What is memory?

How circuits of neurons become the narratives of our lives.

“Close your eyes and think of a lollipop,” instructs Marvin Chun, Ph.D., professor of psychology and neurobiology. “Now, if I test you a week later by showing you a picture of a lollipop and asking, ‘Did you ever see this?,’ you’re probably going to say yes even though you just imagined it.”

Memory is a story. The entirety of our lives, real and imagined, is grounded in memories of learning, thinking, and doing, making our memories inextricably linked to our identities. Every time you recall something, a series of neurons lights up your brain, covering it like a patterned spider web of energy. Remembering the details of where you were, whom you were with, and what an event looked or sounded like causes your brain to activate and occasionally modify this pattern. Like a children’s game of “Telephone,” details can get altered or lost in translation; the brain doesn’t operate like a perfect videotape. But its changeability ensures that we can create and store memories in vast numbers of adaptive neural connections, constantly adding threads to the tapestries of our persons.

Where memories start

There’s a lot that scientists still don’t know about how we form and store memories, but the beginnings of one are easy to trace. All memories start with perception. Through thinking, touching, hearing, seeing, tasting, and smelling, we process our surroundings and things that affect us. “Today I walked into the door and bumped my elbow,” offers Jessica Cardin, Ph.D., assistant professor of neurobiology, as an example. “You have to have a context for that memory. You have to see that you’re in your office.” In the brain, the cortex receives all of this information from various systems and funnels it into the hippocampus, which creates associations throughout the neural network. But from there, everything depends on what type of memory is being formed.

Muscle memory‒learning how to play the piano or ride a bike‒uses a different part of the brain than remembering how you celebrated your birthday last year. Spilling coffee on yourself during a meeting is recalled differently from how to use a knife and fork, which is different from remembering the dates you studied for a history test the night before. Some memories are constantly reinforced by our environments and therefore made stronger‒it’s easier to forget where you placed your keys last night than it is to forget the route from your home to the office.

Mistakes, however, often occur in the perceptual phase. “People are terrible observers most of the time,” says Cardin, who studies the interplay between neurons in both healthy and diseased brains. Eyewitness testimony is notoriously fallible due to missing or imagined details. Psychologists have also conducted experiments in which a subject takes a sheet of paper from someone behind a desk, who then bends down to retrieve something. A different person behind the desk stands up‒and the subject rarely notices. The experiment isn’t meant to embarrass people, or to point out how little they remember about the people they just met, but rather as an example of how they prioritize where they direct their attention. In these cases, were the clients later asked to recall what the person behind the desk looked like, odds are they would offer only general details: gender, or perhaps race.

For items or events to which we do pay attention, we build chemical connections to consolidate a memory. Compared to the number of neurons in the average brain, a single person will have a seemingly unlimited number of experiences. “You can’t assign one memory per neuron,” says Cardin. Even with 100 billion neurons, there simply aren’t enough. Instead, rather than selecting one cell to hold all memories of a family member, we create a pattern involving only a portion of the available neurons, combining them in different numbers and sequences.

A discrete “thing” in the brain

Researchers once envisioned memory as a discrete “thing” in the brain, but now they search for these patterns of replay that indicate information being recalled. Your brain can ignite the pattern for recall in multiple ways, along any part. According to Cardin, smell is one of the best ways to activate a memory. The nerves that react to and process smells in the air connect directly to the limbic system‒which includes the hippocampus. We can also ignite patterns of recall by envisioning specific details or places. If someone asked you to remember something about your childhood, you might start with an image of your elementary school building or your first pet and build from there.

When we reactivate such a neural recall pattern, it becomes malleable. It’s unlikely that anyone would believe they actually saw a lollipop since reading Chun’s instructions, but that might change a week later if subjects were unaware of the test’s purpose. Memory is like a wax seal, Chun says: First, you heat up the wax to press an imprint into it, which then solidifies. But the act of retrieval warms up the wax again. “That malleability‒the fact that it becomes reprogrammable‒that physical fact is what allows memories to become stronger or weaker,” says Chun. “There’s no such thing as reading out exactly what was encoded. It all involves opening the files, making the wax warm again, because we’re just dealing with proteins in the brain.”

