The dialogue in (1) licenses an inference of faultless disagreement: neither party is felt to be wrong/'at fault.' Because a standard semantics predicts that for any proposition p, either p or ~p must be false, faultless disagreement has driven theorizing in both linguistics and philosophy that assigns a special semantics to predicates of personal taste (PPTs such as tasty, fun), epistemic modals, and aesthetic and moral terms. We extend this investigation of faultless disagreement, presenting an experiment that shows faultless disagreement to be sensitive to the expertise of interlocutors, as well as to the experimental participant’s own beliefs about the validity of expertise in a particular subjective domain (e.g. wine, beer, movies, art, ex.2a,b). The takeaway: We suggest that instead of being a fixed property of individual predicates, faultless disagreement is modulated by factors independent of the lexical content of the sentence. This problematizes the status of faultless disagreement as a desideratum for the semantics of individual predicates; any account of subjective predicates needs to account for the ways in which faultless disagreement is sensitive to extra-sentential context. (1) Sam: This wine is tasty. Alex: No, this wine is not tasty. (2a) A conversation between two wine experts. One wine expert says: This wine is {tasty/revolting}. The other wine expert says: No, this wine is not {tasty/revolting}. (2b) A conversation between Andy (who has no expertise in wine) and a wine expert: Andy says: This wine is {tasty/revolting}. The wine expert says: No, this wine is not {tasty/revolting}.