes and practices that are specific to their lived experiences and are therefore contingent on context-specific parameters of taste, class and cultural preference (amongst other factors). Their responses should thus not be taken as representative of Facebook users as a whole. Rather their testimonies are explored here to highlight the ways in which apps intervene in self-performances that are always-already embedded in pre-existing frameworks of socio-cultural and economic norms, negotiations and practices.

Interviews were semi-structured in order to explore participant experiences of Facebook apps, and more specifically participant accounts of autoposting – that is, instances wherein automated status updates were implemented, produced and posted solely by the app on behalf of the user, rather than the participant themselves. As the vignettes highlight, participant accounts of autoposts were almost wholly negative (with the exception of Marc, featured in vignette four), often due to the fact that though the participant had inadvertently consented to autoposting as part of the Terms and Conditions for using the app, they had not realised that the app had autoposted in the specific instances featured in the vignettes. That is not to suggest that participant engagements with apps in general (also covered as part of the interviews) were wholly negative – many participants enjoyed many benefits of engaging with the apps on their phones and networked devices. However, as the relationship between cross-platform autoposting and self-performance constitutes the primary research focus of this paper, specific encounters with autoposting, rather than experiences with apps in general, have here been afforded the most critical scrutiny.

Who do you think you are? Carefully-crafted identity performance on Facebook

Before focusing on in-depth vignettes that explore how apps affect participants’ perceived online identity performances, it is first important to establish how participants constructed and maintained their Facebook personas. After all, if Facebook claims that apps help users ‘express who they are’, who did my participants think they were? What kind of selfhood(s) did my participants seek to articulate through their Facebook profiles?

As Van Dijck recognises, despite Mark Zuckerberg’s claim that ‘you have one identity’, the notion that individuals experience their identities as a single, uniform or coherent selfhood has been long contested. The idea that our Facebook profiles reflect a certain ‘type’ of staged selfhood – rather than an ‘authentic’ or ‘fixed’ self – was reflected in the contributions of many participants. For example, Calum (duty manager, 30) was happy to admit that his Facebook use reflected a ‘version’ of himself – but an exaggerated version. When asked ‘Do you think your Facebook use reflects who you are?’, he replied:

Yes and no – but maybe people see a version of me, a side of me that kind of, meta, hyper, you know, side of me… It reflects an aspect of my identity.

Similarly, Sam (digital communications manager, 29) also suggested that her Facebook use reflected a certain type of self, rather than an ‘authentic’ identity. Sam seemed well aware that her performance on Facebook constituted what she called a ‘constructed public persona’ rather than a ‘true’ self. [5] She explained what she means by her ‘constructed public persona’:

So it’s how I want the world to see me… so for instance, I’ve had depression, and you wouldn’t know that from what I said on Facebook… you wouldn’t know if I was having a really shitty day at work for instance. [My Facebook use] is like me, it’s not a completely different person, it is me, but it’s not all of me. And it’s yeah, it’s like my best self (my emphasis).

Sam and Calum reflect Van Dijck’s observations that ‘users have come to understand the art of online self-presentation and the importance of SNS tools for (professional) self-promotion’ (2013: 200, original emphasis). That is, in performing a ‘best’ or ‘hyper’ self that downplays perceived negative aspects of their personalities, Sam and Calum highlight their awareness that their identity on Facebook is carefully-crafted performance.

Though Van Dijck’s assertion that ‘users… have become increasingly skilled at playing the game of self-promotion’ (213: 210) was echoed in part by some participants, it was clear that for some participants self-presentation on Facebook did not necessarily equate to self-promotion – professional or otherwise. For example, though participants such as Melanie (civil servant, 29), Kevin (accounts executive, 25), Calum and Sara (customer service manager, 29) were acutely aware that they were performing a constructed self, this self was not maintained simply through a desire to perform their ‘promoted’ or ‘ideal’ self (Nagy et al. 2011), it was also contingent on an acute awareness of their ‘invisible audience’ (Sauter, 2013; McLaughlin and Vitak, 2012). ‘Invisible audience’ here refers to the network of friends, family, acquaintances and even strangers that could potentially view their performances of selfhood on Facebook. As Sauter notes, by posting to Facebook users are ‘submitting themselves voluntarily to a panoptic form of constant scrutiny’ (2013: 12) imposed by this audience. Calum explained that:

I’m quite aware that, because I see friends who post lots of political things like all the time, or petitions all the time and you do become a bit exhausted to see that kind of stuff, um so I don’t want to saturate somebody else’s Newsfeed with things that I don’t really think they’re necessarily going to be interested in.

As Calum further states, he was aware that his interest in LGBT politics might not always be welcomed by his Facebook audience:

I could easily just always go on about LGBT policies when people get bored ‘Oh there’s Calum going on about the gay shit again and again’.

Thus, for Calum, posting content to Facebook is not simply about promoting his ‘ideal’ self, it was also about not ‘saturating’ his friends’ Newsfeeds with content that might not interest them. [6] Similarly, Melanie’s controlled performance was also contingent on the eyes of her Facebook network. She states in relation to her Facebook use that, ‘it’s about being able to be selective and thinking about who your audience is’. Sara also recognised that her Facebook use was affected by the scrutiny of her network – she stated that she only posts content that she deems is acceptable to her network and professional colleagues, saying ‘I have to restrict some of my personality I suppose’ in relation to the kind of content she posts, and later added that ‘I know I shouldn’t care what people think of me, but I do’. Calum, Sara and Melanie’s accounts suggests here that the performed self on Facebook does not always equate to a ‘promoted or ‘ideal’ self; rather it is a selfhood also constituted through the perceived desires of their invisible audience.

If the enactment of selfhood on Facebook is a carefully considered and staged with an invisible audience in mind, what role do apps play in this performance? As the following vignettes expand upon, the role that apps play in the construction and presentation of selfhood was revealed as complex, tense and often unwanted – apps disrupted and intervened in these performances in ways that call their status as simply instrumental ‘tools’ (or perhaps ‘props’ would be a more fitting term) for self-expression very much into question.

