summary of co-possession

[Arsenij Roginskij (or possibly Michael Daniel)]

Dear Edith!
I wanted to comment on your summary that the case of Spanish does
not make the tendency to avoid co-possession less universal - it seems
that Spanish simply does not dispose of other adnominal relation marking
device to
disambiguate the construction.

Thus the universal runs as follows -

IF the language dispose of several devices for marking noun to noun
relation,
THEN two different marking devices are used.

1. This works only when the relation of the two dependent nouns
("possessors") to the head is semantically different (like 'part of' VS
'possession', like John's part of the estate); if the same meaning is
intended, the two possessors would be simply coordinated, of course
(like John and Kate's house).
2. Obviously, this universal seems to be more useful for dependent
marking languages, cause in head marking languages co-possessing
constructions (if any) would not be ambiguate in the sense they are in
dependent marking ones.

Michael Daniel