How Many Words Do You Need to Read Books in a Foreign Language?

“How many words do I need to know?” This is the single most practical question in language learning, and it has a surprisingly clear answer. Unlike most questions about language — “How long does fluency take?” or “What is the best method?” — vocabulary thresholds for reading have been studied rigorously for decades, and the numbers are consistent across researchers and languages.

The short version: you need roughly 3,000 word families to read most fiction with some effort, 5,000 for comfortable reading, and 8,000-9,000 for the kind of smooth, uninterrupted reading experience you have in your native language. But those numbers alone do not tell the full story. What matters is what those numbers feel like in practice, how they vary by language, and how reading itself is the fastest way to reach them.

The Research: Coverage Thresholds That Actually Matter

The foundational research comes from two scholars: Paul Nation and Batia Laufer. Their work, spanning from the late 1980s through the 2000s, established the concept of “text coverage” — the percentage of words in a text that a reader knows — and identified the coverage thresholds at which different levels of comprehension become possible.

Laufer (1989) was the first to propose that learners need to know at least 95% of the words in a text to achieve “reasonable” reading comprehension. Nation (2006) refined this, arguing that 98% coverage is the threshold for comfortable, unassisted reading — the kind of reading where you understand the text fluently without needing to reach for a dictionary.

Here is what the vocabulary-to-coverage relationship looks like, based on their research and subsequent studies by Hsueh-chao and Nation (2000) and Webb and Macalister (2013):

These percentages may look close together — what is the difference between 95% and 98%? — but the experiential gap is enormous. At 95% coverage, you are stopping roughly every line to deal with an unknown word. At 98%, you might go an entire page before encountering one that actually matters. That difference is the gap between reading being a struggle and reading being a pleasure.

What These Percentages Actually Feel Like

Numbers are abstract. Let us make them concrete.

Take a passage from a typical novel — say, the opening of a Harry Potter book. A representative sentence might be: “Mr. and Mrs. Dursley, of number four, Privet Drive, were proud to say that they were perfectly normal, thank you very much.”

That sentence has 22 words. At 80% coverage, you would not know roughly 4-5 of them. Imagine reading it like this: “Mr. and Mrs. _____, of number four, _____ Drive, were _____ to say that they were _____ normal, thank you very _____.” You can guess some of those blanks from context, but it is slow, uncertain work. Do that for 300 pages and you will quit by chapter three.

At 95% coverage, you would miss about 1 word in that sentence. Maybe “proud” or “perfectly.” You can probably figure it out from context. The sentence makes sense. But across an entire page, you are doing this guesswork 10-15 times. It works, but it requires effort and focus.

At 98% coverage, you might not miss any words in that particular sentence. Across an entire page, you hit perhaps 2-3 unknowns, and most of them are not critical to the meaning. This is how native speakers experience reading — you occasionally encounter an unfamiliar word, but it does not break your flow.

What “Word Family” Means (And Why It Matters)

Before the numbers above mean anything practical, you need to understand what researchers count when they say “word families.”

A word family is a base word plus all its inflected and closely derived forms. The word “run” as a family includes: run, runs, running, ran, runner, runners. That is one family, not six separate words. Similarly, “happy” includes happy, happier, happiest, unhappy, happiness, happily — one family covering six forms.

This matters enormously for how you interpret vocabulary counts. When someone says “I know 5,000 words,” they might mean 5,000 individual word forms. But researchers typically mean 5,000 word families, which translates to roughly 15,000-25,000 individual word forms in practice, because each family contains multiple members.

The word-family model is optimistic for learners, because once you know a base word, recognizing its variants is relatively easy. If you know “inform,” you can probably recognize “information,” “informative,” “informed,” “uninformed,” and “misinformation” without ever having studied them explicitly. Your brain does the derivation automatically.

This means the actual learning burden is smaller than raw word counts suggest. You do not need to memorize 8,000 separate, unrelated items. You need to learn 8,000 base concepts, and many of their derivatives will come for free.

Nation (2001) estimates that the average educated adult native English speaker knows approximately 15,000-20,000 word families. But crucially, you do not need to match native speakers to read their books. At 8,000-9,000 families — roughly half of a native speaker’s vocabulary — you can read virtually anything written for a general audience.

Real-World Benchmarks: What Can You Read at Each Level?

Here is where the research meets practice. At each vocabulary milestone, certain types of reading become accessible.

500 Word Families

This is the absolute beginner stage. You know basic greetings, numbers, colors, common verbs (be, have, go, want, like), question words, and a handful of concrete nouns.

