College and Research Libraries ity of reference service, about personal qualities and behaviors, the people who offer and use reference services, it has less to say than it might. Unlike its predeces- sors, it is reticent about the affective as- pects of service, the interpersonal dy- namics, pleasures, and satisfactions of the work; in general, the hard glint of clini- cal abstraction lingers in its gaze. Thus, no Bopp and Smith librarian would feel the "interest, amounting to fascination, [the] thrills, amounting at times to ec- stasy" that Wyer sees as the librarian's occupational reward. If capable of it, the Bopp and Smith information hound would raise an ironic eyebrow at Hutchins's narrative of a young librarian who returns "flushed from the periodical indexes" to the desk, where she is flus- tered to encounter a student whom she and colleagues are transforming into a library-competent scholar; nor would the student, days later, feel a pang of disap- pointment in not finding her at her post. Those who lament the absence of theo- retically minded "dead Germans" in li- brarianship will find no comfort here ex- cept perhaps in the rather eccentrically cast chapter 10, which, with chapter 1, might have paid more attention to the economic and political trends that cur- rently threaten egalitarian library service. The editors might have reconciled chap- ter 6 on instruction, and indeed the en- tire history of reference librarianship, with the statement in chapter 7 that "[r ]eference librarians rarely see them- selves as educators." An uneasy tension pervades the text's participation in the transition from print to electronic ser- vices. OCLC and RLIN are still quaintly labeled "nontraditional" reference sources; cards introduce bibliographic control and printed pages periodical in- dexing; the encyclopedia chapter dis- cusses multimedia but gears search strat- egies to printed versions. Granted the difficulties of using electronic interfaces to demonstrate these points, might the text not be reconceived to do so? Book Reviews 93 Bopp and Smith situates its workman- like bulk squarely in the century-old United States tradition of reference ser- vice and is eminently usable in all the ways its predecessors have been. Because Katz (new edition scheduled for 1996) covers similar territory in similar ways, personal preference may ultimately de- termine whether a general reference course requires one or the other. Minor differences of emphasis aside, Bopp and Smith is rather more conscious of itself as a survey of the reference literature, whereas Katz, like Wyer and Hutchins, is more interestingly written. Bopp and Smith smells rather of earnestly cheerless "learning sessions" in airport Ramadas, whereas Katz smells a little more of the reference desk-Robert Kieft, Haverford College, Haverford, Pennsylvania. Literary Texts in an Electronic Age: Schol- arly Implications and Library Services. Ed. Brett Sutton. Urbana-Champaign, Ill.: Graduate School of Library and Information Science, Univ. of Illinois, 1994. 207p. $25. (ISBN 0-87845096-3). This book, which publishes the papers presented at the 1994 Clinic on Library Applications of Data Processing, focuses on fairly recent developments in the area of electronic texts. Its attempt to address the impact of these developments on both scholarly research and library services is not always successful. Although the eleven papers are appropriately wide- ranging, their quality is very uneven. Because a significant number of librar- ies have started to provide access to elec- tronic texts in a serious fashion, a thor- ough examination of the impact of these texts on library services has recently be- come possible and necessary. As a result, libraries have begun to grapple with a number of issues, such as the develop- ment of selection criteria, licensing and copyright regulations, changes within the MARC format to allow for description of and access to electronic/internet re- sources, and fundamental decisions on 94 College & Research Libraries how electronic texts would be made available to the scholarly community. The thinking that has been done recently in this area and the various projects that have resulted from it have clearly ad- vanced our understanding of these is- sues. A number of the better papers in this collection reflect this progress. Susan Hockey offers a succinct histori- cal overview of electronic texts in the humanities since the 1940s, emphasizing the necessity of descriptive markup that recognizes the structural components of texts and demonstrating the ways in which some of the lessons learned have been applied to digital imaging. Hockey quite rightly deplores the lack of progress in the development of analysis tools . Such tools currently constitute a vital missing link for scholars who wish to start using these texts in a rigorous and sophisticated fashion. C. M. Sperberg- McQueen discusses the text encoding initiative (TEl) that has offered a set of guidelines for encoding literary and lin- guistic texts using standard generalized markup language (SGML). A fascinating paper by John Price-Wilkin describes one approach that has used standards (in- cluding use of the TEl guidelines to build SGML-compliant text corpora) and open systems to