re:Re: SUO: On Clay and Vase
Pat Hayes <phayes@ai.uwf.edu> wrote:
>
>>Dear John,
[...snip...]
>>So normally the vase is a sub-state of the piece of clay. Whilst the piece
>>of clay is also the vase the two are of course co-incident.
>>
>>I'm sure you know this, but others may not.
>
>Yes. And if we consider things which consist(in part) of liquids,
>things get even more complicated. For example, Lac Leman is
>considered to be an object in most accounts of geography and by the
>people who live on its shores. But one can also view it as simply a
>wide part of the Rhône (which flows in at one end of the Lac and out
>at the other). There are two ways to cut up the space-time history of
>the lake: one treats it as geographically fixed, with water flowing
>through it; the other identifies particular mereological sums of H2O
>molecules (and assorted detrius, dissolved gases, etc.) which are
>moving through space, ie sloping in space-time. These different kinds
>of history can only coincide for a moment. They are like two
>different spatiotemporal textures drawn at an angle to each other,
>each set of lines covering the entire space, but never parallel.
>(Either of them might qualify as a continuant, by the way; but they
>couldn't both be continuants at the same, er, time.) In order to
>think about the lake with the proficiency of an ordinary Swiss bloke
>(and probably with that of ordinary Swiss fish, for that matter) one
>needs to be able to reason with both kinds of history; and if these
>are both 'things', then they are things that coincide only for a
>moment anywhere.
>
### Pat I don't see the point. What you write about Lac Léman is true of everything, or should I say every Thing. As Continuants, Occurrents, Endurants, Perdurants and what have you are subclasses/subtypes of Thing/Entity/Object/..., and subclasses are not necessarily disjoint (otherwise lotsa people need to be outta here :-) ), their instances in the extension of one's ontology may have "simultaneously" different properties. Which ones are "current" depends on context, user profile, application, moment in history, ...
. As we don't have a query syntax for the Upper Ontology, which conceivably would allow to express linguistically and explicitly a change of context, user profile, application, ... by going "up and down" the Great Tree of Knowledge (cfr. jacket of John Sowa's KR book :-) )--so there seems to be no point in even *trying* to reconcile such "views" on "concepts" --and pardon the frequent quotes, a sure sign of the overloading of the terms involved.
### Sometimes -while this is great fun of course- one could get the impression this group is busy recapturing all of human knowledge in a single framework. In this respect I am learning really a lot here! But. I think for the cause at hand it could be more practical to come up *first* with an "ontological base set" (call this an "ontobase" if you want) of concept relationships (such as taxonomic and meronymic ones), organized in or circumscribed by contexts, then add to this further "linguistic" relationships, and *only then* I repeat *only then* start to define rules, constraints, etc *within* these contexts. If such rules turn out to be more generally applicable, they will just percolate upwards in the ontology (I conjecture *very very* few will --this discussion perhapsproviding empiric evidence).
--Robert Meersman
--Robert Meersman