channarongsds / Getty / The Atlantic

Justice Neil Gorsuch is a proud textualist. According to this approach, what Congress intended, or expected, when it passed a law doesn’t matter. What matters are the words printed on paper. In practice, Justice Gorsuch will strictly follow the text of statutes, no matter what result it yields. Last month, he decided that the 1964 Civil Rights Act has always prohibited LGBTQ discrimination. Everyone had simply missed it for half a century. And at the close of the Court’s term, he determined that an 1833 treaty between the federal government and American Indian tribes was never formally rescinded. Who knew that eastern Oklahoma has been Indian Country all along?

In both cases, Justice Gorsuch insisted he was sticking to the text, the whole text, and nothing but the text. Alas, he wasn’t. His interpretation was shaded by the work of justices who had not been so careful about text. And in both cases, Justice Gorsuch failed to acknowledge that the Court’s precedents were inconsistent with textualism. In doing so, he inadvertently undermined textualism’s justification. One can’t profess to follow the original meaning of a text while in fact following precedents that ignored that meaning. Going forward, he should criticize prior decisions that failed to take text seriously, and either reluctantly follow them, or formally abandon them.

The first of these two decisions was Bostock v. Clayton County. Here, the Court split 6–3. Justice Gorsuch wrote the majority opinion, which was joined by Chief Justice John Roberts, and the Court’s four progressives. Gorsuch parsed the text of Title VII of the Civil Rights Act of 1964. This statute made it unlawful for employers to “discriminate against” employees “because of … sex.” Justice Gorsuch did not begin his analysis by interpreting this text on a blank slate. Rather, he simply assumed that decades of case law had accurately interpreted the crucial phrase discriminate against because of sex. Indeed, he treated decades of precedent as part of the “law’s ordinary meaning” in 1964. This approach built an elaborate textualist framework on quicksand.

All nine justices agree that the Court must determine the meaning of a statute, regardless of whether that meaning leads to a policy that is unwise or unjust. In this case, my view is that the phrase “discriminate against” must inform the meaning of “because of.” In the 1960s, that phrase formed a single linguistic unit. Recently, a biopic was made about the life of Ruth Bader Ginsburg. Its title, On the Basis of Sex, was an homage to the widespread understanding of what it meant to discriminate against a woman on the basis of sex. When these elements are combined, the phrase discriminate against because of sex references discrimination based on bias or prejudice about a person’s sex, and not discrimination based on sexual orientation or gender identity. Justice Gorsuch relied on decades of precedents that focused primarily on “because of” and not the entire clause. Doing so eliminated the requirement that some sort of bias or prejudice exists based on a person’s sex. In effect, he severed the statute in half, and concluded that if sex plays any role in the discrimination, the employers’ actions are unlawful. The Georgetown University law professor Randy Barnett and I have described his approach as “halfway textualism.” Gorsuch followed old precedents narrowing a small component of the statute, even when those cases are at odds with the meaning of the entire statute.

The decision in the second case, McGirt v. Oklahoma, was split 5–4, with Justice Gorsuch joined by the four progressive justices. This case considered whether Congress had formally “disestablished” an American Indian reservation that covers half of Oklahoma. No federal law dictates what precise steps are needed to disestablish a reservation. Thus, textualists have no relevant statute to parse. Instead, the courts have incrementally developed a century of case law about how Congress can eliminate American Indian sovereignty over territory. That framework, alas, is not itself textualist. The Court has never set out some magic words that Congress must utter to wind down tribal authority. Instead, the Court’s approach considers many factors that, when viewed in context, reveal an intent to disestablish the reservation.

Chief Justice Roberts, who dissented in McGirt, described the Court’s “well-settled” approach: “We determine whether Congress intended to disestablish a reservation by examining the relevant Acts of Congress and ‘all the [surrounding] circumstances,’ including the ‘contemporaneous and subsequent understanding of the status of the reservation.’” The Court has acknowledged that “explicit language” is not needed for a finding of disestablishment. Roberts explained, “The appropriate inquiry does not focus on the statutory text alone.” Without question, the Court has adopted a framework that favors Congress, to the detriment of the tribes. These precedents put a thumb on the scale of disestablishment.

This sort of fluid approach is, no doubt, a bitter pill for textualists to swallow. So in McGirt, Justice Gorsuch simply spit it out. Unlike in Bostock, Justice Gorsuch refused to treat the Court’s non-textualist precedents concerning Indian territory as part of the “law’s ordinary meaning.” He did not approach Congress’s entire body of work as the Court has instructed. Over the course of many years, Congress diminished the tribes’ authority, and established a commission to bring the territory under the jurisdiction of the state of Oklahoma. But Justice Gorsuch deemed this evidence too fragmented to establish a unified congressional intent. Rather, he inspected individual congressional actions that concerned the territory in a fragmented, balkanized fashion. Unsurprisingly, Congress did not meet his novel standard for disestablishment. As a result, Justice Gorsuch found that Congress’s 1833 promise to the tribes had not been explicitly repealed, and remained in effect. Congress hadn’t said the magic words. And how could it? Until McGirt, no one knew the precise textual standard that was needed to disestablish a reservation. In this case, Justice Gorsuch’s halfway textualism has literally cut Oklahoma in half.

