Verbal Comprehension · back to the domain
Does your vocabulary keep growing after twenty?
In the best study we could find, yes — steadily, all the way to seventy. Which is the opposite of what the word "aging" usually implies in a sentence about the mind.
What the study actually found
Brysbaert and colleagues (2016) tested a large sample of native speakers of American English on a list of lemmas — dictionary head-words, so walk but not walked — and plotted how many were known against age.
| Age | Lemmas known | Word families known |
|---|---|---|
| 20 years | 42,000 | 11,100 |
| 60 years | 48,200 | 13,400 |
| Difference | +6,200 over forty years | +2,300 over forty years |
Six thousand two hundred words in forty years works out at roughly one new word every two days, sustained across an adult life. And the median in their sample keeps climbing to seventy — the oldest age they plot — rather than peaking somewhere in the thirties.
Why it grows, which is less flattering than it sounds
The paper's own explanation is exposure rather than anything else: older people have simply come across more words. The authors add the part that makes it click — "this suggests that people rarely forget words, even when they don't know the (exact) meaning any more."
So the curve is closer to an accumulation than an improvement. Vocabulary is the one place where having been around a long time is straightforwardly an advantage, because the words stay once they arrive.
This is not a claim about your brain, your health, or how you are aging, and we are not going to turn it into one. It is one narrow finding about one narrow thing — how many head-words a median person recognises — in a sample of other people. It says nothing about memory, speed, reasoning, or anything else on this site, and it is not evidence about decline or its absence.
Why the age tables online contradict each other
Search this question and you will get tables, and they will not agree with each other. We have not verified the individual figures those pages publish and so we are not repeating them here — but the reason they diverge is visible inside the one study we did read at source. None of the tables is necessarily wrong. They are counting different things, and the study read at source shows exactly how far that goes.
- Lemmas or word families? The same person in the same study is "42,000 words" or "11,100 words" depending purely on which you count. That is a factor of nearly four, from one decision.
- Recognition or explanation? Ask someone to spot a word they know and you get one number; require a synonym or a definition and you get a much smaller one.
- Which list? A word list drawn from a dictionary, a school corpus, or general usage will not produce the same total.
Across the published attempts the paper surveys, the estimates "range from less than 10 thousand words known to over 200 thousand words mastered." The domain page goes into that properly — it is the reason we do not publish a vocabulary size at all.
The thing that matters more than age
Education level separates the curves substantially in the same figures, and the authors note that most of that difference appears to originate during the years of study rather than accumulating evenly afterwards.
They also model the outer bounds by language input: at the low end, someone whose input comes only from social interaction; at the high end, someone constantly reading at 300 words a minute. Those two are further apart than twenty and sixty are. [S] Read together, that means the age row you land on is one of the weaker predictors on the page — an inference from their figures rather than a finding they state in those words.
What none of this tells you about your own vocabulary
We cannot place you on that curve, and neither can any test that offers to. Our reference sample is zero people — every number above belongs to other researchers and other participants, and is here attributed, not adopted.
Our own vocabulary test gives you a count out of twenty words with the full item list printed, precisely because the count is checkable and a projected total is not. Twenty words cannot estimate a vocabulary and we do not pretend they can.
Where to go next
- How many words do you know? — the domain page, and the argument for a count instead of a size.
- Take the Word Meaning test — twenty words, untimed, nothing stored.
- What is a normal digit span? — the same question asked of a different domain, with the same refusal at the end of it.
- How everything here is scored, and what Rung 0 rules out.
Sources
- Brysbaert, M., Stevens, M., Mandera, P. and Keuleers, E. (2016), How many words do we know? Practical estimates of vocabulary size dependent on word definition, the degree of language input and the participant's age, Frontiers in Psychology. Read at source, 8 September 2026 — the median counts at 20 and 60 for both lemmas and word families, the steady rise to 70, the education-level separation, the input bounds, the quoted sentence on forgetting, and the range across published estimates.