Art and the Roman Viewer The Transformation of Art from the Pagan World to Christianity J AS ELSNER Courtauld Institute of Art Cambridge UNIVERSITY PRESS CONTKNTS List of Illustrations page ix . \ ckno n-ledge men ts x \ i i Introduction 1 PART I: WAYS OF VIEWING IN THE ROMAN WORLD 15 1 Viewing and "the Real": The Imagines of Philostratus and the Tabula of Cebes 2 1 2 Viewing and Society: Images, the View and the Roman I louse 4<; 3 View ing and the Sacred: Pagan, Christian and the Vision of God 88 4 Viewing and Identity: The Travels of Pausanias; or, A Greek Pilgrim in the Roman World 12s PART II: THE TRANSFORMATION OF ROMAN ART FROM AUGUSTUS TO JUSTINIAN 157 5 Reflections on a Roman Revolution: A Transformation in the Image and Conception of the Emperor 159 6 From the Literal to the Symbolic: A Transformation in the Nature of Roman Religion and Roman Religious Art 190 PART III: EPILOGUE: MODULATIONS OF CHANGE 7 "The Truth within These Empty Figures": The Genesis of Christian Visual Exegesis Bibliography Notes Index ILLUSTRATIONS color plates i. Fresco on lime plaster, from the Villa of P. Fannius Sinister, Boscoreale (now in the Metropolitan Museum), first century b.c. Apse mosaic of the Transfiguration, Monastery of St Catharine at Mt Sinai, sixth century a.o. 3. The Projecta Casket seen from the front, fourth century a.d., Esquiline Treasure, British Museum. 4. I [ercules fighting the I Iydra, right-hand arco-solium, cubiculum N, Via Latina Catacomb, Rome, fourth century a.d. page xxni xxiv xxv xxvi figures 1. Mosaic of fish and octopus, from Pompeii (now in the Naples Museum), first century b.c. 2. Fresco of fish and bread-basket, from the Catacomb of Calistus, Rome, third century a.d. 3. Villa of the Mysteries peristyle, Pompeii, first century b.c. to first century a.d. 4. Villa of the Mysteries peristyle, Pompeii, first centurv b.c. to first centurv a.d. 3 65 illustrations 5. Villa of the Mysteries (after Maiuri). Han with viewing positions around the peristyle: \, inside the central opening of the peristvle; B, inside the portico of the peristyle. 67 6. Villa of the Mysteries, Occus 6, Pompeii, first century b.c. 68 -. Villa of the Mvsteries, Oecus 6, Pompeii, first century B.C. 68 8. Villa of the Mvsteries, viewing positions and perspectives in Oecus 6 (after Corlaita Scag- liarini). 69 9. Atrium of the Samnitc I louse, I Iercula- ncum, first century b.c. 70 10. Atrium of the Sanmite House, Hercula- neum, first century b.c. 71 11. I a\ia's Villa at Prima Porta, garden view and fence (now in the National Museum in Rome), first century b.c. 72 12. Atrium of the Samnitc House, Ilercula- neum, first century b.c. 73 13. Apse and triumphal arch mosaics, Monastery of St Catherine at Mt Sinai, sixth century a.d. io! 14. The Monastery of St Catherine at Mt Sinai, from the east. 103 15. Mosaic of Muses hefore the hurning hush, Monasters of St Catherine at Mt Sinai, sixth century a.d. 106 16. Mosaic of Moses receiving the tablets of the Law , Monastery of St Catherine at Mt Sinai, sixth century a.d. 107 17. Apse mosaic of the Transfiguration, Monastery of St Catherine at Mt Sinai, sixth century a.d. 113 18. Detail of apse mosaic of the Transfiguration, St Peter waking from sleep. Monastery of St Catherine at Mt Sinai, sixth century a.d. i i 5 19. Diagram of the Sinai mosaic programme. 1 19 20. Statue of Augustus from Prima Porta (now in the Vatican Museums), first century b.c. 163 \ 21. Detail of Prima Porta Augustus, breastplate, first century b.c. 22. Reconstruction of the south wall and apse of the Luxor cult room, third century a.i>. (after Deckers). 23. J. Gardner Wilkinsons sketches of the Luxor cult room, nineteenth century a.d. 24. Diagram of the directions of processions, Luxor cult room, third eenturv a.d. 25. Mosaic panel of Justinian and his entourage, presbytery of the Church of San Vitale in Ravenna, north side, sixth century a.d. 26. Mosaic panel of Theodora and her entourage, presbytery of the Church of San Vitale in Ravenna, south side, sixth century a.d. 27. Christ as universal emperor, apse mosaic of the Church of San Vitale in Ravenna, sixth century a.d. 2N. Apse frame with the Justinian panel, the Church of San Vitale in Ravenna, sixth century a.d. 29. Apse frame with the Theodora panel, the Church of San Vitale in Ravenna, sixth eenturv a.d. 30. Arc of cornucopiae, apse mosaic, the Church of San Vitale in Ravenna, sixth century a.d. 31. The Ara Pacis from the west, first century b.c. }2. Aeneas pouring a sacrificial libation, relief from the outer face, west front, south side of the entrance, Ara Pacis, Rome, first century b.c. 33. "Large procession", sacrificial procession from the frieze on the outer face of the precinct walls, south side of the Ara Pacis. Rome, first century B.C. ^4. "Large procession", sacrificial procession from the frieze on the outer face of the precinct walls, south side of the Ara Pacis, Rome, first eenturv B.C. illustrations [6s ■73 '"4 176 .78 '79 [81 182 1H3 [85 '93 '95 i<;6 ■V" xi illustrations 35. "Small procession", sacrificial procession on the frieze from the walls of the inner altar, outer face, north side of the Ara Pads. Rome, first century b.c. 200 36. Diagram of the Ara Pads with the directions of the processions. 201 37. "Italia", relief from the outer face, east front, south side of the entrance, Ara Pads, Rome, first century b.c. 202 38. The Ara Pads from the east, Rome, first century b.c. 203 39. Bucrania, paterae and garlands, frieze from the inner face of the precinct walls, south side of the Ara Pacis, Rome, first century b.c. 204 40. The altar and the inner face of the precinct walls, south side of the Ara Pacis, Rome. first century b.c. 205 41. Acanthus scrolls, detail of the lower reliefs from the outer face of the precinct walls, Ara Pacis, Rome, first century b.c. 209 42. Interior view of the mithraeum at Sta Maria in Capua Yetere, third century a.d. 211 43. Relief carving of tauroctone from the Wal- brook Mithraeum, London. 213 44. Painted tauroctone from the mithraeum at Marino, third century a.d. 213 45. Artemis of Kphesus, Roman copy of the now lost original. 