College and Research Libraries SUSAN A. MERRY The Ontario New Universities Library Project—A Centralized Processing Experiment Completed This paper was written at the conclusion of ONULP, a short-term centralized processing Project carried on at the University of Toronto Library. The prime functions of the Project were the selection, acqui- sition, and cataloging of 34,600 titles in 44,510 volumes for each of •five new academic libraries in Ontario. The paper charts acquisition figures, describes cataloging practice, and concludes with a comment on the book catalog produced by computer for ONULP. A . T THE REQUEST of the government of Ontario, the Ontario N e w Universities Library Project was organized to provide five new universities and colleges with basic and identical undergraduate li- brary collections. These universities were Brock in St. Catharines, Guelph Univer- sity and Trent University in Peter- borough, plus two new off-campus col- leges for the University of Toronto, Scar- borough and Erindale. The University of Toronto library undertook the re- sponsibility for the selection, acquisition, and processing of the the materials and, in addition, for the experimental produc- tion of a book catalog by computer. A staff of approximately thirty-five people was required to design and carry out the project. The last of these libraries to go into operation was Erindale, which opened in September 1967. Erindale Col- lege benefited most as all the books bought through the Project were ready to circulate when the library opened its Mrs. Merry is Head of Technical Services at the University of Toronto Libraries, New University Library Projects. 104/ doors. This unfortunately was not the case for the other libraries, two of which opened in September 1964. The project was completed in three and one-half years. SELECTION The Project was restricted to the selec- tion of in-print monograph material. The goal was seventeen thousand five hun- dred titles in thirty-five thousand vol- umes for each library. Subscriptions and out-of-print materials were avoided mainly because the Project was a short- term venture and also because the main- tenance of the required files would have been too challenging on a temporary basis for a small technical services staff. The selection was based on the Scar- borough College curriculum, which in- cluded twenty-six subjects common to undergraduate arts and sciences pro- grams. CATALOGING The O N U L P collection required little professional cataloging. Some 58 per cent of the 34,600 titles received were in the Ontario New Universities Library Project / 1 0 5 T A B L E 1 ACQUISITION Description Domestic* British European Other Totals Book Expendituret $969,335 $93,400 $167,695 $23,170 $1,253,600 Per Cent 77 8 13 2 100 Volumes Acquired* 172,237 18,792 28,635 2,886 222,550§ Average cost per volumell 5.68 4.96 5.85 Cost (above) incl. discount 10%-34% Cost incl. binding# (av. bind, cost) 2.12 2.20 Per Cent orders unfilled 7 14 19 ( 3 8 5 0 ) 10 out of print ( 1 6 4 3 ) 5 ( 3 6 5 ) 10 ( 7 4 8 ) 14 ( 2 7 5 7 ) 7 cancelled" * ( 6 4 2 ) 2 ( 1 5 2 ) 4 ( 2 8 9 ) 5 ( 1 0 9 3 ) 3 * Three principal dealers used—Co-operative Book Centre, Toronto; Blackwell's, Oxford; Nijhoff, The Hague. This simplified procedures b u t weakened service, particularly with back orders and reports where library control was lost because of lack of automatic claim and cancel period. Such was thought unnecessary with a pre-search in-print acquisition policy. t Budget total $ 1 , 3 0 0 , 0 0 0 . Balance unspent $ 4 6 , 4 0 0 . J A team of five male processors prepared (unpacked, checked, matched order cards, and search reports) 7 , 5 0 0 volumes per month ( 9 0 , 0 0 0 per yr. or 2 5 0 titles i n 1 , 5 0 0 volumes per man per m o n t h . ) § Total volumes processed 2 2 2 , 5 5 0 ; total salaries earned in Technical Services $ 2 3 5 , 1 4 0 . Cost of processing each volume $ 1 . 0 5 . ( S u p p l i e s and equipment not i n c l u d e d ) . || Over-all average cost per volume $ 5 . 1 4 . # Cloth editions pursued always. American and Canadian paperbacks processed unbound. *><> Cancelled—reprinting, n e w ed. in prep., N.Y.P., out of stock, no report. University of Toronto library collection. Also 12 per cent of these were new editions and required subprofessional editing. The next 35 per cent were cata- loged from Library of Congress proof- slips, and the final 7 per cent required professional attention. One professional cataloger, six cataloging assistants, and three revisors, one of whom was pro- fessional, were required for this work. Library of Congress ( L C ) classifica- tion was used with the following modi- fications: Canadian history was classed in F 5000-5949, English Canadian litera- ture in PS 8001-8549, and French Cana- dian literature in PS 9001-9549. (PZ class was not used for fiction.) LC subject headings were used but were modified in length for book cata- log production, a concession to the ONULP machine system. The ONULP Subject Authority List printed by com- puter contained 24,600 name and topical headings, scope notes, and references ( S E E but not SEE ALSO) and was dis- tributed to each of the participating li- braries. LC cataloging was followed closely, with only slight modifications. Cuttering was changed only where a conflict with the University of Toronto library's shelf list was found or where reasonable ex- pansion was not allowed with the LC number. The latest printing date was used in the imprint and the call number. Certain notes were omitted, and as a concession to the machine system the order of title and author's name was re- versed in all contents notes. Multiple copy cataloging required rigid standardization. Where imprint dates varied between one or two copies of a title the date most common to all five copies was chosen for the catalog 106 / College 6- Research Libraries • March 1968 record. No attempt was made to prepare individual records for the libraries ad- versely affected by this rule. Each of the 44,510 volumes cataloged for each library was pocketed and labelled, using the Se-Lin label. ONULP