Help talk:IPA/Danish/Archive 1

Archive 1

Archive 2

Untitled

Thank you to Hhbruun for creating this - I was hoping somebody would. (Now Dutch is the only IPA template without a key, AFAIK.)

A few possible changes - "louver" could be replaced by a more common word, e.g. "law"; [ɔː] could be illustrated with an English example (e.g. "law") since it is present in many English dialects; "bedeck" for [ə] could be replaced by something more common e.g. "about"; and [ɐ] could be illustrated with a non-rhotic example e.g. "but". Do we think these would work? Lfh (talk) 17:51, 12 January 2010 (UTC)

I based the account on the Wikipedia:IPA for Swedish and Norwegian - although the phonologies of these languages are very different. However, using pairs of words with common Norse origin as examples on both pages and using identical English words to examplify the nearest English phoneme enables comparison. I suspect that might interest somebody. In any case, the words "louver" and "bedeck" came from the mentioned page and the rarity argument could be applied there as well. As to the specific suggestions, I wonder if one could come up with a better English example than "law" - initial l in English being much "thicker" than in Danish. I kind of like the rhotic example for [ɐ] - I think the English "runner" phoneme comes pretty close to the Danish; "but" pronounced [bʌt] certainly doesn't, to my mind. Hhbruun (talk) 22:45, 12 January 2010 (UTC)

You're right about "louver" and "bedeck" being used elsewhere - that's interesting, I wonder why they were chosen, but I'm sure it was with good reason. Fair enough. "Runner" varies quite a bit by accent - the er would be R-colored for many speakers, e.g. Americans.

Well, thanks again for starting the page, and good luck with developing it. Lfh (talk) 14:08, 13 January 2010 (UTC)

Missing symbol

I just realized that [ɕ] is missing. Is it missing on purpose? (Danish phonology has it in parentheses). --Zahnradzacken (talk) 08:19, 8 April 2010 (UTC)

I believe that the occurrence of the phoneme [ɕ] in Danish is questionable. Hence, for simplicity or 'parsimony', it was left out. Hhbruun (talk) 18:54, 8 April 2010 (UTC)

The sign is still is used in Wikipedia articles, for instance on en:Zealand. If one clicks on the pronunciation (ˈɕɛˌlænˀ), one gets to this page, so I do not understand why ɕ should not be listed here. Do you have a different way of of pronouncing "sj" in Danish? Beil (talk) 16:12, 29 April 2010 (UTC)

This is not a list of phonemes (written between /slashes/) but phones (written in [square brackets]) used for transcribing Danish pronunciation. Even though /ɕ/ is not a phoneme in Danish, the phone [ɕ] is the typical allophone of /sj/ and should be in the list (which it is now).--Schwa dk (talk) 09:12, 11 March 2011 (UTC)

"Bedeck" not good for schwa

For many speakers, both with and without the roses/Rosa's phonemic distinction, the unstressed vowel in "bedeck" is a schwi rather than a schwa[1]. I know next to nothing about Danish phonology, but if the unstressed vowel in "begå" is ə rather than ɪ, a word like "about" would be better as an English equivalent. --Atemperman (talk) 18:23, 17 April 2010 (UTC)

The unstressed vowel in "begå" is [e] rather than [ə] in modern standard Danish (I've changed the example). I don't think an English equivalent of [ə] exists. Maybe an example from German? Schwa_dk (talk) —Preceding undated comment added 08:54, 11 March 2011 (UTC).

I'm pretty sure that about would work. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:34, 11 March 2011 (UTC)

Well actually in a narrow transcription Danish schwa varies greatly depending on surrounding segments, ranging from [ɪ ʏ ɐ ʊ]. This can't be covered with single example. The prototypical Danish schwa, however, is probably more advanced and rounded ([ə̟̹]) than a proper IPA [ə].--Schwa dk (talk) 09:36, 15 March 2011 (UTC)

Okay, but why would a German example be any better, then? And what are the contexts for each of those realizations? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 12:13, 15 March 2011 (UTC)

Possbile bad example "hel"

I'm learning IPA, so I'm not an expert, but I believe, that the example "hel [ˈheˀl] "whole"" is a mistake. Try go to Close-mid front unrounded vowel i.e. IPA [e] and listen. IPA [ˈheˀl] in Danish would mean "heel". Change the vowel to Near-close near-front unrounded vowel (listen sound file) i.e. IPA [I] and it should be fixed. But people with deeper insight, please validate this.

Don't use the Danish example at Close-mid front unrounded vowel as a reference, as I made it. —Preceding unsigned comment added by EinarSøndergaardRasmussen (talk • contribs) 15:42, 4 May 2010 (UTC)

Danish y-Sounds

These two lines are somewhat confusing to me. First of all, both Danish examples are transcribed as [yː] in their IPA representations, though the sign on the left seems to indicate a difference in quantity. Unfortunately, I don't speak any Danish, so I can't tell which example or transcription is correct.

As for the German equivalents: The difference between the vowel sounds in "über" ['yːbɐ] and "üppig" ['ʏpɪç] may be similar to the described difference in Danish, but there is also another (secondary?) difference in vowel quality ({long} close front rounded vs. {short} near-close near-front rounded). Hence, if there is only a difference in quality, I would omit the German example that corresponds to the short Danish y-Sound. The French example, on the other hand ("vue"), seems fine, as it is a pure {short} close front rounded vowel.

(To make sure that no hesitations arise as to whether my explanation of these vowel sounds in German and French are reliable: I'm a native speaker of both German and French. :-) )

88.77.155.192 (talk) 13:41, 30 July 2010 (UTC)

Yeah that was unclear. I've changed the Danish examples and switchted the French/German equivalents. I agree that "üppig" is not a good example since it's [ʏ], which the Danish sound isn't. I removed it.--Schwa dk (talk) 09:48, 15 March 2011 (UTC)

Stød?

Why isn't the Danish stød shown in the table? —Preceding unsigned comment added by 189.252.40.125 (talk) 05:22, 1 October 2010 (UTC)

Stød has been added. Schwa_dk (talk) —Preceding undated comment added 08:49, 11 March 2011 (UTC).

Bjarne Stroustrup

The phonetic equivalent of the "u" in the name "Bjarne Stroustrop" is missing from this page. It looks like an upside-down Greek omega with a small arc below it. I guess it must be pronounced rather like an English "w"? — Preceding unsigned comment added by Dynamitecow (talk • contribs) 15:45, 8 March 2011 (UTC)

That's true. [w] or [ʊ̯] can be used. The Danish Phonology page uses [ʊ̯] so I suggest using that symbol. Schwa_dk (talk)

"English" equivalents

The headline "Closest English equivalent" should be changed to "Closest equivalent" or the examples should only include examples from English--Schwa dk (talk) 09:45, 15 March 2011 (UTC).

(IP)Alphabetical order?

The ordering of the symbols seems rather random. Shouldn't it be put in some kind of alphabetical order? — Preceding unsigned comment added by Schwa dk (talk • contribs) 18:20, 23 April 2011 (UTC)

Diacritics on plosives

It seems that different contributors use different conventions on whether or not to include diacritics on plosives, ie. [pʰ tˢ kʰ b̥ d̥ g̊] vs. [p t k b d g]. This should be perhaps be streamlined, or at least it should be reflected in the chart (which I have now added) that different conventions exist. I'm not sure what is more suitable.--Schwa dk (talk) 09:16, 7 May 2011 (UTC)

Ok, but I dislike the current table as it implies that the use of IPA for Danish (DIPA) is somehow a separate transcription system from IPA. Since they're both clearly IPA, we should have one transcription column. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 15:54, 7 May 2011 (UTC)

I see, but there need to be some sort of translation between shortcuts used in Danish transcriptions and a narrow IPA transcription. Otherwise the Danish transcriptions would be misleading. Isn't it just a matter of using more suitable column headers?--Schwa dk (talk) 17:21, 8 May 2011 (UTC)

The problem is that all the IPA for X guides have a dissassociation between the transcription we use and narrow IPA transcription (except maybe French). Details about Danish phonetics go at Danish phonology, this is more of a general pronunciation guide for an English-speaking audience. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:05, 8 May 2011 (UTC)

I see your point. The narrow transcription is of academic interest only, which is not the intention of the guide. I'll remove the extra column.--Schwa dk (talk) 21:47, 8 May 2011 (UTC)

Will you add the narrow transcription to the article on the phonology of Danish? — Preceding unsigned comment added by 185.12.47.192 (talk) 13:46, 16 March 2013 (UTC)

I'd say it's there. Peter238 (talk) 12:13, 3 July 2015 (UTC)

The Queen's Danish

Schwa, what are some of the characteristics of the Queen's Danish that set it apart from the other accents of the language? Has anyone studied her pronunciation? I believe one of her sons also has a distinct way of enunciating words: is his accent similar to hers? Are there any books or journal papers which cover the way she speaks? How can such an accent of Danish be learnt? — Preceding unsigned comment added by 185.12.47.192 (talk) 13:42, 16 March 2013 (UTC)

I don't know of any thorough treatment of the queen's Danish. Basically, it's distinctly pronounced conservative Danish, so some older description of Danish will probably be adequate. Some characteristics is [a] instead of [æ], sometimes [a] instead of [ɑ]. Long vowels before approximants in syllables with stød. Consonants in coda should be more fortis, i.e. [ʋ] instead of [w], [ð] instead of [ð̞], [ʁ̞] instead of [ɐ̯] etc. I guess her coronal consonants are more fronted, apicals rather than laminals. Generally her lips seems more protruded and rounded than most speakers'. She is less inclined to use syllabic consonants (known as schwa assimilation in Danish literature), i.e. she will say [nə lə ðə jə] instead of [n̩ l̩ ð̩ ɪ] etc. Likewise with other allegro speech phenomena, but most non-native speakers wouldn't use these traits anyway (sloppy pronunciation takes practice). Her younger son, Joachim, speaks roughly the same conservative sociolect.--Schwa dk (talk) 14:47, 16 March 2013 (UTC)

Oh, I see! Thank you very much! That is fascinating. I will have to dig up some old book on Danish phonology, I wonder how I will manage to find it. Just to be clear, since the sounds here are in square brackets, by [a], for example, you mean a vowel cardinal 4 or close to cardinal 4? — Preceding unsigned comment added by 79.175.66.95 (talk) 15:00, 23 March 2013 (UTC)

ɡ/ɡ̊: Voiceless or voiced?

