I have few memories of being four—a fact I find disconcerting now that I’m the father of a four-year-old. My son and I have great times together; lately, we’ve been building Lego versions of familiar places (the coffee shop, the bathroom) and perfecting the “flipperoo,” a move in which I hold his hands while he somersaults backward from my shoulders to the ground. But how much of our joyous life will he remember? What I recall from when I was four are the red-painted nails of a mean babysitter; the brushed-silver stereo in my parents’ apartment; a particular orange-carpeted hallway; some houseplants in the sun; and a glimpse of my father’s face, perhaps smuggled into memory from a photograph. These disconnected images don’t knit together into a picture of a life. They also fail to illuminate any inner reality. I have no memories of my own feelings, thoughts, or personality; I’m told that I was a cheerful, talkative child given to long dinner-table speeches, but don’t remember being so. My son, who is happy and voluble, is so much fun to be around that I sometimes mourn, on his behalf, his future inability to remember himself.

If we could see our childish selves more clearly, we might have a better sense of the course and the character of our lives. Are we the same people at four that we will be at twenty-four, forty-four, or seventy-four? Or will we change substantially through time? Is the fix already in, or will our stories have surprising twists and turns? Some people feel that they’ve altered profoundly through the years, and to them the past seems like a foreign country, characterized by peculiar customs, values, and tastes. (Those boyfriends! That music! Those outfits!) But others have a strong sense of connection with their younger selves, and for them the past remains a home. My mother-in-law, who lives not far from her parents’ house in the same town where she grew up, insists that she is the same as she’s always been, and recalls with fresh indignation her sixth birthday, when she was promised a pony but didn’t get one. Her brother holds the opposite view: he looks back on several distinct epochs in his life, each with its own set of attitudes, circumstances, and friends. “I’ve walked through many doorways,” he’s told me. I feel this way, too, although most people who know me well say that I’ve been the same person forever.

Try to remember life as you lived it years ago, on a typical day in the fall. Back then, you cared deeply about certain things (a girlfriend? Depeche Mode?) but were oblivious of others (your political commitments? your children?). Certain key events—college? war? marriage? Alcoholics Anonymous?—hadn’t yet occurred. Does the self you remember feel like you, or like a stranger? Do you seem to be remembering yesterday, or reading a novel about a fictional character?

If you have the former feelings, you’re probably a continuer; if the latter, you’re probably a divider. You might prefer being one to the other, but find it hard to shift your perspective. In the poem “The Rainbow,” William Wordsworth wrote that “the Child is Father of the Man,” and this motto is often quoted as truth. But he couched the idea as an aspiration—“And I could wish my days to be / Bound each to each by natural piety”—as if to say that, though it would be nice if our childhoods and adulthoods were connected like the ends of a rainbow, the connection could be an illusion that depends on where we stand. One reason to go to a high-school reunion is to feel like one’s past self—old friendships resume, old in-jokes resurface, old crushes reignite. But the time travel ceases when you step out of the gym. It turns out that you’ve changed, after all.

On the other hand, some of us want to disconnect from our past selves; burdened by who we used to be or caged by who we are, we wish for multipart lives. In the voluminous autobiographical novel “My Struggle,” Karl Ove Knausgaard—a middle-aged man who hopes to be better today than he was as a young man—questions whether it even makes sense to use the same name over a lifetime. Looking at a photograph of himself as an infant, he wonders what that little person, with “arms and legs spread, and a face distorted into a scream,” really has to do with the forty-year-old father and writer he is now, or with “the gray, hunched geriatric who in forty years from now might be sitting dribbling and trembling in an old people’s home.” It might be better, he suggests, to adopt a series of names: “The fetus might be called Jens Ove, for example, and the infant Nils Ove . . . the ten- to twelve-year-old Geir Ove, the twelve- to seventeen-year-old Kurt Ove . . . the twenty-three- to thirty-two-year-old Tor Ove, the thirty-two- to forty-six-year-old Karl Ove—and so on.” In such a scheme, “the first name would represent the distinctiveness of the age range, the middle name would represent continuity, and the last, family affiliation.”

My son’s name is Peter. It unnerves me to think that he could someday become so different as to warrant a new name. But he learns and grows each day; how could he not be always becoming someone new? I have duelling aspirations for him: keep growing; keep being you. As for how he’ll see himself, who knows? The philosopher Galen Strawson believes that some people are simply more “episodic” than others; they’re fine living day to day, without regard to the broader plot arc. “I’m somewhere down towards the episodic end of this spectrum,” Strawson writes in an essay called “The Sense of the Self.” “I have no sense of my life as a narrative with form, and little interest in my own past.”

