I write this chapter at what feels like a turning point in the short yet rich history of social media.

Instagram – a photo- and video-sharing app founded in 2010, which has since attracted near-global popularity – has been accused of being ‘toxic’ to teenagers’ mental health, especially girls’ [Reference Wells, Horwitz and Seetharaman1]. Whistle-blower Frances Haugen leaked several internal documents sourced during her tenure at Meta (the company that currently owns Instagram) to the Wall Street Journal, which included findings from an internally conducted survey of a sample of Instagram’s teenage users. In slides from a presentation posted to the company’s staff message board in March 2020, the survey findings revealed that Instagram ‘make[s] body image issues worse for one in three teen girls’ [Reference Wells, Horwitz and Seetharaman1].

This is one of many high-profile reports about the link between social media and mental health, and it sits within a broader social climate where people are asking pressing questions about the extent to which social media can ‘affect’ or ‘impact’ individuals. It is natural to fear new and fast-growing technologies [Reference Baym2] but regressing to the media ‘effects’ models that have for so long been discredited (remember the hypodermic needle model, anyone?) may not be helpful, or indeed accurate. One of the aims of this chapter is therefore to argue that social media are not one thing and to introduce instead several aspects of the phenomenon to readers, briefly tracing standout phases in their evolution, the characteristics that differentiate them from older media technologies, their (increasingly controversial) business and governance models, and finally their use and non-use among particular social groups.Footnote ¹

The Evolution of Social Media

The phrase ‘social media’ was first used in the English-speaking language around 1994 [Reference Bercovici3], but it wasn’t until the mid-2000s that it gained widespread currency. At that time, US tech companies were recovering from the 2001 dot-com bubble crash, with many embracing the opportunity to start afresh and create what they imagined to be a different Web [Reference Jamieson4]. This new wave of internet history was known as Web 2.0, a term coined by O’Reilly Media’s Dale Dougherty and Tim O’Reilly at a Web industry conference to represent a turning point in the internet’s (albeit short) history [Reference Jamieson4]. However, as Jamieson explains, it’s notoriously difficult to pin down exactly what Web 2.0 is: a business model, a technical development, a social change, or perhaps all the above [Reference Jamieson4]? Many of the features packaged as Web 2.0 – such as user participation and user-generated content – were present in Web 1.0 applications such as wikis, online journals, and blogs, which means Web 2.0 might best be thought of as a discourse as opposed to a set of unique technologies.

Nonetheless, the term caught on in the West, evoking excitement for a new phase of Web history. This sentiment was famously captured when TIME magazine named ‘You’ its Person of the Year in 2006, hailing the potentialities of ‘the new Web’ for enabling ordinary people to create and share media, the supposed actualization of ‘the many wresting power from the few’ [Reference Grossman5]. Around the same time, within academic circles, the term participatory culture was coined to denote the seemingly new opportunities afforded by a move away from the one-to-many communicative style of television shows, magazines, some older websites, and more. The move to a participatory Web, it was said, increased opportunities for average internet users to share user-generated content such as text, images, videos, emoji, and GIFs. The term participation contrasted with older notions of passive media audiences/receivers, hence the rise of hybrid terms like prosumer (producer plus consumer) in the mid-2000s [Reference Bruns6] (though decades of media and communication research teach us that audiences are, of course, never wholly passive).

The history of social media is often told from a Western, US-centric perspective, but this evolution looks very different around the world, and changes according to local infrastructure, politics, economics, and culture. There is therefore no singular rise of social media, and this history depends on how you tell it.

For example, similar technological and societal developments took place in Japan long before Web 2.0 emerged as a discourse in the USA. While Japan was seen as being late to the internet age, partly due to the complexities of displaying non-Western language characters on digital keyboards, the country was a front runner in the development and uptake of ‘person-to-person communication’ [Reference McLelland, Yu, Goggin, Burgess, Marwick and Poell7]. These technologies emerged in Japan in the late 1980s and were known as ‘personal computer communications’ (or pasokon tsūshin), which were ‘basically bulletin board systems that enabled users to seek out information from various news feeds as well as participate in online discussions and send email to other users’ [Reference McLelland, Yu, Goggin, Burgess, Marwick and Poell7]. While these early systems were not taken up by the Japanese population en masse, they found their place in certain communities such as feminist women, those living with chronic illnesses, and among trans people and gay men [Reference McLelland, Yu, Goggin, Burgess, Marwick and Poell7].

Although bulletin board systems might not count as ‘social network sites’ according to boyd and Ellison’s (2007) definition [Reference Boyd and Ellison8], which I discuss in the next section, these developments should remind us that the rise of social media as we know them today was not strictly driven by Silicon Valley. Long before ‘participatory cultures’ were celebrated by Western academics, Japanese writers spoke of the potentialities of ‘revolutionary’ new ‘networks’ (nettowāku) [Reference McLelland, Yu, Goggin, Burgess, Marwick and Poell7]. But what makes social media different from their technological predecessors?

