Reconstruction talk:Proto-Germanic/furskō

Deletion debate
, I see you went and re-created several PG entries that I had moved to PWG ones, including. If a PWG entry has no cognates, particularly when they have an uncertain etymology, it really should be left to a PWG entry. PWG is a nominalization of an unattested verb and could just as well be  + *-u, never existing in PG as. Similar to, , , , etc., all with suffixes productive in PWG.

You also re-created PWG which we had a whole discussion here (and here) on being a WG variant of PG.

-- 15:15, 27 July 2020 (UTC)
 * Agreeing with Victar, back to Proto-West-Germanic. I guess similarly 🇨🇬 was not formed in PIE, there is much time in between. A big maybe belongs to its etymology. Fay Freak (talk) 15:47, 27 July 2020 (UTC)
 * Fay Freak, they still have their PWGmc entries and those remain untouched, so there is no question of restoring to PWGmc, only of deleting PGmc entries.


 * To defend these re-additions... The initial impetus for re-creation is that most of these are actually still listed as etyma in the etymologies of the PWGmc entries without qualifiers or alternate etymologies that do not derive them from PGmc (and some, such as *furskō, are also listed elsewhere on Wiktionary as well), so their re-addition seemed justified and conversely their wholesale move to PWGmc without retaining any PGmc entry did not. Many are also listed in authoritative sources. I mostly checked Kroonen, tbh. *furskō specifically is plausibly (to me at least) explained by Kroonen as representing the PGmc of a pre-Germanic or PIE word. I'm not sure why we should suppose a PWGmc *forskan as a base for PWGmc formation of *forsku to be likelier than his etymology here, even though the existence of *forskan is possible. One of these theories is supported by an authoritative source, the other is not, although there is room for both on Wiktionary.


 * Regarding - I had missed the discussions (I am not sure they really indicate that the word formed in PWGmc, just that its reconstruction is difficult and there appears to be a possible alternation in the stem?), but anyway Kroonen has it which is what I based my re-addition on. Does look like one of the more difficult cases, I'll give you that...


 * Some of the others are also found in Kroonen. I'll admit that for many of these it's dubious when exactly they formed, but a big thing for me here is: how would they be formed in PWGmc specifically? Because if they could not be formed in PWGmc, they must have existed in PGmc. This applies to e.g., , and  - how would they be formed in PWGmc? From what base? That the suffixes found in them may have been productive in PWGmc could clearly indicate PWGmc formation, but there needs to be something to add a suffix to, and I see little indication for some of these words (a good example imo is  *gerstu) that there existed a PWGmc base from which it could plausibly derive synchronically. Inheritance appears to be the likelier explanation.


 * Anyway, perhaps I need to think about this longer, but I think there's a strong enough case to be made for the inclusion of most of these.


 * (As an aside I disagree with the idea that a reconstruction entry needs cognates or even multiple descendants, for the reasons mentioned here by various people. Some such words are very likely to have existed for the simple reason that they cannot plausibly have been formed in the daughter language and must belong to an older language stage.) — Mnemosientje (t · c) 18:12, 27 July 2020 (UTC)
 * I did not imply the PWGmc is gone, I meant “cut it back to it”. “How would they be formed in PWGmc specifically?” For this entry I think Victar has told it. Fay Freak (talk) 18:21, 27 July 2020 (UTC)


 * Just because an author creates a reconstruction, doesn't mean carte blanche inclusion on the project. We reject many things from academics, including Kroonen, who reconstructs quite a few words in PG despite them provably only existing in PWG, like late borrowings from Latin. This very much applies to.
 * The problem with creating ancestral entries for words of uncertain paternity is not just that we've now created a dubious entry, but we would have multiple competing entries for the same descendant. If we have PG, we should also have PWG and PG.
 * -- 16:01, 28 July 2020 (UTC)
 * -- 16:01, 28 July 2020 (UTC)


 * I will admit I am really not sure about *wulbī after seeing the discussions. I wouldn't mind too much if it were to be deleted for that reason, although then I think even then it would be desirable to have a clearer discussion of the difficulties of the etymology at the PWGmc entry. That, or we should go with a common PGmc entry to unite the West- and North-Germanic forms, but like you (judging on the discussions you linked) I am not sure what the lemma form of that entry would have to be.
 * As for *furskō, I can see where you're coming from. Do note that you yourself recently added the PIE pedigree and cognates for the noun over at . Should that then not be presented with more scepticism? I understand that each of these 'cognates' may as well represent an independent innovation without being inherited from PIE (although I don't know enough about Sanskrit and Armenian to judge those cases). That uncertainty is not currently mentioned on the PIE entry (nor on the Sanskrit and Armenian entries btw). But anyway for lack of clear reasoning from Kroonen why PIE inheritance would be more likely than later formation (he doesn't mention the supposed non-Gmc cognates and they seem uncertain anyway) I can't think of a strong argument to privilege his etymology in this case anymore and keep the PGmc I restored other than an argument from authority. Both possibilities rely on hypothetical chains of inheritance spanning a couple thousand years, neither appears particularly much stronger than the other (unlike in the case of *gerstu etc., where a PWGmc formation seems highly unlikely). I think deletion would be reasonable enough, I am not certain enough of the PGmc I created.
 * In cases like *gerstu and the rest it appears far more likely that it formed before PWGmc for reasons mentioned above, particularly the lack of a clear base from which these words would be formed in PWGmc. In these cases a PGmc or earlier origin seems much more likely. That likelihood (coupled with occurrence of the PGmc etymology in authoritative sources) seems to me enough to keep a PGmc entry: no alternatives are offered, let alone plausible ones. — Mnemosientje (t · c) 19:01, 2 August 2020 (UTC)


 * Re: : Right, which would make it a question of where to move PG, meaning that shouldn't exist.
 * Re: : 👍 I've gone ahead and removed the link to it on.
 * Re: The rest: I only RFD'd the above two. As you say, these have clearer PG reconstructions.
 * -- 16:21, 5 August 2020 (UTC)
 * That clarifies a bit, I thought each of the entries you listed was being RFD'd. As for, it can be kept at its current place or moved to some other PGmc form depending on the outcome of the discussion at Reconstruction talk:Proto-West Germanic/wulbi. — Mnemosientje (t · c) 10:30, 12 August 2020 (UTC)
 * No, still needs deleting.  will be moved to whatever we land on as the PG form. --  14:35, 12 August 2020 (UTC)
 * Sure, the end result is the same anyway. — Mnemosientje (t · c) 18:40, 12 August 2020 (UTC)
 * Except one was created in 2011, and the other last month. -- 19:15, 12 August 2020 (UTC)
 * Keep for the reasons Mnemosientje has stated. —Rua (mew) 09:27, 28 July 2020 (UTC)
 * I have deleted *wulbī and *furskō as being created in error, so this can be regarded as closed. — Mnemosientje (t · c) 13:46, 21 January 2021 (UTC)
 * I have deleted *wulbī and *furskō as being created in error, so this can be regarded as closed. — Mnemosientje (t · c) 13:46, 21 January 2021 (UTC)