These metaphysical claims about drift (and selection and other evolutionary processes—but it is just drift that interests us here) set the stage for the statisticalists’ challenge. Matthen and Ariew (2002) challenge the claim that there is a defensible sense in which drift is a force. Aside from the fact that it does not have a predictable or constant direction (as Sober readily acknowledges), they point to a case of two similar populations subject to the same selective pressures, one in which t