Nielsen and Taylor have recently suggested that at least some of what has been previously explained as phonon drag thermoelectric power might in fact be due to electron diffusion. This note examines some of the experimental data on thermoelectric power in the light of this new work. We show that the existing interpretation of very low temperature data survives unscathed, and we investigate the conditions which would be necessary to be able to detect Nielson and Taylor's `phoney phonon drag'.