Help talk:IPA/Korean

aspirated allophones lack mentioning plain b/d/g

aspiration is a word-initial feature in korean, so at the beginning of words there is no difference in pronunciation between <bul> and <pul>, both a simply "phul" with the sole difference being the pitch of that syllable -- with underlyingly voiced consonants carrying a low tone

also, can we finally get rid of that silly outdated classification no modern papers use anymore? there is no "voiceless" "aspirated" "tense" series, there is simply "(underlying) voiced" "aspirated" and "plain" series -- highlighted by the fact that in the pitch mapping they behave the same as vowel onset and r m n (which even devoice/denasalize since aspiration isnt applicable) at the beginning of words, and by the fact that the "tense" consonants are pronounced like english stops in spin/stone/skin while EVEN deriving from the same environment of (s)p/t/k as showcased in older hangeul that instead of doubling the glyph for the consonant place a s-glyph before it 91.65.117.153 (talk) 20:18, 27 March 2025 (UTC)

Derived from?

I'm not very familiar with linguistics. Is it just me or are there a number of different methods for annotating Korean in IPA? Like this website by the National Institute of Korean Language seems to differ from what we do.

I'm currently struggling with getting RS for Hangul#Letters for the IPA transcriptions of Hangul. I've seen different authors transcribe various letters differently.

Is there a reason this page is the way it is? Is it based on any specific sources? grapesurgeon (seefooddiet) (talk) 18:53, 23 September 2025 (UTC)

Update: I've been going through the given sources on the page, and no this is not based on any of the individual sources. I suspect it's mostly OR... Ugh... grapesurgeon (talk) 00:25, 1 November 2025 (UTC)

As a side note, see wikt:Appendix:Korean pronunciation, differs from what is done here and is completely unsourced. Two bits of OR that have raged unchecked for years :/ grapesurgeon (talk) 00:27, 1 November 2025 (UTC)

See also my rant on Talk:Korean phonology#IPA values.

I'm going to give up on fixing all this for now. I have other more visible things to work on and these bits of OR are systemic and have been going on for years, not easy to fix.

Imo the system that's closest to what is currently employed across this page, Hangul, and Korean phonology is the one by Jiyoung Shin, published in Brown, Yeon (2015) and in Shin, Kiaer, Cha (2012). Wikipedians apparently modified it by using ◌͈, instead of * and /ɭ/ instead of /l/. Also the "/y/ ~ /ɥi/" and "/ø/ ~ [we]" I think are possibly unsourced. Also, "/ɛ/" is "/ᴂ/" in Shin's system.

Imo we should just use Shin's system without modification. It's close enough to current practice without all the OR tacked on grapesurgeon (talk) 02:14, 1 November 2025 (UTC)

Incorrect information in cite note 19

Cite note 19 of the article states that "When ㅢ is preceded by anything but ㅇ, /ɰi/ is interchangeable with, or completely shifted to /i/." However, according to exception 3 (다만 3.) in article 5 (제5항) of chapter 2 (제2장: 자음과 모음, https://www.korean.go.kr/front/page/pageView.do?page_id=P000098&mn_id=95) of the official pronunciation guide (표준발음법) by the National Institute of Korean Language (국립국어원), when ㅢ is preceded by a consonant, it is pronounced as [ㅣ]; that is, when ㅢ is preceded by anything but ㅇ, /ɰi/ is completely shifted to /i/. Wanted to clear this up before editing. 118.38.76.56 (talk) 15:05, 18 October 2025 (UTC)

And just to add a question here, should ㅢ be /ɰi/ or /i/ when giving IPA for names like Yong-hui (Korean: 용희)? — Preceding unsigned comment added by AjaxSmack (talk • contribs) 01:27, 26 July 2026 (UTC)

weird transcription...

"ɾ ㄹ [ㄹㅇ] 여드레, 얼음 yeodeure, eoreum Scottish great or American atom" What the? atom sounds like an r? When what t is a liquid? Also most American rs are rhotic, ɹ/, /ɻ/,/ so may be the poorest dialect to pull from. It's a post alveolar ridge r.--KimYunmi (talk) 05:02, 28 April 2026 (UTC)

@KimYunmi: /t/ and /d/ are often tapped/flapped as [ɾ] between vowels in American English, so atom /ˈætəm/ becomes [ˈæɾəm], and latter /ˈlætɚ/ and ladder /ˈlædɚ/ both become [ˈlæɾɚ]. I'm less familiar with the Scottish ⟨r⟩ in great, but the American ⟨r⟩ is broadly /ɹ/, like you said. Snowman304|talk 11:46, 28 April 2026 (UTC)

Yes, so I'm saying that "American 'atom'" is not a good linguistic equivalent. /d/ /t/ is a different sound in American English than Korean liquid r. Korean /r/ is a post aveolar ridge r. So it might be better to pull from say a British dialect or a European language all together which has similar liquids to Korean, more than anything in standard US dialects. KimYunmi (talk) 16:54, 3 May 2026 (UTC)

