Date: Sun, 16 Nov 1997 22:29:18 -0500 (EST) Message-Id: <199711170329.WAA07287@locke.ccil.org> Reply-To: Pycyn@AOL.COM Sender: Lojban list From: "John E. Clifford" Subject: Re: `at least one ' vrs `one or more' X-To: lojban@cuvmb.cc.columbia.edu To: John Cowan X-Mozilla-Status: 0011 Content-Length: 2678 X-From-Space-Date: Sun Nov 16 22:29:20 1997 X-From-Space-Address: LOJBAN@CUVMB.CC.COLUMBIA.EDU BOB: >you have not answered my >question, which is whether {lo} always expands to {da poi}? Gee, I thought I did. Anyhow, the answer is yes, subject to the following point. _lo_ is of itself an article (gadri?) and not a quantifier, so the expansion to _da poi_ is from the implicit default quantifier in front of _lo_. If some other quantifier is actually present, then _q lo broda_ expands to _q da poi broda_ (I think -- I seem to recall that some time ago there was a discussion of this with certain qs suggesting that some of them gave the wrong result here, but I do not remember how that was decided, if at all). bob: >{lo} as defined does not > always refer to all of its referents, whereas {da poi} does. Could you explain what this means, please. Neither _lo broda_ nor _da poi broda_ refers to all the brodas, but (implicitly) to some unspecified ones -- all being one possibility (that _da poi broda_ does not technically refer at all I pass over as irrelevant here -- I think). Both indicate a non-empty subset of the set of aal (presently existing) brodas. Only in this last sense does either indicate all the brodas. I just found your later comments on why you think that "_lo_ does not guarantee to refer to all the referents." But the reason you give is about English glosses (which everyone seems to admit are a problem) and not about Lojban. Further, the claim that "at least one of all those that are" is different from "all those that are," while trivially true (as long as there is more than one whatsis), does not speak to the current issue, since _(su'o) da poi_ is glossed (and means) exactly the same thing (within the limits of its different grammatical function). bob: > {loi} would be the term of >choice when seeing or liking some cats.) I missed this discussion; why is this better than _lo_? Ahah! The negation of _mi nelci lo mlatu_ is not translated "for some cats x, it is not the case that I like x." The particular is inside the scope of the negation and can be brought out (as it regularly is, either explicitly or using "any") only by converting to the opposite type as it passes the negation. So the literal trat is (as expected) "It is not the case that for some cat x, I like x," just what you would get with _da poi_. and like that case it can be converted to "for all cats x, it is not the case that..." I apologize (for myself and everybody else apparently) for jumping over the middle step here and leaving you thinking that 1) the quantifier on _lo_ was somehow outside its context and 2) that it was universal as it stood. Neither is true; it is just that if "some" is false, then "all not" must be true.