Psychiatry Has Never Found Its Penicillin

Lex Fridman opens by asking whether it is fair to call Andrew Scull's view a crisis: some progress over a century, but still no reliable treatment for serious mental illness by drugs or talk, the brain route or the mind route. Scull, who has spent decades on how societies understood madness and how often psychiatric authority was used with false confidence, does not dodge the word. Psychiatry, he says, tries to handle an enormously complicated thing: minds that break with common sense, emotions in turmoil, people who see what the rest of us think is not there.

Over at least the last three-quarters of a century there has been some limited progress, more of it in milder distress than in psychosis. The psychopharmacology of the early 1950s was serendipitous, not planned, and it was an advance on much of what had come before. Parallel changes in psychotherapy helped some patients too. What the field has, Scull insists, are symptomatic treatments, not cures. "We don't have a psychiatric penicillin for any of the conditions we're gonna be talking about."

Help is real and quite limited, and treatment creates problems of its own. Exposure work for PTSD can make many patients worse even as it helps some. Antipsychotics and antidepressants are at best partially effective, fail a significant fraction of the people given them, and leave psychiatrists unable to tell in advance who will respond, who will be harmed, and for whom the side effects and the benefits sit in a tight balance. That is the starting point before the history begins.

The first strand of the crisis is diagnosis. The framework still in use was set in 1980 with DSM-III, after psychiatrists kept failing to agree on what was wrong with the same patient. David Rosenhan's *On Being Sane in Insane Places*, which Scull calls a scientific fraud, and a 1960s professional literature that outsiders did not read, made the unreliability public. Robert Spitzer's task force at Columbia built a tick-the-boxes system: hit six of ten symptoms and you could be diagnosed with major depression. Drug companies, insurers, and frightened families all liked the certainty.

Reliability is not validity. Two doctors can reach the same conclusion and still be naming a constellation the way eighteenth-century physicians named fever. Each revision lengthened the list until some categories were hard to believe as illnesses rather than social constructs. When DSM-5 was being planned, NIMH directors Steven Hyman and Thomas Insel bet that genetics and neuroscience would replace symptom checklists with real pathology. By about 2008 they had thrown up their hands. Insel later said that after thirteen years and more than twenty billion dollars, "the lot of the mentally ill has improved not one bit." He and Hyman denounced DSM-5 as unscientific and useless; Spitzer and Allen Frances joined the attack.

Genetics has not rescued the boxes. Genome-wide association studies can, with hundreds of small variants, account for about ten percent of schizophrenia risk, and the variants that raise risk for schizophrenia, bipolar disorder, and autism overlap heavily. Scull's friend Robin Murray has argued that schizophrenia looks like the extreme end of a continuum. If, in a decade, leading psychiatrists have to say there is no such thing as schizophrenia, public trust will take another blow. Harvard's Leon Eisenberg put the field's swing in one line: he entered a brainless psychiatry and was leaving a mindless one.

Policy is bleeding in parallel, and Scull will not hang that solely on psychiatry. A person with serious mental illness in the United States dies on average fifteen to twenty-five years earlier than everyone else, and the gap has been growing. The three largest centers of inpatient psychiatric care in the country, if they can be called that, are the Los Angeles County Jail, Cook County Jail in Chicago, and Rikers Island in New York. Asylums were abandoned from the late 1960s for something miraculous called community care. Scull calls community care a shell game without a pea: the promise existed, the facilities did not.

Nineteenth-century asylums were born in optimism. Early alienists believed that rescuing people from attics and jail cells, giving them order, work, and decent treatment, would cure sixty, seventy, even eighty percent if they came in early. Scull thinks the early hospitals did some good and that some patients recovered. They were not discharging eighty percent. They were discharging thirty-five or forty, and each year a residue of chronic patients stayed behind. Over time the image of the asylum became the person who spent decades there and left in a pine box.

