You don't link to actual molecular genetic research on behavior and sex differences. I will. Let's do two quick examples.
http://www.nytimes.com/2013/02/15/science/studying-recent-hu...
Gaining a deep insight into human evolution, researchers
have identified a mutation in a critical human gene as the
source of several distinctive traits that make East Asians
different from other races.
The traits — thicker hair shafts, more sweat glands,
characteristically identified teeth and smaller breasts —
are the result of a gene mutation that occurred about
35,000 years ago, the researchers have concluded.
One pleiotropic mutation actually causes smaller breasts, among other things. For added visual impact, take a look at the EDAR transgenic mouse:
http://mathildasanthropologyblog.files.wordpress.com/2008/10...
Wow! Across species, you can actually see that a single variant makes the mouse hair look East Asian. It's so crazily reductionist, it looks like what a troglodyte like Strom Thurmond might sketch if you had him draw an "Asian-looking mouse". Here's another great one, Fruitless:
http://en.wikipedia.org/wiki/Fruitless_(gene)#Function
Although many genes are known to be involved in male
courtship behavior, the fruitless gene has been considered
noteworthy because it exhibits sex-specific alternative
splicing. When females produce the male-spliced gene
product, they behave as males. Males that do not produce
the male-specific product do not court females and are
infertile.[1] ... The fruitless gene locus also controls
the expression of hundreds of other genes,[6] any subset of
which may actually regulate behavior.
We don't yet know the human analog of Fruitless, the master regulator of sex differences. But if it exists, it's a good bet that it's on the Y or X chromosomes.
Point: it's not all super-complex gene-environment interactionism that we'll never understand[1]. Indeed, as Fruitless or the EDAR story show, the genetic origins of complex gender differences and/or behavioral traits can be almost cartoonishly reductionist, attributable to a single master upstream regulator!
So: do you really want to bet it all on black, bet that genetics researchers aren't going to find governing loci for human gender differences and/or sexual orientation, bet that it's 100% social construction? Because if you know anything about where genomics is going, that is a bet you are going to lose.
[1] PS: I love it when people bring up epistasis, for god's sake, as a sort of catch-all neo-Lamarckianism, reached for like a drowning sailor flailing for a life preserver. Methylome much? Didn't think so.
EDIT: You know what? Let's actually go line by line on this thing.
Males and females are treated extremely differently from
birth. Literally, in the very first day. Thus, it is nearly
impossible to actually prove scientifically that a
significant difference comes from genetic and biological
factors.
Science: doing the impossible since ten years ago. Behold, gene expression analysis of innate male/female neurological differences
before birth, known since 2003 (and well before):
http://www.sciencedaily.com/releases/2003/10/031022062408.ht...
Using two genetic testing methods, they compared the
production of genes in male and female brains in embryonic
mice — long before the animals developed sex organs.
To their surprise, the researchers found 54 genes produced
in different amounts in male and female mouse brains, prior
to hormonal influence. Eighteen of the genes were produced
at higher levels in the male brains; 36 were produced at
higher levels in the female brains.
“We didn’t expect to find genetic differences between the
sexes’ brains,” Vilain said. “But we discovered that the
male and female brains differed in many measurable ways,
including anatomy and function.”
How about this little flail-for-a-life-preserver:
There's research to suggest that the difference in spatial
reasoning is societal.
Which in turn links to this paper:
http://www.pnas.org/content/early/2011/08/19/1015182108.full...This is almost a caricature of academia. To come to the conclusion that gender differences in spatial rotation ability are "culturally influenced"...they analyze two tribes in Northeast India on a "spatial puzzle" that has never been administered to anyone before (Figure 1). And in Table 1 the mean puzzle solving time for men in the "matrilineal" society is still actually lower than that for females, even though the medians point the other way.
Forget about the crimes against psychometry that such a test perpetrates. It goes without saying that they did not compute test-retest reliabilities or do a proper battery of spatial tests, let alone estimate IRT parameters for their single idiosyncratic "spatial" item. And this innocence of all aspects of item-response theory is confirmed by their supplementary information (http://goo.gl/ltysD).
The more fundamental point is the lack of thinking. Not only does this study cut against the grain of literally hundreds of other datasets, it is intentionally designed to not be reproducible. Among other things, why else would one use such an obscure population? And why not use the standard Shepard/Metzler test, in which one is actually mentally rotating two objects and seeing if they correspond? One can programmatically vary the rotation parameter to increase the difficulty, and Shepard famously even observed that reaction time varies linearly with rotation angle (http://plato.stanford.edu/entries/mental-imagery/mental-rota...), as if people were actually rotating the objects slowly in their heads.
Anyway. You know what is reproducible? Let's consult this handy meta-analysis of 70 studies...
http://link.springer.com/article/10.1007%2Fs10648-012-9215-x
It shows a huge and reproducible spatial rotation ability difference between men and women. Don't believe me? Put up a web-based test with WebGL, pay participants, and reproduce the result for yourself. But something tells me that you may not really be interested in such studies. Instead:
Furthermore, it ultimately doesn't matter that much.
Western society has decided that we want equality between
men and women. If that's not happening, then we've
committed ourselves to make it happen.
Right. Comrade, here is where I bid you adieu. If men and women have fundamental neurological differences, which all the evidence indicates that they do, attempting to "close the gap" in mathematically & spatially challenging software engineering is a Sisyphean task. It is as futile as encouraging women to make it to the NBA and out-dunk the very best men. While there will be a Lisa Leslie or Cheryl Miller, there won't be very many of them, and that won't be the fault of men.
Because there can be no obligation to do that which cannot be done.