#yr chao

[tcchinese]

As the General Chinese list that Wikipedia linked to has been moved, here’s a link to the list which I have copied from the new source (in case it moves again) and pasted onto Pastebin (with numbers).

[tcchinese]

Two Cantonese tongue twisters with the Cantonese Romanization Museum and TCC, ordered for ease of comparison.

They read respectively:

- Mr postman is skillful at delivering letters; he delivers them swiftly.

- 1 dollar for a catty (around 500g) of chicken, 1 dollar for a catty of turtle, so is the chicken or the turtle more expensive?

Notes:

- "蚊" is a Cantonese word and its TCC equivalent is “元“ (ngiuēn).

- It is not certain (as of yet) whether “定“ (dieng) should be used in this sense in TCC.

[tcchinese]

Basic summary of The Tonal Spelling of Cantonese

Article: The Tonal Spelling of Cantonese (1958) From: T‘oung Pao Article by: Søren Egerod

The article is a review of an earlier article, “The Transcription of Cantonese”, written by Tor Ulving, whose work reviewed Chinese transcriptions and was most critical of Chao’s systems.

The first point Egerod addressed is that Ulving claimed that “u” stood for many vowels, though Ulving mentioned that some are mere mistakes (which I deduced to be based off of Wade-Giles since Ulving criticized that [u] has been written as <huen> when Egerod wrote that it stood for “[χwən]”). Another point is, Ulving listed [ʊ] (though he writes it as “ong”) as one of the inconsistencies, as Chao did not write it with a <u>, while Egerod responded that the sound is in fact [o]. This comes down in the end to different analyses of the same sound.

As for Chao’s Cantonese Romanization, Ulving claimed that Chao’s treatments of the tones were not faithful to Cantonese phonology and were based off of their sounds in Ancient Chinese, when Egerod responded that the terms “bheng, zheug, hoy and yap” are in fact the names used by those who study Chinese dialectology.

The next point. Again (or perhaps consequently), Ulving placed too much emphasis on the actual tones, and criticized that a trailing -h in Gwoyeu Romatzyh (GR) represented a falling tone in Mandarin, and was used to represent a tone that wasn’t falling in Cantonese, and said that a person would need “superhuman concentration” to adjust to such a difference. Egerod rightfully points out that they are tonal cognates and that a similar spelling might even help people recognize that the words in fact have the same meaning in the National Language as well as Cantonese.

Though in the end, Egerod wrote that both Chao’s and Barnett’s systems were used with success and that while Egerod has found some problems with Barnett’s system, it was not horrible. This though possibly means Barnett had a system in use that wasn’t Barnett-Chao, but there wasn’t really much description given in the article.

[tcchinese]

[Reference]

3 JSTOR links which I will likely read in the near future (if possible):

- The Transcription of Cantonese: A Critical Review of Some Current Systems of "Tonal Spelling", and a Presentation of a New Rational Transcription

- A Transcription for Cantonese: Notes on Mr. Yuen Ren Chao's "Cantonese Primer" by Barnett (even though he made a real mess of Yutyux Romadzih, I’d like to look at his reasons and likely find out why Barnett has made such a big mistake)

- Review of A Chinese: English Dictionary in the Cantonese Dialect by Eitel and Genähr by Ball (in Ball’s books, he mentioned how Eitel failed to give accurate Cantonese pronunciations and how Cantonese has been mistakened for a “dialect”. I would expect things such as these to pop up in the review.)

[Extra] The Secret of the Chinese Method of Transcribing Foreign Sounds (Continued)

Edit: [Extra 2] The Tonal Spelling of Cantonese

[tcchinese]

Quick comparison between Romanizations that analyse "vowel length" in Cantonese. These Romanizations generally treat [ɵ] as the same phoneme as /o/ (LSHK "u" before velars). Eitel treats both as /ʊ/.

[tcchinese]

Zeng Zi has noted two things: the EDB Romanization is no longer a choice for looking up information on EDB's website. Since EDB was originally named so as it was recognized by it as the only standard Romanization, Zeng Zi calls it the "ex-EDB Romanization".

