Chapter 1 · Sample
Introduction: Why I Started Listening
This book has a single argument and an honest method.
The argument is that across cultures, eras, and faiths, the words human beings speak in their final hours cluster, with surprising consistency, into seven categories. The categories were not invented for this book. They emerge — independently, in fragments — from the published literature of palliative-care medicine, hospice nursing, oral-history archives, and the historical record of last words gathered over the past four hundred years. What this book does is read that literature as one library and put the seven categories on one page.
The method is synthesis, openly, from sources the reader can themselves consult. I am not a hospice physician. I have not run a chaplaincy ward. I have not attended hundreds of deaths. The cases in this book are drawn from people who have done that work and who wrote it down — Kathryn Mannix, Atul Gawande, Paul Kalanithi, Rachel Clarke, Ira Byock, Christopher Kerr, BJ Miller, Maggie Callanan and Patricia Kelley, Bruce Greyson, Pim van Lommel, Karlis Osis and Erlendur Haraldsson, Peter and Elizabeth Fenwick, the writers and contributors of oral history archives such as StoryCorps and the Library of Congress Veterans History Project, the historians who have catalogued famous last words for centuries. Every named case in the chapters that follow is traceable, by citation, to one of these sources. Where I quote a sentence, I name where the sentence came from. Where I summarise a case, I name the practitioner who first recorded it and the book or paper in which they recorded it. The reader who wants to read further is given the path.
I want to say at the start what I am not doing.
I am not delivering a memoir of my own bedside experience, because that is not the experience I have. I am not a chaplain. The book that this book might have looked like if I were a chaplain — the book Mannix wrote, With the End in Mind — already exists, and it is excellent. The book Kalanithi wrote, When Breath Becomes Air, exists. The book Gawande wrote, Being Mortal, exists. The book Rachel Clarke wrote out of her hospice year, Dear Life, exists. The book this is, instead, is the book that puts those books, plus several others, plus the clinical literature on terminal lucidity and end-of-life experience, plus the oral-history archives, plus the historical compilations, on one shelf and asks: when you read them in sequence, what do they show you that no single one shows you alone.
What they show you is that the dying, when they speak, speak in seven shapes.
I came to this question through reading, which is the honest beginning. I read Mannix, then Gawande, then Kalanithi, in the kind of sequence that happens when one good book teaches you to look for the next. Each one contained, alongside its primary subject, fragments of last words — sentences spoken to the author at a bedside, sentences passed through a family, sentences recovered from a chart. I started noticing that across the three books, certain shapes recurred. A wife was named. A song was named. A request was made. A recognition occurred. A peace was reported. A fear was articulated. A return — the addressing of someone already gone — was performed.
I went looking for the literature beyond those three. It is large and underread by general audiences. Christopher Kerr, a hospice physician at Hospice Buffalo, has spent more than a decade documenting what his patients say and dream in their final weeks; Death Is But a Dream presents findings drawn from a sample of over fourteen hundred interviews conducted under his clinical research program. Ira Byock, in Dying Well and the work that followed it, named the developmental tasks and recurring vocabularies he observed across a career in palliative care. Maggie Callanan and Patricia Kelley, hospice nurses, in Final Gifts, gave a generation of clinicians the phrase "Nearing Death Awareness" and a catalogue of the symbolic language patients use when they begin to approach the threshold. BJ Miller and Shoshana Berger, in A Beginner's Guide to the End, brought the field's accumulated practical wisdom into accessible form. Rachel Clarke, in Dear Life, recorded the year she spent moving from acute medicine into a hospice and the sentences her patients gave her there.
Behind those stand the researchers who have spent careers on the threshold itself. Bruce Greyson, at the University of Virginia, has spent four decades clinically observing the cluster of phenomena that includes the near-death experience, terminal lucidity, and the deathbed vision; After, his 2021 book, summarises that work. Pim van Lommel, the Dutch cardiologist, published in The Lancet in 2001 the prospective study of cardiac-arrest survivors that remains a touchstone of the field, and expanded it in Consciousness Beyond Life. Karlis Osis and Erlendur Haraldsson, in At the Hour of Death, conducted what is still the most ambitious cross-cultural survey of deathbed phenomena ever attempted, comparing American and Indian physicians' accounts of what their dying patients said and saw. Peter and Elizabeth Fenwick, in The Art of Dying, gathered British nursing and family accounts that form a quiet companion to the American and Indian record.
And then there are the documented individuals — the cases that have left the consulting room and entered the literature under their own names. Pam Reynolds, whose 1991 surgery at the Barrow Neurological Institute became the most carefully documented near-death case in the medical record. Harold Owen, whose 1918 vision of his brother Wilfred has been preserved in his own published memoir. Steve Jobs, whose final words are known because his sister Mona Simpson published them in her eulogy in The New York Times. Patients named in Mannix, named in Kerr, named in Clarke, named in Byock — patients whose families consented to the telling, whose stories sit in print, whose sentences can be quoted because someone with the right to quote them already has.
