Modern Chivalry and the Case for Electronic Texts

Related Link: Modern Chivalry Website

Finding editions of particular literary texts for the purposes of teaching or research has always been a problem for literary scholars. Given the current proliferation of electronic versions of texts available on the World Wide Web, it is tempting to assume that the problem is solved. Yet most professors are reluctant to use these sites and do not often recommend them to students. In reflecting on the reasons for this phenomenon, the most obvious causes seem to stem from questions of authority, design, and a general lack of knowledge concerning what is available and where it can be attained.

In a recent search for electronic texts on the web, I found 92 different sites purporting to offer e-texts either free or at cost. More popularly read texts may be available at several sites, and those less often taught or read may not appear at all. Yet even for the most readily obtainable texts, the e-texts are underutilized in teaching and research. The problem is an age-old one: authority. The sites vary considerably in quality and quantity of texts available, and many of them noticeably lack information concerning the methods by which the texts have been generated, the editorial principles and choices made by the editor, or even the identity of the editor. Sometimes there is no "editor" per se; the text is generated through scanning by a computer technician from any available edition without consideration of which edition is authoritative by our own professional standards.

Although in some cases we may take our choice of text for granted in our researching and teaching lives, a well-established critical text is essential to both our teaching and our scholarship. Whatever our choice of theoretical bases, the text itself is the bedrock upon which we build both research and theoretical applications. Without a clear, authoritative text, we are unable to perform a proper analysis. The quest for a good, clear, authoritative text is not a new or rare occurrence for most literature scholars. The proliferation of digital libraries and e-publications in past years, however, has effectively widened our search area, extending it to the World Wide Web. For literary textual studies, as in other areas of research, questions of authority vest with historically traditional published versions. That is, academic study of such works begins with an already clearly established "authorized" text based on earlier scholarship for print publication.

"'Network textual authority' is sometimes embodied in persons [. . .] but is more frequently generated by all the logos, insignia, call signs, and other trademarked sonic and visual imagery that tell us instantly, not which person is speaking to us, but which major institution" (Streeter 1999). Institutional authority of web texts follows the pattern of respectability and reliability previously established in print texts. Scholars search for texts generated by reputable textual scholars and/or well-established presses, with authority reverting to these publishers based upon already established reputations for choosing authoritative scholars and attending to matters of copyright (Hughes). A few examples of editions with already clearly established historical authority are Norton Critical Editions, Riverside Editions, or Oxford University Press Editions. Textual studies scholars who have built solid reputations for carefully researched and detailed texts, such as Thomas G. Tanselle and others, represent a second line of authority for those in search of a "critical" edition.

Aside from these important questions of authority, a second problem discourages the use of electronic versions of texts: that not all sites comply with established standards of text production. Thus the quality of available texts can fluctuate wildly. Some of these e-published texts contain typographical errors and scanner anomalies that make them difficult to read, and often contain no structural breaks in the form of chapters or links to help users read them. Backgrounds can be dark, and font sizes small, making the already difficult task of reading online even more problematic. Their only advantage is that they are cheap, fast, and easy to produce. This being the case, professors are right to be skeptical of either using these texts or encouraging their students to do so.

The solution, then, appears to be entering into a dialog concerning the feasibility of these texts for specific purposes, deciding what we want and need from them, and then making sure that we get what we want. To this end, I would like to begin the dialog by demonstrating what I see as some of the advantages of the electronic text, using my own electronic version of Modern Chivalry, Hugh Henry Brackenridge's 18th century novel, as an illustration. At present, the project is half way to completion. Upon completion, the novel will be available at Crossroads, part of University of Virginia's Electronic Text Center. Part One and my own editorial notes went up last month, and I hope to have Part Two and the Introduction available by August 2002.

Publication History

A little background on the publication history of Modern Chivalry is probably necessary here in support of my own editorial choices, (or non-choices, as I hope to explain a bit later). The entire work was published serially before being collected into its novel form. The first edition of Part I, Volumes I and II, appeared in 1792, published by John McCulloch in Philadelphia. Volume III was published by John Scull in Pittsburgh in 1793. Volume IV did not appear until 1797, and Brackenridge published this volume through John McCulloch again in Philadelphia. For Volumes I and II of Part II, Brackenridge again shifted publishers: these two volumes appeared in 1804 and 1805 respectively, through Archibald Loudon in Carlisle, Pennsylvania. The last volume of Part II was published and sold by Johnson and Warner, Philadelphia and Richmond, Virginia, in 1815.

