Collecting Full-Text CD-ROMs in Literature: Theory, Format, and Selection

COLLECTING FULL-TEXT CD-ROMs FOR a humanities library requires an examination of literary research and the way literature is taught. Radical changes in both theory and technology expanded the sorts of texts literary scholars study. They also revived and reinvented textual scholarship and editorial reconstructions. This means a CD-ROM is worth purchasing if it provides material about the historical context of the literary work, allows for the comparison of successive editions and facsimiles of earlier editions, includes critical work, and points to connections to other arts. At the same time, standard selection criteria for monographs, such as the authority of the editors and good production values, still apply. CD-ROMs should also provide additional “value-added’’ capabilities, such as notetaking, printing, and bookmarking abilities, to justify the cost and the decision to add them to an existing print collection. A checklist of both literary and technical criteria for selection is included. INTRODUCTION The works of Shakespeare, of little-known writers from the Spanish Caribbean, of Miguel de Cervantes, and many others are now available in electronic formats. Selectors in this area however must ask pointed questions about why they would purchase the CD-ROMs, which are often quite expensive. It is hard to imagine that there is a college or university library that does not already own at least one copy of the complete works of both Shakespeare and Cervantes. In addition, the discipline of literature is of Roberta Astroff, Arts and Humanities Library, Penn State University, University Park, PA 16802 LIBRARY TRENDS, Vol. 48, No. 4, Spring 2000, pp. 769-782 02000 The Board of Trustees, University of Illinois 770 LIBRARY TRENDS/SPRING 2000 course firmly rooted in print culture, and only a tiny percentage of its researchers are exploring hypertext writing. As McGann (1995) noted, “the literature we inherit (to this date) is and will always be bookish” (p. 1).So why duplicate the holding by purchasing the electronic version? Is it duplication? Which full-text CD-ROMs in literature are worth collecting? Why choose CD-ROM publications instead of Web-based publications? This article is the first in a series of articles about the integration of literature in electronic formats into library collections and university courses. Future papers will discuss introducing faculty to the new formats, working with teaching faculty to redesign coursework, and the integration of full-text CD-ROMs into subject-specific library instruction. This article will examine how selectors in literature need to take into account how professors teach literature, the type of research they do, and the type of assignments they give students when collecting full-text CD-ROMs for a humanities library. It is within this context, after all, that selectors in literature have to establish selection criteria and develop plans for promoting the use of these new resources. APPROACHES TO LITERATURE TEXTS AND ELECTRONIC The wide range of current theories and practices in teaching literature and the conflicts among them have become notorious as the “culture wars.” Guerin, Labor, Morgan, Reesman, and Willingham (1999) note that “since the mid-1960s we have witnessed a veritable explosion of critical theories, along with a radical expansion and revision of the literary canon” (p.xii). The foundations of traditional literary studies have been challenged by this explosion. For example, traditional textual criticism, defined as the identification of the most authentic text, has been confronted by critical approaches that dethroned authorial intention and challenged the concept of “authenticity” (Greetham, 1994, p. 8).Analyses based on biographical and psychological studies of the author have been countered by complex understandings of the social nature of meaning, theories of polysemy, and a rethinking of the active role of the reader. Nevertheless, the subfields within textual criticism, which include scholarly editing and historical and textual bibliography, provide important histories of the texts scholars study. Similarly, genetic criticism or source research uses related materials such as authors’ manuscripts and notebooks to map the development of a text (Guerin et al., 1999, p. 311). These histories detail the seeming contradiction of the mutability of written and printed texts. The history of specific works details omissions, errors, author’s changes, and all the modifications that befall a printed work even while historians of communication technology point to how print technologyfixed texts (Eisenstein, 1979, pp. 80-88).In this regard, despite ASTROFF/COLLECTING FULL-TEXT CD-ROMS 7’71 a history of relying on authorial intention, psychology, and biography, textual and genetic criticism produce studies that can connect with the work done within cultural studies, at least in terms of the materials used. The divisions between traditional textual