The Moral of a Recent Second Circuit Opinion: Don’t Rely on Commas for Disambiguation

Via this post by Ray Ward I learned of the recent opinion of the Second Circuit Court of Appeals in AIG v. Bank of America (PDF copy here). In the opinion, the court states that whether a modifying phrase following a list of nouns or phrases modifies each item on the entire list, or only the last item, depends on whether the modifying phrase is separated from the last item by a comma.

I won’t get into the statutory language at issue. Instead, I’ll limit myself to the general principle, as stated by the court:

We address first the grammar. The quotation from Barnhart on which Defendants rely does not fully state the principle of construction. The Barnhart opinion, immediately following the sentence quoted by the Defendants, cites and quotes from the Sutherland treatise on statutory construction. See 2A N. Singer, Sutherland on Statutory Construction § 47.33, p 369 (6th rev. ed. 2000). The statement in the treatise, on which the Supreme Court relied, is more qualified and nuanced than the statement the Defendants quote from the text of the Barnhart opinion. The treatise says, “Referential and qualifying words and phrases, where no contrary intention appears, refer solely to the last antecedent.” See Barnhart at 26 (emphasis added).

One of the methods by which a writer indicates whether a modifier that follows a list of nouns or phrases is intended to modify the entire list, or only the immediate antecedent, is by punctuation—specifically by whether the list is separated from the subsequent modifier by a comma. When there is no comma, as in the statute considered in Barnhart, the subsequent modifier is ordinarily understood to apply only to its last antecedent. When a comma is included, as in the Edge Act provision, the modifier is generally understood to apply to the entire series. See, e.g., Sir Ernest Gowers, Fowler’s Modern English Usage 587-88 (2d ed. 1965) (explaining that in the sentence “French, German, Italian, and Spanish, in particular are taught,” the insertion of the comma at the end of the list “show[s] that in particular relates to all four languages and not to Spanish only”); see also Kahn Lucas Lancaster, Inc. v. Lark Int’l Ltd., 186 F.3d 210, 215 (2d Cir. 1999) (“When a modifier is set off from a series of antecedents by a comma, the modifier should be read to apply to each of those antecedents.”), abrogated on other grounds by Sarhank Grp. v. Oracle Corp., 404 F.3d 657 (2d Cir. 2005). For example, the statement, “This basketball team has a seven-foot center, a huge power forward, and two large guards, who do spectacular dunks,” differs from the statement, “This basketball team has a seven-foot center, a huge power forward, and two large guards who do spectacular dunks.” The first statement conveys that all four players do spectacular dunks. The latter statement conveys that only the guards do so.

This analysis exhibits confusion of the sort to which judges are prone: they confuse grammar with arbitrary principles of interpretation. The former involves how people actually write; the latter are driven by expediency. The Second Circuit invokes grammar, but the distinction it draws exhibits the arbitrariness of a principle of interpretation.

In particular, I don’t share the Second Circuit’s confidence that anyone writing, or indeed reading, the two “basketball team” examples would intend, or derive, the alternative meanings offered by the Second Circuit. It’s unrealistic to expect writers and readers to have such a firm distinction in mind. That was the point I made in my affidavit in the Canadian “case of the million-dollar comma.” (Go here for my article about that dispute, but note that MSCD contains my updated analysis of the comma issue.)

That said, my beef isn’t with the courts—it’s drafters who create the confusion, leaving judges to clean up the mess. If you want to avoid having a court apply an arbitrary principle of interpretation when interpreting your contract, make sure you avoid the sort of syntactic ambiguity at issue in AIG v. Bank of America. In particular, if presence or absence of a comma might affect meaning, restructure the sentence. For example, if only the two large guards do spectacular dunks, refer to them and their dunks at the beginning of the sentence. For more detailed guidance, see MSCD chapter 12.

About the author

Ken Adams is the leading authority on how to say clearly whatever you want to say in a contract. He’s author of A Manual of Style for Contract Drafting, and he offers online and in-person training around the world. He’s also chief content officer of LegalSifter, Inc., a company that combines artificial intelligence and expertise to assist with review of contracts.

8 thoughts on “The Moral of a Recent Second Circuit Opinion: Don’t Rely on Commas for Disambiguation”

  1. To say nothing of the fact that “grammar experts” can also be mistaken. In the Gowers example, there should have been a comma after “in particular.” Would that have thrown the court into a tizzy?

    Just as the drafters of the Constitution, in the words of O.W. Holmes, did not enact Herbert Spencer’s Social Statics, drafters of statutes are not enacting Fowler’s Modern English Usage (which, incidentally, is about British, not American, usage).

    • A comma following “in particular” would have refocused that modifier onto Spanish, only, whereas without the comma, the modifier does seem to apply to all four languages.

      • Perhaps this illustrates why using commas for this purpose is dangerous – I don’t see that it has that meaning at all. I don’t think most writers would use it in that way, and I don’t think most readers would interpret it that way. If that is the case, courts surely have no business doing so.

  2. I agree that, for drafting purposes, what matters is to avoid the issue in the first place. Having said that, I think it is very problematic for courts to be using obscure grammatical notions to try to determine the intention of the parties, even if this is done objectively. They have little or no value for these purposes, and someone needs to be pointing this out.

    It is acceptable to counter a specious argument by saying “95 times out of 100, this construction is used to mean X”, because that refers to the likelihood that a construction is being used in a certain way. However, ideas like the “rule” of the last antecedent, and the “rule” referred to in this case, do not reflect what people use in practice, but what someone once considered people should use. “Do” and “should” are not the same thing, and “should” has nothing to do with objectively determining intention.

  3. Ken, I can’t help but think that the US habit of jamming things up into large rectangular blocks of text bears a good part of the shame. This is one of the main benefits of the ‘multi-level indentation’ that is much more common in English and Aussie drafting. By stretching things out, it becomes graphically obvious what the final modifier’s antecedents are.

    • Scott: I note in MSCD that you can use tabulation to eliminate syntactic ambiguity, but I prefer other methods. I use tabulation only when a sentence contains enough separate concepts that breaking them up makes them easier to read. Ken

  4. I agree that the simplest – and best – solution is to avoid the problem in the first place by better drafting; however, I disagree with the point that a comma doesn’t change meaning. I certainly read the basketball players sentence as stated by the court: the second statement definitely limits the spectacular dunking to the two large guards. I would just as certainly expect (a) the writer to understand the difference and (b) a judge to give the difference meaning.
    As for what writers and readers should be expected to have in mind, I would not expect the writer of a local sports column, or a casual reader thereof, to spot the difference; I would, however, expect a highly paid firm of attorneys drafting an agreement, and an experienced judge interpreting that agreement, to understand and appreciate the difference.

    • O: The fact that I don’t share your confidence in the meaning conveyed might shake that confidence a bit. Ultimately, something means what readers think it means.

      There are some nuances that I didn’t get into. I might write a short article about this.



Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.