What you can read:

• Picture books for young children (2-4 year olds)
• Menus, street signs, basic labels
• Very simple dialogues in textbooks
• Social media posts with common slang stripped out

What it feels like: You recognize individual words scattered across a page but cannot follow connected text. A paragraph is a wall of unfamiliar symbols with occasional familiar landmarks.

1,000 Word Families

You know the core structural vocabulary of the language — articles, prepositions, pronouns, basic conjunctions — plus the most common everyday words.

What you can read:

• Fairy tales and folk stories written for children
• Simplified short stories (graded readers at the lowest levels)
• Simple news headlines
• Text messages and casual social media posts from friends (if you can parse the slang)

What it feels like: You can follow the skeleton of a simple narrative. “The man went to the house. He opened the door. The room was dark.” Complex descriptions and abstract ideas are still out of reach. You read slowly and need to re-read sentences frequently.

Example: At 1,000 Spanish word families, you could read a simplified version of a fairy tale like “Caperucita Roja” (Little Red Riding Hood). You would follow the plot — girl, forest, wolf, grandmother — but miss many descriptive details.

2,000 Word Families

This is a significant milestone. You have covered the high-frequency core of the language, and text coverage jumps noticeably.

What you can read:

• Young adult novels with straightforward language
• News headlines and brief news summaries
• Blog posts on everyday topics
• Song lyrics (mostly)
• Graded readers at the upper levels
• Social media fluently

What it feels like: You can follow a story and understand the main events. Character descriptions, motivations, and emotional nuances start to come through, though not reliably. You still miss words, but you can often guess from context. Reading is work, but it is productive work rather than frustrating guesswork.

Example: At 2,000 Spanish word families, you could read “El Principito” (The Little Prince) without much trouble. The vocabulary is concrete and philosophical but not complex. However, you would struggle significantly with “Cien Anos de Soledad” (One Hundred Years of Solitude), where Garcia Marquez uses a rich, literary vocabulary full of botanical terms, archaic words, and elaborate descriptions.

3,000 Word Families

For many learners, this is the sweet spot where reading transitions from study activity to genuine entertainment.

What you can read:

• Most contemporary fiction (thrillers, romance, science fiction, mystery)
• Adapted classics
• Popular non-fiction (self-help, popular science, biography)
• Full newspaper articles on general topics
• Wikipedia articles in your target language

What it feels like: You read for the story, not the words. Unknown vocabulary still appears regularly — roughly once per line — but it rarely blocks comprehension. You can sustain reading for 30-60 minutes without mental exhaustion. This is the level where reading starts to feel like something you do because you want to, not because you should.

Example: At 3,000 French word families, you could comfortably read a Marc Levy novel or a French translation of a Dan Brown thriller. The language is direct, the vocabulary is mainstream, and the plot carries you through the occasional unknown word.

5,000 Word Families

At this level, the vast majority of native-audience content becomes accessible.

What you can read:

• Native novels without adaptation or simplification
• Quality journalism (Le Monde, El Pais, Der Spiegel)
• Non-fiction on specialized topics you are interested in
• Classic literature (with some patience for archaic vocabulary)
• Professional documents in your field

What it feels like: Reading in the foreign language starts to feel similar to reading in your native language. You occasionally encounter an unfamiliar word, note it, and move on without losing the sentence. You can read for hours. You have preferences about authors and styles. You notice when writing is good or bad.

Example: At 5,000 German word families, you could read a contemporary novel by Daniel Kehlmann or a German translation of a Stephen King book without any aids. You would encounter perhaps 1-2 unknown words per page, most of which you could infer from context.

8,000+ Word Families

This is Nation’s threshold for truly comfortable, unassisted reading — 98% text coverage.

What you can read:

• Academic texts and scholarly articles
• Literary fiction with complex, experimental language
• Technical writing outside your specialty
• Poetry (though poetry has its own challenges beyond vocabulary)
• Historical texts with period-specific vocabulary
• Essentially anything written for a literate adult audience

What it feels like: You forget you are reading in a foreign language. Unknown words are rare enough that they feel like encountering unusual words in your native language — you note them with mild curiosity rather than frustration. Your reading speed approaches native levels.

Language-Specific Differences: Not All Words Are Created Equal

The numbers above are derived primarily from research on English. Other languages have structural features that change the equation, sometimes dramatically.

