A long time ago, I took a yearlong course of premed anatomy and physiology. Our professor, Dr. Welton, was tall and bald, wore a white lab coat, and knew a hundred of our names after the first week. I adored him and often attended his office hours, demanding to help him file or dissect.

Once, I asked him about memory. He had devoted a few weeks to the structure and function of the brain, but had barely mentioned memory. “Why are some memories more vivid than others?” I asked. We were in his office, standing in shadow by his desk. At a small lab table along one wall, a grad student prepared petri dishes. I waited for the mechanical answer—some neurochemical explanation for the floods and droughts of time, the way memory buries delight and shears off whole years. He was suddenly still, looking up to the corner where the wall met the ceiling. This man who always had an answer was silent for a long moment.

“I don’t know,” he finally said. It was the only time I heard him say that. “I remember being on a train, and I looked out and saw a window in a building. And it went by.” He looked back at me. “It was gone. And I remember that window as if it was right here.” He looked away. “I don’t know why.”

I am standing in a dry field on a moonless night. Men are lifting heavy boxes, mumbling, grunting, laughing quietly. My father is a broad silhouette against the sky. Then I see a spark, hear the whistle of flight and the whump of a shell exploding, its lucent flower filling the sky. This scene is so clear, and yet so dreamlike that I am not sure it is real. Many of my memories are like this: A single scene, a tableau as still and bounded as a nativity. I am climbing a dirty heap of snow. I am holding a pair of gleaming bronzed baby shoes. I hear the distant cocktail chatter of adults. When I find the edge and pull to see what comes next, the whole thing is jerked from my hand. I am left alone, climbing a heap of snow. I am left with rubble.

Memory is mostly accurate, most of the time. We agree on the general picture of what happened far more often than we don’t. In this sense, what I remember of the past is basically true. But for a writer, the details are everything. What shifts are often the small details.

We are always losing more than we keep. In 1885, the German psychologist Hermann Ebbinghaus created the “curve of forgetting,” a graph corresponding to a mathematical formula that measured how quickly people lose the information they learn. The loss is exponential; more than half of what we lose is gone in an hour. The curve flattens, but within a few days a person will forget about 70 percent of any information acquired, unless she makes a conscious effort to remember. Remembering requires repetition—but repetition, as we shall see, changes what we remember.

People insist that they do remember, of course. In a foreword to his fantastically detailed memoir, Speak, Memory, Vladimir Nabokov bemoaned his “amnesic defects,” the “blank spots, blurry areas, domains of dimness.” But in fact he was proud of his recall. When he found himself at odds with a fact-checker at The New Yorker regarding the color of the funnel on a French ocean liner, Nabokov chose to omit the reference rather than admit he’d gotten it wrong.

When studying memory, neuroscientists sometimes refer to “persistence” and “transience.” Is memory a recording? A storehouse? A catalogue? More like vapor. There are mechanical explanations for the droughts and floods, but researching them is challenging, the operations strange, and the conclusions something of a disappointment. No rabbit emerges from the hat. While reading about how RNA-binding proteins affect the long-term potentiation of neurons, with a jolt I am reminded—that word, re-minded—that this is you and me. This is all of it. Making love. A near-mortal wound. The dried salty grit on your hands at sunset, returning to shore. The heap of snow. All of it.

I have been studying the science of memory, those mechanical explanations, for some time now. I can tell you that procedural memory (how to brush one’s teeth) is different from episodic memory (I brushed my teeth this morning), which in turn is different from semantic memory (I brush my teeth to avoid cavities). I can tell you this, and a little about those binding proteins, but I can’t explain memory. No one can. All these types of memory overlap and blur; little of what we know is purely one or the other. We live in a matrix of events, facts, meanings, contexts, brief jottings on a notepad. As a writer, I am generally concerned with what is called autobiographical memory, the broad tapestry of experiences that are unique to a person and—importantly—linked over time. We are not one self; our selves give way, each to the next in a ceaseless parade. But they are intricately connected. My autobiographical memory includes digging a mud puddle, taking an anatomy exam, sitting by my mother’s deathbed, and, now, drawing the line that ties all these moments together into a package I call me. This is the foundation of the conscious self: I am who I am because this happened. We trust it implicitly. We tend to forget that there is a gulf between autobiographical memory and autobiography itself.

