Hallo!

On Mon, 28 Jun 2010 15:14:36 +0100, R A Brown wrote:

> Jörg Rhiemeier wrote:
> > Hallo!
> > 
> > On Sun, 27 Jun 2010 18:18:17 +0100, R A Brown wrote:
> [snip]
> >>
> >> Well, English has come close to doing that.  But even by the 
> >> 7th cent, much of Brittonic speaking Britain had given way 
> >> either to Gaelic or to various Norse and Saxon dialects. The 
> >> Britons held on only in the western part.
> > 
> > But English was not a trade jargon spoken in a few ports along
> > the southern coast. 
> 
> No - sorry, I had not intended to imply - merely observing 
> that the language a set of immigrant communities has led to 
> the demise of 'native' languages.

Which is self-evident.  Happened many, many times in various parts
of the world, including virtually everywhere Indo-European languages
are spoken.  The Paleolithic Continuity Hypothesis can be easily
dismissed as utterly misguided.  The reconstructable lexicon of
Proto-Indo-European includes words for such things as wheeled
vehicles and metals (copper, probably silver and gold, perhaps even
bronze, though certainly not iron), and that means that the language
cannot have been spoken before about 4000 BC.

> > It was the language of a massive wave of
> > conquerors who came to stay, and brought in their families.
> 
> The Saxons weren't very interested in trading with the Brits 
> - by DNA sampling has shown that in many area there was no 
> displacement of population, but rather the existing 
> population took to speaking the language of their more 
> powerful neighbors and/or overlords.

Right.  I never believed that the Anglo-Saxons wiped out the Britons
whereever they went.

> Something similar must have gone on in Britain (and, I 
> guess, Ireland) before since there is no archaeological 
> evidence the supposed 'Celtic' invasion and settlement. In 
> case of the spread of the Brittonic languages trade does 
> seem have played some part since there is clear evidence of 
> pre-Roman trade between Britain and Gaul.

Trade may indeed have played a role.  It is uncertain, AFAIK, when
the British Isles became Celtic-speaking, and how many waves of
immigration there were, though at least some scholars assume that
Goidelic is from an earlier wave than Brythonic.  But that is
uncertain.

> But as I understand it, some, at least, of the genetic 
> evidence suggests settled population that at some stage gave 
> up their ancestral language in favor of some 'proto-Celtic' 
> form of IE, and later gave this up in favor of Saxon or 
> Middle English (depending upon when the change took place).

Certainly; when a country is taken over by immigrants, the old
and new inhabitants usually mix to some degree.  A catastrophic
collapse of the old population only happens if the invaders are
extremely superior in organization and technology, as when the
Europeans colonized North America and Australia.  That certainly
never happened in the British Isles (except when Homo sapiens
displaced the Neanderthals).

> [...]
> 
> > The Brythonic languages have lost their case system, yes, and
> > have no absolute/conjunct distinction in their verbal inflections,
> > though there are traces of the latter in Old Welsh. 
> 
> Yes, the reduction of case system is not dissimilar to what 
> was going on in Vulgar Latin.  I wonder if this is not to do 
> with creolization so much as simplification which comes 
> about as the language spreads among non-L1 speakers. 

I think there is no hard and fast line separating creolization from
the sort of simplification that happens when a language spreads
among non-L1 speakers.  However, I am not certain that this factor
played a role with the loss of the case system in Vulgar Latin.
In most parts of the Roman Empire, people spoke languages with case
systems very similar to the Latin one before they adopted Latin.
They should have had no problem figuring out how the Latin case
system worked!  What happened was that sound changes, especially
the loss of final /s/ and /m/, eroded the Latin case endings to the
point that the endings became ambiguous.

Also, in Brythonic, AFAIK, the case system did not collapse until
about 500 AD or so.  And as in Latin, I think sound changes (loss
of final syallables) did in the case endings.

