Re: log:forSome (Was: Model-specific identity for anon resources, and its representation: A new issue?

At 02:53 PM 6/21/01 -0500, pat hayes wrote:
>Existentially quantified variables do not bind to multiple values either.

Hmmm, there's something I'd like to better understand here.

Consider something like:

    child(fred) & ( exists(Y): parentOf(Y,fred) )

Now, as I understand, Skolemizing this would yield something like:

    child(fred) & parentOf(Y,fred)

I guess that you're saying that in this expression Y binds to a single 
instance of fred's parent, even though we expect there to be two possible 
such bindings.  So the logical expression does not, of itself, uniquely 
define the binding, even though there may be only one in any instance of 
the expression.

Some problems, it seems to me, call for enumeration of all the possible 
bindings (including problems of the kind RDF is likely to be used 
with).  E.g. Who are fred's parents?  Should I conclude that such problems 
cannot be fully specified within the framework of FOL?

#g


------------------------------------------------------------
Graham Klyne                    Baltimore Technologies
Strategic Research              Content Security Group
<Graham.Klyne@Baltimore.com>    <http://www.mimesweeper.com>
                                 <http://www.baltimore.com>
------------------------------------------------------------

Received on Friday, 22 June 2001 08:35:43 UTC