Mixed script writing in Taiwan, part 2

« previous post | next post »

[This is a guest post by Kirinputra]

To take a step or two back towards the Sad Cripples theme, I had the TV on the other day and the show host — echoing the guest, a dietitian — said this:

“Lán Tâi-oân lâng kóng ‘tāu-leng’; gōa-kok lâng kóng ‘tó͘⁺-nài.’”

(Can't remember if he used Tone 3 or Tone 5 on the last syllable.) This could also be written like this:

“Lán Tâi-oân lâng kóng ‘tāu-leng’; gōa-kok lâng kóng ‘豆奶’ (DÒUNǍI).”

= 咱台灣人講「豆乳」。外國人講『豆奶』(DÒUNǍI)。

Again, he wasn't making a point; he was just summarizing an offhand remark the guest had just made to the same effect. While he seems to be referring to Mandarin speakers as foreigners — and they are, in a meaningful sense — there is no way he meant that. Rather, he & the guest were ultimately both referring to the English word soy milk, but calqued into Mandarin as 豆奶 — also a word in Mandarin, but not generally used in Formosa.

What's interesting is that even though the show was in Taioanese, it never occurred to them to calque soy milk into Taioanese, upon which it would be clear that soy milk and Taioanese 豆乳 line up exactly. The comparison had to be mediated through written Mandarin ("Zhōngwén"), which in turn could only be voiced via spoken Mandarin. (Contrast this with Hongkonger instincts, where such a comparison might still have to be mediated through written … Mandarin, which could then be voiced via Cantonese for a kind of seamless experience.)

At this point we're just looking at this everyday phenomenon where most Taioanese speakers instinctively feel that the Taioanese language is — along with spoken Mandarin — an appendage of written Mandarin, and that it is unfit for a long list of purposes, such as for making sense of words in English.

A closely related phenomenon is translating names into English — which is also how romanization is generally conceptualized in Formosan society. Taioanese speakers instinctively feel that this has to be done via Mandarin — or, rather, formal names (incl. of places, businesses, etc.) & most nicknames are instinctively thought of as being inherently in Zhōngwén; while they can be voiced in Taioanese in local interaction, translation into English instinctively has to be done through Mandarin voicing (spoken Mandarin).

(I'll have to find & send a photo I took this year of a mullet roe shop here in Takow. The shop has its name in Pe̍h-ōe-jī on its sign, under the sinographic name; but there's also a name in English, romanized via Mandarin, as if the Taioanese romanization was unfit for that purpose — which is exactly what most Taioanese speakers instinctively believe.)

There is a martabak shop nearby named "A Liong" after its owner, clearly an Indonesian A-Liông. I "joke" (but not really) that if the owner was a Taioanese A-Liông, with the invisible weight of Chinese modernity on his shoulders, he would've been compelled to name his shop "Ah Liang" or "Ah Lung" for Anglo-Roman consumption (I forget which sinograph is associated, if any).

The other day a Taioanese-literate, not-entirely-unsophisticated friend asked me about a sinograph, which he referred to as the 金 radical with "the khit-chia̍h sinograph." Khit-chia̍h 乞食 is the word for beggar. 釳 came to mind, but I knew nothing about it. Friend B understood that he was referring to 鈣. Soon it became clear that Friend A wanted to discuss words for calcium in Taioanese, and 鈣 was a false start (but again, not entirely). However, he had somehow been mentally constrained to start both the conversation & his orig. inquiry with the Zhōngwén kokuji 鈣*, which is only tangentially related to words for calcium in Taioanese; the general word for calcium is kha-lú-siù-m̀, via Japanese & English. I knew he must have assumed that 鈣 was an ancient sinograph "meaning" calcium, and I tried to explain that this was probably not the case (I hadn't looked it up yet, but I was right), but this overheated their brains and ended that part of the conversation. Since Chinese Taipei trains people from the dawn of their literacy to peg "non-foreign" meaning to Zhōngwén sinographs, it is almost prohibitively brainpower-intensive for them to think of sinographs themselves as dynamic moving parts; even the idea that sinographs can be divorced from Zhōngwén (or that any number of sinographs preceded Zhōngwén, or that Zhōngwén was an early-modern invention) can be hard to grasp, and even harder to incorporate into dynamic thinking.

[*VHM:  Zhōngwén 中文 is one of the many non-specific ways for saying "Chinese (language).  "In Japanese, Kokuji (国字, 'national characters') or Wasei kanji (和製漢字, 'Japanese-made kanji') are kanji created in Japan rather than borrowed from China.]

The most "everyday-exotic" exhibits for this nerve I'm getting at are shop signs where Zhùyīn** are given on the side for sinographs that are not Zhōngwén. They might be Japanese or native Taioanese, or they might playfully require the application of an unschooled Taioanese, non-Mandarin reading in order for the light bulb to go off. (And I'm not talking about the great number of shop signs that require a knowledge of Taioanese, Japanese, Cantonese, etc., combined with a Mando-Zhōngwén reading of the sinographs.)

[**VHM:  All school children in Taiwan learn to read and write with the aid of what is commonly referred to as "Bopomofo ㄅㄆㄇㄈ "), after the first four letters of this semisyllabary.  The system has many other names, including "Zhùyīn fúhào 注音符號" ("[Mandarin] Phonetic Symbols"), its current formal designation, as well as earlier names such as Guóyīn Zìmǔ 國音字母 ("Phonetic Alphabet of the National Language") and Zhùyīn Zìmǔ 註音字母 ( "Phonetic Alphabet" or "Annotated Phonetic Letters").

