College and Research Libraries Research Notes OCLC and RLIN: The Comparisons Studied Jean Slemmons Stratford INTRODUCTION An important frontier of library devel- opment is computer applications in li- braries. Two of the most important institu- tions in this field are Online Computer Library Center (OCLC) and Research Li- braries Group (RLG). The number of pub- lished studies which compare OCLC and RLG or their online computer systems is small. Library Literature was searched from 1970 to 1982 under the headings BALLOTS, BALLOTS project, Ohio Col- lege Library Center, Online Computer Li- brary Center, RUN, Research Libraries Group, and their subdivisions. 1 The arti- cles and their bibliographies were exam- ined to identify studies dealing with both systems or institutions. Though unpub- lished internal library reports must cer- tainly exist, only those few which have surfaced in the published literature are considered here. Through this process twenty-seven English language articles and monographic works were retrieved in early 1983. The purpose of the present paper is to analyze those twenty-seven articles as a body of literature. This body of literature, though small, is of interest to the library community for what it reveals about li- brarians' perceptions of the two systems and their effect upon library programs. Much of the literature appeared in reac- tion to the University of Pennsylvania's switch from participation in OCLC to membership in the Research Libraries Group and concerns the impact of these institutions on networking and interinsti- tutional cooperation. Despite the broad range and degree of opinion represented, the literature does provide a consensus of the major advantages and disadvantages of each system. THE LITERATURE DESCRIBED Chronologically, the pattern of publica- tion forms a marked curve. The first com- parative studies appeared in 1977, as BALLOTS was being marketed on a lim- ited basis by Stanford University just prior to its adoption by RLG and subsequent transformation into RLIN in 1978. Two studies were published in 1977; one in 1978; five in 1979. A disproportionately high number of studies, thirteen, were published in 1980. This peak may be ac- counted for as a reaction to the University of Pennsylvania's switch to RLG in early 1979. Thereafter, publishing dropped off with four articles published in 1981; two in 1982. The concern with the broad implica- Jean Slemmons Stratford has her M . L. S. from the School of Library and Information Science, State University of New York at Albany. 123 124 College & Research Libraries tions that these systems have for libraries is also reflected in the pattern of publica- tion by type of journal. The comparison studies have tended to appear in the main- stream general-interest journals, such as American Libraries and Library Journal, rather than in the specialized journals, such as those dealing with automation. Despite the fact that both OCLC and RLG/RLIN are online cataloging systems, the major area of concern reflected in the literature has been the impact of these in- stitutions and their systems on network- ing and interinstitutional cooperation. Twelve articles focus on this area. Techni- cal and public service applications, such as cataloging, reference, and administrative aspects of the systems, have been of sec- ondary importance; five articles cover ad- ministrative aspects (including system specifications); five are concerned with cataloging and/or technical services; three deal with reference, interlibrary loan. Perhaps in an attempt to justify their choice of system, the RLG/RLIN partici- pants are well represented in the litera- ture, while OCLC members, despite their greater numbers, have been less prolific. Overall, in terms of network affiliation, the authorship divides into four groups: seventeen RLG/RLIN-affiliated authors, nine OCLC-affiliated authors, four unaf- filiated authors, and three authors whose status is unknown. From 1977 to 1979, . only authors not affiliated with either sys- tem (e.g., paid consultants, librarians de- termining which system to join) or RLG/RLIN-affiliated authors were repre- sented in the literature. In response to the University of Pennsylvania's shift to RLG in 1979, six OCLC-affiliated authors pub- lished comparison articles in 1980. As time passed and the University of Pennsylva- nia's action proved an isolated event and not the precursor of a widespread trend, OCLC members again fell silent, with only two studies published in 1981 and none in 1982. This body of literature is defined by its concern with and comparison of OCLC and RLG/RLIN. However, a significant portion of these studies do not develop their own interpretations of the institu- tions and their systems or assert conclu- March 1984 sions from their findings. Thirty-three percent of the studies do not draw signifi- cant conclusions about the two systems. Six of these studies are no more than ob- jective presentations of system specifica- tions. Three studies provide formal quan- titative measures of the two systems, such as hit rates, but provide no interpretations of test results. The remaining eighteen studies draw some conclusions about the systems or their parent institutions, but only two of these base their conclusions on formal test results. Thirteen studies provide commentary on OCLC and RLG/ RUN from personal experience or opin- ion. The remaining three studies are based upon what might be referred to as "infor- mal testing." Danuta Nitecki's study, "Online interlibrary services: An informal comparison of five systems," is based on personal experience and telephone inter- views with selected ILL librarians and sys- tem representatives. 