Monkeys Typing

Last Friday's panel on Demystifying Legal Reasoning, by Larry Alexander and Emily Sherwin, about which I posted earlier, was a rollicking good time for all. While all of us commentators had lavish praise for the book, we all also took issue with, among other things, the authors' commitment to speaker's meaning (with Heidi Hurd delivering the most forceful critique on this point). Here I want to pile on.

Proponents of speaker's meaning (including Alexander and Sherwin) are fond of examples of what appear to be comprehensible statements produced mindlessly: a cloud formation that forms the letters C-A-T; the outline of Alfred Hitchcock produced accidentally by the wandering of an insect on the sand; and nearly everyone's favorite, a complete English language sentence (or all the works of Shakespeare) produced by monkeys randomly pounding the keys of typewriters. Intentionalists say that none of these occurrences has any meaning, because none was intended to communicate anything.

This seems self-evidently wrong. Suppose I am staring out the window of my office, trying to remember the name of a particular student. I look up at the sky and see clouds forming the letters T-O-M. "Eureka," I say to myself. "The student's name is Tom." One need not think that God has sent me a message to see that there was not just meaning in the cloud formation, but useful meaning in it, at least for me.

At this point, the intentionalist will typically object that I have supplied the meaning, but this too seems wrong, or at least incomplete. Suppose another student were in my office and, in answer to the question, "Who was your classmate who made that interesting point about the unitary Exeuctive in class today?", the student volunteered, "that was Tom." So far as I'm concerned, my comprehension of the phrase "that was Tom" supplies meaning in just the same way that the cloud letters "T-O-M" in the sky do. In each case, I must decode (spoken or written) English to produce meaning.

Intentionalists are undoubtedly smart enough to realize all of this, and so we are left with the puzzle of how it is that they can think, as Stanley Fish says, that "apart from intention . . . words don't mean anything." In my example, the cloud "T-O-M" obviously did mean something to me, even though I wasn't searching for anybody's intention.

At this point, one is tempted to say that what we have here is a semantic disagreement. Intentionalists want to preserve the word "meaning" for speaker's meaning, whereas people who believe in utterance meaning (and other conceptions of meaning) have a broader view. It's sometimes exasperating to argue with the intentionalists on whether we are using the concept of meaning correctly, but let's concede the point, and use the word "shmeaning" to refer to the broader category, which includes messages attributed by readers even if not intended by the actual author, and even if no message was intended by any author (or speaker).

Then, it seems, the disagreement over whether to adopt speaker's meaning (what the intentionalists would simply call "meaning") or some other account of shmeaning, can be recast as a debate over whose understandings should count in various enterprises. If the enterprise is "being reminded of people's names," then there is no good reason to prefer speaker's meaning to reader's shmeaning because reader's shmeaning can--albeit rarely--remind one of a name, even absent a speaker.

So suppose the enterprise is "figuring out what the law is." The intentionalist move now is to say something like "figuring out what the law is consists, at least in the first instance, of interpretation, and one can only interpret by retrieving speaker's meaning." But this cannot be a claim about language, for as the cloud example shows, there are reasons to prefer reader's shmeaning to speaker's meaning in some contexts. And so the intentionalist must give a reason peculiar to law why speaker's meaning should be preferred. Claims that speaker's meaning is simply the only shmeaning won't cut it.

Might the intentionalist slightly modify his account to say that it holds wherever we are looking for the meaning or shmeaning of an intentionally created text? Clearly not. Suppose that instead of seeing the cloud formation, I looked at my desk and saw a letter signed "T.O.M.," initials used by a friend named Tanya Olivia McDonald. Tanya didn't intend to convey the (sh)meaning "Tom," and I know that fact. Nonetheless, I derive a valuable shmeaning--the name of the student who made the point about the unitary Executive--from what Tanya wrote, even though it was not what she intended.

One can, of course, marshall decent reasons why there are legitimacy problems with courts inferring unintended (sh)meanings from authoritative texts. But those are reasons rooted in normative political theory, not a theory of language. And so, it appears that intentionalism is, at bottom, an attempt to disguise a (not wholly unattractive) normative political theory as an account--indeed, as the only possible account--of language.

Posted by Mike Dorf