Time to link to another post in Language Log: “The “happiness gap” and the rhetoric of statistics”. It’s more to do with general scientific method than with linguistics, but it expresses really well something that always annoys me no end. Mark Liberman writes: “Most people think in essentialist and non-statistical terms, as if all the members of a category were uniform copies of an invariant prototype.” Yes! That is exactly it. Of course it can be questioned whether actually most people do it – it is certainly the case that it seems that most people do, and that enough people do it for it to be what the doctors call Very Annoying Indeed.
Which is not to say that I would like everybody to walk around thinking in statistical terms. I don’t. I don’t want to think in statistical terms myself, it hurts my brain when I have to. But the point remains: “More women than men like to do X, so she’ll like to do X cos she’s a woman” (or teenager, or any other more-or-less random grouping of people.) It doesn’t actually take any knowledge of statistics or scientific methods to see the basic flaw in that type of reasoning – and yet we see examples of it every day.
So read that post, even though it’s longish and contains statistical diagrams. It’s worth it.