Fried scholar's

« previous post | next post »

Last Sunday, Rebecca Fu went to a Cantonese restaurant called Xǐ yùn lái dà jiǔdiàn 喜運來大酒店 (Happy Fortune Arrives Grand Hotel — actually a modest establisment) in Manhattan's Chinatown.  When she saw the following entry on the menu, she had no idea what it was:  xiāngjiān shìde 香煎士的.  The xiāngjiān 香煎 was not a problem; it means simply "fried" or "pan fried".  But, even though she's a graduate of Peking University, Rebecca drew a complete blank on shìde 士的.  The characters seem to mean "scholar's", but "fried scholar's" just doesn't make sense.  It was only when Rebecca asked the waiter how to pronounce 士的 in Cantonese — whereupon he said "si6 dik1"– that she understood what 士的 meant.

The usual way to say "steak" in Modern Standard Mandarin (MSM) is niúpái 牛排, and that expression is also used in Cantonese, where it is pronounced ngau4 paai4.  But si6 dik1 士的 ("steak") is also used very commonly in speech and even, as on restaurant menus, in writing.  This is a good example of how a parallel vocabulary develops in Cantonese and other Sinitic languages:  one based on indigenous morphemes, and one on loanwords.

Cantonese has been especially prolific in borrowing words from other languages, and many of these words (I'll mention a few of them below) have entered the standard MSM lexicon.  Shanghainese has also borrowed many foreign words directly via their sounds ("clamp," "captain," "last car," etc.), and these too have often passed into the MSM lexicon.  The same is true of other topolects that are not normally written (or are not wholly writable) in Chinese characters.  In contrast, MSM is more apt to create loan translations for new terms and ideas that enter the language from abroad through calquing or through recycling (by redefinition) of old words (e.g., the MSM words for "philosophy", "religion", "economics", "literature", and so forth).  When the latter (recycled and redefined words) pass through Japan (China –> Japan [where the redefinition takes place] –> China), we may refer to such loans as "round-trip words".

It is curious that, if we reverse the order of the characters for si6 dik1 士的 ("steak"), we get dik1 si6 的士 ("taxi"), one of the most successful Cantonese borrowings in the entire Sinitic lexicon.  In MSM, 的士 is pronounced díshì (doesn't sound much like "taxi"), but people know what it means, and they even use it comfortably and flexibly in expressions such as dǎ dì 打的 ("strike [i.e., take] a taxi"), dígē 的哥 ("male taxi driver"), and díjiě 的姐 ("female taxi driver").  In fact, díshì 的士 ("taxi") prevails over the clunky trisyllabic neologisms based on native morphemes, chūzū chē 出租車 (lit., "vehicle for rental" — mainland usage) and jìchéngchē 計程車 ("vehicle that computes / calculates [the distance of a] journey" — Taiwan usage).

Another very successful Cantonese loan that is almost as ubiquitous as dik1 si6 的士 ("taxi") is baa1 si2 巴士 ("bus"), pronounced bāshì in MSM.  This contrasts with clumsy quadrisyllabic neologisms based on native morphemes, gōnggòng qìchē 公共汽車 (lit., "public automobile") and chángtú qìchē 長途汽車 (lit., "long distance automobile" — also often translated into English as "coach").  I still remember back in the early 1970s struggling to get the tones right on both of these monsters.

Finally, to close this blog, I will relate a personal anecdote.  This happened to me when I was teaching at the University of Hong Kong in 2002-2003.  On the ground level of the beautiful apartment building in which we lived, I was intrigued by a door that had the following sign on it:  shìduōfáng 士多房.  Well, that seemed pretty clearly to mean "a house / room with lots of scholars", so I thought that I'd go in and join the club.  When I opened the door and entered, all that I saw was a bunch of janitor's cleaning materials!  Feeling really silly, I asked around and soon found out that this was the si6 do1 fong4 ("storeroom") for the building.

Once again, the priority of spoken sound over apparent written meaning was brought home powerfully to me.  In modern Sinitic languages as in Classical Chinese (Literary Sinitic), if something doesn't make sense when you try to interpret it according to the surface signification of the characters, read it out loud (in the appropriate topolectal pronunciation or historical reconstruction, if you happen to know what that is likely to be) and you might be able to get a hint of what the author really intended.

