Pattern

What Frank Kermode brilliantly diagnoses as 'the permanent need to live by the pattern rather than the fact' [Sense of an Ending, p.11] itself expresses a will-to-pattern ... I mean in the qualification of the need as 'permanent'. Whose to say that need may not alternate with a joy in pattern-disrupting indeterminancy, and do so with no more regularity than is seen in, say, the distribution of primes?