the consortium as learning organization: twelve steps to success in collaborative collections...

5
The Consortium as Learning Organization: Twelve Steps to Success in Collaborative Collections Projects by Christy Hightower and George Soete A project to share translation journals in the physical sciences among nine California academic research libraries achieved many of its goals, yet it encountered major difficulties during implementation. Based on their experience with the project, the authors offer twelve strategies for ensuring success in collaborative collections projects. Christy Hightower is Collection Development Specialist and Engineering Information Specialist, Science and Engineering Library 0175E, University of Ca/ifornia,San Diego, 9500 G/man Drive, La /o//a, California 92093-0175 <[email protected]>. George Soete is an independent library consultant, former/y Associate University Librarian, Collections, University of California, San Diego < [email protected] >. A ccording to David Garvin, “a learning organization is one that is skilled at creating, acquiring, and transferring knowledge, and at modifying its behavior to reflect new knowledge and insights.“’ A critical question for libraries, as they move toward greater interdepen- dence, is whether a consortium can behave like a learning organization, gaining knowledge from both success and failure and modifying behavior accordingly. This article focuses on a case study: a pilot collaborative collections manage- ment project mounted by the University of California and Stanford University Librar- ies to provide for shared access to transla- tion journals in the physical sciences. By some measures the project failed, because of both unexpected developments in the environment and flaws in design and implementation: though it still exists and still saves the libraries more than a $100,000 a year, the project has been put on hold. By other measures, however, it has succeeded, saving scarce resources, ensuring a rational coverage of certain research materials, and providing valuable lessons on which to build future projects. “A critical question for libraries, as they move toward greater interdependence, is whether a consortium can behave like a learning organization, gaining knowledge from both success and failure and modifying behavior accordingly.” The following is an account of a mid- project evaluation that uncovered serious problems in the Science Translation Jour- nals Pilot Project. Though it contains a description of the project’s genesis, objec- tives, and structure, as well as an analysis of the problems encountered, the true emphasis is not on those aspects; nor is it on dollars saved, journals canceled, or other quantitative measures. Rather it is on the lessons that were extracted from this “noble failure,” here presented as 12 strat- egies for promoting success in collabora- tive collections projects. BACKGROUND Collaborative collections management is an established tradition among the libraries of the University of California (UC), comprised of nine campuses spread across the state. Though each library man- ages its own collections program and bud- get, since the mid-1970s they have joined in a shared acquisitions effort, now called the Shared Collections and Access Pro- gram (SCAP).2 SCAP, which derives its funding from a central pot of money repre- senting approximately 3% of the total acquisitions budgets of the UC libraries, has acquired millions of dollars worth of materials for the UC libraries while hold- ing to the objective of avoiding duplica- tion. In recent years, the program has expanded and changed focus: shared funds have been used to acquire and mount databases on MELVYL@ the UC Libraries’ union catalog, and matching funds have been provided by the local campuses for some purchases. Most importantly, perhaps, the key purpose of the program has shifted from enrichment to survival. Shared purchases, which typi- cally focused on microform sets during the program’s early years, now tend to support such things as the “Last Copy” project which provides funding for selected jou- nal titles that have only a single remaining subscription within the UC libraries. Clearly, the last development rises directly from the same decline in acquisi- tions rates that most U.S. academic March 1995 87

Upload: christy-hightower

Post on 30-Aug-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

The Consortium as Learning Organization: Twelve Steps to Success in Collaborative Collections Projects

by Christy Hightower and George Soete

A project to share translation journals in the physical

sciences among nine California academic research libraries achieved many of its goals,

yet it encountered major difficulties during

implementation. Based on their experience with

the project, the authors offer twelve strategies for ensuring

success in collaborative collections projects.

Christy Hightower is Collection Development

Specialist and Engineering Information

Specialist, Science and Engineering Library

0175E, University of Ca/ifornia,San Diego,

9500 G/man Drive, La /o//a, California

92093-0175 <[email protected]>.

George Soete is an independent library

consultant, former/y Associate University

Librarian, Collections,

University of California, San Diego

< [email protected] >.

A ccording to David Garvin, “a learning organization is one that is skilled at creating, acquiring, and

transferring knowledge, and at modifying its behavior to reflect new knowledge and insights.“’ A critical question for libraries, as they move toward greater interdepen- dence, is whether a consortium can behave like a learning organization, gaining knowledge from both success and failure and modifying behavior accordingly.

