Thank you, Adam, for saying something about this. I wasn't paying much attention until I saw your email. This is very concerning. The statement,"In some cases, multiple terms for a single topic may co-exist due to the size of LC’s catalog and issues related to updating the OPAC. When this occurs the authority records for former terms will clearly state “For retrieval purposes only” to ensure that the term is no longer assigned." [Emphasis mine.]indicates that LC is ignoring one of the basic tenets of vocabulary control based on the institution's own convenience. It violates the basic concepts presented in ANSI/NISO Z39.19-2005 (R2010). The definition of controlled vocabulary is clearly stated on page 5:controlled vocabularyA list of terms that have been enumerated explicitly. This list is controlled by and is available from a controlled vocabulary registration authority. All terms in a controlled vocabulary must have an unambiguous, non-redundant definition. At a minimum, the following two rules must be enforced:1. If the same term is commonly used to mean different concepts, then its name is explicitly qualified to resolve this ambiguity.2. If multiple terms are used to mean the same thing, one of the terms is identified as the preferred term in the controlled vocabulary and the other terms are listed as synonyms or aliases. [Emphasis mine.]The plan to use earlier forms of terms in bibliographic metadata and keeping those earlier terms in the list as "related terms," clearly disregards the basic definition of controlled vocabulary. Controlling synonyms and distinguishing homographs are the raisons d'être for authority control. So, how does this decision get made? LC's interest in apply this idea to a controlled vocabulary "due to the size of LC's catalog and issues related to updating the OPAC" appears to be a direct contradiction of a long-standing principle of our profession: "The convenience of the public is alawsy to be set before the ease of the cataloger." -- Charles Ammi Cutter, Rules for a Dictionary Catalog, 4th ed. Washington, DC: GPO, 1904, p. 6. Our patrons should not have to search synonymous terms to find what they are looking for if a controlled vocabulary is employed.I also want to echo Bob's concern that LCDGT "is being used as an experimental guinea pig for this new “principle” and the real goal is to apply it to LCSH (and why not the NAF while we’re at it?)." This seems like a real possibility if this is allowed to happen with LCDGT, and it could very much inflict more damage to a standard (yes, a flawed standard) on which we all depend. I am concerned that this is being considered because, as we've heard so many times about so many things, "linked data will take care of it." I would again like to echo Bob by stating, linked data is not yet here. And, there are plenty of skeptics who are concerned that we are placing all of our eggs into a single conceptual basket. One that may or may not work the way we hope it will.It is still essential for LC to appropriately manage equivalence relationships in our vocabularies, and to honor the principles of unique heading and uniform heading. I am very disappointed that PTPC is considering going in the opposite direction. I would respectfully ask them to reconsider this decision, because it is not a good one.Danny---------------------------------------------------------
Daniel N. Joudrey, Ph.D.
School of Library and Information Science (SLIS)
300 The Fenway, P-205B
Boston, MA 02115 617.521.2863On Wed, Sep 22, 2021 at 7:04 AM Deborah Tomaras <[log in to unmask]> wrote:I agree that having multiple terms listed as "related" but with one not valid anymore for actual usage (but coded as if for actual usage!) causes multiple problems for the LCDGT vocabulary's development, and for multiple library catalogs and authority files.
- Most discovery layers don't allow patrons to see authority records. So having a "for retrieval purposes" note within an authority record that will be invisible to most users will not clarify anything in actual catalog searches done by patrons. One could argue that libraries simply should not assign, or should reclassify, affected terms in their local catalogs to eliminate this issue. However, discovery layers ingest metadata from so many different streams that I doubt it would be possible to locate and eliminate problematic terms entirely--especially if conflicting terms are still "valid" in authority files.
- Many discovery layers have facet display limits, so having former terms that should be see-fors taking up valuable shelf space is detrimental for search faceting purposes.
- This could also cause potential DEI issues, if terms get updated from offensive or outdated versions, yet the old terms remain as "for retrieval purposes" entries and so are still visible in library catalogs for patron viewing and usage (for example a possible change from "Blacks" to "Black people").
