Terminal Delirium

From the perspective of her own experience as a neurologist, Pria Anand critically reads Michael Erard’s “Bye Bye I Love You: The Story of Our First and Last Words.”

By Pria AnandMay 24, 2025

• Science & Technology

SHE HAD BEEN SWARMED, she told me, by a flock of starlings. They had alighted on her arms, and their beaks were now needling her skin. She flapped her hands as she spoke. Beneath the wide sleeves of her hospital gown, the skin of her arms was crinkled like tissue paper, blood darkening into bruises at the crooks of her elbows where someone had tried, and tried again, and again, to sample blood from her tapering veins.

In her 80 years of life, my patient had been in the hospital just twice before, to give birth to each of her adult daughters. Now, she was admitted for a painful broken hip after a fall on ice in her driveway. But her mind was elsewhere, wandering some strange, jungle purgatory in which the drawstring of her hospital gown was a snake coiling heavily around the nape of her neck and the phlebotomists’ needles were the beaks of the furious starlings. She moved ceaselessly in her bed, her ankles crossing and uncrossing, her fingers picking at the loose threads of her sheets—stray feathers, she told me, or fallen leaves, or the filaments of a spiderweb, tangled between the steel bars of her hospital-prison.

I was the neurologist on call, asked to evaluate her for delirium, a condition I see often but continue to find baffling and alarming. In medical terms, delirium describes an acute and profound state of confusion. In its throes, people have vivid hallucinations and lose their ability to focus. During treatment for the metastatic cancer that would ultimately kill him, the polymath neurologist Oliver Sacks chronicled the evolution of his own delirium, charted in the pages of the notebook where he was writing a children’s book about the periodic table, his letters first losing their form, then losing their meaning, and finally disintegrating into chaos.

Delirium is both shockingly common, particularly in elderly patients—affecting 10 percent of older adults evaluated in emergency departments and 70 percent of those in intensive care units—and costly: the US healthcare system spends tens of billions of dollars on the lengthy hospital stays and increased complication rates that come with a diagnosis of delirium. Although it can arise after an illness as minor as a urinary tract infection, delirium also haunts the last moments of life, occurring so commonly that this final confusion has a specific medical designation: “terminal delirium.”

From the perspective of science, delirium is elusively protean, which makes it difficult to study and nearly impossible to explain. It can look and feel different for every person who experiences it. Some sufferers become agitated, seized by an inexorable restlessness, while others are overcome by torpor. In their quest both to map the physical substrate that underlies delirium and to predict who is likely to suffer it, scientists have turned to everything from animal models of delirium—by way of unfortunate mice and rats subjected to the blinding fluorescent lights, nighttime awakenings, and daily cocktails of nameless pills that plague most hospital stays—to predictive AI, which has yet to actually shed light on any of the delirium’s perplexing specifics.

In his new book, Bye Bye I Love You: The Story of Our First and Last Words, linguist Michael Erard eschews both rodents and AI, leaning instead into what is both uniquely human and profoundly personal about delirium. Bye Bye I Love You is billed as a book about words, a linguist’s quest to find parallels between the very first and very last words humans utter in the course of their lives. But its contents are both messier and wider-ranging than this tidy conceit suggests. In his brief prelude to the book, Erard likens the experience of writing it to that of walking back and forth along a shoreline between the first moments of life and the last, listening for echoes in the melody of waves against the rocks. This sounds more straightforward than it is. He finds these echoes not only in the advertised first and last words—those inchoate, “word-like forms [that] wriggle one by one from the phonological mush like proto–land animals crawling from Protozoic seas”—but also in silence, ritual, culture, and grief. In the end, what emerges is a book that is less concerned with words than with how language fails the human experience.

¤

Bye Bye I Love You skims the surface of an array of scientific efforts to understand language, delirium, and dying, but the book’s best moments happen when Erard dives deeper—into, for instance, the physician William Osler’s “Study of the Act of Dying.” I spent my medical training in the long shadow of Osler—he ended his career at Johns Hopkins Hospital, where I was a neurology resident and where the internal medicine residents are still called “the Osler Marines.” Infamously, he conceived of the system of medical residency at Johns Hopkins, first inspiring the terms resident and house staff when he asked his pupils to live at the hospital full-time for the eight-year duration of their training and required them, like novice monks, to remain unmarried until they graduated, a rule that remained in place until World War II. In an essay still given to the Osler Marines on the eve of their training, Osler affectionately described his own residents as “poor, careworn survivors of a hard struggle, so ‘lean and pale and leaden-eyed with study.’”

