What is Painting?

The first words in Western culture concerning man-made images categorically warn us against them:

Thou shalt not make unto thee any graven image, or any likeness of any thing that is in heaven above, or that is in the earth beneath, or that is in the water under the earth: thou shalt not bow down thyself to them, nor serve them.

Why? What makes the God of the Bible forbid all likenesses in the Second Commandment, before pronouncing on killing, adultery and theft?

Images draw us in. The desire to make and attend to likenesses has been powerful throughout history. Their presence in human society has been the rule, and their absence the exception, during the last six thousand years. The Ten Commandments, however, set a control on this desire to make likenesses; but even a people bound to Moses’ law, as the Old Testament relates, needs to be repeatedly restrained from making and contemplating ‘idols’.

How man-made images have been understood in history is the concern of this book. In particular, it examines the images called paintings, and specifically the history of the last two hundred years, during which painting seemed to part company with the age-old desire for likenesses. But this development of modern art needs to be understood within much larger traditions. A contest of feelings about images stretching back to ancient Israel and Greece underlies the complex values we now give to that word ‘painting’.

For the Bible, an image may be any object made by humans so as to look like something first made by God. The prophets of the Old Testament deplore the fact that people, creating these secondary objects, start to revere the work of their own hands. In doing so, they turn their backs on the original maker of objects. Instead, so the Hebrew tradition accuses, they regard what they have made as a visible god. But God is the maker not the made, and is not to be seen; God lies behind sight.

Detail of a relief from the tomb of Ankhmahor, Saqqara, showing Egyptian sculptors working on two statues, sixth dynasty, c. 2345–2181 BCE

A Jewish text from the 1st century BCE, the Wisdom of Solomon, dwells on idolatry with scornful amazement. A carpenter takes an offcut of wood and ‘by the skill of his understanding, fashions it to the likeness of a man’; then, fastening it to the wall to stop it falling, ‘for it is an image, and hath need of help’, he ‘for life, prayeth to that which is dead’. Do not be deceived, the writer adds, by ‘an image spotted with divers colours, the painter’s fruitless labour’.

The writer of these words was very likely casting aspersions on the visual culture which dominated his era and which still influences ours: that of ancient Greece. Greek accounts of image-making and its origins are at first sight quite different. The legend given by Pliny, for instance, is sweetly romantic. A maid of Corinth was bidding farewell to her lover, who was about to cross the sea; noticing his shadow cast on the wall by a candle, she seized a stick of charcoal from the fire and traced its outline. The scene was often depicted by late-18th-century painters, no doubt influenced by the contemporary fashion for silhouettes.

David Allan, The Origin of Painting, 1775

The impulse of the Corinthian maid, in the legend, was enough to initiate likeness-making. Yet that impulse was prompted by an imminent absence, the departure of her lover. For her, the tracing was evidently a second-best, a substitute: ‘Only a picture’, as we say. In fact, for all their dissimilarity, the Greek storytellers and the Hebrew prophets share several common premises about man-made images. They take it that an image is not the original thing, but is – or ought to be seen as – a substitute for it; and that, nevertheless, it has a fascination in its own right, a power to command the eyes. In sum, they believe that the image is initially made as a channel for human desires, but that it subsequently diverts them.

The Vernicle, a Russian two-sided icon, 12th century

These intuitions about images spread very wide in human culture. Turning to the image to answer our wishes, we take it to have a dependent, secondary relation to the original that is its cause. If we rely on it to satisfy our demands, this is often because we believe that it is itself reliably attached to the object for which it is a substitute. This attachment may have a physical status: the camera’s mechanical record of transmitted light is a guarantee of this sort. But even if such a material attachment is lacking, we may uphold the status of the image by a principle of spiritual attachment. For instance, we may say that the likeness of someone holy or divine has been directly imprinted on the working material by an inspiration from above, with the human maker merely functioning as a tool. Thus in the Orthodox Church, icons of the Saviour were produced with minimum deviation century after century because they were intended to be acheiropoietai, ‘things made without hands’. The lack of individualized intervention made them transparent receptacles for their divine content.

