January 2023 (View As PDF)
Stephen Early, Cataloger, CRL
Wen-ying Lu, Head of Cataloging, Santa Clara University
Cecilia Genereux, Continuing Resources Metadata Librarian, University of Minnesota
Cynthia Whitacre, Senior Metadata Operations Manager, OCLC
Hank Young, Principal Cataloger, University of Florida
Dana Jemison, Principal Metadata Analyst, California Digital Library
Sara Amato, Project Manager, EAST
Amy Wood, Director of Discovery & Technology, CRL
Alison Wohlers, Shared Print Program Manager, California Digital Library
Why this matters
Multipart monographs (also known as monographic sets, monographic series, or monographic serials) present a unique set of problems for Shared Print. Cataloging rules and practices allow these resources to be cataloged at the serial/series level or the individual monographic level. Libraries choose to catalog at one or the other or both levels to support access for their users. While helpful for discovery at individual libraries, these differences in cataloging create challenges when aggregating the metadata from many libraries for Shared Print.
Shared Print depends on knowledge and comparison of libraries’ holdings of distinct titles. Monographic series cataloging practices complicate Shared Print analysis, decision-making, and subsequent discovery and access. The challenges associated with monographic series have caused most shared print programs to exclude or avoid this category of content. This is a significant gap.
The Rosemont Alliance and the Partnership for Shared Book Collections – as bodies putting forth unifying best practices and advocates for improved infrastructure – have an opportunity to coordinate the development of resources to support programs and libraries in grappling with this challenging category of content.
The existent problems with multipart monographs are outlined in detail below. Potential solutions and opportunities for handling these challenges are also offered as a starting point for further discussion.
Areas of Concern:
Cataloging Issues / Levels of Cataloging
This section outlines areas of cataloging that can be problematic.
- If the library has only a single volume of an analytic it is likely to catalog the title as a stand-alone monograph and not also keep a set/series/serial record. In support of analyzed volumes, volume level cataloging allows for richer, more accurate subject headings aiding in discovery.
- The 830 series entry in a monograph bibliographic record should ideally match the 130 of the corresponding series authority record entry and either the 245 00 entry or, if present, the 130 0 entry in the corresponding serial bibliographic record. This doesn’t always happen, however, for a variety of reasons, most commonly in the case of serials cataloged according to pre-1977 cataloging practices in which different title change and authorship rules were in effect. Any record with an 830 usually includes a 490 (or 533 $f if the 830 is for a microform series) which transcribes the series as it exactly appears on the piece being cataloged. A 490 and its 830 may match, but not always. Variations in the series title as expressed in the 490 may appear as cross reference entries in the series authority record.
- Ideally, monographs are cataloged such that their 830s can be linked to the series set record or serial parent title.
- Basic problem is that individual monographs aren’t always cataloged for a set/series/serial, or not in a way that can support automated attempts to pull titles together.
- Example of a real world 3 way mismatch in OCLC between a series authority record, serial record for series, and monographic record in the series:
Series authority record (current)
130 0 Research report (University of Sydney. Department of Agricultural Economics)
Serial record for series (pre-AACR2) (as of November 2022)
110 2 University of Sydney. Department of Agricultural Economics.
245 10 Research bulletin.
Monograph record in the series (as of November 2022)
100 1 Fallding, Harold.
245 10 Social factors in serrated tussock control; a study of agricultural extension.
830 0 Sydney. University. Agricultural Economics, Department of Research bulletin ; no. 1.
- There is inconsistency in tracking parent series in analytics and subseries. Parent titles are captured in 4XX (only 490 is currently valid in 4XX range for bibliographic data; all others are obsolete, but may still be present in legacy cataloging records) and 800-830 fields, but this is not always consistently done. There are also other fields 7XX fields that can be used for tracking series/analytic relationships:
- 773 is used by Alma to show the relationship between an analytic and its parent monographic serial/series record. However, 773 is usually lacking.
- 773 was designed for article, chapter level cataloging. Used to be called “In Analytics”. 773 usage for analytics in multivolume sets is unusual and non-standard.
- 773 is similarly structured to MARC 533, reproduction note.
- It can be unclear which MARC fields should be used under which circumstances.
Conversely, it is difficult to determine volume analytic titles for a set/series/serial.
For sets, this can be captured in a 505, but this is spotty and not programmatically feasible for batch or programmatic analysis. For series and serials one can track a series relationship back from an analytic if the data is there, but there is not a uniform way to know what titles comprise the series. While it wouldn’t be realistic to track analytic titles as part of a serial record (as added entries, for example!), there could possibly be some way to track using an intermediary linking data record that could show these relationships. As mentioned above, some institutions are using the 773 for this purpose locally, for example, FLARE and some Alma libraries, though it is labor intensive and not generally used in, and neither recommended nor discouraged, by OCLC.
The MARC 876 field, while defined by MARC formats and standards, is part of the MARC holdings record block and meant to express item level data (item record number, barcode, copy number, volume number, circ status, etc.) not title or other bibliographic level data. See “potential solutions” section for further discussion.
