Finding prominent differences in vocabulary between two ranges

[ct253]

Hello.

 

I am looking into linguistic differences between two sets of authors (for example, the authentic Pauline letters and the pseudo-Pauline letters; or DtrH and the Chronicler). (I have defined ranges for each of these categories.) I would like to see what linguistic/vocabulary differences (if any) are prominent when you compare one set of writings with another. So, for example, does "pseudo-Paul" talk about women more often than Paul? Or slavery? Or obedience? etc etc.

 

What I imagine must happen would be that Accordance (1) looks at one range of books and assesses what words and/or phrases are most common among that range of books (hits per 1,000 words?); (2) it then does the same with the second range of books; and then (3) Accordance subtracts the frequency of each term from its corresponding pair in the second range, and sorts the list by the greatest differences.

 

Any ideas whether or not Accordance can do this? If not, can anybody think of a viable workaround?

 

For extra credit, does anybody think that this could also be applied to grammatical structures as well? So, (to make up an example), if the Chronicler always used long flowery sentences but DtrH always used short sentences with few adjectives. Does anybody think that's something that Accordance could track down?

 

Thank you for your help!

Colin

Edited April 13, 2020 by ct253