Reconstruction talk:Proto-Indo-European/kéh₃tus

How is this proterokinetic? All reflexes seem to be from just *kh₃tus. --Tropylium (talk) 17:40, 20 January 2017 (UTC)

@Tropylium:Look at the descendants from *méntis, *bʰértis, *stéh₂tis, *mértis, and *ǵéwstus. None of their descendants attests a full grade in the root. The classical explanation is that proterokinetic i- and u-stems were generalized to their weak stem already in PIE. There are some i- and u-stems that do show both stems though. The reconstruction of the paradigm though, is mostly based on the idea that there must be always one accented full grade, so the forms *mn̥tís, *bʰr̥tís, sth₂tís, *mr̥tís, and *ǵustús would not be admissible in that theory. –Tom 144 (𒄩𒇻𒅗𒀸) 15:22, 17 March 2019 (UTC)

So evidently this data therefore disproves the "must have a full grade" hypothesis, and per the actual reflexes we should indeed be reconstructing *kh₃tús etc. (Avestan and Lith. do seem to point to full-grade *stah₂tis though.)

(A slightly adjusted views could also be that some instances of *i and *u are full grades, compareable to cases like *muHs.) --Tropylium (talk) 17:58, 17 March 2019 (UTC)

@Tropylium: Yeah, I'm not always sure what to believe, ablaut is a little more complicated than 4 accent paradigms. Maybe wiktionary should be more tolerant to athematic reconstructions that do not conform to mainstream views, since many reconstructions are forced to one paradigm that the evidence does not always support. –Tom 144 (𒄩𒇻𒅗𒀸) 00:32, 20 March 2019 (UTC)

You'd have to make a list of all attested -tis and -tus stems across all languages, and see if there are no root-accented e-grades anywhere. I can already think of one exception: *ferþuz. —Rua (mew) 13:59, 20 March 2019 (UTC)

That would be interesting. These is a brief list of wiktionary u- and i-stems entries and their attestations:

Full grade on the root only

• *gʰóstis
• *wéstis
• *h₂énǵʰus
• *h₁wésus
• *smérus
• *swéh₂dus
• *tégus

Zero grade on the root only

• *gʷʰéntis
• *gʷémtis
• *dʰgʷʰéytis
• *suHnús
• *méntis
• *mértis
• *bʰértis
• *bʰénǵʰus
• *gʷréh₂us
• *h₁léngʰus
• *h₁wérus
• *mréǵʰus
• *pléth₂us
• *ǵéwstus
• *kéh₃tus
• *pléh₃tus
• *ǵénh₁tis
• *h₁n̥gʷnis

Both

• *pértus
• *stéh₂tis
• *déh₂itis
• *dʰéh₁tis
• *déh₃tis

It can be noted that the most prevalent group is the zero grade only, and the smaller is the true proterokinetics. It cannot be denied that there was some alternations, but this is rather scarce tbh.

Also @Rua, there are that include a "hysterokinetic" variant of u- and i-stems in zero grade only attestations, but the inflection table module does not display the plurals and instrumentals properly, should those tables be deleted, or the module adapted? –Tom 144 (𒄩𒇻𒅗𒀸) 00:35, 22 March 2019 (UTC)

You need to take into account the accent as well as the grade. *haþuz, on this page, reflects a root accent per Verner's law. Other Germanic forms reflecting an accent root are *gunþiz, *kumþiz, *burþiz, *kinþiz (also a full grade!). —Rua (mew) 11:42, 22 March 2019 (UTC)