I write memoir sometimes, which is to say I write about the past. For a long time, I didn’t question this—neither the past nor my ability to know it well enough to write about it. Every writer makes a contract with the reader. That contract may be filled with fine print and subordinate clauses, but a few points are there in bold type. The story you are reading now is an essay, which means that it is not fiction. You are trusting me to tell the truth, that is, to use facts and lived events. I promise, pinky-swear, this is true. The same promise is made by memoir and autobiography. There is no firm line separating these forms; most definitions focus on the supposedly neutral, factual approach of autobiography and the personal, more emotional nature of memoir. Yet we are supposed to believe that both forms describe what happened.

Memoir makes a peculiar promise: it offers a true story rooted in the writer’s unique past, known only to the writer. I think of memoir as a revisitation and a retelling of what happened with a delicate and honest overlay of how it felt and what I believed and what it means. I write to evoke—an emotion, a state of mind, the stillness that overcomes one when the past intrudes. Often I am writing to evoke the peculiar, ineluctable sensuality of childhood, the world immense and intimate and present with the child at the center. I evoke, but I want also to invoke, to charm something dead back into life—perhaps that center of sensuality most of all.

I grant that this is an interpretive art. The point of memoir is memory, and memory can be little more than tattered prayer flags. So I am drawn to work like Angela’s Ashes, in which Frank McCourt uses dialogue sparingly, without quotation marks, letting the kind of conversation a child overhears become part of the description, like paint on the walls. The reader is always with McCourt in his memory, knowing it is memory.

I think most other memoirists would agree with my definition, but we often end up in different places. I have always maintained a strict (and, to many writers, too rigid) standard: no composite characters, no re-creating scenes I don’t honestly remember, no invented dialogue, no compressing of time or space or moving events around. I am out of fashion in the present vogue for unfiltered confessions, with my bending over backward to avoid signs of self-indulgence—a gift from my family. And I know, have known for a long time, that this avoidance is part of the story I am writing whether I intend it to be or not. Readers have sometimes taken me for a recluse of some kind. There are many kinds of recluses, of course.

I try to verify every fact I can. I read a lot. I use books, newspaper archives, photographs, maps, interviews, yearbooks, journals, letters, visits to museums and historical societies, just as I do research for an essay like this one. I also use myself: I invade, steal, rewrite, disavow, and in many other ways make use of everything I’ve written, considering it a good primary source. Still, I want to be willing to change my mind about what I remember, to admit that I don’t remember and sometimes remember wrongly.

I try to read new memoirs, but I often set books aside, unable to suspend my disbelief. Jeannette Walls’s The Glass Castle includes a nuanced description of the author’s serious burns at the age of three: the event, the hospital stay, the aftermath. Beautiful Country, by Qian Julie Wang, details a scene of the author’s immigration to the United States at age seven at such length and specificity as to surely be beyond the power of most adults to remember. Maxine Hong Kingston’s much-admired The Woman Warrior opens with a monologue by her mother that stretches more than two pages. Kingston says nothing of how she remembers this. She recounts many lengthy, and perhaps invented, conversations.

It is tempting to substitute today’s psychological truth for history. Memory is wet sand. This is what I want to interrogate: the slipperiness, the uncertainty. What is it about that window?

“Fresnel Lens,” by Patricia Voulgaris © The artist

I have always been bothered by memoir writers who are obviously making stuff up, but I am now also bothered by the possibility that we are all making it up, all the time. I know that I have blended and inverted events over time, perhaps invented one or two, and it startles me to discover this. I long to correct the record. But what have I missed? I started the research for this essay wondering how often my memories are false. As I write now, I wonder whether anything I remember is true.

Some of what we remember didn’t happen at all, and a great deal that did is not remembered. A small child knows faces, names, new words, how to use a toy. But she won’t remember her second birthday party. Children do not form retrievable autobiographical memories until around age three, a phenomenon called childhood amnesia. Such amnesia may exist because of the immature structure of the brain or because small children lack the language with which to form and express a memory. The answer is likely a combination of factors. Not until the age of around ten does a child’s autobiographical memory begin to seem like that of an adult.

