RE: Framing Use Cases (was: Annotating Bibliographic Information)

I have a few remarks and observation pertaining to the recent flurry of messages on this topic.

First, let’s just withdraw the cover art example.  It has been completely distorted in its interpretation and I’ll take full responsibility for that, for not explaining it adequately.  Although I represent libraries and try to understand library requirement I am not a librarian.  Our experts tell me that cover art should be represented as an annotation, but I have to admit, I have never been convinced. That doesn’t mean I don’t believe them, but if I cannot adequately explain it I probably should not have submitted it as a use case and I apologize.   I will substitute a different BIBFRAME use case soon.

Benjamin said  “In my head (at least), I don't feel like a card in a card catalog is an annotation—“

and I am puzzled because I never mentioned a card catalog.  If instead you meant an “online catalog” I’m still puzzled because I did not suggest that a catalog record (or its web equivalent, a resource description) is, itself an annotation.   Perhaps, Benjamin, you could clarify.

Dan said:  “I don’t think annotations have to augment the original source in a way that “enhances” it semantically or provides contextual information.  That is certainly one use of annotation— but not the exclusive one. “

Yes, that’s all I’m saying, it’s one use, certainly not the only use.  I listed three things that annotations are useful for.  Those are the three things that apply in my environment.  I certainly wasn’t suggesting that any annotation has to fall into one of these three categories, only that those are the three categories of interest to my environment.


Dan also said:  “In my mind, there’s no more validity to an annotation that adds to the meaning of it’s target from another that doesn’t”

I completely agree.  I hope nobody thinks that my intention is to impose some sort of requirement that an annotation has to add value to the resource it’s annotation.   On the other hand I hope we don’t preclude annotations that do add value.


Jacob said:  “Just to fog things up a bit; agreeing that a metadata description does annotate the bibliographic resource it describes…..”

We need to be sure we’re using consistent terminology, particularly “bibliographic resource”.  If I have a book and there is a bibliographic record describing it,  the record is the “bibliographic resource”, not the book.   I suspect that you are referring to the book as the bibliographic resource, but if instead I take that at face value – a metadata description of a bibliographic resource – you’re talking about meta-meta data - what we call administrative metadata, although it overlaps with provenance information --  like, when was that metadata created, who created it, and so on.

"….. what then is the proper target for a further annotation of the metadata? For example, let's say I notice an error in a metadata record and craft an annotation that corrects it. Isn't it the case that the bibliographic resource that both annotations describe is also a valid target for the second annotation, i.e., the "correction" effectively annotates both the metadata description and the bibliographic resource?”

So I  am going to consider both possible interpretations of this:

1.      There is administrative metadata for the metadata.  It has been submitted as a provenance annotation. Someone notices an error in it and submits an annotation to correct it.

2.      There is a bibliographic resource (describing a book). Some one sees an error and submits an annotation to correct it.
The first is a whole thread in itself and I don’t want to get too sidetracked so I’ll move on.  The second seems to be to be a general annotation class in itself, a correction annotation.

Bill said: “As for considering a description an annotation: so what about an abstract or a marketing blurb that you want to attach to a document? Those are both descriptions of the document.  ….. I still think they're a valid type of annotation when they are separate from the document itself, particularly when done by somebody not the author of the document. Publishers, for example, certainly want to "annotate a resource with its description," but so might somebody else.

These are valid and important annotatios  for the bibliographic community.

Jacob said: “With an annotation I have to dip into the motivation to know that the relationship between the target and body is A describes B.”

Well I can tell you that the BIBFRAME current plan is to use RDF classes (subclasses of oa:Annotation) rather than motivation.  This, I realize, is (or was, I’m not sure of the current thinking)  a controversial idea, and needs some discussion.

Thanks, all, for the feedback.

Ray