[crm-sig] Re: CRM version 3.2

martin martin at ics.forth.gr
Sat Jul 21 03:40:45 EEST 2001


Dear Nick,

Sorry for answering not earlier. We had a technical problem to bring
CRM version 3.2 up on the Website. This is now done, so everyone
can follow the discussion.

Here my thoughts:

Nicholas Crofts wrote:

> Dear all,
> I have no objections to the modifications in version
> 3.2 needed to avoid the recursion problem.
>
> However, I didn't find Martin's explanation very clear
> so I shall try to reexpress the notions in my own
> terms - Martin please shout if this is not what you
> intended.
>
> The "Beginning of Existence" and "End of Existence"
> classes are high level classes which are linked
> directly to the top level class E1 CRM Entity. This
> means that they can be linked to instances of
> themselves: the beginning of the beginning of the
> beginning of the end of the end of the end.... i.e. a
> recursion. I'm not clear about why recursive linking
> of this sort is a problem, since the recursion is not
> forced. If every instance in the CRM *must* have a
> beginning, then we obviously have big problems, since
> the regression becomes infinite. Talking about the
> beginning of the end or the end of the beginning, on
> the other hand, is sometimes appropriate.

I think it is slightly different than you pose it."Beginning of Existence"
is the event from which on something exists. So e.g. the Beginning of
the Existence of the Beginning of the Existence of Shakespeare's Hamlet
means, that the "Beginning of the Existence of Shakespeare's Hamlet" itself
is something that has existence in the same sense as Shakespeare's
Hamlet itself. This notion is different from the temporal notion of the beginning
of the beginning of the making of Shakespeare's Hamlet, a notion covered by the CRM.

The utility of the concept "Beginning of Existence" may be illustrated
by the following example:
Assume an old copy of an old book referring "...........". An expert can identify
that as a Linnean name. Linne's invention of this name is a terminus postquem
for the text of this book. So, investigation of things that persist with a certain
identity can yield termini postquem and antequem. Items with such a persistent
existence I have called "Existence". May be "Persistent Item" would be better.
Obviously, the Events and Periods themselves are not such items, which may be
found after some indefinite time, and give rise to the question: How was that
brought into existence?
Therefore I regard the application of an event as a result of another "
Beginning of Existence" event as inconsistent.

So far, we have used in the CRM optional properties, but we have
avoided to the best possible, to attach properties to entities where they are systematically
non-applicable.

How is the recursion avoided? The new range value Existence of
"Beginning of Existence.brought into existence: Existence" is no more superclass
of "Beginning of Existence". So it refers no longer indirectly to itself.

"Existence" generalizes over Stuff, Appellation, Actors, Types. What confirms me
of that move is, that at least the Stuff, Appellation, Actors hierarchies overlap with multiple
superclasses. "Existence" is a common root to them, subordinate to "CRM Entity".
It seems to express a common nature, that allows those concepts to be compatible enough
to overlap.  It seems to me a natural abstraction, connected to a basic reasoning
of historians. In the CLIO system, we had introduced this notion to capture things that are created,
but cannot always be detroyed in a natural way, like ideas, etc.

The only properties, that Existence has, are the incoming links from "Beginning of Existence",
"End of Existence".

Please let me know, if this is more comprehensive. Comments, objections welcome.
If this change is regarded as controversial, we return to version 3.1.

best

Martin


>
>
> Be that as it may, I understand that Martin proposes
> as a solution the introduction of a new high level
> entity 'existence' which breaks the recursion loop.
>
> Martin, can you explain please exactly *how* the
> recursion loop is split by the new entity and where it
> sits in the hierarchy. (Reference to the relevant
> property links... etc.)
>
> Thanks
>
> Best wishes
>
> Nick Crofts (pp Disgusted, Tunbridge Wells ;-)
>
>  --- martin <martin at ics.forth.gr> wrote: > Dear All,
> >
> > Just to avoid an outcry of disgust about the change,
> > I send you
> > the version 3.2 for a fast check. We have corrected
> > further errors
> > in the graphics, two missing declarations of what a
> > short-cut is
> > about ("has section", and "took place on or
> > within"). Note that
> > all short-cuts are explained in this document.
> >
> > Now here the remedy for the recursion problem (also
> > in the introduction
> > to version 3.2):
> >
> > The Agios Pavlos Extensions declared two new
> > abstract entities: "Beginning of Existence" and "End
> > of
> > Existence" as fundamental building blocks for
> > temporal and spatiotemporal reasoning. In particular
> >
> > they support the analysis of the concepts "terminus
> > postquem" and "terminus antequem",
> > fundamental to historical studies. The range of the
> > respective links " brought into existence",
> > "took out of existence" has been declared in lack of
> > a more precise notion as "E1 CRM Entity".
> > This causes a recursion on the concept itself, and
> > on all events and activities below.
> > E.g. a Destruction was brought into existence etc.
> > Therefore the abstract entity Existence has
> > been declared in agreement with the scope note of
> > "Beginning of Existence", which seems to perfectly
> > capture the notion, and avoids the recursion by
> > unwanted inheritance, and the
> > underspecification problem of the range entity. Even
> > though the introduction of a high-level
> > entity may seem to be a big step, it is
> > conservative: No current instances are invalidated.
> > It is further an abstract entity, not intended for
> > instantiation. It's effects are local,
> > i.e. restricted to the immediate subclasses and
> > superclasses of "Existence".
> >
> > The entity Existence "exists" since 1992 in the CLIO
> > system at the museum Benaki.
> >
> > Martin
> > --
> >
> >
> --------------------------------------------------------------
> >  Dr. Martin Doerr              |  Vox:+30(81)391625
> >         |
> >  Senior Researcher             |  Fax:+30(81)391609
> >         |
> >  Project Leader SIS            |  Email:
> > martin at ics.forth.gr |
> >
> >         |
> >                Centre for Cultural Informatics
> >         |
> >                Information Systems Laboratory
> >         |
> >                 Institute of Computer Science
> >         |
> >    Foundation for Research and Technology - Hellas
> > (FORTH)   |
> >
> >         |
> >  Vassilika Vouton,P.O.Box1385,GR71110
> > Heraklion,Crete,Greece |
> >
> >         |
> >          Web-site: http://www.ics.forth.gr/proj/isst
> >         |
> >
> --------------------------------------------------------------
>
> =====
> Nicholas Crofts
> DAEL / DSI
> rue David-Dufour 5
> Case postale 22
> CH - 1211 Genève 8
> tél +41 22 327 5271
> fax +41 22 328 4382
>
> ____________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
> or your free @yahoo.ie address at http://mail.yahoo.ie




More information about the Crm-sig mailing list