MARCAuth 5XX relationships in LC names/ resources, rwo/ resources

[briesenberg07]

Proposals

• Do not add LC Names lookups, which yield resources typed as madsrdf:Authority, to property templates which are for creating relationships to non-WEMI resources such as persons, corporate bodies, families, or places.
• Provide training for Sinopia catalogers on obtaining LC RWO resource IRIs using the browser where such resources cannot be searched using a lookup from within Sinopia.

Details

The lookups currently configured for the property template 'has related corporate body of work' include (along with some other sources) the following LC lookup sources:

• LOC person [names (rwo)] (QA)
• LOC organization [names (rwo)] (QA)
• LOC family [names (rwo)] (QA)

These lookup sources yield resources typed as madsrdf:RWO:

Label: Real World Object
Comment: A madsrdf:RWO is an abstract entity and identifies a Real World Object (RWO) identified by the label of a madsrdf:Authority or madsrdf:DeprecatedAuthority.

Such resources are often assigned additional types:

• For persons - bf:Person and foaf:Person
• For organizations - bf:Organization and foaf:Organization
• For families - bf:Family and foaf:Group

Adam S. requested the addition of two lookup sources, to the 'has related corporate body of work' property template, one of these is the LOC geographic [names] (QA) lookup. This lookup yields resources are typed as madsrdf:Authority:

Label: Authority
Comment: A concept with a controlled label.

The argument for adding an LC names lookup here is the desire to select value resources typed as madsrdf:Geographic. Such 'RWO' resources do exist - for example see Quinault Tribe of the Quinault Reservation, Washington, which is typed as both a madsrdf:RWO and a madsrdf:Geographic*:

Label: Geographic Authority
Comment: Describes a resource whose label represents a geographic place or feature, especially when a more precise geographic determination (City, Country, Region, etc.) cannot be made.

Because existing** Library of Congress RWO lookups (given above) are only available for persons, organizations, and families, no lookup can be used to retrieve those RWO resources which are also typed as madsrdf:Geographic. But, again, such resources do exist.

madsrdf:RWO + madsrdf:Geographic resources are available, although not accessible via a Sinopia QA lookup. We know that the Sinopia lookup service is limited. I don't think the lack of lookup functionality for a given LC RWO resource is sufficient reason to abandon our preference for the LC RWO resources over the LC Names resources for relationships such as those being discussed here, a preference which it seems to me has been well-supported in other related discussions.

Adam also requested an LC subject lookup for 'has related corporate body'. RDF modelling for LC subjects needs further analysis before adding as lookup to create relationships to persons, etc. I imagine LC subjects are modeled as concepts and may not meet needs for creating such relationships.

I also think additional evaluation of non-LC lookup sources is required. For example, resources in Getty ULAN are typed as http://vocab.getty.edu/ontology#PersonConcept ***.

Notes

* Setting aside for a moment that this multi-typing -- unlike the assignment of madsrdf:RWO + bf:Person + foaf:Person, for example -- seems somewhat contradictory, and preferring a resource typed as RWO over one typed as Authority
** Lookups available in authorityConfig.json as of 2023-08-24
*** Getty ULAN snippet

<http://vocab.getty.edu/ulan/500011104> a <http://vocab.getty.edu/ontology#PersonConcept> ;
        [...]

ISNI lookups are also configured for PTs creating relationships to persons. Judging by the published data model; seems ISNI resources are typed as either schema:Person or schema:Organization.

[briesenberg07]

@AdamSchiff thank you for providing additional information here! Following up on some of your comments:

When a corporate body changes its name, separate name authorities are created and linked together reciprocally.

You provided some MARC Authority snippets to illustrate (excerpted):

010  n  99804490 
1102 University of Washington. ǂb Department of Anesthesiology
5102 ǂw r ǂi Successor: ǂa University of Washington. ǂb Department of Anesthesiology & Pain Medicine

010  no2020097621 
1102 University of Washington. ǂb Department of Anesthesiology & Pain Medicine
5102 ǂw r ǂi Predecessor: ǂa University of Washington. ǂb Department of Anesthesiology

We will use the name for the creator appropriate to the time period when the work was created. So a law promulgated by Ceylon will have 110 1 Ceylon, $e enacting jurisdiction, but a law after May 1972 will have 110 1 Sri Lanka, $e enacting jurisdiction.

Above I argue against using what I'm calling 'LC Names resources' and for using what I'm calling 'LC RWO resources' based on the way each are modeled.

But what about from the perspective of the successor/predecessor relationships which you'd like to express in some way in our RDF data? Do either of these types of RDF resources model successor/predecessor relationships?

LC Names resources

<http://id.loc.gov/authorities/names/n99804490> a madsrdf:Authority, # Dept of Anesthesiology
    madsrdf:CorporateName,
    skos:Concept ;
    [...]
    madsrdf:hasRelatedAuthority <http://id.loc.gov/authorities/names/no2020097621> ; # Anesthesiology & Pain Medicine
    [...]
    madsrdf:identifiesRWO <http://id.loc.gov/rwo/agents/n99804490> ; 
    [...]

