Reconstruction talk:Proto-Germanic/wiljaną

Aorist-preterite
Shouldn't we reconstruct the preterite as singular *wel-d- but otherwise *wul-d-? It would better explain the -u- vocalism found in West Germanic. Anglom (talk) 04:32, 18 December 2017 (UTC)
 * What would be the source of the alternation? No other weak pasts alternate. —Rua (mew) 17:59, 18 December 2017 (UTC)
 * It would presumably originally have been an aorist relegated to the preterite. It explains the Gothic past 𐍅𐌹𐌻𐌳- against West Germanic *wuld-, with regularization in opposite directions. Anglom (talk) 18:38, 18 December 2017 (UTC)
 * The Gothic past is easily explained as analogical. —Rua (mew) 18:41, 18 December 2017 (UTC)
 * The past stem would still have originally been *wul- then, which would still be irregular. Rather it was probably originally an aorist that was supplied with weak preterite endings because *wel-/*wul- was incompatible with a strong preterite. Anglom (talk) 19:03, 18 December 2017 (UTC)
 * But aorists normally surface as presents in Germanic. —Rua (mew) 23:14, 18 December 2017 (UTC)
 * Usually, but sometimes they can form a preterite, like *stōþ (< *stéh₂-t) and *hlōþ (< *kléh₂-t), where the 3sg. *-t was reanalyzed as part of the root, giving general past *stōd- (< as if *st(e)h₂-t-V́-). I'm not really sure what motivates them to become either present or past in a paradigm, except maybe whether or not a derived present already exists to which they can be associated. Anglom (talk) 02:35, 19 December 2017 (UTC)
 * Actually, many characterised presents were lost in Germanic, and their accompanying aorists took their place. This is why Germanic has relatively few characterised presents; there's no reduplications, no nasal infixes, and no suffixed presents other than j-presents. The aspectual distinction between imperfective and perfective was lost in Germanic, which rendered most characterised presents superfluous. I think those few irregular pasts can be explained another way, namely as imperfective pasts. We know that has its origin in the past tense  of an original reduplicated imperfective verb . It's likely that  comes from . However, perfective/aorist is not distinguishable in form from an imperfective past, and if imperfective and perfective merged, any remaining perfectives were at risk of being reinterpreted as imperfective pasts. We know because of the evidence of  that the imperfective past survived longer in Germanic before the stative/perfect took its place, and that this replacement missed at least one irregular item. So it's feasible that  went the same route, except that it was reinterpreted as an imperfective past rather than already being one in origin. —Rua (mew) 18:25, 19 December 2017 (UTC)
 * That actually sounds very likely to me. And I feel that *dʰédʰeh₁ti is extremely relevant to this conversation, because the expected perfective formation of that root, *dʰéh₁t > *dēþ, could very likely have given rise to the non-singular stem of its preterite: as *stōþ > *stōdun, so *dēþ > *dēdun, but the singular stem *ded- amazingly resisted an analogical change to *dēd- and/or its later transition to *dad- as happened to the class 5 strong preterites.
 * So, rather strangely, the inflection of the verb *dōną in Germanic shows features of at least three different formations: an o-grade athematic present *dō- < *dʰóh₁(ti); an imperfective past singular *ded- < *dʰédʰ(eh₁t); and a (quasi-)perfective non-singular stem*dēd- as if < *dʰ(e)h₁-t-(ń̥t) which could only exist after the model of the class 5 preterite. Anglom (talk) 06:46, 20 December 2017 (UTC)
 * I think the crucial point that prevented from being converted to o-grade is that it ended in a vowel. Statives/perfects of this shape normally belonged to class 7, which had irregular ablaut to begin with. Moreover, the 2nd person form  was completely different from any stative/perfect form.  kept its original inflectional ending  by virtue of it being monosyllabic, but since it was an old relic, this ending was no longer recognisable as such. However, the original 2nd person form would have been, and surely the s-t alternation would have been recognisable to speakers, albeit not as past-tense ending. —Rua (mew) 12:51, 20 December 2017 (UTC)