User:Melroch/Rhodrese

This as yet unnamed Romlang initially has the developments you would expect from a Western Romance language. However at a later time there is vowel mutation by a high or low vowel (a < ă/ā, i < ī, u < ū), with final vowels later being lost, or in the case of -a being reduced to -e [ə]. Somewhat unexpectedly plurals of the first declension show forms as if the a-mutated vowels of the singular had later undergone i-mutation. This most certainly is due to analogy with the second and third declension rather than to a preserved [ai̥] ending — or did an ending pronounced [æ] cause such a 'double' mutation?

Vowels

Stressed vowels

Latin	VL	W. Rmc.	No mut.	a-mut.	i-mut	a/i-mut.	u-mut.
ī	ị	ĭ	i	e	n/a	e	ui > eu /y/
		ī		ei > ai		ei > ai
ĭ, ē, oe	ẹ	ẹ̆	e	n/a	i	n/a	oe /ø/ > e
		ẹ̄	ei > ai
ĕ, ae	ę	ę̆	e				oe > e
		ę̄	ie				ue /yø/ > /ye/
ă, ā	a	ă	a		ae /æ/ > e[6]		o
		ā	ea > ia		ie	e	oa > ua[7]
ŏ	ǫ	ǫ̆	o		oe > e		n/a
		ǭ	uo		ue	oe > e
ŭ, ō	ọ	ọ̆	o		oe > e
		ọ̄	ou > au		eu[5]	ue
ū	ụ	ŭ	u	o	ui > eu	oe > e
		ū		ou > au	eu	ei > ai

Notes

1. ^  Vowel length in Western Romance

   After the Classical Latin vowel length distinction was lost in favor of a distinction of vowel quality there developed a new length distinction in parts of the Vulgar Latin area. In some parts, notably Iberia, all stressed vowels were lengthened, while in most parts, including Italy and Gaul, it was stressed vowels in open syllables that were lengthened; Rhodrese belongs to this latter group. The difference is best seen in the diphthongization of ę and ǫ (from Latin ĕ and ŏ: in Castilian tęrra, pęde became tierra and pié because the first e was stressed in both words, but in Italian they became terra and piede because the stressed syllable was closed in tęrra but open in . In Rhodrese these words become terre and pié (plurals tir and pí).

2. ^  A-mutation

   A-mutation lowered the high vowels i‍̣ and ụ to ẹ and ọ. A-mutation was prior to the secondary diphthongization of ẹ̄ to ei and ọ̄ to ou, so the a-mutated ī‍̣ and ụ̄ diphthongize along with primary ẹ̄ and ọ̄.

   Although a-mutation was certainly later than the diphthongization of ę̄ to ie and ǭ to uo the high first elements of these diphthongs are not subject to a-mutation, whether because of the intervening mid vowel element, or because the diphthongs had already became [jɛ] and [wɔ].

3. ^  A/i-mutation
4. ^  U-mutation
5. ^vowels-ui-eu  The vowels ui and eu
6. ^  The vowels oe and ae
7. ^  Breaking of Western Romance *a
8. ^  I-mutation of ea > ia
9. ^  The diphthong ou < Western Romance ō‍̣ and Latin au

