The Library of Congress >> Especially
for Librarians and Archivists >> StandardsThe Library of Congress >> Especially
for Librarians and Archivists >> Standards
HOME >> MARC Development >> Proposals List
DATE: May 22, 2025
REVISED:
NAME: Defining Field 655 (Index Term-Genre/Form) in the MARC 21 Holdings Format
SOURCE: Yale University; co-sponsored by the Bibliographic Standards Committee (BSC) of the Rare Books and Manuscripts Section (RBMS) of the Association of College and Research Libraries (ACRL), a division of ALA (American Library Association); and by the Cataloging Advisory Committee (CAC) of the Art Libraries Society of North America (ARLIS/NA)
SUMMARY: This paper proposes defining field 655 (Index Term-Genre/Form) in the MARC 21 Holdings Format.
KEYWORDS: Field 655 (HD); Index Term-Genre/Form (HD)
RELATED: 2025-DP03
STATUS/COMMENTS:
05/22/25 – Made available to the MARC community for discussion.
06/25/25 – Results of MARC Advisory Committee discussion: Approved as submitted.
11/26/25 – Results of MARC Steering Group review - Agreed with the MAC decision.
Field 655 is currently defined in the MARC Bibliographic Format as follows:
FIELD DEFINITION AND SCOPE
Terms indicating the genre, form, and/or physical characteristics of the materials being described. A genre term designates the style or technique of the intellectual content of textual materials or, for graphic materials, aspects such as vantage point, intended purpose, characteristics of the creator, publication status, or method of representation. A form term designates historically and functionally specific kinds of materials distinguished by their physical character, the subject of their intellectual content, or the order of information within them. Physical characteristic terms designate historically and functionally specific kinds of materials as distinguished by an examination of their physical character, subject of their intellectual content, or the order of information within them.
Examples of genre terms for textual materials are: biographies, catechisms, essays, hymns, or reviews. Examples of form and physical characteristic terms are: daybooks, diaries, directories, journals, memoranda, questionnaires, syllabi, or time sheets.
May contain terms constructed from a faceted vocabulary. For each term found in the field, an identification is given as to the facet/hierarchy in the thesaurus from which the term came. In addition, identification is given as to which term is the focus term of the expression.
Standard published lists are used for the genre and form terms. The lists are identified either by the second indicator (Thesaurus) or by subfield $2 (Source of term).
This proposal builds upon the ideas and observations presented in MARC Discussion Paper 2025-DP03, which advocated for the definition of field 655 (Index Term-Genre/Form) in the MARC 21 Holdings Format. The discussion paper was shared with the MARC Advisory Committee (MAC) and the broader MARC community in December 2024 and was discussed at the Midwinter 2025 MAC meeting. The feedback from the community highlighted both areas of support and concerns, which this proposal seeks to address.
Discussion Paper 2025-DP03 advocated for the inclusion of field 655 (Index Term-Genre/Form) in the MARC 21 Holdings Format to provide a more precise, copy-specific location for controlled genre/form terms. These terms, frequently used in cataloging rare materials, describe item-specific characteristics such as unique bindings, extra-illustration, annotations, or custodial history evidence. Because field 655 is only defined in the MARC Bibliographic Format, these copy-specific genre/form terms are often recorded at the bibliographic level, using a $5 to indicate the institution to which those terms apply. However, when a bibliographic record represents multiple copies, this method makes it challenging to associate specific terms with the appropriate copy in discovery interfaces and other systems that depend on MARC catalog data.
The discussion paper emphasized that defining 655 in the Holdings Format will address these issues by enabling catalogers to record copy-specific genre/form terms directly in holdings records. This approach aligns with other fields that are already shared between the Bibliographic and Holdings Formats (e.g., fields 541, 561, 562, and 563), which similarly support the recording of copy-specific metadata.
The discussion paper received support from the MARC community, with some agreeing that there was a legitimate use case for defining this new field. One respondent noted that adding a 655 field to holdings records would be "immensely useful for rare materials cataloging" and "greatly help discovery in local systems and discovery layers." Another comment underscored the need for a holdings-specific field to address metadata gaps for rare materials cataloging that are not covered by existing fields like 361 (Structured Ownership and Custodial History). Several institutions also emphasized the importance of expanding support for Item-level description and structured fields in the MARC Holdings Format.
Although there was support for the exploration of this field, respondents also raised valid concerns about its implementation and technical feasibility in library systems. The following section aims to address these concerns and clarify the approach to this new field.
