Close this search box.
Close this search box.

Moving to a Unified Collection

Since the introduction of Tripod, the three college libraries have become increasingly interconnected in the ways they develop their collections and do their work. Today, faced with growing space limitations, the libraries have two choices. On the one hand, they can continue on much the same path as they have been on for the last decade, engaging in cooperative purchasing activities where appropriate but maintaining three independent collections. If the libraries choose this path, then each of them will have a collection that looks much like that of the others. They will continue to buy many of the same books every year, and they will continue to house many of the same older, little-used books.

The second option is to take advantage of the power of Tripod and to treat the three collections as a single collection-one that is capable of providing resources similar to those of a comparably sized university library. Each of the libraries must, of course, retain a core collection that supports a liberal arts curriculum. Circulation statistics indicate that each of these collections should comprise 100,000 to 250,000 volumes. Beyond that core, the libraries have the potential to build their collections into a coherent whole that could support the research needs of students and faculty to a much greater extent than the individual collections currently do.

Developing such a collection means working together to shape the existing holdings with an eye to maximizing the range of titles held system-wide. It also means acquiring new materials in a collaborative manner in order to limit overlapping titles among the colleges, and ensure availability of funds to buy a wider range of more specialized research materials. The Planning Group recognizes that this model for library collections presents many challenges: defining a core collection, buying books collaboratively, and defining new areas in which to acquire materials, among others. At the same time, we believe that this model has great potential for providing the rich library collections that students and faculty need.

Changing the Ways the Libraries Work

As the libraries move toward a unified view of their collections, they are finding it necessary to create new ways of doing their work. In the last few years they have made considerable progress in creating mechanisms for managing collections in a collaborative environment, reaching decisions about new resources, and communicating about collection interests and opportunities. More significant changes are likely to be necessary as the scope of cooperation expands. To assess the scope of these changes, the Planning Group looked at four areas:

  1. Organization of collection development across the three colleges
  2. Development of tools to manage cooperative collections
  3. Development of tools to reduce overlap and regain space
  4. Creation of an environment in which cooperatively built collections can be used effectively by faculty and students

1. Organization of Collection Development.

Each of the three libraries takes a different approach to organizing its collection-development activities. Bryn Mawr has 17 librarians who serve as liaisons to academic departments and programs. Haverford concentrates the duties among eight librarians. Swarthmore has five librarian liaisons. The numbers, however, do not tell the full story. At Bryn Mawr, the liaison program is only four years old. Previously, responsibility for collection development had been largely in the hands of the director of the library, a bibliographer, and the heads of the branch libraries. All the Bryn Mawr librarians now have collection-development duties, but many of them are still learning the work, and collection development is a minor part of their jobs. At Haverford, more than half of the librarians have been involved in collection development for many years, and this has long been regarded as an important component of their work. Swarthmore has the fewest number of librarians in collection development, and most of the work is concentrated in the hands of three people: the librarians for the humanities, social sciences, and sciences. Each library has a head of collection development who is responsible for coordinating the work.

The differing organizational structures at the three libraries present challenges to undertaking Tri-College collection projects. The decision-making processes also vary among the campuses and among disciplines. The burden of work falls most heavily on those responsible for the largest number of disciplines. Experienced bibliographers who have built strong working relationships with their faculty tend to have an easier time reaching decisions than do newer librarians, who are less familiar with both their faculty and the discipline. As collection decisions become increasingly interconnected, more opportunities will need to be found to upgrade collection-development expertise across the libraries through in-house training, release time for further academic work, and opportunities for experienced bibliographers to share their expertise with newer staff. A reexamination of the libraries’ organizational structures for collection development may also be warranted to see whether a closer alignment of structures would help improve the quality and speed of decision making.

Increased collaboration presents an opportunity for moving toward the model of the university library subject specialist. Currently the libraries follow a liberal arts college model, in which librarians are generalists and provide reference, instruction, and collection-development services to a number of departments. In the subject-specialist model, the librarian has advanced training in the discipline and focuses on building strong research collections and supporting the teaching and research in that field.

How far to go in the direction of the subject-specialist model is a matter for more discussion and testing. At one extreme is a replication of the research university model, in which a single bibliographer would be responsible for all collection decisions and advanced reference work in an academic field for all three colleges. But there are also intermediate steps, such as establishing Tri-College subject specialists in fields where expertise is difficult to come by, such as East Asian languages, or as a way of taking advantage of the special expertise of certain librarians. The role of the Tri-College subject specialist is also open to different models. Overseeing all collection development in a field is one possibility; under another model, subject specialists could play more restricted roles that would still help elevate the level of collection decisions and research support that the libraries provide. They might, for example, serve as advisors to senior thesis writers at all three campuses, help with evaluating difficult collection issues, or provide guidance to new librarians on best practices in collection development.

