CHAPTER TEN

RESOURCE SHARING: COLLECTION DEVELOPMENT AND DOCUMENT DELIVERY

Thus far we have considered various bibliographic services and publishing ventures that make information about information and scholarly information itself available in electronic form. The availability of text and data in that form---whether so produced in the first instance or converted retrospectively from print products[1]---in principle permits a degree of resource sharing among institutions that was unimaginable in the past. The extent to which institutions will practice it, however, will depend upon a host of other considerations.

Two broad types of information service, bibliographic and full text, constitute the minimal necessary preconditions for successful resource sharing. Sharing depends on adequate information about the existence of materials and their location, which the various bibliographic services described earlier furnish, and on the availability of full texts in electronic form, which are more easily transmittable than printed texts. What many envision, ultimately, is a situation in which the individual reader at a particular institution is led easily through a series of options culminating in direct access to the aggregate content of the nation's principal research collections, in which local and remote library catalog entries and bibliographic records merge with readily retrievable electronic versions of full texts that can be downloaded and printed locally at one's own workstation.[2] Just as the texts of secondary works in electronic form can be integrated with the underlying base data in a more fully electronic environment, so there might eventually be an unbroken continuum of types of information, from entries in the bibliographic record of monographic collections, to records in various bibliographic, indexing, and abstracting services, to full texts of databases of primary material and studies based on them. The sharp distinctions we now make among these different kinds of information are implicitly challenged by electronic technologies.

Although university libraries have practiced resource sharing for years, for many the preferable option nonetheless remains local ownership of as much of the universe of published scholarly material as resources permit. Interlibrary loan services are thought to be inefficient; the lending library, understandably, attends to its own readers' needs before addressing those of readers elsewhere. The difficult economic circumstances in which research libraries currently find themselves argue for new models, and electronic information technologies seem to hold particular promise.

The degree to which such sharing will necessarily result in cost savings to individual institutions is difficult to determine at this juncture. New paradigms governing local collecting and sharing within consortia will entail new economic relationships among publishers, vendors, and libraries. New pricing schemes will be related to the resolution of copyright issues.[3]

Another important argument for the new technologies is that they will streamline and extend the entire process of scholarly communication in some of the ways considered in the previous section. If such a vision is to be realized, however, it will require reallocation of resources away from expenditures associated with building a self-sufficient collection and toward those associated with cooperative collection development and sharing. The aggregate cost to individual institutions may not be lower, but access to larger universes of material may be facilitated.

EXAMPLES OF DOCUMENT DELIVERY MECHANISMS

In this section we will consider the characteristics of some of the resource sharing and document delivery arrangements various consortia have already attempted. In Chapter 4, we briefly described one such initiative at James Madison University, the University of Virginia, and Virginia Polytechnic Institute and State University.[4] What this initiative and other, similar ones have in common is that they apply existing services and technologies---interlibrary loan, photocopying, telefacsimile---to the problem of access to materials held elsewhere, a problem resulting from the prices of scholarly materials and the consequent inability of any individual institution to provide comprehensive access. Although they assume the medium of print and therefore do not adequately suggest how such sharing might function in a more fully electronic environment, they nonetheless might serve as prototypes of the kinds of infrastructures and organizational principles that might eventually emerge, instances of responses to the current situation anticipating what more might be possible in a fully electronic environment.

Conspectus

The Research Libraries Group has played a critical role in collaborative collection development initiatives from its founding; indeed, cooperative action has been fundamental to RLG's mission from the very beginning. In January 1980 RLG's Collection Management and Development Committee endorsed the recommendation of a subcommittee that

the committee develop an RLG collection policy statement ... to serve as a vehicle for cooperation with the Library of Congress and other major research libraries in developing an eventual national research resource collection of materials held ... by RLG and other major research libraries, with primary collecting responsibilities distributed among those libraries and LC, and with LC acting as a kind of "system equalizer" to minimize the impact of local program change on national research library resources.[5] [emphasis added]

The result was the RLG Conspectus, which permitted participating institutions to build collections complementing those of other institutions, thus ensuring the availability of rare titles.

