The study certainly gores some oxen. A whole medical industry has sprung up based on the mind’s presumed power over the body. The growing popularity of alternative medicines and treatments–everything from Chinese herbs and yoga to acupuncture and faith healing–has been fueled in part by the medical respectability of the placebo effect.

The higher echelon of the placebo industry, as epitomized by Herbert Benson, is more difficult to dismiss. Benson started out as a Harvard cardiologist but became a convert to the power of mind. He helped found the Mind-Body Institute, which develops techniques to help patients of all types cope with stress. His work has been widely imitated. Cancer patients are now given seminars on how to use meditation to mitigate the anxiety of chemotherapy and promote their own healing. Couples undergoing infertility treatment are taught how to reduce stress, which, they are told, may interfere with the reproductive process. Benson insists that the mind-set of the patient and the physician, as well as the quality of their relationship, is crucial to the effectiveness of many medical treatments. This amounts to a somewhat broadened definition of the placebo effect.

Benson is not the only placebo doctor. Studies of psychotherapy have shown that it doesn’t matter whether your shrink is a Jungian or a Freudian or a Roman Catholic priest: all talk therapies seem to work equally well (or poorly). This sounds suspiciously like a placebo. Does this mean psychotherapy is worthless?

Perhaps, but only if you take the Science paper at face value. At first glance, the authors, two Danish medical researchers, seem to have built a strong case. They collated data from 130 previous studies in which two kinds of control groups were used–a group given a placebo and a group given no treatment. Then they analyzed the data to see if there was any statistical difference between the two groups. There wasn’t.

But a closer reading of the paper reveals some wiggle room. For one thing, the authors lumped together studies that dealt with a wide variety of ailments. That’s not fair, say researchers who study the placebo effect. For anemia from heart-bypass surgery, genital herpes and other ailments included in the Danish study, you wouldn’t expect placebo to do any good. These all have very definite physiological causes that are hard simply to think away. Ailments whose causes are less clear are a different story. Asthma, skin rashes and depression come to mind. The authors, in fact, address this distinction in the body of their report: “There was a beneficial effect of placebo… for trials with subjective outcomes, but not for trials with objective outcomes.” But since the final result is an amalgam of the two types, this fine point was lost.

So what, then, are these subjective outcomes? Pain relief, for one. In the late 1970s Howard Fields, a neurologist at the University of California, San Francisco, and his colleagues studied pain relief in dental patients who had undergone surgery. In patients who reported a lessening of pain after being given a placebo, the researchers were able to counter the placebo effect by administering a drug that blocks the action of opiates. This suggested a physiological effect: the placebo triggered the release of endorphins, natural hormones that inhibit the perception of pain. Numerous studies have since confirmed these results. Researchers have been able to substitute placebo for pain medication often with no lessening of relief. “People will do better if they expect to do better,” says Fields, “and they’ll do worse if they expect to do worse.”

Some researchers are trying to refine their picture of what does and doesn’t respond to placebo. In studies of psychiatric patients, Walter Brown, a psychiatrist at Brown University, has found that those who suffer from short-term depression respond well–in the 50 percent range–to placebo, whereas those who suffer long-term depression do not. Brown’s work suggests that placebo may indeed prove medically efficacious–but only for some patients. His research seeks to refine the conventional wisdom that one in three patients respond to placebo–a number, incidentally, that many researchers regard as a straw man. Henry Beecher arrived at it 46 years ago in a review of previous research of dubious value, and it just stuck. The science of placebo has already moved on.