Social Media Is Making Us Anxious and Paranoid

So why can’t we stop using it?

The 2009 documentary We Live in Public told the story of Josh Harris, a late-1990s dot-com millionaire who funneled his considerable fortune into what he considered the future: people broadcasting their lives via internet-enabled closed-circuit television. Harris founded an internet television network with channels like “88 Hip Hop” and “Cherrybomb,” but streaming technology was limited at that point, and only allowed for choppy, frame-by-frame video.

When his internet TV venture failed, Harris built an underground bunker in Manhattan, filled it with television screens and cameras, and invited a collection of scenesters and technologists to move in. The bunker also included a shooting range, random cross-examination of participants, and plenty of recreational substances. (This experiment quickly devolved; the combination of drugs, alcohol, guns, and CIA-influenced interrogation techniques did not produce positive results.)

Finally, Harris and his girlfriend fitted out their apartment with cameras, including one in the toilet, that broadcast to the web twenty-four hours a day, seven days a week. The website weliveinpublic.com included a forum where viewers could weigh in on the couple’s activities and arguments. The relationship, unsurprisingly, did not last.

Smith was eccentric, but his vision of the future has come to pass for a sliver of the population, especially in tech culture. Reality television, Skype, FaceTime, Twitter, Nike+, GPS-enabled cell phones, Instagram, Facebook, Spotify, YouTube, and hundreds of other media have popularized the capturing and broadcasting of personal information to large, networked audiences. While most of us don’t live in apartments with bathroom cams, many of us have tablets or smartphones that make it simple to upload photos and micro-blog entries. The Pew Internet and American Life Project found that 88 percent of American adults own cell phones, and of those who do, more than half use their phone to go online. Of the 44 percent of adults with a smartphone, 90 percent access the mobile internet.

The influence of always-on internet has been rapid and significant. Texting, Facebook, and Twitter are used by teens to remain in nearly constant contact with friends, creating strong bonds of intimacy and togetherness. Celebrities use Twitter to stay in touch with fans by strategically revealing insider information. The technologists who I studied intentionally reached out to followers to increase their visibility and social capital in the scene.

In my research I have found that social software may inadvertently promote inequality rather than countering it. Metrics, like follower count or number of “likes” on a photo, facilitate this process by rendering social status into something that can be quantified, qualified, and publicized. The process of what I call “digital instantiation” works similarly toward quantification, qualification, and publicity by rendering users’ lives in piecemeal fashion, unintentionally creating a whole that is larger than the sum of its parts. Social media tools digitize formerly ephemeral pieces of information, like what one had for breakfast, making it possible to create a bigger picture of a person or community’s actions. Once “breakfast” is captured in a Foursquare check-in or Instagram photo, it can be combined, searched, or aggregated with other pieces of information to create mental models of actions, beliefs, and activities. Within this context, social surveillance, or the monitoring of friends’ and peers’ digital information, becomes normal.

While lifestreaming has plenty of social and emotional benefits, it also comes with costs. Lifestreamers must see themselves through the gaze of others, altering their behavior as needed to maintain their desired self-presentation. This constant monitoring against the backdrop of a networked audience creates anxiety and encourages jockeying for status, even as it brings forth new forms of social information.

Looking at lifestreaming as something a community does makes it possible to evaluate information disclosure beyond platitudes about privacy. In a social context where your peers expect you to share information with them, it’s obvious that sharing has nothing to do with whether you care about privacy or not.

Rather than looking at social media use as an intrinsic privacy violation, lifestreaming needs to be understood as an act of publicity. Lifestreaming can be used to publicize knowledge; to gain emotional benefits, social capital, and information; or to shore up support in an argument, but it is rarely used as a way to disregard or eliminate privacy. Most lifestreamers have sophisticated understandings of what they would or would not share online. They balance their need for publicity with their desire to control their own online image. The necessity of presenting an edited self to the world requires a careful understanding of the risks and benefits of information sharing.

The networked audience

The audience is a key part of lifestreaming, because lifestreaming without an audience is just “tracking.” Lifestreaming involves broadcasting personal data to other people, whether this is anyone with an internet connection, or a filtered subgroup of readers. In a social group of lifestreamers, people place themselves as part of a networked audience in which participants are both sender and receiver. Looking at the collective lifestreams of a group shows that players constantly reference each other, revealing a coherent picture of social actions and connections within a community.

Almost all the people of the tech community I studied contribute to their own lifestream. These lifestreams make up the Twitter stream of people one follows, or the Facebook News Feed of one’s friends. Thus as each person lifestreams a piece of content, they are simultaneously reading the content of others, commenting on it, and adding it to their mental picture of the scene. Audience members watch each other’s actions by consuming their content, and by doing so formulate a view of what is normal, accepted, or unaccepted in the community. This understanding of audience creates an internalized gaze that reflects community norms. Members of the tech scene imagine how the audience will view their own lifestreamed self-presentation, and alter it accordingly. Monitoring of oneself and others thus becomes an expected and normative part of this social interaction.