215 46. Painted tauroctone from the mithraeum at the Palazzo Barberini in Rome, third century a.d. 218 47. Relief carving of tauroctone from Neuen- heim, third century a.d. 219 48. Wall mosaics, north wall, west side, Church of Sant' Apollinare Nuovo, Ravenna, sixth century a.d.: Procession of virgins leaving the citv of Classe, bottom tier; unnamed saints with books and scrolls, middle tier: Christological scenes of healing the paralytic at Bethesda, the Gadarene swine and healing the paralytic at Capernaum, top tier. 224 xii 49- Wall mosaics, north wall, east side. Church illustrations of Sant' Apollinare Nuovo, Ravenna, sixth centurv a.d.: Procession of virgins and Magi to the Virgin enthroned, bottom tier; unnamed saints with hooks and scrolls, middle tier; Christological scenes of the calling of Peter and Andrew, the miracle of the loaves and fishes and the miracle at Cana, top tier. 226 50. Wall mosaics, south wall, west side. Church of Sant' Apollinare Nuovo, Ravenna, sixth centurv a.d.: Procession of martyrs lea vine Theodoric's palace and the city of Ravenna, bottom tier; unnamed saints with books anil scrolls, middle tier; Passion scenes of the women at the empty tomb, the road to Em-maus and the risen Christ before Doubting Thomas and the apostles, top tier. 228 51. Wall mosaics, south wall, east side. Church of Sant' Apollinare Nuovo, Ravenna, sixth centurv a.d.: Procession of martvrs led bv St Martin to Christ enthroned, bottom tier; unnamed saints with books and scrolls, middle tier; Passion scenes of the Last Supper, the Asonv in theCarden at (lethsemane, the kiss of Judas and the arrest of Christ, top tier. 230 52. Detail of the procession of virgins. Church of Sant' Apollinare Nuovo, Ravenna, sixth century a.d. 232 53. Detail of the procession of martyrs, Church of Sant' Apollinare Nuovo, Ravenna, sixth century a.d. 233 54. Christ enthroned. Church of Sant' Apollinare Nuovo, Ravenna, sixth century a.d. 234 55. The Virgin and Child enthroned, Church of Sant' Apollinare Nuovo, Ravenna, sixth century a.d. 235 56. Diagram of the directions of mosaics at Sant' Apollinare Nuovo, Ravenna. 237 57. Sacrificial procession, mithraeum at the Church of Sta Prisca, Rome, third century a.d. 241 xiii illustrations 58. Venus at her toilet and the Christian inscription, Projccta Casket, fourth century a.d., Esquiline Treasure, British Museum. 252 59. The Projccta Casket seen from the front, fourth century a.d. 253 60. Vintaging erotes, floor mosaics from the Villa at Piazza Armerina, Sicily, fourth century a.d. 256 61. \ imaging erotes, ceiling mosaics from the Church of Sta Costan/.a, Rome, fourth century a.d. 257 62. The sarcophagus of Junius Bassus, Vatican, fourth century a.d. 258 63. Inner medallion of Scvso's hunting plate, Scvso Treasure, fourth century a.d. 259 64. Cup 15, from the Boscorealc Treasure (now in the Louvre. Paris), first century a.d. 262 65. Cup 104, from the Boscorealc Treasure (now in the Louvre, Paris), first century a.d. 263 66. Cup 7, from the Boscorealc Treasure (now in the Louvre, Paris), first century a.d. 264 67. Cup 8, from the Boscorealc Treasure (now in the Louvre, Paris), first century a.d. 265 68. The Riha Paten (now at Dumbarton Oaks, Washington), sixth century a.d. 268 69. David presented to Saul, from the David plates (now in the Metropolitan Museum), seventh centurv a.d. 269 70. The Missorium of Theodosius (now in the Real Academia de la Ilistoria, Madrid), fourth century a.d. 269 71. Plan of the Via Latina Catacomb, Rome, fourth centurv a.d. (after Terrua). 272 72. Diagram of cubkulum V Via Latina Cata- COmb, Rome, fourth century a.d. 273 73. Diagram of cubkulum C, Via Latina Catacomb, Rome, fourth century a.d. 2-3 74. Admetus dying, left-hand arcosolium, cubkulum V. Via Latina Catacomb, Rome, fourth centurv a.d. 2-6 \iv 75- 1 [ercules with Athena, left-hand areosolium, illustrations cubiculum Y, Via Latina Catacomb, Rome, fourth century a.d. 277 76. Hercules triumphing Over an enemy, left-hand areosolium, cubiculum \, Via Latina Catacomb, Rome, fourth century a.d. 278 77. Hercules restoring Alcestis to Admetus and holding Cerberus, right-hand areosolium, cubiculum Y, Via Latina Catacomb, Rome, fourth century a.d. 279 78. Hercules stealing the apples of the I lesper-ides, right-hand areosolium, cubiculum Y. Via Latina Catacomb, Rome, fourth century a.d. 281 79. The Arch of Constantine, Rome, fourth century a.d. 283 So. The Brescia Lipsanotheca (now in the Museo dell' Lta Cristiana, Brescia), fourth century a.d. 285 81. The Brescia Lipsanotheca, front, fourth century a.d.: Medallion portraits of Christ and four apostles, top tier (in fact, the edge of the casket's lid); Jonah swallowed and vomited by the whale, upper tier; Christ with Mary Magdalene (Soli me tangere), Christ teaching and Christ at the entrance to the sheepfold, middle section; Susanna and the Elders, Susanna and Daniel, and Daniel in the lions' den, lower tier. 2^ xv ACKNOWLEDGEMENTS This book was originally (a long time ago) a Ph.D. thesis. I must thank my supervisors, Mary Beard, Keith Hopkins and Roger Ling, for their help, encouragement and advice over a period of main vears. Mv friends, John Henderson and Robin Cormack, have read every part of this book in its process of gestation - their support has been invaluable. Anthony Snodgrass and Andrew \\'allace-I ladrill, my thesis examiners, provided a fundamental critique and many useful suggestions. At crucial stages, Peter Brown, Geoffrey Lloyd and Robin Osborne have made important criticisms which led me to restructure elements of the argument or to add substantive sections. Portions of the draft have been much improved at different times by the comments of Mvles Bur-nyeat, John Drury, Peter (Jarnsey, Simon Goldhill, Chris Morrav-Jones, Christopher Rowland and Christopher Walter. In many ways this book is a product of the very exciting times in Cambridge Classics in the second half of the 1980s. I should like to thank all those (then) graduate students and research fellows who collaborated in the shared enterprise of knocking down a few established subject boundaries - especially I amsvn Barton ami Catharine Edwards, who were mv partners in founding an interdisciplinary seminar, and Sue Alcock, Margaret Atkins, Sarah Currie, Penny Glare, Emily Cowers, Valerie Huet, Chris Kelly, Jamie Masters, Sitta von Reden, Peter Singer, Greg Woolf and Maria Wyke. In the latter stages of my work, I have had the benefit of comments from Yun Lee Too and of the detailed scrutiny of Jeremy Tanner. Outside Cambridge, Charles Barber and Martin Kauffmann provided excellent company during a stay in ACKNOWLEDGEMENTS Italy, while Liz James and Ruth Webb have always kept me on my toes. Institutionally, this project could not have proceeded w ith-out the support of three great centres of learning. At King's College, Cambridge, I had the benefit of a wonderfulK open academic atmosphere in writing the original thesis. The Master and Fellows of Jesus College, Cambridge, by their award of a research fellowship, allowed me to complete the bulk of rewriting and rc-rcwriting in the quiet and comfortable atmosphere of Chapel Court. The Courtauld Institute, in appointing me to a lectureship, afforded me the invaluable opportunity to try out some of my ideas on a series of M. \. students. I hope that those who became guinea pigs will one day forgive me! Finally, this book could never have been what it has become without my enjoying the intellectual company and friendship of a number of people outside classics and art history - especially Thupten Jinpa, Simon Coleman, Joan Pan Rubies and Silvia Frenk. On the bibliographic side, I should register my regret that John Clarke's The Houses of Roman Italy, 100 B.C.