Re: P- and Q-Celtic (was Re: Reasonable sound changes.)

Hallo!

On Tue, 11 Jan 2005 18:40:38 +0000,
Ray Brown <ray.brown@...> wrote:

> On Monday, January 10, 2005, at 08:01 , Jörg Rhiemeier wrote:
>
> > Hallo!
> >
> > On Mon, 10 Jan 2005 18:19:55 +0100,
> > Carsten Becker <naranoieati@...> wrote:
> >
> >> BTW, I've always wondered why there is P and Q-Celtic. So
> >> it's because on the Isle, people changed /k_w/ -> /p/ and
> >> on the continent they didn't?
> >
> > The geographical distribution of P- and Q-Celtic is different.
> > They shifted /k_w/ to /p/ in Britain and Gaul, but not in Ireland
> > and on the Iberian peninsula.
>
> Yes, the /k_w/ is the older form inherited form inherited from PIE. The
> shift /k_w/ --> /p/ happened in the central area, leaving the outlying
> areas unaffected.
Exactly.  This also means that there is no "Q-Celtic" node in the
Celtic language family tree other than the root, Proto-Celtic.
Proto-Celtic was indeed a "Q-Celtic" language.  Not even "P-Celtic"
necessarily forms a valid node.

Some scholars assume that the /k_w/ > /p/ shift was connected
with the more advanced La Tène culture and spread because
it was considered a shibboleth of a more sophisticated population.
But that's merely speculation.  After all, it just filled up a gap
in the Celtic stop inventory that was left behind by the deletion
of PIE /p/, and could have happened more than once.

>       We find similar things happening in the Italic dialects
> which is one of the factors that have led some to postulate a Celt0-Italic
> group.
I think Italo-Celtic makes sense, but I know that the evidence
for it is rather weak.

> [...]
>
> >
> > A feature that seems indeed to be confined to the British Isles
> > (with the exception of Brittany, which was settled by British Celts)
> > are initial mutations, which are found in both Goidelic (insular
> > Q-Celtic, i.e. Irish, Manx and Scots Gaelic) and Brittonic
> > (insular P-Celtic, i.e. Welsh, Cornish and Breton), but not in
> > Celtiberian (continental Q-Celtic) or Gaulish (continental P-Celtic).
>
> Absolutely! This feature which some seem to regard as quintessentially
> 'Celtic' developed only in Ireland and Britain; also, tho there are
> superficial similarities, the Gaelic and Brittonic systems of initial
> mutations are different.
Yes, they are.  Even if one assumes that there is such a thing as
"Proto-Insular Celtic", which probably isn't at all anyway,
one could not reconstruct any initial mutations for it (except
perhaps some kind of subphonemic fortis/lenis thing).

> > VSO word order is also an insular phenomenon, it seems.
> > Why that?  It just happened.  Blame the Elves ;-)
>
> Possibly - but it is more often the Semites who get blamed, leading to all
> sorts of wild theories  :)
Yes.

> I have seen it postulated that the 'Insular Celtic' languages developed
> from a creole that evolved in the Cornish peninsular as the result of
> trade contacts between Phoenician traders to the 'Tin Islands'. I am not
> sure how Old Irish would fit into such a theory.
Not at all, I think.

> Others have suggested a substrate population that was related to the
> modern Berbers and migrated up through western Europe in the age of the
> megalith builders (who were certainly pre-Celtic).
>
> Who knows?
In my conworld, i.e. in the framework of the League of Lost Languages,
it is a rather convoluted affair involving *two* different non-IE,
pre-Celtic families, Albic and Pictic.  These two and Insular Celtic
form a linguistic area which is characterized by the "Insular Celtic"
features - initial mutations, VSO order, etc.  What regards the
affiliations of Albic and Pictic, many scholars assume that Albic
is related to IE and arrived on the isles later than Pictic, while
Pictic is linked to Basque by some and to Afro-Asiatic by others.

> Just to add to the fun, other features common to Insular Celtic and the
> Semitic langs are:
> - all nouns are either masculine or fem. (common also to romance langs)
Yes.  This has happened several times in IE, e.g. in Lithuanian.
And I seem to remember that Old Irish still had a neuter gender.

> - adjectives follow the noun (a few excptions in Celtic langs - and
> Romance :)
Most, if not all VSO languages have that order.

> - definite article only
> - prepositions are conjugated
> - common way of expressing genitive, thus:
>         Arabic: beet ir raagil   (double vowels indicate long vowel)
>        Welsh:  ty'r  dyn
>               house the man = the man's house
This is indeed an interesting coincidence.

> Perhaps after all the elves are indeed to blame   ;-)
Perhaps... well, they have a long tradition of being considered a
race with strange, magical powers, and being blamed for just about
everything.

Greetings,

Jörg.

Reply