People twist themselves into knots on the question of whether or not history is a science. I’m not prepared to defend the claim that history is *always* a science, but it certainly can be.
As we all know arguably the most important defining feature of a science is that it is ‘falsifiable’ — it makes predictions about future sensory states which could in principle turn out to be wrong.
One major source of confusion here is that history makes what we might call ‘retrodictions’, i.e. predictions about events that happened in the past. This seems vaguely screwy, somehow.
But the fact that arrowheads or artifacts are thousands of years old shouldn’t concern the historian-scientist anymore than the fact that the light hitting a telescope is millions of years old should concern an astronomer-scientist.
The predictions yielded by a historical theory constrain *future* sensory states in a falsifiable way. If you subscribe to the idea that humans crossed Beringia 20,000 years ago, you should never, in the future, find an arrowhead older than that. If you do, your theory is falsified.
So history passes as least one of the more significant tests by which we separate science from non-science