-2^0 A.D., a book germane to mv subject in Chapter 2, reached the United Kingdom only in mid-1992 (although it was published in the United States in 1991). It arrived too late for my final rewriting - from my point of view a most unfortunate casualty of the difficulties which seem recentlv to have beset trans-Atlantic book distribution. I am grateful to the following societies for permission to reuse material originally published in a rather different form in their journals: a portion of Chapter 3 appeared as "The Viewer and the Vision: The Case of the Sinai Apse" {Art History 17, 1 [March 1994]: 81-102), for which I thank the Association of Art Historians; material from Chapter 4 appeared as "Pausanias: A Greek Pilgrim in the Roman World" (Past and Present: A Journal of Historical Studies 135 [May 1992I: 3-29), for which I thank the Past and Present Society which holds world copyright in the article; a section from Chapter 6 appeared as "Cult and Sculpture: Sacrifice in the Ara Paris Augustae" (Journal of Roman Studies, 81 [1991]: 50—61), for which I thank the Roman Society. In the final straight of producing a book, I am grateful to Cambridge University Press for its support, to its many anonymous referees for their comments and to Beatrice Rehl and her team at CUP - including Janet Polata and my long-suffering production- and copy-editors, Camilla Palmer and Susan Greenberg - for helping bring this book to press. Kim win Howes went through my manuscript with admirable care and acknowledgements detail to correct errors and check references; Jamie Masters and SiKia Frenk came to my aid in compiling the index. In the acquisition of photographs 1 am especially grateful for the help of David Buckton, Ilene Forsyth, Martin Kauffmann and Marlia Mango. At ever) stage, Norman Brvson has proved an unstintingly supportive and encouraging editor. xix COLOR PLATES Plate i. Fresco on lime plaster, from the Villa of I', Kannius Sinisior. Boscoreale (now in the Metropolitan Museum), first century b.c. Photo: Metropolitan Museum of An, Rogers Fund, 1003. XXUl Plate Apse mosaic of the Transfiguration, Monastery <»t St Catherine at Ml Sinai, sixthOentun a.d. Photo: Krncst Hawkins Archive, Couittuld Institute. xxiv Plate 3. The Projects Casket seen from the front, fourth century a.d. Rsuuiline Treasure, British Museum. Photo: Courtesy <>t the 1 rustees of the British Museum. Plate 4. I lercules fighting the I h dra, righthand arcosolium. aibictilum \, \ 1.1 latina C^atacornb, Rome, fourth cen tury a.d. Photo; Pont. Comm. di \rch. Sacra, Rome. wvi INTRODUCTION / want a History of Looking . . . Roland Barthes1 I At the core of this book is a simple proposition: People relate-to works of art in different ways, depending upon different contexts and at different times. Confronted with an image of a fish, a Roman dinner partv host, a modern angler or a cookery book writer of any period might sec a picture that corresponded as closely as possible to the real fish which he or she might eat, catch or cook; each person would notice differences from and similarities to the real thing - how realistic the picture was. Take for instance the various marine animals displayed in any of a number of Roman mosaics, such as the mosaic in Figure i, depicting fish and an octopus, originally from the I louse of the Faun in Pompeii and now in the Naples Museum. Such images evoke associations not only with the sea, by which Pompeii was situated, but also with the sea-food on which Pompeii's inhabitants dined. Confronted w ith the same image, an early Christian might see not the picture of a real fish, but a sign, a symbol for what the fish represented to him or her in the religious system of early Christianity.2 On the walls of the catacomb of Calistus in Rome are two images, each representing a fish with a basket of loaves (Figure 2). Despite their associations with eating, these frescoes in their explicitly Christian funerary context evoke not an ordinary meal, but such scriptural meals as the miracle of the multiplication of the loaves and fishes and such ritual meals as the eucharist. Moreover, beyond the image of dining, the fish evokes Christ himself and his apostles as fishers of men. For a Christian in the early centuries of the first millennium, the symbol of the fish stood as a sign for introduction Figure i. Mosaic offish and octopus, from Pompeii (now in the Naples Museum), first century B.c. Photo: Alinari-\rt Resource, New York. Christ; and the word "fish", in its Creek form of iciithys, was an acronym consisting of the first letters of the formula "Jesus Christ, Son of Cod, Saviour". These arc remarkably different ways in which people choose to understand what may appear to he (from a formal point of view) rather similar pictures. Such different understandings presuppose very different processes of interpreting what one might have thought was the rather unprohlcmatic image of a fish. In effect, to move from the initial perception of a picture to the meaning or referent which one understands the picture to he evoking is rather a complex process which is not the same for every individual. It is a dynamic of interpretation. This book is not about the process or psychology of looking, but rather about the different conceptual frameworks which interpret what is seen to make it meaningful." Different viewers' frames of interpretation, whether literal or symbolic, metonymic or metaphoric, naturalistic or allegorical, are what tend to constitute the kinds of meanings which images bear. In principle, any work of art can give rise, in different observers (and sometimes even in the same viewer), to a multitude of varying and even contradictory responses and meanings. The transformation in the forms of art, which in part this book investigates, was in the first place a