PICTURING HISTORY Series Editors
Peter Burke, Sander L. Gilman, Ludmiilajordanova, i-Roy Porter (1995-2002), fBob Scribner (1995-8)
In the same series
Health and Illness Images of Difference SANDER L, GILMAN
Reading Iconotexts
From Swift to the French Revolution
PETER WAGNER
Men in Black
JOHN HARVEY
Dismembering the Maie
Men's Bodies, Britain and the Great War
JOANNA BOURKE
Eyes of Love
The Gaze in English and French Painting and Novels 1840-1900
STEPHEN KERN
The Destruction of Art
Iconoclasm and Vandalism since the French
Revolution
DARIO GAMBONI
The Feminine Ideal
MARIANNE THESANDER
Maps and Politics
JEREMY BLACK
Trading Territories
Mapping the Early Modern World
JERRY BROTTON
Picturing Empire
Photography and the Visualization of the British Empire JAMES RYAN
Pictures and Visuality in Early Modern China
CRAIG CLUNAS
Mirror in Parchment
The Luttrell Psalter and the Making of
Medieval England
MICHAEL CAMILLE
Landscape and Englishness
DAVID MATLESS
The Thief, the Cross and the Wheel Pain and the Spectacle of Punishment in Medieval and Renaissance Europe MITCHELL B. MERBACK
'Down with the Crown'
British Anli-monarchism and Debates about
Royalty since 1790
ANTONY TAYLOR
The Jewish Self-image American and British Perspectives 1881-1939
MICHAELBERKOWITZ
Global Interests
Renaissance Art between East and West LISA JARDINE AND JERRY BROTTON
Picturing Tropical Nature
NANCY LEYS STEPAN
Representing the Republic Mapping the United States 1600-1900 JOHN RENNIE SHORT
Bodies Politic
Disease, Death and Doctors in Britain,
1650-1900
ROY PORTER
Eyewitnessing
The Uses of Images as Historical Evidence
Peter Burke
Hr*
-
^^^^^^^^
REAKTION BOOKS
2 Iconography and Iconology
ill
[An] Australian bushman would be unable to recognise the subject of a Last Supper; to him, it would only convey the idea of an excited dinner party.
ERWIN PANOFSKY
Before attempting to read images 'between the lines', and to use them as historical evidence, it is only prudent to begin with their meanings. But can the meanings of images be translated into words? The reader will have noticed that the previous chapter described images as 'telling' us something. In a sense they do: images are designed to communicate. In another sense they tell us nothing. Images are irredeemably mute. As Michel Foucault put it, 'what we see never resides in what we say'.
Like other forms of evidence, images were not created, for the most part at any rate, with the future historian in mind. Their makers had their own concerns, their own messages. The interpretation of these messages is known as 'iconography' or 'iconology', terms sometimes used as synonyms, but sometimes distinguished, as we shall see.
The Idea of Iconography
The terms 'iconography' and 'iconology' were launched in the art-historical world in the 1920s and 1930s. To be more exact, they were relaunched - a famous Renaissance handbook of images, published by Cesare Ripa in 1593, already bore the title Iconologia, while the term 'iconography' was in use in the early nineteenth century. By the 1930s the use of these terms had become associated with a reaction against a predominantly formal analysis of paintings in terms of composition or colour at the expense of the subject matter. The practice of iconography also implies a critique of the assumption of photographic realism in our 'snapshot culture'. The 'iconographers', as it is convenient
to call these art historians, emphasize the intellectual content of works of art, their implicit philosophy or theology. Some of their most famous and controversial claims concern paintings made in the Netherlands between the fifteenth and eighteenth centuries. It has been argued, for instance, that the celebrated realism of Jan van Eyck, for example, or Pieter de Hooch (illus. 38) is only superficial, hiding a religious or moral message presented through the 'disguised symbolism' of everyday objects.1
One might say that for the iconographers, paintings are not simply to be looked at: they are to be 'read'. Today, the idea has become commonplace. A well-known introduction to film studies is entitled How to Read a Film (1977), while the critic Roland Barthes (1915-1980) once declared, 'I read texts, images, cities, faces, gestures, scenes, etc' The idea of reading images actually goes back a long way. Within the Christian tradition it was expressed by the fathers of the church and most famously by Pope Gregory the Great (Chapter 3). The French artist Nicolas Poussin (1594-1665) wrote about his painting of Israelites gathering manna, 'read the story and the painting' (lisez I'histoire et k tableau). In similar fashion, the French art historian Emile Mále (1862-1954) wrote of 'reading' cathedrals.
The Warburg School
The most famous group of iconographers was to be found in Hamburg in the years before Hitler came to power. It included Aby Warburg (1866-1929), Fritz Saxl (1890-1948), Erwin Panofsky (1892-1968) and Edgar Wind (1900-1971), all scholars with a good classical education and wide interests in literature, history and philosophy. The philosopher Ernst Cassirer (1874-1975) was another member of this Hamburg circle, and shared their interest in symbolic forms. After 1933 Panofsky emigrated to the United States, while Saxl, Wind and even Warburg's Institute, as we have seen, all took refuge in England, thus spreading knowledge of the iconographical approach more widely.
