« previous post | next post »
IA sent me this article (in Chinese) about a new translation of George Orwell's 1984. It begins:
Yīngguó zuòjiā Qiáozhì Ōuwēiěr de míngzhù `1984' chūbǎn yuē 75 nián, jìnrì yíng lái shǒubù Táiwén bǎn. Yìzhě Zhōu Yíngchéng shuō, zhè shì tuīdòng `Táiyǔ zhèngchánghuà'de chángshì, ràng Táiyǔ mǔyǔzhě bùbì tòuguò Zhōngwén yìběn, yě néng jiēchù shìjiè jīngdiǎn wénxué
英國作家喬治‧歐威爾的名著「1984」出版約75年，近日迎來首部台文版。譯者周盈成說，這是推動「台語正常化」的嘗試，讓台語母語者不必透過中文譯本，也能接觸世界經典文學。
1984, a famous novel by British writer George Orwell, was published about 75 years ago and recently had its first Taiwanese version. Translator Zhou Yingcheng said that this is an attempt to promote the "normalization of Taiwanese" so that native Taiwanese speakers can access world classic literature without having to rely on Chinese translations.
IA points out that, as in the following quotation from the translator, "Zhōngwén 中文" (lit. "Chinese writing"), refers not only to written language but spoken as well:
Tā shuō:`Dāngshí zài guó wài jiǎng zhōngwén, chángcháng bèi dàng zuò zhōngguó rén, yúshì wǒ kāishǐ sīkǎo zìjǐ gēn táiwān de liánjié shì shénme, dé chū de jiélùn shì tái yǔ. Dàn wǒ tái yǔ bùgòu hǎo, yǒu shí wǒmen xiǎng jiǎng qiāoqiāohuà,(jiǎng zhōngwén) pà biérén tīng dǒng, jiù huì qiēhuàn chéng tái yǔ, dàn yòu méi bànfǎ wánzhěngde shuō
他說：「當時在國外講中文，常常被當作中國人，於是我開始思考自己跟台灣的連結是什麼，得出的結論是台語。但我台語不夠好，有時我們想講悄悄話，（講中文）怕別人聽懂，就會切換成台語，但又沒辦法完整地說」。
He said: "When I was speaking Chinese abroad, I was often mistaken for Chinese, so I began to think about what my connection with Taiwan was, and I concluded it was Taiwanese. But my Taiwanese is not good enough. Sometimes when we want to whisper, we are afraid that others will understand (what we are saying in Chinese), so we switch to Taiwanese, but we can't speak it completely."
IA continued:
This is not odd. I've heard high-schoolers also use '中文' in this sense (and/but without there being anything political about it).
How about elsewhere, outside of Taiwan?
I asked Aiong Taigi to comment on the Chinese article. He kindly replied:
I have always heard “Zhōng-wén” to refer to both the spoken and written language, although in Taiwan most common is still the Japanese “Kokugi”. I don't know if it's true but I always assumed the etymology as kokugi* > kok-gí > 國語 > guo2-yu3, which is fairly ironic. In any event, definitely among Taiwanese there is a trend towards using more politically unambiguous names like Tiong-kok-ōe or Pak-kiaⁿ-gí (the latter being more a reference to the government in Peking than the language Pekingese). I think that in any case, it proves the point that I always try to hammer home, which is that languages are fundamentally spoken things, not written. They're so fundamentally spoken that even the literal word “Written Chinese” can come to mean spoken Chinese. Rarely do we see literal words like “Guoyu” come to mean the written form. At least in my experience. Somehow, the idea of primacy of the spoken segues into my thoughts on the article, which are…
I am always happy to see new books published in Taiwanese. I just very often wish they would be written in Lomaji so people could actually read them *in Taiwanese*. Despite what they will tell you, almost nobody can read the Hanji version of a Taiwanese text without extensive (nearly 100%) reliance on their existing Mandarin knowledge. The mental model is a mashup of something like “Mandarin with little bits of Taiwanese sprinkled around for flavor”.** There is a near zero chance that most Taiwanese speakers (who are in truth mostly Mandarin natives) can prevent themselves from parsing the characters as Mandarin, any more than I could read an English text while translating on the fly to French and “not parsing” the English. It simply can't happen.