- While MARC bibliographic Leader 19 might help in identifying analytics/analyzed sets/series/serials, it isn’t normally used in North America. It is used extensively in Germany. Is this something that could be used?
- Not implemented in OCLC.
- Codes defined to indicate if the record is a dependent part (analytic title) or independent part (series/set/serial title.) See “potential solutions” section, below, for more discussion.
- Institutional holdings are not uniformly set at either the series/set or analyzed records. Practices vary widely. In theory holdings could be set in OCLC for both the series/set records and the analyzed records. Note that this would be a policy for individual institutions to decide, as OCLC does not mandate or even recommend a specific practice regarding this issue. Some areas to consider are:
- Individual policy okay by OCLC for setting holdings at any level or both.
- Holdings in OCLC set on both Series and Analytics could be problematic in terms of count and maintenance. Does this cause statistics issues for libraries in terms of how many titles or volumes are held? Does this skew the number of copies held locally/globally? Affect counts? Produce bad statistics?
- Holdings set on both set and analytic level is helpful for discovery.
- Need to consider what happens when items are withdrawn, if the library does not hold both the monograph and serial/series record. How would you know where to remove holdings in OCLC if you only had one or the other cataloged locally, not both?
- Inventory and circulation problems if there are multiple records for one title.
- Potential issues with determining monographic/analytic title scarcity where analytics aren’t marked as “held” in OCLC but the set/serial record is.
- It’s critical that library staff (ILL, reference, etc.) should understand the issues no matter how holdings are set/local cataloging practices are laid out, and make sure that locally held titles are in sync with WorldCat holdings.
- Example of a local practice: In the local catalog, but not always consistently, CRL does both series and title level records which are linked via the item record. In WorldCat the series record has an LHR, but the monograph record does not.
- There are occasional mismatches between series authority records and their serial bibliographic record counterparts. For example, if the serial is cataloged according to latest entry convention so that multiple title changes are covered by a single record, but the SAR (series authority record) is based on successive entry resulting in multiple SARs. Splitting the legacy serial record to match the multiple SARs may be difficult due to lack of access to issues covering certain title changes. Furthermore, there may not be a SAR for every single potential serial title. This can occur with pre-1977 monographic series, and especially 19th Century monographic series.
Interlibrary Loan Implications
- A title may be requested in either its analytic or serial title form. Unless the request matches up with the form, then a false negative may occur where the alternate form is actually available for loan. For example, patron requests serial title v.1, but the library only holds that volume cataloged in its analytic form. Patron is told that serial title v.1 is not available.
- Another example, patron requests analytic title v.1 but the library classified it together with the serial/series title and labeled it as v.314 of the series. Patron is told the analytic title v.1 is not available. Volume numbering assigned by the library for shelving purposes on the spine label/OPAC display can be confusing regarding the actual holdings. Depending on how the holdings data is gathered and/or recorded for the MARC field 583 subfield 3, for example, on an analytic title, it should be v.1-v.2 but it may be recorded as v.314-v.315 if the holdings is gathered (via a script) based on the volume field data in the item records (e.g., Sierra item data converted to Enumeration in GreenGlass). Similar issues also appear for those classified with “etc.” or individual volume designations, e.g., “v. 22, v. 26, v. 34” or “v. 23-v. 25” at the end. When registering holdings in WorldCat with Local Holdings Record (LHRs) Data for the Shared Print programs it is highly possible that the holdings in the 583 subfield 3 are not recorded consistently, depending on whether using the numbering from the analytic title or set/serial title. This creates challenges for determining overlap within a shared print program.
Discovery and Access
Library search and access issues are similar to the interlibrary loan implications described previously (bullet one). With librarian help, the patron may be able to find materials, but only if the librarian on hand understands these types of relationships and knows how to make these kinds of connections within the local catalog. Some Alma libraries use the 773 field to link the analytic record to the series/set. For records that have the 773, users can see that there is some type of relationship and where the volume physically resides with the series/set.
In cases where analytics records do not have a 773, problems may arise, for example, where there is shifting of print materials to offsite storage facilities. When series/sets lacking 773s are moved, the analytic record becomes orphaned, because the analytic no longer points to the location of the series/set record. While these analytic records can be identified and corrected, correcting these records are so problematic and therefore labor intensive, that it may be deemed preferable to catalog the volumes as monographs with no series/set record if it moves to offsite storage.
Also, in an ILS where a single piece can’t be linked to multiple bibliographic records, problems arise in both discovery and access.
ILS/LSP Specific Issues
Differences in how various ILSes/LSPs handle multipart materials can help and/or hinder their discovery and access. Below outlines a few of the issues:
- Problems arise where the barcode can’t be linked with multiple bibliographic records in some ILSs (Alma, for example), though they do offer a workaround by using the 773. Being able to link a single barcode/item record with multiple bibliographic records allows one to associate a single piece with both the set/series/serial and monographic titles.