When the promised cures failed, the explanation blamed the patient. Mental illness was recast as biology: evolutionary throwbacks, defective brains, a lesser form of existence. Degeneration theory said these people breed like rabbits and cannot be released. That story justified locking them up with no therapeutic pretense, then sterilizing them so they might be let out. Scull's own state, California, was a pioneer and kept sterilizing mental patients until about 1960. By the 1960s more than sixty thousand sterilizations had been performed in the United States, California a disproportionate share.

A British psychiatrist said that if his incoming patients had been puppies, he would have tied them in a sack with lead weights and drowned them. In a democracy that language never quite acquired mass support for killing. In Nazi Germany it did. The mentally ill were labeled useless eaters, sterilized, then murdered in the T4 program, named for the Berlin address Tiergartenstraße 4. They were the first victims of the Final Solution; the dead may have numbered a quarter of a million. The technology of the gas chamber, and of disguising it as showers, was developed in that killing. Local people called the buses killing crates.

American narratives of lesser biological beings traveled. Before World War II, federal money barely touched medical research; the Rockefeller Foundation chose psychiatry as a priority around 1930, in part because several trustees had mental illness in their families. Among the geneticists it funded was Ernst Rüdin, who absorbed California's lessons on sterilization and became an enthusiastic supporter of murdering patients. German psychiatry, Scull says, for the most part collaborated. Decent psychiatrists recoiled from becoming jailers or worse, and some began to hunt for biological interventions that might actually help. That hunt produced the desperate remedies of the next half century.

The clearest of those experiments in the public imagination is lobotomy. Portuguese neurologist Egas Moniz began in 1935 by drilling holes in the skull and injecting alcohol or using what looked like a butter knife to break connections. He received the Nobel Prize in 1949, fourteen years of experience later. Scull is sure that is one the committee would like to retract. American neurologist Walter Freeman started in 1936, found the original operation too slow, and, his daughter said, aspired to be the Henry Ford of lobotomy.

Freeman invented the transorbital ice-pick method: two or three electric shocks to knock the patient out, then a pick driven through the bone above the eye socket and wiggled to sever tissue. After the war, with more than half a million people in American mental hospitals, he traveled in a camper he called the Lobotomobile, descending on state hospitals in summer to teach the procedure. He was ambidextrous. When his right hand tired he switched to the left, and he sometimes did twenty or thirty lobotomies in an afternoon. He boasted, "I could teach any damn fool to perform a lobotomy in twenty minutes, even a psychiatrist."

The operation died by generational change, not by a sudden confession. Some lobotomists kept operating into the 1960s and early 1970s. Younger doctors who met the incontinent, emptied people on the back wards rebelled, and by then antipsychotic drugs looked more like ordinary medicine. Women took the brunt. About seventy percent of Henry Cotton's surgical patients were female; hospital records Scull has seen put lobotomy and ECT in the same sixty-to-seventy-percent range. Men were more often tagged with personality disorders or ADHD. Moniz still had the medal. The patients had the aftermath.

At the turn of the twentieth century, as many as a quarter of people admitted to asylums had general paralysis of the insane: gradual loss of walking, swallowing, and speech, paired with the conviction that they were Napoleon, Jesus, or the richest and sexiest man in the world. Rockefeller Institute researchers found the syphilis organism in those brains. This was tertiary syphilis, the infection gone underground for years, then attacking the heart or the central nervous system. Psychiatry had been hunting underlying pathology. Here, for once, it found it.

Austrian physician Julius Wagner-Jauregg had long believed fever could cure mental illness. Rat-bite fever and typhoid vaccine failed. Near the end of World War I he injected malarial blood from a captured Italian soldier into GPI patients and claimed cures. He later admitted the claims were wildly exaggerated, but the treatment spread through Britain, Germany, and the United States. Hospitals kept colonies of malarial mosquitoes; patients were put in straitjackets and left to be bitten. He received a Nobel Prize, one of only two ever given for psychiatric innovations. The other was lobotomy.

What ended malaria therapy was penicillin, a genuine magic bullet. Once you could kill syphilis cleanly, there was no reason to give people malaria. Scull uses the episode as a reference point for the rest of medicine's laboratory turn: Pasteur and Koch, vaccines, Lister's antiseptic surgery. Those are refreshingly clear examples of progress. Psychiatry has had hints, not that kind of leap. The main diseases still have not met their penicillin.

