This is an opinion, but one bias I have seen among academic interpretation is that of assuming conscious motivation from people in the past as though there was a moment where someone stepped back and saw all sorts of alternative ways that the world could be without any bias. There is a reasoning style I see a lot in ancient Greek authors that is contrary to the reasoning that begins with nothing, which seems to me like a more recent way of reasoning in the history of ideas (starting with no preconceived notions). Basically you begin with what exists, and provide explanation for it being that way, you don't conceive of what is as fundamentally flawed or even able to scrap entirely. To an extent, I think we all still do this, less than we ever have perhaps, but there are forms of our way of life that are still largely unquestioned.
In terms of gender, there are bodily conditions that create specific scenarios to work out advantageously or not for a given sex in a community. The symbolism of gender associated with that might not be so complicated and motivated so much as just taking what has occured and slapping a role/symbol on it after. So instead of starting without a concept of how things are objectively oriented, you look at the way things are as indicative of the way they should be. Or rather, you see the way things are as meaningful.
So if a role occurs not due to some symbolic association, but just because it worked out that way, developing a meaningful reason why after isn't that much of a jump
At that point to you, that is just the way things are, seeing meaning in the way things are is something we still do
After that happens, those concepts of role have entered and then it gets reinforced
If you want to get more in depth with it conceptually, you could also say that as symbolic thought grows, a divorce between actor and act appears that allows for "role" to become a meaningful concept very easily, because role is a stand in for "symbol" and gender role is a symbol for the types of actions done by that role, it is a symbol
Language itself encourages identification of roles
So does the sort of pattern recognition humans engage in.
