From lojban@cuvmb.bitnet Tue Apr 02 22:27:52 1996 Received: from punt4.demon.co.uk by stryx.demon.co.uk with SMTP id AA11188 ; Tue, 02 Apr 96 22:27:43 BST Received: from punt-4.mail.demon.net by mailstore for ia@stryx.demon.co.uk id 828403278:08346:1; Tue, 02 Apr 96 01:01:18 BST Received: from cunyvm.cuny.edu ([128.228.1.2]) by punt-4.mail.demon.net id aa08061; 2 Apr 96 1:00 +0100 Received: from CUVMB.CC.COLUMBIA.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP V2R3) with BSMTP id 3387; Mon, 01 Apr 96 18:59:30 EST Received: from CUVMB.COLUMBIA.EDU by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP id 4087; Mon, 01 Apr 96 18:59:50 EDT Date: Mon, 1 Apr 1996 15:44:53 -0800 Reply-To: "John E. Clifford" Sender: Lojban list From: "John E. Clifford" Subject: Re: refgrammar X-To: lojban list To: Multiple recipients of list LOJBAN Message-ID: <828403237.8061.0@cunyvm.cuny.edu> Status: R Some fussy comments on "Nobody.txt" 1. "Exists" talk is liable to create confusions (there is a discussion on one of the logic lists right now about how we can talk about Guernica being painted by Picasso now tht Picasso no longer exists -- being dead and all -- and that is just the simplest problem that arises in these contexts), so I would drop the "there exists" talk in 2.3 ff in favor of "there is" or "for some." Similarly, around 3.5, you talk of "everything that exists" and the like; just drop the "that exists." Same thing at 6.1, where there is in addition a bit about non-existents that just muddles the whole situation further. I don't BTW recommend a discussion of universes of discourse or range of variables or other more correct expressions in these contexts; people seem to work pretty well with the right things for a long time and then, when they do start to raise questions, is the time to lay it out in detail. But I have to admit that using "exist" does make the 8.7-8.9 contrast clear and nothing else seems to work as efficiently. But don't take it too literally. 2. . The "any" in 8.4ff has nothing to do with opacity (and is not even a natural expression in my dialect nor in standard English -- without a contrastive stress). Nor does "any" in English regularly work in the way described, whether universal or particular: "If anyone comes, I'll be surprised" is a more natural pattern and unrelated to opacity. For the same sort of reasons, the "anything" at 9.9 is problematic at least. 3. The explanation of opacity by putting it within an abstraction sumti is okay as far as it goes but does not cover all the cases, as the various threads on the issue over the last couple of years have shown. Nor do the examples suggest how to handle natural language cases -- nor how to spot them for that matter. I think this is an important enough matter to deserve some space (but I think it belongs either in the sumti section or the selbri places sructure section, since quantifiers have nothing special to do with it). 4. While you are running through the moves with negations and quantifiers, two comments: a. The rules with restricted quantifiers are (to xorxes' dismay) different from those for the unrestricted ones, at least when the plenty of the restriction is in doubt (and we have not yet found again the word for the denial of "every"). b. There are a lot of quantifiers other than _ro_ and _su'o_ and they get "inverted" too, so some comments about those may be in order. 5. After 10.11 the discussion has _ga_ and _ge_ (or "both" and "either") reversed. 6. In 6 and 7, _Q broda_ is equated with _Q da poi broda_ (or so). But, a) _Q lo broda_ is not discussed in this connection, either as being the same or as representing _Q da g(something) broda_ . b) Nor is the scoping use of the Q broda/Q lo broda distinction, that & seems to think important, mentioned. c) come to think of it, there is precious little about the use of unrestricted quantifiers to say normal things in this whole chapter. That may be a good thing, of course, give the tendency of some to make these puppies take over where they do not belong, but they do have a legitimate use, especially since you have talked some about "any" 7. The scope note in 14 is not the traditional view in Loglan, where -- based on Peirce -- quantifiers could run on indefinitely, until officially reset by a new-topic marker or (and this may be the reason for the change) requantifying. The new requantifying rule may make for problems anyhow, though it is useful and simpler than what would have been required before, an intervening LUhA, I think it was. But it does make clear long scope cases hard to do and they, the surrogates for pronouns in a lot of cases, are probably more important than the selection sense. At least some more discussion seems called for here (including a good way to do long scopes under the given view). 8. At 8.9, the translation needs the whole of the poi clause spelled out in it. pc>|83