Learning new strategies is called double-loop learning; learning new governing variables is called single-loop learning.
