Authority Control Advisory Committee

Authority Control Advisory Committee Minutes

Date: 2001-07-30

Present: Manon Théroux (chair and recorder), Steven Arakawa, David Faulds, Robert Killheffer, Beatrice Luh, Cécile Mandour, Earl Roy, Richard Sarcia, Dajin Sun.

Absent: Ellen Ellickson, Mickey Koth, Audrey Novak.


Last year, wanting to formalize its terms of membership, the committee decided that members should serve two-year renewable terms, staggered so that no more than half of the committee turned over each year. This decision is now being implemented. Four new members have been appointed: David Faulds (Beinecke), Beatrice Luh (EPH), Earl Roy (History & Social Science Team, SML), and Richard Sarcia (Arts & Sciences Team, SML). Two current members, Steven Arakawa (Arts & Sciences Team, SML) and Robert Killheffer (History & Social Science Team, SML), will rotate off this year but have agreed to delay leaving until the QA of Basefile #2 has been completed. Four former members have already left during the past year: Paula Ball (Medical), Diana Smith (Beinecke), Rebecca Mugridge (Catalog Management), and Martha Conway (Catalog Management). Martha had been a replacement for Rebecca; both have since left Yale. Diana is currently on maternity leave. Continuing members include: Ellen Ellickson (Rare Book Team, SML), Mickey Koth (Music), Cecile Mandour (History & Social Science Team, SML), Audrey Novak (Database Administration), Dajin Sun (Catalog Management Team), and Manon Theroux (Catalog Management Team and Rare Book Team, SML). A list of current members and their terms of service is available on the ACAC website.


Manon distributed a sheet containing the committee's original charge followed by a revised charge that she had prepared for Joan Swanekamp's approval. The new charge broadens the language of committee's scope somewhat to include any issue relating to authority control in Orbis. The language of the original charge suggested that the committee's role was limited to issues of vended authority control, but the committee sometimes assumes responsibility for other projects (e.g. subfield $v implementation presentation). The new charge also specifies the committee's QA responsibilities and incorporates the terms of membership outlined above, including the recommendation that the committee always have representation from the following units: Database Administration (Systems Office); NACO Coordinating Committee; Arts & Sciences Team or History & Social Science Team (SML Catalog Department); and a school or departmental library. The new charge has been posted on the ACAC website.


Last fall, OCLC decided to discontinue its Authority Control Service (ACS), which Yale had been using since 1997. The ACS service was replaced by the MARS service, which OCLC acquired when it bought WLN. MARS performs many of the same basic functions as ACS:

--corrects headings in our bib records;
--sends us associated authority records;
--sends us updates of authority records that we already have; and
--sends us various reports that we use to maintain authority control in Orbis

Like ACS, MARS provides us with heading-level replacements (called "correction transactions" or "transaction records") rather than full bib record replacements. This allows us to update bib records in Orbis during the processing period without losing information due to subsequent record overlay.

Differences between the services include:

--MARS can process conference and uniform title headings (ACS could not);
--MARS sends us authority records for our local 692/693/694 subject headings but does not correct these headings (ACS attempted to correct our local headings)
--MARS does not automatically intercept and send us our NACO records (ACS did); instead MARS sends us our NACO records only when they are "claimed" during processing
--MARS changes 440s with filing indicators to 490/830 combinations even if the only difference is the presence of an initial article (ACS retained the 440s)

A summary of our processing instructions for 6xx headings can be found at:


A "snapshot" of our Orbis bib records (called the "Basefile") was sent to OCLC for processing last fall.  We have received correction transactions for our GPO, in-process, and suppressed records (Basefile #1). We finished QAing these and they are ready to be loaded into production Orbis.  We expect to receive correction transactions for the rest of our bib records ("Basefile #2) soon.

A "snapshot" of our Orbis authority records was also sent to OCLC last fall. These records were compared (using the 010 field) against the "Master File" of authority records that OCLC claimed during the processing of our Basefile records. OCLC sent us back a half dozen files of authority records, corresponding to various parameters. We have QAed the first of these files (which contained authority records that were totally new to Orbis) and it has been loaded it into production Orbis. We still need to QA the remaining files and make decisions about them.

Because our recon records go through MARS processing before being loaded into Orbis, we are able to use full record replacement techniques for them rather than correction transactions. Our recon records had been piling up at OCLC, waiting to be sent through MARS processing, but we are now going ahead with the processing. We just received our recon records from November through May (and their associated authority records). They are currently being loaded into ORBDEV7 and we will QA them ASAP. They are our highest priority right now. If all goes well, we will then load the recon records into production Orbis. Ongoing processing of our recon records should then progress smoothly and no more recon QA by the committee should be needed.

A detailed summary of our implementation progress can be found at:

Steven asked whether Basefile #2 will include Pinyin records. Manon thinks so, but she will find out. Currently our CJK recon records are not being processed by MARS but that may change.


The recon records, both bib records and authority records, are currently being loaded into ORBDEV7.  We will QA them soon. Each person will probably be asked to look at 100 bib records. The purpose of the QA will be to make sure that headings have been changed correctly NOT to evaluate the quality of the descriptive cataloging. Instructions will probably be sent via e-mail once the ORBDEV7 load is complete.

Robert asked how many recon records were in the batch waiting to be QAed. Answer: 358,567 bib records.


The following will need to be done at some point after the Recon QA:

--QA of Basefile #2
--QA of authority records
--revision of both workflow documents (need to change instructions for conference headings and uniform titles; need to provide instructions for retagging 928 when completing cataloging of in-process records that already have 928; need to take into account the time gap between when records are produced in RLIN and when they actually appear in the national authority file and can be used in MARS processing--if the processing occurs before the record is out of pending phase, then we will not receive the authority record because OCLC is no longer providing the NACO intercept service)


We usually meet on the last Monday of each month, from 10-12, in SML Room 411. Because of a room conflict, our next meeting (on Aug. 27) will be in Room 410.

Site URL:
Comments to: Manon Théroux  <>

© 2001 Yale University Library

Top of Page