User talk:KirillW

Ingrian dialect grouping

Hey, I was thinking of maybe grouping the numerous subdialects by the sound changes they undergo. I can see the following sound changes:

1. -ee- > -ii-; -oo- > -uu-; -öö- > -yy- (e.g. Rysymäki, Saarve, Säätinä, Tamminkonttu)
2. -ee- > -ie-; -oo- > -uo-; -öö- > -yö- (e.g. Krivulka, Narusi, Ropsu, Vanakylä, Vääräoja)
3. -i > -∅; -ä > -∅ (e.g. Kotko, Metsäkylä, Viistina)
4. Combination of 1 and 3 (e.g. Reppola)

We could either group these like English does (something like (soo-suu merger)) or we could divide the dialects in groups (something like (Southern Ala-Laukaa)). Do you have a preference? For the second option we'd need to collect more info though.

The primary benefits of both options are that 1) we'll reduce the amount of names in a pronunciation section, 2) we'll be able to predict pronunciations without Nirvi's and Kuznetsova's directly attested data. Thadh (talk) 13:26, 8 October 2021 (UTC)

Hi @Thadh, sorry for the late response. Sound like an interesting idea! There are isoglosses of this kind and I believe I've seen this mentioned in some article. I'll look up the article and discuss this on Saturday with the study group. KirillW (talk) 19:46, 12 October 2021 (UTC)

@Thadh, the short summary from Mehmet, using -oo- as an example:

• For Ala-Laukaa the changes are not really consistent. It is possible to say that the norm could be either -oo- > -^uoo- or -oo- > -uo-, but it can vary even in speech of one person.
• For Soikkola the split is North (almost no changes) vs South (system #1 from your list, i.e. -oo- > -uu-), excluding Viistinä, where it is mixed, but more like in the South.
• Ylä-Laukaa had developed system #1.
• Hevaha had both no changes and system #1.

Some words don't follow this pattern, specifically one-syllable words (soo tends not to become suu even in system #1), recent loanwords and others that have historical reason for this. That said, the process is rather synchronous.

Much more details you can get from Kuznetsova's thesis, in sections 3.3-3.6. In 3.4 you can see that it is still rather difficult to either name the system (there are actually 11+5 of them listed) or precisely specify the geographic distribution. Let alone the exceptions (see 3.5).

I see one way for simplification: omitting specific village names altogether, assuming this is not that important for the dictionary. At the same time the Wikipedia Ingrian phonology article can be updated with the relevant information. This is still on my table, but I hadn't really expected I'd get to it in the summer. I still plan to do it.

One problem with omitting the village names can be that not all villages in Nirvi are actually Ingrian. I remember we were finding some words there, but Mehmet were telling, that they seem to be Finnish words, which is common for some of the villages with significant Finnish population. I can check what these villages are, if needed.

Could you please clarify point (3)? Do you mean reduction of terminal vowels in general for Soikkola dialect or something else?

KirillW (talk) 20:34, 17 October 2021 (UTC)

Point 3 and 4 can be seen at, for example, keeli and meeli: There are five reflexes:

1. /ˈkeːli/ (underlying form)
2. /ˈkiːli/ (ee-ii merger)
3. /ˈkie̯li/ (diphthongation)
4. /ˈkeːl/ (vowel reduction)
5. /ˈkiːl/ (ee-ii merger; vowel reduction)

So, 3 refers to /ˈkeːl/ (vowel reduction, no ee-ii merger), while 4 refers to /ˈkiːl/ (vowel reduction and ee-ii merger). Do I understand correctly that the mergers/reductions are too irregular to be bound to geographical locations and that we're better off just calling the vowel changes by name (like English does)? Thadh (talk) 20:48, 17 October 2021 (UTC)

I have to check on the reduction in Soikkola dialect. But indeed looks like binding to geographical locations would be both imprecise and impractical. I'm not sure about the term "merger" in this case. The actual merge tends to be avoided (see soo/suu example). The term used is "mid vowel rise" -- is it good enough for our purpose? KirillW (talk) 21:07, 17 October 2021 (UTC)

It's a little too technical for my taste, but it's okay if we don't think of anything better... Thinking of huuli (“lip”) and hooli (“worry”) though, it seems there is a merger, but maybe we could think of better names indeed. Would you like to write the appropriate sections at w:Ingrian phonology first? Thadh (talk) 21:28, 17 October 2021 (UTC)

@Thadh, might be too technical, but at least it's quite short and succinct (compared to Russian "повышение гласных среднего подъёма"). Anyway, I don't have any strong preference as long as it doesn't confuse more than clarify. :) Speaking about mergers, I came up with another one-syllable example from text books: in pre-1935 books "loo" was usually spelled as "luu".

I couldn't find any suggestions on geographical distribution of reduction in Soikkola dialect.

Interestingly, for your example word, kieli, on pronunciation in Repola Nirvi reads: "Samassa kylässä (Repola) sana ääntämistapa on erilaista: kḙḙli, ki̬i̬li and keeli". And, you know, there could be a debate if other Ala-Laukaa reflexes should be considered. So probably it's not worth to go down this route...

One thing you might also find useful. Recently I've found out that Fyodor Rozhanskiy is creating a Soikkola dictionary with sound examples: https://ingrian.org/Ingrian-dictionary/. I had some discussions with him, mostly suggesting that we could use sound examples for Wiktionary, which he declined at least until the dictionary is published. He also stated that he'd better focus on recording the current state of Soikkola dialect and is not interested in a dictionary of an artificial Ingrian language. The spelling and declension tables a pretty phonetical (even shows 1,5x long consonants), but doesn't indicate, for instance, mid vowel rise in the word spelling itself, but shows it in sound example transcriptions.