I haven't figured out what to do with such stuff. I think the widespread inability to grasp & grapple with this stuff is at — or close to — the root of why Taioan-Formosa remains in limbo as something like another Hong Kong (yet in some ways less than) rather than another Korea or Vietnam, or another Japan (to give a sinographic example). However, any short treatment of such a topic would tend to miss or fall short of the target, with sympathetic readers pouncing on shallow, shadow side-issues and missing the forest. Long treatments turn people off first by being too long, and secondly by making many perceptive (a good trait in itself) people feel deeply disgusted at seeing somebody try so hard to dig up nameless things from the primordial muck for no reason that they can think of (which might be the problem). But I figure you might know how to make meaningful presentations out of small bites of this material. And it gets us closer to the heart of the matter underlying Sad Cripples, which is that even ("non-foreign") speech is subconsciously felt to be derived from sinographs, which means non-sinographic, "non-foreign" writing (as well as speech-centered sinographic writing when it happens to be "non-conforming") is a copy of a copy, doubly corrupt(ed). (I'll send some photos when I can.)

Selected readings

 



29 Comments »

  1. David Marjanović said,

    May 29, 2024 @ 4:01 pm

    I wonder if this is more or less typical for diglossic situations. Standard German – Schriftsprache! – really is derived from its written form more than vice versa; dialects like mine are not written, and indeed cannot easily be written using Standard orthographic conventions because the sound system is too different; unlike in English, local forms of place names (let alone personal names) are thought to belong exclusively to the local dialect and may not even be known to speakers of mutually intelligible dialects, while the Standard form is closely bound to the spelling; nobody above age 6 is expected to learn a dialect; and the dialects, even individual words from them as far as known and at all practicable, are not used for communication with anyone who isn't speaking a comfortably mutually intelligible dialect.

    I hasten to add that that's about where the similarities to Taiwan end. For example, where I come from, there has never been a campaign or any other social pressure to make people speak Standard German in everyday life; Standard German contains a lot more regional variation than Standard Mandarin even between the mainland and Taiwan; and the last common ancestor between my dialect and the main contributors to Standard German was spoken sometime around the year 700, I guess, compared to maybe 200 for the last common ancestor of Taiwanese and any kind of Mandarin (or, by the way, 600 for Proto-Slavic).

  2. Jonathan Smith said,

    May 29, 2024 @ 5:52 pm

    I remember hearing the remark below from A-Hôa a yearish ago and thinking OK folks in Taiwan are clearer about these issues… but turns out not so much :D
    "Hàn-jī pún-sin kî-si̍t bô ì-sù, Hàn-jī ê sú-iōng sī chin chhìn-chhái ê tāi-chì."
    "Actually Chinese characters don't have intrinsic meaning; their application is a very arbitrary matter"

    A thought: arguably, the representation "gōa-kok lâng kóng 'tó͘⁺-nài'" is unfaithful to the cognitive reality (except if the single-quotes bit is trying to be a strictly phonetic rendering) — the sentence should instead be shown as a code-switch ("…kóng 'dòunǎi'") because there's practically no such thing as a phonetic loan from Mandarin into Taiwanese (among others). Instead, ALL Mandarin items are brought in by (what is felt to be) "reading" characters. The pool is bottomless; such "borrowings" can be and regularly are created on the nonce. Don't know how this compares to (e.g.) the Standard German/dialects situation.

    * BUT the reverse isn't totally true, presumably b/c Taiwanese words have less standardized/known relationships to written forms: so Tw. pa̍t-á 'guava' has been properly phonetically borrowed as Mnd. bā​lè (arbitrarily written 芭樂), thuh-tshàu as tù​cáo (arbitrarily written 吐槽~吐嘈), etc. But this is still rare as it's "better" to work from characters of course. (That is to say, it would be "better" if pa̍t-á had become bázǎi…)

    * If Taiwanese suddenly decided to phonetically borrow Mandarin, and speak of pe̍h-tsing or what not for Běi​jīng 北京 etc. etc., that would be freaking amazing and piss many off i.e. be amazing. Even write "白鍾"!! But not ever happening. (Japanese of course started really doing this for Chinese names and maybe other stuff [?], which I'm not sure Chinese people know how to feel about…)

    * the situation for Japanese > Taiwanese words is again different even given clear and widely-known characters. E.g., you are supposed to write "ba̋i-khín", not "黴菌" in Tw. This may be part of a Korean-writing-like mentality that characters should have regularish-feeling "colloquial" or "literary" "readings." IDK.

  3. Jonathan Smith said,

    May 29, 2024 @ 6:00 pm

    Forgot to mention that on several occasions I have heard people say (in Taiwanese) "In Mandarin we/they say X" where X is the Taiwanese-reading version of the written form of the item in question, NOT the Mandarin word. This point may relate more closely to the central point of the post :P

  4. Benjamin E. Orsatti said,

    May 30, 2024 @ 8:37 am

    David Marjanović said:

    dialects like mine are not written, and indeed cannot easily be written using Standard orthographic conventions because the sound system is too different

    Is åne Bairisch di Mutterspracha vui schenner?

  5. David Marjanović said,

    May 30, 2024 @ 3:52 pm

    Don't know how this compares to (e.g.) the Standard German/dialects situation.

    Both spontaneous unmodified loans and etymologically nativized loans (i.e. some of the regular sound correspondences are applied) are very common; of course there's nothing like the Sinitic tradition of having a system for pronouncing every character in pretty much every topolect.

    Is åne Bairisch di Mutterspracha vui schenner?

    …My own dialect is in the Bavarian group, but western east-central (Upper Austria), not west-central (most of Bavaria, e.g. Munich).* I can't identify your second word. If the last is meant to mean "more beautiful" (Standard schöner), it doesn't work in Munich either: it's /ʃɛnɐ/ with a short /n/ (consonant length is phonemic in almost all Bavarian dialects), so you can't use nn to distinguish /ɛ/ from /e/. The distribution of these two is completely unpredictable from Standard /ɛ/ vs. /eː/, so this would be a beautiful example of Standard spelling conventions just not working in Bavarian dialects.

