carter on sumti-raising

[lojbab (Bob LeChevalier)]

I apparently missed something key in my explanation of sumti-raising,
which (.ue.ianai) jim hit most squarely on the head with:

>If I get your drift, you ask "what does it mean when a non-abstract
>sumti turns up in an abstract slot"?  You say, "hundreds of brivla
>include vague default predicates and the listener is supposed to drop
>the non-abstract sumti into x1 of the default predicate".  I strongly
>recommend that such default predicates be discouraged.  A predicate
>should mean THAT PREDICATE, not some other relation contextually
>defined via poetic license.  The default predicates are too illogical!
>Particularly, major semantic surgery on the language cannot be
>justified by the (negative) utility of interpreting such non-abstract
>sumti.  Also, it's common for someone to use a colorful non-abstract
>metaphor such as le to'ercnici "the mess" to refer to an event.  How
>can you tell whether the speaker wants the default predicate engaged,
>or the event referent of the non-abstract metaphor to be used as such?
>More common sense?  I see no option other than to take the user's words
>to mean just what he said.

sumti-raising is NOT a proposal to allow people to do something new.
It is a pre-emptive strike against something that already occurs in
Lojban usage, and will surely grow - people will do it because they do
it in their other language.  jimc's last line is key.  In an unmarked
situation, one must be able to assume that the speaker means EXACTLY
what he said, which means that the pragmatic  errors people make will
lead to confusion and non-communication, or more likely that the places
of the place structure will get as fuzzy as English prepositions.

People DO sumti-raise, the proposal is to provide a way to mark it so
that we can more effectively teach people how to avoid doing it
unmarked.  It is a Lojban truism that it is better to be overtly vague
than covertly ambiguous, and we are right now in the latter situation.
Since this is a semantic ambiguity, we merely make it easier to be
precise.  IN real conversation, people will NOT expand what they say in
one word in natural language into a double-nested abstraction like I did
with "try the door".  That would be a hopeless language prescription.
The motivation is to make it easier to at least acknowledge some hidden
structure.

The application of this to cleft place structures is that it makes it
easier to live with uncleft structures when people have the natural desire
in many usage situations, to pull the actor out of the abstraction.  We
can do this with a lujvo too, but at the language bootstrap level, people
seem to be unwilling to make lujvo, and of course are very unlikely to
memorize them.

Summary - we are recognizing an actual usage and marking it, not proposing
a new one.  It just conveniently happens that this usage solves a problem.

>preja loi  cinta) cu galfi le bitmu.  YECCH!  All these "unspecified"
>predicates that get specified by imagination -- they bother me.

They bother me too.  That is why I want the speaker to let the listener
know they are really there.  Perhaps this will lead to them less often
being unspecified.

>From another message from jimc:
>2. When the user naturally wants to put a sumti at main level, but logic
>and/or definitions demand that it be buried in an abstract sumti or
>clause, what happens?  E.g. he wants to say "the cat kills the rat".
>
>    a.  We tell the user to say the sub-bridi with explicit words, e.g.
>        le mlatu cu gasnu/rinka  le nu            le ratcu cu morsi
>        the cat     does /causes (abstract sumti) the rat  is dead
>
>    b.  We tell the user to use a tanru and trust to "common sense" to
>        straighten out the meaning:
>        le mlatu cu morsi gasnu fi le ratcu
>        the cat     death does  (to) the rat
>
>    c.  We promise to provide a suitable lujvo for all the *numerous*
>        combinations of this general kind.  The example is in Old Loglan
>        since I haven't seen any official list of Lojban lujvo.
>        le katma ga mordu le ratcu
>        the cat     kills the rat

the gismu "catra" will serve your need.  There are no 'official' gismu
in Lojban; I believe that the language users, not a potentate
'word-makers-council' should make lujvo.  The lexicographer who writes a
Lojban dictionary can choose to implicitly prescribe by leaving out or
correcting mistaken forms.

>    d.  We provide rules by which regularly constructed lujvo (diklujvo)
>        and/or tanru can be interpreted and transformed into "deep
>        structures" as shown in alternative "a".

You are leaving out:

     e.  Allowing people to do what comes naturally, and use the key
         communicative information from the abstract clause, usually a sumti in
         such a way that it elliptically 'stands-for' the 'sub-bridi'.  This
         raising of the sumti of the 'sub-bridi' to the main level is what I am
         calling 'sumti-raising'.  (It is sumti-RAISING because linguists assume
         that our brains work at the deep-structure level, and transform that to
         surface expression forms when speaking.  I offer no opinion on this
         theory - but chose a name that linguists would accept.)

But this usage must be marked, else the logical-structure is obscured.  Hence
"tu'a".

lojbab