PyeongChang: how do you say that in English?

« previous post | next post »

Should we say the name of the host city of the 2018 winter Olympics the way the Koreans pronounce it [pʰjʌŋtɕʰaŋ]?  Or should we say it more in accord with English phonetics?

The following article by Jane Han spells out the controversy clearly:

"NBC, read my lips – it's PyeongChang" (The Korea Times [2/18/18)

For Olympics viewers in the U.S., PyeongChang is not heard the way Koreans know it, but more like PyeongChay-ing.

This is because NBC, the official broadcaster of the 2018 Winter Olympics, has decided to stick to its own pronunciation where the "chang" in PyeongChang rhymes with sang, instead of tong.

Even though this is wrong, NBC's pronunciation policy seems to exist because it sounds "cleaner," as Sports Business Journal quoted Mark [sic] NBC Broadcasting and Sports chairman Mark Lazarus as saying.

But as the network's coverage of the PyeongChang Winter Olympics continues, Asian journalists and Korean viewers are not too happy with the way the correct pronunciation is being ignored.

"It really should not be difficult to pronounce PyeongChang if you put in the due effort to learn the correct way to say it," said Kevin Lee, a freelance journalist based in Los Angeles.

"If NBC is indeed purposely neglecting the accurate pronunciation because the wrong way sounds better, that's even worse because that is just disrespectful."

Dallas resident Jenny Kwon, who has been closely following the Olympics, says the inaccurate pronunciation of PyeongChang is annoying, but that may well be the American way.

"See how Americans pronounce IKEA," she said. "It's well known that the American way of saying the store name is wrong, but they continue to cling to their way. And there are so many other similar examples."

Noticing the inaccuracy and inconsistency in the way the host city's name is handled on air, Asian journalists have tried to promote correction.

Earlier this week, the Asian American Journalists Association (AAJA), the largest organization of Asian American journalists in the U.S., released a video demonstrating the pronunciation of PyeongChang.

"We've been getting a lot of questions about the correct pronunciation of the South Korean Olympic host city," said CeFaan Kim, AAJA's MediaWatch co-chair. "The correct pronunciation is PyeongChang…ang..like when you go to the doctor. Ah."

Despite growing criticism, an NBC Sports spokesperson told The Korea Times that the network was not intentionally mispronouncing PyeongChang.

"We have great respect for South Korea, its people, culture and language," the spokesperson said in an email response.

"When preparing for the Games, we heard a variety of different ways to say PyeongChang. Our commentators have no intent of saying it incorrectly or disrespectfully. We're trying to be consistent across our shows, while acknowledging that our on-air talent has slightly varying pronunciations, just as they do with English words.

"They are trying their best with Korean phonetics based on how they hear it, but inevitably it sounds different than if it were a native speaker."

I have added bold emphasis to the evaluative, judgemental terms used in the article.

Would you rather Americans say Běijīng [pèi.tɕíŋ] or Bayzhing or, for that matter, Peking (my preference, going back to the days of John MIlton, variants of which are still used in many languages around the world, and indeed in many Sinitic topolects?  See "Backhill/Peking/Beijing" (free pdf), Sino-Platonic Papers, 19 (June, 1990), 1-6.

In the interest of linguistic reciprocity, I asked a number of speakers of different languages how they say the names of various cities and places around the world (mostly American).  Their answers are often surprising and, in some cases, downright astonishing.

KOREAN

1.

Seattle 시애틀  Siaeteul

San Francisco 샌프란시스코  Saenpeuransiseuko

Atlanta  아틀란타 Ateullanta

New York 뉴욕  Nyuyok

Cleveland 클리브랜드  Keullibeuraendeu

Philadelphia  필라델피아 Pillidelpia

San Diego    샌디에이고   Saendieigo

San Jose    산호세    Sanhose   (Some people pronounce it as 새너제이 Saeneojei like American, but the official one is 산호세)

I used Revised Romanization.

2.

Seattle: 시애틀 (Shi ae teul)

San Francisco: 샌프란시스코 (Saen peu ran si seu ko)

Atlanta: 아틀란타 (At eul lan ta)

New York: 뉴욕 (Nyu yok)

Cleveland: 클리블랜드 (Keul li beul laen deu)

Missouri: 미주리 (Mi ju ri)

Michigan: 미시간 (Mi shi gan) – according to the McCune-Reischauer Romanization (Mi si gan is the romanization according to the Revised Romanization of Korean)

JAPANESE

Seattle Shiatoru シアトル

San Francisco Sanfuranshisuko サンフランシスコ

Atlanta Atoranta アトランタ

New York Nyūyōku ニューヨーク

Cleveland Kurīburando クリーブランド

Philadelphia Firaderufia フィラデルフィア

MANDARIN

Seattle Xīyǎtú 西雅图

San Francisco Jiùjīnshān 旧金山

Atlanta Yàtèlándà 亚特兰大

New York Niǔyuē 纽约

Cleveland Kèlìfūlán 克利夫兰

Philadelphia Fèichéng 费城

RUSSIAN

New York : Нью-Йорк

San Francisco: Сан-Франциско

Seattle : Сиэтл

Atlanta: Атланта

Cleveland: Кливленд

There is one place in the US that is often mistakenly translated in Russian: Silicon Valley as Силиконовая долина.

The word "Silicon" in English denotes a particular chemical element  Silicium, which should be translated in Russian as Кремний, therefore the correct translation of Silicon Valley in Russian should be Кремневая долина. The Russian word 'Силикон', which sounds similar to English "Silicon", means SiliconE, which is a chemical compound that is different SilicoN.

MONGOLIAN

New York is usually contracted into N’York, with a long ‘o’. The ‘a’ in Cleveland, since it is ae, people either say is as ‘ah’ or more frequently ‘eh’. I know a lot if people find the ‘v’ sound unnatural, especially those who didn’t study Russian, and call Virginia, Birjina/Berjina or Birjinia/Berjinia. Some people find ‘f’ unnatural, so Philadelphia would become Piladelp’.

TELUGU

I remember when I was trying to trace the pilgrimage route of Enugula Veeraswamy, documented by him in Telugu in a narrative entitled Kasiyatra Charitramu, on Survey of India maps.  There was one place in his Telugu narrative that I just couldn't seem to locate–it would have transliterated as Magadanaldu Chatram. Finally, after much difficulty, I realized it was McDonald's Choultry (Chatram and Choultry both are terms for a resthouse).

It's almost always the English words in Telugu script that I have the toughest time understanding!

MARATHI

sīyāṭala, sāna phrānsisko, aṭlānṭā, nyū yārk, klīvhlanḍ

The interesting one is always Newark, which is pronounced nevārk by many.

Marathi speakers have an interesting way of pronouncing v and w in English as vh.

TAMIL

English names are Tamilized in accordance with Tamil phonology. Most of the clusters are declusterized in Tamil.  Some of the place names in Tamil are as below:

siyāṭṭal sāṉ pirāṉciskō aṭlāṇṭṭā niyūyārk cikkākō kiḷivlāṇṭ niyūjerci  okāyō ṭōkkiyō

As you know,  when English words are written in Tamil script, much phonetic information becomes obscure due to the absence of script for voiced letters.

This is how Ohio is written in Tamil, but intervocalically k becomes h in Tamil, while orthographically only k is written.  c becomes s intervocalically, k becomes g before nasal and so on are some of the phonological rules in Tamil. So, Ohio would be written okāyō and pronounced ohāyō.

Phonemic    phonetic

/ciyāṭṭal/ > [siyāṭṭal]

/cāṉ pirāṉciskō/ > [sāṉ pirāṉsiskō]

/okāyō/ > [ohāyō]

aṭlāṇṭā niyūyārk cikkākō kiḷivlāṇṭ niyūjerci ṭōkkiyō are the same phonemically and phonetically.

THAI

We pronounce those names similarly to English speakers with some Thai accent (wrong stress).

