1. Anthony M. Cummings, Marcia L. Witte, William G. Bowen, Laura O. Lazarus, and Richard H. Ekman (Association of Research Libraries, 1992). [BACK]

2. Ekman has overall Foundation responsibility for its work with libraries, university presses, scholarly communication, historical societies, and centers for advanced study in addition to duties in other program areas (including the Foundation's work with Latin American libraries and archives) and for other aspects of the management of the Foundation. Quandt has been responsible for the development of the Foundation's program in Eastern Europe, which has, among other accomplishments, led to the automation and modernization of dozens of university and national libraries; he has also been working closely with another Foundation staff member, Thomas I. Nygren, on the development of similar library projects in South Africa. [BACK]

1. See, for example, Astle and Hamaker (1988), Joyce and Mertz (1985), Joyce (1990), Kingma and Eppard (1992), Chressanthis and Chressanthis (1993a, 1993b, 1994a, 1994b), Noll and Steinmueller (1992), Petersen (1989,1990), Quandt (1996), and Thatcher (1992). [BACK]

2. A small number of additional papers, by authors not involved in Foundation projects, was commissioned on topics deemed particularly relevant for electronic and digital libraries. [BACK]

3. For more detailed discussion of these issues, see Blinder (1997) and Blinder and Quandt (1997). [BACK]

4. As a parallel, note that investment in computers and related equipment accounted for less than 10% of gross investment in the U.S. economy in 1996. For a comprehensive account of the economics of the "computer revolution," see Sichel (1997). [BACK]

5. For example, digitizing in a myriad different ways: see Anne R. Kenney's remarkable benchmarks in chapter 2. [BACK]

6. Technically speaking, the sellers of such products are monopolistic competitors. [BACK]

7. But price discrimination emerging from the difficulty of reselling cannot be the whole story, because it seems largely confined to journals and not to monographs. [BACK]

8. Louisiana State University, among others, has replaced journal subscriptions with a liberal document delivery service, saving hundreds of thousands of dollars annually. [BACK]

9. Bakos and Brynjolfsson (1997) argue that under certain simplifying assumptions, the latter strategy permits the vendor to realize substantial profit while at the same time inflicting only modest deadweight loss on the purchaser. [BACK]

10. The question of the quality of presentation versus the cost is one of the important trade-offs that the academic profession may have to face in the future. See Andrew Odlyzko (chapter 23). [BACK]

11. By "interrelated" we mean that some of these questions cannot be answered in isolation: the answer to one may well depend on the answer to another one. [BACK]

12. See, for example, Chressanthis and Chressanthis (1993a, 1993b, 1994a, 1994b) and Petersen (1989, 1990). For a notable exception that does deal with the issue of which journal prices are increasing, see Noll and Steinmueller (1992). [BACK]

13. Analogies are the Netscape HTML Editor or the Microsoft FrontPage Editor that permit one to write HTML code practically without knowing anything about HTML. [BACK]

14. It is a common perception that the amount of time spent on software installation and on the maintenance of computer and software integrity is substantially larger today than it was in the ''good old days" of DOS, a mere five years ago. [BACK]

15. The vastly different outcomes that may occur when different search algorithms are applied to the same database is well illustrated in Besser and Yamashita (1997). [BACK]

1. Ian Graham's HTML Sourcebook: A Complete Guide to HTML 3.o, 2d ed. (Wiley, 1996), especially the beginning of chapter 3, gives an excellent overview of the characteristics of a book in the context of a discussion of the design of electronic resources. The third edition of this book was published early in 1997. [BACK]

2. Jay David Bolter's Writing Spaces: The Computer, Hypertext, and the History of Writing (Erlbaum, 1991) expands on some of these ideas. See also George Landow, Hypertext: The Convergence of Contemporary Critical Theory and Technology (Johns Hopkins, 1992) and my own Knowledge Representation, a paper commissioned as part of the Getty Art History Information Program (now the Getty Information Institute) Research Agenda for Humanities Computing, published in Research Agenda for Networked Cultural Heritage (Getty Information Institute, 1996), 31-34, and also available at http://www.ahip.getty.edu/agenda/represen.html. [BACK]

3. These terms, among others, have been used by the Model Editions Partnership ( http://mep.cla.sc.edu ). [BACK]

4. This was the planning meeting for the Text Encoding Initiative project. It was held in November 1987. [BACK]

5. C.J. Date, An Introduction to Database Systems, 4th ed. (Addison Wesley, 1986), is a good introduction to relational database technology. [BACK]

6. By far the most useful starting point for information about SGML is the very comprehensive Web site at http://www.oasis-open.org/cover/. This site is maintained and updated regularly by Robin Cover of the Summer Institute for Linguistics. [BACK]

7. The TEI's Web site is at http://www.uic.edu/orgs/tei. It contains links to electronic versions of the TEI Guidelines and DTDs as well as to projects that are using the DTD. [BACK]

8. See Richard Giordano, "The Documentation of Electronic Texts Using Text Encoding Initiative Headers: An Introduction," Library Resources and Technical Services 38 (1994): 389ff, for a detailed discussion of the header from the perspective of someone who is both a librarian and a computer scientist. [BACK]

9. More information about the EAD can be found at http://lcweb.loc.gov/ead. This site has examples of the Library of Congress EAD projects. Others can be found via links from the SGML Web site. [BACK]

10. This example can be seen at http://www.ceth.rutgers.edu/projects/griffis/project.htm. The site also provides instructions for downloading the Panorama SGML viewer. [BACK]

11. See Yuri Rubinsky, "Electronic Texts the Day After Tomorrow," in Visions and Opportunities in Electronic Publishing: Proceedings of the Second Symposium, December 5-8, 1992, ed. Ann Okerson (Association for Research Libraries, 1993), 5-13, also available at http://arl.cni.org: 80/scomm/symp2/rubinsky.html. Rubinsky was the founder of SoftQuad and a leading figure in the SGML community until his tragic early death in January 1996. [BACK]

12. There is a useful set of Frequently Asked Questions (FAQ) on XML at http://www.ucc.ie/xml/. See also the XML section of the SGML Web site at http://www.oasis-open . .org/cover/related.html. [BACK]

13. See note 3. [BACK]

14. The Orlando Project's Web site is at http://www.ualberta.ca/ORLANDO. [BACK]

15. For a more detailed and very useful discussion of these issues, see Liora Alschuler, ABCD ... SGML: A User's Guide to Structured Information (International Thompson Computer Press, 1995), especially chapters 5 through 9. Alschuler shares my view that it is very difficult to make general statements about the value of SGML-based projects since SGML can mean so many different things. Chapter 4 of her book consists of a series of case studies. It begins with a caveat: "anyone looking for a picture of a typical SGML implementation, look elsewhere," and then goes on: "These case studies represent the richness and diversity of real-world implementations, not the mythical norm." Elsewhere Alschuler notes that "ultimately the most profound impact of converting to structured information may be on the products you produce rather than on the methods you use to produce them" (186). [BACK]

16. In order to deal with the problem of overlap, the Wittgenstein Archives at the University of Bergen ( http://www.hit.uib.no/wab ) have devised their own encoding scheme, MECS (Multi-Element Code System). MECS contains some of the properties of SGML, but has simpler mechanisms for structures that are cumbersome in SGML. However, the use of their own encoding scheme has meant that they have had to develop their own software to process the material. [BACK]

17. For a longer discussion of new questions posed by the use of SGML and especially its perceived lack of semantics, see C. M. Sperberg-McQueen's closing address to the SGML92 conference at http://www.oasis-open.org/cover/sgml92p.html. He notes: "In identifying some areas as promising new results, and inviting more work, there is always the danger of shifting from 'inviting more work' to 'needing more work' and giving the impression of dissatisfaction with the work that has been accomplished. I want to avoid giving that impression, because it is not true, so I want to make very clear: the questions I am posing are not criticisms of SGML. On the contrary, they are its children.... SGML has created the environment within which these problems can be posed for the first time, and I think part of its accomplishment is that by solving one set of problems, it has exposed a whole new set of problems.'' [BACK]

1. Stephen Chapman and Anne R. Kenney, "Digital Conversion of Library Research Materials: A Case for Full Informational Capture," D-Lib Magazine (October 1996). [BACK]

2. Currently, scanning is the most cost-effective means to create digital files, and digital imaging is the only electronic format that can accurately render the information, page layout, and presentation of source documents, including text, graphics, and evidence of age and use. By producing digital images, you can create an authentic representation of the original at minimal cost and then derive the most useful version and format (e.g., marked-up text) for transmission and use. [BACK]

3. Michael Lesk, Image Formats for Preservation and Access: A Report of the Technology Assessment Advisory Committee to the Commission on Preservation and Access, Commission on Preservation and Access, Washington, DC, July 1990. [BACK]

4. See Charles S. Rhyne, Computer Images for Research, Teaching, and Publication in Art History and Related Disciplines, Commission on Preservation and Access, Washington, DC, January 1996, p. 4, in which he argues that "with each jump in [on-screen image] quality, new uses become possible." [BACK]

5. Interesting work is being conducted at Xerox PARC on image summarization. See Francine R. Chen and Dan S. Bloomberg, "Extraction of Thematically Relevant Text from Images," paper presented at the fifth annual Symposium on Document Analysis and Information Retrieval, Las Vegas, April 15-17, 1996. [BACK]

