This database is searchable as CSVs. The quickest way to get an addition or correction merged is to fork the repo or clone locally, edit the CSV file and send in a pull request.
The focus is on antedating and amending entries where the lemma is still current in educated speech. Editorial history and argumentation can be accessed via git blame. Triage of items in Yang's bibliography (2026-01-14) is now completed.
This Supplement is designed to be consulted alongside Huang. A solidus ('/') in a cell indicates that the infomation is the same as in Huang or the previous entry.
Lemma
To facilitate cross-checking, the arrangement of lemmas is that of Mair et al. (2003) with the following modifications:
words whose head characters share the same sequence of letters and tone are sorted by subsequent characters;
in case of a tie, they are sorted by the complexity of the characters concerned in terms of (1) the number of strokes in the radical and (2) the number of residual strokes.
The ideographs used are those suggested by the Traditional Chinese character set of iOS at the time of the commit, which tend to be the only form attested throughout the period covered here.
Numerals following a lemma refer to the different senses of a homonymous word, which form a superset of the senses defined in Huang. If a word is absent from Huang, the numbering of Hsien-han (for neologisms) or Han-ta (for traditional words which underwent semantic change) is used.
Obsolete words, phoneticloanwords and anatomical terms are collected in separate appendices. For loanwords, see also On Diglossia; for obsolete words, see also the Stylistic Appendix.
Sense
Glosses serve to disambiguate and are set in roman type.
Domain classification is set in italic type.
Assignment to Word Class follows the analysis of Huang–Shih (2016) as corrected in the Grammatical Appendix.
Year
The first entry for a lemma represents the first known attestation. When a date is given, it is generally earlier than the earliest quotation in Huang except in the case of postdating.
When the publication date and composition date of a source differ, the dating style of the Middle English Dictionary and OED3 is used.
Quotation
The exact variant ideographs as used in the source are reproduced when they are available in the latest version of FSung. Failing that, the closest variant is substituted.
The typography of the source is reproduced to the extent that the baseline features of modern browsers allow.
A blank means we had no access to the source.
For traditional critical symbols, see West (1973).
Source
Dictionary evidence is treated as a primary rather than secondary source, and represents one single attestation instead of a statement about contemporary usage.
When a word is attested in an item carried in multiple newspapers on the same day, the least well-known paper is cited as a way to diversify scholarly sources away from Shen-pao and L'Impartial. Other attestations may be located in 全國報刊索引.
An exclamation mark ('!') preceding a source means the quotation is taken from a modern edition with normalised orthography and we had no access to an early version.
Authorship is given where it is considered desirable to draw attention to the fact that the writer was not a native speaker. Whilst neologisms in such writings can eventually enter the language, scholars should exercise caution when drawing conclusions about the lexicon of contemporary native speakers.
Authorship is not otherwise indicated if it can be easily looked up, although controversial cases are dealt with in the Checklist of Editions. An exception is made, however, for writers of extralinguistic importance.
A plus sign ('+') following a source means the word is attested more than once in that source.
When a shelfmark is added in parentheses, it can be used to identify the relevant digital copy linked to in the Checklist of Editions. Transcription is made from the earliest surviving edition, unless the relevant section in a copy comes from a later completion.
In addition to the studies referenced in the bibliography, the following scholars and institutions have contributed directly to the success of this Supplement:
Entries: Yan Cui, Lok Ching Roxanne Fung, Ruohan Ma (etymology), Theodore Nze (etymology), Jane Xuzhen Tang.
Research: Edward Huang (synonyms), Xiaoyu Jin (orthography), Xingni Li (orthography), Ethan Liu (orthography).
Tooling: Ansel X. Zhang (font).
Corrigenda: 操瑞青, 石雲里, 田野村忠温.
Supervision: Gabriele Tola (MT24).
Books/articles: Adrian Dudink, 黃河清, 菊地恵太, Federico Masini, 繆蓬, 沈國威, SOAS Special Collections, 楊馳, 莊欽永.
Can't edit?
If you are in a place with internet restrictions, contact your local authorities and ask them to unblock the site for you. In the meantime, you can send the contributions to 67616464787179717a6054607b7a7f617a6760787166397b7a39607c713976617a703a777b79.
