Comment by stcredzero

My conclusion is that 100% of these articles are more about "accurately modeling the world" then they are about "providing insight," although they all have some theoretical insight component as well.

Before you can figure out why, you have to make sure you can accurately characterize the what. So there's a lot of science that is focused on coming up with a descriptive tool like an adhoc curve, before the underlying principles are discovered.

I think Chomsky is afraid that statistical models will cause people to stop looking for the underlying principles.