Let’s put these two decisions in perspective. In Bostock, Justice Gorsuch quietly accepted precedent that paid little attention to text. In McGirt, he quietly rejected precedent that paid little attention to text. In both cases, he erected elaborate textualist structures on top of a foundation well worn by the Court’s prior decisions. And in neither case did he acknowledge the relationship between precedent and textualism. In doing so, I believe the justice erred.

Generally, the Supreme Court will follow the doctrine known as stare decisis, which is Latin for “to stand by things decided.” (I say generally because justices of all stripes can always muster sufficient justification to overrule old decisions.) Textualists, particularly those of an originalist bent, have an especially tough time with stare decisis. In many instances, the text of a statute, or the Constitution, has a meaning that conflicts with the Court’s long-standing interpretation. Or the Court has adopted a method of reading a specific statute that requires consideration of subjective extrinsic factors, such as legislative history or policy concerns. What is a textualist to do? There are three general approaches.

First, the textualist can acknowledge the conflict between text and precedent, but maintain that stare decisis compels a result that conflicts with textualism’s basic principles. This approach has the virtue of humility, even if it reaches the “wrong” result, at least by a textualist standard. Chief Justice Roberts spoke to this method in his McGirt dissent: “Unless the Court is prepared to overrule these [old] precedents, it should follow them.”

Second, the textualist can overrule the non-textualist precedent in order to enforce the statute’s natural sense. Here, stare decisis gives way to textualism. This approach has a significant downside, however: It disturbs arrangements that the people and government have come to rely on. Yet at least it has the virtue of candor: The jurist can explain the why and how of a departure from settled law. And I find this sort of honesty refreshing. Justice Clarence Thomas is the member of the Court most likely to take Door No. 2. (Though in McGirt, he joined the chief justice behind Door No. 1.)

Justice Gorsuch, however, favored Door No. 3 in Bostock and McGirt. He professed to apply a form of unadulterated textualism. But he failed to account for contrary precedent. In Bostock, he quietly baked into his analysis decisions from the 1980s and ’90s that were hardly textualist. And in McGirt, he demanded a level of textual precision from Congress that had never been demanded before.

Repeating, over and over again, that Congress can amend the statute if it disagrees with the Court’s decision is not enough. Of course it can. But this argument goes only so far. Congress has been operating under certain presumptions for decades; it thought the scope of Title VII and the boundaries of Oklahoma had been settled long ago. But Justice Gorsuch maintains that everyone was wrong about Title VII for five decades, and that everyone was wrong about eastern Oklahoma for a century. At least in the unique context of Indian law, the Court had established a “well-settled” method of reading tribal law. And this framework is itself subject to stare decisis. (The Court has treated century-old antitrust laws in this common-law fashion.) Departures from that methods are permitted, but should be addressed.

Finally, I am not sanguine about the prospects for the future of textualism. Justices Ruth Bader Ginsburg, Stephen Breyer, Sonia Sotomayor, and Elena Kagan are savvy enough to recognize the limits of Bostock and McGirt. The quartet recognizes full well that Gorsuch’s opinions leave enough wiggle room to avoid a strict textualist holding in the future. They are in no sense tied to the mast of textualism. They are free to depart as needed.

But I don’t think Justice Gorsuch is trying to lock them in. Indeed, I have a newfound respect for him. In these cases, he was utterly disinterested in how his opinions would be received. He doesn’t care. There is no political calculation. There is no long game. There is no three-dimensional chess. Bostock dropped a thermonuclear bomb on the conservative legal movement, and Justice Gorsuch was excoriated by many of his supporters who felt betrayed. The first Trump appointee, no political slouch, must have anticipated that possible reaction. But he followed his principles. I commend his temerity.

My disagreement, then, is methodological. Textualism, like originalism, must start from the blank slate of a statute, without regard to how the Court has interpreted that statute in the past. Justice Gorsuch cannot begin from the 50-yard line. He must start from his own end zone. In its present form, Justice Gorsuch’s textualism is far too fragmented to form a coherent jurisprudence. In the future, he must grapple with the interplay between stare decisis and textualism. When feasible, he should choose Door No. 2, and reject precedents that ignored textualism. If that approach is not viable, he should stay behind Door No. 1, and at least cast doubt on why that precedent is flawed, but follow it anyway. But Door No. 3 is misleading. It preaches textualism, but practices precedentialism. This approach, in the long run, will serve only to undermine textualism. If Justice Gorsuch wants to move the law away from nebulous, flimsy reasoning toward more textualist, neutral principles, he must account for both text and precedent.   

This story is part of the project “The Battle for the Constitution,” in partnership with the National Constitution Center.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.