Module talk:IPA/data/symbols

Multiple XSAMPA symbols?

@Giorgi Eufshi, Dixtosa: Is there any way to associate multiple XSAMPA symbols with the same IPA symbol? There are comments saying that P _R _F _~ are equivalent to v\ _/ _\ ~ respectively; in addition ' is equivalent to _j, and I'd personally like to add , as an equivalent to = since using = within templates like {{x2i}} is awkward. —Aɴɢʀ (talk) 16:04, 30 November 2016 (UTC)

I just moved some of the data from Module:IPA/data. I have no understanding of either data module. --Dixtosa (talk) 17:50, 30 November 2016 (UTC)

@ZxxZxxZ, CodeCat, what do y'all think? Is it possible? —Aɴɢʀ (talk) 18:45, 30 November 2016 (UTC)

@Angr, Dixtosa: It would be easy to achieve this if a separate table of X-SAMPA and IPA were created. I could copy my table from w:Module:Sandbox/Erutuon/X-SAMPA. — Eru·tuon 00:16, 21 February 2017 (UTC)

Done. And now N=, N_0 are correctly transformed to ŋ̍ ŋ̊ rather than ŋ̩ ŋ̥. — Eru·tuon 01:37, 21 February 2017 (UTC)

But I need to make the "with descender" form of the diacritic also be used when there is already a diacritic underneath the letter. — Eru·tuon 01:30, 27 February 2017 (UTC)

Pharyngealisation

@CodeCat, could you please add ˤ as an allowable character? I'm not sure what the right way of adding it is. —Μετάknowledge^{discuss/deeds} 00:06, 21 February 2017 (UTC)

I noticed that character wasn't included in the list. It's actually not a real IPA character: MODIFIER LETTER SMALL REVERSED GLOTTAL STOP. The IPA character is MODIFIER LETTER REVERSED GLOTTAL STOP. The two look identical in some fonts: ˤ vs. ˁ. — Eru·tuon 00:13, 21 February 2017 (UTC)

That's great to know! Then we won't add it, and I'll tell DTLHS to add it to the characters to be globally replaced. —Μετάknowledge^{discuss/deeds} 00:16, 21 February 2017 (UTC)

• @Erutuon, ř is a character that should be added, however, since it seems that the whole Unicode decomposition thing is getting in the way again. —Μετάknowledge^{discuss/deeds} 00:27, 21 February 2017 (UTC)

  Or perhaps the IPA could be decomposed and we could search for combining diacritics (making the list of valid characters much shorter). Hmm. In any case, I've added that character. — Eru·tuon 01:11, 21 February 2017 (UTC)

  @Erutuon: That would indeed be good. We have the same issue with ŋ̍, ɺ̡, ɺ̢, and other combinations as well, it would seem. —Μετάknowledge^{discuss/deeds} 07:32, 21 February 2017 (UTC)

  @Metaknowledge: Ahh, the problem with those is the diacritics. Apparently they're not on the list. The list is frustrating because it's not very organized. I'll do some work on it. — Eru·tuon 07:46, 21 February 2017 (UTC)

  Much appreciated. Category:IPA pronunciations with invalid IPA characters had about 15,000 members when I suggested working on it to DTLHS and now it has 11,000-odd members. I've been cleaning up ones with actual problems by hand, but a large part seems to be IPA that's fine, just not recognised by the module. —Μετάknowledge^{discuss/deeds} 07:51, 21 February 2017 (UTC)

@Metaknowledge: Hm, it seems that the palatalized hook below and rhotic hook below (Unicode hex numbers 0x321 and 0x322) are obsolete or nonstandard, but are not in the list of obsolete or nonstandard symbols. I think that instead of ɺ̡, ɺ̢, ɺʲ,  should be used, but I could be wrong because I am not familiar with any languages whose transcription uses those symbols. — Eru·tuon 01:35, 27 February 2017 (UTC)

I think its only use here is for Pashto, and it follows the use in the Wikipedia article Pashto phonology. I see  as a box, which certainly does not make me want to recommend it, but I don't know what characters are best. —Μετάknowledge^{discuss/deeds} 01:40, 27 February 2017 (UTC)

I see that character as a box unless it is enclosed in templates that tag it with class="IPA". I guess it must be so newly added that my browser (Chrome) doesn't know which fonts to assign it. Gentium, the font that I have assigned to the IPA class in my common.css, does have the letter, though. Oh, now that I look at w:Retroflex lateral flap, the symbol  isn't standard IPA? Huh. I guess I will add the rhotic hook, but not the other diacritic. It's not listed in w:International Phonetic Alphabet, but it is probably semi-valid at least. — Eru·tuon 02:17, 27 February 2017 (UTC)

 is a Private Use character. Don't add it, its appearance depends on the whims of the font maker. —suzukaze (t・c) 02:23, 27 February 2017 (UTC)

@Metaknowledge, Erutuon According to w:Pharyngealization#IPA symbols (as mentioned at Wiktionary:Grease_pit/2019/June#IPA_template), we have the pharyngealization symbols backwards. The official IPA symbol is ˤ (U+02E4 modifier letter small reversed glottal stop), which the module is currently deprecating. The symbol the module is currently preferring is ˁ (U+02C1 modifier letter reversed glottal stop), which WP says "the IPA Handbook does not mention … at all". —Mahāgaja · talk 13:42, 19 June 2019 (UTC)

