Memorising Chinese vocabulary is only half the battle — the other half is not forgetting it. Here's the method that makes new words stick for good.
Chinese vocabulary is the most frequently discussed — and most frequently misunderstood — aspect of learning the language. Learners obsess over word counts, worry about forgetting words they've 'already learned,' and spend enormous amounts of time reviewing flashcard decks that produce diminishing returns.
The obsession is understandable. Vocabulary is measurable. You can count how many words you know (or think you know). You can track your Anki deck stats. You can feel productive reviewing cards. The problem is that this kind of vocabulary study, pursued as the primary learning strategy, produces a specific and frustrating outcome: words that are recognisable on a flashcard but inaccessible in real conversation or reading.
There is a better approach. It is based on how memory actually works, what reading research says about vocabulary acquisition, and what distinguishes learners who reach functional fluency from those who plateau. This article lays out that approach clearly, from the foundations of memory science to a practical daily routine.
The standard approach to Chinese vocabulary learning goes something like this: get an HSK 1 word list, put the words into Anki or a similar app, and drill them until you can recall the English definition for each Chinese word. Repeat for HSK 2, 3, 4, and so on.
This approach has two structural problems.
The first is isolation. Words are not known in isolation by fluent speakers — they are known as elements of a language system, with associations to other words, to contexts where they naturally appear, and to the kinds of sentences they are typically used in. When you learn a word only as a pairing of Chinese character and English definition, you learn a thin slice of what knowing the word actually means. You can translate it but you can't use it.
The second problem is the difference between recognition memory and production memory. Flashcard drilling trains you to recognise the correct answer when you see it on a card. This is a weaker and less transferable form of memory than the associative knowledge that allows you to produce a word spontaneously in the right context. Drilling recognition memory produces words you can define but can't generate.
Neither of these problems is fatal to flashcard use. They become problems when flashcard drilling is treated as the primary vocabulary learning strategy rather than as a supporting tool for something more powerful: reading.
The most robust finding in memory research is the spacing effect: information reviewed at increasing intervals after initial learning is retained far more durably than information reviewed in concentrated blocks (massed practice). This has been replicated hundreds of times across more than a century of research, and it is the foundation of spaced repetition systems like Anki.
The mechanism is related to the difficulty of retrieval. When you review a word shortly after learning it, the memory is still fresh and retrieval is easy. The ease of retrieval produces relatively little long-term benefit. When you review it after a longer interval — at the point where you can just barely recall it — the effortful retrieval strengthens the memory trace significantly more. This is sometimes called desirable difficulty: making retrieval slightly harder in the short term produces better long-term retention.
For Chinese vocabulary, this means that cramming a list the night before class produces very poor long-term retention, and that reviewing the same vocabulary every day is not much better. The optimal review schedule — one day after initial learning, then three days, then a week, then two weeks, then a month, and so on — is exactly what a well-configured spaced repetition system calculates automatically.
The practical lesson: if you use Anki for Chinese vocabulary, configure it correctly. The default settings are reasonable but learners often adjust the interval modifier upward and the new cards per day upward in ways that create huge review backlogs and defeat the spacing effect. Keep new cards per day low (10-15), keep reviews manageable, and trust the algorithm.
The most important insight from reading acquisition research is that encountering a word in context produces richer and more durable memory encoding than encountering it in isolation, all else being equal. And in extensive reading, where you encounter many words in many different contexts across many texts, vocabulary acquisition happens rapidly and durably without explicit study of those words.
Paul Nation, one of the leading researchers in this field, estimates that a word needs to be encountered 10-15 times in varied contexts for it to become part of active vocabulary. This is not 10-15 flashcard reviews — it is 10-15 encounters in real language contexts, where the word appears in a sentence with meaning, surrounded by other words that contribute to disambiguating its sense.
Graded reading creates ideal conditions for this. A well-designed HSK 2 story recycles the same vocabulary across different contexts within and across stories. The Beijing day-in-the-life trilogy is a good example — three linked stories covering the same day that continuously recycle and extend the same vocabulary pool. The word for 'arrive' appears in one story when a character arrives at work, in another when they arrive at a destination after a journey, in a third when someone arrives unexpectedly. Each encounter adds a slightly different contextual dimension. By the tenth encounter, in different situations, the word is known not as a definition but as a language element with natural range of use.
No flashcard system can replicate this, because a flashcard has one context: the card itself. What reading provides — and drilling cannot — is the full, rich, contextual web of associations that constitutes knowing a word in the way a native speaker knows it.
This is not an argument for abandoning flashcard study. It is an argument for understanding what each tool does and using them accordingly.