Entry "ɡ/ɡ̊" links to "Voiceless velar stop," whose article lists only "k" and variants and their corresponding characters, but is listed as having as an example the voiced "[g as in] good." Which is the correct value for this letter? 96.37.67.222 (talk) 00:38, 18 May 2015 (UTC)

I've changed the wrong examples. Peter238 (talk) 09:09, 13 June 2015 (UTC)

'Unstressed'

"ɪ: kage [ˈkʰæːə] 'cake'" Was this supposed to be [ˈkʰæːɪ̯ə] or [ˈkʰæː.ɪ]? 176.221.120.173 (talk) 18:20, 17 June 2015 (UTC)

[ˈkʰæːɪ], not sure whether this is one syllable or two. It's probably two. Peter238 (talk) 18:57, 17 June 2015 (UTC)

Yes, definitely two syllables (in non-allegro speech, of course). Kokoshneta (talk) 19:43, 16 August 2015 (UTC)

Stød

@Lohphat: How is there "a need for clarity" when we're already linking to stød? I object equating it to a glottal stop, since it's not the most common pronunciation, and it seems to falsely imply that it is (or, worse, that it's the only correct pronunciation). Care to actually address that? Mr KEBAB (talk) 22:31, 20 July 2016 (UTC)

EVERY heading on that page is in English and uses English terminology for the header. "Stød" is NOT English terminology; it should have at least an English term. EVERY Danish language course I've encountered refers to stød as the "glottal stop". And your usage of the word "pronunciation" isn't relevant to the issue, we're not talking about pronunciation but definition. Lohphat (talk) 21:18, 21 July 2016 (UTC)

@Lohphat: Please go read stød and then reply to me once again. I'm done repeating myself. Also, list the books you're talking about. "EVERY Danish language course I've encountered" is not a source. Mr KEBAB (talk) 21:31, 21 July 2016 (UTC)

æʊ̯, example

Surely [ˈhæʊ̯ˀ] is not a common pronunciation of hæv.__Gamren (talk) 13:39, 16 August 2016 (UTC)

The strict IPA rendering of that word is [ˈhɛʊ̯ˀ], as the vowel is open-mid. Maybe that's the source of your confusion. Danish front/central unrounded vowels are to a large extent transcribed with phonetically incorrect symbols, as [e, ɛ, æ, a, ɑ] are actually [e̝, e, ɛ, æ, ɑ̈]. However, that's what the sources use and that's what, I think, we need to stick to, however 'wrong' it feels to use such symbols. Mr KEBAB (talk) 15:05, 16 August 2016 (UTC)

Or maybe the source (Grønnum (2005)?) lists an old-fashioned pronunciation. Wiktionary transcribes it [ˈhɛʊ̯ˀ] which, again, in strict IPA would be [ˈheʊ̯ˀ] (the vowel is close-mid). Den Danske Ordbog transcribes the infinitive form of that verb with [ɛː] (which confirms that it is close-mid), so you're probably right. Which word would you like to see there instead? Mr KEBAB (talk) 15:13, 16 August 2016 (UTC)

As a Wiktionarian myself, I am quite sure that we do not replace [ɛ] with [æ] or [e] with [ɛ].
As for an alternative: I have some difficulty in telling apart [æ] and [a] (not least because of these competing notations!), but I would suggest drev, snav!, lav etc.
Also, Grønnum transcribes hæv as [-ɛw] (p. 294), while lav is transcribed [-æw].__Gamren (talk) 15:56, 16 August 2016 (UTC)

No idea whether you do, but I'd expect that to be the case, as [e, ɛ, æ, a, ɑ] is the 'normal' transcription of these vowels, at least as far as the books I've read are concerned. It's not so much "replacing [ɛ] with [æ] or [e] with [ɛ]", as it is a "normalized Danish IPA", or what I sometimes call a "Grønnum-Basbøll IPA-based transcription" - see the first pages of Basbøll's "Phonology of Danish" to see what I'm talking about. If you ask me, transcribing e.g. a close-mid vowel with the symbol [ɛ] is a very bad idea (introduces massive confusion for people who are e.g. learning Danish and German at the same time), yet it's common. Just another inconsistency in this world.

[æ] and [a] (strict IPA: [ɛ, æ]) seem to have merged in Modern Standard Danish (I don't have a source, sorry, just my ears judging DDO recordings), so that it may very well be the case that dictionaries prescribe an artificial, outdated contrast, which is the source of your confusion. Don't quote me on that though.

Changed to lav. Mr KEBAB (talk) 16:26, 16 August 2016 (UTC)

[ʌ]/[ɐ]

These are identical in my idiolect, and, Danish phonology leads me to believe, in most language forms. I have no objection to the differentiation, however, if there is a simple rule for when to use which one. Is it a simple matter of stressed/unstressed?__Gamren (talk) 21:01, 12 February 2017 (UTC)

@Gamren: Basbøll says that the merger of [ɐ] and [ʌ] is probably usual.

[ɐ] can only be unstressed (as [ə ɪ ʊ]), and, phonemically, it is any of the following sequences: /ər, rə, rər/ (as in læger - lære - lærer, all of which are [ˈlɛːɐ]). Thus, all you need to do is to look for 'r' in spelling to find out whether the vowel in question is [ɐ] or [ʌ]. It's an extremely reliable method, but there could be some exceptions.

[ʌ] is phonemically /ɔ/, the short counterpart of /ɔː/, and I guess that the most common spelling of this vowel is ⟨o⟩, but ⟨å⟩ is also possible (as in the case of måtte [ˈmʌd̥ə]). Mr KEBAB (talk) 21:45, 12 February 2017 (UTC)

I have not occupied myself with phonology, so those explanations are largely lost on me. As for your r-rule; do råt, råstof, rondo, rok etc. comply? If not, might it be more prudent, perhaps, to check for the presence of unstressed <e> (this being a sincere question)?__Gamren (talk) 22:05, 12 February 2017 (UTC)

@Gamren: I probably don't know much more about Danish phonology than you do.

They don't - [ɐ] doesn't appear in stressed syllables. I think the only spellings you need to look out for are -er(-), -re(-), -rer(-). The e must be there, and the vowel must be unstressed. Again, this may not be a complete explanation, but it's a decent start.

You know, I'm starting to question whether [ɐ] and [ʌ] are different vowels at all (the fact that they're different on a phonemic level is probably indisputable). Are there any minimal pairs? Are there any words in which [ʌ] is always unstressed? Råstof may be one, but I'm not sure whether there isn't a secondary stress on -stof. There are words like og [ʌ] that are often unstressed when you speak in full sentences (as opposed to words in isolation), but that's also true for many other words that don't necessarily contain [ʌ] (or [ɐ]) at all. Mr KEBAB (talk) 22:46, 12 February 2017 (UTC)

A rant

Can someone explain to me why English æ in "hat" is equivalent to Danish a, English ɛ in "bet" is to Danish æ, English ɪ in "kit" to e, and English e(ɪ) to Danish ɛ? What's the point of IPA if each symbol means something completely different in each language? I have no idea about Danish phonology...

Wait, if you have no idea then why are you commenting?

I thought the point of this chart was to help people who don't know Danish phonology how to pronounce Danish names on Wikipedia. If it's only for people who already know Danish phonology then it's pointless because, well, why do they need a Help:IPA for Danish page if they know Danish? I know IPA so when I saw a name containing /æ/ I tried to pronounce it with /æ/, but then I went to this page and learned that maybe like "e" in English "bet"... So what's the point of IPA. Maybe use SAMPA... Or add some audio files to at least hear how these sounds sound like... — Preceding unsigned comment added by 83.26.139.98 (talk) 19:00, 5 March 2017 (UTC)

Because [a, æ, ɛ, e] is a broad transcription of these vowels. A narrow transcription is [æ, ɛ, e, e̝]. It's misleading, yes, but that's how reputable sources transcribe them, and we just follow that practice. Mr KEBAB (talk) 19:05, 5 March 2017 (UTC)

• In all languages there is some degree of adjustment to the needs of the specific language when using IPA. You should always be aware of that when reading IPA transcriptions, they may be more or less wide/narrow. The problem with Danish is that there are more vowel phonemes than there are vowels in the IPA chart. Therefore some of the vowels are pushed one spot, so to speak, so a vowel may be represented with a symbol intended for a more open vowel. Alternatively you would have to graphically merge two vowels that are phonologically distinct, or come up with some new symbols or some other workaround. --Schwa dk (talk) 13:31, 7 January 2019 (UTC)

Plosives

@Mr KEBAB: I sort of have a problem with your edit because, as much as it's probably accurate just in terms of describing how native speakers produce the sounds in Danish, when someone who speaks English reads Nikolaj Coster-Waldau, for example, goes to this page and tries to recreate the sound, guest could very well mislead them into thinking Nikolaj is pronounced like [nɪgolaɪ]. Also, we're specifically describing the prototypical sounds on these IPA key pages, so removing the notes just because "Danes sometimes use those sounds" I don't think is really helping either. Bring back the notes or restore spare etc. I don't see how that's "overcomplicating". Nardog (talk) 05:45, 10 July 2017 (UTC)

@Nardog: /nɪɡoʊlaɪ kɒsdər ˈvældaʊ/ is a perfectly reasonable approximation of the Danish pronunciation. The problem is that you're deceived by Danish spelling. 'k' in Nikolaj belongs to the /ɡ/ phoneme because it's unaspirated, its voicing is irrelevant. Whether you pronounce that word with a voiceless or voiced /ɡ/ doesn't change anything. BTW, I consider [b̥, d̥, ɡ̊, kʰ, pʰ, tˢ] to be an unreasonable pseudo-narrow transcription. They don't differ in anything but aspiration (or affrication in the case of alveolars). We should switch over to [b, d, ɡ, k, p, t] as some sources have already done. Diphthongs ending in [-ɪ̯, -ʊ̯] should also be simplified to [-j, -w] per Grønnum (2005). We can also simplify [ʋ] to [v] since they're non-contrastive and variable. Mr KEBAB (talk) 07:35, 10 July 2017 (UTC)