transformation of content whereby different viewers, such as a pagan Roman and a religious Christian, might see the same image in such profoundly different ways. For one, it might have a realistic meaning relating to the natural or material world; for the other, a symbolic meaning, relating to the Other World - the Sacred. The following pages explore some aspects of the transformation of Roman art. Part I examines a great variet) of ancient texts and images to discover the ways in which Romans looked at their art. The taxonomy of viewing which is presented divides essentially into realist ways of viewing on the one hand - with the enormous range of psychological, introduction Figure 2. fresco of fish ami bread-basket, from the Catacomb ofCilistus, Rome, third century a.o. Photo: Pont. Comm. di Arch. Sacra, Rome. 3 introduction dogmatic and political implications that these display - and symbolic ways of viewing, on the other, which are primarily (though not exclusively) the domain of religious and initiate cults such as Mithraism, Neoplatonism and Christianity. Part II explores a number of important monuments, dating from the age of Augustus at the turn of the first centuries b.c. and a.d. to the age of Justinian in the sixth century a.d., in the light of how they were viewed. By comparing these monuments - by observing the changes both in their formal characteristics and in the ways they were looked at over the period from the first to the sixth centuries a.d. - we can trace the contours of one of the most remarkable and comprehensive transformations of culture that took place in the Western world before the Renaissance. Part III, by way of an epilogue, turns from this broad taxonomie approach to look briefly at some of the modulations and complexities in what was a very gradual and multifaceted process of change. I look there at the genesis of a specifically Christian exegetic mode of viewing - how it resembled and how it differed from the kinds of initiate viewing in pagan antiquity out of which it arose. 11 What, then, is viewing? One answer is that viewing is one activity in which people confront the world. They themselves may change under the influence of what they see, or what they see may cease to be a neutral object and become something interpreted by them according to the prejudices and associations present in their minds. Viewing is always a dual process of interpretation in which what is seen becomes fitted into the already existent framework of the viewer's knowledge and thereby, very subtly, changes both the content of what the viewer knows (because something new has been added) and the meaning of what is seen (because it is now framed by the viewer's know ledge). The history of viewing, expressed in this way, becomes in principle a historv of how people interpret anything in their world - nature, art, other people. By focussing on art, I have limited the theme to a very narrow range of the totality of objects that influence and are interpreted by human consciousness. However, the viewing of art is a special case that can perhaps elucidate other kinds of viewing too. This is because Creek and Roman intellectuals were more self- 4 conscious about how they interpreted and viewed their art introduction than they were about how they interpreted and viewed other material objects in the culture. They even wrote at considerable length about the subject. One reason tor the importance of the viewing of art is that it is a secondary process. Looking at an image, a beholder is assessing a view of the world already created by an artist. In the viewing of art, the artist's own creative viewing of objects is a critical mediator between the beholder and the world. The Greeks and Romans were particularly interested in the mediating processes of artists, and some - like Plato - were very worried by them.4 By tracing the multiplicity of responses to art in ancient texts, and by examining the presuppositions that give rise to these responses, we can learn a great deal about the different kinds of subjectivity - the many ways of viewing - which existed in Roman times. We can get close to some of the deeper inclinations of Roman thought. In effect, the ways Romans looked at the world were in some sense conditioned by the ways they looked at art and vice versa. To examine the relatively limited area of Roman responses to art gives us a means of approaching the much larger theme of the w ays in which Romans responded to other features in their culture. The focus on viewing offers a precise way of approaching a central issue in the interpretation of Roman art: How are we to understand the change between Classical art, with its emphasis on likeness to the natural world, and the abstraction of late-antique art? This is a problem which has exercised art historians since Yasari in the sixteenth century. Some have looked to a "decline" of technical skills, some to changes in the structure of society, some to religion and the rise of Christianity. Most approaches to the transformation of Roman art or to the rise of Christian and Byzantine art (which is usually another way of describing the same phenomenon) locus not on the total process of change from the Classical period of Roman art under Augustus to the first full flowering of Christian art under Justinian, but on the particular moment of transition. Scholars have tended to cite specific groups of objects from the third and fourth centuries (such as the art of the Dominate, the Arch of Constantine, the church and synagogue of Dura Luropos or the catacombs of Rome) as the key to change. An exclusive exploration of the "moment of transition" is adequate onlv if the differences between the art of Augustus and Justinian are self-evident. These differences - which cer-tainly exist in the style and appearance of works of art - are introduction not self-evident or self-explanatory, unless one has a formalist approach to art In which style becomes the primary criterion of change. Because I am questioning formalism or style as the main criterion for judging art, I do not see the differences between art in the first century and art in the sixth century as obvious or self-cxplanatorv. On the contrary, I believe thev need understanding and explaining. And the arena of explanation must lie in a broader conception of art than formalism, in a conception that gives due weight also to the functions and meanings of art within Roman culture - that is, to the ways in which art was viewed. But to understand such a broad span of change requires a grasp