From LOJBAN%CUVMB.bitnet@YaleVM.YCC.YALE.EDU Sat Mar 6 22:51:50 2010 Received: from YALEVM.YCC.YALE.EDU by MINERVA.CIS.YALE.EDU via SMTP; Fri, 21 May 1993 14:25:45 -0400 Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2) with BSMTP id 9402; Fri, 21 May 93 14:25:01 EDT Received: from CUVMB.COLUMBIA.EDU by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP id 7224; Fri, 21 May 93 14:25:53 EDT Date: Fri, 21 May 1993 18:56:07 +0100 Reply-To: ucleaar@UCL.AC.UK Sender: Lojban list From: Mr Andrew Rosta Subject: Re: Cowan on morphology X-To: lojban@cuvma.BITNET, conlang@diku.dk, Logical Language Group To: Erik Rauch In-Reply-To: (Your message of Fri, 21 May 93 10:52:15 EDT.) <9305211453.AA55147@link-1.ts.bcc.ac.uk> Status: O X-Status: X-From-Space-Date: Fri May 21 19:56:07 1993 X-From-Space-Address: @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET Message-ID: > I hate to complain, but could And enlighten us to what he finds baroque > or ghastly about Lojban morphology. I can'rt think of anything much > simpler than having components that represent each possible root, and > putting them together in the order they would appear in a syntactic > based compound. You can call it complex to have multiple choices for the > morpheme to represent some roots in some positions, but since you can always > use the longest form, and ALL forms of the root are identical in meaning > and completely interchangeable, the effects of this polymorphism is minimized > in terms of langauge understanding, while allowing for a redundancy that > Lojban like most highly regular conlangs, tends to otherwise lack. Rafsi have to be memorized: although there is only a finite number of potential phonological structures for the rafsi of any given gismu, this number can be quite large, and one has to learn which of this large number are actually correct. Even after one has learnt one rafsi, there may still be more rafsi to learn for the same word. Rafsi are (fairly) freely varying allomorphs, so if a lujvo contains 3 rafsi, and the gismu for each rafsi has 3 rafsi, then this lujvo has 27 alternative forms. This, I would predict, could vastly impede word recognition. It is well known that many or most readers recognize words by their visual shape. Quite possibly when listening we recognize words holistically by their acoustic shape. Only with unfamiliar words do we bother to do a morphological breakdown. The rules for what constitues a legal lujvo are also complicated. I don't think that the longest forms of rafsi are the ones generally used. Ju'i Lobypli is read mainly by people with low Lojban competence, but the journal does not have a policy of using only 5-letter rafsi. Also, it might be difficult for a competent speaker to remember to reform familiar lujvo into 5-letter rafsi for the benefit of a less competent interlocutor. As for redundancy, I think that for a long time to come the problems of memorizing the vocabulary will loom larger for Lojban users than the problems of talking in a noisy environment. Also, the risk of two brivla having similar sound and equally plausible meaning in context is less acute than the risk of two cmavo having similar sounds and equally plausible meaning in context. It's the cmavo where lack of redundancy is a problem, though even here I think that the mnemonicity gained is worth the redundancy lost. ------ And.