From lojbab@lojban.org Thu May 11 02:35:51 2000 Return-Path: Received: (qmail 7777 invoked from network); 11 May 2000 09:35:51 -0000 Received: from unknown (10.1.10.142) by m3.onelist.org with QMQP; 11 May 2000 09:35:51 -0000 Received: from unknown (HELO stmpy.cais.net) (205.252.14.63) by mta3 with SMTP; 11 May 2000 09:35:51 -0000 Received: from bob (dynamic125.cl7.cais.net [205.177.20.125]) by stmpy.cais.net (8.8.8/8.8.8) with ESMTP id FAA07064; Thu, 11 May 2000 05:34:11 -0400 (EDT) Message-Id: <4.2.2.20000511050221.00ac9460@127.0.0.1> X-Sender: vir1036/pop.cais.com@127.0.0.1 X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.2 Date: Thu, 11 May 2000 05:37:56 -0400 To: araizen@newmail.net, lojban@egroups.com Subject: Re: [lojban] re: nazycau gerku and najyzme In-Reply-To: <95802547501@out.newmail.net> References: <11.3110cb6.263cc12c@aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed From: "Bob LeChevalier (lojbab)" At 11:14 PM 05/10/2000 +0200, Adam Raizen wrote: >la pycyn cusku di'e > > Is it time for the occasional worry about the lack of redundancy (= the > > packedness of some word spaces) in Lojban? > >I don't really think that Lojban does lack redundancy. I'm not sure >what you mean by packedness of word spaces, but I don't think it's >true if you mean that basically every word is necessary. I interpret "packedness" as referring to the fact that nearly all possible word in Lojban ARE words of some sort, even if their meaning is not always well-defined. (Nearly all CVs and CVVs are words, all consonant finals are names, a high percentage of words in lujvo-wordform space are valid lujvo because nearly all short-rafsi values have meaning, and of course all words in fu'ivla space are presumed valid, if not necessarily meaningful.) Probably he is referring specifically to cmavo, though I myself worry more about lujvo. No, after looking back at the text that inspired this: ><le narju guzme>> it is redundancy in lujvo space that pc was referring to. But I'll address the general redundancy problem anyway. While not all strings of Lojban words are valid Lojban sentences, the lack of features like morphological agreement means that errors in grammar have less secondary information to allow correction. Some areas of the grammar are very highly packed in that you cannot use cloze techniques to accurately predict the word category/part-of-speech of a missing word. Thus in, "le _____ broda", the space can be filled by several cmavo of several different selma'o, and some of those cmavo are very similar is sound/spelling and yet have different meaning and grammatical function. Likewise in "mi _____ broda" (a common language textbook technique is cloze drills, but I found that when writing the Lojban textbook, that such drills were impossible because too seldom could I constrain problems to the grammatical features being taught at that point). Coupled with the fact that no one is expert and fluent in all the cmavo, we tend to assume and allow for the possibility of word-choice error and accept some ungrammatical utterances; but in Lojban things are tightly enough packed that there is fear that the wrong correction might be made. On top of this, minor changes in the wording can lead to major changes in meaning. This is true of all languages to some extent, some seems more true of Lojban. Small wording changes that lead to major changes of grammar while still being grammatical is even more common. I myself am taken aback by more than 2 or 3 cmavo in a row, which are not written as a compound. I know that such forms can be grammatical, but they usually involve cmavo that I am less familiar with and I thus have low confidence that I can correct errors in cmavo usage. (In spoken Lojban I am noted for being worse than most in my ability to understand Lojban that has unusual grammar or minor errors). > Lojban has >many words which define grammatical structure unambiguously but >nevertheless are rarely needed to understand the sentence. Take >for example words like 'va'o', 'ri'a', etc. These are basically always >followed by 'le nu', and thus 'le nu' is redundant, except in a strict >lojban grammatical definition. That is indeed a case where redundancy may exist, though we have Lojbanists that push the boundaries of the language and use non-"le nu" sumti there. > In addition, if the language really >does lack necessary redundancy somewhere, it has enough >machinery that its speakers should have no problem inventing the >necessary redundancy. For example, I find that I sometimes add >redundant FA's, especially when talking about the third place of >'knows/opines/intuits/etc. x2 about x3', maybe because the >redundancy makes it easier to interpret. For lujvo, you always have the ability to use long-form, narjyguzme vs nazbyguzme, but if we have to do this too much, then short-form lujvo space will be shown to be too tightly packed. 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 (newly updated!)