Re: YACL: Thylean (alternate-history)

Thanks to all for positive comments,

As always, by the time one reads the replies to yesterday's posting, one
has already made lots of changes in one's mind. So none of the feedback was
much surprise to me, except perhaps Elliott Lash's Old Latin sample proving
how far back some of the changes date, which I once thought to be post-
Classical.

My sudden interest in Latin is due to the fact that I'm studying it at
school. I've studied it before, but that was in class, where nothing is
ever accomplished. Now I'm a, what's it called, a out-of-school-student,
i.e. I do my studies on my own. That way I can finally use the time I'd
waste in classes to actually learn stuff. So, I'm all crazy for Classical
Latin and wanted to make a conlang out of it that wouldn't just be Y A
Romance Lang. Which could best be accomplished by having the conlang
separate from the Latin community before the VL stage really takes off.
Such as Sardinian kind of did (that's a funny Romance lang - check it out
if you haven't). I wanted a completely different lexical and phonological
development, and preferrably a much more conservative development in the
grammar and morphology. In short, the Romance equivalent of Icelandic. So
which better place for it than Iceland/Thyle?

Before I read the replies to my post, I'd already consulted an excellent
French book on Latin's phonological history, qui s'appelle "La phonétique
historique du latin". There I had my suspicions on the final m confirmed.
So lots of things need to be rethought, though I was quite prepared to drop
the final m from the beginning. It strikes me how deceptive CL spelling
actually was, i.e. that lots of words weren't actually pronounced as they
were written. I don't know what to make of it, if I should start
pronouncing CL more correctly, which might make my fellow students and my
teacher balk, or just read literally anyway.

I particularly liked Cristophe's positive feedback - merci Cristophe! :)

On Mon, 6 Nov 2000 16:23:17 +0100, Christophe Grandsire
<christophe.grandsire@...> wrote:

>Having carefully read this post, yes I do! Welcome in the club of Romance
>conlangs inventors :)
It's an honour :)

>Interesting history. I'm just wondering how many people were part of the
>Republican faction and if Brutus could have really convinced them to leave
their
>homeland for an uncertain future... But if he managed to convince Cicero,
then
>it would be no problem to convince the others... :)
As I noted, the history's kind of far-fetched, though it would have been
possible, technically speaking. It's still more credible than Hollywood...
But really, it was just the only reasonably credible way I could think of,
which would get me a sufficient amount of CL speakers in a place far enough
from the rest of the Empire.

>> - Consonants
>>
>> /p t k/
>> /B D G/
>> /l m n N r/
>> /s S f/
>> /w j/
>>
>
>Interesting, no voiced stops, no voiceless /P T x/, /f/ but /B/. I'm
wondering
>about the stability of such a distribution. What do you all think?
You're right. I've rethought this:

/p t k/
/v D G/
/l m n N r/
/f T s S x/
/j/

As you noted below, /T/ and /x/ should also be phonemes, so they're
included. I decided that /B/ appeared only in transition, but then shifted
to pair with /f/. /w/ subsequently merged with /v/, as it has widely done
in Spanish AFAIK.

>> Allophones:
>>
>> /n/ is [N] before velars /k G/
>> /G/ is [j] before front vowels /i e/
>
>Interesting.
>
>> /k/ is [hj] (unvoiced semivowel) before /i e/
>
>This one too, quite unusual :) .
Kind of unusual, yes. Many Norwegian dialects and undoubtedly some Swedish
ones have the reflexes [j] and [°j] (correct SAMPA?) from palatalized
velars. I wonder why it's not more common, it seems like a logical
development.

Some corrections:

Changes CL > Th:

/b d g/ > /v D G/
/bb dd gg/ > /v D G/
/mn/ > /vn/



>What have become of /T/ and /x/? did they vanish during subsequent sound
>changes?
No, just an oversight :)

>> MORPHOLOGY
>>
>> The major change I have in mind is the merging of the genitive, dative,
>> and
>> ablative into one case (don't know what to call it).
>
>Why not oblique? or simply call it dative, it sounds okay to me.
I'd rather not call it oblique, as acc is not included in it. Dative is
fine yes. Perhaps 'relative'? (seems to describe the common theme of
gen/dat/abl together)

>Pretty much the evolution of VL to Old French.
Interesting...tell me more!

