Wednesday, February 20, 2008

"The Plain Meaning of the Text"

I'd like to inaugurate the substantive portion of this blog by explaining why the commonly used phrase "the plain meaning of the text" is incoherent from a linguistic point of view. Since doctrines that invoke the plain meaning of the text dominate judicial approaches to the interpretation of statutes, contracts, constitutions and other texts, these approaches largely rest on an incoherent foundation. And since discussing approaches to interpretation is much of what judges and legal scholars do, I'd like to suggest that a whole lot of legal thinking is based on an incoherent conceptualization of language, law and text.

Here's the problem: text doesn't have meaning, plain or otherwise. There's no function that takes a string of text as input and returns a meaning as output. Rather, text and meaning are two separate outputs of the syntactic system. They are related, and one of the main activities that preoccupy linguists is trying to figure out the nature of the syntactic computation by studying the restrictions on the relationship between text and meaning. They are separate outputs, and they are fed into different interpretive components of the language system - text into what Chomsky calls the articulatory-perceptual component, and meaning into what he calls the conceptual-intentional component. Properly speaking, the text that we hear, read, or in the case of sign language, see, is the modality-specific translation or interpretation of the textual output of the syntactic system, by the articulatory system in the case of speech and what I suppose could be called the graphical system and the gestural system in the cases of writing and signing, respectively.

Let me try to clarify with a concrete example. Consider the ambiguous sentence I shot an elephant in my pajamas. What makes it ambiguous is that a single text corresponds to multiple meanings. Linguistic theory explains that two distinct syntactic processes, which lead to different meaning outputs, happen to converge on the same textual output (which is then processed through the modality-specific system to be read, heard or seen). So I shot an elephant in my pajamas, as a text, doesn't have a meaning, but rather two meanings, to which the text is indirectly related and mediated by the syntax.

To go a little further, figuring out the meaning of a sentence requires knowing its structure - the way in which the words in the sentence are related to each other. That means that the meaning output of the sentences must include, in addition to the words themselves, information about the structural relationships between the words. In the above sentence, in the meaning in which I was wearing the pajamas, the output contains the information that shot an elephant forms a grouping, or constituent, to the exclusion of in my pajamas. In the meaning in which the elephant was wearing my pajamas, the output includes the information that an elephant in my pajamas is a constituent to the exclusion of shot. This information is not found in the text, which makes no structure apparent, but instead linearizes the words. If interpretation of the text simply involves considering the text and no other evidence, as "plain meaning of the text" suggests, it cannot include the structure which is necessary to determine meaning.

Rather than text corresponding to a meaning, it could be said to correspond to a set of meanings - the meanings that can be generated by the syntactic system consistently with the text. That is,given a text, if you consider every syntactic procedure that generates the text as one output, and put together each meaning generated by those procedures, you'll get a set of meanings that corresponds to the text. Given nothing except the text, the interpreter or constructor has a set of meanings, typically consisting of more than one meaning to select among, and no evidence to assist her with her selection. Finding the "plain meaning of the text," then, first requires finding the set of meanings corresponding the text, which requires considering all of the syntactic processes that can produce the text, then identifying their meaning outputs, and then coming up with a strategy for deciding among them. Hardly plain, and necessarily not restricted to considering just the text.

1 comment:

Anonymous said...

That's okay by me. Weldone. Regards, Emeka