SCIENTISTS WORSHIP AT THE shrine of objectivity, but even the pious occasionally lapse. A century ago archeologists who discovered the great stone ruins of Zimbabwe went through all sorts of contortions to prove that the magnificent oval palace and other structures were built by the Phoenicians of King Solomon's time--or by anyone other than the ancestors of the Bantus. In the 1960s biologists studying conception described the "whip-lashlike motion and strong lurches" of sperm "delivering" genes required to "activate the developmental program of the egg," which "drifted" along passively. The model portrayed sperm as macho adventurers, eggs as coy damsels. And throughout the 1970s and later, ornithologists gathered sheafs of data proving that, in birds, a female's success laying eggs and rearing hatchlings was always enhanced by the presence of a male.
These acolytes of scientific objectivity were spectacularly wrong. The Bantus' ancestors did build the great stone complex. The human egg does play an active role in conception. And in some bird species, particularly the eastern bluebird, the father's presence makes little or no difference to the survival of hatchlings. But why did scientists get it wrong in all three cases, and many others? That question lies at the heart of the "science wars." The combatants are, on the one side, bench scientists who study the biological and physical world and, on the other, sociologists and others who study scientists as if they were exotic Borneo tribesmen. Their battlegrounds are scholarly journals and books where the two sides attack each other. And the issue they're fighting over goes to the heart of the scientific enterprise: is science an objective pursuit?
The critics of science say that the practice of science--the questions it asks, the way it interprets observations, even what counts as data-is subject to the political, cultural and social influences of the times. If society considers females passive, say the critics, then scientists will tend to see the same characteristics in the egg. And if social values mean that an intact nuclear family is best for kids, then most scientists look for, find or give more credence to examples of birds or other species where that holds true. It is not that evil scientists intentionally set out to enshrine the prejudices of the day in their research conclusions. But as mere mortals, they cannot escape their influence. Science, say its critics, is therefore a "social construct," and its discoveries and conclusions have no special claim on truth.
To many scientists, those are fighting words. But they have been slow to react. The early criticism came from academic fields like women's studies and literary criticism that few scientists pay attention to, so they didn't even realize they were under attack in the 1980s. When they finally woke up to the bombardment, they found the criticism so patently stupid-gravity is an opinion, a social construct, and not a fact?-they ignored it, figuring that everyone else would, too. But everyone else didn't. "The science critics began to have an influence among undergraduates and on the curriculum," says primatologist Sarah Blaffer Hrdy of the University of California, Davis. So scientists launched a counterattack. The most notorious counterpunch came last summer. Physicist Alan Sokal of New York University wrote a spoof of the "constructivist" argument, passed it off as the real thing and duped the post-modern journal Social Text into publishing it. The editors' gullibility, Sokal argued, revealed the bankruptcy of the constructivists' ideas. Now some of the more extreme science-bashers have modified their views: the feminist critic who called Isaac Newton's Principia "a rape manual" now regrets her choice of words. Of more than 20 science critics interviewed for this article, every one offered a version of the disclaimer "Science is still the best game in town when it comes to producing knowledge," as philosopher Ann Cudd of the University of Kansas put it.
But a much more important change is also underway. As the extremists on both sides are frozen out, more thoughtful criticisms of science are winning over even some physicists, astronomers and biologists. One of the world's leading journals of science, Nature, editorialized recently that scholars who describe science as a social construction should "not be dismissed," since "fashionable ideas" do affect what "becomes accepted as scientific truth." At Princeton University, an informal group of lab scientists and the scholars who study them is meeting once a month to talk about how science can be skewed by ideology and how to make it more objective. They call themselves Reality Check. And last month at a conference on "Science and Its Critics" at the University of Kansas, working scientists presented compelling examples of how science got wrong answers when social and political values influenced the work. Says physicist Kurt Gottfried of Cornell University, who co-authored a paper on the science wars in the current issue of Nature, "Cultural and other extraneous factors are more important in the creation of science than most people realize."
Most often, such factors are brought to bear on emotionally or ideologically charged subjects, like sex or race. Countless studies have "proved" the intellectual inferiority of women, blacks or immigrants. But values have also skewed other sciences:
Hubble trouble: What could be more removed from politics than the expansion of the universe? But even this subject, as astrophysicist Keith Ashman of the University of Kansas recounted at the science-wars conference, was distorted by subjective factors, in particular loyalty and careerism. Throughout the 1970s and 1980s, two groups of researchers, one in Texas and one in California, consistently found wildly different values for the rate at which the universe is expanding, which is called the Hubble constant. Texas got 100, California got 50 (don't worry what the numbers mean). Each group became set in its view of how to measure distances to galaxies and stars, and how to measure the speed of a receding galaxy. (There is no simple or agreed-on formula for either measurement.) Both are used to calculate the Hubble constant. "These were highly technical issues that outsiders had a hard time judging," says Ashman. "So for 20 years the community was far too influenced by the reputation of these people, and that hindered attempts to find a consensus figure for the Hubble constant." Depending on who a cosmologist's friends were and on whom she or he studied under, the scientist aligned with one camp or the other. The few lone voices suggesting that the correct value might lie between 50 and 100 were ignored. The right value, as determined by the orbiting Hubble Space Telescope, is indeed around 75. With more objectivity, astronomers might have learned that sooner, says Ashman.
