Module talk:ine-nominals

Acc.pl -ms

@Rua, JohnC5, Victar, Per utramque cavernam, Florian Blaschke, माधवपंडित, Chuck Entz, AryamanA, Mahagaja, -sche: Would it be too crazy to revise the nominal desinences given by the module, so we can replace them for more suitable ones in case they're consensual? This table is based solely in one author (Ringe), and I think it'd be good if we discussed some possible flaws on this reconstruction.

More specifically, I'd like to propose two changes, firstly I'd like to change the accusative plural "*-ns" for "*-ms", which is supported by Sihler, Melchert, Kloekhorst, Yates, Lundquist, Pooth, Kortlandt, Kiparsky, and even Schleicher. The main reason for this change is that Hittite "-uš" [-ɔs] must come from "-ms". This ending can also be internally reconstructed as an accusative "*-m-" plus a plural "*-s", while "*-ns" can't. And finally the outcomes that point to "-ns" can be explained with an early assimilation of *m to *n due to the adjacent *s.

Later I would also like to propose taking Jasanoff's views on the "m/bh" problem, which are presented here. --Tom 144 (𒄩𒇻𒅗𒀸) 18:49, 2 June 2018 (UTC)

@Tom 144: I know nothing of Hittite but if your version does account for the descendants better, the changes should be made. I'd like to know what JohnC5 thinks about this. -- माधवपंडित (talk) 05:15, 3 June 2018 (UTC)

@Tom 144: These modules need reworking to include several different modules. Rua just programmed one version, but others exist. I just don't have the time anymore to work on it. If you can make a copy and update it with the several different modules then present it for approval, then do it. —*i̯óh₁n̥C^[5] 06:29, 3 June 2018 (UTC)

@JohnC5: I'm afraid I don't know how to edit modules. --Tom 144 (𒄩𒇻𒅗𒀸) 19:23, 3 June 2018 (UTC)

I'm not opposed to this speciallically, but I am generally weary about applying Proto-Indo-Anatolian reconstructions to Proto-Indo-European declension tables. --Victar (talk) 14:51, 3 June 2018 (UTC)

@Victar: I would support something like that but I don't think other editors would. Also, any reasons why you oppose the "-ms" reconstruction? --Tom 144 (𒄩𒇻𒅗𒀸) 19:23, 3 June 2018 (UTC)

Sorry, I left out the not, mistakenly. Also, my second sentence wasn't a call to any specific action. --Victar (talk) 19:53, 3 June 2018 (UTC)

@Tom 144 I think you jumped the gun in making the change already. You need to give people more time to reply and this discussion should have probably been stated in the Beer Parlour, despite the pings. --Victar (talk) 04:45, 4 June 2018 (UTC)

@Victar: Ok I'll revert it- --Tom 144 (𒄩𒇻𒅗𒀸) 04:49, 4 June 2018 (UTC)

@Tom 144: I'm not knowledgeable enough to contribute, but I trust you. You have a green light from me. --Per utramque cavernam 08:41, 4 June 2018 (UTC)

Oh, and also pinging @Tropylium. --Per utramque cavernam 08:41, 4 June 2018 (UTC)

I haven't heard of the Hittite argument for *-ms, but it makes sense to me. I don't have a strong opinion on the change, so if everyone else doesn't oppose it, I won't either.

As for the dative-ablative and instrumental plural endings, please read this paper (which also refers to Jasanoff's proposal), even if it's flipped to the side. (I have a preliminary version in PDF form which is easier to read.) I find the case here compelling. (I know, I've already promoted this paper several times before. It's really worth reading, though.) tl;dr: The original endings are dat.-abl. pl. *-m(y)os and instr. pl. *-mis, with the *-bʰ- endings originally variants found after stems ending in *-n- and (secondarily) *-nt-, due to a dissimilative sound law in PIE. Interestingly, Anatolian, despite showing in the nouns an apparently simpler ending dat. pl. *-os, is implied to have originally had the same distribution (another strike against Indo-Anatolian). --Florian Blaschke (talk) 09:48, 4 June 2018 (UTC)

@Florian Blaschke: The argument is not very extensive, "-ons" would have given "-aš" as in ḱwō(n) > ḱwōns > kuwaš, and as in the genitive of heteroclitic verbal nouns in "-war ~ -was" < "*-wr̥ ~ *-wéns". In the other hand a comparison can be made with the first person singular desinence "-un" [-ɔn] < "*-m̥", while "n̥" regularly yields "an" (e.g. "laman" < "*h₃nómn̥"). Kloekhorst briefly argues in favor of this reconstruction in the page 1073 of his dictionary.

Concerning the m/bh issue, I think we can agree that assuming "*-m-" in the dative & ablative and "*-bh-" in the instrumental isn't a very good explanation. Though I think Jasanoff's theory is more convincing. It would be great if we had some other opinions to see which solution is the most popular --Tom 144 (𒄩𒇻𒅗𒀸) 00:41, 10 June 2018 (UTC)

I'm aware that *-m̥ appears to have a different reflex than *-n̥ in Hittite. But do we know if Hittite -uš is from the o-stem ending *-oNs (as you assume) or from the consonant-stem ending *-m̥/n̥s (which is also possible)? This complicates your argument. I'm inclined to believe it, but I'm not sure the argument, as presented here, is completely solid. We'd like to have completely fitting examples for *-ons or *-n̥s respectively; I still see a lot of (smaller) holes in your argument.

