Was this petty, snarling man, utterly lacking in presidential temperament and heedless of the established rules and norms of politics, going to get us into a nuclear war? That’s what a lot of Americans wondered about Richard Nixon back in 1956. Chosen as Dwight Eisenhower’s running-mate four years before, Nixon was seen as unscrupulous, mean-spirited and reckless in smearing his opponents. So in the 1956 campaign, with Ike still wildly popular, Democrats made Nixon’s character an issue. The Soviet Union had just detonated its first hydrogen bomb, and Eisenhower had recently suffered a heart attack and a bout of ileitis (a gastrointestinal illness)—putting Nixon, they warned, perilously close to the Oval Office. In the fall Adlai Stevenson, the Democrats’ nominee, invoked what the liberal New York Post columnist Max Lerner called “the triple issue”: should Eisenhower suddenly die, the unreliable Nixon would have his finger on the button.

Today, of course, we’re wondering the same things about Donald Trump, whose recent tweets about big nuclear buttons and his own “stability” have pundits buzzing about his mental health. His behavior—impulsive, erratic, petty and often cruel (to name just a few of the most salient qualities)—especially amid the prospect of a nuclear showdown with North Korea, makes it impossible not to think about his psychology, just as it was during the Cold War.

It’s an understandable impulse: The public needs to understand what moves our leaders to act as they do, including not just “rational” motives—pursuing policy objectives, responding to political pressures, working within economic and international constraints—but also the drives and responses that operate deep in an individual’s unconscious, formed by experiences and relationships long ago.

But more than a century after Sigmund Freud revolutionized the understanding of the human mind, the use of psychology to understand our political leaders has failed to realize its promise. The diagnoses we’re now throwing around for Trump, from incipient senility or venereal disease, are simplistic, facile and unhelpful—and possibly fueled by politics rather than objective analysis. As crucial as it is to recognize the psychological dimension of Trump’s behavior, history suggests that analyzing public figures is a dangerously fraught project—one that tends to descend into subjectivity, politicization and self-caricature.

Applying psychological insights to political leaders’ behavior started way back in Freud’s own day. Stanley Allen Renshon, a political scientist and psychoanalyst, cites as the first such effort a March 1912 New York Times Magazine cover story titled “Roosevelt as Analyzed by the New Psychology,” about Theodore Roosevelt’s relationship with his successor as president, William Howard Taft. In it, Dr. Morton Prince, a professor of “nervous diseases” at Tufts, argued that Roosevelt was acting strangely because of an internal struggle: On the one hand, he consciously wanted to support his hand-picked successor and stand by his avowals to forswear a third term. On the other hand, he had a repressed, subconscious wish to win back the office he had loved so much. Over time, Prince explained, “the subconscious wish to be an active candidate became acceptable to his consciousness.”

But despite occasional invocations of Freudian theory in the press, it wasn’t until the Cold War that the psychoanalyzing of our presidents became a widespread pastime. The impulse had its roots in the work of the World War II-era Office of Strategic Services and its successor, the Central Intelligence Agency, both of which used psychoanalysis to study Hitler, Stalin and other leaders, to better understand their behavior and, perhaps, anticipate their actions. Then, in 1957, the Harvard historian William Langer, in his presidential address to the American Historical Association, urged his colleagues to avail themselves of the insights of “dynamic” or “depth psychology”—approaches to studying the mind, building on Freud, that look at the deep, even unconscious, forces that underpin human actions, feelings, decisions and struggles—in thinking about the past. Langer—who had served in both the OSS and CIA, and whose psychoanalyst brother wrote the OSS report on Hitler—argued that in light of the strides made by psychoanalysis, “the homespun, common-sense psychological interpretations of past historians … seem woefully inadequate, not to say naïve.”

The next year, Erik Erikson published his landmark psychobiography Young Man Luther, about the Reformation leader. It won a generally positive reception—though it was too overtly Freudian for some people’s tastes. Erikson used the harsh and demanding nature of Luther’s father to explain not only the boy’s deep sense of moral inferiority but also the conception of God he ultimately developed that would characterize early Protestantism. Even many critics of psychohistory conceded the book’s virtues.

More than academic fashion, though, it was the Cold War itself that put the mental health of political leaders on the public’s radar. The United States had emerged as a nuclear superpower, engaged in a hostile struggle with a rival nuclear power. Americans readily grasped the need for a president who was sane, stable and mature.

