Tuesday, November 15, 2011

Negotiating multiple identities on the social web: Goffman, fragmentation and the multiverse

(This is a keynote presentation I delivered at this year's webCom Montréal)

Let me start off with a confession: I have multiple personas. No, not multiple personalities! I have multiple personas. And I’m pretty sure you do too! Erwin Goffman, the famous Canadian sociologist once observed that when we interact with others, we enter a stage and take on the role of an actor presenting a character to an audience. We start performing and in this performance, we present desired impressions of selves to others.

Goffman’s idea of social interaction as a performance of identity is not all that different from what happens when we join a social media platform and use it to connect with others. Except that things get a little more complicated when we enter the online world. In real life, the confines of physical space easily identify the situational context in which our performance is to take place. When I drive to campus in the morning, the buildings, the reserved faculty parking lot, even the physical layout of the classrooms (with the chairs facing towards me & the blackboard) all remind me that I am about to step into my role of professor and that that is the front I will be performing for the next hour or so. All these contextual clues make it easy to figure out what stage play I will be enacting, and identifying my audience takes all but a quick glimpse around the room.

Who are we performing for online?
Online though, things start to get messy. Stages merge and audiences become fluid. When I enter the Twittersphere, digital audiences aren’t as easy to define any longer. Yes, there’s my primary audience composed of the people who chose to follow me on Twitter. But to think that that’s my only audience might be a bit naïve. All it takes is one re-tweet for my message to leave the confines of my own Twitter network and to reach new audiences. Of course, my tweets also live on on my Twitter profile, which if set to public, means that my potential audience has just grown to pretty much anyone with an Internet connection. And let’s not to forget the Library of Congress, which in 2010 announced plans to acquire every publicly shared tweet since 2006. Let’s think about this! One day, years from now, my great-grandkids – an audience that doesn’t even exist yet - might be reading my tweets through the Library of Congress archive. With all these potential audiences, how are we supposed to know any longer who we will be performing for?

To make matters even more complicated, thanks to cyberspace, I can also be present in multiple places at once. In Goffman’s terms, I can now perform different plays to different audiences at the same time. Just think of a tweet: I can send it out via Twitter and simultaneously post it to Facebook and pull it into my blog, all of which have different audiences and serve different self presentational needs. That poses a problem though: Facebook Corinne and Twitter Corinne are not the same persona. And they’re also slightly different from Corinne, the blogger. I’m a lot pickier about who I let join my Facebook network and I rarely let mere acquaintances in. If you want to connect with me on Facebook, I have to know you fairly well. As a result, you’d probably get to see a much more unfiltered version of Corinne than you would on Twitter. Twitter Corinne is an engaged professor and researcher, tweets in a number of languages and aside from the occasional (but justified) rant about AT&T’s dismal phone service, tries to present a very professional image. But the point is this: although we may think of Facebook, blogs and Twitter as separate stages with different audiences each, sometimes the performance we stage for one audience gets viewed by an altogether different audience. Social media platforms have forced us to become actors on multiple stages with multiple sometimes overlapping audiences.

And that’s not even taking into account an altogether different audience. When Goffman published his now seminal book “The Presentation of Self in Everyday Life” back in 1959, he probably couldn’t have conceived of a digital world where actors might not just perform for human audiences but also for an audience of computers -- for search engines, or sophisticated algorithms. This additional audience goes back to a fundamental question about who these new social media platforms are really built for: “us,” the user, or “them,” the Internet giant, i.e. Google, Facebook, and company. When Google CEO Eric Schmidt argued that Google + was an identity service which depended on people to use their real names, he clearly answered that question. Google + wants you to use your real name, not because they want to protect you from putting creeps in your circles (although that would be nice), but because it helps them build better products. It helps them better target their ads and personalize your search results.

The fact that Zuckerberg and Schmidt didn’t just built their platforms for the common good without any ulterior motives may not come as a shocker, but the idea that we are no longer just performing for human audiences has important implications for online self-presentation and identity management. If our identities are socially constructed through our stage performances, it matters whether they are viewed through the lense of a human being or an algorithm. It matters because humans and search engines don’t see the same thing when they bump into you online.

My abandoned SL Avatar
Our online identities are fragmented. Partly because the web has become fragmented with walled communities popping up everywhere. As a social media professor, I feel pressured to keep up with these communities, so naturally I set up shop in them as soon as a new one arrives. In fact, I have created identities in so many of these services that I have no idea how many parts of me are floating around the Internet. I know there’s VirtualCori, my Second Life avatar from way back when Second Life was still considered cool. Last I checked, she was stuck on my university’s island chained to a wall in Plato’s Allegory of the Cave (for those of you who are wondering: I had to leave her there at the end of class because I couldn’t figure out how to unshackle her)… And then there’s @corinnew, cweisgerber, and all the other ones whose avatars or screen names I don’t even remember! Google tells me there are 3,470 fragments of my identity strewn all over the Internet. I have a fairly unique name and there are only 3 other Corinne Weisgerber’s on the Internet that I am aware of: a 9 year younger version from my native Luxembourg, a slightly older version from next-door Belgium, and a Twitter spammer from San Antonio (I won't link to her - don't want to give her any Google juice) who tweets under my name and promotes winter jackets in sunny San Antonio… I can say with certitude that most of these 3,470 search results refer to me.

We know that most Internet users will never see all of that. Actually, the typical Internet user doesn’t look beyond the first three pages of search results. But what if an audience could see all 3,470 fragments and piece them back together? That’s exactly what data mining engines do. And that’s what sets them apart from human beings. While we see identity fragments, engines see identity aggregates. That’s not to say that humans can’t sift through all the available identity data, but merely that most of us don’t have the time or inclination to do so. Actually, leave it to the French to do that! In 2009, a French magazine called Le Tigre published an intimate portrait of a randomly chosen Internet user laced with private information garnered from social networking sites around the web. They called it the Google Portrait of Marc L. The idea was to pick a complete stranger and tell his life story based on the digital footprint that person either voluntarily or involuntarily left behind on the Internet.