The concern about the proper level for copy-specific data is well-founded, as such data ideally belongs at the item level. However, MARC lacks a dedicated item format, and item record structures in many ILSs are proprietary and rigid, limiting interoperability and customization. In such cases, recording copy-specific genre/form terms at the holdings level using a 655 field would represent a practical compromise. This approach would enable terms to be recorded closer to the intended level of description, reducing ambiguity in bibliographic records.
Using subfield $3 (Materials Specified) in the proposed Holdings 655 field would provide additional granularity by specifying the copy or part of a resource to which the term applies. This flexibility would be especially useful in scenarios where institutions have multiple copies attached to a single holdings record. For libraries where each holdings record corresponds to a single copy, $3 may not be necessary, but its availability could enhance precision in describing specific features of a resource, parts of a multipart/serial resource, or titles in bound-withs/sammelbands. See the Examples section below for potential uses of the 655 $3 in holdings records.
The potential bifurcation of data – where copy-specific genre/form terms could be recorded in multiple locations – is an important consideration, particularly with respect to its implications for consistent data sharing across institutions. While bifurcation may introduce certain complexities, the proposed addition of field 655 to the MARC 21 Holdings Format aims to offer catalogers an optional way to address specific challenges inherent in the existing cataloging landscape.
It is important to weigh the priorities related to data sharing for copy-specific metadata against the advantages of maintaining clear boundaries between copy-level and Instance/Manifestation-level data. Copy-specific metadata typically pertains to features unique to an institution's individual copy of a resource. As such, it is often relevant only to the holding institution and may not be intended for broader data sharing. In practice, many institutions opt to keep this information local for internal use rather than sharing it with utilities like OCLC. Similarly, the inclusion of copy-specific data in shared cataloging databases such as OCLC has been a subject of ongoing community debate. While some advocate for discontinuing the practice of sharing such data, others support its continued inclusion. Enabling the recording of copy-specific genre/form terms directly in holdings records could address these concerns by granting institutions greater control over where to record this data and whether to share it externally along with their bibliographic records.
Defining this field to record copy-specific genre/form terms in the holdings record would not preclude institutions from continuing to use the bibliographic record for this purpose where local practices or system limitations require it. This dual approach would reflect the diverse needs of the cataloging community while offering the flexibility to adapt to future data models, such as BIBFRAME, which aim to separate Instance/Manifestation-level and Item-level data more cleanly.
The 655 field is uniquely suited for inclusion in the MARC Holdings Format due to its ability to describe "is-ness" – physical characteristics, genre, or form – rather than "about-ness," which relates to intellectual content. Other 6XX fields are fundamentally tied to subject content and typically apply to all copies of a resource, lacking as clear a use case at the holdings level. The 655 field, however, could align with the conceptual level of holdings records by containing metadata that only applies to particular copies, such as "Extra-illustrated copies" or "Hand coloring." The addition of a 655 field would complement existing fields like 562 (Copy and Version Identification Note) and 563 (Binding Information) that focus on item-specific attributes in the holdings record. It would provide a targeted solution to a recognized need without necessarily inviting the definition of further Holdings 6XX fields.
The concern over whether ILSs and discovery layers can adequately handle the 655 field in holdings records – including indexing, display, searchability, and authority control – is understandable. However, certain systems already demonstrate the ability to display and index holdings fields, such as 5XX note fields in Primo- and Solr-based discovery layers (see examples from Harvard and Yale). Likewise, the mapping of multiple MARC fields across bibliographic and holdings records to unified search indexes is possible with appropriate configurations. For institutions where implementing a holdings-level 655 field is not immediately feasible due to system rigidity or limited development resources, the existing practice of recording copy-specific genre/form terms in Bibliographic 655 fields, using subfields $3 and $5, remains a viable option. Less constrained institutions, however, would be able to leverage the benefits of holdings-level 655 fields to improve the search and discovery experience for users.
Additionally, the introduction of controlled vocabulary fields in the Holdings Format is not entirely without precedent. Fields such as 337 (Media Type), 338 (Carrier Type), and 361 (Structured Ownership and Custodial History) already support controlled terms. While these fields do not systematically link to external authority records, they demonstrate the validity of using controlled terms in holdings data. Institutions using systems like Ex Libris' Alma, which allow controlled vocabularies to be assigned to particular fields or subfields, could apply a similar workflow to the Holdings 655 field. While authority control linkage might not yet be universally supported, institutions could nonetheless benefit from using standardized terms in conjunction with subfield $2 for the vocabulary source and $0 for the URI, ensuring metadata consistency and interoperability as systems adapt to support this functionality.
