turchin comments on One-Magisterium Bayes