User talk:Mahagaja/Archive 14

Can Irish pronunciation be automated?

Taking a quick glance through w:Irish orthography, it looked like it could be done. Is this actually true? —JohnC5 21:45, 1 December 2016 (UTC)

w:Irish orthography is pretty incomplete, though. I started writing the full story at w:User:Angr/Irish bh and mh, w:User:Angr/Irish dh and gh, w:User:Angr/Irish l and n, and w:User:Angr/Irish orthography, but I was so overwhelmed by the enormity of the task that I never finished. It probably could be automated to some extent, though some sort of respelling parameter would be necessary for cases where the pronunciation doesn't correspond to the spelling, and there are a lot of these. For example, at sibh, an automated pronunciation template would correctly predict the pronunciations of Munster, Mayo, and Ulster, but the Galway pronunciation would probably require a separate parameter along the lines of {{ga-IPA|Galway=sib}}. —Aɴɢʀ (talk) 21:52, 1 December 2016 (UTC)

So similar to mod:fr-pron in that respect? Just a thought. It'd be an interesting project for me if only to learn the rules of Irish IPA. Thanks! —JohnC5

If you want to try your hand, go right ahead. I have no idea how to write modules, but I'll help any way I can. —Aɴɢʀ (talk) 13:56, 2 December 2016 (UTC)

fáinne

You described the process by which this word got an initial f- as a "back-formation". Would you still consider that correct? If so, I'd kindly ask you to expand the definition because right now it doesn't cover that. Kolmiel (talk) 21:56, 16 December 2016 (UTC)

PS: Wait, I suppose you mean fáinne is a backformation from *fháinne, not the actual áinne. Hmm, okay. I'll try to phrase it a bit more clearly. Nervermind. Kolmiel (talk) 21:58, 16 December 2016 (UTC)

Westrobothnian

Hey, I see you removed the Westrobothnian entry at least one place. It seems to be completely gone now. While it was weird and looked weird and the spelling was likely the work of one guy and not actually used by anyone I'm kinda curious about it. Do we know who added it? Where the words were gathered from?

I don't know. User:Korn know more about it than I do. —Aɴɢʀ (talk) 22:29, 16 December 2016 (UTC)

The work of User:Knyȝt and some IP which might be identical with him. Search 'Westrobothnian' in the recent Request for Verification/Deletion pages, another IP gives hints there as to where the words might come from. Knyʒt is active on sv.Wikt, so he might be a native speaker. Korn [kʰũːɘ̃n] (talk) 22:41, 16 December 2016 (UTC)

Neither of you know anything about this language. That's it. It's a fairly easy language to understand it if you know a lot of Norse, but clearly none of you guys do. Due to ethnocide/systematic oppression by the Swedish state you won't find anything serious written about it either way.

The words come from several different dictionaries (oldest one from 1804) but who cares right? it's gone from this site so there's clearly no interest.

Korn, do you honestly think I use a Danish IP to edit Danish on English Wiktionary and then switch to this account to edit Westrobothnian on Swedish Wiktionary? And then combine both on English Wiktionary for the occasion of messing with you? Are you insane? Do you ever take five minutes to think or to look something up?

You got a specific language deleted for no reason. it happened to be the language of my ancestors. thanks dude. how come a bunch of other languages are allowed but not mine? oh the answer is literally no reason? Nice one. Such quality this site. — Knyȝt (talk) 00:28, 17 December 2016 (UTC)

@Knyȝt: Neither Korn nor I have any objection to the presence of Westrobothnian words on Wiktionary as long as they're written in an attested orthography. If there are Westrobothnian words written in the Swedish Dialect Dictionary or linguistics articles or anything else published, you can include them, as long as they're spelled the same as they were in the publication. The words that were here got deleted not because of some disdain for or lack of interest in Westrobothnian, but solely because they were in an orthography that you invented yourself and that has never been used by anyone else. —Aɴɢʀ (talk) 07:49, 17 December 2016 (UTC)

Do you have anything say about all the other unwritten languages that are accepted here in made up orthographies? This point keeps being ignored. — Knyȝt (talk) 10:31, 17 December 2016 (UTC)

