Backhill, Pekin, Peking, Beijing
« previous post | next post »
Yesterday, while doing research for a paper on medieval Dunhuang popular narratives (biànwén 變文 ["transformation texts"]), I did a Google search for the Peking Library, where some of the bianwen manuscripts are kept. Instead of the national library of China in Peking / Beijing in the PRC, I was led to the Pekin Public Library in Illinois. That prompted me to ponder the fact that this Illinois city followed the French pronunciation, Pékin, of the Chinese capital when it took its name, rather than the English Peking.
Following the official Hanyu Pinyin Romanization of the PRC, English now transcribes 北京 ("Northern Capital") as Beijing (Běijīng [pèi.tɕíŋ]). But until recently this was not always the case for English, much less for dozens of other languages around the world. Thirty-one years ago, in "Backhill / Peking / Beijing" (see "Selected readings" below), Bosat Man wrote (p. 6):
We should not feel guilty for saying Peking instead of Beijing. It is not because we are uncouth foreign devils that we pronounce the name of the Chinese capital the way we do, but because we have inherited a long tradition shared by virtually the rest of the world. Asking around among my friends from other countries, I find the following usages: Pīking (Hindi), Peking (Hebrew), Pekin (Persian), Bīkīn (Arabic), Pekin (Polish), Peking (Czech), Pechino (Italian), Peking (Swedish), Pekín (Spanish), Pekino (Greek), Pékin (French), Pekin (Russian), and Peking (German). [VHM: Pekin ペキン (Japanese), Beijing 베이징 (Korean), Bắc Kinh (Vietnamese).] It is obvious that it is not simply because we are perverse that we insist on maintaining the traditional pronunciation which the northern Chinese have themselves given up [VHM: chiefly through palatalization of the velars] during the last few centuries.
Peking has been a part of our heritage since at least the time of immortal Milton who wrote:
Of mightiest Empire from the destined Walls
Of Cambalu, seat of Chathaian Can,
And Samarchand by Oxus, Temir's throne,
To Paquin of Sinaean Kings…. (Paradise Lost, XI.387-390)
Things may have changed in the decades since Bosat Man wrote the above, but the propensity for "Pekin(g)" over "Beijing" is still unmistakable.
Etymology and pronunciation
běi 北 ("north")
From Proto-Sino-Tibetan *ba (“to carry (on back), shoulder”).
The sense of “north” is derived from “back (of body)”: “back” → “to turn the back to; to retreat” → “north”.
The ancient Chinese value the southern direction and houses are traditionally oriented along a north-south axis, as evident in the fengshui theory and orientation of buildings in Chinese Neolithic sites. North is the direction the back is oriented to when the person is facing south.
- Mandarin
- Cantonese
- (Guangzhou, Jyutping): bak1
- (Taishan, Wiktionary): bak2
- Gan (Wiktionary): bet6
- Hakka
- Jin (Wiktionary): bei2 / bieh4
- Min Bei (KCR): bă̤
- Min Dong (BUC): báe̤k
- Min Nan
- Wu (Wiktionary): poq (T4)
- Xiang (Wiktionary): be6
- Dialectal data
- Middle Sinitic: /pək̚/
- (Baxter–Sagart): /*pˤək/
- (Zhengzhang): /*pɯːɡ/
(source)
jīng 京 ("hill [> capital]")
- Proto-Sino-Tibetan *s-gaŋ (“hill, ridge, mountain”) (STEDT), cognate with 岡 (OC *klaːŋ, “hill, crest”), 鯨 (OC *ɡraŋ, “whale”), Tibetan སྒང༌ (sgang, “ridge; mound”), Burmese ခင် (hkang, “ridge”), Burmese အခေါင် (a.hkaung, “summit”), Lahu qhɔ³³ (“mountain”), Sulung gɹaŋ³³ (“mountain”), Proto-Tamangic *ᴬgaŋ ("hill, mountain"); or
- of Austroasiatic origin, compare Proto-Mon-Khmer *graŋ (“hill, peak”) (Schuessler, 2007), whence Mon ဂြၚ် (krɛ̀aŋ, “peak”) and Old Khmer grāṅ (“hill on flat land”).
- Mandarin
- Cantonese
- (Guangzhou, Jyutping): ging1
- (Taishan, Wiktionary): gen1
- Gan (Wiktionary): jin1
- Hakka
- Jin (Wiktionary): jing1
- Min Bei (KCR): géng
- Min Dong (BUC): gĭng
- Min Nan
- Wu (Wiktionary): jin (T1)
- Xiang (Wiktionary): jin1
- Dialectal data
- Middle Sinitic: /kˠiæŋ/
- (Baxter–Sagart): /*[k]raŋ/
- (Zhengzhang): /*kraŋ/
(source)
There is a world of difference between "Pekin(g)" and "Beijing". They may denote the same modern city which is the capital of the People's Republic of China, but their connotations — culturally, historically, and linguistically — could hardly be more different.
Selected readings
- Bosat Man, "Backhill / Peking / Beijing", Sino-Platonic Papers, 19 (June, 1990), 1-6 (free pdf) — highly recommended for anyone interested in the history and ideology behind the name and concept of "Beijiing", with global comparisons.
- "Surprising Transformations of a Beijing Street Name" (1/29/11)
- "Jingle bells, pedophile" (12/7/09)
- "How they say 'Beijing' in Beijing" (8/18/08)
- "On swallowing and slurring in Pekingese" (5/3/19)
- "OMG moments induced by allegro forms in Pekingese" (1/26/12)
- "Acquiring literacy in medieval Dunhuang" (2/20/21)
J.W. Brewer said,
October 5, 2021 @ 4:24 pm
The spelling "Pekin" for the Chinese capital (under Manchu rule) is quite common in 19th-century books in English, so it would not have seemed oddly French-like when the Illinois town was settled. There are a number of other American states with a "Pekin," presumably from the same era. One 1884 traveler's guide gives "PEKIN, PE-KING, or PE-CHING" as possible spelling options. My best guess would be that usage in English did not definitively standardize around "Peking" until the late-Imperial "postal map" romanization system chose that spelling in the 1890's, and even then it took a while for it to become uniform — with a detour through "Peiping" during the temporary period when the KMT wanted to emphasize that Nanking was the capital and that place up north wasn't.
