How to say "Seoul"
« previous post | next post »
So far as I know, most Americans pronounce the name of the capital city of the Republic of Korea as "soul".
(Received Pronunciation) IPA(key): /səʊl/
(General American) enPR: sōl, IPA(key): /soʊl/
Rhymes: -əʊl
Homophones: sole, soul, sowl
From Korean 서울 (Seoul, literally “capital city”), originally from Claude-Charles Dallet's French-based romanization of Korean, reinforced by the 1959 South Korean Ministry of Education romanization of Korean, which transcribed the Korean vowel ㅓ (/ʌ/) with the digraph "eo" and which was official until 1984.
Note that English Seoul predates the Revised Romanization romanization of Seoul. The two romanization systems simply produce identical forms.
Then I asked many Koreans how they pronounce 서울, the Hangul pronunciation of the name of their capital city. They made it sound as thought it had two syllables. Ross King:
It's two syllables: something like [sɔul], depending on how speakers render the first vowel (for some it's more like a schwa).
For a deeper dive on the name 서울. Here's Wikipedia's take on the toponymy of the city:
Traditionally, seoul (서울) has been a native Korean (as opposed to Sino-Korean) common noun simply meaning 'capital city.' The word seoul is believed to have descended from Seorabeol (서라벌; historically transliterated into the Hanja form 徐羅伐), which originally referred to Gyeongju, the capital of Silla.
Wiryeseong (위례성; 慰禮城), the capital settlement of Baekje, was located within the boundaries of modern-day Seoul. Seoul was also known by other various historical names, such as Bukhansan-gun (북한산군; 北漢山郡, during the Goguryeo era), Namcheon (남천; 南川, during the Silla era), Hanyang (한양; 漢陽, during the Northern and Southern States period), Namgyeong (남경; 南京, during the Goryeo era), and Hanseong (한성; 漢城, during the Joseon era). The word seoul was used colloquially to refer to the capital as early as the 17th century. Thus, the Joseon capital of Hanseong was widely referred to as the seoul. Due to its common usage, French missionaries called the Joseon capital Séoul (/se.ul/) in their writings, hence the common romanization Seoul in various languages today.
Under subsequent Japanese colonization, Hanseong was renamed as Keijō (京城, literally 'capital city') by the Imperial authorities to prevent confusion with the Hanja '漢' (a transliteration of a native Korean word 한; han; lit. great), which may also refer to the Han people or the Han dynasty in Chinese and is associated with 'China' in Japanese context. After World War II and the liberation of Korea, Seoul became the official name for the Korean capital. The Standard Korean Language Dictionary still acknowledges both common and proper noun definitions of seoul.
Unlike most place names in Korea, as it is not a Sino-Korean word, 'Seoul' has no inherently corresponding Hanja (Chinese characters used in the Korean language). Instead of phonetically transcribing 'Seoul' to Chinese, in the Chinese-speaking world, Seoul was called Hànchéng (汉城; 漢城), which is the Chinese pronunciation of Hanseong. On 18 January 2005, the Seoul Metropolitan Government changed Seoul's official Chinese name from the historic Hànchéng to Shǒu'ěr (首尔; 首爾). Shǒu'ěr is a phono-semantic match incorporating both sound and meaning (through 首 meaning 'head', 'chief', 'first').
Since I'm mingling with people of various social levels and educational backgrounds who speak a variety of combinations of Korean, Chinese, and English — and mix the three freely in the same sentences and phrases, it is difficult for me to distinguish them. I will give specific examples in subsequent posts..
Selected readings
- "I.SEOUL.U" (11/15/15)
- "Using Sinitic characters in Korea" (7/3/15)
- "Japanese readings of Sinographic names" (/26/18)
Martin Holterman said,
May 12, 2025 @ 4:58 am
Interesting. I guess that casts new light on my musings, the other day, that South Korea maybe should move its government to one of the southern cities, so that it's less vulnerable to North Korean attack.
jin defang said,
May 12, 2025 @ 6:15 am
I've thought that, too, Holterman-sensei. Any ideas on the name for the new capital? Given the prickly relationship between Japan and Korea, they might be disinclined to copy the Japanese example wherein the old capital Kyoto, meaning capital, was changed to Tokyo, meaning eastern capital. So unlikely that the replacement for Seoul would translated to "southern capital"?
unekdoud said,
May 12, 2025 @ 11:27 am
Beijing is "northern capital", so moving to a "southern capital" would be close to maximally confusing.
Although there is an even more misleading hypothetical scenario where the capital moves to a different place, which is renamed New Seoul, and later moves back.
