Re: Middle Welsh (was Cein)

First of all, a great big thank you to Keith and Sally
who explained non-finite verbal-nominal clauses in
Middle Welsh in such an excellent way. Thanks a lot!
That really cleared things up in my head.

Sally Caves wrote:

> I'm not exactly sure, though, how it is an "active" language.
> Can Daniel be more specific?  I may need some reminding about
> active languages, but I thought it was one rather like Teonaht,
We had quite a discussion about active languages some months ago
while you were offlist. *looking* It was Feb 3rd to Feb 13th.
this year. Look for "THEORY: What is an active language?" and
"Active case-marking natlangs".

Anyway. Since Elliott (and others) asked, the reference for
the article is:

MANNING, H. PAUL 1995. Fluid intransitivity in Middle Welsh:
Gradience, typology and 'unaccusativity'. Lingua (97), 171-194.

That shouldn't be hard to find. Any respectably library
should have the journal "Lingua".

He's a bit hung up on Vendler's "activities, accomplishments"
and such, but generally he comes up with some interesting
points. I wouldn't have analyzed it that way, but that's what
you get when you absolutely have to go by a theory.

Summary:

Subjects in non-finite verbal-nominal clauses are marked
peculiarly from the stand-point of finite clause marking
which is nominative-accusative in that it allows variation
in the marking of intransitive subjects.

In transitive non-finite clauses, *transitive* subjects (A)
are marked with the prep. _o_ 'of, from', and transitive
objects (O) are marked as simple adnominal genitives.

(1)   VN       A              O
    kymryt [o  Arthur] [y  daryan eureit]
    taking  of Arthur   the shield golden
    'Arthur took the golden shield.

Subjects of non-finite *intransitive* verbs vary between being
marked as either the A of above or the O.

(2)   VN         A
     ymlad    [ohonafi] dros vym baryf
     fighting  of-me    for  my  beard
     'I fought for my beard.'

(3)           VN        O
     kynn   diodef    [Crist]
     before suffering  Christ
     'Before Christ suffered.'

_Ymlad_ 'fight' thus marks its subject as A while _diodef_
'suffer' marks its subject as O.

Some other points:

Only [+ HUMAN] subjects can be marked as A. If the intransitive
subject is [- HUMAN] it is marked as O, regardless of which
verb is used. Remember also that this variation between A and
O is only applied to intransitive verbs.

According to Manning's analysis, the semantics behind the
marking is this:

(1) If the verb is an Activity, its subject is marked as A.

(2) If the verb is an Achievement, its subject can be marked
    as either A or P (fluid marking).

(3) If the verb is Stative, its subject is marked as O.

Sample verbs of class 1 (A):
ride, walk, run, fight, preach, eat, study, pray, return,
sin, mount.

Sample verbs of class 2 (A or O):
suffer, sleep, sit down, flee, remain.

Sample verbs of class 3 (O):
be, die, slip, escape, fall down.

Manning is able to describe the variation by dividing the
verbs into "telics" (events with a clear terminal end-point,
like "kick") and "atelics" (no natural end-point, "play").

Other variables he uses are [CONTROL] and as said above
[ACTIVITY].

Summing up the results in a table:

Marking:  Atelic:                  Telic:

A         [+control] activities     --
A or O    [-control] activities     [+control] telics
O         states                    [-control] telics

Or in other words: If the verb is a controlled activity
A-marking is used, if it is a state or a non-controlled
telic, O-marking is used. If it's a non-controlled activity
or a controlled telic, it can be either-or.

He finally notes that some of the either-or verbs have
a bias towards O-marking and thus doesn't have totally free
marking.

Did that clear things up a bit?

||| daniel

--
<> Mad llamgalf! <> daniel.andreasson@telia.com <>
<>     Ond llam! <> www.geocities.com/conlangus <>