Introduction

Project-based learning has figured prominently in discussions of progressive pedagogy for several decades, and according to Phyllis Blumenfeld et al. as well as Laura Helle, Päivi Tynjälä, and Erkki Olkinuora, its roots reach back to the educational writings of American pragmatist philosopher John Dewey. (For more information see also the critical discussion by John Garrick and Stewart Clegg, who warn that project-based pedagogies can be "too wedded to instrumental desires for performativity" [119]. A project course is, perhaps, less susceptible to such a charge than a project-based curriculum). While definitions of project-based pedagogy vary, most display the features articulated by Kenneth Adderly in 1975 and reiterated by Helle, Tynjälä, and Olkinuora in 2006:

(1) [projects] involve the solution of a problem; often, though not necessarily, set by the student himself [or herself];

(2) they involve initiative by the student or group of students, and necessitate a variety of educational activities;

(3) they commonly result in an end product (e.g., thesis, report, design plans, computer programme and model);

(4) work often goes on for a considerable length of time;

(5) teaching staff are involved in an advisory, rather than authoritarian, role at any or all of the stages – initiation, conduct and conclusion. (288)

In the university setting, professional disciplines like Medicine (Barrows) and Engineering (Mills and Treagust) have been at the forefront of project-based pedagogies, while traditional academic disciplines like English have seemed to lag behind.

To some extent this image of the stuffy traditional literary education is a misconception. After all, every academic essay is a concrete product in whose preparation the student exercises independent critical judgement, resolves a set of problems, develops a demonstrable thesis, and (ideally) invests considerable time in research, composition and revision. Hence even the most traditional literature course is more project-oriented than one finds in disciplines where quizzes, mid-terms and final examinations comprise the main assessment tools. But there are important criteria for a project, and important benefits of the project approach, where the traditional literary-critical essay falls short. Most obviously, collaboration is customarily prohibited in undergraduate and graduate student essay assignments. More significantly, most traditional essay assignments fall short of the "authenticity" called for in many discussions of project-based pedagogy.

Though definitions of authenticity in learning activities vary, most insist on "a similarity between the structured learning activities and some meaningful context for that activity" (Barab, Squire, and Dueber 38; Petraglia; Radinsky et al.; of these, Radinsky goes further, distinguishing between approaches to authenticity based on "simulation" of "real world" activities, and approaches based on "participation" in those activities. Since the "real world" of scholarly editing is, predominantly, the university, the approach described here might be seen as a hybrid of the two approaches). Where the learning activity in question is writing, or otherwise preparing a text for readers (as in preparing a scholarly edition), reading provides the "meaningful context": access to readers through some form of publication is the key criterion for authenticity. The course essay is inauthentic in that it seldom reaches (or even envisions) a reader beyond the instructor who grades the paper. A digital editing project published on the World Wide Web using HTML allows students to experience the benefits of project-based pedagogy—collaboration, original research, independent decision making, and preparation of a concrete product with real-world usefulness. In what follows, we describe one iteration of such a course, and the editions it produced, with particular emphasis on the issues of representation that emerged and the students' technological and editorial resolution of those issues.

Course description

The assignment for the 13-week class of eight honours English students was to create electronic scholarly editions of two seventeenth century texts, the anonymous Eighth Liberal Science: or a New-found Art and Order of Drinking [1650], a mock courtesy book for drunkards (electronic scholarly edition seen online here), and Edward Whitaker's Directions for Brewing Malt Liquors [1700], a home-brewing handbook (electronic scholarly edition seen online here). Students worked with the digital facsimiles available through Chadwyck-Healey's Early English Books Online database (EEBO). In this instance the texts were selected by the instructors, though previous 26-week versions of the course had required students to conduct preliminary research, prepare proposals for texts to be edited, and make a final selection. Since neither of the texts handled by this team had been published in a modern edition, transcribed, or selected for transcription by the EEBO Text Creation Partnership (EEBO TCP, the project developed, according to its website, with the intention to create "searchable and readable transcriptions [linked] directly to the corresponding ProQuest image files in the [EEBO] database"), the students were engaged in a truly original scholarly project. Their editions provide open access in conveniently readable (and machine readable) form to public-domain texts that would otherwise be available only in select rare-book libraries or through institutional subscription to costly database or microfilm collections. The students worked in two editorial teams, each team concentrating on one of the two books. There was, however, some movement between teams as distinct roles, interests, and aptitudes emerged among the students. Moreover, the work of each team was presented to the entire group at weekly meetings, and decisions involving editorial policy, design, and mark-up practice were made collectively.

