Kenneth Kirkland, DePaul University, LIBKLK@DEPAUL.BITNET.
1. KEYNOTE ADDRESS: What is the literature for? John Maddox, Editor of _Nature_.
Scientific literature is an essential part of science, not just a by-product of it. Generally, scholars acknowledge that their writings embody their scholarship, almost exclusively. However, scientific literature differs from other scholarship in that new techniques and new technologies are usually less well described than are discoveries and explanations, while the structure of scientific knowledge implies that all authentic contributions to the literature are, in principle, indispensable. Maddox says that scientists do not write well; they are satisfied to assume that it is understood that writing is a mere abbreviation of their thoughts.
There is continual change in the texture of the literature. Research articles have gained precedence over books, while printing and communications technology have made their publication rapid. Though these developments are positive, scientific literature is becoming ever less accessible as literature. The presently admired stylized format and general insensibility in the use of language are more of an impediment than the prevalence of specialized jargon.
The use of publication counts to assess the value of individual researchers in the short term leads at best to nonsense, and at worst to corruption of the literature. _Nature_ has become somewhat chary of the increased number of articles from Spain, where academic salaries are being determined with bonus points for the number of articles published in "good journals." _Nature_ appears near the top of a ranked list of such journals. The cold fusion contretemps is an obvious example of corruption of literature, and Maddox notes the recent rash of cases of scientific fraud.
2. The impact of networks on libraries. John Akeroyd, South Bank Polytechnic
Akeroyd described several network systems used in the United Kingdom, beginning with JANET, "an online computer network with about 50 libraries attached, providing access to the bibliographic database, interlibrary loan, information retrieval, professional communication, and online searching."
He observed that the CD is merely a convenient way to distribute large databases as LANs become very accessible. The advantages are that it is easier to budget for CD-ROMs than for the costs of online searches, and CD- ROMS are easily monitored to analyze level of usage. Disadvantages are that multiple databases with different search engines for each can confuse the users, can put them off.
The chief interface issues are the problem of attachment, lack of standardization, and search complexity. There are four different standards involved: CD-Rx, SFQL, Z39.50 (ISO), and DXS (Silver Platter).
The question and answer session elicited the comments that abstracts are becoming the primary sources for some users, that a competitive environment appears to be emerging in the UK rather than a cooperative one, and that there seems to be a trend towards on-demand publishing.
3. Networking databases: a user's perspective. Godfrey Lance, University of Bristol
BIDS (Bath ISI Data Service) links together online _Science Citation Index_, _Social Science Citation Index_, _Arts and Humanities Citation Index_, and _Index to Scientific and Technical Proceedings_, via JANET. Endusers, (i.e., all students, staff, faculty, or library patrons) have "free" access to search these databases online at the institutions that have agreed to provide the service. One hundred simultaneous sessions are possible; this will expand to 300. CHEST (Combined Higher Education Software Team) has leased these databases on behalf of the British higher education community for 1991-1994, at a cost of L6,000 per site per year, plus VAT. Eventually coverage will reach back to 1981 for most of the databases. The service is run from Bath University Computing Services on an ICL Series 39, Level 80 processor with 3 nodes, and is mounted under STATUS free text retrieval software with a menu driven interface. The user interface is described as "cleverly written" and the printed documentation as "quite clear." Searching is by word(s) in the title, author, institution, journal, Research Front Number, cited patent, or combination thereof. Searching must be done one year at a time, and one database at a time. An account describing BIDS' history and usage appears in an article by S. Scanlon, "BIDS: The Revolution in Database Access," in UKSG's publication _Serials_ (March 1992, vol. 5 no.1) p. 18-28.