Spacing "raised" diacritic

@Erutuon, CodeCat: Can we add ˔ (the spacing equivalent of the "raised" diacritic ̝ ) to the list of valid characters? It needs to be used, for example, at tlakate, since the symbol it modifies already has another diacritic beneath it. Thanks! —Aɴɢʀ (talk) 18:03, 21 February 2017 (UTC)

@Angr: Done! I also added the spacing version of the "lowered" diacritic, which was missing. — Eru·tuon 04:25, 23 February 2017 (UTC)

Superscript parentheses

The majority of the entries with invalid characters are coming from {{ru-IPA}} generating pronunciations with ⁾ and ⁽. Is this absolutely incorrect, or can these symbols be added? DTLHS (talk) 02:48, 22 February 2017 (UTC)

In what context are those parentheses used? Around superscript j? — Eru·tuon 03:01, 22 February 2017 (UTC)

@Benwing2 Can you provide context? DTLHS (talk) 03:59, 23 February 2017 (UTC)

Yes, these are around superscript j, indicating optional palatalization of a consonant, typically when directly preceding another palatalized consonant. Benwing2 (talk) 14:17, 23 February 2017 (UTC)

The raised parens are also used by {{ny-IPA}} for the transcription of the sound formerly spelled with ŵ. —Aɴɢʀ (talk) 15:18, 23 February 2017 (UTC)

I think the superscript parentheses should be allowed; ⁽ʲ⁾ and ⁽ᵝ⁾ look neater than (ʲ) (ᵝ). I've added a rule in Module:IPA that allows these parentheses, but only around superscripts. (Hopefully I haven't broken anything in creating the list of superscripts, though.) — Eru·tuon 22:12, 23 February 2017 (UTC)

Still seeing the invalid character message (абсорбировался for example) DTLHS (talk) 22:38, 23 February 2017 (UTC)

Oops, a regex syntax error. Now it works. — Eru·tuon 22:42, 23 February 2017 (UTC)

Preocclusion symbols

@Erutuon, Octahedron80, DTLHS: Can someone please add ᵇ ᵈ ᶢ to the list of valid IPA characters? They're needed for preoccluded consonants in languages like Manx. Or should we be using b͡m d͡n ɡ͡ŋ instead? —Aɴɢʀ (talk) 15:43, 10 October 2017 (UTC)

@Angr: I'm not sure which is correct according to the standard use of the IPA. The Wikipedia article doesn't mention pre-occlusion. But I've added those symbols anyway. It can be up to you which to use. — Eru·tuon 22:38, 10 October 2017 (UTC)

I thought these are [bⁿ] [dⁿ] [gⁿ]. --WikiTiki89 22:43, 10 October 2017 (UTC)

Those would be nasal release. kwami (talk) 21:02, 11 May 2022 (UTC)

extIPA

@Benwing2, Erutuon, Octahedron80, Rua, Surjection: On Kwamikagami's talk page (User talk:Kwamikagami#𝼆) he and I discussed the possibility of adding some extIPA characters to the list of approved IPA characters here. That way we could use ⟨𝼆⟩, for example, instead of ⟨ʎ̥˔⟩ for the voiceless palatal lateral fricative. What do y'all think? I would support adding only the extIPA characters that represent sounds found in non-disordered speech, namely (per Wikipedia) ⟨ʪ ʫ ꞎ 𝼅 𝼆 𝼄 ¡⟩. —Mahāgaja · talk 20:41, 11 May 2022 (UTC)

For a bit of background, JIPA will accept extIPA characters for its 'illustrations of the IPA' articles. (I specifically asked about <𝼆 𝼄>.) Ladefoged used <ꞎ> in his description of Toda.

There's also [𝼈], found in languages spoken by hundreds of millions of people, and the rarer [ᶑ] and [𝼊] (the latter most commonly rendered with non-IPA <‼>). The IPA supported superscript variants of all three what they called "implied" IPA letters as part of last year's expansion of Unicode support of the IPA.

Some of the extIPA diacritics are also used in non-disordered speech, e.g. unaspirated, alveolar and the parentheses, occasionally others. kwami (talk) 20:45, 11 May 2022 (UTC)

That's true; we already use ◌͈ "stronger articulation" for Old Irish and, I believe, Korean. ⟨ᶑ⟩ is already whitelisted in the module. —Mahāgaja · talk 06:47, 12 May 2022 (UTC)

Chi

@Mahagaja We should be using the Greek letter, not the Latin one. Appendix 2 of the Handbook of the International Phonetic Association specifies the code point for the voiceless uvular fricative to be 03C7, as does the List of symbols and diacritics with descriptions & identifiers published in 2020. Nardog (talk) 22:58, 17 June 2023 (UTC)

Update: I thought Mahagaja had added code that suggested the Latin one for a Greek input, but now I see it was the other way around. So my comment only concerns his summary. Nardog (talk) 11:44, 20 June 2023 (UTC)

OK. Since I knew what I had done all along, I assumed that the comment above was solely in reply to my edit summary and not my actual edit, which is why I didn't bother answering. —Mahāgaja · talk 11:55, 20 June 2023 (UTC)

Invalid characters

@Theknightwho, are your recent edits the reason why there are 13,000+ entries in Category:IPA pronunciations with invalid IPA characters? Most of them are alphabetized in the category under "Ɩ" even though they don't even use that character and never did. —Mahāgaja · talk 07:50, 4 December 2023 (UTC)

@Mahagaja There was a bug a couple of days ago which accidentally put everything into that category - I assume the ones still in there are simply filtering through, since pages don't update immediately. Theknightwho (talk) 15:00, 4 December 2023 (UTC)