Michael Carasik said,
October 5, 2021 @ 4:32 pm
Perhaps worth mentioning that (as I was taught long ago) Pekin, Ill. got its name because it was assumed to be directly opposite the Chinese city on the other side of the globe. I believe Canton, Ohio got its name the same way.
cameron said,
October 5, 2021 @ 6:35 pm
Hmm. That story about Pekin IL, and Canton OH being so called due to the idea that they were located "directly opposite" the corresponding Chinese cities is fishy in the extreme. Obviously the antipodean points for those locations would have to be in the southern hemisphere
John Rohsenow said,
October 5, 2021 @ 6:48 pm
In explaining "Peking/Beijing " over the years I have referenced the complications of the older Wade-Giles romanization system widely used during the late 19th and more than first half of the 20th century, which
(more accurately) expressed the UNaspirated voiceLESS initial stop series in (what was to become) standard Mandarin by using p-,t-,k- (with no additional annotation) vs. the (phonemically contrastive) ASPIRATED voiceLESS initial stops as p'-,t'-,k'-, so that 北 "north" (w/ UNaspirated voiceless initial consonant) was written pei vs. 陪 p'ei 'accompany' (w/ ASpirated voiceless initial consonant). Unfortunately Western typists for simplicity often omitted these aspiration marks, giving us Pe(i)king, Ching (not Ch'ing) dynasty, Tang (not T'ang) dynasty, and Kaifeng (not K'aifeng) City.
Then the final version of Hanyu Pinyin romanization adopted by the PRC in the 1950s decided (for the convenience of English/Western readers/writers?) to represent the voiceLESS UNaspirated initial stop
series using b-.d-,g-, and keep p-,t-,k- for the voiceLESS ASpirated series; thus, Beijing, GuominDang (vs. W-G: KuominTang), etc.– Therefore,the "B" in modern HYPY spelling "Beijing" for native speaker/readers of northern Mandarin Chinese actually does represent the same voiceless unaspirated initial sound as used to be represented by the letter P- in Wade-Giles and other earlier systems. If English speaker/readers happen to mispronounce B- as a VOICED initial consonant, that is irrelevant in China (and in fact may not be perceived by native Chinese speakers for whom the voiced/voiceless distinction is not phonemic anyway :-)
Jim Breen said,
October 5, 2021 @ 7:34 pm
Half a century ago I worked for where there were also staff from the US's NSA. One NSA colleague related how they had been required to call that Chinese city either "Peking" or "Peiping". Use of "Beijing" was discouraged on the grounds that it was "giving comfort to the enemy". (This was before Nixon made his famous visit; I dare say things changed after that.)
Like many people, I am all over the place in the pronunciation of city names. I use the usual English styles for Paris and Tokyo, but I cannot bring myself to say "Lions" for Lyons, or "Kie-o-toe" for Kyōto. And if I used the local pronunciations of Gothenburg or Copenhagen no one would know what I was talking about. I must say that the Japanese tend to do these things fairly well, with their イェーテボリ and ケーペンハウン.
Jim Breen said,
October 5, 2021 @ 7:37 pm
My previous message began "Half a century ago I worked for [censored] where …" but I made the mistake of putting the word "censored" in the wrong sorts of brackets, so it was, er, censored.
Laura Morland said,
October 5, 2021 @ 8:37 pm
I was studying Old English with my mentor, Prof. Alain Renoir, when the change from "Peking" to "Beijing" was put into effect.
Renoir deeply regretted that students in the future would not recognize the word "Peking" in older texts. It's a loss with no perceptible gain. As he pointed out, "An English speaker who says 'Beijng' is no closer to the native pronunciation than 'Peking'."
(Thanks to John Rohsenow, now that I know that "b" is used to represent "the voiceLESS UNaspirated initial stop," it would seem that we English speakers have taken a step backwards on the first syllable.)
Curious: why does the change only affect English? French retains Pékin; have other Western nations been forced to change their spelling of China's capital city?
Claw said,
October 6, 2021 @ 12:02 am
Laura Morland:
I disagree here. In English, while initial "p" and "b" generally distinguish between voiceless and voiced, they also distinguish between aspirated and unaspirated – i.e., "p" represents [pʰ] while "b" represents [b]. Mandarin Chinese, however, does not make a phonemic distinction in voiceless and voiced consonants; rather, the phonemic distinction is in aspiration, so "p" represents [pʰ] while "b" represents [p].
While representing "Beijing" with a "b" initial does cause English speakers to pronounce it as a voiced [b], more importantly it causes them to pronounce it as unaspirated, so Chinese speakers would perceive it as being close enough to [p], since voicing is not a distinction that is made in Mandarin, as John Rohsenow also noted. Continuing to write it with a "p" initial would instead cause English speakers to pronounce it as [pʰ], which because aspiration is a phonemic distinction in Mandarin, would sound obviously wrong to Chinese ears.
For this reason representing the initial consonant as "b" is arguably the better choice for English speakers. It gets them to pronounce the name in a way that is closest to the Mandarin Chinese pronunciation while still following English phonotactics.
I can only speculate on the reasons, but I will note that the initial "p" in French is pronounced without aspiration in French, so it does end up matching the Chinese unaspirated [p].
John Swindle said,
October 6, 2021 @ 1:33 am
@cameron: Indeed, Pekin, Illinois, isn't precisely opposite Beijing on the globe. They should have named their town "Indian Ocean, Illinois," but the notion that China was on the other side of the world was a commonplace.
Thomas said,
October 6, 2021 @ 2:07 am
With regards to the Korean name of Beijing, I am wondering when the name switched from the traditional reading of the hanja as 북경 to the phonetical 베이징. It seems to me that this is the analogous change to the English one, Peking -> Beijing. Perhaps someone familiar with the subject can enlighten me?
Victor Mair said,
October 6, 2021 @ 6:04 am
From Juha Janhunen:
I find the use of the English spelling "Beijing" just another case of Anglosaxon-North American political correctness. It is not only historically incorrect, but it also results in English-speaking mouths in the totally incorrect pronunciation [bei'ʤiɪŋ]. Fortunately, in Finnish the form "Peking", pronounced ['pekiŋ], has survived. Some years ago the Finnish national carrier Finnair tried to introduce "Beijing", pronounced ['beijiŋ] or ['beiʤiŋ] (with the two sounds [b] and [ʤ] both being alien to Finnish), but this practice was never adopted by the speakers, and Finnair has returned to using "Peking".
