Module talk:km-pron

Respelling of កម្ពុជា​ and អូស្ត្រាលី​

@Wyang, Stephen G. Brown: Hi guys, could you please respell phonetically កម្ពុជា (kampuciə) and អូស្ត្រាលី (ʼoustraalii)? --Anatoli T. ^{(обсудить}/^вклад) 22:09, 20 February 2018 (UTC)

The former should have "...oo...", the latter should have "...uʔ...". --Anatoli T. ^{(обсудить}/^вклад) 22:11, 20 February 2018 (UTC)

@Atitarev The former and the latter should be the latter and the former. :) For the former, /uq/ loses its glottal stop in unstressed syllables; e.g. in Cambodian System of Writing and Beginning Reader and Wikipedia (Khmer alphabet): "In a stressed syllable with no written final consonant, the vowel is followed by a glottal stop [ʔ]". The latter is Fixed. Wyang (talk) 22:46, 20 February 2018 (UTC)

Sorry, Anatoli, I'm not much help with this. I have not figured out how to do phonetic respellings. To me, អូ is [’oo], and ពុ is [pu] (generally speaking). I don't think we're writing them like that anymore, so I can't offer any help. —Stephen ^(Talk) 23:38, 20 February 2018 (UTC)

Hmm. The distinction between /ou/ (a-series ូ) and /oo/ (o-series ោ) is made in some textbooks and dictionaries, for example Huffman's Cambodian System of Writing and Beginning Reader (screenshot of its finals list), while some sources merge them into a single /oo/ or /ou/. I think we should accept that /oo/ for ូ in certain dictionaries (e.g. Headley 97, 77) is only a notation difference (and use no respelling on that syllable). Wyang (talk) 00:04, 21 February 2018 (UTC)

@Wyang: Haha, sorry for the former/latter mixup, it happens when you edit something written earlier. Thank you for your edits and for help. I now find that my pocket Cambodian dictionary and Sealang dictionary mismatch in some cases these methods but I need to learn how to map my sources and learn the script.

@Stephen G. Brown: That's OK and thank you, anyway. This is a new effort and we all just need to learn to write Khmer phonetically (respell) using just native methods, if possible, just like Thai, even though some Thai respelling conventions at Wiktionary are un-orthographical. Hopefully all scenarios and irregularities will be covered and we can link Khmer writing to the pronunciation and have a good transliteration. It's a lot of work but we are under no pressure. --Anatoli T. ^{(обсудить}/^вклад) 11:09, 21 February 2018 (UTC)

No worries! For now I will change the respelling back to អូ, and we can treat /oo/ in Sealang (Headley 97, 77) as a notational variant of /ou/, used for a-series ូ. Wyang (talk) 11:18, 21 February 2018 (UTC)

Alternative readings of ទៅ and respellings

@Wyang, Stephen G. Brown, Octahedron80: Apparently, the default/regular pronunciation of ទៅ (tɨw) is /tɨw/ in o-series? Which diacritic is used to convert it to a-series and be pronounced /taw/? Can I used ៊ (៊): ? {{km-IPA|ទ៊ៅ}} doesn't work. I also tried with រុស្ស៊ី (rusii). --Anatoli T. ^{(обсудить}/^вклад) 12:07, 21 February 2018 (UTC)

@Atitarev The actual pronunciation of ទៅ (tɨw) is /tɨw/, in the o-series. This is regular:

Orthographic and Phonemic ទៅ dau WT romanisation tɨw (standard) IPA(key) /tɨw/

since the letter ទ (tɔɔ) (Indic d) is voiced, and thus belong to the o-series consonants. The letter ត (tɑɑ) can be used to respell to produce /t/ with an a-series vowel, as in the variant pronunciation /toh/ of this word: តុះ (toh). រុស្ស៊ី (rusii) is Fixed. Wyang (talk) 12:26, 21 February 2018 (UTC)

That's right. ៊ (៊) converts a-series to o-series, and ៉ (៉) converts o-series to a-series. There is no need for a conversion diacritic on ទ (tɔɔ). You can use ៉ (៉) on these o-series consonants: ង (ngɔɔ), ញ (ñɔɔ), ម (mɔɔ), យ (yɔɔ), រ (rɔɔ), វ (vɔɔ). —Stephen ^(Talk) 12:36, 21 February 2018 (UTC)

Thank you both! --Anatoli T. ^{(обсудить}/^вклад) 12:38, 21 February 2018 (UTC)

