How to notice when you're changing.

You think back five years. You remember a version of yourself — slightly smaller, slightly more anxious, more easily impressed — and you find them gently embarrassing. You'd never make that mistake now. You'd never let that conversation happen now. You can see, very clearly, who you used to be.

You think forward five years. You can't quite picture it. Probably you'll be much like you are now. Older, perhaps a little tired, but mostly the same person, with the same views and the same friends and the same way of moving through the world.

Both of those intuitions are wrong, and the second is wrong in a more interesting way than the first.

The end-of-history illusion

In 2013, a group of psychologists at Harvard and the University of Virginia ran one of the more elegantly designed studies of the past two decades.¹ They asked over 19,000 people, ranging from ages 18 to 68, to do two things: report how much they had changed in the previous decade, and predict how much they would change in the next.

At every age, people reported that they had changed substantially over the previous decade — in personality, values, preferences, friendships. And at every age, they predicted that they would not change much over the next.

This was not a few outliers. The pattern was uniform. The eighteen-year-old expected the twenty-eight-year-old to be more or less the same person. The forty-year-old expected the fifty-year-old to be the same. The sixty-year-old, looking back at fifty, agreed she had changed; looking forward, expected stability.

The researchers called this the end-of-history illusion: the persistent intuition that the person you are now is the final version, and that your previous changes are evidence of your current arrival, not a process that is still happening.

The illusion is so reliable that it suggests something structural about the human mind. We can measure change in retrospect because we have two snapshots — past self, present self — to compare. We cannot measure change in prospect because we have no comparison data yet. The future self is invisible, so we assume it is identical to the present self.

This is the first reason most people don't notice themselves changing in real time. The brain is not equipped to do it.

How much you actually change

The popular notion is that personality stabilises around age thirty and then holds steady. The data say otherwise.

The largest meta-analysis on the question is Brent Roberts, Kate Walton, and Wolfgang Viechtbauer's 2006 review in Psychological Bulletin, which pulled together 92 longitudinal studies covering most of the lifespan.² What they found is that personality continues to change measurably across the whole of adult life. People become, on average, more agreeable, more conscientious, and more emotionally stable as they age — and the largest changes happen not in adolescence but between roughly twenty and forty.

This is not the dramatic change of a personal-growth memoir. It is gradual, low-amplitude, and easy to miss in real time. But over a decade, the cumulative shift is substantial. The person you were at twenty-five is not, in the technical psychometric sense, the person you are at thirty-five.

The same point appears in research on values. Across longitudinal studies of what people care about — including the relative importance of family, achievement, security, and meaning — values drift steadily across adulthood, often without the person noticing the drift.³ You become a slightly different person who wants slightly different things. The wanting feels continuous from the inside. The trajectory is not.

Why you can't see yourself moving

The reasons are roughly three.

Adaptation hides change. Hedonic adaptation — the tendency for sustained emotional states to become invisible — applies to your relationship with yourself, not just to external circumstances.⁴ If you are gradually becoming more anxious, or gradually becoming less reactive, the change happens at a rate slower than your noticing apparatus is calibrated to. By the time you would have called the new state different, it is the new normal.

You don't have your old data. When the friend says you're not the same person you were three years ago, she is doing something you can't do for yourself. She is comparing two snapshots that exist in her memory but not yours. Her version of you in 2022 is preserved in a way your own version of you is not, because she had less to lose by not constantly updating it. Your present self has been overwriting your past self all along, in service of a continuous identity. Hers hasn't.

The narrative wants continuity. The brain is a continuity machine. It produces a coherent first-person story whose central character is you, and any evidence that you used to be a different kind of person is integrated, smoothed over, and folded into the current story as background. I've always been like this. I've always cared about that. These claims are usually false in ways you have no access to from inside.

This is the second reason most people don't notice themselves changing in real time. Even if the brain were equipped to do it, the narrative apparatus would prefer that it didn't.

Writing as comparison data

The single most useful tool for noticing change is also one of the oldest: writing things down and reading them back.

A journal that is genuinely re-read produces what you cannot otherwise have — your old self, in your own words, before your present self had a chance to revise her. This is what most journaling apps quietly fail to deliver. The entries accumulate. They are never read back. The data exists, technically, but the comparison never happens, so the change remains invisible.

What you are looking for in re-read entries is not "have I become a better person." That question is the wrong question, and it tends to invite either self-flattery or self-criticism. The better question is something closer to: what did the person who wrote this care about that I no longer care about? What did this person worry about that I have stopped worrying about? What does this person not yet know that I now know?

James Pennebaker's research on linguistic markers gives one specific signal to watch. Across a large body of work, his collaborators have found that when people are processing something well over time, the language they use about it changes in measurable ways. Causal words ("because," "realised," "led to") and insight words ("understand," "see," "now I think") increase. Pronoun use shifts — first-person singular usually drops as the writer moves out of acute distress.⁵ You can do this analysis crudely, by hand, on your own old entries, and a great deal will become visible.

A second signal: the things you no longer write about. If a name that appeared in twenty entries last year appears in one this year, something has changed in that relationship. The change may not be conscious. The change may not even be welcome. But it has happened, and the absence of the name is the evidence.

What to look for, practically

If you wanted to start noticing yourself changing, you'd want a small, repeatable structure. Three things, roughly:

A semi-annual re-read. Twice a year, sit down for an hour with what you wrote six months ago and read it as if it were written by someone else. Underline the sentences that don't sound like you anymore. Note the worries that no longer feel pressing. Note the people whose names appear and the people whose names are absent. The differences are the data.

A small set of stable questions. A handful of questions, asked of yourself at regular intervals over a year, will track change far more usefully than a long, varied questionnaire asked once. What am I most afraid of right now? What am I most curious about? Who am I trying to be more like, and who am I trying to be less like? The stability of the question is what makes the answer comparable.

The ten-year prediction test. Once a year, write a short paragraph about who you expect to be in ten years' time. Don't try to be accurate; try to be honest. Then put it aside. Each subsequent year, re-read the previous predictions before you write the new one. You will find, after the third or fourth iteration, that you have generated for yourself the very thing the end-of-history illusion denies you: a record of how wrong you have been about your own stability. That record is, in itself, a measure of how much you have changed.

A note on what we're building

This is the question Loopn was built around. Most journaling tools store entries; very few read them back to you. The Sunday Loop is the read-back: a weekly summary that draws on what you actually wrote, including weeks and months ago, and surfaces where the language is moving — about your manager, your father, the project you keep avoiding. The point is not to celebrate your growth. The point is to make the change visible while it's happening, instead of only in retrospect.

A journal that doesn't watch you change is doing half the job. The change was always going to happen. The question is whether you got to see it.

If any of this resonated, you can read more about self-awareness is not the same as self-criticism or why your relationship patterns keep repeating.