Zompist Bboard Again

Frislander wrote: ↑Sun Apr 19, 2020 7:31 pmNukuoro?

no

I’m guessing West Uvean.

(If not that, then my next guess is Mele-Fila, but I think West Uvean is far more likely.)

EDIT: Yes, I’m totally sure now that it’s West Uvean. I just checked the paper I just downloaded about it, and it turns out that it has exactly the same sample text that Nortaneous included in his post!

I think the problem with this thread is that it’s far too easy to cheat, even when you don’t mean to — I was very careful to avoid searching for extended portions of the text, and merely wanted to find all languages which used a particular word, but it turns out that searching for even just one word (in my case, Logōtī) was enough to bring up the exact text that Nortaneous used. We’ll need some way to avoid the original sample text being found with such a simple query. (Idea: Maybe restrict ourselves to sample texts from the Grammar Pile and the Stack? I believe many of those documents aren’t indexed by Google.)

EDIT 2: I tried to compile a sample text myself, and found three techniques that seem to work fairly well to make a submission resistant against searching:

• I repeatedly looked for words in my sample text which looked particularly distinctive; I then deleted the sentence in the sample containing that word if I was able to find the language easily by searching for that word. After a few iterations I had a text where you could safely search for most words in it without revealing the language.
• If I take sentences from various glossed examples within the description rather than taking a full, coherent sample text, I can put them together to get a full sample text which is harder to search for. I think this is because the format of the gloss (with extra dashes and spaces and suchlike) makes it harder for the search to return a relevant result; also, searching for the full text won’t work either, since I actually just put together several sentences which were rather widely separated in the original document.
• The most effective technique I found was to use scanned documents which haven’t been OCRed. After all, you can’t search for text in an image unless it’s undergone OCRs.

I’d be happy to post the three sample texts I compiled using these techniques to see if anyone can figure them out only via Googling them. (I’m reasonably sure you can’t.)

My take on this is that the pleasure — the only pleasure — is in teasing one's memory, and secondarily the reward of realizing someone else out there has read about the same obscure language you have. If it's not something I recognize even broadly, then I just sit back and wait for the next one. "Making it harder" and "stopping/hindering cheating/online searches" sorta seems to miss much of the point, IMO.

Re: Itelmen — oh, I didn't even catch that "Kamchadal" also fits the "two of the same vowel and includes <l>"! Itelmen (that is, the western dialects that are really documented) is pretty distinctive, I agree — all the -knans and -qzens, etc. I had wanted to post something of the southern or eastern dialects, which look really quite different, but realized I only had excerpts of word lists and no continuous text at all

fusijui wrote: ↑Mon Apr 20, 2020 2:55 pm My take on this is that the pleasure — the only pleasure — is in teasing one's memory, and secondarily the reward of realizing someone else out there has read about the same obscure language you have. If it's not something I recognize even broadly, then I just sit back and wait for the next one. "Making it harder" and "stopping/hindering cheating/online searches" sorta seems to miss much of the point, IMO.

Interesting! I see it as exactly the opposite — my knowledge of obscure languages is fairly bad, so the fun for me is in trying to search and figure out what language it is using only the clues which have been given.

bradrn wrote: ↑Mon Apr 20, 2020 1:53 am I’m guessing West Uvean.

yes

unfortunately most of the papers I have are indexed. for less minor languages videos with captions are probably ideal - a while ago I found some videos with a lot of text in Pumi, but they've disappeared

It looks like no-one has posted a new text yet, so I’ll go ahead with one of my own:

Aṭȯṭa uk̭hiḍek ṭimikhek̭el̃ or ṭimikhek̭ȯkȯ. Ṭimikhu tarakamu bel̃o. Ṭȯkho tȯṭṭȯṇ toṭkatal̃ tag̭ew totbek̭ayė birañe. K̭al̃ȯ rak̭ȯṇ phol̃ȯ. Ukuntȯ bol̃ȯl̃ ubul̃iwek tag̭ew tȯk̭ȯre erbul̃in binkel̃ ṭȯtbėyik̭ nȯnk̭ekum. Ṭibura l̃il̃e mel̃ȯṇ uṇk̭ayė phol̃. Bol̃ k̭okbi motombȯye ḍel̃ok̭hara. Bol̃ k̭okbi nȯnel̃ȯwe krėko.

The author of the study I got this from uses a transcription system which I can only call idiosyncratic, so I’ll translate it into IPA as well:

aʈɔʈa uciɖek ʈimikʰeceʎ or ʈimikʰecɔkɔ. ʈimikʰu taɾakamu beʎo. ʈɔkʰo tɔʈʈɔɳ toʈkataʎ taɟew totbecajɛ biɾaɲe. caʎɔ ɾacɔɳ pʰoʎɔ. ukuntɔ boʎɔʎ ubiʎiwek taɟew tɔkɔɾe eɾbuʎin binkeʎ ʈɔtbɛjic nɔncekum. ʈibuɾa ʎiʎe meʎɔɳ uɳcayɛ pʰɔʎ. boʎ cokbi motombɔje ɖeʎocʰaɾa. boʎ cokbi nɔneʎɔwe kɾɛko.