Chinese: Fewer Families, Different Counting

Chinese does not have word families in the traditional sense. There are no inflections — no “runs” or “running” — because Chinese words do not change form. What Chinese does have is a character-based system where individual characters combine predictably to form compound words.

If you know the characters for “electricity” and “brain” (computer), “electricity” and “speech” (telephone), “electricity” and “shadow” (movie), you can often decode new compounds containing “electricity” even if you have never seen them before. This compositionality means that character knowledge compounds: each new character you learn does not just give you one word — it gives you partial access to dozens of compound words.

Research by Liang et al. (2015) suggests that Chinese learners may achieve comparable text coverage with a smaller number of base vocabulary items than English learners, because the compositional nature of Chinese vocabulary means each item unlocks more total text coverage. Roughly 3,000-4,000 high-frequency characters cover 99%+ of characters in general text, though the number of distinct multi-character words you can form from them is much larger.

The practical takeaway: for Chinese, focus on characters and common character combinations rather than counting words. The vocabulary thresholds are lower in terms of base units, but the learning effort per unit (memorizing characters) is higher.

German: Compound Words Inflate the Count

German famously creates compound words by sticking existing words together. “Handschuh” (hand + shoe = glove), “Krankenhaus” (sick + house = hospital), “Geschwindigkeitsbegrenzung” (speed + limitation = speed limit). A German text will contain many unique word forms that are, in practice, combinations of words you already know.

This means that raw word counts for German texts look intimidating — a German novel might contain more unique word forms than the same novel in English — but the actual vocabulary burden is smaller than it appears. If you know the component parts, you can decode most compounds on sight, even if you have never encountered that specific compound before.

Research by Bauer and Nation (1993) and subsequent work on German specifically suggests that learners who understand German word formation patterns effectively get “bonus” coverage. Knowing 5,000 base word families in German might give you the effective coverage that 6,000-7,000 families would provide in English, because your knowledge of components lets you decode compounds you have never seen.

Romance Languages: The Cognate Advantage

English speakers learning French, Spanish, Italian, Portuguese, or Romanian start with a significant head start. Thanks to the Norman Conquest and centuries of cultural exchange, English shares an enormous pool of cognate vocabulary with Romance languages.

In French, an English speaker can recognize “important,” “different,” “possible,” “information,” “communication,” “situation,” and thousands of similar words with minimal adaptation. Estimates vary, but English speakers can typically recognize 30-40% of written French vocabulary through cognates, dropping to about 20-25% for spoken French (because pronunciation obscures the similarity).

Laufer and Goldstein (2004) found that cognate awareness significantly accelerates vocabulary acquisition in related language pairs. An English speaker learning Spanish effectively starts at 1,500-2,000 recognizable word families before studying a single flashcard. This does not mean they know those words perfectly — false cognates (“embarazada” means pregnant in Spanish, not embarrassed) and partial cognates cause confusion — but it dramatically reduces the time to reach reading thresholds.

The practical implication: an English speaker needs to actively learn roughly 5,000-6,000 Spanish word families to reach 8,000 effective families, because 2,000-3,000 come free through cognate recognition. These differences in difficulty across languages are worth considering when setting your vocabulary goals.

Japanese: Kanji Knowledge Compounds

Japanese uses three writing systems, and kanji (Chinese characters) are the key to vocabulary growth. Each kanji character carries meaning, and most Japanese words are written with 1-3 kanji. Like Chinese, this creates a compounding effect: knowing the kanji for “work” and “place” lets you decode “workplace,” “working hours,” “homework,” and other compounds.

Japanese has an additional wrinkle: most kanji have at least two readings (pronunciations), one derived from Chinese and one native Japanese. This doubles the memorization burden per character but also means each character is more productive — it participates in more words.

Research by Mori and Shimizu (2007) found that kanji knowledge is a stronger predictor of Japanese reading ability than word knowledge alone. A learner who knows 2,000 kanji but has a smaller measured vocabulary may actually read better than a learner who knows more individual words but fewer kanji, because the kanji knowledge enables real-time decoding of unfamiliar compounds.

The practical threshold for comfortable Japanese reading is approximately 2,000-2,500 kanji (the list taught in Japanese schools through high school) plus 6,000-8,000 vocabulary items. The kanji investment is steep at the beginning but pays exponential dividends.

Agglutinative Languages: Turkish, Finnish, Hungarian

Languages like Turkish, Finnish, and Hungarian build words by stacking suffixes. A single Turkish “word” might correspond to an entire English phrase: “evlerinizden” = “from your houses” (ev + ler + iniz + den: house + plural + your + from).