A monkey's uncle: About 40 years ago, primatologists began studying savanna baboons in the hope that they might shed light on how human ancestors lived. Baboons form some of the most male-dominated troops of any primate. They also form shifting alliances with other males to get better access to females. The males' behavior was seen as the social glue that held the troop together; the females were widely interpreted as being superfluous to the troop's cohesion. Even though another researcher, as early as the 1950s, had studied gibbons and other primates where the males were less aggressive, it was the baboon studies that "became an instant paradigm of how to understand primate behavior as well as a model for human society," says primatologist Hrdy. "It fit people's conceptions of male and female behavior. The baboons were supposedly telling us what our hunter-gatherer ancestors were like," and therefore how modern humans were naturally inclined to behave.
But then the social winds shifted. Primatologists, influenced by feminism, realized that they have to study every individual in a troop-not just the biggest, strongest or most conspicuous-to understand primate society. They began amassing examples of apes where females rather than males select mates, females as well as males maintain social cohesion, and cooperation rather than aggression characterizes the life of the troop. The behavior of animals other than the alpha males was suddenly deemed worthy of study, and researchers corrected their one-sided view of how primate societies work. How had they misstepped in the first place? "In nature, things do not come with labels saying 'This is data'," says Kansas's Cudd. "In deciding what is data" -whether it's the baboon pounding his chest or also the female grooming her mother- "scientists impose their values. That's what we mean by the social construction of theories."
Scrambled eggs: For decades biologists described sperm as the more active participant in conception. The egg merely drifted and waited; the sperm had "velocity" and "penetrated." But in the mid-1980s biologists measured the force a sperm exerts while swimming. It turned out that sperm are ineffectual swimmers: they flounder around, meander sideways and expend 10 times more power on turning the head of the cell side to side than on moving forward. The sperm was as likely to scoot away from the egg as to reach it. Biologists had hardly considered the possibility that the egg actively grabbed the sperm, but it does. It's not that researchers had no clue that the egg contributes to conception-in 1964 they knew that genetic material in the egg guides the fertilized cell during the hours right after conception. But research indicating an active role for the egg "just sat there," says anthropologist Emily Martin of Princeton University. "No one knew what to do with it."
Peacock's tail: If you're a peacock and your father had the fanciest tail in the jungle, will you inherit his feathered finery? To answer that, evolutionary biologists followed bird flocks for generations and measured how reliably males inherit their father's ornamentation. Their consensus: ornamentation has a "heritability" of $7 percent. (A value of 100 percent means a trait is always inherited, as eye color is; a value of 0 means it never is.) But after 1988, the heritabilities published in science journals suddenly shot up, to an average of 67 percent, finds Rauno Alatalo of the University of Jyvaskyla. What happened? The birds hadn't changed. Fancy tails and other decorations didn't suddenly become more heritable. What did change, says Alatalo, is that a new theory of why females choose the fanciest birds as mates suddenly became fashionable. The new theory holds that ornaments are signs of a male's strength and fitness, qualities he would pass on to his offspring. If this theory is right, the heritability of ornaments would be high. The popularity of the new theory, says Alatalo, skewed the data. He suspects that biologists, enamored of the new theory, were more likely to believe, and publish, findings of high heritabilities.
At one level, the science wars might seem to be an intellectual brawl holding no interest for anyone outside the precious environs of academia. But society is driven by science. The conclusions of researchers answer questions as important to ordinary people as the heritability of alcoholism and the merits of using IQ scores to assign 5-year-olds to "gifted" school programs. Science, in short, matters. It matters if its conclusions are no sturdier than the latest intellectual fashion. If they are not--if the truths of science depend on who's doing the research-then "science forfeits its credibility," warns archeologist Bruce Trigger of McGill University.
That is already a danger, as shown in the ways lawyers and policymakers use and abuse science. If a senator wants a scientist to testify that lead poses little threat to children's intelligence, he can find one. If a lawyer wants an expert to testify that an endangered species isn't, she can find one. Even worse are cases where the actual science that is done-not what laymen do with the science-is pummeled by politics. That happened to epidemiologist Janet Daling of the Fred Hutchinson Cancer Research Center in Seattle. She was struck by findings that early pregnancy releases hormones that might increase the risk of breast cancer, and late pregnancy changes the breasts in ways that may protect against cancer. Might abortion, then, increase the risk of breast cancer? In 1994 Daling published, in the Journal of the National Cancer Institute, a study of 1,806 women indicating that it did. JNCI was evidently uncomfortable with this conclusion: it ran an editorial pointing out several ways the study might have erred. In contrast, a study published this January in The New England Journal of Medicine found no additional risk of breast cancer after abortion. The editorial about it implied that the question had been resolved forever: "A woman need not worry about the risk of breast cancer" rising after an abortion. It did not point out that this study, no less than Daling's, might have flaws. "I'm not saying I'm sure our studies are right," says Daling. "But let's be objective and not let emotional issues into it."
In some cases, then, values rather than facts inflect research. Nevertheless, science differs from, say, literary criticism, in a crucial way. First, it has built-in mechanisms to catch and correct the results of human foibles. After all, researchers did finally recognize that molecules in the egg actively participate in conception, and that baboons are probably not the only model for the behavior of human ancestors. "So yes, scientists are influenced by factors we shouldn't be," says Kansas astrophysicist Ashman. "But despite all this, we are still measuring something real, something that is not a social construct, and with better data the scientific method allowed us to converge on the right answer." Or, as McGill's Trigger puts it, "the constraints of the data are what make the difference between writing a novel about the past and doing archeology." The real trick, for scientists and for those who base public policy on their work, is to tell when the research is still being skewed by social and political values, and when those biases have been recognized and neutralized by the scientific method. That's a war worth fighting.