From jorge@xxxxxxxxxx.xxx.xxx Sun Apr 25 22:52:11 1999 X-Digest-Num: 124 Message-ID: <44114.124.715.959273824@eGroups.com> Date: Mon, 26 Apr 1999 02:52:11 -0300 From: "=?iso-8859-1?Q?Jorge_J._Llamb=EDas?=" What has 'safe' to do with anything? :) Sorry! I have to remember to use different arguments with different peple. I do think masses are easier to use, but I also think that they are in a sense more basic. My own goal is to make the language as usable as possible without sacrificing any of its precision. I certainly don't believe in fudging to make it easier to use. >You seem to mean 'people would find >it easier', but that was not an argument that was given much weight in >the discussion that led to "tu'a"; and my interest in Lojban is that we >explore the intricacies that follow from parts of our design, rather >than sweeping them under the carpet. I'm with you all the way! That's what this discussion is about. My objection to {tu'a} is that, as far as I can tell, it is not a fully satisfactory solution to the issue it was supposed to solve, not that there is no underlying issue to take into account. >You're right, the book does. .ue.oi. I don't like this at all. I accept >that there are going to be occasions when you want them massified (ko >bevri le pipno), but I see individuals as more fundamental, and also >believe that they will more often be useful. I would advocate "lu'oko >bevri le pipno". Apparently I am advocating a change in the language. The reason that I think masses are more fundamental is that, roughly speaking, with masses one sentence refers to one situation, but with many individuals one sentence refers to many situations. lei ci nixli cu dunda lei cukta lei re nanla The three girls gave the book(s) to the two boys. That's one event being described. le ci nixli cu dunda le mu cukta le re nanla Each of the three girls gave each of the five books to each of the two boys. That's thirty events being described, with the five books changing hands over and over again. This is very general: sentences with distributive {le} describe a whole array of parallel situations, and I don't think this is the case with normal sentences in other languages. The reason this is not as noticeable in general as one would expect is that the most common use of {le} is for singular descriptions, in which case this splitting of situations obviously does not happen. >>Aren't quantifiers very explicit? What more explicit than that can >>you get? >> >No, because the grammatical operation depends crucially on the VALUE of >the quantifier - not its form or the words it is made of. Remember "vei >ny" is a perfectly good quantifier. Yes, I remembered {vei} after I sent my reply. My excuse is that I think MEX should be scrapped from the language anyway. I don't see in what real situation this could come up, but I understand your objection. One possible way to deal with that (I admit not a very elegant one) is to make only PA numbers have this mass/unmass property, not quantifiers in general. >>Perhaps we should start another thread about this. Can you explain >>what criterion was used to determine which places were passible >>of sumti raising and which weren't? For example: > >No, I can't (at the moment). I've no doubt there is plenty of fudging >and unexamined cases in the allocation of +kamsucta to terbridi. I am >talking about the principle. My argument is that we haven't dealt with the issue in a satisfying manner. The places marked in the gismu list as specifically abstraction only seem arbitrary. There are places marked explicitly as "object", there are places that specifically allow both, and there are places without any mark at all. And as far as I can tell there are no guidelines to identify when sumti raising is a problem and when it is not. In my experience, I often find myself using objects in places where I know they are specifically forbidden because using tu'a would make the sentence too confusing. co'o mi'e xorxes