Argues that much of the conceptual-change literature does not study change per se, but infers, often from interviews, changes in students' conceptions and discourse during learning. Concludes that what is needed are descriptions of learning trajectories. Contains 17 references. (DDR)