Vignette One: Autoposting Apps and the Invisible Audience

As the previous accounts suggest, the performed self on Facebook is enacted under the gaze of ‘the sprawling mass of contacts most people amass on Facebook’ (Marwick, 2013: 368). Kevin (accounts executive, 25) however, who was interviewed with his housemates Alice (researcher, 28), Rory (sales manager, 30) and Daniel (graphic designer, 29), had such an acute awareness of his ‘invisible audience’ that his Facebook activity was very limited. Kevin called himself a ‘lurker’ and explained that:

I never post anything, I never do it… I feel sort of self-conscious. I feel like I don’t want other people to think that I’m fishing for likes or if I don’t get enough likes I’m like ‘oh that was so embarrassing I shouldn’t have put that one up’ (Kevin laughs).

He explains later in the interview when asked if his Facebook profile reflects ‘who he is’ that ‘I don’t think [people] would really, like get very much from my profile, because I don’t really contribute much’. Kevin’s performance on Facebook is thus very much restricted by a consciousness of how others might see him, and is not contingent on any notion of an ‘ideal’, ‘promoted’ or ‘authentic’ self.

Given that Kevin’s awareness of his imagined Facebook audience leads to a reluctance to perform at all, how does his interaction with Facebook apps affect this limited self-performance? Along with other participants such as Beth, Sara and Alice, it was Spotify that caused the most contention for Kevin in regards to his apps usage. The Spotify app on Facebook currently boasts over 10 million monthly users and as Facebook’s Apps Centre states, by connecting to Spotify via Facebook, a user must agree as part of the Terms of Service to this somewhat ambiguous condition: ‘This app may post on your behalf, including songs you listened to, radio stations you listened to and more‘ (Facebook Apps Centre, 2014, my emphasis). Thus, in signing up to Spotify via Facebook, a user’s songs (‘and more’) have the potential to be automatically posted to their friends’ via the Newsfeed or ticker. Spotify’s current settings allow free-account holders to listen to music as part of either a ‘public session’ – in which a user’s song choices are publicised to Facebook – or a ‘private’ session – in which songs are not publicised. Perhaps unsurprisingly, listening on a ‘public session’ is the default option for users connected to Facebook.

Kevin explains that though he is aware that his Spotify and Facebook accounts are connected, he has occasionally forgotten to switch to a ‘private session’ on Spotify, meaning that his song preferences are then publicised to his Facebook friends’ network. The following exchange between Alice and Kevin reveals the consequences of Spotify’s autoposting of Kevin’s listening choices:

Kevin: If you forget [to switch to a private session on Spotify] then everybody’s like watching every song that you’re listening to, you could be listening to complete trash (Alice and Kevin laugh) really depending on what it is, it’s happened a few times to me, I didn’t even realise it was posting, I feel like, loads of people like it one time, like ‘what is this?’

Alice: And it’s like Dolly Parton

Kevin: Yeah it was Nickleback

Alice: No way – that’s so embarrassing!

Here then, Kevin’s restricted Facebook performance is undermined by the Spotify app; even though Kevin consciously chooses to limit the amount of content he posts to Facebook, Spotify autoposts the songs he listens to his Facebook network without Kevin’s knowledge or consent, at least at the time of posting.

Kevin’s issue with these autoposted songs is clearly apparent: the app is publicising songs that Kevin – and Alice – consider to be ‘trashy’ and ’embarrassing’. Kevin and Alice’s sentiments suggest that listening preferences are here considered to be what ‘symbolic markers’ of identity (Marwick, 2014: 367) or ‘interest tokens’ that constitute a ‘taste statement’ (Liu, 2007). As Marwick and Liu note, identity is in part constituted by ‘interest tokens’ (such as songs) which ‘serve as symbolic markers that signal something about who [users] are’ (Marwick, 2014). Here then Kevin’s music choices are clearly framed as a taste performance that partially signifies ‘who he is’ (in Kevin and Alice’s eyes at least). [7] Crucially however, unlike the symbolic markers of selfhood that Marwick describes, the songs Kevin is listening to on the Spotify app are not consciously ‘displayed’ by Kevin as markers of taste – they in fact function as unwanted markers that are automatically posted by the app, not by Kevin himself. In posting Kevin’s potentially ‘trashy’ or ’embarrassing’ listening preferences, the Spotify app is performing a clearly unwanted utterance of selfhood – a moment of intervention into Kevin’s Facebook activity, wherein the app is revealed as an powerful, algorithmic ‘socio-technical actor’ (Gillespie, 2014: 179). In performing an act of self-articulation on Kevin’s behalf, Spotify thus reveals a power to actively (re)shape Kevin’s intentional representations of identity, rather than functioning simply as a tool for self-expression.

Finally, Kevin’s account exposed the role of the ‘like’ button as a flexible signifier that does not necessarily denote that someone actually ‘likes’ the content that they have acknowledged. When Daniel later states that he makes sure his Spotify app is set to ‘private session’ when listening to embarrassing songs, Kevin states:

Kevin: To be brutally honest I’ve done opposite, I’ve found a really good song and turned it off private and then played it to see who would comment (the group laughs) like five times in a row, like ‘I’ve discovered this amazing music’

Interviewer: right and has it ever had the desired effect? Have you ever had any likes or anything?

Kevin: No it only gets likes when it’s a terrible song

(Both laugh)

Interviewer: Is that because people actually ‘like’ it you think?

Kevin: No, no it’s because they’re ripping the piss, I think, otherwise they’re kind of like ‘yeah whatever, you found some music, I don’t care’.

The fact that Kevin believes that his friends only ‘like’ songs in order to ‘rip the piss’ not only highlights the discursive and affective limitations of the ‘like’ economy that Gerlitz and Helmond identify, it also reveals the complex strategies mobilised by individuals in order to subvert the socio-linguistic architecture imposed by Facebook – according to Kevin, his friends are re-appropriating the ‘like’ button in order to signify their derision of Kevin’s song choices. In using the ‘like’ button to signify a form of ‘dislike’ (or at least derision), Kevin’s friends reveal what Latour calls ‘the risky intermediary pathways’ (1999: 40) subjects follow when assigning meaning to referents. In this case, the rigid logic of positive sentiment enforced on users through ‘like’ button is challenged; the pathway to meaning behind the button is made slippery, playful and ironic.

