Learn & Understand

Counting the Uncountable: How Vocabulary Size Is Measured

The vocabulary size estimator uses statistical sampling to estimate how many words a person knows. Estimating vocabulary presents a genuine puzzle: no one can be tested on every word in a language, so how can total word knowledge possibly be measured? The answer is a clever application of sampling, the same statistical idea behind polls and surveys. Understanding how vocabulary size is estimated, and the surprisingly slippery question of what it means to know a word, illuminates a measurement that is more sophisticated than it appears.

The Impossibility of Testing Everything

A large language contains far too many words to test someone on all of them; the effort would be endless. So directly counting a person's vocabulary by checking every possible word is impossible. This is the same problem faced by anyone trying to measure a property of a huge population, and it calls for the same solution. Rather than examine everything, one examines a well-chosen part and extrapolates. Vocabulary estimation borrows this fundamental statistical strategy to make the uncountable countable.

Sampling and Extrapolation

The method is to draw a random sample of words from a large reference list, test how many of them the person knows, and then scale that proportion up to the full list. If someone knows a certain fraction of a representative random sample, one estimates they know roughly that same fraction of the entire reference vocabulary. This is exactly how a poll estimates an entire population's views from a small sample. The reliability of the estimate depends on the sample being genuinely random and representative, not biased toward easy or common words.

What Does Knowing a Word Mean?

A subtle complication is defining what counts as a word and what it means to know one. Vocabulary is often measured in word families, grouping a base word with its inflected forms and derivatives, so that related forms count as one item rather than many. Whether such related forms are counted together or separately dramatically affects the total, which is why estimates of vocabulary size vary. And knowing a word can mean anything from vaguely recognizing it to using it fluently, so the measure depends on the depth of knowledge being tested. The seemingly simple count rests on contestable definitions.

A Rough But Meaningful Number

Despite these subtleties, vocabulary estimation yields a useful figure for gauging progress and comparing against benchmarks, such as the range of word families a native speaker typically commands. The calculator applies the sampling method, scaling the fraction of a sample known up to a reference list to estimate total vocabulary. The result is an approximation, sensitive to the sample and the definition of a word, not an exact count. But it turns an impossible-seeming measurement into a meaningful estimate, letting a learner put a rough number on the breadth of words they command, a milestone that would otherwise remain entirely uncountable.

Set a vocabulary goal from your estimate with the Words Per Day Calculator, or see how vocabulary relates to text coverage with the Text Coverage Vocabulary Calculator.