6. An interesting conclusion from a project on the use of art and architectural images at Cornell focused on image size guidelines to support a range of user activities. For browsing, the project staff found that images must be large enough for the user to identify the image, but small enough to allow numerous images to be viewed simultaneously-the physical size on the screen preferred by users was 1.25" to 2.25" square. For viewing images in their entirety, images were sized to fit within a 5.5" square; for studying, detailed views covering the entire screen were necessary; and for "authoring" presentations or other multimedia projects, users preferred images that fit in a half-inch square. See Noni Korf Vidal, Thomas Hickerson, and Geri Gay, "Developing Multimedia Collection and Access Tools, Appendix V. Guidelines for the Display of Images'' (report delivered to the Council on Library Resources, April 1996), 14-17. [BACK]

7. A number of leading experts advocate this approach, including Michael Ester of Luna Imaging, Inc. See, for example, Michael Ester, "Digital Images in the Context of Visual Collectons and Scholarship," Visual Resources 10 (1990): 11-24; and "Specifics of Imaging Practice," Archives and Museum Informatics: Hands-on Hypermedia and Interactivity in Museums, Selected Papers from the Third International Conference, San Diego, CA, October 9-13, 1995. [BACK]

8. Roger S. Bagnall, Digital Imaging of Papyri: A Report to the Commission on Preservation and Access, Commission on Preservation and Access, Washington, DC, September 1995; Janet Gertz, Oversize Color Images Project, 1994-1995 Final Report of Phase I, Commission on Preservation and Access, Washington, DC, August 1995; Picture Elements, Inc., Guidelines for Electronic Preservation of Visual Materials, Part I, Report to the Library of Congress, March 2, 1995. Michael Ester argues that an "archival image" of a photograph cannot be benchmarked through calculations but should be pegged to the "functional range of an institution's reproduction sources." See page 11 in Ester, Digital Image Collections: Issues and Practice, Commission on Preservation and Access, Washington, DC, December 1996. For a critique of this approach, see Chapman and Kenney, "Digital Conversion." [BACK]

9. Don Williams, "What Is an MTF and Why Should We Care?" RLG Diginews 2(1) (February 15, 1998), http://www.rlg.org/preserv/diginews. [BACK]

10. Anne R. Kenney and Stephen Chapman, "Film Scanning," Chapter 7 in Digital Imaging for Libraries and Archives, Cornell University Library, Ithaca, NY, 1996, p. 169. [BACK]

11. ANSI/AIIM MS 23 -1991, Practice for Operational Procedures/Inspection and Quality Control of First-Generaton, Silver Microfilm and Documents, Association for Information and Image Management; ANSI/AIIM TR26-1993, Resolution as It Relates to Photographic and Electronic Imaging, Association for Information and Image Management; and Kenney and Chapman, Tutorial: Digital Resolution Requirements for Replacing Text-Based Material: Methods for Benchmarking Image Quality, Commission on Preservation and Access, Washington, DC, April 1995. [BACK]

12. For a description of this verification process, see Anne R. Kenney, "Digital-toMicrofilm Conversion: An Interim Preservation Solution," Library Resources and Technical Services (October 1993): 380-401; (January 1994): 87-95. [BACK]

13. Anne R. Kenney and Oya Y. Rieger, Using Kodak Photo CD Technology for Preservation and Access, Cornell University Library, Ithaca, NY, 1998. [BACK]

14. A fuller explanation of the display benchmarking process is included in Anne R. Kenney and Stephen Chapman, Chapter 2 in Digital Imaging for Libraries and Archives, Cornell University Library, Ithaca, NY, 1996, pp. 76-86. [BACK]

15. Improvements in managing color digitally may be forthcoming from an international consortium of industry leaders working to develop an electronic prepress industry standard. Their "International Color Consortium Profile Format" is intended to represent color consistently across devices and platforms. [BACK]

16. See Peter van Minnen, "Imaging the Duke Papyri," (December 1995) http:// odyssey.lib.duke.edu/papyrus/texts /imaging.html, and Roger S. Bagnall, Digital Imaging of Papyri: A Report to the Commission on Preservation and Access, Commission on Preservation and Access, Washington, DC, September 1995. [BACK]

17. Rhyne, Computer Images, 5. [BACK]

18. The formula for calculating the maximum percentage of a digital image that can be displayed on-screen is as follows:

a. If both image dimensions < the corresponding pixel dimensions (pd) of the screen, 100% of the image will be displayed.

b. If both image dimensions > the corresponding pixel dimensions of the screen, % displayed = horiz. screen pd ³ vertical screen pd ³ 100 image's horiz. pd ³ image's vertical pd.

c. If one of the image's dimensions < the corresponding pixel dimension of the screen, % displayed = image's opposite pixel dimension ³ 100. [BACK]

19. The scaling formula for complete display of image on screen is as follows:

a. When digital image aspect ratio < screen aspect ratio, set image's horizontal pixel dimension to the screen's horizontal pixel dimension.

b. When digital image aspect ratio > screen aspect ratio, set image's vertical pixel dimension to the screen's vertical pixel dimension. [BACK]

20. This formula presumes that bitonal images are presented with a minimum level of gray (3 bits or greater) and that filters and optimized scaling routines are used to improve image presentation. [BACK]

1. Clifford Lynch in "Technology and Its Implications for Serials Acquisitions," Against the Grain 9, no. 1 (1997): 31. This article is based on a talk by Lynch at the November 1996 Charleston Conference. He identifies the key needs in building digital libraries as authentication, printing, individual item addressability, accessibility, and linkage. Lynch concludes with this insight: "The theme I want to underscore here is that we need to be very careful about whether we have technology that can deliver this electronic content for which we are busy negotiating financial arrangements in acceptable ways on a broad systemic basis" [emphasis is mine]. [BACK]

2. The position statement "Fair Use in the Electronic Age: Serving the Public Interest" is an outgrowth of discussions among a number of library associations regarding intellectual property and, in particular, the concern that the interests and rights of copyright owners and users remain balanced in the digital environment. This important position statement was developed by representatives of the following associations: American Association of Law Libraries, American Library Association, Association of Academic Health Sciences Library Directors, Association of Research Libraries, Medical Library Association, and Special Libraries Association. It espouses the philosophy that the U.S. copyright law was created to advance societal goals and well-being and embeds the notion of technological neutrality. It can be found at: http://www.arl.org/info/frn/copy/fairuse.html. [BACK]

3. I have recently had the opportunity to read statements from the international publishing community in two major position papers originating with the International Publishers Copyright Council, the STM group of publishers, and the International Publishers Association. These documents affirm the following:

· Digital versions of works are not the same as print versions, because digital information can be manipulated and widely distributed. (The implication is that manipulation and distribution will happen and that it is happening with copyrighted works, often in an illegal manner.)

· Digital versions of works need even more protection than printed versions.

· Digital browsing is not the same as reading print: the very act of browsing involves reproducing copies (which immediately implicates and possibly violates copyright law).

· There should be no private or personal exemptions from copyright in the digital environment.

· There should be no exceptional copyright treatment for libraries in the digital environment-the exemptions for traditional materials, if carried over into the digital environment, will result in unfair competition with publishers.

· Digital lending (a digital analog to ILL) will destroy publishers.

· Publishers are now poised to offer and charge for electronic delivery of information and therefore they ought to be able to. Such services will replace most of the copying that libraries and individuals used to do in print.

· The role of libraries will be to provide access, select materials for users via what they choose to license, instruct users in the vast array of electronic sources and how to use them, and support users in searching and research and learning needs. [BACK]

4. See the 30 June 1996 decision by the United States Court of Appeals, Seventh Circuit, in ProCD v. Zeidenberg. The question posed was: must buyers of computer software obey the terms of shrink-wrap licenses? The district court had said not. The court of appeals reversed this decision. ProCD (the plaintiffs) compiled information from more than 3,000 phone directories into one database with additional information such as zip code extensions and included their own searching software. They packaged the product, called SelectPhone, as a CD for personal sale in a shrink-wrap box. They also sold it to commercial companies in other formats, such as mailing lists. Mr. Zeidenberg bought SelectPhone at a shop in Madison, Wis., and formed a company to resell the information on the basis that factual information cannot be copyrighted. He made this information available over the Internet apparently quite cheaply. Zeidenberg argued that a person cannot be bound by the shrink-wrap license because the terms are not known at the time of purchase. They are inside the package and the purchaser cannot be bound by terms that are secret at the time of purchase. The judges' decision was that the shrink-wrap license is legal and that a buyer is bound by it. [BACK]

5. See Martha Kellogg, "CD-ROM Products as Serials: Cost Considerations for Libraries," in Serials Review 17, no. 3 (1991): 49-60. The tables in this article, as a basis of comparison between print reference or indexing and abstracting works and their CD equivalents, show a difference of about 30% where resources are comparable. Recent e-mail from the University of Michigan Library suggests that differentials between print and electronic works are as high as 600%. [BACK]

6. Newfour is a joint project of the Yale Library, the University of Pennsylvania, and the University of California, San Diego Library. Its fully searchable archive is located at http://gort.ucsd.edu/newjour/. [BACK]

7. A good summary of the flavor, debates, and progress of CONFU can be found at http:/ / www.utsystem.edu/OGC/IntellectualProperty/confu.htm. The CONFU interim report is available at http://www.uspto.gov/web /offices/dcom/olia/confu/. [BACK]

