Knowledge and Power Seminar Series: "The Marvels of Old in the Digital World. Toward the Methodological Turn in Middle Eastern Studies"

Por Maxim Romanov (Universität Wien)

Sala Julián Ribera 1C

Organización: Adday Hernández (ILC-CCHS) y Jan Thiele (ILC-CCHS)

Secretaria: Aurora González (ILC-CCHS)

Abstract: Over the past two decades, a significant number of printed Arabic books have been digitized in the Middle East. Now scholars anywhere in the world—not only at universities privileged with rich Middle East collections—have thousands of fully searchable volumes of classical Arabic texts at their fingertips. Due to this development, research tasks that used to take years of hard work now can be completed within mere hours. However, the field of Arabic & Islamic studies is yet to realize the profoundness of this change. Almost a century and a half ago, with the appearance of printed editions scholars began to find increasingly more texts which they could study. At the same time, the shift in the form from idiosyncratic manuscripts to normalized prints introduced “distance”—a condition of knowledge, as Franco Moretti puts it—that allowed scholars to focus their attention on the in-depth analysis of multiple texts (close reading). The change in the field went hand in hand with the change in technology. Now we are living through yet another technological shift. Unlike libraries, machine-readable corpora fuse texts into qualitatively new entities and through that promise a new form of “distance” where we will be able to focus our attention on the in-depth analysis of all available texts (close and distant reading). The digital age also brings us new computational methods that allow us to engage with these machine-readable corpora in the most efficient ways. For example, text-reuse identification methods offer us a unique view on how any text in a corpus is connected to all other texts and through that we can get a penetrating perspective on the complex interwoven texture of the Arabic written tradition itself. Making possible the extraction of meaningful data from unstructured texts, text mining methods offer ways of modeling large-scale and long-term historical processes from myriads of bits of information scattered across a corpus. The lecture will highlight significant developments and current results in the said areas and will conclude with the discussion of the issue of resources and infrastructure required for the implementation of this critical methodological turn.