Re: Latin loans in Welsh

At 12:25 pm +0000 9/6/01, kam@CARROT.CLARA.NET wrote:
[snip]
>
>Just to summarise, mainly in response to Ray's post, here's what happened
>to the Latin vowels in Welsh (secondary developments in brackets) ignoring
>various special cases :
>
>Short vowels :
>
>i >> I, (e)
>e >> e, (I)
>a >> a, (ei, ai, e)
>o >> o, (ei, ai, y)
>u >> u <w>, (o, I)
>
>(In W. /I/ is a centralised [I] or [@], it merged with /e/ in Breton)
I take it [I] is SAMPA's [1], i.e. modern, north-Walian _u_ in _punt_.

>Long vowels
>
>i: >> i
>e: (borrowed as [ej]) >> u:j <wy> (Breton <oua, oue> etc.
>a: >> aw, o     (Breton eu,e)
>o: (borrowed as [ow] >> [oy] >> [y] <u>
>u: as /o:/ above.
>
>(In Modern W. /y/ <u> has become unrounded and centralised and has
>merged with /I/ but is still distinguished in writing)
>
>Rarely Latin /o:/ gave W. /aw/
>
>All these developments are the same as for the native sounds.
Quite - which is what I would expect with Latin (or any other language)
words which become assimilated in ancient British; the same happened to
Norman French words in English - once they become part of the language,
they are then subject to all subsequent developments just like native
words.  All it tells us is the probable pronunciation of Vulgar Latin in
Britain at the time when these words were taken into ancient British.

But Brithenig is derived from the spoken Vulgar Latin of the already
Latinized urban areas which, unlike *here*, continued to function *there*
in the parts of Britain that Saxons did not occupy *there*.


[snip]
>
>The point being that WCB all replace the British/Latin quantity
>distinctions by quality distinctions, but I think, in rather a different
>way from Vulgar Latin.
But there was not one consistent system in Vulgar Latin either.  Western
Romance shows, basically (note phonemes, not phones):
  Latin   /i:/  /i/  /e:/  /e/  /a:/  /a/  /o/  /o:/  /u/  /u:/

Early VL  /i/   /I/  /e/   /E/     /a/     /O/   /o/  /U/  /u/
                 \    /                           \    /
                  \  /                             \  /
Late VL   /i/     /e/      /E/     /a/     /O/      /o/    /u/

But this wasn't universal - both Sardinian & Dalmatian Romance show
different treatments.  In any case, maintained itself like this only in
stressed syllables; unstressed show further simplifications. Indeed, to
deal with the development of VL vowels in Old northen French, we have to
consider at least all 12 the different cases below:
                             STRESSED
                        BLOCKED    UNBLOCKED      UNSTRESSED
(a) Independent            1a         2a              3a
(b) after a palatal        1b         2b              3b
(c) before /j/             1c         2c              3c
(d) before a nasal         1d         2d              3d

The Brittonic borrowings obviously occurred early on, but Brithenig must
have developed from the latest stage of VL spoken in Britain, and if there
continued to be trade and interchange with Gaul & the rest of the Empire -
as long as it lasted - the language would've been affected by later
changes.

Short of time travel, I don't think we can be sure how Latin was pronounced
on the streets of Londinium, Venta Belgarum or Venta Silurum on the eve of
the Saxon incursions.  We just have make intelligent guesses and accept
certain assumptions if we are going to compile a fantasy language like
Brithenig.

One telling point, it seems to me, is that only in north Gaul do we find
ublocked, stressed /e/ and /o/ diphthongizing, i.e.
/e/ >> /ei/ >> /Ei/ >> /Oi/ (e.g. habere >> aveir >> avoir)
/o/ >> /ou/ >> /Ou/ >> /9u/ (e.g. nepote(m) >> neveu)

This seems to me remarkably similar to:
>e: (borrowed as [ej]) >> u:j <wy> (Breton <oua, oue> etc.
.....
>o: (borrowed as [ow] >> [oy] >> [y] <u>
So. also, is the shift of Latin /u/ to [y], which has often been cited as
an example of Gallic (i.e. Gaulish) influence.

..and the treatment of -ct- is similar, cf.
Vulgar Latin: lacte  >>
Old French: lait /laIT/ ( >> mod. French /le/)
     Welsh: llaeth
   Cornish: leth /le:T/ (Unified)
    Breton: lêz _or_ laez


>The reflexes of long and short vowels on the
>whole stay separate, only W. /o/ could come from either /a:/ or /o/.
As we might expect if the borrowings were early.  The only odd thing would
be if it is clearly shown that Classical Latin /a/ and /a:/ distinction was
maintained.  In the examples shown so far, as far as I noticed, the change
of /a(:)/ >> /aw/ occurs when /a/ would've received stress in Vulgar Latin,
which is what I'd expect.

But, as I've said above, Brithenig is not derived from the Latin of first
or second generation Britons, but from the Vulgar Latin of the thoroughly
Latinized urban population of the late empire.  It's only if a substantial
part of that population maintained itself that a British Romancelang
could've survived and, indeed, would've almost certainly have supplanted
the earlier Brittonic speech.  As Andrew observed - rightly IMHO - early
on, if British Romance survived, i.e. Brithenig really existed here, then
there'd be no Welsh (or Cornish or Breton) since Brittonic would've
certainly have gone the way of Gallic in Gaul.

In such a world, we'd have to look to later Vulgar Latin and I think Andrew
was not unreasonable in looking at what happened in northern Gaul.

Ray.




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