The Siege of Antioch Project: Digital Approaches – Nicholas L. Paul
Intro:
The Siege of Antioch digital project first began in the summer of 2016 in a series of conversations between the Fordham Center for Medieval Studies and a UK-based team of scholars [SLIDE 2] (Linda Paterson, Simon Parsons, and Carol Sweetenham) who were preparing an edition and translation of the long verse narrative text we are now calling the Siège d’Antioche [1] As it has transpired, this was a very fortuitous meeting. Linda Paterson, Professor of French at the University of Warwick, and our initial point of contact, had in the past years completed the [SLIDE 3] AHRC funded project Troubadours, trouvères, and the crusades [2] a major digital resource to which all students in the field of crusade studies are greatly indebted. This resource was of particular interest at Fordham, where the Center for Medieval Studies has maintained a robust digital presence since the mid-1990s with the publication of the Internet Medieval Source Books [3] [SLIDE 4] and several other digital resources. Among Fordham’s strengths was special expertise with the use of the French language in medieval England. Since 2010, the Center had undertaken a number of digital projects further expanding in this area to encompass French as it was used in Italy and in the eastern Mediterranean in the period of the crusades. Most relevant here [SLIDE 5] were the French of Outremer project [4] and the Oxford Outremer Map [5]. Our shared interests in crusade studies, Old French, and digital humanities made this a potentially strong match with a digital project based on a newly-rediscovered Old French crusade text.
The UK team, which includes Professor Paterson, Dr. Carole Sweetenham, and one of our hosts here today Dr. Simon Parsons of King’s College, all have extensive experience with the editing and translation of medieval texts, and following on from the success of the troubadour/trouvère crusades corpus, were open to the possibility of a digital text. The text in question, the epic Siège d’Antioche lends itself to a digital edition for several reasons. First, and most obvious, is its length. At 16,000 lines of verse, it renders as a major undertaking for many print publishers, and also potentially difficult to access for many readers, for its presumed hefty price and its unwieldy length. [5] On the other hand, the text has very wide potential interest: [SLIDE 6] the study of the crusades is now a massive field that includes scholars from many different disciplines, and is taught at some level in the majority of universities where there are historians of premodern Europe and the Near East. Within this field, the study of the First Crusade, and the flood of narrative histories that it spawned, is one of the largest areas of concentration. [6] The importance, therefore, of a lengthy vernacular work, one of the very few sources associated with England, which remains unedited and untranslated and mainly unknown to the scholarly community, cannot be overstated. [7]
Another feature of this project lends itself to the digital format: the edition being prepared is based on two manuscripts and a fragment, making the display of variants alongside the lemma possible in the context of a simple pop-up window. The small number of MSS also makes theoretically possible, should images become available in IIIF format, of integrating images within the edition. Finally, as an epic of the First Crusade, the Siège d’Antioche has both a very large cast of characters, many of whom are based on identifiable historical actors, and a significant topographical imaginary based on the crusaders’ homelands in Europe and the story world of their journey in the eastern Mediterranean. The sets of both personal and place names can be challenging, even to seasoned experts on First Crusade texts, and so the integration of the text with online glossaries and maps can greatly facilitate comprehension. For the untrained reader, the undergraduate or general enthusiast who might happen upon the text, it is equally critical that clear explanations be available for terms such as “pagan” “Persian” or “Saracen” in describing the crusaders Muslim adversaries. [8] [SLIDE 7-BLANK]
And what about the wisdom of digital editions more broadly? Thirteen years after Peter Robinson asked, in the first issue of the journal Digital Medivalist “do electronic scholarly editions have a future?” the answer is still yes, but it is even clearer today that medievalists operate in a hybrid environment. As experts in on one of the most robust corpora of textual witnesses to survive from the premodern world, medievalists are perhaps most keenly aware of the fragility of most forms of textual media. The possibilities of digital editions must always be weighed against the problem of obsolescence and limited support for the legacies of digital projects. It is not for nothing that a predicted future loss of digital knowledge is colloquially known as a “Digital Dark Age”. [9] With that in mind, our project is envisioned in two stages, first if possible a full or at least partial digital publication of the Siege to be published online for a fixed time, followed by a print edition. [10] This approach has the advantage of placing the text into the scholarly conversation with regard to the narratives of the First Crusade as soon as possible, while also avoiding possibly insurmountable problems arising from the longer-term legacy of a digital project. It is also possible that some aspects of the digital edition, in particular the glossaries, digital maps, and notes compiled by the Fordham team, could be archived and made available to readers of the printed edition when it appears.
