DATE: May 22, 2003

NAME: Defining subfield $2 in fields 155, 455 and 555 of the MARC 21 Authority Format

SOURCE: Bibliographic Standards Committee, RBMS/ACRL

SUMMARY: This paper considers the definition of subfield $2 (Source ) in fields 155, 455 and 555 of the MARC 21 authority format to allow for greater flexibility and consistency with field 655 in the bibliographic format. This may result in better ability for machine validation of headings if the source were in the same subfield in both formats.

KEYWORDS: Field X55 (AD); Genre/form (AD); Subfield $2, in fields X55 (AD)

RELATED: 2002-07 (January 2002); 95-11 (June 1995)


05/22/03 -- Made available to the MARC 21 community for discussion.

06/21/03 - Results of the MARC Advisory Committee discussion - Participants agreed that it is not necessary to define subfield $2 in the X55 fields since the present coding methods adequately identify the source of genre terms. Participants were concerned that defining subfield $2 in the X55 fields would only assist institutions using systems that cannot currently identify the subject heading system/thesaurus used in a record because of specific system limitations. Defining subfield $2 would also confuse catalogers by giving them a third method with which to code the same information. The group did not recommend that this paper come back as a proposal.

Discussion Paper No. 2003-DP04: Defining subfield $2 in fields 155, 455 and 555


When the 155, 455, and 555 fields were defined in the MARC 21 Authorities format, they included most of the subfields authorized for use in the 655 field in the bibliographic format. However, some "numbered" subfields were not defined in the authorities format, perhaps because they were not thought to be candidates for authority control. This may be true of some of them, but not of subfield $2. This paper discusses adding subfield $2 in authority fields 155, 455, and 555 so that it might be controlled along with the headings in the field. Note that subfield $2 is defined in authority field 755 (Established heading Linking Entry-Genre/Form).


The contents of 655 fields in MARC 21 bibliographic records are controlled by a defined set of thesauri, found online at: www.loc.gov/marc/relators/relasour.html#rela655b, or a locally-defined thesaurus may be used. Terms (and their subdivision strings, which are governed by the thesaurus the main term comes from) are therefore subject to authority control, and thus it is appropriate that a corresponding set of X55 fields has been defined for use in the MARC 21 Authorities Format.

Because thesauri sometimes conflict (both as to terms and as to hierarchy), a mechanism must exist to identify in the bibliographic record the thesaurus governing the term. Two separate mechanisms now exist in field 655 for most thesauri, the thesaurus is identified by a code in subfield $2 at the end of the string; or, as of January 2002, six thesauri are identified by values in the second indicator position. Codes are available for LCSH, LC subject headings for children's literature, MeSH, NAL subject authority file, Canadian subject headings, and Répertoire de vedettes-matière.

Thus, for terms from thesauri not represented by a second indicator value, subfield $2 is used in field 655, which contains a code for the thesaurus containing the term. There are more than 30 authorized thesauri, each with a corresponding code (see reference above). Because the code is required in 655 fields derived from all but six thesauri, and because the code is also controlled, it, too, is subject to validation.

However, catalogers are prone to errors in entering codes just as much (if not more so) as they are in entering terms and term strings. Therefore, it is just as important that this portion of the string be subject to a system's authority mechanisms as the earlier portions of the string. Authorizing subfield $2 for use in the MARC 21 Format for Authority Data would allow systems to place this section of the string under authority control.

The following examples illustrate how this coding would work:

Bibliographic record:

  655 #7 $a Science fiction. $2 gsafd

Authority record:

  155 ## $a Science fiction. $2 gsafd

Bibliographic record:

  655 #7 $a Moralities. $2 gsafd

Authority record:

  155 ## $a Moralities. $2 gsafd

Bibliographic record:

  655 #7 $a Morality plays. $2 rbgenr

Authority record:

  155 ## $a Morality plays. $2 rbgenr

The Moralities / Morality plays examples illustrate the need for this coding. GSAFD (1) and RBGENR (2), two of the thesauri authorized for use in 655, use a different term for the same concept. An individual library can choose either to allow only one in their database, or to allow both. If it chooses to authorize both, it needs to distinguish between the two terms in bibliographic records, and the accepted method for these two thesauri is to distinguish by subfield $2. The string "Moralities. $2 rbgenr" would be incorrect, even though the term itself, if taken from GSAFD, might be authorized in the library's database. But without subfield $2 in the authority record(s), the library's system would be unable to detect the error.