Yet this appeal to divine inspiration was used to defend images against the contrary intuition: that they may deflect our attention from what they represent. The visible substitute may gain favour at the expense of the absent, invisible person or entity it stands for. Between the original and the viewer the seductive power of image-making intervenes, a force that may be seen as magic and malign. This suspicion was voiced not only in the Jewish tradition but also, with an added twist, in the early 4th century BCE, by Plato in The Republic. Painters, Plato accuses, distract our attention with their likenesses of the way things look; yet the way things look is itself only a poor likeness of their true nature. For truth belongs with the idea: the permanent, God-given form behind each appearance we perceive. Copying mere appearances, the painter ‘knows nothing worth mentioning about the subjects he represents’, and we can conclude, Plato writes dismissively, that ‘art is a form of play, not to be taken seriously’.

In this attack on image-making, Plato – following his mentor Socrates – characterized the practice as mimesis, a noun formation from the verb mimeisthai, meaning ‘to mime’, a bodily performance in which you tell a story, or bring someone’s presence to mind, without speaking. The insistent, attention-seeking aping of appearances that invited Plato’s scorn may perhaps be seen in the few fragments of sophisticated fresco painting to have survived from his day, such as the tomb mural discovered at Vergina in northern Greece. We might call it ‘painterly flair’, but in terms of Plato’s theocentric approach, it was triviality on triviality, at a double remove from the truth. Mimesis, however, was to become the principle that would underpin the intellectual status of Western painting for centuries.

The development is due to Aristotle, writing a generation after Plato. Rather than attempting to transcend the visible world, the younger philosopher was concerned with giving a coherent account of it, and he saw mimesis as a vital tool in doing so. He used the term to describe something that happens throughout painting, sculpture, poetry and drama, and also in some forms of music. Through these means, wrote Aristotle, people enact the likenesses of other persons or phenomena – that is, they ‘put themselves into them’, as it were – as a way of comprehending them. ‘It is through mimesis that man develops his earliest understanding.’ Children, then, with their make-believe and toys – their ‘early learning aids’, as we now say – were included in Aristotle’s psychology. All such activities, from his point of view, offered a sane way of widening the mind’s access to the world, with the potential to cleanse the emotions rather than to tantalize and warp them. Play might be an alternative to work, yet play still had a purpose.

The Rape of Persephone by Hades (detail), Vergina, c. 340–330 BCE

This acceptance of images by Aristotle on the one hand, and the Platonic and Judaic rejection of them on the other, hinge on an ancient and well-trodden opposition of terms: knowledge versus feeling, logic versus intuition, head versus heart. Aristotle inclines to see images as coming from one side of the opposition: people make them because of their wish for knowledge. Plato suspects them of coming from the other: people make them to indulge their desires – vain desires, from his point of view.

The battle between these two viewpoints has not gone away. Among Christian societies, proponents of Plato’s suspicion have included the original ‘iconoclasts’, who banished images from churches in the Byzantine Empire during the 8th century, and the Puritans who smashed up much of Britain’s artistic past during the 16th and 17th centuries. In their wake, comparable destructions of art were carried out by secularizers during the French Revolution and China’s 1960s ‘Cultural Revolution’. Meanwhile Muslims, from the time of the Prophet onwards, have followed Jews in being wary of figural imagery, and most recently it has been Islamist movements which have courted attention by carrying out spectacular acts of iconoclastic violence. There is an understandable impulse to call their perpetrators ‘barbarians’, but the supposed point of principle behind such actions – that images are made to cater to human desires, and that they can divert them – is one that any form of civilization has to acknowledge, for instance via its policies towards pornography. As long as images continue to have force, so do objections to them.

Despite these resistances, most Christian traditions have highly valued image-making, as have the secular ideologies that sprang up in their wake. And along with Aristotle’s ready acceptance that it was normal for people to make images, cultures that cared for painting, sculpture and poetry accepted his tenet that mimesis was a common rationale for all three arts. Up until the later 18th century, when ideas of painting started to expand into the diversity we now call ‘modern art’, most writers who discussed the practice took it as an assumption that whatever else they might mean to do, painters were satisfying the common and reasonable wish that people had to have ‘mimicked’ on flat surfaces sights that they might not otherwise be able to see.

