By Lorraine Janzen Kooistra (Co-PI, Ryerson University) and Constance Crompton (Project Manager, York/Ryerson)

A version of this paper was collaboratively presented at the “Victorians on the Web” panel hosted by the Congress at Concordia University, Montréal, May 2010. The name of the project was changed from The 1890s Online to The Yellow Nineties Online in December 2010.

With the coming of electronic scholarship to Victorian studies, there has been an epistemic shift in “how we do the biz.” In Lorraine’s case, the changes came incrementally, rather than as a rupture. My digital life began in 2004, when Jerome McGann invited me to join the Victorian Editorial Board of NINES, the Networked Interface for Nineteenth-Century Electronic Scholarship. At the time, NINES existed more as a scholarly vision than a working reality: it didn’t have a website up, let alone the range of digital tools and federated peer-reviewed sites that it does today. And although Jerry assured me that my contributions would fall within my field of nineteenth-century scholarly expertise, and that I wouldn’t need any technological know-how, I knew that in accepting the invitation, I was entering a foreign world without a Baedeker. The first thing to do, I reasoned, was to learn the language. And the best way to do that would be total immersion: I would create a scholarly site to be peer-reviewed by NINES.

At this time, I was also in the process of moving from a small northern Ontario university to Ryerson in downtown Toronto, an institution known for its tech-savvy and innovative research. Here I discovered that a fellow Victorianist, Dennis Denisoff, was also keen to become involved with NINES. Over dinner one night we talked about how our common interest in exploring digital scholarship might combine with our shared areas of scholarly expertise, nineteenth-century visual/verbal relations. By the time we’d reached dessert, we had become co-PIs on The Yellow Book project, which we agreed to submit as a proposal for the inaugural NINES summer workshop at the University of Virginia. Our proposal, “‘The Yellow Book Project’: A Co-Application by Dennis Denisoff and Lorraine Janzen Kooistra for Joint Participation at the 2005 NINES Summer Workshop, July 18-25, University of Virginia,” was vetted by the NINES hosts and became one of twelve successful projects (and the only Canadian one) included in their inaugural workshop. Five years and a name change later, we can still affirm the gist of that proposal: “The first stage of the 1890s Hypermedia Archive, The Yellow Book Project is dedicated to publishing an annotated scholarly edition of the periodical’s 13 volumes [1894-1897], and to developing interactive research tools and an online site for refereed articles addressing fin-de-siècle cultural studies.” What’s changed is our understanding of what we’re doing, why, and how—the implications of electronic scholarship.

In our initial proposal, we had fondly imagined that we would complete the publication of our digital edition of The Yellow Book by 2008: our intention was to imitate the magazine’s quarterly installments and bring out a volume, fully marked-up for searching, every three months for thirty-six months, beginning and ending with an April issue. Instead, we celebrated the official launch of The 1890s Online and the electronic publication of Volume 1 of The Yellow Book only last month, at a Digital Symposium we hosted at Ryerson on Visualizing the Archive. The remaining twelve volumes have yet to be marked up and published, but we’re cautiously optimistic that this may be completed by 2014. In what follows, Constance and I want to bring you behind the scenes of our labours and expose the back end of The Yellow Book Volume One’s epic journey from proposal to online publication to peer-reviewed federation. Along the way, we’ll explore what our experience as makers and users of electronic scholarship has taught us about “The Way We Live Now” (Trollope 1875) as Victorianists: the dynamic nature of our scholarly object, the complexities of collaborative authorship, and the informing implications of temporality.

While our vision of providing open access to peer-reviewed archival and scholarly material has remained unchanged, the site’s various names—The Yellow Book Project, The 1890s Hypermedia Archive, The 1890s Online—signal shifts in our understanding of what it is, exactly, that we’re building, why, and for whom. The 1890s Online is, simultaneously, an electronic edition, a digital archive, and an e-publication of peer-reviewed scholarly essays relating to “the magazine’s production and reception, the periodical industry at the time, and the lives and works of key contributors.” Our vision for “an enriched scholarly environment” for The Yellow Book expands the digital archive to include facsimile editions of other fin-de-siècle periodicals with limited print runs, but wide-ranging significance, such as The Pagan Review and The Evergreen. While still conforming to our initial proposal of 2005, The 1890s Online is clearly an amorphous and malleable scholarly object, constantly changing and evolving in response to local and international circumstances, computer technologies, and human social processes and interactions.

