To get a sense of how edgy a career in sociology can be, consider this scene from the life of William Sims Bainbridge.
The year is 1971. A cavernous room in Cambridge, Massachusetts, lit only by candles and dominated by a goat-headed effigy. Bainbridge and a woman named Isolde kneel before a man in flowing robes. All around them, 50 people dressed in black chant in unison.
Purify me with the Fire,
Satan, test me in Your Pit of Fire, desire ...
Purify me with the Fire,
Tell me how to give my life to You.
Isolde is visibly nervous, shaking and sweating profusely, but Bainbridge appears calm. He has prepared for weeks for this initiation, memorizing texts, meditating, and for the past 24 hours, fasting. To infiltrate the group, he posed as a piano tuner--a skill he actually possesses--without mentioning that he is a sociology Ph.D. student at nearby Harvard University. Bainbridge flawlessly recites his vow to help bring about "the final judgment of humanity and the ending of the world of men." A bearded man with shoulder-length hair passes his hand through a flame and draws an inverted cross on Bainbridge's forehead. He's in.
Back then, official rules for conducting research on human subjects had not yet been established, he says, and such an investigation now faces "a serious amount of ethics review and paperwork." As program director for Human-Centered Computing  at the U.S. National Science Foundation (NSF) in Arlington, Virginia, Bainbridge should know.
As in the vast majority of "deviant" religious groups, Bainbridge found no evidence of criminality, let alone the murder and cannibalism of which Satanists were accused. He characterizes the cult as "the outpouring of the inner fantasies of about 200 very creative people" who broke from the mainstream, "recapitulating the origins of so many other religions, major and minor, throughout the history of our passionate and crazy species."
You could say that Bainbridge, now 67, was initiated that day not only into a satanic cult but also into a unique career. Understanding what drives some of the stranger social dynamics is Bainbridge's passion. "These small-group dynamics seem to be universal," he says, leading to the constant generation and extinction of protoreligious groups even today. Understand these, he says, and "we'll better understand human nature."
Like the people he has studied, Bainbridge has remained far from the mainstream, says Rodney Stark, a sociologist at Baylor University in Waco, Texas. "Mainstream science is about publishing what everyone else is publishing with very small changes," he says. "You'd better at least start off that way if you want to get tenure." But big ideas don't come to those who avoid risk, and Bainbridge started out by taking risks. In his early research on religion, "he was right and his critics were wrong," Stark says.
"He is one of those best of scientists who has courage and does not shrink from difficult issues," says Philip Rubin, a psychologist at Yale University. "Quite the opposite--he often seeks them out."
Bainbridge's latest research is no nearer the mainstream. He has plunged into online culture, studying people's behavior in virtual games and worlds. His interest? Immortality through "personality capture" and "the possibility that science or some kind of cultural movement derived from it might replace religion."
Special feature this week: Courting Controversy
See also in the Courting Controversy feature this week ...
Animal Rights (and Wrongs),  Sarah Webb
For scientists, working with laboratory animals can be a source of public antagonism and private ambivalence.
"My original theory about why I've always been outside the mainstream was that from earliest childhood I was an avid reader of science fiction," Bainbridge says, "and I simply assumed that in adulthood I would make it come true." It led him to become an undergraduate physics major at Yale, "brooding about fusion drives," but soon he was seduced by sociology. At least part of his motivation for studying religious groups seems to be the taboo of it. "Many people have the feeling that science must not evaluate the truth or falsity of 'sacred' religious beliefs," he says. "But, for science, nothing is sacred."
Bainbridge is a secular researcher through and through. "It is clear to me that most religious beliefs are simply false," he says, although acknowledging that they may--or may not--have positive consequences. "Can religion heal the body? The mind? Does religion deter suicide? Does it prevent crime? Are religious people more moral than nonreligious people?" All of these are "empirically testable," he says, and considering the vast number of religious people in the world, "you'd think that this would have been thoroughly studied." But these questions haven't been, partly because "most of the researchers focused on religion are themselves religious." This presents a conflict of interest when the outcome of research may prove that religion yields few clear benefits for its adherents.