I think one interesting city name is Bangkok. We pronounce it บางกอก  [bāːŋ kɔ̀ːk], but the name was changed to กรุงเทพฯ  [krūŋ tʰêːp] around two hundred years ago (by the same king who ordered the translation of The Three Kingdoms. The full name of the city is considered the longest city name in the world:

Krungthepmahanakhon Amonrattanakosin Mahintharayutthaya Mahadilokphop Noppharatratchathaniburirom Udomratchaniwetmahasathan Amonphimanawatansathit Sakkathattiyawitsanukamprasit

กรุงเทพมหานคร อมรรัตนโกสินทร์ มหินทรายุธยา มหาดิลกภพ นพรัตนราชธานีบูรีรมย์ อุดมราชนิเวศน์มหาสถาน อมรพิมานอวตารสถิต สักกะทัตติยวิษณุกรรมประสิทธิ์

We even have a song to remember this name: listen here.

VHM:  Here is an English translation of the full Thai name for Bangkok, which consists of Pali and Sanskrit words:

City of angels, great city of immortals, magnificent city of the nine gems, seat of the king, city of royal palaces, home of gods incarnate, erected by Vishvakarman at Indra's behest.

Source

Most Thais I know have long names, and — no matter what their original ethnic derivation, their surnames are usually Sanskritically based.  When I first met a woman named Pratoom Angurarohita, her surname stymied me, but then I collected myself and looked at its Sanskrit roots and it no longer intimidated me.  Ever after that time, I could rattle off long Thai names without much trouble.

See:

"Bahasa and the concept of "National Language"" (3/14/13)

This post documents the Indianization of Southeast Asia during the Middle Ages.

Returning to the question with which we began this post, after the Olympics, "PyeongChang" will go back to being "Pyeongchang", but we will still be caught in a quandary over how to pronounce it:  accurately / correctly / right or inaccurately / incorrectly / wrong.

[Thanks to Haewon Cho, Lisa Mitchell, Pushkar Sohoni, Shelley Shim, Nikita Kuzmin, Dotno Pount, Jichang Lulu, Pattira Thaithosaeng, and Vasu Renganathan]



99 Comments

  1. Michael Watts said,

    February 19, 2018 @ 6:27 pm

    Would you rather Americans say Běijīng [pèi.tɕíŋ] or Bayzhing or, for that matter, Peking (my preference, going back to the days of John Milton, variants of which are still used in many languages around the world, and indeed in many Sinitic topolects?

    This is kind of oddly phrased; the implication is that běijīng, the reading of 北京 in the popular Mandarin Sinitic topolect, is not a variant of Peking.

  2. S Frankel said,

    February 19, 2018 @ 6:46 pm

    Well, the "wrong" pronunciation of IKEA in English is promulgated by the company itself. I first became aware of the firm in Sweden, before they opened any stores in the US, and was very puzzled when they started their American advertising with the wrong pronunciation. Why? It would have been just as easy to go with an anglicized version of the Swedish (ee-KAY-ah, or something like that, in advertising phonetic spelling). Did they do any market research on this, or did they just not care?

  3. Vicki said,

    February 19, 2018 @ 7:05 pm

    Which Newark? My understanding is that Newark, New Jersey is "Newurk," Newark, Delaware is "New Ark," and Newark, Ohio is "Nork" (rhymes with Yotk). I learned this from someone who lived in Newark, Delaware; "Newurk" is the pronunciation I'm used to, having grown up in New York City..

  4. Uly said,

    February 19, 2018 @ 7:28 pm

    The NBC pronunciation is not just the wrong pronunciation in Korean, but it's a very unusual pronunciation for English. Certainly I've never heard it before.

  5. Victor Mair said,

    February 19, 2018 @ 7:32 pm

    @Michael Watts

    Your comment is oddly phrased. There is no such implication.

  6. Scott McClure said,

    February 19, 2018 @ 7:40 pm

    It reminds me of the Turin/Torino situation…which, of course, Language Log covered:
    http://itre.cis.upenn.edu/~myl/languagelog/archives/002834.html

  7. dainichi said,

    February 19, 2018 @ 7:40 pm

    I understand that the American convention is that PALM/LOT is closer to [a] (shouldn't that actually be more like [ä] here?) than TRAP, but since (AFAIK) PALM/LOT is not available before /ŋ/ phonotactically, is CLOTH really the closest match?
    At least some Americans have quite a low TRAP, and some have quite a high/rounded CLOTH.

  8. Jason said,

    February 19, 2018 @ 7:47 pm

    What I really hate is word-internal capitalization, like 杭州 as HangZhou instead of Hangzhou.

  9. Mark Liberman said,

    February 19, 2018 @ 7:51 pm

    For those who wonder what PyeongChang actually sounds like in Korean, there are dozens of helpful YouTube videos, for example this one.

  10. Pflaumbaum said,

    February 19, 2018 @ 7:57 pm

    Yes, even though my work means I’m used to the eye-dialects of Americans from various regions, I found myself totally baffled by this. If Gen AmE “sang” is the wrong approximation, why is it transcribed /a/? What am I missing?

  11. dainichi said,

    February 19, 2018 @ 8:15 pm

    In the sweetandtastyTV video that Mark links to, it stands out to me that she first says "aaaah", with an [ä] sound, then says "chang" with a noticeably higher allophone, something like [ɐ]. So I would say that in American English, even STRUT (i.e. "chung") could be a contender, especially because its length also seems to match the Korean vowel better.

  12. Jerry Friedman said,

    February 19, 2018 @ 8:50 pm

    Pflaumbaum and Dainichi: Many, I think most, Americans have very different allophones of TRAP before /ŋ/ than in other situations. I'd call them [e] or [eɪ] or things around there. I believe I noticed or even was told as a child that "sang" had the same vowel as "sane".

    Dainichi: A lot of Americans merge CLOTH with LOT/PALM and have a low unrounded vowel in those words. The person who suggested "tong" may be one such. I have a higher and maybe slightly rounded vowel in CLOTH, as you say, but I don't have a problem using my LOT/PALM vowel before /ŋ/ in foreign names.

    Also, the memberships of people's CLOTH classes vary. If I thought about it, I might be able to come up with a word that ends in "-ong" but is in my LOT class.

  13. Jerry Friedman said,

    February 19, 2018 @ 8:52 pm

    Apologies to dainichi for capitalization. Corrections of my attempt at something linguistics-ish are welcome.

  14. Chris Button said,

    February 19, 2018 @ 9:22 pm

    @ dainichi

    … then says "chang" with a noticeably higher allophone, something like [ɐ]. So I would say that in American English, even STRUT (i.e. "chung") could be a contender

    I was thinking the same thing. In fact, The British English /ɐ/ for the STRUT vowel isn't actually far off at all (just slightly higher) given how central the Korean "a" is. Of course for many Americans it tends to merge with schwa – I suppose the closest approximation for American English would be a word final schwa in something like "sofa" which at least may surface as a little lower.

    @ Jerry Friedman

    Many, I think most, Americans have very different allophones of TRAP before /ŋ/ than in other situations. I'd call them [e] or [eɪ] or things around there

    Yes – again I think the mismatch would be more pronounced in American English than in British English here too due to the velar raising. Furthermore, regardless of the conditioning environment, the British English TRAP vowel is generally better represented nowadays as /a/ rather than RP /æ/ which is pulling in a different direction from American English. So with this option as well, British English is going to be closer than an American one to the centralized Korean "a".

    A lot of Americans merge CLOTH with LOT/PALM and have a low unrounded vowel in those words. The person who suggested "tong" may be one such

    However, as dainichi noted in terms of the Korean sound, the many Americans, and I believe all Canadians, who do this make a much longer vowel sound best transcribed as /ɑ:/.

  15. Michele said,

    February 19, 2018 @ 9:31 pm

    “Newark, Ohio is "Nork" (rhymes with Yotk [sic – assuming you mean York])”

    I’m from Ohio and don’t pronounce it as rhyming with york. It’s definitely got 2 syllables, and is something like no-ork. Maybe a very soft w sound between syllables, but not really.

    How would that be represented kn IPA?

