20 June 2019
Islamic Painted Page: Growing a Database
Today's post is by Stephen Serpell announcing the launch of the new version of his online database Islamic Painted Page, now hosted with the University of Hamburg. In a world where individual institutions still maintain their idiosyncratic approaches to locating and displaying digitised images, this resource is a major breakthrough.
Since its launch in 2013, Islamic Painted Page (IPP) has grown into a major online database of Islamicate arts of the book, with over 42,000 references to paintings, illuminations and bindings from over 270 collections around the globe – of which the British Library is one of the most important.
IPP is found at www.islamicpaintedpage.com and it does two things. First, it enables users to locate and compare works worldwide using a single database, displaying images wherever possible; and second, it signposts users onward to more authoritative sources, with hotlinks direct to the specific image pages of collection websites where available, and page-specific references for printed publications.
The website enables users to search by picture description, collection, accession number, date, place of origin, manuscript title or author, or publication – or any combination of these. So it is possible, for example, to find with a single search 77 different interpretations of the famous scene where Khusrau sees Shirin bathing, with IPP itself showing images of 36 of them.
Or one could look into the development of non-figurative illumination and page decoration during the reign of Sultan Ḥusayn Bāyqarā in Herat, 1469-1506 (70 different results); or search under an accession number to locate reproductions of works not currently published online, such as the paintings from the Topkapi Royal Turkman Khamsah H762; or search by a particular classical author, for example to study the star charts in different manuscripts of the Ṣuwar al-kawākib of al-Ṣūfī. And one can even search the contents of a publication, perhaps to check if it contains relevant illustrations, or to cross-check for metadata that was left out of the printed text (IPP is good for filling in missing details).
IPP aims to help users find not just images of works, but also articles and commentaries about them; so its search results list all the publication references it holds on each item, with the collection website location topmost if one exists. This means that well-known works return multiple “hits” in a search; for example the Miʻraj painting in the British Library’s celebrated Khamsah of Shah Tahmasp (Or. 2265, f.195r) is one of the most-published of all Islamicate miniatures and comes up with 25 references. However very few works achieve such fame, and in fact the database currently holds about 42,500 references for its total of about 30,000 separate items - so on average, each item only appears in 1.4 publications.
Multiply published: “The Miʻraj of the Prophet” from the Khamsah of Shah Tahmasp (BL Or. 2265 f.195r). Public Domain
This illustrates a further use of the database; its very large size means that it could be used as a starting point for statistical analysis, for example to chart the production of particular illustrated works against place of production or by date, or how the popularity of certain scenes has varied over time.
Finding needles in haystacks: Islamic Painted Page, main search page
The database originated simply from one individual’s frustration over the difficulties of studying Islamicate miniature paintings and illuminations, since they are dispersed all over the planet and references to them are scattered throughout a daunting corpus of literature; and even though many are now published online, it can still be very laborious to find relevant links. This led to a personal database that soon grew to point where it seemed likely to be useful to others, if only it could be placed online. A grant from the Iran Heritage Foundation made the website possible in 2013 with an initial 12,300 entries. Subsequent support from the Islamic Manuscript Association in 2015 improved the website’s utility for manuscript studies, including proper attention to transliteration. By this time the database had already grown to 20,600 references and had built in item-specific links to VIAF, WORLDCAT and FIHRIST so that users can just click to find fuller, authoritative information on authors and works, print publications, and - for UK items - manuscript details. Needless to say, a private sideline had by then become a mega-hobby.
However the most exciting subsequent step has been adding actual images of the paintings, illuminations and bindings wherever possible. Copyright prevents the database from reproducing illustrations in printed works, but IPP also covers works published online; and in many cases this has enabled IPP to show images that have been published as Creative Commons or Public Domain, or where a collection has given special permission.
It was a particular pleasure in 2018 to receive permission to incorporate images for the British Library, since it houses one of the world’s most important collections of Islamicate manuscripts and has been digitizing many of its finest holdings. Together with coverage of 19 other collections, IPP is now able to display thumbnails and larger images for about 50% of its references so far; and it is the inclusion of images that transforms the usefulness of the site for most researchers. It should be stressed that every thumbnail and every flyout image in IPP acknowledges the collection source and provides a folio-specific weblink to the relevant collection webpage, together with a recommendation to proceed to the collection website for authoritative images and other details.
Along the way, IPP has had to confront some difficult issues. Users need to be able to search efficiently, especially if they are trying to find a painting of a particular scene; but this requires consistent descriptions, whereas different authorities give different titles to the same scene (eg Khusrau sees Shirin bathing; Khosrow spies Shirin bathing; Shirin bathes observed by Khusrau….). To help manage this, IPP uses just one consistent description for each scene, but also holds the corresponding alternative descriptions. This ensures that users who cannot find what they want among the “consistent descriptions” can still search among the “alternative descriptions” if necessary.
The price for this simple-sounding device is that IPP not only has to check for consistent titling across the entire database for every new entry, but also has to maintain entire sub-databases of descriptions listing every scene encountered in each of about 30 of the most popular painting cycles, such as those illustrating the Khamsah of Niẓāmī (where artists have represented over 300 different scenes), the Haft Awrang of Jāmī and the Shāhnāmah of Firdawsī (which extends to over 1,000 scenes and where the work of the Cambridge Shāhnāmah project must be fully acknowledged). Hobbyists, beware!
Four scenes from the Shāhnāmah painting cycle (Royal Asiatic Society MS 239, ff. 7r, 16v, 32v, 44r)
Different authorities also ascribe different dates and places of origin to the same items. IPP respects this but it does result in inconsistent metadata between the relevant IPP references. And even authorities can make mistakes, or fail to provide essential details, and publications can suffer misprints; IPP has filled in a lot of missing accession numbers and corrected a lot of wrong ones.
IPP includes thousands of references to non-figurative illuminated pages and bindings, as well as covering figurative pictures; and an important upgrade is in hand to improve the detail of its 2,500 references to decorated Qurʼan pages.
IPP is an academic resource and its future clearly needs to lie with an academic institution, not with an individual. For that reason, about a year ago IPP began a relationship with the University of Hamburg’s Centre for the Study of Manuscript Cultures that aims to enrich the database’s features and extend the coverage of works published online as well as in print. One of the first fruits of this collaboration has been the re-launch of the IPP website hosted and supported by the University of Hamburg, with a new look and a number of improvements to the user interface.
Meanwhile the database continues to grow and it is planned to include more images, enlarge its coverage of collections and secondary sources from the Muslim world, and extend its geographical scope. In this way, it is hoped that IPP can act as a multi-disciplinary resource and assist not only art historians and manuscript scholars, but also contribute to digital humanities and wider cultural studies.
The author would like to thank Dr. Barbara Brend, Professor Charles Melville and Dr. Teresa Fitzherbert, as well as his own wife Elizabeth, without whose support, encouragement and patience Islamic Painted Page would never have come into being.
Stephen Serpell, Islamic Painted Page
Research Associate, Centre for the Study of Manuscript Cultures (CSMC), University of Hamburg