2 Similarly, Klaus Musmann's "Southern California experi- ence with OCLC and BALLOTS'' is based on visits and interviews with local system participants and representatives. 3 Joseph R. Matthews based his study on inter- views and a random sample questionnaire with a forty-five percent response rate. 4 According to Maurice B. Line, when 'I fewer .than half the sample as selected have responded-the results must be re- garded as insufficient to come to any firm conclusions.' ' 5 CONFLICTING PERCEPTIONS The majority of comparison studies or commentaries, as previously stated, deal with OCLC and RLG and their relation- ship to networking and library coopera- tion. These studies reveal conflicting per- ceptions of the two systems. On the one hand is the vision of OCLC as the "Na- tional Library Network.' ' 6 In this scheme of things, RLG/RLIN detracts from the size and integrity of the one truly demo- cratic network. OCLC' s democratic status derives from the fact that it has not ad- dressed itself to the problems of any one type of library. On the other hand, RLG, with its focus on research library problems and perspectives, is viewed as' I exclusion- ary," ignoring the "wider interests of li- ] braries as a whole.' ' 7 This argument, championed by Michael Gorman, is ele- vated to an issue of ''moral and philo- sophical imperatives.' '8 The alternative vision is represented by the likes of David Starn, Richard DeGen- naro, John Knapp, and Jo Chanaud. The existence of OCLC and RLG is seen as healthy competition, "free enterprise."9 .As Richard DeGennaro puts it: Our greatest success has come from allowing the entrepreneurial forces of the private sector . . . to act in our own best interests unfettered by government .. .. " 10 Contrary to Gorman's exclusionary -view of RLIN, David Starn sees the database as a means of making "the work of the al- .bl h " 11 leged few . . . access1 e to t e many. In response to the declaration of OCLC as the one national library network is the vi- sion of a national library network emerg- ing from the development of a set of communications links and standard protocols which will allow two- way, multi-lateral communications links be- tween our existing and potential computer- based bibliographic services. 12 As a part of this debate, Michael Gorman has predicted that if RLG knows how to achieve quality control in a six-million record database when and if they achieve that size, I will be happy to see it and I am sure that OCLC will embrace the tech- nique.13 According to an article by Julia E. Miller, the RLIN database now contains 6.2 mil- lion records. 14 There seems little hope for OCLC' s adoption of RLIN' s methods in this regard. The exact nature of OCLC is another area where conflicting viewpoints exist. Often, OCLC is described as a library con- sortium turned ''commercial vendor.'' 15 Though OCLC began as the Ohio College Library Center, it ''divested itself of me~­ bership" in 1978 and (as a not-for-profit corporation) began to contract its services to libraries via membership networks .16 As further proof of its commercial attitude, critics cite OCLC' s claim that ''it owns the database" and proposed OCLC contract language [that] Research Notes 125 would have prevented the use of any system other than OCLC by preventing the use of bib- liographic records by anyone other than ... OCLC and the user library .17 Joseph F. Boykin, Jr., president of the OCLC Users Council, represents the op- posing view when he explains that '' OCLC has . . . three classes of member- ship ... [and] representative gover- nance."18 SYSTEM ADVANTAGES AND DISADVANTAGES Despite the conflicting opinions that ex- ist regarding the nature of these two sys- tems, there is a consensus in the compara- tive literature as to their advantages and disadvantages. The studies are strikingly similar in this respect. Though some pre- viously central differences, such as data- base size, are becoming less important with time, the consensus of system differ- ences remains of interest. The majority of OCLC' s frequently cited advantages are in some way a function of its size. OCLC is credited with having a larger database, more participating li- braries, a higher percentage of non- Library of Congress and older records and, as a direct result, a better hit rate. OCLC is also viewed as more stable as a resUlt of its broader financial base. Per- haps as a function of size, OCLC is also ac- cepted to be less expensive. This must be seen as major plus by today's budget- conscious librarian. Other positively viewed Jeatures in- clude a more sophisticated ILL subsystem and a regional structure. As Mary Ellen J a- cobs explains it, ''The majority [of OCLC users participate] through membership in one of the 20 regional networks offering OCLC services." 