The Hamburg group's approach to images was summed up in a famous essay by Panofsky, first published in 1939, distinguishing three levels of interpretation corresponding to three levels of meaning in the work itself.2 The first of these levels was the pre-iconographical description, concerned with 'natural meaning' and consisting of identifying objects (such as trees, buildings, animals and people) and events (meals, battles, processions and so on). The second level was the iconographical analysis in the strict sense, concerned
34
35
with 'conventional meaning' (recognizing a supper as the Last Supper or a battle as the Battle of Waterloo).
The third and ultimate level was that of the iconological interpretation, distinguished from iconography because it was concerned with 'intrinsic meaning', in other words, 'those underlying principles which reveal the basic attitude of a nation, a period, a class, a religious or philosophical persuasion'. It is at this level that images offer useful - indeed, indispensable - evidence for cultural historians. Panofsky was particularly concerned with the iconological level in his essay Gothic Architecture and Scholasticism (1951), in which he explored homologies between the philosophical and architectural systems of the twelfth and thirteenth centuries.
These pictorial levels of Panofsky's correspond to the three literary levels distinguished by the classical scholar Friedrich Ast (1778-1841), a pioneer in the art of interpreting texts ('hermeneutics'): the literal or grammatical level, the historical level (concerned with meaning) and the cultural level, concerned with grasping the 'spirit' (Geist) of antiquity or other periods. In other words, Panofsky and his colleagues were applying or adapting to images a distinctively German tradition of interpreting texts.
Readers should be warned that later art historians who have taken over the term 'iconology' have sometimes employed it in different ways from Panofsky. For Ernst Gombrich, for instance, the term refers to the reconstruction of a pictorial programme, a significant narrowing of the project linked to Gombrich's suspicion that Panofsky's iconology was simply another name for the attempt to read images as expressions of the Zeitgeist. For the Dutch scholar Eddy de Jongh, iconology is 'an attempt to explain representations in their historical context, in relation to other cultural phenomena'.3
For his part, Panofsky insisted that images are part of a whole culture and cannot be understood without a knowledge of that culture, so that, to quote his own vivid example, an Australian bush-man 'would be unable to recognize the subject of a Last Supper; to him, it would only convey the idea of an excited dinner party'. Most readers are likely to find themselves in a similar situation when confronted with Hindu or Buddhist religious imagery (Chapter 3). To interpret the message it is necessary to be familiar with the cultural codes.
In similar fashion, without a reasonable knowledge of classical culture, we are unable to read many western paintings, to recognize references to incidents from Greek mythology, say, or Roman history. If, for example, we do not know that the young man in sandals and a
36
io Detail showing .Mercury and die Graces, from Botticelli's Primavera, c. 1482, tempera on wood. Galleria degii Uffizi, Florence.
11 Titian, The Rape of Lucretia, 1571, oil on canvas. Fitzwilliam Museum, Cambridge.
peaked cap in Botticelli's Primavera (illus. 10) represents the god Hermes (or Mercury), or that the three dancing girls are the Three Graces, we are unlikely to be able to work out the meaning of the painting (even with this knowledge, all sorts of problems remain). Again, if we do not realize that the protagonists in the rape scene illustrated by Titian (illus. 11) are King Tarquin and the Roman matron Lucretia, we will miss the point of the story, told by the Roman historian Livy in order to show the virtue of Lucretia (who
37
wiped out her shame by killing herself), and to explain why the Romans drove out the king and founded a republic.
The Method Exemplified
Some of the most important achievements of the Warburg school concern the interpretation of paintings of the Italian Renaissance. Take the case of Titian's so-called Sacred and Profane Love (illus. 12). At the level of pre-iconographic description, we see two women (one naked, the other clothed), an infant and a tomb, which is used as a fountain, all situated in a landscape. Turning to the iconography, for anyone familiar with Renaissance art it is, one might say, child's play to identify the infant as Cupid, whereas decoding the rest of the painting is not so easy. A passage in Plato's dialogue the Symposium provides an essential clue to the identity of the two women: the speech of Pausanias about the two Aphrodites, the 'heavenly' and the 'vulgar', interpreted by the humanist Marsilio Ficino as symbols of mind and matter, intellectual love and physical desire.
At the deeper, iconological level, the painting makes an excellent illustration of the enthusiasm for Plato and his followers in the so-called 'Neoplatonic' movement of the Italian Renaissance. In the process, it offers important evidence for the importance of that movement in Titian's milieu in Northern Italy in the early sixteenth century. The painting's reception also has something to tell us about the history of attitudes to the naked body, notably the shift from celebration to suspicion. In early-sixteenth-century Italy (as in Greece in Plato's day), it was natural to link heavenly love with the naked woman, because nudity was viewed in a positive light. In the nineteenth century, changes in assumptions about nudity, especially
12 Titian, Sacred and Profane Love, 1514, oil on canvas. GalleriaBorghese, Rome.
female nudity, made it seem obvious to viewers - simple common sense, we might say - that the clothed Venus represented sacred love, while the nude was now associated with the profane. The frequency of images of the naked body in Renaissance Italy, compared with their rarity in the Middle Ages, offers another important clue to changes in the way in which bodies were perceived in those centuries.
