Investigation into Currier’s multiple authorship theory for the Voynich manuscript

Abstract

The Voynich Manuscript (VMS) is a manuscript thought to be from medieval Europe. The VMS acquired its name from Wilfrid Voynich, a Polish book dealer who purchased the manuscript in 1912. The author(s) of the script are unknown. As is the language which the manuscript is written in, Voynichese. Various theories with regards to the manuscript’s origin, language and author have formed amongst the years. Captain Prescott H. Currier, a cryptologist who was fascinated by the mysterious VMS, hypothesized that the script was written in two statistically distinct languages and by five to eight different scribes. Scribes are medieval copyists who physically write the manuscripts. Lisa F. Davis, a paleographer and the executive director of the Medieval Academy of America, took a different route to the aforementioned analysis. Davis applied a digital paleographic tool to be able to differentiate between the handwriting of these scribes. A total of five scribes were established. Davis’ results indicate that Currier’s multiple authorship theory is plausible. This dissertation involves the statistical analysis of a transliteration of the VMS generated from the Extensible Voynich Alphabet (EVA), by using a stylometry-based approach. Stylometry examines text to identify the probable author(s) based on the author(s)’ use of certain stylometric features. Stylometric features are used in this study for measuring the linguistic style of each scribe. The most frequently used words have proven to be good stylometric features for differentiating between the five scribes. Machine learning algorithms are applied to the transliteration to find possible distinctions between the scribes identified by Davis. Unsupervised learning algorithms such as K-Means Clustering and Hierarchical Agglomerative Clustering are utilized to cluster the scribes based on these features. Additionally, supervised learning algorithms such as Multinomial Naive Bayes and Support Vector Machines (SVMs) are also applied to determine the likely scribe or scribes based on the most common words. The results suggest that the possibility of more than one scribe is likely, thus corroborating Currier’s and Davis’ findings