of the art at both ends of that span, as well as an investigation of the period of transition. It is more interesting, therefore, and more revealing, to employ a broad comparative method in which one both grasps and then compares the relationships between objects, their producers and their viewers at different moments within the process of change. Only when we understand the changing principles of viewing and looking at art throughout the period of transformation (from Augustus to Justinian) can we grasp what the differences are between art in the first century and the sixth. The empirical and inductivist creed of academe would have us approach a body of material, for instance, Roman art, with an open and cntirelv unprejudiced mind - prepared to construct an interpretation motivated solely by the suggestions of the evidence. Unfortunately, our minds are not like that. Like the viewers of art, any writer or researcher brings to what he or she studies the history of his or her intellectual training, personal attitudes, unconscious prejudices. Moreover, we must tread carefully through our chosen area, strewn like a minefield with the intellectual wreckage (the views, positions and prejudices) of centuries of scholarship. It is on these foundations that we build; it may often be against them that we find ourselves fighting. In this sense, academic study is no different from the ways of viewing 1 explore in this book. Like the viewing of art, the study of the evidence is prev to the multiple human fallibilities of subjectivity. In looking at the transformation of Roman art, there are, it seems to me, two attitudes in particular by which one might characterize the nature of scholarly approaches. These may be summarized (and parodied) by the sentences: "Everything changed" and "Nothing changed". In a way, I agree with 6 both these views. I accept the position of the Renaissance - a introduction position repeated in this century by distinguished scholars such as Bernard Bcrenson, Frnst Gombrich and Ernst Kit-zinger - that there was a fundamental transformation in the making and understanding of art in late antiquity/ It was part of a fundamental transformation in culture, society and identity. I low ever, I disagree strongly with the characterization of this change as "decline". On the other hand, I accept that almost everything in Christian art (indeed, in Christian culture) is the direct descendant of elements, attitudes and forms present for centuries in Classical civilization. Nonetheless, such elements were clearly transformed by aspects of the Jewish culture out of which Christianitv emerged. From this point of view, one can argue that nothing changed radically, that there was a slow , delicate ami subtle process of modulation and transformation.6 Modern scholars have been particularly distinguished in their careful tracings of the dynamics underlying this process. Indeed, art history has recently been commended for its judiciousness in not portraying the contrast of pagan and Christian themes as a life-and-death struggle. Art historians have traced the delicate process of the rise of late-antique abstraction from numerous angles. Some have explored sociopolitical influences (looking at the impact of social change/ of non-Classical artistic modes from the periphery of or beyond the borders of the empire,9 including especially Jewish art,10 or of class-conflicts between "plebeian" and "patrician" elements within Roman culture)." Some, on a more conceptual level, have looked at the impact of late-pagan philosophical thought, especially that of Plotinus, which advocated a less descriptive or literal discourse so as to emphasise the spirituality of what was represented.Some have turned to the rhetorical and literary background of ancient culture to explore developments in imagination and audience response.13 All these approaches have the considerable merit of not seeing decline as the crucial characterization of change in late-antique art. My problem with them, however, is that they tend - bv emphasising elements of continuity - to downplay the significance of change. Despite the fact that Christian art (indeed. Christian culture) owed pretty well everything to the combination of its Graeco-Roman environment with its Jewish origins, I believe it to have become (very early on) something distinctively, even radically, different.14 In Fart III, I try to examine something of this difference by looking at the origins of Christian visual typology. » 7 INTRODUCTION III This book began as an attempt to look again at the formal and stylistic change in Roman art from what has been called "naturalism" to w hat has been characterized as "abstraction", between the first and sixth centuries a.D.,j My aim was to look at cultural contexts and at the ways art was viewed, to see if a broader historical explanation could help us understand artistic change. But as it progressed, the lxx)k became something a little different. What Art and the Roman Viezver has come to explore arc two radically different conceptual frames within which viewing (and many other aspects of social life) took place in the ancient world. Whether one looks at first-century images of Roman domestic wall-painting (Chapter 2), of sacrificial ritual (Chapter 6) or of silverware from the context of elite dining (Chapter 7), one finds a complex potential for irony, parody and dceon-struction. Like the subtle complexity of many Roman texts of this period (from the poetry of Horace, Ovid or Juvenal to the prose of Petronius and Tacitus), such imagers gently (or cruellv) mocks its own context, subverts am simple or straightforward view of its iconographic and social function and undermines not only w hat it represents but also its Own potential for subversion. Of course, this is only one aspect of the complexity of Roman culture in the early empire. Its "deconstructive" possibilities flourished in an age which was also deeply religious and whose art, art criticism (see the discussion of Cebes in Chapter 1) and view ing (see the discussions of "mvstic view ing" and Pausanias in Chapters 3 and 4) could take on quite un-, indeed, anti-, ironic modes of religious allegory, symbolism and initiation. I must emphasise that I see such "deconstruction" as only one aspect of Roman culture, but I also believe that not to see it (as some scholars do not) is the price (or priz.e) of not w anting to see it. By contrast, the arts and literature of the sixth century exhibit no such "deconstructive" or ironising self-reflections. This does not