We can use DDO for all of these simplifications. Mr KEBAB (talk) 10:10, 10 July 2017 (UTC)

Hmm, on second thought, I'm not sure whether Nikolaj is phonemically /neɡolaj/ or /nekolaj/. According to Danish phonology, it depends on how it's normally syllabified. If the first syllable is Ni-, the correct analysis is the former one. If it's Nik-, the latter analysis is correct. That is because coda /ɡ/ is an approximant [j ~ w] (depending on the preceding vowel). But that doesn't change anything - phonetically, it is an unaspirated, lenis, voiceless velar plosive. Mr KEBAB (talk) 12:45, 10 July 2017 (UTC)

@Mr KEBAB: I guess you have a point insofar as there's no contrast between voiced and voiceless unaspirated plosives in Danish, but as far as the notation [b̥, d̥, ɡ̊] is concerned, it is by definition declaring that they're voiceless, so it is counterintuitive to list ball etc. as "equivalent" because whilst they may be devoiced utterance-initially, it is still /b/ phonologically and that's what most English-speaking people perceive it as.

So you can go ahead and make the change per DDO as far as I'm concerned. Nardog (talk) 13:12, 10 July 2017 (UTC)

@Nardog: I'm not denying that. What I have a serious problem with is the inconsistency and confusion that it introduces. Why write [b̥, d̥, ɡ̊, kʰ, pʰ, tˢ] if we're transcribing the close-mid front unrounded vowels with [ɛ, ɛː]?! It's an effect of a cult-like adherence to the "phonemic transcriptions must be as simple as possible" dogma that obviously also influences phonetic transcriptions. And let's not mix phonetics with phonology - the fact that the main allophones of Danish /b, d, ɡ/ are voiceless (at least in theory, they can be voiced as well) doesn't mean that the rest of them are (and AFAIK... they're all voiced: [w, ð, j ~ w]). That's one of the reasons why they're written /b, d, ɡ/ and not /p, t, k/.

Maybe we should change closest equivalent to English approximation? Mr KEBAB (talk) 13:40, 10 July 2017 (UTC)

@Mr KEBAB: That per se doesn't solve the problem. I guess what I'm getting at is, it is weird to describe the prototypical (i.e. underlying) allophone of a phoneme to English speakers with a non-prototypical allophone of an English phoneme because few speakers, native or otherwise, notice the difference in the first place. So long as you're describing unaspirated voiceless plosives, spy, sty, sky, etc. are about the best you could hope for to compare them with because they are at least more persistent to phonetic environment and therefore less "narrow" than ball, done, etc. Nardog (talk) 14:36, 10 July 2017 (UTC)

@Nardog: Ok, go ahead and make the change. It's too trivial to debate. Mr KEBAB (talk) 17:03, 10 July 2017 (UTC)

@Mr KEBAB: Thanks. And yes, I've been called out for being fussy my entire life. Nardog (talk) 17:26, 10 July 2017 (UTC)

@Nardog: What we (or at least I) didn't consider are words like skinnede [ˈsɡenð̩ðə] which have a mandatory phonetically voiceless onset. Pronouncing skinnede as [ˈzɡenð̩ðə] (with a Russian-like voicing assimilation) does sound ridiculous and non-native, and the examples I chose could lead at least some people to pronouncing it like that. Your examples may be better after all. Mr KEBAB (talk) 17:53, 10 July 2017 (UTC)

The last vowel in Nikolaj is open central [ɑ]. Just so you know, I've just checked Forvo. Kbb2 (ex. Mr KEBAB) (talk) 14:36, 10 January 2019 (UTC)

Sort

"Sort" is used as an example in the table of diphthongs. This is not the best example, since Danish has two words "sort", one meaning black, the other meaning sort or cultivar, with different vowel sounds. If the first is intended, "gjort" would be a better example, if the second, "kort". Note that "bort" has the same problem as "sort". --Klausok (talk) 08:01, 11 July 2017 (UTC)

@Klausok: Done, but if there aren't any objections I want to remove the diphthongs altogether and just list them in a footnote. They're non-phonemic and the corresponding English approximations are often rather awful. English speakers just have to do their homework and learn to put [j, w, ɐ̯] after Danish monophthongs. Mr KEBAB (talk) 10:40, 11 July 2017 (UTC)

Done. Sorry for the delay. Mr KEBAB (talk) 16:31, 2 March 2018 (UTC)

Move discussion in progress

There is a move discussion in progress on Help talk:IPA which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 16:16, 15 July 2017 (UTC)

The set of vowel symbols we use is misleading

Hello. I'd like to propose a major (although consistent) change to the way we transcribe some of the Danish vowels. Currently, we write them with ⟨ɑ, ɑː, a, æ, æː, ɛ, ɛː⟩, ⟨ɶ, ɶː, œ, œː⟩ and ⟨ɒ, ɒː, ɔ, ɔː⟩. I propose that we change them to ⟨a, aː, a̝, ɛ, ɛː, ɛ̝, ɛ̝ː⟩, ⟨œ, œː, œ̝, œ̝ː⟩ and ⟨ɔ, ɔː, ɔ̝, ɔ̝ː⟩. In my opinion, the current set of symbols is just too broad for our purposes and there have been numerous complaints (all of them understandable) in the past regarding its lack of phonetic accuracy. AFAICS, there's actually little point in putting Danish IPA in our articles at all when we use ⟨æ⟩ for what is actually consistently realized as an open-mid vowel (or even worse - ⟨ɛ⟩ that's consistently close-mid) or when we unnecessarily use exotic symbols such as ⟨ɶ⟩ when a simple ⟨œ⟩ would be more than sufficient. I also can't get my head around the idea that sources that use such a simplistic notation would actually bother to put diacritics on all of the 6 Danish plosives (which we can restore if we change the vowel symbols - we can return to writing them with ⟨b̥, d̥, ɡ̊, kʰ, pʰ, tˢ⟩).

The change I'm proposing is based partially on the system used in Haberland (2013) and partially on the narrow transcription of Danish vowels in Basbøll's Phonology of Danish. What Haberland writes with ⟨e̞, e̞ː, ø̞, ø̞ː⟩ I write with ⟨ɛ̝, ɛ̝ː, œ̝, œ̝ː⟩, which mean the same but the diacritic is put on the symbols that are typically used to transcribe these vowels (that's the idea actually - to put the raised diacritic on any vowel symbols that needs to be freed up to be used for a phonetically closer vowel). I'm using ⟨ɛ̝, ɛ̝ː⟩ even though the vowel is consistently close-mid in order to make the transcription consistent with the way we would write the other four true-mid vowels (they actually have about the same height in conservative Danish) as well as with the near-open front vowel which we now write with ⟨a⟩. The raising diacritic also doesn't really tell us how much raising is involved, it could be raising from open-mid to mid but also from open-mid to close-mid or even higher.

He also writes what we currently transcribe with ⟨æː⟩ with ⟨ɛː⟩ - phonetically, this is spot on.

Basbøll narrowly transcribes every single Danish vowel and so we just need to pick the most important symbols and leave out the excessive detail. I propose that we take from his transcription ⟨œ̝, œ̝ː, œ, œː⟩ (for what we currently write with ⟨œ, œː, ɶ, ɶː⟩), ⟨ɔ̝, ɔ̝ː, ɔ, ɔː⟩ (for what we currently write with ⟨ɔ, ɔː, ɒ, ɒː⟩) as well as ⟨a̝⟩ for what we currently write with ⟨a⟩. The last change will allow us to write the open central vowels with ⟨a, aː⟩, which is a very slight OR - basically, Basbøll writes these as [ɑ̈, ɑ̈ː] in narrow transcription and calls them centralized open back unrounded vowels (which basically means central, AFAICS). The equivalent transcription is [ä, äː] and because there are no true open front vowels in Danish (I'm not counting the conservative variety) we can safely drop the diacritics. After all, this is the vowel quality found in so many languages such as German, Spanish, French, Italian, Portuguese, Russian, etc. It makes more sense to write them with ⟨a, aː⟩ and it gives us more reason to write ⟨a̝⟩ as such as recently there was a complaint that [a] is a very broad transcription of that vowel (Danish [a] is much more front and close than the vowels found in the aforementioned languages, though Russian /a/ when it occurs between palatalized consonants comes pretty damn close to it). Note that it can only ever be short, so that transcribing the long open central vowel as [aː] won't be confusing in any way. The traditional analysis is that [a, æː, ɑ, ɑː] belong to one phoneme anyway (AFAICS vowel length is analyzed as a feature of the syllable, as is stød).

If that's not an acceptable solution (for whatever reason), we can write the near-open front vowel with ⟨æ⟩ and the open central vowels with ⟨ɑ, ɑː⟩, but IMO this is a less optimal solution - see below.

I also propose that we don't differentiate anymore between [ʌ] and [ɐ] and write both with ⟨ʌ⟩, per DDO. Native speakers tend to hear no difference between the two.

The problem with the current set of symbols is that not only does it teach to pronounce Danish in an old-fashioned manner but it can also slow learners' progress of learning to understand spoken Danish. I thought this through and if we're going to change anything, I think that this set of symbols is the best. We're basically dropping ⟨æ, æː, ɶ, ɶː, ɑ, ɑː, ɒ, ɒː⟩ as well as ⟨ɐ⟩. The idea behind that is to use the presence of these in any given transcription as an indication that it uses the imprecise set of symbols that is used elsewhere, so that more people will be able to correct the IPA - which, after all, should be helfpul to our readers (already there are multiple editors who dislike the presence of IPA in the lead, so when it's kept there we better make sure that it's accurate).

