Re: [engelang] Xorban: la je cmla nltra

Original:
"Once when I was six years old I saw a magnificent picture in a book, called True Stories from Nature, about the primeval forest."

This is a tricky first sentence!  The noun phrase headed by "book" adds most of the complexity, so let's simplify that to just "book" and see what we get.

Hu'a. So pvo fo la'a mba'a ji'a la nnca ska'aka, le (je li ckti pseki) (je bnle) pxre vska'ake.
P.  Once (I) the myself, when I was a six year-laster, a (book-pertaining) (grand) picture I see.

That's not very complex so far.  In general, the more SOV-leaning a language, the greater the tendency of relative clauses and prepositional phrases to precede the modified term.  Xorban is pretty SOV-leaning, so (IMHO) most naturally, we want to put the book's prepositional phrase before the picture.  Note that I have parenthesized the modifiers of pxre.  In Xorban, adjectives, prepositional phrases and relative clauses are all basically the same thing except for their internal complexity.  They are just formulas preceded by "je".  Here's the stuff (I think) we left out as wff, followed by different options for integrating them:

mi je li clci psaki je jtna lsra = "(i) is called <nature-pertaining true stories>"
la je prmvla rcfla psika = "(i) is about the primeval forest"

1. The most straightforward approach is to just slap these in with "je":

Hu'a. So pvo fo la'a mba'a ji'a la nnca ska'aka, le je li (je mi je li clci psaki je jtna lsra) (je la je prmvla rcfla psika) ckti pseki je bnle pxre vska'ake.
P.  Once (I) the myself, when I was a six year-laster, a (called <nature-pertaining true stories>) (primeval forest pertaining) book-pertaining grand picture I see. 

2. That does seem complex.  Maybe we want to try a non restrictive clause approach (via parentheticals).  We can accomplish that simply by just dropping the "je".

Hu'a. So pvo fo la'a mba'a ji'a la nnca ska'aka, le je li (mi je li clci psaki je jtna lsra) (la je prmvla rcfla psika) ckti pseki je bnle pxre vska'ake.
P.  Once (I) the myself, when I was a six year-laster, a -- it was called <nature-pertaining true stories> and was about primeval forest -- book-pertaining grand picture I see. 

3. That's still heavy.  We might as well move the parentheticals past the main verb and let the implicit binding rule do its thing.

Hu'a. So pvo fo la'a mba'a ji'a la nnca ska'aka, le je li ckti pseki  je bnle pxre vska'ake; mi je li clci psaki je jtna lsra; la je prmvla rcfla psika.
P.  Once (I) the myself, when I was a six year-laster, a book-pertaining grand picture I see; it[book] was called <nature-pertaining true stories>; it was about the primeval forest.

4. I am a little iffy about implicitly binding "i", which is my preferred local throwaway variable, especially when the old binding is deeply nested.  Maybe we should move out "book" to the same level as the picture in the main clause, and give it a discourse topic variable.  That would give us:

Hu'a. So pvo fo la'a mba'a ji'a la nnca ska'aka, la'i ckta'i le je pseka'i je bnle pxre vska'ake; ma'i je li clci psaki je jtna lsra; la je prmvla rcfla psa'ika.
P.  Once (I) the myself, when I was a six year-laster, a certain book, the of-it grand picture I see; it[book] was called <nature-pertaining true stories>; it[book] was about the primeval forest.

That would be my final version (in this stage of my understanding).  Maybe we could also assign "e'i" to "pxr" to make that a little more salient.  You can keep reusing these variable in the story, rebinding them every so often.

As you can see it's complicated, but there are actually tons of options.  I think we just have to use the language's strengths to our advantage, ignore English's structure, and see what we can do.

Since "nx" (I think) would nearly always be used as "nxu'a", I have to say that it seems a little wasteful to set aside a whole CC stem for this.  What I think we need is a word for "paragraph".  Then we can form the sentence "This very sentence begins a paragraph".  Then we assign a particle as a shortcut for that sentence, perhaps "wu'a".  I am not sure how Jorge and others feel about using interjections in this way, so I will add experimental "hu'a" to the list.  Does that work for your purposes?  I am trying to gather all words in use on this page:
 
http://loglang.wordpress.com/xorban/vocabulary/

... keeping in mind that anything can change at any time.

I have no problem with that.  I really don't think that we need more than five generic V at this point, and ten would be plenty for certain, so at least a few others can be assigned implicit restrictions.  (Maybe we should revive Jorge's idea of "e'u" from "prenu" to mean "le'u pre'u" i.e. "someone", which can be useful at times.)
 

I agree that "la'a mslfa'a" (6 syl.) is a mouthful, so I have already added "mb" (a mere cluster) as a possible shortening of "mslf".  Keep in mind that "a'a" is fully grammaticized in Xorban, so you will never have to say "la'a mba'a" (4 syl.) except to add emphasis.  As far as topic-comment, we already have the versatile particle "co" which together with mb can keep things down to three syllables, which is as short as English:

co mba'a klma'a "As for me, I'm going".

Most likely we'd want to quantify over events e.g. "so nmpo fo smo'e."  Some one thing was an event in which something was something = once, something happened.  It's hard to imagine how to shorten that at the moment, so hopefully that works for the present time.
 

That works fine.  It means "Some A & E pertaining to each other are such that ...".  Usually one would say something like "la je psa'ake mlte" meaning "my cat ...".
 

I think that <i> & <u> need lax or near-open allophones to facilitate <yi> and <wu>, and that a lot of people would mix up [I] and [i] even if that weren't the case.  So I would say that buffer vowel has to be either the schwa, or the mid or mid-open rounded front vowel like French neuf.  Another option I think would be the close central vowel found in Russian sometimes written [1].

That's my fault, but it was not random.  I was thinking of this: http://en.wikipedia.org/wiki/Benford%27s_law  I would consider "nl, pv, dv" optional CC variants for 0, 1, 2.  The problem with C roots is that the low amount of redundancy given that we have no vowel differences among roots.
 

That's And's preference, but I think that <'> remains [h\] and that <q> is [?].  However, it would not harm anything to allow <q> to be written <"> I think, which might be a little easier on the eyes.