Truth/satisfaction conditions

[Logical Language Group <lojbab>]

la bob. pu cusku di'e

> > If you do not eat the apple, it is false.  It can only be true if you
> > do eat the apple.  Hence, the imperative is true if and only if the
> > command is carried out.

la .and. cusku di'e

> This seems to contradict what pc recently said. I can't remember
> his exact terms, but it was something like that imperatives have
> "satisfaction conditions" rather than "truth conditions".

Well, yes and no.  Satisfaction is a more general notion than truth: it
applies even to open sentences.  For example, the open sentence:

	A is the mother of B

is satisfied under the set of bindings {A=Marianne Cowan, B=John Cowan},
and:

	2 + 2 = x

is satisfied under the set of bindings {x=4}.  "True" may be interpreted as
"satisfied under the empty set of bindings".  So it may do less violence
to language to say that commands are satisfied, rather than true, iff they
are obeyed.  (There is a pathological case hiding in the woodwork: if I
order you to "head" a football, thus:

	le stedu be ko cu lamji le jmabolci
	the head with-body you [imperative!] is-in-contact-with the foot-ball

and someone else throws the ball at your head, is the imperative satisfied?)

> What are the truth conditions for Lojban wh-questions?

By similar reasoning, a question is true (or satisfied, at least) iff a
correct answer is given.  This also applies to Y/N questions.

The utility of this feature is to make logical connections of imperatives
and/or questions reasonable, as someone else has pointed out.

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