ACRL College & Research Libraries News

College & Research Libraries News

Online databases and book preservation

By Danielle Mihrani

Assistant Curator, Bobst Library New York University

Two topics of increasing interest to literary scholars were discussed at MLA.

This year’s annual meeting of the Modern Language Association (MLA) took place in San Francisco on December 27-30, 1987. The attendance was quite large—approximately 11,000 participants—and the sessions which were distributed over that period reached a bewildering number—721. Even for a single scholar, the attempt to attend all concurrent sessions of interest within a single subject specialty was nearly impossible.

Though much of the program had to do with literary studies, a few sessions proved quite interesting from the librarian’s perspective: those that related to online literary databases, and, for the very first time, to book preservation. There were also sessions dealing with desktop publishing, the ethics of publishing, and the evaluation of educational software. Anyone interested in obtaining a comprehensive view of the program should consult the November 1987 issue of the PMLA. This report will concentrate on a few sessions that brought together teams of librarians and academics in discussions regarding online databases and book preservation.

The main focus of one session (#230) was the question: “Should there be a Library of Great Britain like the Library of America?” while another (#137) concerned itself with “Dating Manuscripts: Current Science Techniques for the Identification of Paper and Ink.” Another session of note (#130) was chaired by Geoffrey D. Smith (Ohio State University) and entitled “Literature on File: Prospects for a National Literary Data Base.” Smith had attended an MLA meeting in 1984 where he had presented the results of his work on the American Fiction Project; this time, the session included presentations dealing with the “Afro-American Novel Project” (Maryemma Graham, University of Mississippi) and the “Black Periodical Fiction Project” (Ilenry Louis Gates Jr., Cornell University). Librarians interested in obtaining information about these two databases are encouraged to contact each of the two speakers.

The session began with MLA’s Eileen M. Mack- esy’s “Overview of Current and Future Development in Online Literary Data Bases.” Mackesy’s presentation echoed many librarians’ concerns about the impossibility of obtaining an accurate tally of existing online databases in the humanities.

Interestingly, one comment from the audience was that the MLA should consider publishing a directory of online databases relating to areas of interest to the membership. Mackesy noted that in the 1960s such databases were not too well structured and that there is now a need to create consortia for collections of databases in areas such as folklore, linguistics, women’s studies, and German literature, to name just a few. She outlined a few of the partnerships with a few vendors that the MLA was currently considering, and acknowledged that such ventures were not easy since certain questions, such as the merging of differently structured databases and rights of ownership and distribution, are difficult to resolve. The MLA is currently working with organizations here and abroad, including Cambridge University, in order to bring such plans to fruition.

ARTFL

Of parallel interest was session #380, chaired by myself, which dealt with the French textual database, American and French Research on the Treasury of the French Language (ARTFL). It is currently available at several libraries (including Berkeley, Columbia, and Yale) but because of its contents and, until recently, rather cumbersome search capability, the database has not received as much attention as it might. The panel consisted of one librarian (myself), two professors, (Paul A. Fortier, University of Manitoba, and Raymond T. Riva, University of Missouri, Kansas City) and a respondent, the director of the ARTFL, Robert Morrissey of the University of Chicago.

The database was conceived 25 years ago in France by Paul Imbs and evolved into the Institut National de la Langue Française in Nancy. The initial aim of the project was to compile a dictionary of the French language by using a “word bank” created by machine-readable texts. This project, financed in 1957 by the French government, led to the creation of a new dictionary of the French language, Le Trésor de la Langue Française. About 1,500 works from the 18th to the 20th century were included.

This database was made available to researchers in North America through the University of Chicago, which contracted with the Centre National de la Recherche Scientifique in 1981. Inaugurated in October 1983, ARTFL became accessible to the public in January 1984 and can be accessed via modem from anywhere in the United States and Canada.

Some universities subscribe to ARTFL through their libraries, while others access it via individual departments. Critics of library subscriptions have asked: Can the library really minimize connect time? What are the equipment needs of a large user community? Can acquisition be justified by the number of potential users? Would only librarians with expertise in French language and literature qualify as search specialists?

Fortier maintained that ARTFL software and its documentation was superior to other products, but that novices might find the manual unclear. He also noted that the compilation of concordances should not be a problem since ARTFL uses the Oxford Concordance Package (OCP), in his opinion one of the best software packages available for this application.