Chris Button said,
October 6, 2021 @ 6:29 am
@ Claw
A French "b" and a English "b" are not usually the same unless the English "b" undergoes intervocalic voicing or the like.
Victor Mair said,
October 6, 2021 @ 6:31 am
From dako-xiaweiyi:
Cambodia became Kampuchea.
Rangoon became Yangon.
Bombay became Mumbai.
Christmas Island became Kiritimati.
In each case it was the foreign rendering of the name that changed. The locals did not change the way they said it.
Jake said,
October 6, 2021 @ 7:47 am
@laura:
> Renoir deeply regretted that students in the future would not recognize the word "Peking" in older texts
Did he have such a low opinion of students that he thought they were fundamentally unable to remember that 'Beijing' used to be 'Peking'?
Coby Lubliner said,
October 6, 2021 @ 8:15 am
I've always wondered why Beijing isn't Hokkyo in Japanese. After all, Tokyo is Dongjing in Chinese.
David Marjanović said,
October 6, 2021 @ 8:22 am
In English, the difference between /p t k/ and /b d g/ is mostly aspiration as you get closer to the beginning of the word, so that /b d g/ are usually completely voiceless there while /p t k/ are reliably aspirated (though still less so than in Mandarin), and mostly voice at the end, where /p t k/ are normally unreleased, so they can't be aspirated.
In French, /p t k/ are never aspirated, and /b d g/ are always voiced.
A French /p/ and a Mandarin /b/ are still not exactly the same, but speakers of both languages generally hear them as the same. That tiny difference is phonemic in a few Upper German dialects, in the Mandarin dialect of Wuhan, and I think that's pretty much it.
Carl said,
October 6, 2021 @ 8:43 am
Cambalu is the same as Paquin.
Chris Button said,
October 6, 2021 @ 9:46 am
@ David M
The distinction between – p and -b at the end in English is more to do with preceding vowel length due to pre-fortis clipping.
David C. said,
October 6, 2021 @ 9:54 am
@ David Marjanović
As a speaker of both languages, I can certainly notice the difference but unable to describe it. What would you say are the distinguishing features?
I find your comment that speakers of both languages hearing them as the same quite fascinating. For me it's one of the most noticeable markers of a Mandarin-speaking L2 learner of French or Spanish (and really, Southern Chinese topolect speakers as well). When intending to pronounce Pékin, many Chinese speakers end up saying what sounds almost like Béguin.
I once saw a video on Youtube where a Mandarin speaker was saying one of the trickier points in French pronunciation was distinguishing "cadeau" and "gâteau".
Rodger C said,
October 6, 2021 @ 9:57 am
@Carl: So say the Chineses.
Antonio L. Banderas said,
October 6, 2021 @ 10:19 am
@Chis
"Is" to do with ? https://en.wiktionary.org/wiki/to_do_with#Preposition
@everybody
https://en.wikipedia.org/wiki/Help:IPA/Mandarin
MattF said,
October 6, 2021 @ 11:53 am
I recall my first trip to Europe— I was in a train station and noted that a significant fraction of the eastward-bound trains were headed to a place called ‘Wien.’ Puzzling, at least at first.
Philip Taylor said,
October 6, 2021 @ 1:07 pm
"a place called ‘Wien’" — I had a similar experience when I booked a flight (online) to Nürnberg and received in return a ticket with destination Nuremberg. I had to call the airline to ask if they had made a mistake …
Victor Mair said,
October 6, 2021 @ 1:37 pm
From Bob Ramsey:
Well, I'd just like to say that I wholeheartedly agree with Juha Janhunen's comment. There is a bit of political correctness, even smugness, about Americans' insistence on using what they think of as the Chinese pronunciation. And it drives me up the wall to hear people in the media pronounce the name of the Chinese capital as [bei'ʤiɪŋ] (using Janhunen's phonetic transcription). Argh! That consonant they put at the beginning of the second syllable exists neither in Chinese OR English–except in the speech of some Americans' attempts to Frenchify their vocabulary. Ugh! A couple of years ago, a major TV news source (NBC?) had a running on-air discussion of how the word should be pronounced, and the anchors even brought in a native speaker to teach them how to pronounce it! And that worked for a while. But even so, it didn't take too long before most went back to their old over-pronouncing habits. As my old colleague and good friend Ron Walton always told people about the pronunciation: Just pronounce that second syllable like the jing in jingle!
Oh–and I might mention that I also liked what Thomas said about the Korean pronunciation. I'd just add that the pronunciation 북경 [VHM: Buggyeong] he mentioned is still around, though admittedly most of the users of that pronunciation are of an older generation. –But then again, a few old-timers and traditionalists give both Chinese AND Japanese(!) names written with Chinese characters Koreanized readings! My old colleague and mentor Lee Ki-Moon did so quite deliberately, especially with distasteful Japanese names–such as Toyotomi Hideyoshi (豊臣秀吉), whom he always called 풍신수길 [VHM: Pungsinsugil]!
Alexander Browne said,
October 6, 2021 @ 3:34 pm
I'm confused…
Isn't that what is written? Are you thinking [ʒ]?
Alexander Browne said,
October 6, 2021 @ 3:38 pm
OK, as I posted that I saw the two vowels, i and ɪ. Is that the issue?
Jake said,
October 6, 2021 @ 3:46 pm
Bob Ramsey:
> [bei'ʤiɪŋ] … That consonant they put at the beginning of the second syllable exists neither in Chinese OR English–except in the speech of some Americans' attempts to Frenchify their vocabulary
As an American who doesn't think he's attempted to Frenchify his vocabulary, it comes as quite a surprise to me that [ʤ] doesn't exist in English.
If, as Alexander Browne suggests, you had meant [ʒ], that's almost as big a surprise, I might say it was 'immeasurable' but I'll need some help in the middle there to avoid being too Frenchie.
Bathrobe said,
October 6, 2021 @ 10:19 pm
Much of this thread seems to me to be somewhat wide of the mark.
* The spelling "Beijing" is pinyin. That's all. The choice of "Beijing" is due to the conventions of the pinyin script. There is no way you can accurately represent the Chinese unvoiced aspirated bilabial in English. The old Wade-Giles choice ofp vs p' was problematic since people kept leaving the apostrophe out, as someone mentioned. b vs p, as adopted in pinyin, seems a reasonable solution.
* What is problematic is the switch from the old spellings to pinyin. This is a more interesting issue.