criticism and criticism based in cultural studies and deconstruction are being blurred as textual scholars confront the challenges mentioned earlier. The set of approaches known as cultural studies has profoundly modified literature studies. Cultural studies ignores or erases boundaries between disciplines, texts, genres, and earlier cultural distinctions such as elite and popular cultures. A professor of literature will now look at narrative and representation in novels, films, fashion, science, news, and other forms of discourse. In addition, researchers taking a cultural studies approach focus on the materiality of culture, the means of production of cultural work, and the institutional practices of literature and its disciplines. Guerin et al. (1999) note that the “new historicism” of the 1980s, influenced by the writings of Michel Foucault, Jean-FranCois Lyotard, and Frederic Jameson (p. 240): concerns itself with extraliterary matters, including letters, diaries, films, paintings, medical treatises, etc. It looks for an opposing tension in a text, then for an opposing tension related in history. New historicists seek “surprising coincidences” . . . that may cross generic, historical and cultural lines previously maintained, highlighting unsuspecting lending and borrowing of metaphor, ceremony, dance, dress, or popular culture. (Veeser, pp. xii, 248) The specific social and cultural contexts of the work, such as contemporaneous mortality rates, the economic structure of the publishing industry, and the social structures of race, gender, and class, are seen as actively shaping the literary work. The work’s intertextuality thus extends beyond its relations to other literary texts to other forms of discourse. At the same time, “textual and editorial work are once again being seen for what they are and have always been: the fundamental ground for any kind of historically oriented intellectual work (McGann, 1996,p. 2). Despite the very real differences in theoretical foundations and despite jeremiads about the absence of literary works in today’s literature departments, a quick look at current research and course syllabi across the literature departments at a research university showed definite patterns and commonalities significant to librarians. First and most obvious is the centrality of a primary text, even as theorists argue about the definition of “text.” In other obvious patterns, researchers need and create concordances; explore the relations between music, art, politics, and literature in specific eras; and compare translations, editions, folios, and manuscripts. Graduate students are asked to evaluate editions, compare them, and create their own critical (and at times electronic) editions. Undergraduates are asked to explore events and conditions of the author’s time 772 LIBRARY TRENDS/SPRING 2000 period and to identify trends in art and music contemporary to the literary work being studied. The electronic text format can be very useful in each of these approaches to literature. As Ellis (1993) noted: “In the humanities, . . .each new paper directly or indirectly works upon a ‘core’ or primary text. Thus as far as the humanities are concerned, core texts (in full-text) must be included in the electronic universe to adequately represent [the] structure of the field” (p. 26). Scholars have used machine-readable texts to produce concordances since the 1960s. Later formats that allowed for Boolean searching made electronic texts useful to research in stylistics, linguistics, and lexicography (Ellis, 1993). Now, the quality of image reproduction on CD-ROMs and some World Wide Web sites means facsimiles of original manuscripts can be used for research purposes. The ability to display two or more editions or translations side by side facilitates close textual analysis that can otherwise require travel to archives. Unsworth (1996) notes that: We can expect to see increasing interest in editing (including the theory of editing), in bibliographic and textual scholarship,in history, and in linguistic analysis, since these are areas in which the new technologyopens up the possibility of re-creatingthe basic resources of all our activities and providingus with revolutionary tools for working with these resources. (p. 5) In addition, the number of documents each CD-ROM can contain means that each disk can be a collection or archive-of literary texts or of texts and related materials-rather than an individual title. M i p e l de Cewantes (GonzPlez Echevarria, 1998), a full-text CD-ROM produced by Primary Source Media, for example, is essentially a Cervantes archive. It contains twenty-three critical editions of works by Cervantes or selections from those works, six translations, seven facsimile reproductions, and reproductions of seven first editions. The CD also provides approximately 180 illustrations, selected from books and the fine arts, including thr Iconografia de Don Quajote by Francisco L6pez Fabra. The illustrations reproduced in the IconopaJa, first published in 1879, are