He’s not too bothered by his dissenters. "Scientists are very hospitable generally, and artists, too, but art critics might feel threatened by some of this," he says. "They may not like the fact that I could say to you that I know that most people will respond to the beauty of the human figure when it is painted in a particular way because of the way receptors are distributed."

I wonder, in this sense, how effectively the brain can lie to itself about what it finds beautiful and what it does not. Does Zeki have data, for example, on how paintings that an observer professes to love, in fact have little neurological effect on him or her in the areas that respond to aesthetic beauty?

Zeki smiles. "That is certainly the kind of experiment that would make art critics nervous," he concedes. The closest he has come to it is an experiment he is currently preparing that will explore the context in which art is viewed. Is the brain more excited, empirically, by a painting that has a National Gallery label compared with one that is computer generated? Do we really perceive a urinal with any more aesthetic excitement if it is exhibited at the Met?

"I think one can deceive oneself and I think we often do. You only have to look at some of the things people choose to put on their walls," Zeki says, laughing. It is a little like the way that one can deceive oneself when in love. From studies he has done of students who are in the full flush of romantic love he has proved how love deactivates many of the critical sensors in the brain; it lets us believe that the person we love is perfect in every way. "It is possible," he says, "that if I had a painting worth a million pounds, the knowledge of that value, which activates a different part of the brain, might make me think it beautiful when otherwise I would not."

The people who should perhaps feel most threatened by such research, Zeki suggests, are the directors of auction houses. "Imagine if you had a hundred million pounds to spend on a painting and you had a priori knowledge of which paintings were actually objectively liked or disliked by people through scanning their reactions, as we may one day be able to do. Values could well change overnight."

Over the years, Zeki has worked extensively with artists, who have mostly been intrigued by his insights. Back in the 1980s, not long after he had done his pioneering work in defining the areas of the brain that perceive color, Zeki met Balthus at a party. He persuaded the artist to collaborate in some experiments and on a book about the neural basis of art. "He was," Zeki recalls, "extremely hostile to begin with, and extremely converted by the end."

One way of looking at art history, in Zeki’s terms, is as the progression of the human brain’s understanding of its own capacity for visual perception. Most artists may not believe that they are engaged in exactly this project, but they demonstrate it nevertheless, in their work "When an artist says: ‘How can I make a great portrait?’" Zeki observes, "what they really mean is ‘how can I represent this particular face on canvas so that it allows the brain to generalize its concept of faces and therefore becomes a great portrait?’" This desire can be tested experimentally; some cells in the brain will only "fire" with excitement when presented with particular views of the face. The greatest portrait painters have, through experiment, intuition, and skill, discovered the rules of this visual grammar.

Zeki’s latest book, which came out earlier this year, is called The Splendors and Miseries of the Brain: Love, Creativity, and the Quest for Human Happiness. The title is borrowed from a short story by Balzac in which an artist’s interior vision of the world constantly fails to match up to his representation of this vision. All great art, Zeki contends, exists in that disjunction. Artists are never satisfied, because the interior and the exterior never become united in their work. We respond often more powerfully to unfinished works — Michelangelo’s Rondanini Pieta is the perfect example — because they both attest to a perfect concept and acknowledge its impossibility.