Jane Fae, responding (2013-01-17) to a comment by 'Troll1' (who doesn't believe in more than two genders or sexes) on her recent New Statesman piece:
I don't "just" believe all that "social-construction-of-gender nonsense". I question the somewhat dogmatic assertion around sex(es).
[...]
The issue of sex, as well as gender, is actually a categorisation issue, which is often highly problematic to a scientific study of an area. Like nuclear particles, before we had the tools to delve deeper, we had a pretty limited view of the number of categories in existence.
Sex has historically been divided into two. However, an alien race, washing up on our shores, would probably not come to that conclusion.
They'd toss all the factors to hand, including gene and chromosomal factors, physical expression, behavioural and what-have-you, churn the requisite dimensional transmogrifier, and see what came out.
And no: i'm not theorising some sort of sci-fi machine: I mean tried and tested methods such as cluster analysis or factor analysis or principal components analysis or....
Oh, dear. There are a lot of methods and while some fit certain data types better than others, the awkward thing (and I've watched this happen across dozens of data sets), the number of categories you extract is often method-dependent.
Sometimes similar categories with argument over whether there are nine or seven or similar: sometimes totally different categories.
In the end, its not just gender, but sex that is socially constructed...and subject the base data to analysis and I'd guarantee you end up with more than two categories."