“And now,” I said to a prospective pupil, 21, with an awe-inspiring thesis on his heart, “I've got to go home and write 750 words on chance. The editor of Tempo wants them by tomorrow.” “Chance?” he asked respectlessly, “that's not your subject. Musical determinism, yes.” He was right; in fact, as will be seen in seven hundred words' time, he had given the reason why I chose the subject. As long ago as 1966, I was struck by the philosophical and indeed musico-analytical naivety of Iannis Xenakis's (otherwise fascinating) piece ‘The Origins of Stochastic Music’ (Tempo 78), which defends the introduction of mathematics into music on the basis of the proposition that “by reason of its complexity, the strict, deterministic causality extolled by the neo-serialists was a lost cause”. I wrote my heart out when total serialism started, diagnosing the lost cause as it was born—but I never observed any complexity: a phrase of Beethoven needed far more complex analysis than a page of neo-serialism, whose saint-like simplicity seemed its only charm. What produced the illusion of complexity was inaudibility: ‘This must be very complex: I can't hear it’. It wasn't complex at all; you just couldn't hear it. Xenakis got nearer the truth when he said, “A contradiction exists between the linear polyphonic system and the result as heard (my italics), which is merely surface and mass”.