• @Knyȝt: 1. No, I never take even a second to look up where a random IP is situated and I feel no obligation to eiher. 2. You were specifically asked in person to verify the entries you made and you were given six weeks time for that, which is even two weeks more than Wiktionary nominally requires. If they come from printed dictionaries, the correct process would have been to provide those citations when you were asked. Then your entries would still be around, in their attested spellings. All languages are 'allowed'. But there is Wiktionary:Criteria_for_inclusion and we apply it to make sure we are not a source of misinformation. You don't really have anyone to blame but yourself. 3. The reason I tagged Westrobothnian and no other language is not that I specifically care about Westrobothnian, it's that while browsing for something, I came across Westrobothnian and noticed that, while situated in Sweden, it used a lot of non-Swedish letters. So how come Westrobothnian got tagged and the rest not? I happen to be Germanic and not Bantu, so I browsed something Germanic and not Bantu. It's called coïncidence. Korn [kʰũːɘ̃n] (talk) 10:50, 17 December 2016 (UTC)

Quote: "No, I never take even a second to look up where a random IP is situated and I feel no obligation to eiher." Then you should in NO WAY assume that User:Knyȝt and I are one person, let alone propose it, considering you admitting that you pulled it out of your ass. I shall repeat it again for you, I am not User:Knyȝt, I don't even speak Swedish (I understand it, though). I'm Danish, and I happen to be interested in Knyȝt's work here. Again, I will not discuss with you, as you've managed to destroy authentic, valuable information because of your stubbornness. Congrats. --80.63.3.167 13:21, 17 December 2016 (UTC)

You're not the IP I was referring to. And if the information was authentic, why didn't you verify it when a request for it was made? You're showing a lot of aggression against the usage of a run-of-the-mill Wiktionary progress which you could have stopped a minute after it was started by simply supplying all the sources you keep speaking of. Korn [kʰũːɘ̃n] (talk) 14:23, 17 December 2016 (UTC)

Fucking tell us what IP you're referring to. — Knyȝt (talk) 09:05, 19 December 2016 (UTC)

@Knyȝt: I'm unaware of any other examples of languages accepted here in made-up orthographies. You've mentioned !Kung in previous posts, but there are published reference works that mention !Kung words, as well as portions of the Bible that have been published in !Kung, and as far as I know we use the same spellings as those published works. If that's not the case, then those words are eligible for RFV too. —Aɴɢʀ (talk) 10:55, 17 December 2016 (UTC)

You have no obligation to look like less of an idiot, it's just that it's so easy to do, you wonder what is going on in your head to make you rather randomly speculate on someone's identity than take the literal minute it takes to look it up. Anyway, is it really proper to delete languages based on coincidence rather than a uniformal application of the rules? I'll just go ahead and add another 20 entries in Norn or Jamtish while you talk about how unofficial Westrobothnian is. — Knyȝt (talk) 11:46, 17 December 2016 (UTC)

I don't think I've mentioned that language. I don't know what it is. But if you take a second to check the Wikipedia article it says there is a myriad of different ways of spelling it, and the same thing is true for Westrobothnian, Jamtish, Gutnish, Norn etc. etc. There are of course many examples. Now here comes the big problem with these published works on Westrobothnian. You have the same exact words and sounds published in different books in different spellings — sometimes using letters that have no coding — and no literature that clearly state what language it really is, just to what villages the exact variety is particular to. So would we then have perhaps eight different spellings of a word that is pronounced in only one or two ways in the same category of language or in separate ones? Or all of them listed under Swedish even though it by no measure is Swedish and will not be recognised by Swedes as Swedish? Or the version preferred by the Swedish state: pretend the language doesn't exist because ethnocide is okay?

