Tap a word, save a word: building vocabulary out of conversation

Everyone has abandoned a vocabulary app. The usual explanation is discipline, and it is usually wrong. The real reason is that the words were not yours — they came from a list somebody else wrote, in an order that had nothing to do with your life, and your brain treated them accordingly.

A word you tripped over in a conversation you were having is a completely different object.

Why context does the work

Memory is associative. A word stored with nothing attached has one thread leading to it, and threads break. A word stored with a scene attached — who said it, what you were trying to do, why you did not understand — has five or six.

This is why "would you like a bag?" learned from a shop assistant survives, while the same phrase in chapter four does not. Nothing about the phrase changed; everything about what it is tied to did.

It also explains the common and demoralising experience of recognising a word on a flashcard but failing to retrieve it in conversation. Your retrieval cue was the flashcard, and the flashcard is not present when you need the word.

What to look at when you tap a word

A translation is the least useful part of a dictionary entry, and it is the only part most people read. Three other things matter more.

The base form. The word in front of you is probably inflected — conjugated, declined, in a polite form. Knowing which dictionary word it came from is what lets you recognise it again in a different shape.

What it goes with. Words travel in company. In English you make a decision and take a photo, and swapping them marks you as a learner instantly. No amount of knowing the individual meanings prevents that error; only seeing the pairing does.

Its close relatives. Knowing that a word sits next to two similar ones, and roughly how they differ, is what stops you from using the formal one in a bar.

Tapping a word in Amiko shows these — base form, forms it takes, similar words, and what it is commonly used with — rather than a bare gloss.

Save less than you want to

The instinct is to save everything unfamiliar. Do not. A saved list of two hundred words is a list you will never review, and its real function becomes making you feel behind.

A better filter: save a word only if you can imagine yourself saying it this month. Not reading it — saying it. That single question removes most of what you would otherwise capture, and what survives is a small list of things you genuinely want to be able to produce.

Five to ten a week is a good rate. Fifty is a symptom.

The saved list is a to-do list, not a trophy case

Saving a word does nothing on its own. The value comes from what happens next, and it should happen within a day or two.

The strongest move is to force the word into a conversation. Start a chat, and steer until you can use it — even clumsily, even if you have to bend the topic to get there. That single deliberate use does more than a week of review, because it is the first time the word has been through the production path rather than the recognition path.

Amiko's practice modes run off your saved words for the same reason: the app asks for at least five before it will drill them, which is a polite way of saying that a list is only worth anything once you use it.

The realistic maths

Ten words a week that you can actually say is 500 a year. That sounds slow next to an app promising a thousand a month, and it is roughly ten times more useful, because those 500 are all in the productive half of your vocabulary and are all attached to a situation you were personally in.

Most people's problem was never the size of the intake. It was that nothing they took in ever came back out.