Also, I'm not sure if this is the longest translation, maybe it is, but it's a bit… presumptuous? to assert that the book「證明了將外文作品翻譯成台語是可行的」, since not only is this not the first foreign language book published in Taiwanese, it's not even the first Orwell. The 5% 台譯計畫 from the 1990s translated a dozen or more foreign works into Taiwanese, and published them, including Orwell's Animal Farm. My own publication of The Little Prince began it's life as an unpublished edition from that project.
Finally, not to rag on anyone's Taiwanese as mine sure isn't perfect, but I really really really wish that more *high quality* native speakers could get involved in these projects. People who come to the language later in life simply cannot capture the subtleties and “flow” as well as a true native can. Especially for those translations written in MOE characters for some reason, the grammar and word usage is generally… not as fluent as I would like. Among my circle of “über Taiwanese” friends, they almost invariably disregard books like this after one or two sentences in with a comment like “it's just Mandarin written with different characters”.
But please note that I have not read this particular translation, so I'm speaking more in generic terms than about this translation in particular.
——–
*kokugi 国技 ("national sport", i.e., "sumo") — both terms have been borrowed into English
**I believe this is true of virtually all so-called topolectal literature written in sinographs.
A'ióng's words are both profound and straightforward.
Afternote
Taiwanese Hokkien or Taiwanese Minnan (臺灣閩南語), also known as Taigi or Taiwanese, is a type of Hokkien language spoken in Taiwan. Many Taiwanese people who came from southern Fujian speak it. It's one of the major languages used in Taiwan as about 70% of Taiwanese people speak it.
(Wikipedia)
Selected readings
April 21, 2025 @ 5:43 pm
· Filed by Victor Mair under Alphabets, Topolects, Vernacular, Writing systems
Permalink
Jonathan Smith said,
April 21, 2025 @ 6:59 pm
Presumably intended was Japanese kokugo '[the] national language', not kokugi 'national sport'. And presumably this term+ideology ('National Language') got "calqued" across East Asia beginning from Japanese — but need not have entered Mandarin from Taiwanese specifically / only.
"Despite what they will tell you, almost nobody can read the Hanji version of a Taiwanese text without extensive (nearly 100%) reliance on their existing Mandarin knowledge."
>> This. To return to a thought from previous threads, (Taiwanese) Hanji/Hanlo texts are only fine in a vacuum~ideal world. Given present circumstances, it's gotta be Lomaji all day, every day. (With exposure to speaking from childhood via some approximation of immersion more important again by many-fold than reading/writing of any kind.)
"People who come to the language later in life simply cannot capture the subtleties and “flow” as well as a true native can."
>> A cruel irony: the people who love doing this work (as they see its value) are young adult *heritage* speakers; the people who should be doing this work (but don't necessarily perceive its value or have opportunities presented to them) are *life-long native* speakers, most of them elderly.
Jonathan Smith said,
April 21, 2025 @ 7:15 pm
Also re: "normalization", I've noticed this term is used often by advocates for Taiwanese, often explicitly along with the suggestion that the situation in other (Western + Japan!) countries with respect to a "national language" is "normal" and represents a standard towards which Taiwan may aspire. But the status quo in Taiwan is perfectly normal — Situation Normal All F**ked Up, as they say.
Chris Button said,
April 21, 2025 @ 7:25 pm
So is this a pun?
Just to check I'm getting it, is the suggestion that Japanese "kokugi" 国技 (national sport) is conflated with Japanese "kokugo" 國語 (national language) because "go" 語 is pronounced something like "gi" in Taiwanese?
Ted McClure said,
April 21, 2025 @ 7:34 pm
Would it be "better" (or even possible) to publish such translations as audiobooks rather than in print?
Victor Mair said,
April 21, 2025 @ 9:11 pm
I asked Aiong Taigi (!) what he meant by "kokugi", and this is his reply:
————-
Good call. There's kind of a lot I intend to convey there and it's probably lost without further explanation.
