the week of march 10

trying to derive the exclusive interpretation of or

After working through some further examples of scalar implicatures, we investigated the question whether the exclusive interpretation of or can be derived in a similar way, using Gricean reasoning and the Horn scale <and, or>. We encountered several difficulties when trying to do so. First, we had to work hard to find examples where it is clear that or by itself is responsible for the exclusive interpretation. In cases where the disjuncts themselves are logically or contextually incompatible, we do not need an exclusive interpretation of or to account for the fact that the disjuncts are perceived to be mutually exclusive. The following example, which is modeled after an example originally due to Barbara Partee, might have the right properties: "Jane is studying or (Jane is) in the library". In this example, we perceive an exclusive interpretation of or, even if we know that Jane is a student and is therefore likely to study in the library. The source of whatever exclusive interpretation we perceive must therefore be or itself.

When we tried to derive the exclusive interpretation of or using Gricean reasoning and the Horn scale <and, or>, we encountered further difficulties. As before, Horn scales and Gricean reasoning all by themselves don't take us straightforwardly from the weaker statement "Jane is studying or (Jane is) in the library" to the stronger "Jane is studying or (Jane is) in the library, but not both". Gricean reasoning alone only allows us to derive the weak 'epistemically modified' implicature that the speaker's evidence doesn't imply that Jane is studying in the library, which amounts to saying that the speaker's evidence is compatible with Jane's not studying in the library. If we wanted to derive the exclusive interpretation of or as a quantity implicature via the Horn scale <and, or>, we would first have to derive the stronger epistemically modified implicature that the speaker's evidence implies that Jane is not working in the library. Under the additional assumption that the speaker's evidence is sound, we could then derive the epistemically unmodified implicature that Jane is not studying in the library, which would give us the exclusive interpretation we are after. How exactly would we derive that implicature, though? The question has received a lot of attention recently, and there is currently no consensus about the correct way of going about this in the case of or. Here are the slides about or we looked at last week, and then again this week.

Gamut's chapter 6 gives you a first taste of some of the tricky issues surrounding the enterprise of deriving exclusive or as a conversational implicature from inclusive or. As the chapter makes clear, a speaker who is obeying the Gricean maxims cannot make a disjunctive claim if she knows of one of the disjuncts whether it is true or not. But that already explains why such a speaker isn't in a position to make the corresponding conjunctive claim. That is, if a speaker who is in compliance with the Gricean maxims claims that Jane is studying or (Jane is) in the library, she can't know whether or not Jane is studying, nor whether or not Jane is in the library. It's obvious, then, why she couldn't have claimed that Jane is studying (and Jane is) in the library. This means that we don't need Horn scales to derive the weak 'epistemically modified' implicature that the speaker's evidence doesn't imply that Jane is studying in the library. That it would even be wrong to rely on Horn scales for this purpose becomes painfully clear once we consider disjunctions with more than two disjuncts. If you want to find out more about this, Jenning's entry on disjunction in the Stanford Encyclpedia of Philosophy is a good startiing point. If you want to know even more, take a look at: Jennings, R.E., The Genealogy of Disjunction, New York: Oxford University Press, 1994.