What would be the problem with variable "a", appearing free in Qa, that was previously bound in a sentence "Sa Ra Pa" being implicitly bound by
[la je Ra Pa]
I like my idea but it doesn't work too well with "na".
na sa xrja vfla. na [la xjra] se sme nlveka "It's not so that pigs fly. It's not so that there is something that is a wing of them [pig]"
"Pigs don't fly. They don't have wings."
... gets clobbered with the stronger restriction:
na sa xrja vfla. na [la je xjra vfla] se sme nlveka
"It's not so that pigs fly. It's not so that there is something that is a wing of them [pigs that fly]"
"Pigs don't fly. ?They don't have wings."
... which is nonsensical because we just stated that there are no pigs that fly. Moving the binding operator in front of "na" doesn't help:
na sa xrja vfla. na [la je xjra na vfla] se sme nlveka
"It's not so that pigs fly. It's not so some thing is a wing of them [pigs that don't fly]"
"Pigs don't fly. ?They don't have wings."
... which is true in the actual world by coincidence but is not the intended proposition.