From a.rosta@dtn.ntl.com Fri Jul 07 08:42:59 2000 Return-Path: Received: (qmail 3167 invoked from network); 7 Jul 2000 15:42:59 -0000 Received: from unknown (10.1.10.27) by m3.onelist.org with QMQP; 7 Jul 2000 15:42:59 -0000 Received: from unknown (HELO relay3-gui.server.ntli.net) (194.168.4.200) by mta1 with SMTP; 7 Jul 2000 15:42:59 -0000 Received: from m75-mp1-cvx1c.gui.ntl.com ([62.252.12.75] helo=andrew) by relay3-gui.server.ntli.net with smtp (Exim 3.03 #2) id 13Aa8D-00007w-00 for lojban@egroups.com; Fri, 07 Jul 2000 16:33:34 +0100 To: Subject: zi'o & otpi (was: RE: [lojban] So, wait til you feel a cold no-nose Date: Fri, 7 Jul 2000 16:42:51 +0100 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) Importance: Normal In-Reply-To: <20000504133953.52290.qmail@hotmail.com> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200 From: "And Rosta" > From: Jorge Llambias [mailto:jjllambias@hotmail.com] > Sent: 04 May 2000 14:40 > > {zi'o} is used to drop a place, but I avoid it as much as > I can, it doesn't feel right. It feels right to me only when it is slightly counterintuitive -- where the resulting predicate is substantially different from the original, e.g. le se gerku be zi'o, for a dog breed that exists independently of actual dogs. The issue of place dropping more generally arises [in my imaginary Lojban usage, at least] in the case of those many gismu where peripheral places are filled by elided zo'e but you don't really want to have to check that you really are claiming that each of these zo'e can be replaced by a da/de/di. Nobody's going to have the energy to replace elided zo'es with overt zi'os. Knowing you, I'd predict that your strategy would be to ignore them in usage, so that usage eradicates them. But (and I think here is a case where the lack of linguistics experience of much of Lojbania shows) a prevalent usage pattern is distinct from but hard to distinguish from a rule of grammar, and the power of pragmatics is such that it is both normal and feasible for people to say something that literally means X but in use habitually conveys Y. The language will either be defined by usage, in which case its grammar will be relatively vague and indeterminate, or it will be defined by formal documentation, in which case usage will largely be irrelevant. (Presumably, until computers are as intelligent as people, computers would have to speak the formally documented version.) So in reality, usage will never "decide" between alternatives; it will merely create doubt among them as to which is correct. So better than zi'oing off unwanted places, or pretending they're not there, is to use some alternative brivla. If VCCV fu'ivla really are kosher then they are an attractive solution, since they're even shorter than gismu, and although they don't have rafsi, in lujvo you could use the too-many-placed gismu, because in lujvo formation you can lose unwanted places inherited from the component gismu. So, for example, if you want a word for "bottle such that something actually is a bottle even when it's empty", then you could use "otpi" (with, in lujvo, the same rafsi as "botpi"). If "otpi" were as well-documented as "botpi", it'd stand a chance of competing against it in usage, and then usage really would tell you which was the more popular. --And.