“As societies grow decadent, the language grows decadent, too,” Gore Vidal wrote in 1986. “Words are used to disguise, not to illuminate, action: you liberate a city by destroying it.”In recent days, the Court’s conservatives have issued one ambitious opinion after another. They expanded President Trump’s powers to fire independent regulators, rescind deportation protections, and turn away asylum seekers; weakened state authority to enact gun control; narrowed the ability of religious minorities to vindicate their free-exercise rights; eroded the due-process rights of green-card holders; and handed big wins to multinational oil and tech companies.Yet anyone not paying close attention would likely miss the Court’s radicalism. The justices’ language in most cases obscured their opinions’ effects; the word decadent fits. Using invocations of precedent to disguise rather than illuminate, the conservative justices pretend to preserve what they are overturning.This duality—sweeping remaking of law presented as continuity—has become a hallmark of the Roberts Court. Precedent matters. The idea is so axiomatic to the legal system that stating this risks condescension. But the basics are worth restating: Precedent—and the legal doctrine of following it, what scholars and judges call stare decisis—constrains a given judge’s discretion. It also fosters predictability, fairness, and stability in the legal system, allowing society to order its affairs with some confidence about the law.[Adam Serwer: The Court that will believe anything is ‘race-neutral’]When the Court does overrule precedent, it is a big deal, as in yesterday’s decision in Trump v Slaughter. The opinion officially overturned Humphrey’s Executor, a 90-year-old case. But the separation-of-powers practice formalized in Humphrey’s Executor goes back at least 50 years before the Court decided it. Relying on Humphrey’s allowed Congress to build the modern federal government, insulating agencies from the whiplash of electoral politics and channeling bipartisanship, expertise, and continuity into technical policy making. These agencies will now be upended in ways their creators never anticipated.Precedent, of course, is not a magical constraint that produces mechanical results. A better way to think of it, as the philosopher Ronald Dworkin put it, is as a “gravitational pull” on judicial reasoning. Respecting it grounds judges in traditions larger than themselves and guards against overreach.Following Humphrey’s Executor served those values faithfully, its overturning reflects the triumph of something else: faddish theories of the moment—here, the theory of the unitary executive—and the hubris of a Court that thinks it knows best. The Court has replaced the gravitational pull of accumulated wisdom with the gravitational pull of itself.In one respect, though, Slaughter stands apart from what is now standard practice. The Roberts majority said the word overrule. It owned its jolt.That was not true in many other significant cases this term. Consider Cisco Systems, Inc. v. Doe, a case brought by followers of Falun Gong—a spiritual movement persecuted by the Communist Party of China—who alleged that Cisco had violated the Alien Tort Statute by helping Beijing surveil and suppress them. The ATS allows federal courts to hear cases brought by foreign nationals for noncriminal harms committed in violation of international law. Congress passed it at the turn of the 18th century precisely to furnish a concrete remedy for such claims.The task before the Court should have been easy. In a 2004 opinion, Sosa v. Alvarez-Machain, the Court had reaffirmed long-standing precedent that the ATS conferred a limited but real cause of action for claims grounded in established international-law norms. The Ninth Circuit had allowed the Falun Gong plaintiffs’ suit to proceed on this basis. There was no need for the Supreme Court to grant review of the case and disturb the active legal proceedings below.[Listen: How to save the Supreme Court from itself]Instead, the majority took the case up and used it to all but eviscerate the ATS, barring claims of this type and therefore making it far more difficult for foreigners to bring human-rights lawsuits in U.S. courts. It did so by citing Sosa—but rather than apply the majority opinion in good faith, the Court turned to Justice Antonin Scalia’s concurring opinion. Scalia had agreed with the majority’s bottom line in Sosa, but the meat of his opinion rejected the majority’s reasoning, arguing that federal courts should be closed to ATS claims except those predicated on three 18th-century wrongs. Scalia’s colleagues in the majority rejected his position, calling it “particularly unconvincing in light of what we know about congressional understanding bearing on this issue.”Yet the Cisco majority makes this “unconvincing” position the new legal rule and does so without acknowledging it, framing the adoption of Scalia’s argument as a straightforward and required application of existing law. Scalia’s “criticisms,” the opinion notes, have “resonated” with Republican-appointed justices in later cases, warranting closing “the door that Sosa cracked to judicially created