John F. Sowa wrote:
> There is a difference between saying that all the referents
> of a proposition are fully specified (i.e., no indexicals)
> and saying that the proposition has the same truth value in
> all contexts.
> JFS>> If all such languages are able to express the "same" propositions
> >> (in whatever sense of "same" seems reasonable), then the simplest
> >> assumption is to assume that the "language-independent meaning"
> >> is converted to a context-independent form by assigning specific
> >> referents to the indexicals.
> vQ> This is reasonable; I agree, and actually hoped for such an answer.
> > However, in your (fascinating) book on knowledge representation, you
> > write:
> > "
> > McCarthy (1993) introduced the predicate ist(c,p), which may be read
> > "the proposition p is true in the context c."
> > "
> > You do not seem to criticize this, though here propositions appear
> > to be context-dependent in their truth.
> Since McCarthy was using a version of predicate calculus in which
> there are no indexicals, the question of context dependence for
> meaning does not arise. But he definitely did want to talk about
> propositions (or sentences) that had different truth values in
> different contexts.
> For example, consider the following sentence:
> "No roses are blue."
> This sentence and its translations to other languages, natural
> or artificial, have always been true. However, some people
> have taken a gene from petunias that encode the color blue
> and inserted it into some roses. They are now breeding blue
> roses. (01)
Yes. But then, you have an implicit reference to time here; the
sentence "no roses are blue" was true at some time, is not (perhaps)
true now, and may become false again. The 'are', with its present
tense, refers to the current moment, at least it can be read this way.
"no roses are blue" is different, not only syntactically, from "no rose
can ever be blue". The latter was never true, as it appears, after the
genetic manipulations have been done. (02)
The question, again, is about propositions. The sentence "no roses are
blue" was true some time ago, and is false now; but does it correspond
to the same proposition in both cases? Is it the proposition that
changed its truth value, or are we really dealing with two distinct
propositions: 'no roses are blue at t1', and 'no roses are blue at t2'? (03)
Pat argued, as far as I understood, that propositions, on his view or
on the view adopted for IKL, are not contextualized in any way, in
particular, not temporally. That is, a proposition cannot be true at
some time and false at another. The truth of a proposition is not
functionally dependent on any parameter, time included, it is constant. (04)
Note that I do not argue for this or other theory of propositions; I am
just curious, and it seems to be an issue that should not be just
neglected (e.g., for the purpose of precise documentation of IKL, a very
practical task). (05)
> The meaning of that sentence has not changed in any way, but
> because the world has changed, its truth value has changed. (06)
If propositions are context-independent, and propositions are meanings
of sentences, then of course the meaning of the sentence has changed, in
the sense of it corresponding to a different proposition now than then.
The world changes, but propositions and their truth do not -- following
the view discussed above. (07)
> vQ> In the above example from IKL [Dead Osama], it seems
> > that the truth of the proposition *is* taken to be
> > context-dependent, and that it includes an unresolved
> > indexical -- a temporal one.
> The predicate isDead is a special case because people (and
> other animals) don't normally switch back and forth between
> a live state and a dead state. But consider the following
> (isSleeping Osama)
> There is nothing unusual about saying that the meaning of
> this statement is context independent, but that its truth
> may change from one context to another. (08)
As above; if a proposition is fixed to a fact, it cannot be refixed to
another fact -- or would this be what you suggest? If the fact is (was)
that Osama slept at t1, the proposition that Osama slept at t1 was true,
remains true, and will always be true. The sentence, though, implicitly
refers to time, and thus it slides over a sequence of propositions, each
true or false, and each fixed to some point or interval in time. (09)
> And by the way, I revised and extended that discussion from
> Chapter 5 of my KR book in the following two papers:
> Laws, Facts, and Contexts
> Worlds, Models, and Descriptions
> These models allow identical conceptual graphs g1 and g2 in
> different contexts C1 and C2 to have all their referents bound
> in some containing context C. Yet the truth value of g1 in
> context C1 can be different from the truth value of g2 in C2. (010)
Thanks, I'll have a look. (011)
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (013)