Reconstruction talk:Proto-Germanic/risiz

Gemination
Something's off about this reconstruction. If it were -sj- then you'd expect gemination in West Germanic, but none is attested. —CodeCat 21:02, 11 August 2016 (UTC)
 * Should it be an i stem ?
 * That seems likely. —CodeCat 21:52, 11 August 2016 (UTC)

Also, I'm very skeptical about the initial w-. You'd expect it to surface in Northern West Germanic and in Continental North Germanic (compare ), but it doesn't. —CodeCat 21:09, 11 August 2016 (UTC)
 * This is not wholly unusual. Compare *vrangr/rangr; vrist/rist. The initial w- connects it (possibly/probably) to extra-Germanic cognates, like . Leasnam (talk) 21:31, 11 August 2016 (UTC)
 * I'm not sure what the / comparison implies here. The w- of those terms is preserved exactly in the descendants where it's expected to be, just like in . However, in this term, there is no such distribution; w- is lacking throughout the direct descendants, and only shows up in a few purportedly derived/related terms. agrees with me as well. Combined with the point I raised above, I think this should be . —CodeCat 21:52, 11 August 2016 (UTC)
 * If it's *risiz, what's its parent ? Would it then be a relation of ? Or would it possibly still be from an earlier PGmc *wrisiz with the current parent as is ?Leasnam (talk) 21:58, 11 August 2016 (UTC)
 * There's no precedent for a PG wr- > r- change, and plenty of counterexamples, so to suggest that would be very unscientific. Equally, if we stick with wr- for Germanic, we'd have to explain why the w- was lost in most descendants, even those where there are plenty of counterexamples preserving wr- (we already have 3 in this discussion alone). The only linguistically sensible way out of this is to posit r- for Germanic and assume that this does not come from older wr-. —CodeCat 22:06, 11 August 2016 (UTC)
 * Ok. Let's move it Leasnam (talk) 22:11, 11 August 2016 (UTC)
 * According to Pfeifer, Old Saxon wrisi (besides wrisilīk "riesenhaft") and Old Dutch wrisil are also attested. (While a variant riso is also found in Old Saxon, this might be a borrowing from Old High German.) If so, only the absence of the w- in North Germanic remains problematic. --Florian Blaschke (talk) 11:54, 1 February 2017 (UTC)

I've updated the entry to reflect the sources. I think this entry should be moved, with 🇨🇬 and 🇨🇬 reflecting a secondary PWG agent suffix form. It's far more likely that the initial *w- was dropped than simply added to OS and OD. To also note, it's been hypothesized that the ethnonym is derived from this word. -- 08:41, 2 May 2020 (UTC)
 * See above. —Rua (mew) 09:15, 2 May 2020 (UTC)
 * I have, and find your arguments uncompelling and you never addressed any of 's points. -- 09:18, 2 May 2020 (UTC)
 * Yeah, no. —Rua (mew) 09:23, 2 May 2020 (UTC)
 * So you're arguing that 🇨🇬 (cf. OHG ),, and 🇨🇬 are just all unrelated? --  09:30, 2 May 2020 (UTC)
 * They must be, because they begin with wr- and this term doesn't. —Rua (mew) 09:31, 2 May 2020 (UTC)
 * I have no idea how to explain the wr/r discrepancy, but from it appears the two forms are regularly related to one another in scholarly literature anyway. If you want to problematize that notion in the etymology section (as you did at *risiz) and/or with qualifiers or whatever elsewhere in the entry, that's fine, but just removing all mention of the wr- forms here as if they don't exist doesn't seem like a good idea. — Mnemosientje (t · c) 10:03, 2 May 2020 (UTC)
 * I think the entry should stay where it is since none of the listed descendants suggests a *wr-; but of course we should continue to list the alternative reconstructions with *wr- since they're sourced and discuss the problem in the Etymology section as we already do. Personally I think it's most likely there are two etymologically unrelated stems risi- and wrisja- that were probably semantically close enough (maybe the former was a noun meaning "giant" and the latter an adjective meaning "huge"?) that they got conflated in Old Saxon and Old Dutch. I certainly don't think it's a matter of a simple phonological rule *wr- > *r- that applies in just this word in languages that otherwise don't have that rule. —Mahāgaja · talk 10:07, 2 May 2020 (UTC)
 * Yeah, I've wondered if they might be two unrelated lexemes myself. If we choose this solution, the difficulty arises how to assign forms with no trace of initial *w- to either stem, given that they are etymologically ambiguous. --Florian Blaschke (talk) 17:53, 2 May 2020 (UTC)
 * Unless someone reliable has actually proposed two unrelated lexemes in print, then we shouldn't have separate entries or anything like that. Just list everything on this page and comment that some forms have an initial w- of unknown provenance. —Mahāgaja · talk 17:56, 2 May 2020 (UTC)
 * And to quote R:nl:NEW, "De woorden te scheiden is op grond van de gelijke betekenis ook bezwaarlijk." -- 05:53, 3 May 2020 (UTC)
 * OHG and OWN aren't the only Germanic languages to dissimilate #wr- to #r-. MLG sporadically exhibited this as well, > MLG, as does Dutch,  > Middle Dutch , and Frisian,  > Saterland . Now you could argue that some of these forms may be borrowed from High German, but then we could say that about this entry as well. --  03:11, 3 May 2020 (UTC)