Also, notice that Westrobothnian was added here along with a group of other languages (in the same situation) by a group of people not including me. — Knyȝt (talk) 11:46, 17 December 2016 (UTC)

If Westrobothnian words are attested in many different spellings, we can include them all. We're not paper, we're not going to run out of space. We just don't want to include unattested spellings. —Aɴɢʀ (talk) 13:02, 17 December 2016 (UTC)

• So would we then have perhaps eight different spellings of a word that is pronounced in only one or two ways in the same category of language [...]? Yes. Let me state this very clearly: The words were not deleted because they were Westrobothnian, you're fighting against windmills here. Nobody ever claimed Westrobothnian is unofficial, we still have all the Westrobothnian codes and categories, they are just empty now because nobody made a valid entry. The deleted words could just as well have been English, Swedish or Hausa and would have gotten the same treatment under these conditions. The conditions being that none of them was attested. If half of the Westrobothnian entries would have been attested, half of them would still be there. If all of them would have been attested, all of them would still be there. And just for the sake of the argument: If Westrobothnian were a spoken-only language, and there was a consensus to include it in a Wiktionary-only spelling, it would belong in an appendix, not the mainspace where regular entries are. Korn [kʰũːɘ̃n] (talk) 13:29, 17 December 2016 (UTC)

Right, so since the dictionaries are all in Swedish, we can't have any definitons on English Wiktionary, since they must be written in English, correct? I wouldn't imagine that I would be allowed to freely translate complex definitions if I can't freely translate phonetic notation (which is how the words are written in the dictionaries). — Knyȝt (talk) 09:05, 19 December 2016 (UTC)

No, where did you get that idea? We write our own definitions here, in English. I have to do that for Lower Sorbian, for example, because all published information on Lower Sorbian is in German. But the entry name has to be in a spelling that's attested, at the very least because of our general rule at WT:CFI: "A term should be included if it's likely that someone would run across it and want to know what it means." If you're using a spelling that has never appeared in a description of the language before, then it isn't likely that someone will run across it. —Aɴɢʀ (talk) 18:08, 19 December 2016 (UTC)

Well, almost nobody would have run into any spelling at all. The majority of regular people would recognise the words purely by what they sound like when they read them, not the shape of the typed word or anything like that. But there are phonetic dictionaries with a variety of different spellings if we are desparate for the closest thing to some sort of spelling. Of course, they use a variety of letters which may or may not be found in computer coding. What do I do when the letters used don't exist on computers? Moreover, nobody is going to have come across a spelling like vɑɳ unless they read the one particular dictionary that spelt it that way, and most people aren't going to know exactly what ɑ or ɳ means. Modern dictionaries all have instructions on how to read them because they aren't readable unless you already know the phonetic system or know exactly what you're reading should sound like. — Knyȝt (talk) 09:33, 23 December 2016 (UTC)

@Knyȝt: How practical would it be to spell Westrobothnian terms using the International Phonetic Alphabet? — I.S.M.E.T.A. 22:20, 23 December 2016 (UTC)

I don't think it would be more practical than anything else. A lot of words vary in exact realisation, so you would probably get even more (unecessary) articles than if you went with the dictionary spellings. Also, the exact phonetics are not very clear from reading the dictionaries, but rather something you sort of figure out with experience, so it's not obvious what a system based entirely on IPA should look like. — Knyȝt (talk) 23:55, 24 December 2016 (UTC)

@Knyȝt: Hmm, well I'm not sure what to think. Because of the way WT:CFI works, we'd have entries for all attested spellings, but because of the purported irregularity in Westrobothnian spelling, it would be very reasonable to have a scheme of standardised spelling for lemmatisation purposes. The question is what to base that scheme upon; there doesn't seem to exist any academic consensus in that regard. — I.S.M.E.T.A. 20:11, 25 December 2016 (UTC)

And then we do the same for Low German, Swiss German, Frisian, Scots, Silesian, all minor Chinese varieties and every other uncodified language? Korn [kʰũːɘ̃n] (talk) 18:45, 29 December 2016 (UTC)

@Korn: Yes. Why? — I.S.M.E.T.A. 18:49, 29 December 2016 (UTC)

Just making sure this isn't one of those instances where a systematic problem could be fixed with a general policy but is then is ignored outside of a single case. My proposal for this situation with regards to Low German was to use Middle Low German (longstanding standard form amongst researchers) as the basis for our lemmatisation. So abstracted: We could make it out default practice to lemmatise uncodified languages according to the most recent available standardisation of one of their ancestors. This seems to be similar to what Knyʒt was doing in the first place too. (For unwritten languages, IPA seems to be a decent choice.) Maybe this is the time to leave Angr in peace and move to the Beer Parlour? Korn [kʰũːɘ̃n] (talk) 19:08, 29 December 2016 (UTC)