Mimesis was generally translated, in the academies where this theory grew from the Renaissance onward, as imitation; in other words, it was accepted that the first business of painters was to make a resemblance of something. That something bore the general name of nature. In what is often termed ‘the classical theory of painting’, the standard epithet for the painter’s art was thus ‘the imitation of nature’. Neither term of this formula, however, was entirely unambiguous.

‘Imitation’ seems to suggest something like the photograph of the English painter Charles Wellington Furse at work in 1903. We see him making a copy of what he sees (more exactly, of what the photographer sees – the photographer having borrowed his viewpoint). With his brush and oils, he makes the canvas show, as much as possible, the same pattern of light that the figures present. A faithful resemblance has been performed by skill.

But of course this account ignores the studio props that help give the photograph its farcical air. On the canvas, the sporty angler and his adoring spouse are being conjoined to a rocky riverbank, presumably rendered from sketches done outdoors. The imitation of one part of visible nature is being stitched together with the imitation of another part. The principle of imitation, as taught in Europe’s academies, not only allowed for this selective approach to nature, it demanded it. Painters should imitate what was most beautiful and most significant in nature; they should not merely copy its every detail without discrimination.

A photogravure of Charles Wellington Furse at work by an unknown photographer, 1903

How should they ascertain what was beautiful and what was significant? By consulting the great works of the past, came the most consistent answer, works such as the sculptures of classical Greece and Rome or the paintings of Raphael. These works would educate the eyes of painters, who would then seek out such beauty in the visible world and relay it selectively in new arrangements. Painting properly practised, by this line of reasoning, was thus a negotiation between two imitations: not only the imitation of nature, but also the imitation of art.

Nature itself, however, was also a double-edged concept. Contrasting art (or ‘culture’, or ‘nurture’) with nature suggests that there are two kinds of material in the world: that which humans have made, and that which God has made. God made things first, and there is a large body of opinion that asserts he made them better: that mountains and flowers are superior to buildings and paintings. The preference for the original as against the secondary, implicit in the Second Commandment and much restated in the Book of Job, is deeply rooted in many religions. And painters themselves, from ancient times, have deferred to this reverence, or nostalgia, by offering views of landscape to the wistful urbanite. Likewise, the nakedness we have from birth (natura is, at root, ‘what is born’) can be seen as finer than the fashions and face paintings we add to it; hence the nude’s centrality in classical art.

But there is more in the world to imitate than landscape and the nude. If ‘the imitation of nature’ is to be a general account of painting, there can only be, in fact, one kind of material in the world: visible material, made no matter how. In this sense, all we see is nature. (After all, if humanity is a work of God, humanity’s own works merely amount to a kind of subcontracted production.) ‘Nature’ thus might mean ‘the visible world’ – and also, by extension, the principle that holds that world together, the ‘great creating nature’ of Shakespeare’s The Winter’s Tale.

The concepts of ‘imitation’ and ‘nature’ could be combined to offer a complex account of painting in classical theory; but they needed supplementing by terms such as imaginatio and idea to cover all aspects of painting practice. The former was required to account for the obvious fact that painters have never been content merely to imitate what is given by nature – they also make things up. Yet, in academic doctrine, this capacity to imagine was ultimately a personal and selective approach to the imitation of nature. The phantasms in Bruegel’s Dulle Griet may look unnatural, but Bruegel would have cooked them up using perfectly natural ingredients: stir in the legs of a cripple with a half-gutted fledgling; pivot a head round an anus; serve up a slice of hell’s horrors. As Albrecht Dürer had previously prescribed, ‘If a person wants to create the stuff that dreams are made of, let him freely mix all sorts of creatures.’ But both concoctions of this kind – ‘grotesques’ – and caricatures, in which nature was deliberately pulled out of shape, showed the imagination serving ‘low’ and trivial ends.

