Assignment 3: Identity in Social Networking Services

Weblog assignment 3: Analyze one or more sites such as Orkut or Tribe.net. This analysis should focus on how the software shapes/limits your definition of self, but bring in elements from the previous assignment about cultural, dramaturgical, organizational or information theory dimensions as they bear upon how a sense of identity and place is constructed.

Most of my investigation involved the Orkut social networking service, but I touched on other services. Discussion here applies to Orkut except where otherwise indicated.

Dramaturgical considerations in Orkut

Brenda Laurel argues that we should design for action, but Orkut and Friendster both seem designed first for “characters” and configuration and exploration, with action as a secondary concern. I’ve spent probably 98 percent of my time on each service figuring out the interface elements and the site structure and exploring other people’s profiles and relations between people; I spend a very tiny portion of my time there doing "active" things such as posting information on my profile and on others’ profiles, messaging people, arranging offline meetings, and so on.

But this isn't a bad thing. Laurel goes too far in suggesting that "action" is the one and only worthy goal in interface design – in many cases playing with and figuring out a tool and its interface can be a worthy goal in itself. For instance, video game designers often intentionally introduce arbitrary difficulties into the game interface that make the game more challenging and enjoyable. Consider Dance Dance Revolution, a game in which players must dance skillfully by stomping on sensors in proper sequence: the game itself is extremely simple but the challenge lies in mastery of the dance-step interface. In the case of Orkut, the value of the interface comes not from presenting enjoyable challenges but through the ability it provides members to create a place for engagement in social exploration and activity. Its browsing and searching mechanisms, the network diagrams, etc. shape the ways in which members construct their identities and determine how they fit into their social networks. Some might say Orkut's interface defines these activites too narrowly.

Jane McGonigal reacts to interface design philosophies that emphasize building for action and transparency by demanding “opaque” interfaces that encourage more exploration and play in her manifesto "The Curious Interface." McGonigal has valuable points but I think her manifesto went wrong in the same way that Laurel went wrong: both take their philosophies to the extreme. They both imply that most interface design problems should be approached using a single philosophy, but there’s a need for both approaches. Games like Dance Dance Revolution obviously call for the playful/exploratory/opaque philosophy. Designers of a fire alarm should follow the clarity/transparency/action approach, making the alarm as simple as possible to operate during an emergency (while making it difficult to mistakenly activate). Context is key. The approach to designing an interface needs to consider the context of the interface and no single philosophy applies in all cases.

In Orkut's case, the designers might have enjoyed more success had they veered farther towards McGonath's side of the spectrum. People tend to actively tinker with and redefine and reinvent their definitions and presentations of self in countless subtle ways; a less compartmentalized and clearly-defined design approach might enable more of the flexibility and variability that are needed here.

Thoughts on the definition and presentation of self in Orkut

Goffman and Crush Lists: Erving Goffman puts forth the concepts of “give” and “give off.” Messages that we “give” to other people are transmitted intentionally and we usually have complete control over them; when I tell you that I didn’t steal your bicycle I’ve given you a message. “Given-off” messages, on the other hand, are harder to control. If I turn beet-red while telling you that I didn’t steal your bike, my blush gives off a message that differs significantly from the message given through my words.

Online we lose a lot of the “given off” subtleties that enrich our face-to-face communications. This might be a stretch, but one could consider Orkut’s “crush list” scheme a crude form of online “given off” communication. Here’s how it works: if I come across any other Orkut members who strike me as particularly attractive I can add them to my crush list. This list of people is visible only to me, but if anyone on my crush list adds me to their own crush list, Orkut notifies both of us that we share this mutual interest. If this happens it’s not full-on “given off” communication like an offline blush, because I don’t have to specifically turn on my blushing capability before I will blush. But it’s not a completely “given” form of communication either because the communication is not fully under my control and for all I know it will never take place.

        A crush list sidenote: The crush list scheme might be subverted by someone who crush-lists everyone she knows and announces this to all her friends. Such a person would learn which friends have crush-listed her after having shirked the personal risk that usually applies in the crush-list transaction. Such activities might tremendously devalue the crush list scheme. But as it stands interface difficulties make adding many people to a crush list very tedious and time-consuming so this might not be likely to happen unless someone writes software to automatically do this dirty work.

        Another crush list sidenote: On Orkut you can't add anyone to your crush list who hasn't filled out the "personal" portion of his or her profile. So in effect, it's impossible to have a crush on someone in Orkut-land unless they allow it by taking an arbitrary action, and most members are probably unaware of this obscure connection between this action and crushes. Orkut invented an arbitrary and artificial constraint on crushes that differs the way such dynamics operate in the outside world. Why did they do this? I can't imagine why; perhaps it's the same bizarre reason why Orkut prevents people from declaring themselves fans of anyone but their friends, when in the real world most fans don't personally know the people of whom they are fans.

Friend Devaluation: In the real world, my closest friends (at least those who live near me and who often hang out with me in public) become an important part of my definition and presentation of self. This important element is not available on Orkut. Just like on Friendster, on Orkut I can’t present to others which friends are my close personal friends, which are vague acquaintances, which are business contacts, and so on. In “Friendster and Publicly Articulated Social Networks” danah boyd points out that on Friendster the artificial, binary friend/nonfriend distinction leads to devaluation of “friend” status because people who use these services realize that many people accept friend invitations from people they barely know. Orkut repeats this Friendster mistake.

Constraints on Multiple Presentations: Orkut also prevents me from presenting myself differently to different groups; I get just one profile that appears the same to all viewers, whether they’re my boss, my grandmother or my drinking buddies. In this sense Orkut limits my presentation more even than Friendster, which now allows me to make my profile viewable only to people within 1, 2 or 3 degrees from me. But on Orkut I can send messages only to my friends, to friends of friends, or to a given community; in that small way I can present some things to some people and groups that others won’t see.

Why does Orkut require that testimonials be at least 30 characters long? Why can’t I publicly express my relationship with a friend in three words if I wish? This constraint is pointless and nonsensical.

Misleading Colors: Orkut’s network-diagram color assignments don’t make sense. In a particular display called the “network diagram,” the names and faces of as many as 35 of a person’s friends appear on the same screen, and most of the faces appear on colored backgrounds. The color assigned to each person (blue, yellow, or pink for “cool,” “trustworthy” or “sexy”) doesn’t really reflect how cool or trustworthy or sexy your friends consider you, as the Orkut documentation claims. Apparently you’re given the color category for which you received the most votes. So if you have 2 cool points and just 1 in each other category, you’ll show up as blue (cool), even though a person right next to you in the diagram who has 200 cool points won’t show up as cool if they have 201 points in another category. (Disclaimer: I’m not sure this is how these diagrams work because in practice they don’t seem to work in the way described in the documentation. If I misinterpreted any of this, please e-mail me and I will correct any mistakes.)

Posted by sean at Duben 16, 2004 11:29 PM | TrackBack (0)