MARC 09X

[jbaxmeyer]

Locally-assigned classification numbers (09X) are not converted.

[rguenther52]

In general we are not converting any local fields. This may need to be discussed.

[kefo]

It will be hard to impossible to predict how institutions used local fields but there should probably be a way to transform this, perhaps as a localNote or, in this case, localIdentifier (there is http://id.loc.gov/vocabulary/identifiers/local ). If libraries cared enough to use local fields, it stands to reason they'll want the information in those fields on the otherside of the marc2bibframe transform.

[robertzero]

When converting MARC to BIBFRAME, 090 and 092 should definitely be retained. They aren't locally made up call numbers, they are valid LC and NLM call numbers that were assigned by an organization other than LC and NLM. An 090 is equivalent to the MARC 050 _4. An 092 is equivalent to the MARC 060 _4.

[danizen]

Why use 092 rather than 060? AFAIK as a software guy at NLM, we maintain only the 060. If that becomes 092 at OCLC "downstream" it is news to me.

On Wednesday, July 16, 2014, robertzero notifications@github.com wrote:

When converting MARC to BIBFRAME, 090 and 092 should definitely be retained. They aren't locally made up call numbers, they are valid LC and NLM call numbers that were assigned by an organization other than LC and NLM. An 090 is equivalent to the MARC 050 _4. An 092 is equivalent to the MARC 060 _4.

— Reply to this email directly or view it on GitHub https://github.com/lcnetdev/marc2bibframe/issues/71#issuecomment-49164032 .

[robertzero]

The "Why" is a very good question and one that I don't have an answer for. However it is true that many libraries did (and continue to) use 090 and 092 to assign perfectly good call numbers. If a record has both an 082 and an 092, then you could convert the 082 and ignore the 092, but if a record only has an 090 or an 092 and you don't convert them, then you lose the call number entirely. If you have access to OCLC, look at ISBN: 1557930201 / OCLC # 35054962. In that record the only Dewey call number is in an 092.

[kefo]

Yeah, the reasons behind "why" could be anything, including cataloger knowledge, system limitation, and even /when/ the item was cataloged.

What I do want to say, however, is this: local fields need to be converted or passed through the transform, independent of another field in the record. For starters, I'm not entirely certain we can assume an 092 is also DDC (the 082). (092 is a locally defined DDC within OCLC's universe, but that is not to say an 092 is so defined in someone else's universe.) Moreover, just like we can never answer "why," there is no way for us to have enough information to make a decision to drop locally defined information. That's just looking for unhappy customers.

[ostephens]

+1 that 09X don't directly map to anything per se, but specific data sources may have conventions. From MARC docs "For interchange purposes, documentation of the structure of the 09X fields and input conventions must be provided to exchange partners by the organization initiating the interchange."

[fallgrennj]

I completely agree with Kevin. At the public library system where I worked, we may have had to fill out a profile when we joined OCLC to explain that on non-fic we used 092 for abbreviated Dewey numbers and that on fic we used 090 for our local shelving convention, but if not, then no one but us knew what we were doing with the 09X fields. Regardless, those fields were necessary to the maintenance and organization of our collection and if they were lost in a data conversion, we'd have been sunk..

To shed some light on 'why?' -- Having been a cataloger in a small public library system and academic institution, the 09X fields are used when those institutions want to assign the classification locally, even if they are using a standard classification system and that standard classification has already been assigned by another institution. It may be that the cataloger doesn't agree with the assigned 'official' classification; or, based on the local collection, s/he may want to place emphasis on another aspect of the "subject' for shelving; or the library may use an entirely different class scheme locally (esp. for fiction); or it may have a policy to shorten the Dewey numbers. So these 09X fields are important locally even if/when an officially assigned class number also is present.

[danizen]

Librarians at NLM who read this list concur that this is completely normal. On Jul 23, 2014 8:46 AM, "robertzero" notifications@github.com wrote:

The "Why" is a very good question and one that I don't have an answer for. However it is true that many libraries did (and continue to) use 090 and 092 to assign perfectly good call numbers. If a record has both an 082 and an 092, then you could convert the 082 and ignore the 092, but if a record only has an 090 or an 092 and you don't convert them, then you lose the call number entirely. If you have access to OCLC, look at ISBN: 1557930201 / OCLC # 35054962. In that record the only Dewey call number is in an 092.

— Reply to this email directly or view it on GitHub https://github.com/lcnetdev/marc2bibframe/issues/71#issuecomment-49868345 .

[kiegel]

The classification portion of MARC 090 is now converting:

090 __ |a QE36 |b .T37 1994

becomes:

http://example.org/99141659680001452classification17 a bf:Classification ; bf:classificationNumber "QE36" ; bf:classificationScheme "" .

Can a classificationScheme be assigned? Something along the lines of what OCLC calls the field: Locally Assigned LC-type Call Number. This would be more useful than the null value now given.