Talk:コッ

Ainu transliteration

User:I'm so meta even this acronym, I didn't realize you have kana-2 in your Babel. This basically traces back to my refusal of providing a manual translit. I'd hoped that I could quickly draft kana-Latin pairs and then ask someone to plug them in into a simple module, but there would be quite a few cases where auto translit simply wouldn't work, e.g., both *otke and *okke could be valid words and kana would spell them the same – o, small tsu and ke.

It appears that manual translit is the way to go for ain. Neitrāls vārds (talk) 15:30, 27 July 2015 (UTC)

@Neitrāls vārds: Sorry about taking out the alternative form; word-final っ, ッ (っ, ッ) doesn't occur in Japanese (AFAIK), so it didn't occur to me that this word might be spelt both コッ and コツ. I think Japanese transliteration uses undisplayed characters like ^ and .; perhaps Ainu transliteration could do the same. For example, オッケ could be transliterated okke, whilst オッ.ケ could be transliterated otke; what do you think? — I.S.M.E.T.A. 15:58, 27 July 2015 (UTC)

A period could be the solution (caret should be avoided though, I guess, as it appears that Ainu Times uses it to indicate suffixes: prefix=word^suffix, TDA uses spaces pref= word =suff to indicate pref. or suff.) I couldn't write anything that sophisticated though (my Lua = 0).

How do you think if I wrote the conditions could someone at some point turn them into Lua? I started something of the like at User:Neitrāls vārds/sandbox. It shouldn't be super complicated but there would be some "cascading," e.g., a + i = ay, but not if a y-row kana follows: ai + yu = aiyu (not ayyu,) but I think this would be as complicated as it would get.

As both Kana and Latin appear to be equally valid orthographies, technically it should go both ways, but Latin to Kana could be more complicated, e.g., TDA appears to keep original Latin spelling of proper nouns in Latin: [Mr.] Setsu (not [Mr.] Secu), but perhaps there's no need to rush ahead of time ("not to bite off more than one can chew" and all that), 99% of wikt's ain entries at present are in kata anyway. Neitrāls vārds (talk) 18:23, 27 July 2015 (UTC)

@Neitrāls vārds: You'd be better placed that me to decide on these undisplayed characters (FYI, ^ is, I think, used to generate initial capitals in rōmaji). My understanding of Lua is also nigh-nonexistent, so I'm afraid I can't help much with the technical aspects. Re Latin-script entries for Ainu, I recommend you write WT:AAIN, where you can prescribe that katakana-script entries be the lemmata for Ainu, with Latin-script ones being soft redirects.

Pinging @Atitarev, Eirikr, Haplology, Jamesjiao, Metaknowledge, TAKASUGI Shinji, Wyang as editors who may be able to help with this.

 — I.S.M.E.T.A. 23:51, 27 July 2015 (UTC)

Thanks for pinging them, this whole thing is nothing urgent, I just wanted to "advertise" it somewhere. Btw, I was mistaken re: caret (^), Ainu times uses it as some weird rendition of kata long sign (which I'm not a fan of, I think if there is ever need of showing long consonants in Ainu, just double letters would be used) and they just plainly connect both prefixes and suffixes with an equals sign without any spaces (which I like.) But all of this is not necessarily relevant, as it's more of punctuation stuff than actually translit'ing terms. Neitrāls vārds (talk) 00:01, 28 July 2015 (UTC)

Ainu is a different language from Japanese, and it just happens to use katakana because of its location. Actually you can’t distinguish ai and ay using katakana: it is rare to distinguish アィヌ and アイヌ, though the former is linguistically correct. アィヌ is aynu, not ainu, because it has two syllables in Ainu, not three syllables as in Japanese. — TAKASUGI Shinji (talk) 00:52, 28 July 2015 (UTC)

Thanks for passing comment, Shinji. I know Japanese and Ainu are unrelated languages, but I would assume there to be some similarity in the ways they transliterate katakana, which is why I pinged you and the others. — I.S.M.E.T.A. 16:00, 28 July 2015 (UTC)

• @ I.S.M.E.T.A.: One of the key challenges that Shinji points out is the issue that the linguistically "correct" spellings, which use the small kana to distinguish non-moraic sounds in diphthongs, are not used consistently. I recently ran into this in a thread on my talk page about the Ainu term オウペカ (owpeka). Strictly speaking, the Ainu term is three syllables and three morae, so the ow *should* technically be spelled オゥ in katakana to clearly indicate the single-mora reading ow, as opposed to the two-mora reading o + u. However, my research so far has uncovered no evidence for the オゥペカ spelling actually being used.

Similarly, the term aynu *should* technically be spelled as アィヌ to clearly indicate two morae, but in practice, the term is much more commonly (almost always?) spelled as アイヌ, which parses as three morae. However, since Ainu has no three-mora word a + i + nu, readers default to reading this as the expected two-mora word ay + nu.

Since the spellings are ambiguous, we cannot create any means of reliably generating automatic romanization in all cases. Much like for historical kana spellings, we will just have to manually supply the romanizations in those cases where the template cannot generate the correct values. ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:18, 15 December 2015 (UTC)

Rather than generating a latin spelling, why not generate a katakana spelling from a latin spelling? I think it is regular. The Japanese Wiktionary uses latin spellings as main entries for Ainu. — TAKASUGI Shinji (talk) 00:27, 16 December 2015 (UTC)

@Eirikr: As you say, Ainu orthography is… unsettled, indeed. — I.S.M.E.T.A. 01:58, 22 December 2015 (UTC)