[YG Conlang Archives] > [ceqli group] > messages [Date Index] [Thread Index] >
On Sun, Apr 21, 2002 at 11:10:21AM -0600, Rex May - Baloo wrote:
> These linking things are new to me, I think. The Loglan system is that
> preds modify following preds. Le kukra prano. The fast runner. And to
> make compounds you have to bust the words up into cockeyed allomorphic
> 'combining forms', giving you something like 'kuapra' for a compound. I say
> 'something like' because you never know what the combining form is going to
> be without looking it up.
This is how it works in Lojban too - "kukra prano" would be what I call
a tanru - but tanru can also have words in between that specify
something about how the words are modified. I think Loglan must have had
these in some form too.
"kukra prano" - let me change this to "sutra bajra" so the result isn't
a horrible mix of languages - could also be expressed as "sutra je
bajra" ("fast and runner", which is more precise, but people avoid it
because it's one syllable longer and it contradicts their English ideas
of what an adjective is), or "bajra co sutra", and if you wanted to say
"the female fast runner" you would say "fetsi bajra bo sutra" to make
sure that bajra still modifies sutra first.
>
> The Ceqli system is that there are no combining forms. On the contrary,
> something like.
> To kala faul.
> Automatically forms a compound
> To kalafaul
> Unless the preds are separated with a 'sa'.
> To kala sa faul.
Right. So what I'm hoping is that, for example, "kala je faul" would be
the same kind of structure as the one formed with "sa", while
"kalajeqfaul" would be a compound.
You can use r instead of q if you want. But it wouldn't be entirely like
Lojban, as Lojban tended to use whatever letter was available, and I
think that CVr could clash with a bunch of existing predicates.
> I presume rafsi are the same as the hideous Loglan allomorphic combining
> forms? If so, of course not. Ceqli's intitial impulse was to do away with
> those. All other changes pale in comparison. Now, I don't mind at all
> deriving predicates from structure words by adding -r (or maybe something
> else). That's an entirely different thing. A good, helpful thing.
Right. And misunderstood me. Incidentally, I wouldn't fully consider
these "predicates", because they can't stand alone (what would "boq"
mean?) and have no place structure. Perhaps there could be another term;
for now I consider them "fake predicates".
--
Rob Speer