End-users' direct access to bibliographic databases may be changing the role of librarians/information scientists. Librarians must do the training of database users, since new users need instruction on how to do searches on various systems. Librarians are or must become the best qualified to teach and assist. With BIDS even computer novices encounter little difficulty in accessing the databases, negotiating the menus, and carrying out meaningful searches. Scanlon says, "The traditional role of the librarian as the custodian of knowledge which he sometimes reluctantly imparts to the masses needs to change to accommodate the needs of the end-user." He notes that the recent debate over the impact and appropriateness of CD-ROM may be obscuring the more profound development, that of direct access by endusers. They are likely to be more demanding of mediated online searching, to be more demanding for other databases, for text retrieval software, and for increasingly more sophisticated search capabilities through both expert and novice menu driven interfaces. Although some librarians and information professionals may fear that they are in danger of becoming de-professionalized as end-user searching becomes more widespread, this is not inevitable. Collateral support in training, guidance, and ancillary services such as ILL requests and reference assistance will become the major focus of librarians. In ruminating on the future, Lance observes that LIBERTAS is currently used to locate material and suggests that ADONIS is an interim solution. Electronic publishing is a long-term solution, where the user will get an article direct and pay for it.
4. The information cornucopia: experiments with an integrated workstation. J. Eric Davies, Loughborough University
The EI Reference Desk, developed by Engineering Information Inc. of New York, is designed to be a one-stop information facility that allows a user to identify and select references, retrieve and deliver documents, and manipulate material. The Workstation set up at Loughborough was described and observations drawn. There is a generation gap in computer use. Students are keener on electronic searching than are faculty. Electrical engineers are more likely to use the workstation than are mechanical engineers. Things have changed since the time of Newton, who exchanged letters with his cohorts and did not need journals, which did not exist yet anyway. In a way, electronic publishing and electronic bulletin boards are going back to that earlier model.
5. ADONIS. Barrie Stern, Director of ADONIS.
"ADONIS is live, ADONIS is well, ADONIS is frustrating." Commercial operation of ADONIS began in 1991 after five years of trial and development initiated by publishers including Elsevier, Springer Verlag, and Blackwell's. In the initial two-year trial stage the contents of 219 biomedical journals were scanned and the images stored on 84 CD ROMs using Group IV (GIV) CCITT facsimile standard compression at a resolution of 300 x 300 pel. At that time nearly one new disk per week required much manual handling by the 13 test sites. During the next period trebling of the CD ROM capacity was attempted, building-in intelligence to a high volume scanner, special procedures for handling print errata and automated audit trials. It is set up as a not-for-profit corporation whose subscription charges are to pay office expenses, to cover printing costs and fees to copyright holders. The software uses unique identifiers to count usage of every article retrieved. Essentially, ADONIS is document delivery of biomedical information on CD ROM and is proving to be cheaper than photocopying. The operation is quite international, with indexing in Amsterdam and daily courier service involving Berlin and Wales. The jukebox is the latest development. Seven chained together handle 700 disks.
6. Electronic journals and newsletters: the "Newsletter on Serials Pricing Issues." Marcia Tuttle, University of North Carolina
Tuttle opened by observing that the first ARL directory of electronic publications listed over five hundred entries and a second edition is just coming out. She traced the first mention of such publications at the 1982 UKSG conference, where BLEND (Bristol Loughborough Electronic Network Development) was described, and IES and IIT which were inspired by Lancaster's 1978 _The Paperless Society_. The earliest attempts failed because articles were not available to readers. Then, microcomputing took off. A newsletter is a serial recording news on a specific topic; in NSPI "news" is interpreted broadly. It is not and will not be refereed. The newsletter is not "scholarly" and is not intended to be. It began as a committee publication of the American Library Association, the first issue appearing in February 1989. There were 50 electronic subscriptions, and 100 paper format subscriptions. More copies than that are distributed by vendors such as Faxon and EBSCO, and Dan Tonkery from Readmore spoke up to say that Readmore sends out 400 copies through its facilities. Electronic subscriptions have passed 1000. Costs of publishing NSPI are hidden. Will the University continue to subsidize it? EBSCO and Readmore get claims for issues when the editor has gone on vacation.
The NSPI tries to avoid publisher bashing but prints many letters complaining about price increases. The paper edition did not last long -- many paper subscribers switched to electronic format. NSPI separated from ALA in May 1991, and discontinued the Faxon Courier version shortly thereafter. After some months of operation SERIALST, the serials discussion group owned, operated and edited by Birdie MacLennan at the University of Vermont, invited Tuttle to be a co-editor. SERIALST provides instant discussion and interchange online. NSPI has the right to re-publish information concerning pricing issues.