If the library chooses not to authorize both, but only one, the system needs a mechanism for recognizing existing and incoming incorrect forms. This could be accomplished through the use of cross references:

Authority record:

  155 ## $a Morality plays. $2 rbgenr
455 ## $a Moralities. $2 gsafd

Such cross referencing could allow machine recognition and "flipping" of unauthorized headings if the library desired. However, without subfield $2 in the authority record, the system would not recognize the entire string in the bibliographic record, and would not be able to replace it with the correct string (including its correct subfield $2 code). Mixing occurrences of subfield $2 from different thesauri within a single authority record would undoubtedly be done only in a local context. Shared authority records for a given thesaurus would only have that thesaurus's code in any X55 field. (Alternatively, instead of using local 455 equivalent terms, institutions could use 755 and allow systems locally to decide what to do with the heading (e.g. suppress it, treat it as a see or see also reference etc.))

Subfield $2 in the authority record is also desirable for systems that allow global change in bibliographic records (currently most systems do this). Recently, for example, RBGENR changed the spelling of the form term "Catalogues" to "Catalogs." The string in bibliographic records reading:

  655 #7 $a Catalogues. $2 rbgenr

needed to be changed to:

  655 #7 $a Catalogs. $2 rbgenr

Systems that use the authority record to effect global change usually look at the whole string needing change. Without subfield $2 in the authority record, such systems would have greater difficulty in effecting the change than with it. Additionally, machine validation requires subfield $2 in the 455 field to allow the unauthorized heading on incoming records to be "flipped" to the valid form.

Another problem that might be solved by the addition of subfield $2 in the authority record is the problem of multiple thesauri within a single file. Although efforts at reconciliation have been made, different thesauri still contain different terms for the same concept, as seen above in the "Moralities" example. There is much interest in the creation of an international shared database of genre/form terms much like the Name Authority File, but conflicts between thesauri remain a barrier. The addition of subfield $2 to the string in the authorities format would allow conflicting terms from different thesauri to coexist within the same authority file.

Multiple thesauri may not only contain differing terms for the same concept, but usually also have differing hierarchical structures (different broader/narrower terms for the same term/concept). Adding subfield $2 to all three authority fields (155, 455 and 555) would allow a system to distinguish between thesauri in displaying the hierarchical structure so that, for example, when a GSAFD term is displayed only the GSAFD hierarchy would display, rather than mixing it with that of another similar thesaurus such as LCSH.

3. Current Situation

The MARC 21 Format for Authority Data does not currently include identification of the thesaurus in the 1XX field because the information is given in a code in 008/11 (Subject Heading System/Thesaurus). Codes are assigned for eight thesauri; if the applicable code is not given, the user is instructed to use a code in field 040 subfield $f (Subject heading or thesaurus conventions). That subfield refers to the MARC Code Lists for Relators, Sources, Description Conventions for the appropriate code. This is the same list used in subfield $2 of field 655.

When the 7XX linking entry fields were defined in the Authority Format, they were intended to include an equivalent heading in another thesaurus. Several discussions in the MARC Advisory Committee in the early 1990s established principles for the use of 4XX/5XX fields versus 7XX fields in authority records. These included the decision that a heading from another thesaurus or authority file does not belong in a 4XX/5XX field, but rather that the 7XX fields should be used to record in an established heading record the equivalent heading (or subdivision) as found in another thesaurus or authority file. The linking entry field identifies the subject heading system/thesaurus in the second indicator or subfield $2 if the thesaurus is not one of the specified values. Thus, the heading and references (1XX, 4XX, 5XX) in the authority record correspond to the syndetic structure of the thesaurus system for which it is intended.

