Reconstruction talk:Proto-Indo-European/ǵéwstus

etymology of *ǵéwstus

As far as I know Latin gustus and Gothic kustus stem from a zero-grade base, *ǵustus, not from the e-grade form *ǵéwstus. *ǵéwstus would have given Gothic *kiustus. I would like to move the lemma accordingly and delete or comment out the declension table, which shows an e-grade nominative. Any objections? --MaEr 20:04, 19 March 2011 (UTC)

The declension table shows that the noun has ablaut, and the nouns in both languages could be based on the zero grade forms. So I think the form could fit after all. —CodeCat 20:15, 19 March 2011 (UTC)

Pokorny (Indogerm. Ety. Wörterbuch) and Philippa (Etymologisch Woordenboek van het Nederlands) assume ǵustus (zero-grade nominative). I'd rather rely on them, not on the table. From where comes the information that the nominative has e-grade? --MaEr 20:51, 19 March 2011 (UTC)

I think u-stems with the -tu- suffix are always proterokinetic, so that would mean they always showed this kind of ablaut in PIE. —CodeCat 20:58, 19 March 2011 (UTC)

I understand. I guess the zero-grade nominative *ǵustus has evolved from the e-grade nominative *ǵewstus under influence of those cases which have zero-grade ablaut, analogy within the paradigm thus. --MaEr 07:59, 20 March 2011 (UTC)

Acc. sg.

@CodeCat Is the accusative singular *ǵéwstwm̥ correct? It makes phonological sense; it's just not what I'm used to seeing. The same question goes for *(é)-tym̥, which I'm more used to seeing cited as *(é)-tim. —JohnC5 19:30, 14 April 2016 (UTC)

That seems to be the result of a change I made earlier, to correct the dual of these nouns. The desired form is -wih₁, but the template turned it into -uyh₁. I'm not quite sure why the rules are different here. —CodeCat 19:32, 14 April 2016 (UTC)

Maybe some word-final rule that states CRR# > CR̥R#? I have no idea about the ordering, though. —JohnC5 19:41, 14 April 2016 (UTC)

To clarify, the module starts off with the stem ending in -u- and ending -ih₁ for the dual, -m̥ for the accusative. So the sonorants at the "edge" are in their vocalic form. Originally, it first looked if the stem ended in a vowel, and if so, it turned the sonorant in the ending into its consonantal form. Thus, ǵéwstu-m̥ > ǵéwstu-m but also ǵéwstu-ih₁ > ǵéwstu-yh₁.

I have no idea if this is a special case word-finally, but you can compare it to the verb ending -mi, which doesn't ever seem to appear as -m̥y. Perhaps there is a preference rule of some kind, that says u and i are chosen preferentially for vocalisation. —CodeCat 19:48, 14 April 2016 (UTC)

I looked at what Ringe says, here's an excerpt:

If the rightmost member of the sequence was adjacent to a syllabic (i.e. a vowel[...]), it remained nonsyllabic, but if not, it was assigned to a syllable peak. The rule then iterated from right to left

But he also mentions some exceptions, including:

Note also that the accusative endings are always nonsyllabic -m, -ns after a high vocalic.

So it appears that the change I made was essentially correct, and these two endings are just exceptions to the rule that need to be manually accounted for. —CodeCat 19:55, 14 April 2016 (UTC)

Fixed now. —CodeCat 20:01, 14 April 2016 (UTC)

Great! I'm glad I wasn't going crazy when I saw this change. —JohnC5 20:01, 14 April 2016 (UTC)