This article focuses on a case study: a pilot collaborative collections manage- ment project mounted by the University of California and Stanford University Librar- ies to provide for shared access to transla- tion journals in the physical sciences. By some measures the project failed, because of both unexpected developments in the environment and flaws in design and implementation: though it still exists and still saves the libraries more than a $100,000 a year, the project has been put on hold. By other measures, however, it has succeeded, saving scarce resources, ensuring a rational coverage of certain research materials, and providing valuable lessons on which to build future projects.

“A critical question for libraries, as they move toward greater

interdependence, is whether a consortium can behave like a learning organization, gaining knowledge from both success

and failure and modifying behavior accordingly.”

The following is an account of a mid- project evaluation that uncovered serious problems in the Science Translation Jour- nals Pilot Project. Though it contains a

description of the project’s genesis, objec- tives, and structure, as well as an analysis of the problems encountered, the true emphasis is not on those aspects; nor is it on dollars saved, journals canceled, or other quantitative measures. Rather it is on the lessons that were extracted from this “noble failure,” here presented as 12 strat- egies for promoting success in collabora- tive collections projects.

BACKGROUND

Collaborative collections management is an established tradition among the libraries of the University of California (UC), comprised of nine campuses spread across the state. Though each library man- ages its own collections program and bud- get, since the mid-1970s they have joined in a shared acquisitions effort, now called the Shared Collections and Access Pro- gram (SCAP).2 SCAP, which derives its funding from a central pot of money repre- senting approximately 3% of the total acquisitions budgets of the UC libraries, has acquired millions of dollars worth of materials for the UC libraries while hold- ing to the objective of avoiding duplica- tion. In recent years, the program has expanded and changed focus: shared funds have been used to acquire and mount databases on MELVYL@ the UC Libraries’ union catalog, and matching funds have been provided by the local campuses for some purchases. Most importantly, perhaps, the key purpose of the program has shifted from enrichment

to survival. Shared purchases, which typi- cally focused on microform sets during the program’s early years, now tend to support such things as the “Last Copy” project which provides funding for selected jou- nal titles that have only a single remaining subscription within the UC libraries.

Clearly, the last development rises directly from the same decline in acquisi- tions rates that most U.S. academic

March 1995 87

research libraries have experienced during the last 10 years. While local situations dif- fer somewhat, the UC libraries as a whole have lost approximately 30% of their col- lections buying power since the mid- 1980s. Thousands of journal subscriptions have been cancelled and monograph acqui- sitions have fallen pre-cipitously. At one campus library, the number of volumes acquired per year per faculty member declined by 33% from 1985 to 1992, a sit- uation fairly indicative of what happened throughout the UC system.

Jolted by the sharp increases in journal prices during the mid-1980s the UC libraries began to take collaborative col- lections management very seriously. Sev- eral pilot projects were developed within a strategic planning context to test collabo- rative project methodologies3 The Sci- ence Translations Project was one of those projects.

“Jolted by the sharp increases in journal prices during the mid-1980s, the UC libraries began to take collaborative

collections management very seriously.”

The purpose of this article is twofold: (1) to describe some typical problems that might arise in a collaborative collections management project and, most impor- tantly, (2) to suggest practical strategies, both preventive and remedial, that can assist other libraries in forestalling and resolving problems during collaborative collections projects.

PROJECT DESCRIPTION

Physical science translation journals were targeted for collaborative collections management because they are a body of expensive material that receive generally low-to-medium use. The goal of this five- year project was to assure that at least one copy of each science and engineering translation journal essential to UC pro- grams was held somewhere in the UC sys- tem or at Stanford. It was not a cancellation project, although reliance on a shared copy would allow campus librar- ies to cancel low-use duplicate subscrip- tions if they wished.

Involving nine universities across the state (Stanford and eight UC campuses) and including 253 journal titles for which the libraries were paying a total of

88 The Journal of Academic Librarianship

$208,000, this was a large scale collabora- tive collections project in every sense of the word. Hopes for the project were high and anticipated benefits were:

1. Improved access:

Guaranteed availability within the UC system or at Stanford, for at least five years, of at least one copy of each science and engineering trans- lation journal essential to UC pro- grams; and

Expedited interlibrary loan of these materials, with expected turnaround time of 48 hours.