- For libraries that rely on automated authority control and updating of terms, having two conflicting preferred terms will cause these authority updates to fail. Libraries would then have to manually update terms, and repeatedly handle "ambiguous" authority headings reports in their local systems.If the purported reason for Library of Congress's decision to follow this non-ANSI procedure is, as it appears, to be based solely on issues related to their individual catalog ("
From: Program for Cooperative Cataloging <[log in to unmask]> on behalf of Diana M. Brooking <[log in to unmask]>
Sent: Tuesday, September 21, 2021 5:06 PM
To: [log in to unmask] <[log in to unmask]>
Subject: Re: [PCCLIST] LCDGT Announcement
But delays in implementing BFM could be the case for any change of preferred term in any library.
Should the structure of a widely used controlled vocabulary be dependent on the BFM of any particular library, even if that library is LC?
University of Washington Libraries
Seattle WA 98195-2900
I expect the LC practice is simply a pragmatic one applicable only in cases where a change of preferred term entails a large amount of LC bibliographic file maintenance. If so, then the existence of two preferred terms for any given concept will always be temporary, persisting only until the related BFM is completed. (At least that’s how I read the LC announcement.)
While it is exciting to finally know that the moratorium on LCDGT will be lifted soon, and backlogged proposals will be processed and decided upon, I have some serious qualms about the model for former terms that LC is using, and I don't believe that it is in accordance with ANSI/NISO Z39.19-2005 (R2010). The standard is at https://groups.niso.org/apps/group_public/download.php/12591/z39-19-2005r2010.pdf.
Section 6.2.3 of ANSI/NISO says:
A History Note is used to track the development of terms over time. These notes provide important guidance for researchers who are interested in a topic covering many decades. It is especially important to indicate when and how a term has changed over time. If appropriate, the history note may also include the date discontinued, the term that succeeded the term, and/or the term that preceded it. History Notes are frequently marked by the abbreviation HN.
Assuming from this example that Venetian windows was once an authorized term, the example shows it as a Used For term. This is further substantiated by 126.96.36.199:
188.8.131.52 Modification of Existing Terms
Indexers and searchers should be able to propose modifications to existing terms or their relationships, explaining the rationale and supplying supporting documentation for the proposed changes. Like candidate term nominations, such proposals may be communicated electronically or via printed forms. Such proposed changes should be considered by the controlled vocabulary editor and board, using the criteria for term selection in sections 6 and 7. If a term is modified, the date of the change should be recorded in the history note (see section 6.2.3), and a USE reference should be made from the old form to the new form. If the controlled vocabulary is used in an indexing system, the date on which an old term was last assigned should be included in the history note. If the relationships are modified, a record of the old ones should be maintained in the history note as well.
(I have highlighted the most important sentence above).
Based on my reading of the ANSI/NISO standard, I don't think these existing LCDGT terms follow it:
Having two separate authorized terms for the same concept does seem to me a violation of ANSI/NISO. My reading of the standard would result in Oceanians being a UF on Pacific Islanders, and there could also be history note in that record stating when the change from earlier form Oceanians was made.
I also don't think it serves users in a faceted retrieval system to be presented with multiple boxes to select for the same concept:
Each group should be distinct. If a user sees a box for an earlier form and misses another box further down the list for a later form, they will miss retrieving items that they would have wanted to see.
Adam L. Schiff
University of Washington Libraries
From: Program for Cooperative Cataloging <[log in to unmask]> on behalf of Cannan, Judith <[log in to unmask]>
Sent: Monday, September 20, 2021 1:50 PM
To: [log in to unmask] <[log in to unmask]>
I am pleased to announce that on October 1, 2021, LC will start a new model for the LC Demographic Group Terms (LCDGT). I have shared this message with SAC and I will be sending it to SACO. This announcement will also be posted on LC website and the PCC website.
--Stephen Hearn, Metadata StrategistCataloging, Metadata, & Digitization ServicesUniversity of Minnesota Libraries170A Wilson Library (office)160 Wilson Library (mail)309 19th Avenue SouthMinneapolis, MN 55455Ph: 612-625-2328Fx: 612-625-3428ORCID: 0000-0002-3590-1242