    * West-central [ui] = east-central [ʏ] = Standard [ɪl] and [iːl] as in viel. MHG [ɪɫ], apparently.

  6. Jonathan Smith said,

    May 30, 2024 @ 7:43 pm

    FWIW it occurred to me that some terms for family members are indeed straight up phonetic borrowings from Mandarin of some period into Taiwanese… e.g., words for 'mom' and 'older brother', rather arbitrarily spelled "má-mah" vs. "ko–koh" at sutian.moe.edu.tw where both spellings achieve a Mandarin Tone 1 + "neutral tone"-like effect (totally unlike Taiwanese, where final stress is the law and reduplicated words can't have this tonal profile.) "Pá-pah" (sp?) 'dad' also exists but oddly I don't see it in this dictionary — it uses the 'mom' tonal profile as opposed to matching Mandarin Tone 4 + neutral. And "thài-thài" 'wife' (mentioned by me on LL recently :/), which for whatever reason does aim to match Mandarin 4 + neutral (Tw. Tone 3 weirdly being high falling in "sandhi" positions, low in "standing" positions.)

  7. Jonathan Smith said,

    May 30, 2024 @ 7:51 pm

    And this German/Chinese anecdote belongs here (paraphrasing):
    Chinese acquaintance: "When traveling in Germany, a local guy pointed out to me that [signage at some scenic spot] was in the standard language as opposed to in his 'mother tongue' — a language which had no written form. How odd!"
    Me: "Your are from as Hakka-speaking household thus your situation is rather similar no?"
    Chinese acquaintance: *brain overheating symptoms* :D

  8. KIRINPUTRA said,

    May 31, 2024 @ 12:51 am

    @ Victor

    I was able to track down the two photos I had in mind.

    A quick note: By "Zhōngwén" I mean literally (& objectively) the national language of modern China (roughly, the two republics of China) in its written aspect, as it is objectively constructed (not as it is subjectively imagined within its nation). By "Zhōngwén kokuji" I mean sinographs coined within — or (as a practical, chronological matter) very nearly within — the modern construct of Zhōngwén; in other (less precise) words, non-traditional, non-universal, PRC/ROC-only sinographs.

    Chiàⁿ Bī Tin — (link to image)

    This is a mullet roe shop in one of Takow's (Táⁿ-káu 打狗) port districts. The sinographic form of the name of the shop is adorned with Pe̍h-ōe-jī furigana. Pe̍h-ōe-jī on signs is uncommon anywhere in Formosa, and vanishingly rare in commercial contexts. The English name is nevertheless given as "Cheng Wei Chen", via Mandarin. Notice that "Cheng Wei Chen" is presented as writing while "Chiàⁿ Bī Tin" serves as furigana. The sign reveals that the sinographs 正味珍 are central in the minds of the owners; the Taioanese & Mandarin "readings" of this name are peripheral, while the Mandarin reading was (subconsciously, of course) deemed more appropriate for translation to English.

    Martabak Manis A Liong — (link to image)

    This is a dessert shop in a modest Indonesian- & Vietnamese-oriented commercial district fronting the main Takow train station. "Martabak Special Bangka" is a stiff translation of Indonesian "martabak khas Bangka"; Bangka is an island off Sumatra with a large number of mixed-descent Chinese-Malay people speaking what has been described as a Hakka creole. To the continental eye, "A Liong" suggests a Teochew or Hokkien background — a possibility, but Bangka Hakka might be the best bet. The sign reveals that the name "A Liong" in Indonesian or some language local to Bangka is central in the minds of the owners; the sinographic form 阿龍 is peripheral; and, as is the custom in Latin-lettered lands, "A Liong" in written form doubles as the English name of the shop if that should be necessary. Unlike with local shop signs (but unremarkably for an Indonesian shop sign, including in Formosa), no Mandarin romanization is given for the sinographic form, nor is the "English" form based on Mandarin romanization. However, since both sinographs & "English" are provided, the sign is "complete" (although exotic) by local standards.

  9. KIRINPUTRA said,

    May 31, 2024 @ 4:41 am

    @ David Marjanović

    Thoughtfully said. A decent if obscure comparison would be to romanized Taioanese (& Hokkien). A few dialects of Taioanese (& many of Hokkien) are not covered by the (unofficial) standard Hokkien-Taioanese romanized script, and local place names can be a mystery even to romanization-literate people from elsewhere. Penang (in Malaysia) is the best example that comes to mind at the moment.

    The current relationship between Taioanese & Mandarin clearly fits some definition of “diglossia”, but not a narrow one. The abstand distance is much, much greater than commonly advertised. (Also, it seems far from certain that Taioanese & Mandarin share a common ancestor at all. But, assuming Taioanese & Hokkien share ancestry with the so-called Eastern Min languages Hokchiu 福州 & Hokchia 福清, the last common ancestor of all these may have been spoken during the late first millennium AD (?). (Jonathan Smith might have more insight, if he sees this.)

    There are phenomena related to 20th century Taioanese-Japanese diglossia that come to mind that support your point that at least some of the phenomena described in my post are general to diglossic setups. (I'm thinking of mid century romanized Taioanese texts where Western names are translated into Japanese rather than Taioanese.)

  10. KIRINPUTRA said,

    May 31, 2024 @ 4:54 am

    @ Jonathan Smith

    > A thought: arguably, the representation "gōa-kok lâng kóng 'tó͘⁺-nài'" is unfaithful to the cognitive reality — the sentence should instead be shown as a code-switch

    You’re right! I’d edit if I could. (For most POJ Taioanese-literate audiences, the unfaithful format arguably works better than the more faithful alternatives for the time being. But this audience is not those.)

    > Instead, ALL Mandarin items are brought in by (what is felt to be) "reading" characters.