In the United States it seems that only Virginia Polytechnic Institute and Ohio State University have resolved the question of whether/how to furnish electronic journals to patrons. Lack of access, the likelihood of tampering with text, questions of how to archive to preserve original text, and retention are all concerns for any electronic publication. A member of the audience noted that indexing of some electronic publications is expected to begin soon by the Modern Language Association and the American Psychological Association.
7. The Scottish Science Library; its role in national provision. Morag Nisbet, The Scottish Science Library.
The Scottish Science Library traces its roots to 1682 with the foundation of the Library of the Faculty of Advocates and by 1925 was part of the National Library of Scotland. As such, it is one of the six legal deposit libraries of the United Kingdom. The SSL opened in mid-1989, with the collection of the Royal Society of Edinburgh and separating out the science collection of the National Library of Scotland. In November 1989 its business information service was launched.
Decisions have been made not to duplicate many services already successfully offered elsewhere, and to pursue acquisition polices and services that fill gaps identified in existing Scottish libraries.
8. BBC. Frances Tait, British Broadcasting Corporation
The presentation answered several questions. When and why do television programme makers use serials? How do they get access to them? How do the libraries cope with a reader population interested in everything, at an academic and a popular level, who must be completely up-to-date and "by the way could I have the information straight away as" a) "my programme's on-air in 10 minutes" or b) "I'm flying to Tokyo first thing tomorrow"?
Tait characterized the use of periodicals in a special library such as that of the BBC as "haphazard." Clippings or cuttings files abound. The major thrust is to provide up-to-date information, news items. There are approximately 300 general interest subscriptions including _The British Medical Journal_ and _The New England Journal of Medicine_, with a retention period of ten years. Some journals are used primarily for picture files to provide historic photographs of actual events, of the 1940s for example, or to illustrate styles of the period for sets or costuming. _Vogue_, _Ideal Home_, and _Faces_ are titles supporting this category. Contacts are absolutely essential to BBC. It is commonplace to telephone editors of journals, or authors writing about computer viruses or any topic of pressing current interest, to get specifics or updates or more background. Full-text databases are used, as in the recent example of discovering what the US term "political correctness" was all about, discovered via Nexus in an _Atlantic Monthly_ article. Scanning is done, especially through databases, to discover items relevant to current BBC programs. Another aspect of library service was called the "quick fix," to satisfy immediate demands for data on hot topics. Reviews are maintained for background, supplementary, or introductory material, for occasions when "I want everything on the effect of electromagnetic waves on health." The well-received presentation ended with a personal expression of gratitude to publishers, distributors, and librarians.
9. Royal Botanic Garden, Edinburgh. Colin Will, Royal Botanic Garden
The RBGE was established in 1670 as The Physic Garden, and has occupied its present location since 1820. (The trees were moved at the time). There are three other branch gardens.
Will noted that the way in which a subject is studied dictates the structure of its literature, and hence the management techniques required within a special library. The RBGE Library is a major collection which, together with the Herbarium and Garden, support international research in plant taxonomy. The institution works intensively on a finite range of plant families, and extensively on the whole range of plants found in five geographical regions. The research leads to publication of specialized types of literature, Floristic and Monographic. Floras are region-specific, while Monographs are taxon specific. The library's many serials are specialized and of limited circulation. Newsletters, small society journals and other ephemera are viewed as significant. Enthusiasts groups exist, a situation which is not common in other sciences. At times this can make it difficult to judge the worth of some small publications such as "Orchidea", "Garden Newsletter," "The Magnolia," "The Orchid Advocate," and "Garden."
Plant names change all the time, and in fact are in constant flux. Hence there is much literature generated as a consequence. The concept of what is a species has changed many times.
10. Goldman Sachs International: The specialist use of serials with an international investment bank. Jane Borcherson, Goldman Sachs International.
The bank was founded in 1869 in York. It is privately owned by 147 partners, with branches in New York, London, and Tokyo. The library uses more than 550 serial titles to support the bank through company, industry, product, and people information. The serials are also employed to ascertain how Goldman Sachs International as an institution is viewed by the press. One particularly interesting description was that of subscriptions to Japanese journals via satellite from Tokyo. The paper copy is then printed out in the UK. This is seen as the only way to get the journals quickly.