[Thanks to Alan Chin for assistance with the Cantonese and to Jing Wen and Zhao Lu for Beijing usages]


  1. Ian F. said,

    August 6, 2011 @ 10:08 pm

    There seem to be a few of these Cantonese-specific borrowings floating around. I heard 士多卑利 in Hong Kong a few times. See if you can figure it out: it's pretty straightforward (hint: it's edible).

  2. Claw said,

    August 6, 2011 @ 10:10 pm

    The usual way to say "steak" in Modern Standard Mandarin (MSM) is niúpái 牛排, and that expression is also used in Cantonese, where it is pronounced ngau4 paai4.

    In Cantonese, 牛扒 (ngau4 paa4*2) is used much more often than 牛排. The Chinese edition of Wikipedia appears to have the explanation:

    "In Shanghainese 排 is pronounced [ba], which the Cantonese mistook for 扒 [pa], leading to the usage of 牛扒."

  3. Jinhui Chen said,

    August 6, 2011 @ 10:28 pm

    Re Ian F.: 士多卑利 means strawberry, even though I don't know how to speak Cantonese.

  4. Rebecca said,

    August 6, 2011 @ 10:57 pm

    To Ian F.: 士多啤梨 is used frequently, and 梨 hints that it is a kind of fruit~

  5. Wu Yong said,

    August 7, 2011 @ 12:18 am

    It seems like Chinese has it's own slightly more complicated version of the Japanese katakana script: there is a set of characters that are mostly used for writing foreign loanwords (i.e. 巴、斯 etc.). I sometimes wish the usage had been standardised in the early days of the republic…!

  6. Chris Barts said,

    August 7, 2011 @ 1:28 am

    Google Translate renders 香煎士的 as "Pan-fried with disabilities" and I doubt anyone could improve on that from an unintentional humor perspective.

  7. Chris Barts said,

    August 7, 2011 @ 1:30 am

    Also, in the name of comedy, 喜運來大酒店 is the "Max Fortune Hotel" to Google Translate. Sounds like a cheesy 1980s action flick.

  8. dom said,

    August 7, 2011 @ 1:40 am

    Although, if it's a Cantonese restaurant, it would actually be called "Hei2 wan6 loi4 dai6 zau2dim3", not "Xǐ yùn lái dà jiǔdiàn"…

  9. Herman Burchard said,

    August 7, 2011 @ 2:42 am

    Similar situations arise in languages written in Roman (Carolingian) glyphs using Google translate:

    bifteck French, bistecca Italian, biffstek Swedish, biefstuk Dutch

    These appear very well in line with si6 dik1 士的 ("steak"). The Dutch word stuk means "piece", stuk maken is to break.

  10. Steve said,

    August 7, 2011 @ 6:58 am

    I wonder if what we see here in these numerous examples isn't a form of pidgin, especially in the Cantonese speaking cultures that had early contact with the West. I recall from my childhood in San Francisco, my Cantonese grandparents on both sides of the family spoke a pidgin English, a language that had to be learned and had rules and standards. Their pidgin was, perhaps, an American reflection of what developed in Hong Kong the nineteenth and twentieth centuries to the present, the transformation of words and sounds into characters. Less a question of "fragrant fried scholars" than "where's the beef." Just kidding.

  11. Trimegistus said,

    August 7, 2011 @ 7:18 am

    Is that Greengrocer's Apostrophe in the original Cantonese?

  12. Victor Mair said,

    August 7, 2011 @ 7:22 am

    @Ian F.

    I thought that I would sleep on your 士多卑利. Then, when I woke up and saw Rebecca's variant, 士多啤梨, all I could think of was "lots of scholars beer pear". BTW, it was very smart of Rebecca to guess that this was some kind of fruit!

    But to tackle your 士多卑利 more directly and seriously, I knew that it must be the transcription of a foreign word, but not being able to figure out right away what 士多卑利 was, I decided to run it through Google Translate, and was bemused when out came "Groceries Peel". Hmmm…. For Rebecca's 士多啤梨, however, Google Translate is very sharp and is right on the target: "strawberry"! Good job, Google Translate.

    My guess is that Rebecca's version, 士多啤梨, has become the standard orthography for this originally Cantonese transcription which, by now, has probably also worked its way into Modern Standard Mandarin (MSM) to a certain degree, where it would be pronounced shìduōpílí. How's that for "strawberry"?! In Cantonese, Rebecca's version would be si6 do1 be1 lei4. Not bad!

    Let's see what Baidu Fanyi comes up with:

    士多啤梨 "strawberry"
    士多卑利 "With many Old Bailey" That's really cute!