mean they are "simple". On the contrary, religious images such as the pilgrimage mosaics of the monasters at Mt Sinai (Chapter 3) or the cycle of martyrs and Christian sacrificial images at Sant' Apollinare Nuovo in Ravenna (Chapter 6), as well as the sixth-century imperial iconography of San \ itale in Ravenna (Chapter 5) and even some seventh-century dining silver (Chapter 7), show remarkable sophistication and polysemic complexity. But this range of s signification essentially aims through iconography to support introduction its subject-matter with the scriptural underpinnings on which Christian culture and identity came to be founded. The reference of such Christian imagers was not to what it appeared to represent, but to the symbolic meanings of a sacred text. In the end, all the arts of Medieval Christian culture came to be based on a brilliant ami symbolic pattern of scriptural typology and exegesis. My argument is that these two cultural contexts were profoundly different frames for the ways viewers formulated their responses to images. In first-century Roman culture, it was always possible to find an ironic or polemical place from which to look askance at the culture's social construction and self-representation. Naturalism, as a style, broadly goes with such a context, for naturalism - the art that imitates as closely as possible hut never is what it represents - always exhibits a gap between the object represented (which the image so desires to be) and the imitation or image. It is precisely this gap, and the analogous gap between the image as imitation of a real object to be desired or possessed and the viewer as one who desires but cannot possess what turns out to be a mere illusion, which all the "dee 3 \ L nder the imaginary table that separates me from my readers, don't Wt secretly clasp each others hands? Bruno Schulzw Those who write about the process of interpretation, ami especially about the ways interpretations change, must be acutely aware that they themselves are interpreters. Just as Roman viewers looked at art, so I too am a viewer of Roman art as well as an explorer of Roman views about their art. And my survey is itself a view which my readers will interpret and review in their turn. This endlessly reflexive process gives birth to a kind of humility. Although our interpretations are what make our world, the only certainty they carry is that they will be superseded. In calling to mind the impermanent, insubstantial nature of all of our views, there is a certain perspective: I low limited are our know ledge and our activity. It is in this spirit that I offer this book to my readers. WAYS OF VIEWING IN THE ROMAN WORLD foreword Even in ancient times, the history of Classical art was written as a history of the rise and triumph of naturalism. For Roman writers of the first century a.o., like Quintilian and Pliny the Elder, the art of painting developed gradually from the monochrome efforts of its earliest exponents to the acme of Apelles, an artist "who surpassed all the painters that preceded and all who were to come after him" {Natural History 35.79).1 Pliny presents the development of art as a passage from winter to spring in which the great artists were those who brought innovations into the practice of painting, allowing it to become increasingly more realistic.2 Eumarus of Athens was "the earliest artist to distinguish the male from the female sex in painting" (35.56). Cimon of Cleonac "first invented \atugrapha\ or profile drawings" (35.56). Panacnus of Athens "is said to have introduced actual portraits" (35.57). Polygnotus of Thasos "introduced expression in the countenance in place of stiff archaic features" (35.58). Apollodorus of Athens "was the first to give realistic presentation of objects" (35.60) and into the gates which he had thrown open, there "entered Zcuxis of I Ieraclea . . . who led forward the already not unadventurous paintbrush . . . to great glory" (35.61). Lest we believe this story too keenly, let us reflect that it represents a Roman synthesis of a long tradition of Hellenistic scholarship and art criticism. This Hellenistic literature, which unfortunately is now largelv lost, was produced at the courts of the successors of Alexander the (ireat. For reasons perhaps more political than factual, the zenith of Greek art -in the persons of the painter Apelles and the sculptor ways of viewing Lvsippus - was constructed to coincide with the reign of Alexander. Apelles was Alexander's court painter {Natural History 35.85—7) and Lvsippus his court sculptor (34.63-4). The storv may of course be true, but it reflects all too clearly a concerted scholarly effort at the behest of Alexander's successors to justify a conqueror from Macedon (a country which had always been perceived to be semi-barbarian) as embodying the apogee of Greek culture. The Plinian pattern represents an artistic relay-race with each artist handing 00 the torch of naturalistic imitation in an ever-increasing crescendo of illusionistic verisimilitude. This description puts an emphasis on the figure of the artist as innovator and on the goal of art as naturalism which Western art since the Middle Ages and the historiography of Western art have never yet been able to shake off. When great painters have defects, these are that they fall short of the naturalistic ideal at the heart of Pliny's picture. "Zeuxis is criticized for making the heads and joints of his figures too large in proportion" (35.64). Parrhasius of Fphcsus, "the first to give proportions to painting and the first to give vivacity to the countenance . . . won the palm in the drawing of outlines" (35.67) but "seems to fall short of his own level in giving expression to the surface of the body inside the outline" (35.60). Moreover, because naturalism is an ideal, a sculptor like Demetrius can be "blamed for earning realism too far", for he "is less concerned about the beauty than the truth of his work" (Quintil-ian, Inst. Or. 12.10.9). Inevitably, given the structure of this account based on rise and fall, the (Roman) painting of Pliny's own day is a decline from the supreme peaks achieved by the Greeks (e.g.. Natural History 35.2 and 28: "thus much for the dignity of a decaying art"). And yet the Plinian pattern always leaves room for innovations in the cause of naturalism. Kven in the days of Augustus, StudiuS "introduced a delightful style of decorating walls" (35.1 16). The Plinian pattern of the art of the past as an ascent to naturalism (with the present seen as a state of decline) is not