Library Committee Handbook

Executive Committee

Report on the use of Exceptional Dewey Classification


Report on the use of Exceptional Dewey Classification scheme for Literature including a

Recommendation for Reclassification


Charge as approved by EC and sent to the group April 16, 2010

Explore options to streamline our application of classification schemes. 

This evaluation will draw on a cost-benefit analysis being conducted by Jessica Efron, Stephanie Baker, Lea Howard and Susan Hill (who will liaison with the Exceptional Classification Routines Working Group).

Make recommendations regarding the timeline and process for adopting the best options identified

Work with the Literature and Languages Implementation Team to determine whether the formation of the proposed Literatures and Languages Library provides a viable opportunity to move away from the current exceptional Dewey classification process, considering whether the long-term benefits in (reduced costs, quicker time to shelf) outweigh the costs (tangible and otherwise) associated with making the conversion.

Share and discuss options, scenarios and cost benefit analysis with the relevant Library and campus stakeholders by May 8th (later changed to early July)

Final report by July, 2010. Report submitted July 11, 2010



Harriett Green, English Library

Marek Sroka, Slavic and Eastern European Cataloging and English Library

Gail Hueting, CAM

Qiang Jin, CAM

Jenn Miller, CAM

Geoff Ross, History, Philosophy and Newspaper Library

Susan Hill as Liaison from the Cost Benefit Analysis Team

Lynn Wiley, Acquisitions, Convener


The unanimous recommendation from this working group is to adopt LC for Literature in Western European languages (including English) for any newly required material and to do a full retrospective conversion for the collection housed in the new Literatures and Languages Library.

            Timeline: to coincide with the move of the merged collections into the new space for the             Literatures and Languages Library.

Cost was a factor in our discussions but guaranteeing user access during this time of budget constraints combined with the significant number of staff retirements were the two main reasons behind the recommendation.

Options discarded: The team was unanimous in agreeing that to move any Exceptional DDC items to published DDC would in effect result in two separate shelving schemes and would not gain the library any significant cost savings and would hinder user access. That left the decision to either retain EDDC along with DDC or to move to LC. Note: one member did suggest there may be potential to streamline the exceptional DDC process but all agreed that would still mean additional labor and so the group did not investigate this further.

If this recommendation is approved, the working group further recommends that a new Implementation and Planning Team be convened ASAP to address the following:


Cost benefit Analysis


This was accomplished as a precursor to the classification review to understand what the transactional costs were in using DDC. The full document is included in the Appendix. A summary is included here; See also the detailed analysis provided by Duke University in their review of DDC which resulted in a switch to LC



An estimate of how that may look as far as annual costs is given below. It assumes a range of volumes added annually per class (LC, EDDC and DDC).


We can assume 50,000 books received in print that require call numbers.

Assume less than one third is Literature or 14,000 items (no Latin American approval books are included as they go direct to high density storage and no Asian items counted because they are LC already) but also add to those the other exceptional areas for art/arch/philosophy etc). bringing the exceptional DDC  non-Asian volume count to 18,000.  The LC is estimated at 5000 for Asian and 2000 for Performing Arts for 7000 total. The rest are published DDC or 25,000 items.










Class Scheme

Annual volumes added

Classification Labor

student work cost per volume(mark stamp strip)

cost per scheme for all






Published DDC





Exceptional DDC
















If Literature Moves to LC


Class Scheme

Annual volumes added

Classification Labor

 student work cost per volume(mark stamp strip)

cost per scheme for all

If Literature moves to LC





remaining EDDC





Published DDC





LC with Lit













To switch to LC for Literature has the potential to be 50k cheaper though the volume count here does not attempt to segregate the literature in Undergraduate Library or the literature in Classics so the savings would be somewhat less if they were not converted.


A switch to LC for ALL newly acquired material



Annual volumes added

Classification Labor

student work cost per volume

cost per scheme for all













And if we moved wholesale into LC for all print books, we would see substantial savings but this report is not suggesting we do that at this point in time. See the section on advantages and disadvantages.