> Clearly this was happening with Latin as spread across the 
> Empire; something similar could have been going on as 
> Brittonic was diffused across Britain.
> 
> It does not, it seems to me, to have to reflect an actual 
> feature in any substrate language - there may, in case, have 
> been more than one pre-IE language spoken in Britain before 
> the diffusion of 'proto-Celtic'.  In any case the diffusion 
> of 'Celtic' through the Britain & Ireland seems to have 
> occurred in two phases.

The loss of cases in Brythonic certainly has nothing to do with a
substratum, I concur with you.  We know virtually nothing about
the languages spoken in the British Isles before they became
Celtic-speaking; we can only make guesses guided by what appear
to be traces of lost languages.

The Isles take part in the "Old European Hydronymy", a network of
recurring river names discovered by the German Indo-Europeanist
Hans Krahe in the mid-20th century.  Those river names are found
throughout most of Western and Central Europe; Krahe assumed them
to be Indo-European, but that is controversial.  Theo Vennemann
ascribes them to languages related to Basque, which I consider
unfounded, especialy as the distribution of the names shows a
gap between the Garonne and Ebro rivers, i.e. exactly where Basque
was spoken in antiquity.  My gut feeling is that the names are from
a language family related to Indo-European but breaking off before
the latter developed the ablaut system.  I call this hypothetical
language family "Hesperic", and associate its spread with the
spread of Neolithic agriculture in Central Europe.

So the British Isles may have been predominantly Hesperic-speaking;
however, the Hesperic of the British Isles appears to have shown
properties distinct from the Hesperic of the Continent, if the
typological aberrancies of Insular Celtic syntax are due to a
substratum effect.  The substratum languages appear to have been
VSO with verbal nouns, a tendency towards running together closely
associated words (leading to the rise of initial mutations) and 
second-position sentence clitics (constructions of that kind are,
according to Cowgill, underlying the absolute-conjunct distinction
in early Insular Celtic verbs); maybe they also were active-stative,
judging from the peculiarities of the behaviour of Insular Celtic
passives, impersonals and deponents.  The latter trait may have been
inherited from the common ancestor of IE and Hesperic, which probably
was active-stative, as can be internally reconstructed from PIE.

But all that is just speculation, useful as source of inspiration
for conlangs, and not much more.

> > The tendency
> > towards analytical verb forms is a fairly recent one, and cannot
> > be blamed on a Pre-Celtic substratum.  
> 
> Tho it is an areal feature as the same sort of thing has 
> been going on in English also - it would have been going on 
> for some time in popular speech before it made its 
> appearance in written documents.

Yes.  Tendencies towards analytical constructions are found in
several languages of Western Europe, including Basque, which is
almost like Kelen in its eschewal of finite lexical verbs.
Generally, in Europe, languages become ever more analytic the
farther west one goes, it seems.

> > Also, I wouldn't consider
> > the simplification of the PIE verb system one can see in Latin
> > and other western European IE languages a case of "creolization",
> > though perhaps a substratum played a role.
> 
> But the spark that began the simplification is likely to be 
> the spread of the language among L2 speakers.

This makes sense.  L2 speakers tend to miss the finer details of
the language's grammar, especially when it comes to categories
their L1 did not have.  One can perhaps conclude from this that
the pre-IE languages of Western Europe were tense-marking rather
than aspect-marking ones, considering how the IE languages of
Western Europe all transformed the aspect-marking verbal
morphology of PIE into tense-marking systems.

>       Interesting,  
> the Latin imperfect indicative (-bam, -bas etc.) and those 
> futures that ended with -bo, -bis etc appear to have began 
> life as a locative of a 'verb-noun' with the verb "to be" 
> derived from PIE bh- forms, i.e. a striking parallel to what 
> we find in Insular Celtic and, of course, English forms such 
> as "I'm a-coming", "he was a-going" etc.

Yes.  Similarly, the weak Germanic preterite is from an
analytical construction with 'to do'.  Another example is the
Romance future based on 'to have'.

> All the tenses (including subjunctives) formed on the 
> perfect stem, except the perfect indicative itself, are 
> clearly formed either with fusions of parts of 'to be' or 
> remodeled on analogy with that verb.

Yes.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html