Talk:Proto-Northern-Romance (MGR)

Pronouns

Personal, reflexive and impersonal pronouns

Many of these pronouns have a short or unstressed and a long or stressed form. Where this is the case the short or unstressed form is shown to the left and the long or stressed to the right with a slash between them. It should be noted that some of these stressed or long forms go back to Latin unstressed forms which became stressed once again in certain contexts. These re-stressed forms with lengthened vowels could then lose stress again, resulting in doublets with a long vowel and a diphthong respectively, and thus three different forms corresponding to three degrees of stress or emphasis. In the further development of the separate languages some or other of these forms usually fell out of use, the cycle of loss and acquisition of stress going on through the centuries. The impersonal pronoun hom is in origin an unstressed form of the noun hoam 'man' < homō, illustrating this process at its extreme.

Personal, reflexive and impersonal pronouns
		Singular		Plural
		m.	f.	m.	f.
1st person	Nom.	jō		nus
	Acc.	mi/mē/mei
	Gen.	mīs	ma	nōster	nōstra
	Dat.	mī		nous
2d person	Nom.	tū		wus
	Acc.	ti/tē/tei
	Gen.	tus	ta	wōster	wōstra
	Dat.	tī		wous		Refl.	Impers.
3d person	Nom.	el	ella	lī/ellī	lē/lei/ellas		hom
	Acc.	lu/lō/lou	la/lā	los/els	las/ellas	si/sē/sei	hom(m)
	Gen.	sus	sa	lōr/lour/ellōr			hommis
	Dat.	lui	lei	līs/ellīs		sī	hommī

---

Revision I (PMC)

The columns have been reaaranged, primarily to reduce the amount of empty space, but also because this layout seems a little more logical to me - paricularly with regard to the reflexive pronouns.

[see main page for revised table]

I've made the following changes to the pronouns:

Unstressed mid vowels

• Hommis > Hommes:
  • Unstressed original /i/ shifts to /E/, not /I/.

According to my understanding all of Latin unstressed ĭ,ē,ĕ shift to a single mid vowel /e/ which gets identified as */ɪ/ in Germano-Romance and then later is lost together with unstressed */ʊ/ similarly from ŭ, ō, ŏ — i.e. corresponding to two mid heights front and back among stressed vowels there is only one vowel each front and back of indeterminate mid height which in PNR gets identified as high lax. Probably they tended towards [ɪ̵], [ʊ̶] as in the OTL precursor to Old French. Of course how you write that indeterminate mid vowel i/e or u/o is of course a matter of taste. BPJ 11:39, 24 July 2008 (UTC)

I think what actually happened in OF was that unstressed /e/, /o/ tended toward [ə] and were eventually lost for that reason, before a shifted to [ə], and was eventually lost centuries later. The whole idea with an [ɪ] stage in PNR was that it might cause umlaut, but I think that's untenable anyway.BPJ 12:17, 24 July 2008 (UTC)

• Hom(m) > Homme:

Uh, yes, we have to determine which stage of PNR — before of after the loss of unstressed /e/ɪ/ and /o/ʊ/ this page is going to represent.

BPJ 11:39, 24 July 2008 (UTC)

• • Unstressed final syllable /E/ is only lost after an obstruent - loss here is therefore blocked by the /m/.

Where did you find such a restriction on the loss of unstressed /e/? After all OF has panem > pain!

BPJ 11:39, 24 July 2008 (UTC)

• el > elle
  • No loss of final /E/, as above.

Same thing: VL salem > OF sel.

Perhaps it is the presence of a geminate which blocks the loss in homme, but I think that there are examples of loss even after geminate sonorants (il is a case in point), and homme is simply an oddball word. BPJ 12:17, 24 July 2008 (UTC)

Ewert on atonic vowels

Ewert's treatment is terse, to the point and clarifying:

Fine distinctions of quality are obliterated, with the result that ɛ and e, ɔ and o falI together. Further, the V.L. quantitative distinction (if it ever existed in atonic vowels, cf. § 21) was not maintained. GalIo-Roman therefore inherited from V.L. the atonic vowels a, e, i, o, u, which were presumably short and were already tending to weaken and disappear.

(ii) FINAL AND COUNTERFINAL VOWELS

29. These persist in V.L. in a weakened form with a tendency to confuse e and i, o and u. In Old French, a remains in a weakened form as so-calIed feminine e ( = ə) (cf. § 61) : BONA> bonne, AMAS> aimes. This change dates from about the end of the eighth century. e, i, o, u generally disappear (about the seventh century), but they persist in the form of the weakened supporting vowel ə in the following cases: (a) before a group of consonants (AMENT>aiment); (b) after a group of consonants requiring a supporting vowel, notably cons. + I, r, m, n, excepting kl, gr, gn, rm, rn (DUPLUM > double, PATREM > peðre >pere, *HELMU (Germ. helm) > helme > heaume, ALNUM> alne> aune). The group may be primary, i.e. inherited as such from Latin, or secondary, i.e. developed subsequently through the loss of a vowel (MASCULUM > MASC'LU > masle> male). In the absence of any supporting vowel an ə is developed (INSIMUL> ENSEM'L > ensemble, MINOR> MEN'R > mendre later moindre ≠ moins). It will be seen that ə persists even after the reduction of the group which originalIy required the supporting vowel (pere, heaume, aune, male). For the apparent exceptions presented by borrowed words, cf. § 500.

I take this to mean:

C.L.	V.L.?	Stage 1	Stage 2	Stage 3	OF
ī	i	i	i	i	i
ĭ	ɪ	e	e	ə	—
ē	e
ĕ	ɛ	ɛ
ū	u	u	u	y	y
ŭ	ʊ	o	o	ə	—
ō	o
ŏ	ɔ	ɔ
ā, ă	a	a	a	a	ə

BPJ 21:13, 24 July 2008 (UTC)

hominem must have been homne at the time of the fall of final /e/, /o/ to retain its /ə/.

3rd person case forms

• 3rd person accusatives:
  • Grandgent posits *sě for the accusative singular and plural in VL (§385).
  • Elle etc is being used for the definite article - would this preculde its simultaneous use as a pronoun?
• 3rd person datives:
  • Grandgent posits *sī (< sibi, and by analogy with mī & tī) for the dative singular and plural in VL (§385).
  • Elle etc is being used for the definite article - would this preculde its simultaneous use as a pronoun?
• Reflexive pronouns:
  • These also need 1st and 2nd person forms.
  • My understanding is, that although CL had forms for all the cases except nomnitive, only accusative forms survived?