Advantages of LC



Disadvantages of a switch to LC







Cost savings: More of our peers use LC (enables sharing records)  Cost is less than DDC: see cost benefit analysis



Funds are needed to convert new Lang/Lit Library to LC, to set up new LC shelving area in Stacks  also for signs to direct users in the Stacks

NOTE: Duke , Purdue SIUC, UMIch all noted significant cost savings in cataloging









Necessitates a split collection at least in new Literatures/Languages Library if that is not wholly converted as well as in the Central Stacks


LC call numbers available widely; Shelf ready may be applied broadly



Will require training in LC classification and with student shelvers; All public service desk personnel will need to understand the changes







Titles will be sent fully classed to the shelves faster



Physical Browsing will be impacted with a split collection


LC will replicate much of EDDC in classing material appropriately together (foreign authors handled differently with EDDC). LC numbers can be much easier to read than DDC with complex cutters



LC will not completely replicate EDDC in that material about an author will be shelved separately. But LC  does provide for appropriate adjacencies for authors works


 Many Faculty and Grad students are familiar with LC already.  SIUC noted greatly improved shelving accuracy with their switch.




LC training much easier and more accurate



necessitates new training for student shelvers and staff in Central circ



User Access: Sampling of delays now seen in applying EDDC and DDC

Sampled 39 records from three categories to see when they were received, when MARC was available, when they were give a classification number, the final step before going to the permanent shelf location. The three categories are:

  1. Domestic approval books with copy records available that required DDC or EDDC; 2. European Blanket order books where we buy full MARC; and 3.those European Blanket orders where we do not buy full MARC.

There were 24 in the first with 4 not yet classed  2-3 months after receipt (all EDDC) and the 20 that were classed at the time of this study took an average of 35 days (at which point they were shelved and available for users).

There were 9 in the second with 2 not yet classed after 2 months (all EDDC) and for the 7 with a call number, the average time for the class assignment was 66 days.

For the last group of 6, none had a call number and were all waiting 5-8 months. 


These books had LC call numbers available at item receipt with the exception of the last group and for this group, full LC Core were available for purchase and all had LC class numbers.



The Library at the University of Illinois at Urbana Champaign uses DDC for item classification except for the following areas: Asian has used Library of Congress since 1964  as does Law (adopted as the K classification was published, previously materials were not classed) and the Music and Performing Arts Library (converted in 1989) is also now LC. In addition, the Library adapted DDC locally for several areas with that classification called exceptional DDC and in the rest of this report referred to as EDDC.  See Gail Hueting and Winnie Chan's document: Classifications Used at the U of I Library (

 The largest exception to Dewey as published was and remains the 800s, Literature.  The reason is to keep works by the same author together and arrange them by broad chronological periods, as the LC Classification does.  In some other areas, UIUC stayed with earlier versions of the classification.   Mathematics, agriculture, and the World Wars were areas in which modified Dewey was used.  (added note: those areas have since updated to a later edition of published DDC).

Literature at the University of Illinois Urbana Champaign is housed in the Main Library Central Stacks, at English and Modern Languages (merging to become the Languages and Literature Library) as well as the Classics and the Undergraduate Library (literature at the high density storage facility on Oak St does not require a call number).  The Area Studies libraries now merging to form the new International Center will have a non-circulating collection with literature shelved in the Central Stacks.

Other local modifications to published Dewey followed at the Library include:

 artists, biography and criticism, philosophers, material about Abraham Lincoln, and the S-Collection. The Library dropped the prefix B for Biography this year.






The group met beginning in late April and through May and June to gather information and to discuss the task. The group discussed the purpose of classifying in an online world; reviewed virtual browsing; and reviewed LC vs. DDC for physical browsing by looking at the current holdings in the English and Modern languages Library. We delegated the task of surveying peer libraries that switched across the group and developed questions to ask. We delegated tasks in gathering statistics and did a quick scan of our current situation regarding the budget and staff retirements. We reviewed the cost benefit analysis and drew up a comparison of the advantages of LC vs. EDDC for Literature with examples.


Rationale for the review of classification schemes:


LC Conversion has come up several times in the past. This review was initiated due to the following:


ARL Peers: Many large ARL libraries have switched to LC recently. We surveyed the following:


All reported that users adapted easily, that the switch resulted in faster and cheaper cataloging and several commented they wished it had happened earlier. Survey notes are available as well as excellent reports provided to us by Duke in particular, See below for their location.


