Wednesday, September 15, 1999

An On-Line 'Quiz Show' Uses the Anonymity of the Internet to Reveal Biases By SCOTT CARLSON It sounds like a sweet way to get a Ph.D.: Joshua Berman, a doctoral candidate at Georgia Institute of Technology, spends his study time in grad school playing games -- specifically, playing, analyzing, and tinkering with the Turing Game, an on-line site that he created with his adviser, Amy Bruckman, "At this point we are comfortable enough to challenge and question and play with issues of gender, but we're not comfortable yet to even begin the discussion about race." an assistant professor of computer science. But Mr. Berman is serious about his project -- and he has his work cut out for him, because his Turing Game has fairly lofty goals. For now, the modest game seems like little more than a low-graphics, chat-room version of To Tell the Truth, the old television show, but Mr. Berman and Ms. Bruckman foresee the game's being used someday in diversity training in colleges and corporations everywhere. The Turing Game, they say, challenges utopian notions of a raceless, classless, genderless Internet -- it shows that you carry all of your biases and some of your attributes on line when you log on. It also works to break down some of those biases, and to bring those various races, classes, and genders to the table. "People are communicating increasingly on line, and there are different theories of what that means," Ms. Bruckman says. "Some people say that when we go on line we leave behind aspects of our personal identity. Others say that's silly -- you're still exactly who you are, and nothing has changed. "The truth is obviously somewhere in between," she continues. "I come to this project as an educational-technology designer, and I see this as a learning experience -- not just for us as designers, but for everyone who participates. The point is to help people question their assumptions." The Turing Game is deceptively simple. It's set up like a chat room -- a player logs on, using an alias, and Georgia Tech Joshua Berman (left) and his adviser, Amy Bruckman, say the game challenges utopian notions of a raceless, classless, genderless Internet. becomes either a member of the panel or a member of the audience. Panel members pretend that they all share a particular identity -- that they're all gay, or all Catholic, or all female. Audience members then lob questions at the panel to determine who's really gay, or Catholic, or female -- and who's an impostor. Audience members can discuss panel members' responses as a group or "whisper" to each other individually. After a round of questions and a vote by the audience on each panel member's truthfulness, the moderator reveals who the impostors are, and proposes a discussion of the game. One of the moderator's main tasks is to filter and edit the questions for the most interesting effects. Ms. Bruckman says vague questions are the most fun and can be surprisingly revealing. In a recent game, someone asked a supposedly all-woman panel: "How many rolls of toilet paper do you go through in a month?" "The women gave these reasonable answers, like two or three," Ms. Bruckman says, "but the men had this notion in their mind that women use a lot of toilet paper, so they answered 14 or 16." The women in the audience picked out the men right away, Ms. Bruckman says, "but the men looked at the 14 and 16 and said, 'Ah, those must be women.'" Skirmishes in the gender war aside, the game can also expose underlying and deeply set biases of the players. "One of the interesting things about the Net is the disinhibitive behavior it causes," said Mr. Berman, who is also studying psychology and sociology. He cites studies -- such as those of Sherry Turkle, a sociologist at the Massachusetts Institute of Technology -- showing that the anonymity of the Internet can allow people to be more open than they would be in face-to-face conversations. "We're leveraging off that -- using that power to get people talking more frankly than they otherwise would in a discussion about race or gender or sexual orientation," he says. "When you sit around in the classroom, even the dumbest student in the room knows that when you're talking about race you don't say, 'Oh by the way, I think black people are stupid.' The people who think that aren't going to say it, because they know they're not protected." Mr. Berman hopes that the "near-perfect" anonymity of the Internet will let players loosen up and blurt out something insensitive -- maybe even downright offensive -- but honest. Because, he says, that's when a game really gets going, when the lessons are learned. He cites a "Who is Jewish?" Turing Game he played recently. One audience member asked, "Is the reason Jews are smart today because the smart ones knew to leave Germany before the Holocaust started?" Of the three "Jews" on the panel, one was aghast, writing: "I don't even know how to respond." But the two others offered longer, more thoughtful replies. One panelist started a response with: "No, even very intelligent people ignore warning signs of disaster." As it turned out, that panelist was Jewish. So was the player who posed the offending question -- he asked it because he thought he could tell from the responses who was Jewish and who wasn't. "As a result of that," Mr. Berman says, "we had an incredibly productive discussion about how Jews react to the Holocaust and how Christians react to the Holocaust." Its creators are the first to say that the Turing Game is no panacea for prejudice, and that it is a work in progress with certain limitations. The chat-style "discussions" are a poor substitute for real, person-to-person conversation: The players' comments can be abrupt and are randomly ordered, since they appear in the order players enter them. To compensate for the absence of non-verbal communication, Mr. Berman is trying to expand the game's reliance on what he calls "emotes" -- notations like "Josh smiles" or "Susan walks across the room." In more-organized game settings, he says, he would add a final person-to-person discussion to wrap up the game. But even in faceless forums on line, players can be squeamish about addressing certain topics. Since creating the Turing Game late last year, Mr. Berman and Ms. Bruckman have been unable to start a race game, in which players would guess who is black, or Asian, or white. Players are eager to watch such a game, but few are willing to step up and pose touchy racial questions, much less answer them. "I think that tells you something profound," Ms. Bruckman said. "At this point we are comfortable enough to challenge and question and play with issues of gender, but we're not comfortable yet to even begin the discussion about race." The game is named after the "Turing Test," which was developed by the mathematician Alan Turing in 1950 to determine whether a computer can "think." In the Turing Test, a person types questions into a terminal, not knowing whether a computer or a human being is answering. If the answers are so persuasive that the person believes a human being is responding when in fact it's a computer, the computer passes the test. Once all the design issues are hammered out, Mr. Berman and Ms. Bruckman say, the Turing Game could be used in corporate human-resources programs and as an intervention tool in situations involving social tension between groups. Mr. Berman says he and Ms. Bruckman are also beginning to talk to colleges about making the Turing Game part of freshman-orientation diversity training. But for now, they both seem to savor the opportunity to learn from their project. It is indeed, for Mr. Berman, a sweet way to get a Ph.D. "I think this is a piece of interactive philosophy," says Ms. Bruckman. "It's a reflective experience for both the scholars and everyone who participates." Easy-to-print version E-mail this article