2. Increased commitment to communi- cate: . Participants agreed not to cancel

these titles even after the project was complete without notifying the other UC libraries and Stanford of its intention at least six months in advance.

3. Cost savings:

l Reliance on the shared copy would allow the libraries, if they wished, to cancel duplicate subscriptions to all but their most heavily used transla- tion titles.

In fact, most of these benefits were achieved. Access was improved. ILL tum- around time was rated as acceptable by participants and was generally less expen- sive and faster than the alternatives avail- able outside the consortium at the time. Though only $33,000 of the existing total cost of $208,000 was met with new SCAP funds allocated for the project, many libraries were able to realize substantial additional savings by canceling duplicate subscriptions to project titles. Cancella- tion of duplicates saved $114,055 system- wide in just the first year, and the savings have been compounded in subsequent years.

DESIGN SPECIFICS

The project was designed in 1989 by a committee consisting of one representa- tive from each library. Initially, each repre- sentative identified translation journals currently held at his or her location, creat- ing a consolidated list of 302 titles. Repre- sentatives then ranked each title owned by their respective library (a library that did not currently own the title could not rank it). Rankings reflected the perspective of the holding library on the importance of each title, and used the following scale:

Rank #1: This rank meant, “Very high importance. Will definitely keep. Must be on this campus and available in California. Library willing to assume retention responsibility.”

Rank #2: “Medium importance. No immediate plans to cancel. Should be available in California in at least one location. Would assume retention responsibility if SCAP funding were available to pay for the title.”

Rank #3: “Low importance. Will prob- ably cancel or have already cancelled. No need for a copy in California. Suit- able for recommendation to Center for Research Libraries.”

The most important titles (those ranked #1) were designated as local library responsibilities, to be shared but paid for by local funds. Titles of medium impor- tance (ranked #2) were to be shared and paid for by SCAP funds. Titles of low importance (ranked #3) were usually can- celled and referred to the Center for Research Libraries as purchase proposals.

Each title in both the high and medium importance categories was then assigned to a specific library, based upon negotia- tions by the committee. The number of assignments varied: one library received none, while another received 53. The library accepting “retention responsibil- ity” agreed to accept official responsibility for acquiring and housing the title and for providing expedited interlibrary loan ser- vices to participating libraries. Libraries accepted these holding responsibilities for a five-year period, beginning with the 1990 calendar year.

REVIEW METHODOLOGY

Despite its apparent success, some par- ticipating libraries called for a review of the project early in 1993. The principal cause was a critical environmental devel- opment: California’s_and the Univer- sity’s-declining resources. With journal buying power having fallen 20-25% over a three-year period, several libraries were questioning the value of their substantial monetary investment in this program, which was supporting mostly low-use titles, given the serious cutbacks they were being required to make in their other jour- nal subscriptions to manage shrinking budgets. This project, like most projects in today’s environment of scarce and dimin- ishing resources, was facing a serious test of both its philosophy and its practice of collaboration.

The project was reviewed by a new committee composed of one representa- tive (usually the physical science bibliog- rapher) from each participating library. The review aimed to: (1) reassess the project’s current value; (2) entertain pro- posals from libraries to re-rank titles (which would include proposals to cancel some titles); (3) verify retention locations and holdings; and (4) identify titles that had ceased publication since the project began. Review results were posted on e- mail and the committee then began exten- sive electronic discussions to determine the project’s fate.

THE PROBLEMS ENCOUNTERED

Most of the news was good. Three and a half years into the five-year project, the review found that the majority of the titles (64%) were still considered valuable by project participants, and these retained their original rankings. On the other hand, 75 titles (27% of the total) were proposed for downgraded rankings: in effect, some libraries were requesting permission to cancel some titles because of low use and/ or perceived decreases in quality (See Table 1). Consensus was going to be difft- cult to achieve, however: not all libraries were equally disenchanted, either with the project as a whole or with those particular 75 titles. Nor was the original project agreement specific enough to settle the question.

Adding to the feelings of disenchant- ment, and further muddying the waters, was the fact that the review also uncovered other unexpected problems caused by either lax project design or unmonitored project maintenance. The most stunning of these was the discovery that 23 titles (8% of the project total) had already been mis- takenly canceled prior to the review. The general explanation for these cancellations was that the titles lacked identifying notes in local online catalog records to indicate their special project status. The lack of notes allowed them to be canceled during previous library cancellation projects when the titles were identified as low use and high cost. Definitely a loophole that needed to be closed!