On the same hospital wards where Osler was said to have invented morning rounds, I first learned about the innumerable, terrifying ways in which a person can become suspended in the endless expanse between life and death. In Osler’s time, death was arbitrated by a mirror held beneath someone’s nose to look for the fog of breath, or by a pair of fingers pressed into the hollow of their neck to feel for the liquid beat of a pulse. But I began my residency in 2014, in an age of miracles, of intensive care units and machines able to do the work of the heart and the lungs long after the body has failed. Death had become much more complicated.

When Osler arrived at Hopkins in 1889, Erard writes, the corpus of published literature on the medical features of dying was scant and anecdotal. Osler turned his attention from autopsies in the hospital morgue to the brief moments that preceded death. “One reason for the shift seems to be that once he turned forty, he realized that he was perilously too close to the limits of his own mortality to know so much about corpses but so little about the human experience of dying,” writes Erard. Osler outlived both of his sons—one died at birth, while the second died during World War I, two years before Osler himself succumbed to pneumonia. Steeped in death, Osler developed a fascination, building a library of books about its medical features, and about ghosts, immortality, and embalming.

At Hopkins, Osler undertook the first systematic study of how we die, asking doctors and nurses to report on the nature of someone’s death along with its quality—“last remarks and especially premonitions and fears,” one physician recalled. Erard describes spending hours in McGill’s Osler Library of the History of Medicine, thumbing through the cards Osler had handed to his colleagues to chronicle death. Each card is printed with a series of prompts to capture details about what Osler terms “the act of dying,” some informed by Osler’s medical practice while others seem more inspired by the metaphysical section of his morbid library. Was the act of dying sudden? Did the lungs stop before the heart? Was there bodily pain? Spiritual remorse?

In total, Osler captured data from 486 deaths. Erard finds that only four patients uttered final words, while most others merely groaned, moaned, or cried out. One patient woke from a six-day coma and turned to his daughter, his eyes wide, before he soundlessly died, while another reached out for his doctor’s hand. In these final moments, Erard notes, “verbal linguistic powers were not, on the whole, very evident.”

Despite this resounding silence, Osler was certain that his tidy, 15 x 17 centimeter cards offered privileged insights. In his 1904 lecture series “Science and Immortality”—his only recorded mention of the study after it was completed—Osler spoke of what he called “the modes of death and the sensations of dying. […] Like their birth,” he said of his subjects, “their death was a sleep and a forgetting.” It is difficult to understand how Osler could be so sure of his conclusions. Discomfort, often physical, was documented on 186—nearly 40 percent—of his cards, while others captured emotions like anxiety and regret. His subjects wished they could go home, asked to be healed, pleaded for time.

Knottier still are the 30 cards in which death is described as “quiet.” These cards distill some of the fundamental paradoxes of both the book and the vantage point of the deathbed observer: Is a quiet death necessarily a peaceful one? And what does observing death truly reveal about experiencing it?

Confronted with the many wordless deaths that fill Osler’s cards, Erard devotes as much time to the moans and groans of the dying as he does to last words. In one chapter, after stumbling on an online nursing forum conversation about silencing the otherworldly cries of the dying with pain medications, Erard wonders, “What if the moan is actually an attempt by a person who can’t summon words to see if anyone is there?” Among Bye Bye I Love You’s strengths is its empathy, and Erard’s visceral horror at this thought is palpable; he describes his own fear of “being buried alive, communication-wise, [his] signal taken as mere noise.” Like words, argues Erard, the moans and gestures of the dying—an outstretched arm, a furrowed brow—are imbued with meaning by those bearing witness; this meaning is most likely utterly divorced from the dying person’s “intent,” representing no more and no less than the last gasps of dying neurons within an unraveling mind.

Reading about the ways a deathbed observer might project his own narratives onto the contortions of death, I could not help but wonder whether Osler’s “Study of the Act of Dying” was as much an exercise in grieving as it was a scientific inquiry. Were his cards—collected first after the death of one son, and then the other—meant to prove that death was peaceful, not because it is but because he wanted it to be?

