[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
--- In jboske@yahoogroups.com, "Jorge Llambías" <jjllambias@...> wrote:
>
> On 9/26/06, John E. Clifford <clifford-j@...> wrote:
> > --- In jboske@yahoogroups.com, "Jorge Llambías" <jjllambias@> wrote:
> > > On 9/25/06, John E. Clifford <clifford-j@> wrote:
> > > > I agree that if
> > > > someone says {lo pavyseljirna} in a primary place, I take it that
> > > > there is at least one unicorn in his universe and adjust the
universe
> > > > I am constructing accordingly (or get him to readjust his).
But this
> > > > does not say anything about the critters John wants: in your mini
> > > > universe, John could want a hippopotamus or a centaur without
changing
> > > > anything.
> > >
> > > John could want those things in addition to wanting unicorns, but if
> > he doesn't
> > > want unicorns, then the model constructed in the discourse, which
> > includes
> > > {la djan cu djica lo pavyseljirna} as one of its true sentences,
> > would not be a
> > > very good model (even if internally self-consistent).
> >
> > I wasn't saying that he does not want unicorns, only that he can want
> > things other than unicorns, things that do not appear in the
> > universe.
>
> And I agree with that. We won't know what else he wants one way or
> the other if all we know is the set of truths of the theory of the
model.
My point is that there is nothing in the the model containing only
John and elephabts to keep {la djan djica lo pavyseljirna} (though
that is not the way I would write it in this case) from being a truth
*of the model*
> > So, indeed, he could want unicorns in a model which contained
> > only John and a bunch of elephants.
>
> That I don't quite follow. One doesn't want things in a model.
I guess I don't understand what you mean by "model;" I thought you
introduced a model with three things i it: John, unicorn, and elephant.
> If you
> mean that the sentence {la djan djica lo pavyseljirna} could be a member
> of the theory of a model whose domain of discourse does not include
> unicorns, then no, that sentence could not be a member of that
theory. If
> you mean that his actual wanting of unicorns won't affect the
validity of the
> model (which cannot have any sentence about unicorns) then I agree, it
> won't.
Of course I do mean that "I want a unicorn" can be a member of the
theory of a model which does not contain unicorns: the thing wanted
need not be in the uniiverse in which the wanting is evaluated (the
narrow scope version of the reference).
> > (He could alsosay truthfully in
> > this smaller model that unicorns do not exist -- indeed that there are
> > no unicorns.)
>
> If there are no unicorns in the domain of discourse of the model, then
> {no da pavyseljirna} could be a true sentence of the model, I agree, but
> {lo pavyseljirna na zasti} I would take as uninterpretable in the model.
OK. What does that do to sentences that contain that expression? I
take it that it makes all primary occurrences false (since there is no
referent for {lo pavyseljirna}, its referent is not in the extension
of whatever the predicate may be). Thus all negations of primary
cases are true. Even if you take the expression to be meaningless and
thus its primary occurrences also, their negations may be true, since
the negate is not true. Further, {no da pavyseljirna} entails {lo
pavyseljirna na zasti} (and for any other predicate as well), so
allowing one forces the other.
> We would need to expand the domain of discourse in order to make
> sense of it. If he wants something that is not in the domain of
discourse
> of the model, then the model won't be useful to express that, a new
> model will need to be constructed that includes what he wants as a
> member of its domain of discourse.
And, of course, something like that does happen. A want opens up a
want-world, in which what is wanted is available to be a referent for
the expression of that want. But what happens in want worlds stays in
want worlds; it does not add anything to the original world. Consider
the proof that there is no greatest number: it begins with the
assumption that there is a greatest number and so takes us into a
world where there is such a thing. Once that world is shown to
contain a contradiction, we leave it behind and go back to world we
came from, without bringing tht greatest number along. So with wants:
that take us to a world where that want is satisfied and where the
wanter has what he wants. We can play around there a while, but
generally, once we determine that he does indeed want that thing, we
come back to the world we started from and we leave the stuff in the
want world behind, In discourse analysis, "want" marks a relation
between a person and a world (or two -- we can argue that), which
world(s) are subordinate to the original world, i.e., what is in or
happens in the outer world affects the subordinate world but the
opposite is not the case -- beyond the holding or failing of the want
condition. So, the world with only John and an elephant, "John wants
a unicorn" is true id, among the wishworlds (subordinate models)
generated by John are some that involve having a unicorn and
satisfaction (and maybe none that involve one but not the other).
This is all standard stuff, "forced" by the linguistic data (quotes
because there are any number of ways of dealing with that data, some
of which describe the situation differently -- though always with the
same result as far as things mentioned in wishes; I used this one
since we seem to be doing something like discourse analysis here).