The Post Hoc Fallacy: Vaccine Fears and Roosevelt-Menacing Indians

April 6, 2015

I recently read Candice Miller’s book The River of Doubt, about Theodore Roosevelt’s 1914 exploration of an unknown river in the Brazilian Amazon. It’s a fascinating story of adventure, misadventure, murder, and more. In the book I also found an excellent real-life example of one of my favorite logical fallacies: post hoc ergo propter hoc, also called faulty causation.

I’ve written about examples of this very common fallacy many times, often in the context of vaccine fears: Many parents came to believe that vaccines caused their children’s autism because the symptoms of autism appeared after the child received a vaccination. On a psychological level, that assumption and connection makes sense; but on a logical level, it is a clear and common fallacy with a fancy Latin name: post hoc ergo propter hoc (“after this, therefore because of it”).

Because the human mind seeks connections, people often misattribute causes, thinking that, “B happened after A did, so A must have caused B.” The child was fine until he was vaccinated, and soon he showed signs of autism. It makes sense-except that it’s not necessarily true. It’s like saying “roosters crow before the sun rises, so the roosters must have made the sun rise.”

First-year students of statistics are constantly reminded that “correlation does not imply causation.” That is, two events that appear to be causally related may not be; there are other possibilities. The appearance of causation may simply be coincidence; or A may have caused B, or B may have caused A; or there may be a third unknown factor relating to A and/or B. Only carefully controlled scientific studies can conclusively tell the difference. As it happens, autism often first expresses itself in children at about the same time that vaccinations are recommended for those children.

But I thought I’d share this excerpt from The River of Doubt because it’s not only dramatic but shows the life-or-death consequences of magical thinking. It’s a related by French anthropologist Claude Levi-Strauss in the 1930s, who witnessed a massacre in rural Brazil rooted in this fallacy:

“A Nambikwara Indian with a high temperature presented himself [at a Protestant mission in Juruena] and was publicly given two aspirin tablets, which he swallowed; afterwards he bathed in the river, developed congestion of the lungs and died. As the Nambikwara were excellent poisoners, they concluded that their fellow tribesman had been murdered; they launched a retaliatory attack, during which six members of the mission were massacred, including a two-year-old child. Only one woman was found alive by a search party sent out from Cuiaba.” (p. 117-118).

Of course the mission had not poisoned the man with aspirin, but since he was seen taking the medicine before he died, the Indians assumed he must have caused his death. Roosevelt and his men encountered this same group.