The Burr: Memory Techniques

You won't hold on to a list of twenty words. Yet the plot of a film you saw once ten years ago you can retell in detail — faces, lines, even what the room in the shot looked like.

The film holds hundreds of times more information than the word list. It's easier not because it matters more, but because it has something to catch on.

The glass bead and the burr

Imagine you have to attach something to a woollen coat.

The first object is a smooth glass bead. Press it against the fabric, let go — it falls. Not because it's heavy: it simply hasn't a single protrusion. That's how dry abstract information behaves: a textbook definition, a number, someone else's phrasing. You honestly press it against memory — it slides off.

The second object is a burr. You don't even have to press it; it catches by itself, and afterwards you have to tear it off. The reason is that its whole surface is covered in hundreds of tiny hooks.

The job of any mnemonic is to turn the glass bead into a burr. Not to learn harder, but to grow hooks: an image, motion, emotion, strangeness, a link to what you already know.

A dry fact slides. A strange image catches.

Why the hooks are pictures

The first and most powerful hook is a visual image.

Allan Paivio described this in the 1970s in his dual coding theory: information represented both in words and as a picture is stored in two systems at once and can therefore be retrieved by two different routes. Hence the well-studied picture superiority effect: images are on average remembered more reliably than the words denoting the same things.

The reason isn't magic but the biography of the species. For millions of years the brain trained to remember what the place where edible things grow looks like, and how the predator moved in the bushes. It began memorising sequences of abstract symbols literally a few thousand years ago. Mnemonics don't hack the brain — they translate a new task into the language of an old, well-tuned system.

The second hook is strangeness. Back in 1933 Hedwig von Restorff showed that an item standing out from a series is remembered noticeably better than the rest. That's why images built for memory are made absurd, excessive, in motion. Not "a doctor with a folder," but "a doctor in rubber boots sliding down the corridor on the folder like a skateboard." Absurdity isn't play; it's a way to stand out against thousands of interchangeable memories from the day.

The third hook is you. The self-reference effect, described by Rogers, Kuiper and Kirker in 1977: information a person has connected to themselves — to their own experience, flat, acquaintances — is remembered better than information processed neutrally. Which is why an image with your own kitchen beats an image with an abstract kitchen.

The fourth hook is the other senses. An image doesn't have to be only a picture: you can load it with sound, smell, temperature, touch and movement. Every added sense is one more independent route to the memory, and this is exactly how working memory champions operate: their images aren't examined, they're lived through — with the crunch, the stink, the weight and the speed. A practical test: if you can say how your image sounds and what your hands feel in it, you have enough hooks.

And the fifth, decisive one: you have to invent the image yourself. Norman Slamecka and Peter Graf described the generation effect in 1978: material a person produced themselves is remembered better than material received ready-made. Someone else's elegant mnemonic is always weaker than your clumsy one. The memorising happens in the moment of inventing.

The memory palace

There's a technique that exploits spatial memory directly, and it's two and a half thousand years old.

It works like this. Take a route you know by heart: your flat, the way from home to the shop, a school corridor. Choose several distinct points along it in a fixed order — the hallway, the kitchen table, the windowsill, the bathroom. Then turn each item you need to remember into a vivid image and leave it at a specific point, making it interact with the surroundings somehow. To recall the list, you simply walk the route in your mind and collect what you left there.

An absurdly simple idea — and it is exactly what every winner of memory championships uses.

There's a telling study about them. In 2003 Eleanor Maguire and colleagues scanned the brains of competitors at such events, expecting to find anatomical peculiarities. She found none: there were essentially no structural differences from ordinary people, and their intelligence was in the normal range too. Something else differed — while memorising, the regions associated with spatial navigation lit up. They didn't remember better. They translated everything into routes.

And the follow-up, which matters more: in 2017 Martin Dresler and colleagues took people with no training at all and taught them the memory palace method. After six weeks of practice their list-memorising scores roughly doubled, and their pattern of brain activity shifted towards the one observed in champions. So it isn't a gift, it's a skill — and one acquired in months rather than years.

Mnemonics for arbitrary material

A separate class of techniques exists for information with no internal logic: terms, dates, foreign words, the order of items.

The keyword method for vocabulary. Richard Atkinson and Michael Raugh tested it in 1975 on English speakers learning Russian words: students found a familiar word that sounded like the unfamiliar one and built a picture with it. For the Russian zvonok ("bell"), for instance, the similar-sounding English keyword oak gave an image of a bell swinging from an oak branch. The group working this way scored substantially better than the group learning by ordinary repetition. The technique works in both directions: for a new foreign word, find a native word that sounds like it and bind them in a single scene.

Acronyms and phrases for order: fold the first letters of the items into a word or an easily spoken sentence. This works exactly as long as the items themselves are familiar to you — the letter only reminds, it doesn't explain.

Numbers into images. Digits are abstract, so they get translated in advance into permanent pictures: 2 is a swan, 8 is a dumbbell, 0 is a plate. Build the set once and you can then assemble any date or number from it.

Where the boundary lies

An important caveat, without which all of this turns into a harmful crutch.

Mnemonics create a label — the same one the article on the warehouse was about. They answer the question "how do I get it out," not the question "how does it work." You can memorise the order of steps in an algorithm with a mnemonic; you cannot understand why it's correct that way.

Hence the rule: mnemonics for the arbitrary, understanding for the structural. Terms, names, vocabulary, sequences, dates — yes. The mechanics of a proof, the workings of a system, the causes of a phenomenon — no: those need a chunk, not a hook.

And second: mnemonics don't cancel repetition. An invented image fades on its own too — brighter than an ordinary fact, but it fades. Send it into the same spaced repetition schedule as everything else.

In practice

Build your first palace today. Take your flat and write out ten points along a route in a fixed order. A palace built once serves for years and gets reused.

Make images absurd, moving, and yours. A quality check: if the picture is funny or awkward to describe out loud, it will work. If it's tidy and neutral, it will slide off.

Add sound, smell and touch. Don't just see the object — hear it grinding and feel that it's sticky. Every sense is a separate hook and costs you a second.

Don't take ready-made mnemonics if you can invent your own. Someone else's saves a minute and loses the generation effect. Your clumsy one is more reliable than their elegant one.

Tie things to the personal. Put your kitchen, your acquaintances, your habits into the images — that adds hooks for free.

Check by retrieving, not by admiring. Once you've invented an image, close your eyes right away and walk the route, pulling everything back out. Otherwise what you've remembered is that you invented something good, not the material itself.

Check yourself

Close the article and answer in your own words:

Why does a dry definition "slide off" while an absurd image "catches" — what hooks hold it?
What did Maguire find in memory champions, and what follows from that for an ordinary person?
Why should you prefer your own clumsy mnemonic to someone else's elegant one?
When are mnemonics appropriate, and when do they become a harmful crutch?
How does the keyword method for foreign vocabulary work?

In short

Mnemonics don't improve memory; they grow attachment points: image, motion, strangeness, sound and smell, a personal link.
Dual coding and picture superiority: a picture is stored by two systems and retrieved by two routes.
The strange is remembered better than the ordinary, your own better than someone else's, the personal better than the neutral.
The memory palace translates lists into spatial routes. Champions don't have special brains but a special strategy; the skill takes weeks of training.
Mnemonics are for arbitrary material. Structural material needs understanding, not a hook. And repetition is still required either way.
A dry fact slides. A strange image catches.