Colloquium: Two Problems about Permission

October 23, 3:30-5pm, Cobb 201
Paul Portner, Georgetown University

We have a philosophical problem and a linguistic problem:

Problem 1: Lewis (1979) pointed out the original "problem about permission": while it is easy to give a formal pragmatic theory, in terms of possible worlds, of commanding, it is not equally easy to give a parallel account of permitting.

Problem 2: The canonical device for commanding in natural language, the imperative, apparently can also be used to give permission, as in (1):

  1. a. Eat some more broccoli.
  2. b. Have some more wine.

Linguists have sometimes responded to the second problem by denying that imperatives ever truly give permission in Lewis's sense. Rather, Sperber and Wilson (1998), Schwager (2005), and Portner (2005, 2007) propose that permission imperatives have the same force as commands, but differ in that the action is something the addressee desires. Thus (1)b gives permission because it is assumed that the addressee would like some wine. Grosz (2008), however, argues that permission imperatives are semantically different from command imperatives. In particular, he studies the occurrence in German of modal particles in both imperatives and explicitly modal statements. It turns out that JA and bloß occur in both command imperatives and modal statements with necessity operators, while ruhig occurs in both permission imperatives and modal statements with possiblity operators. Based on this correlation, he argues that (1)a should be formalized with a null analogue of must, and (1)b with a null analogue of may.

I will argue that both problems can be solved in terms of Portner's (2005, 2007) analysis of imperatives. In these works, I propose that the conversational force of an imperative is to be analyzed as the addition of a property to the addressee's To-do List, a set of properties defining (in combination with the common ground) which actions will be seen as rational. This techology will let me develop in greater detail van Rooij's (2000) solution to Problem 1. Moreover, the concept of permission which arises from this discussion implies that permission imperatives involve precisely the same modal force as command imperatives. I conclude by discussing why the correlation between command and necessity modal, on the one hand, and between permission and possibility modal, on the other, fails to prove that permission imperatives contain an analogue of may.