Our understanding of memory is all wrong

Memory defines us in so many ways, but it’s not exactly what we think it is.

We tend to imagine memory almost like a filing cabinet — a faithful record of the past we can pull from when needed. But according to new research, memory is less about storing facts and more about shaping the story of our lives. It helps us make sense of the present and construct meaning out of chaos.

Dr. Charan Ranganath is a neuroscientist at University of California Davis, and the author of Why We Remember: Unlocking Memory’s Power to Hold on to What Matters. His work has transformed how scientists understand the mind’s most mysterious function. I invited him onto The Gray Area to talk about why forgetting is as essential as remembering, how emotion shapes what we recall, why trauma lingers, and how collective memory can bind — or divide — entire societies.

As always, there’s much more in the full podcast, so listen and follow The Gray Area on Apple Podcasts, Spotify, Pandora, or wherever you find podcasts. New episodes drop every Monday.

This interview has been edited for length and clarity.

You write that the most important message from memory science isn’t “remember more.” So what is memory for?

It’s not a vault that stores every experience. Memory is a resource we draw on to understand what’s happening now, to plan, and to anticipate the future.

When people say, “I have a bad memory,” what do you think they’re misunderstanding?

If someone truly had a “bad memory,” as in clinically impaired, they couldn’t function independently. I’ve tested patients like that. What most people mean is, “I can’t always recall what I want, when I want.” Often these are high-functioning people who expect to remember everything. That expectation is the mismatch.

But surely some memories are “better” than others…or is “better/worse” the wrong frame?

That’s the mistake. People conflate “more” with “better.” Take highly superior autobiographical memory: Some folks can tell you what they ate on March 7, 2011, who won a game, what the weather was. You might think they have a great memory. But they don’t learn a new language faster than anyone else. And many report it’s a burden; they can’t stop replaying minor negative moments. Some even call it a curse. So “more” isn’t necessarily “better.”

So in my case, I’d say that I have a somewhat weird and annoying memory. I can remember whole chunks of certain books or random trivia, but then I routinely forget faces and names — things I actually want to remember. Why?

There are two big issues. First, competition. Memories compete with each other. If my desk is piled high with near-identical papers, it’s hard to find one. Faces are like that: Most have two eyes, a nose, a mouth; they’re highly similar. Names are also similar, and the mapping between a face and a name is arbitrary. “Baker” used to refer to someone who baked bread; now it doesn’t. So you’ve got similar inputs and an arbitrary link. That’s interference.

The other issue has to do with attention. When you meet someone, your attention is split: noise, small talk, your own self-consciousness. If the name doesn’t get a clean “write-in,” retrieval later is shaky. It’s not that you can’t remember names; it’s that the name was never strongly encoded in the first place.

Can you give us practical hacks so we stop embarrassing ourselves at parties?

I’ll give you three quick ones and a couple bonuses:

How does memory change with age? What declines and what holds?

Separate episodic from semantic memory. Semantic is facts and knowledge — history, vocabulary, expertise. Episodic is remembering a specific event in time, like where you put your keys, details from a conversation, the unique texture of a day.

With aging, it’s mostly episodic memory — forgetfulness, names, details — that declines. Semantic memory often stays solid, even grows. Another piece: retrieval control. Older adults often know the actor’s name but can’t pull it up; proper nouns are notorious. That’s tied to executive function in the prefrontal cortex, which gradually declines starting around 30. It’s not just storing memories; it’s regulating attention and using strategies. That’s what slips.

Give me concrete examples of episodic versus semantic in everyday life.

Semantic: you know what a mortgage is, the rules of baseball, your profession’s jargon. Episodic: the first house you toured, where you parked today, the specific at-bat where your kid hit a double. When people say, “My memory is getting worse,” they usually mean, “My episodic memory is less reliable in the moment,” not “I’ve lost my knowledge of the world.”

So the throughline is: Memory isn’t free. You have to set an intention.

Exactly. Ask: What do I want to remember from this? If your goal is to remember names, start with that intention and deploy the strategy. If you assume memories “come for free,” you’ll be disappointed.

There are studies showing the happiness we get from our choices is determined less by what we experienced and more by what we remember. Why is that?

Because we forget most details quickly. Classic findings show you can lose around 60 percent of newly learned details within an hour. What remains are beginnings, endings, highs, and lows — the parts most useful for future decisions. The “remembering self,” not the “experiencing self,” often drives satisfaction later.

Can you give an example of how that plays out?

Think about a vacation. The experiencing self has 100 small moments — waiting in lines, one great sunset, one bad dinner. The remembering self condenses [it all] to a highlight reel: the perfect swim at the end (peak), the travel nightmare (low), the last night’s dinner (ending). That compressed story — more than the raw sum of moments — drives whether you say, “That trip was amazing; let’s go back.”

That suggests the stories we tell ourselves really matter. We’re constantly updating memories — they’re not etched in stone — so maybe we can deliberately remember the “right” things and be happier.

Memory is more like a painting than a photograph. Van Gogh’s church is recognizably that church, but it’s also his perspective: what he emphasized, what he left out, where he stood, the light he chose. Every time we remember, we create a fresh painting — same event, different vantage point. That’s why a terrible experience can become, later, a funny story. The facts don’t vanish; our relationship to them changes.

