Which of the following arguments has recently been made in TV commentaries, newspapers and blogs:
1. Even non-smokers can get lung cancer, so smoking does not put one at risk for lung cancer.
2. Even thin people can have high blood pressure, so being obese does not put one at risk for high blood pressure.
3. Even white people can be arbitrarily harassed and arrested by police, so being black does not put one at risk for arbitrary harassment and arrest by police.
1 and 2 are bad arguments and though I suspect I’ve heard the first from some tobacco defenders, it’s not going to fool most people, one hopes. #3, which has been showing up recently, even on supposedly liberal blogs, strikes too many people as a good argument.
So what’s wrong with it? The premise is about there being some instances of a feature, F, in a population of non-G’s. The conclusion is about the probability that a G will have F. We can show the reasoning is fallacious by producing arguments of the same form with true premises and false conclusions. We can point out that nothing in probability theory supports it; the premise is irrelevant to the truth of the conclusion. I’m not sure it has, however, its own name. Does anyone know of one? If not, perhaps we should think of one. We see the same form in enough contexts; e.g.,
Men can get raped, so women are not especially at risk….
Men can suffer from domestic violence, so women are not especially at risk…
What is going on when these arguments are put forward? We could spill a lot of ink on the topic, but the comment “They just don’t get it,” might be as good as any. Relatedly, the programs on TV I at least have watched have displayed an unnerving ability to find white people who seem just not to get it with Henry Louis Gates’ recent arrest. That might be one sign of the problem the USA still has with race.
My favorite comment in all the reporting: “Now we know that if you want to rob a house, the first thing you do is get a taxi to go there and wait for you.”