Originally published 2017-08-10.

Shutterstock
Forgetting is not only a failure of memory. A growing body of neuroscience suggests the brain actively clears out some information, and that this helps us adapt: we drop an old password so the new one comes to mind, or blur a hundred similar shopping trips into one useful sense of how the shop works. In a 2017 review in the journal Neuron, Blake Richards and Paul Frankland of the University of Toronto argued that a healthy memory system is built to forget as well as to remember.
That does not mean every lapse is a good sign. But the everyday experience of losing trivia, outdated details and the specifics of routine days is normal, and probably useful. Here is what the research says, where it comes from, and where its limits are.
The short answer: memory is for decisions, not trivia
For decades, memory research focused mainly on how information is stored and kept, which scientists call persistence. Richards and Frankland turned the question around and looked at transience, the gradual loss of what was stored. Their review, titled "The Persistence and Transience of Memory," pulled together existing studies rather than reporting a new experiment.
Their central claim is that memory exists to help us act well in a changing world. In the University of Toronto's announcement, Richards put it plainly: "The real goal of memory is to optimize decision-making." On that view, a brain that held every detail forever would be worse at its real job, because stale or conflicting memories would get in the way of choosing what to do now.
He illustrated the point with a sporting example: the purpose of memory is not to recall who won a particular hockey championship decades ago. Quiz champions are impressive, but sharp recall of isolated facts is not the same thing as intelligence.
How the brain forgets on purpose
The review describes forgetting as a process with its own biological mechanisms, not simply as memories fading through neglect. Two are especially interesting.
Weakening connections
Memories are thought to be stored in patterns of connections between neurons. Those connections can be weakened or removed over time, and some of that pruning appears to be an active, regulated process. When links that support an old memory fade, the memory becomes harder to retrieve.
New neurons in the hippocampus
The hippocampus, a curved structure deep in the brain that is central to forming new memories, is one of the few places where new neurons may keep appearing after birth. Research from Frankland's lab found that adding new neurons there can promote forgetting, probably because the new cells rewire existing circuits. This idea may also help explain why most adults remember almost nothing from before about age three or four, a period when the infant hippocampus is producing neurons rapidly.
Whether, and how much, new neurons form in the adult human hippocampus is still debated. Much of the evidence comes from rodents, so the human picture should be treated as an open question.
Two benefits: updating and generalising
Richards and Frankland point to two practical advantages of forgetting.
Updating. The world changes. A favourite café moves, a colleague changes jobs, a road gets closed. If the old memory stayed as vivid as the new one, the brain would keep offering conflicting answers. Letting the outdated version fade makes it easier to act on current information. In one experiment cited in coverage of the review, mice searching for a target in a maze found its new location faster when the memory of the old location had been weakened.
Generalising. If you remembered every detail of every visit to your local supermarket, you would struggle to extract the simple, useful pattern: where the bread is, when it is busy, which checkout moves fastest. Losing specific details helps the brain keep the general rule. Richards compared this to a technique in artificial intelligence called regularisation, which deliberately limits how closely a model fits its training data so that it copes better with new situations.
Earlier evidence: forgetting competitors
The idea did not start in 2017. A widely reported 2007 brain-imaging study led by researchers at Stanford University, discussed at the time by Wired and New Scientist, scanned volunteers while they practised recalling particular word pairs. Practising some memories made related, competing memories harder to recall, and as that happened, brain regions involved in resolving conflict had to work less hard.
In other words, suppressing the competition made remembering the right thing cheaper. Commenting on the work, psychologist Michael Anderson of the University of Oregon, a long-time researcher of memory suppression, described this kind of forgetting as functional: a tool the brain uses, not merely a flaw.
Myths and caveats
- "Forgetful people are smarter." The research does not say this. It says forgetting is part of normal, healthy memory. Being absent-minded is not evidence of high intelligence.
- "Good memory means remembering everything." People with highly superior autobiographical memory, a rare condition, can recall almost every day of their lives, yet they do not consistently outperform others on everyday reasoning, and some describe the constant recall as a burden.
- "Any memory problem is nothing to worry about." Normal forgetting is gradual, patchy and mostly about unimportant details. Memory loss that gets steadily worse, disrupts daily life, or comes with confusion, getting lost in familiar places or changes in personality is different. If you notice that in yourself or someone close to you, talk to a doctor.
- "This was proven in a new experiment." The 2017 paper was a review of earlier studies. Its strength is that it connects many findings, much of it from animal research, into one framework.
What you can take from this
If you blank at the pub quiz on a fact you once knew, there is no need to feel bad about it. Your brain may simply have judged it less useful than what you needed this week.
The research also offers a quiet practical lesson. Because the brain favours what is used, revisited and relevant, the things you want to keep are best protected by using them: practise the language, rehearse the skill, return to the ideas that matter. Information you never touch is the most likely to fade, and for much of it, that is exactly as it should be.
Sleep, regular physical activity and managing stress are also linked with healthier memory in the research literature. None of these will make you remember everything, and that, as Richards and Frankland argue, would not be the goal anyway.
Sources: University of Toronto News: Why forgetting is really important for memory (2017); Richards, B.A. and Frankland, P.W., "The Persistence and Transience of Memory," Neuron, June 2017; ScienceAlert coverage of the review. Image: Shutterstock.

COMMENTS