The instructors provided first, a two-and-a-half hour seminar introducing editorial theory and practice supported by readings from Williams and Abbot's Introduction to Bibliographic and Textual Studies; and second, an intensive five-hour hands-on introduction to HTML using the popular Adobe Dreamweaver HTML editor. HTML mark-up is far from a "cutting-edge" tool for preparing electronic scholarly editions; nevertheless, it has significant virtues for a project-based course in electronic textual scholarship, which Matthew Steggle enumerates in a discussion of electronic journal production:

[M]ore elegant and structured ways of encoding data are certainly technologically feasible, but HTML is readable by even the most basic of web browsing software; it is accessible to search engine spiders; it is robust; and if anything goes amiss, the editor can solve the problem themselves [sic] rather than having to call for technical support. The [. . .] success [of the online journal EMLS], and its continued wide readership, is partly a product of this "tractor technology." (96; see also Lavagnino.)

Pedagogically, the primary goal of the course was to inspire students to achieve the high standards required of scholarly work intended for publication—something most undergraduates never experience. HTML offers students the opportunity to produce an original publication with real (if modest) scholarly value, and it offers undergraduates an introduction to more complex tasks should they pursue more work in this vein. Some students had had little or no previous experience with HTML mark-up, much less stylesheets, ftp, the relationships of file structures in the construction of a website, or usability issues, but they were all soon proficient at creating and editing web pages and uploading them to the server. The seriousness with which they tackled editorial problems was, at least partly, a function of the publication opportunity accorded by the simplicity of HTML, the accessibility of the World Wide Web, and the students' consciousness of a real audience (that these editions do reach an audience can be demonstrated by an internet search for references to As One Phoenix: Four Seventeenth Century Women Poets, the earliest of these student editorial projects, completed in 1998).

Students in this course also prepared and presented reviews of selected web-based scholarly editions as part of their preliminary introduction to editorial procedure and web design. Students completed all their written work in HTML and uploaded their work to a course homepage before each weekly meeting in the Humanities and Fine Arts Digital Research Centre at the University of Saskatchewan. This approach facilitated real-time correcting and updating of transcriptions and annotations, and a growing confidence in independently solving questions about mark-up and file management as the group collectively resolved editorial and design issues. It also created an awareness of the need for version control and the "check-in" and "check-out" features of software for managing web-based content development in a team environment. The instructors provided regular supplementary advice on mark-up and standard editorial procedure as questions arose at weekly meetings. Nevertheless, the key editorial and design decisions were made by the student editors in response to challenges they identified as their work progressed. The syllabus provided a rough breakdown and schedule of tasks (e.g. transcription, transcription checking, annotation, drafting, and revising the critical introduction), and grades were assigned to individual participants based on completion of these tasks (with an added peer-evaluation component). But substantive decisions, such as the topics to be addressed in the critical introduction, were made by the editorial teams.

Editorial decisions

The earliest and most significant editorial decision was to prepare a documentary, or "non-critical" edition based on a careful old-spelling transcription of a single copy of each book's first edition. To some extent this was an arbitrary decision—a concession to the time constraints of a 13-week course, and to the necessity of working from digital facsimiles of the early printed books. Moreover, in the case of the Eighth Liberal Science, only a single copy of the book was available (the British Library copy, reproduced in electronic facsimile in the EEBO database), so no collation of textual variants was practical. In the case of Whitaker's Directions for Brewing, the editors also relied on the British Library copy (again from EEBO). A horizontal collation exercise would have been possible for Directions for Brewing, since EEBO includes digital facsimiles of two copies of the first edition (the BL copy and the Cambridge University Library copy), but it would have compromised the collaboration achieved by synchronizing the two teams' tasks. Also, each of the books contained material that had been previously published, so assembling the thorough textual history required in a critical edition would have been prohibitively time-consuming. The arbitrariness of this approach is difficult to defend except on grounds of expediency; nevertheless, as E. S. Ore argues, "any scholarly edition is better than none, even if it is not a critical edition" (36).