@Korn: I find that policy very agreeable. Please ping me if/when the Beer-parlour discussion gets going. — I.S.M.E.T.A. 20:17, 29 December 2016 (UTC)

ɨ~ᵻ

Should I change all Dongxiang ᵻ into ɨ? Fields uses ᵻ to denote "retroflex apical vowel" whatever that may be (I suspect a Sinologist would know), is this the same thing as ɨ? Crom daba (talk) 22:04, 21 December 2016 (UTC)

For Dongxiang, I really don't know. ᵻ isn't an official IPA letter. The Mandarin retroflex apical vowel (the i in shi, zhi, etc.) is usually transcribed /ʐ̩/ or /ɻ̩/, but I don't know if that's right for Dongxiang. —Aɴɢʀ (talk) 22:11, 21 December 2016 (UTC)

That's probably it, Dongxiang phonology is supposed to be the same as the phonology of local Mandarin dialects (plus uvulars, a trill/tap and a glottal fricative), but I guess I'll use ɨ just in case. Crom daba (talk) 23:12, 21 December 2016 (UTC)

Also if we proscribe ᵻ, we should probably remove it from the IPA toolbar. Crom daba (talk) 23:12, 21 December 2016 (UTC)

I've removed both ᵻ and ᵿ, since they're both not official IPA characters. —Aɴɢʀ (talk) 11:01, 22 December 2016 (UTC)

Aedacan

Why would it be close and not a mid vowel? What you've written would be Aidacan, not Aedacan. ÞunoresWrǣþþe (talk) 21:41, 25 December 2016 (UTC)

Never mind, apparently it's a "long diphtongue". ÞunoresWrǣþþe (talk) 21:42, 25 December 2016 (UTC)

Old Irish

I've seen non-word-initial "c" appear as both /k/ and /g/. What are the rules for that? Why is cerc /kerk/ but clocan is /kloɡaːn/? Is it a question of being lenited intervocalically? Cheers, ThunoresWraethe ÞunoresWrǣþþe (talk) 22:00, 25 December 2016 (UTC)

You really can't tell from the spelling. If it goes back to Proto-Celtic /nk/, then it's /ɡ/ in Old Irish. If it goes back to Proto-Celtic /kk/, or to /k/ after /l/ or /r/, then it's /k/. The easiest way to tell, though, is to look at the modern descendants: Modern Irish has cearc with a /k/ and clogán with a /ɡ/. —Aɴɢʀ (talk) 07:33, 26 December 2016 (UTC)

Thank you. ÞunoresWrǣþþe (talk) 16:44, 26 December 2016 (UTC)

The preceding l or r is still preserved in Old Irish isn't it? So rc and lc would always have /k/, right? —CodeCat 16:20, 28 December 2016 (UTC)

Yes; I tried to make that clear by saying "/k/ after /l/ or /r/", rather than "/lk/ and /rk/", which could have implied the liquid was lost. It occurred to me later that voiced /ɡ/ also appears where Proto-Celtic had /ɡ/ after /l/ and /r/, e.g. *wergā > ferc, ferg, fergg /fʲerɡ/. —Aɴɢʀ (talk) 16:27, 28 December 2016 (UTC)

Question about the etymology of fiach

Feel free to tell me I don't know what I'm doing and should stop before I embarrass myself.

While looking for the origin of the sga pr.n. Fiachnae, I started running into contradictions and confusions. For a start, Fiachnae looks like a variant of fiachán = fiach +‎ -án. But I couldn't figure out what the Ogham would be... *viacagna?

But looking in CISP, I find veqoanai. Would I be right in my suspicion that this is the Ogham form of Fiachna, from a PIE root of *wekʷ- (“to speak”)? So the origin of Fiachna would be something like wekʷ-(...) > veqoanai > *fʲacanae > *fiacʰanae > fiachnae? And that would mean that we might fill in *wekʷos (as "speaker"?) as the etymon of fiach.

