Two Cultures: The art of neurology challenges the science of art criticism reign

You don't have to agree with Constable that painting is a science to accept his conclusion that its pictures are experiments. To get the most out of them, we have always needed good art critics, but we are now entering an era in which our appreciation of art is being given new depth by what might seem like an unlikely profession: brain scientists.

Whenever human beings look at a painting - or do anything else, for that matter - they are using the most complicated object we know of in the entire universe: their brain. Just how the ten billion neurons and the rest of the grey matter between every human pair of ears make sense of the world promises to be a mystery for some years to come, but there's no doubt that neuroscientists are now making breathtakingly rapid progress.

Only last week, the leading science journal Nature reported another advance on how we perceive colour. Neuroscientists have long known that colour itself is not out there in the world about us but is "created" in the eye and in the brain. Now two scientists at the University of California have shown that brain cells in our cerebral cortex engage in continuously dynamic cross-talk to make sense of the light entering our eyes.

Research like this will soon be influencing how we think about painting. Meanwhile, thanks to the enterprise of the National Gallery, we have an opportunity to see the kind of light that well-established science gives to our understanding of art. In Jonathan Miller's fascinating exhibition "Mirror Image", the redoubtable doctor explores how painters use reflections and how we perceive mirror images in art and in the real world.

Miller is at his most engaging best here. By using both his artistic imagination and his formidable analytical skills, he positively teems with insights and helps us to see old pictures with new eyes. In one of my favourite moments in the exhibition Miller analyses Jan van der Heyden's charming View of the Westerkerk, Amsterdam (1660), which features houses reflected from the surface a canal flowing across the foreground. This surface has a sheen which disappears immediately if we block out the houses and leave only the water's surface exposed. What is happening here, Miller explains, is that the sheen is not included in the painting, but is "brought to" the reflected image by our sensory system. Remarkably, our brains add something to the picture it sees in front of itself.

Miller delivers analysis like this with his usual engaging didacticism. One price that we have to pay is that the exhibition is decidedly heavy on words, which bombard us from both the labels and the audioguide. Miller's interpretive style is scientific not only in content but also in tone: he is much more direct and literal-minded than is usually thought seemly in art criticism. This disjunction is, I fear, inevitable whenever scientists comment on art: to be effective, scientists attack their prey, whereas artists stalk it.

You can't help feeling sympathy for Miller: as authentic polymath, equally at home in the arts and sciences, he has to put up with being continually denigrated by cultural Lilliputians. Even if he has a thick skin, he can't have been best pleased by the reaction to "Mirrors in Mind" of some of the bien pensant art critics who have been critical of his style and choice of paintings. Brian Sewell, the London Evening Standard's resident aesthete, dismisses the exhibition with his usual weary condescension. Miller, he sighs, is guilty of follies "not of a fool but of the scientist who knows nothing of the history of art and the sensibilities of painters". He even fears that the exhibition may actually do some harm if some of its visitors "never again look at a painting for what it is but only for what Dr Miller says it is".

Mr Sewell need not fret. I suspect its visitors are far more intelligent than he gives them credit for, and that they will take away from the exhibition insights that will prove valuable whenever they come across paintings that feature mirrors (as Miller demonstrates, there are plenty of them). This is not to deny the importance of the learning and critical skills of the professional art critics: Miller's observations add to the value of their scholarship and together give us a richer appreciation of art.

It's disappointing to hear that after "Mirrors in Mind" closes on 12 December, it won't be preserved in any form apart from its sumptuous catalogue. An exhibition of this quality deserves to be on display for longer than the regulation three months and to be seen outside London. Perhaps the National Gallery could organise a touring version or, better, make a version of it permanently available through the world wide web? Whatever the exhibition's fate, I believe it will be remembered as a landmark in innovative art interpretation. Miller's meaner critics may not be aware that they are firing the first salvos of what promises to be a long battle that will end only when they accept that every painting is an a complex experiment on everyone who sees it. If we are to have the most comprehensive appreciation of a picture, we need the views of both professional arts scholars and neuroscientists. The art critics' monopoly is - whether they like it or not - coming to an end.

Graham Farmelo is Head of Exhibitions at the Science Museum, London.