provided catalog card sets (six cards per set) to three of the par- ticipating libraries, in addition to the computer printed book catalogs and shelflist cards. Card reproducing equip- ment was not located in the department, however, hence catalog card sets were not shipped in their books; the libraries involved complained of shelving delays. There were several aspects to the suc- cess of this centralized processing Proj- T A B L E 2 OPERATIONAL T I M E FACTORS ( I n M o n t h s ) Pr oc ess ing Pe rio d Co op era tiv e Bo ok Ce ntr e Bl ac kw ell 's Ni jho ff Selection-Ordering * % Ordering-Receiving includes preparation for catalogingt 3% 6/2 Receiving-Cataloging includes preparation for shipping to 5 libraries* 2% 3J4 3M Total Time Factor— Selection-Cataloging 5)4 7% 10% " Foreign items were batch ordered. f Certain restrictions were placed on each dealer. First, regular orders were not shipped by any dealer until he had all five (or t e n ) copies of the title in stock. Second, no incomplete orders were accepted unless a report was received in advance enabling O N U L P to choose to accept or refuse less than five copies or incomplete multi-volume sets and to report back. Third, no automatic claim and cancel period or replacement program was planned ( O N U L P was not a continuing service). Fourth, it was not foreseen in the early stages that the budget would provide twice the number of titles originally contracted ( a large number for a f e w dealers). All these factors contributed to high average time lapses between ordering and re- ceiving. t Most foreign material required original cataloging. N o cataloging priorities were established, hence ma- terial with cataloging copy was processed more readily. ect not the least important of which was that it was planned to come to an end. Its policies were not well enough de- fined to have operated successfully on a continuing basis. The advantages it en- joyed were that it served only a uni- versity community, each book collection was new, no subscription files were re- quired, and no individual library idio- syncracies were allowed. Under these conditions the processing was carried out successfully on a steady basis. C O M M E N T ON THE M A C H I N E SYSTEM The computer production of the ONULP book catalogs has been ably described in two or three papers, and does not need repeating here. With the conclusion of the project a f e w features of the system deserve comment. Some good features of the system: 1. Full cataloging information was re- tained in the maih entry records and laid out as in a conventional catalog. 2. Secondary entries were constructed automatically in abbreviated form from the master record. 3. Headings in both catalogs were checked for uniformity by the match- ing programs, and non-uniform head- ings were discarded with error mes- sages to ensure correction. 4. Initial articles in twenty-seven lan- guages were disregarded in filing title main and added entries. (Texts in more than one language were manual- ly controlled.) 5. Qualifiers in name headings, e.g. jt. author, ed., comp., were disregarded in filing. 6. Author and title analytics were gener- ated where required. Some desirable features that were un- available to ONULP for a variety of rea- sons, mostly economic: 1. Longer sort fields (maximum was sixty characters, capitals, diacritics removed) to ensure more accurate Ontario New Universities Library Project / 107 T A B L E 3 S T A F F — S A L A R Y C O S T ANALYSIS S A M P L E Y E A R Job Description Salary Per Cent, Section Work Days Per Cent, Section Per Cent, Total Salary Acquisition: 8 clerical staff— 1963/67 turnover 13 staff: Preparation of orders (bibliographical details; estimate; assign agent), type order forms, reports . $ 4,286 19 295.5 18 5 Books: receiving and processing, invoice preparation, returns . 15,734 65 1,128 72 19 Supervision, revision, statistics, special jobs 3,339 16 178.9 10 4 Totals $23,809 100 1,672.4 100 28 Searching: 5 clerical staff— 1963/67 turnover 11 staff: Preparation of titles for cataloging Preparation of titles for cataloging $14,145 100 1,071 100 17 Cataloging: 12 staff incl. 2 librarians—1963/67 turnover 20 staff: Data sheet preparation and revision $12,172 28 815 30 15 Keypunching and revision, keypunch training 3,278 7.6 248 9 4 Book shipping, book pockets and label preparation . 7,615 16.5 567 21 9 Master card typing and revision—sort xeroxed card sets 6,005 9.6 375 11 8 Filing and revision (master cards, working shelf list, 2 authority files, computer printed shelf list) . 2,893 6.5 236 9 4 Original cataloging . . . . 2,063 4.5 92.5 4 3 Administration 5,045 11.1 181 7 6 Book catalog preparation and correction (monthly machine cycle) 3,058 12.2 156.5 6 4 Secretarial 1,718 4 85 3 2 Total $43,847 100 2,756 100 55 Total Salaries $81,801 100 108 / College 6- Research Libraries • March 1968 filing particularly in the subject cata- log. 2. Consolidation of all titles by one au- thor under his name, listed once. 3. Consolidation of SEE references to avoid waste space. 4. Ability to generate a changed shelf- list record for one library at a time and to suppress the record where other libraries not affected; and to keep such changes from reappearing in the book catalogs. 5. Although unchangeable, the rigid au- tomatic filing of the computer, based on spacing and punctuation, is some- what inconvenient for the catalog user. 6. Although not essential, the lack of non-Roman alphabets, italics, or print- ing in red has been noted by some users of the catalogs. There was general consensus that a book catalog produced monthly was not the most satisfactory index to a working collection, despite its advantage of port- ability. Also, a sample study in 1964 re- vealed that the cost of listing 5,220 titles in the book catalogs (150 copies) was 1.7 times greater than that of listing the same titles in manually produced card catalogs ( 1 copy). The real future for automating bibliographic records ap- parently lies in a more detailed, flexible, and enlarged data format, with inter- national compatibility and accessibili- ty. • •