The present article draws on some parallels between pragmatism and the interactionist sociology of deviance to discuss the quest for liberatory consequences often associated with the Labeling Approach (LA). Both pragmatism and the LA exhibit a tension between their antifoundationalist nominalism and their liberatory meliorism. This tension revolves around the question if the insight that "descriptions are all we have" leads to a possibility to change these descriptions. While many proponents of the LA have thought so, antidualist formulations of pragmatism have mellowed this hope without destroying it: Descriptions can always change, but it is rarely antifoundationalist theory that changes them.