7 Responses to Proposal: High-level subject relationship in RDA

  1. Kathy Glennan says:

    Note that when the proposal refers to J.1.3, J.2.3, J.3.3, and J.4.3, the numbers are all off by one. The reference should be to J.2.3, J.3.3, J.4.3 and J.5.3.

  2. Robert L. Maxwell says:

    Comment from John Attig (SAC RDA Subcommittee)

    First, I believe that it is correct to say that there is nothing in this proposal that contradicts or questions anything in the ALA proposal to add the high-level subject relationship to Chapter 23 of RDA. The definition of the “Subject” relationship (Recommendation 1 on page 2 of the document) is slightly different from the wording that ALA proposed, but the differences seem to be minor.

    Most of the document deals with the “Descriptive Relationships” defined as relationship designators in RDA Appendix J. When Appendix J was drafted it was argued that the descriptive relationships were by definition subject relationships. In spite of this, they were included in Appendix J. Now Gordon is proposing to treat them as subject relationships and suggests some modifications to the existing relationship designators. The main point is that, as subject relationships, descriptive relationships must always exist between a work and the entity that is the subject of the work; this means deleting some designators that exist at the expression or manifestation level. Interestingly, Gordon provides examples that would seem to justify the existence of descriptive relationships from an expression or manifestation. If such relationships exist, why must they be excluded from the model? This is something we might wish to comment on.

    I was able to add to the Working Group document a section dealing with the “described in” relationship — the inverse of the standard “has as subject” relationship. These relationships support (among other things) the relationship to published citations (MARC field 510). I proposed — because the model did not allow for such relationships at the expression or manifestation level — that they be treated as elements instead of relationships (Recommendations 3 and 4). This is based on the assumption that “described in” relationships are in fact subject relationships. I’m not sure that this is the case. This is something the ALA might wish to comment on.

    The proposal does not make clear whether the “descriptive relationships” will continue to be included in Appendix J — where they are limited to relationships between Works and WEMI entities that are the subjects of those works — or whether they belong in a separate Appendix that deals with subject relationships. Because of the limitations of the scope of Appendix J, it seems to me that descriptive relationships — to the extent that they are to be treated as subject relationships — should NOT be covered in Appendix J, but rather in the Appendix L proposed by ALA for designators for subject relationships. This is also something ALA might wish to comment on.

  3. Robert L. Maxwell says:

    Comment from Bonnie Dede (SAC RDA Subcommittee)

    “Described in” relationship in the Working Group document:
    John writes that these relationships “be treated as elements instead of relationships (Recommendations 3 and 4). This is based on the assumption that ‘described in’ relationships are in fact subject relationships. I’m not sure that this is the case.” Very interesting comment because I see that “described in” leads us to both a description relationship (in the case of descriptive bibliography for rare materials) and a subject relationship component (the portion of the description which is about the exact identification of the material in hand).

    Appendix J issues:
    Yes, there are limitations of the scope of Appendix J. John correctly identifies that “descriptive relationships — to the extent that they are to be treated as subject relationships — should NOT be covered in Appendix J, but rather in the Appendix L proposed by ALA for designators for subject relationships.” Both Appendix J and an Appendix L are needed in order to make the differences between descriptive and subject relationships clearer.

    Existence of descriptive relationships at an expression or manifestation level:
    John raises a good point here that “if descriptive relationships exist at an expression or manifestation level, why must they be excluded from the model?” It appears that this area needs further discussion.

    And lastly, is there a word missing in the Working Group document (p. 4, Recommendation 2, paragraph 5, 3rd sentence)?:
    “… and therefore _has a literal containing the text_ …”

  4. Robert L. Maxwell says:

    Comment from Tony Olson (SAC RDA Subcommittee)

    I think the important point that the JSC Technical Working is proposing is as follows. If descriptive relationships are to be considered subject relationships, then the reason why descriptive relations from an expression or manifestation are excluded from the model is stated in Recommendation 2: “Bring the RDA descriptive relationships into line with FRSAD by allowing only Work to be the domain of primary descriptive relationship designators for WEMI entities (and the range of their reciprocal designators) … “. In FRSAD only the “work has subject” or its reciprocal “is subject of work” relationships are allowed. Subject relationships from expressions and manifestations are not defined.

    I also agree with John and Bonnie that subject relationship designators should be dealt with in Appendix L and not Appendix J.

