Machine-Readable Bibliographic Information Committee (MARBI)

This report is in proposal and discussion paper order, followed by a report on business meeting and on task forces. The papers are linked to the agenda at LC usually adds a statement of MARBI action in the months between conferences. As a reminder, approved changes are not implemented immediately but according to a schedule determined by LC and the utilities.

Proposal 2001-01: taxonomic hierarchies in 754 (bib). This proposal was rewritten from 2000-08 to be parallel to 654 where hierarchy code is in $c rather than part of the data in $a through whatever for levels of hierarchy. $x and $z will be added for non-public and public notes. Accepted with changes and some details to be finalized between LC and FCLA (proposing organization). A subfield will be added for common taxonomic name. Michael Fox indicated that Minnesota Historical Society could use this field for taxonomic information about objects in its collection. This led to some discussion among representatives about the long ago discussion of 655 (genre) vs. 755 (physical characteristics) and the abandonment of 755 in favor of just 655 for both genre and physical characteristics.

Proposal 2001-02: non-MARC country codes in 043 and 044 (bib). Subfield c was defined in 043 and redefined in 044 for the ISO 3166 country code, e.g. US for the United States. These codes are used widely in CORC and other metadata projects.

Proposal 2001-03: identification of source in 015 and 017 (bib). These fields for national bibliography and copyright registration numbers were expanded, including $2, to allow for multiple numbers and varying agencies that assign such numbers. 015 was made repeatable.

Discussion paper 2001-DP01: narrators in 508 and 511 (bib). 508 is used for creators and producers and 511 is used for participants and performers. Narrators are sometimes both, and are sometimes present and othertimes voiceover. After discussion, it was determined that the format is ok but that the documentation could be clearer. LC will work on clarifying where to put narrators, normally 511. It is recognized that indexing which separates data from 508 and 511 will cause problems in retrieving from existing records with varying usage.

Discussion paper 2001-DP02: non-MARC language codes in 041 (bib and community info). In some metadata projects, ISO 639 language codes are sometimes used rather than MARC codes. Sometimes they are combined with country codes to differentiate, for example, American and British English. The concept of adding subfields for non-MARC codes was accepted. The proposal for Annual 2001 will probably come back with $2 for type of code and use the existing 041 subfields for uses of language (e.g. languages of text, summary, translation). Also, 041 would become repeatable with a new field being used if a different set of codes ($2 would not be repeatable). There was also some discussion of making $a repeatable rather than stacking codes when more than one language is serving a particular role (e.g. “$a eng $a ger” rather than “$a engger” for a bilingual English-German text).

Discussion paper 2001-DP03: types of dates for electronic resources. CORC and other metadata projects have been pushing the envelope on types of dates needed in MARC records, especially those being mapped from projects with other types like created, valid, available, issues, and updated (Dublin Core qualifiers). While the discussion paper particularly addressed electronic resources, some resources in traditional formats like looseleafs have similar needs to record revision and similar dates. OCLC anticipates that its URL validator might look for modification dates. After considerable discussion, it was determined that a proposal should be written to expand and broaden the use of 046 for a variety of dates. Some aspects of the data to be coded are the source of the data, the syntax of the date, and the nature of the date. We in the art and visual resources cataloging communities should be thinking about and studying the proposal to see if it would be expandable to include such dates as exhibition span that we might want to code. Coded data in 046 might be explained in a note (some 5XX field). Harmonization will be needed between bib and community info formats because ranges of dates are handled differently.

Business meeting: All formats are now available in MARC 21 versions and are available as a package. They will be in Cataloger’s Desktop in first quarter 2001 (early/mid February). The new version of LCCN with four date digits was implemented at beginning of 2001. The additional characters in the ANSEL set and spacing diacritics were added to the MARC character set. MARC 21 organization codes are available in a database using Site Search from the MARC documentation page. LC has developed a mapping between MARC and ONIX (a new publishing industry metadata element set). The proceedings from the BIBCONTROL conference at LC should be available by Annual (June). Issues from the preconference of particular importance to MARBI are multiple versions and semantic interoperability. LC is now looking at the recommendations and determining what can be done short or long term and which need to be done by LC or by others.

Other business: MARBI will co-sponsor a preconference workshop on serials in the Holdings format at Annual 2001 in San Francisco. Suggested topics for a joint meeting with CC:DA at Annual included: recommendations from BIBCONTROL conference; chapter 12 rewriting (LC anticipates some proposals for Annual on seriality). The East Asian Character Set Task Force has almost completed its work. The CJK set includes 26 punctuation, 36 component, 13368 ideograph, 172 kana, and 2228 hangul characters (total of 15730), just in case you were wondering why they didn’t finish overnight. They have established some private use values in order to do roundtrip conversion (36 components, 231 ideographs, and 32 hangul). The task force website will be linked to the MARBI website. The Unicode Task Force is also about done, with some final reviewing and then vote by MARBI.

The Multilingual Records Task Force is working on the specifics of examples. Broadly speaking, they are moving away from the single-record model (everything in one record) to separate linked records. They are looking at how to link records for the same thing in different languages, in different scripts, in different vocabularies, etc. An IFLA subcommittee working on this has also decided to abandon the idea of a single record and suggests using a record-level code for the language of the catalog.