original. We rind exactly the same formula in the architect Vitruvius, writing in the last quarter of the first century B.C. (during the reign of Augustus and the career of the fresco painter Studius). For the latter, in Ik Architectura (7.5), painting arose as an imitation of nature. Its decline is represented by the surrealism of the murals of \ itruvius's own time which imitate not existing things but all sorts of monstra and falsa. This ancient view presents the development of Greek art 16 as an evolution of illusionism. It sees the most triumphant ways of viewing achievements of art to be those statues or paintings which are so realistic that the viewer is deceived into believing thcv are real. In the most famous of Pliny's anecdotes {Natural History 35.65): Parrhasius and Zeuxis entered into competition. Zeuxis exhibited a picture of some grapes so true to nature that birds flew up to the wall of the stage. Parrhasius exhibited a linen curtain which was painted with such realism that Zeuxis, swelling with pride over the verdict of the birds, demanded that his rival remove the curtain and show the picture. When he realised his error, he yielded the victory, frankly admitting that whereas he had deceived the birds, Parrhasius had deceived Zeuxis himself, a painter/ At the heart of this anecdote, the genius of illusionism is ultimately defined by its ability to deceive. The more skilfully art can deceive even those, like the painter Zeuxis, who are skilled in deception, the more realistic and hence the better it is. In the myth of Pygmalion, the supreme myth of Classical art as realism, as it was formulated in Ovid's Metamorphoses (10.245-97), this process reaches its peak. Pygmalion's ivory statue deceives not the non-human birds, nor a rival artist, but the very maker himself into believing that it is real. Pygmalion falls in love with the woman he has created, and eventually, in a miracle performed by Venus, is granted his heart's desire when the ivory woman is turned into real flesh. Clearly these Roman texts present a coherent theory on the nature and history of ancient art. They chronicle an evolution and highlight the complex psychology implicated in realism. But do they tell the whole story? What do we lose in understanding ancient art when we rely principally on the Roman historiographical pattern retailed by Yitruvius and Pliny? Other ancient texts about images, rarely found in any art-historical discussion, such as the Tabula of Ccbes (a well-read work in antiquity, so far as we can tell), give a very different impression about what at least some people wanted to see in art. Moreover, the most famous images from antiquity, such as the Olympian Zeus and Athena Parthcnos of Phidias, were cult statues much larger than human size, executed in ivory and gold, and displayed in artificial lighting inside impressive temples. They do not survive. But there is no evidence that they would have appeared the slightest bit "naturalistic" (whatever that term may mean in the religious context of a cult deity for whom there was no prototype in nature).4 Naturalism may have been less comprehensive and less significant in the ancient world than those who exclusively trust •7 ways of viewing Pliny and \ itruvius might like to claim. But there is a second point. In antiquity, naturalism was not an objective imitation of the natural world, as ancient. Renaissance and modern writers have imagined. Its history, as we have seen, was told by Alexandrian writers w ith a clear political interest in w here the peak of its development should be placed. Naturalism, like all styles propagated by political states, had a politics. It was, as I suggest in Chapter 2, the aesthetic and stylistic standard-bearer for the visual self-presentation of the Roman Empire itself. In the four chapters which constitute this part of the book, I seek to replace the Plinian picture with a portrait of Roman art which emphasises the variety of kinds of viewing in the ancient world. The psychology and the politics of naturalism are certainly part of this picture. But no less important are the kinds of religious images which occasioned the dialogue written bv Ccbes or the pilgrimage through Greece of Pau-sanias in the second century a.d. It was this persistent association of images and religion, of sight and the sacred, throughout pagan antiquity that would be the ancestor of Christian art after the conversion of Constantine. lor those viewers, whether Christian or pagan, who wished to be initiated into the deeper meaning of sacred images, naturalism became explicitly unacceptable. Naturalism, as the Younger Philostra-tus - one of the last of its ancient apologists, writing in the fourth century a.d. - admitted, was inherently deceptive. For Philostratus, the inevitable deception of the naturalist attempt to imitate objects was "pleasurable and involves no reproach".' But this very deception became a sign for ethical deviation in the minds of those who were seeking religious Truth. Deception could not co-exist with Truth. This is why certain Church Fathers, like F.piphanius of Salamis - a contemporary of the Younger Philostratus, writing in the latter part of the fourth century a.d. - would accuse images of "lying". Painters "lie" because they cannot depict Christ or the saints as they actually were and instead present them "in different forms according to their whim"/' Likewise, Augustine, in the Soliloquies, believed art w as "unable to be true". Because of the inevitable deception by which image-making works: "A man in a painting cannot be true, even though he tends towards the appearance of a man".7 In essence, my argument is that naturalism - never the only mode of representation in the ancient world - became an untenable mode by late antiquity. In a culture which subjected the artefacts it produced to increasingly complex 18 symbolic, cxcgeric and religious interpretations, art was ex- ways of viewing pected to stand for symbolic and religious meanings rather than to imitate material things. Not only was the mimetic illusionism of naturalistic art no longer necessary to late-antique culture, but its very attempt to deceive was a barrier for those who sought truth or religious edification in images. Art became "abstract" or "schematic" not because of a decline in taste or skill, or for simple reasons of political appropriation, but because viewers, patrons - in fact, the collective taste and subjectivity of the culture - wanted it that way. In structure, this section of the book is not arranged