American Library Association
Midwinter Meeting (2004 Jan. 9-12 : San Diego, Calif.)
Report on MARBI
(Machine Readable Bibliographic Information Committee)
Saturday a.m. and Sunday p.m.

The agenda for the MARBI meetings is available at and includes links to each of the papers discussed below. The proposals and discussion papers are listed here in numerical order rather than in discussion order.

Proposal 2004-01: Making Subfields $e, $f, and $g repeatable in Field 260 of the MARC 21 Bibliographic Format These subfields are for describing information about manufacture of a resource. While they are rare in modern materials, they are common before the 20th century. The examples show fields recording multiple agents in the manufacturer, e.g. electrotyper, stereotyper, printer. This could be useful for some visual works such as posters and other printed graphics. Many other visual resources are not manufactured in the same sense as printed material, but there are certainly different roles played in the realization of an art work and this might have application if using MARC for VR records. The proposal was approved as written.

Proposal 2004-02: Defining new field link type codes for Subfield $8 (Field link and sequence number) in the MARC 21 Bibliographic and Holdings formats This proposal arose from RLG’s plans for migration of its database. The current RLIN system uses an archival control segment which integrates action information (e.g. received, processed, available for use) and source of acquisition information (e.g. donor, seller). Since these pieces of information occur in at least two different MARC fields (541 and 583), it is necessary to have a linking and sequencing device. Subfield $8 is used elsewhere for this sort of linking. Since some systems reorder fields, a code was also proposed for general sequencing. There was significant discussion that systems should not rearrange certain fields or groups of fields, while at the same time recognizing that some systems do and that fields can go awry during data migration. A sequencing device would ensure field order. The proposal was approved with the action code generalized to provide link/sequence from one 583 field (action) to another field (not specifically 541, also including 561 for custodial history).

Proposal 2004-03: Designating the privacy of Fields 541, 561 and 583 in the MARC 21 Bibliographic and Holdings formats This also arose from RLG’s migration process. The ARC segments have had provisions for privacy of the information beyond and within the custodial institution. OCLC makes these fields private except for the holding institution. Indicators were approved to provide privacy at the field level since some information has value for provenance while other information is confidential. Implementation of this indicator will probably mean that all relevant fields on existing records will be set for private. Prospectively, catalogers and archivists will be able to code for private or public.

Proposal 2004-04: Definition of Field 258 (Philatelic issue date) in the MARC 21 Bibliographic Format This proposal from the National Archives of Canada would provide for recording the issuing jurisdiction and denomination of postage stamps and other philatelic materials. There was significant discussion of where this information belonged and whether it was similar to other information in 25X fields (e.g. Area 3 of ISBD, edition). The Canadian Rules for archival description (RAD) uses Area 3 for this philatelic data. MARC alternatives such as 037 or 65X or 75X were proposed and ultimately rejected. The field as proposed mixes descriptive information and access which has often gotten MARC into trouble (principally 4XX/8XX). On a first vote, the proposal was rejected but the discussion continued. Another vote was held and the proposal passed. The reversal, as so much of MARBI discussion these days, was most amusing.

Discussion paper 2004-DP01: Changes needed to accommodate RISM data -- music incipits This proposal came from the RISM Zentraldirektion and the Music Library Association. RISM uses a scheme known as “Plaine and easie code” to precisely identify musical works by incipit, and they need a place to record this data. The paper proposed Field 031 for this purpose. Discussion included the relationship to FRBR though it is recognized that these incipits may be neither work nor manifestation. There was general support and this will come back as a proposal.

Discussion paper 2004-DP02: Applying Field 752 (Added entry -- Hierarchical place name) for different purposes in the MARC 21 Bibliographic Format Field 752 is currently mostly used for cartographic materials and for production place in hierarchical form. The paper proposed extending coding of hierarchical place name to topical coverage, which is normally covered in 6XX fields without hierarchy. Discussion centered on whether the current hierarchy breakdown is appropriate, e.g. $a United States $b New York (State) $c Allegany County $d Alfred; $a France $d Paris; historical names and hierarchies; relation to thesauri such as the Getty Thesaurus of Geographic Names; use of indicator(s) for content/display. OCLC and RLG expressed concern with the inclusion of topical subject terminology in 7XX. This paper will come back as a proposal with two options (7XX or 6XX/7XX) and with a discussion of authority implications. LC uses authorized forms from NAF at each level of the hierarchy which is included. LC is using the field in digitization projects as well as on cartographic materials. MODS also calls for levels higher than nation ($a).

Discussion paper 2004-DP03: Changing the mapping for the double-wide diacritics from MARC8 to Unicode/UCS from the Unicode/UCS half diacritic characters to the Unicode/UCS double-wide diacritic characters This discussion paper also came from RLG and is related to conversion of data from the existing MARC character set to Unicode. The latter will allow richer display of vernacular characters as well as some symbols such as the section marker and ellipsis. Current practice calls for inputting double-wide ligatures and tildes in halves before each of the two characters in the combination. For example, the cyrillic ia gets the left ligature before the i and the right ligature before the a. Tests in OCLC found 23,000 cases of mismatched double-wide diacritic characters (e.g. one half was there and not the other). Tests in Northwestern, not a particularly strong Slavic collection, also showed significant numbers of mismatched double-wides. The double tilde was used in Tagalog (Philippines) but has not been used extensively since 1898. There was significant support for the proposed use of one double-wide diacritic character rather than the current split characters. A proposal will follow, perhaps in conjunction with other considerations about Unicode conversion.

Following the discussion of this paper, there was a discussion of a report by Jack Cain (consultant to LC) entitled “Assessment of options for handling full Unicode character encodings in MARC 21.” The paper was only released a few days before the conference so discussion will continue. The current MARC8 character set has about 17K characters while Unicode has more than 200,000. Cain recommended Option 4, and it was stated that this was what the industry is following.

Business meeting Update 4 is produced and will be distributed soon. New cumulated editions of the formats will be released this year. New language and country code lists have been issued since summer 2003. Understanding MARC authority records has been published, similar to Understanding MARC bibliographic records; copies are available from CDS. MODS 3.0 was posted a month ago and the guidelines have been updated. A new version of the FRBR experimentation tool was released later in January. Rich Greene summarized the FRBR preconference which will be held at Annual.

Another amusing MARBI moment: when someone mentioned the old slogan “the future is longer than past,” someone quipped “not for all of us.”

... go to other ALA reports ...