I use the term “audience” rather than “public” when describing viewers of a piece of digital content (as in “this picture is public”). “Audience” can refer to the audience one imagines while texting or posting a photo, the actual audience, or the potential audience for one’s content. But while “potential audience” resembles the vernacular sense of “public,” “audience” here means the actual audience, the people interested in a piece of information who actually view it. Just as media professionals do not use the term “public” for people watching a movie or TV show, we should not use it for digital content.

The use of audience also implies performance, because a lot of digital content is created with impression management in mind. While it is never possible to determine who exactly has or has not viewed something online, because the actual audience may be very different from what a creator imagines, keeping in mind the difference between publicity done for an audience and information made public will help us to understand some of the social dynamics I describe.

The networked audience is distinct from the broadcast audience in that the networked audience is connected. The tech scene is a superlative example of the networked audience, because the social element is articulated both on and offline. Unlike many online communities where a small percentage of people create most of the content, people in the tech scene act as both content producers and consumers to maintain status and intimate ties with the community. Lifestreamers read others’ lifestreams and create content with their audience in mind. Their online and offline lives are intrinsically interwoven, meaning that nonparticipation has real social costs.

The networked audience is distinct from the networked public, which internet scholar danah boyd defines as the social space created by technologies like social network sites and the imagined community that thrives in this space. While it is possible to describe a single site like Twitter as a networked public (although I would not do so), the term networked audience is more appropriate for lifestreaming. “Networked public” implies a set of people communicating through a single technology (MySpace, Usenet, World of Warcraft, and so on), while the networked audience moves across sites. The concept of audience implies a specific set of people interested enough to view digital content rather than an amorphous mass of potential readers. Given these properties, what does lifestreaming look like in a social group that uses social media intensively?

Lifestreaming in practice

Lifestreaming is a normal part of the technology scene. People expect their friends to be familiar with the latest social media applications and to connect and engage using blogs, Twitter, and Facebook. As Auren Hoffman, CEO of the reputation management firm RapLeaf, stated in our interview: “If you were an employer, and someone applied and they didn’t have any activity on social networks and that person was 23 years old, you’d think they were the Unabomber. You would be really scared to meet this person without even a bodyguard. I don’t even know if that person exists.”

To people like Hoffman, who are intimately familiar with social technology, not using social media marked unsophistication and backwardness. In Hoffman’s view, the relationship of employer and worker requires the familiarity of common social ties and community involvement; nonparticipation would not only make it difficult to contribute to social and technological conversations, but also potentially limit one’s economic mobility.

Consequently, most people I knew during this period used microblogging technologies, such as Facebook, Pownce, Twitter, and FriendFeed, to lifestream media consumption, location, digital pictures and videos, and the flotsam and jetsam of everyday life. The availability of these streams to an audience varied by individual and service, from entirely publicly accessible Twitter accounts to password-protected digital files. Lifestreaming ranged from piecemeal aggregation like FriendFeed, a trendy piece of software that pulled in dozens of data streams to create a semi-comprehensive picture of what friends were doing across the internet, to personal blogs that dynamically aggregated day-to-day doings. While I did meet people in the technology scene who used social media specifically to track personal data for self-improvement, they were a minority.

Proponents say this type of networked lifestreaming facilitates connections to others, deepens relationships, and creates a source of real-time information. Sharing information through services like Twitter creates an “ambient awareness” of others, a sense of what friends and acquaintances are doing or thinking that builds up over a long period of time. This ambient awareness is akin to a sense of co-presence, even if the participants are not geographically proximate.

At the same time, networked lifestreaming often creates anxieties about creating and maintaining one’s social identity in front of an audience, and the extra layers of social information can result in intense social conflicts and arguments colloquially referred to as “drama.” Drama is what danah boyd and I call “performative, interpersonal conflict that takes place in front of an active, engaged audience, often on social media.”

Drama can be a form of norm policing, where social media is used to call out community members who violate explicit or implicit social norms. While our definition of drama was formulated during a large-scale study of teenagers, it applies equally to other social milieus that display the same networked audience effects. Inferences and implications made visible by social media can reveal connections and actions that are usually tucked away from each other. These difficulties have given rise to a variety of different ways of conceptualizing the “public” and the “private” and of managing how information flows between different entities, websites, and users. This delicate balancing act is made even more difficult in a community where virtually everyone lifestreams.

Somebody’s watching me: Social surveillance

Before the internet, people found out about parties or romantic relationships by gossiping or asking friends. This type of knowledge wasn’t secret, but it wasn’t available to everyone and was rarely written down. Today, any member of the networked audience can peruse a Facebook invite to see who was or wasn’t invited, or look at Foursquare check-ins to see who is spending time together. Social information is digitized and aggregated through the lifestream to create a layer of relational data that lays over the ordinary social graph.

While this information facilitated bonding and personal connection, it also magnified gossip, suspicion, and uncertainty. My informant Jill suspected that her boyfriend was having drinks with Mary, a woman that she strongly disliked. Jill noticed that her boyfriend’s Twitter feed had been silent for several hours. She then saw Mary use Dodgeball to check-in to a bar on his street and subsequently tweet out a photo of the bar. Jill interpreted this information to mean that the two were together, and was convinced that Mary intended her to know about it. Combining information from both people’s lifestreams created a larger social picture that was interpreted through a lens of suspicion. In retaliation, Jill tweeted a message about trustworthiness without naming either party.