Just like Marc L., we may feel comfortable sharing bits of private information online because we think that this one bit of information won’t jeopardize our privacy. We may even comfort ourselves thinking that when looked at in isolation and by its intended audience, it doesn’t reveal much. But what we tend to forget is that taken together, these pieces of information grow much more powerful. Once aggregated, they can draw a cohesive and troublingly intimate picture of our lives, or worse, they can completely misrepresent who we are. And that’s where another danger comes in. Computers can’t compete with human audiences when it comes to inferring meaning from pieces of data – at least not yet. They may not see the apparent irony of a pretend Twitter user from San Antonio, Texas who touts the virtues of fur jackets in the middle of one of the hottest summers in Texas history. Worse, the computer's inability to separate data from multiple owners of the same name may lead to inaccurate online portraits, such as those produced as a form of critique by MIT’s Personas project, which scours the web for information and attempts to characterize the person.

So if the people we perform for only see a fragmented identity, and computers can’t be trusted with making sense of these fragments, where does that leave strategic self-presentation? When we enter the realm of strategic performance of identity, the buzz-phrase that gets tossed around the most is that of personal branding. It’s interesting that the phrase personal branding (although coined over 10 years ago) was popularized only recently when social media started taking off. It's interesting because we just entered a new digital communication era, an era in which the traditional top-down communication model was flipped on its head and the laws of message control no longer apply. It’s what we teach in social media 101. It’s what a lot of businesses had to learn the hard way: they are no longer in control of their messages. Yet, despite all these dramatic changes, we chose to talk about online identity in terms of a branding metaphor. A metaphor that is built on outdated assumptions of message control.

Why would we tell businesses they no longer control their messages but then go on pretending that we control our own personal brand? Sure, it makes us feel better to think that we do. But do we really? Interpersonal communication research tells us that on social networking sites the people in our network actually co-construct our identities. For instance, we know that if your Facebook friends are physically attractive, others on Facebook will perceive you as more physically attractive. If your Facebook friends are unattractive, you in turn, will be perceived as less attractive! An MIT project, referred to as Project Gaydar, even suggests that your Facebook connections can give away your sexual orientation. You may chose not to state your sexual orientation on Facebook, but you can't prevent your friends from identifying theirs. And according to a team of MIT students, that’s enough information for an algorithm to predict whether or not you're straight or gay! The point is this: your online friends contribute to the construction of your identity the same way customer reviews on Yelp contribute to the construction of a brand’s identity.

So why treat our online persona as a brand? We’ve just seen that this marketing concept doesn’t work so well in a world where your friends can inadvertently leak your sexual orientation by identifying theirs. But even this branding metaphor is built on another metaphor: the word brand comes from “to burn” as in firebrand. Egyptians burned ownership onto cattle and slaves, which were seen as livestock and not people. This same practice was later continued in the U.S. In essence then, personal branding enslaves us as property of a strict image or set of ideas. A brand says we are this—not that. This idea is captured in popular personal branding advice such as “be clear about the image you intend to project. If you have more than one message you run the risk of confusing people about what you are all about.” Or “make certain your brand message is consistent across all platforms.”

I argue that being too concerned with branding restricts the self. Just take a look at U.S. leaders who conflate themselves to the ideology of the party even when it’s clear their own beliefs are far more diverse and subtle. This has lead us to distrust elected officials as we see them as merely parroting talking points. Now compare that to a person like Steve Jobs. Jobs refused to be branded. He was not Apple. He was not Next, or Pixar. He was a unique self, full of contradictions and that’s what humanized him. That’s why we saw the outpouring of support when news of his death spread across the Internet.

The branding metaphor may work for a business but it doesn't work so well for an individual. Issues of ownership, distinction, possession, and enslavement confuse the metaphor. The problem is that we think of a brand in the classical way of thinking of the cosmos: it’s either this or that. It can’t be both. It’s all about getting the positioning right. We used to think of a particle the same way: it is either here or there. It can’t be both places at once. It can only have one position. Or can it? Quantum mechanics suggests it can. According to the Heisenberg principle though, once we observe the particle and try to measure it, we disturb the way it behaves. This in turn changes what we see.

Maybe that’s the problem with online identity. If you look in one place you see one aspect of a person’s identity. If you look in another place you find another aspect. What you’re looking for, where you’re looking for it and the instruments you use to do so will determine what you see. Just like in quantum mechanics. Maybe that's why we need a new metaphor to talk about online identity. Maybe the idea of the multiverse with its multiplicity of possible universes could somehow inform our concept of identity.

When you think about it, the Internet literally chops our identities into packets and hurls us piecemeal around the globe. Our digital identities, reduced to subatomic particles or electrons, fly at near light speed through semiconductors, wires, and cables strung across the ocean floor. We mount to the air as waves from satellites, cell phone towers, and wi-fi hotspots. We shoot out as streams of photons from our screens, as waves of sound from our speakers, and glide across the surface of our tablets with the brush of a finger. 

Our very identities have become the indeterminate particles and waves of quantum theory. We do in essence exist in millions of places at once, being observed by a million others who interpret us in a myriad different ways. The Internet defies position, embraces fluidity, and fosters multiphrenia. Whether or not the concept of the multiverse stands the test of scientific rigor, I argue that it is an apt and useful metaphor to inform discussions of identity in our time. We can no longer speak definitively of position, of brand. Instead, we must speak of multiple voices and multiple interpretations, coexisting throughout our physical and digital world. We must embrace our multiple personas.