From @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET Mon Mar 1 04:02:44 1993 Received: from YALEVM.YCC.YALE.EDU by MINERVA.CIS.YALE.EDU via SMTP; Mon, 1 Mar 1993 04:02:42 -0500 Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2) with BSMTP id 5533; Mon, 01 Mar 93 03:59:04 EST Received: from CUVMB.BITNET by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP id 1710; Mon, 01 Mar 93 04:04:55 EST Date: Mon, 1 Mar 1993 09:00:43 GMT Reply-To: I.Alexander.bra0125@oasis.icl.co.uk Sender: Lojban list From: I.Alexander.bra0125@OASIS.ICL.CO.UK Subject: TECH: RE: Goats' legs and counting X-To: cowan@snark.uucp X-Cc: lojban@cuvmb.cc.columbia.edu To: Erik Rauch Status: OR Message-ID: TECH: leg of goat mi pu cusku di'e > > [W]hat does > > > > 5) pa nanmu cu litru pa dargu > > > > mean? "There is exactly one man who travels-via > > exactly one road"? Ever? la djan. cusku di'e > No. There are elliptically omitted sumti here. > See below. > > "The predicate {litru} is satisfied by exactly > > one man as x1 and exactly one road as x2, > > (by some (unspecified) means/vehicle x3, > > at some unspecified time and place)"? > Yes. Oops. I intended this to be an expansion and clarification of the previous version, with the added information about unspecified quantities being purely incidental information. You've obviously interpreted it differently. > Unspecified sumti are quantified in an unspecified way. > ... > Since unspecified sumti have vague quantifiers, > it doesn't matter where they go. At first I didn't know _what_ to make of the former statement, until I realised that there are two obvious candidates, existential and universal quantification, and it could just be ambiguous which is intended. Then I re-read the latter statement, and now I'm totally lost again. sarcu falenu mi pensi de'u > > How might this work in predicate calculus itself? > > I would probably say > > > > 6) E(x, y: nanmu(x) & dargu(y) & litru(x, y, ?) & text(x, y)) > > > > where text(...) is the succeding narrative in which > > and refer each to a single individual. > > I could of course mirror this in Lojban as > > > > 7) da po'u pa nanmu de po'u pa dargu zo'u tu'e > > da litru de .i li'o tu'u > > > > but I can't imagine myself going to this extreme > > very often. > Example 7 means rigorously the same thing as Example 5, modulo further > use of "da" and "de". Aargh*** I was relying on the default {su'o} quantifier on {da} in (7), but I think you're rightly pointing out that {da} is in this context restricted to be {pa nanmu}, so any such default quantifier is irrelevant. Hmmm. I also suspect that you wouldn't translate (6) as (7), but probably as 12) da poi nanmu ku'o de poi dargu zo'u tu'e da litru de .i li'o tu'u where {da} and {de} don't so much stand for a single individual each, as for a number of entities, for each of whom as an individual the bridi is true. This is a less familiar-to-me interpretation of (6), but I suppose it's equally valid. > BTW, you need not use "tu'e" to extend the scope > of the prenex: a prenex has scope until the next prenex. (The grammar > is misleading on this point; attempts to fix it ran into insuperable > problems with embedded sentences). Yes, I was just playing it safe. ta'o do you really mean the next prenex of any kind, or the next prenex which mentions {da} and/or {de}? > On a tangential point, "roroi" and "fe'e roroi" probably > don't mean what we think they mean. Technically, they mean "at all points > in time (resp. space) within the whole of an unspecified interval". There > is no reason to suppose that the unspecified interval is necessarily all > of time (resp. space). That's interesting. I think subconsciously I had a suspicion that we weren't using these quite right, but you've put your finger on it where I couldn't. > Yes, Examples 8-10 work. But I think they are painful. There are always likely to be some things that are painful to express in precise detail. But then we generally only have to use that form of expression in exceptional circumstances. We have other more colloquial ways of putting things in day-to-day use of language. It's still helps to use the more formal expression occasionally for explanatory purposes. > "no'a" repeats the selbri from one level up. That's useful if it does. I was working from the cmavo list, which says "next outer bridi". > You are absolutely right; the statement may be true only at a particular > time and place. But it is >exactly< this vagueness about time and place > that saves your Example 5! > > "pa nanmu cu litru pa dargu" means that exactly one man traveled along > exactly one road AT THE TIME AND PLACE I HAVE IN MIND." There is no > problem with such an assertion being veridical. zu'unai la .iVAN. pu cusku di'e > > In any case, if I have in mind two > > legs by which the goat is belegged, it shouldn't matter whether it is > > belegged by something else as well (by two more legs, as it were). .i la djan. cusku di'e > It matters because quantification is veridical; it is not associated with > in-mind-ness. If you assert that there exist exactly two things which beleg > the typical goat, you are precluded from asserting that there are more things > as well. It seems to me there's an inconsistency between these two statements of yours, John. If the man travels the road at the time and place I have in mind, why shouldn't the goat have legs at the time and place I have in mind? mi'e .i,n.