Pinging @Aeusoes1, Gamren, Maunus, and Nardog:. See User:Kbb2/IPA for Danish for an idea of how the change would look like. I can go through our articles and change the transcription when we reach a consensus to do so, that's no problem. Kbb2 (ex. Mr KEBAB) (talk) 11:00, 7 January 2019 (UTC)

Also @Schwa dk:.·maunus · snunɐɯ· 11:09, 7 January 2019 (UTC)

• I originally started using Dania transcriptions some thirteen years ago, because I was using Nina Grønnum's phonology book. I personally find the use of IPA [ɛ] for the vowel in words like "gade" to be misleading and erroneous, and I don't think the pronunciation with the lower vowel [a] (or [a̝]) is more conservative. I disagree with the idea of transcribing r-colored a as long - especially before stød because if anything it is perceptually shorter. I also think transcribing r-colored a as [a] is wrong, even given your rationale and example, because the English equivalent you give ("art") is in fact transcribed with ɑː according to Help:IPA/English. Similarly [ɛ] for r-colored i in "frisk" strikes me as wrong, since this for me, and I think most commonly, has exactly the same quality as the [a] in "kat" ("kaste" rhymes with "riste"). But aside for my own opinions and preferences here, really I think the only approach we can take is to take one authority and follow it consistently. Basbøll is as good as any. ·maunus · snunɐɯ· 11:28, 7 January 2019 (UTC)

• @Maunus: Why is it erroneous? The vowel is open-mid according to all vowel charts I've seen (and Basbøll himself). Are you perhaps mistaking my change of symbols for a different phonemic identification? Because what's usually written [ɛ, ɛː] in IPA transcriptions of Danish is written with ⟨ɛ̝, ɛ̝ː⟩ here.

I'm just following the sources and the way we already transcribe them.

It's an approximation, not an equivalent. English /ɑː/ has a huge allophonic variation, depending on the dialect. Help:IPA/English isn't even phonemic but diaphonemic. You shouldn't look at the symbols used there and assume that they're equivalent to the canonical IPA values of the sounds. Kbb2 (ex. Mr KEBAB) (talk) 13:04, 7 January 2019 (UTC)

Because it produces a pronunciation that I don't recognize. I don't believe anyone would transcribe the vowel in RP English "cat" as [ɛ], and this is the vowel I have in gade - it is lower than that. I think it is better to do as Schwa suggests below and merge a and æ to æ.·maunus · snunɐɯ· 13:31, 7 January 2019 (UTC)

@Maunus: I genuinely don't know what to tell you. Den Danske Ordbog transcribes the word as [ˈɡæːðə] and the corresponding recording has a very clear [ɛː] vowel. It's not [æː], although that's how you would transcibe it in broad transcription. See [2] and [3]. Basbøll also describes it as open-mid. Kbb2 (ex. Mr KEBAB) (talk) 13:37, 7 January 2019 (UTC)

It is an irrelevant point, I think - the only viable solution from my perspective is to follow a source. I am fine with transcribing it as [ˈɡæːðə] like Basbøll and DDO does. Unless you find an authority that uses [ɛ] instead of [æ] to transcribe this vowel in Danish and which we decide to follow, I am not supporting the proposal. It probably is mostly an issue as sSchwa says of preferring separate glyphs for separate phonemes. But even if I recognized your proposal as better I wouldn't support it without a source. Wikipedia should not be a medium for the circulation of original transcriptions.·maunus · snunɐɯ· 14:48, 7 January 2019 (UTC)

@Maunus: It is relevant, and I've already found such a source - it's Basbøll. I'm saying that we should follow his narrow transcription in specific instances like the ones I've mentioned, save for the close-mid vowel that's traditionally been transcribed with ⟨ɛ⟩ - that we can write with ⟨ɛ̝⟩ instead of ⟨e⟩, although we could also use the latter symbol, in which case the vowel traditionally transcribed with ⟨e⟩ would have to be written with ⟨e̝⟩. If you read my original message thoroughly you'd see that there's very little to no OR there.

The transcription system we use here isn't phonemic and IMO we shouldn't be concerned with phonemes at all in this guide. This guide is supposed to help non-native speakers of Danish pronounce the language and hopefully also facilitate understanding spoken Danish. The current state of it is unacceptable to at least a few Wikipedians (I'm also counting myself) and I'd argue that at least a few symbols are heavily misused here in comparison with how they're normally used in IPA and with the canonical IPA values of these symbols. If we won't reach a consensus to change the guide in some way then I think it'd be a much better idea to replace the pseudo-IPA we have now with the Dania transcription. Because even you until very recently have been misled by the symbol ⟨æ⟩ in Danish transcriptions, which is very different to how ⟨æ⟩ is used in transcriptions of English. Danish [æ] is the same as German [ɛ], or English [ɛ] in many dialects (although many dialects can also have a higher vowel). English [æ] is lower than Danish [æ]; it's either a true [æ] or front [a], close to Danish [ɑ] which is a central vowel (not a back one). In modern RP, it's the latter. Can you see now why we really need to revise this system?

The open-mid phonemes /ɛ, ɛː, œ, œː, ɔ, ɔː/ can be written /ɛ̝, ɛ̝ː, œ̝, œ̝ː, ɔ̝, ɔ̝ː/ in phonemic transcription in order to match the revised transcription of the main allophones. This is why I prefer to write the close-mid front unrounded vowel with ⟨ɛ̝⟩ rather than ⟨e⟩. I see no problem there. Kbb2 (ex. Mr KEBAB) (talk) 15:56, 7 January 2019 (UTC)

No, I can't see that because I don't agree that Danish [æ], if this is the vowel found in Danish gade/kat/ and the letter name "a" has the same sound as German [ɛ], or English [ɛ]. Rather it has something closer to a (open or near open front vowel) as its most common realization in standard Danish - which is also why Basbøll and the DDO transcribes the word [ˈɡæːðə] and not ˈɡɛ̝ːðə - and hence why it would be a bad idea to transcribe it as such since that would exactly mislead the reader to produce a higher vowel. Either we follow Basbøll or we follow someone else, we can't just change his transcriptions because we disagree with them. ·maunus · snunɐɯ· 21:57, 7 January 2019 (UTC)

@Maunus: What you're saying is clearly at odds with Danish phonology. The vowel is open-mid in mainstream in Standard Danish and the fact that you refuse to acknowledge that is probably the proof that you've been mispronouncing some of the vowels in English (certainly [æ]). Also, please don't confuse transcriptions - I'm saying that gade is pronounced [ˈɡɛːðə], not [ˈɡɛ̝ːðə] as [ɛ̝ː] is a different vowel.

Look, I want to stop arguing about this. Basbøll does say that the vowel is open-mid and you're continuously confusing narrow and broad transcriptions. You should do more research.

Here's a screenshot from page 45. Kbb2 (ex. Mr KEBAB) (talk) 03:43, 8 January 2019 (UTC)

Though I don't see why you would need to be rude about a simple disagreement, Schwa has convinced me that the ɛ transcription is more accurate for the vowel I want to transcribe as æ. So sure, carry out the proposal, I won't stand in the way. ·maunus · snunɐɯ· 10:16, 8 January 2019 (UTC)

There is no doubt the vowel in 'gade' most closely corresponds to IPA [ɛ], but there is a tradation in Danish (e.g. Den Danske Ordbog and many other) to use ⟨æ⟩ instead, due to the principle of using distinct glyphs for distinct phonemes, and with the many vowel phonemes in Danish, the other glyphs are already in use.--Schwa dk (talk) 11:25, 8 January 2019 (UTC)

@Maunus: I apologize if I sounded offensive. Perhaps I just got overly frustrated about the need to explain something that I felt was already explained, but perhaps it was me who didn't do a very good job at explaining the issue. Again, my apologies. Kbb2 (ex. Mr KEBAB) (talk) 11:29, 9 January 2019 (UTC)

• I agree that some of the vowel symbols currently used are misleading, and I am in favor of a revision of some sort. I would like to propose the use different glyphs for different phonemes, though. Using only diacritics to differentiate phonemes is understating the importance of the distinction as readers tend to see diacritics as less important or even disregard them altogether. For what it's worth, in my own work I have begun to substitute the following, which may be of inspiration, some of it along the line of what is being proposed:

• ⟨e⟩ > ⟨e̝⟩
• ⟨ɛ⟩ > ⟨ɛ̝⟩
• ⟨a, æ⟩ merged to ⟨æ⟩ (or ⟨æ̝⟩, modern speakers tend to merge these, or at least struggle to perceive the difference)
• ⟨ɑ⟩ > ⟨a⟩
• ⟨œ⟩ > ⟨œ̝⟩
• ⟨ɶ⟩ > ⟨ɶ̝⟩
• ⟨o⟩ > ⟨o̝⟩
• ⟨ɔ⟩ > ⟨ɔ̝⟩
• ⟨ɒ⟩ > ⟨ɒ̝⟩
• ⟨ʌ, ɒ⟩ merged to ⟨ɒ̝⟩ (modern speakers struggle to perceive the difference. [ʌ] as Daniel Jones pronounces it is nowhere near the vowel it traditionally represents in Danish textbooks. Traditionally, ⟨ʌ, ɒ⟩ represent short/long versions of the same phoneme, and the short version (ʌ) might be slightly less rounded than the long version (ɒ). The short (ʌ) is even, if anything, more open than the long vowel, opposite to what the glyphs suggest. I do not think the difference is prominent enough to justify two different symbols)

However, I am not in favor of merging ⟨ɐ⟩ and ⟨ʌ⟩, as these are different phonemes that behave very differently. It is true that if you stress /ɐ/, then it is pronounced like /ʌ/. But it is only stressed in metalinguistic contexts. In natural speech /ɐ/ varies between [a]~[ɒ] depending on context, which /ʌ/ does not, and it is a huge mistake to merge these in my opinion. By the same argument you could propose merging ⟨ə⟩ and ⟨œ⟩, which would be an equally huge mistake.