Also, [tut˥] (嘟) has become a verb in daily Cantonese usage, as he points out. This means that Chao's Cantonese Romanization does not officially cater to this sound, but using his, and my, interpretation, the sound could be written as <twut>.

[tcchinese]

趙元任粵語羅馬字

So, the method of using voiced consonants to represent the yeung tones in Cantonese has already been adopted by Yuen Ren Chao before. The following article describes the original Cantonese Romanization used by Chao, unaltered by Kenneth M. A. Barnett.

Below are the tables of initials.

Like Chao's Gwoyeu Romatzyh, this system uses tonal spelling (the tones are indicated in the spelling). Chao indicates yam tones with voiceless consonants, and yeung tones with voiced ones. The letter "h" has been used here in two senses: to mark aspiration (like in ph, bh and ch), or to imply "voicelessness" (like in mh, nh and lh). In Middle Chinese, the voiced initials are what gave rise to the yeung tones in modern Cantonese. The use of the voiced consonants may thus be considered an homage to that. While the sonorants have theoretically been always voiced, there exist sonorant initials in the yam tone, and these are marked with a following <h>. The exceptions are "i" and "u", which may be interpreted as medials without an initial (or a glottal stop initial), and thus does not carry a voiced initial, hence going into the yam category. While the aspiration mark in the yeung tones is not necessary, since yeung ping and yeung seung are always aspirated while yeung heui and yeung yap are never, it makes the representation more intuitive.

The vowels are a bit more complicated. Chao's analyses is mostly consistent with the later Penkyamp, if you are familiar with the system. The vowel pairs [a]-[ɐ], [ɛ]-[e~ɪ], [ɔ]-[o~ɵ~ʊ] are regarded as contrasting. In addition, [u] and [y] are represented the same way: by <u>, as they do not overlap in Chao's analysis, with the [u] considered as coming after labials (including m-, as in 抹, but not standalone "mu", a point not made clear in his book) or an implied/hidden/elided [w]. Yale yū is written as <iu> (initial i+u) while yīu is <iiu>.

[a] is written as <aa> except when alone, when it is written as <a>, which is used otherwise for the "short" [ɐ].

[ɛ] is normally <ea>, but <e> when alone. <e> is otherwise used for the short [e~ɪ]

[i] is <i>, while <u> represents both [u] and [y]. [u] is used after labials (except for "mu", which lacks a consonant coda) and labio-velars (which creates phonetically [k(ʰ)u], as [k(ʰ)wu] does not exist, considering the lack of phonemic distinction between [ji] and [i] and [wu] and [u]. See also my post where "kwui" is shown in Ball's Romanization). 

[ɔ] is <oa>, but <o> alone, which is also used to represent [o~ɵ~ʊ]. Yale's -eui is written as <oi>, while Yale's -oi is <oai>.

[œ] is written as <eu>, and only exists alone or before -k and -ng.

The syllabic nasals are written with a trailing <u>, which may be considered an homage to the Middle Chinese ngu (which becomes ng in modern Cantonese).

On the tones, the level tone uses the normal spelling convention.

The rising tone is written with a trailing x, unless the coda is -i, -u or -ng, in which case they are made "more open" to indicate the rising tone: -e, -o, -g.

The departing tone is written with a trailing h, unless the coda is a semivowel -i/-u or a nasal. The semivowels are written as their consonant letters -y/-w while the nasals are made "heavier": -mm, -nn, -q

For tone shifts, the spelling for the literary pronunciation is used, but an asterisk is added afterwards to indicate it. For example, even though 梨 is read as "lhee" after the tone shift, its literary pronunciation is "lei", and it is thus written as "lei*".

Also, the high level tone may specifically be clarified by adding a trailing <°>.

Chao's system was however, later drastically modified into what could be read as "Barnett-Chao", which you could read about on Wikipedia. In the new system, Barnett dropped the reference to Middle Chinese, and represented the initials in a way more consistent with Cantonese phonology (each initial is only represented one way) and follows the IPA for some of the initials and vowels, though it represents tones with extra silent letters <r> and <h>, making the resulting Romanization very hard to understand. The name of the system would be rendered as "Jrytjryr Romazri" after Barnett's modification, whereas Chao's original system would write it as "Yutyux Romadzih".

Comments