Tuesday, October 30, 2007

A Possibly Foolish Question on Larry Solum on Semantic Originalism

I've neither the time nor the expertise to discuss the extensive Legal Theory Blog post on semantic and normative originalism in any detail. This, then, ought to be seen more in the vein of a comment (as I am an inveterate blog commenter, but Solum's blog doesn't have a comment field), and particular question about which I'd like enlightenment.

To what extent is it the case that the choices within semantic originalism are non-normative? It seems to me that (a very small part of) the case for normative reductionism is a little stronger than Solum suggests, because the criterion for choosing sentence meaning in the context of determining legal principles is necessarily normative.

I take the part of Solum's argument with which I'm concerned to hinge on the idea that if speaker meaning is impossible, there's no normative choice to be made between sentence meaning and anything else: sentence meaning is all there is. And speaker meaning is impossible unless the speaker knows that the audience has knowledge of the speaker's intentions (Solum's "reflexivity condition"), a condition that does not apply to legal texts and in particular the U.S. Constitution.

[Caveat: I know very little about Grice's theory of meaning. I've basically skimmed some stuff of his and some secondary literature. So I may make major errors. Fortunately, this is the blogosphere, where errors are fine and dandy.]

It seems to me that the reflexivity condition makes the existence of speaker meaning weirdly audience relative in a troubling fashion. For surely the drafters of individual passages knew that their fellow drafters (i.e. the others in the room at the constitutional convention) knew their expressed intentions. It's only the broader audience that was supposed to be ignorant.

But that audience relativity might make it possible for there to be something *between* speaker meaning and sentence meaning. Imagine three people, A, B, and C. A utters a statement to B, fully satisfying the reflexivity condition as to B. (B knows what A meant, A knows that B knows what A meant, and we can even make it stronger and say that all of this is common knowledge all the way down.) Now suppose B transcribes what A said, and, unbeknownst to A, communicates it to C. And suppose C develops a plausible theory of what A meant when A uttered the statement to B. (Perhaps B communicated A's intention too, perhaps C simply engaged in a successful hermeneutic exercise.) Then C knows (or has a justified belief about) what A meant when A uttered the statement, but A didn't know it would happen. (I submit that the ABC model accurately reflects the interpretive universe of the U.S. constitution.) This is something more than brute sentence meaning, which claims no knowledge about the contents of the specific speaker's head except insofar as it matches the standard use of the words. But it's something less than fully reflexive speaker meaning.

That middle kind of meaning (call it transcribed meaning) might exist when speaker meaning does not. And so, even if speaker meaning is out the window, we're owed some kind of argument as to whether we make use of transcribed meaning or sentence meaning. And it seems that the argument we're owed has to be normative, for the distinction between speaker meaning and transcribed meaning is not about what the speaker meant by the statement, but what the speaker meant for us to get out of the statement. And intuitively, the reasons why we might want to respect or not respect that are normative ones. So perhaps (in Solum's terms) we might not have to choose between framers' meaning and clause meaning, but we have to choose between transcribed framers meaning and clause meaning.

There's so much more to discuss in that post, I might (or might not, time presses badly) discuss things further here. But that's a first thought, and hopefully not too foolish for all that.