  16. Arthur Waldron said,

    February 19, 2018 @ 10:14 pm

    Normally all foreign names are either pronounced as if they were French as in bay zhing and Taipeh OR murdered by some plummy Englishman. La bell dayme sans mercy. Which is also classy. My favorite is NPR news in accented English. How would they virtue signal that they had hired say an Indian if she spoke perfect American? She may well in private Some of you may remember Maurice Chevalier with his tank evun for leetle gerls French accent key to his stardom. In fact he spoke fluent unaccented English. He had to learn his stage pronounciation. That said unless we have an American name for the place already Paris Moscow Peking Soul Toe kee you we should try to get the names right while not foreignizing them. We are making them new American words. As the sainted Maggie used. to say “I trust I have made myself clear”. No? Rats (fourth tone). Only on LL kids! ANW

  17. julie lee said,

    February 19, 2018 @ 10:33 pm

    I've heard Chinese people who pronounce "Detroit" and "Florida"
    as Chinese characters, romanized below (with numbers indicating the Mandarin tones:

    Detroit: DI (1). TE (4) LI (3)

    Florida: FU (1). LO (2). LI (3). DA (2) .

    At first I didn't know what DI-TE-LI meant. Had to pause before I realized what FU-LO-LI-DA meant, because of the Mandarin tones, with DA dragged out DA-A-A.

  18. liuyao said,

    February 19, 2018 @ 10:48 pm

    Could the pronunciation be influenced by how Americans say the Chinese surname Chang (= Zhang in Pinyin)?

  19. J.W. Brewer said,

    February 19, 2018 @ 11:19 pm

    I would think most Anglophones would by default pronounce Pyeongchang to rhyme with Pyongyang, which then raises the question of how they pronounce the latter. Wikipedia claims UK: /ˌpjɒŋˈjæŋ/, US: /ˌpjʌŋˈjɑːŋ/, but I'm not sure how completely accurate that is as to AmEng, since I believe I've heard the jæŋ version often enough to think it is also current in the US. AmEng experience with the pronunciation (in an English-speaking context, with the pronunciation potentially Americanized) of Korean-origin surnames like Yang, Hwang, Kang, etc. (not to mention Chinese-origin surnames with -ang spellings, esp. as contrasted with those with -ong spellings) might also be relevant.

  20. RachelP said,

    February 20, 2018 @ 12:25 am

    Just remembering being in Italy trying to get a train to Munich. They call it 'Monaco', which could cause quite a problem if you wanted to get to Monaco.

  21. Tim Friese said,

    February 20, 2018 @ 12:31 am

    'where the "chang" in PyeongChang rhymes with sang, instead of tong.'

    Part of the problem here is precisely that the 'correct' Korean pronunciation does not rhyme with my pronunciation of tong, since I preserve the cot-caught distinction (from St. Louis). Indeed, I don't believe there is a single word in my idiolect with the sequence [aŋ]. I noticed this when I taught English as a second language: it was quite jarring that very advanced students still said that they [saŋ] in choirs, deposited checks at a [baŋk], and hung their clothes on [haŋɹz].

  22. Michael Watts said,

    February 20, 2018 @ 12:34 am

    Which Sinitic topolects don't use a variant of Peking to refer to 北京?

  23. Michael Watts said,

    February 20, 2018 @ 12:39 am

    Tim Friese, I do not have the cot-caught distinction, but all three of your examples would be jarring to me too, as "sang", "bank", and "hangers" all use the vowel of "sang", not the vowel of "song". How jarring did you find the Koreans' pronunciations of e.g. "long" / "strong" / "song" / "wrong"?

  24. AntC said,

    February 20, 2018 @ 1:35 am

    [J.W.B.] would think most Anglophones would by default pronounce Pyeongchang to rhyme with Pyongyang, …

    Before the games started, there were reporters doing background pieces where their greatest worry was that non-asian viewers would confuse PyeongChang entirely with PyongYang. Since that would be the only place in Korea most would have heard of other than Seoul.

    Of course the reporters didn't have in mind readers of LLog.

    The worry arising from anecdotal evidence [youtube stop-people-in-the-street shove-microphone-in-face] that hardly any Americans could place either of the Koreas on a map. Neither could they place Vietnam (some pointed at Vietnam as being Korea), nor Iraq — answers included anywhere between East Africa, Northern Europe, Central Steppes. Call me old-fashioned, but if you're at war (or threatening war) with a place, it'd be good to send the troops and missiles in the right direction.

    Perhaps the Winter Olympics (and the chilling, robotic so-called Cheerleaders) will make some sort of lasting impression?

  25. Keith said,

    February 20, 2018 @ 3:32 am

    For the last couple of months I've been thinking about the way the French pronounce Korean words, too.

    Although the Winter Games hold almost no interest for me, I hear a round-up of the day's events each evening when I get home and switch on the radio.

    French journalists seem almost invariably, in my experience, pronounce Pyeongchang (평창군) as pjɔ̃tʃɔ̃ and Pyongyang (평양시) as pjɔ̃jɔ̃.

    Many regional accents of standard French (by this, I mean where the lexicon and grammar are standard French, but the pronunciation is detectably different, even for non-French speakers) make a clear difference between the sounds written "en", "an" and "on", between those written "ein", "in" and "un".

    The "Parisian" pronunciation is merging "en", "an" and "on" (not complete, there is still some difference in some words and for some speakers) and has completely merged "ein", "in" and "un". This last means there is no difference between "brin" (a sprig, small leaf, pinch of powdered spice) and "brun" (the colour brown), and reinforces the use of "marron" for all shades of brown, not just reddish chestnut brown.

    It is this Parisian pronunciation that is used by almost all presenters and journalists on the stations I listen to (admittedly, not very many).

  26. Victor Mair said,

    February 20, 2018 @ 6:26 am

    @Michael Watts

    It's a question of which Sinitic topolects still retain the velar at the beginning of the second syllable.

  27. David Marjanović said,

    February 20, 2018 @ 7:02 am

    필라델피아 Pillidelpia

    라 is la, not li.

    PALM/LOT is not available before /ŋ/ phonotactically

    *lightbulb moment*

    If Gen AmE “sang” is the wrong approximation, why is it transcribed /a/?

    It's not a specifically American or English transcription. Instead, it tries to use the usual values of the Latin letters as far as is practicable.

  28. Pflaumbaum said,

    February 20, 2018 @ 8:08 am

    @ David Marjanović

    No I didn't mean that – I meant if it's /a/ in Korean, what is so wrong about using AmE's TRAP vowel as an approximation?

    But I understand better now after David Friedman's explanation.

    By the way, are you in communication with Piotr Gąsiorowski? Any idea if he's planning on continuing his blog at any point?

    @ Chris Button

    "Of course for many Americans it tends to merge with schwa"

    Same for many English people – basically everyone outside the South except those who have a full [ʊ].

  29. Pflaumbaum said,

    February 20, 2018 @ 8:09 am

    Sorry, *Jerry* Friedman's. Multitasking fail.

  30. Rodger C said,

    February 20, 2018 @ 8:16 am

    My grad-school housemate (1970s) from Newark, NJ said that it was locally pronounced nʊɾk.

  31. arthur said,

    February 20, 2018 @ 8:50 am

    One of the few enjoyments of mid-afternoon Amtrak train service from Washington to New York City is hearing the conductors correctly pronounce Newark, Delaware and Newark, New Jersey in the same list. I've long wanted to take a bus from Des Moines to Des Plaines, but circumstances have not yet permitted.

  32. Vilinthril said,

    February 20, 2018 @ 8:55 am

    To add pronunciations from my native German, I don't think the examples given in the blogpost are particularly interesting, but a few other things do come to mind: For example, London is pronounced pretty universally with the “wrong” vowels by German speakers (i. e. like English “condom”), and Chicago often is pronounced with an initial [ʧ].

  33. Frank DiBona said,

    February 20, 2018 @ 9:22 am

    We "mispronounce" the capitals of Italy, France (as well as the names of the countries themselves) wrong. The "correct" pronunciations should be Roma (with a distinct rolling of the r and with clear round vowels), and Paree (with a rise of pitch and a singing quality of the second syllable). Of course the country if "Italia". "France" of course should be "française". Then, of course, we could just speak English.

