Here is a list of beliefs in the biomedical and climate sciences that must not be questioned if you're applying for a government grant:

- That global warming is caused by humans;

- That AIDS is caused by a virus;

- That radiation, cigarette smoke and other toxins are dangerous in proportion to their strength, no matter how small the dose;

- That heart disease is caused by saturated fats;

- That cancer is caused by mutations.

This is part of a list offered by a University of Washington professor of surgery, Donald W. Miller, who is a heart surgeon at the VA Medical Center in Seattle. Miller believes that all the above ideas may be false, and ought to be tested. Whether they are false, I don't know. I have thought they were true, but that is only a belief - and it is the business of science to test such beliefs.

But much of science runs on government money. Some people find the stink of bias only in private money, and see government as free of it, but they are mistaken. Government likes certain beliefs. To get its money, you have to get the approval of the scientists it selects, and you are less likely to get it if they think your idea wrong.

What that means, Miller says, is that "If you say low doses of radiation aren't bad for you, or that global warming is due to variations in the sun, you can't get funded."

He says this happened to University of California scientist Peter Dues-berg, who challenged the viral theory of AIDS, and to Harvard's Willie Soon, who challenged the pollution theory of global warming, and to others. In a paper published in 2007 in the Journal of Information Ethics, Miller argued that conformity is built into the system of government grants.

Another critic of the grants system is Gerald Pollack, UW professor of bioengineering. Pollack's work in muscle contraction, cell structure and the molecular properties of water has challenged the reigning view in his field.

In 2005, in the scientific journal Cellular and Molecular Biology, Pollack made an argument similar to Miller's. American science, he wrote, has become "a culture of believers" whose rule is, "just keep it safe and get your funding."

For science, the result has not been good.

"A half-century ago, breakthroughs were fairly common events in science," Pollack said in an interview. But who today are the equivalents of Linus Pauling in molecular biology, Jonas Salk in vaccines, Richard Feynman in physics, or James Watson and Francis Crick in the study of DNA? Said Pollack, "Where are the heroes of the past 30 years?"

In his paper, Pollack wrote, "Einstein's challenge of orthodoxy would probably fail in today's grant system." Today's committees of scientists demand that an individual predict what he will accomplish at the end of year one, year two, etc., all of which amounts, Pollack says, to "an implicit admission that no breakthroughs are to be anticipated."

If science is likened to a skeleton, the grant system sets out to pay a multitude of scientists each to add a tiny bit of flesh. But what if the skeleton itself is misdesigned?

"I think a lot of the skeleton is erroneous," Pollack says.

Thomas Kuhn, the philosopher of science, argued famously that science progresses in revolutionary bursts, in which the "dominant paradigm" is overturned. But what if the supporters of the dominant paradigm are the people vetting your application?

The grant system needs to be changed, Pollack says. Short of that, there are ways around it. One, he says, is "to get the money for something else, and do your work on the side."

Miller predicts that at some point, a major belief like one of those listed above will come tumbling down. "And when it's acknowledged," he predicts, "a lot of other science will be called into question."