Message-Id: <199604101936.PAA03160@locke.ccil.org>
Date:         Wed, 10 Apr 1996 14:23:52 -0400
Reply-To: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      *Re: {ti} (was: Re: *old response to And on fuzzy proposals)
To: John Cowan <cowan@LOCKE.CCIL.ORG>
Content-Length: 1137
Lines: 29

Content-Length: 1106
Lines: 26

Content-Length: 1074
Lines: 23

>From: ucleaar <ucleaar@UCL.AC.UK>
>Subject:      Re: {ti} (was: Re: *old response to And on fuzzy proposals)
>
>> >{dei} & co are more precise, but I don't see why {ti} can't point to an
>> >utterance.
>> How do you know what "ti" refers to.
>
>It has to be something proximate to the word {ti}. Within that constraint,
>reference is established in the usual way. But I appear to be taking too
>broad a view of the meaning of {ti}.

Well, we specifically put in all these alternatives to deixis like
"di'u" and "dei" and "nei" "no'a" and "ri" (most of which could be seen
as a plausible referent for "ti" in a typical textual reference
pragmatics situation) for a reason.  You can coin examples where "ti" is
so limited that which of these is intended is unambiguous, but you can
just as easily coin examples where "ti" is highly ambiguous.

As a result - for this and many other situations where English idiom can
cause ambiguous results, we will be far better in teaching Lojban to
simply NOT teach "ti" ever to refer to a textual object.

lojbab