  34. Vulcan With a Mullet said,

    February 20, 2018 @ 9:56 am

    Every phoneme in the native pronunciation for Pyeongchang is pretty close to a typical English sound. I don't understand why it's so hard to repeat it once you've heard it once.

  35. dainichi said,

    February 20, 2018 @ 10:23 am

    @Jerry Friedman:
    > A lot of Americans merge CLOTH with LOT/PALM and have a low unrounded vowel in those words.

    Sure, but I was under the impression that something like pre-velar raising also applies to merged CLOTH/LOT/PALM. Listening to
    https://youglish.com/search/song/us
    all speakers seem to have some rounding and/or raising on the vowel of "song". Of course, I don't know exactly who of them are cot-caught merged, but searching instead for "stop" I hear some of those same speakers with a clear [ɑ] vowel.

    > I don't have a problem using my LOT/PALM vowel before /ŋ/ in foreign names.

    That's interesting. Do you think that can be said in general for not-cot-caught-merged AmE speakers?

    @Keith:
    > The "Parisian" pronunciation is merging "en", "an" and "on"

    Whoa… this is news to me. In most French material I've been listening to, I still (believe I) hear /ɑ̃/ as [ɔ̃] or [ɒ̃] and /ɔ̃/ as [õ]. So is it that they're actually saying pjɔ̃tʃɒ̃ and pjɔ̃jɒ̃, but merging the vowels?

  36. Gregory Kusnick said,

    February 20, 2018 @ 12:16 pm

    "PALM/LOT is not available before /ŋ/ phonotactically"

    In my brand of US English, "pong" and "kong" are distinct from "song" and "long".

  37. Jerry Friedman said,

    February 20, 2018 @ 12:29 pm

    Gregory Kusnick: I'm not sure about those in mine, but in polysyllabic words such as "bongo", "conga", "Congo", "Congress", "mongrel", "Mongolia", I have the LOT vowel. Also spelled with an "a" as in "Bangalore" and "Tang dynasty".

    dainichi: I think a lot of Americans pronounce those words the way I do, though I don't have any data. There may be some raising, but I can't detect it in my speech.

  38. J.W. Brewer said,

    February 20, 2018 @ 1:10 pm

    My cot/caught are unmerged but the vowel I use in (most?) words ending in -ong feels a bit intermediate. Definitely closer to the CLOTH vowel than to PALM/LOT, but maybe a bit raised compared to my CLOTH vowel in other phonotactic contexts? But what's probably most relevant here is that even without a merger PALM/LOT doesn't contrast with CLOTH in that context. So if I'm going to pronounce (w/o self-consciously adopting a foreign pronunciation) a Chinese-or-Korean-origin name like "Chang" contrastively with "Chong" it's going to default to the same contrast I use for e.g. "bang" v. "bong," (or "sang" v. "tong", to give the examples above — although perhaps unlike certain other AmEng speakers my vowel in Chang/bang/sang is pretty close to my usual TRAP/BATH vowel in other contexts).

  39. dainichi said,

    February 20, 2018 @ 1:11 pm

    @Jerry Friedman,
    Thanks. Yeah, listening to
    https://youglish.com/search/congress/us
    I do hear a lot of [ɑ]s. So it looks like there _is_ a CLOTH-LOT distinction before /ŋ/ after all, and I have been wasting everybody's time… Sorry about that. At least I've gotten wiser.

  40. J.W. Brewer said,

    February 20, 2018 @ 1:37 pm

    dainichi: I would not assume most/all AmEng speakers use the same vowel in the first syllable of "congress" as they would in the monosyllable "cong" or in cong/kong as a final syllable of a polysyllabic word. Not least because if you ask them to say the first syllable of "congress" on its own you'll get "con" (with LOT/PALM) rather than "cong" (just as the second syllable on its own will be -gress rather than -ress). The implicit phonotactic rules that speakers follow w/o necessarily being consciously aware of, much less able to articulate, can be pretty fine-grained indeed.

  41. Chris Button said,

    February 20, 2018 @ 1:45 pm

    "PALM/LOT is not available before /ŋ/ phonotactically"

    I would say that this is true as regards PALM /ɑ:/ which is a relatively "new" vowel conditioned by lengthening before certain consonants (which is behind the "TRAP-BATH" /æ/- /ɑ:/ split that generally didn't occur in American English outside of a few words like "father") so it is phonotactically restricted as a result.

    However, for the American English LOT vowel /ɑ:/, I would suggest this is not quite correct because while some words with final /ŋ/ or /g/ (e.g. "long", "dog" and exceptionally also "gone") resisted a merger with PALM /ɑ:/ and instead have THOUGHT /ɔ:/, this varies sporadically from word to word and from speaker to speaker. Also, the relative lowness of THOUGHT /ɔ:/ in American English (it's generally higher in British English although usually transcribed with the same symbol) means that for an increasing number of Americans they are all merging as LOT /ɑ:/ as they have already done in Canadian English.

  42. Ellen Kozisek said,

    February 20, 2018 @ 2:37 pm

    I wouldn't so much mind hearing -chang pronounced to rhyme with sang if it was done by individuals. Since /aŋ/ doesn't exist in English words, I can see anglicizing to /eiŋ/ or /æŋ/ (whichever one has). Or /ʌŋ/. But, given that /aŋ/ is quite easy for an English speaker to say, for NBC to tell it's announcers to pronounce it differently is wrong. And disrespectful to Korean.

  43. Ellen Kozisek said,

    February 20, 2018 @ 3:40 pm

    I'm surprised by all the mentions in the comments to LOT/PALM. For me (American, middle of the country), those two words have different vowels. "Lot" having an unrounded vowel, and "palm" having a rounded vowel due to the L that follows.

  44. Geoff said,

    February 20, 2018 @ 3:41 pm

    Just curious: In the russian rendering of San Francisco, did your subject give the last vowel full value as suggested by your spelling, or reduced value as would be normal for an unaccounted vowel in this position? Is there a normal practice in Russian spelling of foreign proper names – ie phonetic approach versus influenced by the spelling of the original? Same issue arises in the other direction too (Gorbachev vs Gorbachof)

    And in the Russian rendering of Cleveland, did your subject palatalise the e as suggested by your spelling ?

  45. BZ said,

    February 20, 2018 @ 4:19 pm

    It's interesting that you didn't mention the Russian pronunciation of Philadelphia (Филадельфия) with the soft marker between the l and the f. As an immigrant from Russia my intuition is that "lf" should be easy to pronounce in Russian, but I can't think of any "lf" clusters in any Russian words and all the borrowings seem to become "льф", so my intuition might be compromised after living in the US for so long. Or maybe it's closer to the original Greek.

  46. Matt said,

    February 20, 2018 @ 4:32 pm

    Wait till they hear how we pronounce the name of their country. It sounds nothing like "Hanguk." Also, please don't tell the Albanians (Shqiperi), Armenians (Hayastan), Bhutanese (Druk Yul), Chinese, Croatians (Hrvatska), Egyptians (Masr), Georgians (Sakartvelo), Germans (Deutschland), Greeks (Hellas), Hungarians (Magyarorszag), Indians (Bharat), Japanese (Nihon), Maldivians (Dhivehi Raaje), or Montenegrins (Crna Gora) how badly we mangle the names of their countries in English.

  47. BZ said,

    February 20, 2018 @ 4:32 pm

    @Geoff,
    In Russian unstressed vowels are always reduced. A typical Russian (speaking the standard dialect. There are dialects where the reduction does not occur) would not be able to pronounce it any other way.

    In general, Russian spellings of borrowings tend to try to be phonetic, which fails miserably when sounds that don't exist in Russian are transcribed. There is considerable variation in how those are transcribed from one borrowing to another.

