The original virtual reality

In the temptations of vanity and the pursuit of self-knowledge, mirrors - in art as in life - have hidden depths

Mirrors, those reflective surfaces which produce an image of objects placed in front of them, recur as a pictorial motif in the history of Western art. Painters seem to have been intrigued by the relationship between the virtual reality which spontaneously appears in a mirror and the one which they artificially create by marking an unreflective surface of plaster, paper or canvas. In both cases, the observer sees something which is not where it seems to be. But in contrast to a painted image, which presupposes that the marked surface is visible, the vista which appears in a mirror requires that the reflective surface be invisible. When mirrors are represented in paintings, the situation becomes intriguingly complicated, because the virtual reality of the picture includes a second order of virtual reality in the form of a painted reflection.

The National Gallery houses perhaps the earliest and certainly one of the most striking examples of a painted reflection: the mirror in the optical centre of The Arnolfini Portrait by the 15th-century Netherlandish master Jan van Eyck. Yet it was not this famous picture which first aroused my own interest in reflections, but the behaviour of a familiar cat whom I watched some years ago from the front porch of my house.

An hour or so earlier, an unexpected downpour had left deep puddles of rainwater in the gutter. Normally the cat would have hurried across the street without pausing to examine the ground underfoot, but now it sat hesitantly at the kerbside, reluctant to go any further. From time to time, it cautiously dabbled the surface with its paw and fastidiously shook off the droplets of water. After a while it crept along the pavement to a point where the gutter was dry. Here, it scuttled across the road just as I had seen it do before. Evidently pawing the puddle had confirmed its suspicion that it would be risky to entrust its full weight to such a treacherously yielding surface. But why had the cat hesitated in the first place, and what prompted it to "test" the puddle so cautiously? One answer might be that, in contrast to the dry ground, the surface of the puddle was disconcertingly reflective. But what exactly does that mean? What does a reflective surface look like and how could a cat or anyone else distinguish it from a non-reflective one?

It was no good invoking physical optics, because that's how we explain the difference, not how we experience it. Cats know nothing about the physics of reflection, and neither do most human beings. So it must have been a question of appearances, which is another way of saying that reflective surfaces just look different. In what way though? When you ask people to characterise a reflective surface, they usually insist that it looks shiny. But not in the way the sun shines: it's the surface which gives off the shine. But the more reflective the surface - a mirror being the ideal example - the harder it is to see it, and the more difficult to claim that it looks shiny. How can something invisible be said to have a look, shiny or otherwise? Yes, but ... It would be absurd to claim that the cat, looking at the puddle, had seen nothing. Something it had seen must have caused it to hesitate and what it had seen turned out to have a surface, as the cat soon discovered by dabbling its paw in it.

All right then: although the surface of a puddle, like that of a flawless mirror, is invisible, the area it occupies is not. Pools of water, like mirrors, afford the observer a "view", though as we shall see, in contrast to the "actual" view to be seen through a window or a door, reflected views are "virtual" or apparent: they are not where they seem to be. Looking at the puddle from the opposite side of the street I could see what I knew to be the reflected image of the cloudy sky above. I doubt if the cat saw it in those terms, and that it was disconcerted by seeing clouds and foliage where they had no business to be. Still, in all probability it must have experienced the puddle as a perplexing discontinuity in the surrounding dry ground, and although it may not have recognised what it saw as displaced sky, I suspect that it did see something at a disconcerting depth below the surface of the road. In other words, the cat reacted to the "virtual" view of sky and branches above as if it were the "actual" view of an illuminated cavity.

In that sense, the cat's behaviour was comparable to that of a human infant who will consistently refuse to crawl across the edge of a trompe l'oeil "cliff" painted on the nursery floor. In fact, one of the most important functions of the visual system is to represent the mechanical supportiveness of the ground underfoot and to anticipate dangerous pitfalls. The ability to detect watery depths has the added advantage of protecting terrestrial mammals against the risk of drowning as opposed to merely stumbling.