    Because I'm a fruit and dessert fanatic (even though I'm allergic to strawberries and nearly all other fruits), I long ago learned that the "proper" rendering of "strawberry" in Chinese is cǎoméi 草莓.

    Now, to conclude these supplemental Sunday morning notes:

    "士多啤梨" 1,220,000 ghits
    "士多卑利" 68,200 ghits
    "草莓" 81,300,000 ghits

    From this, we can see:

    1. Rebecca's form has become the standard transcription for English "strawberry" in Chinese characters

    2. Rebecca's form has indeed been adopted into MSM

    3. Ian F.'s form is a minor variant

    4. In terms of the whole sinosphere, 士多啤梨 has a long way to go before it catches up with 草莓. I, for one, know that I will not soon be switching from Mandarin cǎoméi to shìduōpílí. Even though the latter has a nice ring to it, still, it's quite a mouthful, as it were.

    Oh, by the way, in response to Chris Barts' "Pan-fried with disabilities" 香煎士的 from Google Translate, I've been noticing that sometimes when Google Translate is stymied, it spews out "disabilities". It seems to be some kind of more or less frank admission on the part of the service to the effect, "Sorry, I'm not able to handle this" — yet.

  13. M said,

    August 7, 2011 @ 8:51 am

    Trimegistus: it is what the characters say. "的" is a possessive particle. It's not really a greengrocer's apostrophe because it wasn't intended to be a plural marker.

  14. Bruce Rusk said,

    August 7, 2011 @ 9:03 am

    My favorite of these is one a colleague came across reading early 20th c. material: 德律風 Dé lǜ fēng. "Wind of virtuous regulation"? "Trend of German law"? She puzzled over it for a while, then realized: "telephone."

  15. Pflaumbaum said,

    August 7, 2011 @ 9:49 am

    gōnggòng qìchē … I still remember back in the early 1970s struggling to get the tones right on both of these monsters.

    Same here, in the mid-nineties. Then an English girl who was a fluent Mandarin speaker overheard me and told me it sounded incomprehensible but vaguely obscene. She also informed me that when I thought I was saying 'excuse me' to people, I was actually saying 'frog' – which explained why they were so reluctant to give me directions.

  16. Sinkovits said,

    August 7, 2011 @ 3:45 pm

    What I love about cǎoméi 草莓 is that it's a calque: 草 = "grass", 莓 = "berry".

  17. Peter said,

    August 7, 2011 @ 5:34 pm

    The characters seem to mean "scholar's", but "fried scholar's" just doesn't make sense.

    Perhaps it does, considering this.

  18. Trimegistus said,

    August 7, 2011 @ 8:12 pm

    Ah. Thanks. I wondered about it.

  19. Bob said,

    August 7, 2011 @ 11:09 pm

    let me give you the most common 啤; no doubt, you know 酒吧,蛋挞,苹果批, etc., and one you probably have not idea: 基佬, 啊蛇.. ha ha..
    generally speaking, we use phonic Chinese characters to represent new things (to the region, at that time); ((草莓 was foreign to HongKong/Canton)) storeroom as 士多房 was typical of colonial HongKong, and 士多 as store (shop) is still commonly used.

  20. Aaron said,

    August 8, 2011 @ 12:15 am

    Sometimes the word also tugs along the Cantonese pronunciation. In Taishanese, we likewise say 的士 with our own local pronunciation, although not entirely local. The “native” pronunciation of each character is [ɛt ɬu] (or [it ɬu], depending on the accent), but we borrow some of the Cantonese segments and instead say [tɛt/tit si]. The avoidance of velar codas after [i] is preserved, although I also know of certain Four Counties accents, which more faithfully render this word as [tik si].

  21. J. Goard said,

    August 8, 2011 @ 12:54 am

    The usual way to say "steak" in Modern Standard Mandarin (MSM) is niúpái 牛排

    Hmm, that would be 우배 (u-bae) in Korean, which AFAIK has never had currency. (우 'cow[SK]' doesn't seem very widespread compared to compounds with the native Korean 소 (so), the primary exception being the high-frequency 우유 'milk'). I'm a bit thrown by that second character, which to me means 'remove, exclude'. So, 'cow-remove' — well, yeah, I guess the animal doesn't have that part anymore, but seems like a strange choice. Is this widely used in Mandarin for a piece or slice or whatever taken from the thing in the first half of a compound?