Or am I haring off on a wild goose chase, without knowing what a goose should even look like? If I'm right, would that imply a connection with gwyach? --Catsidhe ^{(verba, facta)} 23:46, 27 December 2016 (UTC)

fiach has two syllables, so there was once a consonant between the two vowels. That rules out the Primitive Irish forms you gave. Primitive Irish e (actually ē) results in é, later ía in Old Irish, but the difference between ía and ia is crucial here. —CodeCat 00:25, 28 December 2016 (UTC)

I'm at work right now, away from my reference works, but CodeCat is right that fïach as a 2-syllable word can't come from anything starting *wēk-. A preform like *wisak- is most likely, off the top of my head. If Ogham veqoanai stands for *wēkʷanai or something, I suppose it could be the etymon of Fiachnae, provided the name is really Fíachnae with the /iːa̯/ diphthong that comes from ē. Then it would be unrelated to fïach, of course. —Aɴɢʀ (talk) 09:15, 28 December 2016 (UTC)

OK, so I was likely on a whole series of goose chases, all in different directions and all at the same time. Maybe to reduce it to parts:

1. Fiachnae was the name of a few early kings, and may be the same as, or have been confused with, Fiachra, Fiacha, and Fiach. (When you compare the various Annals against each other, you get people referred to as any or all of these. Fiachna mac Baetain is Fiachna mac Baetain, Fiacha mac Baettain, and Fiachra Caoch, mac Baodain all in the Annals of the Four Masters, and Fiachaidh mac Báetáin in the Annals of Tigernach.) So I would have hoped for at least one attestation in Ogham. I could only find two attestations in the Annals of Fíachrach (being the typical genitive of Fiach(r)a, as in Cenél Fiachach), which is rare enough for me to happily call those mistakes.
2. So: was vegoanai a possible Ogham attestation of Fiachna?
3. Was (or: could it have been) vegoanai derived from *wekʷ-?
4. Is Fiach(n/r/∅)a related to fïach, or to fíach, to each other, or to none of the above?
5. Is fiach cognate to gwyach? (Which might be cool, but I'm just asking for completeness.)
6. Seriously, tell me if I'm just being one of those guys with more enthusiasm than clue.

--Catsidhe ^{(verba, facta)} 10:49, 28 December 2016 (UTC)

GPC says fïach and gwyach are cognate and come from *wesākos, from *wes- (“feasting”) (which is not a root I'm familiar with; I wonder if they mean *h₂wes- (“spend the night”), but getting from there to 'grebe'/'raven' is quite a stretch). What exactly is the Ogham form? First you said veqoanai with Q, now you're saying vegoanai with G. If the Q is right and the E is long, then yes, the Ogham can be the ancestor of Fiachna, but it can't be related to the bird word. I suppose it could theoretically come from *wekʷ-, but you'd have to explain the long vowel as well as the rest of the word. —Aɴɢʀ (talk) 11:19, 28 December 2016 (UTC)

Note that Veqoanai is a genitive. Only one or two ogham stones give us the nominative form of names, which is usually -as for men (o-stem, yo-stem etc...) and -a for women. Fiachnae would be descended from the nominative form. ÞunoresWrǣþþe (talk) 11:35, 28 December 2016 (UTC)

McManus identifies Veqoanai as a possible ancestor of Fiachnae, and identifies it as a yo-stem. That would make the nominative something like *Veqoanaias unless it's irregular. ÞunoresWrǣþþe (talk) 11:40, 28 December 2016 (UTC)

Tyop: veqoanai with a q. (The browser helpfully puts a squiggly red line under words it doesn't know, meaning I can't see at a glance whether it's 'q' or 'g'.)

So in summary (repeating what I think you said, so you can see if I'm misunderstanding): It is possible for veqoanai to be the Ogham form of Fiachnae (with q > "c", and then intervocal lenition to "ch", and "ē" > "ía"), but in that case the apparent similarity with fïach = "raven" is coincidental. Fïach and gwyach are reportedly cognate, derived from *wesākos, possibly from *h₂wes-, with a basic meaning of "feaster"? --Catsidhe ^{(verba, facta)} 11:43, 28 December 2016 (UTC)

Yes, you've understood correctly. The last bit (from "possibly" to the question mark) is the most dubious in my opinion; I'm also having trouble imagining what bird the Proto-Celtic word could have referred to since ravens and grebes (which look a whole lot like ducks but aren't closely related to them) are really extremely different. —Aɴɢʀ (talk) 11:55, 28 December 2016 (UTC)

