Russian Loans in Northeast and Northwest Mandarin: The Power of Script to Influence Pronunciation

« previous post | next post »

Three days ago, I asked the students in one of my classes to tell me all the languages they knew.  One of the female students listed Northeastern Mandarin among her languages.  When I asked her to say a sample sentence in that language, she said something like "Ni de blaji hen haokan" (Your blaji is pretty).  Her sentence surprised me for two reasons.  First, I didn't know the meaning of blaji; second, I was stunned that she used what sounded like a bl- consonant cluster at the beginning of the word with which I was unfamiliar, since Mandarin — at least proper, standard Mandarin — does not have consonant clusters.

At that point, I asked whether anyone else in the class knew the word blaji.  Out of approximately 50 takers and auditors in the lecture hall (probably about half know Mandarin), only one other female student was familiar with the word blaji (N.B.:  two syllables), but she pronounced it very clearly as bùlāji 布拉吉 (three syllables).  The same word is also transcribed as bùlājí 不拉及.

After inquiring, I learned from the two students that the word blaji / bùlāji means "dress" and was borrowed from Russian платье.  Here's what a blaji / bùlāji looks like.

Because I was so startled by the consonant cluster at the beginning of the word as pronounced by the first student and was intrigued by the disparity between the pronunciation of the same word by the two students, I quizzed them both very carefully about how they learned the word.  It turns out that the first student, an undergrad who grew up mostly in America, but with Northeast China roots, is barely literate in Chinese characters, though she is fluent in Mandarin, and had learned the word blaji strictly orally from her grandmother.  The second student is from a village close to the Russian border, had stayed in China her whole life until coming to graduate school in the United States, and is highly literate in Chinese characters, having attended Peking University as an undergraduate Chinese language and linguistics major, then coming to continue the same field of study as a graduate student in America.

It is fascinating how the student who is highly literate in Chinese characters unmistakably says bùlāji for Russian платье, whereas the student who is barely literate in Chinese characters pronounces the same word as blaji.

I have found the same phenomenon among Dungan speakers in Central Asia.  The Dungans are Muslims who fled Northwest China after failed rebellions during the latter part of the 19th century and ended up mostly in what are now Kazakhstan and Kyrgyzstan.  The Dungans are illiterate in Chinese characters, but literate in Cyrillic.  This has had a profound impact on the shape of words borrowed into their language from Russian, Persian, Arabic, and other languages.  Here I shall give only a single example, but it demonstrates exactly the same phenomenon as does the borrowing of Russian платье into non-sinographic Northeast oral Mandarin.

For Russian трактор ("tractor"), the Dungans say трактор, despite the fact that they speak two varieties of Northwest Mandarin.  In Modern Standard Mandarin, the word for "tractor" is tuōlājī 拖拉機 ("drag-pull-machine," but with the first two syllables cleverly chosen to represent the first syllable of the original.

[With thanks to Anne Huang for the website showing photographs of blaji / bùlāji.]



41 Comments

  1. mondain said,

    January 24, 2011 @ 12:27 am

    One of such loans used in the Northeast I know is WEI4 DE LUO2 for 'bucket,' from Russian ведро, also known as BA2 KAI1 SI, from Japanese バケツ (from English). The tones may be not accurate.

  2. Janne said,

    January 24, 2011 @ 12:42 am

    Your first student learned a largely unadulterated dialect, while your second student spent a number of years in the capital, studying language no less. To what extent could the pronunciation change be due to influence from being immersed in a standard Mandarin-speaking environment, rather than the writing system? The implicit social pressure to tone down your dialect – obvious and distinctive features like this especially – is normally quite strong in such a situation.

  3. tiffert said,

    January 24, 2011 @ 1:10 am

    Traveling among rural Kyrgyz in Xinjiang, I found their Mandarin, which was not what they spoke among themselves, to be peppered with straight-up Russian loan words, such as машина instead of 车.

  4. DL said,

    January 24, 2011 @ 3:17 am

    Here's another example from Cantonese: hum buh laang VS humblaang (colloquial version of 'all' that very few people use anymore). Probably part of the reason for the alternation is because no one knows how it's written or how it's supposed to be pronounced.

  5. Kyle said,

    January 24, 2011 @ 4:34 am

    I wonder if this shows up in English. Maybe people who read a lot speak in a way that reflects the written forms of words (fewer reduced forms) — more so than people who are illiterate or read very little.

