In one sense, all media are social, and all media forms have always been social. As a species, we learned to sing and recite stories and poems in a circle, around a fire, as a way of forging social bonds through cultural exchange. Movie theaters have been sites of courtship for more than a century. Films don’t end as credits roll. People continue to discuss them around offices and barbershops. Mail, telegrams, telephone calls, and emails all connect individuals to each other, each media form more conveniently than the last.
So when we refer to Facebook, Twitter, Instagram, or even Tinder as “social media,” what do we mean? How are these services distinct from other, older forms of connectivity and information assimilation? What’s so “social” about social media?
The operative definition of a social media service within the field of social media studies was established in 2008 by two researchers in the emerging field of *internet studies, danah boyd and Nicole Ellison. They defined social media services as bounded systems that allow people to enroll and construct a profile of themselves, let them choose other profiles with which they would interact, and allow them to view material posted by their chosen list of profiles. Ellison and boyd generated this broadly applicable definition when most social media services were websites, before most of the activity moved to freestanding applications that run on mobile platforms like smart phones or tablets. But, just as importantly, they chose not to include one of the most important aspects of the most dominant social media services around the world: algorithmic amplification.
Ellison and boyd emphasized the choices that users make to construct their profiles and to interact with select other members of a service. But as the second decade of the twenty-first century rolled in it became clear that the most important relationship in social media services such as Facebook, Instagram, Twitter, or even the employment service LinkedIn is that of user to algorithm.
Users often unwittingly feed their information to a database. The database thus contains deep and broad records of user preferences, desires, queries, and patterns of interaction with other users’ profiles and with sources of information and commercial service. Through phones social media platforms track users’ movements, locations, and even communication via text message outside of the platform itself. All these data help the algorithms not only predict but also shape what a user might want to read, view, or buy. Social media services design their algorithms to feed users more of what they have habitually expressed interest in, keeping users fully engaged and offering them constant affirmation through a count of “likes” and comments from fellow users. Algorithms select posts that generate significant engagement—clicks, shares, likes, and comments—to spread them faster and further than other posts, landing in front of more users. This selective amplification structures the experience of using Facebook, Instagram, or Twitter. It keeps people using the service. It focuses interest and limits exposure to expressions and opinions beyond what a user has desired. And it positions advertisements effectively and efficiently before likely consumers. Overall, the relationship among users, the data, the algorithms, and advertisements is synergistic and revolutionary. The rise and global scale of Facebook and other social media services is unlike any previous media form.
These platforms have risen in the first two decades of the twenty-first century to dominate the global media ecosystem. By 2020 Facebook had more than 2.5 billion users—nearly one-third of the population of the Earth. WhatsApp and Instagram, both owned by Facebook, enjoyed more than 1.5 billion users each. And WeChat, the largest social media platform in China, had the loyalty of more than 1.3 billion people. Facebook hosted content in more than 110 languages by 2020.
Through these services people share accounts of their lives and desires, clips of videos they had made or watched, links to articles they had written or read, and expressions of emotion ranging from joy to rage. That meant that social media platforms contain and carry the content of previously dominant media forms: text; recorded music; video. But more than that, social media platforms record and relate the feelings and opinions that people express about that content.
While films live on around the water cooler, content on Facebook arrives via Facebook and conversation around and about that content remains on Facebook. The water cooler is now in the theater, and the audience need never leave. Facebook and Twitter first arrived as websites, designed to be used on personal computers attached to keyboards that sat on desks. This model worked well in the early years of these services (2004 through 2010 for Facebook; 2006 through 2010 for Twitter) as they grew in North America and western Europe, where computers and broadband data connectivity were ubiquitous and affordable.
By 2010 most social media services had turned their attention to becoming chiefly mobile platforms, used mainly through telephone handsets like the iPhone, which was introduced by Apple in 2007, or those using the Android operating system introduced by Google in 2008. By 2011 the majority of Facebook usage was through the mobile application rather than the traditional websites. That growth has continued as Facebook expanded around the world and made its service available in more languages. By 2018 more than 75 percent of Facebook usage was mobile. This shift meant that Facebook users no longer considered Facebook a “place” to “go” at particular times of the day, whenever one happened to be sitting in front of a computer. Facebook by 2018 went everywhere that users went. The time that Facebook would not be available to users would be infrequent, and its presence would be constant and ubiquitous.