Other problems also surfaced once the project data were organized into a data- base: three titles had never been purchased because of miscommunication; two titles were erroneously assigned multiple library locations; and two titles purchased for the project were discovered to be out of scope because they were not actually translations.

Table 1 Project Review Statistics

Initial Findings (1993)

Titles Reviewed 276

Titles Already Cancelled 23

Libraries That Had Cancelled Titles 4

Titles Proposed for Downgraded Ranking 75

Titles That Had Ceased Publication 12

Titles Never Purchased 3

Titles Found to be Out-of-Scope 2

Final Data (1994)

Titles Remaining in Project 253

Titles That Changed Library Location 28

Titles Remaining Cancelled 7

Other Titles Dropped 5

Total Funds Committed $208,266

SCAP Funding Requested $ 33,797

(8%)

(44%)

(27%)

(4%)

(1%)

(1%)

(11%)

(3%)

(2%)

DIFFERING VIEWS OF SUCCESS

What emerged during the review dis- cussions were at least three distinctly dif- ferent viewpoints on whether the project had succeeded or failed. From the “up close and personal” perspective of the review committee members, all of whom were front-line bibliographers acutely aware of the effects (both pro and con) of collaborative projects on their ability to meet their users’ needs, the problems encountered and the issues to be resolved loomed large. Some of these bibliogra- phers wanted to: (1) change the project in midstream to make it reflect the realities of budgetary declines; and (2) rationalize the “premature cancellations” and allow some additional titles to be canceled even though the five-year commitment period had not yet ended. Other committee mem- bers felt that the commitments were cru- cially important to maintaining a climate of trust within which to build further col- laboration; for these people changing the rules spelled the failure of the project. A third view was expressed by the Collec- tion Development Officers of the campus libraries, who felt that the problems encountered had been quite normal, that the project had been successful despite the glitches, and that what was learned in the project-a pilot project, after all-should translate into improved design and main- tenance for this project and for others to come. Discovering and acknowledging that there are likely to be significant differ- ences in the perspectives of practitioners

and administrators of collaborative pro- jects was one of the major eye-openers of this review.

RESOLVING THE PROBLEMS

After about six months of intermittent e-mail conversations and votes, a compro- mise agreement was finally reached. It was decided that no further cancellations of titles would be allowed for the duration of the project. However, libraries would be allowed to trade retention responsibilities for project titles at this time, on a purely voluntary basis. This trading would allow some libraries to be relieved of some bur- densome retention responsibilities and, since there were a few duplicate subscrip- tions in the system for titles that had been mistakenly canceled, the trading would restore some canceled titles to the project without requiring libraries to reinstate canceled titles and purchase missing back files. Twenty-eight titles (none of them SCAP-funded) were traded in this manner, thereby restoring 16 of the previously can- celed titles completely as well as relocat- ing an additional 12 titles to sites where they are expected to receive more use. This left seven titles still canceled. It was decided to allow these seven titles to remain canceled, rather than requiring reinstatement at this late date.

In preparing this article, the authors sent a brief survey to the members of the review committee. Responses indicated that participants gave the project, overall, a passing grade in spite of the problems encountered in the review. Most signifi-

March 1995 89

cantly, all participants registered their willingness to participate in further collab- orative projects if the sorts of problems encountered in this one were avoided through more careful design and mainte- nance.

“We may be more familiar with designing successful technical systems for resource sharing, but we must also take care to

develop the associated human systems that promote trust, open communication, and

constancy of purpose. Most important, as consortia we

must “learn to learn.”

DEVELOPING COLLABORATIVE

PROJECTS: HUMAN AND DXHNICAL SYSTEMS

In any collaborative collections man- agement program, there is a need to develop both technical and human sys- tems that will lead to success. Technical systems might include hardware, the materials themselves, funding and accounting programs, shared electronic catalogs, acquisitions procedures, and document delivery. In a real sense, how- ever, it is the human systems that are most critical and might be the least attended to by designers. At the planning level, human systems include setting realistic objec- tives, resolving disagreements creatively, and offering meaningful incentives; how- ever, there are also significant climate fac- tors such as trust, risk-taking, perceived equity, and fear of loss. Success in both systems is achieved through effective training and ongoing communication, as well as thoroughness and care in the design of projects.

