Return-Path: LOJBAN%CUVMB.BITNET@vms.dc.LSOFT.COM Received: from SEGATE.SUNET.SE (segate.sunet.se [192.36.125.6]) by xiron.pc.helsinki.fi (8.7.1/8.7.1) with ESMTP id FAA28234 for ; Mon, 22 Jan 1996 05:51:28 +0200 Message-Id: <199601220351.FAA28234@xiron.pc.helsinki.fi> Received: from listmail.sunet.se by SEGATE.SUNET.SE (LSMTP for OpenVMS v1.0a) with SMTP id C11DDA62 ; Mon, 22 Jan 1996 4:51:28 +0100 Date: Sun, 21 Jan 1996 22:50:47 -0500 Reply-To: Logical Language Group Sender: Lojban list From: Logical Language Group Subject: Re: sera'aku SNU: ki'e doi skot. X-To: ucleaar@UCL.AC.UK X-cc: lojban@cuvmb.cc.columbia.edu To: Veijo Vilva Content-Length: 1162 Lines: 25 >> Lojban content words are going to be PREDOMINANTLY 3, 4, 5 or 6 >> syllables. ideally Zipf will work to make it possible to have the >> shortest words be the most common ones > >What's the mechanism for that? If a long lujvo starts getting used a >lot, how can it be abbreviated? 1. Most likely, people will use a less specific word with fewer terms. 2. People will leave out se/nu/ke terms in lujvo making 3. ultimately, people could ask for a gismu to be assigned to the word or to a major component of it that is used in many lujvo. (or in spontaneous use, of course, they just start using a gismu). These are in order of preference. I really doubtthat many long lujvo will be so overused as to get a gismu assigned. There will probably be some objective criteria to meet before I would accept such a word, say by comparing frequency of word use to average frequency of gismu and 2 term lujvo. But this in particular is one thing I could see being considered after the 5 year baseline ends - consideration to adding to gismu to shorten some high frequency long words (the long words needn't go away, they will just have a short synonym). lojbab