So far, I have spoken primarily of the value of the project to the larger scholarly community. While I very much believe in this as an outcome of the project, however, the main aspect of the project I’d like to discuss with you today is a pedagogical one. It is a long-running tradition at Fordham that graduate student participation in our digital projects has been considered a key opportunity for training. The Online Medieval Sources Bibliography, online since 2004, for instance, is built from the contributions of 67 MA students. [11] Our Center’s former Director Maryanne Kowaleski has co-authored two articles considering the value of previous digital projects in pedagogy together with her students who were involved in their production Morgan Kay and Esther Liberman-Cuenca. [12] More recently, beginning with the Oxford Outremer Map Project, students have not only worked on our projects, but they’ve helped from the outset to conceptualize and shape them. In this case, we saw the advantage for our students in gaining experience and training not only in Digital Humanities, medieval history, and romance philology, but also in seeing how a major project of textual editing and translating worked from the inside. What I think we did not anticipate was the degree of experience and training that the students would receive as project developers and team members. This kind of experience is relatively rare among students in the humanities, rarer still in the USA where humanities funding is less often tied to group scientific grants and network schemes. But this type of training is nonetheless essential for humanities students, especially as doctoral programs in the humanities expand beyond the prospects of the traditional academic job market.
Hence the focus of my talk today. My goal is to explain our approach to preparing a digital edition of this text in collaboration with Simon Parsons and our other UK partners. In doing that, I will discuss not only the technological choices we’ve made and the encodings and platforms we’ve adopted, but I will also suggest the pedagogical value of this kind of project. Remarkably, I believe that this project demonstrates more than anything how digital tools greatly facilitate not only new kinds of scholarly interaction, but also very old-fashioned human collaboration and group activity. Finally, I should note what my role is here. [SLIDE 8]Above all, I am primarily the chronicler of the work undertaken by our team. The real products, innovations, and contributions, as you will see, clearly belong to them, and have very little to do with me. [SLIDE 9]
Tools:
Having earlier experimented with using text encodings to create interactive online transcriptions of medieval Latin charters, there was no question that our team would adopt the framework and encodings of the Text Encoding Initiative, or TEI, for this project. The advantages of TEI are legion and well known, so not worth rehearsing in detail here. [13] With regard to the specific demands and opportunities of this project, however, the TEI provided a framework through which to think about the visualization of an edited medieval verse text, and the text files we produced using the TEI encodings gave us the ability to integrate a large amount of metadata about the text as we received if from the UK team and our own queries, notes on revisions, and reminders to ourselves. TEI is above all a series of guidelines, built through experience by scholars in different fields encoding different types of texts. The guidelines (now in the P5 series) can be enormously helpful in conceptualizing a project, but they can also be overwhelming, presenting hundreds of possible tags that can be used to provide information about a text and potentially also to dictate how it is ultimately going to be rendered). The tags the team selected for common use in the project include tags for the laisse number, tags indicating the language of the work, tags indicating the location of the line breaks, folio (manuscript page) breaks, and editorial information including editorial corrections, and interpolations. The team also adopted tags indicating where variant readings exist between the manuscripts and tags for proper nouns. Before they can be displayed on any website or platform, the TEI files require an ODD (one document does it all) file, which contains the instructions about how each XML element is to be interpreted and rendered (for instance making a popup box appear with a variant reading or showing a particular type of word in a particular color.) When completed, the xml files contain not only the instructions about how the text is to be displayed, but also stand as a record and self-contained project archive of the process for use by scholars who may wish to either replicate or learn from our methods in the future.
Where in previous projects we had faced a somewhat awkward and time consuming process of porting our text files with TEI encodings directly into platforms such as WordPress or Omeka, for this project we opted for a platform more clearly intended for the display of TEI compliant XML texts [14]. A relative newcomer, TEI Publisher was first launched in early 2017 by ExistDB, an XML based open source software project. [15] It was attractive to us right away because, unlike so many other possible platforms we looked at, it allows the creation of digital editions without becoming a programmer. The TEI Publisher platform, while not necessarily intuitive as it does require some familiarity with coding languages, is accessible for non-programmers to create an edition without directly writing lines of code. It was particularly appealing for the Siege project because we have been able to customize it to display parallel text in columns, allowing us to prepare a facing text and translation. We are able to instruct TEI publisher to display alternative readings that pop up when they are moused over, to tag and hyperlink proper names so that the reader can consult a glossary. Our TEI Publisher edition can be embedded in an iFrame, allowing for display within a larger project website, which is housed on WordPress. We use WordPress for the main site as it makes it easier to integrate with glossary webpages and potentially also with digital maps or other supporting work. Finally, as a NoSQL open source software dedicated to creating a straightforward database without the use of code, TEI Publisher conforms in its essence with the approach taken by Medieval Studies to so many of their projects. We like their philosophy.