Idea, however, came from on high. The idea, for Plato, was how the object existed in God’s mind, rather than in ours. How could we hope to visualize such an image, and thus attain to a beauty that was essentially divine? The answer given by 16th-century theorists was through line. When I look before me and draw a line, my mind is performing an act of analysis; for things do not present themselves to the eye wrapped, as it were, in linear packaging. I am discovering, or picking out the form – ‘form’ being the common translation of the Greek idea. The impetus to do so comes from the discriminating, God-directed intelligence; hence the upward gaze shown by the allegorical figure in the engraving Idea illustrated on page 18. Beauty – that which the painter hoped to achieve – would proceed from a well-directed understanding of form. It might be defined through a selective recourse to Plato’s theories: it was ‘ideal’, it was a glimpse of the way that God knew things. (The further implications of this idea of ‘form’ are pursued in chapter 3.)

Pieter Bruegel the Elder, Dulle Griet (detail), 1562–6

It was when theorists began to give imaginatio the value previously reserved for idea, in the later 18th century, that these classical theories of painting started to buckle. The new proposal, for the thinkers of this era – one much stressed in the writings of Romantic poets such as William Wordsworth – was that the imagination itself was divine. And with this conflation of values the other terms of the classical system – imitation and nature – had to shift. As James Jefferys, a young English draughtsman working up turbid nightmares of enchained giants in Rome in the 1770s, put it: ‘I may be excused for putting bones and joints out of their place, as painting in this sense is no longer Imitative but Creative.’ Imitation was out for the forward-looking painters of the era, and its role taken up by ‘creation’ or ‘expression’. Painters no longer saw themselves as copying ‘nature’, reproducing the facts of the visible world; they were now themselves vehicles for the principle of ‘nature’, the creative force that made things visible. Therefore, by this argument – the working premise of painters as different as William Blake and Caspar David Friedrich – we should understand the actual painting not only as a representation of the seen world, but also as an expression of the painter and the unseen feelings and creative power to which he has access.

Simon Thomassin after Errard, Idea, Conceptus imaginatio, Natura and Imitatio sapiens, from Fréart de Chambray’s Parallèle de l’architecture antique et moderne, 1702

Of course, it is true that many paintings produced before the 18th century seem to invite us to do both. Rembrandt’s Woman Bathing in a Stream of 1654, in common with Vermeer’s slightly later The Lacemaker, focuses on a figure who is herself acutely focused on an action she performs. But whereas the calm, limpid glazes of the latter seems to state ‘This is how this woman appears to the eyes, a represented visual phenomenon,’ the brusque thrusts of buttery oil paint with which Rembrandt finishes off his rendering of the figure amount to a declaration: ‘This is how I feel the woman is: these are my hand movements enacting what it is to be her.’

To propose that the painter’s inward experience should form the main burden of the painting, however, required a break with the whole idea of imitation. This shift away from the concept was to have far-reaching implications for the practice of painting, leading to much of what we mean by the term ‘modern art’. In fact, exploring these implications, I will argue, has occupied many painters up to the present day, and explaining them will occupy much of this book. Chapter 2 will trace what happened to the idea of imitation, while chapter 4 will survey the development of the ethos of expression.

James Jefferys, Prisoners, c. 1779

Firstly, however, I need to establish what I mean by painting; since it is painting, rather than ‘images’ or ‘the arts’ in general, that I hope to explain. Arguments about ‘images’ may extend equally to work in either two or three dimensions, and as much to mechanically produced likenesses as to handmade objects. ‘The arts’ are a category we owe ultimately to Aristotle, who sought to account not only for painting but also for sculpture, drama and poetry with his doctrines of mimesis. ‘Painting’ may well fall within these two broader categories, but it also forms its own distinct category, and does so in two ways.

Rembrandt van Rijn, Woman Bathing in a Stream, 1654

Johannes Vermeer, The Lacemaker, c. 1670

There are paintings. People across the world have fixed pigments to all kinds of surfaces – walls, hoardings, panels, canvases, pavements, box lids and so on – aiming to do something more than hide or protect the raw material beneath. When we call the results ‘a painting’, we mean that a set of the resulting marks comes together to claim our attention. Calling it by that name, we think of it as a particular type of human product.