19 The regional adminis- tration of the system is viewed as an ad- vantage responsible for rapid provision of such services as maintenance, training, and support. OCLC is not without its flaws, chief of which is its limited search capabilities. OCLC searching is extremely rigid. Based on search keys, it does not permit precise entry of the search request. For example, exact name searching beyond the '' 4,3, 1'' search key is not possible. The OCLC 126 College & Research Libraries database cannot be searched by subject, perhaps its most unfavorably viewed limi- tation. The corporate name index had been unavailable between 9:00a.m. and 5:00p.m. Monday through Friday. This was a serious drawback for reference ap- plication of the system, as that is the pe- riod ''when [corporate name searches] are most often requested. " 20 The quality of OCLC's non-Library of Congress cataloging is also frequently crit- icized. Some authors even feel that OCLC' s lower rates are offset by the fact that due to the quality of records, OCLC is estimated to require more professional catalogers than RLIN. 21 Another short- coming for catalogers is OCLC' s inability to perform any of the functions of a local online catalog or provide access to a li- brary's own record for a given title. OCLC is criticized for its lack of interest in coop- eration between the networks. There has been considerable apprehension over OCLC's assertion that it owns the data- base. If so, "the RLIN system may prove more advantageous by default. " 22 RUN's most frequently praised feature is, as it is repeatedly referred to, its power- ful searching capability. The literature is unanimous in its positive evaluation of this feature. RUN is searched, similar to DIALOG, by the use of Boolean opera- tors. It can search entire words, trunca- tions, or phrases in fourteen general in- dexes: personal name and exact personal name, title word or phrase, related title phrase, corporate name word or phrase, subject phrase or subdivision, Library of Congress card number, Library of Con- gress and Dewey classification, geo- graphic class code, U.S . government doc- ument number, ISBN, ISSN, Coden, publisher or issuing agency, and pub- lisher/agency number. In addition, there are also ten local indexes which apply to the user library's holdings only. 23 RUN is also credited with higher quality and more detailed member input records than OCLC. The provision of local call numbers is a plus, as is the amount of flex- ibility and choice provided through access to all individual member records online. Additionally, the availability of one's own records online is seen as a major advan- March 1984 tage, making the maintenance and up- grading of records much simpler. The fact that RUN can fulfill some of the functions of a local online catalog is also viewed as a point much in that system's favor. In general, RUN is viewed as more am- bitious, "meeting a more inclusive set of library objectives . ''24 Where OCLC is seen as having the present advantage, RUN is repeatedly praised as the system of the fu- ture. A part of that praise is due to RLG's efforts to promote network cooperation and allow unrestricted access to its data- base. The major disadvantages of the RUN system are a function of its size. It has a smaller database and fewer participating libraries. This is perhaps its most fre- quently cited drawback. Similarly, its growth has been slower, its percentage of Library of Congress MARC records higher, its hit rate lower, and it has less fi- nancial security than OCLC. Another im- portant factor that can be linked to size is that of cost. RUN is substantially more ex- pensive than OCLC. CONCLUSION Overall, librarians have compared these two systems in terms of their effect on li- brary programs. Their function as tools available for the improvement of library operations such as cataloging and refer- ence work is only a small part of the litera- ture. The coexistence of the two networks is viewed as either exclusionary or the positive forces of free enterprise at work. This difference of opinion may account for the emotional tone of much of the litera- ture ·and the high percentage of articles that express an opinion without any quan- titative measurement to reinforce its valid- ity. Only two studies that evaluate the re- spective merits of OCLC and RUN have as their basis any sort of formal testing. Certainly, in many respects OCLC and RUN are not comparable. OCLC's mis- sion is directly linked to its database and the provision of auxiliary systems such as ILL and acquisitions. For RLG, however, the RUN database is only one of several tools developed to support programs for cooperation, preservation, and collection