Most articles and editorials are replete with a common error called the “Post Hoc Fallacy,“ which represents a Latin term meaning, “after this, therefore because of this,” which expresses the often flawed idea that if an event follows another, the first event causes its follower. The classic example of this fallacy is provided by the crowing of a rooster followed by sunrise. But everyone knows that the rooster’s cries do not “cause” the following sunrise. But a hoard of other examples can be just as fallacious but far less clear: For instance “Yesterday I ate a green apple, and today I have a stomach ache.” Questionable conclusion: The apple must have caused this following stomach ache. Or another more common example: “Nearly all heroin addicts used marijuana before they tried heroin. Thus marijuana use must lead to heroin addiction.” This one has been convincingly refuted by massive contrary evidence. But even more common are political examples of this fallacy: President Hoover’s inauguration was followed by the great depression, or, more recently, president Biden’s ascendancy to the presidency was followed by increased inflation. Although highly doubtful, whether any of these single events caused the following “results,” is, at best, unprovable. In virtually all such instances, any blithe expressions of certainty are not only fallacious, but potentially dangerous.
So, how does one “prove” that the antecedent events are causative? In medical science, answers are provided by typical drug trials, wherein a large number of subjects suffering from a given malady are given a trial drug to assess their responses. But this provides only half the answer: A similar number of subjects are given an identically appearing, but inactive (placebo) “treatment,” and by employing sophisticated numerical analyses, the two groups are compared to determine whether the group receiving the active drug fares better (and by how much) than the untreated (control) group. If we lacked such a “control” untreated group, any improvements could be attributed to either a spontaneous recovery or to the placebo (mental?) effect, and in either case, we would be unable to exclude a post hoc, fallacious, conclusion. But some examples cannot provide the direct experimental means just noted, such as the relationship between smoking and lung cancer: In this example, we must amass comparable results reported from many different institutions, all noting a greater frequency of cancer in heavy smokers as compared to light or non-smokers. All this was coupled with data from animal studies and many other sources. All these factors, when taken collectively, allow us to form the sound conclusion that smoking does indeed cause cancer!
But what about the many naturally occurring examples we cannot control, as exemplified by historical or political events? This is far more complicated, and they seldom lend themselves to firm scientific answers as noted by the cited examples above. In this regard, by allowing us to set up contemporary experiments and observations, science has an advantage over history. With regard to historical or political events, scientific inquiry can generally only allow examination of related cases and recognize and compare their common features. In the vast majority of these latter examples, we will never know the definitive answers: For instance, did Chamberlain’s apparent weakness cause world war 2? or did any Western country’s negligence provide a springboard for Putin’s attack on Ukraine? Although many such conclusions seem plausible, they cannot be accepted without many layers of confirmatory evidence, which is usually lacking. Thus trying to form firm conclusions from such fallacious post hoc reasoning is not only foolhardy, but often dangerous!
I conclude with a warning to all: Avoid the post hoc trap with constant skepticism!
Below is a link to access to my books available for purchase (digital or print versions):
For placement on the mailing list for future posts, contact tavelmorton@gmail.com