Drafters can learn three things from courts screwing up analysis of ambiguity. First, you learn what ambiguous text looks like. Second, you learn that fights over ambiguous contract language are messy and expensive. And third, you learn that too many courts are incapable of analyzing ambiguous contract language in a way that makes sense; see this 2020 blog post for more about that.
Yesterday the U.S. Supreme Court issued its opinion in Facebook v. Duguid. Unfortunately, it allows us to learn, yet again, the third lesson.
This case involves syntactic ambiguity, namely uncertainty over what part of a sentence a phrase modifies, or what part of a phrase a word modifies. I have no interest in the details, because in this 2020 blog post I’ve already analyzed the language at issue. What’s of interest to me is the confusion, not who ultimately wins.
But I couldn’t help but notice this passage:
Under conventional rules of grammar, “[w]hen there is a straightforward, parallel construction that involves all nouns or verbs in a series,” a modifier at the end of the list “normally applies to the entire series.” A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 147 (2012) (Scalia & Garner) (quotation modified).
What the court is referring to is rules of interpretation, also referred to, more grandly, as “canons of construction.” They offer quick-and-dirty ways for courts to short-circuit deliberations over ambiguity. The one thing they aren’t is “rules of grammar,” and those courts that can’t tell the difference are demonstrating that they don’t know grammar and linguistics.
O shock, o horror! See the spectacle of Sotomayor trying to shoehorn everything into a Procrustean bed of construction canons, with Alito jibing her about needing to see things in context!
I was just amused to see Sotamayor citing to Scalia.
1/ The canons of construction often overlap rules of grammar, as in treating ‘and’ conjunctively and ‘or’ disjunctively.
2/ In principle, it’s no more wrong to resort to principles of interpretation to resolve ambiguities than to use dictionaries, experts, parol evidence, or background law other than the canons. The real problem with the canons is that for every canon that says widgets are white, there’s another that says widgets are black (I exaggerate), and there are no ‘meta-canons’ to say which canon applies when two or more conflict. If some judges wrongly apply the canons in aid of result-oriented analysis, shame on them, but it’s not the fault of the canons. Also, it’s hard to gin up sympathy for the parties who say the court got it wrong when the parties are the architects of their own predicaments. An ounce of prevention….
They overlap, except when they don’t!
Rules of interpretation (I can’t abide to preposterous canons) are arbitrary. They’re expedient, but they ain’t pretty.
To be fair, the court didn’t rely solely on the series-qualifier “cannon;” it also looked at the legislative context, noting that it was the use of random-number generators that congress was concerned with and noting that, under the plaintiff’s interpretation, any cell-phone speed-dial function would run afoul of the law. It seems as though it should have started with that analysis before diving into the linguistic weeds.
The problem, of course, is that the statute’s language is gibberish, and it can’t be parsed in a way that makes sense. How does one store a number that isn’t produced? How is the storage of that number accomplished by a random number generator?
The courts are hesitant to call out crappy legislative drafting and instead resort to things like official sounding cannons of construction to make it seem like the legislature knew what it was doing when it half-assed a statute. Argument over cannons is often nothing more than arguing over which shade of lipstick looks best on the pig.