How does such a scheme work? The designers of the Conspectus recognized, first, that it had to respect the autonomy of individual institutions, to facilitate planning but not have the prescriptive force of a policy. Each institution was free to base collecting practices on the profile of its academic program, for example, or on the availability of resources locally. The Conspectus was seen simply as a means to facilitate coordination of individual institutional efforts.

The Collection Management and Development Committee envisioned an interactive, online format that would allow one to search the Conspectus database by subject, institution, Library of Congress classification, and so on. If bibliographers at a particular university had to decide whether to purchase a certain title, they could first search the RLIN database to determine if any other RLG library had purchased or ordered it. If no record were found, they could then search the Conspectus database for information about collection strength in the field in question at other institutions. If they discovered, for example, that another RLG library had both a comprehensive collection in that field and a commitment to continue to collect at the current level, they would have the option of choosing to rely on the other institution's collection.

If there were at least three collections in a particular subject at levels termed "research" or "comprehensive," coverage was thought to be adequate; if two or fewer RLG libraries had research-level collections, it was thought that a member should be identified who could accept primary collecting responsibility. The Library of Congress agreed to consider assuming a primary responsibility in fields where there was neither an RLG member with a strong academic program in that field nor one interested in increasing its collecting to a level considered desirable by the membership. In return, the Library of Congress hoped to be able to depend on the collecting responsibilities of other research libraries in the country.[6]

Although the Conspectus was first designed for purposes of collaborative collection development, it has a number of other uses, as the authors suggest: it is being used to assign responsibilities in preservation; it can help identify materials for storage, in that diminishing local emphases in collection development might suggest lower use, a usual qualification for storage; and it can be used for purposes of allocation of staff and materials expenditures, in that it aids in collection assessment and therefore in appropriate allocation of resources.[7]

Collection interdependence remains central to RLG's mission.[8] There is currently an initiative designed to ensure that critical resources in particular journals remain accessible. Titles in chemistry, business, and mathematics considered essential to scholarship have been identified by subject specialists, and one or more institutions have agreed to continue their subscriptions to particular titles, regardless of other claims on the materials budget. Lists of the titles in question, annotated with information about collecting responsibility, are available on RLIN. The initiative will be expanded to include periodical titles in foreign law, geology, physics, and German literature, as well as art exhibition catalogs, German monographic series, and foreign newspapers.

As we have suggested, efficient document delivery services go hand-in-hand with collaborative collecting initiatives. The effort RLG is currently engaged in, for example, also involves the development of access and delivery procedures; the Task Force on Interdependent Collections is responsible for designing a service that would expedite delivery of copies of articles from the core titles to members requesting them.[9]

Before the development of electronic delivery technologies, document delivery relied entirely on methods familiar from traditional interlibrary loan services. Under the terms of the understanding that led to the RLG Conspectus, for example, members agreed to give priority to loan requests from other members, to respond within three days to any request, and to ship materials by United Parcel Service.[10] Electronic technologies, telefacsimile in particular, have greatly expedited transmission of shorter items; for book-length materials, telefacsimile is clearly not an appropriate distribution medium.

UnCover and Ariel

Two organizations in particular---the Colorado Alliance of Research Libraries (CARL)[11] and RLG once again---have prominently featured document delivery by electronic means among their services. CARL's UnCover service, for example, furnishes bibliographic information on articles from some 10,000 periodicals in a variety of disciplines; the information is entered into the CARL database when the latest journal issues are received by the CARL libraries.[12] UnCover2, the companion delivery service, delivers the full texts of articles from the journals indexed in the UnCover service. More than 97 percent of requests are met, all within 24 hours; 40 percent of the articles requested are sent by telefacsimile within the same working day, and optically stored articles are delivered in less than an hour. A copyright royalty fee is collected for each article ordered; publishers are compensated either directly or by arrangement with the Copyright Clearance Center. CARL offers a variety of payment schemes: clients may either pay by credit card or maintain deposit accounts; the average size of an account is $500 to $700, and reports on account use are provided monthly.