The Characteristics of Social Media

In what has become a go-to source for those seeking a definition of social network sites, boyd and Ellison [Reference Boyd and Ellison8] explain that they are

Readers might notice that this is a definition of ‘social network sites’ as opposed to ‘social media’, a term on which this book is based and which we now more commonly use in the English language. This shift in terminology is tough to trace, and it is difficult (and certainly controversial) to decide what counts as the world’s first example of ‘social media’. Instant messaging and bulletin board services evidently count as social forms of media technologies, but if the creation of a profile and the ability to share lists of connections was central to boyd and Ellison’s definition of a ‘social network site’, then one of the earliest examples was SixDegrees.

Founded in 1997, SixDegrees ‘allowed users to create profiles, list their Friends and, beginning in 1998, surf the Friends lists’ [Reference Boyd and Ellison8]. While these features existed on other sites, SixDegrees was the first to combine them [Reference Boyd and Ellison8]. Although boyd and Ellison’s 2007 definition continues to be heavily cited, what we now know as social media platforms have more complex characteristics than their predecessors, and these changes invite us to question the endurance of the profile and friends list to more contemporary definitions of what counts as ‘social media’.

Let’s take Myspace (launched in 2003) as another example. Myspace users were invited to create their own profile, which displayed a profile picture along with their name, age, and location – their ASL, or age/sex/location, as we used to call it (see Figure 1.1). Users could also choose their Top 8 friends and browse each other’s social connections, leave comments on people’s profiles, send messages, and display biographical information. Crucial to Myspace’s success, however, was the editability of the profile: ‘Myspace users needed to learn basic HTML and CSS to creatively customize their profiles’; affordances that are, unfortunately, ‘mostly absent from contemporary social media platforms’ [Reference Miltner and Gerrard9].

Figure 1.1 A static clone of Tom’s iconic 2006 Myspace profile page

(image sourced from https://github.com/wittenbrock/toms-myspace-page)

It’s easy to see why, in 2007, boyd and Ellison centred the profile and friends list as key defining features of social network sites. However, the functionalities and characteristics of social media, as we now know them, have advanced considerably since the days of SixDegrees and Myspace. Current social media users can record and edit videos on TikTok and Douyin (a Chinese equivalent of TikTok), send ephemeral pictures through Snapchat, livestream a video on Facebook, or buy clothes for their dog on Instagram.

What distinguishes ‘social media’ from ‘social network sites’ might therefore be the shift from the profile – and the showcasing of friends and networks – to the creation and sharing of media content. This change is reflected in Burgess et al.’s [Reference Burgess, Marwick and Poell10] more recent definition of ‘social media technologies’, which they define as ‘those digital platforms, services and apps built around the convergence of content sharing, public communication, and interpersonal connection’ [Reference Burgess, Marwick and Poell10]. While social media are still, of course, used to communicate with people, this definition also rightly includes content creation and public communication, core elements of their business models that I now discuss in greater depth.

The Business Model(s) of Social Media

It might be surprising to readers that social media companies don’t always generate profit straight away. In the quote above, taken from The Social Network – a 2010 biographical film chronicling the invention of Facebook – the company’s founder (Mark Zuckerberg) and chief financial officer (Eduardo Saverin) debate the point at which they should start monetizing their then-new site.

Facebook – and plenty of other social media companies – make most of their money by displaying advertisements for external products and services. They make some money by charging sign-up fees, but the big players in today’s social media game tend to offer their sites for free [Reference Baym2].Footnote ² Currently popular platforms find increasingly sophisticated ways to show people ads. On Instagram, for example, users might see ads as they scroll down their Feed, or as they browse through people’s Stories. For those unfamiliar with Instagram, this means paid-for posts are interspersed with content from a person’s social network. Businesses can also pay to advertise their products in more creative ways, such as Snapchat filters, as shown in Figures 1.2 and 1.3.

Figure 1.2 Photo of a sponsored Disney PhotoPass Lenses Snapchat filter,

taken by Dr Phoenix Andrews

Figure 1.3 Photo of a sponsored Ben and Jerry’s Sundae Snapchat filter,

taken by Ysabel Gerrard

To decide which advertisements to show to a particular user, social media companies will analyse that person’s data: their clicks, shares, likes, follows, and even their activities on other sites hosting ‘like’ buttons for social media sites [Reference Maris, Libert and Henrichsen11] – a process called social media data mining. This information is of great value to people paying social media companies for advertising space, as they know their ads will be seen by the people most likely to engage with their products or services. Less discussed academically, however, are instances when social media platforms become incredibly popular but make no money at all. Let’s take the failed app Fling as an example.

