Further research is necessary

An updated version of this post is available at Lay Science.

I found a moment to read Edzard Ernst and Simon Singh's Trick Or Treatment. Ernst is professor of complementary and alternative medicine at Exeter University, and he applies the methods of science and evidence-based medicine to so-called alternative therapies. In this volume, Ernst and science-writer Singh review the evidence base -- or lack of -- for some popular (UK-centric) alternative medicines. It's what you would expect from such an undertaking: a series of chapters along the lines of "[ homeopathy | acupuncture | chiropractic ] (delete as applicable) is supposed to work by ................... (insert magical and demonstrably false hypothesis about how biology, chemistry, or physics is supposed to work). It doesn't." This is followed by a chapter on why CAM is unethical , including for example, why CAM is usually a scam based on arbitrary and discredited unscientific foundations, and why real doctors are wary of lying in order to take advantage of the placebo effect.

One chapter was different: a collection of notes on "herbal medicines". This chapter covered a whole class of treatments, and it concluded that some of them have some value, most of them have no value, and some of them are highly poisonous. This is hardly surprising: many (most?) of our pharmaceuticals are based on (that is, safer and more effective derivatives of) nature's designs: the chemicals produced by plants, and other branches of life. There's still a lot of magical, irrational, and plain wrong thinking associated with herbal remedies, but that doesn't mean that believers can't chance upon a correct idea once in a while. Stopped clock, twice a day, etc.

The general conclusion is that, other than a few herbal remedies, CAM treatments are useless, based on laughable pseudosciences, and are sometimes dangerous. Either they have been demonstrated to have no effect, or studies have failed to show any significant effect. A recurring theme is that studies and trials in CAM tend to be performed incompetently by people who either don't know how to perform research properly, or who know that if they did it right, they wouldn't get the answer they want. (The authors are too polite to put it quite so bluntly.) So since CAM is so widely used, some of it may be useful, and there may be hidden (and not so hidden) dangers, further research is necessary to make up for the shortcomings of the current body of literature. (Of course, academics have to say that about their fields as often as possible these days.)

I don't buy it. Why is further research necessary? There are a select few cases of herbal remedies where further research might not be a waste of time -- you can tell by the fact that, in addition to public institutions , pharmaceutical companies are doing that research. But for everything else, this is just money down the drain -- money that could be spent finding real ways to save and improve lives. Homeopathy, chiropractic, and acupuncture not only fail under the inevitable stacks of negative evidence, they, like most of CAM, suffer the provenance flaw. The supposed mechanisms of these "remedies" were just conjured from the air; and they often contradict what we already know about how the human body, chemicals, and the physical world work. There is simply no reason to think that such absurd fiction might turn out to be true.

When something has a provenance flaw, you can't just say "further research is necessary." I could invent an unlimited number of baseless hypotheses -- in the same way that homeopathy, acupuncture, and chiropractic were conjured -- and, surprise, further research would find only that none of them have any merit. Picking your nose gives you cancer, chocolate digestives cure Alzheimer's, my dad invented space travel, Mornington Crescent Station is really a top-secret government intelligence headquarters, and Stephen Fry is a time-travelling robot controlled by aliens. Further research on these topics is necessary; cheques payable to Joe D, please. No? How about if I first get a few tens of thousands of people to believe in my bullshit?

This is why the new US administration should defund NCCAM.