It is also important to note that adoption of a controlled, holdings-level 655 field aligns with some ILS vendors' growing interest in linked data, such as the Ex Libris Linked Data Focus Group. By integrating this field into cataloging workflows, institutions would encourage vendors to prioritize more robust support for controlled vocabularies and authority control in holdings records, reducing the configuration burden on individual institutions. This proposal could also serve as a catalyst for broader system functionality enhancements to MARC holdings records.
While implementing the 655 field in Holdings would require resources and systems expertise, existing technological infrastructure already offers partial support, with further advancements contingent on vendor engagement and institutional demand. Importantly, the development of MARC standards should not be overly restricted by current system limitations, particularly when such constraints are specific to certain platforms and likely to evolve.
In a Bibliographic 655 field, subfield $5 can specify that a genre/form term applies only to a particular institution, and subfield $3 can associate the term with a particular copy or part of a resource. Some libraries already use a Bibliographic 655 field with one or both subfields for copy-specific genre/form terms, and defining a 655 field in the MARC Holdings Format would not prevent them from continuing to do so. However, embedding copy-specific terms in bibliographic records can present both practical and conceptual challenges.
In general, intermingling copy-specific terms with Instance/Manifestation-level terms in bibliographic records risks creating overcrowded, difficult-to-manage records, which can confuse both catalogers and end users. This practice may also complicate system migrations and data-sharing efforts, increasing the risk of misassociation or over-sharing of data. More specifically, relying on the $3 subfield in Bibliographic 655 fields places unnecessary demands on the free-text subfield. The $3 subfield can provide detailed information about the specific part or copy a genre/form term applies to, as explained in the "Copy-specific data in items vs. holdings records" section above. However, its use in holdings records is a workaround due to the absence of a dedicated MARC item format for recording such terms more appropriately. In contrast, the MARC Holdings record already provides a location separate from the bibliographic record, making it more logical and efficient to record copy-specific genre/form terms directly in a Holdings 655 field. This approach would eliminate the need to include holdings-related information in $3, thereby preserving the subfield's clarity and reducing the amount of information it must hold.
Moving copy-specific genre/form terms to a 655 field in the MARC holdings record would more clearly separate copy-specific data from Instance/Manifestation-level data, aligning more closely with linked data principles and enhancing both human readability and machine actionability. Catalogers and systems alike could more effectively determine which data applies to specific holdings, and which applies to all copies of a resource.
Field 562 (Copy and Version Identification Note) provides some granularity in terms of the types of copy-specific characteristics it describes. However, it is a free-text field that lacks support for controlled vocabularies, which limits consistency and interoperability. Field 655 is designed specifically for structured, controlled terms.
Field 361 (Structured Ownership and Custodial History) serves a different purpose and would complement rather than replace a 655 Holdings field. While 361 includes subfield $f for "ownership and custodial history evidence terms," its scope is limited to features like bookplates, stamps, and autographs related to custodial history, not general genre/form terms. Additionally, 361 is designed to record these features in relation to an agent ($a), whereas many evidence terms and broader genre/form terms can apply without a known agent. For instance, an unidentified armorial bookplate is better recorded in a Holdings 655 field than in a 361 $f without an associated agent. Similarly, copy-specific genre/form terms like "Hand coloring" or "Imposition errors" typically do not require agent attribution. In short, field 361 focuses on custodial history terms tied to agents, while field 655 is better suited for general genre/form characteristics of individual copies.
Field 380 (Form of Work) is currently defined in the MARC Bibliographic and Authority Formats as "a class or genre to which a work belongs". This suggests that this field is limited to Work-level characteristics; and indeed, it was specifically defined to mirror the RDA element of the same name related to Works and Expressions, which was considered core if used to differentiate a Work from another Work with the same title (pre-Official RDA; see 2010-04 and 2009-DP06/3). Its current definition does not accommodate local, Item-level genre/form information, and redefining its scope for holdings use would risk introducing inconsistencies with its established role in the Bibliographic and Authority Formats. Field 655, on the other hand, already has a broad definition and is widely used to record both general and copy-specific genre/form data.