– The switch to pinyin by the Anglophone world (sorry, not just the US) seems to me to be a response to pressure from the Chinese side. "Since pinyin is official in China, you should follow it, otherwise you are just a bunch of colonialists". "Oh, ok, we'll respect you guys…"
– The main reason that non-Anglophone countries don't feel obliged to follow is, I suspect, because the Chinese just aren't so obsessed with other languages. English is the all-important one, and getting the Anglophone world to adopt "Beijing" is the ultimate victory. The other, less important languages can carry on as before; it's not that important to keep pestering them to change.
* Everyone here is commenting on the switch from "p" to "b". The more significant shift is from "k" to "j". That represents a real difference of pronunciation in Chinese. My understanding is that at the time that "Peking" was adopted, the pronunciation /kiŋ/ for 京 was apparently much more current in China, even in northern dialects, than it is now. (This is important, although Bosat Man mentions it only in a very general sort of way, amidst all the other factoids he introduces.) Not being a dialectologist or historian of Chinese I don't have sources for this, but this really does appear to be a case where the pronunciation /ʤiŋ/ has replaced the old-fashioned or regional pronunciation /kiŋ/. It's not just due to the adoption of pinyin.
* The pronunciation of "j" as /ʒ/ by many English speakers is due to the vague belief that "j" should be pronounced the French way. It is actually a step too far. /k/ to /ʤ/ is justified in terms of Chinese historical phonology. /ʒ/ is an unintended side effect of the choice of "j" to represent the Chinese sound /ʤ/. (In fact, in a narrow phonetic transcription, the sound is not /ʤ/ at all. More specialised phonetic symbols need to be used.)
So the issue comes down to 1. Why did the creators of pinyin decide to change the assignment of letters to represent the phonemes of Chinese? 2. Why did the pronunciation of the morpheme 京 change? 3. Why did the Anglophone world decide to cave in and accept pinyin to represent Chinese place names in English? 4. Why do English speakers feel obliged to pronounce "j" as /ʤ/?
Four separate questions.
Bathrobe said,
October 6, 2021 @ 10:29 pm
BTW, I am open to challenge on all four points above. It just seems to me that focussing only on the change from /p/ to /b/ is taking a very narrow view of the matter.
Bathrobe said,
October 6, 2021 @ 10:48 pm
I've always wondered why Beijing isn't Hokkyo in Japanese. After all, Tokyo is Dongjing in Chinese.
You may wonder. So do I. I suspect that:
1. Pekin was adopted from the Western languages. Alternatively, the Japanese had an existing tradition of pronouncing Chinese place names as they heard them. But I'm doubtful about the second alternative. I don't think that pronouncing 廣東 as Kanton in Japanese arose from some kind of tradition. It looks more like influence from English.
2. Japanese has a tradition of pronouncing Chinese characters flexibly. Not only do they have several different traditional readings of Chinese characters, they also have Japanese pronunciations for them. So the Japanese are more amenable to adopting a novel pronunciation to pronouncing 北京 as Hokkyō (or Hokkei).
3. China traditionally had a system of pronouncing characters according to the dialect. This didn't involve mixing different systems together; it meant that each system of reading was independent, allowing multiple systems to coexist without (too much) miscegenation. China has now pretty much done away with this tradition. Everything is standardised in Mandarin. Foreign names are pronounced in Mandarin. End of story.
4. There is a particular mentality in East Asia that "China is the centre; the rest are peripheral and imitative". The Chinese would never be convinced that they have to follow the lead of their followers. The Japanese, on the other hand, are aware that much of their culture is from China. It's easy to convince them to adopt Chinese pronunciations (as they now have for the leaders of China) on the grounds that the Chinese pronunciation is the "original and correct one". (The Chinese did adopt much Japanese-coined vocabulary, but only because they could convince themselves that the Japanese had borrowed it all from China in the first place and were just applying traditional Chinese concepts to the coinage of vocabulary.)
My 2 cents worth.
Andreas Johansson said,
October 7, 2021 @ 2:54 am
In Swedish, both Peking and Beijing is in common use. I'd guess the former is more common but that the latter is catching up, largely due to anglophone influence.
I see that the discussion page on the Sw. WP for another Chinese city largely consists of a discussion whether it should be called Kanton or Guangzhou. I don't believe I've ever seen the latter in a Swedish contexts except as a parenthetical addendum (like "Kanton (ki. Guangzhou)"), but some people there are claiming it's the commoner form in modern Swedish.
Victor Mair said,
October 7, 2021 @ 4:56 am
@Bathrobe:
"The more significant shift is from 'k' to 'j'."
Indeed!
As VHM / I mentioned in a bracketed comment at the end of Bosat Man's remarks, this was due to the palatalization of the velars. This explains the change from -king to -jing, which is both a historical and a regional shift, as briefly explained here:
"General Tso's chikin" (6/11/13)
http://languagelog.ldc.upenn.edu/nll/?p=4682
=====
Naturally, in the countless other Sinitic topolects, jī 雞 / 鷄 / 鸡 ["chicken"] has many other pronunciations. Here I will mention only Cantonese gai1 (sounds like "guy"), which will be familiar to those who frequent Chinatown restaurants. One might well ask how Mandarin jī and Cantonese gai1 are related and can both be legitimate pronunciations for 雞 / 鷄 / 鸡. Perhaps the simplest way to put it is that the evolution of jī from an earlier stage is part of the palatalization of the velars that started in the northeast over three centuries ago and is gradually moving southward, having reached about to the Yangtze River.
=====
See also the remarks by Chris Button here:
"Of horse riding and Old Sinitic reconstructions" (4/21/19)
https://languagelog.ldc.upenn.edu/nll/?p=42523
Frans said,
October 7, 2021 @ 7:57 am
@cameron
The antipode would be in or near Argentina, but if you look at the other side of the earth at the same longitude then at a quick glance it would appear at least plausible that you might end up in or near Illinois.
NB I'm not saying I think the story's true, but I find it bad form to summarily dismiss something obviously wrong when there's a not-so-obviously wrong steelman readily available. ;-)
J.W. Brewer said,
October 7, 2021 @ 7:57 am
I think bathrobe's point that the PRC authorities (and many other regimes as well) just *care* more about how their toponyms are spelled in English than how they are spelled in other languages, because English is now the de facto world lingua franca, is a significant one. Whether Anglophones (esp. "gatekeeper" institutions like the keepers of stylebooks for major publishing houses and newspapers etc.) ought to want to play along with that is a different question, of course.