ATS liability” in Cisco. In words, Sosa and the ATS remain viable. In practice, a precedent has been overturned, and the statute has been gutted. As Justice Sonia Sotomayor observed in her dissent, Cisco marks a “low point in this Court’s esteem for its precedents,” with the Court overruling Sosa “without even acknowledging that it is doing so.”This is not how courts are supposed to operate. A roving tribunal rummaging through past decisions for minority views it likes, then reviving them while studiously refusing to admit what it has done, is not acting like a legitimate court. It is aggrandizing power from the past judges whose reasoning it discards and from future judges who will be bound by its decisions should they take stare decisis seriously.The form of precedent subversion displayed in Cisco would be troubling enough if it were a one-off. It is not. Earlier this year, the Court used the same technique in Callais v. Louisiana, which dismantled the enforcement apparatus of Section 2 of the Voting Rights Act against racial discrimination in electoral maps. Justice Samuel Alito’s majority opinion took his own dissent from Allen v. Milligan—a case decided just three years ago—and turned it into the law of the land, writing that “a vast social change” warranted narrowing Section 2 to elusive circumstances where plaintiffs can show a “strong inference” of intentional discrimination.Alito characterized this constriction as an “updated” application of Thornburg v. Gingles, a 1986 decision that has governed Section 2 litigation for nearly four decades. Alito did not say that he was overturning Gingles, even as he “updated” it into nonexistence. The supposed realignment has rendered Section 2 lawsuits—suits to create majority-minority opportunity districts—impossible to sustain. Gingles was not formally overturned; it is now a dead letter.Alito also cited Rucho v. Common Cause, a 5–4 decision split along ideological lines holding that federal courts cannot hear challenges to partisan gerrymandering. The decision drew immediate criticism for placing beyond judicial reach a practice widely condemned as antidemocratic by allowing politicians to entrench themselves in power by choosing their voters rather than the reverse. The Rucho majority seemed to foresee the weight of that criticism: It went out of its way to acknowledge that partisan gerrymandering was nevertheless incompatible with democratic principle, as if to soften the blow of locking in the practice.Callais reinvents that concession to justify its evisceration of the Voting Rights Act. It cites Rucho not only for the legal proposition that federal courts cannot hear partisan gerrymandering disputes, but as authority for the proposition that partisan gerrymandering enjoys constitutional protective status—that states have an affirmative right to manipulate electoral maps for partisan purposes. The two readings are irreconcilable: Rucho explicitly rejected as contemptible what Callais now presents as a constitutional entitlement.[Leah Litman: The Supreme Court’s era of meaningless rights]With both Gingles and Rucho, the Court has bent live precedents to meanings they cannot bear. Describing the creation of entirely new rules as the compelled requirements of existing ones, the Court’s conservatives have pushed precedents into meanings they cannot sustain—at least not in any intellectually honest reading. Twist a precedent far enough from its original meaning, and it becomes a different kind of tool entirely: no longer a constraint on what the Court can do, but a cover story for what it has already decided to do.The costs of decisions such as Slaughter, Cisco, and Callais are enormous on their own terms. But there is a further cost that goes beyond any particular decision. American courts are, or at least are supposed to be, passive institutions: They wait for disputes to come to them and decide the cases before them. The Supreme Court has extraordinary power in our constitutional democracy in part based on the premise that it operates this way—not as a superlegislature armed with an ideological to-do list, reasoning toward it and filling in the legal details afterward. When a court works backward from preferred conclusions, it violates that underlying bargain.The language of precedent is a mechanism through which courts earn public trust, demonstrating the regularity and fairness of its operating procedures. When the Court says, We are bound by prior decisions, and here is how they apply even if we would have reached a different result originally, it subordinates its own preferences to a rule it did not make and cannot change on a whim.When a court carelessly overturns the precedents of the past, it is not doing that. And when it pretends to follow precedent while covertly replacing it, it hollows out this legitimating function entirely.The conservative majority is wagering that as long as the opinions are long enough, cite enough cases, and avoid explicitly saying We are overturning this too frequently, the public will not notice—or at least will notice less.But you cannot liberate a city by destroying it, whatever you call what you are doing. Words cannot cover up the rubble.