In the Spring 2018 edition of Judicature, Bryan Garner, an old friend, responded to my article in the previous issue,[1] an article that took the form of a mock opinion by Kimble, J., in Lockhart v. United States.[2] He wrote his own mock opinion, with an introduction criticizing mine.[3]
Simply put, Garner and I disagree on whether canons can dispose of cases such as Lockhart with clarity and concision. In my view, they cannot — as the conflicting opinions in Lockhart demonstrated. The case rested on a syntactic ambiguity caused by a modifier following a three-part series. In his own opinion, Garner relies on the series-qualifier canon — which favors Lockhart — as “more directly applicable” than the last- antecedent canon because the ambiguous language involves a “straightforward, parallel construction that involves all nouns . . . in a series.”[4] But he begs the question when he calls the construction “straightforward” and “parallel.” (I realize that a judge writing an opinion can state conclusions as declarations.)
In fact, much of the debate between the majority and dissenting opinions in Lockhart was, in a sense, over how straightforward and parallel the items in the series are. The series, without the trailing modifier: aggravated sexual abuse, sexual abuse, or abusive sexual conduct. Yes, you have three nouns, but the first and third have two modifiers in front, and the second has one. And the justices disagreed on whether the trailing modifier — involving a minor or ward — applied to all three items or only the last one.
Justice Sotomayor, for the majority, invoked the rule of the last antecedent. She said that “it takes more than a little mental energy to process the individual entries in the list, making it a heavy lift to carry the modifier across them all.”[5] She followed with a hypothetical example from baseball: “imagine you are the general manager of the Yankees and . . . tell your scouts to find a defensive catcher, a quick-footed shortstop, or a pitcher from last year’s World Champion Kansas City Royals.”[6] The natural reading, she said, would limit the trailing modifier’s reach to the final item, pitcher.
Justice Kagan, dissenting along with Justice Breyer, countered with these two examples: an actor, director, or producer involved with the new Star Wars movie and a house, condo, or apartment in New York.[7] Here, she insisted, the modifiers apply to all the previous items as a matter of ordinary English, which the series-qualifier canon reflects. But in contrast to Justice Sotomayor’s example, none of the nouns in Justice Kagan’s series had modifiers in addition to the ambiguous trailing one.
Justice Kagan discounted the baseball example above as “not parallel” because pitcher does not have a modifier of its own, as the other two items do.[8] In response, Justice Sotomayor offered the example of a friend’s asking you to get tart lemons, sour lemons, or sour fruit from Mexico, and said you “would be forgiven” for thinking you could bring back lemons from California.[9] Justice Kagan, disagreeing, said there would be “no doubt” that “[y]our friend wants some produce from Mexico.”[10] Thus, the justices were of different minds on a parallel series with a single preceding modifier.
Now, perhaps, you’ll see why in my opinion I provided several baseball examples, taking off from Justice Sotomayor’s baseball example, to show how intuition, common knowledge, and slight variations can affect the possible meaning. Figuring out when to apply the series-qualifier canon is not always as simple as you might think. My sense (without any empirical evidence) is that trailing modifiers can be somewhat dicier than leading modifiers to begin with. A series with internal modifiers can be dicier still: consider adults and young children who are healthy or versatile infielders and durable catchers who can hit.
At any rate, the majority in Lockhart rejected the series-qualifier canon in favor of the last-antecedent canon, which Garner asserts is less applicable because the construction at issue “doesn’t involve a ‘pronoun, relative pronoun, or demonstrative adjective’ — in short, no word has a grammatical antecedent.”[11] He knows, though, that “last antecedent” has become the catchall name that courts use and apply even when the trailing modifier doesn’t technically involve an antecedent. And because I think the canon still gets more credence than it deserves in resolving ambiguity, I revisited the Court’s previous use of it, after having said in my introduction that “readers will notice [in my opinion] an uncommon candor and willingness to consider scholarly opinion . . . .”[12]
Neither syntactic canon was the clear winner in Lockhart. If you were merely choosing between them, you might well side with Justice Kagan (and Garner). But either way, you ought to acknowledge and address plausible contrary arguments. What’s more, you ought to recognize that picking between those canons was not all there was to the disposition. That was the target of my mock opinion: decision by canon alone.
