[YG Conlang Archives] > [romconlang group] > messages [Date Index] [Thread Index] >
Capsicum skrev:
In Rhodrese '-um' is different too. I thought it is because of the 'm'. Is it just because it is at the end of a word. I know Alemannic dialects were 's' can have a similar influece as 'n'. So it could affect '-um' and '-us'. I did not read the whole post. Don't mind if I write stupid things.
Well, in Vulgar Latin as a whole final -m,
except in stressed monosyllables was lost;
in fact the letter represented, by the evidence
of Old Latin, where it is written or not
written a bit randomly, merely a nasalization of
the final vowel, so what was lost was vowel
nasalization. The jury is still out (and will
be forever, as we really can't know unless some
new evidence emerges) regarding whether -um
became -u or -o, and thus also whether the
accusative in the second declension merged
with the ablative/dative phonetically and not
just morphologically. The confusion in the
spelling of these endings
in Late/Vulgar Latin texts may be due to the general
confusion of the letters u/o brought about by the
merger of short /u/ with long /o:/ or to an actual
merger of the two unstressed vowels in this case.
In fact written -VM may have represented spoken
/o~/, being written with V on analogy with -VS.
For Rhodrese I actually assume that -VM represented
-/u~/, which became -/u/ and remained distinct from
-/o/, so th�t later -/u/ could change into something
else ([@]), whereafter -/o/ became -/u/. All this
is to satisfy my linguistic taste: I want first person
present singulars to show 'u-umlaut' but the singular
of nouns and adjectives not to show it.
However I'm ATM reevaluating this: for one thing it mightn't
be a bad thing if AMATUS becomes _amuad_ and so has a
different stressed vowel from _amiade_ < AMATA. Secondly
it might also not be a bad thing if -IS and -US actually
merge with -AS, -ES, -OS and they all end up with an
*-@s/-@z/-@j/-@ allomorphy. The end result may be much the same
anyway. Consider the case and number forms of AMATUS in
Vulgar Latin:
AMATUS AMATI AMATA AMATAS
AMATU AMATOS AMATA AMATAS
and the present singular indicative and imperative of AMO and TACEO
AMO AMAS AMAT AMA
TACEO TACES TACET TACE
In a first stage there is one simple shift:
final unstressed /o/, but not /os/, merges
with /u/ -- probably the vowels in closed
final syllables, other than /a/, were laxer than absolute
final vowels. (I have tacitly[!] implemented
some consonant changes):
ama:dus ama:di ama:da ama:das
ama:du ama:dos ama:da ama:das
a:mu a:mas a:mat a:ma
ta:tSu ta:dzes ta:dzet ta:dze
Note that VL length is different from CL length:
in VL all vowels in open stressed syllables are
long, all other vowels are short. However some
unstressed penultimate vowels get lost early, so
the now length becomes phonemic.
Next /a:/ fronts to [&:] (the vowel of
American English _cat_, but long)
*except* where followed by /u/ -- there
it becomes [Q:] (the _law_ vowel) instead:
am&:dus am&:di am&:da am&:das
amQ:du am&:dus am&:da am&:das
Q:mu &:mas &:mat &:ma
tQ:tSu t&:dzes t&:dzet t&:dze
For the time being this is merely an allophonic
change: the phoneme /a:/ has two variants depending
on context.
Next all vowels in final syllables except /a/ merge as
[@] -- quite possibly /i/ and /u/ merge first as [I\]
which then merges with [@]. Notice how this makes /&:/
and /Q:/ distinct phonemes, since /am&:d@/ and /amQ:d@/
are otherwise identical!
amQ:d@s am&:d@ am&:da am&:das
amQ:d@ am&:d@s am&:da am&:das
Q:m@ &:mas &:mat &:ma
tQ:tS@ t&:dz@s t&:dz@t t&:dz@
All along -/s/ has had two allophones:
[z] before a voiced sound, including across a
word boundary without a pause and [s] elsewhere.
