Quantifying the Quran

This paper will present the conclusions of an interdisciplinary seminar focused on a Seljuq qu’ran from Hamadan, Iran. The manuscript, shelfmark N.E.-P. 27, is dated to 1164 CE by a colophon, and is now held by the University of Pennsylvania Museum of Archaeology & Anthropology. Students with backgrounds in Near Eastern Languages and Culture and Art History collaborated to investigate the production of the book – one of the few complete Qur’an manuscripts dated to this period. Individual students produced focused studies of different features of the book. The main text of the manuscript was probably written by a single calligrapher, although several campaigns of textual corrections are scattered throughout the book. The verse markers and sura headings, however, are the product of at least four different illuminators. The book was significantly repaired sometime in the eighteenth century, perhaps when it was donated as waqf by Amir Ahmad Jawish (d. 1786) to the al-Azhar Mosque in Cairo.

Digital tools played a large role in the art historical analysis of the book. In addition to digitizing the manuscript, targeted pigments were analyzed using portable XRF analysis. Digital photos of the manuscript were enhanced to reveal the complex construction of the frontispiece. Computer vision algorithms, especially scale-invariant feature transform (SIFT) were explored but rejected. Principal component analysis (PCA) and other statistical techniques like co-occurrence analysis proved extremely useful for revealing trends in the complex and varied sura heading decoration. The question for art historians, however, is what exactly these trends represent: individual artists, formulaic models, or some combination of the two? While similar techniques are widely used by archaeologists for the analysis of geochemical data in ceramics, their deployment in art historical contexts is still developing and raises a number of methodological questions that this paper will explore.

Author: Alex Brey (Bryn Mawr)