I mean: the Japanese refer to their language as the “national language”. When they came to Taiwan, the Taiwanese started referring to Japanese as the “national language” but in Taiwanese. And then still later, when the Chinese came, who didn't necessarily refer to their language as the “national language” in their own tongue, the Taiwanese repurposed the name “national language” in their own tongue to mean the new foreign language. All the while nobody has ever stopped to consider why Kok-gí wouldn't be a perfectly applicable moniker for Taiwanese, or how this very strange kind of name transfer is even possible.
As for the latter, I think it's only possible because of the confusion inherent to Kanji, and only happened by routing through the Kanji to get from kokugi to kok-gí and kok-gí to guo2-yu3. Kanji are like a quantum field where all things are possible, and you don't know what something means or even what language it is until you try to say it.
It would be like if modern Zulus called English by the name “African”, borrowing “Afrikaans” and transferring it from the Dutch to the English. Not an exact parallel, but I think it's the same idea.
Very strange.
Peter Cyrus said,
April 22, 2025 @ 4:30 am
When I first went to Taiwan in 1975, "Taiwanese" meant native Austronesian.
Chris Button said,
April 22, 2025 @ 6:35 am
"Taigi" representing 台語 "Taiwanese"?
KIRINPUTRA said,
April 22, 2025 @ 10:14 am
What is "topolectal" literature?
Allan from Iowa said,
April 22, 2025 @ 11:38 am
Does Lomaji refer to the romanization system that Wikipedia calls Tâi-lô ?
Victor Mair said,
April 22, 2025 @ 9:23 pm
@KIRINPUTRA
fāngyán 方言文學 ("topolect literature")
I will write a separate post on this tomorrow or the day after.
KIRINPUTRA said,
April 22, 2025 @ 10:55 pm
@ Victor
Looking forward, since this coinage ("topolect literature") seems to defeat the stated purpose of the coinage "topolect".
KIRINPUTRA said,
April 22, 2025 @ 10:59 pm
(Also, why do you think "The mental model is a mashup of something like 'Mandarin with little bits of [Cantonese] sprinkled around for flavor'" applies to Cantonese-language lit. even today — not to mention 2+ generations ago?)
KIRINPUTRA said,
April 22, 2025 @ 11:17 pm
@Allan from Iowa
Short answer: No.
In this context:
LOMAJI = Lô-má-jī = LATIN SCRIPT
= Tâi-oân Lô-má-jī = LATIN SCRIPT FOR Tâi-oân ōe
= Tâi Lô
These are generic terms; in historical & practical terms, what Wikipedia calls "Pe̍h-ōe-jī" IS Tâi-oân Lô-má-jī. (Confusingly, Nationalist China — or, rather, the spiritual Taiwan province under it — brand-named its new system "Tâi Lô"; that is reflected on Wikipedia.)
Victor Mair said,
April 23, 2025 @ 6:40 am
Yes, "topolect literature" seems to be a contradiction in terms — fāngyán wénxué 方言文學 ("topolect literature"). On the other hand, "oral literature" is a well-established concept / term in global scholarship, the idea being that this is written literature transcribed / derived from oral sources.
More to come.
Jonathan Smith said,
April 23, 2025 @ 10:09 pm
A couple pages are visible on Sanmin —
https://www.sanmin.com.tw/product/index/014090644
The translator (again 周盈成) clearly went hard. What a monumental undertaking; kudos.
For fun, p. 7 has "…你lóng無地知影敢當咧hŏng監視" i.e. POJ "…lí lóng bô-tè chai-iáⁿ kám tng teh hŏng kàm-sī" (original: "There was […] no way of knowing whether you were being watched") — which caused me to think I haven't seen such embedded questions in the older POJ literature. Instead we(=me) expect final (á) bô / bōe e.g. (finding examples via https://db.nmtl.gov.tw/) "Góa m̄ chai i ōe-hiáu-tit á bōe (Gûn Peng-ôe); "m̄ chai i chit sî tī im-kan kòe liáu ū hó–bô (Khan ang-î).