    > there's practically no such thing as a phonetic loan from Mandarin into Taiwanese

    Most. Not all. And this is “academic” in part: The line between phonetic loans from Mand. & code switches is ill-defined for now, and “academics” would tend to have to resolve in favor of code switching for the time being. The same was true of phonetic loans from Japanese before approx. 1980. And a whole Japanese-free cohort had to inject itself into society before we could confirm that the mid-level tone in the first syllable of BAT⁺-TÉ-LIH ("battery", via バッテリ) was tonemic.

    For now, a solid everyday example (down-island, at least) is SEH-SEH for “thanks”, from 谢谢. This word is avoided by people who don’t want to be perceived as Mandarin-challenged, which means you hear it a lot less in the urban North. I’m pretty sure it’s used a lot by folks that speak no Mandarin.

    Question particle *MĀ and final particle *PAH are even more common, but they’re arguably code switches. Final particle MÀ seems to be a solid loan. (I see you’re onto the kin terms.)

    There are a host of partly phonetic loans that your “(what is felt to be)” clause might cover, but consider -KHAH (“card”). One might assume this was brought in via sinograph, but 卡 had no non-obscure reading before the Neo-Chinese takeover (and the obscure reading was CHA̍P). In essence, a “non-Sinitic” Mandarin word was borrowed together with its “dialect” sinograph, obscuring the phonetic aspect of the loan.

    Going back in time (to a time when “Taioanese” as such may not have existed), look at HONG-CHÀ. Look at CHAU-THAT or KAU-LIÂNG or IÀN-KHÌ or LO-SO. If nothing else, the verb CHOÁN should disabuse us of the notion that phonetic loans from Mandarin don’t happen.

    > Even write "白鍾"!! But not ever happening.

    Just like the Japanese occupation was never going to end. (It’s just a matter of education and, indirectly, navies, armies, etc.)

    > the situation for Japanese > Taiwanese words is again different even given clear and widely-known characters. E.g., you are supposed to write "ba̋i-khín", not "黴菌" in Tw.

    Sounds like something a Taiwanese Literature major might say. In the wild, we’re slightly more likely to be told that BĂI-KHÍN should only ever be written as 黴菌. From a sapient Taioanese-centered perspective, BĂI-KHÍN is much preferable to 黴菌, but the same logic would have to apply to loans from Mandarin. (Of course, the working class thinks of all phonetic loans as code switches, and the literati mostly have yet to realize there is such a thing as loans between Taioanese & Mandarin.)

    > on several occasions I have heard people say (in Taiwanese) "In Mandarin we/they say X" where X is the Taiwanese-reading version of the written form of the item in question, NOT the Mandarin word.

    Not unusual. Also happens with Japanese words.

  11. KIRINPUTRA said,

    May 31, 2024 @ 5:11 am

    @ Jonathan Smith

    ŌE-CHONG is another interesting subclass (?) of modern loan from Mandarin. ŌE is clearly not any kind of reading of 化. Rather, ŌE-CHONG comes to us via the spoken form of 化妆 (HUÀZHUĀNG), which some critical mass of Taioanese speakers had internalized w/o reference to the actual sinographs. So Mandarin loans can be neither sinographic nor purely phonetic. (Notice how the Chinese Taipei Ministry of Educ. dictionary pretends this word doesn't exist.)

  12. Jonathan Smith said,

    May 31, 2024 @ 7:58 am

    Very interesting; yes seh-seh is of course a great example. Re: "Going back in time" yeah this is a crucial consideration — it is clear that my characterization is less apt the further back and the further from near-universal-literacy one goes.

    An interesting theoretical question is to what extent borrowings that one might tend to think of in the normal ("character reading") terms can in fact be regarded as regular old calques: i.e., absent literacy, to what extent are words like ha̍k-hāu / tông-ha̍k~tông-o̍h (old people: "tóng​xué") / lāu-su~nóo-su (old people: "lǎo​sī") and all the rest generated just the same. It would seem that they are still generated, but perhaps not "just the same": the process is mediated/regulated/expediated by literacy. What say you?

    Your last class could be seen as a kind of "failed" case of such calque generation. I have heard similar in media… "king-tsè-lâng" given homophony of 'manage' and 'economics' in Mandarin is one that comes to mind.

    A related question is the extent to which names, which you mentioned in the post, might be regarded as calques as opposed to "character readings." I.e., my name is Lovely, thus Tw. bí, Mnd. Měi. But this seems overly hopeful… the arguably weird notion that written forms ARE the names and that they are considered to be READ in different ways appears more successful here. Names before/absent literacy should (have been?) investigated…

  13. Jonathan Smith said,

    May 31, 2024 @ 8:20 am

    "Final particle MÀ"

    do you refer to whiny/plaintive/coy ma, as in Mand. "pleeez give me a little more maa"? But this is so Taiwanese Mandarin… I was thinking it orginated in Taiwanese. Late 19th cent. Peh-oe-ji texts write final "mah" with some frequency in very particular contexts… something like "come now, surely it's the case that such-and-such mah."

  14. KIRINPUTRA said,

    May 31, 2024 @ 11:06 am

    FINAL MÀ & MAH

    MAH seems to have been translationese, via Mandarin. Typically used in rhetorical questions.

    MÀ must be from Mandarin, but — who knows? — maybe it came from non-standard Mandarin or non-native Mandarin.

    > my characterization is less apt the further back and the further from near-universal-literacy one goes.

    A key matter not to be overlooked is that book-borrowing only operates via universal sinographs, from the koine. It's not even supposed to work with regional or "dialect" sinographs, whether or not they can be found in the ancient rimebooks. A more concrete way to put this is that 檔, for example, does not have a genuine Taioanese (or Hokkien) reading; nor does Zhōngwén 拼 (not to be confused with the native sinograph 拼, for PIÀᴺ). The "gap" (as a Chinese nationalist might see it) can be strong-armed shut if the hard power is there, but literacy's got nothing to do with it.

