Pierre Laplace didn't do science any favors when he let Napoleon provoke him.

After reading the French mathematician's opus on celestial mechanics -- the movements of planets -- the emperor asked him why the treatise, unlike the work of Isaac Newton, made no mention of God. Laplace reportedly replied, in a huff, that he had no need of that hypothesis.

Ever since, science has been saddled with the canard that it arbitrarily and a priori rules out the existence of a deity. When the Kansas board of education deleted the words "natural explanations" from the definition of science last year, it seemed like an effort to right that supposed wrong. But those who attack science as anti-God are fighting a mirage, say both secular and religious scholars.

"It is a serious error to arbitrarily insert God or the supernatural as explanations for scientific mysteries," says biologist Richard Colling of the evangelical Olivet Nazarene University, Bourbonnais, Ill. "But it is equally unjustified to claim science excludes God."

As Barbara Forrest, a philosopher of science at Southeastern Louisiana University, Hammond, explains, "Science doesn't rule out anything a priori. Saying it does is false, and makes science look dogmatic."

Even to those who have never heard of Laplace, it's easy to get the idea that science starts with an atheistic, or at least agnostic, presumption. A report by the quasigovernmental National Academy of Sciences says science "is limited to explaining the natural world through natural causes." The National Science Teachers Association says science "cannot use supernatural causation in its explanations" and calls supernatural forces "outside its provenance."

Although both definitions make it sound as though science rules out the supernatural from the get-go, what actually happens is that working scientists simply find that entertaining a supernatural explanation doesn't get them very far. In that sense, argues Thomas Clark, director of the Center for Naturalism, a nonprofit educational group in Somerville, Mass., "Science doesn't presume the natural-supernatural distinction; it generates it" by dividing what works from what doesn't.

The supernatural is a dead end because science strives for testable explanations and predictions: The sun will rise in the east because Earth spins west to east, not because "God wanted it that way." Since only the most arrogant would claim the ability to predict what He will do next (and would likely be struck dead for hubris anyway), supernatural explanations fail as science. It isn't that they don't fit science's preconceptions, but that they don't get you anywhere in either deeper understanding or predictive power.

"What science is is settled methodologically," says Forrest. "It's not that science rules out the supernatural as a precondition. But scientists want to apprehend the world, and there is no procedure for studying the supernatural. God is not a controlled variable."

Although science can consider any hypothesis, natural or supernatural, a scientist who entertains the possibility of the supernatural will quickly reach a dead end. Consider the hypothesis, "Angry gods make volcanoes erupt." It doesn't get you anywhere -- not predictively (how do you know when a god will be mad?) and not mechanistically (how does the angry god make lava and gas explode out of the volcano?). Including unspecifiable processes doesn't advance understanding. As a classic Sidney Harris cartoon showed, the explanation "then a miracle occurs" doesn't cut it.

"Unless you specify the agent, its purposes and characteristics, it's an explanatory dodge," says Clark. "Agents have to be described specifically enough to be verified."

That includes specifying when and how a supernatural agent intervenes in nature. If you want to credit the supernatural with designing human beings, for instance, you have to specify why it built in autoimmune diseases, put remnants of old viral DNA in our genes, spliced in repetitive breakage-prone DNA that causes awful diseases, and took away one enzyme in the biochemical pathway that makes vitamin C but left the rest to hang around uselessly. "Working in mysterious ways" falls short. A scientific explanation must account for why one thing happens and another doesn't.

Colling, a lifelong Christian, argues that foregoing supernatural explanations "should not bother religious folks. God is not a micromanager." Explaining wondrous phenomena naturally "expands our comprehension of the created order." None of this is to deny the supernatural, just to say that it doesn't work in science. Students "are being told that they must choose between scientific reality and God," he says. "Nothing could be further from the truth."

If scientists ever bring the supernatural into science by specifying how it works and predicting what it will do next, the result may not be to the liking of those pushing for science to include God. The supernatural "will then generate reliable, predictive knowledge," notes Clark, and become just another explicable, predictable force of nature, stripped of its awe and mystery.