Supreme Court

Textualism, Title VII, and "Discrimination . . . Because of Such Individual's Sex"

Does the text of Title VII prohibit discrimination on the basis of sexual orientation or transgender status?

|The Volokh Conspiracy |

At the beginning of the term, the Supreme Court heard oral argument in a trio of cases considering whether Title VII of the Civil Rights Act prohibits employment discrimination on the basis of sexual orientation or transgender status. We are still waiting for the opinions in these cases.

Much commentary on these cases noted that the interpretive method generally favored by the court's conservatives (textualism) would seem to produce a "liberal" outcome (such discrimination is prohibited by Title VII), whereas consideration of purpose, legislative intent and legislative history would produce the more "conservative" outcome (such discrimination is not prohibited).

The textualist argument was made well by Professor Pamela Karlan at oral argument:

When a employer fires a male employee for dating men but does not fire female employees who date men, he violates Title VII. The employer has, in the words of Section 703(a), discriminated against the man because he treats that man worse than women who want to do the same thing. And that discrimination is because of sex, again in the words of Section 703(a), because the adverse employment action is based on the male employee's failure to conform to a particular expectation about how men should behave; namely, that men should be attracted only to women and not to men.

Consideration of Congressional purpose and intent, however, shows that sex discrimination and sexual orientation discrimination were understood as different things. Moreover, some would argue it is implausible that Congress sought to prohibit sexual orientation discrimination at a time when homosexual activity was illegal in much of the United States and was a basis for expulsion from the military.

Contrary to the conventional way these cases are often presented, James Phillips, a fellow at the Stanford Constitutional Law Center, argues that there is textual evidence supporting the employers' position that the prohibition of "discrimination . . . because of such individual's sex" does not cover sexual orientation discrimination. Here is the abstract of his short essay:

This short essay focuses on a linguistic (and therefore textualist) principle overlooked in the trio of Title VII cases currently before the U.S. Supreme Court: compositionality. By taking that principle seriously–a principle the Supreme Court has recognized in some form for a century–the essay uncovers corpus linguistic and dictionary evidence contemporaneous to the enactment of Title VII that sheds light on the relevant statutory language. That light provides an answer to the interpretive question the Court is facing.

Title VII makes it unlawful "to discriminate against any individual . . . because of such individual's . . . sex." Many observers seem to drop the word "against" and focus just on "discriminate." Or if they do take "against" into account, they nonetheless fail to read the whole operative statutory phrase—including "discriminate," "against," and the relevant trait (in this case, sex)—as an indivisible whole. The argument for the plaintiffs, in particular, would require us to give "discriminate" and "against" the meaning each of them would have if it existed apart from the rest of the phrase. This "dissection" approach is most obvious in the most precise and careful formulations of the plaintiffs' central textualist argument, as this Essay will show.

But that approach, as I will also show, violates the linguistic principle of compositionality. So it produces a demonstrably inferior reading of the text on purely linguistic grounds, prior to any appeal to subjective intent, purpose, policy concerns, or other modes of legal argument.

That is because, as it turns out, the phrase "discriminate against . . . because of [some trait]" was a linguistic unit (a composite) by the time of Title VII's enactment, which makes the principle of compositionality relevant. And read as a composite, the phrase had more semantic content than one could glean from separately analyzing and then amalgamating its three parts ("discriminate," "against," and "sex"). While a "dissection" reading might suggest that Title VII covers any adverse treatment that even adverts to sex, as plaintiffs suppose, a linguistically superior reading (taking compositionality into account) proves that the operative text refers only to adverse treatment that rests on prejudice or bias—i.e., unfair beliefs or attitudes directed at some or all men in particular, or at some or all women in particular (whether the beliefs be outright falsehoods or just unduly rough or weak generalizations; and whether the attitudes be indifference, discounting of interests, distaste, or outright antipathy). And this defeats the plaintiffs' textualist argument. Whatever the legal merits of their case overall, their textualist case fails because it violates a basic linguistic principle as applied to linguistic data from the era. But as it happens, the prejudice-based conception of discrimination that is required by the text properly read, also fits well with all the Court's precedents on sex discrimination.