This is not the only form of subversion evident in this exchange however – in playing an ‘amazing’ song five times in a row, Kevin attempts to present what he deems to be a publicly acceptable song to his audience. In doing so Kevin is attempting to use the app as a ‘tool’ to perform a revised selfhood; a self-performance constituted by the public disclosure of ‘amazing’ rather than ‘trashy’ songs. Unfortunately for Kevin, his efforts to take back control of his performance falls on deaf ears so to speak – it seems that Kevin’s friends only acknowledge his performative slippage of listening to ‘terrible’ songs in public. Kevin’s attempted redirection of Spotify’s autoposts exemplifies Gillespie’s proposal that algorithmic socio-technical architectures encourage users to ‘orient [themselves] towards the means of distribution through which we hope to speak’ (2014: 184). He writes of Facebook:

Some [users] may work to be noticed by the algorithm: teens have been known to tag their status updates with unrelated brand names, in the hopes that Facebook will privilege those updates in their friends’ feeds… other may work to evade an algorithm. (2014: 184)

Here then Kevin is attempting to be ‘noticed’ by the Facebook/ Spotify connection in the ‘right’ kind of way (by listening to the same song five times in a row) – in order to present a socially acceptable form of selfhood, Kevin orients his actions to suit the algorithmic protocols of the two connected apps.

Kevin’s account not only reveals the power of apps as ‘tools’ for self-writing – more specifically for writing the wrong kind of public self – they also compliment Elwell’s assertion that ‘although the way we present ourselves online is often highly crafted… the construction of online identity is likewise often beyond our control’ (2013: 238). Kevin’s account suggests that ‘selfhood is thus a contingent process that is intricately intertwined in complex networks with other actants and entities’ (Sauter 2013: 4). The Spotify app is one such actant, one that encourages a form of ‘turning to face’ the algorithm (Gillespie, 2014: 184) in order to ‘correctly’ present identity via Facebook.

Vignette Two: Re(writing) and Regulating the Self through Spotify

Like Kevin, Beth (24, teacher, UK) also recounted a number of unwanted autoposts by the Spotify app. She stated as a part of her survey response:

I didn’t realise Spotify automatically shared everything [to Facebook]! It was only when someone ‘liked’ the fact that I added a song to a play list and played a song that I realised. I didn’t care too much, despite having a lot of guilty pleasure songs but I generally switch it to a private session now as it just seems unnecessary.

Echoing the exchange between Kevin and Alice, Beth’s sentiments suggest here that music choice is a symbolic marker of taste; and by divulging her ‘guilty pleasures’, Spotify is unwantedly intervening in her taste performance. Furthermore, Beth expressed later that she felt the autoposting of ‘guilty pleasure’ songs could have an impact on how others see her on Facebook. She states when asked ‘Do you think the Spotify posts affect how others see you?’ that:

Um I guess so, because… when you’re a teenage you get that ‘what kind of music do you like?’ question all the time, it’s a big way to find out if someone’s cool or not, but when you’re a grown up like people don’t really ask so I guess they could see things I’m listening to and be like ‘oh they like that’ or ‘oh they’re listening to this’ and I guess it will affect the way people see you, but not necessarily in a bad way.

In keeping with Beth’s observations that her Facebook use ‘reflects the best bits of her personality’, the autoposting of ‘tacky’ songs on to other people’s ticker’s represents a potential departure form the carefully-crafted selfhood that Beth constructs on Facebook.

Like Kevin’s testimony, Beth’s account so far highlights the power of apps such as Spotify to intervene in self-performance on Facebook. Yet Beth goes on to emphasise that apps hold even greater potential – not only to disrupt the writing of the self on Facebook but to also regulate and restrict the self* beyond* the boundaries of the site. As Beth explains, the Spotify app currently has a function in which it automatically switches from a ‘private’ to a ‘public’ listening session after twenty minutes. She states that:

Half the time on my phone, if I go out and I just have my headphones on if I’ve left [Spotify] for a bit it goes back to the non-private setting, so um, half the time on my phone I don’t do it because I’ve already started walking and you have to like, I don’t know remember how to find [the ‘private’ setting] or whatever.

The fact that Spotify switches from a ‘private’ to ‘public’ session after twenty minutes impedes Beth’s capacity to comfortably remain in the realm of private listening, leading to an anxiety that the type of song she is listening to many not be suitable for sharing:

I might be listening to something and then I’m like ‘oh I want to listen to something else’ and then I’ll think, I’ll remember I’m online, maybe because I want to listen to something that’s a bit more like, I don’t know, that I don’t want anyone to know about.

Beth explicitly states that this disclosure of her songs choices is unwanted, but she does not know how to stop it; she states: ‘I’d rather have it so it’s a private setting all the time, but… I don’t really know how to do that’. To compensate for this lack of technological know-how, Beth has come up with an alternative solution to avoid the unwanted disclosure of the songs she is not comfortable sharing:

If I listen to a playlist quite often I’ll just kind of leave it because then it will just say I’ve listened to that playlist rather than specific songs.

Beth here then recounts that her Spotify/ Facebook connection has actively led to a restriction of the songs that Beth feels she can listen to while she is listening on her mobile – to avoid the risk of publicising a song ‘she doesn’t want anyone to know about’, Beth will only listen to specific play lists. The Spotify app’s connection to Facebook here then works to regulate Beth’s listening habits, redirecting Beth’s self-performance through an architectural framework that encourages her to adhere to symbolic markers of music taste that she feels are publicly acceptable.

Beth’s strategy for coping with Spotify’s autoposting capabilities expose the power of apps not just to perform on behalf of the user but actively redirect – and in Beth’s case regulate – the kinds of performance that can be enacted in relation to the self. Beth’s coping strategies exemplify Agre’s ‘capture model’ of socio-technical organisation; that is, ‘the grammar of action’ implemented and imposed on users causes the individual ‘to orient their activities towards the capture machinery and its institutional consequences’ (1994: 110). In Beth’s case the ‘capture machinery’ is made up of the algorithmic technologies employed by Facebook and Spotify, and the ‘institutional consequences’ are the making ‘public’ (to Beth’s Facebook network at least) of music preferences that would otherwise be private. By forcing Beth to orient her activities and regulate her music choice to adhere to a normative ideal of publicly acceptable music, the idea that Spotify helps Beth express ‘who she is’ is called very much into question. It seems that apps hold the potential to tell users ‘who they are’, rather than vice versa.