Standing back from the interpretations and focusing on the method that they exemplify, three points stand out. The first is that in the attempt to reconstruct what is often called the iconographical 'programme', scholars have often joined together images which events had put asunder, paintings which were originally designed to be read together but are now dispersed in museums and galleries in different parts of the world.
The second point is the need for iconographers to have an eye for detail, not only to identify artists, as Morelli argued (Chapter 1) but to identify cultural meanings as well. Morelli was aware of this too and, in a dialogue which he wrote to explain his method, he created the character of a wise old Florentine who tells the hero that people's faces in portraits reveal something of the history of their time, 'if one knows how to read it'. Again, in the case of Sacred and Profane Love, Panofsky drew attention to the rabbits in the background and explained them as symbols of fertility, while Wind concentrated on the reliefs decorating the fountain, including a man being whipped and an unbridled horse, interpreting them as references to 'pagan initiatory rites of love'.4
The third point is that iconographers generally proceed by juxtaposing texts and other images to the image they wish to interpret. Some of the texts are to be found on images themselves, in the form of labels or inscriptions turning the image into what the art historian Peter Wagner calls an 'iconotext' which may be 'read' by the viewer literally as well as metaphorically. Other texts are selected by the historian in an attempt to clarify the meaning of the image. Warburg, for example, in his approach to the Primavera, noted that the Roman philosopher Seneca had associated Mercury with the Graces, that the Renaissance humanist Leonbattista Alberti had recommended painters to represent the Graces holding hands, and that a number of medals showing the Graces were in circulation in Florence in Botticelli's time.'
How can we be sure that these juxtapositions are the appropriate ones? Could Renaissance artists have known about classical mythology, for instance? Neither Botticelli nor Titian had much formal schooling and they are unlikely to have read Plato. To meet this objec-
38
39
tion, Warburg and Panofsky formulated their hypothesis of the humanist adviser, who drew up the iconographical programme of complex images for artists to execute. Documentary evidence of such programmes is relatively rare. On the other hand, the painters of the Italian Renaissance often had opportunities to talk to humanists, to Marsilio Ficino in Botticelli's case and to Pietro Bembo in that of Titian. Hence it is not implausible to suggest that a variety of allusions to ancient Greek and Roman culture can be found in their work.
The Method Criticized
The iconographical method has often been criticized as too intuitive, too speculative to be trusted. Iconographical programmes are occasionally recorded in surviving documents, but generally speaking they have to be inferred from the images themselves, in which case the sense of the different pieces of a puzzle fitting together, however vivid, is rather subjective. As the unending saga of new interpretations of the Primavera illustrates, it is easier to identify the elements of the painting than to work out the logic of their combination. Iconology is still more speculative, and there is a risk of iconologists discovering in images exactly what they already knew to be there, the Zeitgeist.
The iconographical approach may also be faulted for its lack of a social dimension, its indifference to social context. The aim of Panofsky, who was notoriously indifferent if not hostile to the social history of art, was to discover 'the' meaning of the image, without asking the question, meaning for whom? Yet the artist, the patron who commissioned the work, and other contemporary viewers may not have viewed a given image in the same way. It cannot be assumed that they were all as interested in ideas as the humanists and the iconographers. King Philip II of Spain, for example, commissioned scenes from classical mythology from Titian (c. 1485-1576). It has been plausibly argued that Philip was less interested in Neoplatonic allegories or in representations of specific myths than in pictures of beautiful women. In his letters to the king, Titian himself described his paintings as his 'poems', without any reference to philosophical ideas.6
Indeed, it would be unwise to assume that the classical allusions which Panofsky, a humanist himself, so much enjoyed recognizing, were appreciated by the majority of viewers in the fifteenth and sixteenth centuries. Texts sometimes offer us precious evidence of misunderstandings, of one god or goddess being taken for another by contemporary viewers, for instance, or a winged Victory viewed as an
angel by a spectator who knew more about Christianity than about the classical tradition. As missionaries were at times uncomfortably aware, peoples who had been converted to Christianity retained a propensity to view Christian images in terms of their own traditions, to see the Virgin Mary as the Buddhist goddess Kuan Yin, for example, or as the Mexican mother goddess Tonantzin, or to see St George as a version of Ogum, the West African god of war.
Another problem of the iconographical method is that its practitioners have often paid insufficient attention to the variety of images. Panofsky and Wind had sharp eyes for painted allegories, but images are not always allegorical. As we shall see, the question whether the famous seventeenth-century Dutch scenes from everyday life carry a hidden meaning remains controversial (Chapter 5). Whistler issued a challenge to the iconographical approach by calling his portrait of a Liverpool shipowner 'Arrangement in Black', as if his aim was not representational but purely aesthetic. Again, the iconographical method might need to be adapted to deal with surrealist paintings, since painters such as Salvador Dali (1904-1989) rejected the very idea of a coherent programme and attempted instead to express the associations of the unconscious mind. Artists such as Whistler, Dali and Monet (below), may be described as resisting iconographical interpretation.
