Talk:𐌳𐌰𐌻

Unattested nonlemma forms
Hey, thanks for adding the genitive plural. I was thinking it may be better to avoid creating unattested entries like even if it may be the expected form, because in the end it remains unattested and thus shouldn't really be included. The way I've handled this kind of thing in the past is simply adding the "unexpected" spelling entry as a typical form-of entry (not a soft redirect to the expected spelling), noting its irregularity on the lemma entry under the inflection/declension/conjugation header (and possibly on the form-of entry as well if you wish). That way there are no unnecessary entries and no ambiguity as to what is attested and what is not and the entry remains consistent with other Gothic entries. Note for example how I dealt with the irregular form attested for. (In that particular case, the irregular spelling may reflect a real ambiguity of pronunciation, as I've found the final -þ for that verb form to alternate with -d occasionally in other verbs as well.) Anyway, if you agree with this approach, the unattested expected form would be deleted. — Mnemosientje (t · c) 08:36, 12 March 2019 (UTC)
 * I did debate with myself about whether to include or not, and finally decided I wanted to include it so that I could add the pronunciation, which would have looked weird on . If  is truly a scribal error, a misspelling, then the correct pronunciation is , not *; but I can't quite bring myself to list  as the pronunciation of . So instead I left  without a pronunciation section and instead listed it at unattested . I suppose we could move [[𐌳𐌰𐌻𐌴]] to [[Reconstruction:Gothic/𐌳𐌰𐌻𐌴]], but that feels like overkill. In other ancient languages (Latin, Ancient Greek) we do frequently have pages for unattested nonlemma forms if we're confident as to what the form would be if it were attested, so having a page for  isn't really that different. I wouldn't encourage it for every single unattested nonlemma form in Gothic, but in this case I think it's ultimately easier on the user to have  than not to, especially since it's made clear that it's only an expected form, not an attested one. —Mahāgaja · talk 10:14, 12 March 2019 (UTC)
 * Ah, I hadn't considered the issue of adding pronunciations -- I personally almost never add them for non-lemma forms (too lazy). That makes sense, although personally I dislike adding any unattested form (predictable Latin and Ancient Greek unattested forms have also been deleted following failed RFV in the past). I agree a Reconstruction namespace entry is a bit much for what is ultimately an easily predictable form though. I've added a small note specifying explicitly on the lemma page that dalē is in fact unattested (just to soothe my perfectionism). — Mnemosientje (t · c) 10:49, 12 March 2019 (UTC)