And there is painting. This is the activity that unites these objects, one that makes us think of them as demonstrations of a certain ‘art’, rather than as surfaces that merely happen to have paint on them. Painting is the manner of production that distinguishes these flat things from other sorts of marked surface – from writings, photographs and so on. If the objects are products, then painting – the concept that ties a set of visual experiences together – is a practice.

What does this practice, this unifying factor, consist of? We have seen that until the 18th century, at the least, it had one common, minimal definition in the West: painting was the marking of surfaces so as to represent visible things. To restate this still widely entertained idea: paintings present us with the sight of things we might see otherwise. In particular, they use two dimensions to show us objects we might potentially see in three. What we see on the flat corresponds on a basic level to what we might see in the round. To take an example of the practice at its most vigorously direct: the painting by Hermenegildo Bustos, a self-taught 19th-century Mexican artist, offers you the sight of fruit, the like of which you might pick up and eat. Producing this experience of ‘pictorial representation’, so the idea goes, is the distinctive business of ‘the art of painting’, separating it from ‘painting and decorating’, the application of paint to cover and adorn surfaces.

This definition of painting, which effectively restates the theory of the imitation of nature, no longer looks adequate. It has quite evidently been insufficient since the advent of abstract painting around 1910, but it had already been superseded in the eyes of many painters since the later 18th century, as we have just seen. But to get the full measure of this change, or reformulation, of painters’ practice we need to step back for a while from historical arguments and expand on our definition of the objects called paintings.

A painting is the flat thing formed by marks and a surface, I wrote above. ‘Surface’ seems reasonably unambiguous. Any visible, solid expanse which stops the passage of light could be used as a ‘ground’ on which painters could do their painting; but once it is marked, the surface of the ground will become an integral part of that painting.

Hermenegildo Bustos, Still Life with Fruit and Frog, 1874

‘Flat’ seems clear enough, too, since it seems to be only the surface that really concerns the painter. It might, it is true, be more accurate to say ‘flattish’, if we consider the way paints can cake up and the complex signals our eyes may have to interpret about the under- and over-layers of pigment that the painter has applied. Nonetheless we habitually scan the results of that work as a more or less two-dimensional spread.

The word ‘marks’, however, requires more explanation. Marks are made not only by brushes, pencils and similar graphic (that is, ‘marking’) tools, but also by tyres skidding, boot studs on a pitch, pebbles scuffing one another as the tide rolls them and so on. A mark is whatever we see that we recognize as having a cause – whether that cause is intentional or not. We see it and we see past it, or into it; it is what it is and a reminder of something besides. It is when we see something in that double, ambivalent manner that we call it a mark.

As such, a mark is a sign you can see. A sign, in logic, is something which points beyond itself; something which means. Obviously, the prime example of a sign you cannot see is the spoken word. Semiotics – the study of signs, which has greatly influenced recent theories of art, as we will see in chapter 6 – compares and sorts words, marks and whatever else has significance. From its point of view, the relationship between the thing itself – the sound, the brushstroke or the scuff on the pebble – and the thing it points the mind towards is one of ‘denotation’, or, quite simply, ‘representation’.

This is to say that ‘representation’, as we know from our common use of the word, has a more general meaning than the relationship of ‘pictorial representation’ – that is, between picture and seen thing – described above. It is equally the relationship between the seen thing and meaning. Thus in the work of a painter such as Jan van Eyck, the painted canvas itself represents (beneath the central arch) a vase of lilies; but on another level, the lilies represent the Virgin’s chastity. The former is pictorial representation – the criterion of whether or not a painting is ‘representational’ – while the latter is symbolic representation, on which depends the painting’s ability to communicate meanings.