In 1815, Brackenridge began collecting and editing his work for a novel edition. At that time, he made changes to the volume and chapter system, streamlining it by shortening the number of volumes within the parts and merging shorter chapters into the ones preceding them. In addition, he made some minor editing changes (changes in vernacular spellings and punctuation) and others that appear to be more major. In one instance he revised the name of a particular character (from Miss Fog to Miss Vapor), and he edited out any references to “God” in favor of more secular language. In addition, he eliminated the majority of his initial introduction to Volume III of Part I, excising a long poem from it (approximately 60 printed pages). The 1815 and 1819 editions both state on the title page that these are the last editions for which Brackenridge had any editorial input—he died in 1816.

Since that time, the novel has appeared in several editions, the most inclusive being Claude Newlin’s in 1937 through the American Book Company. For his edition, Newlin chose to publish the work in its entirety from the serialized versions, proofing against the revised text of 1815 only for the purpose of correcting misprints in the first editions. One of the least inclusive is a recent (1965) version, Lewis Leary’s for the Masterworks of Literature Series. That edition reproduces only Part I. Leary’s editorial choice is based upon reproducing the narrative parts of the text (the story of Teague) without the more numerous and longer digressive chapters that form most of Part II. Leary’s version excises half of the original text by eliminating Part II. A later version, a collection of “narrative” sections of the text, appears in Daniel Marder’s A Hugh Henry Brackenridge Reader through University of Pittsburgh Press in 1970. Textual scholar G. Thomas Tanselle rightly points out that Marder’s version of the text makes no attempt to establish a critical text, or even a “best” text: his aim is to make available a portion of what he, as editor, considers the “best of” Brackenridge’s work to modern readers. Discounting the fact that the editor takes it upon himself to determine which “parts” constitute the “best” of Brackenridge’s work, he offers no information about the textual version from which he drew these parts; therefore, a scholar cannot assess with any degree of accuracy the authoritativeness of this version.

Here we can clearly see the amount of variation between three different editors with three quite different editorial strategies, and these differences in strategy produce radically different texts. Without intimate knowledge of the author and text, readers can unwittingly be misled in their ideas about Modern Chivalry itself, and about Brackenridge as an author, early humorist, and historian.

Features of the Chivalry Web

Marcus A. McCorison suggests that the eccentricities of Modern Chivalry’s publication history may account for its relative rarity in terms of complete sets available. After the many hours of eyestrain involved in textual transcription and proofing, I might tend to agree. However, I believe that there are several other reasons why Brackenridge’s humorous and irreverent rendering of the American Frontier post-Revolutionary War is seldom published and even less often read. The first problem is the considerable one of establishing a best text. Which ought a scholar to choose as the “critical” edition? On the one hand, the first edition (serialized) would be the historical, textual scholar’s choice, and was the driving force for Newlin’s 1937 edition. On the other hand, Brackenridge clearly made some major editorial revisions in 1815 in preparing the novel edition. An 1819 edition, the last for which Brackenridge had any input, includes Part II, and incorporates additional deletions, shifts in chapters, and other revisions. Since these revisions are more than cosmetic, a scholar might argue that he envisioned the novel edition quite differently than the original serialized publication. The second problem is the relative age and composition of the text itself. Its topical references are often anecdotal and unclear to those readers unfamiliar with the time period, and Brackenridge’s reliance on classical history and Latin maxims can also be difficult for modern readers. In addition to the contextual problems for readers, the novel is long, and few publishers are willing to expend that amount of page length for a novel that has not historically sold in large quantities.

In order, then, for Modern Chivalry to regain the place in the literary canon that it deserves, its editor must be able to somehow balance the needs of the modern reader with the demands of potential publishers. The first task is to establish a best text that does not compromise the integrity of the novel (as two of the previously mentioned editions clearly do) from the several possibilities. The serial publication of the work contains clearly valuable material that was excised in the 1815 and later 1819 edition, which complicates an editorial decision. The second task is then to establish the number and kind of annotations necessary to make the work accessible to modern readers. The final task is to negotiate a page length with a publisher that will make the text “profitable.” The overall task is a formidable one.

