BRENDAN SMIALOWSKI/AFP/Getty

Tallying up the lengthy list of constitutional harms done by the Trump administration’s extraordinary deployment of National Guard troops and other federal agents to disrupt peaceful protests in Washington, D.C., earlier this month, it is hard not to include the fear that this president has set yet another precedent that could warp the nature of the American presidency. After all, what past presidents have done unquestionably shapes the way the presidency runs today, both in setting norms of day-to-day operations inside the executive branch, and in shaping how the public understands what behavior counts as “presidential.” Indeed, the Supreme Court has noted more than once that interpreting the words executive power under Article II of the Constitution without recognizing “the gloss which life has written upon them” would be a mistake. What presidents do can change the way the people and public officials alike understand what the Constitution permits. Do the recent events at Lafayette Square now count as part of the canon of presidential precedent?

The good news, such as it is, is that Trump’s actions need not necessarily have any such effect. From President Franklin D. Roosevelt’s executive order targeting Japanese Americans during World War II to President Richard Nixon’s decision to fire the special Watergate counsel, there is what we might call an executive-branch “anticanon” of sorts—conduct that has come to be so widely recognized as unacceptable in character, it has not produced any of the common precedential effects. No court relies on it as evidence of constitutional meaning. And neither presidents nor the public sees it as belonging to the defining norms of presidential behavior, save perhaps as part of a case study illustrating how no American president should behave.

So how does some presidential conduct become part of the anticanon, or more to the present point, how might Americans today help ensure that the attack at Lafayette Square comes to join it? As it turns out, unlike the decisions of the Supreme Court, presidential precedent is less born than made—or unmade, as the anticanon shows. And there is more than one way to unmake it. Formal legal rejection, for example, is not required. President Roosevelt’s Japanese American exclusion order was welcomed at the time with broad popular support, congressional inaction, and indeed formal approval by the Supreme Court. Nearly 30 years passed before Congress enacted a law aimed at preventing the future military detention of Americans in similar circumstances, closer to 40 before the civil-rights activist Fred Korematsu personally saw his record cleared, and more than 70 before the Supreme Court formally repudiated its initial decision (in 2018). But whether measured by public officials’ and scholars’ negative references or by the generations of American schoolchildren assigned to read Farewell to Manzanar, race-based internment had clearly entered the executive-branch anticanon long before then.

For the particular conduct to have resulted in punishment is also not necessary. The Nixon administration’s elaborate efforts to keep secret from Congress its air-bombing campaign in Cambodia—including falsifying military records—was the subject of intense constitutional criticism when it finally came to light. But the House Judiciary Committee tried and failed to include the air campaign among the articles of impeachment against Nixon. Instead, Congress enacted a new law, over the president’s veto, making clear that presidents were required to report to Congress all commitments of American forces abroad.  

Just more than a decade later, the Reagan administration attempted a similar brand of foreign-policy concealment, trading arms for hostages and funneling funds to a military venture in Nicaragua. This time, the resulting scandal triggered investigations engaging the public and all three branches of government. The special counsel’s investigation produced more than a dozen criminal indictments; Congress eventually passed legislation that further expanded reporting requirements; and the otherwise popular president saw a record drop in public approval. Significantly, President Reagan didn’t cite Nixon’s Cambodia precedent in his defense. Instead, he announced the creation of his own investigating commission, and later apologized. The Nixon crew had been spared direct punishment for Cambodia. But the presidential practice of hiding foreign military ventures from Congress had become anticanonical nonetheless.

So what relegates something to the anticanon? The most significant factor is that multiple and diverse actors, in government and society, publicly object to the original deed. Congress could take disciplinary action in real time, but it more commonly enacts contrary legislation after the fact. Both parties come to favor nominees for executive-branch positions who condemn the practice—and Senate confirmation committees insist on it.  As important, institutions responsible for professional training and public education, and leaders who shape popular discourse, support and reinforce official condemnation. Long before they find themselves in Cabinet positions seeking the advice of government counsel, government officials see the wrongness of the practice in the cultural landmarks that still define civic life for large majorities of Americans.  

Because presidential practice shapes the public’s understanding of the presidency inside and outside the executive branch, the most effective route to anticanonical status is not one that counts on some decisive blow by a single authoritative institution at once, but one that aims to engage the full range of public and institutional voices over time. Nixon’s effort to disrupt the investigation into his own corruption by having the Watergate counsel Archibald Cox fired by then–Solicitor General Robert Bork was never held unconstitutional by any court. (Bork himself was later nominated to the Supreme Court.) Yet the anticanonical status of that behavior has proved remarkably stable, even in Trump’s extraordinary presidency. Long before the administration’s latest embarrassment surrounding its thwarted attempt to replace the U.S. Attorney for the Southern District of New York with a more beholden successor, Trump’s former White House counsel Donald McGahn risked his own job in resisting Trump’s demands to remove the independent counsel Robert Mueller. McGahn resisted not because the Court had held such removals unlawful, but rather because, as McGahn later put it to investigators, he did not want to be like “Saturday Night Massacre Bork.”

For this reason, Americans should welcome and amplify the outpouring of regret and criticism from retired and even active-duty military leaders in the aftermath of the Lafayette Square debacle, many publicly reaffirming their commitment to keeping the military out of partisan politics, with some writing eloquently about the nature of the military’s oath to defend the Constitution, not the president. It is equally important that Congress succeed in its efforts to block the Pentagon from using funding and personnel against peaceful protesters in the future. And it will be essential to insist that future Defense Department nominees make their views on these events clear, and that future executive-branch officials across the board are oriented to their job not just with briefings on ethics and classification rules, but with constitutional literacy enough to understand how the presidency is supposed to function in times of public unrest.

Concern about the potentially precedent-setting consequences of the Trump presidency is understandable, and entirely right. But assuming the precedent is already set, or waiting for a single institutional solution, is a mistake. Above all, such expectations risk effectively absolving Congress, the courts, and the American people of a responsibility that will soon be more theirs than Trump’s.

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.