From - Thu Jan 16 10:09:12 1997 Reply-To: Claudio Gnoli Date: Thu Jan 16 10:09:12 1997 Sender: Lojban list From: Claudio Gnoli Subject: Lojban's imperfections? X-To: Lojban mailing list X-cc: Chris Bogart , Tommaso Donnarumma To: John Cowan X-UIDL: 229521a618a38210d594120e35b69d75 X-Mozilla-Status: 0001 Content-Length: 1706 Message-ID: > Date: Thu, 12 Dec 1996 17:47:41 -0700 > From: Chris A Bogart > On Thu, 12 Dec 1996, Christian Richard wrote: > > 6. To learn Lojban, I would have to master the order, number, and > > semantics of arguments associated with each predicate. I have notice > > that this dependency on order is also very much like the usual order > > English. Are there again reasons why this order is not culturally neutral > > and wasn't originally determined randomly instead of matching the natural > > order the arguments would have in English? Or is this a direct import > > from the predicate calculus? > I think an attempt was made to place the arguments in order of > importance -- but this task was performed by an English speaker. > I think it was done well enough, but perhaps a native speaker of > something very different from English would see the bias more > strongly than I can. There is a vicious circle involved here, as it emerged while discussing with my friend conlanger Tommaso Donnarumma. To found the definition of argument places on objective data, one could made statistics on the occurrence of each verb-object pair in a given natlang. But that frequency of occurrence is itself biased by the grammatical structure of the natlang. Lojban grammatical structure is no doubt defined more neutrally than many natlangs' ones, but it seems theoretically impossible to avoid any bias at all. In fact, also the frequency of occurrence of a given argument in Lojban can be biased by the definition of the predicate: not grammatically, because any argument can be omitted, but probably still psychologically. co'o mi'e klaudios