This way you gain a transcription that is both more accurate and comparable to the usage of IPA in other languages while still respecting the phonological contrasts in Danish by using distinct glyphs for different phonemes.Schwa dk (talk) 12:31, 7 January 2019 (UTC)

@Schwa dk: I agree with changing ⟨ɛ, œ, ɔ, ɑ⟩ to ⟨ɛ̝, œ̝, ɔ̝, a⟩, but I'm not sure about the rest.

I think that the diacritics on what you propose to write as ⟨e̝⟩ and ⟨o̝⟩ are rather redundant. If we can drop the centralization diacritic from ⟨a, aː⟩ (older ⟨ɑ, ɑː⟩, which also require a centralization diacritic in narrow transcription), we can also drop the raising diacritic from ⟨e̝⟩ and ⟨o̝⟩. My rationale for using the diacritic can be found above.

I disagree with the way you write what we now transcribe with ⟨æ, ɶ, ɒ⟩. I think writing them with [æ̝, ɶ̝, ɒ̝] when ⟨ɛ, œ, ɔ⟩ are available is to rather massively miss the point. This guide is supposed to be helpful to our readers. If they're familiar with German or French, they can safely use their [ɛ, œ, ɔ] for Danish (although maybe not the last one, especially in the case of French - French [ɔ] tends to be rather noticeably centralized, at least in Paris). I think it's unfortunate that there already are phonemes that are written with ⟨ɛ, œ, ɔ⟩, but this can be easily bypassed by writing these symbols with ⟨ɛ̝, œ̝, ɔ̝⟩ in phonemic transcription. It's only partially OR - yes, these symbols are probably never used in phonemic transcription, but they can be used to correctly transcribe the main allophones of these phonemes, especially the long ones and especially the last two (the main allophone of /ɛ/ is close-mid, but I think that writing it with [e] would be an overkill - see above for an explanation).

If readers tend to see diacritics as less important or even disregard them altogether then I bet they're gonna read ⟨æ̝, ɶ̝, ɒ̝⟩ as if the diacritic wasn't there. Which isn't a good thing either.

I don't know whether we should cease to differentiate what we write now with ⟨a⟩ and ⟨æ⟩ as well as ⟨ʌ⟩ and ⟨ɒ⟩. I think [ʌ] has been traditionally considered to be the short variant of /ɔː/ (phonemically /ɔ/), no? Kbb2 (ex. Mr KEBAB) (talk) 16:04, 7 January 2019 (UTC)

I'm in favor of greater phonetic accuracy, but I'm hesitant to use diacritics if we can avoid them. What would be wrong with ⟨a, aː, æ, ɛ, ɛː, e, eː⟩, ⟨œ, œː, ø, øː⟩, and ⟨ɔ, ɔː, o, oː⟩? — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:15, 7 January 2019 (UTC)

@Aeusoes1: You missed some of the symbols. In Danish, there are [æ, æː, ɶ, ɶː, ɒ, ɒː], [ɛ, ɛː, œ, œː, ɔ, ɔː] and [e, eː, ø, øː, o, oː]. [ɛ, ɛː] are actually close-mid and [e, eː] are actually near-close. My proposal is that we write the first set with ⟨ɛ, ɛː, œ, œː, ɔ, ɔː⟩, the second one with ⟨ɛ̝, ɛ̝ː, œ̝, œ̝ː, ɔ̝, ɔ̝ː⟩ and the third one with exactly the same set of symbols as originally ⟨e, eː, ø, øː, o, oː⟩, so that 6 symbols would undergo a rather drastic change, 6 would undergo a cosmetic change (adding the diacritic) and the rest would be left unchanged (save for what we write with ⟨a, ɑ, ɑː⟩, which I would write ⟨a̝, a, aː⟩ - again, one cosmetic change and two drastic ones). There would be 8 drastic changes and 7 cosmetic ones.

Personally, I don't want the changes to be too drastic. Actually, if you look at it, the raising diacritic tells you in each and every case that the symbol that carries it is usually the one used for writing another vowel, save for the diacritic (compare [ɛ̝] vs. [ɛ] - by looking at the guide you'd know that they are normally written with ⟨ɛ⟩ and ⟨æ⟩). That's also the reason I'd write the near-open vowel with ⟨a̝⟩, although ⟨æ⟩ would be a perfectly defensible choice (and it's possible that it's better, I can see your POV - as Schwa dk said, modern Standard Danish tends not to differentiate between the quality of what is traditionally written with ⟨a⟩ and ⟨æː⟩, with both being phonetically open-mid). Kbb2 (ex. Mr KEBAB) (talk) 20:42, 7 January 2019 (UTC)

There are a lot of vowels here. Are these all phonemic contrasts? Danish phonology indicates 21 vowel phonemes and if those are all phonemes that you're listing, then that's a total of 18 plus the 4 high vowels and schwa. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:24, 7 January 2019 (UTC)

With my suggestion all the diacritics are redundant and can be left out without any risk of confusing two distinct phonemes, since different glyphs are used for different phonemes. That is exactly the point. In a broad transcription you leave out the diacritics, in a narrow transcription you keep them in. I do not think it is a good idea to use diacritics for some vowels and not for others. In that case they wouldn't be transcribed at the same level of precision.

I think think there are a few considerations that need to balance: Precision (symbols should be used how they are defined by IPA), comparability (usage should comparable with the usage of IPA in other languages), readability (e.g. less use of diacritics, goes somewhat against the need for precision), distinction (important distinctions, i.e. phonemic contrasts, should be clearly reflected by using distinct glyphs). These are the considerations I have tried to balance in my suggestion above.--Schwa dk (talk) 11:47, 8 January 2019 (UTC)

Can we make a table showing the comparison between what's there now and what's being proposed by the various parties? — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 20:48, 8 January 2019 (UTC)

Also, what's the difference between Danish unaspirated voiceless lenis that appears in the syllable onset: ([b̥, d̥, ɡ̊]]) and English /b d ɡ/ that appears in the syllable onset? If they're identical, I'm not sure if there's much point in using diacritics for those, either. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 20:05, 7 January 2019 (UTC)

@Aeusoes1: It's not a big deal to me, but I'd use the diacritic indicating voicelessness. In words like skinnede, the ⟨k⟩ is always fully voiceless and always unaspirated, much like English /k/ after /s/, as in sky. The problem is that in Danish this is analyzed as /ɡ/. I think that transcriptions such as [ˈsɡenð̩ðə] can be a bit confusing to native speakers of English and especially other languages (Romance, Slavic, etc.) Kbb2 (ex. Mr KEBAB) (talk) 20:42, 7 January 2019 (UTC)

If there's no phonetic difference, between g and sk, we could use [k] for both. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:24, 7 January 2019 (UTC)

@Aeusoes1: But that's the problem - in phonemic transcription and in broad phonetic transcription, ⟨k⟩ is used for the aspirated plosive. This is another thing that makes this system so awful (well, for some people at least). The contrast is one of aspiration (as in Icelandic), yet everyone insist on writing the unaspirated plosives as /b, d, ɡ/ in phonemic transcription, when in reality they should be transcribed /p, t, k/ and the aspirated versions /pʰ, tʰ, kʰ/ (affrication is a phonetic detail and that, if anything, should be omitted).

I have nothing against using [p, t, k] for the unaspirated plosives and [pʰ, tʰ, kʰ] or perhaps [pʰ, tˢ, kʰ] for the aspirated ones. However, Den Danske Ordbog (just to name one source) writes these with [b, d, ɡ] and [p, t, k] and this may be rather confusing. I don't know what to do, honestly. Kbb2 (ex. Mr KEBAB) (talk) 03:57, 8 January 2019 (UTC)

I like [p, t, k] for the lenis plosives and [pʰ, tʰ/tˢ, kʰ] for the fortis ones. I think it's more confusing, as someone unfamiliar with Danish, to use ⟨[b̥, d̥, ɡ̊]]⟩.

There is apparently a linguistic tradition of using fairly divergent symbols and inconsistencies in transcribing Danish so that, no matter what we choose at the end of this discussion, we may want to take a cue from Help:IPA/Irish and show how our system compares to other systems in use. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 04:58, 8 January 2019 (UTC)

@Aeusoes1: Personally I have nothing against that and I like your idea of comparing our system to other systems. We'll see what others think. Kbb2 (ex. Mr KEBAB) (talk) 05:30, 8 January 2019 (UTC)

@Aeusoes1: This should be a topic of it's own. However, I think using [b d g] with or without diacritics is misleading. The Danish plosives are all voiceless, but contrast in aspiration. In my opinion, using only [p t k] and [pʰ, tˢ, kʰ] would be the best way to go.--Schwa dk (talk) 11:18, 8 January 2019 (UTC)

Although I haven't studied Danish phonology enough to approve or disapprove each symbol, I agree with basically everything Aeusoes1 has said. The current conventions are indeed confusing and general IPA is preferable, diacritics should be used as sparingly as possible, and I also concur about plosives. Providing comparisons may also be a good idea, particularly to the conventions used by Den Danske Ordbog. Nardog (talk) 12:08, 8 January 2019 (UTC)

@Nardog, Aeusoes1, and Schwa dk: We can provide comparisons to the DDO, to Basbøll (2005), Grønnum (2005), Grønnum (1998), Haberland (2013) and the Dania system (it's also listed in Basbøll (2005)). I think that should be enough, although I've read only a fraction of the relevant literature. Schwa should know a lot more than me regarding different systems.

They aren't all phonemic contrasts, but we do need to transcribe 3 heights of the mid vowels. This is because (I'll use the current transcription system here) the near-close [e] and the close-mid [ɛ] belong to different phonemes (/e/ and /ɛ/) and so do the near-open and open-mid [a, æ] and the open central [ɑ] (phonemically /a/ and /ɑ/, respectively). Among other mid vowels, the close-mid [ø] and the mid [œ] belong to different phonemes, but [ø] is rendered as mid [œ] after /r/, whereas /œ/ is open-mid [ɶ] both before and after /r/. So we can't really get rid of the third symbol in each case.

