After lights out at the state forensic hospital, a worker would peer through my door’s square window every fifteen minutes. He would quickly flick the overhead light on and off—just long enough to record whether I was breathing and in what position I was sleeping. Sometimes I could hear him whispering the data aloud to a partner making rounds alongside him, recounting our bodily orientations like buried bodies in a graveyard:
Left side. Right side. Back.
They knew the position in which my body slept every fifteen minutes for ten months. But they rarely asked me a personal question about my life.
I spent fourteen months in county jails and a state psychiatric hospital after accepting a not-guilty-by-reason-of-mental-disease plea on eight felony stalking charges arising from a psychotic manic episode. I had a doctorate from the University of Chicago and a book with Cambridge University Press. I had also been misdiagnosed by four different hospital systems, prescribed the wrong class of medications for four years, and carried an undetected neurosyphilis infection that inflamed my brain while every clinician I encountered looked somewhere else. By the time I arrived at Mendota Mental Health Institute, Wisconsin’s forensic psychiatric hospital, in January of 2025, the system had demonstrated an extraordinary capacity to observe and document me and an almost total incapacity to know me.
The maximum-security units were arranged in a hub-and-spoke design. A central core provided sightlines into four radiating wards, each terminating in a nurse’s station enclosed in tinted glass from which staff monitored and documented; from the dayroom, we could often see them behind the glass looking down at their phones. Exchanges with psychiatric care technicians (PCTs) occurred only through a narrow request portal, whose hours had shrunk over the years to ten or fifteen minutes per hour. The portal was a test: could you wait and follow the rules? Could you manage your frustration with the demands of bureaucratic order while practicing the skills of emotional regulation? These were the questions the watchtower was built to answer. Whether you were suffering, or confused, or intellectually alive, or desperately lonely—these were questions the watchtower had no room for.
Goffman argued in Asylums that total institutions mortify the self through a process in which “all acts of resistance by the so-called mentally ill” become “just further evidence, in the eyes of the staff, of their mental incompetence and illness.” Mendota was not a mid-century asylum—it was cleaner, better staffed, and far more humane. But the structural logic Goffman recognized had not vanished; it had just been refined. I intend this less as a polemic than as an honest portrayal—a careful account, from the inside, of a system that has developed a more sophisticated capacity to measure a human being than to know one.
II.
Through the wall of my room at Mendota, my neighbor Johnnie—a man with schizophrenia—was the conductor of a full symphony of voices no one else could hear. In one movement alone, he shifted seamlessly from polite dialogue to laughter to verbal combat. Then came the machine-gun of counting: “One-one, two-two, three-three”—up to twenty, silence, followed by another cartridge. Each number landed twice, as if repetition could nail it to the fabric of reality. In a world where thoughts bled into voices and voices blurred into persecution, counting seemed to provide Johnnie with an anchor—a metronome to ward off psychic freefall. What looked like madness from the outside was, from within, a courageous effort at coherence.
He often cornered me in the hallway to announce an epiphany, “Vocationally, spiritually, politically, logistically, the zebra is the outright leader of the animal kingdom—correct, Patrick?” I learned that polite agreement was the kindest response. Johnnie called me by name, and the institution did not always manage that.
In a psychology group one afternoon, our therapist asked us why it was so hard to be happy. We sat in a semicircle—six men adjudicated dangerous to themselves or others, parsing the nature of happiness like students in a high school ethics class. At the end of the session, Johnnie shot up his hand. “I’d just like to say it’s important to remember the following: we’re not sick, we need a sheikh!” The room burst into scattered laughter. And for a brief instant, the cage dissolved.