  22. Bathrobe said,

    August 8, 2011 @ 8:03 am

    There is 奶昔 nǎi xī, a mixed word, that has made its way into Mandarin. The first character means 'milk'. The second means 'past', but I believe the Cantonese pronunciation is close to 'shake'.

    Brand names also seem to come via Cantonese. Both Lux (力士 = lìshì) and Rolex (劳力士 = láolìshì) betray their Cantonese origins. Hint: there is a 'k' at the end of 力 in Cantonese.

  23. Bathrobe said,

    August 8, 2011 @ 8:06 am

    Also, 力士 means 'sumo wrestler' in Japanese.

  24. Jim said,

    August 8, 2011 @ 1:32 pm

    "I'm a bit thrown by that second character, which to me means 'remove, exclude'. So, 'cow-remove' — well, yeah, I guess the animal doesn't have that part anymore, but seems like a strange choice. Is this widely used in Mandarin for a piece or slice or whatever taken from the thing in the first half of a compound?"

    What we have here is a faux ami. The core meaning in Mandarin is a row or line as in ranks, to set in order. In catnoese the same word forms the expression for spareribs. It's a good description of slicing steaks off a roast or chunk of meat. The meaning you quote, "elimination, to drain away" may or may not be related, but that seems to be the meaning that went along into Korean.

  25. Fluxor said,

    August 8, 2011 @ 4:47 pm

    … they even use it comfortably and flexibly in expressions such as dǎ dì 打的 …
    I don't think 的 morphs to the fourth tone when used with 打. I believe the most commonly pronunciation is still dí.

  26. Fluxor said,

    August 8, 2011 @ 4:51 pm

    Here's another one of the most common Cantonese loan words from English: 忌廉. It's pronounced "gei-lim" and means "cream".

  27. Bathrobe said,

    August 8, 2011 @ 4:58 pm

    I think I've only ever heard dǎ dì.

  28. B.Ma said,

    August 8, 2011 @ 5:29 pm

    士的 is not a usage I have ever seen, and I initially read it as 的士..

    but 27 comments later nobody has thought to mention that 士多 can also be reversed to 多士 "toast", most commonly used in 西多士 "Hong Kong style French toast"

    Wikipedia claims the 西 is actually the last transliterated syllable of "France", which is plausible I suppose.

    Trimegistus / M: probably it should have been translated as "Of the fried scholar"

  29. Derek said,

    August 8, 2011 @ 10:03 pm

    Bill Poser blogged about 芝士士的卷 in 2004.

    As a native Cantonese speaker who grew up in Hong Kong in the 1990s then moved to Vancouver, I can't recall ever running across 士的 in the sense of "steak" in either city. If anything, 士的 seems to be used more often as a transcription for "stick", as in a walking stick or a cane.

    On the other hand, I've seen 百加利 baak3 gaa1 lei6*2 for "broccoli" at numerous Chinese restaurants in Greater Vancouver, in place of the standard 西蘭花 sai1 laan4 faa1. 柏文 paak3 man4 for "apartment" and 燕梳 jin3 so1 for "insurance" are also common among Cantonese speakers here; I've never encountered 百加利 or 柏文 in HK, while 燕梳 may have some currency at least among car-owners there.

    For 的士 ("taxi"), the second syllable undergoes a tone change from tone 6 to tone 2, so it's dik1 si2 instead of dik1 si6, and matches baa1 si2 巴士 (correctly represented in the post).

    As for gōnggòng qìchē 公共汽車, I've seen it abbreviated as 公車 gōngchē.

    Back to Cantonese, there's also 哥士的 go1 si6 dik1 for "caustic", as in caustic soda / sodium hydroxide.

  30. Bathrobe said,

    August 9, 2011 @ 12:11 am

    力士 means 'sumo wrestler' – Most likely from 大力士, the Cantonese word for "Heraklēs"

    Do you have a source for your etymology or are you just shooting from the hip?

  31. Bathrobe said,

    August 9, 2011 @ 12:13 am

    Whatever happened to the comment I was commenting on?

    [(myl) If it was a comment by "Bryan", I deleted it.

    I apologize for disrupting your discussion, but my tolerance for troll-like behavior of the kind that he exhibits has become increasingly small as the years go by. One of his recent comments was "It's Chinese, not Sinitic, fucktard!", and his other comments have been longer but not better informed or (much) more civil.]