    SINOGRAPHIC BORROWING AS CALQUING

    Right. I think it is a kind of calquing.

    > i.e., absent literacy, to what extent are words like ha̍k-hāu / tông-ha̍k~tông-o̍h (old people: "tóng​xué") / lāu-su~nóo-su (old people: "lǎo​sī") and all the rest generated just the same. It would seem that they are still generated, but perhaps not "just the same": the process is mediated/regulated/expediated by literacy.

    Not sure what you mean exactly, but calquing has def. been a way of life. A good share of the Formosan Hakka lexicon evidently wasn't inherited from continental Hakka; many if not most of these words seem to be local loans or calques from Hokkien (at a time when "Taioanese" didn't exist as such). An interesting "failed" calque is CHÒNG-SHÀN 長山, from Hokkien TN̂G-SOAᴺ 唐山.

    A general theme is that modern thinkers underestimate the degree & impact of literacy in pre-modern Formosan society. The many sinographically-induced mutations of place names in pre-modern Formosa testify to a widespread if half-assed literacy. Pre-modern Formosa was not as illiterate as some people want to believe.

    Meanwhile, a word like HA̍K-HĀU reached the masses via the literati, regardless of whether we want to call it a calque.

    Some of the interesting phenomena we see with post-WW2 Mandarin loans in Taioanese are due to the wedge (in non-low contexts) that the Nationalist Chinese state & media were able to drive between Taioanese on one hand & sinographs on the other. Put simply, Taioanese-speaking society became less Taioanese-literate than it had ever been while becoming more literate than it had ever been. Paradoxically, this has curbed the intake of sinographic loans from Mandarin (or any other source) and slowed the erosion of native vocab. & expressions….

  15. David Marjanović said,

    May 31, 2024 @ 2:54 pm

    it seems far from certain that Taioanese & Mandarin share a common ancestor at all

    Of course they do, they're all Sinitic, one branch of Sino-Tibetan = Transhimalayan. But of course you're nonetheless right that the "distance is much, much greater than commonly advertised".

    Some of the interesting phenomena we see with post-WW2 Mandarin loans in Taioanese are due to the wedge (in non-low contexts) that the Nationalist Chinese state & media were able to drive between Taioanese on one hand & sinographs on the other.

    *lightbulb moment*

  16. KIRINPUTRA said,

    June 1, 2024 @ 9:05 am

    > Of course they do, they're all Sinitic….

    Wow. If you're able to share or cite the evidentiary basis for this claim (beyond just "so many experts couldn't be wrong"), I'd love to see it. (I.e., proof that Hokkien, Teochew, etc. are conventionally descended from one of Mandarin's ancestor languages instead of from, say, a first millennium Chinese-lexified creole.)

  17. Jonathan Smith said,

    June 1, 2024 @ 10:43 am

    @KIRINPUTRA

    My interest is (partly) in thinking about to what extent universal phenomena do or might in theory underlie processes hitherto accounted for in "folk" terms — so in this case, many words having entered Taiwanese relatively recently which people talk about in terms of "reading" written forms (i.e., one is instructed to "chiàu jī tha̍k" and the like) *could in principle* be accounted for in terms of calquing.

    To be clear re: calque (since I do some reading in language pedagogy where the term is used shall we say casually), I mean structure-borrowing, where knowledge of a Language A "metaphor" allows the (sufficiently) bilingual individual to rebuild it in Language B. Such a process *could in principle* account for many electric+brain = computer type borrowings within Chinese, such that bi/multilingualism would be the special sauce and literacy totally irrelevant.

    Here your oē-chong category, which might be more fun to call whiffed calques, is actually super interesting as it is proof positive that people are sometimes (often?) working in exactly this way.

    However, the above is insufficent for tons of cases. For 'blood pressure', in a vaccuum, we would expect a God's Honest Calque TM to use a Tw. verb that really means 'press', like the Mandarin version (if such was the immediate source.) Or in 'same study (= classmate)', one would expect the regular word for 'same'. And so on. So even if newish items like these aren't composed of "readings" per se, one has to at least grant that they are Zhongwen-literacy-mediated, i.e. "mediated/regulated/expedi(a?)ted by literacy."

    Like I said kinda in an earlier thread, it is a bit tough to gauge how disinterested your assessments of certain issues are (OK true disinterest is perhaps a scientific fantasy but…) So above, (1) re: "Sinitic," as I pointed out in the other thread, *all* of Min is convincingly united by intricate phonological correspondences across basic vocabulary, not just Hokkien, some items being peculiar to Min. Perhaps we can agree to disagree for now about the significance of the fact that most of these items are further shared with the rest of Chinese and stand in "purdy regular" correspondence thereto. (2) re: premodern literacy in Formosa (and adjoining regions), the topic is of tremendous interest and really deserves some extended academic discussion/presentation — it depends how half-assed you mean perhaps, but the impression one gets from late 19th century sources is very much of hyper Zhongwen literacy in highly specialized circles springing small leaks out into broader communities only in specific, limited contexts. (3) re your MAH (rhetorical questions) yes that's the one — this does not resemble Mandarin and I bet it is the source of Tw. Mandarin 'I (whiningly) entreat you' particle ma… literally bet some martabak, preferrably :D While cold comfort from a Taiwanese advocacy POV, Tw. Mandarin is chock full of Taiwanese blood… after all it largely reflects Mandarin as spoken as a second language post WWII and reminds one what terms like "substrate" mean IRL.

  18. KIRINPUTRA said,

    June 2, 2024 @ 11:43 am

    [For future reference, the following merges a discussion that began at

    https://languagelog.ldc.upenn.edu/nll/?p=64045#comments

    ….]