Oh, and one last note: this document wasn’t OCRed, so searching probably won’t work. (That’s also why the sample text is so small — there was more, but I couldn’t be bothered to type it all.) Good luck!

Judging by the retroflexes and lack of fricatives: Dravidian?

Yalensky wrote: ↑Tue Apr 21, 2020 11:48 pm Judging by the retroflexes and lack of fricatives: Dravidian?

No.

(It’s interesting that you associate retroflex + lack of fricatives with Dravidian… for me, that sounds more like Australian languages. But it’s not Australian either.)

bradrn wrote: ↑Wed Apr 22, 2020 12:02 am No.

(It’s interesting that you associate retroflex + lack of fricatives with Dravidian… for me, that sounds more like Australian languages. But it’s not Australian either.)

Well Pama-Nyungan was my next guess, so there goes that idea...

Andamanese?

Nortaneous wrote: ↑Wed Apr 22, 2020 12:06 amAndamanese?

Very good! How did you get it so quickly?

EDIT: But there’s several language families in the Andaman islands; I assume you mean Great Andamanese.

bradrn wrote: ↑Wed Apr 22, 2020 12:07 am

Nortaneous wrote: ↑Wed Apr 22, 2020 12:06 amAndamanese?

Very good! How did you get it so quickly?

Looked like it. And no fricatives, but Ph P B - what else has that? But I half expected it to be some South American language instead; I'm not really sure what they have there, but there are a few fricativeless languages on that continent. (On reflection, Central African could've been more likely, but the words are too long and there are no tones or anything. I doubted it because I don't recall Jarawa having /ʎ/ and that's the Andamanese language I read about most recently, but I haven't seen much on Great Andamanese.)

So it's Great Andamanese then, rather than Onge or Jarawa?

Nortaneous wrote: ↑Wed Apr 22, 2020 12:11 am

bradrn wrote: ↑Wed Apr 22, 2020 12:07 am

Nortaneous wrote: ↑Wed Apr 22, 2020 12:06 amAndamanese?

Very good! How did you get it so quickly?

Looked like it. And no fricatives, but Ph P B - what else has that? But I half expected it to be some South American language instead; I'm not really sure what they have there, but there are a few fricativeless languages on that continent. (On reflection, Central African could've been more likely, but the words are too long and there are no tones or anything.)

I chose Great Andamanese because I thought it was too obscure for anyone to know anything about it. Looks like I was wrong…

So it's Great Andamanese then, rather than Onge or Jarawa?

Yes.

Great Andamanese is more obscure than Jarawa, at least, but given that Dravidian looks a hell of a lot like it belongs in Australia, one naturally takes some interest.

The source text here is handwritten, so I'm not entirely sure about word division, and the pairs <r v> and <q a> are sometimes hard to distinguish. (But the text definitely uses all of these letters.)

Heua oira kekeriva sa ira reoreoru pirio airari; Aiva, apasuiripa karou. Oira ouhueue. Bohituku taharoro pira oira uririeiva bikorou. Are hauk mae avaea karou. Roarai lavisiop. Uei avioga kekeva, areha bioga uririeva haia. Sa oira Akuhuihue aipo ruvoqe. Areha uka oira kekeivoh. Ara haia. Areha oira tarioiup bopuava ruru opa ovarai avaire sa ovarai rava veriveri iu.

Nortaneous wrote: ↑Wed Apr 22, 2020 12:34 am Great Andamanese is more obscure than Jarawa, at least, but given that Dravidian looks a hell of a lot like it belongs in Australia, one naturally takes some interest.

The source text here is handwritten, so I'm not entirely sure about word division, and the pairs <r v> and <q a> are sometimes hard to distinguish. (But the text definitely uses all of these letters.)

Heua oira kekeriva sa ira reoreoru pirio airari; Aiva, apasuiripa karou. Oira ouhueue. Bohituku taharoro pira oira uririeiva bikorou. Are hauk mae avaea karou. Roarai lavisiop. Uei avioga kekeva, areha bioga uririeva haia. Sa oira Akuhuihue aipo ruvoqe. Areha uka oira kekeivoh. Ara haia. Areha oira tarioiup bopuava ruru opa ovarai avaire sa ovarai rava veriveri iu.

Is it a Southeast Solomonic language?

bradrn wrote: ↑Wed Apr 22, 2020 12:40 am Is it a Southeast Solomonic language?

no

Nortaneous wrote: ↑Wed Apr 22, 2020 12:51 am

bradrn wrote: ↑Wed Apr 22, 2020 12:40 am Is it a Southeast Solomonic language?

no

In that case, is it Austronesian at all?

(I think I’ll need a bit of help with this language, since I don’t recognise it at all… It looks almost Polynesian, but I know of no Polynesian language with ⟨q⟩. Fijian is almost Polynesian and has ⟨q⟩, but it uses ⟨q⟩ far more than your text does.)

Oceanic?

Very obviously a language spoken in the southern part of Bougainville. Looks like Rotokas, but not quite. Konua?

Karch wrote: ↑Wed Apr 22, 2020 5:45 am Very obviously a language spoken in the southern part of Bougainville. Looks like Rotokas, but not quite. Konua?

Minor nitpick: I believe that should be the northern part of Bougainville.