Background Documents: (all documents are available at: G:\Collections Info\Classification review\Exceptional DDC



UIUC Literature LC/Dewey Cost Benefit Analysis


Assume $18/hour staff time




No time involvement - goes directly on shelf





Can assume most bibs already loaded to Voyager at point of order.

2-3 minutes per book (estimate)

May be able to automate to make faster



1.      Pull up bib record and copy LC number.

2.      Create MFHD and paste LC number into 825.

3.      Create label.


If done by hand:  $.90 per book plus label and marking costs

If automated much less. NOTE: Many ARL Libraries are using students to do the classification label, we assumed a staff labor of $18 an hour so with students it would be at least half that or $.45 per book.





5 minutes per book - about 8 to 10 in an hour

Most of these have bibs already in Voyager.

Includes 81X's and 82X's (82X for outside US and Canada).



1.      Search author in Cataloging to see if call number is established yet for that author.

2.      If so, choose appropriate Dewey number (e.g. 811 for poetry or 813 for fiction, depending on the subject of the book) and use established cutter.

3.      If not yet in Voyager, choose appropriate number based on the subject of the book (e.g. 811 for poetry, 812 for drama, 813 for fiction, etc).  Create cutter for that author (or title if no author) with the cutter table.

4.      Add call number to 852 of MFHD.

5.      Create label.


$1.80 per book plus label and marking costs





$1.80 - $6.00+ per book plus label and marking costs


Low: <10 minutes (40%)

Medium: 10-20 minutes (%20)

High: >20 minutes (20%)

Some of these do not have bib yet in Voyager and need to download.  This adds more time.


1.      Search author in Cataloging to see if call number is established yet for that author.


If call number is established: 10-15 minutes. 


If call number is NOT established: 20+ minutes. 


2.      Download bib if it does not yet exist in Voyager.

3.      Add call number to 852 of MFHD.

4.      Create label.


$1.80 - $6.00+ per book plus label and marking costs







Small: 2100 labels = $18 -  ~$0.009/label

Wide: 1250 labels = $18 -   ~$0.014/label


NOTE: There is some significant loss of labels due to the fact that label sheets can only be fed through the printer 1 or 2 times before it risks jamming the printer.  So some sheets are only partially used and then thrown out.


Label Savers: 250 label savers = $11.20 - $0.045/label saver


Student cost:

            5-10 seconds per book = 6 books/minute = 360 books/hour @$8/hour = $0.02/book


Printer cost: hard to quantify, but there is a cost in maintenance, power, toner, etc.


Total label cost per book = ~$0.08

(may be more like $0.10 if you figure the students probably don't really get 360 done in an hour.....)






Materials Cost:

5,000 strips per box @ $180 per box =  $0.036/strip (Plus there is about a 5% loss due to bad strips)


Student Cost:

            2 mins/book = 30 books/hour @ $8/hour = $0.27/book


Total stripping and stamping cost per book = $0.31/book



TOTAL LABEL STAMP AND STRIPPING IS $.41 per book all labor and supplies included
















FROM DUKE REPORTS provided to the team July 2010

Dewey Vs LC


Seven GOOD reasons why the library (Duke) should consider converting from Dewey to LC


1.         Increase delivery of information resources to library users and reduce cataloging costs.

Assigning Dewey Classification numbers is by far the most time-consuming aspect of copy cataloging.  In our current Dewey environment, member copy catalogers (Library Assistant, Sr.) assign Dewey numbers to 64-percent (or 36,000 titles) of all copy cataloging at a cost of $6.69 per title (Table 1.0).  The other 36-percent (or 20,000 titles) that include Dewey numbers is cataloged by rapid catalogers (Library Assistant) at a cost of $1.32 per title and a rate four times faster than member copy and ten times faster than original.  Imagine what it would mean to our users if we were able to catalog 95% (or 53,000) of all copy cataloging titles at a cost of only $1.32 per title and a rate ten times faster than original cataloging.