This point about resistance leads on to a final criticism of the method, which is that it is too literary or logocentric, in the sense of assuming that images illustrate ideas and of privileging content over form, the humanist adviser over the actual painter or sculptor. These assumptions are problematic. In the first place, the form is surely part of the message. In the second place, images often arouse emotions as well as communicating messages in the strict sense of the term.
As for iconology, the dangers of assuming that images express the 'spirit of the age' have been pointed out many times, notably by Ernst Gombrich in his criticisms of Arnold Hauser and Johan Huizinga as well as of Erwin Panofsky. It is unwise to assume the cultural homogeneity of an age. In the case of Huizinga, who inferred the existence of a morbid or macabre sensibility in late medieval Flanders from the literature and the paintings of the period, the work of Hans Memling (c. 1435-1494) has been cited as a counter-example, a painter who was 'widely admired' in the fifteenth century yet lacks the 'morbid preoccupation' of his colleagues.7
In short, the specific method of interpreting images that was developed in the early twentieth century can be faulted as too precise and too narrow in some ways, too vague in others. To discuss it in general terms risks underestimating the variety of images, let alone
40
4i
the variety of historical questions which images may help answer. Historians of technology (say), and historians of mentalities come to images with different needs and expectations. Hence the chapters that follow will focus in turn on different domains such as religion, power, social structures and events. If there is a general conclusion to 'If^; be drawn from this chapter, it might be that historians need iconography but also need to go beyond it. They need to practice iconology in a more systematic way, which may involve making use of psycho- ":Mjj analysis, structuralism and especially reception theory, approaches which will be drawn upon from time to time as well as being discussed > jf !|
more fully and more explicitly in the final chapter of this book. :|§;*:
The Problem of Landscape :;;|y
Panofsky's second and third levels may well appear to have little relevance to landscape, but for this very reason landscapes allow us to see with particular clarity both the strengths and the weaknesses of the iconographical and iconological approaches. I am using the term 'landscape' with deliberate ambiguity, to refer not only to paintings and drawings but also to the land itself as it has been transformed by 'landscape gardening' and other forms of human intervention.
One of the strengths of the iconographical approach is that it has inspired geographers and art historians alike to read the physical landscape in new ways. The iconography of the land itself is particularly obvious in the case of gardens and parks. There are also the typical or symbolic landscapes that represent nations by means of their characteristic vegetation, from oaks to pines and from palm trees to eucalyptus. One might measure the importance of this symbolism by the indignation aroused when the British Forestry Commission, for example, planted pines where traditional English deciduous trees had grown.8 jif;
If the physical landscape is an image that can be read, then the painted landscape is the image of an image. In the case of painted landscape, the weaknesses of the iconographical approach may well appear obvious. It seems to be no more than common sense to suggest that landscape painters want to give viewers aesthetic pleasure rather .> than to communicate a message. Some landscape painters, Claude ./: Monet (1840-1926), for example, rejected meaning and concentrated on visual sensations. When he painted a view of Le Havre in 1872, he called it simply Impression: Sunrise. All the same, what appears in a given culture to be 'common sense' needs to be analysed by historians and anthropologists alike as part of a cultural system. In the case of
ll
landscape, trees and fields, rocks and rivers all carry conscious or unconscious associations for viewers.9 Viewers, it should be emphasized, from particular places in particular periods. In some cultures wild nature is disliked or even feared, while in others it is the object of veneration. Paintings reveal that a variety of values, including innocence, liberty and the transcendental have all been projected onto the land.
For example, the term 'pastoral landscape' has been coined to describe paintings by Giorgione (c. 1478-1510), Claude Lorrain (1600-1682) and others, because they express an idealizing vision of rural life, particularly the life of shepherds and shepherdesses, in much the same way as the western tradition of pastoral poetry from Theocritus and Virgil onwards. These painted landscapes appear to have influenced the perception of actual landscapes. In late-eighteenth-century Britain, 'tourists', as the poet Wordsworth was one of the first to call them, guidebooks in hand, viewed the Lake District, for instance, as if it were a series of paintings by Claude Lorrain, describing it as 'picturesque'. The idea of the picturesque illustrates a general point about the influence of images on our perception of the world. Since 1000, tourists in Provence have come to see the local landscape as if it were made by Cezanne. Religious experience too, as we shall see (Chapter 3), is partly shaped by images.
Given these pastoral associations, it is likely that Monet's The Train (1872), with its landscape of smoking factory chimneys, must have shocked some of its early viewers, while even the tiny trains to be seen in the distance of some nineteenth-century American landscapes may have raised eyebrows. A more difficult question to answer is whether the artists introduced the trains because they were admirers of progress, like the Mexican mural painter Diego Rivera (1886-1957), whose frescos of 1926 celebrated the tractor and the mechanization of agriculture.10
The last point implies that landscape evokes political associations, or even that it expresses an ideology, such as nationalism. Prince Eugen of Sweden was one of a number of artists in the years around 1900 who chose to paint what he called 'Nordic nature, with its clean air, its hard contours and its strong colours'. We might say that Nature was nationalized at this time, turned into a symbol of the mother or fatherland.11 In twentieth-century Britain, the land has been associated with Englishness, with citizenship, and with the 'organic society' of the village, threatened by modernity, industry and the city.12
Again, it has been perceptively observed that eighteenth-century
43
13 Colin McCahon, Takaka - Night and Day> 1948, oil on canvas laid on board. Auckland Art Gallery Toi o Tamaki, New Zealand.