If paintings are a type of mark, representing in either of these senses, then as part of the whole category of signs they are comparable to words. But – and this is one of the concerns of the semiotician – they seem able to attach themselves more strongly to what they represent than words can. The word ‘stone’ does not in itself have anything very stony about it; it is no stonier than pierre, lithos or kő. Tom Walker’s Stone, however, does seem stony; it appears to offer our eyes much the same information as a stone itself would, were it present in all its solidity before us; and this would be the same, whether we spoke French, Greek or Hungarian.

The picture, in this sense, was traditionally defined as a natural representation – one that works by presenting again what nature presented first. The spoken or written word by contrast was a customary or conventional representation – one that works because there is a custom or convention among a certain group of language users that a certain sound denotes a certain thing, and one that works for no other reason.

Jan van Eyck, The Virgin with Chancellor Rolin (detail), c. 1436

Tom Walker, Stone, 1995

We are back to the distinction between what comes from God and what comes merely from humans. And this leads to a certain paradox: the same impulse that leads the Jewish tradition to favour the divinely created wild flower above the humanly fashioned painting directed Greek philosophers to grant precedence to the painting, which is representation using ‘natural’ means, over the spoken word, which is representation by human convention. Thus the wordsmith Plato, for all his scorn of painting, composes a hymn of envy to the art in the dialogue Cratylus, in which he tries through far-fetched etymologies to give words the same stable status of transmitted natural fact that pictures seem to have: as if all speech and writing were like the letter ‘O’, which pictures the shape of the mouth that pronounces it.

Likewise, when Aristotle uses the blanket term mimesis to cover poetry as well as painting, he assumes that even if words do not individually resemble what they refer to, the grouping of them in poetry is aimed to work in the same way that painting does. The poet, as it were, ‘paints pictures’ in the minds of his audience. Thus the evidence of the eyes (from the Latin videre, meaning ‘to see’) helps us to envisage what poetry is like (a tautology, when so much of our vocabulary leans on the idea of vision), and if we seek a prototype for all forms of representation, it is to painting we should look.

Mimesis is imitation is representation. This was the assumption prevalent in Europe for two millennia after Aristotle, the great philosophical categorizer. And if imitation is a skill to be perfected, so therefore is representation. The classic account of excellence in painting is the Greek story of the two rivals Zeuxis and Parrhasios. Zeuxis painted grapes so lifelike that the birds tried to peck at them; but when he wished to inspect Parrhasios’ showpiece he himself was fooled – the curtain he tried to pluck turned out to be the painting. By this standard, trompe l’oeil, in which human work most closely replicates nature’s work, is the definitive mode of painting.

This was an ethos enthusiastically adopted by the Romans, as can be seen in surviving murals from their villas. While illusionistic painting was marginalized for a thousand years after the rise of Christianity, it was reanimated at the turn of the 14th century by Giotto – as much in the eye-teasers he appended to the Arena Chapel in Padua as in the solidly embodied figures of his stories. With these, Giotto established the idea that a painting could be like a window – an idea that would be given systematic form when Filippo Brunelleschi developed the technique of ‘artificial perspective’ around the year 1423. ‘Natural perspective’ consists of the geometric rules that define how light comes to the eye; Brunelleschi showed how these rules could also be used to define what can be shown in a painting, and the way things appear in it in relation to the angle of vision (their ‘aspects’).

Even more important than the practical usefulness of the techniques of perspective – which painters generally used at their discretion, rather than as fixed rules to be obeyed – was the intellectual status they lent to painting. According to Leon Battista Alberti and Leonardo da Vinci, the major Renaissance theorists, this apparent window could open out on to the whole of nature: painting had the potential to offer a complete representation of the cosmos in all its diversity. Analysing and reconstituting all visible phenomena, painting became, for Leonardo, a primary demonstration of humanity’s intellectual dignity. This was an argument that greatly empowered the painter. Vindicated by Leonardo’s own investigations, it also applies to a line of artists before and after him: Van Eyck, Bosch, Michelangelo, Raphael, Dürer, Titian, Bruegel. Each of these Renaissance originals clearly extended the range of what might be represented in paintings. And yet to invoke such a range of European artistic identities is to recognize that representation per se cannot be some simple, unitary endpoint of painting. And the problem only expands if we look further afield.