In thinking about these problems of definition, we’ve come to realize that they’re related to the ways in which an electronic website differs from our traditional scholarly object—the monograph. Usually a single-author work, the monograph represents an individual scholar’s “take” on a given subject at a given time. Once it’s published, the book is complete in itself—a textual event and an object in space whose temporality is henceforth related more to its reception than its production. The author supplies the text, and the publisher builds the paratext, as Gerard Genette reminds us, “precisely in order to present it, in the usual sense of this verb but also in the strongest sense: to make present, to ensure the text’s presence in the world... .” For Genette “the paratext is what allows the text to become a book and to be offered as such to its readers and, more generally, to the public” (1).

In our experience of publishing Volume 1 of The Yellow Book on The 1890s Online, we’ve learned that the paratext is also what allows the text to become an electronic object and presented as such to its readers and, more generally the public at large. The difference is—and here’s the rub—the authors of digital text are also responsible, in various ways and to different degrees, for building the paratext that makes their scholarly material present to the world. And while our initial aim may be, say, to “present” Volumes 1-13 of The Yellow Book in electronic facsimile, we’re immediately confounded by the paradox of trying to present static historical objects (editions) in a dynamic temporal environment (the internet). The tensions of mutability and permanence, while also present in the codex, become amplified in the always-under-construction and never-completed nature of electronic scholarship.

When we write a book, we’re working with a number of codes—including the alpha-numeric code and the codex code—that we’ve naturalized over millennia in ways that we don’t usually reflect on in our writing, just as we generally leave most of the paratextual details to our publishers. When we build a scholarly website, our codes of conduct suddenly become much more complex. The “book” form that is our object of study—in our case, The Yellow Book Volume 1—and its linguistic, bibliographic, and iconic codes (McGann 13)—are suddenly put into dramatic interaction with programming languages and the conventions of website and database design. Some of these languages interpret our object in ways we had not anticipated. These languages include TEI, the standard mark-up language for the humanities, and xslt, html, and aspx, all of which we need in order to paratextually present texts on the web. In addition, the limits of proprietary software and hardware have shaped our database, and by extension our users’ ability to search through the site’s material. And, as we learned, all these codes are not merely vehicles of content, but also interpretive carriers of meaning. In this sense, like any paratext, they co-produce the material and its interpretive point of view.

And that’s not the only way computer coding and the conventions of the web inform electronic scholarship. Because programmers have to take into account “users” and their “functions,” these unidentified individuals of various ages, genders, educational levels, abilities, ethnicities, orientations, geographic locations, and temporal periods become enormously important in the development of the site. With the launch of The 1890s Online and its demo model of Volume 1 of The Yellow Book, we’re at the “beta-testing phase,” trying to figure out how users interact with the site by adapting some Google-metrics’ software for user-tracking purposes—with a view, of course, to changing what’s needed. Programming and site development are iterative processes. You never get it exactly right, because you never know what you don’t know in advance. And so you constantly build and tweak and refine, in a never-ending social, collaborative, and technological process that plays out over time.

For this reason, an electronic website has a lot in common with periodicals—a comforting sort of analogy for a research team engaged in the study of a Victorian illustrated magazine. Like periodicals, websites have a fixed beginning but no known end—theoretically, once begun, they carry on to infinity, irrespective of any changes in editors, contributors, publishers, and locales. Because of their design features and make up, digital sites also promise what Laurel Brake and Julie Codell have described as the periodical’s appearance of a “false unity... as if the journal itself were by a single author” (1).

Instead, as we know, The 1890s Online is co-edited by Dennis Denisoff and Lorraine Janzen Kooistra; the content comes from a host of contributing scholars around the world; the mark-up is provided by a team of Ryerson-based research assistants; the programming is in the hands of various research associates, collaborators, consultants, and technicians, from a number of North American universities; and the overall criteria are established and vetted by NINES, with its own host of collaborators, associates, and editors. Like a periodical, a website is a publication broken up into discrete units of design and information, produced within the social processes of collaborative authorship, and engaged in the productive tensions between ephemerality and permanence. And like the illustrated periodical that is our particular object of study, a website also interacts with its users through visual and verbal modes of communication, offers multiple points of entry, and permits a wide variety of user-generated activities. What Margaret Beetham says of the reading practices encouraged by the periodical form, extends, by analogy, to scholarly websites: “readers can to a unique degree construct their own text from the printed [or in our case html-encoded] version...the form invites us to flip [or, alternatively, click] through, read in any order, omit some sections and read others carefully” (13).

On the other hand, a scholarly website is also not like a periodical, or any other form of print culture. It is its own unique thing. While the analogy to the periodical is useful for people like us who like to think in metaphors, we’ve also learned, over the last five years, that electronic scholarship requires us to reconceptualize the production process, the means of circulation, and the division of labour in creating scholarly works.