Social surveillance is the process by which social technologies like Facebook, Foursquare, and Twitter let users gather social information about their friends and acquaintances. As privacy researcher Christina Nippert-Eng writes, “Humans are constantly scanning, constantly receptive to and looking for whatever they can perceive about each other, for whatever is put out there.” Eavesdropping is a very human action, and people are resourceful at combining information from disparate sources to create a “bigger picture” of social activities. This picture is augmented by information provided on social media sites like Twitter or Flickr.

Social media has a dual nature whereby information is both consumed and produced, which creates a symmetrical model of surveillance in which watchers expect, and desire, to be watched themselves. The presence of the networked audience not only enables connection, it encourages performances of intimacy and conflict to elicit reactions from others. Social media creates a context in which people are constantly monitoring themselves against the expectations of others—a context that can provoke anxiety and paranoia.

In the absence of face-to-face cues, people will extrapolate identity and relational material from any available digital information. Jennifer Gibbs and her colleagues found that online personal ads were constructed with a hyper-aware self-consciousness because users knew that misspellings, cultural references, and even time stamps were likely to be scrutinized by potential suitors.

Similarly, in early social media like IRC or Usenet, people would infer identity information from e-mail addresses, nicknames, signatures, spelling, and grammar. Digital traces and nuances are often interpreted incorrectly, but the act of interpreting becomes normal. Privacy scholar Helen Nissenbaum writes that the value of aggregation is in extracting “descriptive and predictive meanings from information that goes well beyond its literal boundaries.”

Social media users are practiced in the extraction of nuance through ongoing analysis of the lifestream. While each piece of information by itself may not mean much, it creates a larger picture when combined with others. For example, knowing that Julie visited a local bar on Tuesday night is not, in isolation, particularly interesting. The bar is publicly accessible, Julie can expect to be seen there, and she will probably tell her friends where she is. If she tracks, codifies, and broadcasts this information using social media, however, the information can undergo a transformation. If analysis of the lifestream reveals that Julie’s best friend’s ex-boyfriend was also at the bar, and this is the third night in a row that they have been in the same place, a new picture emerges.

The accessibility and persistence of personal information tracked and broadcast through social media create an extra layer of relational data that is not easily explained by the dichotomy of “public” or “private.” It is very complicated to manage self-impressions and relationships with others when faced with this phenomenon.

People in the scene recognized these complexities and shared strategies on how to handle them. For example, two Digg employees, Aubrey Sabala and Joe Stump, proposed a (rejected) panel at South by Southwest called “Is the internet killing your game?” which described how relationships were affected by the lifestream. Digital pictures posted on Twitter, Facebook, or Flickr were open to interpretation, meaning that someone who wasn’t present when the picture was taken could jump to the wrong conclusion. As shown in the earlier example, “radio silence,” or “dropping off the Twitterverse” for a day, was noticeable and questionable.

They also mentioned what they called the “right hand vs. left hand problem,” which described situations where “not everyone knows not to Twitter something out.” This occurs when a group of people have different information boundaries, and someone lifestreams something that other group members want to keep private. These practices reveal intensive attention to detail and monitoring of other people’s lifestreams, which from my observation was common among San Francisco technologists.

All of this extra information, and the additional meanings it sometimes implied, made the people I spoke to anxious. Since it was possible to keep close tabs on virtually anyone with a lifestream, people in my study spoke of trying, and failing, to resist the temptation to monitor ex-boyfriends and girlfriends, rivals, or partners. Some people installed browser software that blocked them from looking at specific Facebook profiles or Twitter feeds so that they would not be tempted to “cyber-stalk” exes or their new partners.

But nothing was foolproof. If someone they wished to avoid was connected to the networked audience, their username or picture would pop up in retweets, @replies, and other people’s Facebook messages. This created endless social conflicts, and I frequently saw someone get upset because they saw a picture of their ex in their Flickr stream, or noticed when a trusted friend checked in with a sworn enemy. Because the networked audience includes indirect connections (for example, someone connected to a friend or friend-of-friend), it makes visible those interactions that one could otherwise avoid.

While social media helps people stay in touch with loved ones, find emotional support, and get up-to-the-minute information, it can also create problems. In social contexts where virtually everyone lifestreams, social media brings its own set of anxieties and difficulties, which emerge without clear methods of handling them. While the use of Twitter or Instagram should be a matter of personal choice, people often feel pressure to participate in networked public life; they fear being left out, or even that their career may be compromised without visibility. We live in a world where, for the first time, everyday people can commandeer the huge audiences once available only to politicians and celebrities. Celebrating social technologies uncritically, or condemning them piecemeal, does us a disservice. Instead, we must recognize both the great opportunities and new challenges brought forth by these new tools.

From Status Update: Celebrity, Publicity and Branding in the Social Media Age (Yale 2013), an ethnography of the “Web 2.0” era of San Francisco technology.