Friday, October 30, 2009

A Brief Objection to the Conjunction of Schaffer's Contrastivisms

To begin, I want to consider a sentence that I think I understand.

S. Jones knows that Suzy’s throwing the rock caused the window to break.

By claiming to understand this sentence, I do not mean to suggest I have complete and fully worked out views providing informative analyses of knowledge or causation. I simply claim that the sentence makes sense to me. A helpful comparison is to a sentence like this:

S’. Jones knows that Suzy’s throwing rather than tossing the rock caused the window to break rather than shatter, rather than that Billy’s throwing rather than tossing the rock caused the window to break rather than shatter.

While I don't claim that I am outright unable to parse such a sentence, S seems comprehensible to me in a way that S' does not. And this is the basis of one concern I have surrounding the conjunction of Jonathan Schaffer's contrastivisms.


Schaffer is a contrastivist about knowledge. Specifically, rather than taking knowledge to be a binary relation obtaining between an agent and a proposition, Schaffer takes it to be a ternery relation obtaining among an agent, a proposition, and a contrast proposition. As part of the linguistic story required to make this view of the knowledge relation jive with apparently binary knowledge ascriptions (e.g. "Jones knows that grass is green"), Schaffer maintains that such apparently binary sentences are contextually supplemented with a contrast proposition. Essentially, the assertive utterance of "Jones knows that grass is green" in a given context expresses a proposition equivalent to the content of some sentence like "Jones knows that grass is green rather than blue." The question at issue in the context of assertion determines which contrast proposition is contributed, but the explicitly contrastive claims are not themselves context sensitive.

Schaffer is also a contrastivist about causation. Specificially, rather than taking causation to be a binary relation obtaining between pairs of events, he takes it to be a quaternary relation obtaining among one event, a class of events contrasting that first event, a second event, and a class of events contrasting that second event. As part of the linguistic story required to make this view of causation jive with apparently binary causal claims ("Suzy's throwing the ball caused the window to shatter"), Schaffer maintains that such sentences are contextually supplemented with the contrast events. Essentially, the assertive utterance of "Suzy's throwing the ball caused the window to shatter" in a given context expresses a proposition equivalent to the content of some sentence like "Suzy's throwing rather than tossing the ball caused the window to shatter rather than crack". Again, the contextualist element of the view only has to do with superficially binary causal claims, while explicitly contrastive claims are not themselves context sensitive.

Now, on the conjunction of Schaffer's views, assertive utterances of S ("Jones knows that Suzy’s throwing the rock caused the window to break.") express a proposition along the lines of that expressed by S' ("Jones knows that Suzy’s throwing rather than tossing the rock caused the window to break rather than shatter, rather than that Billy’s throwing rather than tossing the rock caused the window to break rather than shatter."). But since I generally understand assertive utterances like S and have a lot of difficulty parsing assertive utterances of S', it seems as though this is a rather significant problem for Schaffer's view. The complexity of embedding a causal claim (or, for that matter, a knowledge ascription) within a knowledge ascription, is much higher, on Schaffer's view than on a binarist approach to causation and knowledge.

I am sure that there are things for a proponent of the views to offer, but it does strike me as something that requires an answer.

3 comments:

Jonathan Jenkins Ichikawa said...

Your objection is approximately of this form:

According to theory T, sentence S expresses, in a given context, proposition p.
But S is much easier to understand than is S', where S' is a sentence that expresses, in all contexts, p.
Therefore T is false.

I don't think this argument form is a good one. There's no prediction in general that complicated sentences that specify more contextual parameters will be as easy to understand as the simpler context-sensitive sentences.

Joshua said...

Some might distinguish between understanding S and understanding what S says. On Schaffer's view, what S says just is what S' says (in the appropriate context). Lewis might take his key premise to be that he finds what S says comprehensible in a certain way and he does not find what S' says comprehensible in that way. This seems to be a better argument than the one attributed by Jonathan.

Unknown said...

I'm going to opt to accept Joshua's help to avoid Jonathan's objection.

I guess my thinking was that, on Schaffer's view(s), there is a lot more complexity to the proposition asserted by "Jones knows that x causes y" than to "Jones knows that x loves y" (assuming Schaffer is not a contrastivist about love). But this doesn't seem to be the case.