Among the back vowels, it's very similar: the close-mid [o] and the mid [ɔ] belong to different phonemes (/o/ and /ɔ/), though the mid [ɔ] is rendered as open-mid [ɒ] before /r/ with which it fuses to produce a long open-mid monophthong [ɒː], just as [ɑ] fuses with the following /r/ to produce a long open central [ɑː].

Among the front vowels, there also are differences in quality when a vowel is in contact with the /r/: the close-mid [ɛ] is lowered to open-mid [æ] before /r/ (after /r/ it's near-open [a] for those who differentiate between the two qualities), whereas after /r/ the near-close [e] is rendered as close-mid [ɛ] (before /r/ there's no difference).

What we could do however, is to stop differentiating between the qualities of the near-open [a] and the open-mid [æ] (both tend to sound like the latter and differ only in length) and to transcribe them with the same symbol. We can do the same with the near-open [ʌ] and the open-mid [ɒ], which tend to sound like the latter anyway.

Take a look at User:Kbb2/IPA for Danish and tell me if you like the system as it is now. Basically, I got rid of all of the diacritics. What we now write with ⟨e, ø, o⟩ I changed to ⟨ɪ, ø, o⟩; what is now transcribed with ⟨ɛ, œ, ɔ⟩ I changed to ⟨e, œ, ɵ⟩ and what is written with ⟨æ, ɶ, ɒ⟩ I changed to ⟨ɛ, ɶ, ɔ⟩. I've gotten rid of [a] and [ʌ], which are now rendered the same as the short [ɛ] and [ɔ] ([æ] and [ɒ] in the traditional transcription). The open central vowels are now simple [a, aː], consistent with the consensus that we already have regarding them (I think? Correct me if I'm wrong).

In my system, the unstressed vowels [ɪ, ʊ] are rendered the same as the main short allophones of /e/ ([ɪ]) and /o/ ([o]) because that's how they can be pronounced (or not, they can be slightly different - the point is that we list so many vowels that we just have to make some compromises, AFAIK our readers can safely think of them as the same phones). Their phonemic analysis as /jə, və/ is made clear by retranscribing the unassimilated variants after the assimilated ones: [ˈkʰɛːɪ, ˈkʰɛːjə] (if the word has more than two syllables than we can just retranscribe the last syllable of the more distinct pronunciation ([-jə])). Obviously, [ˈfɪðˀ] can't be transcribed as *[ˈfjəðˀ] (which is a transcription which would make zero sense in Danish) and so that's enough in my view.

The near-close front vowel is written with ⟨ɪ⟩ as that's an alternative symbol to ⟨e̝⟩. Its close-mid counterpart is written with ⟨e⟩. The mid back vowel is written with ⟨ɵ⟩ as it's audibly centralized and AFAICS there's no appreciable difference between it and the Swedish /ɵ/. Compare the Danish pronunciation of ost ([ˈɵst] in my transcription system) with the Swedish pronunciation of pust [ˈpɵst]. Can you hear any big difference between them (save for the initial /p/ in Swedish)? Because to me they sound almost exactly the same. The difference between Danish close-mid, mid and open-mid back vowels is pretty much defined by rather strong centralization of the mid ones. I suspect that Danish vowel charts may be somewhat mistaken in that what I write [ɵ] may be actually a fully central vowel (rather than a near-back one, as it's often shown), exactly the same as Swedish /ɵ/ (save for the type of rounding).

This shows how different IPA notations of almost the same phones can be: in Danish it's normally written with ⟨ɔ⟩, in Swedish and Russian with ⟨ɵ⟩ and in Dutch with ⟨ʏ⟩ or ⟨œ⟩ (rarely with ⟨ɵ⟩) when all of them are almost the same freaking sound. In Danish it's slightly more back than central, in Swedish it's more central and has a different type of rounding and in Russian as well as Dutch it's slightly higher (in Russian it's also slightly diphthongized). I don't want to go off-topic, I'm just bringing attention to a rather obvious fact.

The open-mid front unrounded and the open-mid back rounded vowels IMO have to be transcribed with ⟨ɛ, ɔ⟩ as [ɛ, ɔ] is their actual pronunciation. The open-mid front rounded vowel can be still transcribed with ⟨ɶ⟩ as [ɶ] is a possible pronunciation for some speakers. This allows us to keep transcribing the mid front rounded vowel with ⟨œ⟩.

I don't think that we should pay attention to the way Danish vowel phonemes are transcribed in this case. Those symbols are based on an outdated pronunciation. What is transcribed in phonemic Danish IPA with /e/ should be written with ⟨ɪ⟩, /ɛ/ should be transcribed with ⟨e⟩ and /a/ should be actually /ɛ/, just as /ɑ/ should be /a/. I'm basing this, of course, on the pronunciation of the main long allophones of these vowels.

We can bypass this problem by not mentioning phonemes in any way in this guide. Rather, we can just compare how different phones are transcribed by different sources. Transcription used on Danish phonology is probably a separate issue and I think that it should be dealt with on Talk:Danish phonology. Transcriptions such as ⟨æ̝⟩ (or worse, ⟨æ⟩) for what is a simple open-mid front unrounded vowel certainly won't be very helpful to our readers. Kbb2 (ex. Mr KEBAB) (talk) 12:50, 10 January 2019 (UTC)

Forgot to ping @Maunus:. Kbb2 (ex. Mr KEBAB) (talk) 12:59, 10 January 2019 (UTC)

Hmm, on second thought, the Danish mid back vowel is more like a vowel that is in-between [ɵ] and [ɔ] than closer to [ɵ]. Maybe the contrast is enhanced by the fact (well, a supposition at least) that the Swedish speaker who pronounces pust on Forvo actually uses a genuine close-mid [ɵ] instead of the mid one that can be seen on most vowel charts of Central Standard Swedish. Still - this is a detail and I still think that the symbol ⟨ɵ⟩ is one of the most suitable ones for the vowel, which Basbøll describes (on p. 47) as an advanced and raised cardinal [ɔ]. This is confirmed by the vowel chart on page 48, which is probably exactly the same one that we use in Danish phonology.

Perhaps ⟨ɞ⟩ would be a better symbol, but who uses it (I mean in any other language, save for a few papers)? ⟨ɵ⟩ is more commonly used in transcriptions and correctly shows that the vowel is close to Swedish /ɵ/. Kbb2 (ex. Mr KEBAB) (talk) 18:29, 10 January 2019 (UTC)

I like many of the changes. However, I disagree with using ⟨ɪ, ɵ⟩. I appreciate the point of avoiding diacritics, but the use of these symbols to accomplish that goal is not the best solution. The vowels contrast in apperture, not centralization. Therefore I find ⟨ɪ, ɵ⟩ misleading. Also, ⟨i/ɪ⟩ is often associated with a tense/lax contrast, which is also not the case in Danish. Also, ⟨ɪ⟩ is used for the 'i-schwa', i.e. the result of merging schwa + /ɪ̯/, and these vowels should be distinctly represented.

I can not see any logical reason, why we should not use ⟨æ⟩ rather than ⟨ɛ⟩ for the traditional ⟨a⟩. ⟨æ⟩ is used for approximately the same vowel in e.g. English, so ⟨æ⟩ would be helpful for readers used to English IPA.

I can not see a more fitting solution than using standard symbols, ⟨i e ɛ æ a y ø œ ɶ u o ɔ ɒ⟩, then modify them with diacritics in narrow transcriptions, while omitting diacritics in broad transcriptions. This is how IPA is used in other languages.--ə.dk (talk) 13:58, 17 January 2019 (UTC)

@Kbb2: Is this still happening? Nardog (talk) 14:58, 17 May 2019 (UTC)

@Nardog and Schwa dk: I'm all for using ⟨pʰ, p, tˢ, t, kʰ, k⟩ for plosives and for transcribing [a] (current transcription) and [ʌ] (ditto) the same as what we now write with ⟨æ⟩ and ⟨ɒ⟩. But I don't think that we should change ⟨ɑ, ɑː⟩ to ⟨a, aː⟩. Learners of Danish are probably unlike to encounter ⟨æ⟩ and ⟨ɒ⟩ (current transcription) very often (well, not really in the case of the former as it forms various types of diphthongs with the semivowels, unlike [ɒ] which only appears in [ɒw]), but the same cannot be said about [a] (current transcription), which means that this change can cause them to confuse what we now write with ⟨a⟩ and ⟨ɑ⟩. It is also much less incorrect to write an open central unrounded vowel with ⟨ɑ⟩ than to write a close-mid front unrounded vowel with ⟨ɛ⟩, which is what we're doing.

I also don't think that we should use any diacritics. If there's no consensus to change symbols (e.g. to use ⟨ɪ⟩ instead of the current ⟨e⟩, or whatever) the IPA should stay as it is.

If there's no objection to this I think that we can implement the changes soon. Kbb2 (ex. Mr KEBAB) (talk) 18:10, 17 May 2019 (UTC)

Yeah, I didn't account for pre- and post-/r/ allophones in much of one of the previous posts. On second thought, I see little reason not to change ⟨ɑ, ɑː⟩ to ⟨a, aː⟩. The change will be problematic only in the case of the short vowel - then again, the change from ⟨p, b, t, d, k, ɡ⟩ to ⟨pʰ, p, tˢ, t, kʰ, k⟩ is also somewhat problematic, but only partially so. Kbb2 (ex. Mr KEBAB) (talk) 13:38, 25 May 2019 (UTC)

@Nardog, Schwa dk, Aeusoes1, and Maunus: Ok, the guide and all transcriptions linking to it have been updated. However, I'm not sure what to do with Danish orthography because there's both phonetic and phonemic transcription in it. Honestly, I think that we should just use proper phonetic IPA in all of the Wikipedia. This means changing the original ⟨e, ɛ, œ, æ, ɶ, a, ɑ⟩ and ⟨ɔ, ʌ, ɒ⟩ to ⟨e̝, e, ø̞, ɛ, œ, æ, a⟩ and ⟨o̞, ɒ, ɔ⟩ (and therefore, perhaps, returning to distinguishing between [æ, ɒ] and [ɛ, ɔ] (conventionally transcribed with ⟨a, ʌ, æ, ɒ⟩)). I prefer to use ⟨ø̞, o̞⟩ instead of ⟨œ̝, ɔ̝⟩ because the fact that ⟨ø, o⟩ are "simpler" symbols makes the diacritics somewhat more prominent. I see using ⟨e̝, ø̞, o̞⟩ as no different than using the dental diacritic on Help:IPA/Tamil, which is used out of necessity - the IPA doesn't provide separate symbols for dental and alveolar stops just as it doesn't provide separate symbols for mid front and back rounded vowels. If you want to be even remotely precise when transcribing Danish, you have to use diacritics on vowels, same as in the case of Kensiu.

