Le vendredi 30 mai 2014 12:24:38 UTC+9, Martin Bays a écrit :
As for ways to specify that a {zo'e} is constant: an alternative to your
suggestion of introducing new rules for prenexes would be to pull tricks
like:
su'oi da zo'u ro de broda lo du be da
(which may or may not be equivalent to
lo du be su'oi da se broda ro da
).
A bit long-winded for something so important, though.
Although the truth value is the same as Ax B(x,f), {su'oi da zo'u...} is obviously not Skolemized. The idea of {zo'e} in prenex is intended to be similar to Skolemized form of logic.
Meanwhile, a question. Under these semantics, the second (and only the
second!) {zo'e} in
ro zo'e zo'e broda
depends functionally on the quantifier. But in
ro zo'e ro zo'e broda
it doesn't make sense to say that each {zo'e} depends functionally on
the quantifier on the other. This seems to complicate matters?
Martin
I have no idea about that. Because the quantification is implicit in {zo'e} without outer quantifier, I have a feeling that Lojban users would be less attentive to the span of {zo'e} than that of {ro zo'e}. The difference of their usage is quite distinct, and it would not be very complicated.