Dear PCC Colleagues,

I am writing to tell you about some recent processing changes at Cornell
University Library.  Some of you may have heard about these changes before,
but because their impact extends beyond our walls here, I thought you might
be interested in some further information about what we are doing and some
of the Cornell records you may see in OCLC and/or RLIN.

Following the example of Stanford University, Central Technical Services
(CTS) at Cornell University Library introduced classification-on-receipt
(COR) in early January 2003.  Under class-on-receipt, newly-acquired
materials are reviewed by catalogers as they arrive from acquisitions.
High-priority items and non-book formats are given original cataloging at
the appropriate level. The remaining materials receive a level 3
bibliographic record with a full LC call number and uncontrolled index
terms (where necessary) for keyword searching access.  We then send them to
the stacks.  With the exception of some CJK items, which have a distinct
workflow, no materials are being backlogged.

Using the RLG/MARCLink Marcadia service, the resulting level 3 records are
being run against the RLG Union Catalog four times over the course of 24
months.  As fuller copy becomes available, most of the level 3 records will
be overlaid, though classification numbers will not be changed.  At the end
of the 24-month cycle, we anticipate that fewer than 10% of all of our
receipts will have a level 3 bibliographic record.  We currently have no
plans to retrieve items from the stacks and upgrade them should they lack
better copy after the 2-year Marcadia cycle.  Items that fail to be
overlaid will be represented in our catalog with their COR records
indefinitely.  Because we are exporting these records to both OCLC and RLG,
the level 3 COR will also be available to users of those utilities.

Those records are very similar, though not identical to, minimal-level
records.  Catalogers and classifiers are instructed to verify or update
several of the fixed fields  (e.g., encoding level, date code(s),
language.)  They are also expected to verify or update variable fields
critical for access and the Marcadia matching routines.  These include the
010, 020, 245, 260 and 300.  Finally, in many cases, uncontrolled index
terms are added to the 653 field to provide keyword access to the
items.  The complete procedure is available for viewing from our Web site
at http://www.library.cornell.edu/cts/53cor.htm

We are focusing our COR efforts on those materials most likely to have copy
available from other sources, and concentrating our original cataloging
efforts on the materials that are most likely to be unique to
Cornell.  Thus the level of COR cataloging varies among disciplines,
languages, and collections.  We continue to create PCC records for a subset
of most of our materials, though we are doing fewer records during the
phase-in period of this workflow (24 months) than was previously the case.

We chose to encode these records with a value of 3 because their headings
do not necessarily follow established forms, nor do the records necessarily
meet the national level bibliographic record minimal-level cataloging
specifications.  Again, because we are exporting our COR records to both
OCLC and RLG, you or your staff may encounter them in the course of
searching the utilities.  They are there to facilitate resource sharing via
interlibrary loan and for use by any technical services center that may
wish to upgrade them to full copy.

We have adopted this model to meet our goal of sharply reducing or
eliminating our cataloging backlogs while simultaneously improving physical
access to materials.  Despite years of increasing productivity and
processing workflow changes, we still have many thousands of items in our
backlogs.  However, our first six months under the new COR workflow have
brought very encouraging results.  Our overall cataloging productivity
increased by nearly 36% in the fiscal year ended June 30, and our backlogs
have dropped by nearly 40%.  If we are able to maintain or increase our
current productivity, we expect to meet our goal of a near-zero backlog of
print materials by December 2004.  We will continue using COR indefinitely
to prevent another backlog from materializing, to reduce our print
processing costs, and to provide more timely physical access to materials
for our users.

Please let me know if you have any questions or would like more information
about our COR initiative.

David Banush







David Banush
Head, Bibliographic Control Services
Central Technical Services
Cornell University Library
110D Olin Library
Ithaca, NY 14853

Voice: (607) 254-8031
Fax: (607) 255-6110
[log in to unmask]