English landscape painters disregarded agricultural innovations and ignored recently-enclosed fields, preferring to show the land as it was supposed to have been in the good old days.13 In similar fashion the landscapes of John Constable (1776-1837), painted during the Industrial Revolution, have been interpreted as an expression of anti-industrial attitudes because they leave out factories. Factories were not of course part of the landscape of Constable's Essex or Wiltshire, but the coincidence in time between the rise of landscape painting and the rise of factories in England remains both intriguing and disturbing.
The same period saw a new enthusiasm for wild nature, marked by the increasing popularity of tours in search of mountains and woods and the publication of a shelf of books on the subject such as the Observations Relative to Picturesque Beauty (1786) by the writer William Gilpin (1724-1804). It seems that the destruction of nature, or at least the threat of its destruction, was a necessary condition for its aesthetic appreciation. The English countryside was already taking on the aspect of a paradise lost.14
More generally, in the West at least, nature has often symbolized political regimes. The conservative thinker Edmund Burke (1729-1797) described the British aristocracy as 'great oaks', and contrasted the British constitution, which grew naturally like a tree, with the artificial, 'geometrical' constitution of revolutionary France. For liberals, on the other hand, nature represented freedom, defined against the order and constraint associated with absolute monarchy and represented by the symmetrical gardens of Versailles and its many imitations. Forests and the outlaws who live in them, notably Robin Hood, are an ancient symbol of liberty.'5
The landscapes of empire evoke another theme, the theme of dispossession. The absence of figures in an American landscape, for instance, has been said to carry 'a more loaded meaning than in Europe'. In the case of New Zealand, it has been suggested that 'the evocation of an empty landscape ... cannot be seen as a purely pictorial or aesthetic statement' (illus. 13). Consciously or unconsciously, the artist has erased the aborigines, as if illustrating the idea of 'virgin' soil or the legal doctrine that New Zealand, like Australia and North America, was a 'no-man's-land'. In this way the position of the white settlers has been legitimated. What the painting documents is what might be called the 'colonial gaze' (Chapter 7).16
Even in the case of landscape, then, the iconographic and icono-logical approaches do have a role to play, helping historians to reconstruct past sensibilities. Their function is more obvious in the case of religious images, to be discussed in the following chapter.
44
45
power might be to encourage students of history to take control and to make their own films as a way of understanding the past. In the 1970s, for example, some students at Portsmouth Polytechnic were encouraged by their history teacher, Bob Scribner, to make films about the German Reformation. Critical reviews of films in historical journals, a practice which is gradually becoming more common, are a step in the same direction. A collaboration on equal terms between a historian and a director, along the lines of the collaboration of anthropologists and directors in some ethnographic films, might be another way to use the cinema to stimulate thought about the past.
Despite Panofsky's interest in the cinema, exemplified by an article on 'Style and Medium in the Moving Pictures' (1937) the problems of interpreting film seem to have taken us a long way from the iconographical method associated with him, the method which was discussed in Chapter 2. The extent to which it is necessary for historians using images as evidence to go beyond iconography - and in what direction - will be the theme of the final chapters of this book.
10 Beyond Iconography?
I read texts, images, cities, faces, gestures, scenes etc.
ROLAND BARTHES
After examining different kinds of image in turn - images of the sacred, images of power, images of society, images of events, and so on - it is time to return to the problems of method originally raised in the chapter on iconography. Erwin Panofsky published a famous essay on the iconography of 'Hercules at the Cross-Roads', confronted with the decision which would determine his later career. A recent symposium adapted his title to a discussion of 'Iconography at the Cross-roads', the problem whether or not historians of images should continue to follow Panofsky's path,1
Some criticisms of the Panofsky method have already been mentioned (Chapter 2). The question to be discussed here and in Chapter 11 is whether there is any alternative to iconography and iconology. There are three obvious possibilities; the approach from psychoanalysis, the approach from structuralism or semiotics and the approach (more exactly, approaches in the plural) from the social history of art. All these approaches have made their appearance more than once in earlier chapters and all of them have parallels in the history of literary criticism. I call them 'approaches' rather than 'methods' on the grounds that they represent not so much new procedures of investigation as new interests and new perspectives.
