"Categories" flailing notes action

Today I thought that it might make sense to return to the blegg/rube parable to
introduce the "It's profitable to make a new category for things that are
_cheaply identifiable_ as coming from a different distribution" point before
switching context to the two types—but maybe that's too repetitive and boring,
and I should just skip to what I actually want to say about trans women (the
part that I'm continually failing to be brave enough to write)?

On the other hand ... I probably _should_ stick to the rule of establishing
philosophical points with neutral-valence examples before getting into the
mind-killing applications: if you try to do both at once, people who can't
handle the political implications won't learn the philosophy, either. If you
successfully teach the philosophy first, they will learn something before
latching on to some other rationalization to deny the political implications.