Module talk:cau-cir-translit

Capital letters

Capital letters also need to be included. --Anatoli ^{(обсудить}/^вклад) 08:40, 14 July 2013 (UTC)

I'm waiting for User:Adamsa123's approval of what I've done so far before I spend more efforts on capitals. --Vahag (talk) 08:44, 14 July 2013 (UTC)

Latin Extended-E

@Theknightwho: Why use those codepoints? They don't even display on my end. I'd expect transliteration to use more common codepoints, otherwise it kind of defeats its own purpose.. კვარია (talk) 15:52, 14 November 2022 (UTC)

Which codepoints are you referring to? I've actually decided to undertake a more general review of transliterations for the Caucasian languages, as the TITUS system looks awful with so many diacritics. Theknightwho (talk) 16:02, 14 November 2022 (UTC)

This little bugger: ꭓ. Everything else is fine so far. კვარია (talk) 16:10, 14 November 2022 (UTC)

Thanks. I'm going to spend the next few days trying to put together something a bit more reader-friendly, because the current way of doing things just feels like a bad replica of IPA. Theknightwho (talk) 19:33, 14 November 2022 (UTC)

@კვარია I've done a thorough revision, and have come up with a novel romanization based on an extensive review of the phonology of all the literary Caucasian languages (NE and NW). My intention has been to find a way to romanize both families in a way that is phonologically consistent within each family, is intuitive to an English speaker, and is consistent with existing romanization systems wherever possible (primarily Russian, Turkic and Arabic - the first two of which obviously influenced the existing systems anyway). I've drawn a little bit from the IPA, but I wanted to avoid having a jumble of nonstandard characters, so have only used those which seemed intuitive to me (a native English speaker). Where I've had to do anything novel, I've used superscript letters to give secondary characteristics to a letter (e.g. lˢ = /ɬ/). I've tried to only do this where there were no other options, and they're only supposed to be indicative.

Please do let me know if you have any thoughts. Theknightwho (talk) 10:27, 20 November 2022 (UTC)

@Theknightwho: well done, this looks really reasonable. While you're at it, you should disallow the usage of English I/i instead of palochka (ӏ). For example this case shouldn't just silently ignore the "false palochka", thus creating an incorrect transliteration. Though to be frank I don't know if disallowing this stuff is supposed to be a transliteration module's work. კვარია (talk) 10:45, 20 November 2022 (UTC)

Beep-bop. кӏэп (kʼɛp) you just broke. :D კვარია (talk) 10:55, 20 November 2022 (UTC)

@კვარია Ugh. Someone's optimised all the translit modules by using native gsub instead of the Unicode-compatible one, so it broke. Anyway, I've fixed it, so it'll work properly if someone uses capital "I" or lowercase "l". It's possible to set the language module to do character substitutes in links, too, so that the link will still work if someone enters a false palochka. We usually use it to ignore stress marks in links, but I'm pretty sure it'll work for this as well. I'll also find a way to shove them into a maintenance category as well, because they're a nightmare to spot. Theknightwho (talk) 11:09, 20 November 2022 (UTC)

@Theknightwho: what was wrong with the dot to mark the ejectives? That is how it's done in Kartvelian. The apostrophe looks like the spiritus asper used in Armenian to mark the aspiration, which is very confusing for me: WT:HY TR. Also, it is difficult to see the apostrophe beside other superscript symbols. Vahag (talk) 16:50, 23 November 2022 (UTC)

@Vahagn Petrosyan To me, it felt a lot less intuitive in comparison to the IPA system. I also preferred to use dots to represent pharyngeal consonants (similarly to Arabic romanization). Pharyngealisation is very common in Archi and (to a lesser extent) Ubykh, which meant that finding a consistent scheme proved tricky; the one I came up with accounts for the ~200 consonant clusters I identified across about 30 languages, but it's had to involve compromises like this. I do agree that apostrophes next to superscripts can be difficult to read (and I'm sure there's a better way to do it), but I also found that dots combined with other diacritics were either difficult to see, didn't display well or (at best) were hard to parse, which is why I've tried to only used dots where other diacritics aren't involved. That isn't possible if we use them for ejectives.

I will create an explanatory page, so that any changes we make will keep things consistent within the schema. Theknightwho (talk) 17:37, 23 November 2022 (UTC)

Addendum: of course, we don't need to use the same schema for NE and NW Caucasian languages, but I felt that would be undesirable. Theknightwho (talk) 17:41, 23 November 2022 (UTC)