Finally, Beth’s negotiations with her Spotify/ Facebook connection were further complicated by another account of autoposting. She explains how a few weeks prior to the interview that Spotify had ‘added an event from her past’:

[Spotify] sent me this completely random thing that came up on my phone the other day that said um, ‘Spotify has added an event from your past’, and I was like ‘what is that?’ and it was just that I’d listened to this completely random song like, several months ago… it just popped there, and it kind of annoyed me because it didn’t ask me if I wanted to put it on there, it just added it on there.

Beth expanded on her reasoning for being annoyed by this unwanted addition to her ‘past’. When asked ‘was it a song that you were happy to be added?’, she explained:

It wasn’t one that I minded, no it wasn’t like a cheesy, it was just a random album song… I felt a bit indifferent about it, about the song choice, but if felt like [Spotify] was trying to make it significant and it wasn’t, because I was just listening to it as part of the album you know, it wasn’t like a special thing or anything.

Here then it is not so much the ‘tackiness’ of the song that registers as annoying for Beth – rather is the fact that an ‘insignificant’ song in Beth’s listening habits has been suddenly and non-consensually demarcated as ‘significant’ to Beth’s ‘past’. When asked where she deleted the unwanted autopost, Beth replied:

Well no, because when I actually went on to my page I couldn’t see it, but then somebody liked it, so it must’ve been somewhere but I couldn’t find where it was… you know when it shows just [notifications] on the iPhone, but then it wasn’t like on my page or on my, it was kind of just an isolated, so I don’t know where it is, or if it’s still there, I don’t really know.

Spotify’s utterance of selfhood on Beth’s behalf here takes on both an ephemeral and archival quality – it has been added but Beth does not know where it is, rendering action against the offending autopost impossible. As Beth states, the song is not an identity marker deemed important enough for her to consider it as ‘significant’ to her archived selfhood on Facebook, yet Beth in this instance is powerless to become editor of her written historical identity. The limited visibility of such posts is an attribute unique to autoposting, as outlined further in vignette four. Here however it is important to note that Beth’s negotiation with her Facebook-Spotify connection thus constitutes a site of struggle between the autonomy of the user and the autonomy of the app – a struggle in which the Spotify app holds the performative power to quite literally rewrite Beth’s Facebook history to suit the operational imperatives of Spotify.

Vignette Three: ‘You Have One Identity’? Context Collapse Caused by Apps

Not all moments of identity performance slippage via apps exposed a ‘guilty pleasure’ for participants. For example, Sam (digital communications manager, 29) reported that she had only experienced one instance of autoposting – by an app called Slideshare – a tool for designing and creating professional presentations. Sam explains that:

so it turns out when I upload something to Slideshare [a presentation app] it posts a picture of it on Facebook… that’s why I don’t like things that autopost, because I don’t, I don’t really use, I don’t use my personal Facebook profile for works things, I use Twitter for it, so my Twitter profile is like my ‘work me’.

Sam thus alludes to the fact that her identity performance changes depending on the platform – her Twitter account presents her ‘work me’, while her Facebook account does not. Furthermore, as Sam herself admitted during the interview, publicising your professional presentations does not necessarily constitute a disclosure of a ‘guilty pleasure’; why then was she bothered by this unintended posting of professional content? She explained that:

I think for me I guess it goes back to the like, the persona thing because I don’t really talk about work on Facebook… it just didn’t really fit with the sort of stuff I do, whereas with Twitter I’d more than happily say, in fact probably will say, this is a presentation that I did because that’s where I talk to people about work stuff I do, and I have people who follow me for work stuff.

Thus in this instance the autoposts of Slideshare do not disrupt the boundary between public and private – a boundary crossed in Beth and Kevin’s negotiations with Spotify for instance – but instead a boundary between online social contexts. Mark Zuckerberg’s claim that ‘you have one identity’ (a claim Van Dijck seeks to refute) does not apply for Sam – Sam uses different platforms to perform different identities. Sam’s sentiments thus exemplify Van Dijck’s assertion that: ‘each construction of self entails a strategy aimed at performing a social act or achieving a particular social goal’ (2013: 212).

In disrupting the boundary between Sam’s platforms, the Slideshare app’s actions epitomise what Marwick and Boyd call ‘context collapse’ (2008) – this is ‘the theory that social technologies make it difficult to vary self-presentation based on environment or audience’ (Marwick, 2014: 368). As Marwick notes, ‘people have developed a variety of techniques to handle context collapse’ (2014: 368), and in Sam’s case this entails having separate Twitter and Facebook accounts that represent different facets of Sam’s selfhood. In autoposting symbolic markers of her professional selfhood to the wrong context (that is Facebook rather than Twitter), the Slideshare app brings about a collapse between contexts that Sam has worked hard to avoid.

In creating this context collapse for Sam, the Slideshare app highlights the negative consequences of Facebook’s ambition to become a ‘cross-platform platform’. That is, in attempting to provide a commercial viable bridge between platforms, Facebook apps also disrupt the context-specific identity performances enacted by users. In doing so, autoposts by apps highlight that users do not have one identity that can be ‘expressed’ across all platforms to all audiences. The function of apps to apparently ‘express who you are through all the things you do’ thus works to foreclose the possibilities of enacting multiple identities across different platforms. Apps may express* elements* of identity performance, but the ubiquitous connectivity of the ‘like’ economy works in tension with Sam’s desire to enact different performances in different contexts. As such, the cross-platform connectivity of apps may enhance Facebook’s scope for commercial development, but in doing so they negate the potential for users to perform multiple selfhoods in exclusive online environments. In Sam’s words, ‘the apps I choose probably do tell people about me. But I am not my Facebook app permissions’.

Vignette Four: Sexually Suggestive Content and Exploiting the Connectivity of Apps

Kevin, Sam and Beth’s engagement with apps and autoposts discussed so far revolves around the apparently innocuous leakage of ‘interest tokens’ that unintentionally intervene with their self-performance on Facebook and elsewhere. Calum’s experiences of apps however involved the disclosure of slightly more sensitive material. Calum explains:

So what happened was, on Instagram you know, I follow all sorts of things, mostly friends but you know sometimes the occasional celebrity who’s interesting on Instagram… but in this instance it was a porn star

Calum explains that he was ‘liking’ (on Instagram) pictures from this porn star, some of which were sexually suggestive, and in doing so these pictures were appearing as part of his Facebook activity:

of course these [pictures] were coming up on my Newsfeed, which I didn’t, which I wouldn’t have been made aware of, only for I think another friend had actually liked it on Facebook.