arranged to follow the sequence of the narrative. GonzPlez Echevarria (1998) notes in his “Introduction to the Illustrations” that the oldest image of Don Quijote that L6pez Fabra found and reproduced dates from 1618. One dating From 1613 starts the collection of illustrations on the disk. The availability of the images on this disk is significant. While L6pez Fabra may have been the first to bring together illustrations of Don Quijote into a book devoted to the subject, the topic is still of interest, with recent studies compiling representations in artefacts of contemporary popular culture and analyzing the relationships of Cervantes, Raphael and the Classics (Allen, cited in GonzPlez Echevarria, 1998; de Armas, 1998). ASTROFF/COLLECTING FULL-TEXT CD-ROMS 773 Another full-text CD-ROM produced by Primary Source Media, Literature of the Spanish Caribbean (Benitez-Rojo, 1998), brings together over 20,000 pages in fifteen genres, ranging from journals and autobiography to fiction to medical works, written in Cuba, Puerto Rico, and the Dominican Republic between 1492 and 1900. Users can choose to order the table of contents by genre, date, author, or title, which facilitates various types of comparison and analysis. Researchers can also search for keywords and phrases across the collection of documents, a useful tool for the researcher looking for discursive elements shared by documents of a particular era. While all academic libraries may have a copy of the Quijote, few have a collection of Caribbean documents and literature to match that of this CD-ROM. Similarly, Chadwyck-Healey’s Teatro Espafiol del Siglo de Or0 [Spanish Theater of the Golden Age] brings together more than 900 pieces written by sixteen dramatists of the sixteenth and seventeenth centuries, including all front matter. The editorial board chose first editions when possible. Their supporting documentation says that no library contains a complete collection of this material. The selection of such disks then comprises significant development of most Spanish-language collections. SELECTION CRITERIA The growing literature on how to select CD-ROMs for purchase provides useful general advice for librarians (Bosch et al., 1994). For example, Nicholls (1993), citing LaGuardia and Huber (1992), says the librarian’s criteria for a good CD-ROM product are ease of installation, reasonable pricing, good technical support, and reasonable licensing restrictions (pp. 41-43). But most of the existing literature about CD-ROMs for librarians refers to CD-ROM indexes. Lowry (1992) identified six criteria for electronic text, “text quality, software, markup, medium, documentation, and price” (cited in Ellis & Fisher, 1997, p. 20). Ellis and Fisher explore how complicated these criteria can be in practice. In addition to negotiating the fundamental technical issues, the literature librarian also has to apply selection criteria that are identical to those used when selecting books. The selector is evaluating the CD-ROMs as literary texts and as literary archives. And since the library often already owns the print version of these works, the CD-ROM also needs to provide additional features-i.e., “value-added” features, beyond simply the reproduction of the text, that will be useful to the study of literature as described earlier. Criteriafrom the Perspective of Literature Studies In order to choose CD-ROMs wisely, the selection librarian has to identify the practical uses the CD-ROM offers in this context. This also helps clarify selection criteria. The first and perhaps most obvious 774 LIBRARY TRENDS/SPRING 2000 criterion: Librarians should look for authoritative editors, important editions, and good production values. As Murphy (1998) notes, “hypertext makes possible the . . . complicated juxtaposition of materials” appropriate for close historicist and textual scholarship (p. 411). But what is possible or imaginable is not always available on a particular CD-ROM. Different editions or facsimiles of manuscripts and later print editions can be displayed side by side on a computer screen but only if the editors of the electronic product included those editions or chose to design such possibilities. Several online sites and CD-ROMs that provide the works of Shakespeare do not include different folios or editions nor even any discussion about the edition made available (Murphy, 1998, pp. 411-12). In fact, Murphy quotes the opening screen of the MIT online Complete Works of William Shakespeare as telling users not to “worry” about variations in Shakespearean texts. The Voyager Mucbeth (1994) addresses issues of variations in critical articles included, but the only text of the play that is provided is a modernized variant. Murphy (1998) notes though that “the past decade has witnessed an increasingly keen awareness of the importance of variations among early editions of Shakespeare” in print editions, and this awareness is now visible in some electronic publications as well (p. 412). The Chadwyck-Healey Editions and Adaptions