In concert with most similar projects, our team also adopted the use of open source ATOM text code editor for the preparation of initial xml text files and the oXygen XML editor for reviewing of our XML files. Team members worked on ATOM on their own personal machines because it was free. As oXygen subscriptions required a fee, the Center for Medieval Studies acquired the program for one laptop in the Center for Medieval Studies office, so that the files could be validated. [SLIDE 10]
Methods:
The Siège d’Antioche text is divided into sections or laisses of roughly 35-45 lines each (though some are over 120 lines!!). [SLIDE 11] As they are completed, the UK team sends the laisses over in three separate Microsoft Word documents: a transcription, an English translation, and a list of textual variants. Both the transcription and translation often include editorial footnotes. [SLIDE 12] The Fordham team devised a workflow in which each laisse would be converted into an XML text file and then marked up based on the agreed TEI encodings which were preserved in a Codebook housed on a Team Google Drive. [SLIDE 13] The encodings begin at the level of the simple line break and line number tags, but progress to include the variant readings, editorial apparatus, and proper nouns. Proper nouns are categorized as placeName, personName, roleName. In addition, a “term” category was chosen to address the fact that users of the text, especially those unfamiliar with First Crusade narratives, may be confused or misled by designations in the Old French such as “pagan” “Persian” “Turk” “Saracen” etc. Each proper noun and term is assigned a unique ID. This unique ID allows for all of these tags to be searchable in both the transcription and translation, regardless of spelling.
Once the individual team member has completed the text file, they mark the team spreadsheet, which is housed on the Team Google Drive, to indicate that it has been encoded. This file is then checked by one another team member. When new tags are introduced a team leader tests the code in Oxygen to ensure that the encodings are properly formatted. While all of these elements of the process are conducted individual, and require little in terms of contact between [SLIDE 14]the team members, the team meets once per week in person. During these meetings team members review the status of encodings, to discuss difficulties that might have arisen with categories such as “roleName”, and to discuss approaches to the glossary entries for proper names. Once a month, this meeting becomes a review session in which the whole team works collectively to check the encodings for all of the completed laisses. This “final edit” allows the team to finally agree on solutions to any lingering questions and places at least three pairs of eyes on the whole text, with all the encodings, making it extremely unlikely that any errors will slip through. Completed laisses are added to the larger master file in TEI Publisher. [SLIDE 15]
[SLIDE 16] A secondary subfield of the work that team members undertake for the project is researching brief entries for the glossary pages. These have begun for some of the major figures to have appeared in the narrative so far. Each entry, no more than a single page, is prepared for the character. [SLIDE 17] These personography files can be cross–referenced with other existing databases of crusaders, like the one maintained at the University of Leeds. [16] In future, we hope to expand this work to students in our graduate and perhaps advanced undergraduate courses. These entries represent another pathway for students and researchers to become involved in the project, even if they do not want to be part of the text encoding process.
The final aspect of our process is a regular video call between the Fordham and UK teams, usually about 1-2 times per semester or term. This gives the UK team a chance to offer feedback on the digitally rendered text, and for the Fordham students to ask questions about decisions they are seeing in the text files. The teams are both in regular contact over email. In this way, the final product is always something that the whole project is in general agreement about, whether that concerns layout or content. [SLIDE 18]
The Siege of Antioch as Advanced Medieval and Digital Pedagogy:
Working closely with a medieval text, examining its language, manuscript witnesses, arguments, and the historical knowledge that it transmits, are the cornerstones advanced training in Medieval Studies. Essentially allowing our students to look over the shoulder of experts as they make decisions about textual variants, readings, and the meaning of a medieval text as it emerges from manuscripts for the first time is by itself a masterclass in medieval philology. This is greatly facilitated by the fact that the UK team is generous enough to answer questions and explain decisions that they have taken to the students. [SLIDE 19] As one team member writes: “this Project allowed me to engage with the mind of an editor, as we attempted to understand their choices on a level that we could translate into a digital format. All the same, as a very junior scholar, it is incredible to be able to communicate with such brilliant and respected scholars from a position of relative confidence.”
The process of identifying and categorizing people and places in the narrative provoked discussions between the team members and long consideration of the nature of the historical and literary subjects under consideration: “we really had to think about why and what we’re placing in each category” one said. Writing the short personographies for the major figures in the narrative is a task that involves research into the latest scholarship but also synthesis and above all digestion and writing for a non-expert-audience. While all of this work is more or less analogous to the type of work that might be undertaken in a graduate seminar, as a pedagogical exercise, the Siege of Antioch Project also achieves several aims which cannot be addressed in traditional forms of graduate classroom work.
