RE: [lojban] A or B, depending on C, and related issues

["And Rosta" <a.rosta@ntlworld.com>]

pc:
> a.rosta@ntlworld.com writes: 
> 
>   not "le tcima", but some abstraction -- I don't know which -- containing 
>   "ma kau tcima" or "le tcima cu ckaji ma kau". "depnding on the weather" 
>   = "depending on what sort of weather there is".
> 
> But the sort of weather there is is just {le tcima}, what else would it be? 

If there's rain, then according to you it should be equivalent to
"depending on (there being) rain". But that is not the meaning we
want.

> Thinking of these things as "indirect questions" leads to writing them as 
> indirect questions without any semantic evidence that they have anything to 
> do with questions (or interrogatives or what have you) beyond an English 
> resemblance that may be totally superficial for all the evidence so far 
> presented. 

Let me try to run through the issues and reasoning.

1. "X depending on the weather" means "Outcome1 if the weather has
property1 (or, if thing1 is weather); outcome2 if the weather has
property2; etc.".

2. A number of diverse semantic contexts require us to refer to
different (sets of) values for some property (or different (sets of)
members for some set).

3. We don't yet know a logically explicit way to represent most or 
some or any of (2)-type cases, and we certainly don't have a single
logically explicit way that covers them all.

4. All (2)-type cases can in English be represented by a certain
distinct and identifiable syntactic construction that is standardly
called an 'interrogative clause'.

5. At least some English interrogative clauses are standardly 
translated into Lojban using Q-kau.

6. (3), (4) and (5) combine to lead us (with Jorge originally leading
the way), to represent all (2)-type cases by using Q-kau.

7. Use of Q-kau is a kind of stopgap solution until we work out the
underlying logic of (2)-type cases. Once we have done this, we can
then both work out appropriate ways of representing (2)-type cases
and state explicitly about what (Q-)kau means, which will then
determine where in logically proper usage (Q-)kau is and isn't
appropriate.

In addition, my own view is that:

7. (2)-type cases constitute a natural class, and hence it must be
the case that some generalizable logically explicit solution is
waiting to be discovered. (According to my hunch, it is no 
coincidence, then, that English interrogative clauses have the
distribution that they do.)

--And.