How to Expand Your Vocabulary: The Three Jobs Most Advice Skips

Most vocabulary advice answers one question, which is where to find new words. That is the easy part, and it is almost certainly not where you are stuck. The words you have already met and quietly lost outnumber the ones you are about to meet.

There is a version of this advice you have read before. Read more. Keep a notebook. Learn a word a day. Use it in a sentence. None of it is wrong, and most people who follow it end up in the same place a year later, reading more comfortably than before and still reaching for the same four hundred words when they speak.

The reason is that expanding a vocabulary is not one job. It is three, they fail in different ways, and the popular advice only addresses the first.

Job one: meeting words worth having

New words arrive through input, which means reading and listening to material you can mostly follow. The evidence behind this is unusually consistent: language is acquired from input you understand, which is the comprehensible input idea, and volume matters more than intensity, which is what extensive reading is built on. Ninety pages of a novel you enjoy will hand you more usable vocabulary than ninety minutes with a word list.

The selection problem is where this goes wrong. Left alone, most people pick material that is either too easy, where nothing new appears, or so hard that every second sentence stops them and the reading dies by chapter two. The useful zone is narrow: you follow the story, and a handful of words per page are new. If you are looking words up more than about five times a page, the book is not too advanced for you as a person, it is too advanced for this month.

Word lists have a place, but a smaller one than they are usually given. They are good at telling you what a level actually contains, so you can find the holes below where you already read. Our vocabulary lists by CEFR level are worth a scan for that reason. They are poor at telling you which words you personally need next, because that depends on what you read, who you talk to and what you are trying to say.

How much is enough, roughly

Vocabulary size research gives a usable sense of scale. Paul Nation, whose work on vocabulary size and text coverage is the standard reference here, has argued that a few thousand word families cover the great majority of everyday spoken language, while reading an unsimplified novel comfortably takes something closer to eight or nine thousand. Those are approximations across studies rather than a target to tick off, but the shape is the point: conversation arrives sooner than reading does, and the gap between them is large.

Set against a realistic daily intake, that shape stops being intimidating. Ten to fifteen new words a day, kept, is a serious rate, and it is roughly the ceiling most people can sustain alongside a job. The how many words per day question is worth its own answer, but the short version is that the number you keep matters and the number you meet does not. Fifty words a day for a fortnight and then nothing is a slower route than twelve a day for a year.

Job two: knowing a word well enough to use it

Recognising a word on a page and producing it in a sentence are separate abilities with separate training, which is the active and passive vocabulary distinction. Almost everyone has a passive vocabulary several times larger than their active one, and most of the frustration people describe as "my vocabulary is not growing" is really an active vocabulary that has stalled while the passive one quietly expanded.

Knowing a word is also not a single thing you either have or do not. Applied linguistics usually breaks it into form, meaning and use, and the third one carries most of the difficulty. You can know that "commit" means to promise and still not know that in English you commit to a plan, that it takes a strong register in a working context, and that it collocates with words like fully and firmly. A translation gives you the meaning and none of the use.

This is why learning words in chunks beats learning them alone. Store "make a decision" rather than "decision", "heavy rain" rather than "heavy". Native fluency is largely a matter of having thousands of these ready-made pieces, which is the argument in our piece on collocations and chunks.

Job three: not losing what you already met

This is the job almost nobody plans for, and it is where the vocabulary goes. Meeting a word once does close to nothing; the forgetting curve is steep in the first days and shallow after that, so what you do in the week after the first encounter decides whether the word survives.

Two things flatten it. The first is retrieving the word rather than reviewing it: trying to recall the meaning before you check strengthens the memory in a way that rereading does not, which is active recall. The second is spacing those retrievals out, so each one lands near the edge of forgetting rather than while the word is still fresh. Put together they are spaced repetition, and they are the difference between a vocabulary that accumulates and one that churns.

The practical implication is unglamorous. A capture step at the moment you meet a word, and a short review habit that brings it back later, will do more for your vocabulary than any change to what you read.

ApproachWhere the words come fromWhat it trainsWhere it breaks
A word a day appSomeone else, at randomRecognition of words you may never meet againThe words are not yours and never come up
Reading a lot, nothing elseReal material at your levelReading fluency and passive vocabularyMost new words are met once and lost
Grinding a frequency listA list, in orderBreadth, fast, in isolationNo context, no use, and it is dull enough to quit
Read, capture, review, useWhatever you actually read and hearAll three jobs at onceIt requires the capture step to be cheap

What feels productive and is not

A few habits survive because they feel like work while producing very little.

A routine that compounds

None of this needs an hour a day. It needs the three jobs to each get something.

Where MindDory fits

The capture step is the one that decides whether any of this survives contact with a normal week, so MindDory is built around making it cost nothing. Photograph a page of the book you are reading and the words worth keeping become cards with the sentence they came from. Ask Claude or another AI assistant about a word mid-conversation and it lands in the same place, because MindDory connects over MCP. Nothing about that requires you to stop what you were doing and build a deck.

After that it is the third job. Each card carries a memory cue rather than a bare translation, and review sessions are assembled from the words closest to slipping instead of whatever you added last, so the repetitions land where they are useful. If you want the mechanics behind that, the guide on how to memorize vocabulary fast covers the technique side in more detail.

The honest summary is that your reading is probably fine. What is missing is a way for the words you already meet to come back before they fade, and a reason to use them once they do.