Johnnie’s disorganized brilliance was not something the system knew how to recognize. It knew how to medicate and monitor him, document his behavior in hourly notes on electronic tablets, and keep him safe from himself. But it made no effort to know the man beyond the pathology. And Johnnie was not alone in inhabiting an abyss the system could observe but not cross. Gabriel—gregarious, competitive, the undisputed life of whatever room he entered—could charm you senseless and eviscerate you in the same conversation. When I challenged him in front of the group for calling his father about the Vegas odds before our weekly football picks, he turned to another patient and spoke as if I were not there: “This failed old man, who thinks he’s so smart, walking around reading books and acting like he’s so educated—and he has no family or friends. Who does he think he is?” The institution could record his behavior and adjust his privileges, but it could not help him understand why losing at Gin Rummy felt like someone was wreaking havoc on his psyche—because understanding why requires searching for the shape of someone’s soul, and the system had never ventured on such a quest.
The hospital’s inability to know Johnnie, Gabriel, or me was not simply a failure of bedside manner. It was the final act of a longer failure of inquiry—one that began years earlier, when clinicians rubberstamped a previous diagnosis instead of investigating the contours of a tangled life.
III.
In the summer of 2016, my outpatient psychiatrist in Chicago diagnosed me with major depressive and generalized anxiety disorders. Over the following four months, I was evaluated by four additional hospital systems. Not one questioned the inherited diagnosis or conducted a comprehensive life history. Not one, despite my open identity as a gay man in a period of high-risk sexual behavior, ordered an STI panel. Each institution accepted the labels it inherited, added its own prescriptions, and discharged me.
Through my reading at Mendota years later, I learned that diagnostic anchoring—the tendency to cling to an initial explanation even as contradictory evidence accumulates—had functioned as a form of diagnostic inertia. In the care I received, symptom management repeatedly displaced any diagnostic curiosity. And the medications prescribed to treat my supposed unipolar depression were the very class of drugs that major guidelines caution against in bipolar disorder because of the risk of triggering mania. Whether that is exactly what happened to me cannot be adjudicated from my memory alone. But the possibility should have prompted a more nuanced reconstruction of my complex history—and it never did.
When the blood test was finally drawn—after my fourth suicide attempt left me in a medically induced coma for six days—the syphilis diagnosis arrived as a dark awakening. An astute medical fellow approached my mother with hope in his eyes, stating: “This could be the answer.” But our siloed health care system did not integrate what it had learned. The infectious disease team treated the syphilis while the psychiatry team continued treating “depression.” Consider what a different clinical framework might have made possible: if any of those hospital systems had taken a comprehensive life history—one that recognized the sudden behavioral shift from optimism to despair, the impulsive rupture of a long-term relationship, the escalating substance use, and the unscreened sexual conduct—they might have seen the mania before the depression, might have pinpointed a bipolar disorder years before the system finally did, and might have given me lithium instead of antidepressants. The human cost of what the system did not ask is a decade I will never recover.
IV.
For people who are becoming psychotic, manic, or unable to recognize their need for care, the American system regularly offers meaningful intervention only after catastrophe—when police, courts, or jails have already entered the story. As the journalist Alisa Roth documents in Insane: America’s Criminal Treatment of Mental Illness, jails in New York, Los Angeles, and Chicago each house more people with mental illness than any hospital in the country, and by some estimates as many as half of all incarcerated Americans have a psychiatric disorder. America has, in the eyes of Roth and many others, made mental illness a crime. In my case, a booking number, not a medical chart, became my primary identifier. I had to become dangerous enough to others in order to access the vital help I needed for myself.
At the Dane County Jail, I waited weeks for a psychiatric evaluation. Within a week I caught myself standing at attention before any instruction had been issued. The gaze of perpetual surveillance had settled underneath my skin. When I finally met with the jail’s psychiatrist, she had no access to my medical records and stated she could not diagnose bipolar disorder in one session. Instead, she offered me an SSRI. When I protested, noting they can induce mania for those with bipolar, she countered: “But we don’t even know if you have bipolar disorder.”
At Mendota, I had to sign up to shower. I had to sign up to go outside to a courtyard ringed by two high fences—the first laced with concertina wire, the second electrified at ten thousand volts. Coffee was rationed to two scoops of freeze-dried per day. We were issued urinals for the night. On French toast mornings—the best breakfast the institution offered—I noticed a pep in my step. In the dayroom, patients paced around a picnic-style table—the only exercise available on maximum-security units. Coordinating the pacing required a choreography sensitive to pharmacological side effects: some men moved slowly, dulled by medications, while others shuffled briskly and reversed direction, colliding with the herd—a brotherly waltz inside institutional confinement.