In these languages, the concept of “word family” breaks down because the number of possible word forms is astronomical. A single Turkish verb can theoretically generate hundreds of forms. Text coverage calculations need to account for this — a learner who knows 3,000 Turkish base words has access to far more distinct word forms than a learner who knows 3,000 English base words.

Practical research on Finnish (Mustonen, 2015) suggests that learners need approximately 2,000-3,000 base lemmas to reach 95% text coverage in general fiction, compared to 3,000-5,000 for English. The per-lemma learning burden is higher (you need to learn the morphological system), but the total number of items to memorize is lower.

How Reading Itself Builds Vocabulary: The Virtuous Cycle

Here is the most important insight in vocabulary research: reading is simultaneously the thing vocabulary enables and the most powerful way to build vocabulary. This creates a virtuous cycle — reading builds the vocabulary that makes more reading possible, which builds more vocabulary, and so on.

Incidental Vocabulary Acquisition

When you read, you pick up word meanings without deliberately trying to learn them. This is called incidental acquisition, and it has been studied extensively.

Research by Waring and Takaki (2003) and Pellicer-Sanchez and Schmitt (2010) found that readers learn the meaning of 5-15% of unknown words they encounter in a text through context alone. That number sounds low, but it compounds. If you read a 60,000-word novel that contains 600 unknown word families, you might incidentally acquire 30-90 of them. Read 10 novels, and you have picked up 300-900 word families without any flashcards, word lists, or deliberate study.

The acquisition rate depends on several factors:

• Context quality. Words that appear in informative contexts — where the surrounding text strongly suggests the meaning — are learned faster than words in ambiguous contexts.
• Frequency of encounter. A word you see once is unlikely to stick. A word you see 6-12 times across a book has a much higher chance.
• Existing vocabulary. The more words you know, the better you can use context to infer unknowns. This is why the virtuous cycle accelerates — your 5,001st word is easier to learn than your 1,001st.
• Reader engagement. Words encountered in compelling, emotionally engaging text are retained better than words in dry, informational text. Novels beat textbooks here.

The Multiple Encounter Threshold

How many times do you need to encounter a word before you truly know it? Researchers have converged on a range of 6-12 meaningful encounters (Nation, 2001; Webb, 2007).

“Meaningful” is the key qualifier. Seeing a word flash by in a list does not count the same as encountering it in a sentence where you actively process its meaning. Reading provides the meaningful encounters that the brain needs — each time you see the word in a different context, your mental representation becomes richer and more stable.

This has practical implications. A novel by a single author typically reuses vocabulary intensively. Stephen King’s “It” uses the word “sewer” more than 100 times. If you start the book not knowing what “sewer” means, you will know it thoroughly by the end. Authors have characteristic vocabularies, and reading multiple books by the same author is an efficient way to get repeated exposure to a consistent word set.

After that, switching to a different author introduces fresh vocabulary while reinforcing the common core. This alternation between depth (re-encountering familiar vocabulary) and breadth (encountering new vocabulary) is the natural rhythm of extensive reading.

The Annual Acquisition Rate

How fast does vocabulary grow through reading? Nation and Waring (1997) estimated that a learner reading extensively at an appropriate level acquires roughly 1,000 new word families per year. More recent estimates by Webb and Chang (2015) suggest 500-1,500 families per year, depending on the amount of reading and the learner’s level.

At the B1 level (roughly 2,500-3,000 word families), a learner reading 30 minutes per day in their target language could expect to acquire approximately 800-1,200 word families per year through reading alone. Combined with other learning activities — conversation, media, deliberate study — the total might reach 1,500-2,000 families per year.

This means a learner who starts reading seriously at the 2,000 word family level can expect to reach the 5,000 word family level in roughly 2-3 years. Reaching 8,000 families takes another 2-3 years beyond that. These timelines are consistent with the US Foreign Service Institute’s estimates for reaching professional proficiency in Category I languages (600-750 hours of study, spread over 2-3 years).

The key point is that the timeline is realistic. Reaching comfortable reading proficiency in a foreign language is not a decade-long project. It is a 2-4 year commitment of consistent daily reading, and the experience itself is enjoyable if you choose the right material.

Practical Strategy: From 0 to 3,000 Words Through Reading

Knowing the numbers is useful. Having a plan for reaching them is better. Here is a concrete, stage-by-stage approach to building the vocabulary you need to read books in a foreign language.