The syphilis finding fed a worse sequel. Emil Kraepelin, who drew the schizophrenia and manic-depressive split still in use, began to suspect infection more broadly. One of his American trainees, Henry Cotton, took over New Jersey's Trenton State Hospital in 1916 with Adolf Meyer's backing. Cotton believed focal sepsis released toxins that poisoned the brain. Without antibiotics he practiced what he called surgical bacteriology: pull teeth, then tonsils, then stomachs, spleens, and colons. He claimed an eighty percent cure rate, lectured at Princeton, and was hailed by *The New York Times*. Forty-five percent of the people who received abdominal surgery died within a year. He wrote that stomachs were like cement mixers on a construction site and could be dispensed with.

Cotton dropped dead of a heart attack in 1933. Pupils dropped the abdominal work, kept colonic irrigation, and went on pulling teeth and tonsils. Scull interviewed the dentist who arrived in 1916, pulled several hundred thousand teeth, retired in 1960, and still thought Cotton should have won a Nobel. Why did smart people go along? White coats, intimate violence, and a promise of cure make a powerful placebo, for the patient who wants to believe and for the doctor who sees what he wants to see. That problem, Scull notes, has not left contemporary psychiatry.

Eighty percent returned with insulin coma. Manfred Sakel, who had used mild comas in a German addiction clinic, moved to Austria and pushed schizophrenia patients into hours- or days-long hypoglycemic comas, then woke them with glucose, up to sixty times in a course. He claimed eighty percent cured. Mortality ran one to five percent, with brain damage and obesity. A randomized trial in the 1950s killed the treatment. John Nash, later the subject of *A Beautiful Mind*, received insulin comas at Cotton's old hospital in Trenton and was at risk of lobotomy. In the same Austro-Hungarian world, another psychiatrist decided epilepsy and schizophrenia were antagonists, so an artificial seizure might drive psychosis out. Camphor abscessed; Metrazol produced a death-like terror and then fractures of spine and hip. That brutality is how electricity entered the story.

Italian psychiatrists Ugo Cerletti and Lucio Bini first ran current from a dog's head to its anus and killed the animals. Someone sent them to a Rome slaughterhouse, where pigs were stunned with electrodes on the head, convulsed, and had their throats cut. Current through the brain did not have to be fatal. They picked up a homeless man at the Rome train station. The first shock was too weak. He heard them discuss raising the current and said, "No, another one, that's deadly." They did it anyway. He seized, stopped breathing, started again, and came around in contact with reality. Cheap, fast, and free of injections, the method spread.

Unmodified ECT ruled into the 1950s and in some places the 1960s. It did little for schizophrenia, the original rationale, but seemed to help severe and suicidal depression. In the hospitals of the 1940s and 1950s it was used more as discipline than as therapy: patients feared it, so they controlled themselves. Without muscle relaxants, convulsions fractured spines and hips; memory loss was expected. CIA-funded Ewen Cameron in Canada gave multiple shocks a day until people could not walk, talk, or feed themselves, and many were left permanently damaged. Ken Kesey's *One Flew Over the Cuckoo's Nest*, filmed in 1975, fixed Jack Nicholson's McMurphy under the electrodes and then under the ice pick. Scull says almost every student in his Madness in the Movies class had seen it. The film helped sideline ECT for decades.

Modern ECT, with anesthesia and muscle relaxants, no longer breaks bones. Recent trials, Scull grants, give decent evidence that for some people with treatment-resistant depression it works, and it often has to be repeated as maintenance. Fridman mentions a friend who had tried everything and whose life changed for the better; Scull says that story is not uncommon. The treatment remains purely empirical. In California it can almost never be given to involuntarily confined patients. Few medical procedures are fenced by law that tightly. Horror stories and lifesaving stories sit side by side, and the mechanism is still a blank.