  48. Pflaumbaum said,

    February 20, 2018 @ 6:22 pm

    @BZ

    A large number of English loans in Russian seem to take the soft sign after /l/. My assumption was that this was a tradition from when contact was with English spoken in England. But that can’t be right as even for an English person the second /l/ in ‘Philadelphia’ would be “dark” (while the first would be “clear”, so my theory would expect it to be palatalised in Russian).

  49. Eli Nelson said,

    February 20, 2018 @ 6:33 pm

    @BZ: My impression was that the situation is something like this. The sound "ф" does not really occur in inherited native Russian words. The Russian "л" sound is velarized, and historically coda /l/ in the main source languages for Russian borrowings (e.g. French) has been non-velarized or at least less velarized, so it has been borrowed as ль. English does have velarized coda /l/, so I'm not sure why "льф" is used in Филадельфия, unless as you said it is supposed to represent a more Greek-based pronunciation, or it might be based on generalizing the practice to something like "coda /l/ in loanwords should always be borrowed as ль". The Scottish name "Balfour" seems to be Russianized as Бальфур.

  50. Nardog said,

    February 20, 2018 @ 6:35 pm

    The word "Silicon" in English denotes a particular chemical element Silicium, which should be translated in Russian as Кремний, therefore the correct translation of Silicon Valley in Russian should be Кремневая долина. The Russian word 'Силикон', which sounds similar to English "Silicon", means SiliconE, which is a chemical compound that is different SilicoN.

    That reminds me of 聖林 ("holy forest"), the archaic kanji/calque form of ハリウッド Hollywood, mistaking "holly" for "holy".

  51. Bloix said,

    February 20, 2018 @ 8:40 pm

    If the correct pronunciation is Chong then why would it be transiliterated as Chang? You’ve got no right to complain when people say it the way you’ve spelled it.

    And the company decided to Americanize the pronunciation of IKEA because they are selling to ordinary people who want cheap furniture, not to the kind of people who take pride in knowing the rules of pronunciation in six languages.

  52. Ellen Kozisek said,

    February 20, 2018 @ 8:52 pm

    Bliox, because English isn't the only language that uses the Roman alphabet. It's not being transliterated into the English alphabet, but into the Roman alpahbet. A, not O, is the appropriate vowel for the sound used in Korean.

    Also, "chong" only works for those who have an unrounded vowel in "chong". For the rest of us, pronouncing the 2nd syllable of PyeongChang like "chong", rhyming with "song", wouldn't reflect the Korean pronuciation, which doesn't have a rounded vowel.

  53. Ellen Kozisek said,

    February 20, 2018 @ 8:53 pm

    Pardon the typo in your name, Bloix.

  54. Jerry Friedman said,

    February 20, 2018 @ 9:55 pm

    Ellen Kozisek: John R. Wells named that lexical set PALM, so we're stuck with the name (or at least I'm not the one who's going to change it). It's unfortunate but true that "palm" doesn't have the "PALM vowel" for a lot of Americans, as you say. Would "spa" have been the best choice?

  55. Keith said,

    February 21, 2018 @ 2:40 am

    @dainichi

    Hmm… I didn't include some other information, that might be useful. In a few words, such as "on" (2nd person pronoun) and "en" () there is a clear distinction, and careful speakers make a difference between the sounds in more words.

    Radio commentators, however, are often speaking off the cuff, and especially during interviews are reacting and forming questions based on the other person's speech, with all the false starts, halts, pauses, and so on that have been so often discussed on LL with regard to Trump, Pence, SCOTUS judges….

    If the French material you're listening to was specifically recorded for language learning, or if the speakers are being particularly careful (or are just careful speakers in general), then there may well be a clear distinction between "brin" and "brun" and there will almost certainly be a clear distinction between "empreinte" and "emprunte" (but that's another story, from the 1990s).

    Then again, it could just be my bias, and hearing the Korean names butchered has poisoned my recollection of other pronunciations… but to my ear there does seem to be a merging of "en" "an" and "on" going on.

  56. Frédéric Grosshans said,

    February 21, 2018 @ 5:57 am

    @Keith

    For me (native parisian French, but non-linguist), the merging of nasals you describe is wrong on one point: /ɔ̃/ remains distinct from the other two nasals /ɑ̃/ and /ɛ̃/ (/œ̃/ has indeed merged into /ɛ̃/ in my dialect). And I think Marron (brown) is very different from marrant (funny) for any French speaker.

    The only cases of /ɔ̃/ /ɑ̃/ merger I can think of is 1ɑ̃. the word non (no) which is sometimes said /ɑ̃/ 2. mispronunciation by non native speakers e.g. native English and German speakers.

    To come back to the main subject of this discussion, the typical pronunciation of French journalists of these two korean cities are pjɔ̃gtʃɑ̃g for Pyeongchang (평창군) and pjɔ̃gjɑ̃g for Pyongyang (평양시). (The g might be ŋ )

  57. Pflaumbaum said,

    February 21, 2018 @ 6:16 am

    @ Frédéric

    How do you pronounce /ɛ̃/? Presumably it varies depending on accent/context but to my ear it generally sounds a lot more like [ã] than [ɛ̃].

  58. Coby Lubliner said,

    February 21, 2018 @ 9:48 am

    The tradition of pronouncing names ending in -ang (or even -ange), regardless of origin, as /æŋ/, is quite old. Think Fritz Lang, Chiang Kai-shek or Dorothea Lange.

  59. Chris Button said,

    February 21, 2018 @ 10:34 am

    @ Ellen Kozisek

    I'm surprised by all the mentions in the comments to LOT/PALM. For me (American, middle of the country), those two words have different vowels. "Lot" having an unrounded vowel, and "palm" having a rounded vowel due to the L that follows.

    I'm interested that you are putting this down to the "l". While syllable final "dark-l" /ɫ/ (spelled "ll") did indeed cause rounding which is why /æ/ can never occur in such an environment but instead rounds to /ɔ:/ as in "ball" (which as has been noted above then has later unrounds for many Americans to /ɑ:/) as opposed to "bat", I wasn't aware of it happening when the /ɫ/ is not in syllable final position where I just assumed the vowel had been stretched to give /ɑ:/ instead of /æ/. I'm guessing you pronounce "calm" with /ɑ:/ ? If so, perhaps your /ɔ:/ in "palm" might simply be a word specific reflex conditioned by the combined rounding effects of the bilabial onset, bilabial coda and the "dark-l".

  60. Tim Friese said,

    February 21, 2018 @ 10:41 am

    @Chris Button

    I am an American who distinguishes cot/caught and I certainly pronounce "calm" with /ɔ:/. As far as I am aware, I have this in basically all 'a's before coda 'l', with the only exception that I remember noticing being 'alto' /æltoʊ/ (though the environment conditions the first vowel to something like [a]).

  61. Coby Lubliner said,

    February 21, 2018 @ 11:47 am

    @ Tim Button: you may be confusing spelling with sound. I don't think that the final /l/ in "my pal Hal went to Cal" is any different from the one in "call". Then there is the British pronunciation of "mall", and The Dalles, OR.

  62. David Marjanović said,

    February 21, 2018 @ 12:12 pm

    By the way, are you in communication with Piotr Gąsiorowski? Any idea if he's planning on continuing his blog at any point?

    I haven't talked to him, but he still comments on Language Hat occasionally, and still deletes spam from his blog once in a while. My guess is that he's still occupied with the ongoing political battle to save Białowieża.

    To add pronunciations from my native German […] London is pronounced pretty universally with the “wrong” vowels by German speakers (i. e. like English “condom”), and Chicago often is pronounced with an initial [ʧ].

    Yes – these are spelling-pronunciations based on perfectly cromulent English spelling rules. What you didn't mention is that the second vowel of London becomes reduced like German -en endings, usually to nothing.

    In most French material I've been listening to, I still (believe I) hear /ɑ̃/ as [ɔ̃] or [ɒ̃] and /ɔ̃/ as [õ].

    The on vowel has been [õ] for many decades, and is staying put. The en/an vowel and the in/ein/ain vowel (the latter including un in Paris) are still moving, however: [ɒ̈̃] > [ɔ̃] and [æ̃] > [ã], respectively. I expect en/an to merge into on in the generation that is being born now. Occasional misunderstandings are already happening.