People argue about this. (Some write memoirs.) They do remember, they say, of course they do. But many of our earliest memories are based on photographs or stories. One group of researchers describes this as a “scaffold” on which we construct a past. People recall details of things they cannot know, and know things they cannot recall. This is called belief without recollection. You may know that you had an operation when you were a baby. I know that my mother had a miscarriage when I was three. I know that I took swimming lessons, but I don’t remember them—I just remember being able to swim. Freud said of the confusion of memories he encountered in analysis: “It is difficult to find one’s way about in this.” In crucial ways we are made before age ten. So the conundrum appears: How can I possibly know who I was? I can’t, and yet I do.

A single memory is roughly the result of three separate actions: encoding, the creation of a pattern; consolidation, the storing of that pattern; and retrieval, the re-creation of that pattern. Imaging can now show us a little of this activity, but it can’t tell us what is happening in the mind. Our understanding of how memory works is so rough as to seem less like “imaging” than a crude drawing.

A man lifts a box, murmuring. The complexity this image represents is mind-boggling. He murmurs, a shadow in a shadow, and in a barely measurable instant, a surge of activity pulses across a few of the millions of miles of nerve connections in the brain. A complex of pathways is laid, much of it scattered in the medial temporal lobe. The hippocampus is critical, but memory isn’t filed any one place in particular; it shatters into spring rain. Memory is a kind of ceaseless remodeling. The pattern won’t stay without consolidation; it slides down the steep Ebbinghaus curve. Consolidation is a cascade of processes that may work by changing membrane strengths and voltage gradients in neural cells. Consolidation takes seconds or minutes, sometimes days or longer; many studies show that sleep aids consolidation. Then newly traced pathways are linked to earlier ones, and together these are organized in obscure ways that change over time as new traces are created, forming roads with many entrances and exits. And after the pattern is laid and patted into place, what remains? Potential. Capacity. Despite the microscopic shifts, memories don’t exist, exactly; they are not fixed objects. They are more akin to templates or molds—it is their empty space that counts.

How do we find the pattern again? In 1932, the great psychologist Frederic Bartlett performed a series of simple, ingenious experiments. He reported the results in his book Remembering. The work upended the field of memory science. Bartlett was able to demonstrate retrieval as a kind of creation. “Remembering is not the re-excitation of innumerable fixed, lifeless and fragmentary traces. It is an imaginative reconstruction, or construction,” he wrote. “It is thus hardly ever really exact, even in the most rudimentary cases.”

Bringing a memory into awareness is the work of an elaborate machine across a vast but measurable space. The machine does this over and over again, but never in exactly the same way twice. Each retrieval must be unique because the machine itself is always changing, updating, breaking down, repairing. A conscious memory is a combination of the specific pathway encoded at the time of the event and a lot of subsequent knowledge, inferences, beliefs, and experience. Inference but also interference. We get new overlays all the time—experience in the same place or with the same people, similar events, similar views, new knowledge. The brain sweeps this flotsam into the pattern every time we retrieve it. Siblings reconstruct memories from similar fragments but different experiences. Thus our memories are rarely exactly the same. Much diverges with time. In 1962, Daniel Offer interviewed seventy-three fourteen-year-old boys about their lives. He asked about relationships, punishments, families, dating, school. In 1996, he interviewed sixty-seven of the surviving seventy-one men, now forty-eight years old, using the same questions. Comparing the answers across the distance of more than thirty years, Offer found that virtually all their memories had altered so much that their “recollections were about the same as would have been expected by chance.”

Twenty-eight years ago, I wrote an essay called “The Basement,” about my brother and sister and me being required to play in our grandmother’s dim and empty basement when we visited her. I described what I saw as my older brother’s bravery with our unpredictable father, and mentioned that he hated football but knew quitting wasn’t an option. I called my sister “squeamish, chubby, pale, and black-haired—she’s the one left out, the baby.”

I sent my brother the essay, and he told me a few months later, “I liked football. Don’t know why you thought that.” Years later, my sister, angry with me about several old wrongs, complained about my description of her. “I was not,” she said. “Not like that.”