Roman wall painting in the cubiculum (bedroom) from the Villa of P. Fannius Synistor at Boscoreale, 1st century BCE

Giotto, An illusionistic window, Scrovegni Chapel, Padua, c. 1307

Artificial perspective (drawn by Julian Bell)

Consider two paintings of monkeys: one painted in 12th-century China, and one in 18th-century England (see pages 32–33). Everyone can agree that Mao Sung and George Stubbs had both looked long and hard at these animals and were confidently in command of their respective crafts. Would anyone care to say, however, whether one painting or the other should have priority as a representation of a macaque? If we choose the spare, suggestive concision of Mao’s pen and brush, we miss a certain sharp truth that Stubbs is able to dramatize: that the monkey knows that humans are looking at it and that picturing it, we have a kind of power over it. If we opt for the Englishman’s highly modelled, anatomical analysis, we don’t attain to the more contemplative truth that Mao is able to dwell on: that there is something it will always feel like for the monkey to be a monkey, that there is an inner quality and a depth to its life.

Each representation is necessarily partial. Accepting this and appreciating both, we may accept that every representation is grounded in a certain style. Style at once enables and sets limits to the imitation of nature. The style traditions of China and of Europe, each of them complex and supple, have enabled and set limits in differing ways.

Or do we wholly accept this? Don’t we tend to think that style is a human limitation typical of handmade products such as paintings, but that this limitation can be progressively overcome the more we entrust the task of picturing to neutral machines? At each incoming stage of image technology – the photograph, the holograph, the computer simulation – there is the expectation that the representation will be more full of the subject itself and less tinged by the observer’s presuppositions: ultimately, style-free. Then looking back, we realize that, say, an 1880s albumen print of one of Ottomar Anschütz’s photos of monkeys remains interesting largely because it reveals a period style of its own – as, in due course, will the CGI monkeys of the 2010s.

The more we concern ourselves with style, the less we expect representation to be perfected. During the century after the arrival of perspective, the role of the artistic connoisseur emerged – the sophisticated eye capable of distinguishing one artistic style from another, a Raphael from a Titian. Art studies in Europe grew up on the basis of such discriminations (much as they earlier had in China). Very gradually, European curiosity about style extended in scope to encompass more distant eras – medieval art for instance – and more distant cultures, such as China and Japan.

If style was seen as inescapable, and representation could no longer be seen as a perfectible task of imitation, then it needed rethinking. A shift in analogies opened up a way. The ancient philosophers’ term ‘mimesis’ had implied that painters represent by miming – impersonating – whatever in nature they were painting. By the 17th century, the word ‘representation’ had been extended to cover another form of impersonation. Thomas Hobbes, writing in 1651, noted that people could now be said to represent one another ‘as well in tribunals, as in theatres’. Representation, therefore, began its career as a political concept at this stage. But do the lawyer representing the plaintiff, or the ambassador representing the king, or the member of Parliament representing the constituent imitate their clients? No: these are not relationships of resemblance, but of agreed authority.

How does this relate to pictures? Later, in his Leviathan, Hobbes discusses the worship of images, commenting on the frequent unlikeness of the designated object to the imagined deity – the many different faces given the Virgin Mary, for instance, or the fact that ‘a stone unhewn hath been set up for Neptune’. Yet, he points out, these various crude idols ‘serve well enough for the purpose they were erected for; which was no more but by the names only, to represent the persons mentioned… And thus an image in the largest sense is either the resemblance, or the representation of something visible; or both together, as it happeneth for the most part’ (my emphasis).

Attributed to Mao Sung, Monkey, early 12th century

George Stubbs, Portrait of a Monkey, 1774

Ottomar Anschütz, Monkey, 1886

In other words, on the model of a political relationship between people, representation in images can be seen as being a matter of custom or convention, just as representation in words is. The lawyer is accepted as representing his client because there is an agreed relationship of authority within a certain society. The unhewn stone can be accepted as representing Neptune because there is an agreed relationship of naming within a society, says Hobbes. It is thus an arbitrary relation – arbitrary in the 17th-century sense of submitting to arbiters and settling matters under a common rule.