Founded in 2014, Fling invited users to send – to Fling – photos and videos to complete strangers around the world. The app was an instant hit, but it was soon overrun with pornographic images and sexual harassment, leading to its removal from Apple’s App Store. Anonymous app Sarahah met a similar fate. Part of the ‘honesty app’ trend, Sarahah, ‘which means “frankness” or “honesty” in Arabic’ [Reference Bell12], allowed users to ask a question for anyone with a link to answer anonymously. Sarahah quickly rose to the top of app stores in several countries, becoming ‘particularly popular in Arab-speaking regions and also among English-speaking teenagers’ [Reference Bell12]. However, the app’s founder – who had originally designed it for corporate settings, as a tool for workers to give feedback to their employers – hadn’t imagined the site would become so popular with teens. Predictably, Sarahah was removed from app stores and its founder never made a penny.

Apps like Fling and Sarahah are what I call popular by surprise [Reference Gerrard and Gillespie13], and apps that fit into this category typically make little if any profit as their popularity is too fleeting for founders to meaningfully monetize them. In essence, they failed because they were too popular, leaving their founders unprepared to govern them safely.

The Governance of Social Media

Fortunately (or unfortunately, for some), we can’t just say what we want on social media. Every space facilitating user-generated participation – not just social media but, for example, comments sections of online newspapers, or buyer reviews on shopping websites – is governed by a set of rules dictating what people can and cannot say. But it’s not just about what people say via text; it’s also about what they post – videos, images, emoji, and so on. Let me give you an example.

Emoji – ‘small digital images used in online communication’ [Reference Matamoros-Fernández14] – might seem like harmless, playful forms of digital communication. Emoji are immensely popular, with over five billion used every day on Facebook Messenger alone [Reference Matamoros-Fernández14]. But their harmless facade can be used to ‘cloak everyday microaggressions in humour and play’ [Reference Matamoros-Fernández14], as online abusers weaponize smiley emoticons to mitigate their abuse. Emoji are also sometimes used to display opposition to certain religions or to incite hatred to those of certain ethnicities.

As Matamoros-Fernández [Reference Matamoros-Fernández14] explains, emoji pose unique challenges for content moderation: they cannot be switched on and off (unlike, say, switching off the comments underneath a post), and platforms can’t filter out keywords or previously banned images/videos, like they can with other kinds of social media content. It is also extremely difficult for human content moderators – known within the tech industry as commercial content moderators [Reference Roberts15] – to decipher the context of their use. While academic and public discussions focus mainly on the moderation of text, images, and videos, significantly less attention is paid to the moderation of emoji.

This example is one of many to highlight the failures of currently popular social media giants to effectively regulate their content. As Gillespie [Reference Gillespie16] explains, the relatively recent popularization of the term ‘platforms’ by tech giants signals a shift in social media’s history, particularly in terms of their reputation. In stark contrast to the excitement surrounding Web 2.0 technologies, social media companies face growing distrust and backlash from their users, for everything from their failures to act on online harms to their opaque data collection processes. Rather than helping the average, everyday media user to wrest ‘power from the few’ [Reference Grossman5], social media have instead given new, indeed quite frightening, forms of power to a small handful of tech giants. McChesney puts it best by saying: ‘[I]t is supremely ironic that the Internet, the much-ballyhooed champion of increased consumer power and cutthroat competition, has become one of the greatest generators of monopoly in economic history’ [Reference McChesney17].

The emergence of the term social media platform therefore does not mark a technical shift in the way Web 2.0 perhaps did but is instead a discourse popularized by online content providers in their efforts to make ‘strategic claims for what they do and do not do, and how their place in the information landscape should be understood’ [Reference Gillespie16]. By describing themselves as ‘platforms’ in press releases and other public-facing communications, Gillespie argues that this carefully chosen term does the discursive work of helping tech giants to elide responsibility for the often-problematic content they host. The term platform implies ‘a kind of neutrality towards use – “platforms” are typically flat, featureless and open to all’ [Reference Gillespie16], but social media are, of course, anything but. Interestingly, in recent years we have seen a growth in nostalgia for Myspace and similar social sites like Bebo, BlackPlanet, and Neopets; a nostalgia that is, in part, connected to a discontent with failures in current social media governance, and with their inflexible features for representing the self [Reference Miltner and Gerrard9].

The Self on Social Media

Social media platforms often share similar features, many of which I have already described. But each platform fosters its own culture, partly due to the identities people are and are not allowed to maintain. As Gibbs et al. [Reference Gibbs18] note, each social media platform has its own ‘vernacular’: its unique combination of style, grammar, and logic which distinguish user experiences from platform to platform. This is perhaps best represented by comparing two platforms with contrasting approaches to identity.