What about trauma? Those memories can feel impossible to shake. How are traumatic memories different? Are they different at all?

Emotionally significant events get biochemical priority…There’s also a distinction between details and the visceral punch — that surge of fear or pain. [These memories] rely on partly different systems. What makes trauma feel “photographic” is usually the emotion, not perfect detail. The goal isn’t to forget. You want to remember without the punch. You want to keep the information that keeps you safe, without reliving the threat. Sleep can help “de-charge” memories; sometimes it doesn’t, and that’s where PTSD arises and professional care matters.

Music seems to have a special power over memory. Why does a sad song cue sad memories, and a happy song cue happy ones?

Episodic memories are organized by context — place, time, internal state. Emotion is part of that context. If you’re sad now, you’re effectively searching the “sad” wing of the library. Music is a potent emotional cue, and it’s tied to identity across time. We pick music that matches who we are at different stages of life, so it becomes a vehicle to revisit who we were.

You write about memory shaping identity, and how our “self” is a narrative stitched together through memory. Is the self just a memory system holding a story together?

In some ways, yes. We have schemas — blueprints for weddings, basketball games, restaurant scripts — and we have self-schemas: stories about who we are. When we remember, we don’t just list details; we stitch causes and effects: I felt angry, so I left. Or: She left because of an emergency. Those causal links are how we make sense of the past and plan the future. The self-schema organizes those links across time.

If someone loses their memories, can they still be themselves?

People with dense amnesia can retain a thin sense of self — preferences, traits — but there’s often a flatness. Without episodic memory, there’s little ability to “mentally time travel” — to enliven who you were or could be. Many amnesic patients are “stuck” at an earlier age internally; they look in the mirror and don’t recognize the older face. The self can persist, but it becomes emptier without the episodic wellspring.

Do political movements do the same thing — stitch collective memory into a collective identity?

Absolutely. Families have a shared story; nations do too. When we tell each other those stories, everyone’s memory changes — yours, mine, ours. That’s powerful for transmitting wisdom, and dangerous because memory is selective and malleable. Authoritarian regimes know this: change the statues, rewrite the textbooks, restrict archives, and you reshape identity by reshaping memory.

As we become more segregated culturally and informationally, we end up with different memories of the same events. That feels pretty unsustainable.

It’s one of the biggest threats I see, especially as AI ramps up scale and speed. We used to share a limited set of fact-checked sources. That was imperfect, but also constraining. Now it’s easy to believe whatever feels good. Long-form podcasts, Substacks, YouTube — “let it all air and the truth will emerge” — doesn’t account for how memory works. We remember what fits our prior beliefs and emotional states, and we forget base-rate statistics because “nothing happened” isn’t memorable. Vivid anecdotes drive causal beliefs; numbers don’t.

How does that play out with misinformation?

Two ways. First, salience bias: The vivid story of a vaccine side effect outweighs 100,000 uneventful vaccinations. Second, social reinforcement: When a higher-status voice in a group asserts a confident — but wrong — detail, it spreads like a social contagion. People remember the confidently expressed version later. The fix isn’t one magical source of truth, it’s cultivating habits that counteract memory’s weaknesses: multiple sources, wait-time before sharing, explicit uncertainty, and exposure to disconfirming evidence.

How malleable is collective memory? Can a country rewrite its story as an individual can?

Collective memory magnifies the strengths and weaknesses of individual memory. That’s bad news and good news. The bad news is that it’s easy to steer with simple, repeated narratives. The good news is that groups can also change course if they build norms for cross-checking, include diverse perspectives, and keep an audit trail of sources. That slows the “contagion” effect and creates a richer, more accurate shared story.

So which comes first: memories or beliefs? Do memories form beliefs, or do beliefs curate memories?

Both. We build beliefs from what we’ve heard, done, and remembered. And beliefs filter what we can retrieve and how we reconstruct it. People tend to recall their past more positively than it was and remember themselves more favorably. Scale that up to a nation and you get, “We were great; we were wronged; we must be great again.” The narrative selects the memories, and the memories entrench the narrative.

There’s a question here about memory and forgiveness. Do we have to forget to move on — individually and socially? Or is that price too high?

“Forgiving is not forgetting; forgiving is remembering without pain.” That line captures it. Truth and reconciliation processes can help — or re-traumatize — depending on how the remembering happens. Just dredging up the past doesn’t heal. The work is to change your relationship to the memory: retain the information, reduce the toxicity. That’s true in therapy and in society.

What does “changing your relationship” look like in practice?

There are a few elements we see across approaches. You can contextualize causes. Understanding why something happened doesn’t excuse it, but it can reduce the sense of present threat. You can have different vantage points; telling the story from a future self, or from a compassionate narrator, shifts emphasis. You can add new meaning. Extracting what you learned or how it shaped your values rewires the “why this matters” tag. And you can aim for safe retrievals. Recalling in safe contexts, then returning to the present, teaches your brain the memory is not the event.

Any parting advice?

I’d add two things. First, the intention principle: memory isn’t free, easy, or absolute. Decide what matters before the moment — names at a party, details of a family trip, ideas from a book — and use strategies accordingly. Second, I’d love to make a plug for science. If we want progress on Alzheimer’s, trauma, learning, then we need sustained support for research. Call your representatives, express your support, make your voice heard. Those small actions matter.