Jongseong Park said,
May 12, 2025 @ 11:29 am
The Korean language youtuber Hyangmuncheon (향문천/響文泉) has argued in his book (Hyangmuncheon-ui Hangugeo Bisa 《향문천의 한국어 비사》, 2024) that 서울 Seoul cannot be a direct descendant of 서라벌/徐羅伐 Seorabeol, though he does not dismiss the possibility that they share elements that have the same etymology and are thus cognate. This is based on the argument made earlier in the book that Modern Korean is not descended directly from the language of Silla.
Instead, he points to 사비/泗沘 Sabi, the last capital of Baekje, as the likely source of the name. He draws on the proposal of linguist Huisu Yun (윤희수/尹熙洙) that the water radical (氵) represents *-r of Old Korean in the Baekje transcription of the period after the capital was moved to Sabi. By this analysis, 泗沘 can be reconstructed as representing *sirpir.
The same place was previously transcribed as 소부리/所夫里 Soburi, reconstructed as *sɛrpɛrɛ. The anterior portion *pɛrɛ is presumed to be the same element as the 비리/卑離 Biri that frequently occurs in the Mahan (마한/馬韓) place names recorded in the Account of the Eastern Barbarians in the Book of Wei, in the Records of the Three Kingdoms (三國志魏志東夷傳). The same element appears in Silla transcriptions as 벌/伐 Beol *pɛr and later on as 火 *pɔr (here the Chinese character is used not for its usual reading but for the sound of its semantic match in Korean, the word for 'fire' which is 블 pul in Middle Korean with Yale romanization).
Hyangmuncheon admits that tracing the path from *sirpir to Seoul is speculative, but based on personal communications with Huisu Yun, he proposes that Old Korean *ɛ raised to *i around the year 500, causing a chain shift of original *i to *ʲi; this then weakened to *yo (the iotated version of the arae-a vowel ㆍ of Middle Korean written o in Yale romanization) and finally became ㅕ ye in Late Middle Korean.
In short, *sirpir became *sʲipir in the transitional period to Middle Korean, later to be attested as 셔〯ᄫᅳᆯ syěWùl in Late Middle Korean leading to Modern Korean 서울 Seoul.
The book also says that 首里/しゆり Shuri in Okinawa, once the royal capital of the Ryukyu Kingdom, derives from an old form of the Korean name Seoul.
Modern scholarship on Old Korean has made tremendous progress in recent decades, but it's a shame that there is very little information available on the topic for the general audience and virtually zero for those who don't read Korean. So the book is really helpful in providing an overview of recent developments in the field, and you can tell that the author really knows the stuff. From what I can tell, the book generally reflects the latest scholarship (even where it overturns previously widely-held assumptions). But being for a general audience, it doesn't always cite claims meticulously, so it's difficult to separate the author's own speculation from the conclusions of other scholars. The claim about Shuri deriving from an earlier form of Seoul is mixed in with other examples of Koreanic borrowings in Old Okinawan and the whole section only cites Serafim & Shinzato's The Language of the Old-Okinawan Omoro Sōshi (2021), so I assume that the etymology of Shuri comes from Serafim & Shinzato although it's not entirely clear.
Thomas said,
May 12, 2025 @ 1:29 pm
I think I read somewhere that the revised romanisation of Korean was purposefully designed so that Seoul would still be written Seoul. And that is such a silly idea that it might just be true.
David Morris said,
May 12, 2025 @ 3:15 pm
I remember that at the 1984 Olympic games, after an Australian swimmer won an event, an Australian coach excited referred to the next games in 'See-all'. Over breakfast I can't immediately find a video. There is a 1 hour documentary which I don't have time to watch.
Graham said,
May 12, 2025 @ 4:01 pm
Although it isn't "inherent", certainly "京" ("서울 경") deserves a mention. Thinking of 경부고속도로, 경인선, etc. Perhaps that's in the next part?
Anyway, thanks for the interesting post.
KWillets said,
May 12, 2025 @ 4:13 pm
A related question is the history of 올라가다 (going up) to Seoul and 내려 (coming down) when leaving. It's independent of physical altitude.
My take on the 2 vs. 1 question is that the vowels are opposite in brightness, so they don't combine, and vowel harmony with nearby syllables may heighten this distinction.
Jonathan Smith said,
May 12, 2025 @ 4:34 pm
Re: wiktionary, there is no "inherently corresponding" Chinese character for anything anywhere.
@Jongseong Park thanks and wow esp. re: the idea about "氵" — there are more such devices hiding in Sinographic writing of other times/places but they get practically zero attention. Yes it is a real shame so little is available in this area not in Korean as this seems to be the last area language "Sinologists" etc. get around to….