In response to evidence that users may have several different accounts, and that use of pseudonyms is common, some platforms, like Facebook, want you to use your ‘real’ name and to tie you to a form of government-issued ID if necessary. But others, like Reddit and 4chan, actively discourage the use of legal names. The founders of Facebook and 4chan famously went head-to-head with their differing views on social media identities, with Facebook’s Mark Zuckerberg claiming ‘the default is now social’, with users required to bring their ‘real’ identity [Reference Patelis, Lovinik and Rasch19], and 4chan’s Chris Poole arguing ‘anonymity is authenticity’ (emphasis added) [Reference Hogan, Hartley, Burgess and Bruns20].

Zuckerberg’s argument resonates to some extent with recent academic theories proclaiming the ‘embeddedness’ of ‘online’ and ‘offline’ lives, spaces, and selves [Reference Hine21]. While these theories hold true in many contexts – particularly around the near-global uptick in smartphone and smart-device users – the distinction between online and offline identities is enduringly meaningful to people who want to use social media to be a different version of themselves, and this is precisely Poole’s point. One example comes from my own research, led by Anthony McCosker, which found that people often run pseudonymized meme accounts on Instagram as a way of talking about their experiences of depression [Reference McCosker and Gerrard22]. This form of dark humour and identity concealment could only be made possible on platforms that allow users to maintain pseudonymous identities, separating themselves from legal names and documentation. For these individuals, the online/offline distinction is complex. Further, pseudonymity can offer safety to particularly marginalized or outlawed identities, such as LGBTQ people in unsafe environments. Pseudonymity is therefore not just a tool for privacy and identity play; it has real, bodily stakes for many people around the world.

A platform’s vernacular is crucial to user experience, and Goffman’s work on performativity can help us to make sense of this. Goffman broadly explains that people present different versions of themselves according to their audience; a theory that has been used to inspire more modern research about how difficult it is to know who your online audience is, and therefore how to ‘perform’ [Reference Goffman23]. On a given social media platform, especially one with large cross-demographic uptake (such as Facebook), different social groups – family members, co-workers, neighbours, acquaintances, ex-partners, friends of friends, people from hobby/interest groups, past and present students, etc. – are now collapsed into one: the friends list. This phenomenon is known as ‘context collapse’ [Reference Marwick and Boyd24]. Because of this, Hogan argues that people define the lowest common denominator of what is appropriate to post on a certain social media platform (or, in other words, the least controversial thing to say) [Reference Hogan25]. These complexities mean people present different facets of their identity according to the vernacular of a given platform; something van der Nagel calls the ‘compartmentalisation’ of the self [Reference Van der Nagel26].

Expanding on Goffman’s theories, Ditchfield proposes the ‘rehearsal stage’ as a phenomenon unique to social media through which a person prepares their communication and carefully tends to their imagined audiences (thereby putting to bed any grand claims that the quality of communication has declined because of social media) [Reference Ditchfield27]. What’s particularly interesting about this theory is that it allows for people to choose not to be social. Put differently, using social media is not necessarily about being social, or even about disclosing any aspects of the self to a public audience. Sometimes people want to be present on social media but not say or do very much, and this is called lurking.

The participants in Ditchfield’s research often agonized over how much (or little) to say in a Facebook Messenger chat, or whether to abstain from replying altogether. Some people go further than this by avoiding social media altogether; downloading apps to restrict their use of certain platforms, deleting their accounts, or keeping their accounts but avoiding them. In short, there are myriad ways of avoiding social media; a concept that has received significantly less academic attention than social media ‘use’. While it might seem odd to include points about avoidance in a section on the self, I would argue that in certain pockets of society – where smartphone uptake is high and using social media commonplace – a decision to not use social media still means you have a relationship to them, however distant, or troubled.

Writing about Social Media: A Final Thought

While this chapter has outlined several key aspects of social media – their evolution, core characteristics, underpinning business models and governance structures, and capacity to enable explorations of the self – it should perhaps end by acknowledging that social media platforms are not ‘everywhere’, as some might dramatically claim.

Not everyone has access to the Internet or internet-enabled devices, and not everyone chooses to use them. Some social media platforms become globally popular (Facebook), and others are regionally popular (VK); some attempt to appeal to all (YouTube) where others focus on particular demographics (BlackPlanet); and some are not available to everyone who wants to use them. Famously, China’s ‘Great Firewall’ screens out many social media platforms popular in the West, such as TikTok and YouTube. This means social media are not experienced or understood universally, and there are still many, many people who are yet to use them. Some individual platforms, like Instagram, have so many features and nuances – Posts, Reels, Stories, Direct Messages, Comments – that it’s inaccurate to even describe Instagram as one ‘thing’ anymore.

In short, and as the writers included in this book will tell you: social media are very, very complicated.