See Hill, p. 181f., for problems with Jasanoff's proposal: it needs a lot of additional assumptions (the instr. pl. ending *-is postulated only on theoretical grounds is another). Hill's doesn't need so many additional assumptions at all, and in fact explains additional evidence that Jasanoff appears to have missed; it is more simple and elegant. --Florian Blaschke (talk) 18:35, 10 June 2018 (UTC)

I personally believe "-us" was generalized from the consonant stem. However Melchert and Kloekhorst think final accented "*-óm" yielded "-un" [-ɔn] in Hittite. They claim this on the basis of some accusative singular monosyllabic pronouns that ended in "u-un" [-ɔn] (irregular, the expected ending would be "-an"). Once accepted the claim that PIE "*o" can yield Hittite "u", Melchert concludes that "-us" is the regular outcome of "-óms", "-ums" and "-m̥s".

The phonological development of the n-stems is not certain of course, but the nominative "š" was obviously analogically restored, and it'd be a little strange to restore the "*s" without recovering the "*n" too. So the most likely development is the one cited above, although ḱwō > ḱwōs > kuwaš is still possible. According to Kloekhorst "*n" and "*m" assimilated to "*s" word internally as well, though his examples aren't compelling enough to be worth mentioning.

In my opinion it's unlikely for "*n̥" to give Hittite "-us", but the desinence is probably original because it's pronounced as [-ɔs], which must mean that it was once nasalized. It makes sense for "m̥" to yield "u" since they both have a labial element, but "*n" does not. I would propose a development as "*-m̥s" > "*-ũms" > "*-ũs" > "-ɔs".

A couple of objections to Hill that came to my mind:

1. Hittite "kuwapi" contradicts the proposed m/bh distribution
2. In my view the enclitic third person dative plural "-šmaš" is not inherited from the reflexive *se + -mós, since it's inflection was constrainted to the singular number. So "-šmaš" must instead be related to the enclitic third person plural possessive particle -smi-/-sme-/-sma-.
3. I'm skeptical of the HLuwian evidence for "*-mi", since the final "*i" was preserved in the verbal endings, in the dative-locative, and in the instrumental. So I see no reason for the "*i" to be deleted there.

However, I do believe that it is possible that the m/bh endings were competing variants of each other, which were phonetically-determined. That idea is way more compelling than the assumption "-bh-" in the instrumental and "*-m-" elsewhere. --Tom 144 (𒄩𒇻𒅗𒀸) 22:07, 10 June 2018 (UTC)

Hm, I see. But wouldn't *ḱwóns be a more plausible result of analogy than *ḱwōns, and also fit better into your argument?

There are a lot of smaller niggling uncertainties and possible nitpicks, especially about the (phonetical and morphological) developments in Hittite/Anatolian, that make me hesitate a bit, and keep me from being 100% confident about this line of argument, but overall I think the basic argument is correct. I just wish we'd have more certainty about the details, since even the researchers have divergent ideas. But maybe I'm too demanding.

OK, but none of your objections are decisive. I think they're roughly on the same level of my nitpicks of your argument. Again, we agree that the basic idea is probably correct. --Florian Blaschke (talk) 14:15, 11 June 2018 (UTC)

I just realised that your objections are already addressed inside the paper:

1. Hill points out that kuwapi < *kʷo-bʰí need not be PIE.
2. It's not the reflexive pronoun, it's the demonstrative pronoun *só.
3. Hill argues that -i could have been lost after diphthong + m already in Proto-Anatolian and later analogically restored in the verbal system.

So nothing really prevents us from accepting his proposal. He makes very remarkable points. --Florian Blaschke (talk) 21:40, 19 February 2019 (UTC)

Well, if I’m honest, I’ve been more or less convinced by Hill’s proposal, so in case action was taken to adhere to his views, I would support such change. Although I must note that assuming *i was lost after a diphthongs + m, and later analogically restored sounds quite ad hoc tbh. —Tom 144 (𒄩𒇻𒅗𒀸) 10:03, 21 February 2019 (UTC)

Neuter gender in adjectives

@Rua, could you add a parameter to specify the stems of the neuter gender in athematic adjectives? Given that neuter n-stems are proterokinetic while masculines are amphikinetic or hysterokinetic. Compare acc.sg.m युवानम् (yuvānam) < *h₂yuHónm̥, with acc.sg.n युव (yuva) < *h₂yéwHn̥. Also, nt-stems take different stems as well. ––Tom 144 (𒄩𒇻𒅗𒀸) 00:18, 25 October 2018 (UTC)

Genitive plural -óoHom

@Victar: Is the o-stem genitive plural -óoHom correct? Isn't it -óHom? —Mahāgaja · talk 16:39, 23 December 2020 (UTC)

@Mahagaja: Typo? Fixed. --{{victar|talk}} 04:05, 25 December 2020 (UTC)