The organizing principle at Mendota was not biology, practiced rigorously, nor psychology, applied with care. It was closer to what I came to realize was an administrative reductionism: the patient existed insofar as he could be measured, categorized, and acted upon. Awais Aftab, the psychiatrist and philosopher who writes the online column Psychiatry at the Margins, has painted a portrait of a profession “in contradiction with itself”—struggling “to make sense of the divide between what it aspires to in practice and the actual state of care provided to the average person.” The divide, Aftab elucidates, exists “due to broad social, political, and economic structures” including “managed care bureaucracies that make it more and more difficult for clinicians to spend time with patients” and “reimbursement structures that do not incentivize psychosocial care.” At Mendota, I lived inside that contradiction. My lithium level was countable. My behavioral note was countable. My medication compliance was countable. My risk level was countable. But my level of narrative development, trust, grief, or identity construction fell outside the realm of the countable.
And this administrative epistemology—the logic of the documentable—explained virtually every failure I encountered. The fifteen-minute appointment existed because fifteen minutes can produce documentation without overburdening the care team. The behavioral notes mattered because behavior can be observed, charted, and defended in forensic proceedings. The security ladder functioned because risk can be operationalized into privilege levels, incentivizing compliance at each rung. Treatment teams rotated with units because it was administratively simpler to attach clinicians to a ward than to coordinate the same team across a patient’s journey through the hospital—even though that continuity was precisely what recovery required. My mother was excluded because she carried longitudinal knowledge that could not be easily converted into institutional metrics. The textbook I read alone at night went unrecognized because patient-generated expertise does not fit the hierarchy of professional authority. What could not be counted could not be seen, and what could not be seen could not be valued.
V.
My treating psychiatrist—a man I came to think of as the Robot for the mechanical efficiency of our monthly fifteen-minute check-ins—occupied the top of the treatment hierarchy without seeming to notice the expertise beneath him. He was not unkind, and I do not believe he was indifferent by nature; his manner had the reflexive productivity of a clinician whose role had been narrowed to a set of recurring operations. When I mentioned that I had read large portions of the leading clinical textbook on manic-depressive illness—hoping he might be impressed and eager to initiate a dialogue—he gave me a blank stare and asked no follow-up questions. When I asked for more information about bipolar disorder, I was handed a three-page printout. But when I later started the antibiotic doxycycline for a routine infection, the pharmacist delivered me eighty-nine pages. While a commonplace antibiotic got a hefty tome, the psychiatric diagnosis that had restructured my entire life got a pamphlet. I was left to find the textbook on my own, ordering it through the hospital’s interlibrary loan and reading it in the early morning hours on the locked ward while the institution slept.
The next meeting, I tried again and brought him a hypothesis: that neurosyphilis, substance use, and an underlying vulnerability to bipolar illness might have interacted in ways no clinician had ever helped me reconstruct. He told me what I had outlined “certainly makes a lot of clinical sense.” But no matter what happened in the past, he stated, I should realize I was at Mendota and needed to appreciate the “gravity” and “seriousness” of the situation. And then he moved on, never to return to the question.
One afternoon I again raised the relationship between the neurosyphilis and my bipolar disorder. He interrupted me. “Regardless of how it appeared, what you have is brain damage. That’s what matters.” When I later read his summary in my patient file, he wrote tersely that he had “provided psychoeducation about bipolar disorder and lithium.” In his account, our meetings were a one-way transmission from expert to pupil, a far cry from the searching conversation I had sought. In the House of Mendota, knowledge flowed in only one direction.
Gardner and Kleinman identified the conditions that manufactured the Robot. In a 2019 essay in the New England Journal of Medicine, they contended that medicine’s “era of high throughput” had been “especially deforming” for psychiatry, which still faces “substantial diagnostic and therapeutic uncertainty.” The field, they observed, had “largely abandoned its social, interpersonal, and psychodynamic foundations, with little to show for these sacrifices.” What was needed was not a further contraction into biology but an expansion into “comprehensive, relational care.”
