I'd be shocked if more than about five or so people actually fully understood what I'm talking about here, and unsurprised if nobody did. I suspect that those who grok the maths and physics type bits will not grok the gender-theory type bits, and vice versa. However, I'm going to post it anyway, because I'm sure that you all really need more proof that I'm a complete and utter dork.

So, we have the traditional view of gender, which is discrete. There are two distinct states into which everyone falls. Go a little bit more progressive and you get a line with "male" at one end, "female" at the other, and "androgynous" in the middle. Go a little further and you get what amounts to a two dimensional vector space, with male on one axis and female on the other, which allows for differentiation between "both" and "neither". Then, you can add an extra axis, perpendicular to the first two, and labeled "other", giving gender as a three dimensional vector space.

So far so good. I've seen this sort of model, or slight variations upon it, proposed numerous times. The problem is that "other" is a distinctly catch-all type of term. There's no reason why there can't be a second other which is independant of and orthogonal to the first one (and to male and female). Or a third. Or a fourth. Or any number. So let us add an infinite number of dimensions, each with infinitesimal contributions, and move over to a Hilbert space. Let us also remove the labels "male" and "female" from this basis since giving any significance to individual dimensions becomes utterly pointless.

So, each person has a certain inate gender, which is, essentially, some infinitely dimensioned vector within our hilbert space. It would then seem to make sense that we never actually observe the gender directly, but rather that we observe how certain operators act upon it -- or, to put it another way, how the person in question responds to a given set of stimuli. Now, if we allow our operator to be an eigenfunction to the gender's eigenvector then we allow for our actual observable to be a simple eigenvalue.

I'm just about managing to convince myself that it's plausible for their to be whole families of gender-functions which could have broadly similar eigenvalues for many operators, but maybe one or two divergent behaviours, and that this would, essentially, depend upon the exact form of the operators. There could be, for instance, two such broad families, which we could label "male" and "female" and then for the rare gender-function that falls outside these families, we could have chaotic behaviour.

It's simple, really.

So, did anyone understand any of that? And if so, does any of it sound like a remotely useful model, or was it just the pseudo-intelectual wangst that I believe it to be?

- Gender is just like quantum physics

## Re: senji pointed this page out on irc...

rhoTypically, you can have a matrix M. This can then have an eigenvalue n associated with a eigenvector lamda. This gives you M*lamda = n*lamda. When I said "eigenfunction" I was refering to the (matrix representation of) the operator M. I think you may well be right that I've fluffed my nomenclature though; but that's what I was intending to refer to.

## Re: senji pointed this page out on irc...

ewx