Reconstruction talk:Proto-Indo-European/h₁réh₁

@Victar What is the logic behind Proto-Balto-Slavic **rēˀtas? There is no evidence for long vowel in the proposed descendants. Is Latvian ręts different from Latvian rets? I can't find it anywhere. The form *h₁réh₁-to-s which Beekes talks about in the etymology of ἐρῆμος (erêmos) is in regard to Lithuanian rēte (“net”), not the Baltic data. The 0-grade *h₁r̥h₁-ró-s for Latin rārus is also dubious. De Vaan talks about etymology from Proto-Indo-European *h₂erH- (“to perish”), without proposing possible derivation from Proto-Indo-European *h₁reh₁-. He does propose, though, a 0-grade *h₁r̥h₁-ti-s for Latin ratis (“raft”), which is with short grade.

PS Proto-Slavic *rědъkъ does not require Proto-Balto-Slavic *rēˀdkas. Normal u-stem *rēˀdus suffices. In general, u-stem adjectives in Slavic were regularly extended with Proto-Slavic *-ъkъ. This mechanism, however, may not have stretched back to Balto-Slavic times. I have not checked all possible research on the topic but it may be better to stay away from Proto-Balto-Slavic *-C(u)kas in place of Proto-Balto-Slavic *-Cus as early as Balto-Slavic. Bezimenen (talk) 17:21, 15 October 2019 (UTC)

@Bezimenen: You first have to start with Pokorny, which is probably the earliest source to group many of these words. The connection to the "net" terms have been since been rejected. Even De Vaan admits his *h₁r̥h₁-ti-s > ratis (“raft”) reconstruction is pretty weak. Schrijver is the one who gives *h₁r̥h₁-ró-s, and if it's good enough for Schrijver and Beekes, it's good enough for me.

I've added some more sources for the BSL forms, which I hope help in some way. I believe I've also seen it suggested that the -dh- in BSL is secondary, which I don't think that's necessary, but feel free to clean those up to what you think works best. Thanks for the help. --{{victar|talk}} 19:30, 15 October 2019 (UTC)

@Victar I'm asking about the grade in Lithuanian retas and Latvian rets. Neither of these exhibit a long grade in order to justify Proto-Balto-Slavic *rēˀtas. The opposite holds for *h₁r̥h₁-ró-s. I cannot tell how accurate is the etymology of Latin ratis analysed by de Vaan, but at least the development of the cluster *h₁r̥h₁- according to him yields a short grade. The contrary happens with *h₁r̥h₁-ró-s > Latin: rārus. Of course, if we treat h₁r̥h₁-C as a full fledged CRHC, then a long grade is to be expected, so I'll not argue too much about it. Definitely, though, the Baltic reconstruction needs further justification. I'll look for another potential de-lengthening of this sort. It may be reasonable, I'm not sure. The simpler solution, though, is just to reconstruct Proto-Balto-Slavic *retas and to look for reasons why the Baltic forms do not reflect -eh₁-. Bezimenen (talk) 21:21, 15 October 2019 (UTC)

PS I'm presuming that Latvian ręts which is given in the entry is the same as Latvian rets. I cannot find ręts on its own.

@Bezimenen: Well, just because it yields a long vowel in Latin, doesn't preclude it from yielding a short vowel in BSL. That said, I'm still working on a parent entry, which could possibly be a better fit for these BSL forms, but it's still very much a work in progress. Something to keep in mind though. --{{victar|talk}} 21:33, 15 October 2019 (UTC)