Re: [lojban] Re: Holiday Present from the BPFK: The gadri Proposal Has Been Completed

["And Rosta" <a.rosta@v21.me.uk>]

Apols for posting this to Lojban list, but the twiki is not
practicable for me -- wd be happy to post to jboske if that
is preferred.

xorxes:
> --- And Rosta wrote:
> > Is there somewhere I can read more about outer quantifiers of
> > lo/le? 
> 
> Outer quantifiers work in exactly the same way for every sumti:
> 
>   PA <sumti> = PA da poi ke'a me <sumti>
> 
> i.e. they quantify over the referents of the sumti.
> 
> The referents of {lo broda} are the things that are/do broda. No more
> specification than that is given. In some context, these may be each
> and every thing that exists in the world and does broda, in another
> context, it may be a single thing ("Mr Broda") or it may be the 
> only relevant broda around, or a number of relevant brodas. A sumti
> can have several referents at once without this being a reference to
> the single group entity. When {lo broda} refers to more than one thing, 
> it is not specified whether the things will satisfy the bridi they 
> fill distributively or collectively.
> 
> The referents of {le broda} can be more than one in the same way, 
> without this being a reference to a single group thing. 
> 
> > (I'm curious about how the subtype/instance/member
> > distinction got handled.) 
> 
> When there is reference to a group, (for example {loi bakni},
> and {lo gunma be lo bakni}) then you need to use {lu'a} or 
> {lo cmima be} in order to get to the members of the group, and 
> then quantify. A direct quantifier in this case will quantify 
> over groups, because the referents of those sumti are the groups, 
> not the members.

Presumably lu'a or cmima are needed to say "3 members of lo vo
nanmu", too, right? 

> Types are not handled in any special way, the referents of
> {lo broda} can be types of broda in a given context. When the
> distinction is important and not clear from context, you have
> to use a selbri (lo klesi be} vs {lo mupli be}, or something
> with the appropriate place structure if the place structure of
> these is too weird. {lo'e} could also eventually be used for
> this, but this gadri was excluded from the holiday present.

What about things like "this is a picture of two snakes" (or
"every snake depicted by this is..."), or "we drank two wines"
(or "every [kind of] wine we drank was...")? It is not klesi
that are depicted or drunk, is it? 

I suspect your answer will be that the type/instance distinction
is not made. In that case, presumably you would agree that
"No snakes are depicted by this and two snakes are depicted
by this" (and "Exactly 20 wines were drunk by us and exactly
two wines were drunk by us") can be true, if quantification in the 
first clause is over instances and in the second clause is over 
subtypes.

On another point, xorlo says "An outer quantifier can be used to 
quantify distributively over such groups. A fractional outer 
quantifier can be used to select a subgroup and indicate its 
cardinality as a fraction of the cardinality of the group." Does
"ro" count as fractional? And how would one say "2 out of every
3 things that are", as opposed to "2 out the three members of"?

--And.