<p>The paper explores two prominent approaches to the view that linguistic meaning is the product of normatively constrained use: Brandom’s scorekeeping model and Dummett’s assertion-conditional account. I note that the scorekeeping approach faces a problem in securing the repeatability of normatively constrained moves, and that this problem doesn’t feature in the assertion-conditional approach. I conjecture that Brandom opts for the scorekeeping approach as a way of explaining objectivity in perspectival terms, but I go on to argue that Brandom’s arguments here are unconvincing. Instead, I illustrate how an adaptation of the assertion conditional approach—one which appeals additionally to retraction conditions—can motivate a distinction between mere adherence to norms—blamelessness—and getting it right—faultlessness. So I tentatively conclude that the assertion-conditional approach both provides a better locus for the operation of linguistic norms—namely the practice, not of scorekeeping, but of performing speech acts—and delivers a noteworthy notion of objectivity.</p>