Flashcards (and specifically spaced repetition systems like Anki) excel at one thing: getting a word into your memory efficiently so that you will recognise it when you encounter it in reading. This is the bootstrapping function — building the initial vocabulary threshold that makes graded reading viable.
At HSK 1 level, you need roughly 150 words before reading flows. Anki can get you there in three to four weeks of fifteen-minute daily sessions. Without this foundation, reading is a constant stop-and-look-up exercise that breaks the flow of meaning-processing and defeats the purpose.
Once you have the initial threshold, reading takes over as the primary acquisition mechanism. Words that appear in reading get contextual reinforcement. Words that appear in Anki reviews get spacing reinforcement. The two systems work together: Anki ensures you encounter each word at the optimal review interval; reading provides the contextual encounters that convert flashcard recognition into real language knowledge.
The practical balance: at HSK 1-2, spend roughly equal time on Anki and reading. From HSK 3 onwards, let reading do the majority of the vocabulary work, and use Anki for targeted vocabulary — specific words that you know you should know but keep forgetting, or new vocabulary outside your graded reading range.
One of the most demotivating aspects of Chinese vocabulary learning is the apparent enormity of the task. Native Chinese speakers know tens of thousands of words. Surely reaching any kind of functional fluency requires years of grinding through word lists?
The reality is more encouraging. Research on reading comprehension and vocabulary coverage consistently finds that understanding 95% of a text requires knowing approximately 98% of the words in it. For everyday written Chinese, this threshold is reached at around 2,000-3,000 words. For newspaper reading, you need somewhat more — roughly 4,000-5,000. For literary texts, more still.
But the key insight is that the most common words cover a disproportionate share of real text. The 1,000 most frequent Chinese words cover roughly 85-90% of everyday written text. The 2,000 most frequent words cover around 95%. This means that investment in the most common vocabulary pays enormous dividends very quickly, and that chasing low-frequency vocabulary at early and intermediate stages is much less efficient than consolidating high-frequency words through extensive reading.
As a practical benchmark: at HSK 4 (1,200 words), you can read simplified Chinese texts with a dictionary for unfamiliar words rather than needing a dictionary for almost everything. This qualitative threshold — where reading is possible rather than impossible — is achievable by most dedicated adult learners within 18-24 months.
Not all vocabulary is equally valuable to learn. The word for 'computer' will appear in your reading far more often than the word for 'centipede,' and the investment in learning it is correspondingly more efficient.
Frequency-based vocabulary learning — prioritising the most common words in the language — is well-supported by both research and experienced learner practice. The HSK vocabulary lists are, in fact, broadly frequency-based: HSK 1 covers the most common 150 words; HSK 2 covers the next 150; and so on. This is why the HSK framework is a reasonable guide to vocabulary learning order, even for learners who are not preparing for the HSK exam.
For learners using Anki, pre-built frequency-ordered decks exist and are a better starting point than random topic-based vocabulary lists. Topic-based lists (food vocabulary, colour vocabulary, direction vocabulary) have the advantage of coherence and the disadvantage of not prioritising frequency — you end up knowing the word for 'chartreuse' before the word for 'how.'
A distinction that vocabulary researchers make, and that language learners often overlook, is between passive and active vocabulary.
Passive vocabulary comprises the words you can understand when you encounter them in reading or listening. Active vocabulary comprises the words you can produce spontaneously in speech or writing. Passive vocabulary is always larger than active vocabulary in any learner's target language, and the gap is particularly wide in Chinese because learners spend most of their practice time reading and listening rather than speaking and writing.
This is not necessarily a problem. Reading fluency and listening comprehension drive the majority of useful Chinese learning outcomes for most adult learners, and both depend primarily on passive vocabulary. If your goal is to read Chinese content, watch Chinese media, and understand spoken Chinese, a strong passive vocabulary is exactly what you need.
For speaking fluency, however, some deliberate production practice is necessary to convert passive vocabulary into active vocabulary. The most effective methods are output-focused: speaking practice with a tutor or language exchange partner, writing short summaries of what you have read, or using sentence composition (writing new sentences with target vocabulary) as part of your Anki workflow.
One important caveat for English speakers: Chinese vocabulary acquisition is harder in one specific way and easier in another, compared to European languages.
The harder way: no cognates. English shares thousands of words with French, Spanish, German, and other European languages. An English speaker encounters a French word they've never studied and often has a reasonable guess from the shared root. In Chinese, there are almost no such shortcuts. Every word is new. The vocabulary acquisition phase is longer and requires more deliberate effort.