  6. Yao Ziyuan said,

    January 24, 2011 @ 4:48 am

    The remaining curiosity is: You're a male, and how can you wear a dress in a class…

  7. Jacob said,

    January 24, 2011 @ 4:54 am

    @Yao: Are you from either Liaoning, Jilin or Heilongjiang? If not, you've probably not heard someone compliment your or your daughter's 布拉吉 (Russian) or maybe you've never played 嘎啦哈 (from Manchu) or heard any of the number of words of Russian or Manchu origin that are in use in this part of the country.

    Here is the first article that came up in a simple Baidu search of Dongbei words of Russian origin, and there's no doubt lots more.

  8. Jean-Sébastien Girard said,

    January 24, 2011 @ 5:58 am

    That tuo-la-ji example reminds me of phonosemantic matching. When one think about it, it makes sense that Mandarin would have a lot of that.

  9. Victor Mair said,

    January 24, 2011 @ 6:21 am

    @Yao Ziyuan "The remaining curiosity is: You're a male, and how can you wear a dress in a class…."

    LOL!!! (First time I ever wrote that on LL!!) Thanks for a good laugh this happy morning.

    She wasn't saying that MY blaji was pretty!! She was just citing a sentence that she might say to another girl, or that another girl might say to her.

  10. Rick S said,

    January 24, 2011 @ 9:16 am

    @Kyle: As a child, I read voraciously. From books I learned the word "awry", which I pronounced phonetically as [ˈɐɹi], and deduced its meaning from context. Separately, I heard the word "awry", pronounced [ʌˈɹaɪ], in fixed phrases such as "to go awry" and learned its meaning. It wasn't until I was in my 20s that I realized these were the same word, when an episode of the sitcom That Girl used this exact confusion for comedic effect.

  11. Erik said,

    January 24, 2011 @ 9:35 am

    I can't think of any theoretical reason why this couldn't happen in English, but I think it would be difficult to find examples of such.

    You need to find two groups of people with significantly different orthographies, but similar spoken languages. I think English is pretty much always written in the same alphabet, so this is hard to find. You could look into regional or dialectical differences in spelling, I suppose, but without the dramatic difference between a syllabic orthography and a… an alphabetic (?) one, it's harder to identify differences that are linked strongly to the orthography.

  12. Ellen K. said,

    January 24, 2011 @ 10:37 am

    Although none come to mind at the moment, I believe English has plenty of borrowed words whose English pronunciation is affected by spelling, as well occasional other cases of word pronunciation being affected by spelling.

    Also, surnames get their pronunciation in English affected by spelling. German names that begin with a W get pronounced with [w], rather than [v] as in German, for example.

  13. John Cowan said,

    January 24, 2011 @ 10:37 am

    Erik:

    The languages with multiple orthographies are typically cases where there is some other separation, usually geographical/political, and the difference in script is ancillary to that, as in the case of traditional Han (Hong Kong and the ROC) vs. simplified Han (the PRC and Singapore). Here you'd have a pain sorting out which effect is due to what cause. Other causes are languages in transition, or ones in which one script (often Arabic) is confined to a particular sphere, whereas the other is general. Only Serbian is fully biscriptal, but there can hardly be any such effect there, as the two alphabets are perfectly isomorphic, and every Serb learns both scripts from day one of school.

    But if someone wants to investigate multi-scriptal languages generally, here's a list: Azeri (Arabic, Cyrillic, Latin), Central Morocco Tamazight (Tifinagh, Latin), Hausa (Arabic, Latin), Javanese (Arabic, Latin), Kurdish (Arabic, Latin), Mongolian (Cyrillic, Mongolian), Punjabi (Arabic, Gurmukhi), Serbian (Cyrillic, Latin), Tachelhit (Tifinagh, Latin), Tajik (Arabic, Cyrillic), Uighur (Arabic, Latin), Uzbek (Arabic, Cyrillic, Latin), Wolof (Arabic, Latin).

  14. Fluxor said,

    January 24, 2011 @ 10:47 am

    @DL: People do know how "hum buh laang" is pronounced and how it's written in characters: see here. Your second pronunciation variation is just a fast contracted version of the first.

  15. ray said,

    January 24, 2011 @ 12:13 pm

    @Rick S – I also made the "awry" mistake, never having heard it said. But I also memorized Hamlet's "to be or not to be" soliloquy on my own, thinking it would be cool to know it. How amusing it was (embarrassing?) when I first heard the soliloquy actually acted all the way through by a competent performer and realized I'd been mispronouncing "awry" to myself this whole time. Good thing I had never volunteered to recite it for anyone but myself and the dog!

