Broken heuristics: you arrive in a city, perhaps for university—the people one year ahead in your halls of residence say, “Don’t go down there, it’s a bad area.” You say, having trained yourself to think in a scientific way, “That’s a sweeping generalisation, perhaps it would be better to say, ‘Most of the time it’s a bad area, between certain hours—and at other times it’s safe’.” “It’s a bad area.” Well, you could go there and it could be fine—and then your scientific approach will have been vindicated, another foolish superstition banished by scientific thought (perhaps we could get down to quantifying how bad or how good it is as an area).
This way of thinking is considered high-status—there’s a book that was standard issue for about fifty years, went to dozens of reprints, called Straight and Crooked Thinking. It’s full of examples like this: “Some people say black people are less intelligent than white people; now, if we take that to mean ‘Intelligence Quotient’, we can arrive at the statement ‘IQ statistics show that while, on average, black people have a lower IQ than white people, there remains a significant number of black people with higher IQs than the average white person’.” There you go—neatly scientific; and the author extols scientific thought and “experts” throughout. The author, a Cambridge don, really had a liberal bias—so the book is actually about how to think in a liberal way; specifically, he uses heuristics like “it’s a bad area” as examples of “crooked” thinking because it’s not logical, it’s emotive, and it doesn’t flow from evidence.
Heuristics are really useful and usually correct enough for all practical purposes (like when you don’t want to get mugged). This pedantic “but of course, technically, 95.5% of trips to that area are safe” is precisely the “rational scientific thought” that gets people killed—because it’s not “prejudiced”, you see (no sweeping generalisations). Hence scientific thought destroys common sense and, ultimately, society itself.