If your career depended on it, would you have the courage and confidence to attach your name to your beliefs?

Most of the world's most brilliant and innovative scientists in physics, chemistry, molecular biology and medicine do not.

It's not that these scientists are gutless. It's that the cynical wardrobe of scientific "tradition" allows them to cloak their opinions in masks of anonymity. When a scientist submits a research paper to a prestigious journal for publication, that research paper is usually given to another scientist to critique. That's called "peer review."

These critiques frequently determine how the paper should be rewritten and, indeed, if the paper should be published at all. Almost overwhelmingly, these "referees" prefer to remain anonymous. Their comments could be helpful, insightful or insulting -- but the researcher may never know who wrote them.

Why? Because journal editors equate anonymity with honesty. They fear that reviewers would censor themselves if they knew their identities would be revealed. Named reviews, they argue, would thus undermine the quality of research.

"It would absolutely collapse the whole system," asserts Daniel Koshland, editor of Science, America's most important research publication. "People would not be as honest as they should be."

"I'm in favor of anonymous referees," agrees Arnold Relman, who is retiring as editor of the influential New England Journal of Medicine, "because I don't believe that known referees can produce fair opinions when the authors are people who know them and have relationships that are influential or otherwise. It defies human nature to expect that a junior researcher will say exactly what he thinks about a senior person with power. It just won't happen."

"I believe the reviews we get from anonymous reviewers are vastly better," says George D. Lundberg, editor of the Journal of the American Medical Association.

It's easier to criticize anonymously -- but is it really better? Often, knowing the source of a criticism can be as valuable as the criticism itself. Precisely because science is a very human endeavor, researchers should know who critiqued their work. It's ironic that science -- which prides itself on the free flow of knowledge and moral courage of a Galileo and a Sakharov -- feels that its quality control mechanism works best in the dark.

Would science sputter with uncertainty and grind to a halt if scientists actually signed their work? Or might it just be possible that anonymity is a convenience that editors and reviewers are reluctant to let go? Given that the role of a referee is to challenge the hypotheses and methodologies -- not the character -- of the researcher, isn't it just possible that signed reviews would be more likely to change the tone -- and not the content -- of the critique?

"There are occasions where anonymity allows a kind of vituperativeness that is unfair," acknowledges Edward Huth, the recently retired editor of Annals of Internal Medicine. "I think Koshland and Relman are being a little extreme. ... Most of us work on some premises of peer review that really haven't been adequately tested. This policy, which is very much entrenched in journal reviews, is based on opinions rather than any evidence."

Indeed, the evidence that does exist seems to suggest that scientists are not quite as craven about their critiques as some people think. Lundberg's JAMA published a research paper on peer review earlier this year indicating that "there was no association between signing and quality" of review.

In fact, the whole question of peer review has been drawing more scrutiny from journal editors recently. Next month, for example, there will be a meeting of journal editors sponsored by the National Library of Medicine in Bethesda to set up an agenda on how best to explore some of the critical issues surrounding peer review.

Robert H. and Suzanne W. Fletcher, who have succeeded Huth at the Annals of Internal Medicine, plan to use their journal to experiment more aggressively with signed reviews.

"I think there's a fair chance it would run better than Koshland and Relman might think," asserts Robert Fletcher. "On the other hand, I'm not sure about that. ... I guess we come out of a more pragmatic tradition: Would we get better reviews? I think it ought to be tested. The concern for principles is secondary."

My concern for principles is primary. I think it's bad not to attach your name to your critique. It would be professionally unacceptable for a theater critic or a movie reviewer or a columnist to shield their names. And who doesn't become a bit warier of a story that quotes an anonymous "source"?

Yes, scientists are human with all the weaknesses that implies. But it also implies strength -- and one of those strengths is the willingness to have the courage of one's convictions and expertise. Indeed, it should be the editor's role to assure that the reviewer's criticisms are constructive. The very existence of anonymity makes people feel that they are gravely at risk if they choose to be both open and honest.

Peer review should not be a Star Chamber. On the other hand, people shouldn't be compelled to disclose who they are if they don't want to. I do have a suggestion, though: Reward openness. Give funding preference to scientists who have the courage to attach their names to their work.

Science should literally put its money where its mouth is and support the values of an open, energetic and honest exchange of ideas. Those exchanges are where great breakthroughs come from. You'll never have them if some of the most important participants are wearing masks.

Michael Schrage is a columnist for the Los Angeles Times.