    @ Jonathan Smith

    CALQUES

    > many words having entered Taiwanese relatively recently which people talk about in terms of "reading" written forms … *could in principle* be accounted for in terms of calquing

    Of course!

    This is a tricky but poss. rewarding area of study. There are lingering prescriptions (“one is instructed …”) & secret filters shrouding this stuff. (For one thing, most scholars do not see loans between Taioanese & Mandarin as loanwords in the universal sense.) One will want to note these factors w/o letting them skew one's process & output.

    > For 'blood pressure', in a vaccuum, we would expect …

    To me, it’s clearly a question of the actual mechanics & sources of the borrowing. (But you might have to come to the dark side to access this clarity.)

    Some X factors to keep in mind, not exhaustive:

    — Unprescribed kunyomi readings concretely at work in society.

    — A fuzzy system of sound-to-sound conversions that operate independently — at least in part — of both meaning (words) & sinographs. (This system, applied on the fly, plays a key role in the mechanics of how most Taioanese speakers under 60+ actually read sinographs into Taioanese. Hence a brand name like 燒肉衆.)

    > it is a bit tough to gauge how disinterested your assessments of certain issues are

    This doubt seems prejudicial, but kudos to you for voicing it!

    See, FWIW, I sleepwalked into this field, mesmerized by tables of “character readings” (often spun as lexical cognates) for county seats across the empire. I informed my professor in a seminar on language contact that there were no grammatical contact phenomena to speak of in (or right around) the Sinitic zone b/c all varieties were grammatically uniform (if not formless). Reality has obviously changed my views since then. Frankly, I’ve probably drunk way more from this cup than most.

    1 (your numbering) —

    > *all* of Min is convincingly united by …

    Again, the oneness of “Min” is not at issue.

    > most of these items are further shared with the rest of Chinese and stand in "purdy regular" correspondence thereto.

    Not at issue either. However, a broad palette of shared basic lexicon & “phonological correspondences” is (or do you implicitly deny this?) consistent with either shared ancestry or formative contact (such as if “Min” were descended from a Chinese-lexified creole). Details aside, it doesn’t exclude either possibility. This is straight logic.

    Even as far as basic lexicon goes, getting into the details a bit: You’re ignoring the view from the “Min” side. To use Teochew examples again, “person” is non-“Sinitic” NÂNG. “Father” is non-“Sinitic” PẼ. (These etyma are pan-“Min”, right?) If this is a family of “Sinitic” languages with non-“Sinitic” words for “father” (arg. the weightiest word in Sinospheric civilization) & “person”, where is the riveting narrative of how this came to be?

    Then there’s syntax. Was Hainanese conventionally spawned by continental “Southern Min”, or is it in effect a “Southern Min”-based creole the way Papiamento is an Ibero-Romance-based creole? Lexical & phonological comparison are necessary but not sufficient; syntactic & semi-syntactic patterns (like run-stand tone patterns) have to be examined too. Why the conviction that syntax should be waived (?) in the case of “Min” vs “Sinitic”?

    2 —

    Literacy in Zhōngwén — the Neo-Chinese national language centered on Mandarin — reached Formosa in the 1920s. (As for high literacy in the book koine in the late 1800s, Formosa must have lagged behind “Southern Min”-speaking regions of the continent, which in turn lagged behind much of the Sinosphere, incl. the Red River region, evidently. But literacy at the level of the trader or seafarer was not as rare as some people let on.)

    3 —

    I can taste it already. :) The two particles are incontiguous in every way. (I trust your instinct that this is not a Standard Mandarin word, though.)

  19. David Marjanović said,

    June 2, 2024 @ 3:00 pm

    Wow. If you're able to share or cite the evidentiary basis for this claim (beyond just "so many experts couldn't be wrong"), I'd love to see it. (I.e., proof that Hokkien, Teochew, etc. are conventionally descended from one of Mandarin's ancestor languages instead of from, say, a first millennium Chinese-lexified creole.)

    1. "Proof"?!? This is science we're talking about, not mathematics. The best we can do is a parsimony argument.

    2. So… what evidence do you have that such a creole ever existed? The most parsimonious hypothesis is that it didn't – that we're looking at ordinary descent of Proto-Min from "Late Hàn Chinese", substrate influences notwithstanding.

    3. As a matter of terminology… creoles are very much descended from their "lexifiers". Sinitic creoles are Sinitic, Romance creoles are Romance, Germanic creoles are Germanic.

    Even as far as basic lexicon goes, getting into the details a bit: You’re ignoring the view from the “Min” side. To use Teochew examples again, “person” is non-“Sinitic” NÂNG. “Father” is non-“Sinitic” PẼ. (These etyma are pan-“Min”, right?) If this is a family of “Sinitic” languages with non-“Sinitic” words for “father” (arg. the weightiest word in Sinospheric civilization) & “person”, where is the riveting narrative of how this came to be?

    The East Slavic languages don't even have the same word for "father" (and they don't have the same word for "dad" either). And "person"… two hundred years ago, the ordinary word for "person" in English was man. Yet, English preserves the Proto-Indo-European word for "father" practically unchanged except for regular sound changes, while Proto-Slavic had already replaced it, as has e.g. Welsh, as had e.g. Gothic almost completely, even though "father" has arguably been the weightiest word in Indo-European-speaking cultures for most of the last six thousand years.

    As I keep saying in the context of biology: you can't do phylogenetics with one or two characters. In historical linguistics, the same holds.

  20. KIRINPUTRA said,

    June 4, 2024 @ 1:51 am

    @ David Marjanović

    Thank you. BTW, how do you format quotes that way?

