On the microscopic origin of Stokes’ law