When I first went to teach at the Bartlett School of Architecture, Peter Cowan asked me what my research would be. I answered diffidently: ‘meaning in architecture’, and he grinned sceptically, and said: ‘your first job will be to find something to measure’. But I had been seduced by reading Stendhal's De l'amour into believing that measurement wasn't necessary for insight; and persuaded by reading Wellek and Warren on Theory of Literature that theory wasn't to be restricted to physical science. So I was soon giving a course called Meaning in Architecture, without measuring anything, but using lots of comparisons (twin slide projectors had become all the rage).