From @gate.demon.co.uk,@uga.cc.uga.edu:lojban@cuvmb.bitnet Fri Jun 09 22:05:32 1995 Received: from punt2.demon.co.uk by stryx.demon.co.uk with SMTP id AA3332 ; Fri, 09 Jun 95 22:05:27 BST Received: from punt2.demon.co.uk via puntmail for ia@stryx.demon.co.uk; Thu, 08 Jun 95 04:55:28 GMT Received: from gate.demon.co.uk by punt2.demon.co.uk id aa15547; 8 Jun 95 5:55 +0100 Received: from uga.cc.uga.edu by gate.demon.co.uk id aa29500; 7 Jun 95 21:08 GMT-60:00 Received: from UGA.CC.UGA.EDU by uga.cc.uga.edu (IBM VM SMTP V2R2) with BSMTP id 3774; Wed, 07 Jun 95 16:06:28 EDT Received: from UGA.CC.UGA.EDU (NJE origin LISTSERV@UGA) by UGA.CC.UGA.EDU (LMail V1.2a/1.8a) with BSMTP id 9258; Wed, 7 Jun 1995 14:24:42 -0400 Date: Wed, 7 Jun 1995 13:59:18 -0400 Reply-To: Logical Language Group Sender: Lojban list From: Logical Language Group Subject: pa remna, quantifiers X-To: lojban@cuvmb.cc.columbia.edu To: Iain Alexander Message-ID: <9506072108.aa29500@gate.demon.co.uk> Status: R >> Well, you *can* say that pa remna has one head, two arms, two legs. > >You can say anything you want, but if you say that you are saying >something that is not true, since more than exactly one human have that >number of members. (Unless you want to argue that you mean at a certain >time, in a certain place, but I don't think that's fair.) This bothers me. I haven't thought it out thoroughly, but perhaps it is desireable to have "pa remna" = "pa lo remna" be a subselection from remna which makes no claim about other members of remna. This would be a distinct difference from having "lo remna" = "da poi remna". I'm not capable at the moment of thinking through the implications of this on actual usage, logic, the "any" issue, labels like (-specific, -definite) etc., and someone would probably correctly disagree with me on whatever I said anyway since logical implications are slippery to me (logical soap, right? zo'o). So I'll let you guys tell me why I'm wrong before I try to defend it. Or have you guys gotten to the same point as me? >In any case, I have reconsidered the case of general quantifiers and I'm >now inclined to take your view, which really does seem much more >intuitive. Some examples: > >(1) so'i prenu cu klama so'i da > Many people go to many places. > >(2) so'i da se klama so'i prenu > Many places are gone to by many people. I think these would have more clearly showed your issue if you had said "so'i stuzi" instead of "so'i da". Otherwise, I get distracted by the issues of quantificational logic. Do you intend that your conclusion be the same for "so'i stizu" in the above. >In English, those two mean different things. The most natural meaning >(I think) is: for (1) that each of many people goes to many places, but >since everybody can go to different places, each place might be gone to >by very few people; and for (2) that each of many places are gone to by >many people, but each person maybe goes to one place only. > >What do they mean in Lojban? That depends on how are general >quantifiers to be interpreted. > >I thought {re prenu} was to be interpreted as: "There exists an x that >is a person and there exists a y that is a person and x is not equal to >y:" and whatever was claimed was claimed for x and for y. > >But I think And's interpretation is better: "There is a set of two >persons, such that for every x of that set:" whatever. This sounds exactly like "ro lo re prenu". And yet, somehow I think you are intending to say what I said above. That "re prenu" identifies a set of two people out of all who are people, and makes a claim only about those two. Maybe "re prenu" = "ro le re lo ro prenu" which preserves the veridicality and non-specificity of lo at the time of selection but then seems to make the two people definite thereafter. >(Actually, it has to be supplemented by "and no set of more than two >persons", if the exactness of numbers is to be preserved.) > >This would mean that general quantifiers (almost anything except {ro} >and {su'opa}), really hide one existential and one universal quantifier, >rather than some indefinite number of existential ones. This causes (1) >and (2) to mean different things. But if they were to mean the same >thing, it would be that each of the many persons goes to each of many >places, which is not the most useful meaning. > >I couldn't find a single example with more than one general quantifier >in the reference grammar, so I don't know if there really is a policy on >this. My impresion was that they were supposed to be generalized >existentials, but I may well be wrong. I better let John answer. Even if we had a policy, it would have had to be rethought in the face of the "any" issue and the issue quantification scope of implicit "su'opa", since I suspect the quantification issues of those you call the non-general ones impact those of the general ones. If order does not matter for su'o and ro quantifiers on "lo" then order doesn't matter for the general quantifiers either. I wonder if the question you raise is made fuzzy by the use of fuzzy numbers. How do we compare: ci remna cu se tuple re tuple 3 people have 2 legs? vs. re tuple cu tuple ci remna 2 legs are legs for 3 people? Are the English sentences similar in meaning (I think both have typical meanings that differ, but they are sufficiently ambiguous that in some cases the meanings could reverse; e.g. replacing "2" with "6" in the above - especially the first one.) And: >I of course agree. BUT we must make sure we won't be lacking a simple >grammatical means to say: > > There is a set, X, and there is a set, Y, such that for > every V, V in X, and for every W, W in Y, V goes to W. > >(= your "each of many people goes to each of many places"). > >I tentatively propose that, slightly contrary to what you suggest, this >should be the meaning of > >> (1) so'i prenu cu klama so'i da >> (2) so'i da se klama so'i prenu > >While "For each of many people there are many places that they go to" >should be: > > sohi lo prenu cu klama sohi da > (= ro lo sohi lo prenu) > >That is, {lo broda} is equivalent not to {suho lo [suho] broda} (or to >{da poi broda}) but to {ro lo suho lo [suho] broda}, while {suho broda} >is still equivalent to {suho da poi broda}. > >What do people reckon to this? Sounds like you ended up where I did, but backwards (and you have "lo" instead of "le", which may be not much more than aesthetics). I am of the opinion that "quantifier broda" should be the same as "quantifier lo broda" as it is now, but expanding as you suggest; if you want a "da poi" you say "da poi". Note that something is inherently wrong with your formulation, since it is infinitely recursive: every instance of "lo" expands into a nested pair of "lo" which in turn expands into 4 nested "lo", ad nauseum. lojbab