Now somewhen parallel with the above vowel changes
two more allophones of /s/ emerge: [j] before a
voiced consonant and [h] before a pause. Of course
this [j] ceases to belong to the /s/ phoneme, since
there is already a /j/ phoneme! So in fact we have
the following forms:
amQ:d@s am&:d@ am&:da am&:das
amQ:d@z am&:daz
amQ:d@j am&:daj
amQ:d@h am&:dah
amQ:d@ am&:d@s am&:da am&:das
am&:d@z am&:daz
am&:d@j am&:daj
am&:d@h am&:dah
Q:m@ &:mas &:mat &:ma
&:maz
&:maj
&:mah
tQ:tS@ t&:dz@s t&:dz@t t&:dz@
t&:dz@z
t&:dz@j
t&:dz@h
Next there happens one assimilatory change:
-@j and -aj both become -ej -- probably only
a change of allophone as far as -@j is concerned,
since [@] can well be seen as an allophone of /e/.
Also the [h] allophone of /s/ simply disappears:
amQ:d@s am&:d@ am&:da am&:das
amQ:d@z am&:daz
amQ:dej am&:dej
amQ:d@ am&:da
amQ:d@ am&:d@s am&:da am&:das
am&:d@z am&:daz
am&:dej am&:dej
am&:d@ am&:da
Q:m@ &:mas &:mat &:ma
&:maz
&:mej
&:ma
tQ:tS@ t&:dz@s t&:dz@t t&:dz@
t&:dz@z
t&:dzej
t&:dz@
Reduction of final unstressed syllables goes on:
-[@] disappears and -[ej] becomes -/i/. Also final
stops and affricates are devoiced, perhaps except before
voiced sounds within a phrase:
amQ:ts am&:t am&:da am&:das
amQ:di am&:di
amQ:t am&:da
amQ:t am&:ts am&:da am&:das
am&:di am&:di
am&:t am&:da
Q:m &:mas &:mat &:ma
&:mi
&:ma
tQ:tS t&:ts t&:tst t&:ts
t&:dzi
Probably at this time the highly distinctive
plurals and second persons in -/i/ began to
replace the indistictive plurals and second
persons in zero and -/a/, though the zero m.n.pl.
and 2p.sg. and f.pl. in s/z lingered on, the latter
as alternatives, since third declension feminines
all had -/i/ plurals but no -as/-az plurals:
amQ:ts am&:t am&:da am&:di
am&:das
amQ:t am&:di am&:da am&:di
am&:das
Q:m &:mi &:mat &:ma
&:mas
tQ:tS t&:dzi t&:tst t&:ts
t&:ts
Somewhere along the road all long vowels except
/i:/ and /u:/ have diphthongized, so what we have
now is (admittedly the transcriptions become
somewhat hard to read!):
amQ@ts am&@t am&@da am&@di
am&@das
amQ@t am&@di am&@da am&@di
am&@das
Q@m &@mi &@mat &@ma
&@mas
tQ@tS t&@dzi t&@tst t&@ts
t&@ts
The next change is possibly the most important in
the whole history of Rhodrese: i-umlaut
(metaphony), which in the case of front vowels
like [&] means that they become [i], and
then -/i/ disappears. Probably around the same
time -/a/C* is reduced to [@]. Final devoicing was
by now a rule of the language, so the forms which
had lost -/i/ were devoiced:
amQ@ts @m&@t @m&@d@ ami@t
@m&@d@s
amQ@t ami@t @m&@d@ ami@t
@m&@d@s
Q@m i@m &@m@t &@m@
&@m@s
tQ@tS ti@ts t&@tst t&@ts
t&@ts
These are in fact the forms of Old Rhodrese, which
look as follows in 'orthography':
amoatz ameat ameade amiet
ameades
amoat amiet ameade amiet
ameades
oam hiem eamet eame
eames
toatz tietz teatzt teatz
teatz
/BP 8^)>
--
Benct Philip Jonsson -- melroch atte melroch dotte se
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"C'est en vain que nos Josu�s litt�raires crient
� la langue de s'arr�ter; les langues ni le soleil
ne s'arr�tent plus. Le jour o� elles se *fixent*,
c'est qu'elles meurent." (Victor Hugo)