Advocates of so-called alternative medicine (i.e. unproven treatments) don’t always use the best critical thinking skills. False assumptions and logical fallacies abound. Three that I encounter constantly are (1) “ancient wisdom,” (2) “it works for me,” and (3) “it’s natural.”
The argument usually goes something like this: Our wise ancestors used X to treat condition Y. They understood the natural world and had access to knowledge that we no longer have. People wouldn’t still be using X if it didn’t work. X must be safe or we would have found out about the dangers by now.
Wrong. Our ancestors weren’t any more intelligent than we are, and they didn’t know about the scientific method. They had to go by guesswork, trial and error. Sometimes they got it right and sometimes they got it terribly wrong. They got it right when they noticed that willow bark relieved pain: that led to the discovery of aspirin. They got it wrong when they invented bloodletting.
Back before the dawn of history, someone observed that sometimes people got better after they lost blood (this is still a treatment for iron overload in hemochromatosis and it certainly lowers blood pressure). They developed a whole philosophy around the idea of taking blood to help “balance” the four humours and it lasted for around 2 millennia. During that time, there were plenty of patients and doctors who “knew” it worked and could provide testimonials. Bloodletting killed George Washington, and it wasn’t until quite recently that we realized it did more harm than good.
If there is ancient wisdom, there are also ancient superstitions and old wives’ tales. All of them are tenacious.
People wouldn’t still be using it if it didn’t work? Yes they would. They’re still using astrology, and it doesn’t work. They’re still dowsing and it doesn’t work.
Herbalists have used the same plants for centuries. If they have never seen any side effects or dangers from a plant, can’t we assume there probably aren’t any? That might seem logical until you really think about it.
You may remember the Fen-phen scandal of 1997. This combination treatment for obesity had been used for about 5 years. Many thousands of patients took it. Eventually a small number of women were found to have valvular heart disease connected with Fen-phen, and it was taken off the market.
Just to pick a number, let’s say one woman out of 100 who took an herb developed a fatal side effect. The average herbalist might prescribe the same herb to 100 patients over a long period of time, perhaps over his entire lifetime. If one patient developed a fatal illness, why would he have any reason to suspect the herb did it? Probably during the same period 2 or 3 other patients out of the 100 would have died of unrelated causes. He didn’t keep records. He didn’t compare his statistics with those of his colleagues in other villages. It would be nothing short of miraculous if he did make a connection with the herb. It was hard enough for cardiologists in the 1990s to recognize the problem with Fen-phen because it was so totally unexpected – there was no reason to think any drug could harm heart valves. No one doctor saw enough cases to make the connection; it was a technician chatting with the patients of different doctors who first noticed it. Ancient herbalists didn’t have the tools and communications we have today.
Ancient wisdom may actually be ancient folly, and the only way to tell the difference is with scientific testing. The treatments that claim “ancient wisdom” have not been tested adequately, or they wouldn’t be claiming ancient wisdom, they’d be citing scientific studies.
Natural has only good connotations. It’s natural to think natural is good. But it’s a false assumption. When you really stop to think, is there any reason a remedy should be better just because it’s natural? If you have a religious belief that God created the perfect natural remedy for every ailment, I suppose that belief is not open to argument. But absent any such beliefs, the evidence shows that natural is quite often objectionable.
There are plenty of natural poisons, from hemlock to strychnine. Plants have to protect themselves from predators, and they have invented quite an array of chemical defenses. We go organic out of concern for artificial pesticide residues in our food, but natural pesticides make up 99.99% of the pesticides in our diet. Plants have no intention of helping us. If there’s a chemical in a plant that’s medically useful to us it’s not natural: it’s unnaturally good luck.
If a plant turns out to be a chemical factory for something we can use, it’s a rather unreliable manufacturer. The active ingredient may be plentiful if we harvest the plant in April and practically absent in July. It varies with factors like soil and weather. And the process we use to prepare it and choose the dose adds another layer of uncertainty.
When foxglove (digitalis) was first used for heart failure, it was fiendishly difficult to achieve a high enough dose to help the patient without overdosing and killing him. Scientists began by standardizing extracts, then they isolated the active ingredients, then they synthesized an artificial form of digitalis. Today we have a safe, standardized product with a controllable dose, a blood test to measure how much is in the bloodstream, and an antibody antidote that can instantly reverse an overdose. It’s totally artificial, but only an idiot would go back to using natural digitalis leaf today.
When a useful medicine comes to us from a plant, it brings its friends. It comes mixed in with all kinds of other chemicals we don’t need, chemicals that may be harmful or may diminish the effect of the medicine. Herbalists try to tell us there is some kind of synergistic effect where the combination of ingredients in the natural plant interacts for our benefit. This may sometimes occur, but very rarely. The evidence usually shows just the opposite. When an isolated active ingredient has been tested against the whole plant extract, there has usually been no difference in effectiveness. Any of the other, unnecessary ingredients are more likely to hurt than help.
It Works for Me
This is the one I hear when I tell someone there is convincing evidence from multiple randomized, placebo-controlled trials showing that something doesn’t work. They don’t care what the science says, because they know it works for them.
They know no such thing.
What they mean is that they tried it, and their symptoms went away, perhaps repeatedly. What they mean is that they have observed a correlation between using the treatment and getting relief from their symptoms. What they mean is that they have jumped to the unwarranted conclusion that there is cause and effect where there may be only correlation. They may “believe” but they don’t “know.”
My friend Dieter tried vitamin E oil for his toenail fungus and the fungus went away. He said it worked. I said he didn’t know that. We had a long discussion. Part of our disagreement was in the meaning of “worked.” There is a practical, everyday trial and error sense in which it did work. Dieter would be fully justified in using vitamin E oil again if the fungus returned. I’d be the last to dissuade him; I’d probably do the same thing. If it “worked” a second time, I’d be just that more likely to try it a 3rd time.
But there’s a big difference between deciding what practical action to take and stating a universal truth. I would say “My fungus went away when I used vitamin E oil, so I think I’ll try it again.” I wouldn’t presume to know that “Vitamin E oil cures toenail fungus.”
The logical fallacy here is post hoc ergo propter hoc. That’s Latin for A was followed by B, therefore A caused B. It’s probably the most common human error. Our brains are designed to find patterns and detect cause and effect, and we’re so good at it we get carried away. That’s the root of all superstition: I wore my lucky hat and my team won, so I’ll wear it every time to help my team win.
The rooster crows every morning and then the sun comes up, but that doesn’t mean his crowing raised the sun. We can see how that doesn’t follow; but when you take a pill and your pain goes away, suddenly it seems compellingly logical. We must remember that the pill may be just another rooster.
It all boils down to remembering to ask if there could be any other possible explanation besides the one you are thinking of. Remember the man who taught a flea to dance when it heard music? He cut off the flea’s legs, the flea stopped dancing, and he concluded the flea’s hearing organs must be in its legs.
Maybe Dieter’s fungus would have gone away with no treatment at all. Maybe he got some kind of placebo response. Maybe he started eating a new brand of spaghetti sauce that contained a fungus-fighting herb. Maybe he didn’t really have a fungus. We need to think of other possibilities and rule them out with good scientific studies before we can claim “it works.”
Many natural remedies are sanctioned by centuries of use and seem to work, but they haven’t passed scientific tests. Those remedies may or may not work; these three fallacious arguments definitely don’t.
This was originally published as a SkepDoc column in Skeptic magazine.