On Thu, Dec 27, 2012 at 9:08 AM, v4hn <me@v4hn.de= > wrote:

On Thu, Dec 27, 2012 at 07:55:11AM = -0700, Jonathan Jones wrote:
> On Thu, Dec 27, 2012 at 7:36 AM, v4hn <me@v4hn.de> wrote:
>
> > On Thu, Dec 27, 2012 at 07:39:14AM -0500, Ian Johnson wrote:
> > > ko dunda pa plise mi is "make it true that there exists= exactly one apple
> > > that you give to me."
> >
> > > If you have an apple in mind and say that, then you're n= ot conveying
> > > that you have one in mind at all.
> >
> > Completely understood and absolutely right.
> >
> > However, as far as I can see {ko dunda pa plise mi} does not expl= icitly
> > say whether or not you have a specific apple in mind.
> > It's just, that I didn't specify it, if I have one in min= d.
> > In my opinion it's still a valid counterquestion to ask "= ;Which one?" here,
> > don't you think so?
> >
> > Also, if you answer with {pa plise} again, it might be a bit stup= id, but
> > still
> > it's a reasonable answer to say "I got that part, but wh= ich one do you
> > want?"
> > in my opinion.
> >
> > The question is, whether there is a way to _spell out_ that you _= don't_
> > have one in mind instead of omitting further information and wait= ing
> > for the listener to recognize the conversational implicature.
> >
> > No such way, which does not explicitly state {ko cuxna} or {mi na= pensi lo
> > selsteci},
> > was mentioned up to now _as far as I understood everything_
> > and even the {ko dunda da poi plise ku'o mi} seems to rest on= a very strong
> > conversational implicature instead of explicitly stating that you= don't
> > care/know which one you'll get.
> >
> > That is because it doesn't seem to be a problem to me to add<= br> > >
> > {ko dunda da poi plise ku'o mi .ije mi djica lonu do dunda lo= zunle traji
> > mi}.
> >
> > However, it is weird to say something like "Give me any appl= e. I want the
> > leftmost one."
> > or "Gib mir irgendeinen Apfel. Ich will den ganz Linken.&quo= t; (in my mother
> > tongue).
> >
> >
> > Any opinions?
> >
> >
> > v4hn
> >
>
> I do't don't why you's want to explicitly state non-specif= icity. Why is
> extremely over the top vague not good enough for you?

Because, as I said earlier, the mechanism that makes you unders= tand the "over the
top vague" expression is a conversational implicature and it's a r= ather harsh field
of study to extract these from a sentence. Also they can always be cancelle= d afterwards
as I've tried to demonstrate, and you therefore don't really /commi= t/ yourself to
them.

> In any case, the best I can up with is the lujvo {nalterte'i}.

Don't you mean {nalselte'i}?

I suppo= se you could do that instead, yes. I meant {nalterte'i}, however, as th= at refers to a set within which no member has anything special about it.

This would look as follows then?

{ko dunda lo plise poi nalselte'i ku'o mi} or {ko dunda lo nalselte= 'i plise mi}

I didn't really make up my mind about these up to now...

v4hn