Applying Anthony Kenny’s method from Stylometric Study of the New Testament to the Greek Septuagint. Basic approach: identify some number of (boolean) features, and count them. Kenny used 99 features from the Friberg morphology. The subset of interest here: part of speech, aggregated case, number, gender counts, and verb tense/voice/mood.
First problem: comparing by chapters doesn’t provide consistently-sized segments of text. Kenny breaks text into 50-word chunks. [numerous graphs and charts follow] Conclusion: there are some definite concentrations of future tense verbs that don’t obey the expected statistical properties of a binomial distribution.