Games and the Reason-Knowledge Principle 博弈与理性知识原理.pdfVIP

Games and the Reason-Knowledge Principle Brian Weatherson John Hawthorne and Jason Stanley (2008), defend what they call the “Reason- Knowledge Principle”. Where one’s choice is p -dependent, it is appropriate to treat the proposition that p as a reason for acting iff you know that p . (578) There have been many attempts in the literature to show that this leads to implau- sible actions. As Jonathan Ichikawa (2011) shows, most of these attempts rest on further, and arguably false, assumptions about the connection between reasons and action. Relatedly, most of those responses concern the role of knowledge and reasons in decision-making. I’ll argue that we can formulate a sharper prob- lem for the principle if we focus on game-playing, and say exactly which extra assumptions we are making. The Reason-Knowledge Principle should have the following implications, at least for cases where S ’s aim is to produce the best outcome. (1) If S knows that and will produce the same outcome, and S must choose or , then it is rationally permissible for S to choose . (2) If S knows that and will produce the same outcome if p , and will produce a better outcome if p , then it is rationally permissible for S to choose iff she knows p . The point of (1) is that S can use her knowledge that and will produce the same outcome to justify making an arbitrary choice between and . And the point of (2) is that the Reason-Knowledge Principle suggests only knowledge that p could justify ignoring the fact that does worse than if p . Define a symmetric game as having these features: The game is purely co-operative; each player gets the same payoffs; Each player knows nothing about the other save that it is common knowl- edge the players are rational, and hence know what each other’s rational requirements are; Each pl