Linguistic Dating of Biblical Texts using Supervised Machine Learning

Abstract

One of the major concerns in contemporary Bible scholarship is the problem of dating biblical texts. Our aim is to use an algorithmic approach known as supervised machine learning to utilize linguistic features of a biblical text to assign it either to the category of Early Biblical Hebrew or Late Biblical Hebrew, thereby informing us as to whether the text was written before or after the Babylonian exile. In addition to using this methodology to classify every book in the Bible, we perform a detailed case study on the book of Joel and demonstrate how our approach can be used to solve the long-debated question as to whether the book was written in the pre-exilic or postexilic period. We also use the algorithm to individually classify each psalm in the book of Psalms , as most scholars believe that the book of Psalms was written over a long period of time by a variety of authors.