WikiDiscuss

WikiDiscuss


Wiki page BPFK Section: Subordinators changed

posts: 14214

On Wed, Aug 18, 2004 at 06:38:47AM -0700, Jorge Llamb?as wrote:
>
> > ! Formal Definitions
> >
> > (AKA conversion formulas)

In this episode, xorxes rips Robin's conversion formulas a new ganxo.

Not that I mind.

> > || noi | PA broda noi brode cu brodi | PA broda cu brode .i je PA
> > broda cu brodi
>
> Hmm... There's something wrong here. The problem is that a
> quantified term does not provide referents outside the scope of
> the quantifier, so you can't really take the noi-clause out.
>
> ci prenu noi melbi cu klama
>
> Exactly three people, who are beautiful, came.
>
> does not say that exactly three people came and exactly three people
> are beautiful. It says that exactly three people came and that *those
> same three people that came* are beautiful.

Point.

> So {ko'a noi brode cu brodi} = {ko'a brode ije ko'a brodi}, but it
> won't work like that for quantified terms. When you have a quantified
> term, first you have to take the quantifiers to the prenex and only
> then apply this transformation, and you can't take the noi outside the
> scope of the quantifier in such cases.

I only barely followed that. Can you give an example, and do you have a
solution? Does zo'u work with tu'e?

> > voi, another way | PA broda voi brode cu brodi | PA broda poi mi
> > skicu lo ka ke'a broda cu brodi
>
> Shouldn't that be {poi mi ke'a do skicu lo ka ce'u brode}?

Actually, I'm going with just {poi skicu ke'a fo lo ka ce'u brode}.

But yes, your basic point is correct.

> > ne | PA1 broda ne PA2 brode | PA1 broda noi ke'a srana PA2 brode
> > pe | PA1 broda pe PA2 brode | PA1 broda poi ke'a srana PA2 brode
> > no'u | PA1 broda no'u PA2 brode | PA1 broda noi ke'a du PA2 brode
> > po'u | PA1 broda po'u PA2 brode | PA1 broda poi ke'a du PA2 brode
> > po | PA1 broda po PA2 brode | PA1 broda poi ke'a se steci srana PA2 brode
> > po'e | PA1 broda po'e PA2 brode | PA1 broda poi ke'a jinzi ke se steci srana
> > PA2 brode
>
> I would define these much more generally:
>
> ne sumti = noi ke'a srana sumti
> pe sumti = poi ke'a srana sumti
> no'u sumti = noi ke'a du sumti
> po'u sumti = poi ke'a du sumti
>
> etc. It is not necessary to restrict the definitions to a particular
> form of sumti, or to a particular point of application of the clause.

Good point.

> > vu'o | PA1 broda [JOI / A] PA2 brode vu'o [relative] | PA1 broda
> > [relative] [JOI / A] PA2 brode [relative]
>
> This does not always work like that. The relative clause need not be
> distributive.

Example? Solution?

> > zi'e | PA1 broda [relative] zi'e [relative] cu brode | da poi du
> > PA1 broda zo'u da [relative] .e da [relative] cu brode
>
> This one doesn't work in general either.
> You could define it something like:
>
> noi subsentence1 zi'e noi subsentence2 = noi ge subsentence1 gi subsentence2
> poi subsentence1 zi'e poi subsentence2 = poi ge subsentence1 gi subsentence2
> voi subsentence1 zi'e voi subsentence2 = voi ge subsentence1 gi subsentence2
>
> ne, pe, etc. can be put into noi/poi form and then this conversion
> will also apply.

Cool, thanks.

> Mixed cases are a separate issue.

Yes; they should die in the ares.

> > goi, unassigned | [sumti 1] goi [sumti 2] | [sumti 1] poi du
> > [sumti 2] ku'o
> > goi, both assigned | [sumti 1] goi [sumti 2] | [sumti 2] poi
> > binxo da poi mintu [sumti 1] ku'o ku'o
>
> These are not true equivalents.

You mean that "du" is wrong, or that he formula are wrong?

> > * I have no idea why only pe and ne can be used with
> > sumtcita clauses.
>
> For a general tcita, {fi'o broda}, we would have:
>
> pe fi'o broda fe'u sumti = poi ke'a jai broda fai sumti
>
> ne fi'o broda fe'u sumti = noi ke'a jai broda fai sumti

Why 'jai'?

> Once we know what po and po'e are, we could probably have
> similar conversion formulas for them,

I think po'e as it stands is fine; do you have a problem with it?

> but I don't know what it would mean to use tcita with goi,
> po'u, or no'u.

Neither do I.

-Robin