  32. perspectivehere said,

    August 9, 2011 @ 9:50 am

    Victor Mair wrote: "In MSM, 的士 is pronounced díshì (doesn't sound much like "taxi"), but people know what it means, and they even use it comfortably and flexibly in expressions such as dǎ dì 打的 ("strike [i.e., take] a taxi"),…"

    In my opinion, "strike a taxi" sounds implausible as an explanation for "打的".

    In Cantonese, "take a taxi" is expressed as “搭的士” (daap3 dik1 si2):

    It seems pretty likely that “搭的士” evolved into "打的", but it's not known exactly how.

    According to the online Hudong Encyclopedia entry for "打的” (打的):

    "的士: 这个词语并不是现代汉语中本来就有的词语,而是源自粤方言。出租车在经济比较发达的香港和广东地区较早出现,于是粤方言中也就相应地有了“搭的士”的说法。随着经济的发展,出租车这一新事物也成为内地社会生活的需要,“搭的士”的说法就传到了内地。不过“搭的士”的说法传到内地却成为“打的”,那么,这“搭”和“打”的一字之差又是怎么产生的呢?这可能与“搭”、“打”的音义有关。"

    (My translation: "Taxi [的士]: This phrase is not native to modern Chinese, rather, it originates from Cantonese dialect. Cars for hire [i.e., taxis, “的士”] appeared relatively early in economically-more-affluent Hong Kong and Guangdong areas, and as a result there appeared the expression "take a taxi" [“搭的士”] in Cantonese. As the economy developed, this new phenomenon of cars for hire [i.e., taxis, “的士”] also became a necessity of life in mainland areas, and the expression "take a taxi" [“搭的士”] was also transmitted inland. However, when the phrase "take a taxi" [“搭的士”] was transmitted inland, it became “打的”. How did the single character change from “搭" to “打” come about? It may have had something to do with the sound and meaning of “搭” and “打”.)

    It would be interesting if someone could find evidence of the first use of “打的” on the mainland.

  33. Áine ní Dhonnchadha said,

    August 9, 2011 @ 12:08 pm

    In Beijing, we always took the miandi "loaf-taxi": a small van taxi common at the time (and perhaps still). This was in the 90s.

    On the other hand, we definitely also used "gonggong qiche", which admittedly is seriously syncopated by Lao Beijingr folks. You barely hear the off-syllables in normal speech so it's sort of GONG'ng QIch'

  34. Victor Mair said,

    August 11, 2011 @ 6:02 am


    "I don't think 的 morphs to the fourth tone when used with 打. I believe the most commonly [sic] pronunciation is still dí."

    Not sure what you mean by "morphs to the fourth tone", but the situation is certainly much more complicated and different than your proposal that the most common pronunciation "is still dí." The dictionary entry form of 的士 is dìshì (see, for instance, the alphabetical index to the Hanyu da cidian and also in the dictionary itself); though I grant you that some dictionaries do give díshì. However, for dǎ dī 打的 ("strike [i.e., take] a taxi"), dīgē 的哥 ("male taxi driver"), and dījiě 的姐 ("female taxi driver"), the vast majority of native speakers are not using the second tone for 的, but rather the first tone.

    Nearly all native speakers (including several from Beijing, but also individuals from other parts of the north and northeast) whom I consulted stated that they pronounce the 的 of 的士 as dī. Now, this is very strange, because the Hanyu da cidian and other authoritative dictionaries that I consulted all list only dí ("certainly"), dì ("target; goal"), and de (particle) as possible pronunciations of 的. So what is going on here? Where did dī come from?

    Several possibilities spring to mind. It may be part of a general tendency for tones to collapse into the first tone that we have observed upon several occasions, e.g., g., 期. Another possibility is that, since dìshì is not a possible spoken form (it must yield to sandhi constraints), speakers somehow just ended up adopting dī as a compromise between dí (spoken realization) and dì (dictionary entry form). Once they became comfortable with dī, they just stuck with that in all situations, since it wouldn't be affected by sandhi constraints. Yet another possibility is that, since 的士 is a borrowing into Mandarin from Cantonese — and a fairly recent one at that (from the 80s, so it would seem) — there were no hard and fast conventions for how to stabilize the tones in Mandarin.

