Re: [engelang] Re: [jboske] LoCCan3 development ideas. - engelang messages

On Wed, Aug 15, 2012 at 9:46 PM, And Rosta <and.rosta@hidden.email> wrote:

Brett Williams, On 03/08/2012 18:22:

> Hmm, this is an interesting thread. I think there's room for lots of
> loglangs! I wish we would explore more of the possibilities.
>
> The loglang I'm working on is presently named "Babebabo"! Its
> segmentation and basic level of parsing is all in the vowels, it uses
> five:
>
> A: Continues a root, so just "b" is a very short root, and then "bab"
> is another root, or "t" is another very short root, and also "bat"
> "tat" "tab" "batatatatab" "batabatababababababat" etc are all distinct
> roots infinitely. So there's plenty of room for roots, though of
> course there's advantages to being short.
>
> E: Combines roots into compounds, so the root "t" plus the root "b"
> makes the compound "teb", or the root "batatatatab" plus the root "t"
> makes the compound "batatatatabet", or "bab" plus "tat" plus "bat"
> makes "babetatebat".
>
> I/U: Together a root+i and a root+u form a pair of parentheses.
> Inside them can be nothing, or one or more more such pairs. So for
> instance: "babi babu" (babi babu) or "babi babi babu babu" (babi (babi
> babu) babu) or "babebabi babi babu babi babu babebabu" (babebabi (babi
> babu) (babi babu) babebabu) or "babi babebabi babi babu babebabu babi
> babu babu" (babi (babebabi (babi babu) babebabu) (babi babu) babu)
>
> O: Just a shortcut-- if there's nothing inside of an I/U pair,
> instead of saying the word twice it can be shortened to just root+o
> once, so "babi babu" becomes just "babo" and "batabatabababababababati
> batabatabababababababatu" becomes just "batabatabababababato"!

There are less cumbersome alternatives to this root-doubling method.

Brett's system isn't the tersest, but it is simple and has the benefit of being immediately recognizable to human ears what exactly is being terminated, which can be helpful in parsing lengthy, heavily modified phrases. I once had a left-headed, right-branching language with a convention of repeating a root with a prefix as an similar, optional method of terminating a constituent, something like

constituent := {stem} ( {modifiers} ( {terminator-prefix} {repeat-stem} ) )

I got the idea from the HTML syntax i.e. <element> ... stuff ... </element>, where the terminator prefix was sort of like the HTML slash. As in HTML, but not as in XML, the terminator was optional.

In Brett's system, it might be nice to have a shortcut though. Perhaps allowing batabatabababababababati to be replaced with bi or something when there is no ambiguity, which should never arise, because it seems like everything is explicitly terminated.

The thread beginning with <http://listserv.brown.edu/archives/cgi-bin/wa?A2=ind0905b&L=conlang&F=&S=&P=4004> discovered that given a binary tree, a four-way inflectional contrast on terminal nodes is sufficient to yield an unambiguous parse.

So for example if you were prepared to add a sixth vowel, Y, you could use I,U,O,Y as bracketing markers. (I'd use I,E,U,O for bracketing, and Y for the compounder.) -- This presupposes the trees are binary branching, tho.

It took me a little while to figure out their system; it's very interesting, but I don't think it's going to be possible for humans to choose correctly among the four tones (or vowels) on the fly. It appears to require a highly abstract algorithm to determine where to the mark the edges of internal groupings.

My conlang, Livagian, is (nowadays) strictly left-binary-branching, except for quasi metalinguistic stuff like quotatives. So its syntax is so tediously simple as to be virtually nonexistent.