User talk:Mr KEBAB/Archive 15

Hans Jørgen Uldall

I stumbled across the name of Hans Jørgen Uldall, which made me wonder how it was pronounced. I found a Forvo recording, but I honestly have no idea how to transcribe the last name ([ˈuldælˀ] or something?). I also wanted to look up a word ending in -all in Den Danske Ordbog, but I can't seem to find one that is not a borrowing from English. I'd like to know how you would transcribe it. Nardog (talk) 14:17, 21 February 2018 (UTC)

@Nardog: My guess is [ˈuldalˀ], but I can't hear stød very well. To me, the final consonant sounds awfully like [j], but Danes don't spell it ⟨ll⟩!

The second vowel is definitely [a] (=[a̝] / [æ] or [ɛ] in normal IPA) and not [æ] (=[ɛ] in normal IPA) nor [ɛ] (=[e] in normal IPA), as Danish ⟨a⟩ is never used for those (unless I'm missing something? Ask User:Schwa dk to confirm that).

I also don't know about the stress of the first two words. Mr KEBAB (talk) 14:29, 21 February 2018 (UTC)

Thank you! Yeah, the diphthongal quality threw me off too. I wonder if Danish /l/ can be vocalized or if prepausal consonants can be elided.

Really? I heard it to be pretty much like the cardinal [ɛ]. The F1 is around 450 Hz, which is to suggest it's actually even closer than the typical cardinal [ɛ]. I don't know if it's creaky either, but the final syllable has the lowest pitch and is noticeably darker on spectrogram, so it seems to be (which may be the reason the F1 is counterintuitively low, I don't know). Danish orthography says ⟨a⟩ is /a/ or /aː/, and Danish phonology says /a/ is [a~æ], so I thought it was nothing strange, but you know better than me.

Danish phonology#Stress says "[i]n names, only the surname is stressed", so we probably need not worry. Thank you also for suggesting contacting Schwa dk. Nardog (talk) 15:06, 21 February 2018 (UTC)

@Nardog: He could be speaking Standard Danish with some regional flavor. Or have a slight speech impediment he's unaware of. Who knows. Either way, [j] for /l/ doesn't look like textbook Standard Danish, but again - I could be wrong.

AFAIK, Danish stød can only appear on one sonorant (a long vowel or a consonantal sonorant), not on syllables. If it could appear on syllables, proper long diphthongs (such as [eːˀɐ̯]) would have the same type of creaky phonation as the shortened long ones (such as [eɐ̯ˀ]) and that just isn't the case. Natives can hear the difference between those two, even if they're in a more or less free variation (though [eːˀɐ̯] is more formal).

But there's the crucial difference between the quasi-IPA used for Danish transcriptions and proper IPA that is used to denote actual sounds in comparison with Jonesian cardinal vowels. When I write [a] in Danish transcriptions, it needs to be read as [a̝] or [ɛ]. I strongly suspect that there's a full merger between [a] and [æ] to [æ] (proper IPA: [ɛ]) in Modern Standard Danish. It's a bit strange that Danish phonology doesn't mention it, you can clearly hear it on recordings from Den Danske Ordbog.

The vowel does sound similar to cardinal [ɛ] and apparently ⟨a⟩ can also denote [æ] (again, Danish IPA). Sorry for the confusion. Apparently I didn't know that better than you at all ;)

What about Eli Fischer-Jørgensen? Her name was transcribed by Schwa dk himself. Mr KEBAB (talk) 16:17, 21 February 2018 (UTC)

@Nardog: Hmm, I could be partially wrong about stød though. In [eːˀɐ̯] it's definitely only on the long vowel, but in [eɐ̯ˀ] (and therefore also in [alˀ] or [ælˀ], whatever it is) it could actually be on the whole diphthong. I don't know. Mr KEBAB (talk) 17:06, 21 February 2018 (UTC)

I didn't know /a/ could also be [ɛ], so yes, you do know better. Aren't diphthongs phonologically long vowels too?

I was also having a hard time trying to figure out where and to what extent stød can occur. For example, Grønnum (1998) says For stød to occur in a stressed syllable, one of two conditions must be satisfied: either the vowel must be long, and the stød then tends to coincide with the end of the vowel, or the short vowel must be succeeded by a phonetically voiced consonant which then carries the stød. Syllables with short vowels succeeded by unvoiced consonants do not satisfy this condition and accordingly cannot have stød. Does she mean, say, different conditions apply in unstressed syllables? Does a short vowel succeeded by a consonant with stød also carry the stød, or just the consonant? (Which I assume she means the latter.) Basbøll (2005) spends most of its pages on phonological history rather than phonetics or phonotactics, so the answer may be there but I don't know where to look (and most of the content is way above my head!).