8. For a list of the consortia that participated in the St. Louis meeting and descriptions of their activities, see the COC home page at Yale University Library: http:// www.library.yale.edu/consortia/. [BACK]

9. Ann Okerson, "Buy or Lease? Two Models for Scholarly Information at the End (or the Beginning) of an Era," Daedalus 125, no. 4 (1996): 55-76. This special issue on libraries is called "Books, Bricks, and Bytes." I suggest that one possible outcome of the new trend to scaled-up consortial licensing activities is that the library marketplace will gain significant power and that publishers of scholarly information could find themselves in quite a different position than they are in the captive marketplace of today. It is possible to argue that such an outcome is very healthy; on the other hand, even librarians and scholars might find this outcome undesirable in that it would put today's specialized scholarly publications, with their attendant high prices, out of business. The Daedalus piece can also be found at http://www.library.yale.edu/~okerson/daedalus.html. [BACK]

10. For example, an especially rich resource is the site of the University of Texas Office of General Counsel's Copyright Management Center. The center provides guidance and information to faculty, staff, and students concerning applicable law and the alternatives available to help accomplish educational objectives. A large number of materials, organized by topic, is accessible through this Web site. Some important documents are stored directly on the Web server. The principal author is Georgia Harper, copyright counsel for the University of Texas System. The URL is http://www.utsystem.edu/ogc/intellectualproperty/cprtindx.htm.

The higher education community is indebted to, among others, Indiana University's Kenneth Crews, an important voice in CONFU (see, for example, the CETUS Fair Use document at http://www.cetus.org/fairindex.html ); the University of North Carolina Law School's Lolly Gasaway, also a leader in CONFU and contributor of many important resources (see, for example, "When Works Pass Into the Public Domain" at http://www.library.yale.edu/~okerson/pubdomain.html ); and Karen Hersey of the MIT Counsel's Office, a leader in crafting university-producer electronic license agreements and a frequent workshop presenter on this topic. [BACK]

11. See "Licensing Electronic Resources: Strategic and Practical Considerations for Signing Electronic Information Delivery Agreements" at http://www.arl.org/scomm/licensing/licbooklet.html. [BACK]

12. See LIBLICENSE: Licensing Digital Information-A Resource for Librarians. This Web resource contains license vocabulary, licensing terms and descriptions, sample publishers' licenses, links to other licensing sites, and a bibliography about the subject. The URL is http://www.library.yale.edu/ ~llicense/index.shtml. [BACK]

13. LIBLICENSE-L is a moderated list for the discussion of issues related to the licensing of digital information by academic and research libraries. To join the LIBLICENSE-L list, send a message to listproc@lists.yale.edu. Leave the subject line blank. In the body of the message, type: subscribe LIBLICENSE-L Firstname Lastname [BACK]

14. A LIBLICENSE-L message of February 12, 1997, enumerated a dozen different pricing models for electronic resources, and correspondents added several more in subsequent discussion. [BACK]

15. Several reasons are advanced for the higher cost of electronic resources versus comparable print resources: (1) the producers are making new R&D and technology investments whose significant prices are passed on to the customer; (2) producers of journals generally offer a package that includes print plus electronic versions, giving the customer two different forms of the same information rather than one only; (3) the functionality of electronic resources is arguably higher than that of the print version; (4) electronic resources are not marketed as single journals or books but as scaled-up collections, often of substantial heft (consider the corpora of humanities full texts marketed by Chadwyck-Healey, the large backfile collections of JSTOR, the full collection of Academic Press titles available under its IDEAL program; it seems that there is little incentive for producers to create and sell one electronic item at a time); and (5) the publisher, in becoming the source or site or provider, is taking on many of the library's roles and costs. [BACK]

16. A LIBLICENSE-L message of March 14,1997, defined aggregators in the following way: 'Aggregation' as used on this list means the bundling together or gathering together of electronic information into electronic collections that are marketed as a package. For example, DIALOG@CARL aggregates 300 databases; Academic Press's IDEAL aggregates 170+ journals; Johns Hopkins's Project MUSE is an electronic collection of 40+ journals, and so on. But the term 'aggregator' is more usually used in describing the supplier who assembles the offerings of more than one publisher, so one is more likely to hear Dialog, OCLC, Information Access, and UMI spoken of as aggregators, than The Johns Hopkins University Press." [BACK]

17. License negotiations between libraries and producers now do take into account the matter of electronic archiving, or at least the parties pay lip service to perpetual access. For example, it is common for an electronic resource license to offer some form of access or data if the library cancels a license or if the provider goes out of business. However, while the license addresses this matter, the underlying solutions are far from satisfactory for either party. I leave the matter of archiving, a huge topic and concern, to other venues; clearly the whole underpinnings of libraries and culture are at stake, depending on the outcomes of the archiving dialogues that are in place now and will surely outlast our lifetimes. [BACK]

18. At Yale, for example, after close discussions on this matter with the library to make sure that points of view were in synch, general counsel delegated library content licensing to senior library administration and it is now done by the associate university librarian for collections with considerable support and backstopping by Yale's public services and collections librarians in effective and productive teamwork. [BACK]

19. In fact, the software development was funded by CLIR (formerly the Council on Library Resources) in June 1997, and its product is now available at http://www.library.yale.edu/ ~ llicense/software.shtml). [BACK]

20. The case, Princeton University Press v. Michigan Document Services, Inc., asked the question: does a copy shop infringe on publishers' copyrights when it photocopies course pack materials? This material comprises book chapters and articles for students of nearby colleges and universities. The owner of Michigan Document Services argued that he was copying on behalf of the students and exercising their fair use rights. The recent appeal in the United States Court of Appeals for the Sixth Circuit found for the publishers. For extensive documentation on this matter, see Stanford's Fair Use site at http://fairuse.stanford.edu/mds/. [BACK]

21. For the journals available through Stanford's HighWire, see http:/ /highwire.stanford.edu. [BACK]

1. See, for instance, Tefko Saracevic and Paul Kantor, "Studying the Value of Library and Information Services," Journal of the American Society for Information Science 48 (June 1997): 527-563. The terms used by readers to describe values attributed to the use of libraries include "convenience" and closely related concepts, such as "effort required" and "frustration.'' Saracevic and Kantor observe that users rarely assign monetary value to library services but often refer to their time saved. [BACK]

2. "How Libraries Can Help to Pay Their Way in the Future," Logos 7, no. 3(1996): 238. See also Bowen's paper, "JSTOR and the Economics of Scholarly Communication," presented at the Council on Library Resources conference held in October 1995 (available at http://www.mellon.org/jstor.html ). [BACK]

3. The Andrew W. Mellon Foundation Report, from January 1, 1994 through December 31, 1994 (New York, 1995), 19; The Andrew W.Mellon Foundation Report, from January 1, 1995 through December 31, 1995 (New York, 1996), 48. [BACK]

4. A good general account of productivity is provided by the National Research Council, Panel to Review Productivity Statistics, Measurement and Interpretation of Productivity (Washington, D.C., 1979). [BACK]

5. The productivity gains expected of information technology have sometimes been slow to appear, creating the so-called productivity paradox. Federal Reserve Board economist William Wascher argues we have had unrealistic expectations of information technology, which accounted for less than 8% of equipment expenditure in 1994 and was a "relatively minor input in the production process" (BNA Daily Report for Executives, 10 September 1996, p. C3). [BACK]

6. For a fuller account of the difficulties, see Morton Owen Schapiro, "The Concept of Productivity as Applied to U.S. Higher Education," in Paying the Piper: Productivity, Incentives, and Financing in U.S. Higher Education, by Michael S. McPherson, Morton Owen Schapiro, and Gordon C. Winston (Ann Arbor, 1993), 37-68. [BACK]

7. Libraries are no strangers to concerns about productivity. Indeed, the successful application of information technology to library cataloging is one of the great if not widely acknowledged productivity success stories of higher education. At Yale, the largest library investments we are now making-ranging from preserving and shelving the collections to the retrospective conversion of catalog records and the on-line delivery of information-are fundamentally motivated by productivity issues. The productivity issues that inform decisions to build off-campus shelving are particularly interesting. They involve productivity trade-offs between traditionally impermeable operating and capital budgets and appear to put the productivity of library operations in competition with reader productivity needs. [BACK]

8. See, for instance, Frank Guliuzza III, "Asking Professor Jones to Fix the Crisis in Higher Education is Getting More and More Expensive," Academe (September/October 1996): 29-32. [BACK]

9. See, for instance, Robert Zemsky and William F. Massy, "Expanding Perimeters, Melting Cores, and Sticky Functions: Toward an Understanding of Our Current Predicaments," Change (November/December 1995): 41-49; William F. Massy, Resource Allocation in Higher Education (Ann Arbor, 1996); and Charles T. Clotfelter, Buying the Best: Cost Escalation in Elite Higher Education (Princeton, 1996). [BACK]

10. Arthur Levine, "Higher Education's New Status as a Mature Industry," Chronicle of Higher Education, 31 January 1997, A48. [BACK]

11. This description of the Scully Project and parts of my commentary on it draw heavily on Max Marmor's Preliminary Report on the project of 6 September 1996. I am much indebted to Mr. Marmor, Christine de Vallet, and Donald J. Waters for their collegial and critical review of the ideas advanced in this paper and to Elizabeth Owen for her neverfailing and generous willingness to describe the practices of the Teaching Fellows in Professor Scully's course. [BACK]