Psychoanalysis
The psychoanalytical approach to images focuses not on conscious meanings, privileged by Panofsky, but on unconscious symbols and unconscious associations of the kind which Freud identified in his Interpretation of Dreams (1899). This approach is indeed a tempting one. It is difficult to deny that the unconscious plays a part in the
168
169
creation of images or texts. Freud did not often provide interpretations of specific images - apart from his celebrated and controversial essay on Leonardo da Vinci - but his concern with small details, especially in the Psychopathology of Everyday Life, resembles that of Giovanni Morelli (Chapter i), as Carlo Ginzburg has noted.2 Some of Freud's remarks on dreams offer clues to the interpretation of paintings. For example, the concepts of 'displacement' and 'condensation', which Freud developed in the course of analysing the 'dream work', are also relevant to visual narratives.3 The idea of the phallic symbol has obvious relevance to some images. It has been argued by Eddy de Jongh, for example, that the birds, parsnips and carrots which make such a frequent appearance in Flemish and Dutch genre paintings of the sixteenth and seventeenth centuries should all be interpreted in this way.4
Faced with the examples discussed in Chapter 7 in particular, a psychoanalyst might well suggest that some stereotyped images, such as the harem, are visualizations of sexual fantasies, while others -images of cannibals, for example, or of witches - are projections onto the 'other' of the self's repressed desires. It is hardly necessary to be a committed Freudian to approach images in this manner. As we have seen (Chapter 2), attitudes and values are sometimes projected onto landscapes (either the land itself or its painted image), just as they are projected onto the blots of the famous Rorschach test. The discussion of sacred imagery also raised the questions of unconscious fantasies and unconscious persuasion. Again, the discussion of advertising in the chapter on material culture noted the 'subliminal' approach, in other words the attempt to create associations between products and the viewer's more or less unconscious dreams of sex and power.
All the same, even if we leave on one side the controversies over the scientific status of psychoanalysis and the conflicts between different schools of analysis, from Carl Gustav Jung to Jacques Lacan, serious obstacles remain in the way of historians who wish to follow this approach to images. On what criteria does one decide whether an object is a phallic symbol? Can the phallus not be used in its turn as a symbol of something else? The nineteenth-century Swiss philologist Johann Jakob Bachofen regarded it as an image of the sacred, at least in classical art.
There are two obstacles in particular to this kind of historical psychoanalysis, problems which are not confined to images but exemplify the general difficulties of practising what has become known as 'psychohistory'. In the first place, psychoanalysts work with living individuals, while historians cannot place dead artists on the couch
170
and listen to their free associations. We may, like the Spanish director Luis Bufluel, view Bernini's St Teresa (Chapter 3) as an interpretation of religious ecstasy in sexual terms, but all the evidence we have is contained within the marble itself. The sources which de Jongh used in his famous article on sexual symbolism in the art of the Netherlands came principally from proverbs and poems, in other words from consciously-expressed attitudes. However different his conclusions may have been, he did not diverge from Panofsky in his methods.
In the second place, historians are primarily concerned with cultures and societies, with collective desires rather than individual ones, while from Freud onwards, psychoanalysts and other psychologists have been less successful, or at any rate more speculative, in this domain. Freud, for example, devoted his essay on Leonardo to the relation between the artist's 'mother fixation' and his paintings of smiling women, without taking into account the nature of fifteenth-century culture. For example, he based conclusions about Leonardo's personality on his representation of St Anne, the mother of the Virgin Mary, as more or less the same age as her daughter, without realising that this was a cultural convention of the period. Hollywood was described as a 'dream factory' by an anthropologist, Hortense Powdermaker, in 1950, but the processes of production and reception of these fantasies still await analysis. Relatively little has been written on the history of images as expressions of collective desires or fears, although, as we have seen (Chapter 3) it might be illuminating to examine changing images of heaven and hell from this perspective.5
The conclusion seems to be that so far as historians using images are concerned, the psychoanalytic approach is both necessary and impossible. It is necessary because people do project their unconscious fantasies onto images, but it is impossible to justify this approach to the past according to normal scholarly criteria because the crucial evidence has been lost. Interpreting images from this point of view is inevitably speculative. There is of course an irreducibly speculative element in all attempts at iconological analysis - and in much of iconographical analysis as well — but the element of speculation is even greater when the unconscious meanings of images are under discussion. The best thing to do is probably to go ahead and speculate, but to try to remember that this is all that we are doing.
Structuralist and Post-structuralist Approaches
The approach with the best claim to be regarded as a 'method' in a reasonably strict sense of the term, is structuralism, otherwise known
171
as 'semiology' or 'semiotics'. These last terms were coined to describe the general 'science of signs' of which some linguists dreamed at the beginning of the twentieth century. The structuralist movement became more widely known in the 1950s and 1960s, thanks in particular to the anthropologist Claude Levi-Strauss and the critic Roland Barthes, both of whom were extremely interested in images. Levi-Strauss, for instance, wrote about the art of Amerindian peoples such as the Tsimshian of Canada, especially on the phenomenon of 'doubling' in which one side of the picture of an animal, say, is a mirror image of the other.