A “TWELVE STEP PROGRAM” FOR MORE SUCCESSFUL RESOURCE SHARING

The following suggestions, focusing on both the human and technical systems needed to make collaborative collections management successful, derive from our experience with the Science Translations Project.

1. Incentives need to be developed of sufficient strength to promote individual behaviors contributing to project success.

Moreover, since resource sharing is an entirely new mode of working for many staff, attention needs to be paid to a broad range of incentives. Librarians need to understand the benefits their users are deriving from the project; they need to be able to project fiscal savings and, if possi- ble, reallocate at least some of the savings to other service priorities. Matching fund- ing is a particularly effective incentive, since it is a concrete signal of support. (One of the reasons the Science Transla- tions Project ran into problems was that SCAP funding was withdrawn during its last year). Moreover, staff who develop agreements should be acknowledged for their creativity and hard work, both infor- mally and in formal performance reviews. Such incentives are likely to reinforce pos- itive, committed attitudes toward the projects and other participants.

2. Training that focuses on the human dimensions of collaboration as well as on technical and procedural matters is an indispensable component. Staff who are developing agreements need to be trained in the key skills of project planning, inter- institutional communication, and conflict resolution. They need to know what to expect in a collaborative collections man- agement setting and how to cope with problems. Issues such as trust and fear of loss of control need to be confronted squarely before projects are developed; and skills in negotiation and communica- tion need to be stressed and practiced in a training setting. Though we provided ori- entation sessions for bibliographers, the focus was on program guidelines. In effect, we did not stress the interpersonal dimension enough.

3. It is critical that participants under- stand the importance, purpose, need, and permanence of collaborative collections management. Most libraries are collabo- rating because their budgets are taking severe cuts and/or they view collaboration as a means of assuring-perhaps even expanding on-an adequate base of resource materials. Some participants, however, will view pilot collaborative projects as not serious-perhaps as a momentary diversion from the real busi- ness of acquiring and building local col- lections. Such misunderstandings can lead to broken agreements and feelings of betrayal. Establishing the serious purpose and enduring change represented by col- laborative collections management is another critically important step at the beginning of any project.

“In any collaborative collections management

program, there is a need to develop both technical and

human systems that will lead to success.”

4. The University of California expe- rience bears out the importance of retain- ing a focus on the positive. At the same time, participants need clearly to under- stand the consequences offailing tofollow through on commitments for colleagues in other libraries and for the customers served by the consortium. Cancelling a shared journal is not merely a local deci- sion: it is the business of the whole consor- tium. Collections managers need to be encouraged to view agreements as com- mitments to people, not just technical exercises that are subject to quick alter- ation as environmental factors change.

5. A customer focus is critical as projects are developed. In a profession like librarianship, where motivation is likely to be largely of the self-actualizing variety, a strongly held sense of benefit to customers will promote effective collabo- ration. It can be, for many of us, its own reward. In a consortia1 setting, however, we are invited to stretch the concept of customer to include all of the customers of the consortia1 libraries.

6. Once again, effective communica- tion between administrators andpractitio- ners provides the indispensable basis for project success. One of the real surprises of this experience was that bibliographers and collections officers saw the project so differently. Setting aside the many grada- tions of perception for a moment, collec- tions officers tended to focus on bottom line savings and gross indicators of effec- tiveness: for them, the project was a resounding success. Bibliographers tended to focus more on details and were quicker to see the problems that developed as indicators of failure. Had administrators and practitioners communicated expecta- tions and benchmarks more clearly from the beginning, some problems might have been avoided.

7. While a little ambiguity may be helpful in getting a project up and running quickly, too much ambiguity can create serious problems. Make your written col- laborative plan specific and complete. Use specific dates and times (project starting and ending dates, acceptable interlibrary

90 The Journal of Academic Librarianship

loan [ILL] turnaround times, etc). List each important responsibility (ordering, claiming, binding, housing, online record keeping, etc.) in specific terms, being par- ticularly careful to take written note of the quantities and qualities desired. State the project goals clearly, and the criteria by which the project will be judged. For example, are use data relevant or irrele- vant to the project? If relevant, state how and when to measure use from the start. A checklist format may be helpful.