Note that Proposal No. 2002-07 (Definition of Additional Second Indicator Values for Specific Subject Systems in Field 655 in the MARC 21 Format for Bibliographic Data) was approved in January 2002, which made the indication of the source of thesaurus in field 655 consistent with the other 6XX subject added entry fields by defining the second indicator and values. Previously, only subfield $2 was used to indicate the source in this field.

Proposal No. 95-11 (Definition of X55 Fields for Genre/Form Terms in the MARC Authority Format) defined fields X55 in the MARC Authority Format in June 1995 to allow for the creation of records for genre/form terms. The full range of fields (155, 455, 555, 755) is available and the thesaurus is coded in 008/11 or 040 subfield $f as with other authority records.

4. Suggested Changes

This paper suggests the addition of non-repeatable subfield $2 in MARC 21 Authorities format fields 155, 455, and 555, with the label "Source of term," and with an explanation similar to that found in the MARC 21 Bibliographic format field 655 for the location of the official code list. The rationale given by the Bibliographic Standards Committee of RBMS/ACRL is that the code in subfield $2 in the bibliographic record is a part of what needs to be validated by the authority record. Because of this, it would be more convenient if the subject term and the source code were in the same field in the MARC record. This applies not only to the authorized heading (155) but also to see also references (555). The inclusion of subfield $2 in the 555 field would also support the possibility of relating a term in one thesaurus to the conceptual hierarchy in another thesaurus. (This is the sort of relationship that exists, for example, between the Annotated Card headings and LCSH, the former being a subset of the latter and the AC terms often being related to broader terms defined in LCSH.)

It is recognized that most of this is possible in the current format structure, so long as the same code applies to all of the heading/reference fields in the record. As a matter of convenience and flexibility, it is proposed that carrying this data at the field level would do the job better. Libraries that place genre/form terms under authority control would benefit because it will make the process of correcting bibliographic headings and making changes easier, allowing catalogers more efficiency in their operations. Note that subfield $2 has been in use in authority records at Brigham Young University for four years now, with the desired results. The cost of implementing the code itself has been minimal, limited to authorizing the subfield for use in the system, which took only a few minutes.

Note that this proposal is directed only toward X55 authority fields. However, other 6XX bibliographic fields (e.g., 650) also require subfield $2 for identification of terms from thesauri other than those identified by indicators. Therefore, the MARC Advisory Committee might want to consider broadening the proposal to authorize subfield $2 in other authority fields corresponding to bibliographic 6XX fields. However, if this would be a stumbling block to approving the change for X55 fields, where it is needed now, discussion of broadening the proposal should be put off until it is needed elsewhere.

5. Questions for Further Discussion

5.1. Is the potential benefit worth the impact of this change?

5.2. Will it be necessary to change existing records if this proposal were approved? Or, since few libraries are not yet creating and using many genre/form authority records will the changes be minimal?

5.3. If approved would the current means of identifying the thesaurus in 008 and 040 be made obsolete?

5.4. Is there any value to changing the other fields in the authority record (X50, etc.) to allow for identifying the thesaurus at the field level in subfield $2? Should form/genre headings be handled differently than others, especially since field 655 now is consistent with the other 6XX fields since the addition of the indicator values for thesaurus?

5.5. If only X55 used the technique proposed and not all the subject fields, what impact would this have on systems that validate 655 in the same way as other 65X fields?

5.6. Is the perceived need to add subfield $2 to these fields in the authority format a result of system limitations that cannot use the current coding for the desired purpose?

  1. Guidelines on Subject Access to Individual Works of Fiction, Drama, Etc., 2nd ed. (Chicago: American Library Association, 2000)

  2. Genre Terms: a Thesaurus for Use in Rare Book and Special Collections Cataloguing, 2nd ed. (Chicago: Association of College and Research Libraries, 1991)

Go to:

Library of Congress Library of Congress
Library of Congress Help Desk ( 07/23/2003 )