Return-Path: Received: by snark.thyrsus.com (/\==/\ Smail3.1.21.1 #21.19) id ; Fri, 14 Feb 92 10:46 EST Received: by cbmvax.cbm.commodore.com (5.57/UUCP-Project/Commodore 2/8/91) id AA13479; Fri, 14 Feb 92 10:32:07 EST Received: from cunixf.cc.columbia.edu by relay1.UU.NET with SMTP (5.61/UUNET-internet-primary) id AA27809; Fri, 14 Feb 92 02:30:05 -0500 Received: from cuvmb.cc.columbia.edu by cunixf.cc.columbia.edu (5.59/FCB) id AA04451; Fri, 14 Feb 92 02:30:09 EST Message-Id: <9202140730.AA04451@cunixf.cc.columbia.edu> Received: from CUVMB.COLUMBIA.EDU by CUVMB.COLUMBIA.EDU (IBM VM SMTP R1.2.1) with BSMTP id 8294; Fri, 14 Feb 92 02:28:48 EST Received: by CUVMB (Mailer R2.07) id 6955; Fri, 14 Feb 92 02:28:21 EST Date: Thu, 13 Feb 1992 23:04:47 -0800 Reply-To: "F. Schulz" Sender: Lojban list From: "F. Schulz" Subject: learning lojban X-To: uunet!cuvmb.cc.columbia.edu!lojban@uunet.UU.NET To: John Cowan , Eric Raymond , Eric Tiedemann Status: RO X-From-Space-Date: Fri Feb 14 10:47:05 1992 X-From-Space-Address: cbmvax!uunet!CUVMA.BITNET!LOJBAN I really like the diagrammed sentences document. I was able to learn a lot about the grammar without getting bogged down in details. This is my favorite introductory writing on lojban. I believe the document does not need to be changed much, other than the correction of minor errors. The introductory grammar rules should never generate illegal sentences. Covering the full is not necessary. I like learning grammars which generate subsets of the language. The sumti ne sumti pattern is good enough for this level of detail. What I would really like to see, would be follow on documents written in the same delightful style, going into more detail. I have read 3 of the 6 draft lessons. The major thing I did not like was mixing in the vocabulary sentences with the explanatory text. I have a lot of trouble learning the gismu and drill in the vocabulary by constantly rereading sample text. Having sample lojban text isolated in a separate part of the book would be very helpful. The way I learn the gismu is by making lists and memorizing the lists. This way I can do vocabulary drills while doing other activities such as driving or walking. I suspect you need to have a fairly good memory for logflash to work. I memorize gismu 3 at a time, and then assemble clusters of 3 into lists. This minimizes demands on memory during the initial learning. I do not find the gismu to be much easier to learn than words from unknown languages. I believe the linearity assumption that JCB makes is totally useless. For example, take "cidni" which has 100% recognition score for the English component since all of "knee" is present. My personal recognition scoring is 0%. This was the hardest word for me in the lesson 1 body parts. I spent a lot of extra time to learn this word. The scoring is position dependent. The final two letters provide almost no recognition value. Even worse is the "cisma" , "cmila" pair, which I guarantee many English speaking lojban speakers will swap. I tend to swap them even after intense drills. This pair gets very high negative recognition scores for English. Some words are easy. I see "tcati" as "ocha" from japanese. The "ti" at the end is useless baggage. In spite of this criticism of word recognition, I believe the word construction was done well, giving good phoneme frequency counts. Frank Schulz ( fschulz@pyramid.com )