Consonants

Consonant correspondences

VL	Initial	Geminate	Final		Intervocalic
			Primary	Secondary	Pretonic	Posttonic
p	p		n/a	p	b
t	t		t	t	d
ć [c]	c [ts]	x [tʃ] > tx^	n/a	z > ç [ts]	z [dz]
c	c, ch [k]		Ø	c	g, gh [g]
b	b		n/a	f	v
d	d		Ø		d	Ø
ǵ [ɟ]'	g [dʒ]		n/a	i, y [j]i, y	g(i) [dʒ] (> [ʒ]?	i, y
j	i > j [dʒ]	n/a
g	g, gh g		Ø	g		Ø
pj	(pj pi)		n/a	uj, bi >be
tj	c(i) [ts]	c(i)		z > ç	c(i)
kj		x
bj	(bj bi)			uj, vi >ve
dj	z [dz]			z > ç [ts]'	g(i) (> j)
gj	g(i)			x > tx
m	m		Ø		m
n	n		Ø		n
mb	n/a		n/a	m	mb
nd				n	nd
ng				ng [ŋ]	ng [ŋg]
mj	gn(i) [ɲ]			(i)gn [ɲ]	gn(i)
nj, gn, cĭn > ć’n
lj, gl, gĭl, gĕl, dĭl, dĕl	gl(i) [ʎ]	*ggl > gl, ghl(i)		(i)gl [ʎ], gl(i)
cĭl, cĕl, cĭl, cĕl		*ccĺ > cl, chl(i)
cŭl, cŏl, căl > c’l	cl, chl(i) [kl]			gu > go	gl, ghl(i), (i)ghl
tŭl, tŏl, tăl > t’l
gŭl, gŏl, găl > g’l	gl, ghl(i) [gl]	*ggl > gl, ghl(i)		l		l
dŭl, dŏl, dăl > d’l
s	s [s]	ss [s]	Ø	s	s [z]
sj	sc(i) [ʃ]		n/a	-is, -sz > sç (all [ʃ])	si [ʒ] (> j ?)
sc´, scj	sc(i)	n/a
ct	n/a			x > tx
gĭd					g(i) > j
cĭt, cĕt
tĭc, tĕc
pt		t		t
pĭt, bĭt, vĭt		ut		ut	ud
pĭd, bĭd, vĭd		uj		uj	ud	uj

Random notes

Intervocalic b > v; g' > j i; p, t, k', k > b, d, g', g but tj, kj > voiceless ts, gj > [dʒ] gi. Unpalatalized original g tends to disappear.

Initial k', kj, tj > ts c(i); initial g', gj > [dʒ] (g)i.

dj in all positions > [dz] z.

ct > [tʃ], while actual x = ks > s or undergoes metathesis to sk, which may be palatalized to [ʃ] 'sc(i) along with original sc.

Voiced stops and affricates ending up final after the loss of final vowels become devoiced, spelled p, t, z [ts] (later ç), (i)x [tʃ], c.

Final [ʃ] is spelled V(i)ss/Vis, later sç, with occasionally sz/sx in early texts.

Word initial ie, ue, uo are spelled hie, hue, huo, since the normal medieval pronunciation of initial prevocalic i, u was j, v.

The vowels [æ] ae and [ø] oe were found in the earliest texts, but later merged with e, while ui' [y] and eu [øy] merged as /y/ spelled eu! The spelling ue remained, but probably shifted from [yø] to [ø].

A k, g, sk before a front vowel that arose through mutation wasn't palatalized. They are spelled ch, gh, sch.

^  Due to the sc/x merger it is likely that x was primarily a spelling for /ʃ/ all along, while /tʃ/ was normally spelled tx, or in early texts alternating with cx. It is however likely that Classicism created a vougue for using x for /ks/ — and against the 'barbaric' cs (or rather, at the time, cſ) — in learnëd words, this making sc^i,e /sç the normal spelling for /tʃ/ in all cases — even though sçt for /ʃt/ < scĭt/xĭt will look a tad bizarre! It is also likely that the 16^th

grammarians disagreed on these points.  I can envisage this 'pattern':

Rhodrese sibilant spellings

	Medieval	Pirrí	Grieur	Modern
/dz/ > /z/	z, dz	z	z	z
/ts/ > /s/	ci,e , cea,o,u /cia,o,u , tz, cz, zz	ci,e , tz, (cia,o,u )	ci,e , ç	ci,e , ç
/ʃ/	sci,e , scea,o,u /scia,o,u , x, sz, sx, ss, iss, xx	sci,e , x || sci,e , sç, (ixt /ʃt/)	sci,e , sç
/tʃ/	tx, cx, cci,e , ccea,o,u /ccia,o,u , cc	tx	tx	tx
/dʒ/	gi,e , iV , gea,o,u /gia,o,u , gg	gi,e , gia,o,u , gg	gi,e , iV	gi,e , j
/j/	i, y	i	y	y
/ks/	cſ, cx	cſ	x	x

Pirrí and Grieur may be expected to disagree along their usual pattern, with Pirrí being a rationalist and populist and Grieur a classicist/Latinist.

