From lojbab Fri Oct 7 02:29:51 1994 Received: from access4.digex.net by nfs1.digex.net with SMTP id AA19750 (5.67b8/IDA-1.5 for ); Fri, 7 Oct 1994 02:27:06 -0400 Received: by access4.digex.net id AA09626 (5.67b8/IDA-1.5 for lojbab); Fri, 7 Oct 1994 02:26:57 -0400 Date: Fri, 7 Oct 1994 02:26:57 -0400 From: Logical Language Group Message-Id: <199410070626.AA09626@access4.digex.net> To: WEISS@UCLAUE.MBI.UCLA.EDU Subject: Re: Sumti raising Cc: lojbab@access.digex.net Status: RO Actuallly this is a good message for the List, but I'll answer briefly - ask on the List for more details. The general term in linguistics is 'object raising', and it seems to be endemic to natural languages. What we tend to do is take one object in a subordinate clause, and let it 'stand for' that clause in the sentence without giving the other objects and sometimes even the verb. At the conclusion of a meal: "I am done" standing for "My eating of the meal" is done. In reality "I" am 'done' probably when my life is over, or never if you have a religion that involves eternal life %^) In the recent discussions, the debate has been over whether "He needs a book" is really a raising of "He needs 'he possesses a book' to be true." or "He needs to have a book" in more colloquial English. Lojbanists are reluctant to say that these mean the same thing: "A book" is not the same as "the possession of a book" in any logical framework. This is obvious because you cannot substitute the expression in other sentences that use "the book" (e.g. 'I read a book'), so it is really a feature of the predicate "need". Lojban is trying to eliminate the transfer of semantics from the predicates to the objects, at least by requiring that transfer to be marked. Because 'raising' IS so endemic to natural languages, and because it is often very clumsy to phrase the intent in a clear unraised sentence, we have put in a marker that says more or less "I am sumti raising here - deal with it, or ask if it isn't obvious". This marker is the word "tu'a". There are other side effects of raising that affect natural languages and Lojban alike. A classic is the NRA's response to gun control advocates: "Guns don't kill people; people kill people." But when you recognize raising, the more correct statement is "People firing guns is what kills people". You then get into a situation where all causality is between events - events cause events. This is clumsy to express in natural languages, but only mildly so in Lojban and then mostly because it is wordier. But the language becomes pervaded with flows of events, rather than objects being related. At some point, though, dealing with such abstractions starts to break down. So while it is not entriely possible to eliminate 'sumti-raising', it is something Lojbanists, especially those who are pushing the envelope of the langauge in the List discussions, tend to focus on and argue about (politely of course). Hope this helps. lojbab ---- lojbab lojbab@access.digex.net Bob LeChevalier, President, The Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273 For the artificial language Loglan/Lojban, see ftp.cs.yale.edu /pub/lojban or see Lojban WWW Server: href="http://xiron.pc.helsinki.fi/lojban/"