To repeat: Garner (or anybody else) can write a short opinion declaring that the canons, as he and Justice Scalia describe them in their book Reading Law, resolve the case. But his opinion exudes a confidence and certainty that is unjustified. The same goes for textualism in general.[13]
As for the rule of lenity in criminal cases, he is of course right that it’s one of the canons included in Reading Law. When I said in my subtitle that “Canons Are Not the Key,” I meant the syntactic canons at the heart of the case. Commentators had, after all, billed Lockhart as a contest between dueling canons.[14]
Beyond that, Reading Law’s criterion for invoking lenity is this: “whether, after all the legitimate tools of interpretation have been applied, a reasonable doubt persists.”[15] I consider legislative history a legitimate tool of interpretation,[16] and so did the justices in Lockhart.[17] I consider a statute’s framework — its place in the scheme of related statutes — a legitimate tool of interpretation, and so did the justices in Lockhart. Garner’s opinion ignores all this. It says: “If an ambiguity in a criminal statute is genuinely debatable, the defendant wins.” But his opinion (favoring Lockhart) does not go to the trouble of exploring the ambiguity. Except for dismissing the last-antecedent canon in one sentence, the opinion does not touch on any of the arguments that actually led to the government’s winning, including intuitive arguments.[18]
A last point about the case, for anyone who still thinks that Lockhart should have won easily on linguistic grounds alone. Of the five circuits that had explicitly addressed the issue, none applied the trailing modifier to the entire series; they came to the same conclusion — for a similar mix of reasons — as the Supreme Court majority.[19]
Garner criticizes my opinion as “dictum-filled” and “bloated with hand-wringing dicta that only obscure the law.”[20] He points out that his is “some 88 percent shorter.” Shorter, yes, mainly because it pronounces one canon as controlling and (as I just mentioned) says nothing at all about contrary arguments. Decision by fiat, you might say.[21] Underlying my opinion were obvious rhetorical and jurisprudential purposes: to register the importance of considerations that are not strictly textual; to call attention to the inherent dangers of modifiers with a series; to cast doubt on the strength of the last-antecedent and surplusage canons; to trace the history and highlight the subtleties of the series-qualifier canon; and to address courts’ differing definitions of ambiguity, as well as the Supreme Court’s own less-than-consistent standards for invoking lenity. I wasn’t just deciding; I was trying to face these issues head-on, even while expressing points of view.
In the introduction to my opinion, I called it “a flight of fancy.”[22] However unlikely its style or even its content may have been, I believe this: decision- making generally demands more than an exercise in parsing.
Footnotes:
[1] Joseph Kimble, How Lockhart Should Have Been Decided (Canons Are Not the Key), 101 Judicature 40 (Winter 2017).
[2] 136 S. Ct. 958 (2016).
[3] Bryan A. Garner, How Lockhart Really Should Have Been Decided: Canons of Construction Are Key, 102 Judicature 56 (Spring 2018).
[4] Id. at 57 (quoting Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 147 (2012)).
[5] Lockhart, 136 S. Ct. at 963.
[6] Id.
[7] Id. at 969; see also id. at 972 n.2, setting out five more examples, from cases, of similarly uncomplicated series.
[8]Id. at 970 n.1.
[9] Id. at 966.
[10] Id. at 972.
[11] Garner, 102 Judicature at 57 (quoting Reading Law at 144).
[12] Kimble, 101 Judicature at 41.
[13] See Joseph Kimble, What the Michigan Supreme Court Wrought in the Name of Textualism and Plain Meaning: A Study of Cases Overruled, 2000–2015, 62 Wayne L. Rev. 347, 376 (2017) (“Textualists exaggerate the number of appellate cases in which the text alone yields a singular or self-evident meaning. They figure that if they study hard enough all the various and often conflicting textual clues, they will discover the intended meaning. And they largely discount the value of intuition, common sense, legislative history, real-world consequences, and sensible policy in deciding cases.” (citation omitted)).
[14] See, e.g., Neal Goldfarb, Coming to SCOTUS: Battle of the Dueling Interpretive Canons, LAWnLinguistics.com (Oct. 27, 2015) (URL omitted).
[15] Reading Law at 299 (citation omitted).
[16] See Joseph Kimble, The Doctrine of the Last Antecedent, the Example in Barnhart, Why Both Are Weak, and How Textualism Postures, 16 Scribes J. Legal Writing 5, 37–41 (2014– 2015) (listing 12 arguments for considering legislative history).
[17] But see Digital Realty Trust, Inc. v. Somers, 138 S. Ct. 767, 783–84 (2018) (Thomas, J., with Alito & Gorsuch, JJ., concurring in part and concurring in the judgment) (rejecting the majority’s use of legislative history).
[18] See, e.g., Lockhart, 136 S. Ct. at 964 (“We therefore see no reason to interpret § 2252(b)(2) so that ‘[s]exual abuse’ that occurs in the Second Circuit courthouse triggers the sentence enhancement but ‘sexual abuse’ that occurs next door in the Manhattan municipal building does not.”).
[19] United States v. Mateen, 764 F.3d 627, 631–32 (6th Cir. 2014); United States v. Lockhart, 749 F.3d 148, 151–56 (2d Cir. 2014); United States v. Spence, 661 F.3d 194, 197 (4th Cir. 2011); United States v. Hubbard, 480 F.3d 341, 350 (5th Cir. 2007); United States v. Rezin, 322 F.3d 443, 447–48 (7th Cir. 2003); cf. United States v. Hunter, 505 F.3d 829, 831 (8th Cir. 2007) (assuming the contrary without discussion); United States v. McCutchen, 419 F.3d 1122, 1125 (10th Cir. 2005) (same).
[20] Garner, 102 Judicature at 57.
[21] See Bryan A. Garner, Interview with Justice Elena Kagan (Part 4) (video) (URL omitted) (so describing an opinion that, among other things, does not “respond[] to the arguments that you think are losers” and thus disregards the Court’s responsibility to “tak[e] all the parties seriously,” as well as “to show the American public . . . how we reason about cases”).
[22] Kimble, 101 Judicature at 41.