Victor Mair said,
May 12, 2025 @ 6:12 pm
From Bob Ramsey:
Your Korean friends are pronouncing it with two syllables, because, well, how else would they pronounce it in Korean? Hangul may be an alphabet with those much-glorified features visually represented, but it’s still very much syllabic in the Chinese mold. And those are two Hangul syllables there in the word… Oh, and there’s a Silla place name which many think it’s linked to, and it apparently had three syllables.
Victor Mair said,
May 12, 2025 @ 6:15 pm
From Bill Hannas:
It is definitely not the one-syllable pronunciation "soul" that Americans give it. As its hangul spelling shows, the spoken word has two syllables but they are run together because there is no (spoken) consonant between the vowel ending of the first syllable and vowel initial of the second syllable. So it ends up sounding like two syllables compressed into one long syllable–if that makes sense.
S + uh (as in "duh" without the "d") for the first syllable, u + l (as in "cool" without the "c") for the second. The "eo" of "Seoul" romanizes the hangul vowel symbol ㅓ — it is a simple vowel, not a diphthong. "U" romanizes the symbol ㅜ of 울, sandwiched between ㅇ and ㄹ — the circle ㅇ, which doubles as the symbol for consonant "ng" elsewhere, is obligatory in hangul for all syllable initial vowels but has no sound when in this position; it just means "vowel follows".
Chris Button said,
May 12, 2025 @ 10:53 pm
Bill Hannas's point makes me think of how many English-speaking journalists pronounce Kyiv as one syllable instead of two. So instead of "key.iv" we get "keev".
KWillets said,
May 12, 2025 @ 11:58 pm
My previous comment about opposite vowel harmony was wrong, as I misread the vowel table. Both ㅓ and ㅜ are dark.
AntC said,
May 13, 2025 @ 1:42 am
pronounce Kyiv as one syllable instead of two.
It might be a recency illusion, but I've noticed that change only in the past ten years or so:
So kiev was pronounced as two syllables; kyiv gets pronounced differently, to express anti-Russian solidarity(?)
Laura Morland said,
May 13, 2025 @ 2:31 am
Re: KWillets' comment: "A related question is the history of 올라가다 (going up) to Seoul and 내려 (coming down) when leaving. It's independent of physical altitude."
FYI, Paris has the same relationship to the rest of France. You "monter" to Paris or "descendre" from it, no matter in which cardinal direction you're coming or going from. And although it has at least three hills in it — Montmartre and Montparnasse, not to mention St-Etienne-du-Mont — Paris in general is "relatively flat, with an elevation of 35 m (115 ft) above sea level."
So the linguistic turn of phrase for Paris, as for Seoul, is also "independent of physical altitude."
J.W. Brewer said,
May 13, 2025 @ 6:39 am
In the posh university-affiliated register of BrEng one "goes up" to Oxford (regardless of change in actual altitude from ones starting point) when one matriculates (or maybe even returns after a summer break?) and "goes down" from Oxford when one graduates. Or gets "sent down" if one leaves involuntarily due to disciplinary problems with the authorities. I'm not entirely clear on whether the same phrases are used for Cambridge, or for other university towns of less antiquity. Nor do I know how one would characterize "lateral" travel between Oxford and Cambridge.
One reads in the Bible of going "up to Jerusalem" but Jerusalem is situated at about 2,500 feet above sea level (higher than that on the Mount of Olives), so in the context of many/most potential starting points for travel elsewhere in the Holy Land it's a fairly topographically literal idiom.
Victor Mair said,
May 13, 2025 @ 6:55 am
From Min Jung You:
서울 is pronounced with two syllables in Korean: [suh-ool].
Here’s a breakdown of the pronunciation
서 (seo): This sounds somewhat like “saw” or the “su” in “sun,” but with a more neutral vowel, similar to [ʌ] in IPA.
울 (ul): This sounds like “ool” as in “pool” or “school.”
So in Korean, 서울 is pronounced roughly as [suh-ool], not like the English “soul,” and certainly not the Chinese-style transliteration “Shou’er 首尔.”
Even though it’s written as one word, native speakers pronounce it as two syllables with a smooth transition. That’s true for the pronunciation of most Korean words. Even when there are no spaces between words, each Hangul syllable block is distinct and typically pronounced individually——The syllables may be connected either very smoothly or with slight separation, but they remain recognizable as separate units in speech.
Victor Mair said,
May 13, 2025 @ 7:09 am
In reading texts from the early medieval period in China, I was struck by the fact that, regardless of altitude or direction, it is common to refer to going "up to" or "down from" the capital or other large / important city of various states.
