Korean "tense" consonants

[Chengjiang]

What's the general consensus on what these are? I'm getting rather confusing info from the different sources currently at my disposal (especially those geared toward learning Korean). To me they sometimes sound like ejectives (which pretty much everyone says they're not, and I assume they're right), sometimes like slightly longer versions of their non-tense counterparts, and sometimes like they are not necessarily longer but have more pressure built up before release. I do note that they're pretty consistently and clearly unaspirated while both the "tenuis" and the "aspirated" consonant series appear to have some level of aspiration word-initially. (They sound more like "less aspirated" vs. "more aspirated" to me, and I've heard them described as such a few times. Somewhat unusually, even non-tense /s/ appears to be aspirated.)

I ask about this in part just because it's been bugging me for a while and in part because I have a set of consonants in one of my conlangs (Common Gnomish) that I'm currently pronouncing similarly to how the tense consonants sometimes sound to me.

[Kereb]

everything I've read seems to agree it's "you ... wouldn't understand."

[Nortaneous]

I vaguely remember listening to some Korean years ago and deciding that it was unaspirated / less aspirated / more aspirated, but I have no idea which was which. (I later ripped this off for Amqoli: there are lenis plosives, which can be anything from German-style lenis consonants to voiced fricatives; modal plosives, which are lightly aspirated; and plosives with strong aspiration that's usually realized as velar/palatal. Though it wasn't intentional, and the aspirated plosives started off as Ch clusters.)

But yes, Korean does have an aspirated fricative.

[Hydroeccentricity]

My Korean is partly influenced by family accent, so it might be out of date in the "motherland," but... I say tense consonants with more "force" if that makes any sense. It's not aspiration. It's almost like an ejective, but I'm pretty sure I'm not closing my glottis. There's just a little more air power. It's like if someone does something stupid, and you say "Duh!" really harshly. You build up air behind your tongue before you let the D go.

God, that wasn't a very good explanation, was it? I guess just watch TV on Naver or Daum and you'll get a better idea.

EDIT: oh, and don't ask me about ㅆ. I don't even f***ing know what's going on there.

[Hallow XIII]

the difference between ㅅ and ㅆ is the simplest one: the former is sʰ and the latter is basically normal s (or at least this consonant will be understood as ㅆ by Koreans, idk if native speakers have any of that tenseness jazz going on).

EDIT: in fact most tenuis voiceless plosives will be interpreted as tense consonants by Koreans, and they are taught in Korea (or so I am informed) to pronounce mandarin <b d j g> as ㅃ ㄸ ㅉ ㄲ. Given this the challenge seems to be less finding out what the tense consonants are (simply interpreting them as /p t ts\ k/ ought to suffice for most purposes) and rather just what the hell ㅂ ㄷ ㅈ ㄱ are supposed to be. Several sources say they are aspirated at least in the Seoul dialect but those Seoul persons I have had exposure to do not seem to do this; what is clear is that the following syllable is generally pronounced with lower pitch than it would be after a tense or aspirate consonant.

Given also that these consonants voice or nasalize in some environment I think regarding them as phonemically voiced in some way is probably not the worst way to tackle this. Maybe you even need something ridiculous like

> b̥~b̥ʱ d̥~d̥ʱ ɟ̊~ɟ̊ʱ ɡ̊~ɡ̊ʱ

For my part I will just treat Korean as having /b p pʰ/ unless and until I find a good reason for not doing this.

Addendum: if the non-tense consonants are analyzed as phonemically (in some dialects breathy) voiced, where does ㅅ fit? With the voiced consonants or the aspirated ones? It definitely does not voice intervocalically, and I don't think it has the depressor thing going on, so I'd say the latter, but I'm not sure about the pitch thing; plus, typology bla bla bla crosslinguistic tendencies bla bla bla ideology bla bla rhabarber etc etc

[----]

FWIW when I was learning some Korean from a classmate last year, I pronounced the tense consonants as ejectives and he thought it was fine.

[Karinta]

Jesus, how does this even make sense? I've been trying to find sources on this for ages.

[Chengjiang]

@Hallow XIII: That makes sense. That said, I'm primarily asking about what the tense consonants are in an exact phonetic sense, not what they are in either a loose or a phonemic transcription.