>Why not the common /ae/ -> /as/ replacement of the 1st decl. nom. pl.?
Wasn't it
>already appearing in Caesar's time in spoken Latin? Or maybe /ae/ -> /e/, I
>don't remember seeing a sound change of /ae/ in your phonology.
Actually, I didn't have any sound change for /ai/, except I merged /oi/
with it. But you're right, I don't like /ai/ as an ending. Nom. pl. in -as?
That might just be it...yes, it's all forming in my mind now... nom. pl. of
1st and 2nd declensions shall be -as and -us, the same as the accusatives.
No plural will make distinction between nom and acc...ok, read on, I'll
summarize below...

>> Thylean declensions have had their doze of analogy. The 2nd decl. nom.
>> sing. -os was rejected in favor of the vocative -e, because of the
>> former's
>> similarity to the established plural endings. This further led to acc.
>> sing. -om becoming -em, by analogy. 3rd decl. pl. dat/abl -ibus becomes
>> -
>> ios, not -ebos, through the elimination of the b, which was
>> unconventional
>> and outnumbered in the noun declension system.
Cut, cut, cut...

I kind of like the super-vocative idea, but only for animates. Seeing an
inanimate in vocative form just looks too silly to me. Final m is gone, so
let's review:

                             Sing.        Pl.
2nd decl (animate)

nom                         serve       servus    'male slave'
acc                         servo       servus
rel                         servu       servis

2nd decl (inanimate)

nom                         ladio       ladius    'sword'
acc                         ladio       ladius
rel                         ladiu       ladis

1st decl

nom                         serva       servas    'female slave'
acc                         serva       servas
rel                         serva       servis

3rd decl

nom                         neave       neavis    'ship'
acc                         neave       neavis    (/æ/ from CL /a:/
rel                         neavi       neavios    is spelled 'ea')

Obviously, the acc is by now only rarely distinctive from the nom, in fact
only in 2nd decl animates. Yet those aren't that rare at all, and I think
the distinction could reasonably thrive, as mostly an animate-inanimate
distinction.

I find that by far the most useful case in CL is the ablative, which is why
I like this system; through the distinctive 'relative' case, one can
maintain much of the intelligent CL use of the ablative, which I love.


>> 4th decl. merges with 2nd.
>> 5th decl. merges with 3rd.
>>
>
>The last one sounds unlikely to me. In Republic's times the nouns of 5th
decl.
>were already merging with the 1st declination (with doublets like
luxuries -
>luxuria, materies - materia), except dies, fides and res. Those last three
could
>well be put in the 3rd declination though.
True. 'Dies' and 'res' were the only 5th decl. nouns that came to my mind
at the moment, which explains me merging them with the 3rd.

>> CL iam 'already' > Th iam 'now'
Or rather,

Th ia 'now'
   iadro 'by now/already' (from CL iamdudum; dissimilation of latter 'd' +
                           syncope deletion of middle 'u')

>> CL nihil 'nothing' > Th nil 'not'
>>                         nilquam 'never'
>>                         nilque 'none, nothing'
Or rather,

Th nil 'not'
   nilqua 'never'
   nilqui 'no one'
   nirre 'nothing' (nil + rei (rei < CL res 'thing'); assimilation and
                    simplification of the unstressed diphthong)
                    (pronounced [niDre])

>> CL ipse 'him/her/itself' > Th epse 'the'
Rather,

Th esse 'the' (CL ipse was reportedly pronounced [isse], warranting my
               correction)
              (Sardinian also uses/used a reflex of ipse as an article)

>Neat semantic shifts :))) . For an "as for my conlang" digression, I also
have a
>very interesting semantic shift in "Roumant": The adverbs "mãg": very,
much,
>many and "mais": more ultimately derive from "magne" (adverb derived from
>"magnus": big) and "majus" (adverb derived from "major": bigger). The
>resemblance with Latin "magis" (more, gave "mais": but in French
and "más": more
>and "mas": but in Spanish) only helped this shift, especially as the
original
>meaning of "magnus" was taken by "grandus".
A nice addition to my linguistic knowledge, on the origin of Fr "mais", and
Sp "más", quelque chose que je savais pas.

>I think it would be wiser to drop it from the declensions, unless some
kind of
>hypercorrection (maybe due to some partisan of purity of language like
Cicero)
>would put it back in use. That's unlikely, but not impossible.
I intented to make this a puritan kind of Latin, but I'm kind of digressing
from that, with really simple declensions and all...