We can construct an evolutionary thought experiment where drift is absent, showing that drift is not constitutive of all evolving systems and is therefore separable

Brandon (2006), however, argues that “eliminating heterozygosity” is not sufficient to show that drift is directional, given that (as Stephens would readily acknowledge) if there were two alleles at a locus, beginning at equal frequencies, we could not predict which of the two alleles would go to fixation, only that one of them would; Brandon likens this to saying that “a 20-Newton force is acting on object A”, (2006: 325), which, he seems to imply, is not a directional claim. Moreover, he argue