Help talk:IPA/Danish

Symbols for the open central vowels

Can't we just use ⟨a aː⟩ for the open central vowels? Our system of transcription is very similar to that found in [1], which uses ⟨a aː⟩ for them. ⟨ɑ ɑː⟩ explicitly suggest a back quality, close to [ɒ, ɒː] which they are not. ⟨a(ː)⟩ is the most usual symbol for the open central vowels because it's a Roman letter and we use it in Help:IPA/German, Help:IPA/Spanish, Help:IPA/Italian etc. (and also Australian English phonology and New Zealand English phonology) - and that's ignoring the fact that the Danish vowels seem to be even closer to cardinal [a] (being near-front, rather than central) than those. The difference between the Danish pronunciation of the given name Lars and the Swedish one is very noticeable (see [2]). The Swedish one sounds like Lårsj in Danish, whereas the Danish pronunciation is very similar to the German one (as long as the /r/ is vocalized in the latter: [ˈlaːs], rather than [ˈlaɐ̯s] which is not possible in either German or Danish). It's somewhat like Läs in younger Stockholm pronunciation, though obviously more open/central.

(The same question applies to Danish phonology, obviously). Sol505000 (talk) 08:11, 16 February 2023 (UTC)

I really prefer the current way. ⟨a⟩ is commonly used for what we transcribe with ⟨æ⟩, so keeping ⟨ɑ⟩ means one less thing to worry about when comparing transcriptions in sources. If the IPA had a symbol whose canonical value was open central, I'd support using it in a heartbeat, but given the ambiguity that plagues ⟨a⟩, I don't find your suggestion to be an improvement. We deviate from established ("normalized") symbols only where they are patently misrepresentative and confusing. This is not one of them. Nardog (talk) 12:17, 16 February 2023 (UTC)

In that case we can at least specify the dialect before the example word, and that's what I've WP:BOLDly done. The vowel in art, if it's too back (especially in dialects such as Geordie and Pittsburgh English), can be easily mistaken for Danish /ɔ/ ([ʌ]). Sol505000 (talk) 13:11, 16 February 2023 (UTC)

I agree with you, and disagree with Nardog. In IPA, ⟨a⟩ is generally used for any open vowel if there is only one, and ⟨ɑ⟩ is added for languages that contrast two open vowels. In Danish, ⟨a⟩ is used for [æ] for traditional reasons only in order to avoid confusing it with the letter æ that corresponds to [ɛ]. I have examined the use of ⟨æ a ɑ⟩ for 146 Illustrations of IPA, and only Danish deviate from this use of ⟨a⟩ and ⟨ɑ⟩. Research from 2016 even indicate that the Danish ⟨ɑ⟩ is much closer to [a]. ə.dk (talk) 07:57, 10 May 2023 (UTC)

American English is one in which ⟨ɑ⟩ commonly represents a central value. If such research exists, then please cite it in Danish phonology; we generally do this from the ground up (though the key is currently based on SDU/DDO/Basbøll, so if we're going to incorporate developments into the key in any one aspect, we should do so in all and overhaul it). Nardog (talk) 09:00, 11 May 2023 (UTC)

I totally agree with that, and of course I also agree that it should be based on published research. I obviously recommend the phonetic notation used on https://en.udtaleordbog.dk/ipa.php. This is what I use, teach and preach. The research behind it is forthcoming, so it may be a bit early for an overhaul here. But if you are interested, feel free to contact me for a pre-print (check my bio for contact options). ə.dk (talk) 18:29, 11 May 2023 (UTC)

References

Syllabic consonant equivalents in English

I have not been able to find any word other than nightingale actually pronounced by me or in online samples as /ŋ̩/, not /əŋ~ɪŋ/. I previously changed the example for /m̩/ to "Topham", which was misguided, as it is an uncommon proper noun, but rhythm, as can be heard in that page's sample pronunciation, is not guaranteed to contain a true /m̩/, and album or the like would be better suited. If there is somehow an objection to assimilated homorganic obstruent-nasal sequences here, sudden should also be switched out. Célestine-Edelweiß (talk) 22:43, 26 April 2023 (UTC)

You seem unaware of the difference between phones and phonemes. As discussed in English phonology#Consonants, what are phonologically /əl, ən, əm/ are all susceptible to becoming [l̩, m̩, n̩] given the right environment. (⟨l̩, m̩, n̩⟩ in phonemic transcriptions are either synonymous with /əl, ən, əm/ or written by someone who doesn't know what they're talking about. And Wiktionary is not a reliable source.) To my knowledge post-fricative is the only environment where prominent dictionaries unanimously permit/indicate the possibility of [m̩] (rhythm, prism, etc.). Nightingale is not a good fit because it may be realized with unassimilated [n]. Washington is given as the exemplar of [ŋ̍] by the documentation for the TIMIT corpus, and it's about as good as it can be because there's no (near-)obligatory [ŋ̍] in English and there's hardly a difference between [ʃɪŋ] and [ʃŋ̍]. Nardog (talk) 04:25, 27 April 2023 (UTC)