[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >


[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Logical Structure vs. Syntactic Structure



John Cowan <jcowan@r...> wrote:

> The grammars are not intended to produce "logical scope" parses.
> Scope is defined in the prose part of the definition.

This is not quite clear to me.  If something is "defined in 
the prose part of the definition", why would it not be 
considered to be produced by the grammar?  Would it not seem
more accurate to say that the grammar produces two parses,
a syntactical one and a logical one?



> > In other words, the grouping and precedence of logic and 
> > grammar simply don't line up.  What is troublesome to me is 
> > not working out some trivial example such as <ro nanmu cu 
> > prami lo ninmu>, but rather what happens when we start trying 
> > to represent much more complex expressions.
> 
> In fact this parses as (ro nanmu) (cu prami (lo ninmu)), but this
> does not mean that "lo ninmu" is bound more closely to the
> predicate than "ro nanmu".  The deductions you are trying to draw
> are simply illegitimate.

Well, perhaps I misunderstand something, but it appears to me
that the parse you give here directly contradicts the one 
implicit in Chapter 2 of the RG, which is the way I had analyzed
Lojban grammar up until now.  Chapter 2 shows a tree in which
the parse of the above example must be given as:

(cu prami (lo ninmu) (ro nanmu)) if we adopt a convention that
places the header first, or just as well as:

((lo ninmu) cu prami (ro nanmu)) if we opt to maintain the 
word order.  Either way works fine.

In no case can we get ((ro nanmu)(cu prami (lo ninmu))), 
assuming that the predicate always governs its arguments,
as I understand they do in Lojban.

Interesting however, if ((ro nanmu)(cu prami (lo ninmu)))
*is* indeed the correct parse, I really have to wonder why 
you would *want* to argue that <lo ninmu> is not more closely 
bound to the predicate.  First, such an assertion is just plain 
wrong, scopewise; second, insofar as it resolves the whole 
logic-vs.-syntax confusion nicely, why would you reject it?

BTW, to be clear, I am not trying to argue *for* anything at 
this point.  I am strictly engaged in a fact-finding mission,
which means trying to understand the issues and points that 
And and pc and yourself are raising.  This will be an ongoing
process for me until I get up to their and your speed.


Regards,

---   Mike