A more fundamental question hooked Bainbridge early: How does religion arise in the first place? Newly emerging cults are a natural laboratory, but studying them is no walk in the park. "I have been amazed by [Bainbridge's] ability to secure permission to study groups who shun attention," says Roger Finke, a sociologist at Pennsylvania State University in State College. One of those was the "Children of God" cult that thrived in the United States in the 1970s. Bainbridge says he gained entrance to this one by being "honest and transparent" about his interest. Eventually, a member of the cult approached him directly. His survey of 1025 of the 10,000 members revealed that they were far more like the average U.S. population than people assumed. Aside from some unusual views on sex, the group was rather normal.
If potential religions are constantly being created, what draws people to stick to and sustain a few of them? Several theories have emerged to explain what people get in return for their religious adherence. Bainbridge favors social exchange theory, an economic view that assumes that people choose to take part in religions because they believe the benefits outweigh the costs. (Benefits that are intangible or questionable, such as life after death, still have value for believers.) To test the idea, Bainbridge created computer models of the spread and evolution of religions. "The major lesson learned is that religions certainly have mechanistic explanations behind their origins and growth. We can capture their patterns on computers." But "religious people do not like to hear that."
Like most scientists, Bainbridge bopped between U.S. universities for nearly 2 decades before settling on a permanent home. Fresh out of graduate school, he overlapped with Stark at the University of Washington, Seattle. He made a strong impression. "It was clear from the start that there's a lot of ham in the man," says Stark. Bainbridge's sociology lectures became legendary due to his theatrics, showing up in the robes of a satanic cult, for example, or chopping off the head of a fake rabbit to illustrate a point about the mind-body problem. "He's quite a performer."
Bainbridge turned down offers of tenured positions at three universities before deciding to roost at NSF in 1992. After running NSF's sociology program, he moved to the computer science directorate in 2000 to work in artificial intelligence and virtual worlds.
Overseeing research funding keeps him plugged deeply into the field, but it's "very different from an academic career," he says. "Scientists in most fields really cannot continue their research careers if they are government employees." Luckily, the nature of his research--requiring neither laboratories nor big funding--has allowed Bainbridge to continue his investigation of culture. He is doing that mostly on the Internet.
In one of his most recent experiments, Bainbridge hosted the first scientific conference within Azeroth , the online virtual world inhabited by the millions of people who play World of Warcraft (see Science, 20 June  ; this author was a co-organizer of the event). Over the past year, Bainbridge has conducted more than 2000 hours of ethnographic research in Azeroth, studying the attributes and behaviors of more than 4000 virtual characters. An equivalent study in the real world would require a large team and millions of dollars, but in this virtual world Bainbridge can run the study alone and for free. The goal is to test people's altruistic behaviors online and the social hierarchies that emerge among them.
True to his nature, Bainbridge continues to explore topics further from the mainstream than most other scientists are comfortable with. For example, he is spearheading a multidisciplinary effort to achieve the science-fiction dream of uploading human personalities in digital form. "Our bodies are too frail for space travel," says Bainbridge, "so this will be the only way we can truly visit other stars." Aside from speculative articles, he is using the tools of psychology to try to better define what personality is. With his study of players in Azeroth, he hopes to get hard measurements of how flexible our identities really are. "This brings my interests full circle," Bainbridge says. "I foresee a future battle for survival between religion and science-based technologies of immortality."
But Bainbridge doesn't expect to make it to that future. "I plan to retire feet first," he quips. For young researchers contemplating a career outside the mainstream, Bainbridge stops short of advocacy. "There is a fundamental dilemma for creative people: Do what other people already admire and you are competing with the whole world; do something novel and you may find no audience for your accomplishments." It's a risk either way.
Photo (top): Guido Alvarez