The extent to which the library realizes the long-term benefit of switching to LC will depend largely on how the library redefines local cataloging practices.  Table 1.0 provides the most optimistic scenario, one that is built on the assumption that the library would join several of its peers in greater acceptance of member-contributed copy.  Switching to LC positions the library to redefine cataloging practices to the benefit of its users. 



DDC System

LCC System

Copy Cataloging

#Titles Cataloged


Total Cost1

Cost/ Title

#Titles Cataloged


Total Cost1

Cost/ Title

With Classification









Without Classification



















§ Table 1.0. Classification systems influence on copy cataloging costs and productivity

1Total cost based on 2002-03 salary figures for all copy catalogers.

2.         Expedite conversion and minimize impact on other user services by using third-party reclassification services.

Previous assessments of the advantages and disadvantages of converting from Dewey to LC always looked to an in-house operation that would take eight to ten years to complete.  A number of vendors now include reclassification among their list of services (MARCIVE, LTI, Library Associates, etc.)   Depending on the funding available, vendors can take a file of our records, generate call number labels, provide LC numbers for records containing only a Dewey number, and, in some cases, send staff to affix the labels to books.  A recent article in the MARCIVE Newsletter (October 2002) describes a vendor-assisted reclassification project at Grove City College Library, a small college in Pennsylvania with a collection of 118,000 titles. The project was completed in fourteen weeks. Although Duke's monograph collection is 6.5 times larger than Grove's, it is no longer unreasonable to contemplate a timeframe of 1 to 2 years for a collection of 1.3 million monograph volumes.

3.         Free human and/or financial resources to focus on other critical library priorities.

As indicated in no. 1 above, switching to LC stands to reduce cost of cataloging by 66-percent (or $177,000) and FTE staffing levels by 59-percent (or 4.75 FTEs).  Some of the potential projects that could use more resources are converting the manual serials holdings information to the online catalog to facilitate transfer of serials to the LSC, eliminating gift backlogs that are now in excess of 10,000 volumes, completing serials recon for over 8,000 titles, resolving post-reclassification fallout (e.g., recon, database corrections, trouble-shooting, etc), and expediting clean up following conversion to a new ILS.  These represent just a few examples of current and future resource needs.

4.         Position the library to take full advantage of third-party cataloging products and services.

Major vendors such as OCLC and YBP only recently incorporated Dewey into their mainstream cataloging services.  A recent investigation of OCLC's most cost-effective outsourcing option, PromptCat, revealed that the library's use of double cutters and the year of publication would not permit the use of this service, without introducing more confusion in shelving, browsing, and users' ability to find materials. Fortunately, YBP is equipped to handle local practices, but it comes with an additional cost of fifty cents per title. As an LC library, we would remove these limitations and added costs. Newly acquired materials could arrive and be delivered for use within days instead of weeks or months.

5.         Achieve and maintain a backlog-free environment.

Backlogs are the antithesis to delivering quality service.  Even as catalogers steadily increase production, thousands of volumes in gift collections and from new acquisitions remain unavailable to our users.  Eradicating this problem in our current Dewey environment would require additional funding to hire more copy catalogers or outsource the work to third-party providers at a cost almost equal to that of in-house cataloging.  As indicated in nos. 1 and 4 above, converting to LC will not only quicken the pace of cataloging, but also allow the library to take advantage of the most cost-effective outsourcing options.  A long desired goal of a backlog-free environment would be realized and the rate in which are materials made available to our users would increase substantially.

6.         Inventory our collection and increase users' ability to locate materials on the shelves.

A conversion from Dewey to LC would achieve the library's strategic goal to inventory the collection.  Once conversion within each Dewey Classification range is complete and the reclassified items removed, we would find ourselves in a unique position to identify and address a host of problems that interfere with users' ability to locate materials on the shelves.  For example, misshelved and lost items would be discovered, items still classified in Dewey would be reconned, and call number conflicts between the item and the record would be resolved.

7.         Reduce conversion cost and disruption to users by integrating conversion with building renovation requirements to handle volumes no more than once.

Disruption to users during the building renovation will be unavoidable.  Merging reclassification with the movement of materials in the stacks to their new shelving location will permit us to re-label materials with their new LC number prior to the move. Then, materials can be sorted, according to their LC numbers, and shelved in their permanent location.

DUKE 2002