From pycyn@aol.com Fri Oct 06 17:00:39 2000 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@egroups.com Received: (EGP: mail-6_0_3); 7 Oct 2000 00:00:39 -0000 Received: (qmail 31106 invoked from network); 7 Oct 2000 00:00:39 -0000 Received: from unknown (10.1.10.27) by m1.onelist.org with QMQP; 7 Oct 2000 00:00:39 -0000 Received: from unknown (HELO imo-r19.mail.aol.com) (152.163.225.73) by mta2 with SMTP; 7 Oct 2000 00:00:39 -0000 Received: from Pycyn@aol.com by imo-r19.mx.aol.com (mail_out_v28.26.) id a.44.7c92e53 (658) for ; Fri, 6 Oct 2000 20:00:31 -0400 (EDT) Message-ID: <44.7c92e53.270fc19e@aol.com> Date: Fri, 6 Oct 2000 20:00:30 EDT Subject: Re: [lojban] Re: Why place structure? To: lojban@egroups.com MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Mailer: AOL 3.0 16-bit for Windows sub 41 From: pycyn@aol.com X-Yahoo-Message-Num: 4515 The semantic cases that aulun mentions always look very inviting when confronted with learning the place structure of Lojban predicates. However, Loglan-Lojban has considered changing to such a system at least four times and has always returned to the present system. The problem is that either we end up with almost as many cases as their are predicate places taken altogether (roughly 3000) and the problem of remembering kernel each addition can go with meaningfully or we have a more manageable list of cases and the problem of remembering what each means when attached to a particular kernel (is the motive for an act the terminus a quo or the terminus ad quem? for example). We have to learn to express these things somehow and, however it is encoded, we have to learn the code for the whole bunch -- the same amount of information to learn. Now, what Lojban (and Loglan generally, so far as I can tell) could do better is design teaching aids that eased the strain somewhat, namely teaching words in classes with very similar place structures. In fact, some work has been done on this off and on and there are probably a dozen or so such classes that take care of the far greatest part of the vocabulary (and with the help of a few generalizations Agent -patient, goal-source-path, etc., almost all). There will be a few oddities (and more as lujvo and fu'ivla come along, perhaps -- though maybe not), but these can be learned as a separate class. Alas, so far as I can remember, none of these studies have resulted in teaching materials; gismu tend to be taught in order of their usefulness in some classroom situation or imagined "real world" one. Happily, we are (as always) in the process of starting to think about designing the beginning of a plan to work on teaching materials and may be able to incorporate some of this into the materials that eventuate (6 pu'o, I think, with no guarantee of an outcome). As for the scheme aulun mentions, I don't know it in detail but it looks like another one based on someone's conceptualization of the necessities of any event. These conceptualizations tend to be either officially a priori (that is, based on a little logic and a lot of the proposer's native langauge) or a posteriori (based on the possibilities in some agglutinative language's nouns: I-forget-who-tilla's use of Finnish is typical). Or mixed, of course (probably a better description of the Finnish case). They are generally as flakey as the ideas behind Loglan, so we decided to stick to just one mishugash at a time.