By Alex Butterworth and Christophe Leclercq
The seventh day of the Institute once again fell into two halves, a morning devoted to discussion in our base in the Liddell-Hart Room and an afternoon visiting the British Museum’s Research Space team. Both took a round table form, with the morning focused on collective reflection on where we had got to over the previous week and where we were headed, both in relation to the most useful categories of output towards which we should work, and the most effective reformulation of groups to deliver them.
Three different broad objectives were agreed, of various ambition and immediate utility, that were further grouped into deliverables to which different interest groups could might devote themselves. These included vision documents, prototype interface design, protocols, and cookbooks to guide practitioners in starting to work with art historical linked open data. Participants were regrouped and again assigned arbitrarily to these groups, without specific designations of group focus, to allow untrammelled exploration. It is an approach that promises an intensive concluding discussion at the end of the week, when we will bring all the ideas generated back into alignment, to define the speculative tool for which Stuart Dunn has coined the title, the ‘Itineraries Generator’.
If the Institute group needed any example of ambition and dedication harnessed in a greater cause, it was provided by the British Museum team of Dominic Oldman and Diana Tanase and their ResearchSpace project. The ResearchSpace was conceived to address the need for a shared digital research infrastructure that superseded inherited practices of classifying information, using Linked Open Data, and has been developed through various iterations over a ten year period, with support from the Andrew W. Mellon foundation. The current iteration, the fruit of three years work, provides an elegant interface for the exploration of entities within the British Museum collection, leveraging a vast knowledge graph of semantically linked metadata, encoded using the CIDOC-CRM schema.
The team were generous with their time, offering a long presentation that covered the principles that underlay their approach to Linked Open Data, the practicalities of working with curators whose rich implicit knowledge the ResearchSpace could capture and preserve, and the promise of a new way to construct narrative and argumentation. These were presented as superstructures of knowledge, constructed of generic components, that remained rooted in the data and could be contested or elaborated, with specific reference to sources, within the digital environment. After much talk this week around representing the provenance of an object, Oldman drew our attention to an equivalent requirement that ResearchSpace addresses: preserving ‘knowledge provenance.’
A primary concern must be to address what art historians and curators can now do in the digital space that they could not previously in the physical one, given the specific constraints of the latter. However, it is important to remember the biases and constraints that may be in implicit in the digital tools currently used by art historians. For their new digital methodologies to respect but extend practices rooted in their disciplinary traditions, it is crucial that art historian’s own research questions and the conceptual frameworks that they employ to address them are fully reflected in the design of those tools. Only then can the processes of art historical enquiry develop in directions that offer genuinely new possibilities, whilst remaining rigorously grounded.
On first arriving in the museum, we had been taken into the Enlightenment Gallery, which recreates an eighteenth century organisation of material evidence in the service of knowledge generation, in part to encourage what Neil MacGregor described at its opening in the early 2000s as a “dialogue between civilisations.” There Oldman had shown us a wooden aboriginal shield, supposed to have been present at Captain Cook’s landing in Botany Bay and bearing a hole that may have been made by a bullet from a British gun. The caption for the shield offered one articulate, authoritative but static explanation, situated close to the object itself, which sat in a glass cabinet in a long wall of such cabinets, each containing bewilderingly diverse objects, within a vast room housing encyclopedic eclecticism.
Within the ResearchSpace digital environment the record of the shield could carry within its metadata all the curatorial knowledge offered by the caption but far more too, situating it within a knowledge graph that might also hold the record of its situated biography, from creation through the impact of the bullet to its current home behind the vitrine. Our Institute group had spoken in our morning session about our collective ambitions for the future tools of art history that would enable the myriad geographical and conceptual traces through datascapes, and their constellations of possible construction. To see the same object in both its physical and semantically modelled context, in quick succession, brought home the extent and significance of the challenge, and the fresh perspectives that exhibition studies might offer.