I'm trying to imagine the need for this subject (beyond the need to
use it as an example to clarify things.) As Ross says below, if you
want to express that your resource is ABOUT composers, you have a
subject heading for that. But let's say you want to express that your
resource is about the relator:composers. This is directly analogous to
every FRBR Group 3 entity EXCEPT FRBR:concept. All subjects except
concept are a re-use of other entities, and even ignoring FRBR it's
pretty clear that in theory anything that you identify can be the
subject of some other resource. However, in that case, I see no
problem with:
>> (2) <http://opera.stanford.edu/composers/> dcterms:subject
>> <http://id.loc.gov/vocabulary/relators/cmp>
as long as the topic of http://opera.stanford.edu/composers/ is indeed
the LOC vocabulary term "relators/cmp." Otherwise, you need to use a
URI that expresses the true subject of the subject URI. And that is
clearly not "relators/cmp", IMO.
kc
Quoting Ross Singer <[log in to unmask]>:
> On Thu, Jun 3, 2010 at 6:01 AM, Bernard Vatant
> <[log in to unmask]> wrote:
>
>> Using the property is straightforward, as in the initial example of the
>> thread :
>>
>> (1) <http://dbpedia.org/page/Don_Giovanni>
>> <http://id.loc.gov/vocabulary/relators/cmp>
>> <http://dbpedia.org/page/Wolfgang_Amadeus_Mozart>
>>
>> Now if I want to use the skos:Concept aspect of
>> <http://id.loc.gov/vocabulary/relators/cmp>, it will certainly be in an
>> indexing/classification context, such as
>>
>> (2) <http://opera.stanford.edu/composers/> dcterms:subject
>> <http://id.loc.gov/vocabulary/relators/cmp>
>>
>> (1) and (2) respect the "punning" logic as long as they live at different
>> levels. But one could be tempted to classify either or both resources in (1)
>> under the "Composer" concept, with the logic that anything connected by a
>> "Composer" relationship has something to do with the "Composer" concept,
>> hence entail from (1) the following
>>
>> (3) <http://dbpedia.org/page/Wolfgang_Amadeus_Mozart> dcterms:subject
>> <http://id.loc.gov/vocabulary/relators/cmp>
>>
>
> I think where things get really clunky about reusing reusing the
> relator as a concept is when you realize that LC already *has* such a
> concept: http://id.loc.gov/authorities/sh85029386#concept
>
> and we have http://id.loc.gov/authorities/sh2007002284#concept,
> http://id.loc.gov/authorities/sh85000744#concept,
> http://id.loc.gov/authorities/sh99011240#concept, etc.
>
> What makes *more* sense is to reuse the resources they've already got
> which were intended for such a use and create a relationship between
> the relator resource and the subject heading resource.
>
> -Ross.
>
--
Karen Coyle
[log in to unmask] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
begin_of_the_skype_highlighting 1-510-435-8234 end_of_the_skype_highlighting
skype: kcoylenet
|