11. WORKSHOP: The impact of the electronic journal on the serials industry. Sharon Bonk, SUNY Albany; Arlene Moore Sievers, Case Western Reserve University
Electronic journals have no past. They are burgeoning but are not replacing established paper format journals. Interest is spurred by librarians' HOPES of doing something about pricing and the overflow from the information explosion. Researchers have hopes for faster exchange of information and data. However, publishers are not taking over any established electronic journals or starting any new ones.
Three variant definitions illustrate the difficulty of pinning down exactly what an electronic journal is: 1) exists only in a digitized format and is accessible via an electronic database and/or network. (This would exclude "NSPI" which has existed also in paper.) 2) Produced, published, distributed nationally or internationally via an electronic network such as BITNET or the INTERNET. (This excludes commercial publications.) 3) A periodic publication of machine-readable files transmitted through a telecommunications system. (This excludes diskettes).
General observations included Sievers' opinion that the electronic format is most suitable for newsletters, not so great for journals. Bonk remarked on the prevalence of ASCII format although OCLC's Hypertext will be an innovation once it begins. SUNY Albany is not in the forefront, but has _USA Today_ on its OPAC. Case Western Reserve also has _USA Today_. Case is the first fiber optically wired campus. At present they are using, forwarding, but doing nothing systematic with electronic publications. Eventually they will be on the university mainframe.
Virginia Polytechnic Institute is one of the few US libraries actually doing something to handle and distribute the new format. Gail MacMillan presented a paper at the 1991 NASIG Conference and has published descriptions of VPI's work. Their policy proceeded from VPI's philosophy -- to change the method of scholarly communication. Planning started with a task force of faculty and library staff, to determine how to access and store the data. They would have preferred to use the library LAN, but that just was not possible, so they are using their mainframe, which begs the question of who controls and archives the electronic journals. Procedures are based on the old ways of handling serials. They just wanted to DO something. Ohio State University is probably second in comprehensiveness, with ten titles or so. Stanford is downloading to disk, biding their time to put titles on the mainframe.
_The Chronicle of Higher Education_ is trying to digitize and is looking into hypertext. OCLC's AAAS journal is intended to start up on the INTERNET any day now, and it is to be extracted in BIOSIS. A sophisticated work station is required to handle this, IBM compatible, graphics in typeset quality, modem, and Windows. It is a big experiment.
Bonk distributed to all a reprint of "Models of Information Flow" by Czeslaw Jan Grycz of the University of California, Office of the President (ARL 160, January 2, 1992) which she recommended highly. Of the seven models presented in this paper, Bonk noted that the site-license model is the one that seems to come up most often.
Sievers tackled the question of the marketplace and how it affects the publishing industry. She conducted an informal inquiry of Elsevier and Springer and found no specific plans to start an electronic journal or convert to electronic format. Elsevier is experimenting with five medical journals made available through BRS to hospitals and clinicians but not to libraries. These are in subject areas such as cardiology and thoracic surgery and are tied to print subscriptions. The idea is to set up models of how they would charge. Springer is looking at journals targeted to the chemical industry, _Applied Catalysis_ and three others. Again, a specific subject with article access is involved, based on having a print subscription. Beyond the sciences, Bryn Mawr is about to begin a classics review in ASCII base only.
What is a library to do? It must involve many people, including the computing staff, not just one bibliographer and one faculty member. "There is not a lot of demand" for electronic journals, so pick what is applicable for your campus. The 2nd edition of the ACRL directory of electronic publications is 45% larger than the first. An infrastructure must be developed within the library, and although there is danger of heading into a blind alley, libraries must be prepared.
The question and answer session provoked a number of remarks. Publishers are just now accepting articles on diskette and catching on to data-processing. "ASCII is not good enough." One library has an SGML expert on the staff who needs to connect with others on campus, not remain isolated. A publisher asked if electronic journals will be peer reviewed -- the answer is that a few are already. Bonk feels that libraries may have to tailor service more -- have a science services librarian, a humanities services librarian, etc. Gillian Page said the US represented just 1/3 of her subscribers, which will not cause an overthrow of print editions. Publishers may, however, offer parallel formats. She brought up the question of who pays the costs, and said that a publisher must sell more copies of print to make a profit.