Also lurking out there is the notion that for a given toponym written locally (i.e. in the standard script for the dominant local language) in a non-roman script, there ought to be one and only one romanization used worldwide. That often seems like an unexamined premise that could stand some examination. AFAIK, an earlier Communist regime was not particularly bothered by the fact that сове́т came out as "Soviet" in English/French but "Sowjet" in German and various other things (Sovjet, Sowiet, Sobiet, etc.) in various other Latin-scripted languages. As that example shows, different Latin-scripted languages may use the same basic set of glyphs but can have different conventions in how they match glyphs to phonemes, meaning there will never be a one-size-fits-all romanization system that is equally natural/intuitive to everyone worldwide regardless of their L1.
JOHN W BREWER said,
October 7, 2021 @ 8:12 am
What you might call the "semi-antipode" of Peking, i.e. the same latitude north of the equator but the opposite side of the world from a longitude perspective, is gonna be on the floor of the Atlantic roughly south of Nova Scotia and east of Cape Cod. So pretty close to North America but still wet. That said, the folkloric assumption that if you dug a hole straight down you would eventually come out in China was still current (if not necessarily taken literally) when I was a kid in the 1970's, and there are plenty of places in the U.S. where that sort of "semi-antipode" will give you a dry-land location within the current territory of the PRC. The semi-antipode of my house just outside New York City looks to be in Inner Mongolia, maybe an hour's drive west of Bayannur.
J.W. Brewer said,
October 7, 2021 @ 10:21 am
And just as a helpful reminder that orthographic diversity also affects less prominent toponyms, far to the west of Peking/Beijing is a sleepy little border town that may be rapidly becoming of greater economic and geopolitical significance because it is one of two places where the PRC's railroad network now connects to Kazakhstan's part of the former Soviet network, thus making it possible to send freight from China all the way to Europe by rail rather than by ship. Available romanized spellings of that toponym include (per wikipedia) Khorgas, Korgas, Chorgos, Gorgos, Horgos, Khorgos, and possibly Qorǵas. And those are just attempts to romanize the Kazakh name (sometimes as filtered through Russian) – the actual Mandarin variation of the name gets pinyinized as Huò'ěrguǒsī.
David Marjanović said,
October 7, 2021 @ 11:16 am
That's in any case an important factor that I should have mentioned, thanks.
Well, um. The IPA has neither signs nor terms for this. The only available terminology is "fortis" vs. "lenis", though those have been used for several different things and were originally coined for a pure length distinction in Swiss German…
It's a spectrum, not a binary feature, but I bet no language distinguishes more than two points on it.
The "fortes" are made with greater air pressure. The audible difference is entirely in the release; unreleased plosives cannot support this distinction. Both voiceless pulmonic and ejective plosives can support it (Caucasus-area ejectives are lenes, Navajo ejectives are pretty extreme fortes). Voiced plosives apparently cannot. Fricatives can, even voiced ones (the /v/ of those Dutch accents that haven't merged /v/ into /f/ seems to be a voiced fortis). Voiceless affricates can, both as a unit and on the stop alone.
I think I got my two or three anecdotes confused, actually. The person who really heard my /d/ as her /t/ was Russian, not French. The native French speaker(s?) I got into that situation were confused instead.
I'm surprised that's unknown to Wiktionary; it's widespread in Britain. It's a reanalysis, not a misspelling (I've seen "will be nothing to do with" or something like that). Probably it started from "that's nothing to do with", but it may have been made easier by general h-dropping.
Too simple for what we're talking about.
However, the explanation of Pinyin g as the plosive in English skin is accurate: in most Englishes, that's reliably a voiceless lenis, not a fortis. Aspirating northern German accents do that, too.
European railway custom is to never translate the names of stations. I think they just couldn't agree on which exonyms to use and how many to use.
In many Englishes b is accurate word-initially or at least utterance-initially before a vowel – it's just not quite reliable.
Frans said,
October 7, 2021 @ 11:54 am
@John W Brewer
That far out? I did a bad job quickly eyeballing it then; I thought it seemed to be in the vicinity of Pennsylvania or New York. Thanks! (Is there by chance a website that quickly does a calculation for you? I took a quick peek but only found ones focusing on true antipodes.)
My point exactly. You can clearly see on any globe that the US is at the very least in the ballpark.
Chris Button said,
October 7, 2021 @ 2:48 pm
@ David M
Fortis and lenis is like tense and lax on vowels. The common thread comes via a length distinction. It almost certainly (in my opinion) is what lies behind the Old Chinese bifurcation of syllables into type A and type B. The Mon-khmer evidence marshaled by Ferlus and Pulleyblank is instructive in that regard. However, confusion over how it works and what the terms mean has resulted in that not being properly recognized in Chinese historical linguistics.
Kenny Easwaran said,
October 7, 2021 @ 3:26 pm
This question of whether to keep a traditional spelling and pronunciation of a foreign place name in English, or to replace it with (the transliteration of) the current local spelling and pronunciation comes up in some other places as well.
I seem to recall that just over a decade ago, when the winter olympics were there, there was a campaign to change "Turin" in English to "Torino", but when I look at Google Maps right now, it still shows "Turin" along with "Rome", "Milan", "Florence", "Venice", and "Naples". However, for the small town on the coast near Pisa, it shows "Livorno", where English maps a century ago might have shown "Leghorn". I think "Syracuse" for "Siracusa" is probably the smallest Italian city that still has its own English spelling, probably because there's a city in upstate New York that preserves the anglicized version of the ancient Greek name.
It seems notable to me that there are very many examples of this with German and Italian city names, but very few with French or Spanish (apart from the pronunciation). I guess Chinese names are making the switch from one category to the other. I don't know if there's an explanation for why place names in one language go one way rather than the other. ("Oporto" for "Porto" and "Lisbon" for "Lisbõa" seem to be examples in Portuguese.)
Scott P. said,
October 7, 2021 @ 3:35 pm
I think "Syracuse" for "Siracusa" is probably the smallest Italian city that still has its own English spelling, probably because there's a city in upstate New York that preserves the anglicized version of the ancient Greek name.
Croton would also fall under that category.
