[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
On Thu, Oct 11, 2012 at 4:13 PM, selpa'i <seladwa@hidden.email> wrote: > > 2. Transitive compounds > > These are common in Lojbanic languages. One example are the -gau lujvo, > e.g. "kakygau" decomposes into "gasnu lo nu klaku", that is, the first > part of the lujvo gets inserted in a place of the right part of the > lujvo. Right now, these can be handled using the usual Xorban syntax, > but it's a bit clumsy sometimes. > > la mmtaka'a le fe klka'a gsnake > My mother brought about that I cried. We have unary operator b- for this, which introduces an agent, so in this case that would be just: la mmtaka'a ba klka'a > Note that instead of the usual gasnu definition x1 brings about x2 (fa), > we could also use a slightly different one: x1 makes that x2 does/is x3 > (fa). Trying to use this makes me note that it's not possible, since > Xorban requires every formula to have a desinence, and there is > currently no variable that refers to a ce'u-like thing, if there were, > however, then: > > la mmtaka'a le fe klke'u gsnaka'ake > > Here, klke'u is a sort of abstract infinitive, and is filled > automatically by whatever is in the x2 of the (altered) gasnu. > > This solution is yet longer than the first, though, so maybe this kind > of place structure isn't the best for Xorban. The way I would do this would be to define a predicate as "x1 makes x2 be x3", so: la mmtaka'a le klke glfaka'ake When we introduce aspect, we can distinguish "one who is (now) crying" from "one who (habitually) cries". > Anyway, these are fine as normal sentences, but I think they could be > shortened a little bit by introducing a new binary operator that creates > transitive compounds. If possible, I would propose "jo", otherwise, pick > one that would work. > > The semantics of "jo" would be that one of the left formula's places > gets filled with the right formula, depending on which ones have the > same type (event of concrete), and it would also be hard-coded into > place structures which places get filled in such compounds. > > Here's an example: > > la mlta jo gsna'a plpa > The cat, I make that she jumps > I made the cat jump. > > The other sentence becomes: > > la mmtaka'a jo gsna klka'a > My mother, she made me cry Is there a difference betweeen "jo" and "ju"? > Here is another example to round it off: > > la bdna jo tca kkta > Bananas are very delicious. > > (I think tc- should be the root for "x1 is very x2 (fa) (unless you're > planning on introducing lambda calculus à la ka-ce'u into Xorban)). It could be just "x1 is very x2" without any "fa". la glka tca'aka Happy ones, I'm very much one of them. > 3. Loose, metaphorical compounds > > These are just to fill the gaps that remain. Sometimes, a human being > won't want to care about perfect logic and just throw together some > words freely. I had some other solutions for this, but again they seem > to be meant for a language other than Xorban. This suggestion seems to > work: > > cpno'e zfra'a > I am birdly free. > I am free as a bird. > Some other vague meaning. > > Maybe it can be made even more referent-less, by using the > aforementioned -e'u variable: > > cpne'u zfra'a Those are two sentences though. You need either a coordinator or a binder to make them into one sentence: je cpno'e zfra'a li cpni zfra'a co ma'a xrxe