The debates over the problem of faultless disagreement have played a major role in shaping the landscape of today’s semantic theories. In my paper, I argue that even though the existent contextualism-friendly proposals explain a lot of disagreement data by specifying various ways in which speakers may use subjective predicates, neither provides a satisfactory account which would explain what all the subjective disagreements have in common. In particular, what is lacking is an explanation of the persistent autocentric cases (Lasersohn 2004), i.e., disagreements in which each speaker utters a subjective sentence while openly and knowingly occupying his or her own perspective. In my paper, I offer a solution which consists in supplementing the standard contextualist semantics with an explanation of this most problematic class of cases, which is possible due to redescribing the phenomena in speech act nomenclature.