Over the past decades, logicians interested in rational agency and intelligent interaction studied major components of these phenomena, such as knowledge, belief, and preference. In recent years, standard 'static' logics describing information states of agents have been generalized to dynamic logics describing actions and events that produce information, revise beliefs, or change preferences, as explicit parts of the logical system. [22], [1], [12] are up-to-date accounts of this dynamic trend (the present paper follows Chapter 9 of the latter book). But in reality, concrete rational agency contains all these dynamic processes entangled. A concrete setting for this entanglement are games - and this paper is a survey of their interfaces with logic, both static and dynamic. Games are intriguing also since their analysis brings together two major streams, or tribal communities: 'hard' mathematical logics of computation, and 'soft' philosophical logics of propositional attitudes. Of course, this hard/soft distinction is spurious, and there is no natural border line between the two sources: it is their congenial mixture that makes current theories of agency so lively. We will discuss both statics, viewing games as fixed structures representing all possible runs of some process, and the dynamics that arises when we make things happen on such a 'stage'. We start with a few examples showing what we are interested in. Then we move to a series of standard logics describing static game structure, from moves to preferences and epistemic uncertainty. Next, we introduce dynamic logics, and see what they add in scenarios with information update and belief revision where given games can change as new information arrives. This paper is meant to make a connection. It is not a full treatment of logical perspectives on games, for which we refer to [13].
展开▼