WikiDiscuss

On 6/3/05, Robin Lee Powell <rlpowell@digitalkingdom.org> wrote:
>
> How about adding this to the Preface section (and, of course, it
> would need to be stuck in the CLL somewhere; ideas?):
>
> !!! Implicit sumti Raising
>
> If a gismu place specifies that it normally takes an abstraction, or
> a particular type of abstraction (ka, nu, etc), then if a concrete
> object appears in that place, it should be considered to have "tu'a"
> in front of it. This also applies to BAI and SE BAI sumti tcita in
> which the underlying gismu place requires an abstraction.

What does it mean "if a gismu place specifies that..."?
Is that about how the current gi'uste deals with that, or are
you proposing a thorough revision of place structures?

For example, consider {xamgu}:

x1 (object/event) is good/beneficial/nice/acceptable for x2 by standard x3

lo djacu cu xamgu lo spati
"Water is good for plants."

That could be interpreted as {tu'a lo djacu} (receiving water, for example),
and also as {tu'a lo spati} (for the growth of plants, for example):

lo nu cpacu lo djacu cu xamgu lo nu lo spati cu banro

But why stop there? That could also be interpreted as {tu'a lo nu
cpacu lo djacu},
(that receiving water happens frequently, for example)) and also as {tu'a lo nu
lo spati banro} (that the growth occurs intensely, for example):

lo nu lo nu cpacu lo djacu cu cafne cu xamgu lo nu lo nu lo spati cu banro
cu mutce

When do we stop? And the markings in the current gi'uste are rather
haphazard: x1 of xamgu is marked (object/event), x2 of xamgu is
left unmarked. Other places that would seem to work just like
these are marked as event only. Often the markings were constrained
by the definition being required to not exceed a certain number of
characters.

mu'o mi'e xorxes