[YG Conlang Archives] > [romconlang group] > messages [Date Index] [Thread Index] >
At 17:05 11/02/2004, Muke Tever wrote:
On Wed, 11 Feb 2004 15:54:26 +0000, Carl Edlund Anderson <cea@hidden.email> wrote: > My initial thought is that the "proper" nominative form ought > to be Ios, on the model of bos, bovis, with a derivation something like > PIE *dyeu-s > diow-s > dios > ios. Ab origine, *_diu:s_ (cf. nudiustertius) or possibly *iu:s. u: is the standard reflex of IE *ew. Technically, _die:s_ is the survivor (remodelled from the accusative). The paradigm of *dyews split into two words in Latin, similarly to *deywo- > deus and divus...
I hauled out my quick'n'dirty guide to PIE>Latin sound changes (Palmer's _Latin Language_ in this case) to try to get a closer look at what's going on. I've summarized my findings and thoughts below.
Firstly, for reasons I do not understand, Palmer indicates the stem as /*dye:u-/ (with a long diphthong) when talking about the nominative and accusative forms, but /*dyew-/ when talking about the genitive. I'm not sure why; I had thought the stem was a simple short diphthong: /eu/. Was there a different length in the nom. and acc.? It looks like /*gwou-/ also had a different diphthong length in the nominative .....
In fact, /*gwou-/ seems to have gotten a very similar treatment to /*dyeu-/ overall. Palmer gives the nominative form as /gwo:us/ and says the diphthong's second vowel was lost in the accusative to produce /*gwo:m/. A new nominative, he says, was formed on this vowel, with /*gwo:s/ giving a dialect form /bo:s/ (with /gw-/ > /b-/ instead of /v-/), thence borrowed into standard Latin. However, Palmer gives the original genitive form as /*gwowes/ and the dative as /*gwowei/, identifying these as the ancestors of /bovis/ and /bovi/.
Similarly, Palmer gives the nominative form /*dye:us/ and says the diphthong's second vowel was lost in the accusative to produce /dye:-m/, then creating a remodelled paradigm as Muke indicates with nom. dies, acc. diem, etc. Meanwhile, Palmer give the example of the genitive form /*dyewes/ going on to produce /iovis/ (apparently although PIE /ew/ generally goes to /u:/ in Latin, the presence of /w/ near /e/ can cause /e/ to become /o/ as for example, in the case of /*newes/ > /novus/, /*newem/ > /novem/). So /Iov-/ from /*dyewes/ must have them formed the model for the new paradigm.
But, I suppose had the original nominative and accusative been preserved, the paradigm should have gone along the lines of:
nom: dye:u-s > dyu:s > iu:s gen: dyew-es > dyowis > iovis dat: dyew-i > dyu:i > iu:i (the proper original dative ending /-ei/ being, apparently, replaced by /-i/) acc: dye:u-m > dyu:m > iu:m abl: dyew-ei > dyowe > iove (I'm guessing about whether the /e/ in /-ei/ would actually cause /eu/ > /ou/)And a paradigm that messy would be enough to make anyone pick a new stem :) Another problem perhaps was the nominative /i:us/ would have been too much like /*yewes-/ > /i:us-/ ("law")? I guess the dative and accusative were rebuilt on /iov-/, though /Iupiter/ must have been indeed fossilized from a form like /iu:s/ ... from original /*dye:us/? Alternatively, perhaps /Iov-/ was used in the nominative but a /Iovs/ nominative form would be inadmissable, and the /ov/ was collapsed to /u:/?
Also I suppose the example of /na:vis/, originally a diphthong stem /*na:u-/ recast as an i-stem, suggests extending the paradigm remodelling to include a nominative /*Iovis/ would actually be a reasonable reconstruction option. I almost certainly want to keep the separation out die:- and diov- from original dyeu- or I'll have to invent a new word for "day" as well ;)
Further thoughts and comments? Cheers, Carl -- Carl Edlund Andersonhttp://www.carlaz.com/