On 2 October 2012 01:26, la gleki <gleki.is.my.name@gmail.com> wrote:

=
On Tuesday, October 2, 2012 12:15:29 AM UTC+4, selpa'i wrote: Am 01.10.2012 22:04, schrieb Arnt Richard Johansen:
> On Sun, Sep 30, 2012 at 12:56:59AM -0700, la gleki wrote:
>
>> These examples are pretty clear. No ambiguity. Now let's o= pen Chapter 11.4
>> = http://dag.github.com/cll/11/4/
>>
>> 4.9) =A0 la djan. cu zmadu la djordj. le ka mi prami ce'u
>> =A0 =A0 =A0 =A0 John exceeds George in-the property-of (I love= X).
>>
>> This is something very strange.
>> Let's imagine that I'm a boy and I meet a girl. I tell= her
>>
>> {do melbi mi lo ka ce'u clani}
>>
>> Does {ce'u} refer to {do} or {mi}?
> Neither. {ce'u} does not refer to anything, and that's sor= t of the whole point. The clause refers to a =93property=94 of being tall *= in the abstract*, not that of someone in particular being tall.

Yes, but the ce'u place gets filled later by one of the sumti in th= e=20
parent bridi, and that is what the question was about.

> * lo ka mi clani
> =A0 =A0my height
>
> * lo ka do clani
> =A0 =A0your height
>
> * lo ka lo penbi cu clani
> =A0 =A0a pen being long

These are very non-standard, and probably many would call them=20
incorrect. The only way I can see to make these work is to say that &qu= ot;lo=20
ka mi clani" is to be interpreted as "lo ka mi no'u ce= 9;u clani", or else=20
the "ce'u" would end up in a different sumti place and th= e meaning would=20
change drastically.

I believe tha= t {ce'u no'u ko'a} is wrong. The bridi in which the property-ab= straction appears provides the value for the ce'u-place. Personally fil= ling a ce'u-place changes the abstraction type from ka to du'u.

=A0

> * lo ka clani
> =A0 =A0being long (OR being a dimension of length OR being a stand= ard of length)

Right, but much more than with "ke'a", where this is hand= led much more=20
loosely, the convention is that "ce'u" fills the first em= pty slot.

So if there are no {ce'= ;u} =A0specified we must understand it as the first and only the first slot= is filled with omitted {ce'u}. If two or more of the slots of the =A0n= ested brivla are filled we must specify all of them (like in {mi e do simxu= lo ka ce'u ce'u prami}), right?

That would make sense however this is something that mu= st be included into CLL 2.0.
IMO=A0{ce'u}-izing gimste is als= o a must.=A0

btw, dont you think that we can use {= ce'u} in {mi djica lo ka/nu *ce'u* citka lo plise} instead o= f {vo'a/mi}?

Sure, but what of wanting things that don't involve djica1 = at all?

I've thought long and hard about this, as evidenced b= y the abstractions paper that I wrote on the wiki. In general, when an abst= raction place doesn't make sense if you exclude the broda1 from it, the= n it requires ka.=A0

*{.i mi kakne lo nu do citka lo plise} (example from an= old mriste thread of mine)

is nonsense, which prompts me to beli= eve that kakne2 is a ka or equivalent.

However {.i= mi djica lo nu do klama lo zarci} makes perfect sense, even though the dji= ca1 is not inside the djica2.

The gimste being more prescriptive about abstraction ty= pes would be great, but good luck finding enough people who support the cau= se of more rigid types. (AFAIK there are at most two or three, myself and p= robably you included.)

.i mi'e la tsani mu'o