Rooster Tales
By Phil Hansten | November 13, 2012
What we can’t think about: We humans are lousy in dealing with issues of cause and effect.
It’s 5:00 AM and he flutters up onto the roof of his pen. Then he lets out a magnificent “cock-a-doodle-do”… and then another, and another. Minutes later the first rays of the sun stream over the top of a nearby hill. He has done it again! Every single morning as far back as he can remember his mighty crowing has caused the sun to rise!
This is “the rooster taking credit for the dawn”—the classic example of the “post hoc fallacy.” As many of you know, the phrase post hoc, ergo propter hoc means “after this, therefore because of this.” It is often abbreviated to post hoc and in my view it is one of the most pervasive and problematic thinking errors of humankind. This fallacy contaminates discourse in science, economics, education, business, sports, politics, as well as our casual conversations; nobody is immune.
Given our evolutionary history, perhaps it was inevitable that we should suffer from this error. After all, when our cave-dwelling ancestors saw a guy eat some nice red berries and moments later clutch his throat and collapse, well… it made sense to avoid eating those berries. This is post hoc reasoning of a rational kind.
In science, of course, we try to avoid the post hoc fallacy. We study drug efficacy using double-blind placebo-controlled studies, and in epidemiological studies we try to control for every possible factor that may interfere with detection of a true causal relationship. We are not always successful, however. Case reports published in the medical literature notoriously suffer from post hoc fallacies. The “blood thinner” warfarin, for example, is affected by all sorts of dietary, pharmaceutical and other influences, so over time the blood thinning effect may fluctuate. But since people on warfarin start new drugs periodically, just by chance the new drug is sometimes started right before one of these fluctuations. Naturally, the new drug gets blamed, even though it may have had nothing to do with it. This is the post hoc fallacy, and there is much nonsense in the drug interaction literature as a result.
Social scientists need to be especially vigilant to minimize the post hoc fallacy. Several years ago a study found that couples who lived together before marriage had a greater likelihood of divorcing than couples who did not. They concluded that couples should not cohabitate before marriage. I wondered, however, if they considered the very real possibility that people who lived together before marriage may have substantial differences (other than their living arrangements) from those who chose not to live together. It seems almost impossible to avoid post hoc in this case; couples might object to being randomly forced to live together or not for the sake of research!
It is easy to fall prey to the post hoc fallacy. Most of us are no better than the rooster, especially if we are not overly given to reflective thought. Human nature may foil our efforts to untether completely from post hoc thinking, but we must try nonetheless!