Riva reported on some of the frustrations he encountered in his own online searching. He found his 300-baud modem far too slow; his unfamiliarity with ARRAS led to problems in his search strategies; and his frustration was heightened by the fact that there were long waits for tapes to be mounted—in some cases tapes were changed only at night, and sometimes they were unavailable on weekends (I was told by some researchers that a 24- hour notice was required for access to some texts). Fortier said that he has been able to experiment with different baud rates when data interrupts became a problem. He also urged users to create a notebook entitled “This Works” in which successful logon and save procedures can be kept at hand.

In view of the various problems it is not surprising that several of the librarians whom I interviewed both before and after the conference expressed dissatisfaction with the database as a library resource. A paper, presenting one librarian’s experience with the database (J. Spohrer, University of California, Berkeley) is scheduled for the WESS conference in Florence this April.

Robert Morrissey announced that annual subscription rates to ARTFL were reduced to $500, effective in January 1988, and that in March a new system, PhiloLogic, will replace ARRAS and greatly expand its research capability. Because PhiloLogic will run on dedicated computers owned by the ARTFL Project rather than the University of Chicago’s mainframe, the annual fee will provide for a large block of free computer time so that most subscribers will pay only for connect charges. Finally, PhiloLogic will permit immediate access to any work in the database, both by dumb terminals or by a new Macintosh terminal interface.

The database continues to grow as individual scholars donate their own machine-readable texts and it can serve as a central national repository of French literature, history, and political science, accessible to scholars and researchers on demand.

MRTH

The next logical step is bibliographic control. Is there a National Union Catalog of French machine-readable texts? The answer is no, not a complete one. Machine-Readable Texts in the Humanities (MRTH), compiled by Rutgers University and available for searching on RLIN in its MDF file, originated about 1984 under the direction of Marianne I. Gaunt, librarian at Rutgers’s Alexander Library. At its inception the file included infor- mation on 50 items; today it includes information on 661 items. For a complete listing of the MRTH, select command: fin cp rut#; the file may also be searched using all other regular RLIN commands.

MRTH includes information on some ARTFL texts as well as machine-readable texts in humanities from elsewhere in the world—Dutch literary texts, Shona proverbs, ancient Greek and Roman texts, Italian linguistic works, and literary texts and ballads from Oxford. Approximately 88% of the information in MRTH consists of works in ARTFL, although it is incomplete. I tried a personal name search for Victor Hugo and obtained nothing, although one user did get a concordance of specific terms in Hugo’s works by using ARTFL. The file also contains typographical errors, so that it is possible to miss a title if the search is done by key words in the title field or if an author’s name is misspelled during cataloging.

Preservation

An important session dealing with the topic of book preservation was entitled, “Brittle Books: The Preservation Crisis in the Nation’s Libraries and What’s Being done About It.” It was arranged by the MLA Department of English Programs and was chaired by George Farr Jr., of the National Endowment for the Humanities. The panel consisted of three speakers: John W. tlaeger, Research Libraries Group; Barclay W. Ogden, University of California, Berkeley; and Phyllis Franklin, executive director of the MLA.

The meeting’s goal was to explore ways in which scholars might contribute to the preservation of books, and it followed the screening of the 28- minute version of Slow Fires: On the Preservation of the Human Record, shown two days earlier. In an editorial in the Winter 1987 issue of the MLA Newsletter, pp.4-5, Franklin discussed the problem of brittle books and the topics raised in the film, and she urged conference-goers to view the film.

John Ilaeger noted that “the only viable solution to the preservation of books is microfilming.” To conservators this is a sweeping statement and, regrettably, throughout the session no other methods of preservation were discussed. However, Haeger pointed out that microfilming is an expensive and very labor-intensive operation that remains at a cottage-industry level. As a result, RLG is making available in machine-readable form the bibliographic record of existing material microfilmed by any RLG member institution. The first target of this cooperative microfilming project are American imprints that are in the greatest state of peril. The second targeted set is material from the late 19th and early 20th centuries; individual member institutions can choose a field that they wish to target for microfilming (history, literature, etc.). Acoording to Haeger, this is a vacuum-cleaner approach, because no effort has been made to decide which items are more important than others; once the field has been chosen, all brittle books within it become candidates for filming.

This procedure was chosen as a result of the findings of a project conducted by the American Philological Association, which appointed an editorial board of scholars who worked from shelf lists and bibliographies to create a list of key titles printed between 1850 and 1918. The books which appeared on the list were filmed by Columbia University.