Meanwhile, a psychologist at Mendota I will call Dr. Sally asked me about shame. She perceived how my intellectual persona fueled my relentless drive for validation. She began to see the broad contours linking neurosyphilis, bipolar disorder, and the blindness of anosognosia. In the solitary act of rumination, I had rarely yielded insight—but in the presence of another, causality demanded articulation.
The division at Mendota was not simply between disciplines but between regimes of knowledge the institution ranked differently. Psychiatry held authority over medication, diagnosis, and discharge. Psychology and substance-use treatment made room for shame, memory, identity, and narrative. These modalities of care existed beside one another but seldom informed one another. The substance use program I attended—led by a skilled clinician I will call Carrie who used schema work, defense identification, and life-history mapping—never once came up at a treatment team meeting with the Robot. The Robot saw a diseased brain as he looked down into his laptop or checked his watch. Carrie confronted a person whose inner life was in revolt and told him he wasn’t alone as she stared into the back of his eyes. But no one built a bridge between them.
VI.
There is a critical distinction in the clinical literature between what Rachel Aviv has called “clinical insight”—which measures the degree to which a patient assents to his doctor’s interpretation—and what Slade and Sweeney call “narrative insight,” or a person’s effort to craft a meaningful story about one’s experiences. Clinical insight asks: Do you accept the diagnosis? Narrative insight asks something harder and more personal: Can you construct a story about what happened to you that is honest, flexible, and your own?
The institution measured clinical insight. What I was building in my room at night—writing longhand with a tremor after years of relying on a backspace key, reading clinical textbooks, fine-tuning an etiological hypothesis in long conversations with my mother—was narrative insight, and the system gave it no recognized place. The theory I had sought to explore with the Robot—that neurosyphilis, substance use, and a genetic predisposition to bipolar illness had combined in ways no clinician had ever helped me piece together—was the product of that labor. I do not claim my account can be proven retrospectively. I claim that no one made space to explore the question with me.
The institution was not preventing me from speaking, but it was determining which speech counted as insight and which speech could be read as defiance or denial. And there was a disturbing paradox I could feel but not always articulate: my desire to map the complexity of my illness could actually jeopardize how the institution saw me, if that self-understanding conflicted with the institution’s tidy narrative. Mendota’s narrow definition—do you agree with us that you are ill in the cursory way we believe?—had the effect of predetermining the shape of my narrative insight by deciding what counted as meaningful knowledge about myself.
Arthur Kleinman has delineated the gap between disease—the clinically legible condition that appears in the chart—and illness, the patient’s lived experience of being sick, which includes meaning, fear, identity, and the social world the sickness has distorted. Every institution I passed through managed my disease with reasonable competence. But none expressed much interest in addressing my illness. Half a century ago, George Engel proposed the biopsychosocial model as a corrective to medicine’s tendency to conflate illness with biology alone. In many of the psychiatric settings I encountered, Engel’s model was invoked more readily in theory than it was enacted in practice.
A third form of insight—the one most absent from the House of Mendota—I have come to call relational insight: the type of understanding that emerges not from a checklist or a chart but from the awkward and uncomfortable work of actually knowing another person. It cannot be written on a form or added to a bill. It was, in my experience, the form of insight that mattered most to my recovery.
VII.
But relational insight requires something Mendota was incapable of providing: continuity. I had been doing acceptance and commitment therapy with a psychologist I had started to trust for several months when I transferred to a new unit, and the therapeutic relationship was severed overnight—replaced by a stranger who practiced a different model entirely.