In building The 1890s Online, we’ve learned that the framework of an online environment shapes the creators’ production processes, including the texts they write and that users read. Here our analogy with the periodical holds, but there are significant differences. In their discussion of Victorian periodical fiction Linda Hughes and Michael Lund remind us that the “reading process [is] embedded in a specific material framework that shape[s] the response” (9)—to which we might add that the production process is also embedded in its framework, whether material or virtual. As we mentioned earlier, the form of a book or periodical shapes our writing process, but we’ve so naturalized the codex format that we may not notice its imposition on our construction of text.

Scholarly internet sites have inherited these codex conventions: even the journals that only publish online adhere to the structure of print editions. But internet sites also have their own paratextual conventions that makers must adhere to if users are to “read” their sites effectively. For example, navigation bars conventionally start with a “Home” button; users will likely look for a “search” field or box on every page; users expect to find consistent fonts, menus and headers as they navigate through each page on the site.

The scholarly site’s relationship to temporality also distinguishes it from the serials of print culture. A periodical has a specific temporal existence in which each issue is a node connecting the reader to the past and future of its print run. A site like The 1890s Online might try to reproduce that temporality by updating the site or adding new content at fixed intervals; however, the site itself—for all the creators’ interest in transparency—works against this nodal, yet progressive, relationship to time. The site is a discrete textual event. It may have gone through many iterations, it may have a new tone, or visual display, but the reader can’t access the site’s “back issues” in order to grasp the site’s progression through time. Some scholarly sites provide visitors with access to site development, or editorial decision-making processes, but these documents tend to produce self-conscious corporate history. In contrast, the content and paratextual elements of a journal’s back issues reveal more about its editorial and ideological stances than purposefully crafted editorial histories do. There are, or course, independent attempts to archive internet websites, which can be searched through portals like The Internet Archive’s Way Back Machine. Beneficial as The Internet Archive’s snapshots of the web may be, the name “Way Back Machine” suggests that the user is traveling to some remote past, accessible only as the digital equivalent of an archeological dig or virtual museum preserving a curated selection of former websites. Unlike the periodical’s march of printed issue after issue and volume after volume, The 1890s Online’s past has been erased in the launching of the newly designed website, and there’s no link to former pages posted from April 2005 - April 2010.

The motivation and method of circulation for scholarly websites and online archives also distinguish them from periodicals. Periodicals persist so long as they are economically viable. This is, of course, also true of non-commercial, open-access scholarly sites. However, the means of production and the monetary exchange between the reader and the producer were much better defined for the Victorian periodical. The scholarly site, as Susan Brown observed last April at our Digital Symposium: Visualizing the Archive, tends to receive funding as though it were a monograph—in other words, as if it were a bounded, discrete, and finished product. Lack of stable of funding for the dynamic, ever-evolving, and never-to-be completed website is a characteristic of electronic scholarship, and the five years it took to launch The Yellow Book, Volume 1 is surely testament to this.

Another challenge is our distance from our users. An open-access site such as The 1890s Online can’t rely on traditional circulation and subscription information to gauge the gap between the site’s intended and actual audience. Leaving surveys and live trials aside, online archivists do have some access to more sophisticated metrics tools—ones that can tell us how each user found the site, how long they stayed, what they looked at, even what demographic they represent. We have yet to decide which metrics tool to use on our site. But even when we have this information, we may not know how to interpret it. How much traffic do we need to validate our efforts? How should we interpret and respond to the searches that lead to our site? How do humanists harness quantitative data effectively for their scholarship?

We’ve learned to recognize that a website’s electronic publishing medium is much more dynamic and contingent than the print medium of the periodical. The scaffolding for the texts and images (be it paper or a screen) gives the content a distinct type of ephemerality. As an object that exists in space, each periodical becomes book-like the moment it’s bound. Scholarly sites, on the other hand, are uniquely generated in the user’s browser with every new visit. The temporal existence of the site displayed on any screen is limited to the length of the visit. While the periodical may decay though frequent handling, the browser-generated site will degrade if the makers don’t continue to make: old fonts and colours, formatting and even programming languages will work against the maker’s intentions for graphic design as browsers and hardware change.

Finally, in building The 1890s Online we’ve had to engage new modes of production, with divisions of labour unfamiliar to many literary scholars. Collaborators on a scholarly site work in a manner that is both more and less Fordist than the producers of a periodical or book do. The production of a website isn’t a cottage industry—no one person has mastery over all the steps in the publishing process. While collaborators perform their own specialized tasks, however, they must all have a sense of what the other creators are doing. As mentioned earlier, the constraints imposed by computer programming languages, mark-up languages, servers, institutional structures and funding shape site content in a way that is unfamiliar, if not jarring, to those of us who have naturalized the monograph-production process. In order to make sense of these restrictions, collaborators have to be able to explain their specialized labour to one another—often by resorting to analogy and metaphor (as we’ve done today with our web site/periodical analogy). As a result each person working on the site will have a very literal sense of their own task, and a rather metaphorically mediated sense of how that work relates to the whole.