John Urquhart asked how back runs will be archived. On floppies? How does one resolve what to keep online versus what to put on tape. There may need to be a special CRL type of electronic archives. Another person noted that NASA lost the instructions manual on how to run the data cards they had compiled. Therefore, there could be some question of viability of electronic archives.
12. Newspapers on CD-ROM. Geoffrey Smith, British Library Newspaper Library
Smith provided a snapshot of the current situation, saying that any comments would be provisional because CD ROM practice is still evolving. The British Library has mostly British newspapers in this format.
(Price, annually in L) CD MF Index Paper Times. 1990- 595 700 450 165 Guardian. 1990- 495 330 470 145 Independent. 1988- 499 910 x 170 Financial Times. 1990- 900 585 570 170 Northern Echo. 1989- 100 x x x
Smith maintains that the choice of medium is based on institutional or user needs. Up to now, the commercial sector has required on-line access and cuttings in hard copy. Public and national libraries have required hard copy, microfilm, and indexes. On CD-ROM recreational material is omitted -- cross word puzzles, television schedules, birth, death, and marriage notices, and copyrighted material. These are the same exclusions as in online versions, and this limits the value for long-term research uses.
Other limitations of CD ROM are that each title and year is separate, requiring much disk swapping. Jukeboxes are not a complete solution to this problem, because of another major limitation -- no standardization of interfaces and search strategies. Furthermore, the physical life span may be as short as 10 years, and equipment may not be available for older material if earlier technology is superseded.
Smith's conclusion is that CD ROM is not being used as a replacement for microform or hard copy, but rather as complement. For the future, greater standardization is needed. Prices seem to be going down. Multitype interfaces may be supplied with some databases. There may come a splitting off of the interface from the data and the search engine. Increased storage density will allow multiple years per disk and provide for storage of more graphics. Facsimile may become more approximate, with digitization of page images, or storage of page make-up information, so that an actual page may be recreated at the time of retrieval. CD ROM may be only an intermediate stage of development, merely an alternative form of electronic publishing.
13. Networking CD-ROM. Nicky Whitsed, Charing Cross and Westminster Medical School
Software makes it possible to dial anywhere from a PC, which is exciting and fast but may be expensive. Networking saves multiple subscriptions, but the overall cost of networking may be higher than stand-alones. A WAN, Wide Area Network, is many LANs strung together.
Why network? 1) Cost saving; 2) multiple access; 3) off-site access; 4) better resource management; 5) improved disk performance.
Whitsed's evaluation of Silver Platter is that with 44 disks you can't see everything on one screen. There is provision for passwording but not by individual or group. The management reports on usage are a nice feature. The most negative thing is the lack of standardization, and the increasingly higher cost of licensing agreements.
14. PANEL DISCUSSION: What if.... The "what if" questions were gathered in advance from the audience. Gordon Graham, editor of _Logos_, chaired the panel.
1. What if UK subscription agents were reduced to 5 only?
Dan Tonkery of Readmore said the Big 5 will stay around, maybe become the Big 4 or Big 3. He noted that they are prevalent universally. Fred Friend of University College London does not think the number will be reduced.
2. What if librarians continue to cancel relying on the just in time idea for data retrieval services and document delivery?
Friend: Libraries are forced to justify costs, but in actuality do not have an efficient document delivery system in place now (i.e., one that takes minutes or hours vs. days). Friend has not cancelled much yet, but that day may come.
Tonkery forecasts that suppliers will become involved in document delivery. Agents will not build mammoth libraries themselves. Can't really put ALL the world's literature on CD ROMs -- too many disks would be required.
3. What if network access became a charged service, and commercial entities were eligible to use the networks? (Sharon Bonk)
Friend: Commercial organizations are already beginning to use JANET, but as a service to an academic constituency. If the use were for their own ends, however, they should pay.
Tonkery: Back home the federal government is paying. You can bet that eventually there will be charges, and IBM and others will be behind it. In business, a company has many computers, but only ONE will be on the INTERNET, not the whole business, and not the mainframe.