12. To achieve these results, the project managers enlisted the help of a senior administrator and two Webmasters at Yale's Information Technology Services, a graphic designer from University Printing Service, and student assistants for HTML markup and image manipulation. For copyright reasons, access to this Web site is restricted to members of the Yale community. [BACK]

13. Mistakes in the administration of a survey left us without statistically meaningful measures of student use of the Web site or opinions about it. [BACK]

14. Some, but not all, of the Teaching Fellows made use of the Web site in preparing for their classes and in making assignments. The e-mail addresses for Teaching Fellows provided at the Web site encouraged some exchanges with instructors outside of class sessions. [BACK]

15. Dr. Carl Jaffe is director of the Center for Advanced Instructional Media at the Yale School of Medicine. The center was founded in 1987 to explore the educational and communications potential of new interactive multimedia computing technology. Through the center, the Yale University School of Medicine is one of the leading developers and publishers of multimedia medical education programs. The center is a recognized leader in information design, medical illustration, and interface design for networked information. Center projects have won many national and international awards for excellence in educational design and technological leadership. More information about the center is available on the Web at http://info.med.yale.edu/caim/. [BACK]

16. Absent from the cost model is any consideration of copyright management costs. This is a significant and unsettled matter in the use of digital images for instructional support. The presence or absence of such costs in a full-scale digital image database would have a significant impact on the bottom-line results of the cost model. [BACK]

17. However, estimating the cost of maintaining the Scully Project data over time does deserve some discussion. The primary experience of universities in doing this estimating lies in maintaining on-line catalog data and-more recently-in maintaining institutional business data as we adopt new administrative software to solve the Year 2000 and other problems. We have learned that it is very expensive to maintain bibliographic and business data over time as we move from one computing environment to another. Institutions generally do not budget for these costs. Recognizing that past practice is an unacceptable model for the preservation of library materials, Yale's Open Book Project endeavored to model the cost of reliably preserving large bodies of full-text library material in digital form over long periods of time. The Open Book cost model, applied to the Scully Project, suggests that the cost of maintaining a small Web site is $470 a year or $2,049 for six years. Focused attention to preservation is required because digital media are inherently the most unstable of information storage media, and hardware and software have high obsolescence rates. If we do not heed Preserving Digital Information: Report of the Task Force on Archiving of Digital Information, with its sobering account of the technical difficulties and significant costs involved in the long-term preservation of digital information, we will find the wonders of our silicon-based technologies to be houses built on sand. The Report was commissioned by the Commission on Preservation and Access and the Research Libraries Group, Inc., and published by the commission in May 1996; its work was led by Donald J. Waters and John Garrett. [BACK]

18. This comparison assumes these are alternative means of support rather than duplicative. In fact, both means of support were provided in the fall 1996 term, and students made use of both. [BACK]

19. That is, the cost model shows a -$703 balance over 16 years, or essentially breakeven performance on 16-year expenditures of $59,000. Of course, it is quite unlikely that the facts, estimates, and assumptions used in the cost model would remain static for as long as 16 years. [BACK]

20. Teaching Fellows conduct most of their discussion sessions in the Yale University Art Gallery, to relate Professor Scully's lectures to actual museum objects. They conduct conventional classroom discussions, supported by slides, only three times each during the term. For the sake of simplicity, the cost model deals only with these conventional classroom sessions and does not try to model the cost of bringing digital images to locations throughout the gallery. Teaching Fellows often bring photographs to the gallery to enrich the discussions, so access to digital images there would actually be quite valuable. [BACK]

21. See National Research Council, Measurement and Interpretation of Productivity, 33-34. [BACK]

22. See, for instance, Kenneth H. Ashworth, "Virtual Universities Could Produce Only Virtual Learning," Chronicle of Higher Education, 6 September 1996, A88. [BACK]

23. Another "what if" scenario, not developed here, for increasing the productivity of the investment in digital images is to imagine Professor Scully's Introduction to the History of Art being taught on the Web, with a large increase in enrollments beyond the 500 that the lecture hall now permits. This distance learning scenario raises a host of questions, not least that of how the rights to the image content of such a course would be secured and managed. [BACK]

24. This finding is most tentative, given that the cost model is sensitive to the highly variable costs of equipping a classroom for digitally supported instruction. [BACK]

25. This lack of attention is evident in the Chronicle of Higher Education account of Professor Stephen Murray's use of video media to teach architecture at Columbia University (see "Video Technology Transforms the Teaching of Art History," Chronicle of Higher Education, 14 February 1997, A20-22). Having described Murray's remarkable video presentation of Amiens Cathedral that cost some hundreds of thousands of dollars, the Chronicle reporter Lisa Guernsey describes Murray as insisting "the projects are worth it. 'What really convinces us is how they inspire students in the classroom,' [Murray] says. 'They are captivated to study further'" (A22). [BACK]

26. Scott W. Blasdell, Michael S. McPherson, and Morton Owen Schapiro, "Trends in Revenues and Expenditures in U.S. Higher Education: Where Does the Money Come From? Where Does It Go?" in McPherson, Schapiro, and Winston, Paying the Piper, 17. A recent General Accounting Office report found that tuition at four-year public colleges has risen nearly three times as much as median household income over the past 15 years; see Chronicle of Higher Education, 6 September 1996, A59. [BACK]

27. And in observing comparative measures of productivity gain in libraries and other operations of the university, we will also have created a powerful tool for helping to decide among competing technology investment strategies. [BACK]

28. See "Why Colleges Cost Too Much," a so-called investigative report on higher education costs (as exemplified by those at the University of Pennsylvania) by Erik Larson in Time, 17 March 1997. [BACK]

29. The Western Governors' University has been widely reported over the past two years. See, for instance, the account in the New York Times, 25 September 1996, B9. The Chronicle of Higher Education for 27 September 1996, A35-36, reports a similar initiative among twelve Scandinavian institutions. [BACK]

30. New York Times, 25 Sept. 1996, B9. [BACK]

31. See, for instance, Theodore Marmor and Mark Goldberg, "American Health Care Reform: Separating Sense from Nonsense," in Understanding Health Care Reform by Theodore Marmor (New Haven, 1994), 15-18. [BACK]

32. Richard Lanham, "The Electronic Word: Literary Study and the Digital Revolution," in Richard Lanham, The Electronic Word: Democracy, Technology, and the Arts (Chicago, 1993), 23. [BACK]

1. Shirley Baker, talk at Washington University, November, 1996. [BACK]

2. Robin Frost, "The Electronic Gutenberg Fails to Win Mass Appeal," Wall Street Journal, 21 November 1996, B6. Project Gutenberg was a 25-year effort led by Michael S. Hart at the University of Illinois to create, store, and make accessible ASCII files of public domain materials from the Constitution, the Bible, Shakespeare, and beyond. [BACK]

3. A large part of the Project Gutenberg files were moved to the library at the University of Maryland after this essay was written. See http://www.inform.umd.edu/EdRes/ReadingRoom/. [BACK]

4. Stephen Burd, "President Pushes Tax Breaks to Help Families Afford College," Chronicle of Higher Education, 17 January 1997, A33. [BACK]

5. www.ei.org [BACK]

6. http://xxx.lanl.gov/ [BACK]

7. http://scriptorium.lib.duke.edu/papyrus/ offers 1,373 images of Egyptian papyri with a significant database of descriptive textual material. [BACK]

8. http://fairmodel.econ.yale.edu/ [BACK]

9. http://gdbwww.gdb.org/ [BACK]

10. The headquarters publishes Job Openings in Economics (JOE) seven times a year with nearly 1,500 job announcements. In 1995, JOE had about 4,000 subscribers and generated about $41,000 of revenue with a base rate of $15.00 per year ($7.50 for students, $25.00 for nonmembers and institutions). The sum of monthly printing and mailing costs was associated with the number of copies produced and the number of pages per copy for 1995 and 1996 as follows (with t-ratios in parenthesis):

print & mail = -1,129.57 + 0.875 # of copies + 76.725 pages per issue                (-2.83)    (7.35)              (17.2)

This relationship is estimated from data on each of 14 issues over the two years and has an adjusted R-square of 0.957. During this period of time, JOE averaged 25 pages per issue (the range was from 11 to 51). With seven issues per year, this equation forecasts total printing and mailing costs of $30,019 for 4,000 copies.

JOE became available without charge in 1994 ( http://www.eco.utexas.edu/joe/ ). The JOE site generated about 25,000 hits per month in 1996, and the subscription list of the printed JOE dropped to 1,000. The print and mail relationship estimated above forecasts a cost of $11,645 for 1,000 copies. The association will move from a net revenue position of $11,000 ($41,000 - $30,019) in the all-print regime to about a zero net ($15,000 - $11,645) with print subscription sales at about 1,000. Of course, the association incurs fixed costs in producing JOE that may be similar under both regimes.