Using ⟨ɛ⟩ for a close-mid vowel as well as ⟨æ⟩, ⟨ɶ⟩ and ⟨ɒ⟩ for open-mid vowels, ⟨ɑ⟩ for an open central vowel and especially ⟨ʌ⟩ for what is essentially a mid-centralized cardinal [ɒ] is at odds with the recommendation of the IPA to always use the simplest symbols available (yes, they also advise us not to use diacritics when they're not needed - in this case, they are), which would be ⟨e, ɛ, œ, ɔ, a, ɒ⟩ (occupied by vowels that are better written ⟨e̝, e, ø̞, o̞, æ, ɔ⟩).

We already diverge from the way Danish is transcribed by most if not all other sources. No source I'm aware of writes the unaspirated plosives with ⟨p, t, k⟩ and only older sources use ⟨a, aː⟩ for the open central vowels. And we very rarely transcribe Danish on WP using phonemic transcription, even in Danish phonology. I think that we can safely adjust it wherever we use it, so that the traditional symbols ⟨e, ɛ, œ, a, ɑ, ɔ⟩ are rendered as ⟨e̝, e, ø̞, ɛ, a, o̞⟩. Having to write /sø̞ns/ for søns (instead of /sœns/ - ⟨œ⟩ would be reserved for phonetic transcription of a pre-/r/ allophone of /ø̞/) is a small price to pay when there are so many benefits in switching to proper phonetic IPA. We can and probably should do this switch on the whole WP, including the phonology page and maybe also Wiktionary.

In turn, we could transcribe the fortis alveolar stop with a plain ⟨tʰ⟩ (affrication is automatic for millions of speakers from England and plain aspiration will still cause native speakers of Danish to correctly identify the sound as belonging to the /tˢ/ phoneme) and stød with the glottal stop sign ⟨ʔ⟩, making some aspects of the IPA simpler. AFAICS, pronouncing stød as such is mostly fine - except, perhaps, before non-syllabic sonorants such as /n/, which would then have to be realized as very short in order not to be confused with the sequence /ən/. We already use rather broad transcription of the labiodental approximant and the non-syllabic portions of diphthongs such as [aj] and [aw].

Pinging @Kwamikagami and LiliCharlie: who are responsible for the change of the representation of Colognian tonemes on Help:IPA/Colognian so that they're transcribed in proper IPA. The issue with Danish is similar: IMO, we need to use proper IPA to transcribe Danish on WP. Also pinging @Erutuon, IvanScrooge98, and Thathánka Íyotake:. Kbb2 (ex. Mr KEBAB) (talk) 13:55, 1 June 2019 (UTC)

Regarding the fortis /tʰ/ I would prefer we keep ⟨tˢ⟩ for the simple reason that it conveys the idea of affrication, a feature which does not occur with /pʰ/ or /kʰ/, however phonemic this guide may be. Also, it is primarily thought for English-speaking non-specialists, and since the English approximation given uses an example that does not undergo affrication in quite a number of accents, I would at least keep the symbol as it is, possibly with a side note. When it comes to the stød, I am neutral. イヴァンスクルージ九十八_（会話） 10:01, 2 June 2019 (UTC)

Great, I think it is much better than before. A few remarks: I do prefer ⟨tˢ⟩ or even ⟨ts⟩ to ⟨tʰ⟩, but I can see arguments for both. I don't like using ⟨ʔ⟩ for stød, since I am constantly fighting the common misconception that stød is a glottal stop. It is not. It is not a consonant, but a modification in phonation or a change in pitch in the syllable. It depends very much on dialect, but in any case it is some glottal gesture. Therefore I think the glottalization sign is more appropriate. I would also prefer [aɪ̯] and [aʊ̯] to [aj] and [aw] etc. for the diphtongs. That would be in tandem with the use in other languages. It would also make the transcription of non-syllabic vowels consistent with [ɐ̯]. We do have marginal contrasts between [j] and [ɪ̯] in connected speech, e.g. [haɪ̯ɛpə hajɛpə] 'hej Ebbe!, har Jeppe?' (hi, Ebbe! Did Jeppe?). Also it makes transcriptions with or without schwa assimilation more transparent, e.g. [mæːʊ̯ə] vs [mæːʊ] for 'mave' (stomach).--ə.dk (talk) 07:13, 3 June 2019 (UTC)

@Schwa dk: I've reintroduced [ɪ̯] and [ʊ̯] to the guide (and I'm updating the transcriptions so that they match the guide). Is Frejas pronounced [ˈfʁajæs] or [ˈfʁaɪ̯æs]? It looks more like a diphthong to me, but I'm not sure. Also, what about Kirke? Is it [ˈkiːˀjə] or [ˈkiːˀɪ̯ə]? Kbb2 (ex. Mr KEBAB) (talk) 08:02, 4 June 2019 (UTC)

@Kbb2: Great, thanks. Frejas should be [ˈfʁaɪ̯æs]. Kirke should be [ˈkʰiɐ̯kə] --ə.dk (talk) 08:33, 4 June 2019 (UTC)

Well, this is all a bit of a minefield to enter. After all, differing regional and individual pronunciations can account for many of the nuances that we can disagree upon. For example, [tˢ] versus [tʰ] is basically a dichotomy between Insular Danish and Jutlandic, and differing vowel pronunciations can also be regional or individual. To me it is perhaps more urgent to address the fact that the vowel in e.g. op is, following a tradition from Grønnum and even further back, consistently transcribed with [ʌ]. This vowel may be somewhat fronted but it is certainly not unrounded, except in some northern suburbs of Copenhagen. I cannot wrap my head around the fact that Grønnum has chosen this transcription. It may make sense for the variant of IPA traditionally used for Danish but such language-specific variants really should not be used. Maybe we should start here and then proceed to the other vowels. --Thathánka Íyotake (talk) 13:40, 3 June 2019 (UTC)

@Thathánka Íyotake: Thankfully, the vowels representations have already been reworked here to fit with IPA.

I believe the use of ⟨ʌ⟩ stems from the use of the Dania phonetic alphabet in Danish dialectology. Dania distinguishes more vowel qualities than IPA. So in the conversion to IPA, in order to have the same distinctions as Dania a lot of vowels ended up having inaccurate IPA symbols. While this may be convenient in a stricly Danish context, it is confusing when you compare with other languages, and it misses a lot of the points of using an international convention. That's just the historical background. I believe more and more Danish linguists are ready to part with this traditional use.--ə.dk (talk) 08:52, 4 June 2019 (UTC)

Can someone make the comparison table we were talking about? Nardog (talk) 14:45, 30 August 2019 (UTC)

What is the correct IPA for "dum" vs "dom"?

Hi, I'm a beginner learning IPA. In this WT page the terms bum and dum appear. This seems plain wrong compared to the sound in Open-mid_back_rounded_vowel is correct. See IPA of dum: https://en.wiktionary.org/wiki/dum#Pronunciation_2 and IPA of dom: https://en.wiktionary.org/wiki/dom#Pronunciation_2. I would really like these two added in the help page to clarify the distinction. WDYT?--So9q (talk) 17:27, 11 October 2019 (UTC)

@So9q: The first one ("dum") has the mid back rounded vowel and the second one ("dom") the open back rounded vowel. "IPA for Danish" as it is used in the literature is an unscientific abomination that should be reformed ASAP. Kbb2 (ex. Mr KEBAB) (talk) 17:35, 11 October 2019 (UTC)

Bad examples

Hi. I find solen, vinden, and ryggen to be bad examples, as they can also be pronounced with an [ə] before the n (which typically happens in more formal situations). Not sure which examples to use instead, though. Anyone got some ideas? —Biscuit-in-Chief :-) ^{(Talk – Contribs)} 17:30, 7 November 2019 (UTC)

References

Aspirated alveolar plosive

I don't know why we're transcribing it with ⟨tsʰ⟩, not ⟨tˢ⟩ or ⟨ts⟩. Both Basbøll and Grønnum transcribe it phonetically as [d̥͡s], so the reason it's described as "aspirated" in addition to affrication is clearly phonologically motivated—so that it's in line with /pʰ, kʰ/—not phonetically. ⟨ˢ⟩ has the advantage of looking similar to ⟨ʰ⟩, but given it's non-IPA and now that we're transcribing /tʰj/ with ⟨tɕ⟩, ⟨ts⟩ seems the most logical choice to me. Nardog (talk) 15:31, 12 March 2020 (UTC)

@Nardog: Because ⟨ʰ⟩ denotes devoicing of the following sonorant consonant or vowel, which does occur (see e.g. Basbøll). ⟨ts⟩ doesn't, and ⟨tsʰ⟩ is at the same level of narrowness as ⟨tɕ⟩, which is mostly transcribed with ⟨tˢj⟩ or ⟨tj⟩ in the literature (AFAIK). Kbb2 (ex. Mr KEBAB) (talk) 16:34, 12 March 2020 (UTC)

Oh right, I didn't think of [t.s], which I assume does occur intervocalically and contrast with [tsʰ]. Nardog (talk) 16:48, 12 March 2020 (UTC)

@Nardog: I don't think that contrast is very common, but yes, they do contrast (maybe in subminimal pairs?). Kbb2 (ex. Mr KEBAB) (talk) 16:54, 12 March 2020 (UTC)