ទូរស័ព្ទ

@Wyang, Stephen G. Brown: Hi. Still not winning. In ទូរស័ព្ទ (tuurĕəʼsap, “telephone”) I want to produce /tuurəsap/. My Tuttle's "Practical Cambodian Dictionary" romanises it as "dtoo-ra-sup". I'm pretty sure it's the sound /ə/ after /r/ and that matches the dictionary's romanisation description. My last attempt was {{km-IPA|ទូរ៉ិស័ព}} but it doesn't work. Trying to use ិ (ិ) to get /ə/ and ៉ (៉) to convert រ to a-series. How should this be respelled? --Anatoli T. ^{(обсудить}/^вклад) 11:58, 22 February 2018 (UTC)

Hmm, Sealang romanises ទូរទស្សន៍ (tuurĕəʼtŭəh, “television”) as /tuureaʔtʊəh/ but Tuttle as "dtoo-ra-dtoa-a". I'm just getting more confused at the moment. --Anatoli T. ^{(обсудить}/^вклад) 12:10, 22 February 2018 (UTC)

@Atitarev This should be respelt as ទូរៈសាប់ in my opinion. The word has two forms: ទូរស័ព្ទ (tuurĕəʼsap) (with ័ (័) U+17D0 KHMER SIGN SAMYOK SANNYA) and ទូរសព្ទ (tuurĕəʼsap) (without), pronounced the same. Chuon Nath (67) respells it as ទូរៈសាប់, and Headley (97) gives the pronunciation of /tuureaʔsap/, consistent with that respelling. The middle syllable is respelt using ៈ (ៈ) (U+17C8 KHMER SIGN YUUKALEAPINTU), similar to Thai -ะ, but it has two pronunciations instead: a-series /aʔ/ and o-series /ĕəʔ/ (same as Headley's /eaʔ/). In actual conversation the middle syllable is very reduced and indistinctly heard ([1], [2]), but I think the respelling in Chuon Nath and the pronunciation in Headley are correct in representing it as an o-series vowel. Wyang (talk) 12:49, 22 February 2018 (UTC)

@Wyang: Thanks, Frank. Are there online versions of these resources? --Anatoli T. ^{(обсудить}/^вклад) 13:07, 22 February 2018 (UTC)

@Atitarev Yes, Anatoli. Both Headley (97, 77) and Chuon Nath (67) are on Sealang. :) If you search for ទូរស័ព្ទ (tuurĕəʼsap) on Sealang you can see "PRONUNCIATION: ទូរៈសាប់" in [chuon67], which is the respelling. (Don't be afraid to read a wall of Khmer text :))) Beware though, that respellings in Chuon can be missing a few extra signs/symbols (to convert to a/o-series, etc.) sometimes, but here I think Chuon has it correct. Besides, this site also has a collection of Khmer dictionaries, including (often low-quality) audios and what appears to be a slightly different version of Chuon Nath (although manual says version is the same). In most cases, the Tovnah site gives the same results as Sealang. Wyang (talk) 13:16, 22 February 2018 (UTC)

Making it easier?

@Wyang, Stephen G. Brown, Octahedron80: I will continue to edit Khmer entries and add new ones but I sometimes find it very difficult to find the phonetic respellings and when I do, I don't have 100% certainty if the found method is correct or is considered good practice. With the Burmese it was easier with the possibility to "cheat" by sometimes using the IPA or Roman letter to generate the correct respelling and for Thai, which I can read a little easier, the respellings are often available in online dictionaries.

@Frank I don't know, which one would be easier for you - an easy to follow guide, more examples, an easier input method or just more Khmer entries using the new module. Would you be willing to go through some terms, which I found difficult (often loanwords), and add suggested respellings? If you think that will be quite time-consuming for you as well, I will continue to do it myself but a "peer review" on my edits would be appreciated, LOL. --Anatoli T. ^{(обсудить}/^вклад) 04:48, 26 February 2018 (UTC)

@Atitarev Hehe, no problem. I've made a list of the entries in Category:Khmer terms with IPA pronunciation and will gradually check those. Also created Category talk:Khmer terms with IPA pronunciation to track incoming ones, and will keep an eye on that. It is difficult, but I think it will get better over time. All the books use different systems to annotate pronunciation, and sometime the books just copy each other for certain words and leave inconsistencies within itself. I will add another column on Wiktionary:Khmer romanization to list the Sealang (Headley) equivalents of the consonants and vowels later. I would highly recommend Franklin Huffman's Cambodian System of Writing and Beginning Reader (1970) if you haven't read it already; I don't think I can write a better script guide than that... it's only a relatively short read (~150 pages), and explains the ins and outs of the script excellently. Please just let me know if you have any uncertain respellings (keep them in a list, add to testcases, ping on talks, etc.), I will see if I can help. Wyang (talk) 05:16, 26 February 2018 (UTC)