Seems like far too many sound changes to apply to a single word attested soon after the previous form, to be honest. Especially if you subscribe to Koch's idea of Old Irish and Primitive Irish co-existing during the Early Middle Ages. ÞunoresWrǣþþe (talk) 11:56, 28 December 2016 (UTC)

You forgot ai > ae, apocope and elision (veq-o- and -a-nai). ÞunoresWrǣþþe (talk) 11:59, 28 December 2016 (UTC)

Old Irish and Primitive Irish are convenient names for two different stages of the same language; saying they coexisted makes as little sense as saying that Old English and Middle English coexisted. The one gradually morphed over time into the other. And *wēkʷanayos > Fíachnae is perfectly regular and expected; the only unexpected thing is the "o" in the Ogham. —Aɴɢʀ (talk) 12:03, 28 December 2016 (UTC)

It's what Koch says. They co-existed.

And it makes perfect sense. Scots and English coexist, so did Yola and Middle English, so do two different Greek languages.

Old French was still spoken by some at the same time as Middle French, and Latin and Vulgar Latin co-existed. ÞunoresWrǣþþe (talk) 12:25, 28 December 2016 (UTC)

Scots and English are contemporaneous regional variants, as were Yola and Middle English. The only way the claim could make sense is if what he means is that the sound changes that characterize Old Irish didn't happen simultaneously everywhere, so that some dialects of Old Irish were more conservative (and thus more Primitive Irish-like) than others. Which is perfectly plausible, but considering how scantly attested both pgl and sga are I have to wonder on what basis he comes to that conclusion. —Aɴɢʀ (talk) 12:49, 28 December 2016 (UTC)

He puts forward the theory that when Latin was brought to Ireland, Primitive Irish mutated rapidly into Old Irish, but scholars, written texts (Ogham stones) and the higher classes used the more formal Primitive Irish, while lower class people speaking informally would use Old Irish, which drowned out Primitive Irish around 800. It wouldn't be unlike the two languages of Greek spoken today, a formal language in newspapers or used by "noblemen", and Greek that's spoken casually (Katharevousa and Demotic Greek, respectively). It's even more of a good example since the two have the same amount as differences as Primitive and Old Irish have between each other. ÞunoresWrǣþþe (talk) 13:12, 28 December 2016 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘

He uploaded an open-source version of his theory, if you're curious. I'm not really doing it much justice by paraphrasing it using examples he didn't use himself. ÞunoresWrǣþþe (talk) 13:19, 28 December 2016 (UTC)

The about page

I saw how nice the Proto-Celtic about page was, and I tried to add some relevant info to the basically empty Primitive Irish page. For now it's very much still in progress, but I wondered if you agreed with the sound changes I've listed so far? https://en.wiktionary.org/wiki/Wiktionary:About_Primitive_Irish#Developments ÞunoresWrǣþþe (talk) 13:49, 28 December 2016 (UTC)

I'll take a look later when I have more time. —Aɴɢʀ (talk) 16:29, 28 December 2016 (UTC)

No problem. There were some things I was unsure of, like if Primitive Irish /qʷ/ lenited in Old Irish as /x/ or /xʷ/. ÞunoresWrǣþþe (talk) 16:59, 28 December 2016 (UTC)

/x/. The kʷ > k rule applies to the lenited forms as well. —Aɴɢʀ (talk) 17:01, 28 December 2016 (UTC)

I see, thanks for your assistance. Do we know when palatalisation came about? I've noticed that Proto-Celtic k is always Primitive Irish c when next to an e or i, which sounds a lot like Old Irish palatalisation, where consonants next to e or i are palatalised. Is that correct? Also, do we know when that sort of palatalisation started happening? For example, would Primitive Irish <inigena> be /iniɣena/ or /inʲiɣʲenʲa/? ÞunoresWrǣþþe (talk) 17:24, 28 December 2016 (UTC)

Palatalization presumably started during Prim.Ir., but it wasn't phonemic until syncope and apocope removed the vowels that had triggered it. So Prim.Ir. inigena was phonemically /iniɣena/ but phonetically [inʲiɣʲena] (no palatalization before a); after the loss of the second i and the a, Old Irish ingen was phonemically /inʲɣʲen/. —Aɴɢʀ (talk) 18:55, 28 December 2016 (UTC)