  16. linda seebach said,

    January 24, 2011 @ 1:56 pm

    My father, a lawyer but not otherwise much exposed to highly literate company, liked to tell about how he learned that the word he thought was pronounced "mizzled" was spelled "misled." (I used to hear about a city called Neuroshell when I was growing up on Long Island.)

  17. the other Mark P said,

    January 24, 2011 @ 4:48 pm

    I believe English has plenty of borrowed words whose English pronunciation is affected by spelling

    "Garage" and "herb".

    I say these two because the English say them one way, based on spelling, and most Americans the other way, based on origin.

  18. Lareina said,

    January 24, 2011 @ 5:06 pm

    布拉吉 is such a cute word, much cuter than 连衣裙!

  19. Chandra said,

    January 24, 2011 @ 5:16 pm

    @Rick S: I had a similar experience. I knew the spoken word /əˈpɪtəmi/ and the written word "epitome", and it wasn't until one day when I was trying to spell out /ˈɛpətom/ that I realized…

  20. Chandra said,

    January 24, 2011 @ 5:18 pm

    Rather, I was trying to spell out /əˈpɪtəmi/. /ˈɛpətom/ is how I was pronouncing the written word in my head.

  21. Mike P said,

    January 24, 2011 @ 9:37 pm

    I had "askance" in addition to "awry". It was probably a result of too many fantasy books growing up.

  22. Victor Mair said,

    January 24, 2011 @ 9:49 pm

    From a correspondent:

    The lines on the map make people forget that language can be a continuum. I was stunned, visiting Alaska recently, to discover Americans in remote areas speaking a kind of macaronic Russian. I heard a flight dispatcher say the "samolyot will be there in the ootrom." I shouldn't have been surprised, considering that the Americans have only owned the territory for 140 years, and it's not like the existing Russian colonists had a chance to go home.

    So interesting to read about how that Russian influence wraps around in the other direction. I would love to read about Russian influence in Khmer or Japanese, too.

  23. Amos said,

    January 25, 2011 @ 4:01 am

    I would often Chinese speakers in Urumqi (both Han and Uyghur) pronounce 'Kazakhstan', which in Chinese is 哈萨克斯坦 ha sa ke si tan, as something like [xa.sa.ke.stan], with the vowel after 's' elided, (this is perhaps a 'voiceless' vowel, as in Japanese with the lips in the right position, but no voicing / audible sound).

    But I would definitely hear the consonant cluster 'st', which struck me as odd. My hypothesis then was that people would hear the name more in speech than see it written down, resulting in the pronunciation of the consonant cluster.

  24. Nathan Myers said,

    January 25, 2011 @ 5:11 am

    English is filled with examples of mistaken "spelling pronunciations". Almost every day I hear one or other ignorant BBC newsreader mispronounce "often" as "off-ten". It grates.

    [(myl) The OED says:

    Several orthoepists of the 16th and 17th centuries, including Hart, Bullokar, Robinson, Gil, and Hodges, give a pronunciation with medial -t-. Others, including Coles, Young, Strong, and Brown, record a pronunciation without -t-, which, despite its use in the 16th cent. by Elizabeth I, seems to have been avoided by careful speakers in the 17th cent. (see E. J. Dobson Eng. Pronunc. 1500–1700 (ed. 2, 1968) II. §405). Loss of t after f occurs in other cases; compare soften v., … The pronunciation with -t- has frequently been considered to be hypercorrection in recent times: see for example H. W. Fowler Mod. Eng. Usage (1926), s.v.

    So the pronunciation with -t- is simultaneously a spelling pronunciation and a historically valid variant.

    As for that grating sensation, if it persists for more than a few hours, or is accompanied by flatulence, blurred vision, or involuntary clenching of the jaw, you should consult your doctor about modern therapeutic alternatives.]

  25. Matt McIrvin said,

    January 25, 2011 @ 8:02 am

    Jim "Ernest" Varney used to use the "epitome" confusion for comic effect in commercials.

    The BBC "often" variant seems less an erroneous eye-pronunciation than a deliberate stylistic choice. Nobody thinks the two pronunciations are different words.

  26. Marc said,

    January 25, 2011 @ 8:41 am

    Couldn't the US-raised speaker simply have been pronouncing it wrong? My guess is her grandmother pronounced it like the China-raised speaker.

