Discussion paper by DNB: Hidden relationships in attributes (examples: RDA, 9.13, 10.6, 11.3, 16.2.2)

Tagged , , , . Bookmark the permalink.

7 Responses to Discussion paper by DNB: Hidden relationships in attributes (examples: RDA, 9.13, 10.6, 11.3, 16.2.2)

  1. John Myers says:

    I have written across the top of the printout “Brilliant! Good questions.” To the questions posed: a) YES; b) Hopefully; c) YES; d) YES, especially with regards to further implications under implementation scenarios 2 or 3

  2. Matthew Haugen says:

    I second John’s assessment.

    I support this proposal and imagine it will make even more sense once we get out of MARC. For now, recording affiliations (9.13) in MARC 373 doesn’t communicate the nature of the affiliation (without relationship designators) and doesn’t currently actually link anywhere, but maybe that would happen in the post-MARC future based on $2 source coding we’re currently including.

    On the other hand, the ability to include dates of affiliation in 373 isn’t something that can be coded in a 5xx relationship. Dates of affiliation don’t appear to be defined in RDA for affiliations as attributes or as relationships (373 $s and $t don’t map to anything, and RDA 32.2 puts this information in an unstructured explanation of relationships) I wonder if defining dates of the relationship as an attribute of the relationship could be possible and useful in future encoding, as many such relationships are temporary, and dates of incumbency, residency, etc. are sometimes very useful for identifying people, dating works by someone in a particular office, etc.

    I imagine corporate hierarchies, associated institutions, and access points for officials, heads of state, popes, etc., with names and dates of incumbency etc, relating the person to the office, would all fit this category too.

  3. Dominique Bourassa says:

    This discussion paper is an eye-opener. Yes to all questions. I hope the JSC encourages the DNB to explore this question more in depth.

  4. Kathy Glennan says:

    I note that some of the instructions in question are core or core if (but that doesn’t explain everything).

    However, nothing in the cited relationship chapters is core. Would that need to change to make this re-conceptualization work?

  5. Robert Bratton says:

    Yes to all questions! I think this will be a potential mess in MARC, but it is right thing to do for the future. All attributes that can be stated as relationships (an RDF triple) that can link to something, should indeed do that.

  6. Steve Kelley says:

    I’m going to chime in with a “yes” on all four questions.

  7. Tracey L. Snyder says:

    I support the paper. The point that “an entity can be equally well identified by a relationship” (versus an attribute) is well taken. I will include this sentiment of caution, though, from another MusLA member who was otherwise generally supportive:
    “…if you go looking hard enough, what *isn’t* a relationship? I can certainly see a case for ‘Prince, consort of Beatrix, Queen of the Netherlands’ being more accurately recorded as separate title and relationship, but then by the same logic ‘Queen of the Netherlands’ is also a title/relationship pair. At what point does the distinction between attribute and relationship become so minute as to be functionally useless?”

Leave a Reply