Property talk:P1721

@Giftzwerg 88, Zolo, Innocent bystander, Pasleim, Nikki: We now have Lexemes. Do we intend to use this property on Lexemes? Or should we create a new property? Deryck Chan (talk) 16:23, 20 December 2018 (UTC)Reply

Oh, I missed this ping. I intend to use it on lexemes and I can't imagine a new property would be approved because there's no semantic difference between the pinyin transliteration of a string or monolingual text statement and the pinyin transliteration of a form. - Nikki (talk) 09:01, 30 December 2020 (UTC)Reply

I added as main value (Q54828448) to the property scope in Special:Diff/1331509337 because it's needed to avoid constraint errors on lexeme forms. This unfortunately makes the constraint fairly useless but we don't have a way to restrict constraints to particular entity types. If you think we should (or shouldn't), perhaps you'd like to comment on the ticket I created recently for it - phab:T269724.

- Nikki (talk) 09:11, 30 December 2020 (UTC)Reply

Although Library of Congress (Q131454) has replace Wade-Giles (Q208442) with pinyin (Q42222) as the default romanization system since about 1990, but many historical documents still use Wade-Giles (Q208442) to romanize Chinese proper noun (Q147276) and many people outside mainland China (Q19188) may use Wade-Giles (Q208442).

For example, Hu Shih (Q47667)'s English tag is Hu Shih, Zhang Zhizhong (Q197347)'s alias name Chang Chih-chung.

I think Wikidata should have a Wade-Giles (Q208442) property like Hanyu Pinyin transliteration (P1721).

@Nikki, Deryck Chan, Zolo: Kethyga (talk) 06:09, 13 April 2022 (UTC)Reply

While Hanzi to pinyin transliteration is sometimes tricky to automate, pinyin can be converted to Wade-Giles using simple deterministic rules. I think using only one transliteration system in Wikidata is simpler. Others can be automatically generated by the client when needed. Zolo (talk) 22:09, 14 April 2022 (UTC)Reply

The mapping between Wade-Giles and Hanyu Pinyin technically isn't one-to-one, as they represent different points in time: WG for late 19th century Beijing; Pinyin for today's Beijing. The convoluted mapping explained in e.g. w:Wade–Giles#Vowel_o is best explained as a historical sound shift. I'm ambivalent about a WG property though, as you've rightly pointed out that it mainly serves a historical purpose, and a dedicated property might encourage spurious creation of transliterations of Mandarin words that did not exist until the 21st century. Deryck Chan (talk) 23:27, 18 April 2022 (UTC)Reply

I'm in favour of dedicated transliteration properties. People are already adding Wade-Giles transliterations using transliteration or transcription (P2440), but that property is problematic because it requires a qualifier but is often used as a qualifier itself. - Nikki (talk) 08:51, 27 April 2022 (UTC)Reply

I have seen several Anglosphere (Q541576) countries document historical material with Wade-Giles (Q208442), Library of Congress (Q131454), National Library of Australia (Q623578). Take Song Jian (Q9184985)'s Library of Congress authority ID (P244) webpage [1] as an example, without strenuous search, one won't easily know Chung-kuo kʻo hsüeh chia tzʻu tien is 中国科学家辞典/zhōngguó kēxuéjiā cídiǎn, a series of books published in 1982. Basically, most Chinese people can't read the book title. A property is deserved to record the historical material. Kethyga (talk) 10:23, 27 April 2022 (UTC)Reply