SUO: RE: A proposed SUO content outline
Pat,
See my response below.
-Ian
> >On the basis of Pat Hayes' and Chris Partridge's comments, the classes
> >'Occurrent' and 'Continuant' were replaced with the classes
'OccurrentType'
> >and 'ContinuantType', which are classes of classes of occurrents and
> >continuants, respectively. This change was motivated by the fact that
the
> >distinction between occurrent and continuant is more stable with respect
to
> >type descriptions than with respect to individual things themselves.
Note
> >that 'OccurrentType' and 'ContinuantType' are both immediate subclasses
of
> >'Entity' in the merged ontology.
>
> Ian, I am left somewhat at a loss why you think any comments of mine
> would have prompted this structure. In fact I really don't quite
> follow what it is supposed to mean, I confess. What would be an
> example of an occurrent type, for example (as opposed to an
> occurrent)?
>
I quote the following exchange between us (lifted from a message from you
dated 2/6/01). You said there that you thought the type level distinction
between occurrents and continuants was an improvement, and, as I read you,
you implied that the only thing that was problematic about the new
distinction was an implication that it was exhaustive. In fact, you went on
to sketch two possible axioms involving these two new classes
'ContinuantType' and 'OccurrentType'. To accommodate your reasonable
criticism, I removed the exhaustiveness condition on these classes and
assumed that you were on board. Please let me know how I have
misinterpreted you here. I really do want the merged ontology to refect
everyone's comments and criticisms.
Pat> The continuant/occurrent distinction, while popular, has many
Pat> problems. Many things can be seen in either way, and indeed for many
Pat> ordinary reasoning processes *must* be thought of in both ways: an
Pat> ocean wave and a flame are two of my favorite examples, but others
Pat> include weather phenomena, many chemical and biochemical processes,
Pat> and various dynamic aspects of working industrial machinery
Pat> (especially those involving moving liquids, such as combustion or
Pat> hydraulics). If we are forced into a rigorous high-level distinction
Pat> between 'things' and 'processes' as mutually exlcusive, then this in
Pat> turn forces our practical ontologies to make many artificial
Pat> distinctions which just get in the damned way, eg between a
Pat> cloud-as-a-thing and a cloud-as-a-process, or between things and
Pat> their lifetimes in a 4-d ontology. There is no need to do this, other
Pat> than to satisfy a rather naive philosophical doctrine which is based
Pat> on very shaky linguistic intuitions.
Ian> Well, we could distinguish ContinuantType from OccurrentType in much
the
Ian> same way that Cyc distinguishes StuffType from ObjectType -
ContinuantType
Ian> would be a class containing all the classes of continuants and
OccurrentType
Ian> would be a class containing all the classes of occurrents. The
advantage of
Ian> this sort of structure is that it allows us to maintain the
disjointness of
Ian> the high-level classes - there is no overlap between ContinuantType and
Ian> OccurrentType. Furthermore, we can accomodate the observation that
some
Ian> things are, e.g., process-like in some respects and object-like in
others.
Ian> Furthermore, this approach would embody your general suggestion above
about
Ian> categorizing aspects of things rather than things themselves, since
classes
Ian> (or ways of describing things) are categorized as
continuants/occurrents,
Ian> rather than the things themselves.
Pat> Yes, this would be an improvement and would be workable, but I think
Pat> it isnt terribly helpful. At least it should be possible to have
Pat> classes that are in neither category.
Pat> In the other message I sent you about temporal formalisms I noted how
Pat> the occurrent/continuant distinction can be thought of as two ways of
Pat> splitting a 4d thing into pieces. If we regard Types as
Pat> classifications of predicates into those that assume one or another
Pat> kind of 'part' divisions (temporal or spatial) then I think this will
Pat> be the most useful way to think about them. Thus for example a
Pat> Continuant-Type predicate would have the property that it if is true
Pat> of something with a temporal extent and if it is inherited by
Pat> parthood, then it will be true for all the spatial parts of the thing
Pat> during that interval; an Occurrent-Type predicate would be inherited,
Pat> on the other hand, by temporal parts. Examples might be respectively
Pat> being made of a certain material, and happening in a certain place.