The easier way: once you know a base of Chinese characters, new vocabulary is often compositionally transparent. The word for 'telephone' (电话) is composed of the characters for 'electricity' and 'speech.' The word for 'train' (火车) is 'fire-vehicle.' The word for 'airport' (机场) is 'machine-field.' At intermediate level, encountering an unfamiliar compound word whose component characters you know gives you an immediate rough sense of the meaning. This compositional transparency accelerates vocabulary acquisition in ways that are not available in non-morphological vocabularies.
The practical implication: invest in learning the most common Chinese characters and their individual meanings, not just as parts of compound words. The 500 most common characters appear in thousands of compound words, and knowing them as building blocks gives you a head start on any unfamiliar compound you encounter.
Pulling everything together, here is a practical daily vocabulary routine that reflects the principles above:
Morning (15-20 minutes): Anki reviews first, then new cards if reviews are complete. Keep this focused and uninterrupted. This handles the systematic spacing function.
Core session (20–30 minutes): Read a graded Chinese story. If you're at HSK 1–2, the Huangshan Adventure (HSK 1) or the Beijing trilogy (HSK 2) are ideal. When you encounter an unknown word, use tap-for-definition or Pleco to check it — then continue reading. If the word appeared twice in the story or seems important, add it to your Anki deck. Otherwise, let it go and let recurring encounters teach it naturally.
Vocabulary notes (5 minutes, optional): After reading, note 2-3 words you want to remember from the session. Write them in a sentence from the story where you encountered them. This brief production practice converts recognition to more durable encoding.
Total: 40-55 minutes. Done consistently, this routine produces measurably faster vocabulary growth than either pure flashcard drilling or pure reading, because each method is doing what it does best in service of a whole that is greater than either part.
Vocabulary and grammar are often treated as separate domains in language learning, but they are deeply intertwined in Chinese, and understanding their interaction changes how you should think about vocabulary acquisition.
In Chinese, many grammatical functions are expressed through specific vocabulary items rather than through morphology. The particle 了 (le) — explained in Dumplingo's grammar guide — is technically a grammatical marker indicating a completed action or a changed state — but learning it is a vocabulary problem as much as a grammar problem. The same is true of aspect markers like 过 (guo), potential complements, and a range of adverbs that carry specific grammatical meaning. Learning these items as vocabulary — encountering them in context, building a sense of how they function in sentences — is often more effective than studying them as grammar rules.
The implication for vocabulary study: when you encounter a vocabulary item that seems to be doing grammatical work, don't just note its rough translation. Note the sentence context and, if possible, two or three different sentences where it appears. This contextual set will give you an intuition for its grammatical behaviour that a definition cannot.
Input — reading and listening — is the primary driver of vocabulary acquisition. But output — speaking and writing — plays an important secondary role in consolidating vocabulary that has been acquired through input.
When you produce a word — use it in a sentence you construct yourself, rather than recognising it in a sentence someone else has written — you engage a different set of cognitive processes than recognition requires. You must access the word, not just respond to its presence. You must fit it into a grammatical structure. You must produce the correct tone and pronunciation. These additional demands create additional memory encoding, which strengthens the word in your long-term vocabulary.
For learners focused primarily on reading, a simple output practice is to write a sentence using each new word you decide to add to your vocabulary notes. Not a translation of the example sentence from the dictionary — your own sentence, using the word in a context you create. Five minutes of this after a reading session adds meaningful consolidation without disrupting the input-focused core of your practice.
A question that often arises for intermediate learners is when to reduce or stop systematic vocabulary drilling and rely primarily on reading for vocabulary growth. There is no universal answer, but a practical heuristic is useful.
When you find that most of the words on your Anki deck are words you have already encountered in reading — when the reading is teaching you vocabulary at roughly the rate you can add new cards — you have reached the point where reading is the more efficient acquisition mechanism. At this stage, you can let your Anki deck mature (reviewing old cards but adding fewer new ones) and trust that graded and eventually authentic reading will continue to expand your vocabulary faster than drilling can.
For most learners, this transition happens around HSK 3-4 level. Before that point, the vocabulary you encounter in reading is often too dense with unknowns to acquire efficiently without Anki scaffolding. After that point, Anki becomes a supplement rather than a foundation. Making this transition consciously — rather than either abandoning Anki too early or clinging to it too long at the expense of reading time — is one of the marks of a learner who understands the tools they're using.
Dumplingo's graded stories are designed to recycle vocabulary across connected episodes, accelerating contextual vocabulary acquisition. Read your first story free at any HSK level. Start reading →