Hospital psychiatry had gone biological, but large numbers of people were sad, confused, and socially stranded without meeting the threshold for incarceration. Nineteenth-century America answered with religions and rest. Christian Science, Mary Baker Eddy's invention, treated disease as insufficient faith. Neurologist Silas Weir Mitchell prescribed the rest cure: bed, high calories, no intellectual work, aimed mainly at women. Virginia Woolf said it nearly drove her mad. Mitchell's bestsellers *Wear and Tear* and *Fat and Blood* blamed the telegraph and the railway for exhausting the nerves. A hundred and twenty years later, Scull and Fridman note, the same complaint is aimed at social media.

In 1909 Freud, Jung, and a colleague crossed the Atlantic for Clark University's twentieth anniversary, an afterthought among some thirty speakers. Freud lectured in German; the American scholarly audience had learned German to read the medical literature. Mainstream American psychiatry ignored him and later dismissed talk therapy for a biological disease as ridiculous. Freud returned the contempt, suggesting the country be renamed Dollaría. Writers, painters, and playwrights embraced the unconscious, repression, and slips of the tongue. Freud himself lamented that his case histories read like short stories, which is why they sold in the arts and never quite earned the stamp of strict science. Circa 1930 there were perhaps three hundred psychoanalysts in North America, a few thousand patients against hundreds of thousands in hospitals.

War changed the arithmetic. Shell shock in 1914–18 was first blamed on exploding ordinance shaking the brain, then recognized as psychological trauma, and Freud's ideas of the unconscious gained new weight. Hitler drove analysts, and Freud himself, into exile. In World War II the U.S. screened out 1.75 million recruits as psychologically unfit and still saw as many as twenty-five percent of combat soldiers break down; American psychiatric casualties were two to three times World War I. Before the war there were about two thousand psychiatrists in the United States; by the end there were more than two thousand in the military alone. William Menninger put a watered-down Freud into crash-trained army doctors. By 1958 about eighty percent of psychiatrists worked in outpatient offices, not hospitals.

Psychologists drafted into treatment discovered they liked the work. They built packaged, testable therapies; cognitive behavioral therapy is the type specimen. Analysts told the Rockefeller Foundation they were reconstructing whole personalities and could not produce easy proof. Clinical psychologists treated symptoms, assigned homework, and learned how to write federal grants. Aaron Beck, trained as an analyst, showed in the 1960s how unreliable psychiatric diagnosis was, then joined the symptom camp. NICE in Britain now lists CBT as first-line for milder depression; Cochrane rates the evidence as low to medium confidence, and poor for schizophrenia. Closer to ordinary unhappiness, the data firm up. For psychosis they wobble. Psychoanalysis, Scull says, is a niche product for the rich.

Hardly anyone had imagined a drug that would attack mental illness rather than merely sedate a ward. French firm Rhône-Poulenc had an 1880s German antihistamine, chlorpromazine, and no idea what to do with it. Drug testing was a Wild West: hand samples around and see. Naval surgeon Henri Laborit used it as an anesthetic booster; patients awaiting surgery stopped being anxious. He wrote to a relative in a Paris mental hospital that it worked like a chemical lobotomy. Delay and Deniker at Sainte-Anne raised the dose until patients stopped smashing furniture. The first two American companies passed. SmithKline & French bought the rights, put its best salesmen on it, showed state legislatures before-and-after films, and within two years two million people were taking the drug. Psychiatry became a market, and large pharma's era in the field began.

Renamed from major tranquilizer to antipsychotic in the early 1960s, the class reduces hallucinations, delusions, and agitation, the so-called positive symptoms, for a significant number of patients. It does little or nothing for apathy, poverty of language, and social withdrawal. Side effects were ignored for about twenty years: restlessness that never stops, Parkinsonism, and tardive dyskinesia, the late, jerky face-and-tongue movements that make strangers cross the street. What they are seeing, Scull says, is often the drug, not the illness. George Crane finally said so in *Science*. Clozapine and the second generation were sold as the answer. They traded some tardive dyskinesia for other harms, including destroyed white cells in clozapine's case.