Field 842 (Textual Physical Form Designator) currently lacks support for controlled vocabularies and would require significant modifications to its definition to align with the use case outlined in this proposal. These modifications would include adding a second indicator value of "7" and introducing subfield $2 to accommodate thesauri or code lists, and making the field repeatable. More critically, however, implementing these changes would require changes to the field's scope beyond "the physical medium or form, or type of material of the unit held". This is further complicated by the field's established correspondence with the ANSI/NISO Z39.71 Physical Form Designator data element. If such modifications resulted in misalignment between Field 842 and the Z39.71 standard, it might prompt the need to decouple Field 842 from the ANSI/NISO framework, which may be undesirable.
According to ANSI/NISO Z39.71 Holdings Statements for Bibliographic Items, the Physical Form Designator "indicates the physical medium or type of material of the bibliographic unit" (page 17), which aligns closely with the definition of corresponding MARC field 842. Table 3.3 lists coded and natural language values for this designator, including terms such as "Cartographic material, map", "Sound recording, sound cassette", and "Text, large print". The standard additionally states that, "when values in textual form are used, this designator is usually omitted if the unit is regular print text" (page 28). Even if terms from external controlled vocabularies were used, this guidance suggests that the intended scope of Z39.71 Physical Form Designator differs substantially from that of the proposed Holdings 655 field. These designators describe the entire "bibliographic unit", and focus on broad, Instance/Manifestation-level characteristics that apply uniformly to all copies of a given resource. In contrast, a Holdings 655 field would provide a mechanism for describing unique features of specific copies. Compare potential Physical Form Designators as described above or in Table 3.3 with the following potential holdings genre/form terms: "Authors' annotations," "Armorial bookplates," or "Extra-illustrated copies." These terms do not necessarily apply at the level of the "bibliographic unit", nor do they apply to all exemplars of a "bibliographic unit". While the ANSI/NISO Z39.71 standard and its associated Physical Form Designator serve an important purpose in general collections, they are insufficient for describing unique, copy-specific attributes. The proposed 655 field in the MARC Holdings Format is not intended to replace or replicate the Physical Form Designator, but to enhance the descriptive capabilities of holdings records by capturing data beyond the scope of this element.
Current usage in bibliographic record:655 #7 $a Extra-illustrated copies. $2 rbmscv $0 http://id.loc.gov/vocabulary/rbmscv/cv00241 $5 CtY-BR
(no holding specified)
Or
655 #7 $3 BEIN 2019 13: $a Extra-illustrated copies. $2 rbmscv $0 http://id.loc.gov/vocabulary/rbmscv/cv00241 $5 CtY-BR
(holding specified)
Potential usage in holdings record:
852 80 $b beingen $h 2019 $i 13
655 #7 $a Extra-illustrated copies. $2 rbmscv $0 http://id.loc.gov/vocabulary/rbmscv/cv00241
Potential approaches to the use of $3 in a 655 Holdings field. This will ultimately depend on local practices.3.2.1. Holdings record with two associated copies:
852 80 $b beinycal $h Zab P414 $i 978C
655 #7 $3 Copy 2: $a Armorial bookplates. $2 gmgpc $0 http://id.loc.gov/vocabulary/graphicMaterials/tgm0004893.2.2. Holdings record for serial:
852 81 $b beinycal $h Za $i Zar54
866 41 $8 0 $a no.2-no.3
655 #7 $3 No. 3: $a Annotations. $2 rbmscv $0 http://id.loc.gov/vocabulary/rbmscv/cv001023.2.3. Holdings record for monograph with black and white plates:
852 80 $b beingen $h 2025 $i 1091
655 #7 $3 Plates 1-3,6: $a Hand coloring. $2 aat $0 http://vocab.getty.edu/aat/300133555
The BIBFRAME Initiative has not been focused on conversion of the MARC Holdings Format to linked data and so there would be no immediate support for this in the LC BIBFRAME conversion specifications.
Define field 655 (Index Term-Genre/Form) in the MARC 21 Holdings Format with the same indicators, subfields, and guidelines as in the MARC 21 Bibliographic Format (as shown in Section 1. Background). The one exception is subfield $5 (Institution to which field applies) (NR), which should not be defined for a 655 field in the MARC 21 Holdings Format as it is not relevant when the 655 is located in a holdings record. Since holdings records are already institution-specific by design, using $5 would be redundant.
HOME >> MARC Development >> Proposals List
| The Library of Congress >> Especially for Librarians and Archivists >> Standards (11/26/2025) |
Legal | External Link Disclaimer | Contact Us |