Some remarks on the Typological procedures in social research