Calum, like Sam, Kevin, Beth and all other participants who had experienced unwanted autoposts, acknowledges that though he may have consented to some form of autoposting as part of the Terms of Service for using the app [8], he was not aware that this particular instance of autoposting was going to occur. As Calum puts it ‘I wasn’t aware of what [Instagram] was going to be sharing… I understood it more as that if I took pictures and wanted to share them, they would share to Facebook’; it did not occur to him that simply ‘liking’ a photo on Instagram would trigger an automatic post to Facebook. Calum’s experience highlights a subtle but important distinction: though Calum had consented to ‘the app posting on his behalf’ at the time of installing the app, he felt he had not consented to the specifics of autoposting with which he had subsequently been confronted. As scholars such as Gillespie and McStay have noted, the lack of specific information, use of opaque and vague terminology in Terms of Service combined with the lack of skill, knowledge and will that surround usage and consent mean that understanding of the socio-technological conditions which users commonly accept can at times be ‘vague, simplistic, sometimes mistaken’ (Gillespie 2014: 185). It seems then that for the Calum – as well as many of the other participants – merely consenting to autoposting as part of the terms and conditions of app use does not equate to a sense of unconditional consent in all circumstances.

Furthermore, the fact that Calum realised these pictures were being publicised on Facebook only after his friend had ‘liked’ them highlights a form of opacity unique to autoposting: that is, unlike other posts that are consciously written by the user themselves, autoposts by apps do not always appear on a user’s Timeline or Newsfeed. The fact that these autoposts are invisible to the very individual that has supposedly ‘written’ them renders action against such autoposts impossible – unless the autopost is made visible by another user’s acknowledgement of it. This characteristic undermines the control usually afforded to users in regards to self-writing on Facebook; the other ‘technologies of the self’ available to users (status updates, photos uploads, ‘shares’) can be edited or deleted. The invisibility of autoposts to the users that ‘write’ them thus again throws the notion of ‘informed consent’ (McStay, 2010) into question. The lack of control that users enjoy over autoposts is one of the reasons why app autoposts are apparently being phased out by Facebook (Zuckerberg et al., 2014). For now, however, it seems that the unintended taste performances disclosed by apps can have tangible consequences for current Facebook users.

A key theme that emerged from Calum’s account, as well as from other pariticpants, was thus the perceived control afforded to them concerning the posting of content by apps on their behalf. For example, Marc, who had no specific negative experiences of autoposting, did acknowledge that he enjoyed publishing his running activities to Facebook through the ‘Sports Tracker’ app – he stated that ‘I get a few likes now and then if it’s a particularly long run’. The key difference here between Calum’s and Marc’s experiences is that Marc actively instigates posts from the app, rather than the app acting in his stead. Thus, in Marc’s account, the app functions more like a tool than an actor, and thus works to support his intended identity performance, rather than disrupt or work against it.

As Calum explains, the unintended disclosure of the suggestive content was soon rectified – unlike Beth and Kevin, who at times felt compelled to regulate their self-performances to suit the operational protocols of Spotify/ Facebook as well as the scrutiny of their Facebook audiences, Calum found a way to disconnect his Instagram account from Facebook, though he admits that finding the settings to action this disconnection ‘was a bit of a job’. Though Calum’s engagement with Instagram had some unintended outcomes, his negotiation with the Tinder app gave rise to another set of circumstances that worked very much in Calum’s favour. He explained how Tinder works:

Right how it works is, it’s a dating website for gay, straight, bisexual, whatever, you know it’s just a pan-dating website shall we say, um, and, you cannot communicate with anybody on Tinder unless you like each other… the only information people see is what it takes from your Facebook profile.

He notes, however, that there are loopholes in the functionality of the Tinder app which mean a user can exploit the system’s connectivity to Facebook, thereby giving a user increased access to other users’ profile information:

If you have mutual friends, depending on your privacy settings… if you’re clever enough you can go to that person’s Facebook friends list and then suddenly see all of their Facebook information.

Calum notes that this illicit access to another user’s personal information is only possible ‘if they haven’t got their privacy settings amended’, but he acknowledges that, in relation to the Tinder app at least, his own privacy settings are in order. Here then Calum is exploiting the connective capabilities of the Tinder app through loopholes in privacy settings that, on Instagram, had caused a moment of performance slippage for Calum himself. Calum’s manipulation of the Tinder/ Facebook connection thus exemplifies the potential for users to exploit, rather than be exploited by, the connectivity embedded in the social web. Of course, it is weaknesses in other users’ privacy settings that allow for such exploitation; and as such it is those users who, as Calum notes, ‘haven’t got their privacy setting amended’, that are exposed to disruptions of self-performances that can be caused by Facebook apps.

Vignette Five: Autoposts as Spam and Game Posting as ‘Chavvy’

This paper has so far centred on the interventions of apps into participants’ own self-performances – yet many participants also noted the presence of apps in their friends’ Facebook activities. Somewhat surprisingly given their popularity on Facebook, it was friends’ posts regarding game apps that were most frequently cited as irritating, frustrating or annoying. As Melanie states of game posts by her friends:

It’s advertisements as far as I’m concerned… it’s people I know that are advertising these things and it’s crafty and I don’t like it.

Melanie’s observations that games posts are ‘advertising’ exists in clear tension to Facebook rhetorical framing of these same posts as ‘sharing’ – yet her sentiments highlight that the like economy is built on a system of social connectivity that can be efficiently monetised by the platforms that support it (Gerlitz and Helmond, 2013). For example, players of ‘Candy Crush’ can either pay for ‘tickets’ to proceed to the next level of the game, or they can ask their Facebook friends for tickets via Facebook. Instances such as these render the value of social connectivity as profoundly apparent – the connection between three Facebook ‘friends’ (the amount of friends needed to get a ticket) is quite literally worth 79p.

