I usually solve problems by letting them devour me.

In order to change an existing paradigm you do not struggle to try and change the problematic model. You create a new model and make the old one obsolete.

A dull mind … is rarely able to retain the impression that the notion from which [an] inference started was purely problematic.