I found a brief note on Wikipedia that since 2020, HSK is going to be restructured. It links to a tweet, which shows a page in Chinese. Is there any more information compared to what Wikipedia currently states? If not, are there any educated guesses that can be made based on prior two changes to HSK structure and the way Hanban works?

  • I saw two close votes claiming I'm asking for learning resources here. How's this question about HSK2020 structure different than others? – d33tah Jun 21 at 10:17
  • 2
    Yeah, you're right. Doesn't look like a resource request to me. I've protected it now, but that doesn't necessarily stop close flags. – Mo. Jun 21 at 11:16
  • I don't see anything about this on chinesetest.cn. – imrek Aug 4 at 19:07
  • It's also astounding that there's very little information on this change available in English from Hanban and related organizations. For those advanced it shouldn't be a problem to read the Weixin posts and that paper, but for those just trying prepare for the entry levels (1-2-3...) it must be infuriating. – imrek Aug 4 at 19:44

At the time of writing, the best guess is that the HSK 3.0 vocabulary list is in the recently-published book:

A Practical Manual of Tone Patterns and Formats of Stressed and Unstressed Syllables in Mandarin Words for the Application of Teaching Chinese to the Speakers of Other Languages
刘英林, et al.

A redditor posted images of some of the pages from this book on Imgur (from this post; I transcribed some of it here). I bought this book from JingDong (also available here), and posted about it on Reddit. The guess is primarily based on a tweet from an account called HSK Official:

HSK is about to be reformed. #HSK In 2020, the Chinese Proficiency Standards will usher in a new change: a hybrid paradigm of “Three Stages and Nine Levels” characterized by integration and all-in-one.

There are also reports on QQ: here and here. (Note the tweet links to the first QQ post, which mentions 刘英林 and other authors.) These posts use screenshots from the recently-published paper:

刘英林, et al. (available here)

This paper has the same lead author as the book. From this paper (p.152) we have the vocabulary breakdown:

enter image description here

(Here e.g. 140/608 implies new = 140, cumulative = 608. The second last column is vocabulary 词汇.)

I'll reproduce my Reddit post in full, since I go into a fair amount of detail about the aforementioned book:

The book is almost nothing more than two long lists. The first list contains exactly 11092 terms ordered by pinyin, and occupies 253 pages. The second list contains most (not all) of the 11092 terms ordered by level, then by tones, and occupies 214 pages. It appears that the second list omits all single-character terms: e.g. 2057 一① 多 appears in the first list, but not in the second list.

Both lists contain precisely the following:

  • 序号: sequence number (the first list begins with 1 阿拉伯语 [on page 1] and ends with 11092 做生意 [on page 253])

  • 等级: level (see below)

  • 词语: the Chinese characters

  • 轻重格式: which syllables are emphasizes, e.g., 左边 is ZQ (重轻; heavy, light) implying that the stress on 左 is heavy and the stress on 边 is light

  • 声调组合: the tones, e.g., the tones for 昨天 are written 21. 0 is used for the neutral tone.

  • 拼音: pinyin (with two tweaks: (a) a · indicates there's multiple pronunciations, and the example they give is 道理 which can be either dàolǐ or dàoli; (b) a // indicates that the word is separable, and the example they give is 见面 written "jiàn // miàn" [e.g. 见面 in 见你面])

  • 词性: the type of word: 名词 (noun, e.g. 飞机), 动词 (verb, e.g. 听说), 形容词 (adjective, e.g., 安全), 数词 (number words, e.g. 一半), 量词 (measure word, e.g. 公斤), 代词 (e.g. 有一些), 副词 (e.g. 当然), 介词 (e.g. 通过), 连词 (conjunction, e.g. 所以), 助词 (auxiliary), 叹词 (interjections), 拟声词 (onomatopoeia). Many don't fall into any of these categories.

There are 3 major levels: 一 (普及化等级, 2245 terms), 二 (中级, 3211 terms), and 三 (高级, 5636 terms). Level 一 splits into 一① (505 terms), 一② (837 terms), and 一③ (903 terms). Some of the terms (1461 terms) within level 三 are classed as "附录" (appendix). It seems probable that the new HSK levels 1 through 3 respectively have the vocabulary in the lists 一①, 一②, and 一③, and probable that the new HSK4-6 and HSK7-9 contain those in 二 and 三 respectively. However, it's not clear how they are divided.