    Is there a normal practice in Russian spelling of foreign proper names – ie phonetic approach versus influenced by the spelling of the original?

    There are conflicting traditions that tend to be heavily influenced by the spelling – doubly written consonants are always transcribed as double, for instance.

    The Scottish name "Balfour" seems to be Russianized as Бальфур.

    It looks completely French, though.

    It's unfortunate but true that "palm" doesn't have the "PALM vowel" for a lot of Americans, as you say. Would "spa" have been the best choice?

    Probably. Wells didn't choose "father" because that one has the FACE vowel in a few dialects in England.

    (And yes, I've heard American accents where palm, calm have the only rounded open vowel in the whole system, with LOT = THOUGHT being [ɑ:].)

  63. Chris Button said,

    February 21, 2018 @ 12:15 pm

    @ Coby Lubliner

    I think you have me confused with my philosopher brother "Tim Button" :)

    Cases like "pal" are outside of the basic system (as often indicated by the curious spelling with a single "l" in word final position). A word like "shall" is a notable exception with "ll" however along with the "mall" variant you note.

    Phonetically, the vowel generally rounds because of the association of the back feature of the "dark-l". The next step is to round completely. Compare /bɾaziw/ for "Brazil" in Brazilian Portuguese, or the syllabic "l" (i.e. no longer in coda position) in /bɒʔʊw/ for "bottle" in Cockney/Estuary English.

    @ Tim Friese

    I neglected to consider all the words with "lk" like "talk", chalk" etc in which the rounding does occur with a non final "dark-l". As such, rounding with "lm" makes equally good sense of course! However, it didn't seem to happen with "lf" as in half, calf etc. I'm half expecting you to tell me you pronounce "half" as /hɔ:f/ now :)

  64. David Marjanović said,

    February 21, 2018 @ 12:16 pm

    …Actually, I'm not completely sure about THOUGHT, but LOT is definitely [ɑ:] in the accent I have in mind. It's been a while since I last heard it.

    syllable final "dark-l" /ɫ/ (spelled "ll") did indeed cause rounding which is why /æ/ can never occur in such an environment but instead rounds to /ɔ:/ as in "ball"

    It does in more recent borrowings like alcohol, fallacy and the abovementioned pal, as well as in words where the /l/ has only more recently become syllable-final (Hal and Cal as mentioned, from Harry and California).

  65. Vilinthril said,

    February 21, 2018 @ 12:19 pm

    @David: “What you didn't mention is that the second vowel of London becomes reduced like German -en endings, usually to nothing.”

    I don't agree there, at least not in my experience and in Vienna. (FWIW, Wiktionary seems to agree with my impression: https://de.wiktionary.org/wiki/London)

  66. David Marjanović said,

    February 21, 2018 @ 12:20 pm

    I have no idea what happened to shall. It even has rounded vowels on the mainland (German sollen, Dutch zullen….). Talk, chalk, walk are rounded all the way to THOUGHT; half, calf are BATH, which I suppose might count as "better than nothing". :-)

  67. David Marjanović said,

    February 21, 2018 @ 12:23 pm

    I grew up mostly in Vienna, where the teacher told us to say [ˈlʌnˈdʌn] with two stressed syllables (!) to avoid [ˈlɔnd̥n̩]…

  68. Chris Button said,

    February 21, 2018 @ 12:23 pm

    However, it didn't seem to happen with "lf" as in half, calf etc. I'm half expecting you to tell me you pronounce "half" as /hɔ:f/ now :)

    Nor did it lengthen to /ɑ:/ instead of /æ/ in American English (in British English the /ɑ:/ can be put down to the TRAP-BATH split instead).

    There must be a very interesting etymological discussion underlying all of this…

  69. Tim Friese said,

    February 21, 2018 @ 12:26 pm

    @ Chris Button

    You asked if I have "half" as /hɔ:f/ but I'm sad to tell you I do not! That group of 'l's are silent for me and the 'a's are /æ/: /hæf/, /kæf/, etc. Though you know /hɔ:f/ would be a nice analogy to /tɔ:k/!

  70. David Marjanović said,

    February 21, 2018 @ 12:27 pm

    …actually, what's going on here is that the vowel-reduced pronunciation makes use of an English spelling rule, but then fills in the southern German sound system which lacks a schwa (there's nothing available between [ɛn] and [n̩]). Going for -[ɔn] as Wiktionary has it is the outcome of only applying German spelling rules, according to which only e can be reduced (with the partial exception of Doktor).

  71. Jerry Friedman said,

    February 21, 2018 @ 1:35 pm

    Tim Friese: I am an American who distinguishes cot/caught and I certainly pronounce "calm" with /ɔ:/. As far as I am aware, I have this in basically all 'a's before coda 'l', with the only exception that I remember noticing being 'alto' /æltoʊ/ (though the environment conditions the first vowel to something like [a]).

    Algae, calcium, calculate, salvage, salvation? Proper nouns such as Albert, Calvary, Galveston?

    I'm not clear on whether words such as "gallon", "mallet", and "rally" count as having coda "l".

  72. Anglicization said,

    February 21, 2018 @ 3:10 pm

    What about [paɪ.ˈɒŋ.tʃæŋ]? "ye" as in "bye," "ong" as in "long," and "ang" as in "hang."

  73. Chris Button said,

    February 21, 2018 @ 4:34 pm

    @ Tim Friese and Coby Lubliner

    It also just occurred to me that in the -lt cases where the dark "l" is retained along with its associated rounding as in words like "salt", British and American English may be seen to be slowly converging around the COT vowel (so /ɒ/ and /ɑ:/) respectively, since the /ɔ:/ vowel appears to be gradually merging in such instances with /ɒ/ in British English and /ɑ:/ in American English.

  74. RP said,

    February 21, 2018 @ 5:38 pm

    I don't know whether the /ɑ:/ in BrE "half", "calf" can be put down to the TRAP/BATH split, but I'm a BrE speaker who doesn't have the TRAP/BATH split (I pronounce both of them with the TRAP vowel) – however, I use the PALM vowel (/ɑ:/) for "half", "calf" as well as for "palm", "father", "spa".

    Afaik, this isn't unusual. According to dialectblog.com: 'there are words like calf, half, rather, can’t, banana, etc. that have the “broad-a” everywhere in England. So you’ll hear people from Northern England who otherwise don’t have the TRAP-BATH split using the same vowel as PALM in these words. J.C. Wells says of words like these (and I agree with him on this), “if we were considering their pronunciation in England alone, and leaving North America out of account, we should place half, can’t, banana, etc. in the PALM set rather than in BATH.' ( http://dialectblog.com/2011/03/10/mastering-the-trap-bath-split/ )

  75. Ellen Kozisek said,

    February 21, 2018 @ 6:04 pm

    @Chris Button. No, " calm" also has a rounded vowel.

  76. Ellen Kozisek said,

    February 21, 2018 @ 6:12 pm

    @Christ Button

    There's no L sound (for me) in half and calf. Here, the L seems to have dropped out. As it has for some speakers in palm. Then again, there's also no L sound in chalk and talk, which do have a rounded vowel.

  77. Charles in Toronto said,

    February 21, 2018 @ 7:47 pm

    I'm sure this topic has been beaten to death. But it is fair to say that the variance in how Americans pronounce their phonemes is wide enough that advising them on how to say "PyeongChang" using English words is not going to land on the right pronunciation for everyone.

    Speaking as a Canadian (Toronto, Vancouver) – this comes up when we compare our pronunciation of "pasta" to Americans. Most Americans I've met, regardless of their accent, pronounce the first vowel in "pasta" /ɑ/ (the "cot" vowel). Most Canadians I've met are pronouncing it /æ/. Our /æ/ is probably a better approximation for Italian /a/ than our /ɑ/, but many Americans have an /æ/ that's WAY too far away from the mark and they're better off with /ɑ/. However Americans who sound "nearly Canadian" still use their /ɑ/ for "pasta" and I always find that jarring.