If I may, I would like to ask why respelling in the Khmer script is better than directly inputting transcription... —suzukaze (t・c) 05:37, 26 February 2018 (UTC)

@Suzukaze-c: For consistency. I personally prefer the respellings, that way it's insured that we use accurate and consistent transcriptions and native methods at the same time - if you copy-paste or enter by hand, even if it's accurate, there's no way of telling if it's consistent with other entries and no information about the characters - you see the new Thai and Khmer (and Burmese to some extent) modules demonstrate the difference between the written and pronounced, quite neat! It's similar to how kana, phonetic hangeul, Arabic vocalisations are used to render the correct pronunciations to both native speakers and learners. The native respellings are also quite educational, they are sometimes the common misspellings or even alternative forms, the respelling tells you, which part is read irregularly or if a term is pronounced as expected, even if the reading rules involved are complicated. --Anatoli T. ^{(обсудить}/^вклад) 05:46, 26 February 2018 (UTC)

As Anatoli said. We all love a bit of intellectual exercise don't we... 🔨🔨🔨 Wyang (talk) 10:16, 26 February 2018 (UTC)

Transliteration and pronunciation of លោក

Moved from Talk:លោក

Is the IPA for this accurate? I feel like I've heard this word pronounced with the diphthong /ou/ rather than /o:/. Or is this just allophony that I'm unaware of? 24.124.64.19 02:50, 19 May 2019 (UTC)

@24.124.64.19: The person who mostly designed the system has left Wiktionary. I can respell the word as "ល៉ូក", which will give the pronunciation /louk/. I wonder now if the final ក (k) should be /ʔ/ instead. @Octahedron80: What do you think? Headley97 gives /louk/, Headley77 /look/. [3] --Anatoli T. ^{(обсудить}/^вклад) 11:16, 19 May 2019 (UTC)

@Atitarev: I don't know much at all about Khmer script, but would it be correct to respell it as "ល៉ូក"? And about that final /k/, I think word-final /k/ is typically pronounced as [ʔ], but this is a regular allophonic rule -- I don't know what common practice is here with regards to narrow/broad transcription. Interesting that both /look/ and /louk/ are listed in that dictionary... I wonder if there is some kind of alternation, free variation, or dialectal variation going on there. Just to clarify, I brought up the possibility of correcting the IPA to /louk/ because I heard the [ou] diphthong in the audio from Google Translate and from a Khmer-speaking acquaintance. I am not proficient in Khmer myself -- would definitely like to know whether both [oo] and [ou] are native pronunciations for this word, though.Localdenizen (talk) 21:46, 24 May 2019 (UTC)

@Localdenizen, Octahedron80: I have moved the question here. --Anatoli T. ^{(обсудить}/^вклад) 06:53, 27 May 2019 (UTC)

Pronunciation of េ

@Octahedron80, Atitarev

In 2018, Wyang decided to follow Headley 97 in noting /ee/ the first series value of េ, and /ei/ its second series one. This was one weird move from Headley, against any other descriptions of the language. The most recent ones by Anthelme (dictionary), Filippi (pronouncing dictionary + a few papers), both using a narrow and more faithful IPA notation, do not notate a diphthong there either. Headley 97 also swapped around a few other sound values (like ូ) but for some reason, this one Wyang didn't follow.

Granted, there is considerable allophony in Khmer pronunciation, but the "implicit" convention by most scholars was to notate second series vowels with monohtongs (/ii/, /ee/, /ɛɛ/, /ɨɨ/, /əə/, /oo/, /uu/), and first series vowels with diphthongs (/əy/, /ei/, /aɛ/, /əɨ/, /aə/, /ao/, /ou/). It's a little bit like English, "bed" isn't pronounced [bed], but that's what stuck around.

So, since that move by Wyang, the IPA of entries containing a េ have become very confused, with many overrides to force the /ee/ rendering, using lots of series converters (ម៉) or artificial shifts between series (ត​ for ទ,​etc...)​. មេរៀន (mei riən) is one example, and ទេសភាព (teisaʼphiəp) another very recent one.

Would you be ok with me changing back to 1./ei/ 2./ee/, and clean-up the IPA overrides as I see them? Thx. Sitaron (talk) 19:37, 6 June 2021 (UTC)

I'm moving forward with this change then, as it still creates confusion among editors.

Sitaron (talk) 15:38, 17 July 2021 (UTC)