These very problems, however, make Modern Chivalry a good choice for electronic publication. Writing and publishing in a hypertextual environment offer flexibility and branching qualities impossible in print media. Jay David Bolter's explanation of "topographical writing" (Writing Space) and George Landow's articulation of hypertext theory (Hyper/Text/Theory) clearly demonstrate several advantages to web publication: the ability to offer several versions of the text simultaneously, allowing the reader to choose to read one version, or all of them; the ability to define the site architecturally, using linking branches to create a reader-friendly structure; and the ability to (virtually) eliminate the remaining problem of text size (page length).

One of the reasons that I took on the Chivalry project was that I wanted to use the text as one of the readings for my upper-division Early American literature course. Modern Chivalry is a particularly good example of the development of the early novel because it clearly demonstrates the American novel’s debt to its British predecessors. It is also valuable as a study of the development of American humor and to students interested in political and social issues in post-Revolution America. But the most readily available edition (and even this one was only marginally “available”) was the Leary edition, which contained only Part I. Thus, one could say that my quest for a useable edition became my current research project.

Because it was the most complete, I began my transcription with the Newlin edition as copy-text. Although it contains no annotations of any kind other than a small editorial note concerning his choice of copy-text as the first edition, it was the most promising. In proofing against the 1815 edition, however, changes to the text surfaced that Newlin had ignored in creating his historical version. It seemed clear to me that Brackenridge himself wanted the novel edition to take a different direction; it was equally clear that if I chose to use Brackenridge’s revisions, much valuable material related to his writing style, revisionary practices, and contributions to American humor would be lost. It is possible to keep both versions of the text in a critical edition by using appendices, bracketing or italicizations within a hardcopy text, but doing so would add considerably to the page length and the necessary apparatus often defeats creating a clear, readable text. An electronic version, however, can incorporate both texts without compromising clarity of reading. In designing the web page, I chose to use Brackenridge’s 1819 revisionary one as his final word on Modern Chivalry in its novel form, but created links within the text so that readers retain the ability to view changed or deleted material through a pop-up box. Using the box form allows readers to see the revisions if they choose to, without losing their place in the reading. In this way, students have access to a clear text, but scholars at the same time have access to both versions. In the case of the long, hudibrastic verse that originally made up the introduction to Volume III of Part I, the pop-up box was unwieldy, so I created a box that describes the change containing a link for those who might wish to read this 60-page section in its entirety. This flexibility of edition can be an improvement over hard copy texts, since page length ceases to become an issue and readers can effectively choose their own edition to suit the needs of both general reader and scholar.

Web publishing requires a rethinking of traditional means of navigation and orientation for readers. Since essentially the page does not turn, editors need to adapt the more traditional chapter, volume or section breaks to the web medium. Jakob Nielsen’s several web and hard copy articles clearly demonstrate that users read differently on screen than they do in a hardcopy book. The texts, no matter their length, are simply one big page as far as the web browser is concerned. When produced in this fashion, readers need to scroll until they reach the end. In the case of a short story or a poem, this may not present much of a problem, but for novels, this becomes an onerous, repetitive task. It tires the mouse-hand and the eye quickly, and does not make effective use of the web’s linking abilities.

For Modern Chivalry I was able to keep Brackenridge’s own structural volume and chapter divisions, making use of them for a more “natural” read, by adding back-links so that readers can move quickly between chapters or to the table of contents should the need arise to reread or find a particular section quickly. The other addition I made in editing the work was to add sentence-length “summaries” of the chapters to the contents page. While this could be considered inconsistent from a purist, traditional textual studies approach, it does not change Brackenridge’s own text, and I make it clear that this material is my own, and added as a means of orientation and navigation only. In addition, these chapter summaries remain consistent with a practice often used in eighteenth-century texts.

Once the text was established, the second critical task for Modern Chivalry was that of annotation. While most all scholars will agree to the necessity of this apparatus for general readers especially of very old texts like Modern Chivalry, there appears to be little agreement on how much and what kind of annotations best facilitate reading for a varied audience. In the case of Brackenridge’s novel, it is clear that the many Latin quotations and maxims need annotation for modern readers. While a facility with Latin and Greek was once the hallmark of a classical education, this is no longer the case. Brackenridge himself loosely translates some of them; for others, the careful reader can glean the meaning from context. But both general and scholarly readers could benefit from a good, clear translation. As a classical scholar of the times, Brackenridge was thoroughly familiar with ancient historians, Lucien and Thucydides in particular, and he makes reference to them frequently in his political analysis of the times. These authors are also less commonly read in modern times, and should be annotated for modern readers. Finally, many of the author’s references to locally popular Pennsylvania political issues and even some national ones would benefit from annotation for readers less well-versed in early American history.

