This page documents cataloging related questions and issues that affect OCLC Data Sync implementation, and the decisions that have been made in consultation with Metadata Standards working group and other stakeholders.
Manuscripts and Archives
OCLC Data Sync guidelines for manuscripts and archives (OCLC criteria)
(Email from Carrie Morrison, OCLC Metadata Operations, Consulting Database Specialist on Oct. 17, 2019). Criteria for Archival/Manuscript material include:
- Archival Control (Leader/08) value “a”
- Record Type (Leader/06): Can include a, t, c, e, f, d, p
- Type t – Manuscript language material, including all unpublished. i.e. not commercially published, theses
- Type f – Manuscript cartographic material
- Type d – Manuscript notated music
- Type p - Mixed materials
040 $e values for manuscript and archival descriptive rules of cataloging
- See the code list for Description Convention Source Codes.
- For 040 $e codes that have been superseded, if the cataloging of the record has been updated to follow the new code, then the 040 $e should be updated. If not, then the code should stay as-is.
- General guidelines from OCLC’s Bib Formats:
Material | Guideline |
Archival and rare material | Determine the form of material and use the appropriate code. For example, for a rare book, use code a. For a manuscript letter, use code t. For all unpublished theses, use code t. For a collection of mixed materials, use code p. For online resources that are reproductions of manuscripts, use codes a, c, and e (the original format code). Code Ctrl to indicate archival control status of the material. |
Recommended Best Practice for Harvard's MARC bibliographic records for manuscripts and archival collections to sync with OCLC's Worldcat:
- Original cataloging and record enhancements should be done in OCLC Connexion. Archival and manuscript bibliographic records will be eligible for own record replace, which allows our manuscript records to replace matching records in WorldCat provided these three criteria are met:
- Matches on OCLC number
- Matches on Date Entered (008/00-05)
- OCLC symbol in 040 $a and $c of the WorldCat master matches the OCLC symbol of the library represented in the 852 of the incoming record.
INCLUDE THE FOLLOWING FIELDS: - Type of Record (Leader/06) values d, f, p, t NOTE: If the material is a reproduction and has appropriate 008 and 533 fields than the record will NOT replace the OCLC master record.
- Type of Control Leader/08) value “a” (Archival) The material is described according to archival descriptive rules, focusing on the contextual relationships between items and on their provenance rather than on bibliographic detail. All forms of material can be controlled archivally.
- 040 subfield $e dacs or appropriate manuscript and rare materials cataloging conventions (dcrmmss, dcrmb, dcrmc, dcrmg) The specific set of rules for description may be found at http://www.loc.gov/standards/sourcelist/descriptive-conventions.html
Summary of past and current coding practices for Manuscripts and Archival Collections at Harvard
| Questions (hl-cat 11-22-2019) | Archives | Baker Library | Berenson | Divinity | Houghton | Law | Schlesinger |
|---|---|---|---|---|---|---|---|
| Which descriptive conventions are used? | Current practice is DACS and sometimes dcrmg. About 50% of records have no 040, or no 040 $e. Many have $e dacs or $e local (prior to introduction of DACS in 2005). | Majority of archival and manuscript collections are coded appm or dacs. We also use dcrmmss and amremm for manuscript material in our rare book collection. | Current practice is dacs for collections and single item manuscripts. Previously used appm. | Currently use appm, but many manuscript records lack $e in 040 | Current practice is dacs for collections and single item manuscripts. Previously used appm. | ||
| If using the rare book descriptive conventions (like dcrmb), are the record coded "a" in Leader 08 (archival control) | Do not use rare book descriptive conventions. | Generally no. | No | No | |||
| Do you code "a" in Leader 08 for manuscript materials (archival control)? | Yes, for current cataloging. Older records pre-date this byte definition. | Generally yes. It is the current best practice for materials cataloged according to dacs. | Yes. | No, very little use of archival control byte | Yes | ||
| Is original cataloging done in Connexion? | No, original cataloging is done in Alma. | Yes | No, original cataloging is done in Alma | Yes, for the past two years | Yes | ||
| Are enhancements done in Connexion? | No, all work is done in Alma. | Generally yes, but it's a new best practice. | Not currently. | Yes, for the past two years | Yes |
Cataloging reproductions of Manuscripts
Facsimiles
- Describe as format of reproduction, not as manuscript
Leader 06: (Bks/language material) a
Leader 06: (Scores/Notated music) c
Leader 06: (Maps) e
If record for original ms exists (not always or even often), ADD:
77508 $i Facsimile of (item): $a [Main entry heading]. $t [Title]. $d [Place, publisher, and date of publication]. $h [Physical description] $w (OCoLC) [Record control number] (Use insert from cited record option)
Microforms and CD-ROMS:
- Describe as the original manuscript
- Leader 06: Coded as manuscript material (t, d, f)
- 008 Record type: a Microfilm (use the correct record type code for the format)
- 533 $a Microfilm. (+ subfields for publication and extent)
Former practices (Multiple formats are on 1 bibliographic record): - Physical Description Fixed Field (007): is 'h' (Microform or appropriate code for form of item)
- Reproduction note (843) in holding record: $a (Microfilm, Microfiche, Electronic reproduction, etc.)