In 2005 NIMH paid for CATIE, published in the *New England Journal of Medicine*, because industry owned most of the data. The new drugs were not more effective than an old, cheap one. Depending on the drug, sixty-seven to eighty-two percent of patients dropped out because of inefficacy or side effects. The signature new harm was tens of pounds of weight gain, then metabolic syndrome, diabetes, and heart disease. "There is no free lunch with medicine," Scull says. Every treatment is a bargain between benefit and cost, and at the moment you walk into the office there is no biological marker to say which end of the distribution you occupy.

Antidepressants began as another accident. In the 1950s, advanced tuberculosis patients given a new drug started dancing; their mood had lifted. First-generation agents had two lethal flaws: tricyclics killed in overdose, a grim gift to someone already suicidal, and MAOIs could turn cheese or cured meat into a medical emergency. In the late 1980s SSRIs, led by Prozac, arrived with a simple story: depression is too little serotonin, and the pill puts it back. Celebrities and political spouses helped. SSRIs remain the most prescribed antidepressants. Depression, Scull notes, has become the common cold of psychiatry, a change that is partly understanding and partly marketing.

Every controlled study beats placebo by a statistically significant margin, not necessarily a clinically significant one. One or two points on a sixty-point rating scale can be enough for the ads. For most patients the advantage is marginal and much of the effect is placebo. More than forty percent do not respond at all, which means they take the risks for nothing. People describe emotional numbing, a flattening of highs and lows; loss of libido that sometimes does not return after stopping; and withdrawal so severe that many stay trapped on the pills for years. A recent *Lancet* study, in Scull's reading, splits patients into three groups: good response and bearable side effects, no response, and a middle where benefit and harm cancel. Nobody can tell you in advance which group you are.

The same commercial story feeds diagnostic creep, a term Scull takes from Nick Haslam: start from an unambiguous core and keep drawing the circle outward. DSM-5 argued over whether grief counts as depression. Allen Frances, who ran DSM-IV, is convinced the explosion in autism diagnoses owes more to loosened criteria than to more cases. A diagnosis unlocks insurance, school support, and services, so family groups defend the boundary even when relatives' interests and patients' interests are not the same. Historian Michael MacDonald's line is the one Scull keeps quoting: madness is "the most solitary of afflictions for the sufferer and the most social of maladies" for everyone around them.

Fridman asks the practical question: if you are psychotic or depressed, what are you supposed to do? Scull's answer is not pretty. Intense suffering still makes drugs worth trying, with a clinician who watches closely, because every trial is a trade with side effects. For milder depression, try CBT first; NICE ranks it ahead of medication on the benefit-risk balance. For suicidal depression that drugs have not touched, modified ECT is one of the few options with evidence, and a 1975 film should not make the decision. Beyond the clinic, social connection is a real medicine. Loneliness has epidemiology behind it, and giving people some agency and some relationships is routinely underestimated.

He is very worried about ketamine and psychedelics as miracle cures for depression. The evidence, he says, is enormously weak, and he has seen this movie too many times: eighty percent for Cotton, eighty percent for insulin coma, eighty percent for the early asylums. Progress usually comes in small steps; penicillin is the exception. Fridman notes that psilocybin was demonized for a long time and that Johns Hopkins is running serious trials. They agree on the method: test each new treatment with scientific rigor, not enthusiasm, and ignore the click-driven hype. Negative findings keep us from repeating mistakes and almost never make the front of *Science* or *Nature*.

Two larger fears sit behind the clinic. Major drug companies have left psychiatric research, which Scull finds depressing even though they have been bad actors; the people with the most money are not hunting better treatments. And public trust in science, once smashed, is brutally hard to rebuild. Aborted trials, untrained young scientists, four years of lost cohorts: the losses are counterfactual, invisible, and real. Psychoanalysis is a rich person's niche; AI talk therapy is an open question. Public policy for serious illness is a mess he is pessimistic about fixing, because the honest pitch is not that we can turn tax eaters into taxpayers tomorrow. Hope, for Scull, is a less technocratic profession that still refuses cynicism: well-meaning people who see the suffering face to face, and a culture that asks for data, controls, and a decade before it believes a breakthrough.