[crm-sig] Editorial Proposal

martin martin at ics.forth.gr
Fri Oct 11 22:29:02 EEST 2002


Dear Nick,

Regarding Matthew's argument
for populations as Actors, and yours as Persistent Items, I have the impression,
that we confront a typical problem of polysemy:


Let's regard:

The city of Heraklion as an aggregate of buildings. I think we can take it alltogether
as Physical Stuff, be it subject to changes as it might be.

In the 10th century, Nikiforos Fokas conquered the city back from the Arabs,
destroyed the city, moved the population inland to Temenos to better protect them
from attacks from the sea, but they resettled some years later in Heraklion.

One may regard, that the Actor continued, but the Physical Feature had a discontinuity.

So there is an aspect of population that, if acting collectively in some sense,
must be regarded as an Actor.

Heraklion has changed names, like : Candia, Chandax, Chandakas, etc.
The initial settlementis not explored or excavated.
The political system changed at least from Minoan, Mycenean, Dorean, Roman, Byzantine,
Sarazene, Byzantine, Venetian, Turkish, Cretan, Greek, German, Greek.

Another aspect is the "mereological sum" of all activities and events in the city.
This aspect is dominant to my opinion, when I say: "This object is Herakliotic",
meaning, that its production is part of the activities in Heraklion. In that sense,
the settlement makes up a Period for me. (in the gerund sense of settlement,
"settling"). We have used this aspect successfully in the CLIO system.

Crofts Nicholas (DAEL) wrote:

 > Concerning Martin's proposal to add settlements and populations to the scope
 > of E4 Period.
 >
 > Martin observes that scope notes should "remind us our common concepts..."
 >
 > The proposal to include settlements and populations as examples of periods
 > does not concur with my common-sense view of what these things are. I would
 > argue that E77 Persistent Item is more appropriate and that it is already
 > intended to deal with the identity questions Martin raises.
 >
 > As I understand him, Martin gives two reasons for regarding settlements as
 > instances of periods. i) that satisfactory identity criteria are difficult
 > to establish and that ii) their existence has temporal limits.


That is not precisely the point. I mean identity criteria others than being
"ongoing", which is the identity criterion for all temporal entities. They are not
found again, they are still ongoing. The may cease to exist without any
violant actions. Young people leave, old people die, the activity dies out.
This is not a normal end of a persitent item. Everything is of course limited
in time, Persistent Items and Temporalities.


 >
 > Taking these point in turn:
 > i) Martin notes, correctly, that the composition of a city may change over
 > time and that it is therefore difficult to use as a criterion of identity :
 >
 > " What is the city? people change, buildings change, borders change.
 > The diachronic identity can neither be Stuff, Actor, [nor] Place."
 >
 > However, the difficulty of providing precise identity criteria based on
 > material composition has already been discussed in the scope note for E77
 > Persistent Item:
 >
 > "The conditions under which an object can be deemed to maintain
 > its identity are often difficult to establish ... material objects in daily
 > use ... undergo material changes due to maintenance etc. without changing
 > identity. Identity in these cases would seem to depend more on *continuity*
 > rather than the presence of any particular physical state or component."
 >
 > I would suggest that the identity criteria for settlements and populations
 > can be established in terms of continuity and that, therefore, they should
 > be regarded as instances of E77 persistent item.


For me identity criteria in terms of continuity are characteristic for temporalities.


 >
 > ii) Martin elsewhere has noted that the fact that a material object has a
 > limited life-span is not a good reason for regarding it as a period: "I have
 > never seen a reasonable documentation structure, that puts
 > objects under events because of their limited life-span... The evidence in
 > data structures is that they are disjoint."

Of course, see above.


 >
 > I would argue that the same reasoning applies to settlements and
 > populations.
 >
 > Broadening the scope of E4 Period to include settlements would entail
 > including other many other similar objects, for which precise criteria of
 > "diachronic identity" are difficult to establish.


For instance? That is interesting.


Let us turn the argument upside down. To a settlement belongs a settlement period.

Actually I disagree with Matthew. the English term Period may be about time, but we
have clearly defined that:
"A period is characterized by a coherent set of phenomena and or manifestations
(explicitly intended or not), which are assumed to have taken place over a
certain space and time." Every Period in the sense of the CRM has an associated
Time-Span, which is what Matthew refers to, if I am not mistaken.

Now, the settlement Period (E4) is the sum of activitities of their people, and other
related events like houses breaking down due to bad construction etc.

Those events seem to be the source for the culturally relevant products and the substance of
its history.

Then we need the people, which are Actor, and may be completely exchanged, as e.g. with the fall
of Constantinople.

And we have the built and transformed substance at the settlement place, which is
Stuff.

The three facets: People, stuff and activities play a semiautonomous role.

I think, there is no more abstract notion to a settlement,
and the above should be sufficient for our purpose.

I rephrase my proposal to make a FAQ out of that.

best,

Martin


 >
 >
 > Best wishes
 >
 > Nicholas Crofts
 > Conseiller en systèmes d'information
 > DAEL / DSI
 >
 >
 >
 >
 > -----Message d'origine-----
 > De : martin [mailto:martin at ics.forth.gr]
 > Envoyé : samedi, 14. septembre 2002 10:28
 > À : crm-sig at ics.forth.gr
 > Objet : [crm-sig] Editorial Proposal
 >
 >
 > Dear All,
 >
 > I propose to add the following to the scope note of Period:
 >
 > Settlements and populations in their diachronic identity are
 > also regarded as Periods.
 >
 > Argument:
 >
 > E.g. "This object is from Heraklion"
 > can have two interpretations:
 > A) It was found within the city limits at the time of finding.
 >      In that case Heraklion denotes a Place, but only, if identified
 >      at a specific instance of time.
 > B) It is a product or trading good of the city.
 >      What is the city? people change, buildings change, borders change.
 >      The diachronic identity can neither be Stuff, Actor, Place.
 >
 > Therefore I regard it as a contiguous set of phenomena, a Period.
 > I have so far not found any conflict with this interpretation.
 >
 > A "community" would be an Actor. I would however not identify the
 > Greek, Sarazene, Byzantine, Venetian, Turkish, and again Greek
 > community of Heraklion as one Actor. But despite discontinuity in
 > will, sef-understood identity and influence, there is continuity
 > in settlement.
 >
 > I think this explanation has been used by us, but not been formulated.
 >
 > Martin
 >


-- 

--------------------------------------------------------------
   Dr. Martin Doerr              |  Vox:+30(810)391625         |
   Principle Researcher          |  Fax:+30(810)391609         |
   Project Leader SIS            |  Email: martin at ics.forth.gr |
                                                               |
                 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/isl               |
--------------------------------------------------------------






More information about the Crm-sig mailing list