I also found a recording of the name of another person with the same last name, where I hear [l] at the end. When I input Uldall in Google Translate and listen to the synthesized speech, I don't hear [l] and instead a glide (my impression is the glide has an [e]-like quality in both the Forvo for HJU and GT rather than [i]). When I add a vowel to Uldall on GT, the glide becomes [lʔ], suggesting stød on [l], based on whatever technique it uses (I wish GT just right away gave the IPA, whether phonetic or phonemic; and obviously GT cannot be trusted at face value). Nardog (talk) 09:46, 24 February 2018 (UTC)

@Nardog: Is that a general question or are you asking about Danish? If the latter, I'm not sure if Danish has any phonemic diphthongs. [eːˀɐ̯] and [eɐ̯ˀ] are something like /eːˀr/ phonemically. If the former, that obviously depends on the analysis. It's probably true of English diphthongs.

In her Fonetik og Fonologi, I've seen Grønnum ascribe secondary (tertiary, etc.) stress to syllables with stød. So I wouldn't be so quick to judge the second syllable of Uldall as unstressed. It could be [ˈulˌdalˀ] or [ˈulˌdælˀ].

On second thought, no, I don't think that stød in [eɐ̯ˀ, alˀ, ælˀ] extends to the preceding vowel. If that was the case, I think that the alternatives with glottal stop [eɐ̯ʔ, alʔ, ælʔ] would be less likely (if not impossible) to occur.

Fun fact: cockney has phonetic stød. Compare better as [ˈbɛʔɐ] and as [ˈbɛˀɐ], with a creaky-voiced vowel. Maybe this makes cockneys more likely not only to correctly realize Danish stød (correct me if I'm wrong, but using a full glottal stop all the time sounds strange) but also to hear it like natives do (at least on vowels)? I'd like to see some research on that.

The Forvo guy sounds like he's saying [ˈuldæjˀl] (= [ˈuldɛɪ̯ˀl] in proper IPA). The same is true of the pronunciation you can hear in GT. The second syllable is a major mindfuck to me and I'd rather wait for Schwa dk's opinion.

Coda [j, w] are obviously not as close as [i̯, u̯]. Basbøll transcribes them [ɪ̯, ʊ̯], which are more or less synonymous with [e̯, o̯]. I wish IPA would just use [e, ø, o] instead of [ɪ, ʏ, ʊ], but that's another story.

Can GT be trusted at all? For what it claims to be capable of doing it still looks like a failed project to me. The fact that it gets the pronunciation of Uldall right could be simple luck. Mr KEBAB (talk) 15:45, 28 February 2018 (UTC)

I meant the latter. But I also forgot Danish didn't have phonemic diphthongs! (So I support your edit at the Help:IPA/Danish key, now it's much clearer.)

Even if creaky phonation could affect two segments in a row, would that necessarily mean that when stød is realized as a glottal stop it must also affect two segments?

It's worth noting [ɪ, ʏ, ʊ] (or [ɩ, ʏ, ɷ]) used to be defined as "lowered varieties of [i, y, u]", i.e. without centralization, until in the 1979 chart they appeared in the current positions. It's a bit strange that they are not mentioned in the 1970s JIPA articles discussing the revision of the IPA. Nardog (talk) 17:23, 4 March 2018 (UTC)

@Nardog: Thanks. I should improve the notes though...

Well, the fact that consistent realizing of stød by a glottal stop is considered pedantic tells you that it's highly probable that [ʔ] is the form from which stød evolved to the creaky voice we can hear nowadays (although proper glottal stops are still in use). AFAIK this has zero effect on phonemic analysis, because stød (and vowel length too) are considered to be features of Danish syllables, rather than individual segments on which stød appears. But I'd be happy to be proven wrong.

