Jonathan Schaffer presented a paper (just a few hours ago) on a view he favors called contrastivism.

Contrastivism is the view that knowledge ascriptions express a ternary relation - S knows that P rather than Q.

The more traditional view is that knowledge ascriptions express a binary relation between a person and a proposition.

According the Contrastivist an utterance like "I know that is a canary" really expresses some proposition like "I know that's a canary rather than a raven"

What that contrastive proposition is will vary according to the context. Jonathan Schaffer presented an argument for contrastivism today. Here's a quick and dirty presentation of the argument. This will be overly simplified.

An Argument for Contrastivism (quick and dirty version)

1. Whether or not it is permissible to assert P depends on the context, and the permissibility will depend on some relevant contrastive proposition that the assertion is ruling out.
2. Knowledge is a norm of assertion.
3. If (1) and (2), then contrastivism is true.
4. Therefore, contrastivism is true.
   

This doesn't quite do justice to the more complicated way of fleshing out the argument, but it captures the basic spirit of his argument (in an overly simplified way). I don't expect these premises to make any sense to someone not familiar with the knowledge is the norm of assertion literature (or this business about contrastive propositions), but I wanted to get this argument down somewhere. I thought this would be a good place to get it down.

I may come back to this and discuss his motivations for these premises, but I won't go into that now. I'm off to dinner. Philosophy conferences are awesome.