I'll consider carefully what I can actually write in the Wikipedia article: something to the point that I can finish quickly. It took me more than 1,5h to write this comment, so I come back to this later :)

KirillW (talk) 18:53, 18 October 2021 (UTC)

Thanks for the link, that's a very nice dictionary! There's no rush with the article by the way, take your time. Thadh (talk) 19:25, 18 October 2021 (UTC)

reflexive pronoun

Hi, I've done my best, but couldn't find any mention of the paradigms of the reflexive pronouns. Nirvi gives a nice list of the genitive personal forms, and I could find some forms here and there (3SG PTV itsiätä, 2SG ALL itsellees...), but not much more. Would you happen to know anything about it? Thadh (talk) 18:24, 15 February 2022 (UTC)

Hi! I have to check this with Mehmet, because I know this is different between dialects. I'll do it on Saturday. KirillW (talk) 18:41, 15 February 2022 (UTC)

Thanks, I appreciate it! Thadh (talk) 19:46, 15 February 2022 (UTC)

Hi Thadh, here's what I've found out. Indeed the situation is different between dialects. In Ala-Laukaa possessive suffixes are long gone, so there are no possessive forms. For Soikkola is seems these were still in use at the Junus' time. Unfortunately I don't have a specific source to cite, so at this moment I can only retell what I've learned from Mehmet.

The basic Soikkola paradigm: N itse, G itsen, P itsiä, I itsee; Ala-Laukaa: the same, but P itsettä, I itsee/itsesse. Possessive forms we discussed only for 3SG: N itse, G itseen, P itsiää, I itsehee, TRANLS itseksee, others have elongation of the last syllable, depending of whether complete reduction of the ending has happened or not. The difference in illative and translative forms is because the historical forms were preserved in possessive forms. That said, in the two books we were reading recently, I've found a lot of variation. There are seem to be different versions for illative (itsehe, itseessee) and for other cases (itsees vs itsessää, itseest vs itsestää). There is even a case in "Loonnontiito 3 kl." on page 36 where there are 2 very similar sentences, one with "itsen" and another with "itseen". I've also found "itsiätä" form which is apparently an alternative form for "itsiää". I've checked other books in the search for other personal forms and found only "itseen" for 1SG. Where have you found "itsellees"?

Examples for all persons in Nirvi are probably date back to even earlier time (and his translation of "itsiätä" as "itseäsi" seems wrong), so this seems to be a difficult issue with no good source to rely on. Even Rozhanskij doesn't have the whole declension paradigm in his dictionary. If only it was possible to agree on the 3SG possessive forms, others should be deducible from those.

So I'd keep this as you've put it (probably adding some alternatives), until we find a reliable source, if any exists.

KirillW (talk) 22:10, 19 February 2022 (UTC)

Thank you, this is great!!

itsellees is from Nirvi, in the sentence toizelle hautā kaivaᴅ, a itsellēz lūkse mittāᴅ, which I take to mean you dig a grave for another, and you'll measure it for yourself, I guess the Ingrian equivalent of не рой другому яму, сам в неё попадёшь. By the way, that sounds like it could be a great entry one day!

Do you suppose we have to change the 3SG form in the inflection table to itseen? The itseehe form Nirvi gives sounds very suspicious, but on the other hand so does having the same form for both 1SG and 3SG GEN forms... Thadh (talk) 00:05, 20 February 2022 (UTC)

Ah, I've seen this in Nirvi, but thought you mean there's an example in a textbook. This seems to be an interesting example!

Coincidence of 1SG NOM and 3SG GEN doesn't seem suspicious to me, as -Vn suffix is expected in both situations. But -he suffix for the nominative seems indeed suspicious, but that I need to double-check, so I'll follow up on the form for 3SG NOM.

KirillW (talk) 10:37, 20 February 2022 (UTC)

@Thadh, we've revisited this topic today. Mehmet told me, that in Ingrian Finnish dialects possessive illative forms were sometimes observed instead of possessive nominative and genitive. This could explain NOM-3SG itseehe. We can't tell though, if it's mainstream or not. KirillW (talk) 21:27, 26 February 2022 (UTC)

What do you think we should do? We could maybe omit the 3SG-NOM altogether and just give the 3SG possessive paradigm at itse... Thadh (talk) 22:37, 26 February 2022 (UTC)

We could keep it as given by Nirvi, but add a note that this lacks attestation. KirillW (talk) 11:38, 27 February 2022 (UTC)

Okay, I've added two notes! Thadh (talk) 11:49, 27 February 2022 (UTC)

Ingrian pronunciation module

Hey, if you have time, could you take a look at User:Thadh/sandbox? Another editor and I have been working on creating an automatic pronunciation module, and I would value your imput, particularily on the vowel reductions (I've based them on your overview on Github). I'm fairly certain the Soikkola pronunciation are correct (I've checked them against audio in Rozhansky's dictionary), but since I don't have any Ala-Laukaa audio I couldn't check those. Thanks in advance! Thadh (talk) 13:32, 24 August 2022 (UTC)

@Thadh sure, I'll comment there KirillW (talk) 18:19, 24 August 2022 (UTC)