Yeah, these symbols are a bit strange, but maybe they are necessary after all. German has [i, iː, ɪ, e, eː] and their rounded front and back counterparts, Dutch is said not to have phonemic vowel length in native words, so we can write [i, y, ɪ, ʏ, e, ø] if we want, etc. But [ʏ] and [ø] sound literally completely the same to me. They're more similar to each other than [ɪ, ʊ] are to [e, o], probably because [ø] (like [y] and [œ]) is near-front in almost all languages I'm aware of. Mr KEBAB (talk) 21:17, 9 March 2018 (UTC)

@Nardog: Bingo! The source for the [a]-[æ] merger is Fonetik og Fonologi, pages 58 and 268. Grønnum says that it is one of the characteristics of lavkøbenhavnsk and that it's probable that Standard Danish is also in the process of merging them. Mr KEBAB (talk) 03:05, 11 March 2018 (UTC)

@Nardog: My most recent guess is [hanɕɶɐ̯n̩ ˈulˌdæːˀl] (I'm not sure about the middle name). You can read in Basbøll (2005) that long mid vowels are diphthongized in certain environments. By that logic, [ˈulˌdælˀ] would be a possible stylistic variant of underlying /ˈulˌdaːˀl/. But then, would speakers keep the closing diphthong in the latter variant? I'm pretty sure that it could be only interpreted as a variant of [æː], not its shortened counterpart.

There are so many question marks here. Mr KEBAB (talk) 16:17, 18 March 2018 (UTC)

Even Hans is dodgy, as Hans Christian Ørsted and Hans Basbøll transcribe it without stød and Hans Christian Andersen with it. Anyway, I hope you didn't mind my asking Maunus about this. Hopefully he can provide some insight. Nardog (talk) 17:09, 18 March 2018 (UTC)

@Nardog: Hans Basbøll is transcribed correctly, it's either Hans Christian Ørsted or Hans Christian Andersen that has wrong IPA (there's no reason for which they should be treated differently, Danish pronunciation is complicated and irregular but not that irregular). AFAIK, Swedes would treat Hans Christian as if it were one name (Hans-Christian or something like that, I'm pretty sure it'd receive the second toneme ([²hansˌkrɪstjan]) like other compounds). If Danes do the same, then the correct IPA of Hans Christian in the contex of full names is [hans kʁæsdjan], without stress and without stød. Hans Jørgen shouldn't be treated differently than that.

No, of course not. I'm curious to know the answer myself. Mr KEBAB (talk) 17:20, 18 March 2018 (UTC)

stød (and vowel length too) are considered to be features of Danish syllables, rather than individual segments on which stød appears. Then isn't that all more the reason to place the stød symbol after the syllable coda (perhaps unless it ends with a voiceless consonant)? But it's not unusual to see it even before a (non-syllabic) coda sonorant. Basbøll's moraic analysis is an attempt to justify this, if I understand correctly. (Also Grønnum 2015 talks about higher F0 in the beginning of stød syllables, so stød does indeed seem to have an effect on an entire syllable, including its first elements.)

Your recent contributions prompted me to check out Grønnum's publications, and—though I bet you've already seen it—I found Grønnum & Basbøll (2001) to be the most accessible (in English!) and succinct account on in what circumstances stød may or may not occur. But frankly the topic is so complex I begin to lose interest as soon as I try to understand... Nardog (talk) 03:01, 19 March 2018 (UTC)

@Nardog: I think one of the reasons for using the stød symbol instead of the primary stress mark is being consistent with the tone numbers used in Swedish and Norwegian. I think I like that approach.

I actually haven't, so thanks for the links.

Are you learning Danish? That should provide some motivation and, over time, an intuitive knowing of which syllables carry stød. Maybe try learning the rules then?

See also Volhardt (2011). The guy really did his research and I'll try to improve Danish phonology and Icelandic phonology using that source (I've already started). I read a little Danish and I know a few people that can confirm my (or GT's) translation, so that shouldn't be a problem. Mr KEBAB (talk) 04:13, 19 March 2018 (UTC)

Your ANI report

Just a friendly reminder that when you report a user to ANI you must notify them on their talk page. I have done so and noted it in the ANI thread. Cheers, Jip Orlando (talk) 12:19, 16 March 2018 (UTC)

@Jip Orlando: Thanks. I did notify him though, just in a discussion we had rather than in a separate thread. Mr KEBAB (talk) 12:24, 16 March 2018 (UTC)

A vowel quadrilateral is an "acoustic" chart?

At Talk:Australian English phonology#Vowel charts you said The cited article has only formant charts, not acoustic charts ... you need to find a source that has acoustic charts, but AFAIK all that formants represent is the acoustics of sounds, so any formant chart is by nature an acoustic one, whereas a vowel quadrilateral usually represents tongue positions. Are you sure you didn't mean to say articulatory, not acoustic? Nardog (talk) 20:17, 26 March 2018 (UTC)

@Nardog: I actually meant auditory. Mr KEBAB (talk) 20:42, 26 March 2018 (UTC)