Estimates distinctiveness in speakers' (authors') style. Fits models that can be used for predicting speakers of new texts. Methods developed in Spirling et al (2018) <doi:10.2139/ssrn.3235506> (working paper).
