"Pereira et al. (1993) ... have adopted techniques from mechanical
statistics ..."
(I suppose it would not be out of place to also mention Shannon's
remark, "Information theory has perhaps ballooned to an importance
beyond its actual accomplishments.")