Reconstruction talk:Proto-West Germanic/gant

Reconstructable?

I really don't think this can be reconstructed based on Old High German alone... —CodeCat 20:40, 27 December 2012 (UTC)

Well...I know what you're saying...; but if we really think about it, it isn't OHG alone--it's OHG and the PIE root supported by the Lithuanian and Old Greek cognates... (I'm trying hard to save it!!!) Leasnam (talk) 20:46, 27 December 2012 (UTC)

I thought that there had to be at least one attestable descendant (?). True, this is usually the early attested Gothic, but isn't OHG pretty conservative? Leasnam (talk) 20:48, 27 December 2012 (UTC)

There isn't a requirement as such, but the broader the attestation becomes the stronger the reconstruction is. If a form is attested only in German it doesn't mean much because there is no way to rule out borrowing. A form that is found in both Old Saxon and Old High German is still somewhat dubious as those two languages are closely related. A form that occurs in both Old English and Gothic on the other hand is pretty much a guaranteed reconstruction, because Old English and Gothic are so far apart linguistically. Think of it in terms of common ancestors... OHG by itself only has itself as common ancestor, OHG and Old Saxon have continental West Germanic as the ancestor. Add Old English and you have West Germanic as a whole, then add North Germanic and you have Northwest Germanic. But Gothic was the first to split off, so an attestation in Gothic and any other Northwest Germanic language is almost certainly reconstructable to Proto-Germanic since that is the common ancestor then. —CodeCat 20:54, 27 December 2012 (UTC)

To add to that... there is a way to guarantee a Proto-Germanic reconstruction even with one attestation only. I think you could call it a "meet in the middle" reconstruction. If a term is reconstructible for Indo-European as a whole, and the single Germanic attestation is definitely descended from it (because it has all the Germanic sound changes and no un-Germanic sound changes) then you know that it must have existed in Proto-Germanic too. (An example would be *hlefanan) I suppose there is another way too: if you can find a term that was formed through a derivational process that was no longer productive in Proto-Germanic, such as strong verbs or the nouns in -tis. —CodeCat 20:59, 27 December 2012 (UTC)

Plus, several reliable sources also reconstruct *gantaz for PGmc (so it can be assumed that they have ruled out borrowing (?)). TORP and Koebler both have it. Leasnam (talk) 21:04, 27 December 2012 (UTC)

The only possible attestation outside of OHG for this term is in Middle Dutch genten (“to heal”) which is rare but demonstrates umlaut and could therefore be a verb *gantijanan derived from the adjective. Gothic gansjan is sometimes cited but that can't be a cognate because of the -s-. I would be convinced if we could find cognates outside IE that have -d- instead of -t-, which would then demonstrate Grimm's law in this word and therefore that it is Proto-Germanic. —CodeCat 21:09, 27 December 2012 (UTC)

The -t- appears to be a suffix appended to the root of *gʷʰen- (>*gan-t-a-z). There are some Persian words with a -d- (e.g. agandan "to fill up"), but I am not familiar with Persian word formation, so cannot speak to where those stem from. Leasnam (talk) 21:25, 27 December 2012 (UTC)

I also find the lack of a labiovelar a bit suspicious. Germanic is a centum language so you'd expect *gʷʰ to appear as one of the outcomes of *gʷ in Germanic. *g is one of those, but only appears because of delabialisation next to *u; otherwise the regular reflex word-initially is *b (*bidjanan) or *w (*warmaz). The Dutch etymological dictionary suggests that this is a Baltic loanword. —CodeCat 21:31, 27 December 2012 (UTC)

Hrmmm, in such a case, the Middle Dutch cognate would still implicate a Pgmc term, irrespective of origin (whether directly inherited from PIE, or via a sibling)? Leasnam (talk) 21:38, 27 December 2012 (UTC)

Is that really true? It could have been borrowed from Baltic after Proto-Germanic times. The High Germans (Irminones) were the only West Germanic people that were in contact with the Balts, and as they spread south towards the Alps around 2000 years ago they must have taken the word with them. It could have filtered through to Dutch in that early period, but it is only attested once in all of Dutch so it must have been a very rare word that wasn't used in everyday language (much as thou might be nowadays). —CodeCat 21:46, 27 December 2012 (UTC)

I see your point. It's not looking good for *gantaz at the moment for sure...Leasnam (talk) 21:54, 27 December 2012 (UTC)

But we have to decide which was the case: 1). It was a rare word in PGmc, which only survived in OHG; or 2). that is was as you say above picked up by the forebears of today's German speakers and carried south. Are there many other sources which hold to the Baltic loan theory? Leasnam (talk) 21:56, 27 December 2012 (UTC)

The justification of the entry as such should be dealt with now that it has been converted into Proto-West Germanic. It's clear that the word is of West Germanic age, even if it was restricted to High German at the time. Generally, when a word is attested in only one branch or sub-language, it can still be reconstructed for Proto-Germanic if there is a reliable derivation from PIE. But that doesn't seem to be the case here, so I'm adapting the etymology. 178.4.151.244 05:40, 20 May 2023 (UTC)

PIE *gʷʰen- doesn't mean "to flourish, abound"

The proto-Indo-European root *gʷʰen- doesn't mean what it is proposed to mean (currently it is "to flourish, get full, abound"). It means "to chase, prosecute" or "to strike". Furthermore, in Germanic usually PIE *gʷʰ- > proto-Germ *b-. A more likely reconstruction in my opinion is *gʰen- extended with -t- which gave Proto-Slavic *gǫstъ (“dense”) < *gʰont-tis.

The proposed *gʷʰen- is a homonym of the above mentioned. Homonymy does exist in PIE. And yes, it has been brought up previously that the normal outcome of *gʷʰ- in Germanic is b-; and that this term is a borrowing from Slavic. Leasnam (talk) 17:48, 13 February 2018 (UTC)