There’s generally a divide between the people who produce humanities scholarship and the people who have the technological expertise required to build a website. At our Digital Symposium’s Q & A in April, Alan Galey identified the tendency of humanities scholars to contract out technical labour rather than collaborate with computer science experts. The 1890s Online team has learned that outsourcing can sometimes undermine the collaborative process. We’ve generated our most productive relationships when team members with humanities training have acquired technical knowledge in order to bridge the digital divide. Developing our computational literacy means that team members understand the perspectives that motivate humanities scholarship, and can use them in both their computing and non-computing tasks on the project. Even though this method narrows the digital divide, it can’t be deemed a complete interdisciplinary success, since it circumvents, rather than promotes, collaboration with computer scientists and engineers. To produce ideal collaborative projects we need to do more humanities outreach. Just as English scholars on the project have increased their computational literacy, we need to provide an environment that makes the principles and motivations for humanities work accessible and engaging for programmers. The reciprocal training that will turn both literary critics and computer engineers into digital humanities scholars adds temporal pressure to any digital project. Training takes time. And time costs money.

Institutional structures are the final, if inadvertent, collaborators on digital projects. The 1890s Online is currently catalogued by the Ryerson University Library, and thus has presence within the Canadian university library system. Its next step is to be vetted by the Networked Infrastructure for Nineteenth-Century Electronic Scholarship, but this isn’t the project’s first (or last) formative interaction with NINES. Federating groups, like NINES, shape digital projects’ initial formation, provide them with that academic credibility of peer-review, and contextualize the meaning of a site’s content by bringing it up side by side with the content from other sites. NINES has specific programming and content requirements that projects must meet in order to join the federation. Once these initial requirements have been met, the peer-review process initiates more changes to the project. If incorporated with NINES, juxtaposition with items from other projects will also shape the meaning of The 1890s Online’s content as users mine data and collect objects from the federated sites of this electronic consortium.

Our question to ourselves about what The 1890s Online is, has garnered a concurrent question about what our external collaborator, NINES, is. Collaborators on our project tend to communicate through analogy and metaphor, so we’ve tried to describe the work of our silent partner, NINES, to one another through analogy and metaphor. Is NINES a library? Is it more catalogue than library? Is it an enhanced Jstor? A Jstor, that isn’t Jstor at all, by virtue of its inclusion of semantic mark up, bibliographic records, current scholarship, community forums, teaching tools and the highly contingent nature of the independently generated sites in the NINES federation? Is NINES like a meta-periodical?—a form, to borrow Margaret Beetham’s term, “marked by radical heterogeneity” (12)—that is to say, a system containing multiple items with numerous structural variations: at last count, 688,180 peer-reviewed digital objects in 94 federated sites.

We’re looking forward to hearing Susan Brown’s take on the nature of NINES in her upcoming presentation. Meanwhile, we’d like to close by reflecting on the epic saga of launching The Yellow Book, Volume 1, on The 1890s Online, and the epoch it represents in terms of our own electronic scholarship. When co-editors Henry Harland and Aubrey Beardsley launched Volume 1 of The Yellow Book in April 1894, they issued a Prospectus that announced their intention “to depart as far as may be from the bad old traditions of periodical literature, and to provide an Illustrated Magazine which shall be beautiful as a piece of bookmaking, modern and distinguished in its letterpress and its pictures, and withal popular in the better sense of the word.” It would be fair to say that The 1890s Online team shares with our historical predecessors and their print-culture object the desire to be as “modern and distinguished” in our paratextuality as we are in our textuality—in our user interface, page designs, and functions as well as our contents. And we also want to be “popular in the better sense of the word”—that is, to generate as many international “hits” as a well-designed scholarly site freely accessible to the public allows. But, even as we move on to the next stages of our digital project—creating more primary periodical material in facsimile, incorporating the kinds of visualization tools that will allow users to view and understand these historical objects in new ways, and publishing more related scholarly material to enrich the reading context— we also wonder where it will all end. While the original Yellow Book’s print run lasted a mere three years, Harland and Beardsley did achieve the goal outlined in their Prospectus. Their Yellow Book was indeed “a book—a book beautiful to see and convenient to handle” (1, 3). Our twenty-first century electronic edition, on the other hand, will always be in the process of achieving a beautiful and convenient textuality on the digital bookshelf even as it continues to be read, and read again, by users across the world-wide web.