These decisions, and others relating to content, design, and presentation, were driven by discussions of the most likely audience for editions of these texts, and the use to which they might be put. The editors determined that their editions should contain the apparatus (explanatory notes and critical introductions) likely to be helpful to senior undergraduate and graduate student users who might lack access to the microfilm facsimiles in the Early English Books collection published by UMI, or the digital facsimiles available through EEBO. Indeed, even readers with access to these resources, or with more advanced scholarly agendas, might benefit from readable and machine-readable transcriptions if not necessarily from the student-oriented introductions and annotations. Also, despite adopting the conservative approach to orthography appropriate to documentary editions, the editors decided to correct egregious compositors' errors (e.g. inverted characters), to modernize typography (i, j, u, v, w, and s), and to expand compositors' contractions and abbreviations (i.e. those which serve primarily as aids to line justification). This limited regularization of the text facilitates both reading and electronic text analysis.

From the decision to prepare a documentary edition, and from the related discussion of principles and customary applications of documentary editing, developed a desire among some of the student editors to represent the printed book as an artefact rather than to produce merely a reliable text. As they scrutinized facsimiles of the early printed books with increasingly minute attention, several of the editors became convinced, to quote G. Thomas Tanselle, of "how interconnected are verbal works and their tangible representations in physical objects" (ix). In short, they became reluctant to sacrifice more than was absolutely necessary of the information contained in the page design of the early printed book. Paradoxically, part of the rationale for this view emerged from the students' critical analysis of electronic scholarly editions. The editors argued that rendering a long document as a single scrolling page of text (a common procedure in e-texts) is a barrier to reading, and inconsistent with good web-design principles. As Christian Vandendorpe writes,

"kilometric display" is quite inadequate for serious reading. In fact, as evidenced by many studies it incites the reader to skim through the text rather than to read it attentively . . . . Moreover it does not give the reader the white space of the margins … [which] give shape to the text and allow the eye to regenerate from the tension produced by the innumerable saccadic movements of the eye during the reading process. With the disappearance of margins, text is reduced to its content. (211)

In a more ambitious project, the elegant solution (and one briefly considered by our teams) might be to supplement the transcription with high-resolution images, as does the Internet Shakespeare. But acquiring the rights to images would have been impossible or prohibitively expensive in the context of a student project. In the end, the student editors favoured a digital variation on the traditional "type facsimile" edition.

At the same time, the editors recognized the advantages of unbroken text for some users and some applications. Users might, for example, wish to print the text in its entirety, or cut and paste substantial excerpts into a word processor document. An electronic type facsimile would frustrate these applications. Moreover, because of concerns for accessibility, the editors refrained from trying to find ways to prevent readers from resizing the fonts. Without such limitations, carefully designed facsimile pages could render incorrectly in any given user's browser. Furthermore, without paged-media capability in a browser, these pages would be likely to break incorrectly when printed. Ultimately, some of the editors became convinced that adherence to the arbitrary features of the early printed book is an anachronistic gesture, driven more by nostalgia than by any real increase in information. This case gained force as the difficulty of achieving a true electronic type-facsimile using simple HTML became increasingly apparent.

Two views of the text

Recognizing the merits of both positions, the team resolved to produce dual editions, which would provide two views of the same transcription and use a script to switch between style sheets. The print view (Figure 1) offers continuous scrolling text:

The page view (Figure 2) is designed to approximate the page design of the printed book: Moreover, the designers of the Directions for Brewing Malt Liquors opted for a page-spread display in their page view (Figure 3 and Figure 4). Facing verso and recto pages are displayed on screen, as if the book were open on the reader's desk. In both editions, elements required only in the page view are marked off with a <span class> so as to display only in the page view and not in the print view. This approach provided the editors with insight into the way textual variants and multiple states of a text might be represented in a more ambitious electronic critical edition.