    There is one way to find out if Chinese now has consonant clusters: find monolingual Chinese speakers who say blaji.

  27. Victor Mair said,

    January 25, 2011 @ 11:24 am

    @Marc Her spoken Chinese is quite fluent. I quizzed her on BLAJI repeatedly, especially in contrast to the BULAJI of the highly character-literate student, and she insisted that her grandmother had said BLAJI, not BULAJI. Moreover, some of the other commenters above point to similar phenomena in spoken Cantonese and far northwestern Mandarin (in Urumchi), and my original post mentioned the difference between sinographic TUOLAJI and Dungan Cyrillic TRAKTOR. You must remember that what the local northeastern Chinese would have *heard* from the Russians would be closer to BLAJI than to BULAJI, and that many of them would have been sinographically illiterate, so there would initially (before someone decided to write the word down in characters) have been no need to insert a -U- between B- and -L.

  28. stormboy said,

    January 25, 2011 @ 12:08 pm

    As the 'blaji' variant violates the phonotactic constraints of standard and perhaps other varieties of Mandarin, I'm surprised that even in the north-east the consonant cluster wouldn't be simplified most of the time.

  29. Marc said,

    January 25, 2011 @ 1:52 pm

    Right, but the fact that she has "bl" as part of her inventory makes me suspicious. I don't doubt that consonant clusters might be real in that Chinese dialect, but as I say, I'd have to see an entire village of monolinguals doing it, for example, before I concluded that Chinese now has consonant clusters.

  30. Peter said,

    January 25, 2011 @ 3:59 pm

    > I was stunned, visiting Alaska recently, to discover Americans in remote areas speaking a kind of macaronic Russian. I heard a flight dispatcher say the "samolyot will be there in the ootrom." I shouldn't have been surprised, considering that the Americans have only owned the territory for 140 years, and it's not like the existing Russian colonists had a chance to go home.

    This example is intriguing, though: while самолёты certainly existed 140 years ago, this meaning of the word didn't. Presumably there must have been sustained cultural connections well into the 20th century for 'samolyot' to have been borrowed in preference to 'airplane'?

  31. Jongseong Park said,

    January 26, 2011 @ 5:26 am

    I would love to read about Russian influence in Khmer or Japanese, too.

    Central Asian Korean, spoken by ethnic Koreans in the former Soviet republics of Central Asia, of course has a large amount of Russian loanwords. In North Korea, too, there are lots of Russian loans, often where English loans would be used in South Korea. I believe they use something like 뜨락또르 tteurakttoreu from Russian трактор, whereas South Koreans would use 트랙터 teuraekteo from English 'tractor'.

  32. CortxVortx said,

    January 26, 2011 @ 4:03 pm

    I was in my mid-20s before I realized that what I'd been reading as EN-you-eye was in fact pronounced ahn-WEE. Pardon my French, indeed! (And don't get me started on uncommon proper names like Penny-Lope and Beat-rice!)

  33. Sally Thomason said,

    January 26, 2011 @ 5:37 pm

    Russian loanwords don't necessarily mean Russian phonological influence, in Mandarin or English or any other language; and even a sizable number of loanwords doesn't count as a "macaronic" mixture of Russian & English (or any other Russian + X pair). English, for instance isn't a macaronic mixture of English & French (& Latin, & Greek, & …), in spite of its huge number of loanwords from French &c.: the grammar of English is still solidly Germanic.

    But there can certainly be phonological influence resulting from language contact, and it sounds as if that's what's happened, or happening, in some dialects of northern Mandarin — including the introduction of syllable-initial consonant clusters. Ditto, I bet, in some Mandarin/Uighur contact situations. This is a very common phenomenon in intensive language contact situations. It's the major theory of how Vietnamese became a tone language, for instance: first a whole bunch of Chinese loanwords coming in with their Chinese tones, then various phonological changes in native Vietnamese vocabulary that left tone distinctions on vowels in native words. It's also one reason, probably the main reason, English has word-initial voiced fricatives: they came in on French loanwords, e.g. very and zeal.

  34. Jongseong Park said,

    January 27, 2011 @ 4:50 am

    One example of the phonological influence of neighbouring languages I remember is from Slovene, a language with considerable dialectal variation. I summarize from memory, but the gist should be accurate enough:

    Like the majority of Slavic languages, Standard Slovene exhibits final devoicing and regressive voicing assimilation. But in dialects geographically closer to areas where German is spoken, all syllable-coda consonants are devoiced as in German; in dialects closer to areas where Croatian is spoken, voice distinction is preserved word-finally as in Croatian.