The relationship between those who engage with social media platforms and the platforms themselves is complex and unlike that between any previous media industries. Film viewers are a paying audience for the experience of watching a film. Commercial television viewers are the product that stations and networks sell to advertisers. Telephone and telegram companies relied on paying customers to provide the connections and content that made paying for the service worthwhile in the first place. For social media companies, all these relationships complement each other. Users work for social media platforms, as they post all the text, images, and video that constitute the content of the medium. Users also are a commodity to social media companies, as they provide the target for advertisements. And social media users choose the fellow users with whom they wish to communicate. But there is one aspect of the relationship that is new to social media and absent media forms that predate 2000: social media companies engage in total surveillance of users.
This surveillance, the ability to record every user’s location, movements, internet activity, and preference—as well as the patterns of interactions between and among users—allows these companies to map social relationships and predict which products, services, ideas, and expressions a user might desire. This allows for two revolutionary aspects of the commercial activity of social media platforms. They can ensure that advertisements only reach those who have expressed interest in a type of product, service, or political candidate. Surveillance enables a revenue-generating process that not only made Facebook one of the most highly valued corporations in the history of the world by 2020 but subtracts that revenue from some older advertising-driven media industries such as newspapers, magazines, television, and radio broadcast companies. The second aspect is just as important. Data accumulation lets these companies predict which types of content are likely to generate responses from particular users, thus encouraging them to come back to the service frequently.
Therefore the “social” in “social media” refers not merely to the fact that people assemble their groups of friends and followers and contribute almost all the content that makes the services worthwhile. It more properly describes the ways in which computer algorithms deploy vast collections of user-created data to forge the experience of interacting with the mediating service, rather than with other people directly. Every interaction within a social media service is governed by algorithms. These services choose what users will see, read, and with whom they will most frequently interact. Data run through algorithms dictate the terms of those experiences. It turns out that what makes social media distinct from previous media forms is the opposite of what we usually think of as “social.” It’s not about relationships among people. It’s about relationships between people and machines. Social media are more antisocial than social.
The phenomenon of social media arose in the first decade of the twenty-first century only because of a confluence of technological, economic, and ideological forces, each of which amplified the other. The *World Wide Web was conceived in 1990 as a collection of documents, linked within text, and indexable by tracking relations among terms. By the late 1990s many who contributed to the development of the web considered ways to make it “social,” by emphasizing connections among people rather than among documents. Much of this work was inspired by the work of social scientists like Stanley Milgram, who explored the “small world” phenomenon and posited that everyone is separated from everyone else by six degrees, and Mark Granovetter, who mapped relationships and described how networks of people work together. The idea of making the web social was to enhance collaborative creativity. Much of the infrastructure of the internet and many of the most impressive software projects of the 1990s had been built along *“open source” principles. Open source meant that the computer *code was open to revision by any member of a coding community. It would not be controlled by any single company or person, but firms could build on open source software to engage in commercial activities. This idea, that dozens or hundreds of authors could do better, cleaner, more efficient, more elegant work than any single creator could, grew to be a canonical belief in Silicon Valley and beyond. If networks of people could find each other and collaborate, amazing things would emerge.
Many of the early developers of what would become the major social media platforms emerged from a “hacker culture” that celebrated such open source styles of development. They applied such principles easily from software to society. People, it seemed, were objects to be tagged, classified, arranged, and managed.
An influential strain of psychological and social science also influenced the idea that electronic networks could bond and manage people, not just pages. Sociologists and psychologists worked to extend some fascinating early work on the makeup of social cohesion, including the famous “six degrees” experiments in the 1960s by Milgram and others. This work inspired popular thought about what the internet could do if people could communicate and collaborate over long distances at no marginal cost.
Not coincidentally, the earliest web platform that explicitly asked members to create a profile and choose their connections was called sixdegrees.com. It debuted in 1997, but it developed no way to generate revenue, and its technical infrastructure failed as its popularity grew. The same issues condemned the next few platforms that arose in the last years of the twentieth century: LiveJournal; AsianAvenue; BlackPlanet; and MiGente. By 2002 there was enough broad interest in social media interactions that a service called Friendster managed to generate significant attention and enrollment, only to collapse by 2008 just as MySpace and Facebook rose to dominate the industry. The Friendster collapse was the result of weak infrastructure that could not handle its rising popularity along with a lack of any source of dependable revenue. Frustration with fake profiles and frequent service interruptions drove users in North America to MySpace and Facebook, but Friendster continued to thrive for several more years in the Philippines, Singapore, Malaysia, and Indonesia. Meanwhile, a Google project called Orkut grew to dominate the social media market of both Brazil and India, only to be displaced in both countries by Facebook by 2013.
