<p>This paper argues that we cannot let the common ground do double duty. If we believe that both Stalnaker’s and Clark’s projects, on the one hand, and that of Peirce, Hamblin, Brandom and Geurts, on the other, capture important aspects of communication, two sets need to be distinguished: the set of common beliefs to capture the fact that we usually formulate our utterances based on what we assume is common belief between us and our interlocutors, and that of commitments to capture the idea that our assertions commit us to their contents. In discourse-semantic frameworks such as Farkas and Bruce’s very influential Table model, the two are not distinguished as strictly as they should. This point is remedied in the context structure proposed in this paper. By distinguishing common beliefs and commitments, we are capable of dealing with the fact that in actual life we are often committed to things we don’t believe and we often commonly believe things to which we are not committed. The proposed context structure also allows us to do justice to the fact that with our speech acts, we change the world. This is done by adding a Kaplanian context to the context structure. We can now separate the informative update as an update of the common ground (which importantly includes and even starts from an update of the performative aspects of the speech act) from the real performative change. The proposed context structure is suited to illustrate the multifaceted effects of speech acts. It thus also clarifies how certain pragmasemantic strands of research relate to each other in what they do and aim to do.</p>