Stage 1: The First 100-250 Words (Week 1-2)

Before you can read anything meaningful, you need a minimum base of the most frequent words. These are the structural glue of the language — articles, pronouns, prepositions, basic verbs, numbers, question words.

Method: Study a high-frequency word list. Many language learning apps and resources offer curated lists of the 100 or 250 most common words. The Lingo7 app includes pre-made word collections organized by frequency, which give you exactly this foundation.

Goal: Recognize (not necessarily produce) the 100-250 most frequent words on sight. At this stage, flashcards and word lists are actually efficient — the words are so common that you will encounter them constantly, and the initial recognition is all you need. Reading will do the rest.

Stage 2: First 500-1,000 Words with Parallel Translation (Week 2-8)

Now you start reading, but with heavy support. At 250-500 word families, you do not have enough coverage to follow even simple texts independently. You need a translation alongside the original.

Method: Read with parallel translation — the foreign language text on one side, a translation in your native language on the other (or available on tap). Read the foreign language version first, sentence by sentence. When you get stuck, check the translation. Then go back to the foreign language and re-read the sentence with your new understanding.

This is not cheating. It is scaffolded learning. The translation keeps you engaged with the meaning while your brain processes the foreign language patterns. Over time, you check the translation less and less.

What to read: Start with short, simple texts. Children’s stories, very short stories, or the opening chapters of familiar books. Familiarity helps — if you already know the plot of Harry Potter, reading it in French removes one layer of difficulty and lets you focus on the language.

Goal: Build to 500-1,000 word families. You will acquire many of these passively through repeated encounters in your reading. Supplement with deliberate review of words you see frequently but cannot figure out from context.

Stage 3: Graded Readers and Simplified Content (Month 2-6)

At 1,000 word families, you can start reading material designed for language learners — graded readers, simplified novels, and adapted classics.

Method: Read extensively. The goal is volume, not perfection. Read for 20-30 minutes per day. Do not look up every unknown word — only the ones that appear repeatedly and seem important. Keep a small word list (5-10 words per session) for later review.

What to read: Graded readers from publishers like Oxford, Penguin, or language-specific publishers. Simplified versions of classics. Young adult fiction in languages where it is available.

Goal: Reach 2,000 word families. At this level, graded readers become easy (a good sign — it means you are ready to move on) and you can start approaching simpler native content.

Stage 4: Native Content with Support (Month 6-18)

This is the critical transition — from learner materials to real books written for native speakers. It is also the hardest stage psychologically, because real books are harder than graded readers and the jump can feel discouraging.

Method: Read native-level books with support tools — parallel translation for when you lose the thread, word lookup for when a recurring word blocks comprehension, and audio narration to reinforce pronunciation and reading speed. The key is to stay in the foreign language as your primary mode and use the support only when needed.

What to read: Contemporary fiction with accessible prose. Thrillers, romances, and young adult novels are ideal because they use direct language, short sentences, and modern vocabulary. Avoid literary fiction, poetry, and historical novels at this stage — their vocabulary and style add difficulty that is not productive yet. For more specific recommendations, see our guide on choosing books by language level.

Goal: Reach 3,000 word families. By the end of this stage, you should be reading native content with minimal support — perhaps checking the translation once every few pages rather than every few sentences.

Stage 5: Independent Reading and Spaced Repetition (Month 18+)

At 3,000+ word families, reading becomes self-sustaining. Each book you read makes the next one easier. Your vocabulary grows through the reading itself.

Method: Read widely. Rotate genres. When you encounter a word that seems useful and interesting, save it for spaced repetition review. The combination of encountering words in context (reading) and reviewing them at spaced intervals (SRS) is the most effective vocabulary acquisition strategy documented in the research.

Goal: Progress from 3,000 to 5,000 and eventually 8,000+ word families. This happens over years, not months, but the process is enjoyable because you are reading books you genuinely want to read.

Common Misconceptions

”I Need 10,000 Words Before I Can Start Reading”

This is perhaps the most damaging myth in language learning. It creates a false prerequisite — study vocabulary for years, then start reading — when the reality is almost exactly the reverse.

You can start reading productively at 500 words with the right support. Parallel translation, word lookup tools, and graded readers make reading possible at vocabulary levels that would be impossible for unassisted reading. And starting early is critical, because reading is the fastest way to build the vocabulary you need for more reading.

Waiting until you “know enough words” is like waiting until you are fit before going to the gym. The activity is the method.

”Knowing More Words Always Means Better Reading”

Vocabulary size is the strongest single predictor of reading comprehension in a foreign language, but it is not the only factor. Depth of word knowledge matters as much as breadth.