First, of course, is the question of Digital Pedagogy. As a sustained project, students may spend at least one year with the project, the Siege of Antioch project allowed time for students to become truly expert with the TEI encodings, writing XML markup language, and the digital platforms where the text will be displayed. Every team member now has the ability to fully encode a medieval text they come across in the archives and publish it online to internationally recognized standards. Because TEI can be daunting, and time consuming, it may not be that any of them will actively encode the long source texts that they will use in their doctoral work, but whether they do or not, the skills that they acquired may not be as important as the process of acquiring them.
The Fordham team did not learn TEI, XML markup, or any of the other digital skills they used here from a class or a professor at Fordham. Rather, with encouragement from members of our community who had used TEI before, they investigated the TEI framework and taught themselves and each other the relevant encodings, engaging directly with the online community who authored the encodings and who are building the software like TEI Publisher.
Perhaps the most important self-directed aspect of the project, however, from the Fordham point of view, is its organization and leadership. The students themselves, through their regular weekly meetings, established their own workflow, critiqued each others’ contributions, and selected and evaluated the software platforms. At the beginning of each new semester, they offered training sessions for potential new team members, explaining the concept, workflow, and methodology of the project and developing exercises to help new team members come on board quickly. In short, the Fordham students effectively built and ran a digital project from the ground up using skills that would be relevant to any workplace, inside or outside of the academy. One student responded specifically that he was struck by the ways the project made him think about the experience of the end user, the consumer, using the finished product– this is somewhat different from thinking about how a mentor, a committee, or even a scholarly audience will respond to a thesis or article. The team had to navigate many different challenges, including differing levels of fluency with digital tools at the outset of their work. They had to, in the words of one team member: “get a sense of what talents each person had that we could draw upon, and to foster confidence in our own visions for the project.” [SLIDE 20] Presenting the project conception and execution both at Fordham and at conferences related to medieval studies and Digital Humanities, [SLIDE 21] they gained experience advertising not only the project but also their own skills and methods as digital humanists, team leaders, and as educators. [SLIDE 22]
Outcomes and Resources:
Students involved in the Siege of Antioch Project have responded enthusiastically about the benefits of the project to their graduate school training. It is too early to see exactly where this training may lead them after they have completed their degrees (of our emeriti, one student has graduated, with the MA in Medieval Studies, in May of this year and the other is pursuing a PhD). [17] If this project is deemed to be worthwhile, what does it cost? What resources are necessary for a project like this to succeed? [18] The investment in infrastructure, over and above what students need more generally in the university environment, is relatively small: team members must have access to machines, and we must buy licenses to the oXygen XML editor. We are greatly facilitated by having our own dedicated meeting space in Medieval Studies where the weekly meetings, monthly editorial review, and quarterly video calls can take place (lunch is also often provided for the marathon sessions). More expensive is time. All central members of the project are funded graduate students at Fordham. The History Department and the Center for Medieval Studies consider the work that students undertake on the project to constitute five hours of their scheduled weekly work for each respective department. Without this time, the project would be essentially impossible, as the students would not have time for it between their coursework and work duties, and within this framework the students are in theory compensated for their labor. An even greater investment still is the university’s support for expert support staff, specifically in this case our collaborators Shawn Hill and Katherina Fostano, who can provide advice and support for many of the most technical aspects of the work undertaken by the students.
These investments are absolutely essential for the functioning of the project, and I would argue that they are rather modest compared to the potential benefits we are reaping from this experience. Perhaps, to take the broadest possible perspective, the real costs of this project, which we are still attempting to surmount, are cultural: recognition for the importance of digital training and scholarship within, first, the field of medieval studies (I think we now have that) and within our departments (this is trickier, since we battle uphill against colleagues who are not themselves invested in digital humanities). At the level of the university, in inter-university working groups, and at conferences like this one, it is exciting to hear about all of the ways that digital tools have the potential to have real and measurable impacts on the world around us. Sometimes –and I feel this too– the projects and prerogatives of medievalists can seem more distant and divorced from these lofty goals and the urgency that lies behind much contemporary digital scholarship. But it is nonetheless important that we are able to frame digital work in such a way that is not automatically or implicitly exclusive of those of us who are working on the premodern world. Medievalists hear, understand, and join their voices to calls for the relevance of digital scholarship and the urgent sense of its ability to confront contemporary injustice and inequity. As experts who study a world in which access to the written word could be socially and culturally restricted, and from which such a staggering amount of knowledge has been lost, medievalists must frequently and effectively remind our colleagues that the very act of drawing out the words from manuscripts, unread and unstudied for many centuries, and making them available online and in translation is, in itself, a restorative and ethical (if not to say righteous) act, making heard the voices that have been buried, lost and unsung.