In the page view, the editorial teams set out to emulate the original page on the screen while maintaining readability. Many decisions that flowed from this principle were relatively straightforward. In the page view, annotations appear individually as the mouse is scrolled over the note reference. In the print view, annotations appear as end notes. Reproducing relative font sizes, font changes (between roman and italic type), printer's catchwords and page signatures involved little sacrifice of readability. Hence, even if no compelling argument could be made for their inclusion, none was made for their omission or regularization. Fonts, of course, could not be specified, since HTML fonts are limited to those available on the end user's computer. The editors of The Eighth Liberal Science resolved, where the main text is set in a slightly archaic (for 1650) blackletter type, to use a modern sans serif font in place of the blackletter, and modern roman and italic fonts where the 1650 edition uses their early-modern equivalents (see Figure 5 and Figure 6).

Other decisions were only slightly more complicated. For example, in the printed text of the verse "Satyr upon Brandy" appended to Whitaker's Directions for Brewing, long verse lines are occasionally truncated, with the extra word dropped to the following line. Hence the opening couplets of the "Satyr":

Farewel damn'd Stygian Juyce, that dost bewitch,
From the Court Bawd, down to the Country Bitch;
Thou Liquid Flame, by whom each firey Face
Lives without Meat, and blushes without Grace,

are printed as follows in the first edition, because the ornamental capital consumes space on the second line: Though some pointed out that the printer's procedure here is analogous to the sort of abbreviations which the team had decided to regularize, the transparency of the compositor's expedient, and the availability of the print view, where the couplets would be rendered correctly, led the editors to render the text in the page view as it was printed in the 1700 first edition; here, as in several other instances, student editors exercised the autonomy implied in the project-based approach.

Nevertheless, the concern for aesthetics and readability in the page view did lead to compromises, the most significant of which was the decision not to retain the original lineation for the main prose text of both books. Several failed attempts to design a pleasing HTML text with hard-coded line breaks, full justification, and reasonable spacing convinced the editors to abandon the lineation of the first printed editions of both texts. Constraining font sizes or abandoning the justified right margin (or perhaps both) might have led to a more faithful electronic type facsimile, but in the context of a class project, the time invested in web design might well have compromised the time available to attend to crucial editorial issues. In the classroom this problem led the teams to a delicate negotiation among editorial, web design, and learning goals; in the page view of both editions, it led, finally, to a compromise between an electronic type facsimile and a diplomatic transcript, for passages of continuous prose.

Other frustrations in the attempt to emulate the printed page using HTML provided students with important insights into the versatility of moveable type and the limitations of HTML (or perhaps, the limitations of the simple HTML they could master in a 13-week course). For instance, one page of The Eighth Liberal Science (Figure 8) includes text oriented vertically on the page:

However, the editors were unable to discover a browser-independent method of representing this feature of the printed page. As a result, the page view of this page (Figure 9) does not emulate the layout of the printed page in every browser: Here, as with the lineation of continuous prose, the result is a page view that falls slightly short of the electronic-type facsimile the editors had hoped to produce. These compromises led some of the editors to question the value of a near facsimile, though most remained committed to the ideal.

Pedagogical issues and reflections

The major challenge (and benefit) of this project-based approach involved the students' simultaneous engagement with the principles and practices of scholarly editing and web design. At every stage, the student editors were forced to weigh the value of their investment of time in solving a particular design problem or capturing certain features of the original printed books. The student editors' initial enthusiasm for the seemingly limitless possibilities of electronic text inspired them, as it has inspired many electronic editors, to offer users a choice of presentations rather than prescribing a single view of the texts. While they were not able, in the time available, to consider and represent different states of the texts, they were able to explore both the reasons and means for doing so. The process thus resulted not only in a published product, but also in a valuable exercise in team problem solving, time management, and responsibility division. Obviously these students have not become expert textual scholars or web designers, but they are better equipped now than most new graduates to begin the next phase of an education in electronic textual scholarship—the complex and multifaceted task of preparing an electronic critical edition employing state-of-the-art mark-up techniques as a thesis or dissertation project. As readers and critics, these students have a richer sense of the texts they read as constructed artefacts, and as artefacts constructed not just by authors, but by many hands at different stages in their textual histories.

