At 4:22 pm -0700 19/9/00, Barry Garcia wrote:
>- between vowels and finally, d gets dropped; comprado > comprao,
>libertad > libertá. In the first example for this rule, the a and o
>combine to form a diphthong, /au/. The second, the accent over the final a
>keeps the stress in the last syllable.  (a book I have says that

Between vowels Latin [d] became [D] in Old French and then got dropped, e.g.
ui'dere --> *veDeir --> *v@Doir --> *v@oir, and thence mod. voir.

>A few more:
>- Medially, before i and e, c is pronounced as /ts/. Initially, it's /s/
>(but I have been thinking of /tS/).

Double treatment of Vulgar Latin /kj/ is odd.  In all the Romance langs
AFAIK its treated normallt the same whether initial or medial.  If it
wasn't someone surnamed Garcia writing, I'd suspect that the initial sound
was being treated differently to accomodate anglophones who generally find
initial [ts] difficult but who, of course, can pronounce initial [s] or
[tS] quite easily.  FWIW I would advise a re-think on this.

>- Initially, au remains, but in other positions it becomes /o/. Not sure
>how natural that is.

Not very, I think.  It does seem that /au/ remained until the early Romance
period.  But AFAIK all the Romance langs treat the diphthong the same way
in all positions.  The only precedent for different treatment of _any_
vowel in a Romance lang is if it occurs in a stressed syllable and, in some
Romance langs, the treatment of that vowel can vary depending ypon whether
the stressed syllable is open or blocked.

>- when in VL, there was an initial s + consonant, as in Spanish, an e is
>put in front of it, but, the s is dropped: strictu > estreço > etreço (c
>cedilla represents the /ts/ sound the way I am doing this)

It happened in French, but as part of a sound change whereby syllable final
/s/ became /h/ and was then dropped.  To restrict the dropping of -s- just
to st-  --> est-  --> et- would be very strange.

>- ct becomes /ts/, represented by c cedilla again: nocte > noçe, strictu >

Is a possible result, I think.  VL /kt/ turns up over in Romania as /pt/
and central & southern Italy as /tt/, thus Latin _nocte(m)_ becomes
_noapte_ and _notte_ respectively. Elsewhere the /k/ became palatalized and
we get /jt/, cf. French _nuit_, Catalan _nit_, Portuguese _noite_.  In
Portuguese the palatalization tends to affect the /t/ so that it sounds
more like IIRC /nOitS@/ (I'm sure Portuguese speakers here will correct
this).   In Spanish the shift from [jt] to [tS] has become complete, thus
_noche_.  A similar change whereby the sound became [ts] is not implausible.

>>From thinking about examples as I tweaked the rules a bit, it sounded to
>me in my mind as if there's elements of all the different romance-langs in
>it, sound wise.

It depends what you want, and only you can decide.  If you just want a "fun
language" made up indiscrimately of bits & pieces of different Romance
langs, then it doesn't matter a great deal what you do.  But tho the result
may be fun, do not expect it to look like (or sound like) a 'natural'
Romance lang.  If you want to derive a plausible 'alternate' Romancelang
then you have to use consistent 'rules' based on those actually attested.


A mind which thinks at its own expense
will always interfere with language.
                   [J.G. Hamann 1760]