create a wide-area electronic text service. This approach ensures the reusability of texts, the possibility of ap- plying a variety of analysis tools, and the separation of data and software. Further- more, the textual analysis systems that implement this approach at Virginia and Michigan allow for speed, phrase search- ing, truncation, indexing without stop words, structure recognition, and fine- level result displays (KWIC). These are requirements that many text retrieval sys- tems cannot deliver but which are essen- tial for the researcher. The two other papers that report on experiences with electronic texts within libraries are severely limited by a nuts- and-bolts, "how-we-did-it" approach that is not particularly enlightening. In January 1996 her paper on the information arcade at the University of Iowa Libraries, Anita Lowry does not achieve the distance from the project that would be necessary to draw valuable insights. Mark Day's too- extensive reportage on the electronic text resource service at Indiana University is a peculiar combination of a somewhat naive plunge into postmodern theoreti- cal waters combined with personal and institutional history writing that ulti- mately does not advance our thinking on the topic. Three papers introduce issues that go beyond the delivery and analysis of elec- tronic texts. Rebecca Guenther gives a very thorough overview of some of the problems that arise when providing bib- liographic control of and access to elec- tronic texts. She offers a history of some of the important modifications to the MARC record, including the electronic location and access field (856). Guenther also touches upon the work done by the Internet Engineering Task Force (IETF) on uniform resource identification, and by the Library of Congress on the electronic cataloging in publication project. The compatibility of SGML and MARC are also explored, suggesting the possibility of some promising interactions between the two standards. The copyright issues that all Internet users are confronted with are visited by Mary Brandt Jensen in a readable and realistic fashion. She pro- vides for the uninitiated a fine introduc- tion to a complicated legal combat zone. Finally, a paper by Lorrie Lejeune from the University of Michigan Press argues for inclusion of publishers in their tradi- tional role as guarantors of quality and credibility within the new electronic en- vironment. Although this role is defi- nitely desirable, the shift from print- based to electronic publishing is not one that centers only around secure online cost recovery. Lejeune glosses over some of the basic economic shifts that are al- ready taking place: e.g., the replacement of a purchase transaction by a license or royalty transaction. Such shifts have pro- found implications for libraries, chang- ing the cost factors of prolonged or re- peated consultations, as well as the lo- cus of ownership of the artifact. Nor are issues of cooperative collection develop- ment or interlibrary use addressed. One would also hope, at least from a library perspective, that the cost of perpetuat- ing the role of the publisher as guaran- tor of quality is not decided by asking, "How much will the consumer pay?" but rather by asking, "How much does it cost to perform the functions of high-quality peer review and editing?" Conversely, the scholarly implications, of incorporating electronic texts in hu- manities research are far from clear. First of all, relatively few scholars have used electronic texts and textual analysis sys- tems extensively in their research; and second, even for these electronically so- phisticated researchers, the querying of electronic texts has only informed one particular aspect of the research. Thus, although some researchers might enter the proclaimed electronic age at certain moments, they are not full-time inhabit- ants. Given this state of affairs, it is per- haps too early to have valid pronounce- ments on this topic. At any rate, it ap- pears that the scholars who were invited to this conference were not prepared to do so. Jay David Bolter's keynote paper re- counts the history of writing as a move- ment from the printed word-a stable representation governed by authorial or corporate control guaranteeing the accu- racy and fixity of the text-to the elec- tronic text, displaying qualities of fluid- ity, multiplicity, and dispersed control. Bolter asserts the somewhat dubious point that authorship and writing in a print environment have been creative ac- tivities, whereas in a hypertextual envi- ronment they've become merely connec- tive. Bolter expands on this assertion by claiming that "copyright is incompatible with ... electronic writing" because copy- Book Reviews 95 right theory is embedded within the no- tion of author as creative agent. Although his analysis is theoretically sophisticated, it really does not move the reader beyond generalities that have by now become daily journalistic fare. James Marchand's paper, which abruptly re- turns us to techno nitty-gritty, is an im- pressionistic and eclectic account of a scholar's trek through technology. Al- though Marchand correctly identifies cer- tain problems, such as the lack of stan- dardization and of consideration of user needs, these observations