2. Tools to Manage Cooperative Collections.

As the libraries move toward increasing cooperation in building and maintaining their collections, it is essential that effective mechanisms for sharing information and making decisions be created. The libraries have already taken a number of important steps in this direction. For example, the three heads of collection development meet regularly, as do other groups with common interests, notably the science librarians. In addition, the following data-gathering and decision-making structures are in place:

  • Collection statistics. The data gathering required for this report produced the most comprehensive view to date of the libraries’ print collections and the way they are used. All of the subject bibliographers have access to these data through a Web interface. Steps are being taken to ensure that these statistics are updated annually and to eliminate discrepancies in the way the libraries record data.
  • Electronic collections data. The technical services departments of the three libraries have developed the ERTS, a database that records critical information such as price, renewal date, and access restrictions for all the electronic collections to which the libraries subscribe, whether individually or collectively.
  • Electronic Resources Group. To manage the acquisition of electronic collections, the libraries have formed the Electronic Resources Group, a committee consisting of two librarians from each campus and reporting to the heads of collection development. A member of this committee has been the chief negotiator for Tri-College electronic purchases for the last two years. To track information about resources being considered, the libraries have set up a “trials database” that lists the products currently under trial, the terms of the purchase, and the date the trial ends. The database has a comment board where librarians can post their assessments of the resource.
  • Analysis of curriculum. During this study project, the Planning Group tried to compile systematic data on the curricular interests of the three colleges in order to quantify potential demand for collections in different subject areas. We obtained course lists from the colleges’ registrars, converted them into a database, and began to catalog them. The project was too large to complete during the course of this study; however, preliminary findings demonstrate its potential for drawing a clearer picture of each campus’s interests.
  • Last-copy policy. The libraries have recently adopted a policy governing the weeding of collections. A key purpose of the policy is to ensure that weeding projects do not eliminate materials that are likely to be needed on other campuses.
  • Specific projects. Over the last three years the libraries have undertaken a number of projects that have brought together bibliographers from the three campuses to work on common problems. These projects have included identifying Web sites for inclusion in Tripod, creating Tri-College subject guides for the Web, canceling standing orders held by more than one library, and agreeing on electronic journals to acquire through ScienceDirect.

As a result of these efforts, the librarians at the three colleges are building good working relationships, gaining an appreciation for the potential richness of the libraries’ combined holdings, and developing tools to improve the understanding of existing collections and concomitant decision making. These structures and working relationships have been necessary initial steps toward building an integrated Tri-College collection, but more will be needed if the work is to progress.

3. Tools for Reducing Overlap

The two most important goals are (1) to reduce the overlap in existing collections to create shelf space for new books and journals and (2) to reduce the overlap of current publications to make funds available to strengthen collections in the colleges’ fields of interest. To accomplish these goals, the libraries will need to develop more systematic and coordinated methods for weeding their collections and acquiring new materials.

  • Deaccessioning. A high percentage of the Tri-Colleges’ volumes show little or no circulation over the last 11 years, and a significant number of these low-use volumes are held by more than one library. If low-use overlapped and outdated texts can be weeded, the libraries stand to gain substantial amounts of expansion space without reducing the depth of the shared collection. To coordinate a large-scale weeding project, the librarians at the three campuses will need to work closely with faculty to gain a clear understanding of what books need to be close at hand and what can be housed off-site. The mechanics of making large-scale withdrawals in a collaborative and efficient way need to be worked out. The copy to be retained must be the one that is in the best physical condition, and no library should withdraw books that are of potential interest to either of the two others. During the most intensive period of weeding, additional Tri-College staff will be needed to manage the withdrawal process in order not to overwhelm the libraries’ regular staffs and to ensure that the process moves forward in a timely fashion with appropriate communication in place.
  • Approval plans. The libraries purchase more than half their monographs through approval plans. The purchases made through these plans total more than $500,000 yearly. Approval plans are a way for the libraries to receive new publications from major scholarly publishers automatically, thereby giving the libraries a high degree of confidence that they are acquiring the most important new works while substantially reducing the cost of acquiring them. Bryn Mawr and Haverford have had a joint approval plan since the early 1970s. By pooling their book-buying dollars and keeping their overlap rate to about 15 percent, these two libraries have been able to build substantially broader collections than would have been possible if they had worked independently. Swarthmore began using an approval plan five years ago. In the last year, approximately 80 percent of the books acquired through the Swarthmore plan were also acquired by one of the other two libraries. The value of the Swarthmore acquisitions was approximately $170,000. If the three libraries can coordinate their approval plans, a substantial amount of money can be freed to invest in materials not currently being acquired.

