Linked Data and Ancient Wisdom

The SAWS Project, Sharing Ancient Wisdoms: Exploring the Tradition of Greek and Arabic Wisdom Literatures, was a joint project funded by HERA (Humanities in the European Research Area) under a call to explore Cultural Dynamics. The aim was to examine Greek and Arabic collections of 'sayings', and relate them both to their sources (largely in Greek and Arabic literature), to one another, and to texts which made use of them with these links making up the core of the project.⬈#p1

The gnomologia, or collections of sayings, tended to be authored by the scribe of the manuscript in which they survive. One reason for the linked data approach was that the authoring scribe was free to modify a given collection as he thought fit resulting in similar, but distinct, variations on the collection. As these works were translated and re-translated between languages they provide an interesting entry point into both the philosophical traditions and the cultural interactions that they represent.⬈#p2

Our unit of analysis, therefore, was a 'saying', as identified and presented by the scribe of a particular manuscript. Our first challenge, therefore, was to label every such item. Over the last 18 months we have worked very closely with the CTS project, to generate unique ids; this has been slow work, but, we believe, beneficial both for us and for CTS. See http://www.ancientwisdoms.ac.uk/method/using-cts/.⬈#p3

In linking to sources, we were constrained by what is available at present. We have been able to make extensive use of texts in Perseus, but we identified many other items for which there is not, at present, a text to which we could link. We had hoped to be able to link to items in the TLG canon, but this is not yet possible.⬈#p4

We were, however, able to enrich our texts with other kinds of linked data. For authors, we were able to include links to the new Perseus catalog, http://catalog.perseus.org/, or, failing that, to VIAF, http://viaf.org/. For places we were able to use Pleiades, and even add some new locations; this meant that we were able to join Pelagios. For people named as actors, rather than authors, who occur in our 11th century 'destination' text, we were able to use links to the Prosopography of the Byzantine World (http://pbw.kcl.ac.uk) and to the Paregorios list of Roman emperors (http://www.paregorios.org/resources/roman-emperors/). For the latter, we used the long URI for the document: http://www.paregorios.org/resources/roman-emperors/about-alexios-i-komnenos rather than the DBpedia URI http://dbpedia.org/page/Alexios_I_Komnenos, since for the average reader a DBpedia page may be rather too unfriendly. We also tried to enrich the bibliography http://www.ancientwisdoms.ac.uk/library/bibliography with as many links as possible to online versions of the materials being cited. We recorded, but did not have time to include, VIAF references for many of the authors cited.⬈#p5

One challenge for us was to create a resource which makes the fullest possible use of links in machine-readable form, but also offers an environment which is welcoming for human readers. In due course it is to be hoped that CTS references can easily be resolved back into conventional representations: we did not have time to develop this. We explored this aspect in the presentation of the commentary on our 11th century text.⬈#p6

It will be very interesting to see how readers respond. One interesting outcome is that working in this way, when it is possible to go directly to the text to which one is making reference, can show up incorrect or mistaken references: the linking of data should, it can be hoped, lead to a slow steady improvement in accuracy.⬈#p7