This new interest had consequences besides the psychological scrutiny of Richard Nixon. In the early 1960s, the Cuban Missile Crisis, along with doomsday movies like Dr. Strangelove and Fail Safe, brought home the importance of presidential sanity in a nuclear crisis. One nightmare scenario envisioned a hot-headed or irrational president starting a war; another had a Russian missile strike killing the president and others in the line of succession. As historian Rebecca Lubot, a recent Rutgers University history PhD, has argued, these worries spurred support for the 25th Amendment, which establishes the presidential line of succession. In June 1963, Nixon, now out of office, said as much to Tennessee Senator Estes Kefauver, who was planning hearings on a presidential succession bill. “With the advent of the terrible and instant destructive power of atomic weapons,” Nixon wrote, “the nation cannot afford to have any period of time when there is doubt or legal quibbling as to where the ultimate power to use those weapons resides.” Kefauver died weeks later, and with him the bill. But the next chairman of the Subcommittee on Constitutional Amendments, Birch Bayh of Indiana, took up a presidential succession and inability amendment—the 25th—which gained support after Kennedy’s assassination. Section Four of that amendment, which allows the vice president and Cabinet to remove a president for reasons of “inability,” has been getting attention amid questions about Trump’s competence, though to try to invoke it right now seems profoundly unwise.

In 1964, the Republicans nominated Arizona Senator Barry Goldwater for president. He soon got the Nixon treatment. Somewhat like Trump, Goldwater scared Americans with his unfiltered comments touting radical, intemperate or violent policy measures—he suggested using nuclear weapons as a defoliant in the Vietnam War—as well as for his alleged connections to neo-Nazi or other far-right extremists. (Daniel Schorr of CBS reported a segment from Bavaria, “Hitler’s old stomping grounds,” stating exaggeratedly that Goldwater “will be starting his campaign here.”) President Lyndon Johnson exploited his opponent’s reputation as rash and unpredictable. His campaign aired the “Daisy” television ad that showed a girl plucking a flower’s petals before seguing into a nuclear countdown. In a ridiculously unscientific study that year, the publisher-pornographer Ralph Ginzburg, using survey responses from members of the American Psychiatric Association, declared in his magazine Fact that the mental health professionals considered the candidate to be paranoid. (Goldwater won a libel suit, and the APA later adopted a “Goldwater Rule” declaring it unethical to publicly diagnose someone whom the psychiatrist has not examined. Some psychiatrists now want to reconsider the rule.)

Even after the Goldwater lawsuit, scholars and journalists continued to analyze politicians in psychological terms. More and more Americans were availing themselves of the benefits of psychotherapy, and, besides, the obvious neuroses of Johnson and Nixon all but cried out for deep interpretation. “Whether it is disagreeable coincidence or masochistic gratification, we have thereupon in quick succession elected the two most capricious and compulsive presidents we have ever had,” wrote the historian Arthur Schlesinger in a magazine essay entitled “Can Psychiatry Save the Republic?” That these men presided over Vietnam and Watergate—disasters in which the presidents were seen as having visited their own deep-seated afflictions on the nation—confirmed the apparent wisdom of probing our leaders’ psyches.

All the while, the nuclear issue hovered in the background—and occasionally came to the fore. During the 1973 Yom Kippur War, Nixon raised the military alert to Defcon III, just short of nuclear war—but it was reported that he was distraught, distracted by Watergate, possibly drunk and in no condition to lead the nation in a nuclear showdown. These and other anxieties about Nixon’s competency led Arnold Hutschnecker—a specialist in psycho-somatic illness who had treated Nixon over the years—to call for an official “board of physicians and psychiatrists” to make sure that presidential candidates “are healthy in mind and body.” “I cannot help think,” Hutschnecker wrote, “that if an American president had a staff psychiatrist, perhaps a case such as Watergate might not have had a chance to develop.”

Of course, when a politician did admit to seeking psychiatric help, everyone ran for the exits. When in 1972 George McGovern’s running-mate, Missouri Senator Thomas Eagleton, admitted to having undergone electroshock treatment for depression, a scandal ensued, and McGovern dropped Eagleton from the ticket. (Eagleton hadn’t previously disclosed his illness and treatment to McGovern.)