  78. Anglicization said,

    February 21, 2018 @ 9:16 pm

    Just pronounce it like this: P as in Peter, ye as in bye, ong as in long, ch as in chair, and ang as in hang. How simple.

  79. Jerry Friedman said,

    February 21, 2018 @ 10:08 pm

    Ellen Kozisek: I believe the L dropped out of "calm", "palm", "balm", "qualm", and "psalm" a long time ago and is now reappearing widely in the U.S. (and Canada?). "Almond" has a slightly different story.

  80. Chris Button said,

    February 21, 2018 @ 10:20 pm

    @ Ellen Kozisek

    Thanks – I was actually just talking with someone originally from Arizona or thereabouts now living in Maryland who has the COT-CAUGHT merger as /ɑ:/ but then still makes a THOUGHT /ɔ:/ vowel in words like "calm".

    @ RP

    I don't know whether the /ɑ:/ in BrE "half", "calf" can be put down to the TRAP/BATH split, but I'm a BrE speaker who doesn't have the TRAP/BATH split (I pronounce both of them with the TRAP vowel) – however, I use the PALM vowel (/ɑ:/) for "half", "calf" as well as for "palm", "father", "spa".

    The TRAP-BATH split applies to words like "half", "calf", "laugh" since it was conditioned by the lengthening of /æ/ (retained in GenAm) to /ɑ:/ (in RP) before fricatives and in a few other cases like "can't" etc. This also applies to "father" even though GenAm has /ɑ:/ while some words like "mass" do retain /æ/ in RP too.

    The long /ɑ:/ in "palm" and "spa" is due to compensatory lengthening either due to the loss of the "l" in the former or the open syllable in the latter. I wonder if the /ɔ:/ option in "palm" discussed above might be due to an artificial reinserting of the "l" in the pronunciation (how "almond" is often pronounced nowadays with an "l) that then conditioned the rounding only to then be dropped again?

  81. PeterL said,

    February 22, 2018 @ 12:52 am

    Similar comment to "Charles in Toronto" … there's an Alma Street both in Vancouver BC and in Palo Alto California.
    A Venezuelan coworker told me that my Canadian pronunciation of "Alma" was almost perfect but the American pronunciation is way off. To my ear, Americans say something like "Olma" … I'm guessing that it's an attempt to overcorrect the flat "a", which is why it's puzzling that they don't do the same overcorrection for "Chang".

    Nobody has mentioned that English speakers have trouble with the first syllable of PyeongChang (turning it into a disyllable "Pi-yong"), just as they mess up words like "Tōkyō". (I knew an exchange student named Ryūichi … nobody could pronounce his name even approximately correctly.)

  82. RP said,

    February 22, 2018 @ 3:43 am

    @Chris Button,
    What was I wondering was (1) what basis there is for considering the /ɑ:/ in "half", "can't" as part of the TRAP-BATH split, (2) if it is indeed part of the TRAP-BATH split, why are there a large number of BrE speakers who have the split in a small number of words ("half", "can't"), but use the TRAP vowel for "bath", "pass", "glass", "last", "after"?

    On point (1), you've argued that "half" participates in the TRAP-BATH because it has to do with the lengthening of the vowel before fricatives such as /f/. Against this, though, someone might point out that you've argued that the reason for the longer vowel in "palm" is in compensation for the loss of /l/. Well, evidently, "half", "calf" have a loss of /l/, so could equally well fit that explanation, whilst "laugh" has a TRAP vowel among the speakers I'm talking about (which fits well with its lack of /l/).

    Finally, it is clear that Wells's delineation of the BATH and PALM classes isn't based on an analysis of whether, etymologically, a word took part in the TRAP-BATH split or not, but is purely pragmatic. Otherwise, why would he have written that “if we were considering their pronunciation in England alone, and leaving North America out of account, we should place half, can’t, banana, etc. in the PALM set rather than in BATH"?

  83. David Marjanović said,

    February 22, 2018 @ 4:57 am

    The TRAP-BATH split applies to words like "half", "calf", "laugh" since it was conditioned by the lengthening of /æ/ (retained in GenAm) to /ɑ:/ (in RP) before fricatives and in a few other cases like "can't" etc.

    Before fricatives and nasals: can't, plant, banana?

  84. David Marjanović said,

    February 22, 2018 @ 4:58 am

    …but not in can, stand, land

  85. dainichi said,

    February 22, 2018 @ 8:29 am

    @J.W. Brewer, Jerry Friedman, Chris Button
    So rephrasing and summarizing, it seems to be the case that LOT/PALM is possible before [ŋ], but not before /ŋ/, at least not if you analyze the [ŋ] in "congress" to be an underlying /n/ which only _surfaces_ as a [ŋ] due to the following velar. Would that be fair to say? So my question about whether (not-cot-caught-merged) Americans can naturally produce [ɑŋ] word-finally still stands.

    Getting back on topic, I have to admit I find it hard to believe that the Asian viewers who are "not too happy" are more than a minority. As it's been stated, TRAP for [a] in loans is well-established, so it can't really be a surprise, more like a "Yeah, Americans do that to [a]s, don't they, ha ha". After all, I believe only a minority of speakers would say Shonghai, Pyongyong, Bongkok or Yongon. Sorry to anybody who's genuinely offended.

    @David Marjanović
    >I expect en/an to merge into on in the generation that is being born now. Occasional misunderstandings are already happening.

    Got examples? My impression coincided with Frédéric Grosshans'. The height of /ɔ̃/ stands out. Sometimes it almost sounds like [ũ].

    > …but not in can, stand, land…
    Just in case you haven't already found this:
    https://en.wikipedia.org/wiki/Trap-bath_split#In_Received_Pronunciation
    So much for "sound laws have no exceptions".

    @Vilinthril
    >London is pronounced pretty universally with the “wrong” vowels by German speakers

    Japanese usually matches STRUT to /a/, but sometimes to /o/ when spelled "o", so it's "Rondon", not the expected(?) "Randon". No idea if it's because it's influenced by spelling, if it depends on the version of English it was borrowed from, or whether pronunciations were different when words were borrowed.

    Who knows the whole story behind STRUT spelled as "o", by the way? I think I've read that before "v" it was to avoid "uu", since "v" used to be written as "u". How about before "n"? Was it a genuine sound change here?

  86. Jerry Friedman said,

    February 22, 2018 @ 12:58 pm

    dainichi: Gregory Kusnick mentioned having [ɑŋ] in "pong" and "kong". I'm still not sure what I say in "ping-pong", "King Kong", and "Viet Cong". Both LOT and CLOTH sound fine.

    I don't know whether there's an /n/ or a /ŋ/ in "Congress", "mongrel", "bonkers", etc. As maybe I should have said earlier, I have grave doubts about J. W. Brewer's claim that Americans asked for the first syllable of "Congress" will say something that ends with [n], not [ŋ] or [ŋg].

  87. Ellen K. said,

    February 22, 2018 @ 1:19 pm

    I don't know what J.W. Brewer's nationality is, but I'm American and I certainly would consider Congress to be syllabified as Con-gress.

  88. Jerry Friedman said,

    February 22, 2018 @ 1:24 pm

    And if you were shown the word "Congress" and asked to say the first syllable?

    J. W. said "most/all Americans". I'd certainly say "cong" (and "gress" for the second syllable), but that's still consistent with his "most/all".

  89. Ellen K. said,

    February 22, 2018 @ 2:36 pm

    I can't with certainty say what I'd say. Neither can you. But I can with certainty say how I perceive the word (and I do mean the word when spoken, even if indicated by letters in this written forum) to be split into syllables.

  90. Jerry Friedman said,

    February 22, 2018 @ 6:32 pm

    Ellen K.: I'm pretty sure of what I'd say. (Or did you mean I can't be certain of what you'd say? That's certainly true.) Anyway, thanks for your response, which surprises me, but language is full of surprises.

