Reconstruction talk:Proto-Indo-European/tn̥néh₂

Please stop

@Victar, several things:

• Words stressed on a reduced syllable are extremely uncommon, to the point that we would a assume a shift in the accent before reconstructing them.
• The *तना (*tánā) is not actually attested. Only the forms तने (tane, dat.sg.) and तना (tanā, instr.sg.) are found but without accentuation.
• Mayrhofer (and even MW) reconstructs a root noun, and MAyrhofer isn't even sure it comes from *ten-.

I'll give you a bit to see and respond to this, but I'll be moving this page back in a bit. —JohnC5 22:14, 23 July 2017 (UTC)

@JohnC5: Did you see the discussion I pinged you in? The point being that the noun suffix *-neh₂ actually put the stress on the root, not the suffix. --Victar (talk) 22:26, 23 July 2017 (UTC)

Victar, I only said that it appeared that way. No source actually specifically said so. Greek actually has many end-stressed forms with this suffix. Moreover, the presence of zero grade all but guarantees end stress, as JohnC5 noted. —CodeCat 22:30, 23 July 2017 (UTC)

@Victar: Again, there is no evidence for *तना (*tánā) being a feminine a-stem. Most authors prefer a root noun. I'm also not convinced we currently should be going full speed ahead on these *-neh₂ nouns. Many of your reconstructions of late have been quite speculative and ignore much modern consensus about PIE phonology and morphology. —JohnC5 22:36, 23 July 2017 (UTC)

Monier-Williams is the one who reconstructed the stress on *तना (*tánā). Not saying he's right, but if he was right, that would mean the stress is on the root. --Victar (talk) 22:48, 23 July 2017 (UTC)

Sorry, yes, I notice the stress now (I accidentally had the notation turned off). Regardless, MW reconstructs a root noun and the prohibition against *ń̥ still stands. —JohnC5 22:51, 23 July 2017 (UTC)

I was not aware of a "prohibition" on *ń̥. Is there some reading material you can direct me to? I've seen it reconstructed in various books and papers. --Victar (talk) 22:58, 23 July 2017 (UTC)

So, in most systems, *í, *ú, *ŕ̥, *ĺ̥, *ḿ̥, and *ń̥ are all considered very rare (normally considered to appear in later PIE or to be QIE). You'll notice that they very rarely occur in our reconstructions. This is because *y, *w, *r, *l, *m, and *n are non-syllablic phonemically but may become syllabic phonetically. There are certainly reconstructive systems that allow this (Fortson for instance), but more conservative systems prefer *é, *ó, *ḗ, and *ṓ. —JohnC5 23:13, 23 July 2017 (UTC)

As well as laryngeals holding stress. Thanks for the refresh. I'm wondering though where the line was drawn that where *ń̥ is "prohibited" in reconstructions. I know for one, *ŕ̥ is an essential part of PII, so rare as they may be, I don't see how they can be prohibited. --Victar (talk) 23:35, 23 July 2017 (UTC)

The normal explanation is that these stressed zero-grades are created via leveling (whether this is a good idea or not). You'll notice that in the descendants of *-tis, *-tus, and *-us, the full-grade nominatives occur infrequently. The normal explanation is that the ablaut pattern *é-∅ ~ *∅ ~ é leveled to *∅ ́-∅. Also, later innovation introduced types of zero-derivation and compound formation that could shift the pitch accent leftwards onto a normally unstressed syllable. @CodeCat, did I miss anything? —JohnC5 23:58, 23 July 2017 (UTC)

Soooo.... —JohnC5 05:19, 24 July 2017 (UTC)