On Fri, Aug 24, 2012 at 12:03 PM, And Rosta
<and.rosta@hidden.email> wrote:
Mike S., On 24/08/2012 03:13:
>
> There would be relatively little homonymy and no need for a kludge if
> there were a way to transliterate names from the vast array of
> natural languages with relatively little change. I gather that
> transliteration is not a primary aim of the language, but
> nevertheless I believe it would be useful and convenient for Xorban
> speakers to have a relatively faithful and easy method of
> transliteration. Using a designated phoneme like the glottal stop
> (properly spelled) {'} would allow that. That way, recognizable words
> like {'spageti'-}, {'betxoven'-}, {'joxanezberx'-}, {'kimono'-},
> {'tiranosawrus'-} are possible that cause little need for fussing or
> guessing. To xorbanize, simply drop/alter the vowels and drop
> glottals -> spgt-, btxvn-, jxnzbrx-, kmn-, trnsrs- and
> modify/compound if there is a collision with an existing predicate.
>
> Or, maybe the first {'} could be a CV as you suggested, and the
> second {'} always preceded by a vowel. That would be a little longer,
> but easier to pronounce.
Yes, <'> is /?/. I would say that name quotatives enclosing a name X produce a two-place name predicate expressible in metalanguage as "x"(y, z) which means "Y (is known as X to) Z".
Or, you could have two name-initial particles, e.g. <'i> for names where the stem ends before the next vowel, and <'u> for names where the stem ends at the next <'>.
That would be worth it, I think. If tone and stress are free, then you could capture names pretty accurately after initially transliterating the segments.
> > The only question is about using the commonly used variables of shape
> > "V" - if we are also doing things like using free variables to create
> > the passive voice so to speak, we have to be careful about that. The
> > sentence "na la xrja nlceka" intended as "pigs aren't liked", under a
> > certain context might mean "the woman with the red hat doesn't like
> > pigs". So maybe we should set aside or some "V'V" or "Vy"or "aw" for
> > the purpose of an explicitly unbound variable indicating something
> > like "zo'e".
>
> {zo'e} is a terrible idea, because it's so unhelpfully vague. But a V'V for "le co'e" -- "him/her/it/them" might be a good idea. Maybe the rule would be that V'a is interpreted as a definite reference unless explicitly bound.
I saw you and Xorxe working on csna'e-type bindings last night, but I haven't figured it all out, but "what we're discussing" seems to match my idea for a set of assignable discourse topics.
> That's the whole point though. You aren't too concerned that it's
> vague; you only want to fill the next slot. I might have opened up a
> can of worms by mentioning to {zo'e}. I am not sure what the
> difference is in Lojban, but I imagine "aw" (now "o'e" by xortermi'e)
> might mean be more equivalent to "lo ku" in Lojban if that were
> grammatically possible. In Xorban "o'e" would be automatically bound
> by "lo'e smo'e"; I believe "sm-" to be the generic predicate such
> that "ra Pa sma" is true for all predicates P in Xorban.
Ah, like Lojban {lo du} rather than {zo'e}. Yes, fine. But I would assign this interpretation to any variable not explicitly bound, which distinct unbound variables having distinct reference, and (tho probably this creates too many problems) repeated unbound variables having the same reference.
I think that most implicit unbound variables get the "o'e" treatment at the very least.
> As far as definite references, I think all "reused" (i.e. previously
> bound but now free) variables are definite references. Free variables
> are automatically bound as "l- R-" where R was the restriction under
> the most recent binding.
Can you reexplain what you mean here?
I believe that I misspoke when I said "definite reference", because it now seems to me that a previously bound free variable will not be a definite reference, though it will often be interpreted that way. What I understand at the current time about how this works in Xorban is this (and please don't mind the details, I'm trying to work this out for myself too):
(sa xrma (lu bjrafu vska'aku)) (xkra)
That can be reasonably be translated as "(Some horse, (I see it running).) (It's black.)"
In the first sentence, "a" is bound by the binding operation "sa xrma" where "sa" is the binder (quantifier) and "xrma" is the restriction; "la xrma" has scope over the sub-sentence (lu bjrafu vska'aku) so the "a" in "bjrafu" is explicitly bound.
In the second sentence "xkra", "a" is free. The rule given by Xorxe is that if an explicit unbound variable appears that was bound in a previous sentence, then it is implicitly bound again using the same recent restriction but with "l-". So "xkra" can be expanded as:
[la xrma] xkra
I was tempted to imagine that "a" here is a definite reference to the horse seen running in the first sentence, and while that interpretation is left open by the rules, others are not closed. It could mean also "They [horses] (in general) are black". "a" doesn't necessarily mean "the horse that I saw running", because that's not part of the previous restriction.
==================================
What would be the problem with variable "a", appearing free in Qa, that was previously bound in a sentence "Sa Ra Pa" being implicitly bound by
[la je Ra Pa]
Then (xkra) would mean
[la je xrma (lu bjrafu vska'aku)] xkra
"It [horse(s) that I see running] is black"
That is not perfect, but it does tighten the restriction on previously bound free variables considerably. In order to derive the implicit restriction for a free variable, however, [la je Ra Pa] has to clean up its own free variables; this is already a known issue even with the implicit [la Ra] binding operation.
I think the steps are:
1. "Sa Ra Pa" is rewritten "la Ra Pa".
2. Next, depending on what is binding "la Ra Pa", move "la Ra" in front of the previous binding "Se Re" if "e" is free in the original "Pa", alternately move "la" in front of the previous binding "Si Ri" if "i" is free in "Ra". We can probably safely eliminate most je-branches except ju. Repeat zero or more times until no variable is free in the expanded "la R'a P'a".
3. Finally, "la R'a P'a" is rewritten as "la je R'a P'a".
(sa xrma re prne (lu bjrafu vskeku)) (nlceku)
"Some horse, everyone sees it running. They like it.
nlceku = [le RPe] [lu RPu] nlceku
To get [le RPe] i.e. ["le R'e P'e"] start with [Se Re Pe]
0. re prne lu bjrafu vskeku
1. le prne lu bjrafu vskeku
2. le prne sa xrma lu bjrafu vskeku //fronting to bind "a"
3. le je prne sa xrma lu bjrafu vskeku
"person that sees some horse running"
To get [lu RPu]
1. lu bjrafu vskeku
2a. lu bjrafu re prne vskeku
2b. lu sa xrma bjrafu re prne vskeku
3. lu je sa xrma bjrafu re prne vskeku
"event of some horse running that everyone sees"
The final result:
sa xrma re prne lu bjrafu vskeku
"Some horse, everyone sees it running."
nlceku
= [le RPe] [lu RPu] nlceku
= [le je prne sa xrma lu bjrafu vskeku] [lu je sa xrma bjrafu re prne vskeku] nlceku
"They [person that sees some horse running] like it [event of some horse running that everyone sees].