Truth and Reference in Context

In communicative exchanges, sentences are uttered against a background of shared beliefs or attitudes which helps the audience to determine the content of what the speaker has said. Unfortunately, different agents may have different views of this common ground. From this standpoint, one of the most familiar phenomena is accommodation, which enables the addressee to incorporate the speaker’s presuppositions into her own view of the common ground. As suggested by definition (URR), this phenomenon is analyzed here as a case of global context shift, since the missing piece of information, which the addressee is willing to share, will be an integral part of the revised context. A less familiar, but equally important, phenomenon is what I call discommodation, whose main feature consists in the fact that the missing piece of information, although essential to the comprehension of the utterance, cannot be shared by the addressee because it sounds problematic or even false to her. This is not a marginal aspect of communicative exchanges, which are often characterized not only by different views of the common ground, but also by incompatible views. In such cases the addressee opens a “presuppositional slot” to take into account the assumptions which serve to select the reference of the noun phrase, but which are not incorporated into the revised context because of the addressee’s disagreement. Thus, such a process is analyzed as a case of local context shift, which affects only the noun phrase and which, unlike global context shifts, allows us to keep considerations about reference distinct from considerations about truth. Starting from the problem of the truth-value of sentences uttered in contexts containing false assumptions, one of the main purposes of the paper is to propose a definition of truth (with respect to a presuppositional apparatus) which does not ignore the role of discommodation when different views of the common ground are involved. The relevant truth conditions are given in (TC). Finally, this definition is used in the last two sections of the paper to justify the idea that, in the cases of misdecription made popular by Donnellan, one can recognize the semantic role of the speaker’s reference without assuming the ambiguity of the definite descriptions.