J.W. Brewer said,
October 7, 2021 @ 3:48 pm
I think it's correct (and maybe is evidence of something or other) that there don't seem to be too many Spanish toponyms with their own English versions, although Seville and Saragossa are examples of the phenomenon. Wikipedia claims that "Minorca" is merely "the older English name" for "Menorca," suggesting a transition. I don't think I've noticed that latter spelling in English although maybe I don't read too many contemporary texts on Balearic subjects. See also Majorca v. Mallorca.
Bathrobe said,
October 7, 2021 @ 6:45 pm
Available romanized spellings of that toponym include (per wikipedia) Khorgas, Korgas, Chorgos, Gorgos, Horgos, Khorgos, and possibly Qorǵas. …. the actual Mandarin variation of the name gets pinyinized as Huò'ěrguǒsī.
I assume the Chinese would prefer us to write "Huoerguosi"?
In order to right past "wrongs" China boldly asserts to the world that it should respect "China" and "Chinese", which actually means trampling over a lot of other languages and ethnicities under Chinese control. Similar issues exist with regard to places like Erlian (Ereenhot) and Ganqimaodu (Gants Mod), not to mention Kashi (Kashgar) and Xigazê (Shigatse).
David C. said,
October 7, 2021 @ 10:43 pm
I would distinguish ignorance by translators, officials and the like from official policy. There is often no authoritative translation of place names in China, so Hanyu Pinyin is the easy way out. I am speaking about Roman alphabet transliterations here, rather than the treatment of ethnic minority languages in practice.
中国地名汉语拼音字母拼写法 (Hanyu Pinyin Phonetic Spelling of Geographical Names in China), policy in force since 1976, actually refers to a separate document for place names of ethnic minority areas, the 少数民族语地名汉语拼音字母音译转写法. The document states that transliterations are not bound by the name in Putonghua.
A couple of examples come to mind – it's almost always Ürümqi, Hohhot, Inner Mongolia, Lhasa, Nyingchi, Tibet even in English-language state media.
Here's a Xinhua article from earlier in the year where they use Horgos, not Huoerguosi.
Jonathan Smith said,
October 7, 2021 @ 11:27 pm
@David Marjanović
Did not know this contrast in (some) Upper German… interesting to listen to. But surely not in Wuhan?
Across languages there are of course endless examples of small differences like French "p-" vs. Mandarin "b-"… word-initial "b-" in (much) English vs. Mandarin "b-" are not a bad match as you say, but also plainly different with the latter more "fortis".
John Swindle said,
October 7, 2021 @ 11:35 pm
Why do we write or try to say Beijing instead of Peking? Is it Chinese Communist Party's fault, like the evil government of Ivory Coast making us say Côte d'Ivoire? Are we bending over backwards to please? If so, why do we say Moscow instead of Moskva? And is it Taipei, Taibei, Taipeh, or …? If the Canadians decide to call Port Arthur and Fort William (or is it the other way around?) Thunder Bay, do we have to do the same thing? Some of my ancestors and relatives changed their foreign-sounding names on coming to America. Others did not. Is there a theory here? A doctorate to be won?
Philip Taylor said,
October 8, 2021 @ 5:45 am
John — "Some of my ancestors and relatives changed their foreign-sounding names on coming to America. Others did not. Is there a theory here? A doctorate to be won ?" — I asked that very question of an American professor who has a surname which is virtually unattested as she spells it (without diacritics) but frequently attested in its diacriticised form, where it is the surname of a well-known athlete. She did not reply.
Jim M. said,
October 8, 2021 @ 6:03 am
There's a Pekin restaurant in Detroit—been there for ages.
John Swindle said,
October 8, 2021 @ 6:48 am
@Philip: I suspect that there are patterns that will continue to elude us for a while and will prove less than earth-shaking when resolved. Probably every place and every person has been called more than one name at some time or another, whether in one language or many, not even counting pronouns ("they") or expletives ("#!@*&%").
J.W. Brewer said,
October 8, 2021 @ 8:33 am
I guess maybe the way to harmonize Bathrobe's cynicism and David C.'s accurate description of the status quo for toponyms in Sinkiang* is to assume that the CCP may at any moment change its mind with regard to any policy (including but certainly not limited to how local toponyms ought to be romanized) involving the supposedly "autonomous" regions whose historical populations are of non-Han ethnicity. And then the question becomes whether if and when the Beijing regime announces it now prefers Wūlǔmùqí to Ürümqi the journalists/academics/bureaucrats of the Anglophone world will do as they're told and change their own usage.
*Note that the older postal-map romanization of the Mandarin name for East Turkestan features the same k-versus-j issue when compared to current Communist usage as is seen in Peking/Beijing and Nanking/Nanjing.
Robert Ramsey said,
October 8, 2021 @ 12:42 pm
First, a belated note: "Jake" is quite right on both counts. Yes, when I lifted Juha Janhunen's transcription [bei'ʤiɪŋ] to cite in my original message to Victor, I did indeed mean the use of the fricative [ʒ] and NOT the affricate [dʒ]. And yes, that fricative does occur in lots and lots of English words, just as Jake points out. Though originally an innovative pronunciation borrowed from French, that voiced fricative is of course very much a part of English today. So, apologies to all! –But Jake, I'm betting that you yourself don't use it in the pronunciation of "Beijing".
In any case, my point was to raise a voice in support of Juha Janhunen's conjecture about "political correctness." I suspect in fact that all those (non-Chinese speaking) Americans who use the fricative [ʒ] in that word are trying to be politically correct and make the word sound less offensive, because they think pronouncing it like the "jing in jingle" (as Ron Walton suggested) sounds too much like old racist slurs like "ching ching Chinaman" for them to be comfortable using it. For them, [beiʒɪŋ] sounds ever so much more "cultured". And yes, more like the sounds of that most cultured foreign language of all, French.
Now, I admit that this suspicion about political correctness is completely impressionistic and speculative. Still, it does make me grind my teeth to hear [beiʒɪŋ], because that voiced fricative sound is so unlike anything in Mandarin, while the jing of the English word "jingle" is so much closer to the sounds of the actual Mandarin syllable–and so more natural for an English speaker to use!
–Oh, and on a slightly different tack: Did anyone mention the fact that a number of the years ago, the faculty of Peking/Beijing University voted overwhelmingly to use the term "Peking University" as the English name of their institution?