The question was posed, how will commerce access electronic journals? "With difficulty," per Gillian Page, Pageant Publishing, but she declared that every publisher needs to see Marcia's newsletter.
Friend replied, "the same as the paper journal."
4. What if universities establish their own scholarly distribution channels? What happens to commercial publishers? After all, _Nature_ showed a loss for its first 30 years.
Jo Haythornthwait of Glasgow Polytechnic suggested that the commercial publishers will stick to highly commercial items, such as recreational stuff. Academia will do its own esoterica. Universities have not been good publishers, particularly. This could be a disaster.
5. What if academics were paid royalties? Perhaps there would be fewer articles published. Book publishers have cut back, and their prices have not doubled in real terms as they have for journals over the last 13 years. (John Urquhart)
Tonkery: If the economics of journal publishing switches to charging by the article, the amount of information out there will also change. At present, much is subsidized.
6. What if Labor or Liberal Democrats win the election?
Haythornthwaite: Under Labor, education would not change much. The polytechnics are becoming universities, but they still won't be stocked as well as university libraries. Some libraries will be research libraries, some teaching libraries, some both. There will still be a two- or three-tier system of education regardless of who's in power. Labor might put more money in students' pockets. Labor would have a marginal impact.
Page: "If you're starving and standing in line, it doesn't matter who the cook is." If the Liberal Democrats won, there would not be much difference. Universities would still be searching for income. They are hiring researchers and getting patents for royalty income, and the university owns you 24 hours a day.
7. What if cancellations occurred for all subscriptions over a certain unacceptable increase?
That might reduce the size of journals, but would probably increase the number of journals, with no over-all price or cost decrease. Friend applauds the people at Princeton, but the action was not entirely realistic. Will you look at each subscription one-by-one to decide if it is good value for money? Haythornthwaite would like to see more ruthlessness in the cancellations. Tonkery expects that there will be more sharing, more co-operating. Libraries haven't really got there yet. This year in the US was the first time the science titles were really seriously cancelled, and this will continue next year.
CIVIC RECEPTION. Hosted by the Lord Provost and Council of the City of Edinburgh. Members were piped into City Chambers and greeted by the Lord Provost (mayor). Red and white wine and whisky were served. There was a buffet of dressed whole salmon (a very Scottish specialty, and a very special specialty), roast turkey breast, roast rib of beef, baked gammon, and aubergine provencale; salads: mixed; potato; lettuce, celery, apple, pineapple and vinaigrette; pasta; cheese, carrot, peach and sultanas (raisins); and Waldorf. Dessert, before coffee, was fresh fruit salad and cream, and Scottish trifle.
BANQUET. The food was quite good this year. The Banquet, in addition, had the noteworthy tasting of the haggis. The haggis was piped into the hall on a tray held high by the chef, between two pipers, and, after Robert Burns' "Address to a Haggis" was declaimed, duly piped back to the kitchen before serving. The dinner was followed by rollicking Scottish country dancing and Auld Lang Syne sung at a very early morning hour.
Readers of the NEWSLETTER ON SERIALS PRICING ISSUES are encouraged to share the information in the newsletter by electronic or paper methods. We would appreciate credit if you quote from the newsletter.
The NEWSLETTER ON SERIALS PRICING ISSUES (ISSN: 1046-3410) is published by the editor as news is available. Editor: Marcia Tuttle, BITNET: TUTTLE@UNC.BITNET; Faxon's DataLinx: TUTTLE; Paper mail: Serials Department, CB #3938 Davis Library, University of North Carolina at Chapel Hill, Chapel Hill NC 27599-3938; Telephone: 919 962-1067; FAX: 919 962-0484. Editorial Board: Deana Astle (Clemson University), Jerry Curtis (Springer Verlag New York), Charles Hamaker (Louisiana State University), James Mouw (University of Chicago), and Heather Steele (Blackwell's Periodicals Division). The Newsletter is available on BITNET. EBSCO and Readmore Academic customers may receive the Newsletter in paper format from EBSCO and Readmore, respectively. Back issues of the Newsletter are available electronically free of charge through BITNET from the editor. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++