From lojbab@lojban.org Thu Sep 20 14:44:35 2001 Return-Path: X-Sender: lojbab@lojban.org X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-7_3_2_2); 20 Sep 2001 21:44:35 -0000 Received: (qmail 20614 invoked from network); 20 Sep 2001 21:44:35 -0000 Received: from unknown (10.1.10.142) by m8.onelist.org with QMQP; 20 Sep 2001 21:44:35 -0000 Received: from unknown (HELO stmpy-5.cais.net) (205.252.14.75) by mta3 with SMTP; 20 Sep 2001 21:44:33 -0000 Received: from bob.lojban.org (209-8-89-124.dynamic.cais.com [209.8.89.124]) by stmpy-5.cais.net (8.11.1/8.11.1) with ESMTP id f8KLiU092764 for ; Thu, 20 Sep 2001 17:44:31 -0400 (EDT) Message-Id: <4.3.2.7.2.20010920172233.00e33bb0@pop.cais.com> X-Sender: vir1036@pop.cais.com X-Mailer: QUALCOMM Windows Eudora Version 4.3.2 Date: Thu, 20 Sep 2001 17:41:27 -0400 To: lojban Subject: Re: [lojban] META : Who is everyone (and what are they saying) In-Reply-To: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed From: "Bob LeChevalier (lojbab)" At 12:31 PM 9/20/01 +0100, And Rosta wrote: > >>> "Bob LeChevalier (lojbab)" 09/19/01 11:43pm >>> >At 01:33 PM 9/19/01 +0100, And Rosta wrote: >#>I know this, and it's clear that a pioneering loglang can't foresee >everything >#>that needs to be zipfed down. What I mean is that even when usage shows >#>what needs to be zipfed down, we then have no way to do the zipfing. >#>The morphology affords us no spare class of short cmavo, ># >#What do you think was the point/intent of the experimental cmavo space, if >#not precisely to deal with this. There aren't a lot of "short" cmavo, but >#"short" is relative here - there are 4 xVVs and 25 xV'Vs and we ultimately >#have the option of using the last few unused cmavo in regular cmavo >#space. It hasn't been made clear that there are that many Zipfean >#shortenings that are so commonly used as to warrant the shortest forms in >#the language. > >My sense is that people avoid cumbersome locutions because they are >cumbersome, so on the whole you'd only get usage of them if we actually >said to the community "In your usage, please don't try too hard to achieve >conciseness -- say what you wish to say ignoring its clunkiness, and then >in the light of this we will add to the language zipfean shortenings". Rather, I would expect "there is no way to say this that isn't mildly verbose or hopelessly confusing. I want to say it and be understood, so I will say it verbosely." >#No doubt, if usage proved such to be the case, then >#postbaseline revision would find a few of the lerfu and Mex words that >#Jorge despises, being displaced by things that have proven more useful. > >This is quite a radical change. Some of those may have been learnt and >used already. I doubt that our conservative constituency would stand >for it. That will have to be seen. I suspect that there are a few Mex words that only a couple of people have ever taken note of, and none have used. lerfu shifts even more so; they were intended to generalize JCB's goals to represent alphabets more completely with a minimum of cmavo (he uses all the [C]y s AND the [C]ei s for lerfu, so we did a good deal better), but the bottom line is that almost no one spells things out, we have a surplus of anaphora, and other uses for lerfu that exist don't crop up enough in usage to make it worth knowing them. I don't regret including them, because the original goals for the language included covering all features that JCB thought important, so as to be true to it being his invention. But I would be the last to say that they should tie up cmavo space post baseline if no one uses them and other things in experimental space are more used. I just want to save any discussions of specifics until then to give what there is a sound chance. >#Of course we have. That is exactly the sort of thing that I expect will be >#discussed after the baseline period ends (discussion not necessarily >#leading to change, but certainly considering it). I just don't want to >#discuss it until then, because we would need the usage information to make >#such decisions, and I want the decisions based on a more widespread Lojban >#use than would likely exist if we continued to talk about tinkering. > >I think that our usage is likely to be guided by whether or not we view it >as testing the language prior to the consideration of further revisions. That could be, and is the reason why I chose such a long baseline period, and talk down the subject of changes. I realize that they could be desirable or even necessary, but want people not to keep it in mind, and figure time is on my side in enforcing that not keeping it in mind (I think it would be hard to maintain the "this is only a test" attitude for 5 years). >Most >people quite legitimately care more about stylistic elegance than about >precision, and unless the notion is prominent in the culture that for >the time being it is a virtuous and necessary phase in the language design >process to ignore stylistic considerations pertaining to brevity, we will not >get the necessary evidence for where zipfing is required. Why not? The evidence will not be limited to actual verbose and clunky usages, but will also include places where people ellipsized and were/weren't understood, as well as experimental cmavo that will have been tried to deal with clunkiness. Stylistic elegance that fails to communicate something is still evidence. >Either that or >we encourage people to experiment with using their own zipfing experimental >cmavo, possibly with concomitant grammar changes, and then see which >experimentals catch on. Experimental cmavo have no defined grammar, so using them is not a "grammar change". However you have to be skilled enough at the language to be able to communicate without having a jbofi'e to parse and translate your experiments, since jbofi'e won't likely be able to incorporate all of them even if Curnow wanted to try (some might not pass LALR1) >#I have little >#respect for wild suggestions that aren't actually used, especially since >#"terki" and a couple other old Loglan creations (there was one gismu coined >#for x1 encroaching on x2's personal "space", as I recall). We might also >#encode gismu for all the unique Laadan concepts as well by the same logic. ># >#But proposals without actual usage to back them are empty. > >Having them up there in black and white (or blue and beige) on the wiki >gives them a kind of lexicographical legitimacy that might encourage >their usage. If you wanted to see it enter usage, the most effective strategy >would probably be to put it on the wiki and use it and add to the wiki a >full corpus of its usage. You might, and that seems a practical approach to experimental cmavo. But the wiki, especially growing as fast as it is, becomes a place of record of diminishing returns, because the odds that a random Lojbanist will read a given page becomes small. I doubt that I've read 5% of it, and I at most skim the pages changed in the last week (which is the most often I've been able to read it - I've only added to it once and it took me all night earning Nora's chastisement). lojbab -- lojbab lojbab@lojban.org Bob LeChevalier, President, The Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273 Artificial language Loglan/Lojban: http://www.lojban.org