- Data Sync publishing: Form of item is coded a or b depending on 843 $a and 007
- Data Sync publishing: Form of item is coded q when 007/00-01 is 'co'
If record for original ms exists (not always or even often):
77508 $i Reproduction of (item): $a [Main entry heading]. $t [Title]. $d [Place, publisher, and date of publication]. $h [Physical description] $w (OCoLC) [Record control number] (Use insert from cited record option).
Theses
Theses are not sent in the Data Sync collection for manuscripts regardless of Leader Type coding provided a 502 Dissertation Note is present in the record. The 502 note is used to exclude records from the manuscript collection.
NET holdings
| Location code | Data Sync status | Comments |
|---|---|---|
| COLIN | Do not send | Holdings for eresources formerly represented by holdings and now primarily represented by portfolios. Per Lauren Syer do not include in Data Sync |
| INPROC | Do not send | |
| BLH | Do not send | |
| RDM | Send |
Provider neutral records - background
- Email from Robin Wendler, 11-22-2019: Steven Riel estimated that for e-journals and continuing e-resources, 75% of new work is provider-neutral in line with the LC-PCC Policy Statement 1.11. In their normal processing today, the bibliographic record in Alma will contain the OCLC pn #, but holding will not have any OCLC #, if that affects how we plan for the export/import of these records.
- Excerpts of correspondence between Isabel Quintana and Jay Weitz (OCLC) regarding provider neutral on 12-12-2019.
Isabel: I have, what I hope is, a quick question about OCLC merge and provider neutral records.
Here’s the situation. We created a provider neutral record for a print reproduction following RDA PCC PS 1.11. It is OCLC no. 1019825281. Then we went into our ILS and brought in this OCLC record and added our holdings with an 843 with the provider information. I believe that holdings eventually got taped out to OCLC., and another OCLC record was created = OCLC no. 1019898138. There is an 040 $d OCLCQ on the 2nd record. I’m wondering if there is an OCLC quality control process that reads the 040 $e pn and removes the 533, which would have been sent to OCLC from our 843 in our local ILS, that has the print reproduction provider information. If not, I’m not sure why our 843 doesn’t appear as a 533 on the 2nd OCLC record.
In any event, the records are now duplicates. In order to avoid creating duplicates in OCLC, Harvard was thinking of not taping out to OCLC any records from our ILS with 040 $e pn. However, I don’t think that’s a good solution because, if we find a pn record in OCLC and update the cataloging on it and bring it into our local system, our holdings will not get added to OCLC, unless they get added when we tape out our holdings each month. I know that Harvard hasn’t taped out our holdings in a long time, but when we will start again at some point. I’m trying to figure out what will happen with our pn records.
Do you have any information you can share with me about this? Can you tell me how the OCLC dup record 1019898138 was created and whether it had another 533 at some point?
If any of this is not clear, or if you need more information from me in order to investigate this, please let me know. Thanks, in advance, for any advice you can provide!
Jay Weitz, Cc: Morrison,Carrie <morrisoc@oclc.org>; Luo,Ling <luol@oclc.org>; Chirakos,Anthony <chirakoa@oclc.org>; Savage,Michael <savagem@oclc.org>; Bremer,Robert <bremerr@oclc.org>; Ramsey,Laura <ramseyl@oclc.org>
#1019825281 was added manually through Connexion on 20180117; #1019898138 was added two days later (as ELvl M) on 20180119 through the old Metadata Capture (MDC) system that was in production at the time. They did not match at that time because #1019825281 had the “Print reproduction” 533 field, whereas #1019898138 had the full 533 note until it was edited a few days later on 20180122 via the Connexion client. The different 533 notes on January 19 resulted in the records not matching.
The current Data Sync system works differently. My colleague Carrie Morrison (who has been copied) notes that we’re working with Harvard on setting up a Reclamation project, which will involve testing. We can include this particular example and/or similar ones to see what would happen now under Data Sync and to improve the results.
We’re not exactly sure why the records were not subsequently matched after #1019898138 was edited on February 13, 2019, but we’re looking into that.