Reconstruction talk:Proto-Germanic/ampraz

Out-of-date back-projections

@Victar: So the descendants given for Old English and OHG–NHG mean sorrel. The page claims them “from a substantivation *amparô”, but this is probably not how the language works—to give a plant the name “the sour one”. The suggestion at the English entry amper is more sensible, positing a Proto-Germanic *ampra-, *amprōn- (“dock, sorrel”), apparently meaning an ōn-stem like *madarǭ, save that it would Proto-West Germanic, for your plant collection. The Sanskrit comparison of course does not at all vouch for an Indo-European word for sorrel, and may be not even be a parallel formation for formal reasons and better derivations and interpretations of the word. If amārus belongs to it, why have we missed it on Proto-Indo-European *h₂eh₃mós we derive the Latin from? Instead, there, we give a better-known word आम (āma) as the direct cognate, with copious other Indo-Iranian cognates, while from experience such an obscure word, of unknown usage and cognates, tends to be too good to be truly a cognate. That’s why we missed it, the equation is probably deprecated—it is from the 19th century.

It has been added by the rare user @Nin-TD, but this derivation is even parroted nowadays. In the cited Kroonen the Sanskrit comparison is juxtaposed with a mutton dash, as well as Latvian amuols (“yellow-sorrel”, glossed there), which could suggest they themselves do not even believe this reconstruction of PIE plant names; formally it would be according to them *h₂ém-ōl, gen. *h₂m-l-ós, which looks like *h₂ébōl (“apple”), for which similarity alone one is inclined to discard. But the glossing also lapsed by this point, Latvian āmuls, amuls, amuols, Lithuanian ãmalas is “mistletoe”, cognate to Russian оме́ла (oméla) etc., secured by toponymy (the relation of the Balto-Slavic mistletoe word to the Ampfer word in Vasmer and Trubachev is of course very wild, also relying on confusion about the origin of Mistel). Maybe there is something else nowadays known to replace with—forms have to be replaced. Fay Freak (talk) 00:56, 14 May 2021 (UTC)

@Fay Freak: *ampr (“sour”) +‎ *-ō (agent suffix) => *amprō (“sorrel”) seems like a sensible reconstruction to me. Compare *sūrā (“sorrel”) and *habbjō (“barm”). That said, taking the Sanskrit into consideration, the meaning of sour/sorrel could date back to PIE and PWG went and split the term into two separate forms. --{{victar|talk}} 06:34, 14 May 2021 (UTC)