- ‘Bound with’ records, while not multipart monographs per se, are also problematic and handled differently across different platforms. These can be especially problematic in collection analysis if all titles are not represented in records that are used or exported for analysis.
- Holdings on both the set and analytic level cause problems for inventory as some pieces may be counted twice and for circulation, since in many systems barcodes/item records may be linked to only one bibliographic record rather than multiple records.
- Problems can arise when portions of analytics and the set/serial are housed in multiple places. For example, where some analyzed volumes are moved to storage, while others are still housed on campus as part of the set/serial, a barcoded piece can only reflect one or the other location, but not both.
Shared Print Implications
- Selection for collection analysis and the collection analysis process becomes difficult when a monographic set, series, or serial has not been analyzed in the case of monograph shared print, or where analytics are cataloged in lieu of cataloging and summarizing as part of a monographic set, series, or serial. Libraries may use a hybrid approach, where the monographic set, series, or serial is received on the monographic set, series, or serial title record, but then volumes are sent to be analyzed without also updating summary holdings/item data for the serial title. In this case, summary/volume holdings for the monographic set, series, or serial title can’t be parsed into a MARC holdings record without intermediary scripting, creation and mapping of data outside the ILS.
Collection analysis ramifications on counts of existent copies depending on which level title (monograph vs. set/series/serial) has been cataloged. For Shared Print programs this can result in multiple copies being retained (over retention) as a collection analysis may not catch that the title is already retained, or the title may appear unique to the group when in reality it is contained within a parent set.
- Libraries may disclose/register materials at the individual title level or at the series level, or both. There is no agreed upon best practice for this. Registering at both the title and series level would be a considerable effort using the tools we have available at this time, and may not be desirable given the issues with counts outlined previously in this document.
- Problems can arise when a library re-catalogs a title. For example, an institution committed to retain a series and disclosed it to WEST, PAPR, and OCLC as a series, but later re-cataloged it as individual monographs. They are still disclosing it as a series title using a dummy record that’s suppressed in their local discovery layer, but this is problematic for resource sharing as individual monographs.
- In cases of subseries, that is, series within series, or series within serials, multiple numbering schemes may be expressed. For example, Series A volume designation = v.243 and Subseries B volume designation = v.1.
MARC holdings fields 583$3 and holdings 86X contents should express the correct numbering for the bibliographic title being described:
245 \\$aSeries A
022 \\$a<Series A ISSN>
035 \\$a<Series A OCLC#>
245 \\$aSubseries B
022 \\$a<Subseries B ISSN>
035 \\$a<Subseries B OCLC#>
Problems arise where Library I has submitted Series A and Library II has submitted Subseries B, because in effect, these are duplicate volumes. That said, the shared print program may desire title representation for both Series A and Subseries B, or one or the other only. Ideally only one volume is submitted, but both titles and both volume designations are captured; Library I submits v.243 of Series A, but, we also know that this is also v.1 of Subseries B.
- Could OCLC come up with a solution to setting holdings such that holdings could be set for both the analytic and set/series/serial record but without the drawbacks? Some sort of automated linking “thing” that does this work in the background, and would also remove these relationships when withdrawing, or even allow the library to “override” the default such that they still only set holdings on either the analytics or the set/series/serial? OCLC would have to be able to correctly interpret LHR holdings in order to set analytics correctly, and/or library staff would have to have the means to “check off” which analytic volumes are held, for example.
- Could a linking data record be used to show the relationship of titles in an analyzed set? (Like an authority record for set.) Serial records have no way to know individual titles, though could be at the institutional level with linked items.
- Could MARC 876 be expanded to include OCLC number (and/or other unique identifiers) in order to build links from series/serial/set title holdings item data to individual item bibliographic data? This would allow one to see the relationships, if any, between a title set/series/serial record and associated monograph titles.
- Could OCLC retroactively assign LDR/19 to identify analyzed volumes where safe to do so (Though it is an open question how to determine this). Currently, there is no fixed value available for input in OCLC records. Note that the use of this element would be a major change for North American monographic cataloging practice and would likely take years to penetrate the cataloging world. OCLC has looked at retroactively assigning it and they believe it would be very difficult. If there is a major push by the cataloging community to use this, OCLC would certainly try. However, with Bibframe on the horizon, it’s suspected that coding this in MARC would not rise to a high priority. Finally, how to encourage folks to use this value? PCC? LC?
19 – Multipart resource record level:
# – Not specified or not applicable
a – Set
b – Part with independent title
c – Part with dependent title
- Regarding disclosure at the individual title level vs. series title, ideally it would be best to do both and link between them. Staffing shortages have made that very difficult at this point. Possibly there are alternative solutions that involve working with vendors, so that when you’re receiving a serial or buying an individual piece for a series, the vendor also provides the analyzed records. Another solution might be sending serial holdings to a vendor and they identify what the analytics are and package those up for libraries.