One monkey, then, gets represented according to a certain agreed relationship between images and things in China, and another under an alternative arbitration in Europe. This type of approach had become inbuilt into the discipline of art history by the 20th century. It sidelines the question of how, sight for sight, the depiction may resemble the object depicted. It stresses instead the principle of consent, through which the one is agreed to stand for the other as its substitute. Representation, from the point of view of the art historian, was no longer seen as imitation: it was substitution.

But one obvious conclusion of this approach by arbitration is that – as Hobbes states – anything can stand for anything, if there is an agreement to that effect. It is a conclusion humourists have repeatedly reached, from the 18th-century cartoon A Connoisseur Inspecting a Very Dark Night-piece – shown as a pure black canvas – to Alphonse Allais’s First Communion of Anaemic Young Girls in Snowy Weather, a joke exhibited in the Paris Salon des Incohérents in 1883 and retold in a thousand variations over the succeeding century.

On a more serious level, you could say that this is also the logic of the Holy Mass: bread and wine are Christ’s flesh and blood when, and only when, both the priest pronounces them to be so and the congregation have faith in his authority. In this light, representation becomes essentially an act, a decisive movement made at a certain time by the priest – or the painter. The painter, making a mark with the brush, believes in its meaningfulness: that is enough, for him and for those who believe it too; this shared belief is the moment of representation.

Alphonse Allais, First Communion of Anaemic Young Girls in Snowy Weather, 1883. Reconstruction

This is effectively what has happened, during the past century, to the idea that paintings express the painter’s experience. The prevailing tendency, in both private appreciations and public appraisals of the art, has been to regard ‘painting’ as a verb and nothing but: a creative action. Painters have been awarded recognition for their vigour in creativity. Their products might be adjuncts to that award, but otherwise we have not been too certain how to place these objects in relation to the world in which they exist. Outside the charmed circle of the painter’s admirers, the mark left by the creative action becomes no more defined in meaning than a stone. It is not that ‘It is, and it is something else’: it simply is.

This statement corresponds to a common experience among viewers of paintings: that of disengagement. Walking through any big gallery or art fair, it can come upon you: you get tired, you lose the inclination to give the exhibits any credit. The paintings before you may as well be so many stones. On occasion, however, you may get past this reaction and get a second wind. It may come to you that stones are things of wonder in themselves. They do not need to mean something definite: their meaning is potentially infinite. That is how nature is. And this paint too, you may say, this stuff on the canvas, is part of nature – infinite, mesmerizing. The Chinese do not set ‘scholars’ rocks’ on plinths as deities, but simply as singularities, objects whose oddness commands fascination. That is how paintings are: self-contained specimens of nature, to which we should attend.

This is the experience that philosophers, since the 18th century, have singled out as ‘aesthesis’ or ‘aesthetic contemplation’: looking on a thing, whether made by art or nature, for its own sake – a mode of thought to be explored further in chapter 5. Aesthesis is a way of seeing singly, not doubly: it is a mode of experience in which we are no longer concerned with the object’s status as a mark. Philosophers have often upheld aesthesis as an ultimate value. If, however, it means that paintings are no longer marks – that they are not what we settled on in our first definition of painting – then our conclusions have tripped over our premises.

Where, then, has this introductory survey of our feelings about likenesses and paintings arrived? Not yet at any satisfactory formula in which to encapsulate the practice of painting. Rather, it has opened up a space in which to think further about that practice. It is a space defined by a number of polarities: know / desire, represent / express, divine / human, natural / artificial, to name just some of the most obvious oppositions that have cropped up in the course of discussion. These polarized terms can be aligned and realigned and made to cross each other in an indefinite number of combinations, like a cat’s cradle. We can look at this crossing and recrossing of inherited terms as a matrix, out of which the practice of painting was to expand and proliferate in myriad directions over the course of the past two centuries.

Chinese scholar’s stone