Yes she was, I thought. I pulled out old photos, their scalloped white edges beginning to curl. She is crying in many of them, forever left behind when my brother and I took off. But she’s not chubby. Why did I remember it that way? I was wrong. And I was right, for she was what I meant by that word—fragile, weak in a world where weakness was lethal. And why did I say my brother hated football? Because I felt the pressure he was under, the steady squeeze of our father’s demands. The psychologist David Pillemer says that we can, “from a functional perspective,” think of memory not as a mechanism for truth but “as a belief system.”

Perhaps, as many researchers think, our memories are as good as they need to be. Memoir writing aside, we function just fine with all this forgetting, because reconstruction is adaptive. And there is social merit in reminiscence; it strengthens intimacy and community. But reconstruction also allows us to imagine the future. Our false and shifting memories of the past don’t matter to anyone but ourselves. The future only cares about what we learn from them.

Even though I know that memory is not a recording, not a single thing, it feels as though it is. I smell chlorine’s damp tang and my brain flings me back into the deep end. I am in the dim locker room, wet concrete cool against my bare feet. A stretch of water flares in the light. The diving board sings its comic bounce. Autobiographical memories seem to be linked to a kind of network, so that the sound of the diving board brings the river to mind. And that links to the strange afternoon when we found a giant fish hovering under a boulder, and my brother and I tried to lasso it. I am treading slowly in ten feet of water. My brother is anchored on the beach, holding one end of the rope. I take a deep breath and turn over, swimming down to where sound thickens like porridge, the loop of rough rope in my hand. The fish hangs under the sunken boulder, one clear eye on me. I kick closer, rope at the ready. I am who I am because this happened. Because all this happened.

I think this happened.

Jules Feiffer says that writing is “mainly an attempt to out-argue one’s past.” We have a “self-enhancing bias”—that is, our memories gradually make us look better to ourselves. We arrange scenes in the most flattering way. The brain edits, gently but insistently shaping the wisps into coherent scenes. We may not even notice this shifting of perspective. One tends to remember a recent event from the position of the protagonist, living it again. But over time this perspective shifts to that of the observer. We watch ourselves.

One of my earliest bright memories comes from when I was six. My family was at the boardwalk amusement park in Santa Cruz. My little sister and I rode electric cars along a track, pretending to steer. Then our car stopped in the middle of the ride. I knew I was in charge of my sister and so I climbed out of the car onto the track and helped her out. I was rescuing her. I could hear distant shouts and see my mother waving from the bridge. Then a man ran out and met us, and that is the end of the memory. I see it only from above. I am standing beside my mother, watching two small girls in danger. I am yelling with the rest of the adults. And surely this memory is only a fragment, a scene built out of the many times I heard the story told.

Many memoirs are born of vibrant memories like this one. The phenomenon known as flashbulb memory usually refers to recollections of consequential events like assassinations or the attacks of 9/11, our memories of which may seem almost photo-realistic in retrospect. (Though they are sometimes wrong.) Most people also have flashbulb memories of ordinary events with no clear significance. You remember a window, and you don’t know why. I vividly recall perching in a cedar tree, swaying safe and high above an uncertain world. I can recall, as though it is happening now, the back patio on a summer evening as I take my turn at the handle of a big wooden ice cream maker. The cold, salty water pools under my feet. Adults mill about, cocktails in hand, and I lean in to the joyful work in the sweet twilight. And in one breath, I am again sitting at the dinner table when my father starts yelling, pulling my little sister out of her chair, spanking her; she is hollering and my mother is crying and I am leaping up and shouting at him to stop, to leave her alone, to leave us all alone. Then the film breaks, and there is nothing else—after the ice cream, after the shouting, what?

And so we mold our pasts into a story that may bear little resemblance to the genuine mess of actual life. When I write from memory, am I writing a history or a story? Isn’t it both? In the Eighties, the influential psychologist Jerome Bruner popularized what came to be called narrative theory, writing of memory that “we seem to have no other way of describing ‘lived time’ save in the form of a narrative. . . . We become the autobiographical narratives by which we ‘tell about’ our lives.” The story seems to come from the self, but in fact the self is partly constructed from the story.

