[lojban] Closing the discussion on indirect questions

[pycyn@aol.com]

    While the syntax of indirect questions is pretty much under
control with the combinations of <du'u> and <kau>,  the meanings
remain unclear, especially how to expand out into ordinary quantifier
language.  Attempts to do this generally have fallen afoul of the
peculiarities of different contexts in which these expressions occur,
mainly -- but not always -- intensional ones.  
    Even the instant case, "What I have for dinner depends
upon what is in the fridge." was never completely solved.  In this case,
however, it was because of the variety of dependencies that might
arise: sometimes, whether I had it or not might depend on whether it
was there, at other times it might depend on what alternatives were
there to going out and getting it (I got me a hamburger because there
were only brussel sprouts and liver in the ice box) .  And of course,
given that there are a lot of things in the fridge, which one gets picked
-- if any -- seems to reach beyond what can be easily sorted out.  And
then there is the effect of an empty box and the possibilities of no
dinner.  
    One fairly direct paraphrase -- after "The answer to the
question "What is in my refrigerator?" ...."  is "The fact that the
content of my refrigerator is what it is ... " or "the fact that the actual
content of my refrigerator is the content of my refrigerator..."  But the
natural first reading of that, whether with <lo> or <loi> (both plausible
in this case, though slightly different) is the tautologous "the fact that
my refrigerator contains the content of my refrigerator"  and the same
for the "the fact that my dinner consists, of my dinner" or some such. 
We can save this by exporting the sumti, which changes the meaning
even though "the fact that..." is not normally intensional:  "As for the
contents of my refrigerator and the menu for my dinner, the fact that
my refrigerator contains the former "affects" the fact that I eat the
latter for dinner."
    Getting into prenexes suggests a general solution to indirect
questions, related to the "question as set of answers" position in logic.
(what preda) predb = "for x, if preda x, then predb x.   The quantifiers
probably need, we have noted, a relevance condition (as "all" usuallly
does) to eliminate never-occurring cases (the number four, Julius
Cawsar and anchovies in my refrigerator).  In at least some cases, the
negatives also play a role: the absence of some item from my
refrigerator may also affect the menu for dinner.   So the present case
is something like "for every relevant x and y, if x is a food item in my
refrigerator and y is a potential menu item, the fact that x is in my
refrigerator either encourages or discourages the fact that y is on the
dinner menu."   The wishy-washy predicate that attempts to expand a
bit on "affects" needs much spelling out for particular cases and
particular people, but perhaps comes close to the general idea. 
Instantiations, it turns out, are still best done in prenex position,
though in the general case, the move to the interior is not too
problematic: "the fact that there is rutabaga in the refrigerator
encourages that hamburgers be on the dinner menu."
    When intensional contexts are involved, as the naturally
frequently are with indirect questions, the prenex instantiation seems
necessary to avoid the problems that arise about identifying someone
in an intensional context.  If John knows who came and Paul came
then John knows that Paul came, but that last claim is ambiguous, for
John may never have heard of Paul and have him identified in some
totally wrong way (the drummer rather than the vocal), but it will still
be true (the other claims accepted) that as far as Paul is concerned,
John knows that he came, i.e., of the person who is in fact Paul, John
knows that he came (albeit he would describe it totally differently).  In
this case again, negative cases seem relevant -- at least that some list
John gives is complete.  Although there is some claim that being able
to give partial lists are also knowing who came.  
    It is not clear whether this is a complete solution -- or even a solution 
at all
(though no major problems were cited) to the question about indirect 
questions.  Any
further discussion to finish this topic off?   This time 'round.

------------------------------------------------------------------------
Failed tests, classes skipped, forgotten locker combinations. 
Remember the good 'ol days
http://click.egroups.com/1/4053/3/_/17627/_/959723210/
------------------------------------------------------------------------

To unsubscribe, send mail to lojban-unsubscribe@onelist.com