Jack Balkin: jackbalkin at yahoo.com
Bruce Ackerman bruce.ackerman at yale.edu
Ian Ayres ian.ayres at yale.edu
Mary Dudziak mary.l.dudziak at emory.edu
Joey Fishkin joey.fishkin at gmail.com
Heather Gerken heather.gerken at yale.edu
Abbe Gluck abbe.gluck at yale.edu
Mark Graber mgraber at law.umaryland.edu
Stephen Griffin sgriffin at tulane.edu
Bernard Harcourt harcourt at uchicago.edu
Scott Horton shorto at law.columbia.edu
Andrew Koppelman akoppelman at law.northwestern.edu
Marty Lederman msl46 at law.georgetown.edu
Sanford Levinson slevinson at law.utexas.edu
David Luban david.luban at gmail.com
Gerard Magliocca gmaglioc at iupui.edu
Jason Mazzone mazzonej at illinois.edu
Linda McClain lmcclain at bu.edu
John Mikhail mikhail at law.georgetown.edu
Frank Pasquale pasquale.frank at gmail.com
Nate Persily npersily at gmail.com
Michael Stokes Paulsen michaelstokespaulsen at gmail.com
Deborah Pearlstein dpearlst at princeton.edu
Rick Pildes rick.pildes at nyu.edu
Richard Primus raprimus at umich.edu
K. Sabeel Rahmansabeel.rahman at brooklaw.edu
Alice Ristroph alice.ristroph at shu.edu
Neil Siegel siegel at law.duke.edu
Brian Tamanaha btamanaha at wulaw.wustl.edu
Mark Tushnet mtushnet at law.harvard.edu
Adam Winkler winkler at ucla.edu
Why the Supreme Court Should Stop Fetishizing Dictionaries and Start Caring About Words
Alison L. LaCroix and Jason Merchant
As the current Supreme Court Term draws to a close, the decision in one of the most closely watched cases of the year, NLRB v. Noel Canning, remains to be announced.
The case concerns the president’s recess appointment power under Article II, sec. 2, cl. 3 of the Constitution. The Recess Appointments Clause states:
The President shall have Power to fill up all Vacancies that may happen during the Recess of the Senate, by granting Commissions which shall expire at the End of their next Session.
The specific issue in the case is the validity of an NLRB decision against Noel Canning. The company argued that the NLRB lacked constitutional authority to issue the order because three members of the NLRB were not properly appointed under one of a variety of theories: (1) the Senate was not in recess when they were appointed, so the Recess Appointments Clause did not apply; (2) the vacancies they filled did not originate during a recess, so the Recess Appointments Clause does not apply; or (3) when the appointments were made, the Senate was meeting in pro form session, so the Recess Appointments Clause did not apply. The case thus raises classic president-versus-Congress separation of powers questions, with a customary soupçon of politics thrown in, insofar as the recess appointments power is widely understood as a limited basis for presidents to install their own appointees in the face of Senate opposition.
But NLRB v. Noel Canning also presents a useful test case for a new approach to legal interpretation based on linguistic analysis. In an ongoing project, we are developing a theory and a methodology of historical semantics and legal interpretation. The methodology brings together research in historical jurisprudence and in theoretical and computational linguistics in order to understand the meanings of words and phrases in context. Across the spectrum of constitutional interpretation, theories from originalism to textualism to common law constitutionalism, require the interpreter to articulate which written sources, and which words, count for purposes of determining constitutional meaning. At the same time, the Supreme Court justices seem more interested than ever in using dictionaries to elucidate the meanings of constitutional and statutory texts.
Our historical semantics approach is valuable for two reasons. First, it demonstrates that much of what passes for linguistic analysis on the part of judges and advocates is incomplete at best, and pseudo-scientific at worst. We can rely neither on modern intuitions nor on valuable but limited technological tools, such as the Google Ngram Corpus, to reveal what the founders would have viewed as common usage. We cannot rely on our intuitions, or on our algorithms, to read back to the eighteenth century. Carefully applied, however, linguistic analysis can tell us which meanings were possible and probable at a given time, and what their relative distribution and frequency were – and most of the time, many more meanings were possible than a quick glance through a single dictionary would suggest. While less able to tell us categorically that a given meaning was not possible at a given time, careful linguistic analysis can nonetheless guide a sophisticated interpreter to likely possibilities, and away from highly or more unlikely ones.
To see how our methodology can help guide legal analysis, consider the key issues in the recess appointments case:
1. “The” versus “a” recess.
Noel Canning argued that the Senate was not in recess when three of the NLRB members were appointed, so the Recess Appointments Clause did not apply. Failing proper appointment, the NLRB lacked a quorum to act, and therefore the order to the company was invalid. At the appellate level, the D.C. Circuit agreed with Noel Canning and held that the Senate was not in recess at the time of the appointments. Writing for court, Judge Sentelle rejected the NLRB’s argument that the president may use the recess appointments power during recesses that fall within the Senate’s official sessions. “As a matter of cold, unadorned logic,” Judge Sentelle wrote, “it makes no sense to adopt the Board’s proposition that when the Framers said ‘the Recess,’ what they really meant was ‘a recess.’” For the D.C. Circuit, the recess appointment power comes into play only when the Senate is in “the Recess,” defined as “the period between sessions of the Senate.”
