TECH.GRAM.PROPOSAL: require KU after free-floating tense/modal

[Logical Language Group <lojbab>]

la lojbab. cusku di'e

> I of course will defer to pc if he disagrees, but I think that in the issue
> of tense and negation attached to the selbri using grammar rules 130/131,
> that both have bridi scope exported to thhe prenex in the order they are
> attached to the selbri.  reroiku, on the other hand, will act like naku at the
> same location as far as scope, if I understand/recall our previous rulings.
> 
> This seems to be the most consistent way to explain what we put into the 
> grammar, including the interchangeability of tense and negation attached
> to the front of the selbri.  At that point we had clearly decided that such
> attachment of the negation would have full bridi scope, and I think the
> interchangeability demands that the tense have similar scope.

I now believe that this is correct, and I therefore propose the following

CHANGE 36:

PRESENT LANGUAGE:

When a tense or BAI particle is not attached to the selbri, and not governing
a sumti, but rather floating freely about the bridi, an elidable KU follows it.
When a NA is similarly floating freely, the KU is required.

PROPOSED CHANGE:

Require the KU in all cases.

RATIONALE:

When both a tense and a negation are present, it turns out to be important
to make the scope clear.  Tenses and negations attached to the selbri have
bridi scope, whereas floating tenses and negations have scope from the
point where they appear.  Thus:

1)	lo tirxe na klama le zarci
	some tiger ~not goes-to the store
	It is false that some tiger goes to the store.
	No tiger goes to the store.

is not the same as:

2)	lo tirxe naku klama le zarci
	some tiger doesn't go-to the store
	For some tiger, it is false that it goes to the store.

for Example 1 has the prenex form:

3)	naku [su'o] da poi tirxe zo'u da klama le zarci

whereas Example 2 has the prenex form:

4)	[su'o] da poi tirxe naku zo'u da klama le zarci

Similarly,

5)	lo tirxe reroi klama le zarci
	some tiger two-times goes-to the store
	On two occasions, some tiger goes to the store

is different from

6)	lo tirxe reroiku klama le zarci
	some tiger two-times goes-to the store
	There is a tiger which on two occasions goes to the store.

However, because the "ku" is elidable, some combined tenses don't mean what
they appear to mean.  Thus "puzu'a cusku" means "expression going on to my
left and in the past", but "zu'apu cusku", with the space before the time,
means "zu'aku pu cusku", and gives the "pu" bridi scope, but the "zu'a"
scope only from the point given, thus:

7)	lo nanmu zu'a[ku] pu cusku 

which in prenex form is:

8)	puku da poi nanmu zu'aku zo'u da cusku

In the case of "pu" and "zu'a", which are effectively singular terms, there
is no problem, but messy things happen with "-roi" tenses and possibly with
some others.  Requiring "ku" means Example 7 ungrammatical without it, and
so requires a clear signal that there are two tenses in use here.

This makes nothing unsayable, but does remove some rather useless and potentially
confusing strings from the domain of what's grammatical.

-- 
John Cowan		sharing account <lojbab@access.digex.net> for now
		e'osai ko sarji la lojban.