RE: [lojban] Preliminary notes on indirect questions

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

pc:
> So, I am starting from scratch, mainly to keep the rust from getting too
> thick and because Nick and And have more important tasks.

(Yes, but I'm doing lots of thinking about Lojban instead ;-)

> 1) I don't know how deep the identity of interrogative and relative pronouns
> goes.  In I-E it seems present from the get-go, but I don't know how much
> farther back it goes.  From what little I know about languages beyond the
> wildest grasp of Nostratists, the question sems not even to make sense there
> -- relatives and questions seem to be handled by totally different devices,
> neither really pronouns.

A Derzhanski question, this one.

Anyway, the logic of relatives is unproblematic, yet it doesn't seem to
help much in understanding interrogatives. (If anything, it hinders, by
sowing confusion.)

> Anyhow, in English they are often hard to
> distinguish (indeed, rarely unquestionably so), so, before we try to find an
> indirect question form for some expression, we might be well advised to try a
> relative clause one instead.  It works amazingly often.

I'd like to see some examples, because I can't think of any at all.

> And, in Lojban, it
> has the advantage of giving something more obviously grammatical than some of
> the offerings involving {makau}.
>
> 2) The standard theory (not the only one nor necessarily the most
> satisfactory in some particular respects, but the one all others start off
> from) is that questions are sets of answers, claims, of some sort.  The
> various theories start to diverge on what sort of answers get in, but for the
> most part the set cannot be limited to the correct answers (else the puzzle
> about wondering and doubting -- though these can be dealt with).  It is also
> standard that the referent of a thing in indirect context is the its direct
> sense.

This may be so, but it's easier in some ways to derive direct questions from
indirect questions rather than vice versa ("What is the time?" = "I hereby
I ask you what the time is").

> Thus, the referent of an indirect question is either the sense of a
> set of claims, a property of claims, then, or a set of senses of claims, a
> set of propositions, then.  Lojban seems to go for the latter, though some
> moves tend to suggest the former as well (and the relation between the two is
> so unclear as to leave the question of a middle ground open).
>
> 3) If an indirect question refers to a set of propositions, then the standard
> way of talking about it, {le du'u}, is probably wrong in most cases, for the
> speaker often does not know which proposition the knower, for example, knows
> and, since this is oblique context, the various possibilities are not
> reducible to a single archetype.  John may be the tallest boy in the class
> and the Suzy know that the tallest boy in the class went to the store but
> that does not mean that, when she knows who went to the store, she knows that
> John did -- for she may be unaware or indifferent to the fact that John is
> the tallest boy.  So, it is safer to go with {lo du'u} "some apparopriate
> answer" (correct in the case of {djuno}, say).

I'm always glad to see an attack on excessive le-usage. But also interesting
here is the idea that a du'u selbri can denote a (nonsingleton) class of
propositions. I'm 75% agnostic and 25% skeptical about this, but 100% keen to
hear further justification for the idea.

--And.