ofShakespeare (1997), available on CD-ROM and via Chadwyck-Healey’sLiterature on Line series, includes a wide selection of the early quartos, the first folio, the major eighteenth-century editions, the apocryphal texts added in 1664, and sequels and adaptations created by other authors (Murphy, 1998, p. 413). No twentieth-century editions, though, are included. Editors often confront copyright issues when they want to include significant contemporary editions and thus only include older versions. Murphy (1998) compares the Chadwyck-Healey Shakespeare to the Arden Shakespeare CD ROM. He praises this CD-ROM for a remarkable range of original and supporting materials provided, as well as a useful search facility and flexible screen layout. But he notes that the choice of the Arden 2 editions, rather than the Arden 3, has been controversial, in part because a great many of the volumes included in the Arden 2 date back to the 1950s. Murphy notes: Most of these editions predate the critical movements that have served to shape literary scholarship in the last decade or sopoststructuralism, cultural materialism, new historicism, feminism, queer theory, and so on-and thus have a rather quaintly old-fashioned cast to them . . . . [Tlhe texts in this series are (inevitably) for the most part uninflected by the theoretical concerns that have given rise to this undertaking in the first place. The Arden CD scems very much a product of a post-revisionist, post-structuralist textual culture in which great stress is laid on textual variation and pluralism. ASTROFF/COLLECTING FULL-TEXT CD-ROMS 775 And yet the texts that lie at the core of the package remain ineluctably unaware of the deconstructive dispensation that is yet to come. (p. 415) Nevertheless, at times, little information about the contents of a disk can be found on the box or even in the accompanying documentation. Primary Source Media’s documentation for both the Major Authors series and the Literary Eras series provides useful instruction on how to navigate through the work but says nothing about the content. In contrast, Chadwyck-Healey’s Teatro Espafiol del Siglo de Oro [Spanish Theater of the Golden Age] includes a complete printed spiral-bound bibliography as well as one on the disk itself. This is obviously useful both for the selection process and for promoting the use of the disk by library patrons. At the same time, Primary Source Media’s printed documentation of the disk’s interface comes in both English and Spanish, while Chadwyck-Healey’s print documentation is in Spanish only. During installation, the Teatro Espafiol del Siglo de Oro offers the user the option of installing in English or in Spanish. While it can be assumed that the Teatro Espafiol del Siglo de Or0 will be used by patrons who read Spanish, some librarians who are given the responsibility of collecting in literature are required to select material in languages they do not read. The library’s technical support staff, who might be loading these databases, might also be unable to read on-screen instructions in Spanish. Recently a technical support staff member came looking for the librarian for Spanish since he was “pretty sure” he knew what the Spanish screen instructions for the dictionary he was loading meant but wanted confirmation. Most CD-ROM producers have been willing to let librarians test the disks on the library’s machines. Those explorations by the selectors, and reviews of the CD-ROMs, appear to be the only way to really examine the content of some full-text CD-ROMs. The production values of course can only be noted if the CD-ROM can be loaded and tested on the library’s equipment. Value-Added Features The advantages of an electronic format go beyond the reproduction of many documents in a small space. The current capabilities of electronic formats provide new tools for scholars of literature. Thus, when librarians test full-text CD-ROMs, they need to have a set of criteria for these valueadded features in addition to literary ones. Some primary considerations involve the search capabilities of the disk, its display capabilities, and provisions made for annotation and glossing. Search Capabilities. It appears that all CD-ROM search functions allow for keyword, author, and title searches. The best search functions on fulltext CD-ROMs also permit Boolean searching, meaning “and,” “or,” and nested searches. It is also helpful if the user can select what sections of 736 LIBRARY TRENDS/SPRING 2000 text-front matter, names of speakers, or stage directions in dramawill be searched. Researchers tracking the development and uses of discursive elements also need phrase and “within” searching. On disks that include translations as well as editions in the original language, multilingual searching is needed. On the Literature of the Spanish Caribbean (Benitez-Rojo, 1998), for example, combining such a search using an “and” command would be much more efficient and less frustrating than having to do separate searches for “slavery” and “esclavitud”or “sugar”