Reconstruction talk:Proto-North Halmahera/kalubati

Descendants
Reconstructions need at least two descendants or a parent language with two descendants in order to make sense. I'm sure there are more descendants for this word (otherwise Holton and Klamer wouldn't reconstruct it), but it's still important to give these... Thadh (talk) 18:16, 26 December 2021 (UTC)
 * Ah ok! My plan was to fill them in as I go, but I'll try and find something for this one. Thanks for the heads up! Alexlin01 (talk) 18:20, 26 December 2021 (UTC)
 * Filling in as you go would be okay once you have the first two given, so that it has an etymological basis to exist. Anyway, good work!! Thadh (talk) 18:24, 26 December 2021 (UTC)
 * Makes sense! And thanks :D Alexlin01 (talk) 18:29, 26 December 2021 (UTC)

Austronesian Loan?
Hello! According to Blust (p. 35), there's a *qali/kali prefix or morpheme representing "creepy-crawlies", and on page 36, entry 66, he reconstructs PMP and PWMP *X-wati ("earthworm"), where the X represents all possible variant forms of *qali/kali. And, apparently Greenhill, Blust, and Gray specifically reconstruct PMP *qali-wati. If the *qali/kali morpheme was productive in PMP, it would seem like PNH *kalubati could be a loan.
 * You're certainly up-to-date on the newest reconstructions for PMP. Is this *qali-wati still being reconstructed? Are there forms with *qalu or *bati? Holton and Klamer reconstruct PNH to have both *w and *b, so a direct borrowing from *qali-wati should look like PNH **kaliwati. Alexlin01 (talk) 18:24, 27 December 2021 (UTC)
 * And! Do you know any PMP/PAN forms that could have been the donor for 🇨🇬? Alexlin01 (talk) 18:27, 27 December 2021 (UTC)
 * Yes, the recostruction is still upheld by Blust: https://www.trussel2.com/acd/acd-s_w.htm#28756. And here's a can full of worms: https://abvd.shh.mpg.de/austronesian/word.php?v=107. I haven't checked all entries yet, but in the area that I am most familiar with (Sulawesi), there is no potential source for *kalubati with medial *b, nor for Ternate kolotidi. In Saluan and Balantak (which are recipients of quite a few Ternatean and Moluccan Malay borrowings), tidiʔ means "urine", so this is just a (partial) sound-alike of kolotidi. –Austronesier (talk) 19:40, 27 December 2021 (UTC)
 * Thanks! I looked through the can, and none of the forms show the medial b. In fact, looking through our PMP/PAN reconstruction pages here, I haven't found a single case where medial PMP/PAN *w develops into a medial b in any Austronesian language. I guess I'll hold off on calling it a loan for now. Alexlin01 (talk) 20:06, 28 December 2021 (UTC)
 * Actually, there are two languages that have *w > b: Mentawai (e.g. *sawa > saba "pyhton") and Bambam (e.g. *tawa > -taba "laugh"). The first is of course far away, and also the latter is an unlikely source, being spoken in the interior of West Sulawesi ("worm" is kalindoho in Bambam). But still, the borrowing hypothesis is fully plausible, even without an obvious immediate donor language.
 * A similar case is *gate, clearly from PMP *qatay. Nearby languages have a zero reflex of *q, and the only dorsal reflexes of *q in eastern Indonesia are Watubela k and Muna gh [ʁ]. –Austronesier (talk) 23:48, 28 December 2021 (UTC)
 * Oh, thanks! I hadn't come across those ones before. Would you recommend I put something like "Ultimately from PMP *qali-wati"? Alexlin01 (talk) 17:44, 29 December 2021 (UTC)
 * In PMP *qatay, is the dorsal reflex reconstructible to Proto-Celebic or Proto-Central Maluku (or some lower-level reconstruction)? And you know, regarding PMP *qatay, I think it might could also (possibly) be a case of *N- prefixation, which becomes *g- in front of a glottal stop. Similarly, Holton and Klamer don't reconstruct anything for Ternate, but it really feels like it could be from NH *N- + PMP . But then, the only NH cognate with a third syllable is Modole gahili. Alexlin01 (talk) 17:44, 29 December 2021 (UTC)