Prioritized dynamic update and retraction functions selecting on inconsistent information ∗ Giuseppe

In this paper a model for updates on belief sets and retractions thereof is introduced using the standard format of Adaptive Logics. The core of the update retraction procedure is represented by abnormal expressions derivable in the language: they express updates with information contradicting previously derived contents. The adaptive strategy aims at restricting the validity of these formulas by focusing at each decreasing degree on the update which is the most rational to retract in order to restore consistency as soon as possible. This work is related to the standard operations of retraction and withdrawal from the AGM-paradigm and the effects of dynamic operations such as public announcement in Dynamic Epistemic Logic.