From nicholas@uci.edu Thu Aug 23 16:18:22 2001 Return-Path: X-Sender: nicholas@uci.edu X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-7_3_1); 23 Aug 2001 23:18:21 -0000 Received: (qmail 79613 invoked from network); 23 Aug 2001 23:12:08 -0000 Received: from unknown (10.1.10.27) by m8.onelist.org with QMQP; 23 Aug 2001 23:12:08 -0000 Received: from unknown (HELO e4e.oac.uci.edu) (128.200.222.10) by mta2 with SMTP; 23 Aug 2001 23:12:08 -0000 Received: from localhost (nicholas@localhost) by e4e.oac.uci.edu (8.9.3/8.9.3) with ESMTP id QAA07882; Thu, 23 Aug 2001 16:12:08 -0700 (PDT) X-Authentication-Warning: e4e.oac.uci.edu: nicholas owned process doing -bs Date: Thu, 23 Aug 2001 16:12:08 -0700 (PDT) X-Sender: To: Cc: Nick NICHOLAS Subject: soi vo'a: partial backflip Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII From: Nick NICHOLAS X-Yahoo-Message-Num: 10005 I've had a further think on lenu... soi vo'a, which xod brought up, and I'm doing a backflip. It is clear from my survey of Lojban usage that Lojbanists want a long-distance vo'a. It is also clear that in a couple of contexts, they want it to be short-distance. Those contexts are (a) when the long-distance interpretation is nonsense, because the embedded clause is itself the x1 of the outer bridi (so long-distance vo'a would lead to dumb recursion); (b) soi vo'a vo'e, where you'd have to be a masochist to want long-distance. (Robin, in fact, used vo'a twice on the mailing list: once long-distance --- which is why he was right in the lessons on pointing out that vo'a is long-distance, when I thought he was wrong; and once in lenu... soi vo'a --- where he used it short-distance.) I would prefer vo'a to be unambiguous in all cases; but usage has not, and will continue to not respect that, and it's better to at least encode these usage tendencies as conventions. Moreover, the fact that the cmavo list and the refgramm contradict each other means this is now up in the air; why not take account of usage in cleaning this up? So I propose: * vo'a is by default long-distance * when context overwhelmingly allows it, it can be short-distance instead * That context *should* be syntactic rather than semantic (people are allowed to say semantic nonsense, after all), and the two cases I've pointed out, as illustrated by {.i lenu mi tavla vo'a cu xamgu}, and {.i mi jundi lenu la myl. tavla la nyl. soi vo'a}, should be instances of such contexts. I will post this on the wiki. I will also change the discussion in the lessons of vo'a in embedded bridi into a WARNING: DIRTY LAUNDRY subsection, outlining (briefly!) the various points of view. For the hardliners, as Jay and And have rightly pointed out, there's always {lenei} and {leno'a}/{leno'axiro}. -- == == == == == == == == == == == == == == == == == == == Nick Nicholas, Breathing {le'o ko na rivbi fi'inai palci je tolvri danlu} nicholas@uci.edu -- Miguel Cervantes tr. Jorge LLambias