On Tue, 5 Jun 2001, Roland Schwaenzl wrote:
> you seem to worry about where to put new stuff in case there is more than one slot.
> Three types of gadgets we currently have:
>
> Top-level Properties/ Elements
> SubProperties/ElementQualifiers + DC defined names for third party maintained ordering systems + relations.
> DC defined ordering systems (dctype1 list, points, periods....)
That is true, but are we sure the division between the first two levels
(above) will completely make sense five years from now? The reasons
for my doubts are reflected in the difference between the concepts
Property/Sub-Property as opposed to Element/Qualifier. One could
reword the distinction between the first two namespaces above as
follows:
Namespace 1) Top-level properties
Namespace 2) Sub-properties of top-level properties, plus encoding schemes
Given the equivalences above, if an Element Qualifier is a Sub-Property
of another Property, it is also by definition a Property
("sub-properties are also properties") and therefore also an Element.
By this logic, a newly-approved Property would go:
1) into Namespace 1 as an Element (if that element is
considered to be "top-level");
2) into Namespace 2 as an Element Qualifier (if it is a
sub-property of an existing top-level property); or
3) into neither namespace (if it is neither a top-level
element nor a refinement of a top-level element).
Moreover, we have been loosely using the notions of "top-level",
"core", and "cross-domain" as if they were all the same. However,
"top-level" implies (to me) a set of primitives for an ontology, which
(as Carl has frequently pointed out) is not an appropriate way to view
the current Dublin Core. For example, "core" is not "top-level" with
regard to a hypothetical Agent that could someday subsume Creator,
Contributor, and Publisher. And "Core" implies "cross-domain" -- but
not vice-versa! I continue to argue strongly that we should leave our
current Core alone for now and focus on other issues, but clearly there
are some longer-term issues here to be sorted out as our understanding
evolves.
I regard our current Dublin Core as a strawman pidgin -- a very helpful
start. But over time a living pidgin should emerge as the set of
elements most commonly used for interoperability. Over time I would
expect to see some of the properties in the "lower-level" namespace to
become more popular and salient than one or two of the properties in
the "top-level" namespace. Recognizing and ratifying such natural
evolutions in status should not require us to move terms unnecessarily
between namespaces. Nor should the status of a term at the moment it
is added to the DCMI vocabulary be fixed in the namespace URI it was
assigned at the time -- while that status should change, the unique
identifier should not.
Tom
_______________________________________________________________________________
Dr. Thomas Baker [log in to unmask]
GMD Library
Schloss Birlinghoven +49-2241-14-2352
53754 Sankt Augustin, Germany fax +49-2241-14-2619
|