The headquarters also publishes a directory of membership biennially. The directory became available on-line at the University of Texas in 1995 and is getting about 4,600 hits per month. Because the directory comes with membership, we have no measure of the rate of decline in the demand for the print version. [BACK]

11. At some point in the future, membership ballots might be solicited and received by the Internet. [BACK]

12. The AER's reviewing process is double-blind, with author's names withheld from reviewers and reviewer's names kept from authors. When nearly all working papers are posted on the World Wide Web, the refereeing may become single-blind de facto. Anyone who wants might search the title listing in the working paper file and so identify the author. When working papers are generally accessible on the Net, they would seem to be usable in the editorial process with some saving in cost but with some loss in anonymity. [BACK]

13. The fixed costs of a print run (but not typography) would be eliminated entirely if print were abandoned completely. The fixed costs of electronic distribution would replace them in part. Presumably, the more sophisticated the electronic files submitted by authors, the lower the fixed cost of production at the publisher. [BACK]

14. Since 1995, the association has made the JEL available in CD-ROM format instead of print for the same price. The CD-ROM costs about the same to produce on the margin per subscriber as a printed issue of a large journal. The CD-ROM contains the page images of the published journal and is distributed by mail. Its advantage is not reduced cost but increased subscriber benefit: it adds the power of electronic searching. Therefore, this version is gaining popularity. More than 10% of the AEA's members opted for the CD-ROM version of JEL in 1996. [BACK]

15. The annual meeting contributed a net of about $125,000 in 1995. [BACK]

16. Assume the current library subscription rate of $140 from 5,500 subscribers yields 20% of the AEA's gross and that membership plus ads from 21,000 members yields $70, about 40%. Assume the shift to electronic distribution lowers total expenditures by 20%. Doubling the library rate to $280, if all libraries subscribed, would keep AEA revenues constant. [BACK]

17. The notion of doubling the library subscription rate in setting a rate for the campus intranet license is meant to define the association's probable revenue goals, but not to define the rate structure. The rate structure will need to be tied to something more substantial, like enrollment and total research dollars. Alternatively, the rate could be set on the basis of a forecast of the hit rate. OCLC's electronic journal service sets rates on the basis of the number of simultaneous users. The level of rates would likely be set so as to yield about double the current library print subscriptions unless other revenue is forthcoming as discussed in subsequent paragraphs. [BACK]

18. Here is part of the language the AEA prints on the copyright page: "Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or direct commercial advantage and that copies show this notice on the first page or initial screen of a display along with the full citation, including the name of the author." [BACK]

19. Jared Sandberg ("Cash Advances Aid Electronic Commerce," Wall Street Journal, 30 September 1996, B8) reports an offering from CyberCash, a firm working with Visa and several banks. Cybercash put the cost of a transaction between $.08 and $.31 for purchases between $.25 and $10.

http://www.millicent.digital.com describes the protocols and tools developed by Digital Equipment Corporation to facilitate Web transactions in fractions of cents. "The key innovations of Millicent are its use of brokers and of scrip. Brokers take care of account management, billing, connection maintenance, and establishing accounts with vendors. Scrip is microcurrency that is only valid within the Millicent-enabled world." [BACK]

20. Draft essay at ftp://alfred.sims.berkeley.edu/pub/Papers/dlib.html. [BACK]

21. See Malcolm Getz, "Petabytes of Information," in Advances in Library Administration and Organization, XII (JAI Press, 1994), 203-37. Here are some features that might be added to the network working paper service: Each association member might receive a private password and encryption key. When the member submits a paper with the password and key, the service would return a time-stamped digital authentication message. This message and the posting would establish ownership to the working paper at the time of submission. The working paper service might include a more elaborate system of tagging papers, including the author's sense of the target audience, degree of originality, sophistication, empirical content, and revision number. The service might include links to comments. [BACK]

22. E. Jacquelin Dietz, "The Future of the Journal of Statistics Education," North Carolina State University, mimeo, 1996. [BACK]

23. The issue of optimal pricing for three products that share a fixed cost and in which cross elasticities are not zero should be explored formally. [BACK]

24. David Carpenter and Malcolm Getz, "Evaluation of Library Resources in the Field of Economics: A Case Study," Collection Management 20, no. 1/2 (1995), 49-89. [BACK]

25. See Malcolm Getz, "Information Storage," Encyclopedia of Library and Information Science 52, supplement 15 (1993), 201-39. High-density off-site storage might yield an annual cost of $.30 per volume and so about $3.00 of capital cost. [BACK]

26. OCLC's Electronic Journals Online (EJO) preceded the Web-based program. With EJO, OCLC charged publishers for mounting their journals, much like printers charge for printing. This approach did not attract many publishers. The OCLC Web site ( www.OCLC.org ) lists several titles. Here is a sample of subscription rates.

The Online Journal of Current Clinic Trials from Chapman & Hall, distributed by OCLC: Institutional, $220; Individual, $120; Student (with ID), $49; Network (unlimited access), $3,000.

Online Journal of Knowledge Synthesis for Nursing from Sigma Theta Tau International, distributed by OCLC: Individuals, $60; Institutions, $250. [BACK]

27. OCLC, "Bringing Your Publications Online With OCLC" (Dublin, Ohio, ca. 1996) and OCLC, "A Complete Electronic Journals Solution for Your Library" (Dublin, Ohio, ca. 1996). [BACK]

28. Malcolm Getz, John J. Siegfried, and Kathryn H. Anderson, "Adoption of Innovations in Higher Education," The Quarterly Review of Economics and Finance 37(3) (Fall 1997): 605-31. [BACK]

29. http://mitpress.mit.edu/journals.tcl. SNDE is one of six electronic journals offered by the MIT Press in 1996. The library rate includes a license to store the journal on a campus facility and make it available in library reserve services. The MIT Press puts the subscription rate at $30 for individuals and $125 for libraries, with a $12 fee for downloading an individual article. [BACK]

1. Anne Kenney is the associate director of preservation at the Cornell University Library. She has also contributed to this work; see chapter 2, ''Digital Image Quality: From Conversion to Presentation and Beyond." [BACK]

2. We did agree to work with three scholarly associations-the Ecological Society of America, the American Economic Association, and the American Political Science Association-to provide access to current issues through JSTOR. We stand by our commitments to these organizations, but our goal is to learn more about the technology required to make linkages between current issues and the archive, not to build the capability for JSTOR to become a publisher of current issues. [BACK]

3. In the not-for-profit context, a revenue-generating market need not consist solely of paying customers; it could include other types of indirect funders like government agencies or foundations. [BACK]

4. To refer to these costs as one-time costs is not precisely accurate. Not all of the production costs are one-time. We add another volume of each journal title to the database as each year passes, so there is an ongoing element of the production costs, but that element represents a small fraction of total production expenditures. [BACK]

5. There is a caveat here as well. Some of the administrative and overhead costs are higher because JSTOR is adding titles. Negotiating agreements with publishers is a timeconsuming task, as is overseeing the production operation converting 100,000 pages per month. It is not practical, however, to allocate exactly the portion of general administrative and other costs that pertain directly to production. [BACK]

6. For a more complete description of these estimates, see "JSTOR and the Economics of Scholarly Communication," a paper by William G. Bowen, which is available at http://www.mellon.org/jsesc.html. [BACK]

7. For a more complete description of the evolution in the development of JSTOR's library license terms, see Sarah E. Sully, "JSTOR: An IP Practitioner's Perspective," D-Lib, January 1997. [BACK]

1. Marie Hansen, ''Pricing Issues for Electronic Journals," unpublished, 1996. [BACK]

2. Piracy is a real threat. According to the Software Publishers Association, about $13 billion in sales was lost due to piracy in 1996. See the SPA home page against piracy: http://www.spa.org/piracy/homepage.htm. [BACK]

3. Adorno, Minima Moralia, translated by E. F. N. Jephcott (London: New Left Books, 1974), section 30. [BACK]

4. The Gazette (Johns Hopkins), February 24, 1997, 9. [BACK]

5. Skeptics of electronic communication are not convinced that digitization will lead to Utopia with a spell checker. See, for example, Kurt Andersen, "The Digital Bubble: Waking Up from the New-Media Pipe Dream," New Yorker, January 19, 1998, 30. [BACK]

6. An article in Upside forecast that Internet customer services could save businesses 25% to 50% of the cost of traditional telephonic customer support (David Kline, "Reshaping the Way America Does Business," Upside Online, August 5, 1996). But in 1998, mailing lists carry alerts that the telephone companies are seeking ways to add new charges for Internet use. [BACK]

7. The size of the Web market has grown as rapidly as the Web itself. In 1997 CyberAtlas estimated between 5.8 million and 35 million users. The CyberAtlas estimate in October 1998 is 100 million users. See http://www.cyberatlas.com/market/size/index.html. [BACK]

8. See, for instance, http://www.stud.ifi.uio.no/~vebjornk/bookmarks/aow.html http:// www.geocities.com/Athens/4884/; and http://www.kimsoft.com/polwar.htm. [BACK]

1. At the time of this study, the Department of Economics at the University of Michigan maintained an extensive departmental library with support from the central library. This departmental collection is no longer supported. [BACK]

2. This figure combines the 44% of the faculty who were unaware of JSTOR with the 23% of the faculty who were aware of JSTOR, but did not use it. [BACK]

1. BMMR has, as of May 7, 1997, become The Medieval Review (TMR). [BACK]

2. The output by month (4/95-3/97) is as follows:

3. Since May 7, 1997, BMMR (that is, TMR) has been on the Web, which will eventually provide valuable data to compare with the BMCR gopher data. [BACK]

