Stylometry Notes

Plummer and his contemporaries believed that word frequency analysis showed significant difference between writers and they therefore compiled lists of rare words. Today it is more likely that studies will examine conjunctions like de, oun and kai. Why it that the usage of these type words is more revealing? Writers use these words without thinking about them and it is very difficult to imitate the details of sentence-level connectives. For example, Matthew and Luke do not use kai with nearly the same frequency as does Mark.

A recent author attribution study used the word recurrence interval based method, Trigram Markov model method and the multiple discriminant analysis of function word frequencies to determine the author of the Epistle to Hebrews. The purpose of the study was to demonstrate the effectiveness of three new advanced text authorship detection methods. Needless to say, stylometry now involves sophisticated math. This study, after testing their formulas on known authors, determined that two of the three methods are valid author attribution tools and the third method needs to be refined. The study further determined that Matthew, Mark, Luke, John or Paul did not write Hebrews and that it is likely that Barnabas wrote the Epistle to Hebrews.

I suspect that we will see more author attribution studies of Biblical literature. My July reading list will include three stylometric books. After I do a little more reading, I may explain the new methodology being employed to identify the author of an anonymous text.

Copyrighted 2006

Religion

Christianity

Gospel of Luke