  35. Atmir Ilias said,

    January 27, 2011 @ 5:26 pm

    1.布

    Read phonetically: bù
    Meaning:
    布 remnant.
    Remnant Primitive, a hand holding a cloth 巾 – cloth
    English Senses For: bu
    cloth / textiles / …

    2. 布

    Read Phonetically: lā.
    Meaning:
    扌手 shou3 hand palm
    Phonetic meaning, something done with the hand 扌手 – pull
    English Senses For: la
    outstretch/to hold / to seize / to lengthen / to elongate / extended…

    3. 吉
    Read Phonetically: jí
    Meaning:suffix

  36. Atmir Ilias said,

    January 27, 2011 @ 5:29 pm

    1.布

    Read phonetically: bù
    Meaning:
    布 remnant.
    Remnant Primitive, a hand holding a cloth 巾 – cloth
    English Senses For: bu
    cloth / textiles / …

    2. 拉

    Read Phonetically: lā.
    Meaning:
    扌手 shou3 hand palm
    Phonetic meaning, something done with the hand 扌手 – pull
    English Senses For: la
    outstretch/to hold / to seize / to lengthen / to elongate / extended…

    3. 吉
    Read Phonetically: jí
    Meaning:suffix

  37. Yao Ziyuan said,

    January 29, 2011 @ 11:07 am

    Beware! She is a Manchurian Candidate.

  38. jk said,

    January 31, 2011 @ 12:17 pm

    Another for the confused pronunciations list: I sat baffled through a college lecture in which the professor made frequent reference to a word I'd never encountered, pair-a-dimes. Only when rereading the chapter he was discussing did the penny drop; I'd been pronouncing it pair-a-dijjums in my head for several years.

  39. jk said,

    January 31, 2011 @ 12:36 pm

    I have a tiny smattering of Polish, learned aurally from my first-generation parents and exposure to Polish-American radio shows and singers. I have had difficulty matching up some of the words and phrases I know to written versions because of the same kind of mismatches as your student.

    For example, I have words with an internal "N" sound that, I find, are spelled with an "M." Conversely, what I have as vowel + plus M sounds often turn out to be nasal vowels instead.

    Like your student, I would swear that this is how my parents pronounced the words. It is possible that these are actual dialect differences. My father certainly was aware of differences between some of his pronunciations and those of recent Polish immigrants (though he saw this as evidence that they were losing the "true" Polish).

    But it is equally possible that my parents' pronunciations were influenced by their bilingual childhood (all four of my grandparents were immigrants, and my father's spoke Polish exclusively, but my parents attended schools taught largely in English), or that what I heard was influenced by my own English-only upbringing. Adding weight to the latter possibilities, I had teachers from similar backgrounds as my parents, but originating in different parts of Poland, and to my ears their pronunciations were the same as I heard at home.

  40. David Marjanović said,

    March 14, 2011 @ 1:37 pm

    It's probably too late to matter, but I'd like to mention that Dungan has had such strong Arabic and Persian influence for centuries that they borrowed not only consonant clusters but even the sound [r] from these languages. When Dungan was written at all, a version of the Arabic script was used where all vowels were indicated.

    Like the majority of Slavic languages, Standard Slovene exhibits final devoicing and regressive voicing assimilation. But in dialects geographically closer to areas where German is spoken, all syllable-coda consonants are devoiced as in German

    I'm not sure if that's the case in the Carinthian dialects, which have taken most of their sound system from their Slovene substrate. Unlike other Bavarian-Austrian dialects, they have voiced plosives.

  41. Ted said,

    August 29, 2012 @ 3:42 pm

    I'm amused, coming upon this thread much later, to see that I'm not the only one to have been fooled by awry, which, like Rick and ray, I thought as a child rhymed with hoary. And, Linda, I feel great affinity for your father — I'm also from Long Island, and never understood why people were confused when I used forms other than the past participle of the verb "to misle," although when I said it it near-rhymed with rifle. My father, for his part, grew up believing that wholly rhymed with trolley.

    The next town over from Neuroshell, of course, was Mama Ronnick, according to my grandmother. She was a native German speaker, though, which may account for it.

RSS feed for comments on this post