Charlie—a young man in his late twenties with schizophrenia who had been detained nearly a decade—told me simply, “They don’t care about me.” When his voices once told him to escape, he climbed the chicken wire of the high fence and tore his hands open. When his voices persisted, he was told by the treatment team to try harder to control them, as if willpower alone could silence torment. Then, miraculously, he transferred to a unit with a different team. They spent forty-five minutes with him in their first meeting. They listened and lingered in conversation. In the days afterward, Charlie told me, his voices became more muted. I cannot know what else contributed—medication adjustments, a calmer environment, or perhaps just the passage of time. But Charlie’s account made visible what the institution seldom measured: the possibility that personal attention itself can alter the conditions of psychic distress. When I saw him, he seemed less guarded and more willing to sit in the dayroom rather than pace. He spoke about the new team with something close to surprise—as if he had not known that being listened to was something the hospital could offer.
My mother, too, could have supplied the continuity the system disrupted. She had known me for decades—not only as a supportive presence but as a carrier of longitudinal knowledge. She knew my baseline personality before the illness, the behavioral changes that preceded each crisis, my substance-use patterns, my early warning signs, and the difference between the confidence of my ordinary dispositional optimism and the manic grandiosity that had taken hold. While the hospital could tell you I was lying on my left side at 2:15 a.m., my mother could tell you when I had stopped being myself. Privacy law was frequently invoked as a shield, whether or not it required one. She was told she would be “responsible” for my safety after release following my suicide attempts, yet was excluded from clinical conversations under an interpretation of HIPAA that treated privacy as a wall. She could be summoned to assume liability, but she could not be invited to contribute understanding. She was expected to catch me if I fell but was not permitted to see me standing.
Not every family is safe, informed, or welcome in a patient’s care, and no family member should displace the patient’s own voice. Mania, in its cruel lucidity, can make a person suspect that the concern of those closest to him is a form of surveillance or envy or control—suspicions abound among the most intimate precisely when a person needs the most intimate to recover and heal. But when a patient wants a trusted person included, the system should have the capacity to welcome that collaboration as ordinary rather than exceptional.
My mother came every week to Mendota anyway. She always bought me a Diet Dr. Pepper, and in that sterile place it tasted like liquid gold. She sent clinical guides she was reading alongside me so we could discuss them on the phone. She had joined NAMI. “How can I support you if I don’t educate myself?” she asked. One Sunday morning I called and caught her breathing hard at the fitness center. She was seventy-three, though she could never quite believe it—“I’m not seventy-three!” she’d push back, then do a quick calculation and concede the point with a grin. She was the opposite of my father, who had retreated inward with Parkinson’s. She expanded, and now she had turned that same energy toward mental illness. If the institution knew me in fifteen-minute fragments, my mother knew me across decades. The system that exiled her from clinical conversations was exiling the most revelatory diagnostic instrument it had.
VIII.
There was a recreational therapist on one unit—I will call her Ashley—who demonstrated that the problem was not only the persistence of attention (which could easily bleed into surveillance) but the kind of attention the institution had administrative permission to provide. Ashley looked us in the eye and remembered what we had said the day before. When I jogged laps around the courtyard, she joined me for the final ten, matching my pace and talking as if I were simply another person she might pass on a neighborhood trail.
A few times a month, she used the vegetables we grew in the courtyard and made us salsa and chips. To eat something we had grown ourselves, prepared by someone who saw us as capable of cultivation rather than only containment, felt unusually radical. After my first month, she said in front of the treatment team: “Patrick, I’d like to commend you for your active participation in group. I think the rest of the patients benefit from your insights.” It was the first time my presence in the House of Mendota was publicly acknowledged as something of value rather than something to be documented or managed or contained. I felt it physically—a loosening in my chest, then a quick frisson of pride.
Her attention was, by the institution’s own metrics, unbillable. It did not appear in the treatment plan and occupied no line on any chart. But it was, next to the rekindled love of my mother and the reading and writing I conducted on my own because no one on my treatment team ever thought to ask what I was doing with my hours, among the things that actually addressed what was broken in me. She attended not to the disease but the illness, not the lithium level but the person to whom the lithium level belonged.
IX.