In the natural conditions under which the visual system evolved, horizontal sheets of still water would have been the only surface whose reflectiveness represented an interpretative challenge. In nature there are no vertical surfaces offering deceptive views of a "virtual" beyond. So the invention of artificially reflective materials complicated the scene considerably, and although human beings have learnt to live in a world which now includes mirrored walls and plate-glass windows, it's easy to underestimate the perceptual problems created by surfaces which are reflective enough to provide misleading views. We take it for granted that the "virtual" view to be seen in a mirror is automatically distinguishable from the "actual" view through a doorway or a window, but the distinction requires cognitive work. It takes some time for the relevant skills to develop, and it's not until the second year of life that the human infant fully grasps the "grammar" of reflection.

It is hardly surprising that artists, who deal in representational ambiguities, should have been attracted to the motif of the mirror, and exploited it in many different ways. The mirror in van Eyck's The Arnolfini Portrait is paradoxically shiny (!), its reflected imagery rendered in invisible brush strokes and turns our attention in the opposite direction to the "actual" scene represented.

Velzquez, who would have been familiar with van Eyck's painting since it hung in the Spanish royal collection in Madrid, developed the pictorial paradoxes of reflection still further. In addition to the represented mirror, he teasingly implies an unrepresented one, without which it is difficult to imagine how he could have shown himself painting the picture we now see, Las Meninas. For to reflect the appearance of the otherwise invisible self is one of the mirror's best-known functions.

Most things which appear in a mirror duplicate what can be seen in its immediate vicinity. And even when the reflected objects are temporarily invisible in reality because they are behind us, we can see them by turning around. But for each of us, there is one item whose appearance is inescapably confined to the mirror, because there is no way of seeing it except in a mirror. Until we see ourselves reflected, we haven't the faintest idea of what the most recognisable part of us looks like. And pools of water are less helpful in this respect than the myth of Narcissus implies. In order to see your face in the horizontal surface of water you have to crane over the top of it, and from such an angle the reflected image is compromised by what can be seen through the surface. Apart from the fact that mirrors are optically insulated, so that the reflectiveness of glass is not confused by its transparency, they can be set up to afford a more convenient and often full-length view, one which corresponds to the view which others have of us.

But in order to take advantage of this amenity, it is necessary to understand what we are looking at. When a polar bear, for instance, stares at his reflection in water, it is unlikely that he would recognise himself. The chances are that, being a bear of very little brain, he, like the equally dim-witted Narcissus, thought he was looking at someone else! As it is, human beings and chimpanzees appear to be the only animals capable of identifying themselves in their reflection. Which is what we might expect, considering how anomalous the experience must be. There is something paradoxical about seeing ourselves from the viewpoint of another person, and, in the first instance at least, it must be difficult to accommodate the idea of being in two places at once. And yet by the time we are two years old, we scarcely give it another thought; and a significant proportion of human culture is based on the reflected visibility of the personal self. Apart from its obvious and indispensable use in monitoring and modifying our facial appearance - cosmetics and coiffure would be inconceivable without it - the skill of the ballet dancer depends on being able to see herself from the viewpoint of an independent observer, at least during rehearsal. For all these reasons, the mirror has assumed complex metaphorical significance, epitomising both the vice of vanity and the virtue of prudent self-knowledge.

The surfaces which furnish such experiences are ideally invisible, since anything which can be seen on them gets in the way of what is to be seen in them. But there are other artefacts whose reflectiveness is contrived to draw attention to their surface, and although this may also reflect a recognisable image, that is not the point of the exercise. The reflectiveness of jewellery, armour, ornamental tableware and "important" furniture is carefully enhanced, not to allow a view of something else, but because the resultant lustre is delightful in its own right.

The extent to which a surface reflects a recognisable image varies enormously, from the diffuse sheen of burnished copper to the representational realism of silvered glass. The depiction of such variously reflective surfaces has challenged the virtuosity of artists for more than 2,000 years, and the exhibition will, I hope, allow the visitor an unprecedented opportunity to reflect upon the somewhat neglected subject of reflection.

This is an edited extract from Jonathan Miller's introduction to `On Reflection' (National Gallery Publications, pounds 25). `Mirror Image: Jonathan Miller on Reflection', is at the National Gallery, WC2 (0171 839 3321), from Wednesday to 13 December. Four accompanying 10-minute programmes will be shown on BBC2 each Tuesday at 10.20pm until 6 October.