They're separable by experimental verification. The laws of physics are multiple orders of magnitude more accurate than the laws of economics, and the physicists are certain that at least one of their models is wrong.
Try getting an admission like that out of an economist!
As economists we are certain all of our models are wrong! But some can be useful in a constrained environment to answer specific questions, so thats why we have so damn many of them!
Economics are defined by humans.
They’re taught to each other by humans.
It’s all wrapped in how we treat each other at scale.
Separating the two is impossible except as a semantics game.