Given the difficulty of obtaining analytical solutions for the diffusion of interacting geminate pairs of (ion) radicals in liquids, it is common, following the original treatment of Mozumder, to "prescribe" this diffusion. A demonstration is given that such a prescription is impossible for any interaction potential other than the Coulomb potential. This demonstration suggests the inadequacy of this common approach to modeling geminate pair and spur dynamics in the largest emerging class of organic solvents: room-temperature ionic liquids.