    That something like one of these scenarios may actually occurred is borne out by three things:

    1. When I explained to my informants that only dí, dì, and de are possible MSM pronunciations for 的, they understood the implications of what I pointed out to them, but replied simply thus: "I don't know why, but somehow that's the way it comes out", meaning that it comes out in the first tone, not in the fourth or second, much less in the neutral tone. It's very interesting that several of my informants said almost the same thing, something which seems, on the face of it, to be very peculiar.

    2. When I queried them repeatedly how they pronounced 的士, 打的, 的哥, and 的姐, they were all emphatic that in each case the 的 is pronounced DI1 — in the first tone — despite the fact that a couple of them had initially told me DI4SHI4 for 的士 and a couple had said DI2SHI4.

    3. My surmise that a new tonal form had arisen for 的, namely dī, has actually — and much to both my surprise and gratification — been documented in a Baidu entry on taxis, under the section on pinyin:


    的”字的读音存在争议,根据2005年更新的现代汉语词典,应读dí shì, 但根据2011年更新的新华字典,应读dī shì,目前更多的人都会说dī shì。

    Here is the link:


    So we can see that in recent years there definitely has been ambivalence about what tone the 的 of 的士 should be, but that most speakers of MSM have, for some reason that is not altogether clear even to themselves, settled on dī, a tone that previously did not exist for 的.

    My recollection is that throughout China most taxis, esp, on the lighted sign on top the roof, have the word TAXI on one side and 的士 (much less often, at least in past years, 出租车) on the other side. So, since the 的士 appears in writing all over the place, surely the people must have some more or less "standard" way of pronouncing it.

    I'm absolutely certain that I used to hear people saying DA3DI1 or DA3DI4 ("take a taxi"), not DA3DI2. Now, it seems as though most people are saying DA3DI1.

    I myself have never heard anyone say DA3DI2.

    Just before posting this, I received a note from a native informant who hails from the Northeast. She says that she pronounces 的 as DI2 in all of the expressions related to "taxi" discussed above. Nonetheless, the overwhelming majority of my native informants, including especially those from Beijing, insist that they say DI1 in all of these expressions.

    Another late respondent just told me that she USED to say DI2SHI4, but recently she has begun to say DI1SHI4.

    From a correspondent, Julie Wei, who grew up speaking both Cantonese and Mandarin:


    Yes, very interesting problem. I use the various dictionary sounds for 的 that you give , and I also say di2shi4 or di1shi4 for "taxi". I think this is influenced by Cantonese dig1si for "taxi". I write dig1 or di'1 for the Cantonese tone because it sounds like Mandarin first tone. Since Cantonese is dig1 or di'1, it's easy to say di1sh4 in Mandarin. The characters come from Cantonese 的士。 In other words it's a foreign word (Cantonese) embedded into Mandarin and pronounced somewhat like a foreign word (Cantonese), at least that's why I say di2shi4 or di1shi4.


    From my learned friend, Endymion Wilkinson:


    As to your question regarding the pronunciation of the first syllable of 的 士, I suppose the answer is "level tone." The reason being that the two earlier transcriptions for taxi in Cantonese were 德士 and 滴 士, whose first syllables (second and first tone, respectively) conform to the rule that disyllabic collocations are typically composed of a level tone (pingsheng 平 聲) followed by an oblique one (zesheng 仄聲).


    From another learned friend, Cornelius Kubler:


    I don't know about 的士, which is really written Cantonese and not Mandarin anyway, and seldom said in Mandarin (in my experience). But I totally agree with you that in 打的 and your other examples, 的 is first tone. So it's definitely DA3DI1 (in fact, some Beijing speakers will even say DA3DIR1 or DA3 GE DIR1).

    As for the reason, it's interesting to note that in the direct source language, Cantonese, 的 is pronounced DIK with a high level (short and clipped, rusheng) tone. Not just in 的士 but in all other uses of 的 that I'm aware of.

    This would then make it the second character or DUOYINZI / POYINZI that can be pronounced with 4 tones! (FA3 has 3: mei FAR1 "no way," mei FA2ZI "no way," fangFA3 "way," and FA4guo "France" (for those speakers that don't say FA3guo).


    From Jeremy Goldkorn, long-term resident of BJ and co-host of Danwei:


    I asked a few Beijingers. They all say the 'di' is first tone.

    I think the reason may be that the word comes from Hong Kong Cantonese and was not previously used in northern Chinese


    From Yao Dehuai, head of the Hong Kong Language Society:


    Yes, these problems have been much discussed among linguists. I think variation of tone of “的”(變調, tone sandhi) arises from 順口 (“fluency”for saying), and such variations are common in many Chinese dialects.