• In Medieval spelling prevocalic and final i was ambiguous between /dʒ/ and /j/. Pirrí's solution was to never use plain i for /dʒ/, while Grieur's was to use i only for /dʒ/ and always y for /j/.

Modern spelling is fool-proof in never using i in these positions (except the recognized diphthongs and triphthongs ie, ia, ieu, iau), but j/g^i,e for /dʒ/ and y for /j/.

^  Medieval scribes tried to insert hs to diambiguate, but that could get very messy.

• The 16th century grammarian Pirrí didn't like the letter ç -- in fact he didn't mention it at all, but gladly wrote /tsia/ and /tsie/ as ciia, ciie as well as giia, giie, sciia sciie, while e.g. cia spelled /tsa/. Word finally he showed devoicing in spelling so he needed only s, tz, tx, x as final sibilants and no non-prevocalic /dʒ/. His fellow grammarian Grieur lambasted the ii and tz spellings, as well as x for /ʃ/ as barbarisms. What is really puzzling is that Grieur didn't criticize or question the spelling tx for /tʃ/. Did he feel he could not come up with an improvement, or did he in fact propose something like ţ for /tʃ/ and ş for /ʃ/? — although taking this to its logical conclusion we'd get çţ and çç for /tʃ/, and likely some instances of ģ! Of course we must not expect either gentleman to have been scientific or consistent!

• z does not occur before /i/ in normally developed native Rhodrese words. All occurrences are due to borrowing, analogy or ahistorical spelling. The most common item is the name Clauzine altered from Claudine under the influence of Clauze < Claudia

• Augustu > Aost Augustinu > Aoesti [aøst'i] > Aesti > Aiesti.

• molliatu > molleat > molliad. Pirrí: "La C molliade sempre se prononciet TS et la Z sempre se prononciet DS."

• In modern spelling of Rhodrese stress falls unless otherwise indicated on the vowel before the last consonant except in words with final Vt which is stressed on the penultimate. Thus most 3d person singular verb forms need not be marked with an accent. OTOH a few words like amít 1st person singular of amiter 'admit need to be marked. What about final -n < -nt? What about 2.sg. -s? Might it be restored by analogy with perfect -isti > -is, or is that unrealistic? NB -isti would be -st at the time -s was lost.

• Which is the correct Vulgar Latin form of molliare, mollare, mollicare? Participle will be molliad/moliad/molghiad.

• How common will /ɡn/, /ɡl/ < gVn, gVl be? What happened to (initial) gl in Old French? Will the spellings gn, gl for the palatal nasal and liquid and ghn, ghl for the clusters be OK? Since cin, gin, cil, gil will become palatal sonorants they will be quite common. I definitely don't like word initial nn and not really ñ or initial ll either. The spellings gl, ghn, ghl would give Rhodrese its own character.

• The superordinate goal of Grieur's orthography was to preserve the spelling of Latin / learned words, and he accepted quite a high level of ambiguity in order to make R3 spelling less 'barbaric'. For example he preserved the spellings ei, ou despite himself admitting that they were pronounced ai, au, because they were closer to the Latin spellings with e/i, o/u, and at the same time he wrote au in words descended from Latin words with that spelling. As for /ɲ/, /ʎ/ vs. /ɡn/, /ɡl/ he favored ign/gni, igl/gli as the spelling for the palatals and stated that the limited set of words having /ɡni/, /ɡli/ etc. could be learned by rote.