Maybe one of the ways the young(er)s are changing things up…
Re: topolect, the original conception was (1) calque "fangyan" and (2) be historical-comparatively neutral ~ democratic — but the thing is this and cognate terms in their original contexts (cf. Tw. hong-giân) are wielded to subjugate.
KIRINPUTRA said,
April 24, 2025 @ 9:07 pm
@ Jonathan Smith
That sentence is unwieldy & poss. ungrammatical, and may have natural Mandarin underlying it: 你都无法知道是否正在被监视. (HÔNG is 5th tone, not 9th. That's on the editors.)
For the past _ years, the translator — whose Mandarin is phenomenal — has been campaigning on social media for greater acceptance for Taioanese prose with "non-traditional", "modern" (largely Mandarin-fusion) grammatical structures.
Jonathan Smith said,
April 25, 2025 @ 3:35 pm
These grammatical changes are super interesting (including apparently to heritage speakers seeking a kind of "purity" of language…) Kiám~kám seems to begin rhetorical-ish and retains this sense in many modern contexts (e.g. kám án-ne? = Seriously!? I doubt it…), only relatively recently regarded (by some? wrongly?) as a straight-up interrogative particle — and even if plain interrogative, it feels odd/newfangled to use it in embedded questions. This (among much else) deserves a serious statistical look though…
Re: such changes generally, it seems we must grudgingly accept that to the extent Taiwanese survives into the next generation or two it will be massively relexified and restructured to align with Mandarin. The kinda weird thing is that even if such happened, the language could remain entirely independent and tied to its community in part because of the preference for Kanji-driven-calquing as opposed to properly phonological borrowing.
Good catch re: that tone — this item seems often to be written this way based on some (wrong) generalization that contractions require "Tone 9". Unless some folks actually pronounce it such…
KIRINPUTRA said,
April 25, 2025 @ 11:20 pm
@ Jonathan Smith
Have you come across any of these?
As it happens, the translator also flogs “purists” furiously. I’ve wondered whether such adversaries exist in reality.
I guess there are those who … agree that Taioanese, Hakka, & Mandarin (etc.) are distinct languages in theory, but recoil when the three are treated as distinct languages in practice.
KIRINPUTRA said,
April 25, 2025 @ 11:22 pm
+ I doubt kanji-based calquing is categorically more nutritious than “proper phonological borrowing”. More likely that’s a straight-up belief, derived from the Original Fantasy [of a subcontinent where folks talk different but write the same].
FWIW, no “preference for Kanji-driven-calquing” seems to exist in practice. If you hear “hôe-siu” more than “hŏe-so͘”, you’re in a bubble. A tiny, apologist one!
Jonathan Smith said,
April 26, 2025 @ 9:47 am
Re: "purity," I have no stance; I'm just gesturing at people's commonly-voiced preferences (or suggestions to others) to use a more 'authentic' pronunciation or construction rather than some other, to model writing on early as opposed to on contemporary prose, etc., etc. It makes sense to think about and discuss such issues. If someone is being stridently factional on the internet, then sure I agree may they "touch grass."
Same re: "kanji-based calquing": it's just an observable thing that occurs wrt Japanese in the early 20th century and continues in earnest wrt Mandarin to the present day. If you are claiming that phonetic borrowing has occurred on a comparable scale over the last few decades, then I would simply say I doubt it and someone should build a corpus and do the math.
KIRINPUTRA said,
April 26, 2025 @ 8:45 pm
Not sure if there be strident factions. There are those that *prescribe* treating Mandarin vocab. & structures as honorary Taioanese vocab. & structures, often in the name of “just being realistic / descriptive”…. You resemble them, right down to the inchoate “contemporary prose”, but they lack your embittered glee.
I’m not, although that doesn’t sound unreasonable. I do wonder if the facts justify the claimed “preference for Kanji-driven-calquing”, or that kanji-based calquing is more salubrious than phonetic borrowing.