Chinese Vocabulary: Lists, Context, Theme or Frequency?

Neither wins on its own. A list tells you what is worth learning and in what order, and context shows you how a word behaves. The best arrangement uses a levelled list as the map, lessons and reading as the way you meet the words, and spaced repetition to keep them. In Taiwan there is a ready-made map: the TOCFL levelled word list.

Here is how the four common approaches (lists, context, theme, frequency) compare, and how to use each at the stage where it helps most.

Lists: the map

A vocabulary list answers the question “what should I learn next?” Without one, you learn whatever you happen to meet, and you cannot tell what is missing.

Lists have a known weakness. A word learned from a list is a word with a one-line English gloss. You may recognise 辦法 (bànfǎ, method, way) on a card and still not know that it turns up in set phrases such as 沒辦法 (méi bànfǎ, there is no way / cannot be helped). List learning gives recognition without much feel for use.

Use lists to decide what to learn next.

Context: the feel

Learning a word inside a sentence, dialogue or text shows you its partners, its register and its typical grammar. Chinese needs this more than many languages: measure words, verb and object pairings and context-dependent readings (長 as cháng or zhǎng, 行 as xíng or háng) are all invisible in a bare list entry.

Context has its own weakness: it is unsystematic. A learner who only reads what interests them will meet 手機 (shǒujī, mobile phone) a hundred times and 辦法 only occasionally, and will not notice which common words they have skipped. Context also tends to teach words slowly, because each word appears rarely.

Frequency: the return on effort

Frequency means studying the commonest words first. The logic is simple: a word you meet constantly pays back faster than one you meet once a year. In the first months this matters most, since a small set of very common words appears in a large share of everything you will read or hear.

Frequency lists have a cost. They are drawn from a corpus, and the corpus shapes the result. A list built from mainland text will favour mainland vocabulary, and a list built from news will underrepresent the words you need to order dinner or talk to your landlord. Some of the commonest words are also grammar words that are hard to learn as isolated entries.

For Taiwan-bound learners, the frequency list should come from Taiwan usage. That is one reason to work from a Taiwan-developed list rather than a generic one.

Themes: how courses actually teach

Textbooks group words by topic: food, transport, family, work. A Course in Contemporary Chinese (當代中文課程, “Dangdai”) is organised by lesson topic, with each lesson’s words introduced through dialogues and texts around a situation. The Dangdai Book 1 guide shows how that looks in practice.

Theme-based learning works for a reason that is easy to underrate: related words arrive together. If you are learning how to order at a restaurant, 菜單 (càidān, menu), 點菜 (diǎn cài, to order dishes) and 買單 (mǎidān, to pay the bill) are met in one scene, and you can use them immediately. You meet the words in the situation where you will use them.

The weakness is coverage. A theme gives you the words for that situation, but the units are chosen by editors, not by how often words occur. Some very common words appear only incidentally, or never.

TOCFL’s levelled list: both at once

The TOCFL 8,000-word list (華語八千詞表) is organised by level, from Novice up to Level 5. Each level adds entries that the developers judged suitable for learners at that stage. In the 2023-07 edition the list has 7,517 entries in total, despite the name, with cumulative totals of 394 after the Novice stage, 741 at Level 1, 1,226 at Level 2, 2,399 at Level 3, 4,741 at Level 4 and 7,517 at Level 5. The TOCFL word count guide explains those numbers.

A levelled list sits between a frequency list and a topic list. It is organised by learner level rather than published as a frequency ranking, it moves from earlier-stage words to later-stage words, and it belongs to Taiwan’s own test. You can browse the entries on the TOCFL vocabulary pages and see how the levels are laid out on the TOCFL levels page.

It does not teach words, it names them. That is why it combines well with a textbook: the textbook supplies the themes and context, and the list tells you whether you are covering what the test and everyday use assume.

A combined approach

A plan that takes the strengths of each:

1. Early on, follow the textbook units. Dangdai’s theme order gives you words you can use in conversation straight away.
2. Check coverage against the levelled list. Every few weeks, look at your level’s entries and find the common words your lessons have not yet covered.
3. Fill the gaps with sentences. Learn each missing word in a short example, not as a bare gloss.
4. Read. Graded readers repeat common words in varied contexts, which is what lists cannot do. See graded readers in Traditional Chinese.
5. Review with spaced repetition, using word cards by default and sentence cards for the words you keep misusing, as described in sentence cards vs word cards.

What to avoid

• Memorising a long list straight through with no use of the words. You will recognise them for a while and forget them.
• Learning only from content that interests you, with no check on what you have missed.
• Treating a frequency ranking from an unknown corpus as authoritative for Taiwan.
• Learning words in Simplified form from a mainland list and then relearning them in Traditional. Our Traditional vs Simplified comparison explains why the script choice should come first.

Related reading

• How many Chinese words do you need? By TOCFL level
• Dangdai Book 1 guide
• Sentence cards vs word cards
• TOCFL vocabulary by level