My story: A girl with a stout heart. She holds her ground. She is lonely and fierce and does not believe everything she is told. A feral child, at home with lizards and trees. An odd girl anchored to a suffocating and unpredictable family. We rarely touched, never talked plainly. And I became a writer who prefers implication to explanation. I used to call my story one of self-reliance and independence: I left home at sixteen, my father’s predictions of disaster trailing me out the door, and made my way forward. But this story has a deeper theme: you have to take care of yourself, because no one else will. It is a story I have spent decades rewriting.

The psychoanalyst Donald P. Spence proposed that we carry two kinds of truths, one historical (verifiably true) and the other narrative (perhaps not true at all). For analysis, Spence thought, the latter is much more important. The many proponents of narrative theory today seem to assume that we are all sifting through debris to find the structure underneath—and not just that we do this, but that we have to do this, that sifting debris is a fundamental human need, that a self has to narrate to exist. But what if all the narration is a dream? We may know our history, the timeline of events, the key experiences, but many of us still seek a throughline. A unifying meaning or moral. We become both narrator and protagonist—because there has to be an explanation for all this. Doesn’t there? The trick is that sooner or later you have to climb down from the cedar tree.

Many people have written detailed accounts of the past, without comment: Virginia Woolf, Jean Rhys, Isak Dinesen, John Steinbeck. A few admit to imperfect recall: Janet Malcolm, Howard Norman, early Mary McCarthy. Others explore it: John Updike, Zora Neale Hurston. Some say they don’t care about the imperfections: late Mary McCarthy. Martha Gellhorn: “I forget places, people, events, and books as fast as I read them. . . .the situation is hopeless.” She wrote copious memoirs anyway. John Berger thought the autobiographer was freer than a novelist. “What he omits, what he distorts, what he invents—everything, at least by the logic of the genre, is legitimate.”

In her book on memoir, Mary Karr writes of McCarthy’s Memories of a Catholic Girlhood, published in 1957, and of McCarthy’s concern for the truth. To Karr, McCarthy pursued a standard appropriate for “histories and biographies and journalism.” Perhaps people of her time “were more gullible or more secretive or the standards more rigorous.” These days, the peculiar promise that memoir is true is more a wink and a nudge than a solemn handshake. “My own humble practices wholly oppose making stuff up,” writes Karr, who then lists what she actually does: invent dialogue, change names and details, compress time, and describe details that she didn’t observe in the moment. How Karr reconciles this dissonance is never clear. (It is not even explored.) The gap between my picky fact-checking and her pages of imaginary dialogue does not seem to me a matter of degree. I cannot remember what the adults milling around our ice cream socials said. I can guess, but a guess is all it is. Without it, I lose a little of the delicious intimacy of a scene, but otherwise, I am—and there is no way around it—making stuff up. Perhaps I am just as dissonant as Karr in the story I’ve come to tell. But I have tried to simply tell it, without adornment. Karr, seemingly at ease with the central conflict of our life’s work, adds that “deceit in memoir irks me so badly.”

If any one thing distinguishes current memoir from its progenitors (besides a concern with accuracy), it is the claiming of victimhood. In his study of American autobiography, Herbert Leibowitz proposes that its “grand theme . . .almost its fixation, is the quest for distinction.” Today autobiography seems to be a litany of injury, the recounting of loss and harm caused by abuse, racism, abandonment, poverty, violence, rape, and struggle of a thousand kinds. The reasons for such a shift in focus, a shift we see in every layer of our social, cultural, and political landscapes, are beyond my scope. One of the pivotal purposes of memoir is to unveil the shades of meaning that exist in what we believe. This is the problem of memoir; this is the consolation of memoir. Scars are fine; I have written about scars; it is the focus on the unhealed wound that seems new.