8. We certainly live in times of change. Build a plan for change into the project contract. If the project is to remain inviolable during its lifetime, state that clearly in writing before participants sign on the dotted line. Collaboration is more likely to be successful, however, if you plan for some change. But make the con- tract specific on these points, or it will not be useful! State what changes can be made in the project and what cannot change dur- ing the project’s lifetime. Specifics might include: If the project is not “successful” as judged by the criteria set forth in the collaborative plan (see above), can the project be changed before it is over? If there are significant environmental changes (budgetary calamities, etc.) that would justify changing the project, is there an “escape clause” by which a participat- ing library may withdraw completely? Who may call for a change in the project, and when? If votes are to be taken, are they to be majority or unanimous? The best time to discuss modifications to the project is before change is necessary, while heads are still cool. Rather than thinking of this process as focusing on the negative, think of how reassuring it will be later for everyone to know there is already a plan to deal with change.

9. If possible, do not assign anything mechanically. While it is important to dis- tribute responsibilities and costs equitably among the participants, it is also important to make a good “marriage” of responsibil- ities, costs, titles, and holding locations. Good marriages consider the interests and emotions of the participants. Mechanical or strictly numerical methods of choosing titles, or of assigning titles to particular locations, may be scrupulously fair but these methods do not always inspire long- term faithfulness.

10. Appoint a single “guru” to moni- tor project maintenance. The level of detail in large collaborative projects is sig- nificant, yet even the smallest of details can lead to problems if overlooked (as our experience with missing online catalog

notes shows!). Appoint a coordinating “guru” from among the front-line practi- tioners, one who is good with details, to coordinate the annual maintenance reports. Make the smooth running of the project the responsibility of the guru, whose job it is to know everything about the project as it progresses. Acknowledge the importance of this job with appropriate recognition and clerical support, and make it a priority to fill this position promptly should it become vacant.

11. Verify maintenance steps and mid-project modifications with initialled written check sheets, deposited with the project guru. E-mail is great for discus- sions, but its informal and etherial nature makes it a poor tool to verify that neces- sary actions have actually been taken. Develop a quick check-sheet format (using your database, see below) for annual project verification, and have each library’s representative initial each step. Ask for only that data you really need to know to verify project health, but do not sacrifice quality control just to save a little time. There is just no getting around the need to spend some time every year on project maintenance. An ounce of preven- tion really is worth a pound of cure!

“Now chronic economic

malaise is forcing many of us

not only to develop models of

interdependence, but also to

take these models from theory

to practice.”

12. Use electronic tools. Use a spreadsheet program or relational data- base to automate your record keeping, to sort the data (useful for finding duplicate or missing subscriptions), and to generate reports periodically throughout the project. The initial time investment in data entry is well worth the price for the level of quality control a database can provide. Similarly, online catalogs (both local and union) are vital tools for communicating the project intentions to every level of staff involved in the process. As we found out, its worth the time to take full advantage of this tool. Note the project starting and end- ing dates and indicate the special protected nature of these subscriptions in the order records of the affected titles.

POSTSCRll’T

It has taken libraries centuries to develop the self-sufficient collections management models that most still work within. Now chronic economic malaise is forcing many of us not only to develop models of interdependence, but also to take these models from theory to practice. To succeed, such models must be carefully designed and the resulting practice must be closely monitored and adjusted as required throughout their lifetime. We may be more familiar with designing suc- cessful technical systems for resource sharing, but we must also take care to develop the associated human systems that promote trust, open communication, and constancy of purpose. Most impor- tant, consortia must “learn to learn.” The Learning Organization presents a model in which consortia continually try new strat- egies, view each trial as a learning experi- ence, and adjust as they learn.

NOTES 1. David A. Garvin, “Building a Learning Organization,” Harvard Business Review 71 (July-August, 1993): 80. 2. Marion L. Buzzard, “Cooperative Acquisitions within a System: the University of California Shared Purchase Program,” Coordinating Cooperative Collection Devel- opment: A National Perspective, edited by Wilson Luquire (New York: Haworth, 1986), pp. 99-113. 3. George J. Soete and Karin Wittenborg, “Applying a Strategic Planning Process to Resource Sharing: The Changing Face of Collaborative Collection Development among the University of California Libraries,” Advances in Library Resource Sharing II, edited by Jennifer Cargill and Diane J. Graves (Westport,CT: Meckler, 1991), pp. 51-59.

March 1995 91