Annotating a text of this length on the scale I have outlined could easily add from 20-60 pages to the length of a hardcopy edition, an addition that most publishing houses would be unwilling or unable to provide at a cost that would make the book salable. But adding these annotations to the web version would add nothing to the “cost” of production other than the time and energy of the annotator. The necessary apparatus can be supplied in the same manner as the textual variants, in pop-up boxes, without distracting the reader in the process.

The final feature of electronic texts I’d like to discuss is their searchability. While not all e-texts have this feature, it can be of invaluable assistance to scholars in their research. A careful reader will pick up recurring metaphors, images, and language within a text under study. When studying a short poem or story, good readers can probably be expected to locate all of the references within that text. However, in longer works such as epic poems, novellas, and novels, even the most astute and experienced reader may miss a number of such references or be misled as to the total number of such allusions, inferences, or word choices. If a scholar notes what she/he perceives to be a trend within the text, a search function can find all such references without the painstaking hours of study such an exhaustive search might take a reader.

Eric Johnson documents the accuracy and speed of such searches in “Electronic Shakespeare: Making Texts Compute,” and the later web version of his original article, “Electronic Texts and their Use for Literary Research.” His specific example is drawn from Jane Austen’s Pride and Prejudice. He notes that in a rereading of the novel, he “got the idea that the words ‘love’ and ‘affection’ were similarly used in a very precise way” (3). He also asserts that it is extremely difficult if not impossible for a human reader to collect all of the more than 100 such references, but that “a computer using the electronic texts of novels brings all such occurrences together in their context for a researcher to examine” (3). The value of a good search engine, then, is that it can do the drudge work of collecting each and every reference within minutes, saving the researcher’s eyes and sanity for the task of interpreting what she/he has found.

Expanding the Chivalry Web

Thus far, my editorial discussion has not strayed too far from the kinds of tasks performed by traditional text scholars, and the differences between a hardcopy version of Modern Chivalry and the web version are merely mechanical. However, at this time, I would like to suggest some less standard possibilities that could—and possibly should—be incorporated into a web text. The present version of part one that appears on the University of Virginia server is bare text only. That is, it contains only Brackenridge’s text and my editorial note. One might argue that I have “published” an incomplete version of the novel. In fact, what is available immediately is the “best” text that has been sporadically available for the past 35 years—albeit not the optimal version. Upon completion of the full text, it will be a best critical edition. If this were a hardcopy text, the most optimistic turnaround time on a new edition would be 12-18 months, but I will be able to FTP Part II to the server within minutes of its completion and can continue to add annotations over time. An electronic editor can make a text available to readers while at the same time completing the external apparatus.

Additionally, it is possible for readers themselves to "add" to the available text in the form of background material concerning the author, the text's critical reception, a bibliography of Brackenridge research, or interpretive essays/articles related to the text. Full discussions of reader-based texts of this type are available in Bolter's Writing Space and in Landow's Hypertext 2.0 and Hyper/Text/Theory. They encourage active reading of texts by students: effectively, students become reader/writers, "altering" the text by their own direct manipulations and contributing to the existing scholarly discussion in some cases. A group of students from my English A-360 class for the Spring, 2002 semester worked on one such project. With the help of a graduate student assistant, they helped me to proofread Part II and worked on an annotated bibliography that will eventually be included on the web. In this way, they are actively participating in ongoing scholarship in their field while still in school. Their names will be added to the web publication when the project is completed satisfactorily.