4. Naturally, new reviews are visited often; we are trying to isolate those of enduring value. [BACK]

5. Likewise, the .mil domain dropped from 310 to 186; the .gov domain, from 819 to 409. [BACK]

6. The explosive growth in 9/95 and 10/95 was only temporary. [BACK]

7. The difference would be even more pronounced had I not excluded books that appeared on the list only once. In 1996 the gap virtually disappears: 31 medieval titles (total number of titles 53) made the list 126 times (avg. 4.1) while 93 classical titles (total number of titles 169) made the list 360 times (avg. 3.9). [BACK]

8. As is, our response rate is only in the 30-40% range. [BACK]

9. Unfortunately the survey was worded as if only for BMCR subscribers, but even so the response rate was about 35%. [BACK]

10. We found similar results in a pilot comparison of qualitative differences between the two journals that was done by two advanced graduate students (one a classicist, one a medievalist) in the summer of 1995. The students concluded that the major differences stem from the scholarly orientation of either discipline, not from their media (i.e., classicists criticize at a microscopic level, assuming in-depth acquaintance with a given text). The reviews are longer and the number of typographical errors is much greater in BMMR, but other differences seemed to be personal (tone of the review, footnotes and additional bibliography, organization, amount of direct quotation). [BACK]

11. The values given are those of the Pearson chi-square test, but we also ran Continuity Correction, Likelihood Ratio, Mantel-Haenszel test for linear association among the chi-square tests, and also Pearson's R and Spearman Correlation. [BACK]

12. Combined BMCR and joint figures are 912 for 1993, 1,342 for 1994 (+47%), 1,734 for 1995 (+29%), and 2,264 for 1996 (+30%). [BACK]

13. Combined BMMR and joint figures are 518 for 1993,958 for 1994 (+85%), 1,464 for 1995 (+53%), and 1,765 for 1996 (+21%). We have already seen an increase since BMMR relocated (3/97 = 1,985, about 30% annually), and we may expect a considerable bump after official unveiling of TMR at the annual conference in May 1997 (and the introduction of the Web site). [BACK]

14. Printing and mailing costs are about $5,000 and are covered by the subscription price of the paper version. [BACK]

15. BMMR has found that it takes 35 minutes on average to code a review in SGML. [BACK]

16. We received our first fast-track reviews in June and now find that well over half are submitted in this form. [BACK]

17. So far only Princeton and Michigan (of the eight university presses contacted) have signed up for e-advertising. [BACK]

18. Unfortunately, the University of Michigan, where our Web site is located, does not at present allow any advertising. [BACK]

1. Unless otherwise noted, the following discussion pertains to the situation as it appeared in the latter part of 1996 and the first part of 1997. Understandably, a year or two later many of the issues raised will have taken on a new significance. [BACK]

2. These and other features should be available shortly. As for the size of the database, of the total collection only a fragment is presently on-line. Prior to the introduction of the new system, libraries had been cataloging in ISIS. These records have been converted with little or no loss to the UNIMARC format, which meant-in the case of the National Library-that, from the outset, several hundred thousand records were ALEPH-ready. New acquisitions are cataloged directly into the new system and more records are made available through retrospective conversion.

The Web addresses (with access to the catalogs) are as follows: National Library of the Czech Republic (Prague), www.nkp.cz; the Moravian Regional Library (Brno), www.mzk.cz; the Slovak National Library (Martin), www.matica.sk/snk/snk.html; the University Library in Bratislava, www.ulib.sk. [BACK]

3. For additional details on this project, see the LINCA proposal presented to The Mellon Foundation (LINCA 1994). [BACK]

4. As of January 1997, the two campuses of UPJS have been turned into two universities. The move was political (divide and conquer), playing off existing institutional rivalry. What the consequence of this division will be on the project is not yet clear. The details of the original project can be found in the KOLIN proposal to The Mellon Foundation (KOLIN 1995). [BACK]

5. For a more detailed account of the compatibility and conversion problem as well as of the solution, see Appendix H of the MOLIN proposal (MOLIN 1996). TinLib is the most widely used library system among Czech universities (the Czech vendor is located at Charles University in Prague). [BACK]

6. More precisely, out of the approximately 1.5 million volumes deposited in the central library (the Klementinum), about one-fifth were unshelved. Because these volumes were new acquisitions-most in demand by users-most requests went unfulfilled. [BACK]

7. Hence also the symbolic significance of including the call number on the electronic record-it actually corresponds to a retrievable object! What a treat! [BACK]

8. With the series of austerity measures introduced by the Czech government in spring 1997, it remains unclear how the plans will be realized. [BACK]

9. Equally important in the area of rare book and manuscript preservation is the direct digitalization project at the National Library in Prague, in which early medieval illuminated manuscripts are being scanned and made available on CD-ROM (in 1995/96: Antiphonarium Sedlecense and Chronicon Concilii Constantiniensis ). This project is a UNESCO-sponsored Memory of the World project. For more on this project and its latest developments, see A. Knoll (1997), http://digit.nkp.cz/Structure_Proposal/navrhIII.htm. [BACK]

10. There is much more to this fascinating and complex project, well worth a separate study. The interested reader may wish to look at the original text of the project as it was presented to The Mellon Foundation for Funding (RETROCON 1994) and at a special publication of the National Library devoted to this topic (Bare&scarlon;s and Stoklasová 1995). [BACK]

11. With the proliferation of the cellular phone network in Eastern Europe, it is possible that many connections-such as user access-will take the wireless route. [BACK]

12. "External" constraints are often found embedded inside the organization. I wish to exercise caution in using this term since it is often quite difficult to pinpoint where an organization ends and the external world begins. [BACK]

13. The role of time management and, in particular, of delays in the implementation of the library project is a topic of a separate study (Lass 1997). [BACK]

14. According to a recent document (issued by the Slovak Ministry of Culture), the library's new mandate would include, among other things, the issuing of ISBNs and the development of the national bibliographic records. This mandate has put the library in the situation, apparently desired, of having to demand from the Slovak National Library the transfer of positions, computer hardware (mostly CASLIN Mellon purchases), and existing databases, without which it cannot do the job, although how it would gain the expertise remains unclear. While there are other examples in which institutional rivalries have adversely affected the CASLIN project, in only some of them does the rivalry reside with the libraries themselves. In several instances it is the libraries that are caught in the middle of a battle. Such is the case in the KOLIN project discussed above (see also footnote 4). [BACK]

15. On several occasions, librarians of one institution would express concern that "the other" library received more funds than they did. The symbolic significance of foreign (Western) funds is not to be underestimated nor should the role that this phenomenon has on the actual implementation of the project (covered best by the anthropological studies of cargo cults or witchcraft). As for the politics of institutional positioning, the situation under review was made transparent (and more complicated) by the breakup of Czechoslovakia and, following that, by the surfacing of other regional tensions. As a result, the relationship across the border is more amicable (there is nothing to compete over), while the relationship between the two libraries in each of the countries is much less so. [BACK]

16. One of the surprises was the funding of automation at the National (then University or State) Library during the 1980s in Prague, which resulted in the development of a local machine-readable record format (MAKS) that became the accepted system among a majority of Czech and Slovak libraries. The grounds for technology transfer were therefore prepared (contrary to some who maintained that there was no expertise in place) when automation arrived in earnest after 1990. [BACK]

17. Ironically, the library became one of the safe places to hide politically discredited intellectuals (from the post-1968 purges). [BACK]

18. If the purchase of foreign (especially Western) books and periodicals was restricted for mostly political reasons, it is now actually stopped altogether due to zero (!) funding. [BACK]

19. As a result, organizational behavior retains its characteristic sluggishness. It is reasonable to predict that as the constraints on the budget continue to increase, so will the familiar ability to trick the system. Under these conditions, teaching new management skills has been close to impossible, and introducing a new record structure and cataloging rules has been very slow. These conditions account, to a large extent, for the continued backlog of uncataloged books. [BACK]

20. The fact that "planning" is a discredited term doesn't help. And trying to explain that socialist planning and strategic planning may be quite different things doesn't seem to work. [BACK]

21. On the Czech side, the Ministry of Culture has decided to support library automation projects throughout the country, in the form of capital investment grants (no funds for salaries) that meet CASLIN standards. [BACK]

22. As of early fall 1997, the prospects for a functioning CASLIN Union Catalogue look promising. With the purchase of an Oracle software license and access to a Digital Alpha server, the plan is to make access to the database available to participating libraries by early 1998. [BACK]

1. Richard M. Dougherty, ''A 'Factory' for Scholarly Journals." Chronicle of Higher Education 38/41 (June 17, 1992): b1-b2; Bert R. Boyce, "Meeting the Serials Cost Problem: A Supply Side Proposal." American Libraries 24/3 (March 1993): 272-273. [BACK]

2. Stevan Harnad, "Post-Gutenberg Galaxy: The Fourth Revolution in the Means of Production of Knowledge." Public Access Computer Systems Review 2/1 (1991): 39-53; Andrew M. Odlyzko, "Tragic Loss or Good Riddance? The Impending Demise of Traditional Scholarly Journals." International Journal of Man-Machine Studies 42/1 (January 1995): 71-122. [BACK]

3. Frank Quinn, "A Role for Libraries in Electronic Publication." Serials Review 21/1 (1995): 27-30. [BACK]

4. Charles A. Schwartz, "Scholarly Communication as a Loosely Structured System: Reassessing Prospects for Structural Reform." College and Research Libraries 55/2 (March 1994): 101-117. [BACK]