In focusing on the difficulties of producing both a continuous electronic text and an electronic type facsimile based on a particular copy of a particular printed book, the student editors were forced to grapple with important textual questions: How does an editor distinguish, in practice, between the text and its physical representation? Which features are properly aspects of the text, and which are artefacts of its reproduction? Which features of a document does a conscientious editor have a duty to render faithfully? There was, for example, considerable debate among the student editors about the status of contractions and abbreviations. Abandoning the original lineation reinforced the editors' decision (initially made on grounds of readability) to expand compositors' contractions and abbreviations, which serve primarily to regulate line length. But after making this decision, retaining other contractions on grounds of documentary fidelity became harder to defend. These experiences exemplified to the students one of the central paradoxes of textual scholarship: editorial principles are in a certain sense arbitrary (inasmuch as good arguments can always be mounted for different principles), but they must therefore be articulated, and individual decisions made by consistent application of those principles.

The students also experienced the human dimension of scholarship in ways the conventional academic seminar course seldom permits. Although English classes, especially senior seminars, do cultivate class discussion, there is a tendency (valuable in many ways) to leave critical and interpretive issues unresolved. Whether The Merchant of Venice is an anti-Semitic or an anti-racist play is a problem no honours seminar discussion will solve, though individual participants may well arrive at their own firmly held conclusions. Advocates of humanist education customarily claim that such debate over perennial and ultimately insoluble questions is central to the preparation of a democratic citizenry. But humanities students have few opportunities to engage in the kind of debate that must end in a decision. In editorial matters, as in many matters of policy and procedure, a course of action must be chosen collectively, and the grounds for choosing that course of action must be articulated. Such decisions usually have consequences that extend beyond the moment, and beyond the private convictions of the participants. Participating in this kind of debate and discussion is also essential preparation both for citizenship and for professional careers. This was the first opportunity many of the student editors had to participate in such a decision-making process about academic issues. Humanists are fond of pedagogies that emphasize process over product, but the requirement to produce a product informs and transforms process in important ways.

Finally, though the body of this paper has concentrated on the students' engagement with issues of textual scholarship and web design, it should be noted that preparing the critical introductions to these to these texts required student editors to apply their skills in literary and historical research, cultivated in other classes, in ways they had not done previously. Working on texts for which little or no secondary scholarship exists, they were compelled to locate and draw conclusions from primary historical materials such as court records, and to track down archaic terms, classical and literary allusions, and references to obscure people, places, and events. Investigating the biography of Edward Whitaker and the social history of brewing, distilling, and drinking in the seventeenth century led them to explore the Glorious Revolution, English Protestantism, Catholicism, mercantilism, absolutism, Louis XIV, William of Orange, corn growing trends and prices, brandywine embargos, and so forth. The need to explicate such cryptic references as a glancing allusion to one Shr––Sh–– in the "Satyr on Brandy" appended to Whitaker's Directions for Brewing (Sheriff Shute, a reputed drunkard and Whig partisan, Sheriff of London and Middlesex for 1681-2), or more humorously the unfathomable reference to pickledisters ("pickled oysters") in Eighth Liberal Science, gave these students their first taste of what may be the quintessential scholarly experience—spending agonizing hours locating the information that might occupy a single note. For some, this is the "make or break" experience in a scholarly career. Would-be researchers either savour the "eureka moment" (as co-author Joel Salt clearly did, emailing his colleagues at 2 a.m. to announce his discovery of Sheriff Shute), or they resolve to pursue some enterprise where the recognition is more evidently proportionate to the investment of time and effort.

The self-knowledge and disciplinary knowledge thus gained is, perhaps, the most compelling reason for incorporating a component of project-based learning in an honours English curriculum. Students trained primarily in close reading of canonical literary texts (however much the canon may have expanded in recent decades), graduate, and often enter graduate school, with a distorted perception of what original literary scholarship entails. The notoriously poor completion rates in humanities doctoral programs may be partly attributable to the disappointment and disaffection arising from such misperception. A project-based course in electronic textual scholarship can provide crucial information and experience, both for students who choose to pursue scholarly careers, and for those who choose to pursue other professional avenues.