Reflections

There are obvious benefits to the editing/publishing methods we have described here. As we have mentioned earlier, the standards-based digital representation of source documents makes it easier for us to query, to preserve, and to share the texts we study, and it facilitates publication across different media (both digitally and in print) and independently of individual computer programmes. The edited texts exist in an encoding which preserves the scholarly interpretation over their visual presentation, and which allows us to produce multiple visualisations of the same underlying material in both book and web form, as in the Henry III Fine Rolls project, from one single set of edited documents which are not dependent on individual software.

Meanwhile, the system used to capture entities - such as people, places, offices or subjects - within the textual sources – not only theoretically feeds structured views of content such as indices and search functionality, it also potentially allows us to create a network of associations which we can traverse from any starting point, in any direction we choose. Moreover, this does not limit us to linear enquiry, but also allows us to extract dynamic and multiple results in response to a given query, so that we can group and sort any linguistic, diachronic, onomastic or topographic information which we have captured. There is much potential in the kind of new visualisations which this kind of structured modelling enables, such as this geospatial comparison of given textual features over a given time period.

The kind of historical research represented by this project, which involves detailed editing and complex publishing requirements (which are often not fully clear at the point of editing), is well matched to a structured digital editing approach. So what are the limitations and challenges?

There are numerous potential barriers and limitations, starting with financial and technical resources. The tools and methods used here mostly followed open-source principles, and were nominally free or cheap, but in practice building a complex research platform of this kind requires a certain level of resources, both in setting up (to match particular research requirements of a project) and to sustain (although some aspects are easier to sustain than others).

The second barrier is social/pedagogical. While we found our historical collaborators open to new challenges and ideas, there is a learning process involved in learning new tools, and going from learning the initial basic technical aspects, to having enough knowledge to be able to integrate that into scholarly thinking about how to design a piece of research can be a lengthy process.

In 2011 I was invited to give a presentation at the Gerard Aylmer seminar, organised by The National Archives, The Royal Historical Society and the Institute of Historical research.  In that presentation, which centred on the calendaring projects discussed here, I argued for: greater use of common editorial schemes (where appropriate); greater co-operation between academics, lone scholars, record societies regional learned societies, digital humanists and publishers in constructing digitally mediated editing (and publishing) resources; greater work to facilitate cross-searching across corpora and greater support for training in digital methods.

The funding landscape has been altered substantially since 2011, and there has been much progress in some areas (notably with cross-searching, on resources such as the Connected Histories project https://www.connectedhistories.org/, which covers a later historical period), but there is still room for greater collaboration between the various stakeholders listed above in producing digital resources for training and research. What can we learn from the numerous digital platforms produced in the last 10-15 years, including the platform produced for the Gascon Rolls project?

There have been numerous approaches to editing and publishing historical texts and there are various practical aspects to consider here:

• General vs specific. General tools by definition can cater for a wide range of fields, but their application may be blunt for particular research questions. Specific tools generally respond well to the particular uses for which they were designed, but may have limited potential for re-use
• Ease of use vs research sophistication. Generally speaking, people prefer to use tools which are easy to learn (‘intuitive’ is a buzzword here), and digital tools produced by the research community, unlike commercial tools, often do not pay enough attention to consulting their potential users (or to getting feedback from them after they have used the tool). On the other hand, tools which are cited as being easy to use often do not offer enough sophistication for academic research of the kind discussed here.
• Experimental vs sustainable. Tools are designed to offer innovative new approaches to research questions, and for their creators the innovation is often a key part of their research contribution. But from a user perspective, it is usually important to know that a tool will be available in the future, and that there will be support beyond the initial phase of excitement around its launch.
• Standalone vs cloud. Standalone tools may be more powerful, but require updates to different platforms and operating systems over time. Tools in the cloud obviate the need for this, but require an internet connection and tend to be more limited in what they can do
• Commercial vs open access. Commercial tools sometimes provide some guarantees in terms of ease of use and support, but will the research be locked into particular frameworks? It is quite common for commercial tools to stop being supported too, or to change in ways which make export/migration difficult. Equally, open access tools provide guarantees in that both code and content are open and freely re-usable, but there are often hidden costs, for example in developing the technical apparatus required to support an open approach.

This (rather simplified) account of the practical challenges has led to a number of different approaches has led to a number of different models for how to construct research frameworks, but the most effective are often community-driven. This is the case, for example, of the Pelagios project (https://pelagios.org/), which brings together facilities for annotation, collaboration, discover, gazetteer services (to provide global identifiers for historical places) and visualisation. There has been a recent push for less resource-hungry approaches to scholarly edition – such as the Ed text editor based on a ‘minimal computing’ approach https://elotroalex.github.io/ed/. This approach has its appeal in times of increasingly limited resources, but I would argue that for the kind of research being discussed here, a community-driven collaboration (including potentially researchers, libraries, publishers and professional associations) may be the most effective. The contribution of the Gascon Rolls was to create a potential prototype model for such a project - calendaring historical records in a manner which could drive both print and digital publishing, facilitate archiving/deposit in other historical resources (an approach we followed for the Inquisitions Post Mortem project) and allowing for future experimentation/visualisation in a standards-friendly approach.

Paul Spence and Emma Tonkin August 2019, based on contributions from Paul Caton and other colleagues at King’s Digital Lab

Drucker, J. (2011a) Humanities Approaches to Graphical Display. 5 (1), . [online]. Available from: http://www.digitalhumanities.org/dhq/vol/5/1/000091/000091.html (Accessed 18 November 2019).

Drucker, J. (2011b) Humanities Approaches to Interface Theory. Culture Machine. 12 (0), . [online]. Available from: https://culturemachine.net/wp-content/uploads/2019/01/3-Humanities-434-885-1-PB.pdf (Accessed 18 November 2019).

Santoro, N., Quattrociocchi, W., Flocchini, P., Casteigts, A., & Amblard, F. (2011). Time-varying graphs and social network analysis: Temporal indicators and metrics. arXiv preprint arXiv:1102.0629.