In one sense, this means that the project is never “finished,” in the traditional sense. However, in another, the project has unlimited capacity for expansion. I intend to use the text as a part of my Spring 2003 early American literature course because the full text will be available by that time. But a project like this one can also fuel scholarship for graduate or postdoctoral research. Upon completion of the annotations, which can themselves become excellent research projects for both undergraduate and graduate students, giving them valuable insight into researching in general, the page can then be expanded to include scholarly articles about the text. Many traditional hardcopy editions, most notably Riverside and Norton Critical Editions, contain such supporting material. Most often though, considerations of length dictate that these editions can publish only summaries of scholarly articles, whereas a web version, because it need not consider page length, can publish the articles in full-text. A true web of any text could be designed to include a significant amount of scholarship. Many hardcopy articles are presently available in full-text from databases like J-STOR, and could conceivably be linked to the text, although the obstacles inherent to this kind of linking are beyond the scope of the present study. It is not beyond imagination that in the near future, scholars might publish directly to such a site, which would make the full text under study available to readers of scholarly essays. Web texts of this kind and quality could then make good on the promises of Internet designers and providers: a worldwide database of easily accessible information at the click of a mouse.

Conclusions

Admittedly many roadblocks remain on the information superhighway. The metaphor is apt in more ways than one. Until we can resolve the thorny issues of copyright, usability, compatibility, and intellectual property, our “superhighway” will probably resemble the interstates in spring and summer, and be plagued with construction, lower speed limits, and blocked off lanes. But the intuitively obvious starting point is to be certain that the texts we deliver are clear, clean and authoritative. To make that first step a reality, textual scholars and researchers must be the ones to establish the texts for web use. Then we can be certain that as scholars and teachers we are standing on bedrock rather than quicksand.

• In particular, copyright issues pose the most formidable roadblocks to e-publication. Intellectual property rights, less tangible than rights to actual, physical property, have yet to be clearly defined and established.
• Existing tensions between "the reward for the creator and the need for the public to gain access" (Oppenheim 97) make e-text editions of print-based copyrighted material problematic. For a text such as Modern Chivalry, for which the hardcopy demand was low to flat for much of the twentieth century, this issue may well be moot; however, placing other, more popularly printed texts on the web involves costs of obtaining permission that vary widely in terms of time and cost.
• As yet, relatively little protection exists for material once it has been uploaded (Kling and Lamb 1996). Robert Kling and Roberta Lamb cite in particular the problems of obtaining the appropriate permissions to publish a recent edition of Owen's poetry in 1996. In this case, the costs of obtaining permission from the author forced them to consider other, less recent and perhaps less authoritative editions for the web. "Permission was available at a reasonable cost, and the edition [Day-Lewis's 1960s one] considered reasonably accurate" (Kling and Lamb 18). The italics here are mine, and I believe they underscore the problem under discussion clearly. Although a more recent authoritative edition was preferable, the web editors were forced to choose one less optimal based purely upon the costs involved in copyright issues.
• Clearances on copyright are also traditionally slow, and have become increasingly complex according to Kling and Lamb, and this cumbersome process often nullifies one of the web's most promising assets: the shorter turnaround time on a completed electronic text's availability to readers.
• While some permissions are granted for use of copyrighted material, particularly in the case of texts used for teaching and research, protecting such texts for pirating once on the web can be difficult. At present, steps may be taken, such as "watermarking" pages and clearly confirming the owners of the original copyright, but these steps are inadequate protection to the material's misuse (19).

Additional questions related to copyright issues remain:

• What constitutes the actual e-text?
• Do minor alterations of text and style effectively create a "new" or "different" version? How can editors protect sites from tampering?
• How can we guard against misrepresentations of authors' intended meanings in texts where the very flexibility can create such misreadings or misunderstandings?

Debate on these intellectual property issues so far has yet to define a consensus or resolve these issues.

In spite of the monumental problems still to be resolved, many sites are providing the kinds of texts scholars and serious literature students want and need to use. The Center for Electronic Texts in the Humanities (Rutgers and Princeton Universities) and the Georgetown Center for Text and Technology advise and catalog the production and distribution of electronic texts. The Text Encoding Initiative has devoted both time and effort to delineate exactly how the Standard Generalized Markup Language (SGML) can be used to describe the nature and features of an electronic text (Johnson). University of Virginia's Crossroads Web (which houses Modern Chivalry and many other texts ) is another excellent site. Below is a partial list of sites where other text projects can be found:

• Alex: A Catalogue of Electronic Texts on the Internet from Oxford University
• The Gutenberg Project from the University of Illinois
• Electronic Text Center from the University of Virginia
• The English Server from the University of Washington
• Project Bartleby from Columbia University, including the 1901 edition of Bartlett's Familiar Quotations
• Center for Electronic Texts in the Humanities
• University of Chicago Philosophy Project
• The Etext Archive
• Virtual Library of Virginia
• Project Libellus from the University of Washington
• The Dartmouth Dante Project
• Online Book Repository via The Online Book Initiative
• The Online Books Page from the University of Pennsylvania
• Online Books FAQ
• Skidmore Electronic Reading Room
• Making of America from the University of Michigan
• The Internet Classics Archive from the Massachusetts Institute of Technology
• Labyrinth Medieval Studies server
• CCAT's Text Libraries at University of Pennsylvania
• Cyberstacks science and technology reference resources
• Hypertext Fiction on the WWW
• The Master Works of Western Civilization
• Civil War Archival Collections
• Perseus Project from Tufts University
• Wiretap Electronic Text Archive

Links to all of these sites are provided by the Library of Congress Internet Resource page at http://www.loc.gov/global/etext/etext.html.

Even as we study and debate these important questions and problems posed by e-publication, educators and researchers continue to benefit from ready access to information provided on the web. According to John Unsworth, "we can expect that, whatever happens, in the larger cultural sphere, electronic scholarly publishing will alter our profession in significant ways [. . .] electronic forms and practices offer a new field of opportunities for theorizing signification, communication, literature, and culture." (3) He sees e-texts as "re-creating the basic resources of all our activities and providing us with revolutionary new tools for working with these resources" (3). In the years ahead, the availability of such texts will depend upon how these legal and moral questions are resolved. Obviously the debate will continue. But the potential benefits for scholarship e-publications present, especially in the case of a text like Modern Chivalry, make this debate and its outcome a worthwhile endeavor.

Works Cited

Bolter, Jay David. Writing Space: Computers, Hypertext, and the Remediation of Print. 2nd ed. Mahwah, NJ: Erlbaum, 2001.

Engell, John. "Brackenridge, Modern Chivalry, and American Humor." Early American Literature 22 (1987): 43-62.

Hughes, Lorna. Discussion Log. "Re: Electronic texts." 15 November 1993. http://www.ku.edu/~medieval/melcher/matthias/old/log.started931115/mail-5.html (1 May 2002).

Johnson, Eric. "Electronic Texts and Their Use for Literary Research." 1993. http://www.dsu.edu/~johnsone/etextand.html (17 February 2000). First published as "Electronic Shakespeare: Making Texts Compute." Computer-Assisted Research Forum, 1.3 (Spring/Summer 1993): 1-3.

Kling, Robert, and Roberta Lamb. "Analyzing Visions of Electronic Publishing and Digital Libraries." 1996. http://www.slis.indiana.edu/kling/pubs/EPUB6.htm. (1 May 2002). Also available in Scholarly Publishing: The Electronic Frontier. Ed. Gregory B. Newby and Robin P. Peek. Cambridge, MA: MIT Press, 1996.

Landow, George P. Hypertext 2.0; Being a Revised, Expanded Edition of Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins UP, 1997.

—. Hyper/Text/Theory, Baltimore, MD.: Johns Hopkins University Press, 1994.

McCorison, Marcus A. "The Rarity of Modern Chivalry." Proceedings of the American Antiquarian Society 85 (1975): 309.

Oppenheim, Charles. "Copyright in the Electronic Age." Textual Monopolies: Literary Copyright and the Public Domain. Ed. Patrick Parrinder and Warren Chernaik. Oxford: Office for Humanities Communication Publication, 1997. 97.

Rice, Grantland S. "Modern Chivalry and the Resistance to Textual Authority." American Literature 67.2 (June, 1995): 257-81.

Streeter, Thomas. "Reflections on Textual Authority Beyond the Printed Page." 1999. http://english.uq.edu.au/mc/reviews/features/ejournal/authority.html (1 May 2002).

Tanselle, G. Thomas. "Two Editions of Eighteenth Century Fiction," Early American Literature 6 (1971): 274-83.

Unrue, Darlene Harbour. "Brackenridge's Modern Chivalry: A Reassessment." History and the Humanities: Essays in Honor of Wilbur S. Shepperson. Ed. Francis X. Hartigan. Reno: U of Nevada P, 1989. 271-83.

Unsworth, John. "Electronic Scholarship, or Scholarly Publishing and the Public." 1994. http://www.iath.edu/~jmu2m/mla-94.html (1 May 2002). Appears in The Literary Text in the Digital Age. Ed. Richard Finneran. Ann Arbor: U of Michigan P, 1995.