<http://id.loc.gov/authorities/names/no2020097621> a madsrdf:Authority, # Anesthesiology & Pain Medicine
    madsrdf:CorporateName,
    skos:Concept ;
    [...]
    madsrdf:hasRelatedAuthority <http://id.loc.gov/authorities/names/n99804490> ; # Dept of Anesthesiology
    [...]
    madsrdf:identifiesRWO <http://id.loc.gov/rwo/agents/no2020097621> ;
    [...]

LC RWO resources

<http://id.loc.gov/rwo/agents/n99804490> a <http://id.loc.gov/ontologies/bibframe/Organization>,
    madsrdf:RWO,
    foaf:Organization ;
    [...] # no madsrdf:hasRelatedAuthority
    madsrdf:isIdentifiedByAuthority <http://id.loc.gov/authorities/names/n99804490> .

<http://id.loc.gov/rwo/agents/no2020097621> a <http://id.loc.gov/ontologies/bibframe/Organization>,
    madsrdf:RWO,
    foaf:Organization ;
    [...] # no madsrdf:hasRelatedAuthority
    madsrdf:isIdentifiedByAuthority <http://id.loc.gov/authorities/names/no2020097621> .

For these examples:

1. Relation between predecessor and sucessor is generalized to 'has related authority'.
2. This relation is present in the triples for 'Names resources,' but not for 'RWO resources.'

Note that for the two examples above, description sets for Name resources = ~140 lines (including, as has been remarked on elsewhere, a reproduction of the corresponding RWO resource description), while those for RWO resources = ~30 lines.

Descriptions for these two example RWO resources are less rich and missing relationships which we think are important. Other kinds of RWO resources (persons, orgs, families) are likely the same?

NOTE that at present we still use the available QA lookups for RWO (not names) resources in many places, and that the suggestion to add the LC geographic Names (not RWO) lookup was initially only made because no lookup was available for geographic RWOs. Based on previous discussions I believed that when creating relationships to things like people, corporate bodies, and places, we preferred resources which are not typed as madsrdf:Authority (such as, but not limited to, what I call 'LC Name resources') as a general rule.

❓ Does the above information change this preference, if such a preference existed?

[AdamSchiff]