5. Anthony W. Cummings, Marcia L. Witte, William G. Bowen, Laura O. Lazarus, and Richard H. Ekman, University Libraries and Scholarly Communication: Study Prepared for The Andrew W. Mellon Foundation. Washington, D.C.: Association of Research Libraries for The Andrew W. Mellon Foundation, 1992. [BACK]

6. Newsletter on Serials Pricing Issues. Edited by Marcia Tuttle. Chapel Hill: University of North Carolina, 1989-1997. Available electronically by subscribing to PRICES from listproc@unc.edu. [BACK]

7. Deana Astle and Charles Hamaker, "Journal Publishing: Pricing and Structural Issues in the 1930s and the 1980s. Advances in Serials Management 2 (1988): 1-36; Charles Hamaker and Deana Astle, "Recent Price Patterns in British Journal Pricing." Library Acquisitions: Practice and Theory 8 (1984): 225-232; Deana Astle and Charles Hamaker, "Pricing by Geography: British Journal Pricing 1986 Including Developments in Other Countries." Library Acquisitions: Practice and Theory 10 (1986): 165-181. [BACK]

8. David Lewis, "Economics of the Scholarly Journal." College and Research Libraries 50/6 (November 1989): 674-688; H. Craig Peterson, "Variations in Journal Prices: A Statistical Analysis." Serials Librarian 17/1&2 (1989): 1-9; Bruce Kingma and Philip Eppard, "Journal Price Escalation and the Market for Information: The Librarians' Solution." College and Research Libraries 53/6 (November 1992): 523-535; Michael A. Stoller, Robert Christopherson, and Michael Miranda, "The Economics of Professional Journal Pricing." College and Research Libraries 57/1 (January 1996): 9-21. [BACK]

9. Henry H. Barshall, "The Cost Effectiveness of Physics Journals." Physics Today 41 (July 1988): 56-59; H. Craig Peterson, "The Economics of Economics Journals: A Statistical Analysis of Pricing Practices by Publishers." College and Research Libraries 53 (March 1992): 176-181; John O. Christensen, "Do We Know What We Are Paying For? A Comparison of Journal Subscription Costs.'' Serials Review 19/2 (Summer 1993): 39-61. [BACK]

10. Edward A. Dyl, "A Note on Price Discrimination by Academic Journals." Library Quarterly 53/2 (1983): 161-169; Patrick Joyce and Thomas E. Merz, "Price Discrimination in Academic Journals." Library Quarterly 55/3 (1985): 273-283; Patrick Joyce, "Price Discrimination in 'Top' Scientific Journals." Applied Economics 22/8 (1990): 1127-1135. [BACK]

11. George A. Chressanthis and June D. Chressanthis, "A General Econometric Model of the Determinants of Library Subscription Prices of Scholarly Journals: The Role of Exchange Rate Risk and Other Factors." Library Quarterly 64/3 (1994): 270-293; George A. Chressanthis and June D. Chressanthis, "The Relationship between Manuscript Submission Fees and Journal Quality." Serials Librarian 24/1 (1993): 71-85. [BACK]

12. Roger Noll and W. Edward Steinmueller, "An Economic Analysis of Scientific Journal Prices: Preliminary Results." Serials Review 18 (Spring/Summer 1992): 32-37. [BACK]

13. George A. Chressanthis and June D. Chressanthis, "The Determinants of Library Subscription Prices of the Top-Ranked Economics Journals: An Econometric Analysis." Journal of Economic Education 25/4 (Fall 1994): 367-382. [BACK]

14. S.J. Liebowitz, "Copying and Indirect Appropriability: Photocopying of Journals." Journal of Political Economy 93/5 (1985): 945-957. [BACK]

15. For a fuller exploration of the issues that prescribe the study reported here, see Richard W. Meyer, "Monopoly Power and Electronic Journals." Library Quarterly 67/4 (October 1997): 325-349. [BACK]

16. Edward Chamberlin, The Theory of Monopolistic Competition. Cambridge: Harvard University Press, 1935; Jan Keppler, Monopolistic Competition Theory: Origins, Results, and Implications. Baltimore: Johns Hopkins University Press, 1994. [BACK]

17. Directory of Electronic Journals, Newsletters, and Academic Discussion List. Washington, D.C.: Association of Research Libraries, 1991-1996. [BACK]

18. Steve Hitchcock, Leslie Carr; and Wendy Hall, "A Survey of STM Online Journals 1990-1995: The Calm Before the Storm." In D. Mogge, editor, Directory of Electronic Journals, Newsletters, and Academic Discussion List. 6th Edition. Washington, D.C.: Association of Research Libraries, 1996, 7-32. [BACK]

19. Stephen P. Harter, "The Impact of Electronic Journals on Scholarly Communication: A Citation Analysis." The Public-Access Computer Systems Review 7, no. 5 (1996): 5-34. URL: http://info.lib.uh.edu/pr/u7/n5/hart7n5.html. [BACK]

20. Adonis is a product of Adonis USA, 238 Main St., 5th Floor, Cambridge, MA; a wholly owned subsidiary of Elsevier. [BACK]

21. Personal conversation with Karen Hunter, vice president of Elsevier, 15 February 1997. [BACK]

22. George A. Chressanthis and June D. Chressanthis, "Publisher Monopoly Power and Third-Degree Price Discrimination of Scholarly Journals." Technical Services Quarterly 11/2(1993): 13-36. [BACK]

23. This theory is based on the classic work by Abba Lerner, "The Concept of Monopoly and the Measurement of Monopoly Power." Review of Economic Studies (June 1934): 157-175. [BACK]

24. John R. Hayes, "The Internet's First Victim?" Forbes 156/14 (December 18, 1995): 200-201. [BACK]

25. The Associated Colleges of the South includes: Birmingham Southern, Centenary, Centre, Furman, Hendrix, Millsaps, Morehouse (Atlanta University Center), Rhodes, University of Richmond, Rollins, Southwestern, University of the South, and Trinity. [BACK]

26. Two variations in the dependent variable will be used: the net difference and the index of monopoly power after the work of Lerner. [BACK]

27. Chressanthis and Chressanthis, "Publisher Monopoly Power," 13-36. [BACK]

1. The book could be used entirely in an on-line format or scholars could choose to acquire a print version of all or part of the book once they had browsed the on-line version. Alternatively, at least at some point in time and for some forms of books such as textbooks, an electronic format such as a CD-ROM might be better-for technical, cost, or market reasons-than either the on-line or the print format. Malcolm Getz addressed some of the format issues well in chapter 6. [BACK]

2. In effect, funds that would have been spent on interlibrary loan activities, i.e., staff and mailing costs, would be redirected to the producers of the scholarly knowledge, thus supporting the production and dissemination of such scholarship. [BACK]

3. Details on early project activities and findings are available in Summerfield, Mandel, and Kantor, Online Books at Columbia-Measurement and Early Results on Use, Satisfaction, and Effect: Interim Report of The Andrew W. Mellon Foundation-Funded Columbia University Online Books Evaluation Project, July 1997. This report is available at http://www.arl.org/scomm/scat/summerfield.ind.html. Background information on project design and collection issues is provided in the project's Analytical Principles and Design document of December 1995 and in its Annual Report of February 1997. Both are available at http://www.columbia.edu/dlc/olb/. [BACK]

4. A few reference books were already on-line on a text-based campuswide information system. [BACK]

5. As noted earlier, this document is included on the Web page for the project. Questionnaires and other research methodologies have been fine-tuned after pretests and early use, but the general concepts remain in place. [BACK]

6. For Tables 17.2-17.7, Web data exclude hits by project staff. These hits were excluded because they can be substantial in number, as resources are in design phases, and do not reflect the scholarly use that we are studying. [BACK]

7. Columbia's Academic Information Systems designed a Web version of the OED that has various analytical capabilities, but that version requires more server resources than we can devote to this single work. [BACK]

8. Clearly, these measures are not absolutely equivalent. Scholars can access and use books from the regular collection without checking them out. (However, to use a book in the reserve collection, a scholar must check it out.) [BACK]

9. This type of skewed distribution, or Bradford law, is typical of use of all types of library collections. [BACK]

10. Each on-line book is comprised of several Web files-one for the table of contents, one for each chapter, one for the index, and so on. [BACK]

11. For use data to show revealed preference, the collection must contain books that would repeatedly draw users to the collection-either books that users want to look at often or an assortment of books that pulls scholars to the collection for a variety of purposes. [BACK]

12. Other students in these courses may have used the book earlier in the semester. Data for the fall 1997 semester will be complete. [BACK]

13. The School of Social Work is strictly a graduate school. Many of its students have residences in New York before they begin studying at Columbia. Virtually all are involved in various internships that take them away from the campus. In fact, they are likely to be on campus for only two or three days a week. Another group of students take most of their courses at another campus about 20 miles away. All these factors led to an expectation that this cohort would particularly value on-line books and the ability to use them away from campus. [BACK]

14. In January 1998, that count had grown to 2,700. By May 1998, it is likely to be at about 3,000. The residence hall network covers 15 buildings and has a total of 4,500 ports. [BACK]

15. Software allowing annotation of an electronic document is available, but few people are aware of it. The project will seek to bring such software to the Columbia community as feasible. [BACK]

