How many words do you need to know to understand a language?
People usually ask how many words it takes to be fluent. Fluency is not a number: it is speed, pronunciation, grammar and confidence as much as vocabulary. Understanding, though, can be measured, and researchers have done it carefully enough that there are real figures to work with. Here is what they found, language by language, and what it means for where you start.
The idea that makes the question answerable
The key measure is coverage: the share of the words in a text that you know. At 90% coverage, one word in ten is unfamiliar, which on a typical page means one or two unknown words in every line.
How much coverage is enough? In a study by Marcella Hu and Paul Nation, learners read a story in which some words had been replaced with made-up ones. With 80% of the words known, none of them understood the story adequately. The researchers estimated that most readers need about 98%, or one unknown word in fifty (Hu & Nation, 2000). A later study of 745 students in Israel proposed two thresholds: 95% coverage as a minimum for reading with some help, and 98% as the level for reading independently (Laufer & Ravenhorst-Kalovski, 2010).
English
Paul Nation measured coverage in novels, newspapers, films and conversation, counting word families: a base word with its forms, so play, played and player count once (Nation, 2006).
- The 1000 most frequent families cover 78% to 81% of written text and 81% to 84% of speech.
- 3000 families, plus names, cover over 95% of conversation.
- Reaching 98% takes 8000 to 9000 families for novels and newspapers, and 6000 to 7000 for speech.
Films sit between the two. In the scripts of 318 films, the 3000 most frequent families, together with names and words like “oh” and “hmm”, covered 95.8% of the words, and 6000 families covered 98% (Webb & Rodgers, 2009).
Spanish
Mark Davies counted a 20-million-word collection of Spanish. The 1000 most frequent words cover about 88% of speech and 76% to 80% of writing; the first 3000 cover 94% of speech and 88% to 90% of writing (Davies, 2005). He counted dictionary forms rather than families, so pintar and pintor are two words, and still the figures come out close to English.
Italian
Tullio De Mauro’s Nuovo vocabolario di base lists about 7000 words. Its core, about 2000 “fundamental” words, covers 86% of the words in a collection of almost 19 million words of newspapers, essays, literature, entertainment, online writing and speech. The next 3000, words of “high use”, add another 6% (De Mauro, 2016).
French and Modern Hebrew
We found no comparable coverage study for French or Modern Hebrew that we could check, so we will not quote numbers for them. What is known is that word frequency follows the same steep shape everywhere it has been measured: a few very common words make up most of any text, and a long tail of rare words makes up the rest (Piantadosi, 2014). The exact figures differ; the strategy does not.
And native speakers?
Far more than any of these. A large online study estimated that an average 20-year-old native speaker of American English knows about 42000 dictionary words, from about 11100 word families (Brysbaert et al., 2016), and “knows” there can be as shallow as recognising that a word exists. Nobody needs that many to live comfortably in a second language.
Milestones that mean something
- 1000 words: the skeleton of every sentence. Not enough to follow much, but nothing works without it.
- 3000 words: most of everyday conversation, and a solid base for films and series.
- 4000 to 5000 words: real articles and books, with a dictionary nearby.
- 8000 to 9000 words: novels and newspapers on your own.
Each extra thousand adds less coverage than the one before. That does not make it worthless: in the Israeli study, reading scores kept rising with every thousand words, even where coverage barely moved. A rare word can be the one a whole paragraph turns on.
A quick way to see where you stand: take a text you would like to read, count 100 running words and mark the ones you do not know, ignoring names. More than five, and you are below 95%. Two or fewer, and you are around 98%.
Where to begin
Begin at the top of the frequency list, and make each word stay. Our courses are built that way: words in order of how often they are really used, up to 3000, in Spanish, French, Italian and Modern Hebrew. For the remembering part, see how to remember Spanish words; the same techniques work in any language.
Sources
- Hu, M., & Nation, I. S. P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13(1), 403–430.
- Laufer, B., & Ravenhorst-Kalovski, G. C. (2010). Lexical threshold revisited: Lexical text coverage, learners' vocabulary size and reading comprehension. Reading in a Foreign Language, 22(1), 15–30.
- Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? The Canadian Modern Language Review, 63(1), 59–82.
- Webb, S., & Rodgers, M. P. H. (2009). The lexical coverage of movies. Applied Linguistics, 30(3), 407–427.
- Davies, M. (2005). Vocabulary range and text coverage: Insights from the forthcoming Routledge Frequency Dictionary of Spanish. In D. Eddington (Ed.), Selected Proceedings of the 7th Hispanic Linguistics Symposium (pp. 106–115). Cascadilla Proceedings Project.
- De Mauro, T. (2016, 23 December). Il Nuovo vocabolario di base della lingua italiana. Internazionale.
- Piantadosi, S. T. (2014). Zipf's word frequency law in natural language: A critical review and future directions. Psychonomic Bulletin & Review, 21(5), 1112–1130.
- Brysbaert, M., Stevens, M., Mandera, P., & 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, 7, 1116.