Of special interest and importance is RLG's new document transmission system called Ariel.[13] Any printed material can be scanned, stored, transmitted, and printed, including material containing photographic plates, charts, formulae, and tables; the system affords its users rapid, error-free transmission and print images of high quality and is designed for transmission of images over the Internet from one user's workstation to another and local printing on laser printers (the telecommunications charge incurred when telefacsimile is used instead is thus avoided).

What many of these initiatives and services involve is the prior conversion of printed material to electronic form; if scholarly materials are increasingly produced and distributed in electronic form in the first instance, sharing of materials will be easier still. In the case of distribution by telefacsimile, an existing technology was exploited almost immediately for purposes of resource sharing among libraries. The Ariel system offers advantages over that technology. Documents of any size up to 8 1/2 x 14 inches can be scanned directly; there is no need to photocopy them first. The transmitted images can be printed on bond paper of various sizes. Scanning, transmission, and printing are more rapid, and multiple transmissions of the original document are permitted. The user has access to the stored documents, so that any of them can be selected for transmission to other destinations; for copyright purposes a count of the number of times the document is transmitted to different destinations is displayed on the screen when the file is accessed.[14]

It should also be noted that many of the commercial bibliographic services described earlier also deliver full texts in electronic form. Dialog Information Services, for example, in addition to the abstracting and indexing services described earlier, also offers the full texts of such publications as the Atlantic, the Boston Globe, Consumer Reports, Harvard Business Review, Scientific American, Time, and the Washington Post, among a great many others; most of the full-text databases are machine-searchable.[15] Although such services are examples of document delivery services, they are obviously different in kind from the sorts of services considered thus far in that they are provided by commercial vendors; they are not the sorts of services, therefore, that would complement collaborative collection development efforts within library consortia. Moreover, the kinds of texts they deliver represent only a small portion of the kinds of materials academic libraries collect. They do, nonetheless, give some sense of how the full texts of materials of various kinds are increasingly available to individual scholars at their own workstations.

Faxon Finder

Of potentially far greater significance for academics (because of the nature of the material) is a service OCLC and Faxon Research Services, Inc., have recently initiated. Called Faxon Finder, it provides bibliographic information through OCLC's FirstSearch and EPIC services; alternatively, institutions may purchase a site license and load the database locally.[16] The companion delivery service, Faxon Xpress, delivers copies of journal articles located in the Faxon Finder database. The scanned document, which may contain images, illustrations, and other nontext material, is delivered to the user's telefacsimile machine or to a computer facsimile board; orders received by 6:00 P.M. are shipped by 6:00 A.M. the following day. Faxon offers a variety of accounting and billing plans: institutions may arrange for a deposit account; individuals may have their credit card accounts charged. The user is charged a fixed price and applicable copyright fees, and to ensure compliance with copyright provisions Faxon retains the bit-mapped image of the article only so long as to complete the transaction.[17]

At present, users of the Faxon service are able to receive copies of articles from a limited number of journals; eventually, the service will extend to all journal titles. Further, "[i]n five years," according to K. Wayne Smith, OCLC's president and chief executive officer, "you will look up a journal through OCLC databases, punch a button on the computer, and get the document in your hand."[18]

These various current initiatives permit one at least to envision what kinds of creative, productive resource sharing arrangements and document delivery services might eventually emerge in the electronic age, even if the current services will ultimately be seen as only first approximations of what is possible. The services of the Colorado Alliance in particular illustrate the kinds of advantages consortial arrangements offer; no one institution in the consortium could envision offering the full range of information services the members acting together can offer one another. The alliance could serve usefully as a prototype of similar arrangements other institutions might attempt. Only when there is fuller experience with such arrangements, moreover, might institutions undertake careful analyses of the contrasting cost implications of either attempting to build self-sufficient collections or instead distributing resources and delivering copies of materials not owned locally to fellow consortium members.[19] Among the costs of the latter model are those of electronically sharing bibliographic information on one another's holdings to permit collaborative collection development; disseminating copies of materials, either by telefacsimile or some other means; and compensating publishers in accordance with copyright provisions. Fuller data on these different elements of the cost structure should permit at least some preliminary analyses of the relative cost to individual institutions of each of the two principal models of scholarly communication considered in this study.