Your Brain Might Only Have One Way Of Remembering
Are there different forms of memory? Psychology seems to think so, but the evidence is getting muddier.
Think back to your gran handing you a biscuit when you were five. You can probably see the kitchen. You might even remember which tin it came from.
Now think of Pythagoras' theorem. You know it. You have no idea which Tuesday afternoon it went in.
For about 50 years, psychology has said those two memories belong to different systems. A study published in January 2026 suggests the brain may not have got that memo.
A misconception of memory
The idea goes back to the psychologist Endel Tulving in the early 1970s. He split long-term memory into two types.
Episodic memory is memory for events. It comes with a time and a place attached, like the biscuit, or the day you fell off the swings at the park. Researchers sometimes call it mental time travel, because you can step back into the moment.
Semantic memory is memory for facts and general knowledge. It has come loose from wherever you first met it. You know Paris is the capital of France, and you know a² + b² = c². The lesson where you learned them is long gone.
The split made sense, and it stuck. It stuck so well that researchers mostly studied each type on its own. Very few experiments put both kinds of memory through the same test, with the same people, on the same day.
Forty people, a scanner and some logos
A team from the University of Nottingham and the MRC Cognition and Brain Sciences Unit at Cambridge ran exactly that kind of test. You can read the paper yourself here: Nature Human Behaviour.
They put 40 people in an fMRI scanner and asked them to recall pairings between logos and brand names. There were two versions of the task.
- The semantic version used real pairings. You see a logo you already know and recall the brand that goes with it.
- The episodic version used pairings the participants had learned in an earlier phase of the study. The only reason they knew them was that session.
The clever part is how alike the two tasks were. Same kind of picture, same kind of answer, same scanner. Earlier studies had compared tasks that looked nothing like each other, so any difference in the brain could have come from the task instead of the memory.
What we saw
When people remembered successfully, their brain activity looked almost the same in both tasks. The team found no measurable difference between episodic and semantic recall, and a lot of overlap in the regions involved.
Nobody on the team saw that coming. In the university's write-up, Tibon said: "We were very surprised by the results of this study." Decades of research had pointed towards clear differences. She describes the differences they did find as very subtle.
So the two boxes may turn out to be one box with two labels on it.
But, before we get carried away. I spent years in science, so I can't write about a study without a list of caveats. Here are four.
- It's one study with 40 people. That is a normal size for brain imaging, and the result still needs repeating.
- A scanner can miss things. fMRI tracks blood flow, which is a slow and fuzzy signal. Small or fast differences between the two memory types could slip past it.
- Logos are a narrow kind of memory. Recalling a brand name is a long way from remembering your wedding day or how photosynthesis works.
- The study looked at recall. It tells us about getting memories out. It says much less about how they get in, which is the bit teachers and designers lose sleep over.
The finding still stands. The caveats tell us how much weight it can carry.
Why this matters outside the lab
Dementia and Alzheimer's could be directly affected by this rebranding of memory. If recalling facts and recalling events draw on much of the same brain, then support for memory loss could be designed with the whole brain in mind.
Anyone who has spent time with a person living with dementia may recognise something here. Photos of old friends, favourite places and childhood homes often bring a person back into the room for a while. We tend to file that under episodic memory. Those pictures are also full of facts, like names, streets and who was married to whom.
That last part is my own speculation and the study did not test it. It does make me look at a memory box in a care home a little differently.
So what do we do with our courses?
People who design learning have a favourite belief. Wrap the content in a story, an emotion or a memorable moment, and it will stick better than a plain fact ever could. Many of us quietly treat "episodic" as the premium route into memory and drill-and-practice as the budget option.
This study should start to make you question that belief. If facts and experiences come back out through much the same machinery, the gap between "meaningful" learning and "repetitive" learning may be smaller than we assume.
Here is where I have to be careful, because this goes well past what the study tested. Nobody in the scanner was sitting through a compliance module. With that warning on the label, here is what I'm taking from it.
- Retrieval practice deserves more room. Quizzes, flashcards and spaced recall already have strong evidence behind them. This is one more reason to stop treating them as the boring bit.
- Stories are still worth it. They hold attention and give people a reason to care. They may simply have less of a special hold on memory than we like to claim.
- The two can work together. A scenario followed by a few rounds of recall or application is cheap to build and covers both bets.
- Our own assumptions need testing. If you have always believed story beats repetition, try both on a real module and look at the results.
One filing cabinet
Neuroscientists have suspected for a while that the wall between episodic and semantic memory was thinner than the textbooks said. This study gives that suspicion some solid experimental backing with a fair test (finally).
But it is only one study, and the textbooks can stay on the shelf for now. I still think it's a healthy reminder that the tidy categories we teach are our way of organising the brain.
I'll keep the stories in my courses. I'll also stop apologising for the quiz at the end.
Sources
- Tibon, R., Greve, A., Humphreys, G., Quent, J. A. and Henson, R. (2026). Neural activations and representations during episodic versus semantic memory retrieval. Nature Human Behaviour.
- University of Nottingham press release, via ScienceDaily: This brain discovery is forcing scientists to rethink how memory works.