Dear Christian-Emil,
This topic is very interesting. With the CRM, with the intention
to standardize, we always try to be on the safe side of what is
undebatable.
I think it is worthwhile to start a general discussion on family
relations, as an extension of the CRM. I agree with your points below.
I see three cases:
a) a biological relationship through birth and genetic fathershood.
All related family relation terms can be modelled as deductions or short-cuts
of relationships through birth.
I think it is advisable in a real applications to use
some short-cuts in the data store. There is a trade-off between the
complexity of paths for grand cousins or whatever, and the reasoning
needed to resolve shortcuts, such as parents being childs of grand-
parents. The denser the database, i.e. the more complete about some
family, the more the CRM solution is advisable.
I think all cases of genetic motherhood without giving birth herself
can be modelled as specializations of E67 Birth and P96 by mother (gave
birth).
P96 can be interpreted as superproperty of the role of the genetic mother
and the
birth-giving woman.
I would also model under these the "pater est" rule, once it is assumed
that this is the biological father. I'd argue that the assumption is more
relevant than reality for our purpose, in the sense that the fatherhood
stays unchallenged
and socially relevant/active until better knowledge is acquired.
Tricky is always how to make a model that is monotonic in the cases
where knowledge about reality is incomplete.
b) Family relations by legal declaration: What you call a "speech act" I'd see
as a legal
act. There is a formal declaration, ceremony or whatever with legal
consequences, and a following state of validity of this relation.
This might be covered by an Attribute Assignment, but this might be
overstretching the concept. I'd propose to investigate that in more detail.
It may be necessary to also include more informal relationships such as
engagement.
May be a thing like a wedding and most other such relation-establishing
events must be seen as simple activities with different roles of
participants?
c) Social relations by activity: People can act as father, mother, lovers,
concubine,
friend, protector, mentor etc., typically in a temporal bound. I think this
is
covered by the CRM and can be specialized down. One could a priori put
the case b) here, and would not violate anything, except for cases where
someone actually never acts as expected.
What do you think?
Martin
Christian-Emil Ore wrote:
Dear all,
We (Unit for digital documentation, University of Oslo) are designing and
implementing
a database for the historic student matricle at the University of Oslo We want to make the student/person database as CRM-compatible as
possible, but my collegue Jon Holmen (who are repsonsible for this project and doing the work) pointed out that CRM is very operational
and biologically focused with respect to the core relations between family members (siblings, parents):
A person is in CRM connected to his/her parents through a birth event. That is in priciple perfect and it is easy to convert the
information about a student's parents in the registry to a birth event connecting the student to the parents. But in more than very few
cases on of the parents can be step parent or the student can be adopted. In a theoretical cultural study sence on may think of a symbolic
birth here (eg. giving the birth event a type "symbolic birth" through the has-type property), but I done like that solution. Instead it is
more reasonable to think of an adoption as a kind of speech act, like marriage, though mostly in writing. Thus it seems more natural to
model the establishment of a formal parential link between persons as an attribute assignment with a special type with a proper shortcut if
the information about this speech act is not available or out of scope.
In principle the relation "father child" has until the DNA-test become available, always been a juridical relation, e.g. the "pater est"
rule used as a default rule to declare the child's mother's husbond as the child's father in many contries and cultures.
Any comments?
Christian-Emil
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Principle Researcher | Fax:+30(2810)391638 |
Project Leader SIS | Email: [email protected] |
|
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 |
--------------------------------------------------------------
