From jorge@xxxxxxxxxx.xxx.xxx Mon Apr 5 12:26:20 1999 X-Digest-Num: 107 Message-ID: <44114.107.587.959273824@eGroups.com> Date: Mon, 5 Apr 1999 16:26:20 -0300 From: "=?us-ascii?Q?Jorge_J._Llamb=EDas?=" While I am reluctant to reopen the notorious le/lo debate, I think we really >need to clarify this point. I think it's a good thing to discuss this from time to time. Maybe some people joined the list after the latest rehash and were never exposed to the debate. And besides, I always learn something new from the new participants in the discussion. >The veridical/non-veridical distinction is >semantic, while the known/unknown distinction is pragmatic. We need to >decide which takes precedence, and unfortunately the book is not terribly >helpful here. I'm not sure we have a choice as to which takes precedence. First of all, there is a semantic distinction even if we ignore the veridicality property. {le cipni} is by definition the same as {ro le su'o cipni} = "each one of the at least one bird which we're discussing". {lo cipni} is by definition the same as {su'o lo ro cipni} = "at least one of all the birds that there are". You strongly need pragmatics in the case of {le} to determine what are _all_ the birds, and much more weakly in the case of {lo}, but aside from that, there is an important semantic difference: the implicit quantifier! Once you have used pragmatics to identify the complete set of birds (all those under discussion, in the case of {le}, often only one, and all those that there are, in the case of {lo}), usually a large number, we are still left with the quantifier: using {le} you refer to all the members of the set, and therefore once you've identified the set you need no further identification. Using {lo} you only say some property applies to at least one of the members of the set, but you never identify which member. This is a semantic difference, not pragmatic. Your claim in one case applies to an identified referent and in the other case to an unidentified one. >> If there are no clues in the context >> about which apple or which table I mean, then I should >> not have used {le}. > >This is hard to justify in terms of Lojban semantics as they currently >exist, which enable one to use {le} for anything except for cmene. Not really. For example, I can't use {le} here: le ci verba cu citka lo plise Each of the three children eats an apple. In that example with {lo} each child might be eating a different apple. Had I used {le plise}, the meaning necessarily would have been that each child ate the same apple (or apples, but each child eats them all). The most likely meaning is that each child ate their own apple, so I couldn't have used {le}. >> i le zunle cu se pritu >> i le gapru cu se cnita >> i le xamgu cu se xlali >> >It is not clear which left-person, high-person and good-person is referred >to, and in fact, since these are Lojbanic proverbs, noparticular referent is >intended. I'm sure you know that proverbs are particularly problematic for this kind of analysis, certainly they don't always follow the same grammar as normal speech in many languages. Maybe I should have used {lo'e} there instead of {le}? Or maybe {ro pritu}, {ro gapru} and {ro xamgu} would make for a stronger saying? In any case, the way I would interpret it is something like "imagine any situation where you have two or more people, then we can say that {le zunle cu se pritu} = the one on the left has someone to the right." In that context, {le zunle} indeed does serve to refer to an identified referent, even if it's an imaginary one. On the other hand, if I say {lo zunle cu se pritu}, I'm saying a trite nothing, there's someone on the left who has someone to the right, but there would be nothing to lead you to the proverb's conclusion. Maybe someone else on the left doesn't have someone to the right and so maybe being good is not always bad for you after all. >Perhaps, post-baseline, the best thing to do is scrap {lo} altogether. I'm not sure we can at this point. In any case, if we were to choose one single article the best one would be {lei} rather than {le}, but that's for another discussion. co'o mi'e xorxes