In my own limited experience with Korean, which has mostly come from hearing L1 Korean speakers in the US and hearing K-pop singers (awesome sources, I know ), I definitely hear people aspirate ㅂ ㄷ ㅈ ㄱ word-initially at least some of the time, although it's nowhere near as consistent as the intervocalic voicing. I haven't really picked up the effect on pitch, but that's probably just because I wasn't listening for it.

[kodé]

Phoneticians have still not found a reliable auditory/acoustic correlate for Korean tense consonants. IOW, we have no idea what the hell's going on.

[finlay]

They sounded like straight-up unaspirated plosives to me.

[Bristel]

Looking it up on The Wiki offers no real advice on it, ofc.

But I didn't know that Korean's phonology was quite that complex, I'll have to borrow some ideas from it.

[sirdanilot]

^ The phonology is not complex. A three-way voicing contrast for stops is not 'complex', it's actually fairly common.

What is """"complex"""" is the phonetic realization. At least, it is difficult to parse the situation in Korean into our western-minded IPA phonetics system (for Koreans it's probably not even slightly complex). The phonation of the fortis consonants is hard to describe in our terms; is it ejective? Sort of... Is it voiceless? Yes, but that's not the unique feature here. Same goes for aspiration, gemination... I think I've seen them described as 'stiff voice' or something, whatever on earth that may be.

Perhaps 'emphatic' would be a nice term; though this is often used for pharyngeal consonants in Semitic languages, these historically descended from ejective or at least 'fortis' stops. Using vague terms leaves room for different realizations of these phonemes (I suppose there are differences within Korea with the realization of these stops?). But if you want an exact linguistic term to describe this, you're out of luck. Someone would have to invent a new one.

[Nortaneous]

The term is 'tense'.

I suspect this sort of thing is much more common than is currently realized -- plosive MOA contrasts (MOA in the sense of voicing/glottalization/etc; is there a more precise term for that?) are probably much messier than we think. It's already known to be the case in the Germanic languages (in English, the contrast is "aspirated except when they're something else vs. basically anything else" -- preglottalization/ejectivization is common for the fortis stops, and some dialects turn some of them to fricatives in some positions, and the lenis stops can be voiceless, voiced, implosive, approximants, or even weakly ejective), German allegedly has click allophones of certain plosive clusters, French voiced plosives can be implosive, every village in Armenia has a different realization of its plosive MOA contrast, and so on.

But languages with more than one set of plosives generally have a pretty clear fortis/lenis distinction of some sort, and (at least English-)speakers will interpret the fortis and lenis consonants of the language they're hearing as the fortis and lenis consonants of their own language. And between underdescription and many common languages (Spanish, Mandarin) probably actually having pretty clear-cut, easily-described contrasts, this isn't something that's been noticed.

Another problem, of course, is that, when there's sufficient variation within the language in the realization of the thing at hand, linguists will just invent a term for it ('alveolar approximant', 'sje-sound', 'tense consonant') and leave it at that.

[sangi39]

plosive MOA contrasts (MOA in the sense of voicing/glottalization/etc; is there a more precise term for that?)

Phonation?

[Nortaneous]

Phonation, airstream mechanism, and whatever is marked on the consonant but phonetically is distinguished mostly if not entirely on the vowel -- which is already the case for some phonations. (Especially breathy voice, and Javanese 'slack voice', which is realized as breathy voice on the following vowel. Are there any languages with a plosive series whose phonetic realization is creaky voice on the following vowel? Do any languages with ejectives have glottalization spreading from the ejective onto surrounding vowels?)

Another example of one of those terms is 'ballistic syllable'.

Canepari says:

The glottalized sequences /pʔ, tʔ, kʔ, tʃʔ/ are realized either with simultaneous glottalization and lengthening, (p̉:, t̉:, k̉:, t̉ʆ:), or with the laryngealization of subsequent vocalic elements ((V̦), including possible voiced C). /s/ (s) becomes [z̥] between V or between N and V; we also have the glottalized phonemic sequence, /sʔ/, which is realized as (s:) (or, if word-initial, as [sh] + (V̦).

But I don't trust Canepari and neither should you. He makes a point of not liking data.

