One of their [compositional norms] main uses—and purposes—is revealed precisely in departures from them…. In other words, norms are not laws that must be adhered to simply for the sake of coherence or intelligibility, although that is their primary purpose. Absolutely unchallenged “normality” is perhaps the most boring mode of discourse. One rarely finds it in Haydn, or in any imaginative or interesting composer. Rather it is the existence of norms that allows departures to become meaningful—and thereby expressive. In that sense, rules are made to be broken. [Vol. II, 532-533]
And yet our system of analysis is built around the activity of noticing similarities—not deviations—in different works. What would an alternative procedure of analysis look like?