When MySpace launched in 2003 it was just another one of dozens of social media platforms, many of which attracted venture capital investments, but few of which created sustainable revenue. MySpace differentiated itself from others by allowing users to customize the design of profiles. This appealed both to musical acts and to young people who followed those acts. Youth interest drove enrollment in MySpace into the millions and ultimately attracted Rupert Murdoch’s News Corporation, which purchased the service in 2005 for $580 million. MySpace, which focused heavily on young users and thus limited the extent of its “network” in contrast to its rivals was worthless and out of operations by 2011. It had been replaced by an insurgent that had learned from the mistakes of Friendster and MySpace: Facebook.
Facebook started in a Harvard University dormitory in 2004 and served mainly US university students for the first two years of its existence. Its founder, Mark Zuckerberg, resisted growing too quickly before upgrading servers and connectivity so that the service would rarely crash even at peak usage times. By 2007 the general public could use Facebook in English. By 2010 Facebook was available in more than twenty languages. And by 2011 Facebook had spread to much of the world.
In the first few months of 2011 Western coverage of popular uprisings in Egypt, Tunisia, Morocco, Syria, Libya, and Bahrain inspired an assumption, largely unsupported by data, that Facebook and Twitter had sparked them. Journalists in Europe and the United States assumed, because they learned of these demonstrations through Twitter, and because a small group of Egyptian expatriates gathered on a Facebook group, that these services were essential to what were a series of complex political events. Most analysis of these movements at the time ignored long-developing tensions, labor unrest, civic organization, the role of the Al Jazeera news network, and the role of soccer fan clubs in rallying public opposition to an array of authoritarian governments across many Arabic-speaking countries. The focus was on the new: social media. Once governments fell in Egypt and Tunisia (despite surviving in most of the countries in which uprisings sprouted) the myth was sealed. The misnamed “Arab Spring” was a social media phenomenon. The assumptions spread that Facebook and Twitter energized movements that favored democracy and free expression. They were the tyrant’s foes.
Social media research revealed a very different phenomenon. As early as 2010, researchers saw how authoritarian governments used Facebook and Twitter to monitor and surveil dissent. Facebook groups were easy to infiltrate. Many dissidents and activists urged confederates to avoid Facebook and Twitter if they feared arrest or surveillance. By 2014 authoritarian leaders across central and Southeast Asia discovered that the algorithms that drove Facebook favored content that sparked strong emotions. Nationalism, xenophobia, racism, and other forms of bigotry traveled fast and furiously across the platform, making Facebook the best possible propaganda tool. In 2014 Narendra Modi used Facebook to drive his Hindu nationalist electoral victory in India. Soon others took note. By 2016 Facebook’s targeted advertising system and algorithmic amplification contributed centrally to the election of Rodrigo Duterte in the Philippines and Donald Trump in the United States. Facebook also played a central role in the surprise victory of the “Leave” campaign in the referendum on the future of the United Kingdom’s membership in the European Union. Also by 2016, Russia had mastered the art of using Facebook to inject divisive propaganda into democratic electorates, first in Ukraine and Estonia and later in the United States. The Russian Facebook-driven campaign to undermine civic trust continued from 2017 well into 2019 across Europe, with significant success in Poland, Spain, and Italy. Meanwhile, Facebook-driven campaigns helped promote nationalist leaders’ rise to power in Brazil and Indonesia. In 2019 Modi won reelection in India using Facebook and WhatsApp just as effectively as he had in 2014.
As social media platforms, chiefly Facebook and WeChat, grew to dominate the attention of more than half of the population of the world by 2020, they structured many social, cultural, political, and economic relationships. People began to purchase goods and services through these services. And by 2019 Facebook began developing its own encrypted currency in an effort to control, monitor, and monetize even more human activity. These services became more than just media through which we shared photos of children and pets. They became the operating systems of our lives. Not coincidentally, awareness of and resistance to this concentrated power grew as well. Governments around the world considered measures to weaken or dismantle Facebook and its affiliate services, Instagram and WhatsApp. As Facebook fought off such proposals, its leaders invoked the specter of its only remaining competitor, WeChat, and its close relationship with the government of China. The prospects for democracy remained unclear by 2020 for many reasons. The remarkable power of Facebook and WeChat are among those reasons.
Siva Vaidhyanathan
See also algorithms; censorship; cybernetics/feedback; data; databases; files; information, disinformation, misinformation; media; money; networks; platforms; programming; public sphere; surveilling