[Hallow XIII]

nb: the whole fuss I made before was a direct consequence of Koreans interpreting my [p t tɕ k] as tense consonants; unless the Mandarin stops (which I think I used given that Asia and all) have something weird going on the fact that I am apparently incapable of producing Korean lax stops leads me to conclude that those are significantly weirder than the tense ones are (assuming that I am not using the phonetician-baffling tenseness feature subconsciously, and I don't know how I could) -- the salient feature appears to be less [+tense] than [-anything else].

It doesn't help that there doesn't seem to be a single good study of Korean phonology and/or phonetics around (why is the front rounded vowel there? who came UP with this?).

[Chengjiang]

(why is the front rounded vowel there? who came UP with this?).

IIRC the front rounded vowels of Korean are relatively short-lived monophthongizations of older diphthongs that have turned back into (different) diphthongs for younger speakers. (Specifically, [oj] > [ø] > [we] and [uj] > [y] > [ɥi].) From the sound of things they were never really part of the core vowel system. Notably they don't occur in diphthongs, unlike all the other vowels.

[Nortaneous]

Right, that's accurate, but also Korean's vowel system is pretty much what you would expect an alien to come up with, who knows the basic ideas behind vowels but not how they actually work. Middle Korean apparently once had only one front vowel and four central vowels, what the hell

[Hallow XIII]

I was referring specifically to charts of the Wikipedia variety where /2/ is included without /y/ (I have not yet found a single other paper that makes such assertions)

Also yeah Korean vowels are something I refuse to comment on because I found like three papers all complaining of the idiocy of other vowel analyses, including other papers calling people idiots. If you assume that eo is actually some sort of middish unrounded thing then at least Seoul Korean might come out as /a e i o u 1 @/ but of course it's never that easy (several papers point out that eo tends to be rounded and that is also the impression I get so it actually looks more like /a e E i 1 o O u/ except e E are merged by most people because they really can't get over the whole have less front vowels than other vowels thing can they).

[Chengjiang]

so it actually looks more like /a e E i 1 o O u/ except e E are merged by most people because they really can't get over the whole have less front vowels than other vowels thing can they).

Wikipedia claims many speakers merge short /e/ and /ɛ/ but keep long /eː/ and /ɛː/ distinct. Meanwhile, most of what I've read on the subject outside Wikipedia doesn't mention a length distinction in the vowels. What's going on?

Also, I hear you on people calling each other idiots. I think I got one of those papers when I tried looking. "An impertinent account of Korean vowels", yes?

Since we're just sort of talking about Korean phonology in general now, an additional question: What's with the hangul representation of /e/ and /ɛ/? It writes them as if they had been diphthongs /ʌj aj/ at some point, but that leads to, as has been pointed out, a very lopsided vowel system. "An impertinent account" claims that they were never diphthongs, and that digraphs were simply used for these to reduce the number of separate vowel glyphs required.

[linguoboy]

Wikipedia claims many speakers merge short /e/ and /ɛ/ but keep long /eː/ and /ɛː/ distinct. Meanwhile, most of what I've read on the subject outside Wikipedia doesn't mention a length distinction in the vowels. What's going on?

Length is regressive feature. It disappeared from the Seoul vernacular about a generation ago and the loss has been spreading outward from there. Merging /e/ and /ɛ/ is originally a feature of Gyeongsang dialect which found its way into the Seoul standard due to the influence of Gyeongsang-speakers there.

[Chengjiang]

Length is regressive feature. It disappeared from the Seoul vernacular about a generation ago and the loss has been spreading outward from there. Merging /e/ and /ɛ/ is originally a feature of Gyeongsang dialect which found its way into the Seoul standard due to the influence of Gyeongsang-speakers there.

Ah, OK.

Also, is it just me or are an awful lot of papers on Korean written by Altaicists?

[linguoboy]

Also, is it just me or are an awful lot of papers on Korean written by Altaicists?

That's surprising why?

[Chengjiang]

That's surprising why?

Maybe I'm out of touch, but I thought the Altaic hypothesis had fallen out of favor and was a decided minority view at this point.

[linguoboy]

That's surprising why?

Maybe I'm out of touch, but I thought the Altaic hypothesis had fallen out of favor and was a decided minority view at this point.

Maybe so, but you'd still expect Altaicists to be disproportionately interested in Korean.