    > YOU CAN’T DO PHYLOGENETICS WITH ONE OR TWO ETYMA

    (Etyma, right?) Of course. (Surely you didn’t think I….) There are many more where NÂNG & PẼ came from. What do you think of CHŌI (“many”)? Or CHE̍K (“one”)? At some point — the Swadesh lists are only a couple hundred items — the exercise ceases to be the slam dunk you might have envisioned. To put this in context — and correct me if I err — “Min is Sinitic” is based entirely on core lexicon, and phonological correspondences therein, in the first place.

    1. Point taken. (Thank you.)

    2. Onus probandi incumbit ei qui dicit, non ei qui negat. One need not prove Hypothesis B to debunk — let alone question — Hypothesis A.

    A parsimonious hypothesis is one that makes the least assumptions. Does “Min is fully descended from an ancestor of Mandarin, and v.v.” make the least assumptions? I don’t see that it does at all; what’s your reasoning? If you don’t have direct knowledge of “Min” & “Sinitic”, which actual work(s) convinced you?

    “Min as Sinitic” does “minimize the number of language families”, but that’s clearly not what we mean by “parsimony”.

    3. > Sinitic creoles are Sinitic, Romance creoles are Romance, Germanic creoles are Germanic.

    So syntax, morphology, etc. are less relevant to “descent” than lexicon? That’s a valid … opinion.

    As usage reveals, neither linguists nor the modern world at large consider Romance (and so on) creoles to be full-fledged Romance (and so on) languages anyway. Google “Asturian is a Romance _____” (but don't include the blank), and then Google “Papiamento is a Romance” or “Chavacano is a Romance”. Walk through a forest of Romance family trees; how many have Chavacano or Papiamento hanging off them?

  21. Philip Taylor said,

    June 4, 2024 @ 5:42 am

    Totally unqualified to enter into this debate, Kirinputra, but I can say that the block quotations that David M. (and others, including myself) use are delimited by <blockquote></blockquote>.

  22. Jonathan Smith said,

    June 4, 2024 @ 8:13 am

    @KIRINPUTRA

    I think 'person' as in Tw. lang5 unites only Coastal Min… while correspondences for 'one' are irregular. But other distinctively Min items ('many' [Tw. che7], 'house' [chhu3], 'foot' [kha1], 'mouth' [chhui3], 'daughter-in-law' [sin1-pu7], ‘rice cake' [kue2], 'don’t have' [bo5]…) show regular phonological correspondences across the group: thus genetic "Min", unity of which I thought you had expressed doubts about earlier. I note again that no contemporary interest faction is positively disposed towards this construct… so maybe it’s just science.

    OTOH all such factions seem either to be very positively or very negatively disposed towards the idea of unified Chinese, so good reason to be cautious. From a scientific point of view, though, that Min is part of Sinitic is just the simplest way of accounting for the fact that the large majority of what looks like the "Proto-Min" lexicon is not Min-specific but is shared with Mandarin et. al. Yeah certain details re: correspondences are a work in progress, and the notion certainly doesn’t call for religious – more literally ethnic chauvinist – commitment. Instead, regard this hypothesis as operational: it motivates and orients work that could strengthen the idea… or conversely, lead to its abandonment in favor of a neutral alternative of the kind you favor (creole / mixed language / other…) — the kind of position motivated not by positive evidence but by the absence or insufficiency of such.

    Syntax is of course interesting, but yeah less interesting in relation to the above hypothesis… we watch Taiwanese, etc., become syntactically more like Mandarin on a literally daily basis, and parallel convergences have always been happening. Morphology is of course also interesting and could in principle speak directly to the question of descent, but it's rather scarce in this area… there are ideas floating around that Min reflects remnants of derivational affixes lost in other Chinese, but this looks doubtful to me.

    Re: mah I find these items interestingly similar – but absent proof positive, it’s my treat on principle of course :D

  23. David Marjanović said,

    June 4, 2024 @ 9:10 am

    Oh, I'm very sorry! I used "character" in the biological sense and completely overlooked that it might be mistaken as "sinograph".

    In that sense, "character" refers to a, well, feature that occurs in at least two "states"; this can be a feature of anatomy (…I almost wrote "morphology"…), heritable behavior, or a position in a DNA/RNA/protein sequence, for example. The analogs in linguistics could be etyma as well as phonetic or grammatical features.

    There are many more where NÂNG & PẼ came from. What do you think of CHŌI (“many”)? Or CHE̍K (“one”)?

    These are not terribly stable meanings elsewhere. Many, and much for that matter, have nothing to do with German viel(-); the first must be related in some way to German manch- "some", but there's a whole suffix there that doesn't make sense in German (or English). For "one", Germanic, Celtic, Italic, Baltic and arguably Albanian have what looks like PIE *(h)óih-no-, but Indic goes for *(h)óih-ko-, Iranic for *(h)óih-wo-, Slavic has a mysterious *edīn- (meaning "sole, only" in addition to "one"), which I've seen explained as *(h)e-dʰé (h)ih-nó- ~ "this one here" in a rather complex paper; Greek goes for a completely different root, *sem-, which appears in Latin semel "once", while keeping *(h)óih-wo- for "alone" and *(h)óih-no-, if at all, specifically for the ace on dice.

    To put this in context — and correct me if I err — “Min is Sinitic” is based entirely on core lexicon, and phonological correspondences therein, in the first place.

    Actually, I don't know; given that there isn't much morphology and that syntax is rather labile, it may actually be so. But that would make it impossible, as far as I can see, to distinguish "Min developed from 'Old Chinese' in the usual way" from "Min developed from a Sinitic pidgin that creolized", except by parsimony as follows:

    A parsimonious hypothesis is one that makes the least assumptions. Does “Min is fully descended from an ancestor of Mandarin, and v.v.” make the least assumptions?

    Yes, because it doesn't need to assume a pidgin and a creole. Pidgins develop under pretty specific circumstances that are not terribly common, and creoles develop from pidgins under similarly specific circumstances that aren't all that common either, so, as far as I can see, you should propose a fairly specific historical scenarion in which this could have happened.

