The NELC Department will welcome Dr. Hythem Sidky from the University of Chicago to deliver the talk, "One Muhammad or Many Muhammads? What Stylometry Can and Can't Tell Us About Quranic Authorship"
Abstract:
Stylometry, which is the statistical analysis of literary style, has been successfully applied to literary works as a means of determining authorship attribution, authorship verification, and stylochronometry. A recent study has used stylometric analysis to argue for single authorship of the Quranic corpus, along with a new chronology. At the heart of this argument is the so-called criterion of concurrent smoothness. This principle claims that a concurrent-and-smooth evolution of multiple independent style markers in a corpus is best explained as chronological development by a single author; the Quran is demonstrated to be an example of such a corpus. Numerous objections have been raised towards this criterion, some of which are due to a fundamental misunderstanding of statistical principles. Others, however, are legitimate, and worth pursuing further. Surahs traditionally ascribed to the early Meccan phase, for example, are seen as stylistically distinct and thus the product of a different author – perhaps the only truly Muhammaden-era material.
In this talk, Dr. Sidky will explore the limits of what stylometry, and more broadly statistical analysis, can and cannot tell us about the authorship of the Quranic corpus. He will draw parallels with the Homeric Question, which is a longstanding debate over the authorship of the Illiad and Odyssey. While previous statistical investigations into Quranic authorship have relied on purely lexical markers, He introduces semantic features as well. This improves corpus resolution enables us to quantify the degree of internal dissimilarity in the text. He will also present this semantically-aware model as a tool for identifying possible redactions. Despite my more nuanced approach, definitive answers are frustrated by the brevity of the Quranic corpus which pushes the limits of statistical certainty. Dr. Sidky will conclude with a broader discussion of computational analysis: can they replace a philologist's intuition? Can they quantify the features informing this intuition, and if so, are they actually significant?