A learner who “knows” 5,000 word families but only at the level of a single translation for each (happy = feliz) will read worse than a learner who knows 4,000 families but has deep, contextual knowledge of each one (knowing that “feliz” and “contento” and “alegre” are all “happy” but in different ways, knowing which verbs collocate with each, knowing their register).

This is another argument for reading as a vocabulary building method: reading builds depth. You encounter each word in multiple contexts, with different collocations, in different registers. A flashcard gives you one connection (L1 translation). A novel gives you dozens.

”All Languages Require the Same Vocabulary Size”

As discussed in the language-specific section above, this is demonstrably false. An English speaker learning Spanish needs to actively learn fewer words than an English speaker learning Finnish, because cognate recognition effectively gives Spanish learners thousands of “free” word families. A Chinese learner can achieve high text coverage with fewer base vocabulary units than an English learner, because Chinese characters combine compositionally.

The 8,000-9,000 word family threshold is a useful benchmark for most European languages, but it should be adjusted for the specific language pair. For Romance languages, subtract 2,000-3,000 for cognates. For character-based languages, focus on character counts instead. For agglutinative languages, count lemmas rather than word families.

”Frequency Lists Are All You Need”

Frequency lists are valuable for identifying the highest-priority vocabulary — the 500-1,000 most common words that form the structural backbone of a language. But beyond the first 2,000 words, frequency-based study has diminishing returns.

The reason is that low-frequency words are domain-specific. The 5,000th most common word in French might be a term related to law, medicine, cooking, or politics. Whether you need that specific word depends entirely on what you read, not on its statistical rank. A learner who reads French crime fiction needs a different set of low-frequency words than a learner who reads French cookbooks.

After the high-frequency core, vocabulary is best acquired through reading in the domains that interest you. The words you need will find you — because they are the words that appear in the texts you choose.

”Children Learn 10 Words a Day, So Adults Should Too”

The often-cited claim that children learn 10 new words per day (or 3,000-4,000 per year) is based on research by Nagy and Herman (1987) on L1 (native language) acquisition. Applying this to adult L2 (second language) learning is misleading for several reasons.

First, children have 12-16 hours of waking exposure to their language every day. A language learner with 30 minutes of daily study does not have comparable exposure. Second, children’s word learning is cumulative over years of immersion; the “10 words per day” figure is averaged over their entire childhood, including periods of very fast and very slow growth. Third, children are learning their first language, which involves cognitive processes that are partially distinct from adult second language acquisition.

A more realistic target for an adult learner reading 30 minutes daily in a foreign language is 2-4 new word families per day, or roughly 800-1,500 per year. This is slower than the childhood figure but is achieved in a fraction of the daily time, and it compounds — after 3 years, you have 2,400-4,500 word families from reading alone, putting you well within the comfortable reading range.

Building Your Reading Vocabulary with Lingo7

The strategy outlined above — starting with high-frequency words, reading with parallel support, gradually reducing the scaffolding — is exactly the approach Lingo7 is built around. The app gives you books in 90+ languages with sentence-level parallel translation, native audio narration, and tap-to-translate word lookup. You read in the foreign language with a safety net that keeps comprehension from collapsing, even at low vocabulary levels.

Words you encounter while reading can be saved to your vocabulary with one tap, captured in the sentence context where you found them. Lingo7’s spaced repetition system then schedules review at optimal intervals, so words you meet in books move from recognition to long-term memory. Pre-made word collections organized by frequency give you the initial 100-250 word foundation before you start reading.

The result is a seamless progression from “I know 200 words and cannot read anything” to “I just finished my third French novel and only checked the translation twice in the last chapter.” The vocabulary thresholds described in this article are not theoretical milestones — they are stages you pass through while reading actual books, and you can feel each one.

The Bottom Line

The question “How many words do I need?” has a clear, research-backed answer. But the more useful answer is not a number — it is a realization. You do not need to reach a vocabulary threshold before you can start reading. You need to start reading so that you can reach the threshold. The two processes — reading and vocabulary growth — are not sequential. They are simultaneous, and each one accelerates the other.

Start with 250 words and a parallel text. Read something short and simple. Save the words that recur and review them. Read something slightly harder. Repeat. Within a few months, you will look back at the texts you started with and realize they look trivially easy. That is the virtuous cycle in action, and it is the most reliable path from “How many words do I need?” to “I just forgot I was reading in another language.”