    Chris Button: I wonder if the /ɔ:/ option in "palm" discussed above might be due to an artificial reinserting of the "l" in the pronunciation (how "almond" is often pronounced nowadays with an "l) that then conditioned the rounding only to then be dropped again?

    The other possibility is that the /ɔ:/ is by analogy to "talk", etc. I grew up with an accent where "calm", etc., had no /l/ and were in THOUGHT, and that made perfect sense to me because of the analogy. "Half" and "calf" were just weird.
    perfect

  91. Doug said,

    February 23, 2018 @ 8:10 am

    Dainichi:
    " So my question about whether (not-cot-caught-merged) Americans can naturally produce [ɑŋ] word-finally still stands."

    I'm having trouble following this, because I can think of several "-ong" words with the unrounded vowel:

    (ping) pong
    bong
    gong
    (King) Kong
    (ding) dong

    all have the unrounded vowel, unlike the rounded one in song, long, strong, wrong …

  92. Chris Button said,

    February 23, 2018 @ 9:57 am

    @ Dainichi, Doug

    The regular pronunciation of a single spelled "o" in North American English is /ɑ:/. I believe this has the following general exceptions (hopefully I'm not missing anything):

    – low /ɔ:/ before /r/ (e.g. "for")
    – low /ɔ:/ before voiceless fricatives (e.g. "cloth" in accordance with the LOT-CLOTH split) for speakers without a COT-CAUGHT merger
    – low /ɔ:/ before some words with /ŋ/ and /g/ codas (and with "gone" too) for speakers without a COT-CAUGHT merger. My understanding is that the choice of words (particularly the less common ones) can vary from speaker to speaker.

    @ Dainichi

    Who knows the whole story behind STRUT spelled as "o", by the way? I think I've read that before "v" it was to avoid "uu", since "v" used to be written as "u". How about before "n"? Was it a genuine sound change here?

    This accounts for things like sum/some, son/sun etc. I believe it was indeed largely a spelling convention designed to avoid confusion which now just adds to the confusion.

    @ RP

    if it is indeed part of the TRAP-BATH split, why are there a large number of BrE speakers who have the split in a small number of words ("half", "can't"), but use the TRAP vowel for "bath", "pass", "glass", "last", "after"?

    North American English has/æ/ in "half", "calf" and "laugh" which shows the origin in the TRAP vowel that lengthened to /ɑ:/ before the fricatives (and also in "can't) in accordance with the TRAP-BATH split. These words are different from "calm" which does not have /æ/ in North American English either. Beyond that I would put it down to individual speaker variation since the split was not uniform in any case.

  93. RP said,

    February 23, 2018 @ 1:19 pm

    I take Chris Button's point that the vowel split, like many, was not uniform.

    It seems to be that there is none the less a case for considering "half", "calf" as something separate from the TRAP-BATH split, if the following from Wikipedia is correct (it is referenced to Dobson, E.J. (1968). English pronunciation, 1500–1700):

    "half and calf, which had been pronounced with [half, kalf] in early Middle English before developing around the early 15th century to [hauf, kauf] by L-vocalization.[18] In accents of England the development was subsequently the same as that in words such as palm (see below). The North American development to [æ] as in trap seems to be the result of shortening from [hauf, kauf] to [haf, kaf], although there is little evidence of this development" ( https://en.wikipedia.org/wiki/Pronunciation_of_English_⟨a⟩ )

    So, whereas the TRAP-BATH doesn't seem to have started until the 1500s, "half" already had a different vowel from "trap" in the early 1400s. It subsequently shifted back to the TRAP vowel in AmE.

  94. Chris Button said,

    February 23, 2018 @ 2:29 pm

    @ RP

    I think it goes back to the whole "laughter's daughter".

    Both words clearly rhymed and then vacilated over which way to evolve before eventually settling in different directions.

    Spellings like "caulf" or "cauf" for "calf" show a similar influence.

    So the comment you cite may well be correct, but it does then raise a good question about where the modern N. American pronunciations would have come from.

  95. David Marjanović said,

    February 23, 2018 @ 8:14 pm

    The en/an vowel and the in/ein/ain vowel (the latter including un in Paris) are still moving, however: [ɒ̈̃] > [ɔ̃] and [æ̃] > [ã], respectively.

    Correction: in/ein/ain isn't becoming [ã], but [ɐ̃] or thereabouts. Some rounding seems to have already set in, at least in Burgundy.

    Got examples?

    Not offhand. I haven't heard on vowels higher (or lower for that matter) than [õ]; beau and bon are a minimal pair for nasality.

    So much for "sound laws have no exceptions".

    Thanks for the link, but Received Pronunciation is received, not naturally developed – it's an attempt to make sense of a dialect mixture. Other examples abound around the world, like the syllables ending in plosives ("entering tone") developing predictably in each Mandarin dialect except in Beijing/Standard, where they're distributed among the "other" four tones almost at random, or the pronunciation of g as a plosive everywhere except in -ig (but not -ige(-)!) in the German stage pronunciation.

    Who knows the whole story behind STRUT spelled as "o", by the way? I think I've read that before "v" it was to avoid "uu", since "v" used to be written as "u". How about before "n"? Was it a genuine sound change here?

    Would surprise me; honey and money have FOOT in northern England where the FOOT-STRUT split didn't happen. Rather, the spelling convention seems to have been to use o instead of u whenever too many short vertical strokes would occur in writing.

    Contrariwise, /on/ > /un/ (regardless of surrounding consonants) is regular in several Austrian dialects like Viennese.

  96. Vilinthril said,

    February 24, 2018 @ 1:55 am

    /on/ > /un/ in Viennese? I'm trying to think of examples, but failing – I'd rather have put that sound change somewhere in Carinthia, Salzburg (and maybe Tyrol, Upper Austria).

  97. Chas Belov said,

    February 24, 2018 @ 5:38 am

    For San Francisco, I'm much more used to 三藩市 (Cantonese: Saam Faan Sih, three wall city, presumably for being at the tip of a peninsula). We also have Hung Muk Sih, literally Redwood City for the city of the same name (Climate Best By Government Test), and Yau Saan Mei Deih (Excellent Mountain, Beautiful Earth) for Yosemite. There are quite a few other localizations, but it's been so long since I watched the Cantonese News at Nine that I don't remember them.

  98. Eli Nelson said,

    February 25, 2018 @ 4:44 am

    My impression about the use of o for strut in English is that it does seem to have come to be used at a certain point in the general context of "minim" letters in particular, but it seems doubtful that it originated from some process like somebody thinking "hmm, u looks too confusable next to these letters, so I will take all these words that used to be spelled with u and spell them with "o" because that usually represents a similar sound". (I don't know if anyone has seriously proposed this: I'm kind of narrating my own confused initial impression of the minims theory here.) It seems more like there was pre-existing variation in the representation of the STRUT vowel, and following that, the presence of adjacent minim letters came to influence the distribution of variants or the retention of particular variants in particular contexts.

    The use of "o" for STRUT seems to be partly derived from French spelling conventions; in Old French, the sound /u/ cound be spelled o in some contexts. Words like "colour" and "dozen" indicate that (in words from French at least) o can correspond to STRUT even when it isn't next to any letter written with x-height minims.

    Some English vowel changes may have also contributed to the development of the correspondence: "brother" and "other" show shortening of OE /oː/ before the dental consonant /ð/ (this seems like it could be related to the shortening seen before /d/ in words like flood, blood, rudder, or (with other vowels) ladder, bladder), and glove, dove, and shove seem to have had /uː/ immediately after the Great Vowel Shift (maybe related to the spelling of "prove", "move", which still have /uː/) which was subsequently shortened and developed to STRUT.

    English "London" is related to French "Londres"/Latin "Londinium", and it seems likely to me that the Romance spelling had some influence on the present-day English spelling.

  99. goofy said,

    February 28, 2018 @ 11:41 pm

    Some people said Beckham's Devanagari tattoo was misspelled because it has vh in a transliteration of "Victoria". I think it is Marathi
    http://bradshawofthefuture.blogspot.ca/2008/12/beckhams-tattoo_05.html

RSS feed for comments on this post