In this paper, I challenge the popular belief shared by cognitive-minded theorists and therapists that the treatment of "inappropriate" or "dysfunctional" emotion should primarily be about challenging the eliciting cognitions. Although I acknowledge that sometimes therapy should proceed in this way, I point out that in some cases it is clearly the case that therapy should not proceed in this way—namely, in those cases where there are no eliciting cognitions, or in those cases where our concern lies with the kinds (...) of emotions that are being elicited, and not with the eliciting cognitions. I also consider, and reject, two responses that might be made to the arguments that I outline in the paper, and I conclude by describing what I believe to be some of the important practical or clinical implications of the arguments that I make. (shrink)