CuConnacht said,
October 8, 2021 @ 1:03 pm
The ducks are still White Pekins. The dogs seem to changing from Pekinese to Pekingese, at least in North America.
David Marjanović said,
October 8, 2021 @ 4:53 pm
Length is often (in many languages as well as in many environments within a single language) used to enhance fortis consonants and tense vowels, but by no means always.
Xigazê is not Chinese, though; it's from the Pinyin-based orthography of actual Lhasa Tibetan.
(For specifically the Lhasa topolect, the Pinyin-based orthography seems to work very well apart from ignoring the tones. What I don't know is how much it's actually used. The traditional solution of writing Written Tibetan and pronouncing it in the modern Lhasa sound system, which is similar to the worst excesses of English and French spelling, seems to be much more popular based on the very limited evidence I have.)
If you mean it would be interesting to listen to, I'll look for samples.
In this video, j is the same as elsewhere, but q is, to my great surprise, not aspirated. Yet it remains distinct from j. That's a pure fortis-lenis contrast.
Write the scream down as Dschadi! Dschadi! Tschömbusö dschadi!, give it to Schwarzenegger to read, and there's a good chance he'll get it exactly right.
Chris Button said,
October 8, 2021 @ 7:36 pm
@ David Marjanović
Regarding fortis/lenis, it’s a question of surface realization rather than “enhancement”.
On fortis/lenis obstruents, you might want to take a look at Ladefoged & Maddieson’s discussion in “The Sounds of the World’s Languages”.
Separately, you might also consider what tense/lax or fortis/lenis means for sonorants (which include “vowels” since consonant/vowel is a largely meaningless distinction) versus obstruents.
Jonathan Smith said,
October 8, 2021 @ 11:48 pm
@ David Marjanović I did some instructive youtube listening to Zürich German speakers… I would guess the issue re: arguably idiosyncratic quanbu in the Wuhan recording is your Zürich ear as opposed to a feature of the Mandarin dialect in question :D
David Marjanović said,
October 10, 2021 @ 4:27 am
I'm not from Zürich and have never been there, apart from changing trains at the train station there.
Swiss German has a length distinction (even word-initially) corresponding to my Central Bavarian fortis-lenis distinction.
I consider tense/lax an overarching description that describes how languages treat the organization of their sound systems; phonetically, the distinction may be composed of various combinations of actual vowel quality, length, tongue root position or whatever. Fortis/lenis, in contrast, is a matter of actual objective phonetics as far as I can tell, seeing as my native distinction involves neither length nor voice nor aspiration nor glottalization nor even necessarily prefortis clipping (there doesn't need to be a continuant preceding the consonant in question). It does not apply to voiced plosives, unreleased plosives, or any sonorants.
Chris Button said,
October 10, 2021 @ 5:48 am
@ David Marjanović
Tense/lax does not seem to directly correlate with things like advanced tongue root (so I have read in the past, but never looked into myself)
The fortis/length distinctions in Bavarian appears to have a length distinction as a salient characteristic (full disclaimer: I googled this since I have never looked into it in the past).
The terms “fortis/lenis” manifestly don’t seem to be used “objectively” as you claim. It’s rather a highly subjective term that is unfortunately thrown around a lot. Ladefoged & Maddieson state that fortis/lenis “may sometimes provide useful phonological labels for specifying a dichotomy when used language specifically.” Note how they say “phonological” not “phonetic” there. They do however talk about length and how it may be seen as the primary factor involved in many cases (most notably even when analyzed a secondary factor by others). That’s the unifying thread that I use.
So, I apply tense/lax or fortis/lenis in Old Chinese to refer to the syllable (moraic) weight that led to the A/B dichotomy in syllable types. I usually follow Pulleyblank in marking the distinction with a falling or rising accent, but it would be clearer in terms of salient surface characteristics to just mark length on the nucleus or the coda (for the coda, it might help you to think of how Japanese treats “n” as one mora). So for example:
亭 ⁿdaɲː
成 ⁿdaːɲ
To be clear, a final stop coda can’t have length specified, and the length distinction is not assumed to be the result of any compensatory lengthening.
Any fortis/lenis distinctions specifically on the onset in Old Chinese don’t directly pertain to the moraic approach to syllable weight I adopt from Pulleyblank. However, Michel Ferlus does make some interesting Mon Khmer comparisons in that regard, which tie in with some of the observations made by Pulleyblank on how a Mon Khmer register distinction has a similar effect on the nucleus as his notions of syllable weight in the evolution of old to Middle Chinese had on the nucleus there.
Scott P. said,
October 10, 2021 @ 12:00 pm
I think it's correct (and maybe is evidence of something or other) that there don't seem to be too many Spanish toponyms with their own English versions, although Seville and Saragossa are examples of the phenomenon.
There are some others: Cordova/Córdoba. Corunna/La Coruña.
Moa said,
October 10, 2021 @ 5:07 pm
In Swedish, Beijing is the most common spelling in general use. Peking is more common in specialized contexts. It's not a very new change either, I have a book from 1990s that refer to the city as Beijing but spells it Peking in Peking opera and Peking man. Shanghai is spelled like that. I'm not sure about Kanton. In historical contexts Kanton is the most common, but in modern contexts Guangzhou is also very common. So perhaps it's changing?
I doubt it's a big problem in Sweden that readers only recognize one of the spellings. If nothing else, European readers are very used to European cities having different spellings and pronunciations. If we can learn Copenhagen and København, Vienna and Wien, surely Beijing and Peking is a piece of cake.
I also suspect distance has something to do with changing spellings. It's easier to change the name of a place fewer people know well, like Kanton/Guangzhou, as opposed to more familiar places like Köpenhamn/Copenhagen/København,
Jonathan Smith said,
October 11, 2021 @ 7:32 am
Sorry D M, your Central Bavarian ear. The distinction you note between onset length and fortis/lenis per se is interesting but not drawn in the literature I looked at, e.g., this paper ("voiced plosives are lacking in Zurich German (as in all Alemannic varieties spoken in Switzerland). Instead of the feature [±voice], these dialects display a so-called fortis/lenis distinction" which is "based on closure duration." @Chris Button I like writing the OC distinction this way as well, but of course the devil is in the details of development towards the (more) attested Han/MC contrasts, which I don't think are dealt with great-ly on Ferlus's model.