The project showed that such editorial boards are not really effective: scholars differed in their choice of titles needing preservation and did not agree on a final list. Furthermore, this method is time-consuming, costly, and cannot be replicated on a large scale; too many titles need preserving and time is running out. In response to a suggestion by Phyllis Franklin that perhaps the collections of major libraries are themselves a proxy collection of titles to be saved, Haeger pointed out that important material remains underrepresented in the best of collections. As an example he indicated that in his own field of specialization, medieval Chinese history, the collections in most libraries are underrepresented; local and regional history, trade journals, and ephemera are also sparsely collected by most major libraries.

Phyllis Franklin’s presentation centered on the question of financing major preservation projects. She outlined two sets of options.

First, if the Council on Library Resources’ Commission on Preservation and Access can generate funds to microfilm the best collections of books held by major research libraries in the United States, then the target period would be 1860-1920. Future scholars would need to ensure that no gaps exist in those collections. However, this option remains very costly: $384 million is her estimate for filming one-third of the volumes considered important, and this figure represents only 20 % of the total number of brittle books.

A second option requires selective measures. Franklin suggested that instead of a title-by-title selection, a range of material could be chosen as representative of titles found within a field. It then becomes relatively easy to identify major subject areas, as well as minor authors, and preserve examples typifying each subject or author. Another project would be to compile descriptive bibliographies covering both extant and lost material, thus providing future researchers with more detailed information than an unannotated bibliography or a publisher’s trade list.

Franklin identified three geographical areas that she felt should be targeted first by MLA: Germany, Italy, and the United States, particularly because the period to be covered (1860-1920) is so important historically for those countries. She added that if these projects prove worthwhile, MLA would turn to the literature of other countries. Her plan is to approach the divisions and discussion groups in those specific areas and urge them to include preservation concerns in the planning of sessions for their next annual meeting in December 1988.

Barclay Ogden, head of the preservation program at Berkeley, noted that the library is the scholar’s laboratory and that it behooves scholars to be involved in preservation programs. He indicated that one of the major priorities of a preservation program should be collection security. He claimed that, though many libraries believe in democratizing access to their collections (to the point of compromising security), his own inquiries among scholars have shown that they do endorse restrictive access, if the reasons given are well- documented and reasonable. He also encouraged scholars to ensure that their library has a well- established disaster plan, as protection against the consequences of natural disasters.

The normal wear and tear on a collection, even for non-brittle books, necessitates a hard look at the question of book replacement. In some cases, Ogden cautioned, some books which are not in heavy demand are never reprinted, and yet they merit preservation. He asked scholars to get involved in the preservation process by suggesting titles in the collection which should be preserved. He added that new trends in scholarship create shifts in interest, and that scholars can spot these trends and alert their library.

George Farr concluded the session by outlining the current goals of the National Endowment for the Humanities. Though preservation plans have been in place since 1979, two are currently being considered:

1) grants for research in the humanities (microfilming projects and evaluative programs which would bring together scholars and librarians in an effort to identify fields and specialties needing preservation); and

2) projects that would help promote preservation activities, training for preservation managers (such as the Columbia program), grants for the preservation of regional collections, grants for research in the technology of preservation, and grants aimed at raising the public awareness of the need for preservation.

The ensuing discussion elicited a few additional details: a) we need to know how other countries are dealing with preservation problems and establish cooperative programs, especially since so much of the scholarly interests of the MLA include world literature; b) scholars must actively lobby for acid- free paper as a standard in humanities publishing; and c) public awareness could be raised if government agencies would issue frequent text and video news releases about collections in the nation’s libraries, featuring them as part of the national heritage to be preserved for future generations, and providing suggestions about possible contributions of talent and funding from the private sector.

Regrettably, attendance at this session was not very high: there were no more than about 10 or 11 (including the speakers). The time of the session was most unfortunate—it was held on the very last day, during the very last set of early afternoon sessions, and it was the 714th of 721 sessions at the conference. Clearly the session was held to establish a possible agenda for future action, but the importance of the subject calls for more than one session, more panelists (including conservators), and a wider spectrum of points to consider. At the conclusion of the session there was no consensus on points to be considered for future MLA meetings, other than general discussion.

Librarians interested in furthering the exchange of ideas on databases in the humanities, preservation, or any other that might prove of mutual interest to academics and librarians, may submit proposals for special sessions at the next meeting (the deadline for the proposal of sessions is in April 1988). Details regarding the preparation of such proposals appear yearly in the preceding November issue of the PMLA. Hopefully by next year’s conference, teams of scholars, librarians, and preservation experts can be formed to contribute in even greater numbers to the annual deliberations. ■ ■