Hi!

Benct Philip Jonsson writes:
>...
> I say go for it!

Ok. :-)

>...
>>     'serve > ?se'reve 'to serve'
>
> I say what's wrong with non-stop + r? Especially in a lang
> which has many other freak initial clusters. You might
> insert stops or turn v > b to get more conformant clusters:
> _mbrekat, brukan, screve/streve_. FERVERE could certainly
> become _freve_ as _fr_ would occur otherwise anyway.

'brukan' is especially cool, I think, and 'mbrekat' and 'streve' fit
phonology perfectly.  Very good ideas!

>>  In these cases, we might need another resolving strategy
>>  (e.g., drop of 'r' or 'v' in the above case). I do not
>>  want to introduce (unstressed) schwas.
>
> What about having rC > Cr in such cases: _mekrat, vukran,
> sevre/sebre_. This might even be the preferred strategy, or
> CVRC > CrVC *or* > CVCR in cases where both are possible.

I had though of the rC > Cr shift, but did not like it much,
especially since Sardinian tends to shift the r to the front across
syllables (CVCr > CrVC), and I tend to like the result like 'kraBa' <
CAPRA.  (I thinks there are cases where 'r' crosses two syllables in
Sardinian, but I can't come up with an example.  Would have to search
word lists (printed in a book, not on hard disk for easy processing).)

>...
> No no, _protu, prale, Krale, etrenitat_ are *way* too cool.

I think so, too. :-)

> BTW wouldn't _protu, prale, Krale_ lose their final vowels?

It would be 'protu' ('u' is kept due to my u-declension fetish) but
the others would drop the final vowel, yes: 'pral' and 'Kral'.

>>  This is a bit like Modern Dutch's 'kerruk' for 'kerk' and
>>  'melluk' for 'melk' and 'errug' for 'erg'
>
> So pedestrian (and happening in young females' Swedish ATM:
> _ett, tuvå, turé; kulara, starruka färrejer_). This kind
> of drawl isn't half as appealing as metathesis, IMHO.
> T*kunan has a 'snappiness' which you shouldn't give up.

Hehe, yes. :-)

I would probably keep it 'Terkunan', though.  The language name is an
import anyway (like 'Español' is, too).

>>  (only a schwa is inserted here, and the l is involved
>>  too, which might also be thinkable in Terkunan). Words
>>  with lC in Terkunan include:
>>
>>     kalde 'warm' Alman 'German'
>
> Why not _krad(e), Laman/Amlan_?

I will have to run a few tests to decide about -l-, I think.

> Or that the metathesis precedes syncope in *TERRACONA.

Ok, I'll also run tests here and see what the result will be.


Thanks for all your input! :-)

**Henrik