We tend to get a lot of things backwards when considering the hows and whys of behavior. The underlying nature is diffuse and chaotic, hard to describe or point at. These superficial differences are built on top of or used to leverage our inclinations and so provide perfect looking reasons.
True.
That said... the term "underlying nature" may be part of that backwardsness.
We intuitively model human behavior as underlying beliefs and stuff leading to a rationale, leading to behavior. But really, it's often the other way.
There is an underlying behavior, behavioral pattern or whatnot. The rationale, beliefs and suchlike are overlying.
We do know these things exist, but tend to think of them as pathlogies and abhorations... like motivated reasoning. But, conscious reasoning following an intuitively reached conclusion is probably the standard model for human reasoning.