Andrew Usher said,
October 11, 2021 @ 7:41 am
As for changing the English version of place-names: surely racism has something to do with it. That is, those countries' (correct) perception of our desire to avoid seeming 'racism'. That's why it's always countries we'd consider non-white that make a big deal out of it – they know they can use it to demonstrate power over us, which we ourselves concede by 'racism phobia'.
Another reason the Soviet Union never cared what their places were called in English was that they refused to ever recognise English as the global language: Russian always had to be at least equal for them, and they could (or thought they could) enforce that.
k_over_hbarc at yahoo.com
Stephen Reeves said,
October 11, 2021 @ 6:34 pm
When English Language PRC publications such as China Reconstructs and Peking Review, changed to Pinyin in the lat 70's , they also had Xiangang for Hong Kong and Xizang for Tibet, but changed back a few years later ,at least Taiwan is the same in old and new
Chris Button said,
October 11, 2021 @ 9:41 pm
@ Jonathan Smith
Thanks, yes I think the length marking is a more useful way of representing the distinction too. I suppose tense vs lax vowel symbols could work too (as I opted for when looking at Kuki Chin languages—Sizang Chin being the basis for Pulleyblank’s analysis), but the codas don’t have any such counterparts without without resorting to diacritics.
Regarding the shift to Middle Chinese, Pulleyblank’s insertion of /ɨ/ in type B syllables with its associated changes provides a solid phonological account. The issue is where the /ɨ/ came from. The good thing is that Pulleyblank’s and Ferlus’ proposed analogies with Mon Khmer mutually reinforce each other there, albeit from a very different starting point.
David Marjanović said,
October 12, 2021 @ 3:42 pm
As far as I know, tense/lax is a phonological descriptor that has been applied to different things in different languages. In English, the tense vowels are more peripheral, generally longer, and more likely to be diphthongs than the lax vowels.
No, that's Alemannic, not Bavarian: the Alemannic length distinction, which I find very conspicuous, corresponds to the Bavarian fortis/lenis distinction (or what's left of it after a few lenition phenomena)
Bavarian also has a length distinction, because consonant length has not been lost in Bavarian, unlike in Central and Low German (…though the mentioned lenition phenomena mean that minimal triplets for short lenis/short fortis/long fortis are not available). But that corresponds to nothing in Alemannic.
(Long lenes, which sound like the Alemannic "fortes", occur only across morpheme boundaries in Bavarian.)
Well, yes; but (unlike tense/lax) it can be, and occasionally is, used for a distinction that very few lects use without backing it up with voice, aspiration, length or glottalization, and that's how I use it, even though that's not what it was originally coined for – that, funnily enough, was the Alemannic length distinction.
Your proposal of nucleus vs. coda length is very interesting. But in any case it's unrelated to the way I use fortis/lenis.
Belarus.
German is changing from Weißrussland to Belarus as we speak.
Terry K. said,
October 12, 2021 @ 6:04 pm
@Robert Ramsey
I suspect in fact that all those (non-Chinese speaking) Americans who use the fricative [ʒ] in that word are trying to be politically correct and make the word sound less offensive, because they think pronouncing it like the "jing in jingle" (as Ron Walton suggested) sounds too much like old racist slurs like "ching ching Chinaman" for them to be comfortable using it. For them, [beiʒɪŋ] sounds ever so much more "cultured". And yes, more like the sounds of that most cultured foreign language of all, French.
Given the fact that people learn how to pronounce words from how they hear them pronounced, that seems exceedingly unlikely. In my case, I heard it pronounced "bayzhing", and so that's what I understood it's pronunciation to be. Sometimes, it's that simple. (Also I had no familiarity with those slurs.)
As for those who got their pronunciation from the spelling, rather than from what they heard, there's the issues that the [dʒ] sound is one we don't expect to find in words from other languages, and that it's quite ordinary for J in foreign words to not be pronounced like an English J, which makes the [dʒ] quite unintuitive to those of us not familiar with Chinese.
I suppose you could be correct about some people. But certainly not all, nor anything reasonably close to all.
Chris Button said,
October 13, 2021 @ 5:03 am
@ David M
Fortis/lenis and tense/lax are interchangeable terms. Over time the former has become associated with consonants and the latter with vowels, but that is neither here nor there.
+/-ATR is apparently separate,
As for Bavarian, I have no idea.
J.W. Brewer said,
October 13, 2021 @ 1:41 pm
It only takes a fairly modest amount of exposure to Chinese toponyms romanized in pinyin to notice that the "zh" digraph is frequently employed, which ought to suggest that it represents a different sound than the "j" does and thus that "pronounce 'j' like 'zh'" may be a bad guess, even if it may end up happening that the two different sounds are both a bit outside the consonant repertoire of English. Although it then also turns out that "zhing" isn't something you see in pinyinized Mandarin because of various phonotactic constraints, so you don't have a jing-v.-zhing minimal pair. But you do have e.g. the city Zhengzhou, formerly Chengchow.
All of that said, I think Terry K. is right that most Anglophones just pronounce "Beijing" the way they (think they) have heard other Anglophones pronounce it, with the range of possibilities they are likely to hear all being plausibly close enough to a foreign-looking orthography that none of them are going to be ruled out for that reason.
OTOH, I am puzzled by Terry K's assumption that people would affirmatively shy away from pronouncing "j" as dʒ because the latter isn't foreign enough. That's how it's pronounced by Anglophones in Japanese-origin loanwords (judo, ninja, emoji …), plus toponyms like Iwo Jima, and you have to have pretty specialized knowledge (for Anglophones) not to assume that Chinese and Japanese are closely related languages.
Andreas Johansson said,
October 14, 2021 @ 6:57 am
David Marjanović wrote:
German is changing from Weißrussland to Belarus as we speak.
Swedish is similarly shifting from Vitryssland to Belarus. It bugs me, because the usual reason given is to disassociate the place from Ryssland "Russia", which doesn't work very well when I know that the -rus part also means "Russia".
David Marjanović said,
October 14, 2021 @ 7:43 pm
Oops! Only for plosives. For fricatives, nasals and /l/, it corresponds to the Bavarian length distinction, because the extra dimension never existed there.
Fatally, both of them are affricates, and the closest English has to both of them is /dʒ/. Zh differs by being retroflex (and reliably voiceless), j differs by being dorso-palatal (and reliably voiceless). Z is the third affricate in the series (laminal, and maybe sometimes voiced at the end, but otherwise voiceless). The corresponding aspirates are ch, q, c.