Emergency archeology 2: prototyping a digital archive of Islamic architecture
Sir Keppel Archibald Cameron Creswell (known as Archie) spent a lifetime investigating Islamic architecture: photographing and surveying sites in Egypt, Syria, Lebanon, Iraq, Palestine, Tunisia and Turkey; writing essays and books including his two-volume Early Muslim Architecture; and, in the 50 years to 1973 when he finally returned to England, he also built an extensive archive of his own and related works at the University of Cairo .
Creswell’s photographic negatives and prints, his notes, drawings and books are held across many institutions, including the Victoria & Albert Museum in London. (See the first part of this series — Emergency Archeology 1: a dig in the V&A’s photography archive). Today, the archive is stored in cupboards and shelves, tucked away in rare and hard-to-find books or in public digital files in mid- or low-resolution on grimly hard-to-search websites. We thought that this vividly accurate cultural record should be made more freely available, and especially to the scholars, architects and archeologists actively working to conserve Islamic heritage. And so we made a prototype to explore how the different parts of the archive could be connected together intelligently and opened up at the highest resolution. The goal was to demonstrate the value of a digital archive, to raise funds for digital scanning and data mark-up, and also to explore the design and technical challenges involved.
The prototype began as the work of a small combined team from the V&A and Made by Many; we approached the task according to the methods and goals described here in Experiments in Collective Curation. We set out to experiment with a better connected digital record of Creswell’s work, one that created, managed and published a digital archive that brought together the photographic record in the V&A with those in other institutions, and also with his sketches, notes and books — as well as with documents and objects in the museum that had a relationship to the sites he recorded.
A simple structure that serves the subject matter
Creswell observed and recorded sites, so we used the site as our atomic unit of data: everything relates to a site — images, plans, notes, descriptions. However as well as the individual building, the collection lends itself to organisation by building type, period and location, eg. Madrassa / Ayyubid / Aleppo (or for that matter Church / Byzantine / Aleppo etc) and so we made these filters clearly visible in the design. This approach also fits well with results of qualitative audience research by the museum which shows that people prefer to search by theme.
We modelled integration of the Creswell collection with Google image search to facilitate identification of related images of subjects not in the V&A collection but available online. A multiplicity of archives is the best feature of the archnet website, unfortunately spoiled by a recent site redesign that makes it harder to use. To the same end, we modelled API integration with other collections (eg. American University in Cairo, Harvard Library, Ashmolean etc.).
What we made is a set of concepts rendered in high res from original sketch designs; this is not however intended to be a finished design and has plenty of missing pieces, but we think it gives a good sense of what we are trying to achieve. Here’s a video showing the core elements of the prototype:
Below are further samples, reproduced as pngs and animated gifs (the prototype is made in Keynote with animations). These show an archive discovery page and two sites recorded by Archie Creswell: the Madrasa al-Halawiyya in Aleppo, and Salar & Sanjar in Cairo.
1. Overview page: Creswell archive home and discovery
2. Site page: Madrassa al-Halawiyya
3. Site page: Salar & Sanjar
4. Detail: individual photograph (enhanced museum object)
Our designs assumed that we would exploit open source (principally iiif) viewing applications and standards, but we also explored pairing images for ready comparison which as far as we know is not yet available as a standard in iiif but which we found to be a powerful feature. In every case we assumed the availability of high resolution colour files, global zoom and point annotation of key features within images.
5. Detail: zoom and point annotation
6. Detail: pair comparisons and timeline
We combined Creswell’s photographs with non photographic content from the V&A collection (prints, documents, books, objects, cameras) including Creswell’s own works on Islamic Architecture, now out of print, and other works in the V&A National Art Library. We assumed integration with Google books and the library online ordering system as ways of achieving this. We also assumed the presence of relational data within the improved V&A catalogue, combined with an existing data structure in the new V&A webite, to make it easy to link catalogue objects to each other and to website articles or exhibitions in which they are included or which have connected subject matter. The latter already exists in the data structure.
In sum, this was not a particularly difficult design challenge (our prototype demonstrates a good way of making the archive accessible) nor a complex technical challenge. An improved web platform (disclosure — created by Made by Many and the V&A digital team) will soon be joined by an improved digital asset management system. There are also plenty of open source and third party enablers to choose from (iiif and Google, largely) to reduce the technical development work required. New APIs will need to be written to connect other institutions with the V&A, but for a competent software engineer this is not hard to do well or quickly. Completing the scanning of the collection to a conservation standard (ie. recto and verso, print and mount, at a resolution sufficient to extract all of the image detail) is also relatively straightforward with good equipment, process and handling; and while it’s easy to over-complicate the making of the data, as long as one takes the approach of experimenting, testing and learning first, it is again easy enough to get this right and to build the data around the object necessary to make it (the object and data) both accessible and useful. The really wicked problem, however, is that each of these relatively trivial pieces of work involve combining together different teams and departments and — in total — a considerable resource over time. The challenge is all about people and institutional cooperation, and finding the leadership and commitment to enable people and institutions to work together to a common purpose.
This article is number two in a series:
For more detail on the Creswell archive at the V&A see
Emergency archeology 1 — a dig in the photographic archive.
For more on the investigation into online visual data on the Madrasa al-Halawiyye see
Emergency archeology 3: a century of change at Aleppo’s Madrasa al-Halawiyya
With thanks to Omniya Abdel Barr, Erika Lederman, Ella Ravilious, Bill Sherman, Chris Pearson and Richard Palmer at the V&A and to Ross Burns. Prototypes by Kristof Goossens @ Made by Many.
Also see: Culture in Crisis: from Cairo to Aleppo at the V&A 3rd November.