As for Barthes, the essays collected in his Mythologies (1957) comment on a wide range of images, including films about ancient Rome, advertisements for soap powders, photographs of shocking events and the illustrations in contemporary magazines, including what he called the 'visual myth' of the black soldier saluting the tricolour on the cover of an issue of Paris-Match (25 June-2 July 1955)- 'I am at the barber's', Barthes tells us, 'and a copy of Paris-Match is offered to me' (presumably a self-respecting French intellectual of the period would not have allowed himself to be seen buying a copy of this popular paper). 'On the cover, a young Negro in a French uniform is saluting, with his eyes uplifted, probably fixed on a fold of the tricolour.' Barthes read the image - which he did not reproduce -as signifying 'that France is a great Empire, that all her sons, without any colour discrimination, faithfully serve under her flag'.6
From the point of view of this chapter, two of the structuralists' claims or theses are particularly important. In the first place, a text or an image may be regarded, to use their favourite phrase, as a 'system of signs', emphasizing what the American art historian Meyer Schapiro calls the 'non-mimetic elements'.7 Such a concern diverts attention from the relation of the work in question to the external reality it may appear to represent and also from its social context, as well as from the elements which iconographers claim to decode or interpret. On the positive side, regarding an image or a text in this way means focusing attention on the work's internal organization, more especially on the binary oppositions between its parts or the various ways in which its elements may echo or invert one another.
In the second place, that system of signs is viewed as a sub-system of a larger whole. That whole, described by linguists as 'langue' (language), is the repertoire from which individual speakers make their selection ('parole'). Thus the Russian folklorist Vladimir Propp (1895-1970) analysed Russian folktales as permutations and combinations of 31 basic elements such as 'The Hero acquires the use of a
magical agent'. Structurally, according to Propp, it is the same function (no. 14), whether the princess gives the hero a ring or the king gives him a horse.
What are the consequences of approaching images as 'figurative texts' or 'systems of signs'? Among other things, the structuralist approach encourages sensitivity to oppositions or inversions. Images of 'the other', for instance, may often be read as inversions of the observer or the painter's self-image. The binary oppositions between pairs of images, as in the case of Cranach's 'antitheses' between Christ and the pope (illus. 18), or within a single image, as in the case of Hogarth's Calais Gate as noted earlier (p. 134), or Pieter Brueghel's Carnival and Lent, take on a new importance when one is wearing structuralist spectacles.
It is particularly illuminating to analyse visual narratives in structuralist terms, whether they are tapestries, engravings or films. To return to Bertolucci's Novecento (Chapter 9; illus. 80), its depiction of two families, one of landowners and the other of agricultural workers, is a complex combination of similarities and oppositions. The protagonists, Alfredo and Olmo, were born the same day, grew up together and are deeply attached to each other, yet destined for conflict. Their relationship is in some ways a replay, but in others the exact opposite, of the relationship between their grandfathers, Alfredo senior and Leone.
A structuralist approach is also concerned with the associations between one sign and another, a car and a beautiful girl, for instance, created in the mind of the viewer by means of frequent juxtapositions of the two elements. As for the structuralist emphasis on system, advertisements have been analysed, as we have seen (Chapter 5), to show how each new example refers back to earlier ones and in turn
80 Poster for the Bernardo Bertolucci film 1900 (Novecento) {1976).
I72
173
adds something to the common treasury. A similar point might be made about other ensembles of images. For example, the paintings, sculptures, engravings, medals and other images produced in the seventeenth century in order to glorify Louis XIV formed a self-referential system. A medal was struck to commemorate the erection of a statue to the king, an image of the medal was published in a book of engravings, and so on.8
As a single concrete example, we might take Umber to Eco's structuralist analysis of the Camay advertisement already discussed in Chapter 5 (illus. 45). Eco describes the woman as beautiful ('according to current codes'), nordic ('a sign of status', since this is an Italian advertisement), rich and cultivated (since she goes to Sotheby's); 'if she is not English she must be a high-class tourist'. The man is virile and self-confident but 'does not have an English appearance'. He is an international traveller, rich, cultivated, and a man of taste. He finds her fascinating, and the legend suggests that the brand of soap advertised is the source of the fascination.9
Michel Foucault was also a kind of structuralist, though not along the lines laid down by Levi-Strauss. He was interested in systems of 'representation' just as he was interested in systems of thought. By a 'representation', Foucault meant a verbal or pictorial image of some object, made according to a certain set of conventions, which interested him more than the greater or lesser fidelity with which the object was described or depicted. His famous analysis of a painting by Velazquez, Las Meninas, followed these lines, describing it as 'the representation ... of classical representation', at a time when the traditional links between signs and the objects they signify had been broken. In the wake of Foucault's work in the sixties and seventies, the idea of representation was taken up by art historians, literary critics, philosophers, sociologists, anthropologists and historians. The success of the term doubtless contributed to the success of the interdisciplinary journal Representations (founded in 1983), and vice versa.10
Another aspect of the structuralist approach deserves to be noted here. The concern with the act of selection from a repertoire not only underlines the importance of visual formulae and themes (Chapter 8), but also focuses attention on what is not chosen, what is excluded -a theme which was particularly dear to Foucault. In the course of this study we have already had occasion to note the importance of such blind spots, the equivalent of silences in oral discourse; the absence of children from medieval imagery, for instance (Chapter 6), that of the indigenous inhabitants of New Zealand from McCahon's landscape (Chapter 2), and the lack of the traditional royal attributes of crown
174
and sceptre in the portrait of Louis Philippe (Chapter 1). These blind spots should be distinguished from the 'blanks' which the image-maker consciously leaves the viewer to fill in, like the absent tricolour which the viewer infers from the salute in the case of the cover of Paris-Match analysed by Barthes. Interpreters of images need to be sensitive to more than one variety of absence."
