There’s been a lot of discussion recently about the threat that enemies of science represent to contemporary American culture. That concern, is well founded. In fact, it is better founded than I believe most people suspect. Until recently, most victims of the anti-science contagion were religious fundamentalists, or those on the far right of the political spectrum. Recently, however, a more virulent form of the anti-science contagion has appeared. It’s spreading like wildfire and reducing to imbecility even those one would hitherto have counted among the intellectual elite. A case in point Marie McCullough, a science writer for the Philadelphia Inquirer.
McCullough argues that cases of shingles outbreaks in people who had received recently received COVID vaccines cannot have been caused by the vaccines as early speculation had suggested (“Shingles is not caused by COVID-19 vaccines. Here’s the science”). “Science,” writes, McCullough, “is littered with post hoc fallacies.” The inference, she argues, from the fact that an outbreak of shingles followed closely upon a COVID vaccine to the conclusion that it was caused by the vaccine is an example of such a fallacy.
A post hoc, or more correctly, a post hoc, ergo propter hoc fallacy refers to the inference from the fact that one event followed another event, to the conclusion that the first event caused the second event. More is necessary to support such a conclusion than mere temporal succession. McCullough is right about that. That’s about the only thing she gets right, however, in an otherwise very misleading article that is marred by its own equally egregious informal fallacy.
Correlation, explains McCullough, is not sufficient for establishing a causal relationship. She compares the inference that shingles might be a side effect of the new COVID vaccines to the infamous post hoc, ergo propter hoc inference that “ because the rooster crows at dawn every day, he makes the sun come up.”
Again, McCullough is right that a correlation is not proof of a causal relationship. Failure to prove a causal relationship, however, is not proof that there is no causal relationship. That inference is also a well-known informal fallacy. It’s called the “fallacy of appeal to ignorance,” or argumentum ad ignorantiam. It occurs when someone interprets the failure of an argument to prove its conclusion as proving the opposite conclusion. If you argued, for example, that alcohol has a sedative effect on the basis of the fact that you felt more relaxed after you had a glass of wine, someone could legitimately point out that you had committed a post hoc, ergo propter hoc fallacy. They would commit a fallacy of argumentum ad ignorantiam, however, if they purported, by having exposed the problem with your argument, to have proven that wine does not have a sedative effect.
We know wine has a sedative effect because years of observing changes in behavior of people after consuming alcohol led scientists to assume there was a causal relationship between alcohol consumption and sedation so they set about looking for the causal mechanism responsible for those changes. That’s the nature of the scientific method. You observe a phenomenon, such as a correlation between two events —e.g., alcohol consumption and sedation — come up with various hypotheses concerning what causal mechanisms might be at work there, and then test those hypotheses until you find one that passes the tests.
Correlations, while they don’t in themselves prove causal relationships, are crucial to science. They are our first indications of causal relationships. We observe a correlation between two events and then proceed to investigate whether there might be a causal relationship.
McCullough’s comparison of the inference that shingles might be a side effect of the new COVID vaccines to the inference that a rooster’s crowing makes the sun come up is misleading. There’s no prima facie reason to suppose a relation between animal behavior and the movement of planets. There is a prima facie reason, however, to suppose a relation between a vaccine and a physical ailment. That is, both relate to the body, and, more specifically, both relate to the immune system.
The COVID vaccines, which have received only Emergency Use Authorization, not the standard FDA approval, use novel medical technology that we are only beginning to understand, hence we need to be particularly vigilant in our collection of information concerning possible side effects. That’s why the Department of Health and Human Services has created VAERS, a website for the reporting of what people may suspect are adverse reactions to these vaccines. Many of what may appear initially to be side effects may turn out, on closer inspection, to be unrelated to the vaccine. We won’t know that, however, for some time to come, hence no physical ailment experienced shortly after receiving the vaccine should be dismissed out of hand.
Finally, while it would be wrong to conclude from a rooster crowing and the sun rising that the former caused the latter, there is a causal relationship between the two. It just goes the other way. That is, it’s the sun rising that causes the rooster to crow. It doesn’t do this directly. What scientists now believe happens is that the rising and setting of the sun sets the rooster’s internal circadian rhythm to alert the rooster to the approach of dawn. Of course we would never have figured that out if we’d dismissed the correlation between the rooster crowing and the sun coming up as purely “coincidental” in the manner McCullough argues the relation between the COVID vaccine and shingles is.
We shouldn’t simply dismiss the possibility of a causal relationship between two events because we can’t immediately confirm what it is. There is no more extreme expression of antipathy toward modern science than that.
McCullough should know that given that she is a “science writer.” In fact, every educated person in our society should know that.