Several approaches could be taken to coordinating approval plans. The first is a Tri-College version of the current Bryn Mawr-Haverford plan. With this approach, the three colleges would agree on a profile designed to acquire most new academic press books automatically. If the plan were to follow the existing Bi-College model, bibliographers from the three colleges would examine each week’s shipment of books and then meet to decide where each new book should go and which books should be duplicated. One disadvantage of this model is the amount of travel and discussion that would be required of the bibliographers. Certainly some communication can take place by e-mail and conference call, but decisions on location and duplication can be difficult if the books are not examined firsthand. This approach could become much more practical if publishers and approval vendors provided substantive information about their new books in advance. The approval plan vendor for all three colleges, Academic Book Center/Blackwell, has expressed an interest in opening discussions toward creating a “virtual approval shelf” that could eliminate the need for much of the physical examination of the books.

The libraries should also explore other options for managing their approval plans, in the event that the virtual approval shelf proves impractical. One possibility is for the three colleges to have a single approval plan profile, but rather than the librarians discussing each book as it arrives, the books would be distributed automatically into the three collections according to a predetermined formula. Another option is to continue the current arrangement of separate Bi-College and Swarthmore approval plans, but with coordinated profiles that would reduce the amount of overlap.

4. Effective Faculty and Student Use of Cooperatively Built Collections

An integrated collection for the three colleges is desirable only if it is readily usable by faculty and students and if it provides them with a richer set of resources than they currently have available.

  • Browsing in Tripod. The concern most frequently raised by students and faculty about the unified collection concept was the loss of the ability to browse the shelves. Tripod has a mechanism for looking at cataloging records in call-number order, but it is neither easy to find nor easy to use. Improving this system must be a priority.
  • Enhancement of recent cataloging records. Even if browsing by call number in Tripod becomes easier, the lack of information in most cataloging records makes it difficult to determine whether a book is worth consulting. In order to make Tripod a more effective tool, the libraries have been buying table of contents information and adding it to the cataloging records of new books. This information is now available only for books published since 1995. More enhanced cataloging information is available commercially, however, including tables of contents for books published between 1991 and 1995, book reviews, summary notes, and portions of first chapters. The libraries should purchase as much of this additional information as seems appropriate to make Tripod a reasonable alternative to physical browsing.
  • Enhancement of older cataloging records. Catalog enhancements are available commercially only for books published fairly recently. If the libraries are going to eliminate overlap copies of many older books, it will be important to find ways of helping faculty and students evaluate the remaining copies through Tripod. A possibility worth considering is digitally capturing the tables of contents, indexes, and first chapters of such books and linking the images to the catalog record.
  • Expansion of the range of materials provided. If the libraries could reduce the amount of money spent on acquiring multiple copies of books, funds would become available to acquire a broader range of materials than the libraries are currently buying. To determine how this money should be spent, librarians will need to consult with faculty to identify areas where the collections could be strengthened. Further studies of collection use will also help to indicate areas needing additional support. The most important measure of unmet need is the amount of borrowing the students do from PALCI and other libraries beyond the Tri-Colleges. Meaningful data on interlibrary borrowing was not easily available for this study, and further investigation is warranted. The libraries should also compare their holdings with those of comparable libraries to determine areas of relative strength and weakness.

Key Findings

  • Each of the libraries organizes its collection-development activities differently, resulting in different approaches to collections and different methods of decision making. In a collaborative environment, the libraries might look at adopting other models for organizing their work, including the model of the research university subject specialist.
  • In the last few years, the libraries have developed a number of tools for managing collection data and new structures for making joint collection decisions. Additional tools and structures will be necessary as the collaborative work increases.
  • Large-scale, coordinated weeding projects are an essential part of any long-term strategy for recovering library space. Undertaking such projects without weakening the overall quality of the collections will require careful communication among librarians and faculty and, at least for a time, a central project staff to manage the process.
  • Reducing the amount of overlap in new acquisitions has the potential to free funds to spend on materials not currently being acquired. Since a major portion of the libraries’ book budgets is spent through approval plans, coordinating these plans is critical.
  • The greatest faculty and student concern about the move to an integrated collection is the loss of the ability to browse the shelves. The libraries will need to find methods of making the shelf-browsing function in Tripod more usable and of enriching the cataloging records so that virtual browsing is an acceptable substitute.


Skip to content