Correspondence Analysis of Luke

Abstract

This paper applies multivariate statistics to the Gospel of Luke. The method most used is correspondence analysis. The paper tests source theories current in conventional New Testament scholarship. The text is divided into 500 word samples. A variety of criteria are used: the variables include function words, parts of speech, and some letter variables. The Infancy narratives are distinct from the rest of Luke. Within the main text the Q material is fairly distinct from the Markan material and the L material. Distinguishing the Markan material from L is not so clear. Where good separation was achieved an attempt was made to replicate this with fewer variables to identify a smaller variable set that would still effect separation.