on 05-07-2013
Carolina Blasio & João Marcos, IFCH - UNICAMP - Brazil & LoLITA - DIMAp - UFRN - Brazil
July 5, 2013, Friday, 16h15m.
Abstract: Semantically, if an agent asserts or refutes some sentence it is commonplace to identify these attitudes with truth-values. We propose instead a useful model for representing and reasoning about information that accommodates distinct cognitive attitudes concerning acceptance or rejection by an agent. From that model, we show a generous four-place notion of entailment, henceforth called B-entailment, that generalizes the well-known approaches of multiple-conclusion entailment. One of the advantages of such model is that it provides a natural interpretation for non-classical logical values characterizing `the unknown': some sentence unbeknownst to a given agent might be a sentence which the agent has reasons to accept and also reasons to reject; alternatively, the sentence might be unbeknown to the agent if she simultaneously has reasons not to accept it and reasons not to reject it. As we will show, yet another advantage of our generalized notion of entailment and of its underlying many-dimensional structure of truth-values is that it provides a simple and appealing framework for the uniform representation of many known logics. The talk reports on joint work with João Marcos.
Room: 3.10, Mathematics
Support: SQIG/Instituto de Telecomunicações with support from FCT and FEDER namely by the FCT project PEst-OE/EEI/LA0008/2013.