To support this proposition, the D.C. Circuit cited Samuel Johnson’s Dictionary of the English Language from 1755, which defined “the” as “the article noting a particular thing.” But even so estimable a work as Johnson’s Dictionary is not a terribly reliable or insightful source into the meaning and usage of “the” in 1755. Actual usage, especially of very frequent and flexible words such “the” (itself the most frequent word in our corpora), is much more difficult to delineate than Johnson’s short paraphrase would indicate. Johnson does not discuss, nor does Judge Sentelle mention, the fact that the definite article “the” also has and had a generic use, as in “The lion is a fearsome hunter”, where no particular individual lion is intended, but rather the class of lions—any lion. Likewise for the use of “the” in expressions such as “During the day, I wear glasses,” where again no specific day is picked out or intended: rather a generic reference to any day is made.
Worse still, our modern intuitions about usage do not apply well to even the same words used in an 18th century context: our and previous research shows that use of the definite article has changed over the course of the history of the English language. In particular, generic uses of the definite article “the” were relatively more common in the 18th century than they are today. The frequency of expressions such as “without the consent of” (relative to its counterpart with the article, “without consent of”), or “without the knowledge of” (versus “without knowledge of”), has been in steady decline over the past two centuries. An eighteenth century writer was much more likely to use a “the” in such contexts than a writer would today. The founders may well have used “in the recess” generically, even though we no longer do so today; indeed, “in recess” was barely used at that time. Notwithstanding its confidence, the D.C. Circuit thus made an analytical error in 2013 in relying on Johnson’s descriptively inadequate definition from 1755.
2. Does “during the recess” modify “fill up” or “happen”?
The clause applies to “all Vacancies that may happen during the Recess of the Senate.” Noel Canning argued that the relevant NLRB vacancies did not originate during a recess, and therefore that the Recess Appointments Clause does not apply.
This question points to what linguists term attachment ambiguity: uncertainty as to where the prepositional phrase “during the recess” attaches within the sentence, and therefore what phrase it modifies. Does it attach to, and hence modify, the verb phrase headed by “fill up,” or to that headed by “happen”? Much psycholinguistic research has shown that hearers and readers processing such sentences show a preference for taking such phrases to attach to the verb phrase headed by the most recently encountered verb – in this case, to that headed by “happen.” Linguistic analysis thus points toward the conclusion that the vacancies themselves must happen during the recess (whatever “the recess” is – see above), a conclusion consonant with the unambiguous attachment of the similar phrase in Art. I sec. 3.2 (“and if Vacancies happen by Resignation, or otherwise, during the Recess of the Legislature . . .”).
A final question then arises: what is the status of “happen”?
3. “Happen”: an ongoing state or an inchoative?
At oral argument, several justices focused on the meaning of “happen” – what does it mean for a vacancy to happen during the recess? Must it begin during the recess, or must it only be occurring during the recess? In linguistic terms, the question is whether “happen” refers to a state or activity (which can be presented as an ongoing eventuality) or to an accomplishment or an achievement (which necessarily encode endpoints). Justice Kagan invoked the Oxford English Dictionary. Justice Breyer pressed Solicitor General Verrilli to provide an example in which the word “vacancy” was used in an ongoing, stative sense. Verrilli did not offer an example for “vacancy,” but Justice Breyer did point to the word “battle” as a possible subject for a stative usage of “happen.” “Of course battles happen,” Justice Breyer observed. “That’s because battles happen over time.” The conclusion we draw, following much linguistic work on aspectual classes, is that the status of the verb “happen” depends both on the nature of its subject and on the aspect in which it occurs. In sentences such as “While the celebrations were happening, the thieves broke into the back office,” the celebrations were ongoing, a background state framing the time of the event of the break-in. In the text of Article II, sec. 2, the simple aspect of “happen” used in the modal relative clause does not permit a definitive reading: as a matter of linguistic analysis, either reading is possible.
4. The payoffs.
What are the payoffs of the historical semantics approach? For the specific case of recess appointments, our research suggests that the phrase “the recess” was not as narrow as modern intuitions would suggest. For constitutional interpretation more broadly, the point is that one must look to other sources besides the text, including historical practice and linguistic context. At what point must legal interpreters acknowledge vagueness? Even if one is a strict textualist, the answer cannot be “never.” Dictionaries appear to supply certain answers. As the oral argument in Noel Canning suggests, the justices want dictionaries to function in this way. But even Dr. Johnson’s Dictionary cannot bear this weight. Instead, interpreters must have knowledge of what the terms and phrases are a proxy for, and in what context they emerged and developed, and apply a judicious understanding of their relative usages. Historical semantics provides a means of doing just that: using linguistic analysis, not just dictionaries, to understand the range of possible meanings available at the founding, and at other important moments of law creation. To analogize to another interdisciplinary methodology, doing textual analysis with only dictionaries is like doing law and economics with only a pocket calculator.
Alison L. LaCroix is Professor of Law and Ludwig and Hilde Wolf Teaching Scholar, and an associate member of the Department of History, at the University of Chicago. Jason Merchant is Professor of Linguistics and Deputy Dean for Languages at the University of Chicago.