False memories can be created by something as simple as a leading question. “You can do it experimentally by having people witness an event,” says Cardin. A subject might observe a brown-haired person bumping into someone, with an experimenter afterwards asking the witness what time the blond person bumped into the other. A week or two later, when asked to describe the event, the witness is more likely to say that a blond-haired person bumped into someone. The immediate follow-up by the experimenter is crucial; when recalling the event later, the witness’s brain activates the pattern of neurons and therefore heats up the wax again, allowing the memory to be modified by the suggestive question.

Change occurs to some types of memory more easily than others. Memory of our personal timeline, referred to as “episodic memory” or memory with a date and time stamp, is most susceptible to alteration or decay. The memory of what you gave a friend for a birthday or who attended the party is less resilient than the memory of what purpose a chair serves. Similarly, forgetting that the sky is usually blue or that a zebra is striped would be unlikely without some type of brain damage. These learned memories, ingrained factual information outside our personal experiences, are categorized as “semantic memory.” Psychologists still debate whether the brain handles the information in one system or two and how much reinforcement the brain needs before a fact becomes part of semantic memory, but no one argues that a distinction exists.

Trauma as a research tool

Traumatic brain damage and neural disease provide critical information about what memory is and how it works. Some of the brain’s memory systems are particularly robust. While episodic memory is often the first to go, skill memory‒the eponymous “muscle memory”‒is one of the most resilient. Even amnesiacs don’t forget how to walk. Those who have difficulty forming new personal memories can still learn new motor skills, including complex tasks like writing backwards on a mirror. The brain stores skill memory in a different region from plain information.

When humans do forget things, the process may occur in a few different ways. The act of trying to remember is akin to searching for something on a computer hard drive or in a messy room. If the computer files are organized, or the room has been recently cleaned, the search is easier‒but with time and entropy, finding things becomes difficult. “Memory is the exact same thing,” says Chun. “Sometimes, it’s hard to find because of all the other stuff.” Whether memory patterns, if formed through proper attention, can ever be truly “deleted” from our brains is still an open question in psychology. However, they decay and can be overwritten by more important information.

This notion of forgetting or lost memories has long been a source of fascination for pop culture. Movies across all genres deal with the topic: The Notebook for Alzheimer disease; Memento or 50 First Dates for anterograde amnesia, in which the main characters cannot form new memories; The Bourne Identity and Eternal Sunshine of the Spotless Mind for a more science-fiction take on lost memories. “We don’t really understand forgetting that well,” says Chun, but he points out that an abnormal rate of memory loss is one of the most debilitating situations in which patients can find themselves: “It robs you of who you are and how people relate to you.”

Clinically, forgetting as a process receives significant research attention. Even though memory patterns are spread throughout the brain, conditions like Alzheimer disease hit memory the hardest of all the brain functions. “What we can do to stave off and understand that disease is probably our most pressing challenge,” says Chun. “The more you can understand memory and the more you can map it onto the brain, and the more you can understand when it goes bad and when you can preserve it‒what all memory researchers do is helpful for that effort.”

But our brains have also created forgetting as an adaptive function, and researchers are harnessing it to treat other disorders. With chemicals that block neurotransmitters, scientists are researching memory alteration as a potential therapy for post-traumatic stress disorder, drug addiction, delusions, and phobias. By activating a memory and then using an agent that weakens connections in the brain, researchers may be able to remove the emotional attachment or craving that accompanies the memory pattern and stop it from being reinforced.

Memory as a higher brain function defines who we are. We rely on our memories for experience, wisdom, relationships, flexibility, and any ability to learn. Even in Greek mythology, the Muses‒goddesses of inspiration and knowledge in the arts and sciences‒were daughters of Mnemosyne, the goddess of memory. But when our identities are predicated on a temporal continuity, what does it mean when it turns out we’re not exact documentaries but some flavor of creative, narrative film? Our brains simply do the best they can‒the rest is up to us.

  • Tags:
  • Faculty
  • Neurobiology
  • Research

Related People

Marvin M Chun

Dean of Yale College, Richard M. Colgate Professor of Psychology and Professor of Neuroscience

Jessica Cardin

Associate Professor Term