In the new issue of the journal Science, Judith Donath, a professor at MIT's Media Lab, worries what will happen to social trust as our physical interactions move to online virtual worlds like Second Life (the article is not online). Donath is mainly interested in the set of behaviors that we know, colloquially, as "body language." The avatars that people build in online graphical worlds are increasingly incorporating sophisticated human behaviors such as, for instance, eye contact -- but what if we use these techniques to fool each other?
Eye contact is an important real-world cue. When someone holds your gaze, you tend to assume a certain level of engagement and credibility -- there's no evidence suggesting that people who look away from you while they're talking are lying scum, but we nevertheless assume so. But computers can give us eye contact without any level of engagement; your Second Life avatar may be programmed to look directly at another avatar you're talking to, but you may actually be paying attention to something else -- another avatar, your cat, the TV.
Donath points to an even more fascinating phenomenon. In an experiment, the researchers Jeremy Bailenson and Andrew Beall showed several students a picture of a political candidate that was digitally morphed with a photo of each student's own face. The morphing was subtle; of 36 subjects in the experiment, only two detected that the photo contained a bit of their own features in it. (When asked to name someone the candidate looked like, people said things like, "He looks like my grandfather" or "He really looks familiar but I'm not sure who he is.") But even though they didn't consciously recognize the morphing, the subjects reported finding the adjusted candidate's arguments more persuasive than those of a candidate who didn't resemble anyone they knew. (Read more about these experiments in "Avatars at Work and Play: Collaboration and Interaction in Shared Virtual Environments," which is searchable on Google Books.)
Donath's arguments are hypothetical; she notes that in today's graphical worlds, "behavioral sophistication lags behind rendering skill, so we have avatars whose appearance raises high expectations of humanlike behaviors but whose gaze and gestures are relatively primitive." It's conceivable, though, that in a few years' time, "avatars whose behavior is nearly imperceptible from humans' will be available," she says. Already, politicians, business and others who would seek to influence us are setting up shop in online worlds. You can bet they'll be using behavioral techniques -- eye contact, face morphing and other things we haven't yet thought of -- to trick us soon.