Redudant Historical VaR benefits?

CFAI text says 2 benefits of historical method are: 1) not constrained to normal distribution and 2) non-parametric (i.e. avoid probability distribution assumptions). Aren’t these implicitly the same thing?