@Kbb2: Where does Basbøll say it's aspirated? On p. 60 it's only described as affricated, but not aspirated. I know [ʁ] gets devoiced when preceded by it, but that to me is all the more reason to regard the aspiration as something phonological, not phonetic. Or does one see a hiatus between the frication and a vowel in [tsʰa] etc. in waveforms? Nardog (talk) 04:43, 13 March 2020 (UTC)

@Nardog: On pages 213 and 259. But on page 32 he says otherwise, I think. Grønnum (2005) also seems to say that varieties that affricate /tʰ/ do it instead of aspirating it. /sj/ becomes [ɕ] just like [tɕ] is used instead of [tsj]. This is also a feature of Dutch, which doesn't aspirate the fortis stops. So I think you're right, the aspiration diacritic is superfluous if not "wrong". The fricatives seem to devoice sonorants as well, e.g. in flaske [ˈfl̥æskə]. Kbb2 (ex. Mr KEBAB) (talk) 11:06, 13 March 2020 (UTC)

And while we're at it, isn't /ts/ a better phonemic transcription if the affricate is really unaspirated? After all, the phonemic close-mid and mid vowels are also kinda messy when it comes to transcription. If we're not gonna use ⟨tʰ⟩ anywhere in phonetic IPA, maybe we should get rid of it, like we got rid of ⟨r⟩. Kbb2 (ex. Mr KEBAB) (talk) 16:38, 13 March 2020 (UTC)

@Kbb2: I don't think so. Didn't we just establish it can contrast with [t.s]? Anyway, I was in the wrong to think it wasn't aspirated: The fricative noise is followed by a real aspiration... (Fischer-Jørgensen 1954:52). You can indeed hear it in tak, tal on this page (but not in tirsdag, so it might still depend on speech rate/carefulness/vowel height/etc).

So my only remaining concern over using ⟨tsʰ⟩ is its appropriateness when used before a consonant (which I believe can only be [ʁ] or [v]). Now that I think about it, Basbøll and Grønnum are probably using the tie bar in narrow transcription to distinguish it from [t.s]. And when linguists use ⟨tˢ⟩ instead of ⟨ts⟩, ⟨tsʰ⟩, or what have you, I think they're trying to kill two birds with one stone: it's aspirated, both phonetically and phonologically, forming a natural class with /pʰ, kʰ/; it's affricated before a vowel, but not before a consonant (or coalesced before /j/), further indicating that it's underlyingly one segment. So I'm circling back to favoring ⟨tˢ⟩, in both IPA-da and phonemic transcriptions. Nardog (talk) 01:58, 14 March 2020 (UTC)

It is affricated before /ʁ/ (listen to træt and traktor, which feature an alveolar affricate followed by a voiceless uvular fricative), and apparently also before /v/ ([4]). If /v/ is devoiced here then this follows the behavior of /v/ (or /ʋ/) in Swedish and Dutch (Polish too :P), which is also devoiced after /t/. So, the way we analyze it on Danish phonology, this is an affricate, in all positions. It's just that it has a dialectal realization as an aspirated stop.

Perhaps - but note that we don't differentiate between affricates and stop-fricative sequences in Help:IPA/Polish and a few other guides. (@Aeusoes1:, what do you think?)

If we switched back to ⟨tˢ⟩ then we'd use two rather similar diacritics: the affrication diacritic in ⟨tˢ⟩ and the aspiration diacritic in ⟨pʰ⟩ and ⟨kʰ⟩. ⟨tsʰ⟩ conveys the same information (save for aspiration, which isn't covered in ⟨tˢ⟩) and it's probably a better choice.

Danish doesn't feature phonemic affricates, that's true - the fortis counterpart of /t/ belongs to the aspirated series. Phonetically though, the contrast is between a plain stop and an affricate, and in that sense it's way more similar to the contrast between /d/ and /ts/ in Standard German. Danish /tsʰ/ is just like SG (or Polish, Russian, Italian) /ts/ in that affrication is mandatory (AFAICS). Kbb2 (ex. Mr KEBAB) (talk) 07:48, 14 March 2020 (UTC)

If ⟨tsʰ⟩ is phonetically representative even preconsonantally like you say, then I have no problem. As for phonemic notation, I'd drop ⟨s⟩ before I would ⟨ʰ⟩, again because of its relation to /pʰ, kʰ/. Nardog (talk) 13:13, 14 March 2020 (UTC)

Vowels with stød aren't long

How come ˈtˢiːˀ is transcribed with a long vowel? I assume it's the IPA for ti (10), and I've never heard anyone pronounce that with a long vowel. —Biscuit-in-Chief :-) ^{(Talk – Contribs)} 11:10, 12 January 2020 (UTC)

@Biscuit-in-Chief: Haberland (p. 318 on Google Books) says that vowels with stød are about 15% shorter than the ordinary long vowels. That's not much, IMO. Basbøll (p. 272) says that vowels with stød are about as long as the long vowels. Why should we drop the length mark in this context? Kbb2 (ex. Mr KEBAB) (talk) 13:05, 12 January 2020 (UTC)

No offence to Haberland or Basbøll, but I find that to be complete BS, at least in modern Standard Danish. Compare ti with, for example, Tine. They're definitely not the same length. —Biscuit-in-Chief :-) ^{(/tɔk/ – /ˈkɒntɹɪbs/)} 15:43, 12 January 2020 (UTC)

@Biscuit-in-Chief: The vowel in barn sounds pretty long to me. Either way, we need at least one source that says something else than Haberland and Basbøll. WP:OR is the reason. Kbb2 (ex. Mr KEBAB) (talk) 17:40, 12 January 2020 (UTC)

Well, some parts of this seem OR to me anyway. Like the examples. —Biscuit-in-Chief :-) ^{(/tɔk/ – /ˈkɒntɹɪbs/)} 17:44, 12 January 2020 (UTC)

@Biscuit-in-Chief: Either way, you're not the first native speaker to bring this up. If 85% of length of the long vowel is enough of a shortening that you perceive it as a short vowel (or at least "not a long vowel", so to say), then we could drop the length marks, following e.g. DDO. Even if those vowels are phonemically long, the final consonants in dåb, hat and tak are also phonemically aspirated and phonetically unaspirated (at least in normal speech). This wouldn't be the first case of the same allophone (a phonetically short vowel) belonging to two different phonemes (to simplify things a bit). Both types of transcription (⟨iːˀ⟩ and ⟨iˀ⟩) are found in the literature; Haberland writes vowels with stød like this: ⟨iˑˀ⟩, with a half-long sign (or its non-IPA counterpart, actually). That would be an overkill in my view. Kbb2 (ex. Mr KEBAB) (talk) 12:41, 14 January 2020 (UTC)

Wow ... Why are there so many different ways of doing stuff! Can’t everyone just follow exact IPA? :( —Biscuit-in-Chief :-) ^{(/tɔk/ – /ˈkɒntɹɪbs/)} 13:33, 14 January 2020 (UTC)

@Biscuit-in-Chief: I don't know. I support the removal of the length marks because of words in which a stød-bearing vowel receives a mere secondary stress, e.g. Rådhuspladsen [ˈʁɒðˌhuˀsˌpʰlæsn̩] (or [ˈʁɒðˌhuːˀsˌpʰlæsn̩] in the current transcription). A vowel that's too long would sound really awkward in that position, if what you're saying is correct. Native speakers of English tend not to be very good with vowel length (though that depends on the variety of English) and so it's probably better to drop the length mark in that context. Kbb2 (ex. Mr KEBAB) (talk) 19:48, 17 January 2020 (UTC)

@Biscuit-in-Chief: I'll be WP:BOLD and remove the length marks. The table underneath the list of symbols is clear about the fact that transcription without the length marks is widely used. You're at least the second native who's concerned about this (as far as I'm aware of - the first one was Maunus when I asked them to transcribe Hans Jørgen Uldall for me). Kbb2 (ex. Mr KEBAB) (talk) 14:08, 21 January 2020 (UTC)

I fully support that change. —Biscuit-in-Chief :-) ^{(/tɔk/ – /ˈkɒntɹɪbs/)} 14:52, 21 January 2020 (UTC)

@Kbb2: Why did you remove length marks in unstressed syllables? My understanding is that phonologically long vowels with stød become longer again when unstressed. If not I'd like to see evidence of it. I bet if vowel length is contrastive in Danish it is in unstressed syllables too. Nardog (talk) 00:03, 22 March 2020 (UTC)

@Nardog: If that's true then maybe removing the length marks wasn't the best idea. Maybe we should restore them, given the fact that stød-bearing vowels are phonologically long and stød also tends to shorten consonants (at least /n/, AFAIK), not just vowels. This seems like too fine a distinction for this guide. Kbb2 (ex. Mr KEBAB) (talk) 06:54, 22 March 2020 (UTC)

@Kbb2: The very example we're using to illustrate first and middle names being unstressed in Danish phonology#Stress contains a long unstressed vowel, as does in the source, Grønnum (1998: 104). Pronunciations of first names like Peter and Martin on Forvo clearly retain the vowel length.

What's more, I was surprised to find many examples of Carl/Karl were clearly pronounced with stød. It seems the case that, while vowel length and stød are frequently lost in spontaneous speech when unstressed, this is by no means obligatory (compare the examples of "distinct" and "normal" speech here, and search for "stress reduction" in Basbøll etc.).

I think we should restore not only vowel length but stød and stress in first/middle names etc. In IPA-xx transcriptions we normally try to capture as many lexically contrastive elements—which stress, stød, and vowel length clearly are—as possible, rather than the result of an utterance after syntax and intonation are applied. We do trasncribe some allophonic information that only arises in connected speech, like [siˈmom boˈliβaɾ] and [ˈtʃe ɣeˈβaɾa], but these are easily reconstructable to /N, ɡ/, while Nikolaj transcribed as [nekolaj] cannot be reconstructed to [ˈnekoˌlajˀ]. Nardog (talk) 17:28, 22 March 2020 (UTC)