Participants’ framing of posts by games as ‘advertising’ or ‘spam’ suggest that the monetary value of the social web does not go unnoticed by those users implicated in it. These posts by game apps highlight the advertising value of Facebook apps for third party stakeholders (app developers, app owners, data aggregators etc) in terms of advertising. Yet these same posts, according to participants, also hold negative value for users in relation to self-performance on Facebook. For example Sophie (publishing assistant, 28) stated:

My biggest reaction when I see people post gamey kinds of status things is just like I can’t believe you play those stupid games, and people actually go down in my esteem.

Similarly, in their group interview with Daniel and Kevin, Rory and Alice took up the idea of game invites as ‘spammy’ and annoying. Like Sophie, Alice and Rory believe that autoposts on Facebook do affect how others see them:

Alice: Yeah I think I just think people are probably just a bit stupid that’s really harsh isn’t it, I just think people are bit stupid … I mean it’s slightly hypocritical me saying this because I’ve clicked through terms and conditions without looking at anything, but I think it’s just a sign of people not really paying attention to what they’re doing, or not really having the foresight to think oh hold on maybe I should check this because games are really dodgy on Facebook

Rory: I guess there are some people who are just, so [pauses] I don’t even know how to describe it

Alice: Were you going to say chavvy?

Rory: Well I can think of somebody who I would class as being chavvy who does, who everything comes through and you think, ah, typical

Alice: I hadn’t thought of it as a generalisation but I can immediately think of some people who would fit that bill

The class connotations in this somewhat hesitant and self-conscious exchange are clear – both Rory and Alice agree that inviting people to play games and posting game posts is ‘chavvy’. As Liu points out, social network profiles as taste performances can be analysed through societal paradigms relating to cultural and educational capital (2008) – and here then it seems that allowing a game to post on your behalf can be framed as a ‘taste expression’ that carries low cultural capital. Thus it seems that apps have a performative capacity to shape and redirect ideas of the classed self on Facebook.

Another participant, Rebecca (lecturer, 27), made an important distinction regarding the publication of game posts. She states:

When I see people post stuff or you know sharing stuff about games, I don’t think you’re an idiot for playing the game, I think you’re an idiot for sharing it.

Thus for Rebecca, it is not the playing of the game in itself that matters, it is making the game play public that is seen as detrimental. Here then, Liu analysis of ‘destructive information’ in relation to the performance of the self becomes especially relevant. As he states:

Any outlier of interest tokens in [user] profiles – such as the inadvertent mention of something tabooed or distasteful – could constitute destructive information and spoil the impressions that users are trying to foster. (2008: 258)

Autoposts by game apps are framed by Sophie, Rebecca, Alice and Rory as pieces of ‘destructive information’ – the sharing of game achievement is connoted to be detrimental or distasteful. The question of why game posts are so detrimental to these users – as compared to other autoposts or actions by other apps – is a complex one that requires critical examinations of class, gender and socio-economic circumstance that lie beyond the remit of this small study. However, as some participant responses suggest, distasteful or detrimental identity performances via game apps seem to involve breach of expected norms and practices on Facebook. For example, when asked why she finds frequent posts from her gamer friends annoying, Sam states:

I think the fact that they don’t seem to have any self-control about sharing, so whether or not it’s that the app’s too tricky or forces you to invite people… but I think it’s because I’d see it as being a little bit impolite, or it’s just not my version of internet etiquette to spam people with this stuff.

Game app posts by friends for Sam then breach her expectations of ‘internet etiquette’. As McLaughlin and Vitak note, online norms are dependent on the technology that facilitates them. They state that though the offline norms that regulate behaviour tend to be ‘ingrained into children from an early age’, on SNSs sites ‘norms evolve with the technology’ (2011: 3). Autoposting by apps is a developing socio-technological practice that is yet to be accompanied by a fixed set of norms. As such, it seems game autoposts can breach the established socio-cultural norms held by some of the individuals exposed to them. Yet, as mentioned earlier, autoposts by apps may soon be phased out of Facebook structural architecture altogether, due to their unpopularity. It seems then that my participants are caught up in socio-cultural moment that may not last – their negotiations with apps, and their reception of their friends’ performances via apps, signify a transient moment of technological flux that may soon be ‘fixed’ by Facebook.

Conclusion: Regulating Self-Performances via Apps

In his Butlerian analysis web profiles, Cover states that ‘social network activities are performative acts of identity which constitute the user’ (2012: 178). This paper has sought to consider apps as one such ‘social network activity’ capable of performatively acting on the user’s behalf. As I have explored, autoposts by apps constitute utterances of selfhood for those that use apps as part of their Facebook activity. In constituting utterances of selfhood, apps have the capacity to intervene in and disrupt the staged performances of those that use them. The performative acts of apps – such as adding an event to Beth’s ‘past’, causing context collapse for Sam, or disclosing ‘destructive information’ and unwanted ‘taste performances’ for a number of participants suggest that apps are not only tools that ‘help users express who they are’, as Facebook claims. They are also technological actors that hold the autonomous potential to write, and therefore perform, acts of selfhood on behalf of users. Furthermore, if we consider, as Cover writes and as Butler as suggests, that identity performances actively constitute the subject, then apps to do not simply articulate pre-existing taste expressions of identity on behalf of the user – they hold the potential to constitute* *facets of identity for the uses implicated in their connective functionalities. Though apps are not the only technological actors to be imbued with such ‘autonomous existence’ (2008: 15), as Software Studies theorists such as Goffrey (2008) and Gillespie have noted, the struggle for autonomy between technological actants and users is not always made so explicit. Furthermore, Beth’s negotiations with Spotify reveal that apps can intervene not only in present articulations of the self but in past ones too. By adding an unwanted event to Beth’s Facebook history, the Spotify app has the power to quite literally rewrite Beth’s ‘past’ selfhood on Facebook.

