Because we have easily observed societal norms that have to account for some percentage of the difference, without measuring differences absent those norms we have no conceivable way to know that hereditary traits contribute to any percentage of the differences.

> we have no conceivable way to know that hereditary traits contribute to any percentage of the differences.

Right, that explains why we can't know what % heredity contributes. It doesn't explain why we should assume that % to be small.

Because many historical cases of what were thought to be fundamental differences between the genders turned out to be strictly socially informed. I think it's more likely that the current suite of stereotypes is also incorrect than that we just happened to finally get it right.

I disagree that stereotypes really vary over time or space as it seems like you're implying. Many of the gender stereotypes today are more or less the same as they were 2000 years ago, either here or on the other side of the world.

> we have no conceivable way to know that hereditary traits contribute to any percentage of the differences.

That's exactly what cross-cultural studies do. The evidence is pretty suggestive.