1. Battin, Patricia. "Libraries, Computers and Scholarship." Wilson Library Bulletin, April 1982, 580-581. [BACK]

2. University Libraries and Scholarly Communication: A Study Prepared for The Andrew W. Mellon Foundation. Published by the Association of Research Libraries for The Andrew W. Mellon Foundation, Washington, D.C., 1992. [BACK]

3. Hawkins, Brian. "The Unsustainability of the traditional library and the threat to higher education." Paper presented at the Stanford Forum for Higher Education Futures, The Aspen Institute, Aspen Meadows, Colo., October 18, 1996. [BACK]

4. "Price Comparison of STM Journals 1996/1997." Harrassowitz News: Press Release, November 1996; URL: http://www.harrassowitz.de/news/9611pro1.html. [BACK]

5. Brueggeman, Peter. "Journal Subscription Price Increases." URL: http://scilib.ucsd.edu/sio/guide/prices/. For similar information from the University of Wisconsin-Madison, see University of Wisconsin-Madison. "University Library Committee Proposed Resolutions on Faculty Concerns on Copyright and the Role of Libraries (as adopted and amended May 6, 1996). Faculty Document 1214a, May 6, 1996. URL: http://www.library.wisc.edu/libraries/issues/ulc-ipr.htm. From Cornell, see Atkinson, Ross. "Summary of the 16 November 1995 Presentation to the Science Faculty." URL: http://scilib.ucsd.edu/sio/guide/prices/prices4.html. [BACK]

6. Chrzastowski, Tina E., and Brian M. Olesko. "Chemistry Journal Use and Cost: Results of a Longitudinal Study." Library Resources and Technical Services 41, no. 2, 1997, 101-111. [BACK]

7. Okerson, Ann. "A Librarian's View of Some Economic Issues in Electronic Scientific Publishing." Paper presented at the UNESCO Invitational Meeting on the Future of Scientific Information, Paris, February 1996. URL: http://www.library.yale.edu/~okerson/unesco.html. [BACK]

8. ARL Statistics, 1995-96. Association of Research Libraries, Washington, D.C., 1997, 6. [BACK]

9. See, for example, Drucker, Peter. The Age of Discontinuity: Guidelines to Our Changing Society. (New York: Harper and Row, 1969). [BACK]

10. Castells, Manuel. The Rise of the Network Society. (Cambridge, Mass.: Blackwell, 1996). See especially chapter 2, "The informational economy and the process of globalization," which discusses "the productivity enigma." [BACK]

11. Massy, William F., and Robert Zemsky. "Using information technology to enhance academic productivity." URL: http://www.educom.edu/program.nlii/keydocs/massy.html. [BACK]

12. Kingma, Bruce R. "The Economics of Access versus Ownership: The Costs and Benefits of Access to Scholarly Articles via Interlibrary Loan and Journal Subscriptions." (New York: Haworth Press, 1996), p. 37. [BACK]

13. Bowen, William G. "JSTOR and the Economics of Scholarly Communication." URL: http://www.mellon.org/jsesc.html. [BACK]

14. Cooper, Michael. "A Cost Comparison of Alternative Book Storage Strategies." Library Quarterly 59, no.3, 1989, 239-260. [BACK]

15. ARL Statistics, 1995-96. Association of Research Libraries, Washington, D.C., 1997, 11. [BACK]

16. Cooper, "Cost Comparison." [BACK]

17. Bowen, "JSTOR." [BACK]

18. Lemberg, William Richard. A Life-Cycle Cost Analysis for the Creation, Storage, and Dissemination of a Digitized Document Collection. (Ph.D. dissertation, School of Library and Information Studies, University of California, Berkeley, 1995). Also available on-line at URL: http://SIMS.berkeley.edu/research/publications/DigtlDoc.pdf. [BACK]

19. Costs for production of electronic journals in this paper are based on the experience of the University of California Press in its SCAN (Scholarship from California on the Net) project, funded by The Mellon Foundation. For details regarding these costs, seeAppendix A. [BACK]

20. See The Astrophysical Journal at http://www.journals.uchicago.edu/ApJ/ , Earth Interactions at http://earth.agu.org/ei/ , or any humanities journal with lots of multimedia. [BACK]

21. Braman, Sandra. "Alternative Conceptualization of the Information Economy." Advances in Librarianship 19, 1995, 99-116. [BACK]

22. Hammer, Michael. "Reengineering Work: Don't Automate, Obliterate." Harvard Business Review, July-August 1990, 104-112. [BACK]

23. University Libraries and Scholarly Communication reports, however, that the percentage of university budgets allocated to library budgets is gradually declining rather than consuming ever increasing proportions of institutional resources. [BACK]

24. Hawkins, Brian. "Creating the Library of the Future: Incrementalism Won't Get Us There!" Serials Librarian 24, no.3/4, 1994, 17-47. [BACK]

25. Battin, Patricia. "New Ways of Thinking About Financing Information Services." In Organizing and Managing Resources on Campus, ed. Brian L. Hawkins (McKinney, Texas: Academic Computing Publications, 1989), 382. [BACK]

1. Plato refers to a storehouse of wisdom (sophias ... thesauron), at Philebus 15e. Similarly, Xenophon alludes to the "treasures, which they left behind written in books" (thesaurous ..., hous ekeinoi katelipon en bibliois grapsantes), at Memorabilia 1.6.14. [BACK]

1. Peter F. Drucker, The Age of Discontinuity (New York: Harper and Row, 1978). [BACK]

2. Walter Wriston, The Twilight of Sovereignty (New York: Charles Scribner's Sons, 1992), xii. [BACK]

3. Erik Brynjolfsson, "The Productivity Paradox of Information Technology: Review and Assessment," Communications of the ACM (December, 1993),36(12)67-77. [BACK]

4. See, for example, William F. Massy and Robert Zemsky, "Using Information Technology to Enhance Academic Productivity," a White Paper for the EDUCOM National Learning Infrastructure Initiative (NLII), 1995. [BACK]

5. Anthony M. Cummings, Marcia L. Witte, William G. Bowen, Laura O. Lazarus, and Richard H. Ekman, University Libraries and Scholarly Communication. (Washington D.C.: The Association of Research Libraries, 1992). [BACK]

6. See Stanley Chodorow and Peter Lyman, "The Responsibilities of Universities in the New Information Environment," in The Mirage of Continuity, ed. Brian L. Hawkins and Patricia Battin (Washington, D.C.: Council on Library and Information Resources, 1998), 61-78. [BACK]

7. See Manuel Castells, The Rise of the Network Society (London: Blackwell Publishers, 1996), 56. [BACK]

8. Geoffrey Nunberg, "The Places of Books in the Age of Electronic Reproduction," in Future Libraries, ed. R. Howard Bloch and Carla Hesse (Berkeley: University of California Press, 1993), 21-22. [BACK]

9. The term "user" is applied to the consumption of digital documents the way the term "reader" has been applied to the consumption of printed works. "User" is a semantic strategy for pointing out that engagement with a technology mediates between the reader and the text, allowing for the direct control over content and format that Nunberg describes. [BACK]

10. See, for example, the report on a workshop held at the Hawaii International Conference on System Sciences in January 1995: M. Stuart Lynn and Ralph H. Sprague Jr., eds., Documents in the Digital Culture: Shaping the Future (Honolulu, Hawaii: A HICSS Monograph, 1995). [BACK]

11. On knowledge management, see, for example, Thomas H. Davenport, Sirkka L. Jarvenpaa, and Michael C. Beers, "Improving Knowledge Work Processes," Shan Management Review (Summer 1996), 53-65. [BACK]

12. The term "Cyberia" reflects the anthropologist's approach to analyzing the Internet as a site for culture and community and is best summarized by Arturo Escobar, "Welcome to Cyberia: Notes on the Anthropology of Cyberculture," Current Anthropology (June 1994), 35(4)211-231. [BACK]

13. John Seeley Brown and Paul Duguid, "The Social Life of Documents," Release 1.0:Esther Dyson's Monthly Report (New York: Edventure Holdings Inc., October 11, 1995), 7. [BACK]

14. John Hagel III and Arthur G. Armstrong, net.gain: expanding markets through virtual communities (Boston, Mass.: Harvard Business School Press, 1997), 5. [BACK]

15. Brown and Duguid, 5. This argument is derived from Benedict Anderson's Imagined Communities. [BACK]

16. The concept of virtual community was introduced by Howard Reingold, writing about the social relationships and sense of community that inter-relay chat, MOO, and MUD technology sustained. The term "community" is used provisionally, because the participants use it to describe their experience, not because virtual community has any necessary resemblance to more traditional meanings of the word. [BACK]

17. Note that Latour chose the genre of the novel to discuss this phenomenon. Bruno Latour, Aramis, or The Love of Technology, trans. Catherine Porter (Cambridge, Mass.: Harvard University Press, 1996). [BACK]

18. See, for example, Robert D. Putnam, "The Strange Disappearance of Civic America," The American Prospect (Winter 1996), 24(34-48). In the same issue, see also Sherry Turkle, "Virtuality and Its Discontents: Searching for Community in Cyberspace." [BACK]

1. This figure neglects fixed costs such as marketing, overhead, and so on, which would typically be included in a publisher's costs calculation. [BACK]

2. Odlyzko [1997] and Harnad [1997] have similar cost estimates. [BACK]