@KimYunmi: Do you have a published source that says that /ɾ/ is post-alveolar, whether that's [ɾ̠] or [ɽ] or something else? The sources I'm looking at consistently use /ɾ/ (e.g., doi:10.1080/23311983.2024.2384708) or /l/ (e.g., doi:10.1017/S0025100300004758; discovery.ucl.ac.uk/id/eprint/1324541/1/1324541.pdf) as the main allophone, with the other as an alternate. I'm seeing [ɭ] as a syllable-final allophone, but that's covered in a different cell of this chart. Snowman304|talk 18:02, 3 May 2026 (UTC)

And none of what you linked is a /d/ but as noted on the article [ɾ̠] that might sound a lot like the American /d/ even though it's argued to be a different sound (within said wiki article noting tongue position on curl). The problem with ㄹ is it changes sound based on 1. position 2. is there are 2 in a row, 3. (I'd argue emotionality as well, as noted more in the research gate paper) none of which the papers you cited seems to go into, with the last one going more into nasalization. (Going over how Koreans, as a Korean, handle scholarship is a pain. So many problems sometimes.) Oddly, language textbooks do go over this linguistic sound change, while linguistic papers don't, which is aggravating.

ㄷ is closer to d/t (softer t) of [American] English, but Koreans don't particularly distinguish the sounds well with it sometimes floating back and forth in normal speech. I've never, ever seen any paper argue that ㄹ sounds like ㄷ, though. There is a noted flattening of ㅐㅔ with contemporary speakers and only people over 40 distinguishing the sounds... but never that ㄹ sounds like ㄷ.

https://www.researchgate.net/publication/350238620_Phonetic_variation_in_the_Korean_liquid_phoneme

This one says there is more variation between the two sounds listed in common Korean academic papers. As I said, Korean consonants tend to be a little more liquid in natural speech, but the problem with Korean academia is that they often give the "idealized" version of the speech, especially when publishing in English. (It's a beef I have with Korea academia: The narrowing down to the idealized, rather than honoring the range, often to lock out outsiders. I'd argue, though no one will post papers about it, for example, that the ㄹ in Jeolla saturi tends to *also* have some rolled sounds to it in natural speech, especially under emotional stress, but Korean linguistic papers are entirely unwilling to note anything beyond gross sound changes between Seoulmal and saturi. Meh, half of my relatives are from Jeolla... I definitely hear it in a natural Jeolla speaker versus a Seoul speaker trying to imitate Jeolla saturi. It's one of the ways I catch Seoul actors...)

(Not a reliable source, but a demonstration)

Tiktok

https://www.tiktok.com/@korean_summer/video/7625305758321708295

You can hear the Youtuber use both sounds for ㄹ here:

https://www.youtube.com/watch?v=VIQyW1nSgS0 despite saying it's (l with the loop, sorry can't type it...)

ㄹㄹ== more of an /l/ sound.ㄴㄹ, as noted, also produces an /l/ sound.

ㄹ "traditionally" when taught sounds something like /ɾ/ [ɾ̠], there was a phoneme shift as well between Middle Korean and Contemporary/Modern Korean in initial position and something like /l/ in final position.

But as noted in the previous paper, the "ideal" version of ㄹ in language learning classes and in linguistics versus the reality... the range is much wider than the Korean papers are willing to admit in public.

BTW, I would, personally also argue this for ㅂㄷㄱ as well, where the "ideal" doesn't meet the cultural linguistic reality of dialectal behavior and everyday use. (None of the papers you cited cited cross sections of the mouth during natural speech, but more gave specific phrases to the only Seoul participants and used wave forms, rather than examined the phonemes themselves.) It's sometimes like a /l/~/ɾ/ or floating sometimes to [ɾ̠] (much like the ㄷ floats a bit in natural speech between d~t, but linguistic papers from Koreans aren't going to list that). TT When I speak Korean American Korean and am precise in the US, I don't get flagged as much. But in SK I do until I relax it. So I know it's a thing. (Linguistic papers need to catch those variations though)

BTW, noted here is a sound shift in nasalization of ㄹ between Middle and current Korean: https://journal.kci.go.kr/koreall/archive/articleView?artiId=ART002268223 Neither here nor there, but might be useful to note.

Either way there is no way a ㄹ is going to sound like a /d/ or a soft /t/ due to how the tongue curls and the clear differentiation between ㄹ and ㄷ. So I stand with the idea of using a different dialect of English /r/ instead. American English does not have to be mentioned to make the linguistic sounds legitimate in the article. (Another weakness of Korean academia is deferring to SAE and also not recognizing that US English has some weird diphthongs in natural speech that do match Korean sounds.) KimYunmi (talk) 14:45, 10 June 2026 (UTC)