Benjamin E. Orsatti said,
May 13, 2025 @ 7:57 am
The spelling of "서울" as "Seoul" has bothered me since my very first Taekwondo class twoscore years ago. In a very neurotic, irrational kind of way; like the way some people are bothered by the sound of cereal being eaten or the sight of blood. There's a rule running across the ticker tape in my brain that says: "Non-final 'e' after a consonant must always be pronounced!", and so whenever I see "Seoul", my brain gets to the first two letters, outputs /sʌ/, and then crashes.
Why is that, psycholinguists?
David Morris said,
May 13, 2025 @ 8:36 am
The video I mentioned this morning is not a documentary but an extended recent interview with the swimmer. I found videos of the event, interviews with the swimmer and another part of the coach's reaction, but not that part.
I note that Seoul is the only South Korean city in which the second syllable starts with a vowel, so we can't compare it to anything else.
Victor Mair said,
May 13, 2025 @ 9:07 am
With the additions by Orsatti and Morris, this has become even more arcane.
Psycholinguists (and morphologists) do please help.
Jongseong Park said,
May 13, 2025 @ 10:01 am
I hope Benjamin E. Orsatti doesn't have to deal with any Irish names then.
As explained in the Wikipedia article quoted at length in the post, Seoul is originally a French-style spelling. The Korean vowel ㅓ eo was written as ⟨e⟩ in nineteenth-century French material (ㅔ e was written as ⟨ei⟩ in case you're wondering). ㅜ u is of course written as ⟨ou⟩ in French-style romanization.
The Dictionnaire coréen-français from 1880 by the Société des missions étrangères actually spells it as Sye-oul, with the Korean given as 셔울 Syeoul. I'm guessing the glide y might have already dropped out in actual speech by then accounting for the popular spelling Seoul/Séoul, though I'll have to check.
So it was actually Se-oul to begin with. Later, it was reanalyzed as Seo-ul, and the Revised Romanization of 2000 chose to write ㅓ as eo and ㅡ as eu, meaning that the spelling produced by the new rules matched the spelling Seoul that had been entrenched for over a century by then (even if the way they got there was different).
Since ㅓ /ʌ/ and ㅡ /ɯ/ can be thought of as unrounded versions of ㅗ /o/ and ㅜ /u/, there is a certain logic to rendering them as eo and eu respectively. But it's interesting to think that this may have originally been motivated by a mistaken reanalysis of French-style Se-oul as Seo-ul.
David L said,
May 13, 2025 @ 12:19 pm
I'm not entirely clear on whether the same phrases are used for Cambridge, or for other university towns of less antiquity. Nor do I know how one would characterize "lateral" travel between Oxford and Cambridge.
One indeed goes up to Cambridge and down from it. Other universities need not apply.
As for your second question, one goes, if absolutely necessary, to the Other Place — best not to say anything about it, though.
David Morris said,
May 13, 2025 @ 3:11 pm
The closest equivalent is Ulsan, which doesn't help.
Philip Taylor said,
May 13, 2025 @ 4:22 pm
I do mentally wince whenever I hear Seoul pronounced "soul"/"sole" (/soʊl/), just as I wince when I hear Київ pronounced "Keev" (/kiːv/), but in my mind (and clearly incorrectly, as I now learn from this thread) I have always thought of Seoul as "Sea-ole" (/siː oʊl/). It is good to finally learn how it should be pronounced.
David Morris said,
May 14, 2025 @ 6:49 am
At the back of my mind was the place name Noeun, which is a suburb and metro station in Daejeon.
I listened carefully to my wife's pronunciation of Seoul, and I would say that it has one and a half syllables.
Jongseong Park said,
May 14, 2025 @ 10:47 am
Noeun in Daejeon is 노은 No-eun, but could also be the romanization for 뇌운 Noe-un or even theoretically for 노에운 No-e-un for that matter. I prefer to put in an apostrophe and write e.g. No'eun whenever a sequence like ae, eo, eu, oe, or ui occurs in romanization that actually represent two sounds, not one.
Chas Belov said,
May 15, 2025 @ 3:53 am
Yeah, it took me a while to integrate that eo represented the schwa sound of ㅓ. I used to pronounce the city as Soul but now, hopefully not too far off from the correct sound, prononunce it Suh ool.
