Brian Ellis has argued that the assigning of forces is, in the final analysis, a matter of convention. This conclusion is backed by the premises (1) that forces and force-effects are necessary and sufficient for each other, and (2) that the classification of some state of affairs as a force-effect is at least partly conventional. We argue that the first premise is false, that the second premise is ambiguous as between several senses of "conventional," and finally that he has not (...) established that force-effects are conventional in the sense required for the conclusion he wishes to draw. (shrink)