From pycyn@aol.com Mon Oct 23 02:00:18 2000 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@egroups.com Received: (EGP: mail-6_1_0); 23 Oct 2000 09:00:18 -0000 Received: (qmail 89375 invoked from network); 23 Oct 2000 09:00:17 -0000 Received: from unknown (10.1.10.27) by m8.onelist.org with QMQP; 23 Oct 2000 09:00:17 -0000 Received: from unknown (HELO imo-r17.mail.aol.com) (152.163.225.71) by mta2 with SMTP; 23 Oct 2000 09:00:17 -0000 Received: from Pycyn@aol.com by imo-r17.mx.aol.com (mail_out_v28.31.) id a.c5.a5063ef (3959) for ; Mon, 23 Oct 2000 05:00:13 -0400 (EDT) Message-ID: Date: Mon, 23 Oct 2000 05:00:12 EDT Subject: Re^n: literalism To: lojban@egroups.com MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Mailer: Windows AOL sub 41 From: pycyn@aol.com X-Yahoo-Message-Num: 4664 Damn, I hate to get behind. So, working backward a bit. maikl: <> An arguable point, I think. But, in particular, unless the explanation already pushes beyond the old concepts, all you have is an old concept, a potential that has already been covered. Hopping up to something Lojbab said, adding 1,000 lujvo doesn't add 1,000 concepts. If you have a word "cat" for something which has a color and a color word "black," using "black cat" (or even "blact," assuming this is something one can do regularly) doesn't give a new concept. But suppose you want to talk about a racoon, in alanguage which doesn't have a word for it or any notion of it up til the first confrontation. You can, of course, probably work out an explanation in a sentence or two and then convert that to a noun phrase. But you can't use it often, or, I would contend, really grasp it. So you need a short expression for it. You can borrow a word from someone who has a longer acquaintance, {nimlyraknu}, say, for Lojban. Or you can create from somewhere a new word for it, just {raknu}, say, or, to avoid problems, something from space: {xnuka}, maybe. Or you can take a legitimate existing form and twist it "robber cat" maybe, or "washing cat" or some other vocables. And ultimately the last is the most efficient and generally acceptable to speakers. Of course, it changes the meaning of a word somewhat "cats" now includes some things that aren't cats, and thus it opens the way for another bunch of words -- for otter and weasel and ... <> I didn't say it was metaphorical, only that it was non-literal, which it clearly is. But it works beautifully for all that (indeed, largely because of that, in this case). And that is the point I keep trying to make about non-literal lujvo (and tanru, for that matter): if they work, use'em and don't just dump on them because they don't fit this or that or any rule, though, as Lojbab reminds us, JCB used to think of a lujvo -- in his earliest versions -- as a free-drop zone, capable of having any place structure at all, regardless of its components, so long as it had the right places for what he had in mind and he liked the look of it, and that may be going too far. ivan: <<>Since the `sky scraping' metaphor is very widespread, its >carrying over to Lojban would appear less scandalous than >some other `naturalisms' do, but what does it do that >{tcergaldi'u} or {galgaldi'u} do not?>> Do I need to remind folks that I don't necessarily think that some "sky scraper" calque is good Lojban; all I said was that it was very good in English originally and that it presents a good model for good Lojban. For one thing, ^sky scraper^ isn't a new concept, it is already inherent in ^building^ at the beginning of the 21st century.. So {tcergaldi'u} works fine. I am less clear about {galgaldi'u}, on good literalist grounds : {galgaltu} doesn't mean "very high" and {galtu galdi'u} needs a grouping mark. <, there will be just as little left to rejoice about. I'm willing to accept the argument that literal, transparent compounding has its limitations, and that at some point one does have to resort to metaphor. I am, however, in favour of keeping metaphor as a last-resort technique.>> I don't see how non-literal lujvo, by themselves, threaten lojbanic purity, though I admit that they can be misused to make a mere code for something else. That would be objectionable -- but not because of the way the compounds were formed. As noted, I think there are places where non-literals have to be the technique of first resort, simply because they are the only practical way to get the result needed. I thank aulun for his support and his examples on "skyscraper." I don't think that Lojban is likely to become Germanic in compounds, which is another reason why literalism (which often requires it and often needs empty syllables to do it) is not the wave of the future. lojbab: <> Note, again, I am not advocating nonconventional lujvo in the sense that the place structure has to go against the rules or that the connection between the components has to be paradoxical or any other thing like that. Too be sure, they make nice test cases, but so would "robber cat" for racoon. What I am objecting to is simply the rejection of proposals because they do not happen to fit a set of rules that the objector has canonized -- or even that The Book and its background data have. The first question is, does it work. After an affirmative answer to that, the rest doesn't matter (consider good ol' {le} for example -- 999 times out 1,000 it refers to a specific critter of the type named, the last time it doesn't but in the context no one doubts what it means. Good non-literal lujvo are like that last case, regardless of any fiddling that they may need for "the rules."