J.W. Brewer said,
May 15, 2025 @ 7:19 am
Why not pronounce "Seoul" with its standard (monosyllabic) English pronunciation when speaking English and with its standard Korean pronunciation when speaking Korean? A half-century ago when I was a boy living in Tokyo I learned how to pronounce Tōkyō* in a more authentically (I don't say perfectly …) Japanese fashion. At various points in my life I have pronounced it that way while speaking English, if I was speaking carefully and self-consciously. But I eventually came to believe that that was an unhelpful affectation in at least most discourse contexts, so I reverted to using the (or rather a) conventional English pronunciation of Tokyo when speaking English.
*The spelling difference is intentional, because one of the steps toward the "authentic" pronunciation is learning to treat vowel length as phonemic, which is not natural for most AmEng speakers.
Yves Rehbein said,
May 15, 2025 @ 4:11 pm
Seems to be almost a regular phonetic process. Instead of meow you can have mew, and the bird mew reflects *maiwī, that is very similar to soul from *saiwalu (Proto-West Germanic reconstruction according to Wiktionary and further etymology usually uncertain with potential of internal borrowing).
If I am unsure about that, I am completely clueless about Korean.
CCH said,
May 15, 2025 @ 10:53 pm
I've been learning Korean for around 15 years and I say the single syllable "soul" in English, and the two-ish syllable "서울" in Korean. It doesn't bother me when people say it to rhyme with "soul", because that's just how it's pronounced in English, which is a different language from Korean and allowed to treat proper nouns how it wants. I don't think that pronouncing (or spelling for that matter) toponyms the "native" way is better, it just makes it less anglicised and I think words should be allowed to be anglicised as much as words are sinicised. In my opinion, the fact that this language has adopted words from everywhere shouldn't mean that we can't do what we want with them after. I don't expect people outside the UK to know how to pronounce British place names; even after collectively trying to tell Americans that "shire" doesn't rhyme with "ire" they still get it wrong! I don't think a Korean person struggling with the initial ㄹ of 런던 (London) becoming closer to an English /r/ is that different from an English speaker struggling with the standardised orthography and secret second syllable of Seoul.
Also the Kyiv thing is slightly different because Kiev is historically the Russian name for the city, just like saying "the Ukraine" has gone out of style because of historical Russian/USSR implications. That was still being said recently – I'm in my early thirties and I have definitely said "the Ukraine", though I am more conscious of that now as I don't want to imply that Ukrainian sovereignty is in question. As it stands, I don't think any English speaker really knows how to pronounce "Kyiv" so /keev/ has been adopted as the standard English pronunciation in lieu of Kiev.
Jongseong Park said,
May 16, 2025 @ 1:37 am
I don't think any pronunciation that uses English speech sounds will get close to the Korean pronunciation of 서울 Seoul [z̥ʰʌ.ul]. But the biggest obstacle in fact is the first consonant.
The voiceless lenis s of Korean Seoul is difficult for English speakers to produce. A typical English [s] would be heard as ㅆ ss by Korean speakers, not ㅅ s. Even if an English speaker could pronounce the Korean s, it will feel wildly out of place in English speech.
On the other hand, the English GOAT vowel (in the dialects where it is realized as a diphthong) sounds reasonably close to the sequence of vowels in Korean Seoul even if it is a single syllable. As some commenters have touched upon, in rapid speech Korean Seoul doesn't necessarily come across as two clearly distinct syllables anyway.
There is no reason not to pronounce Seoul identically to "soul" when speaking in English. Korean speakers themselves will use this pronunciation when speaking in English, and will expect English speakers to do so as well.
I pronounce Kyiv as two syllables, [ˈkiː.ɪv]. But I'm guessing that many people don't know that the y in that spelling is supposed to represent a vowel.
Jarek Weckwerth said,
May 18, 2025 @ 5:38 am
@Jongseong Park — Thank you for the measured and clear posts in this thread! Much appreciated.
Also, I fully support @CCH and @J.W. Brewer in advocating for English pronunciations in English ;P
Chas Belov said,
May 21, 2025 @ 12:01 am
My bestie questioned my pronouncing Vacaville with Spanish a's.
And I pronounce Los Gatos using Spanish pronunciation.
But I pronounce San Jose as Sannazay like most other Bay Area residents.
Benjamin E. Orsatti said,
May 21, 2025 @ 7:47 am
Chas,
That's interesting; raises the question of when, say, /dey mwã/ becomes /d'moyn/ for Iowans.
And I wonder if there's a "taboo effect" in favor of nativizing certain pronunciations. For example, do Floridians avoid respanifying "Boca Raton", lest the semantic connection to "rat mouth" be strengthened?
Jongseong Park said,
May 22, 2025 @ 8:35 am
I should point out that Des Moines is pronounced [de mwan] in contemporary Metropolitan French, and historical North American French pronunciation would have realized Moines as something like [mwɛn(ə)].