MSNBC.com

July 2-9, 2007 issue - Humans have 23 pairs of chromosomes containing about 20,000 genes, DNA is the molecule that carries hereditary information in every living cell, matter is made of atoms that are built of protons and neutrons and electrons and ... Alan Leshner isn't buying it. CEO of the American Association for the Advancement of Science, which publishes the journal Science and promotes science literacy, he agrees that people "need, at minimum, a rough understanding of the core concepts of science—the more the better." That would keep people from rejecting genetically modified food because, as they tell pollsters, it "contains genes" (all living cells do).

The real problem today, however, is not ignorance of the fact that Earth revolves around the sun once a year (something 25 percent of adult Americans do not know). "It's that people don't understand what is and isn't science," says Leshner.

Science observes and measures the natural world. From those data it infers the empirical laws that govern physical and biological proc-esses. Explanations of large classes of phenomena must make testable predictions and be falsifiable. That is, there must be a way to make an observation that could disprove the explanation. (Scientists call that overarching explanation a theory; the term does not mean, as it can in everyday parlance, somebody's off-the-cuff guess.) The requirement of falsifiability rules out supernatural explanations; you cannot disprove, for instance, the claim that God scattered fossils throughout rock strata to make it look as if species had evolved over millions of years. God may have done that, but we'll never know and there is no way to disprove it. In that way, faith is fundamentally different from science.

Despite the face it sometimes presents to the world, science is humble, recognizing that all findings are tentative (although in many fields the weight of evidence would be pretty tough to overturn) and only as good as the next experiment. It labors to distinguish true effects from random chance. Experiments have "control" groups to make sure that an effect thought to come from, say, taking a new drug does not also show up in people who did not take the drug.

Good science distinguishes correlation from causation. If kids who play violent videogames commit more violence, before you blame the game you'd better be sure that violence-prone kids are not more drawn to violent games than other kids. If so, then violent behavior causes the playing of violent videogames, and not the other way around.

—Sharon Begley