As previously mentioned, despite their potential marketing value for external third parties, autoposts are being phased out by Facebook due to their unpopularlty with users (a sentiment clearly reflected throughout the vignettes). Where then does this leave user engagement with Facebook’s cross-platform apps? The vignettes featured highlight the fact it is unwanted autoposts – posts by apps that are published without the knowledge or instance-specific consent of the user – that hold the potential to undermine and disrupt the identity performances of the users in question. It is important to reiterate, as highlighted by Marc’s testimony, that not all posts by apps are detrimental to self-presentation on Facebook: given the right level of consent, control and understanding, apps can and are used by users to display wanted – rather than unwanted – taste articulations. The popularity of apps also suggests that many users willingly and enjoyably engage with apps on a daily basis. It thus seems it is the unconsensual nature of autoposts – wherein the app as tool becomes the app as unwanted actor – that is resisted by participants, and now partly by Facebook itself. Though autoposting (in its current form) may soon be a thing of the past, the monetary value of apps and their capacity to express some element of selfhood – especially in relation to perceptions of taste and class – suggest that the relationship between self-performance and apps use is open to further critical scrutiny.

Finally, this paper has sought to highlight that the intervention of autoposting apps in users’ staged self-performances at times transcends the boundaries of Facebook. For example the self-regulation of Beth’s listening preferences suggests that apps hold the capacity to reconstitute the self in ways that affect identity performances outside of Facebook. Cover’s and Agre’s arguments that users’ selfhoods can be performatively regulated and even restrictively constituted by SNSs profiles and computational frameworks thus become particularly relevant; by performing the self on behalf on the user, Facebook apps expose users to ‘the violence of a normative truth regime that excludes post-modern, post-structuralist ways of configuring identity, self complexity or doing subjectivity otherwise’ (2012: 183). In doing so, autoposts highlight the performative power of Facebook apps to constitute particular kinds of selfhood – selves that adhere to the normative expectations of taste, public acceptability and cultural interest, and that suit not only the logic the algorithm but also the monetised, ubiquitous social connectivity that currently pervades the ‘like’ economy.

Biographical Note

Tanya Kant is a doctoral candidate and associate tutor at the University of Sussex, Brighton, UK. Her research interests include the socio-cultural implications of system-initiated personalisation practices for web users, online identity performance and constitution, online privacy and political economy of SNSs. Recent publications: ‘Giving the viewser a voice? Situating the Individual in Relation to Personalization, Narrowcasting and Public Service Broadcasting’, Journal of Broadcasting and Electronic Media 58:3 (2014): 381—399.

Notes

• [1] As the paper will highlight, although apps form a component of Facebook user’s profiles, apps function beyond the confines of a user’s Timeline. In fact, some utterances of selfhood as performed by apps never actually appear on a users own profile, but only on their ‘friends’ Newsfeeds.
• [2] The Facebook ticker is a real time stream of ‘friend’ activity that appears on the upper right-hand side of a Facebook user’s Newsfeed.
• [3] Two group interviews were undertaken as part of the project – Kevin, Alice, Rory and Daniel were interviewed together, as were Rebecca, Audrey, Sophie, Terry and Steve. The participants taking part in these interviews were not strangers to each other – they were housemates, most of whom were friends on Facebook and therefore constituted part of each other’s Facebook networks. Their ‘offline’ connection as housemates thus added a valuable additional dimension to their interviews, in the form of dynamic exchanges between participants (as exemplified in Kevin’s accounts of autoposting) that highlight how the intervention of apps into users’ Facebook profiles does not just affect the user themselves – it also impacts on their network.
• [4] There are various ways to expand the ‘reach’ of a Facebook page – through publicising the page on other Facebook pages, other websites and alternative social networks, as well as through ‘promoting the page’ through Facebook’s pay-to-use marketing mechanisms. The ‘Plugged-in Profiles’ page was publicised through external alternative sites, pages and social networks with a small degree of success (in terms of visibility). However these page ‘likes’ did not always necessarily translate to interested interview participants. The other option – to pay to promote the page – though a tempting prospect, was not implemented. I felt that paying for ‘likes’ was at best counter-productive and at worst unethical in relation to my research aims, and as Emiliano De Cristofaro et al (2014) have highlighted, using such Facebook’s pay-to-use promotional techniques would most likely have led to ‘fake likes’ generated from legally and ethically dubious ‘like farms’, rather than viable (human) research participants. By far the most successful form of ‘reach’ thus transpired to be page ‘shares’ mobilised by other individuals in or close to my own ‘friends’ network; the outcome of which being that interviewee participants were recruited from networks in close proximity to my own.
• [5] Sam preceded her use of the term with the question ‘is it ridiculous if I say pretentious media studies words?’, suggesting that her mobilisation of this somewhat complex phrase can be explained by a background knowledge in theories of identity construction.
• [6] Perhaps Calum was right to be wary. As Beth explained in her interview, the only friend she has blocked on her Newsfeed was a friend who posted too much political content.
• [7] The notion that your music and listening preferences tell others something about ‘who you are’ is not only reinforced in other participants experience of identity performance on Facebook, it is also encouraged by Facebook itself, as evidence in their statement concerning ‘social apps that let you express who you are through all the things you do.’
• [8] Sam, Beth and Kevin also acknowledged that that autoposting may have been a prerequesite to using the app. For example, Sam stated: ‘I must have just got excited and pressed the button and not looked’ in reference to signing the terms and conditions for using the Slideshare app, and Beth stated ‘maybe it was like in [Spotify’s] Terms and Conditions or I just didn’t read it properly, but I felt it should have asked me first [before autoposting].’ Beth’s sentiment highlights an important distinction in terms of consent: though participants recognise that they were required to consent to the possibility of autoposting in order to use the app, they did consent to the specific instances of autoposting that they have had to negotiate.

