On 5/25/17 8:33 AM, Svensson, Lars wrote:
> Karen,
> 
> On Wednesday, May 24, 2017 3:25 PM, Bibliographic Framework
> Transition Initiative Forum wrote:
> 
>> Yes, it looks like BnF has no need of BIBFRAME, which is fine. What
>> is interesting is WHY they have no need of BF - which is that they
>> already have linking. Which we do not.
> 
> This is extremely interesting. It never occured to me that someone
> could implement BIBFRAME in order to _get_ linking. To me BIBFRAME
> (like MARC or any other transportation format) is a way of
> transporting the data you have. How to get that data is to me a
> totally different (perhaps even orthogonal) problem.

Lars, thanks. My reading of BIBFRAME is not that it is a transportation
format, but is being implemented as an actual systems format. If I'm
wrong about that, I'd love to hear from implementers on their view.

That said, I've generally thought of linked data as a way of interacting
in the public data space, not necessarily as an internal data view,
although there is no reason it could not also be the latter. I think
OCLC shows us this with their  use of schema.org.

US MARC doesn't have the ability that UniMARC (and various local
versions) seem to have to directly connect, for example, a series with
the series members. (I can't find an open copy of the UniMARC standard
so I'm going on rumor here.) As UniMARC seems to show, that could be
possible without a move to RDF, but few (if any?) of US MARC systems are
designed to implement such links. Obviously one doesn't need to move to
RDF to get links between bibliographic records, but I suppose it makes
sense, if you are going to require a significant system overhaul, to go
with the latest technology. Personally, I'd prefer it if there were
multiple proposals for a future bibliographic data format rather than
just one.

> 
>> So the question is not BF or not BF, but would our catalogs benefit
>> from linking? And if the answer is yes, then how do we want to
>> achieve that? Clearly it can be achieved in ways other than
>> migrating to BIBFRAME, so we should be looking at all of the
>> options.
>> 
>> I don't know what the LRM has to do with this.
> 
> LRM is about expressing links, too. We should keep in mind, though,
> that LRM (like FRBR or CIDOC-CRM) is a _conceptual_ model, whereas
> BIBFRAME is an _implementation_ model.

Yes, LRM is about expressing links (in part). But as most proposed
bibliographic models aren't actually implementing the concepts in the
FRBR model, it seems unlikely that they will implement the concepts in
FRBR-LRM. Having read through FRBR-LRM, I actually have some intense
questions about what is proposed there, but no place to ask those
questions. Some of the linking that they propose makes no sense to me.
And while both FRBR and FRBR-LRM are presented as conceptual models,
they also provide some fairly detailed technical specifications, which
blurs the line between concepts and implementation.

Note: Although it may look like BIBFRAME is "FRBR-ish" I would argue
that if FRBR had never been published, BIBFRAME might look very much
like it does today because the bibliographic data model is pretty
obvious - content, carrier, agents, topics. Add the data needed for
inventory purposes and you have the data concepts that can be found in
nearly any bibliographic product of the last century and a half.

kc

> 
> Best,
> 
> Lars
> 

-- 
Karen Coyle
[log in to unmask] http://kcoyle.net
m: +1-510-435-8234
skype: kcoylenet/+1-510-984-3600