The medication helped, though the relationship between the lithium and my recovery was more complicated than the House of Mendota’s narrative permitted. I had arrived at Mendota already stable—months of forced sobriety in county jail having calmed the mania that substance abuse had been igniting for years. The lithium provided a chemical floor I was grateful for, but I could not be certain whether it was the lithium or the sobriety that had done the foundational work. And to raise that uncertainty in the room with the Robot and the treatment team, where the institution was deciding my fate, would be to introduce a nuance the evaluation had no space to hold. If I noted that my stability had preceded the lithium, I risked sounding non-compliant with the drug the institution believed was holding me together. On my first day at the treatment rehabilitation unit, the charge nurse asked which psychotropic drugs I was taking. When I told her lithium, she gave me a coy smile and said, “What’s amazing is we still don’t know how it works!” There was delight in her voice—wonder rather than defensiveness. She was among the few clinicians at Mendota who betrayed anything resembling intellectual humility about the tools she was using.
But whatever the pill accomplished, it stabilized the platform. Everything else was built by people the hospital either shunned or dismissed. Fellow patients were often prescribed ten or fifteen medications. Long-term residents bore metabolic syndromes, tardive dyskinesia, the thorazine shuffle—the snowballing weight of years of pharmaceutical tinkering by clinicians who came and went. Gardner and Kleinman’s diagnosis bears repeating: biological psychiatry has “failed to produce a comprehensive model of any major psychiatric disorder, any tests that can be used in a clinic to diagnose clearly defined major psychiatric disorders, or any guiding principle for somatic treatments to replace the empirical use of medications.”
Recovery, as I came to understand it, was not the absence of a behavioral note. It was the capacity to read the heat lightning of mania on the horizon before the full storm arrived: the shrinking of sleep, the acceleration of thought, the conviction that every coincidence contained a message, and the reckless arrogance that wore a mask of freedom. It was the ability to speak with my mother without shame, to accept medication without experiencing it as a verdict on my humanity, and to imagine a life organized around work, friendship, and ordinary responsibility rather than crisis and surveillance. None of that can be reduced to a risk score. Yet those were the capacities that would determine whether I could live outside the House of Mendota.
Forensic hospitals cannot abandon vigilance; they must protect patients, staff, and the public. But vigilance becomes corrosive when it is allowed to displace every other clinical obligation—when surveillance supplants the need for human attention, and containment for interpersonal recovery. I do not know what an adequate system would look like in every detail, but I can name what was missing in mine: an intake process that includes not only a symptom checklist but a biography; routine use of collateral history, with consent processes that invite family into the clinical conversation; diagnostic time-outs after repeated admissions or medication failures; continuity clinicians who remain attached to the patient across unit transfers; recovery-oriented measures alongside risk metrics; and avenues for patient self-education—so that a patient reading the textbook on his own illness is invited to dialogue as a partner rather than as a complication.
This is not a call for clinicians to become best friends with patients. It is an argument that psychiatric systems should be organized around knowing the person, not merely measuring the patient.
The modern mental health system has constructed an elaborate apparatus for observing, classifying, medicating, documenting, and containing people while steadily losing the institutional capacity to know them—diagnostic inertia in outpatient care, fragmented silos in the emergency room, absent psychiatry in the jail, ruptured continuity at the forensic hospital, all of it flowing from the same institutional logic, a system that has steadily replaced the capacity to know a person with the capacity to measure one, and has mistaken the measurement for the knowing.
The Robot was not the problem, but he was the symptom. The failures I experienced were not aberrations of individual clinicians but the predictable consequences of a design that was never built to hold a person whole.
That design kept me safe, and I will always be grateful for that. But it never knew me. And there are nights now, months after my release, when I lie on my left side in the dark of my own home and think about the worker who recorded that position without knowing what I was dreaming. He did not know that I was dreaming of my father, who died during the years the system could not find me. He did not know that the man on the other side of the door had spent a decade trying to understand what was happening inside his own mind while every institution he entered looked at the outside and called it sufficient. He did not know my name and he did not need to. His job was to confirm that I was breathing, and I was. That is what the system was built to verify—that I was alive—not who I was, not what I had lost, not what I was trying, in the lonely quiet of a room where no one asked, to become.