To establish the above points about the text, this Essay appeals only to (1) dictionaries from the time of Title VII's enactment, and (2) systematic data on linguistic usage from the same period (which this Essay draws from the Corpus of Historical American English, containing 24 million words' worth of naturally occurring text that provides a balanced snapshot of American English usage).

Ryan Nees, a rising 3L at Stanford Law School, takes issue with Phillips account, and has authored an extensive post on the blog of the American Constitution Society. Here is a taste, summarizing the response:

Phillips's intriguing argument is worth carefully considering, and is an instructive application of the larger project urging judges to make greater use of corpus linguistics to derive contemporaneous meaning. He argues, in summary, that the petitioners' textualist reading depends upon "separately analyzing and then amalgamating . . . three parts" – namely, the words "discriminate," "against," and "sex." Analyzed in that way, the LGBTQ petitioners may have a point, and  Justice Gorsuch seemed to acknowledge as much at oral argument. The better approach, Phillips says, would be to assess the distinct meaning of the phrase "discriminate against," which has its own highly specific connotation entailing prejudice as a motivation, especially when the phrase is paired with a suspect class.

Corpus linguistics could be a useful resource to identify idiomatic meaning of this sort. But the surprising consequence is that, the more idiomatic the meaning uncovered, the more purposivist the reasoning starts to appear. And for reasons I explain, Phillips's analysis is unpersuasive on its own terms, relying on a limited linguistic corpus while overlooking any possible legalistic meaning that prevailed at the time. It seeks to create a new term of art even as textualists have traditionally disfavored doing so. As the corpus-linguistic method rises in popularity as a tool of statutory interpretation, the Title VII case study demonstrates its risks.

Phillips, for his part, has posted a reply on the Originalism Blog.

NEXT: Is the Stock Market's Response to COVID-19 Optimistic or Just Plain Nuts?

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. Accepting Karlan’s view that Title VII requires “unfair beliefs or attitudes directed at some or all men in particular, or at some or all women in particular,” perhaps the Harris funeral home should lose because they had an unfair belief directed at Stephens about which gender biological men should present and behave as?

  2. If the law treats a man who hires a male employee worse than a man who hires a female employee, is it discriminating against the man on the basis of sex?

    What’s the analytical difference between dating behavior and hiring behavior? Why should the two be treated differently?

    1. If the law treats a man who hires a male employee worse than a man who hires a female employee, is it discriminating against the man on the basis of sex?

      The premise is flawed, because Title VII indicates an employer — not a man — doing the hiring.

      1. I don’t think one can get rid of this argument this easily. One could similarly argue that the beneficiary is an employee, not a woman.

        Employers can also be beneficiaries. The same definition of discrimination also applies to government agencies, including government agencies dealing with employers. For example, courts.

        In that context, employers are beneficiaries. And when they are, their sex (and hence sex-stereotyped behavior), and discrimination based solely on it, is relevant.

  3. I’m a ‘linguist’ — at least my PhD says so. My sense is that corpus linguistics is good stuff, very useful for understanding original intent. There are contexts where corpus linguistics statistics can be fairly convincing as evidence.
    But as I read Adler’s citation of Phillips, I immediately had the sense that it was probably too cute by half, because he’s deliberately pairing ‘discriminate against’ as a possible compound. I would have to see a whole lot of references to uses of ‘discriminate’ in the absence of ‘against’, before I would accept statistical evidence that ‘discriminating against’ is different from an ordinary language meaning of ‘discriminate’. That’s just my intuition and hunches talking. We need to look at those, before we attach significance to ‘discriminate against.’
    The obvious use for corpus linguistics in this case is the question of what ‘sex’ means — as Adler notes in his opening. And so I applaud the student Nees for taking up the challenge. But since IANAL I won’t try to follow all the pleadings, but will be interested to a summary later.

  4. This is all quite interesting, and the Court has to do something, but I really wish it was explicitly prefixed (as opposed to implicitly):

    The Court has to interpret Title VII as written, which presents questions of construction and language. Regardless of how we rule, the best outcome is for Congress to amend Title VII to unambiguously endorse or reject the plaintiff’s claims.

    1. the best outcome is for Congress to amend Title VII to unambiguously endorse or reject the plaintiff’s claims.

