Manslaughter, Violent Crimes, and Legal Theory: Corpus Linguistics Arrives at the Second Circuit

In corpus linguistics by James Heilpern2 Comments

This past week, corpus linguistics made an unexpected appearance in Judge Menashi’s concurring opinion in United States v. Scott, an important en banc case out of the Second Circuit which asked whether first-degree manslaughter in New York constitutes a “violent crime” for sentencing purposes under the federal Armed Career Criminal Act.  I say unexpected because Judge Menashi did not use corpus linguistics to actually interpret the ACCA, but to instead show why his colleagues in the majority were incorrect to characterize their interpretation as being based on the statute’s “ordinary meaning.” In doing so, he demonstrated a point I made in my last post: that corpus linguistics is forcing judges to confront (and plug) glaring holes in our modern legal theory.

But first, some necessary background . . .

For a state crime to be considered “violent” under the ACCA, it must have “as an element the use, attempted use, or threatened use of physical force against the person of another.” In making this determination, courts don’t look at what the defendant actually did (so the fact that Scott shot one victim in the head at point-blank range and stabbed another is irrelevant), but the minimum he could have done and still been convicted of the same crime. Under New York law, a person can commit first-degree manslaughter through an act of omission, so long as (1) he had the requisite intent and (2) that act of omission actually resulted in someone’s death.  But when that happens, is the act of omission itself a “use . . . of physical force”?

A majority of the Second Circuit said no. Relying on United States v. Smith, the Court concluded that “the ordinary, natural, everyday meaning of the word use requires only that the person make use of the violent force, convert such force to one’s service, employ it, avail oneself of it, utilize it, carry out a purpose or action by means of it, or derive service from it” (cleaned up). The majority then drew strong parallels to United States v. Castleman, where the Supreme Court held that the “knowing or intentional causation of bodily injury necessarily involves the use of physical force” to support this broad interpretation.

Judge Menashi’s concurrence

Judge Menashi agreed that first-degree manslaughter was a violent crime but took umbrage with the majority’s characterization of its interpretation as the ordinary, everyday meaning of use of physical force. Instead, he claimed that this was the phrase’s specialized or legal meaning, mandated by precedent and the common law, not how Americans typically understand the phrase. To demonstrate this, he relied on corpus linguistics: “The Corpus of Contemporary American English—the most widely used corpus of American English—contains forty-seven non-specialist instances of ‘use of physical force’ and all refer to physical contact; none plausibly refer to ‘deriv[ing] service from’ a preexisting physical force.” His use of corpus linguistics demonstrates how judges are using this methodology to help our legal theory evolve: He first uses corpus linguistics to show that the court’s interpretation is not the ordinary meaning and then argues that it is only by “[r]ecognizing the distinction between ordinary and legal meaning [that] . . . the absurdity or inconsistent results that sometimes afflict applications of the categorical approach.”

In fact, Judge Menashi’s concurrence highlights at least three other underdeveloped legal questions:

  • What is the relevant search term? The majority (like the Supreme Court in Smith) focuses principally on the meaning of the word “use.” But Judge Menashi focuses on the full phrase “use of physical force.” That decision makes a difference. A phrase is often more than the sum of its parts. Dictionaries and other traditional tools of statutory interpretation can’t show that empirically. A corpus can. But how do we decide what the relevant phrase is?
  • What is the “ordinary meaning” of “ordinary meaning”? Law and corpus linguistic scholars have pointed out that there are at least five variations of the ordinary meaning canon currently being used by federal and state judges. It’s the danger of shared vocabulary. No one even realized that judges were talking past each other on this point until we started to and quantify the determination. Judge Menashi makes clear that he thinks ordinary meaning means prototypical meaning. Regardless of whether you agree or not, the fact that he clearly defines his terms is a welcome development. To my knowledge, this is the first opinion to respond to this research and clearly stake out a position on the actual meaning of ordinary meaning.
  • When does a term become a legal term of art? The majority’s reliance on Smith is a little disturbing. “Use” is one of the 100 most common words in the English language, and as such can take a wide-variety of meanings depending on context. Yet, based on Smith the majority concludes that whenever “Congress employs the word ‘use’ in a statute, its intent is to sweep broadly and not to cabin legislation only to those uses that most immediately come to mind or that manifest a defendant’s active, i.e., physical, use.” (cleaned up). Judge Menashi appears to agree that this is the result mandated by precedent, but essentially argues that use of physical force has become a term of art. But are we comfortable with a rule that says that when a word—especially a malleable word like use—is interpreted by a court of last resort, that word must bear that meaning for all statutes in the code, regardless of the subject matter or the possibility of linguistic drift?  Doesn’t that impede the common man’s ability to understand the law?

I’m sure there are others theory questions implicated by this case that I haven’t thought of. But bottom line, I think this case is good example of what I think will be corpus linguistics’ largest contribution in the long run: helping judges refine their jurisprudence and trigger needed development in our overall legal theory.


  1. Looks like you linked to the panel opinion in Scott, rather than the en banc.

Leave a Comment