The Marriage of Mercury and Philology: Problems and outcomes in digital philology
e-Science Institute, Edinburgh, March 25-27 2008.
(Event website; programme wiki; original call)
I was asked to summarize the third session of papers in the round table discussion this afternoon. My notes (which I hope do not misrepresent anybody’s presentation too brutally) are transcribed below.
Session 3: Methodologies
1. Federico Meschini (De Montfort University) ‘Mercury ain’t what he used to be, but was he ever? Or, do electronic scholarly editions have a mercurial attitude?’ (Tuesday, 1400)
Meschini gave a very useful summary of the issues facing editors or designers of digital critical editions. The issues he raised included:
- the need for good metadata standards to address the problems of (inevitable and to some extent desirable) incompatibility between different digital editions;
- the need for a modularized approach that can include many very specialist tools (the “lego bricks” model);
- the desirability of planning a flexible structure in advance so that the model can grow organically, along with the recognition that no markup language is complete, so all models need to be extensible.
After a brief discussion of the reference models available to the digital library world, he explained that digital critical editions are different from digital libraries, and therefore need different models. A digital edition is not merely a delivery of information, it is an environment with which a “reader” or “user” interacts. We need, therefore, to engage with the question: what are the functional requirements for text editions?
A final summary of some exciting recent movements, technologies, and discussions in online editions served as a useful reminder that far from taking for granted that we know what a digital critical edition should look like, we need to think very carefully about the issues Mechini raises and other discussions of this question.
2. Edward Vanhoutte (Royal Academy of Dutch Language and Literature, Belgium) ‘Electronic editions of two cultures –with apologies to C.P. Snow’ (Tuesday, 1500)
Vanhoutte began with the rhetorical observation that our approach to textual editions is in adequate because the editions are not as intuitive to users, flexible in what they can contain, and extensible in use and function as a household amenity such as the refrigerator. If the edition is an act of communication, an object that mediates between a text and an audience, then it fails if we do not address the “problem of two audiences” (citing Lavagnino). We serve the audience of our peers fairly well–although we should be aware that even this is a more hetereogenous and varied a group than we sometimes recognise–but the “common audience”, the readership who are not text editors themselves, are poorly served by current practice.
After some comments on different types of editions (a maximal edition containing all possible information would be too rich and complex for any one reader, so minimal editions of different kinds can be abstracted from this master, for example), and a summary of Robinson’s “fluid, cooperative, and distributed editions”, Vanhoutte made his own recommendation. We need, in summary, to teach our audience, preferably by example, how to use our editions and tools; how to replicate our work, the textual scholarship and the processes performed on it; how to interact with our editions; and how to contribute to them.
Lively discussion after this paper revolved around the question of what it means to educate your audience: writing a “how to” manual is not the best way to encourage engagement with ones work, but providing multiple interfaces, entry-points, and cross-references that illustrate the richness of the content might be more accessible.
3. Peter Robinson (ITSEE, Birmingham) ‘What we have been doing wrong in making digital editions, and how we could do better?’ (Tuesday, 1630)
Robinson began his provocative and speculative paper by considering a few projects that typify things we do and do not do well: we do not always distribute project output successfully; we do not always achieve the right level of scholarly research value. Most importantly, it is still near-impossible for a good critical scholar to create an online critical edition without technical support, funding for the costs of digitization, and a dedicated centre for the maintenance of a website. All of this means that grant funding is still needed for all digital critical work.
Robinson has a series of recommendations that, he hopes, will help to empower the individual scholar to work without the collaboration of a humanities computing centre to act as advisor, creator, librarian, and publisher:
- Make available high-quality images of all our manuscripts (this may need to be funded by a combination of goverment money, grant funding, and individual users paying for access to the results).
- Funding bodies should require the base data for all projects they fund to be released under a Creative Commons Attribution-ShareAlike license.
- Libraries and not specialist centres should hold the data of published projects.
- Commercial projects should be involved in the production of digital editions, bringing their experience of marketing and money-making to help make projects sustainable and self-funding.
- Most importantly, he proposes the adoption of common infrastructure, a set of agreed descriptors and protocols for labelling, pointing to, and sharing digital texts. An existing protocol such as the Canonical Text Services might do the job nicely.
4. Manfred Thaller (Cologne) ‘Is it more blessed to give than to receive? On the relationship between Digital Philology, Information Technology and Computer Science’ (Wednesday, 0950)
Thaller gave the last paper, on the morning of the third day of this event, in which he asked (and answered) the over-arching question: Do computer science professionals already provide everything that we need? And underlying this: Do humanists still need to engage with computer science at all? He pointed out two classes of answer to this question:
- The intellectual response: there are things that we as humanists need and that computer science is not providing. Therefore we need to engage with the specialists to help develop these tools for ourselves.
- The political response: maybe we are getting what we need already, but we will experience profitable side effects from collaborating with computer scientists, so we should do it anyway.
Thaller demonstrated via several examples that we do not in fact get everything we need from computer scientists. He pointed out that two big questions were identified in his own work twelve years ago: the need for software for dynamic editions, and the need for mass digitization. Since 1996 mass digitization has come a long way in Germany, and many projects are now underway to image millions of pages of manuscripts and incunabula in that country. Dynamic editions, on the other hand, while there has been some valuable work on tools and publications, seem very little closer than they were twelve years ago.
Most importantly, we as humanists need to recognize that any collaboration with computer scientists is a reciprocal arrangement, that we offer skills as well as receive services. One of the most difficult challenges facing computer scientists today, we hear, is to engage with, organise, and add semantic value to the mass of imprecise, ambiguous, incomplete, unstructured, and out-of-control data that is the Web. Humanists have spent the last two hundred years studying imprecise, ambiguous, incomplete, unstructured, and out-of-control materials. If we do not lend our experience and expertise to help the computer scientists solve this problem, than we can not expect free help from them to solve our problems.