      That would involve Congress doing its job rather than hiding and letting courts deal with it, and any anger it generates.

      1. Sad but true. The Court can’t make Congress do its job, but it can call them out over it.

    2. Why would that be the best outcome? If Congress agrees that the court has interpreted the statute correctly, what is the purpose of passing legislation saying so?

      1. Because it puts the law’s actual implementation on firmer democratic grounds, and it obviates the non-stop complaining of the parties on what the statute really means.

        Consider Mass v EPA. If Congress wrote unambiguously: “CO2 can be considered a harmful pollutant under 42 U.S.C. § 7521(a)(1)” that would be far better for us as a republic than basing the EPAs regulatory authority on an ambiguous law from 1963, well before climate change was a thing.

        And, FWIW, if I were a Senator I would vote in favor of such a law. But I would likewise consider that if Congress passed the inverse law (“CO2 cannot …”) and I voted against it and it passed anyway, it would still be better than the status quo IMHO. The legislature ought to legislate, not force the courts to parse minutiae.

  5. Roberts will declare “sex” as it pertains to Title VII as a “tax”.

  6. “But in focusing on this point, Mr. Nees’s analysis overlooks the usefulness of a corpus: the ability to drill down on context.”

    Is that posh for: if I can fix the reference class, I can show pretty much whatever I want?

    As shown in his own example further down:

    ”In COHA, the word appears 237 times per million words. In the Supreme Court corpus Mr. Nees used, “court” appears 80,447 times per million words, or 339 times more frequently than in the ordinary corpus. But no one would think this shows that “court” has a special legal meaning that diverges from the ordinary meaning.”

    So court jesters are barristers, Hampton court is a place where justice is dispersed, and your court yard is where the defendant waits?

  7. How would Phillips’ analysis weigh in a typical race-conscious or sex-conscious affirmative action case? Is any such affirmative action acceptable because it is not based on unfair beliefs or attitudes about whites or men (or whatever the disadvantaged group is) as a whole?

    1. If affirmative action is not permissible, it is because of the 14th Amendment’s guarantee of equal protection. Unlike in Title VII, there is no mention of “discriminated against.”

  8. >When a employer fires a male employee for dating men but does not fire female employees who date men, he violates Title VII. The employer has, in the words of Section 703(a), discriminated against the man because he treats that man worse than women who want to do the same thing.

    You’re giving this argument credit as “well stated”? This argument is ridiculous. It’s clearly describing behavior-based discrimination, not sex-based discrimination. The policy can result in both men and women being hired, as well as the declining to hire of both men and women. In other words, there is no discrimination based on sex. The restricted behavior is equally restricted for both men and women.

    Secondly, even if that were what the law stated, it would be unconstitutional, as it would clearly conflict with religious liberties. Last time I checked, the constitution outranked statutes passed by congress.

    1. The restricted behavior is equally restricted for both men and women.

      This argument was considered and rejected in Loving v. Virginia (1967). It is unwise to attempt to resurrect a losing one here.

      1. Loving v. Virginia struck down an unconstitutional law imposed by the state. Here we are referring to the actions of private citizens.

        1. I suppose you’re welcome to try it again then. We’ll see how it goes.

  9. I think transgender discrimination is an even stronger argument for coverage because, even if a biological male who identifies as female is considered a man, he is still being discriminated based on gender stereotypes because he’s fired for identifying as a female.

    The sexual orientation argument has some logic (a man who sleeps with a man is treated differently from a woman who sleeps with a man), but orientation is arguably broader than such conduct and includes men who would prefer to sleep with men but haven’t actually slept with anyone. It would be odd if Title VII covers one group and not the other.

  10. The timing of this is very suspicious. Did Gorsuch’s vote leak?

  11. Oh, and by the way:

    “adverse treatment that rests on prejudice or bias—i.e., unfair beliefs or attitudes directed at some or all men in particular, or at some or all women in particular (whether the beliefs be outright falsehoods or just unduly rough or weak generalizations; and whether the attitudes be indifference, discounting of interests, distaste, or outright antipathy). And this defeats the plaintiffs’ textualist argument.”

    Does anyone seriously believe that this sort of discrimination is not motivated by antipathy towards some women, i.e., trans women?

Please to post comments

Comments are closed.