Clearly, the public vetting of politicians for mental health issues was not going to be simple. Part of the problem was that even some of the best analysts had a hard time bracketing their own politics when diagnosing public figures. Psychological profiles often read like political attacks—suggesting, for example, that Nixon bombed Cambodia because he needed to replenish his self-esteem. “We are danger of having the insights of psychotherapy,” warned the political journalist Godfrey Hodgson in a 1981 book review, “used as a tool for character destruction, certainly for libel, potentially for revenge.” Compounding the problem, people were using psychology to highlight a politician’s negative traits—such as, in Nixon’s case, his lying, his narcissism, paranoia, his rage—but rarely what one analyst called the president’s “many ego strengths and adaptations” that might explain how he had risen as high as he did.

Commentators in the media also failed to distinguish among varieties of mental distress. It was one thing to be insane or deranged; quite another to display neurotic behavior like extreme narcissism or paranoia; another to be functional yet still temperamentally ill-suited to the responsibilities of global power. And, of course, just as many perfectly sane people came to seek out therapy to cope with common neuroses or sources of suffering, so a politician might exhibit certain unhealthy tendencies while still mostly coping well in his or her job and daily life. But in the popular discourse, all of these distinctions tended to get lost.

In time, any pretense of Freudian or Eriksonian depth fell away. While journalists continued to harp on the question of “character,” their understanding of it became shallow and even silly, focusing on whether someone had avoided military service, smoked pot or slept around.

Still, when hotheaded candidates emerged, so did the public trepidation about whose finger would be on the button. Trump was actually on the mark last weekend when he tweeted, however vaguely, about “the old Ronald Reagan playbook.” Although nostalgia has softened Reagan’s image, during his 1980 presidential bid voters worried enormously about his penchant for angry, strident outbursts against his political opponents. “If it takes a bloodbath, let’s get it over with,” he had said of student protests at Berkeley in the 1960s; he later lashed out at the Soviet Union, calling it “the Evil Empire.” He also spoke openly of his belief in Armageddon, and his belligerent rhetoric and promises of an arms build-up raised the specter of nuclear war. To placate wary voters, Reagan spent the fall presidential debates against Jimmy Carter striking a gentle pose and reiterating his commitment to peace.

After a few harrowing years of renewed nuclear tension, Reagan pivoted and embraced Mikhail Gorbachev and arms reduction. Since then, nukes have loomed less large (although the fear that terrorists or “rogue nations” will get hold of them has underlain the Gulf Wars and conflicts with Iran and North Korea), and we’ve heard less talk about our presidents’ psychology. Whatever their failings, our recent ones have generally seemed grounded. (One exception was the controversy about candidate John McCain’s stability in the 2000 and 2008 campaigns.) Meanwhile, pinterest in psychodynamic therapy has given way to quick-fix treatments like pharmaceuticals, while the vogue for brain science has reduced the complex interactions of unconscious and conscious drives to the mechanical firing of neurons. Political scientists and historians, meanwhile, have by and large veered away from explanations rooted in individual human psychology in favor of those rooted in social and institutional structures.

With Trump, it seems necessary to retrieve the rich conceptual vocabulary of psychoanalysis. His aberrant behavior seems to demand deeper understanding. Why does he feel such a need to belittle and humiliate anyone who challenges his primacy, be it Jeb Bush, Hillary Clinton, Mitch McConnell or Kim Jong Un? What explains his habit of not simply denying criticisms made of him, but completely inverting them—to claim, for example, that he’s “the least racist person” of all, or that it was Hillary Clinton who colluded with the Russians in the 2016 election? How can he, without apparent awareness, embrace people one day, turn on them with fury the next, and then perhaps welcome them again as if nothing happened? These actions go beyond standard political spin and hypocrisy, beyond most politician’s self-aggrandizing or inconsistent behavior. Perceptive and psychologically conversant analysts should be able to contribute to our understanding of these sorts of patterns where political science and journalism cannot.

But there’s also a risk of repeating the errors of the Goldwater- and Nixon-era psychologists, who too often filtered their readings of those men’s personality through their own politics and failed to notice their high-level functioning in many areas. Trump, too, for all his juvenile, obnoxious and cruel antics, is a successful businessman and politician, one who functions in his job, however erratically. Moreover, as the psychiatrist Richard Friedman has argued, the character traits that make him outrageous aren’t necessarily symptoms of any deep psychological disorder; they may simply be the human traits of selfishness, pettiness, nastiness and shamelessness, in very high doses. Labeling Trump unstable or deranged does nothing to help us understand why he acts as he does. On the other hand, the dormant tradition of psychobiography, if practiced with due modesty and sensitivity, just might.