How characters carry sound and meaning.
5,274 characters with historical readings · 2,623 imported component relations · every claim has a source.
Start with a connection
One dictionary · distinct evidence
CC-CEDICT supplies word definitions. Unihan supplies character properties and selected readings. BabelStone describes graphical structure. These scholarly layers add attributed component roles and historical phonology.
Character formation and word etymology are different questions. A phonetic-series connection does not establish that two words share an origin.
Middle Chinese readings
7,127 retained readings on 5,274 exact character matches from PHONO-ML · Jacques & Delaporte / CRLAO. No readings are copied across simplified/traditional variants.
Transcriptions, fanqie spellings, category strings and identifiers remain aligned. In this notation, X and H mark the rising and departing tone categories; they are not Mandarin tone numbers.
The dataset derives readings from Guangyun using the methods of Baxter (1992) and Jacques (2021). It was generated automatically and checked on a subset by its authors. Source errors remain possible.
CC BY 4.0 · snapshot 103cd735 · Normalized selection · Original CSV · License
Phonetic components
2,623 relationships retain explicit single-character 聲 labels from CJKVI / Kawabata · component analyses 18.02.20. Source rows, comments and reference numbers remain available on each entry.
Hill & List use this analysis layer in their formation-graph study. They distinguish direct components from broader phonetic series and caution that traditional analyses are not invariably correct. This import is attributed source data, not a claim that every pair has been individually reviewed.
No semantic role is inferred from leftover graphical parts. Variant chains, 亦聲 and 省聲 labels are outside this import.
GPLv2 · Adapted data · Original source · Original notice · License · Transformation script
Selected Old Chinese records
Baxter & Sagart · Old Chinese reconstruction v1.1 (2014) · version 1.1, 20 September 2014, printed p.109 / PDF p.111. These entries share the GSR 0812 catalogue group, which is distinct from a direct-component claim.
| Character | Mandarin | Middle Chinese | Old Chinese | GSR |
|---|---|---|---|---|
| 青 | qīng | tsheng | *[s.r̥]ˤeŋ | 0812c' |
| 清 | qīng | tshjeng | *tsʰeŋ | 0812i' |
| 情 | qíng | dzjeng | *[dz]eŋ | 0812l' |
| 請 | qíng | dzjeng | *m-tsʰeŋ | 0812k' |
| 請 | qǐng | tshjengX | *[tsʰ]eŋʔ | 0812k' |
| 晴 | qíng | dzjeng | *N-tsʰeŋ | 0812- |
| 精 | jīng | tsjeng | *tseŋ | 0812g' |
| 星 | xīng | seng | *s-tsʰˤeŋ | 0812x |
| 醒 | xǐng | sengX | *s-tsʰˤeŋʔ | 0812b' |
* indicates reconstruction; brackets and other notation retain uncertainty in this model. This small selection is not a full historical dictionary. No Old Chinese reading is supplied for 靜 here.
Literature & method
Nathan W. Hill & Johann-Mattis List (2019), Using Chinese Character Formation Graphs to Test Proposals in Chinese Historical Phonology, Bulletin of Chinese Linguistics 12(2), 186–200. DOI · Author preprint
Six explicitly attributed component claims are drawn from §1, preprint pp.1 and 3. Graph edges keep direct phonetic components, semantic components and catalogue series membership separate.
Original explanations and source comparisons are AI-assisted; independent expert review is still outstanding. Editorial method ↗
Further reading supplied for this project
Gam P. Go, Understanding Chinese Characters by Their Ancestral Forms, third edition, 1995, ISBN 0-9623113-4-0.
This illustrated introduction informs the learning perspective. Its explanations require individual historical cross-checks before becoming evidence records. The book and its drawings are not redistributed here.
Put the evidence to work.
Explore connected reading guides with exact Classical passages, explicit comparisons, and questions that test an interpretation.
Open the Classical guides ↗ · Compare texts ↗ · Read eight complete witnesses ↗