Reconstruction talk:Proto-Germanic/wulgī

Old Swedish
In Old Swedish, the word for she-wolf is ylva, which is now a common forename in Swedish for women, but I believe this might be a later construction based on ulver, the word for wolf, since it contains no G. Could this be correct? --Lundgren8 (t · c) 12:49, 21 June 2011 (UTC)
 * Yeah, I thought so, as well. Except for the forename, the word "ylva" isn't in actual usage in modern Swedish, a she-wolf would be a "varginna" from varg and feminine ending -inna. Wakuran (talk) 01:03, 23 November 2018 (UTC)

Form of reconstruction
How about Middle High German wülpe "she-wolf" (Old High German wulpa)? In light of this form and of the equation between ylgr and Sanskrit vrkí:s (sorry, somehow the special character insertion doesn't work anymore), shouldn't the reconstruction be *wulgwi:z? In fact, I seem to recall this is what Kuiper reconstructed in his article on Gothic bagms and Old Icelandic ylgr (published in NOWELE). --Florian Blaschke (talk) 15:36, 6 September 2013 (UTC)
 * The Sanskrit evidence doesn't really have much merit because it doesn't necessarily have anything to do with the Germanic situation; neither Proto-Germanic nor Sanskrit necessarily reflect the PIE origin. In fact, the PIE origin is normally reconstructed as *-ih₂ ~ *-yéh₂-, without any ending *-s. I don't know about the Old Norse evidence, -r is the regular ending for all long ō-stems in ON, which would include this noun. But it's well established that Western ON took some of the ō-stem endings from i-stems, so it's quite likely that this happened here too, and the real origin of the -r is the Germanic i-stem ending *-iz. So there really isn't any conclusive evidence for it at all. 15:47, 6 September 2013 (UTC)
 * Then, please, explain where the Sanskrit -s comes from; otherwise we need to treat it as an inherited feature, all the more because it is odd (unsystematic and unexpected), not despite of that fact (Meillet's principle, IIRC). Moreover, you've still not explained how you get from *wulgī to wulpa/wülpe, and why it is not listed. --Florian Blaschke (talk) 17:20, 6 September 2013 (UTC)
 * I don't know where the -s came from, but analogy is a likely origin. The nominative of most masculine nouns ended in -s, but it didn't for many feminine nouns, although quite a few did have -s (i-stems, u-stems). So it may have been added in certain languages to make the declension more regular. I can't think of any way that *wulgī would become wülpe. Through i-mutation and the High German consonant shift, it can be derived from Germanic *wulbī (ī/jō), *wulbijō (jō) or *wulbijǭ (jōn). Any further history must be Pre-Germanic. Possibly, it is actually the same stem as *wulgī, which may have split into two because of sound changes.
 * Originally, the stem was *wulkʷ- before Grimm's law and Verner's law took place. The masculine was then *wúlkʷos and the feminine *wúlkʷī (nominative) ~ *wulkʷiā́s (genitive).
 * An irregular change then happened to the masculine which changed *kʷ to *p (before Grimm) or *hʷ to *f (after Grimm), which resulted in *wulpos or later *wulfos, but that change may not be relevant here. The feminine form, after Grimm and Verner's laws, was *wulhʷī ~ *wulgʷijās by this stage, with consonant alternation in the root as is common in Germanic. There doesn't seem to be any trace of the stem *wulhʷ- in Germanic, and stem alternations as a result of ablaut or Verner's law are usually regularised away in Proto-Germanic nouns by generalising the genitive stem into the nominative. So for that stage we can assume *wulgʷī ~ *wulgʷijās.
 * This is where things become difficult though and I'm not sure if there is really a full understanding among linguists about this either. During this time, Germanic developed a kind of "conspiracy" to eliminate most traces of the phoneme *gʷ from the language. Word-initially, it usually shifted to *b, while it became *w in many word-internal positions, particularly before consonants. It also became *g in some instances. By the late Proto-Germanic period, *gʷ remained only after *n and had shifted to some other sound elsewhere. Exactly which conditions trigger which sound shift is still a bit unclear though. It's quite possible, though, that the difference between the two stems *wulgʷī and *wulgʷij- somehow affected these sound shifts so that one stem shifted to *wulb- while the other shifted to *wulg- and these two remained in alternation until one stem or the other "won out" but differently in the different descendants.
 * There's also another possibility, though. If we go back to the point that *kʷ became *p / *f in the masculine word, then you would have ended up with two nouns with rather different-looking stems: wulf- for the masculine and *wulgʷ- for the feminine. At this stage, Verner alternations between different noun forms must have still been regular and common (much as umlaut was in Old English), so speakers must surely have realised that the alternation *f ~ *gʷ was unusual and irregular. This could have been a reason to create a new stem for the feminine noun, with the "regular" voiced alternant of *f, giving *wulb-. At this point, the two stems *wulg- and the newer *wulb- could have existed side by side for a while as alternative forms or synonyms, until eventually one or the other "won" and the other disappeared. This would then be the stage seen in Old Norse and Old High German.
 * I hope that helps a bit? 17:52, 6 September 2013 (UTC)