In this book, there's an emphasis on "stress" (not just "tones") which I haven't seen elsewhere previously. In particular they write:


Here, 音节 ("syllable") is regarded as the combination (声母, 韵母, 声调) = (initial, final, tone). The book emphasizes how this is not enough for authentic Chinese pronunciation.

The 4-character words are:

  • 一①: none [page 260]

  • 一②: 公共汽车 共用电话 一路平安 一路顺风 不好意思 [page 274]

  • 一③: 中华民族 高速公路 电子邮件 另一方面 [page 291]

  • 二: 这就是说 西班牙语 酸甜苦辣 绝大多数 五颜六色 一般来说 一模一样 自言自语 意想不到 [page 351]

  • 三: 知识分子 东张西望 阿拉伯语 伊斯兰教 千方百计 依依不舍 粗心大意 亲朋好友 官僚主义 翻来覆去 哄堂大笑 多种多样 相比之下 资本主义 丰富多彩 鸦雀无声 公共场所 千变万化 说干就干 无家可归 长期以来 提心吊胆 没完没了 除此之外 时好时坏 难以想象 由此可见 无可奈何 从早到晚 实话实说 无论如何 一事无成 毫不犹豫 实事求是 一动不动 随处可见 犹豫不决 不大不小 十字路口 总的来说 小心翼翼 改革开放 与此同时 喜怒哀乐 讨价还价 也就是说 与众不同 乱七八糟 一帆风顺 大街小巷 兴高采烈 大吃一惊 或多或小 一天到晚 至关重要 未成年人 大名鼎鼎 帝国主义 一年到头 一长一短 自由自在 异口同声 不可避免 二氧化碳 四面八方 必不可少 社会主义 这样一来 断断续续 各式各样 不怎么样 讨人喜欢 一塌糊涂 [pages 434-436]

  • 附: 忽高忽低 息息相关 ...omitted... 高新技术 有两下子 [pages 460-468]

This indicates the new HSK1-6 (levels 一 and 二) will not contain any idioms, i.e. chengyu (成语). (If you want to compare, a while back, I made a list of the HSK chengyu.)

(The book came with a CD labelled "MP3", but I don't have a CD player, so I don't know what's on it. Presumably pronunciations of these words.)

(Note the per-level vocabulary numbers in the book are not an exact match for the numbers in the paper. Although the total number is the same: 11092.)

In this comment, I also analyzed some of the vocabulary. In particular:

The new level 1 appears to be formed from: 36% of the HSK1, 23% of the HSK2, 12% of the HSK3, 1% of the HSK4, and 3 HSK5 words (工人, 日期, 平安), and many non-HSK words (forming 51% of the total words in the new level 1). [This ignores single-character words.]

The new level 2 appears to be formed from: 7% of the HSK1, 19% of the HSK2, 40% of the HSK3, 28% of the HSK4, 6% of the HSK5, one word from HSK6 (事业), and many non-HSK words (forming 32% of the total in the new level 2). [This ignores single-character words.]

(I may have made some transcription mistakes, so these numbers might not be perfect.)


| improve this answer | |
  • 4
    The usage of a dot to indicate 轻读 is quite common in printed dictionaries, such as 现代汉语词典. To be more precise, the introduction in the dictionary says that it means "一般轻读", meaning that even though a word is listed with a full tone, it's common to reduce it anyway. – Olle Linge Jun 21 at 15:49
  • I'd put a warning at the top of your post to remind people to wait for Hanban/Confucius Inst. to confirm the vocabulary lists before rushing to buy a book that could cost up to $50 with shipping, fees, etc. depending on the location. I don't see anything official on the new test on chinesetest.cn, but a lot of speculation on forums and social media. Let's wait until Hanban bureaucrats finally post something thorough, concise and complete, and not just drop some selective info on social media. – imrek Aug 4 at 20:23
  • So more is expected for the exam than what is being taught? That seems odd. – mic yesterday

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.