References

• Agre, Philip E. ‘Surveillance and Capture: Two Models of Privacy’, The Information Society 10.2 (1994): 101—127.
• Butler, Judith. Bodies that Matter: On the Discursive Limits of “Sex” (London: Routledge, 1993).
• Clough, Patricia. ‘The Philosophical Carpentry of the App: Criticism and Practice’, in Paul D. Miller and Svitlana Matviyenko (eds.) The Imaginary App (Cambridge: MIT Press, 2014), 35—50.
• Cover, Rob. ‘Performing and Undoing Identity Online: Social Networking, Identity Theories and the Incompatibility of Online Profiles and Friendship Regime’, Convergence 18.2 (2012): 177—193.
• Dyson, Frances. ‘Wireless Affections: Embodiment and Emotions in New Media/ Theory and Art’, Convergence 11.4 (2005): 85—105.
• De Cristofaro, Emile, Arik, Jourjon, Guillaume, Kaafar, Mohamed Ali, Shafiq, M. Zubair. ‘Paying for Likes? Understanding Facebook Like Fraud Using Honeypots’, CMI 2014 (2014), arXiv:1409.2097v2
• Elwell, Sage. J. ‘The Transmediated Self: Life Between The Digital and the Analog’, Convergence: The International Journal of Research into New Media Technologies 20.2 (2014): 233—249.
• Facebook. ‘Introducing Timeline’ (2014), httpss://www.facebook.com/about/timeline
• Facebook Apps Centre. ‘Get Spotify’ (2014), httpss://apps.facebook.com/get-spotify/
• Facebook blog. ‘News Feed FYI: Giving People More Control Over When They Share from Apps’ (2014), https://newsroom.fb.com/news/2014/05/news-feed-fyi-giving-people-more-control-over-when-they-share-from-apps/
• Fagerjord, Anders. ‘Toward Apps Studies’, Proceedings from the Internet Research 13.0 Conference University of Salford, Oxford (2012), https://fagerjord.no/downloads/FagerjordAppstudiesIR13.pdf
• Foucault, Michael. ‘Technologies of the Self’, Essential Works of Foucault 1954—1984: Ethics (London: Penguin, 1994).
• Gerlitz, Carolin and Helmond, Anne. ‘The Like Economy: Social Buttons and the Data-Intensive Web’, New Media & Society 15.8* *(2013): 1348—1365.
• Gillespie, Tarleton. ‘The Relevance of Algorithms’, in Tarleton Gillespie, Pablo Boczkowski, and Kirsten Foot (eds.) Media Technologies: Essays on Communication, Materiality, and Society (Cambridge: MIT Press, 2014), 167—194.
• Goffrey, Andrew. ‘Algorithm’, in Ed Fuller (ed.) Software Studies, (Cambridge, Mass.: MIT Press, 2008) 16—18.
• Karppi, Tero ‘FCJ—166 “Change name to No One. Like People’s Status”: Facebook Trolling and Managing Online Personas’, the Fibreculture Journal 22 (2013): 278—300.
• Killoran, John. B. ‘Under constriction: Colonization and synthetic institutionalization of Web space’, Computers and Composition, 19 (2002):19—37.* *
• Kitchin, Rob and Dodge, Martin. Code/ Space (Cambridge: MIT Press, 2011).
• Latour, Bruno. Pandora’s Hope: Essays on the Reality of Science Studies (Cambridge, Mass.: Harvard University Press, 1999).
• Liu, Hugo. ‘Social Network Profiles as Taste Performances’, Journal of Computer-Mediated Communication 13 (2008): 252—275.
• Manovich, Lev. ‘The Language of Media Software’ in Svitlana Matviyenko and Paul D. Miller (eds.) The Imaginary App (Cambridge, Mass.: MIT Press, 2014), 1—20.
• Marwick, Alice E. ‘Online identity’ in Hartley, John, Burgess, Jean and Burns, Axel (eds.) The Companion to New Media Dynamics (London: Blackwell, 2013), 365—374.
• Marwick, A. and Boyd, d. ‘I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse and the Imagined Audience’, New Media & Society 13.1 (2011): 114—133.
• Matviyenko, Svitlana. ‘Introduction’ in Paul D. Miller and Svitlana Matviyenko (eds.) The Imaginary App (Cambridge: MIT Press, 2014) 1—20.
• Mellamphy, Dan and Biswas Mellamphy, Nandita. ‘From the Digital to the Tentacular, or from Ipods to Cephalopods: Apps, Traps and Entrées-without-Exit’ in Paul D. Miller and Svitlana Matviyenko (eds.) The Imaginary App (Cambridge: MIT Press, 2014), 231—50.
• McLaughlin, Caitlin and Vitak, Jessica. ‘Norm evolution and violation on Facebook’, New Media & Society, 14.2 (2012): 299—315.
• McStay, Andrew. ‘I Consent: An Analysis of the Cookie Directive and Its Implications for UK Behavioural Advertising’, New Media & Society, 15. 4 (2012): 596—611.
• Nagy, Peter and Koles, Bernadett. ‘The digital transformation of human identity: Towards a conceptual model of virtual identity in virtual worlds’, Convergence (2014): 1—17.
• Richardson, Gemma ‘Feeling Suicidal? There’s an App for That’, unpublished manuscript Apps and Affect Conference 2013, University of Western Ontario, London, Ontario.
• Sauter, Theresa. ‘”What’s On Your Mmind?” Writing on Facebook as a Tool for Self-Formation’, New Media & Society – first online (July 18, 2013): 1—17.
• Statistic Brain. ‘Facebook Statistics’ (2014), https://www.statisticbrain.com/facebook-statistics/
• van Dijck, José. ‘”You Have One Identity”: Performing the Self on Facebook and LinkedIn’, Media, Culture and Society 35. 2 (2013): 199—215.
• van Zoonan, Liesbet. ‘From Identity to Identification: Fixating the Fragmented Self’, Media, Culture & Society 35. 1 (2013): 44—51.
• Walker, Rettberg Jill. Seeing Ourselves Through Technology: How We Use Selfies, Blogs and Wearable Devices to See and Shape Ourselves, (Basingstoke: Palgrave MacMillan, 2014).
• Williamson, Ben. ‘Algorithmic Skin: Health-Tracking Technologies, Personal Analytics and the Biopedagogies Of Digitized Health And Physical Education’, Sport, Education and Society 20.1 (2015): 133—151.
• Wilson, Julie Ann, Yochim, Emily Chivers. ‘Mommy Apps: Keeping the Family ‘Happy’ in Digital Media Culture’, unpublished manuscript, Apps and Affect Conference 2013, University of Western Ontario, London, Ontario.
• Zuckerberg, Mark, Sukhar, Ilya, Archibong, Ime & Liu, Deb. ‘F8 2014 Keynote’ (2014), httpss://www.youtube.com/watch?v=0onciIB-ZJA
• Zuckerberg, Mark, ‘F8 2011 Keynote’ (2014), https://www.youtube.com/watch?v=ZLxlJbwxukA