    Another strange phenomenon is“variation of Hanzi”. Cantonese“埋單”( =“結帳”, paying bill in restaurants ) is now being borrowed in Putonghua, but is written as“買單” ( to buy a bill )!


    From Yu Li:


    The word "dishi"/taxi only came to the Mandarin vocabulary in late 1980's, when many trends (including for example words and the favorite number of "8") from the Canton area became new fashions in major cities of the whole country.

    Taxi in Cantonese is something like "tek-sii" with the first syllable has a checked vowel and sounds like the first tone in Mandarin. Since then, "di1shi4" replaced "chu1zu1che1" in colloquial Chinese. Nobody gives it much thought since this is a loan word and it probably never enters a formal dictionary.

    Another interesting thing I have noted which is similar is that in dictionaries and in 99% of the language textbooks I have seen, the character "喂" has the pinyin of wei4 (hey). But in real phone conversations, people always say "wei2." So far, I have only seen one textbook marking the second tone in this situation. In the actual classroom, I have always taught my students to say the rising tone despite the fact that our textbook indicates the fourth tone.

    So I guess the lesson here is that dictionaries do not respect colloquial Chinese, or they cannot follow closely the changes of the Colloquial Chinese.

    Maybe in some newly compiled dictionaries that focus on Colloquial Chinese, you will be able to find "di" being marked as "di1" in "di1shi4."


    To conclude, it is fascinating to contemplate that the tone which native Mandarin speakers have settled on for the 的 of 的士 is dī, a tone that previously didn't exist in the repertoire of possible tones for 的. If we can figure out how and why this happened, I think we will have a much better grasp of the nature of Chinese than heretofore. Of course, we cannot just focus on 的, but need to look at a whole range of similar manifestations of real, living language, language in the context of interaction with other languages, not just as it is described in isolation.

  35. Neil Kubler said,

    August 11, 2011 @ 9:52 am

    Two more quick comments about DI as in DADI: (1) I worked extensively with native Beijing speakers in the fall of 1994 and spring of 1995 in developing new Mandarin instructional materials and already then, all my informants were saying DADI with Tone 1 on the DI. A few preferred DADIR with the -R ending, still Tone 1. So it's been around for a while. (2) Another thought about the origin of Tone 1 on DI: DI Tone 1 is a kind of CHE ("vehicle, car"), which is also Tone 1. Could the tone of a semantically related word (the semantic category) be in the mind of the speaker and also exert a certain amount of influence? I have a small collection of errors by native Mandarin speakers where they imposed the tone of a semantically related word on another word. Example: pronouncing 找不到
    with a Tone 2 on the 到 (clearly because at a deep structure level they were also considering using 找不着 — and 着 is Tone 2). One final note: My original speculation in an earlier posting that the high level (even if very clipped) tone of DIK in Cantonese DIKSI 'taxi' had something to do with the Tone 1 of DI in DADI this, as others also have noted, could still be part of the equation. But I think one must admit that the borrowing of DIKSI into Mandarin originated at the written level, i.e., Chinese characters, and from there spread into Mandarin speech — which makes my speculation about the influence of the Cantonese high level tone less plausible.

  36. Ahkow said,

    August 11, 2011 @ 10:07 am

    To quote Victor (who quotes Yao Dehuai):
    "Another strange phenomenon is“variation of Hanzi”. Cantonese“埋單”( =“結帳”, paying bill in restaurants ) is now being borrowed in Putonghua, but is written as“買單” ( to buy a bill )"

    Not that surprising, actually. The tone (陽平) in 埋 in Cantonese is a low-flat/low-falling tone, which is very similar to the low tone in the Mandarin third (上) tone. Basically, “埋單” [Cantonese] is homophonous with "買單" [Mandarin], even in tone. Add to that the similarity between the concepts of bill-paying and buying an item.

    In Singaporean and Malaysian Mandarin, the standard local Mandarin Chinese term for taxi is de2shi4 德士 (cf., among others). Note that the first syllable is second tone (rising) in standard speech.

    This can be attributed to 德士 being pronounced as [teksi] in the Min dialects commonly spoken in the region, with corresponding tonal reflexes. 打的, 的哥, and other forms based on 的士 are uncommon and immediately identifiable as being foreign.

  37. Observation said,

    August 23, 2012 @ 5:11 am


RSS feed for comments on this post