RSS Feed

Tag Archives: persian manuscript

The role of computer technology in analysing art

Whilst working at the Chester Beatty Library with the astonishing Ruzbihan Qur’an, I spent some time wondering whether there were ways in which computers could assist with our analysis of the manuscript. Some things I wanted to find out seemed almost within reach, and certainly could be done manually, such as checking to see whether the jadval (frame lines) were all exactly the same dimensions on each page. A minor thing, but it might help us to understand whether the same mistara (ruling device – see here for some very clear examples) was used throughout. Equally, I wanted to compare other divisions of the layout on each page – the size of the onvan heading panels, and the size and position of the decorative side panels. Any anomalies might help us to understand if different quires were worked on by different people, or using different tools, or maybe just at different times. However, I’m limited by my own knowledge and skills when it comes to this kind of analysis – I can manually take measurements of each page but the book is over 800 pages long, so it would certainly be laborious, and then I would have to pick out of the measurements those which matched and those which differed. I’m fairly competent in Excel so could find a way to do this, but I felt sure there must be easier ways…

…And of course there are, and already some exciting methods are being exploited to interesting ends in the fields of art history, bibliography, and codicology. Since I began thinking about this many examples have crossed my path, including a paper given at TIMA last year by PhD student Alex Brey, who used the statistical programming language R to look for variations within certain elements of a Seljuk Qur’an: for example, to find differences between the verse markers, which may help to identify the presence of different hands. His research aims to find out exactly the things which most intrigue me about Islamic manuscript workshops – how many people may have been involved with producing a manuscript; how tasks were divided up within the studio; and how (or, indeed whether) a consistent design aesthetic was achieved when undoubtedly more than one artist was involved. Alex’s blog also delves into other areas where technology interacts with art history. This post is a very interesting discussion on whether computer vision algorithms can effectively look for compositional similarities in paintings (and indeed, what benefits exist from training computers to do this).

Image matching software is another area of interest for art historians and scholars, and already has some very exciting applications.  This Ukiyo-e Search website looks for similar prints across multiple online collections, meaning that comparing impressions, colours (and fading) and so on of the same print is exceptionally fast and easy. You can also upload your own photo of a print, and search to find matches. The Broadside Ballads Online from the Bodleian Libraries has a similar function called ‘Image Search’- you can clip an area from one of the digitised ballads, and search to see where else it appears. As woodblocks for these ballads were used over and over again, by gathering together all the similar images we can also start to see how the woodblocks age as wormholes and cracks appear as white gaps in the printed image. There are a couple of very good images of this in this blog post: Of ballads and worms.  The degradation of woodblocks can be instrumental when it comes to dating and analysing print impressions: here is an excellent poster from the University of Oxford Department of Engineering Science’s Visual Geometry Group which demonstrates this, and the role computer aided image-match technology can play in understanding the order in which prints were made. In fact, if you don’t follow any of the other links in this post I would still recommend you look at this one!


Love and Devotion at the Bodleian

This week I visited the Bodleian Library’s current exhibition Love and Devotion: From Persia and Beyond, which is dedicated to illuminated manuscripts from the Persian tradition combining poetry and images. The exhibition is the perfect size for this kind of material: small enough to allow one to spend time looking closely at each object, but just big enough to leave one feeling satisfied and better informed.

Some beautiful manuscripts are on display, and the exhibition shares some wonderful stories perhaps less well-known in the Western world, such as the love stories of Layla and Majnun, Khusraw and Shirin, and Yusuf and Zulaykha. My previous post on Persian illuminations at the Chester Beatty Library features illustrations of some of these characters. I was also particularly interested to see a frontispiece signed by the master illuminator and calligrapher Ruzbihan, as whilst at the Chester Beatty Library the conservation studio spent a lot of time conserving a Qur’an signed by him. Ruzbihan came from a family of well-renowned artists and calligraphers, and as well as their skills in book arts they were also responsible for the design of various calligraphic inscriptions on buildings across Shiraz. It was said that artistic competitors of the family were “mere eaters of crumbs from their table”.

The team I worked with at the Chester Beatty are keen to continue research into the Ruzbihan Qur’an, so perhaps in the future I will have more of interest to post here. In the meantime, here is a poster I displayed recently at the Institute of Conservation’s triennial conference, which details a little about the ongoing treatment and research into this beautiful manuscript and the workshop that produced it.

The exhibition at the Bodleian finishes this weekend (April 28th), so please rush to visit it! It is also worth pointing out that their online information about the exhibition is extensive and includes high quality images, as well as a very useful interactive map and timeline.

Persian illuminations

 The three loose folios below are from the Khamsa of Amir Khusraw, an illustrated Persian manuscript in the Chester Beatty Collection, dating from 1485 and produced in Herat. The Khamsa is a collection of five classical romances from the Persian tradition. The romances featured in these particular folios are Shirin and Khusraw (left and centre), and Laila and Majnun (right). 

Persian illuminations are meticulously crafted paintings on paper. The Islamic-style paper used is surface-sized and burnished heavily to provide a smooth, glossy surface for the calligraphers and painters to work upon. The painting is then built up in layers of pigment, burnished between applications in order to create opaque, jewel-like lustrous colours. However, the layered structure of the pigment strata can cause problems with flaking and losses, whilst the characteristics of certain pigments used can cause specific problems: lead white has a tendency to flake, copper and lead- based pigments can cause corrosion and discolour, whilst silver tarnishes quickly (visible in the river in the image above right).

The materials and techniques used therefore mean that flaking pigments can be a problem inherent in this kind of art. Conservation of such objects involves examining the folio methodically under magnification in order to identify flaking, cracking,  and areas of loss, and to ascertain whether there is a risk of further losses being incurred. If at-risk areas are found, the problem pigment can be ‘consolidated’ – this means applying tiny amounts of adhesive to strengthen the bond between paper and pigment. Using a very fine brush, tiny amounts of a cellulose-ether based adhesive are introduced next to the crack or lifting flake. The paper substrate should then absorb the adhesive, and using capillary action drag it through the fibres, so that the adhesive locates itself between pigment and paper without the conservator having to touch the pigment layer directly.

© courtesy of the Trustees of The Chester Beatty Library

Khusraw entertains Shirin at his palace in Armenia (Per 163.82)

This particular folio  exhibited several areas of flaking pigments, visible in the red railings, white tiling, and the purple robe of Shirin herself. The purple pigment proved to be particularly problematic: flakes were exceedingly cupped and lifting, and introducing adhesive around the edges did not seem to encourage the pigment to flatten and re-adhere.
© courtesy of the Trustees of The Chester Beatty Library

Purple pigment under magnification

This image shows the extent to which the flakes curl away from the paper, and how blue underpainting is visible where losses have occurred.   An additional problem was that the small amount of moisture from the adhesive appeared to be causing some mobilisation of the purple pigment (a reaction that has not been seen before), suggesting the presence of an unstable organic red in the make-up of the purple.  This has prompted further research and reading on my behalf into the purples used in Persian manuscripts. Unfortunately, another method attempted (using a nebuliser to introduce a very fine mist of consolidant) did not appear to stabilise the area or secure the lifting flakes either, so in this instance it was decided that the folio is too fragile to be displayed, as further handling and movement could result in pigment loss. Further investigation into what the purple pigment might be may allow us to reassess the treatment of the folio at some point in the future. 

Conservation of the other two folios was much more straightforward. Per 163.120 (Leila and Majnun in the desert) is now on display in The Arts of the Book gallery at the Chester Beatty Library.

All images © courtesy of the Trustees of The Chester Beatty Library