Monday, March 17, 2008

Linguists on DC v. Heller

Tomorrow, the Supreme Court will hear oral arguments in DC v. Heller, a gun control case that will call on it to interpret the Constitution's controversial Second Amendment:

A well regulated militia, being necessary to the security of a free state, the right of the people to keep and bear arms, shall not be infringed.

A group of linguistics and English professors have submitted a friend of the court brief ("the linguists' brief") on behalf of petitioner, the District of Columbia, in which they analyze the likely original meaning of the amendment. DC is trying to maintain what is in effect a ban on handguns (and in fact, a requirement to register handguns in a District where the means to register them don't exist). I don't know how frequently linguists submit briefs to courts on issues of interpretation, and I don't know whether judges find them influential. I certainly think courts should welcome briefs by linguists on issues of language, to help check bad judicial assumptions and reasoning about meaning and language.

The big issue appears to be how to interpret the amendment in light of the relationship between the absolute clause and the main clause of the sentence. Mark Liberman posted about this issue on the Language Log a couple of months ago. There seems to be widespread agreement about how to analyze the sentence syntactically: the sentence divides into two, with the first part ending in "state" and the second part starting with "the right of the people." The first part of the sentence is what's known as an "absolute clause," and modifies the second part, which is the main clause.

There is one exception to the syntactic consensus, apparently not taken seriously by anybody, based on the unusual punctuation in the amendment. The Language Log post references a group of anti-gun academics who have argued, based on the first comma, that "a well armed militia" is the subject of the sentence. Everyone else seems to think that the first and third commas should be ignored. The linguists' brief suggests that in the eighteenth century, punctuation was more a matter of style than grammar, adding that people were taught to put a comma where there ought to be a pause for breath. [They also point to awkward comma insertions elsewhere in the Constitution, including the 26th Amendment (1971) and the 27th Amendment (1992, but drafted a couple of centuries earlier)].

There also appears to be general agreement on the semantic analysis of the sentence, in which the absolute clause, the first part of the sentence, states a principle that is the cause for the principle stated in the main clause. The discussion in the amicus brief suggests that absolute clauses in which the verb has the -ing suffix usually have a causal relationship when their verb is stative, and a timing relationship when their verb is active. For example, in "the ship having arrived, we all embarked," the absolute clause indicates the time on which the embarking took place, i.e. the time of the ship's arrival. I would add from my recollection of Quirk et al's writings on absolute constructions that sometimes the semantic relationship is conditional. "Standing on the chair, John can reach the ceiling" means that if John stands on the chair, he can reach the ceiling. [EDIT March 28: the more appropriate reference is to Gregory Stump's work. See Barbara H. Partee's comment.]

An exception to the semantic consensus comes from Nelson Lund, who has a forthcoming article in the George Mason University Civil Rights Law Journal. According to the linguists' brief, Lund takes the position that because the absolute clause is grammatically independent of the main clause and especially because the main clause is a command, there is no semantic relationship between the absolute and the main clause. The authors of the brief show this conclusion to be empirically wrong.

As a side note, sentences with syntactic imperative force (as opposed to declaratives with an instruction or command illocutionary force, like the Second Amendment) are at best very awkward when modified by absolute constructions:

??A well regulated militia being necessary to the security of a free state, do not infringe on the right of the people to keep and bear arms.

The awkwardness is not semantic; a different modifier clause used to convey the same meaning is fine:

Since a well regulated militia is necessary to the security of a free state, do not infringe on the right of the people to keep and bear arms.

This shows that there is no rule against a command being semantically modified by a clause stating a reason or cause, though there may be syntactic restrictions involving particular reason/cause arguments. [Having written a masters thesis on the syntax of imperatives, I feel fairly confident saying that the linguistics community does not have a good understanding of the syntax of modifiers in imperative sentences; at least, it did not have a good understanding of it nine years ago when I wrote the thesis, and no facts suggesting that the understanding has improved have come to my attention since then.]

Leaving aside Lund and the anti-gun academics, the main point of disagreement seems not to be a question of linguistics but of rules for interpretation: how should the Court interpret the law in light of the agreed-upon syntax and semantics? Is it, as the DC Circuit Court of Appeals found, a general individual right to own guns, in spite of the narrower purpose articulated in the absolute clause? Or is it not an individual right but one restricted to those belonging to well-regulated militias, as most other federal circuits have held?

Besides discussing constitutional punctuation and the semantics of the absolute construction, the linguists' brief takes the following positions:

- The "well regulated" modifier was consciously used to distinguish between militias, and in particular between the more ad hoc kind of militia, consisting of ordinary citizens, referenced elsewhere in the Constitution, and the well regulated militia, which was a force trained and disciplined by the state. This addresses the DC Circuit Court of Appeals' critical holding that the Second Amendment applied to all citizens who were subject to being organized by the states into militias, as opposed to actually being in a militia. If the framers had meant the right to extend to all individuals subject to being organized into militias, it would make little sense for them to use the restrictive phrase "well organized."

- "Bear arms" is, and was at the time, an idiom meaning to fight as a soldier. But the idiomatic meaning can be suspended in favor of a more general "carry weapons" meaning if additional language is added. The linguists' brief cites the following language as an example:

“bear arms for the defence of themselves and their own state, or the United States, or for the purpose of killing game”

They add, citing eighteenth century authority, that "arms" generally meant objects used in war, such as firearms and swords, while "weapons" excluded firearms and included objects used for offensive purposes on other occasions.

The linguists' brief's overall conclusion: the DC Circuit Court of Appeals erred by failing to give the amendment its most natural interpretation, which is that people have the right to serve in a well-regulated militia and to keep arms for that purpose. To reach its conclusion that there is a general right to own weapons, the appeals court ignored the words "well-regulated" and wrongly interpreted "bear arms" as "carry weapons," when in fact it is an idiom meaning "engage in hostilities." Accordingly, the Supreme Court should reverse the appeals court.


bhp1 said...

Hi, Uri!
You wrote "I would add from my recollection of Quirk et al's writings on absolute constructions that sometimes the semantic relationship is conditional. "Standing on the chair, John can reach the ceiling" means that if John stands on the chair, he can reach the ceiling." I think the best source on this is Greg Stump's dissertation and his published book, Stump, Gregory. 1985. The Semantic Variability of Absolute Constructions : Synthese Language Library. Dordrecht: Reidel. He's the one who discovered that when the absolute phrase involves an "individual-level" predicate, roughly one that is atemporal (like "a well-regulated militia being necessary ..."), it's interpreted as 'factive' (presupposed to be true), and its relation to the main clause can be any reasonably inferrable relation that can hold between two true propositions ('because', 'in spite of' (!), ..., or simply 'and' plus "relevance"); whereas when absolute contains a "stage-level" predicate ('standing on a chair', 'with that coat on', 'having arrived early', ....), then the absolute phrase serves to restrict some temporal or modal or other operator explicit or implicit in the main clause.
-- None of this affects your observations, which I agree with. I just wanted to add that reference. Stump's work has been discussed in some of Angelika Kratzer's work on stage-level and individual-level predicates (that distinction first came from Greg Carlson's dissertation), and in some of mine. More references on request.
-- best, Barbara

Uri said...

Thanks for the reference and the summary, Barbara. Stump's work was what I was hoping to cite, but working from memory and having never actually read Stump, I couldn't remember the proper reference and cited Quirk et al, which I actually looked at at some point.

price per head service said...

Please one more post about that.I wonder how you got so good. This is really a fascinating blog, lots of stuff that I can get into. One thing I just want to say is that your Blog is so perfect