In 1944, psychologists Hedwig von Restorff ran a series of experiments at the University of Berlin in which she presented participants with lists of items - mostly similar, but with one odd one out. The isolated item was recalled dramatically more often than the others. The finding was so consistent that it now carries her name: the Von Restorff effect. It describes something your memory does automatically and ruthlessly: it ignores the ordinary and clings to the strange.
Your vocabulary has a Von Restorff problem. When you try to memorize a new word by reading its definition and repeating it a few times, you are asking your brain to store something it has been specifically designed to filter out. The definition is abstract. The word is unfamiliar. There is nothing strange, spatial, or emotional to grab. So the brain files it in working memory, where it lasts about 20 seconds before vanishing.
This is not a willpower problem. It is a mismatch between how most people try to learn words and how human memory actually works.
What your memory is built for
Your brain did not evolve to store dictionary definitions. It evolved to navigate space, recognize faces, remember where the danger was, and recall emotionally charged events. These are the categories of information it stores effortlessly and retains for decades. You can probably describe in detail the layout of the first home you lived in, the face of someone who frightened you once, or the smell of a particular kitchen. You stored all of that without trying, because those inputs matched the formats your memory runs on: spatial, visual, emotional, and vivid.
Vocabulary words arrive in none of those formats. They are arbitrary sounds attached to abstract meanings. You have to translate them into a format your memory can use, or they will not stay.
The mechanism for doing this translation is called the keyword method, and it has been validated in language learning research since the 1970s. The basic logic is simple: you take the sound of the new word, find a word in your own language that sounds similar, and construct a vivid mental image that links the sound-alike word to the meaning. The image serves as a retrieval hook - when you later encounter the word, the image surfaces, and the meaning comes with it.
Building the hook in practice
Take the word "lachrymose," which means tearful or prone to weeping. On its own, it is a string of syllables with no obvious visual anchor. But say it out loud: lach-ry-mose. You can hear "lake," "rim," and "moss." Now build a scene: a lake whose rim is covered in soft green moss, with water weeping slowly over the edge. The moss is crying. The whole lake is lachrymose.
That image is strange. It is slightly absurd. And that is exactly why it works. Your memory treats strange images the way Von Restorff's participants treated the odd item in the list: it notices, it flags, it retains.
The hook does not need to be a perfect phonetic match. It needs to be good enough to fire the association when you hear the word again. A rough sound-alike that produces a vivid image outperforms an exact phonetic replica that produces nothing interesting.
Key Point: The keyword method works because it translates an arbitrary sound into a spatial, visual, and slightly bizarre scene - which is exactly the format human memory is built to retain. You are not fighting your biology. You are using it.
The three requirements for an image that sticks
Not all mental images are equally memorable. Research on the keyword method and related mnemonic techniques consistently finds that images with three properties outlast images without them.
First, action. A static image is easily forgotten. An image where objects are interacting, colliding, or transforming is much harder to lose. The crying moss weeping into the lake is doing something. The lake is not just a picture - it is a small event.
Second, scale. Exaggerated proportions catch attention. A normal-sized lake is forgettable. A lake the size of a city block, with moss the height of trees, dripping water the size of buckets - that version lodges somewhere. Do not restrain the image. Make it bigger than it needs to be.
Third, personal relevance. If you can anchor the image to something from your own life - a specific lake you know, a specific person who cries often, a smell that matches the scene - the memory gains additional hooks. The more it connects to existing information, the more paths lead back to it.
The abstract gap
The trickiest vocabulary to encode is not the emotionally neutral but visual words. It is the genuinely abstract ones: words that describe states, relationships, or qualities that have no obvious physical form. "Equanimity" means a calm, composed mental state. You cannot point at equanimity. You cannot photograph it.
For words like these, you personify. You build a scene around a person or creature demonstrating the quality, rather than representing the quality directly. Equanimity sounds like "equal enemies." Picture two bitter rivals standing perfectly still on a balance scale, each as calm as the other, neither willing to tip the scale. They maintain equanimity to avoid losing. The scene represents the concept through a situation rather than a symbol.
This approach takes slightly more effort than encoding a concrete noun. But the principle is the same: convert the abstract into a scene with action, scale, and enough strangeness to trip the Von Restorff filter.
Key Point: Abstract words require an extra step - converting the concept into a scene or situation that represents it, rather than an image of the thing itself. The more vivid and slightly absurd the scene, the better the retention.
The keyword method is how you make the word stick in the first place. What happens after you build the image - how you schedule your reviews, how you manage forgetting - is covered in Lesson 3. For now, practice the construction. Pick five words you have been trying to learn. For each one, find a sound-alike and build a scene that is a little strange, involves some action, and connects the sound to the meaning. It takes about three minutes per word. That time is not wasted on repetition. It is invested in a hook that will hold.