Are all differences in definition between OpenWEMI and FRBR really motivated?

[aisaac]

This is a split from https://github.com/dcmi/openwemi/issues/87 and it is a stub. I will try to felsh it out with examples following the discussion in 87.

Some of the OpenWEMI resources, especially the properties, are semantically different from the FRBR ones, obviously. But for others, the definition is different but it doesn't feel clear whether this is really needed.

And maybe in some cases where there is a really needed adaptation of the original FRBR defintion this could be reflected in a way that would be less elegant but clearer about the intention. Something like New OpenWEMI definition of X = [Original FRBR definition for X] + "In addition in OpenWEMI we consider that X is [blah] / X includes [blah]"

[kcoyle]

It sounds like we need a section or appendix that is a comparison of openwemi with frbr,-wemi and more about how openwemi overcomes frbr constraints. This would be mainly for readers who are familiar with frbr, especially people who create library catalog data. They would be thinking: why not just use frbr? I also think we need more than just the class definitions in the primer. I just looked at the original frbr document and there are nice explanations - some of which would not be wrong for openwemi. We used some of the wording from FRBR but that is what some usage board members are wanting to change. Next I'll do a table comparing the two.

[kcoyle]

OpenWEMI and LRM definitions

Element	LRM	OpenWEMI
Work	The intellectual or artistic content of a distinct creation	An abstract notion of an artistic or intellectual Endeavor
Expression	A distinct combination of signs conveying intellectual or artistic content	A perceivable form of an Endeavor
Manifestation	A set of all carriers that are assumed to share the same characteristics as to intellectual or artistic content and aspects of physical form. That set is defined by both the overall content and the production plan for its carrier or carriers	A format in which the Endeavor becomes available
Item	An object or objects carrying signs intended to convey intellectual or artistic content	An instantiation of an Endeavor

A major difference between these is the superclass above WEMI:

• LRM uses Res, "Any entity in the universe of discourse"
• OpenWEMI uses Endeavor, "A creation"

I'm not sure if the "universe of discourse" would include things like buildings, sculptures, statistical data, and other non-"sign-based" creations. OpenWEMI's "A creation" could be construed to be something concrete, and the only other way I can think of that would be to say "Something created" but the "thing" there bothers me, as does "entity" in LRM.

Note also that the Expression definition in LRM is in terms of "signs" which again might be interpreted too narrowly to include non-linguistic, although it seems to be broadly defined via semiotics: "The term “sign” is intended here in the meaning used in semiotics." Therefore the use of "sign" may be broad enough. This use of the term, however, may not be known to the majority of the audience for which the LRM is intended. The OpenWEMI use of "perceivable form" is not in contradiction to the LRM use of "signs."

I'm not in favor of the LRM definition for Item because it uses "object or objects" which to me violates the "item-ness" of the term. It should be, instead, defined as a single instance of the created endeavor being described. It seems that the intention is that an item can have more than one part, but to my mind it is still a single item, and the use of "object or objects" here is less than ideal.

That said, I see nothing in LRM that would violate the OpenWEMI classes if (albeit very unlikely) the LRM WEMI entities would be subclassed to OpenWEMI classes.

[aisaac]

Thanks for the investigations, @kcoyle ! I am not sure that these LRM definitions make the discussion on definitions simpler, indeed. We might as well focus on the difference with the original FRBR ones, no? Or are these LRM definitions supposed to replace the FRBR ones? I'm sorry I'm a bit struggling to appreciate the positions of all these proposals floating around. Unless "LRM" is actually "LRMoo" mentioned in issue 88? In this case your analysis could also be copied there!

[kcoyle]

@aisaac, I used the LRM ones because the library community that follows IFLA standards would consider FRBR superseded. We agreed that we would add a comparison of openWEMI and the library standards for our library community readers. So this is the beginning of an analysis for an appendix to the Primer. I've also found a LRM diagram that changes some of what I thought in the above text. This is getting complicated and long for an issue so I will add it to the wiki.

Yes, I can also add a comparison with the original FRBR. I think that non-library folks still see FRBR as the model, while IFLA folks have made clear that LRM is the current model.

[aisaac]

@kcoyle thanks for the answer! So this clarify that we're in the case "Or are these LRM definitions supposed to replace the FRBR ones?". So let's use the LRM ones as reference point, then. My only problem with that is that I might find the LRM definitions harder to understand than the FRBR ones, but anyway maybe this is a matter of my personal background.

As for "We agreed that we would add a comparison of openWEMI and the library standards for our library community readers": yes, this is why I suggested to copy the comparison into issue #88, which is the placed where I had asked if a comparison could be reported!