next up previous
Next: Counterfactuals Up: DISCUSSION OF LITERATURE Previous: Logics and Theories of

Other Topics


There are two other areas where some analysis of actions has been necessary: command-logics and logics and theories of obligation. For the former the best reference is Rescher's book (1966) which has an excellent bibliography. Note also Simon's counterarguments to some of Rescher's theses (Simon 1965, 1967). Simon proposes that no special logic of commands is necessary, commands being analyzed in the form `bring it about that p!' for some proposition p, or, more generally, in the form `bring it about that P(x) by changing x!', where x is a command variable, that is, under the agent's control. The translations between commands and statements take place only in the context of a `complete model', which specifies environmental constraints and defines the command variables. Rescher argues that these schemas for commands are inadequate to handle the conditional command `when p, do q', which becomes `bring it about that tex2html_wrap_inline1497 !': this, unlike the former, is satisfied by making p false.

There are many papers on the logic of obligation and permission. Von Wright's work is oriented in this direction; Castañeda has many papers on the subject and Anderson also has written extensively (his early influential report (1956) is especially worth reading). The review pages of the Journal of Symbolic Logic provide many other references. Until fairly recently these theories did not seem of very much relevance to logics of action, but in their new maturity they are beginning to be so.

John McCarthy
Mon Apr 29 19:20:41 PDT 1996