Perhaps Peter will grow up to be an episodic person who lives in the moment, unconcerned with whether his life forms a whole or a collection of parts. Even so, there will be no escaping the paradoxes of mutability, which have a way of weaving themselves into our lives. Thinking of some old shameful act of ours, we tell ourselves, “I’ve changed!” (But have we?) Bored with a friend who’s obsessed with what happened long ago, we say, “That was another life—you’re a different person now!” (But is she?) Living alongside our friends, spouses, parents, and children, we wonder if they’re the same people we’ve always known, or if they’ve lived through changes we, or they, struggle to see. Even as we work tirelessly to improve, we find that, wherever we go, there we are (in which case what’s the point?). And yet sometimes we recall our former selves with a sense of wonder, as if remembering a past life. Lives are long, and hard to see. What can we learn by asking if we’ve always been who we are?

The question of our continuity has an empirical side that can be answered scientifically. In the nineteen-seventies, while working at the University of Otago, in New Zealand, a psychologist named Phil Silva helped launch a study of a thousand and thirty-seven children; the subjects, all of whom lived in or around the city of Dunedin, were studied at age three, and again at five, seven, nine, eleven, thirteen, fifteen, eighteen, twenty-one, twenty-six, thirty-two, thirty-eight, and forty-five, by researchers who often interviewed not just the subjects but also their family and friends. In 2020, four psychologists associated with the Dunedin study—Jay Belsky, Avshalom Caspi, Terrie E. Moffitt, and Richie Poulton—summarized what’s been learned so far in a book called “The Origins of You: How Childhood Shapes Later Life.” It folds in results from a few related studies conducted in the United States and the United Kingdom, and so describes how about four thousand people have changed through the decades.

John Stuart Mill once wrote that a young person is like “a tree, which requires to grow and develop itself on all sides, according to the tendency of the inward forces which make it a living thing.” The image suggests a generalized spreading out and reaching up, which is bound to be affected by soil and climate, and might be aided by a little judicious pruning here and there. The authors of “The Origins of You” offer a more chaotic metaphor. Human beings, they suggest, are like storm systems. Each individual storm has its own particular set of traits and dynamics; meanwhile, its future depends on numerous elements of atmosphere and landscape. The fate of any given Harvey, Allison, Ike, or Katrina might be shaped, in part, by “air pressure in another locale,” and by “the time that the hurricane spends out at sea, picking up moisture, before making landfall.” Donald Trump, in 2014, told a biographer that he was the same person in his sixties that he’d been as a first grader. In his case, the researchers write, the idea isn’t so hard to believe. Storms, however, are shaped by the world and by other storms, and only an egomaniacal weather system believes in its absolute and unchanging individuality.

Efforts to understand human weather—to show, for example, that children who are abused bear the mark of that abuse as adults—are predictably inexact. One problem is that many studies of development are “retrospective” in nature: researchers start with how people are doing now, then look to the past to find out how they got that way. But many issues trouble such efforts. There’s the fallibility of memory: people often have difficulty recalling even basic facts about what they lived through decades earlier. (Many parents, for instance, can’t accurately remember whether a child was diagnosed as having A.D.H.D.; people even have trouble remembering whether their parents were mean or nice.) There’s also the problem of enrollment bias. A retrospective study of anxious adults might find that many of them grew up with divorced parents—but what about the many children of divorce who didn’t develop anxiety, and so were never enrolled in the study? It’s hard for a retrospective study to establish the true import of any single factor. The value of the Dunedin project, therefore, derives not just from its long duration but also from the fact that it is “prospective.” It began with a thousand random children, and only later identified changes as they emerged.

Working prospectively, the Dunedin researchers began by categorizing their three-year-olds. They met with the children for ninety minutes each, rating them on twenty-two aspects of personality—restlessness, impulsivity, willfulness, attentiveness, friendliness, communicativeness, and so on. They then used their results to identify five general types of children. Forty per cent of the kids were deemed “well-adjusted,” with the usual mixture of kid personality traits. Another quarter were found to be “confident”—more than usually comfortable with strangers and new situations. Fifteen per cent were “reserved,” or standoffish, at first. About one in ten turned out to be “inhibited”; the same proportion were identified as “undercontrolled.” The inhibited kids were notably shy and exceptionally slow to warm up; the undercontrolled ones were impulsive and ornery. These determinations of personality, arrived at after brief encounters and by strangers, would form the basis for a half century of further work.

By age eighteen, certain patterns were visible. Although the confident, reserved, and well-adjusted children continued to be that way, those categories were less distinct. In contrast, the kids who’d been categorized as inhibited or as undercontrolled had stayed truer to themselves. At age eighteen, the once inhibited kids remained a little apart, and were “significantly less forceful and decisive than all the other children.” The undercontrolled kids, meanwhile, “described themselves as danger seeking and impulsive,” and were “the least likely of all young adults to avoid harmful, exciting, and dangerous situations or to behave in reflective, cautious, careful, or planful ways.” Teen-agers in this last group tended to get angry more often, and to see themselves “as mistreated and victimized.”