Re: Nested preposed relative clauses

[John Cowan <cowan@locke.ccil.org>]

> The problem is in the rule for sumti_tail_111. Two of the rules
> referred to sumti_G_97 in grammar.235 - here they refer to F_96.
> Replacing F_96 with G_97 will remove the reduce conflicts.

Thanks.  That solves the problem.

> When I read Colin's comments concerning nested relative clauses,
> it occurred to me that a different type of preposed relative
> clause would be possible. If we add the following rule to the rules
> for sumti_F_96
> 
>     | quantifier_300 pre_rel_clauses_124 sumti_G_97
> 
> we get a preposed relative clause which behaves like an outer
> quantifier:
> 
>     ro poi klama le tcadu ku'o le nanmu
>     (all who come to the city) of the men
> 
> The nested case has explicit gadri at all positions:
> 
>     re poi klama le tcadu pe'e le nanmu se viska ku'o le nanla
> 
> This type of clause has the further advantage that it can be preposed
> to any type of sumti:
> 
>     pa poi klama le tcadu ku'o do 
> 
> The leading quantifier makes it unnecessary to introduce any new cmavo.
> Most often, the quantifier would be just the implicit outer quantifier
> made explicit. Many people might find this construction more natural
> and easier to comprehend than the present one.

Are you proposing this as an alternative to the current kind of preposed
relative clause, or as a replacement of it?  (I would tend to favor the
latter: two kinds of preposed relatives seems excessive to me.)

Note, also, that we have "LAhE/NAhE+BO pre-rel-clause sumti /LUhU/"
as a valid type already under Change 20, so "je'abo poi klama le tcadu
(pe'e le nanmu cu se viska) ku'o le nanla" works.  Maybe this should be
the only kind of preposed relative clause for sumti?  (The other occurrences
are in vocatives and in "LA pre_rel_clause CMENE"., which are semantically
different.

-- 
John Cowan					cowan@ccil.org
		e'osai ko sarji la lojban.