William Gass at his most disenchanted said of the “vulgar copulation” of history and fiction that “nowhere would one find the blend better blended than in autobiography.” To misuse history in this way “can only be to circumvent its aim, the truth, either because one wants to lie or now thinks lying doesn’t matter.” Or, he added, “because an enlivened life will sell better than a straightforward one.” More even than the wink, I find a dismissal, as though none of this is worth noting. Of course, the writer seems to say, I don’t remember this. But it’s a good story and it feels true. Many writers no longer pretend otherwise: of course the story is manipulated. Our ordinary lives are pretty messy, with a lot of filler, so let’s dispense with that. In which case, anything goes. Geoff Dyer: “Everything in this book really happened, but some of the things that happened only happened in my head.”

Things have gone exactly as Gass feared: mundane history has gleefully mutated into a grand reveal, an almost genetic modification of memoir into autofiction. This is personal history as an imagined journey, a felt truth in a manipulated past. Is autofiction the opposite of memoir or its evil twin? Do they stretch in opposite directions until they meet? This is one of the fundamental excuses of the memoirist: it may be untrue, but that doesn’t mean it isn’t correct.

Amnesia, while annoying, is less interesting to me than the honest mistakes, the ones we don’t even know we are making. I pulled out one of my own books recently, searching for a reference I could barely recall. Suddenly I was reading the story I was trying to write. I had forgotten that I’d written it already.

We swing between the poles of persistence and transience, and we all suffer from what one scientist refers to as a “proneness of memory to error.” As soon as a memory is activated, it is suddenly fragile again, subject to interference. It must be reconsolidated every time. With each pass, tiny deformations appear. You can’t tell what has changed, because each time you recall a memory it feels correct. Neruda: “Many of the things I remember have blurred as I recalled them, they have crumbled to dust, like irreparably shattered glass.” The act of studying our own past—sorting family photos, trading stories with siblings, writing memoir—destroys it.

I wrote memoir when I was young. It didn’t occur to me not to, that the events were too close. But of course all those stories are different now. That’s the thing. The story never ends, it is retold and remade, over and over, minor characters stepping forward, a major aria fading out. And the gun lying on the table in the first act will inevitably go off toward the end.

Most wrong memories are accidental. The brain mixes things up, the way our cousin shows up in a dream about the office. Reaching for peach, it finds apricot. Both carry the peculiar tingle of the familiar, and most of the time either will do. Neuroimaging studies of the past few decades have shown a little success in distinguishing false memories from true, based on a faint neural sensory signature. But we don’t live inside an MRI machine. The brain does not monitor itself for truth; once encoded, a memory is simply there, no matter how false it is, and a false memory can be as vivid, detailed, and laden with emotion as any other. The same neural networks are at play, and most of the time, we can’t tell the difference.

Neuroscientists use the term “imagination inflation” to describe one way that falsehood creeps into memory. Simply imagining an event several times can create a memory in essentially the same way as having the experience. The same voltage spark, the same synaptic tinsel. Such inflation is strongest when a person questions the memory and then writes about it in detail. Add fake evidence and a little persuasion, and you have “false feedback.” With relative ease, researchers have convinced people that they have had accidents, met famous people, nearly drowned, gotten lost, and committed crimes. In 2015, Elizabeth Loftus and Lawrence Patihis were able to convince more than 30 percent of research participants that they had seen a video of the crash of Flight 93 in Pennsylvania on 9/11. No such video exists. The participants maintained this false memory over time, even under rigorous questioning. The researchers concluded that “the misleading information has irrevocably replaced” the truth.

We can believe without recalling. But we can also recall without belief. A false memory can be undermined. With enough time, hard evidence, and supporting testimony, I could convince you that you have remembered an event wrongly. This is not easy to accomplish; people hold on tight. And I can’t take the memory away. Perhaps the most important point I’ve absorbed from months of reading about memory is that even when a memory is proven wrong beyond all doubt, a person still remembers it.

Suggestive responses help to create false memories, and it gets easier if the suggestions come from an authority figure. Every time I am rewarded for a memory—It’s true—I brighten the memory a bit. Every time I write about a memory, I am telling myself, That’s right. That’s what happened. Reassured, I strengthen the scene.

I used to think that I would be a good eyewitness. Now I no longer trust eyewitnesses at all. I am struck by how autobiographical memory is formed in the first place. Children live in a whirlwind of imagination inflation. Our parents and teachers have their own unreliable memories, their own constantly shifting contexts and biases, which they are only too happy to share. “You were so fussy when you were little. Don’t you remember all those tantrums?” “Wasn’t it exciting when Santa came last year?” All families have myths, and many families are fond of recalling them, passing stories about your childhood and your Uncle Joe and how Maria broke her leg around the dinner table like gems. Perhaps false memories feel true because of how often we visit them. Pillemer, among others, has studied how the ways parents reminisce with their children have “profound implications for the types of people that they will become.”

My parents and grandparents weren’t tale-tellers. We heard a few, the same ones told now and then: Grandpa driving a truck on the set of Errol Flynn’s The Adventures of Robin Hood. Our Uncle Gus panning for gold up in the hills. But never a word about my father’s service in the war. Not a word about how his father had died when he was just sixteen. I was endlessly curious, but too reserved (too proud) to ask questions; questions tended to be met with blank looks. Instead I explained things to myself and grew up to be a writer who likes to leave trails of breadcrumbs instead of explanations. Parental narration is not neutral. Parents make mistakes. Parents lie. The policeman will get you if you don’t behave. Boys don’t cry. We believe it all.

Since our memory is both true and false, so must be our memoir. Why not write the story that supports the inner truth—the narrative truth? Plenty of writers say that memoir is not about memory at all; that it is an attempt to explain oneself, to create coherence. To win an argument. I am casting shadow and light all the time; you have no idea what I’ve left out. But the rubble is what interests me, the fact that our pasts are unreliable. The fact that I am not sure and can never be sure. I want to explore what it means not to know, not to ever be able to know. Life is dead ends, conflict, dissonance, gaps, great clouds of confusion and misunderstanding. Do I tell a story, or do I tell you how it feels to have only the remains of one? The first is certainly a better story. But the second is better history. Which do I really want?

The philosopher Galen Strawson rejects the idea of life-as-story out of hand; he is antinarrative. Strawson, like me, doesn’t believe in a pearl of self traveling intact through time. We are each watching what seems a ceaseless parade of selves. He doesn’t experience his past or future as belonging to the self who is thinking about it. Saying This happened to me is a fallacy, and therefore I am who I am because this happened to me is a gross fallacy. The impulse to organize our lives into stories is “essentially a matter of bad faith, a radical (and typically irremediable) inauthenticity.” If everything we remember changes each time we remember it, if our efforts to remember are influenced by every comment and story and photo we encounter, if all that effort of sorting and talking and writing about our pasts moves us ever more toward watching an imagined self, how can we know ourselves? It is difficult to find one’s way about in this.

Did my father really throw the dog in the river that time? I think he did. I can see it happening, his meaty arms scooping up the shivering dog, the dog scrambling in the air, slamming into the water, climbing out onto the rocky beach. My father laughing. When I try to investigate this, when I question it—that beach wasn’t rocky, it was sand—all I find is an old photograph of my father squatting beside the dog on the riverbank, watching the camera with a curiously sad gaze. All I find is an unhappy man, a cowed dog, and the tapestry of living with him, with his sorrow and anger and power, for sixteen years.

Dr. Welton, one of the most sober people I have ever known, remembered a window with a shiver of pain and desire. I expected him to give me a kind of mechanical explanation, yet he had nothing. What I remember decades later is his face, suddenly soft. His brisk speech slowing. His lambent nostalgia. It knocked me sideways somehow, to see his wistful, wishing glance at an empty corner that only he found full. In a very rich year of human anatomy and physiology, that moment taught me as much about being a human as anything he said in his lectures. The cedar tree, the shouting. Mere belief.

This writer’s self can’t stop telling stories, but I may never write memoir again. At least, I won’t make the same promise. I can’t. This doesn’t feel like a loss or a change in the script; I am working on a book about the past right now. But the interrogation has changed. Lived life is past and present and future all receding at once. What we long to hold on to, we lose; what we remember is often what we would just as soon forget; the future is always bearing down, an endless distraction. I know myself as a glitter of synaptic activation, a flimsy thing easily swept aside. A ceaselessly increasing sum materializing out of nothingness, each integer instantly flung behind me. I am persistent. I am transient. Memory is not a fixed object, and neither am I.