are clouded by generalized statements that are contradic- tory or overly simplistic. Robert Alun Jones advocates effective collaboration between scholars and "tool developers," which he illustrates using three instances of such cooperation at his home institu- tion. It would have been extremely useful to have included in this collection anum- ber of papers by scholars who have ac- tively used the electronic text services described by Price-Wilkin, either at Vir- ginia or Michigan. Scholars with solid experience in this area could have ex- plored the possibilities and limitations of these systems; the ways in which elec- tronic texts can further current research; new areas of humanities research that now can be effectively tapped by using electronic texts; the searching capabilities and effectiveness comparing the SGML- encoded text corpora to the ARTFL data- base (Chicago) of French texts that does not use SGML; the usefulness and/ or limitations of "minim a l" structural markup and the possible need for more sophisticated but expensive higher-level markup; innovative ways of incorporat- ing electronic texts in day-to-day teach- ing, their requirements of and needs for analytical tools; and the impact of these systems on humanities research method- ology. An in-depth and thorough examination of all these issues is needed. Some of the papers in Literary Texts in an Electronic Age 96 College & Research Libraries are valuable starting points, but the book as a whole fails to accomplish this goaL- Kurt De Belder, New York University. Collard, Claude, Isabelle Giannattasio and Michel Melot. Les Images dans les bibliotheques. Paris: Editions du Cercle de la Librairie, ยท1995. 390p. (ISBN 2- 7654-0577-8, ISSN 0184-0886.) All the world, by some accounts, is fast becoming images. Text has long been digitized. Sound has long been digitized. Visual images themselves, which have been available as bits as long as have the others, now at last are becoming avail- able to the general online world thanks to the World Wide Web. The end result of all this effort has been the creation of images: images of text, images of other images, multimedia applications using sound and text and graphics, and much more. If nothing else, the hours that we all spend now in front of video screens are making us very aware of images. There is an accompanying suspicion/ however, that a world preoccupied with image is a world become superficial. An image, some feel, somehow bears less validity or significance than "reality." Much criticism of the entertainment value of images stems from this suspi- cion. The hold of television and video games on the consumer imagination is criticized for contributing to such super- ficiality: entertaining images that should-but many feel do not-have something more real underneath. The controversy that accompanies the collection of images, particularly their collection by an icon of established cul- ture such as a library, is more understand- able if it is considered in conjunction with this suspicion of images. Libraries peren- nially have assembled vast image collec- tions: illustrations accompanying text inside printed books and illuminated manuscripts, images standing alone as- sembled into books or preserved in vari- ous other media. By the modern, digital definition, library contents-text and January 1996 all-might even be seen to be nothing more, or less, than giant collections of images. Michel Melot presents a masterful analysis of the many issues involved in this complex and convoluted picture in Les Images dans les bibliotheques. Melot is the former curator of the prints collec- tion of the Bibliotheque Nationale, was the first head of the Pompidou Center's Bibliotheque Publique d'Information, and currently is President of the Conseil Superieur des Bibliotheques. He is an ac- complished author and a recognized ex- pert on prints and images. This book, which contains the contributions of three authors, is at once a handbook of library procedures for the treatment of images, an essay on the particular French ap- proach which views library problems as processes rather than objects (French li- brarians think of documentation and col- lection more than they do of books and serials), and, thanks to Melot's essays, a trenchant piece of sociological and semiological investigation into the pre- cise meaning of the term image. Melot's introduction sketches the dis- tinctions that others have drawn around the concepts of image, sign, language, and writing, and the differences and re- lationships he sees among them. His eru- dition on the subject is impressive: fans of semiology, structuralism, linguistics, and cognitive studies will not be disap- pointed, yet readers unfamiliar with or usually uninterested in these arcane ar- eas also will have their understanding of "images" much deepened by Melot' s analysis. The first part of the book, "Documents and Their Users," covers image repro- duction, uses and users, and "the great collections." The second part of the book, "Managing the Fixed Image," addresses the practical topics: collections, conser- vation and restoration, reproductions, documentation, communication and ser- vices. The book's third part concerns the particular problems of motion pictures: