What's the Worst Problem in Human Thought?

Overconfidence, says Daniel Kahneman. Stereotyping is unavoidable in spite of its tragic nature:

“Think of it this way. A form of stereotyping is involved in understanding the world. So I have a stereotype of a table, I have a stereotype of chairs. Now when you start having stereotypes of social groups, it’s the human mind at work. It’s not a different mind. It’s what you need to get around in the world.” You can slow down and become aware of this, Kahneman believes, but the underlying mechanism isn’t going to change.

Not that he's in favor of it: the article goes on to relate his experience as a victim of the Holocaust. But compared to overconfidence, stereotyping is at least mostly useful most of the time. It's absolutely necessary to understanding the world, even though it leads us to error now and then.

It's an issue we discuss here from time to time. It's what I mean when I talk about physical reality being analogical rather than logical. If "human" was a logical object, you could make claims about a human that would be absolutely certain to hold for any human. It turns out that you can't do this with logical certainty: even claims that hold almost all the time ("a human being has two eyes," "a human being is either male or female," "a human being will have exactly one parent of each sex") turn out to admit of exceptions. We are blindsided by the exceptions because we think of the category as logical, but it's really a stereotype -- or, as I prefer to explain it, the members of the category are members by a relationship of analogy. It's not that they really belong to a hard, logical category. It's that we've found ways in which they seem to be alike, and have made analogies between them in our minds.

Recognizing that should do something to undermine the problem of overconfidence, as you suddenly become aware that much of your understanding of the world is built on shaky ground. It is characteristic of analogies that they always fail. Sooner or later, your analogies will break.

By the way, his examples are interesting because tables and chairs are artifacts rather than naturally occurring things like human beings. With artifacts, you really can make at least some logical claims because they were created by a human mind for a particular purpose. You can say with certainty that "any artifact that is a table will be potentially be capable of holding objects up off the floor or ground" because that's what a table is. These are actually the best case scenario for stereotypes about physical objects that hold universally.