This is quite interesting to see (that LC has dumbed down the relationships in id.loc.gov, which doesn't make sense to me when we explicitly code the nature of the relationship with RDA elements). Here are some agent relationships from the LC/NACO file: ``` 010 no2019112479 040 NL-UtRU ǂb eng ǂe rda ǂc NL-UtRU 046 ǂf 1917-12-20 ǂg 1983-01-12 ǂ2 edtf 1001 Aalberse, Han B., ǂd 1917-1983 370 Bruinisse (Netherlands) ǂb Bruinisse (Netherlands) 374 Publishers and publishing ǂa Novelists ǂa Translators ǂ2 lcsh 5001 ǂw r ǂi Real identity: ǂa Keulen, Johan van, ǂd 1917-1983 ``` ``` 010 no2019112558 040 NL-UtRU ǂb eng ǂe rda ǂc NL-UtRU 046 ǂf 1917-12-20 ǂg 1983-01-12 ǂ2 edtf 1001 Keulen, Johan van, ǂd 1917-1983 370 Bruinisse (Netherlands) ǂb Bruinisse (Netherlands) 374 Publishers and publishing ǂa Novelists ǂa Translators ǂ2 lcsh 4001 Keulen, Johannes van, ǂd 1917-1983 5001 ǂw r ǂi Alternate identity: ǂa Aalberse, Han B., ǂd 1917-1983 ``` ``` 010 n 87128107 040 DNAL ǂb eng ǂe rda ǂc DNAL ǂd DNAL ǂd DLC ǂd WaU 0247 141986822 ǂ2 viaf ǂ1 http://www.viaf.org/viaf/141986822 1112 AFIA Nutrition Symposium 368 Symposium 368 Congresses and conventions ǂ2 lcsh ǂ0 http://id.loc.gov/authorities/subjects/sh85031115 372 Animal nutrition ǂ2 lcsh ǂ0 http://id.loc.gov/authorities/subjects/sh85005216 372 Animal feeding ǂ2 lcsh ǂ0 http://id.loc.gov/authorities/subjects/sh87007869 373 American Feed Industry Association ǂ2 naf ǂ0 http://id.loc.gov/authorities/names/n86131110 ǂ1 http://id.loc.gov/rwo/agents/n86131110 377 eng ǂ0 http://id.loc.gov/vocabulary/languages/eng 4112 American Feed Industry Association Nutrition Symposium 4112 Nutrition Symposium, AFIA 4112 Nutrition Symposium (American Feed Industry Association) 5102 ǂw r ǂi Sponsoring corporate body: ǂa American Feed Industry Association ǂ4 http://rdaregistry.info/Elements/u/P60676 ǂ0 http://id.loc.gov/authorities/names/n86131110 ǂ1 http://id.loc.gov/rwo/agents/n86131110 ``` ``` 010 no2021046227 040 NNU ǂb eng ǂe rda ǂc NNU 0247 0000000446480070 ǂ2 isni ǂ1 http://isni.org/isni/0000000446480070 0247 Q5476713 ǂ2 viaf ǂ1 http://viaf.org/viaf/Q5476713 046 ǂs 2013-06-28 ǂt 2019-03-20 ǂ2 edtf 1102 21st Century Fox (Firm) 370 ǂe New York (N.Y.) ǂ2 naf 372 Television production companies ǂa Motion picture studios ǂa Press ǂa Mass media ǂa Broadcast journalism ǂ2 lcsh 377 eng 4102 21CF (Firm) 5001 ǂw r ǂi Founder: ǂa Murdoch, Rupert, ǂd 1931- 5102 ǂw r ǂi Predecessor of split: ǂa News Corporation 5102 ǂw r ǂi Hierarchical subordinate: ǂa Twentieth Century-Fox Film Corporation 5102 ǂw r ǂi Absorbing corporate body: ǂa Walt Disney Company ``` ``` 010 no2012154179 040 MsU ǂb eng ǂe rda ǂc MsU ǂd MsU ǂd DLC ǂd UPB ǂd WaU ǂd IaU ǂd WaU 0247 294008977 ǂ2 viaf ǂ1 http://www.viaf.org/viaf/294008977 1003 Baroni (Family : ǂc Natchez, Miss.) 370 ǂc United States ǂ2 naf ǂ0 http://id.loc.gov/authorities/names/n78095330 ǂ1 http://id.loc.gov/rwo/agents/n78095330 370 ǂe Natchez (Miss.) ǂ2 naf ǂ0 http://id.loc.gov/authorities/names/n79042247 ǂ1 http://id.loc.gov/rwo/agents/n79042247 376 Families ǂ2 lcdgt ǂ0 http://id.loc.gov/authorities/demographicTerms/dg2015060157 376 ǂb Baroni, Marjorie Rushing, 1924-1986 ǂ2 naf ǂ0 http://id.loc.gov/authorities/names/no2012154119 ǂ1 http://id.loc.gov/rwo/agents/no2012154119 5001 ǂw r ǂi Progenitor: ǂa Baroni, Marjorie Rushing, ǂd 1924-1986 ǂ4 http://rdaregistry.info/Elements/u/P60683 ǂ0 http://id.loc.gov/authorities/names/no2012154119 ǂ1 http://id.loc.gov/rwo/agents/no2012154119 ``` ``` 010 no2012154119 ǂz n 2007164560 040 MsU ǂb eng ǂc MsU ǂe rda ǂd DLC ǂd MsU ǂd UPB ǂd DLC ǂd UPB ǂd DLC 046 ǂf 1924-08-16 ǂg 1986-03-03 ǂ2 edtf 1001 Baroni, Marjorie Rushing, ǂd 1924-1986 370 Mississippi ǂc United States ǂf Natchez (Miss.) ǂ2 naf 372 Women--Religious life--Mississippi ǂa Civil rights movements--Mississippi ǂa Mississippi--Race relations ǂ2 lcsh 373 Copiah-Lincoln Junior College 373 Mississippi Council on Human Relations ǂa University of Southern Mississippi ǂ2 naf 374 African American women civil rights workers ǂa Newspaper editors ǂ2 lcsh 377 eng 4001 Baroni, Marge Rushing, ǂd 1924-1986 4001 Baroni, Marge, ǂd 1924-1986 4001 Rushing, Marjorie, ǂd 1924-1986 5003 ǂw r ǂi Descendants: ǂa Baroni (Family : ǂc Natchez, Miss.) ``` Are these all generalized to "related authority" too? Regarding your question, ❓ Does the above information change this preference, if such a preference existed?, I guess we need to have a good discussion about this. All I know is that we need to be able to express a relationship as creator or contributor to works and expressions for geographic entities that are also jurisdictions, and these entities reside in the LC/NACO Authority File tagged as geographic names. Interestingly, the MARC 21 Authority Format just refers to them as "geographic names":

**[151 - Heading-Geographic Name (NR)](https://www.loc.gov/marc/authority/ad151.html)** Field Definition and Scope: Geographic name used as a heading in an established heading record, an established heading and subdivision record, a traced or an untraced reference record, or a reference and subdivision record. In an established heading record, field 151 contains the established form of a geographic name. In records for references, this field contains an unestablished form of a geographic name. X51 - Geographic Names-General Information Guidelines: Jurisdiction names (including those that are ecclesiastical entities) alone or followed by subject subdivisions are X51 geographic names.

Adam L. Schiff Principal Cataloger University of Washington Libraries (206) 543-8409 ***@***.***

[briesenberg07]

Thank you @AdamSchiff for sending more test data! I would like to query the RDF representation of these authorities when I'm back so we can be more confident that this is in fact the pattern.

to do @briesenberg07 - query a test set including Adam's records above along the lines of previous queries on LC resources