  24. David Marjanović said,

    June 4, 2024 @ 9:11 am

    extra -n on scenario not intended

  25. David Marjanović said,

    June 4, 2024 @ 9:16 am

    a neutral alternative of the kind you favor (creole / mixed language / other…)

    If by "mixed language" you mean something like Michif or Media Lengua or Mednyy Aleut, those are extremely rare, much more so than creoles.

    If you just mean "a language with lots of influence from other languages", like English, sure, go for it; I'd love to see a study of non-Sinitic influences on Min.

  26. Jonathan Smith said,

    June 4, 2024 @ 9:53 am

    @David Marjanović

    I guess I mean "whatever might make sense within the hypothetical scenario envisioned by KIRINPURTA or whomever" — "bilingual mixed language" (Sally Thomason's term?) or other. Ye much more remains to be done with the plentiful non-Chinese material in Min both at the proto- ane lower levels…

    More generally it seems to me that MSEA inclusive of Sinitic typology presents theoretical problems here: if one views phonological correspondences in core lexicon at scale alone as insufficient to distinguish creole/lexifier situations from "conventional" transmission, one is going to have issues not only re: Min vis-a-vis other Chinese but re: pretty much everything…

  27. KIRINPUTRA said,

    June 6, 2024 @ 1:30 pm

    @ Philip Taylor

    Thank you!

  28. KIRINPUTRA said,

    June 6, 2024 @ 1:45 pm

    @ David Marjanović

    I used "character" in the biological sense

    I see! That crossed my mind. (And I should’ve given more weight to that thought.)

    Yes, because it doesn't need to assume a pidgin and a creole.

    Just to be clear, I don’t (assume that).

    Moreover, I’m not sure an intervening creole or mixed language is the sole scenario where “Min as Sinitic” would prove false. Again, onus probandi….

    Still, an intervening 1st millennium Sino-creole seems very possible. (This is a modest claim, but surely sufficient to question the implicit blanket assumption that such a creole could not have existed.) As shown in “The Chinese Colonization of Fukien Until the End of T’ang” (Bielenstein), the Chinese presence on the coast of what became “Fujian” 福建 was concentrated for centuries in one maritime colony located near modern Hók-ciŭ 福州; this colony was a (or the) nexus from which Chinese power spread through the region.

    As for core lexicon, the examples I’ve given are plenty stable across “Coastal Min”. I’m working off the Swadesh lists; do you propose an alternative list (or paradigm)? (On another tangent: What do you make of the shallow distribution in “Coastal Min” lexicon — in its southern reaches, at least — of reflexes for the pan-“Sinitic” roots often represented as 人 and 多 and 一?)

    “Min as Sinitic” is largely built (1) on the notion that “Min” is lexically “Sinitic” at heart, and (2) on the presumption that core lexicon stays put. But is “Min” actually lexically “Sinitic” at heart? One must stand & deliver at some point, or show where that has been done by others. (Or if quantity is the issue, we could continue with KIÁᴺ 子 “son; child” and KHA 脚 “foot”.) Alternatively, if core lexicon is permeable, how much is left of “Min as Sinitic”?

  29. KIRINPUTRA said,

    June 6, 2024 @ 2:07 pm

    @ Jonathan Smith

    Well said, on many counts.

    creole / mixed language / other…

    As for “other”, I wonder if archaic “Sinitic” loans in “Min” have ever been addressed (i.e., teased apart from non-loans). Even if “Min” is simply “Sinitic” with non-“Sinitic” substrate, that substrate would have borrowed “Sinitic” elements in it, no less than a Tai-Kadai or a Hmong-Mien.)

    if one views phonological correspondences in core lexicon at scale alone as insufficient to distinguish creole/lexifier situations from "conventional" transmission, one is going to have issues

    Now we’re really talking. We can’t just not face the dragon, though. And I think we’re far from unarmed for the task.

    thus genetic "Min", unity of which I thought you had expressed doubts about earlier

    I tend to agree that (at least, “core”) “Coastal Min” is a conventional family. As for “Inland Min”, I’ve heard it said by “Sinitic” — if not “Min” — specialists that the two “Mins” don’t reconstruct to an exclusive common ancestor. But I plead no contest pro tempore on this.

    no contemporary interest faction is positively disposed towards this construct… so maybe it’s just science.

    Yes. (Now, what if there had happened to be an interest faction in that space? I hate to think….)

    that Min is part of Sinitic is just the simplest way of

    regard this hypothesis as operational: it motivates and orients work that

    Maybe, at one point. There’s a lingering question of bias, and questions of lingering bias. Most practitioners seem hostile to the idea of alternatives, but “Sinology” has clearly bonsaied & isolated “Minology” in ways that are exhausting to list. “Min as Sinitic” may have facilitated the study of “Min” from certain angles, but those angles are long done…. In light of the progress being made in, say, Mienology, “Minology” is stagnant on this front, and not b/c of this or that scholar, but for paradigmatic & institutional reasons….

    less interesting in relation to the above hypothesis… parallel convergences have always been happening

    Surely true to some extent, but — is syntax really exceptionally uncorrelated to linguistic descent in MSEA+? Has this been “scientifically” assessed?

    Partly at random, I have 潘悟云’s “温處方言和閩語” on my mind. (There may be better specimens.) Right at the top, he runs through a prelim. list of syntactical tendencies that “Min” & “Southern Wu” share with each other but not “Sinitic” at large. Most are strikingly recognizable in modern Hoklo. So there is an iceberg (substantial, consistent) of syntax — laced with INTRICATE tone phenomena — that we have somehow ignored to this point in the context of reconstruction & linguistic history, with Mandarish interlanguage playing various facilitative meta-role(s) in this ignoring.

RSS feed for comments on this post

Leave a Comment