In some ways, the age of miracles has led to more wordless deaths—punctuated now by the mechanical whir of a ventilator forcing air into unyielding lungs and the high-pitched shrieking of an emptied medication pump. In other ways, it offers a glimpse into the dying mind that Osler could only guess at: the testimonies of those who have experienced it, whose deaths—the stopped pulse, the fogless mirror—were overturned by doctors and machines.

People who have experienced delirium describe dreams bleeding into waking life and distant memories seeping into the present. In one study Erard mentions, doctors interviewed patients about their experiences with delirium after surgery. One woman remembered seeing rats spill from the sterile surface of her bedside ventilator, while another heard funeral music swelling from the bland hospital artwork. One man was certain that he was a lame racehorse, destined for slaughter, while another thought he was a whale, trapped in an aquarium. The world became misty, unreal. Sometimes, words for the experience eluded these patients entirely. Their accounts speak to the futility of Osler’s exercise, and perhaps even Erard’s: the impossibility of distilling something as rich and diverse as the phenomenology of the flailing mind into a tidy card or a collection of words, to be tallied, counted, and filed away.

¤

I spent my first nights on call as a doctor with a woman in the last days of her life. When it became clear that she would soon die, the woman called her entire family to the bedside—three daughters, seven younger brothers and sisters, uncountable nieces and nephews and grandchildren. She spoke with them each alone, and then together, hours steeped in family lore and encouragement, and when she felt ready, she sent them all away and waited, quietly, to die.

But in the final two nights of her life, peace eluded her. Her body began to writhe and thrash, contorting into impossible shapes. She picked at her bedsheets, then her own limbs, kicked her legs until they were bruised by the plastic rails of her bed, and jerked her arms into the air as if in some perpetual plea. We treated her pain and her hunger for breath, but this final torment—constant and excruciating motion—didn’t subside until she died. It would be years before I had words for the woman’s symptoms: terminal restlessness, the inexorable drive to move that often animates terminal delirium.

When my own trainees are troubled by an agitated death, I sometimes tell them: “It’s just terminal restlessness,” as though that label says something about the quality of the death. In reality, the words are a contrivance that reveal little about the phenomenon. I have no idea whether terminal restlessness indicates pain or joy, whether the movements of the restless are intentional or reflexive, why the dying brain manifests itself in this way.

In the prelude to Bye Bye I Love You, Erard remembers, “Someone asked me if I had a theory to connect first and last words. No, I said, just an intuition. Of a resonance. A rhyme. Of a metaphor to be made.” This framing—intuition rather than theory, and resonance rather than analysis—seems to be his rationale for flitting between anecdotes and studies without excavating them too deeply. The book leans into metaphor (far more than into any hardcore science), at times leaving the reader with an illusion of insight that vanishes with deeper reflection. For instance, a chapter provocatively titled “The First First Words” skims genetics, self-domestication, and the Tower of Babel without ever truly grappling with the meaty question of when and how language first emerged in our evolutionary history. In lieu of an explanation, Erard offers a coinage, as imprecise as “terminal restlessness”: “the interaction engine,” defined by another linguist as “various abilities, instincts and motivations, which work together to make possible the miracle of human communication.” These are words, to be sure, but they don’t offer much insight.

In another chapter, Erard compares the dying mind first to a collapsing house, then to a citywide power outage. “The structure doesn’t come down all at once but in bits and pieces,” he writes. “Some parts may seem intact, but this actually reflects their isolation. After all, a door can swing in its frame even in a house with no roof.” These metaphors nod at the patchwork way cognition can fail, allowing some language functions to remain intact even as the dying lose their grip on reality. In his quest to find resonances between developing brains and dying ones, Erard hints that the order in which these functions fail may resemble the order in which language first emerges. Erard stays with the metaphor just long enough to outline the sequence of the failures—first the cortex, the slim sheet of gray matter that enfolds each of the brain’s dips and folds like a blanket, then the deeply buried limbic system—but his comparison between evolving brains and dying ones begins and ends with the observation that the development of the cortex was essential for the emergence of language.

At times, Erard’s reflections can feel more akin to the chaotic flights of the delirious mind than to the fastidious lexicography promised by the title. Of the physical underpinnings of delirium itself, Erard has only this to say: it is “the result of a neurochemical commotion.” The unfortunate rodents and predictive AI have not yielded insight. Erard doesn’t do much better. His book, rather, is a creative and resolutely interdisciplinary endeavor that raises questions about death, delirium, and language. No answers are on offer.