These are the people the chapters that follow are about. When this book speaks of an individual, that individual is named, and the source is given. When it speaks at the level of pattern — when a chapter says that a particular shape of utterance recurs across the literature — the literature is cited. There are no anonymous bedsides in this book that the author personally attended, because the author did not attend them. There are no composites stitched from several patients into one. There is no "Mr. K." There is no "a woman I will call Sarah." Where this book wants to illustrate a pattern that does not have a single named exemplar attached to it, it says so explicitly — "consider a person who," "imagine a family in which" — and the reader knows they are reading a hypothetical, not a report.
I labour the point because the genre invites the opposite. Death writing drifts, almost gravitationally, toward the unsourced vignette: the wise patient whose name has been changed, the dying grandfather whose dialogue is reconstructed, the composite chaplain. There are honourable reasons writers reach for those forms — confidentiality, narrative compression, the shielding of grieving families. But the seven-category claim this book makes is a claim about a pattern in a real corpus, and a pattern argued from invented examples is not argued at all. So the book holds itself to the harder rule. If a sentence is in this book inside quotation marks, a real human being said it, and the citation says where they said it and who heard them.
The seven categories will arrive in their own chapters. I will not preview them in detail here, because the work each chapter does is the work of letting the sources speak in sequence until the shape becomes audible. But I will name them, so that the reader knows the architecture of the building they are walking into.
The dying speak, in their final hours, of love. They speak of recognition — of the people, sometimes living and sometimes long dead, whom they appear to see in the room. They speak in the form of request: a thing they want done, a person they want fetched, a permission they want granted. They speak as witnesses — telling, often for the first or last time, the truth of what their life was. They speak of peace, which is its own register and not to be confused with the others. They speak of fear, which the literature does not hide and this book will not hide either. And they speak in what I have come to call return: the addressing, by name, of someone who is already gone.
Seven. Not five, not nine. The number is not mystical; it is what the corpus produced when I stopped imposing a schema and let the categories assert themselves. Other curators reading the same library might draw the lines slightly differently. Callanan and Kelley collapsed several of mine into the single capacious category they called Nearing Death Awareness. Byock organised his around developmental tasks rather than utterance-types. Kerr foregrounded the dream and the vision over the spoken word. Each of those framings is defensible, and where this book differs from them I will say so and say why. The argument is not that seven is the only count. The argument is that when you read the published bedsides side by side, the count converges, and the convergence is itself the finding.
A word about the reader I am writing for.
I assume three people. The first has lost a parent or a spouse in the last two years and is in the part of grief where the mind goes back, again and again, to the last thing the person said, trying to work out what it meant. The second has not lost anyone close yet but has become aware, perhaps suddenly, that they will. The third is the one I think about most often: the person reading at three in the morning beside a hospital bed or a hospice bed or a bed that has been moved into the living room, trying to work out what they should listen for in the hours that remain.
For the first reader, the book offers a frame in which the sentence they cannot stop replaying may belong to a category, and the category may have a literature, and the literature may grant their sentence its dignity by showing that it was not random. For the second reader, the book offers an orientation — a map of the country they will eventually have to enter, drawn by the people who have spent their working lives there. For the third reader, I have tried to write each chapter so that any one of them can be opened first, on the page that matters tonight, without requiring the others to have been read.
That is also why the chapters do not need to be read in order. They can be. The order I have placed them in has its own logic, moving from the most common categories outward to the rarer and stranger ones. But each chapter is a self-contained reading of one shape, and the reader sitting up at three in the morning has the right to skip to the chapter whose title sounds like the sentence they just heard.
A final word about register, because the genre's other gravitational pull is toward the inspirational, and the inspirational is the failure mode this book has to refuse in every paragraph.
The dying do not speak in order to inspire the living. They speak because they are dying and because something in them, sometimes against considerable physical odds, still has something it needs to say. Treating their sentences as material for uplift is a category error. It mistakes a report for a sermon. The clinicians whose work this book draws on understand this — Mannix's prose is calm, Gawande's is clinical, Clarke's is plain, Kerr's is restrained, Byock's is measured. None of them write to make the reader feel better about death. They write to make the reader see it more clearly. The hope, where it appears in their books and where it appears in this one, is a by-product of accuracy, not a goal of style.
I have tried to hold the same line. The author, in the chapters that follow, recedes. The dying speak. The clinicians who recorded them are named. The pattern is allowed to assemble itself. Where a chapter ends in something that resembles consolation, the consolation is the consolation the sources themselves report — that people, at the end, are very often more articulate, more loving, and more lucid than the living expect them to be — and not a consolation imported from outside the record.
This is the book I wanted to read after I had read the others and found that none of them, on its own, contained the chorus. Putting the chorus on one page is the whole of the project. The seven shapes are what the chorus turns out to be singing.
I started listening because the books I was reading kept handing me sentences I could not put down. I kept listening because the sentences, laid alongside each other, began to rhyme. The chapters that follow are an attempt to set the rhymes down in the order in which I came to hear them, with the names of the people who first wrote them attached, so that the reader, if they wish, can go and listen for themselves.