development. Yet, their ability to perform similar functions makes comparative eval- uation necessary to informed library plan- ning and decision making. The library community must venture into formal quantitative measurement of OCLC, RUN, and the other networks as a basis for future development. The present body of published literature is insufficient to ad- equately support the decision-making process faced by many libraries at this time, and those libraries that have pre- pared evaluative internal documents should consider publication of their re- search. AsKazuko M. Dailey points out: Research Notes 127 By making our internal documents available to the profession, we hope to encourage other li- braries to come forth with their analyses of the bibliographic utilities and perform analyses, where before we had only assertions or as- sumptions . The ultimate purpose of "going public" is not to criticize, but to comprehend the bibliographic databases. 25 Only through the availability of such for- malized comparisons will the networks be seen in their proper perspective, as tools available for the work at hand rather than forces beyond our understanding or con- trol. REFERENCES 1. Library Literature (New York: Wilson, 1970- ). 2. Danuta Nitecki, "Online Interlibrary Services: An Informal Comparison of Five Systems," RQ 21:7-14 (Fall1981) . 3. Klaus Musmann, "The Southern California Experience with OCLC and BALLOTS," California Librarian 39:28-39 (Apr. 1978). 4. Joseph R. Matthews, The Four Online Bibliographic Utilities: A Comparison (Chicago: American Li- brary Assn., 1979). 5. Maurice B. Line, Library Suroeys: An Introduction to the Use, Planning Procedure and Presentation of Suroeys. 2d ed. (London: Clive Bingley, 1982), p .111. 6. Michael Gorman, "Network! or I'm Rational as Hell and I'm Not Going to Take It Anymore," American Libraries 11:48 Qan. 1980). 7. Ibid., p.50. 8. Ibid. . 9. Jo Chanaud, "One for the Record," American Libraries 11:279 (May 1980). 10. Richard DeGennaro, ''Computer Network Systems: The Impact of Technology on Cooperative Interlending in the USA," Interlending Review 9:43 (Apr. 1981). 11. David Starn, "Networks and the Common Cause," American Libraries 11:278 (May 1980). 12. John F. Knapp, "Requirements for the National Library Network: A View from the Local Net- work," Journal of Library Automation 10:136 Qune 1977). 13 . Michael Gorman, "Gorman Responds, " American Libraries 11:279 (May 1980). 14. Julia E. Miller, "OCLC and RUN as Reference Tools," Journal of Academic Librarianship 8:273 (Nov. 1982). 15. DeGennaro, "Computer Network Systems," p.40. 16. Susan K. Martin, "OCLC and RLG: Living Together: Some Basic Questions and Answers," Amer- ican Libraries 11:270 (May 1980). 17. Ibid., p.271. 18. Joseph F. Boykin, Jr., "OCLC Still Has Members," American Libraries 11:411 Quly 1980). 19. Susan K. Martin, ed ., Online Resource Sharing II: A Comparison of OCLC Incorporated, Research Li- braries Information Network and Washington Library Network (San Jose: California Library Authority for Systems and Services, 1979), p .10. 20 . Miller, "OCLC and RUN," p.272. 21. Kazuko M. Dailey, Grazia Jaroff, and Diana Gray, "RUN and OCLC Side-by-Side : Two Compari- son Studies," Advances in Library Administration and Organization 1:119 (1982). 22. Paul G. Knight, Jr., "Cataloging with OCLC and RUN: A Comparative Analysis," North Carolina Libraries 38:56 (Summer 1980). 23 . Sharon Cline Farmer, "RUN as a Reference Tool, " Online 6:17 (Sept. 1982) . 24. Knight, "Cataloging with OCLC and RUN," p.49. 25. Dailey, Jaroff, and Gray, "RUN and OCLC Side-by-Side," p.73 . . ~ll"'b~;.1 ... o-~~ll····~ss''~ ll·~!~ \l • Backllles from 1975 to the present. · • The annual subscription price for the A Ualqv.e lefenaoe 8JRem The topic could be Latin America or Mrica, National Security or the Kennedy Assassination. Whatever the information, The DeclassJJJed Document Beference·Sys- tem is the on1y comprehensive compilation of its kind. Documents are microtllmed by . Research Publications as they are released under provisions of the Freedom of Infor- mation Act Amendments in 1974. Issued in quarterly installments, each annual subscription contains on microfiche · 3,000 of the actual government documents which helped mold U.S. history over the last 35 years. Printed quarterly abstracts and indexes, and an annual cumulative index, are included with each subscription. The abstracts and indexes may also be pur- chased separately. Allo Available • A Retrospective Collection containing 8oo0 documents in two hardbound vol- umes of abstracts plus a hardbound sub- ject inqex. complete collection (microfiche and abstracts and indexes) is on1y $718. (Prices slightly higher outside the U.S. arid Canada.) ·To order, or for more information on either the Decla8sit1ed Document Reference System or RetrospectiVe Collection, write or call: ~arch publications 11 LuiW' Drive/Dnwer AI WooOrldCe, 0! 08818 (101) 117 ·1800 !1r.l:710.t88·8148 lAX: 101·117·1811 0Uu14e •onhad SouUl America: P.O.Ba'8 leacliDC, BCU 8 BliDClllld DL: 0716·8811'7 . DLU: 868118 UDL G 1