Problems remain, as some of the most distinguished practitioners of the structuralist approach themselves admit. Is the idea of the 'language' of images, or of paintings as 'texts' anything more than a vivid metaphor? Are there 'disanalogies' as well as analogies between art and language? Is there one language or 'code' for images, or are there a number of different ones, the equivalent of English (say), Arabic or Chinese? Is the code conscious or unconscious? If unconscious, is it so in the strict Freudian sense of what is repressed, or in the ordinary-language sense of what is taken for granted? To some critics the structural approach appears to be intolerably reductionist, with no place for ambiguities or for human agency. In one of the best-known and most forceful of these criticisms, the American anthropologist Clifford Geertz concluded that 'To be of effective use in the study of art, semiotics must move beyond the consideration of signs as means of communication, code to be deciphered, to a consideration of them as modes of thought, idioms to be interpreted."2
My own view on this controversial issue is that the practice of the structural analysis of images as if it were an alternative method to iconography is indeed open to the criticisms summarized above, but that the structuralists have made an important contribution to the common treasury of interpretation by their emphasis on formal parallels and oppositions. A point which brings us to the question of the novelty claimed for this approach. Structural analysis is surely more innovative - and more shocking - in the case of literary narrative than it is in the case of images. Literature, as the German critic Gottfried Ephraim Lessing explained in his Laokoon (1766), is an art of time, but the structuralists deliberately ignore this point and read narratives against the grain, as in the case of Levi-Strauss's analysis of the Oedipus myth, which reduces it to a single point repeated over and over again.
In the case of painting, on the other hand, an art of space, a concern with internal relations, with what artists and critics call 'composition', is traditional, a reading with the grain rather than against it. If the structure lies beneath the surface in literary works, which we read or hear word by word, it lies on the surface of images, at least if they are seen from a distance. A concern with internal rela-
175
tions was indeed the 'formal' or 'formalist' analysis in vogue around the year 1900, the approach against which Panofsky reacted by stressing meaning (he entitled a collection of his essays 'Meaning in the Visual Arts'). Like the formalists, the structuralists differ from Panofsky in the sense of showing less interest in the decoding of specific elements of the image than in the relation between them. They emphasize what the critic Hayden White has called 'the content of the form'.
In any case, insofar as they do analyse specific elements of images, Levi-Strauss, Barthes and Eco might all be described as doing iconography rather than breaking with it. Bernardette Bucher's structural analysis of a series of engravings of the New World was inspired by Levi-Strauss and Panofsky alike. For his part, Levi-Strauss once described Panofsky as 'a great structuralist'. Again, imagine what Panofsky might have said about the Camay advertisement. How different would his iconography and iconology have been from the semiology of Eco? Barthes's idea of reading culture, brilliantly exemplified in Mythologies in his celebrated essay on wrestling as a performance of suffering and justice, has its parallel within the hermeneutic tradition in Clifford Geertz's at least equally celebrated reading of the Balinese cockfight. Both interpreters treat sporting occasions as texts and compare them to drama, yet one of them is supposed to be employing a structuralist approach, the other a hermeneutic one.13
As we have seen, the structuralists have been criticized for a lack of interest in specific images (which they reduce to simple patterns), and also for lack of concern with change. In reaction against their approach there has developed a movement known as 'post-structuralist'. If iconographers stress the conscious production of meaning, and the structuralists, like the Freudians, emphasize unconscious meanings, the focus of the post-structuralists falls on indeterminacy, 'polysemy' or what Jacques Derrida has called the 'infinite play of significations'. They are concerned with the instability or multiplicity of meanings and the attempts by image-makers to control this multiplicity, by means, for example, of labels and other 'iconotexts' (discussed above, Chapter 2).14
Like despotism and anarchy, the structuralist and post-structuralist approaches might be said to have opposite strengths and weaknesses. The weakness of the structuralist approach is the propensity to assume that images have 'a' meaning, that there are no ambiguities, that the puzzle has a single solution, that there is one code to be broken. The weakness of the post-structuralist approach
is the inverse, the assumption that any meaning attributed to an image is as valid as any other.
Another question to ask about the emphasis on ambiguity in the post-structuralist approach is whether it is really new, or more precisely, to what extent and in what ways it differs from earlier movements. Some at least of the practitioners of the 'classic' icono-graphical approach have long been aware of the problem of polysemy or 'multivocalityV' So indeed was Roland Barthes, despite the fact that accepting polysemy undermines the structuralist decoding of images, or at the very least the grander claims made for this approach. Again, studies of propaganda have long paid some attention to the use of inscriptions - on Roman coins or Renaissance medals, for example - as a means of leading viewers to 'read' the image in the correct way.
What is new in our day is essentially the emphasis on indeterminacy and the claim that makers of images cannot fix or control their meaning, however hard they try to do so, whether by inscriptions or other means. This emphasis fits in well with the post-modernist movement in general and in particular with the analysis of the 'reception' of images, an approach that will be discussed in the following chapter.
I
177