At the annual conference of the John Donne Society in February of this year, we declared phase one of the John Donne Society’s Digital Prose Project complete. We also started a conversation about next steps in the development of the Prose Project and other digital assets related to the study of John Donne. Future blog posts will look forward to this planning process; this one looks back at what we accomplished in phase one.
The principles and objectives of the project were laid out in the 2013 article “Radiant Donne: A Case for the Digital Archive and the John Donne Society’s Digital Prose Project.” This article was a prospectus of sorts and, as it turns out, a rather ambitious one. A full assessment of how closely we have followed the direction laid out there is needed, but that is a topic for future consideration. Here I want to focus on what we accomplished and how we did it.
The goal for phase one was to produce “a complete set of transcriptions of every witness of every prose work by Donne.” This work included production of “fresh digitization of exemplars (original documents); completion and correction of EEBO-TCP transcription against these exemplars; [and] XML markup of the transcriptions using a simplification of the ‘TEI-lite’ guidelines” (197-198). All this was to be achieved without application for major funding, relying instead on a “group-sourcing” model where members of the Society volunteered their own time or sought resources from their own university to support the work in some other way (196-197).
With respect to completion, we did not quite meet the threshold for phase one anticipated at the outset of the project. As the project progressed, after an initial surge of enthusiasm and flurry of activity, it became clear that a “complete set” of transcriptions of every available witness was beyond our immediate reach. Again, more needs to be said about the merits and viability of this group-sourcing model and how it might be managed going forward, but in brief, it was a moderate success. What we have been able to accomplish is a transcription of at least one witness of each of Donne’s prose works in English and, in a few cases, multiple witnesses of some works, i.e. the Paradoxes and Problems. In most cases, we began with transcriptions provided by the Text Creation Project of EEBO. These were checked against new, high-quality images acquired for the project (with permission to publish them) through the generosity of several libraries and archives, sometimes at the request of a contributing scholar. I note in passing that academic libraries and archives are often willing partners in donating their time and resources if they are approached with a plan for making their cultural and scholarly assets widely and openly available, a mandate of many if not most research libraries. They are even more willing if approached by one of their own faculty members on behalf of such a project as ours. For example, Jesse Sharpe produced the images and secured permissions for Devotions (1624) from the St. Andrews University Library where he was employed while a grad student. Greg Kneidel requested images of Conclave Ignati (1611) from his library at University of Connecticut; and Erin Kelly acquired images of Biathanatos (1648) for the project from her library at University of Victoria. Other images and rights to reproduction came to us from the Cushing Memorial Library and Archives via the Donne Variorum project and DigitalDonne, and several print editions were acquired by David Bindle in Special Collections at the University of Saskatchewan Libraries specifically for this project. In one instance, University of Alberta, a neighbouring library (down the road on highway 16), generously loaned us their copy of LXXX Sermons so we could digitize it and use the images. Other libraries either allowed us to digitize and use their copies or in some cases provided the image and permission to use them upon request. The University of California, San Diego Library was happy to produce and supply images of LXXX Sermons and Fifty sermons (both ex libris Don Cameron Allen) and permission to use them. In a few cases where we couldn’t acquire fresh images of our own, we relied on print facsimiles that were clearer than those provided in EEBO. In the next phase of development, we will continue to seek fresh images of the seventeenth century printings and manuscripts, ideally multiple copies.
Most of these TCP transcriptions (when they were available) were quite accurate but contained errors owing to content that was unfamiliar to the transcribers (Latin), print that was small and therefore difficult to read (marginalia), or poor-quality images. The high-quality images we produced were necessary to the work of improving these transcriptions. One transcription, that of the Letters 1654, came to us from DigitalDonne and required translation of the markup system used for the Donne Variorum into TEI-compliant XML. The other major element in all cases was bringing the XML markup (available later in the project) to our project standards.
TranscriptionsFor the transcriptions themselves, we produced some 1,824,500 words in TEI-compliant, XML-encoded transcription through four kinds of user contribution. I have grouped these contributions into four classes, counting the amount of transcription accomplished by each:
At the outset, we expected that the work would be completed by scholar volunteers, and to begin, that was the case. Transcription checking was undertaken by volunteers at all stages of career, from the most senior and award-winning scholar to the most junior. Twenty-seven faculty members contributed transcriptions: of these some contributed a lot of transcription work, and others produced less. I add into this category “citizen scholars,” when in fact, we had one—a very prolific one who came to the project by way of a spouse who is a member of the John Donne Society. She produced nearly 90,000 words of transcription and accounted for eleven of the eighty sermons in LXXX Sermons.
Student Volunteers
With one exception, all our student volunteers were PhD students; and in most cases, they were assigned very limited portions, under advisement that they should not be contributing a lot of their time and effort on work that would have limited impact on their CV. Graduate students were enthusiastic contributors from the outset. Happily, of the ten PhD volunteers, five went on to find faculty positions.
Students in courses
As time went on, volunteer fatigue set in. When the Prose Project began, the Donne Variorum was at the height of its volunteer efforts, and several society members were or became heavily involved in that work. Moreover, not everyone felt that transcription checking played to their strengths (indeed, not all scholars are equally prepared for and adept at this kind of work). Some volunteers contributed a significant amount of time to this work which really has no clear place on a cv or academic report. One area of growth was the involvement of students in curricular work. Society members assigned transcription in their courses and used these assignments as opportunities to introduce textual scholarship and the digital humanities to their students, and this occurred in both graduate and undergraduate courses. All students, whether in courses or as research assistants, are credited in the metadata of these transcriptions, as are all contributors. For more on student involvement, see Nelson and Robinson 2023.
Paid Student Research Assistants
In the end, the most efficient means of completing this work was with paid student research assistants. In this mentor relationship, students could be trained, developed, and supervised, and as they became more and more expert, they required less and less supervision. Most of this funding we acquired internally at the University of Saskatchewan or was undertaken as part of an adjacent research project. Colleagues at other universities were also able to secure and donate research assistant time.
Looking forward
The forthcoming consultation and planning process will determine what transcription work we should prioritize in phase 2 and the best use of our volunteer resources. A full set of transcriptions of every available edition or manuscript witness of Donne’s prose works would amount to approximately another 950,000 words of transcription work. This, however, includes about 500,000 words that are available in the OUP Sermons Project’s transcriptions under a Creative Commons Attribution-Non Commercial 3.0 License. Moreover, some of the remaining print editions may not be crucial to project stakeholders (e.g., multiple editions of Ignatius His Conclave), who will likely (and, perhaps, rightly) prioritize transcriptions of the remaining manuscript witnesses of Donne's prose (~250,000 words of content).In any case, we now have at our disposal a base set of transcriptions covering all of Donne’s works, both poetry and prose, upon which we can further build and develop our digital resources.
Works Cited
Nelson, Brent. “Radiant Donne: A Case for the Digital Archive and the John Donne’s Society’s Digital Prose Project.” John Donne Journal 33 (2013): 175-200.
Nelson, Brent and Peter Robinson. “Curricular Contexts for Real World Research in Textual Studies.” In Digital Pedagogy and Early Modern Material Textual Studies. New Technologies in Renaissance Studies. Eds. Scott Schofield and Andie Silva, 273-294. Toronto: Iter; Tempe, AZ: Arizona Centre for Medieval and Renaissance Studies, 2023.