  5. Robert L. Maxwell says:

    Comment from Stephen Hearn (SAC RDA Subcommmittee)

    The FRBR Review Group’s Working Group on Aggregates defines an aggregate in its report as both “a manifestation embodying two or more expressions” and as an “aggregating work,” noting also that “The aggregating work may, or may not, be deemed important enough to be recorded.” (page 6-7). As long as it’s clear that when a subject is assigned to an aggregate per se, it’s being recorded for the work aspect of the aggregate, that should all be consistent with the Technical Working Group’s recommendation.

  6. Robert L. Maxwell says:

    Comment from Shelby Harken (SAC RDA Subcommittee)

    I also agree that subject relationship designators should be dealt with in Appendix L and not Appendix J. It would be too confusing when J allows expressions and manifestations and subjects are only for Works.

  7. Matthew Haugen says:

    On behalf of DCRM and Standard Citation Forms for Rare Materials Cataloging Editors:
    We thank the JSC Technical Working Group for its attention to citations in RDA. The practice of making references to published descriptions is vital to rare materials cataloging (cf. DCRMB 7B14), and specifically, we believe recommendations 3-4 will greatly facilitate this practice.

    Re: Recommendation 2: We encountered substantial resistance to characterizing the relationship between a descriptive resource and the WEMI it describes as a subject relationship, rather than that of instantiation or identification of the WEMI of which the description is a surrogate (much like the bibliographic records we create which contain these references). We affirm this recommendation’s acknowledgment that references must often be to a specified expression, not only to the work. Expressions are generally the fundamental unit of reference in the existing Standard Citation Forms for Rare Book Cataloging (SCF) and its forthcoming online successor Standard Citation Forms for Rare Materials Cataloging. If the domain is restricted to Work as a subject relationship, then these relationships would be excluded.

    Re: Recommendation 3: We support the development of RDA elements for references to published citations for works, expressions, manifestations, and items. For descriptive cataloging, the addition of elements Reference to Published Citation (Manifestation) and Reference to Published Citation (Item) would be very welcome. They would formally provide RDA elements for citations of the kind that we record in MARC bibliographic field 510. While Reference to Published Citation (Work) and Reference to Published Citation (Expression) are generally out of scope of DCRM, they would be logical additions as well.

    Re: Recommendation 4: Current community practice (as embodied in SCF) places an emphasis on citations as attributes, conveyed in notes; however, we look forward to the possibility of expressing these citations as relationships, especially in a linked data environment. We support the development of designators to express the relationship between the WEMI being described and the source of the description. If characterized as WEMI-to-WEMI relationships rather than subject relationships these would be appropriate in Appendix J along with other indexing, abstraction, etc. relationships also recorded in MARC 510, rather than Appendix L.

    General comments:
    Our community’s citation use case frequently includes recording the location of the citation within a published descriptive resource. It is not clear whether the recommendations in this proposal will accommodate this. Citation numbers often serve as an identifier for the description (and even the described WEMI). More precise references assist with identification of WEMIs, differentiation of variants, or justification of information (e.g. dates) supplied elsewhere in the description. While location within a resource is sometimes self-evident due to alphabetical arrangement, etc., omitting location information often renders a citation useless. For example, in a typical reference like 510 4 ESTC $c R1023, ESTC is the full resource, while R1023 is the unique citation number within it. The two pieces of information would not be very useful in isolation. This citation structure identifies the description R1023 as a part of the English Short Title Catalog as a whole, and differentiates it from descriptions of potentially similar manifestations within the ESTC, or descriptions with the same citation number in a different descriptive resource. While citation numbers are generally constant at the expression level, volume /page numbers or URIs (e.g. 510 4 ESTC $c R1023 $u http://estc.bl.uk/R1023) are also used, which would potentially be specific to the “published” manifestation. (Note: In the forthcoming online version of SCF, many abbreviations like ESTC are being replaced with fuller forms like English Short Title Catalog).

    As an aside, there is also precedent for noting details on these references, e.g. “Date of publication from Evans.” or the absence of such references when they would be expected, e.g. “500 Not in ESTC.”

    Whether or not such references are given explicit accommodation in RDA, we suspect it may be useful to define MARC bibliographic 510 $5 (institution to which field applies) for references which describe an item, not a manifestation.

Leave a Reply