chronologically. My aim rather is to elaborate and explore syn-chronicallv two motifs, which were highly significant for viewing throughout antiquity, interwoven in a sort of fugue. In Chapter i, I introduce two themes or modes of viewing which I suggest were always present in Graeco-Roman culture. One (which I expose through the elegant discourses on pictures delivered by the Elder Rhilostratus) involved the viewer's flirtation with the deceptions and trompe Foci I of naturalistic images as representations of the material world. The other (which I explore through the religious interpretation of a picture by Cebes) engaged viewers in an allegorising exegesis of images which may have resembled items in the material world, but were explained in a religious context as referring to the Other World. These two ways of viewing, which may be said to correspond broadly (but not precisely) to the "secular" and the "sacred" as social or cultural definitions, were both implicated in the ideologies and politics of ancient society. The next three chapters address the theme of viewing and subjectivity. How does the act of looking at art condition and create the identity of individuals? I explore this problem by taking three case-studies from the ancient world - viewing in a domestic context, viewing in a sacred context and viewing in the complex situation of a pilgrim whose journey is through a social and secular world, but whose goal is religious. These three studies expand the basic dichotomy examined in Chapter i between Cebes and Rhilostratus, between reality as seen to be something otherworldly, open only to allegorical description, and reality as located firmly within the material grasp of the desires of this world. Chapter 2, on Roman domestic housing, aims to illuminate some of the many constraints and contradictions - individual, social and political -which inevitably press upon and compete in the production of an identity in the secular and day-to-day world. Chapter 3, on mvstic viewing and the mosaics at Sinai, presents ways OF VIEWING admittedly an elite and extreme version of the formation of subjectivity in a sacred context within Christianity. Chapter 4, on the traveller Pausanias, illustrates through an exploration of a single individual, as he looked at art and recorded his views, some of the problems involved in bridging these two antithetical worlds. The cases I present in these chapters are exempli gratia. I do not mean to imply that socio-political viewing of the kind I evoke in first-century wall-painting did not exist in later periods; nor to suggest that mystic viewing of the sort I explore in the Christian mosaics at \lt Sinai did not flourish in, say, the Julio-Claudian period. On the contrary, the first-centurv date of Cebes is proof that such sacred viewing did exist in the pagan Roman world. Moreover, the theological language of Christian discourse should not disguise the often deeply political content which such language came to subsume. Nonetheless, I do suggest that the development of Christianity was attendant upon, perhaps even w as to some extent caused by, a shift ol emphasis towards the sacred over the socio-political, towards the mystagogic oxer the secular. If these two strands have always existed side by side in human experience, then to some extent different periods are marked bv the predominance of one over the other. My argument is that the change from the ancient to the Medieval world was precisely attended by such a shift of emphasis, and that in such rare and fascinating works as the book of Pausanias we can observe that shift as it was happening. By presenting the poles of secular and sacred through objects from the temporal limits of my Study, I risk misrepresenting the complexity of the process by implying that Augustan art is secular and Justinia-nie is sacred. I do not mean to suggest this. The positive point of mv strategy is to underline the poles (both conceptual and temporal) in an attempt to make clear what exactly the differences were (despite all the numerous caveats and exceptions). For I do believe, perhaps unlike some scholars, that there reallv was a fundamental change not only in art forms and social structures but also in the very notion of identity and in the frameworks governing people's subjective responses to the world. The period of late antiquity, notwithstanding such late and passionate apologists for naturalism as the Philostrati and Callistratus, saw a general move towards initiate and exegetie modes of interpreting art (essentially religious modes) which gradually came to dominate, often eventually to exclude, the emphasis on seeing images as referring naturalisticallv to the material world. chapter i VIEWING AND "THE REAL": J?JZf THE IMAGINES OF PHILOSTRATUS AND THE TABULA OF CEBES This chapter examines two ancient texts about the viewing of art which appear to have been somewhat influential in antiquity. As a twin exploration of two parallel but different texts, it tells a complex story. Both the Imagines of Philostratus and the Tabula of Cebes arc ancient interpretations of art whose consummate rhetorical sophistication makes them difficult but rewarding works to analyse. The problems of relating words and images have never been adequately resolved, because it is impossible to reconstruct fully a visual and essentially non-verbal experience (that of looking at a picture) in a text, however imaginative and creative the text may be. Although significantly different in the kinds of expectations which they presuppose about images, the Imagines and the Tabula both treat art as a means of education. They demonstrate to their readers how to create a contextualizing narrative (a very different kind of narrative in each case) through which the viewer may assimilate a picture into his or her experience. This process of assimilation is reflexive; it changes the painting bv subjecting it to the interpretative eye of its beholders, but it also changes the beholder by the educational process of his or her confrontation with a didactic image. What nobody has examined in terms of its formal implications for the history of art is the rctlcxivity of viewer and object each constructing the other. This reflexivity is highly complex: It is the aim of this chapter to show some of the ways in which it works in the cases of both these texts. Once we understand how this rerlexivitv operates, we can use-it to analyse both an ideology of viewing (the extent to which certain viewer strategies of assimilation are prevalent 2 i ways of viewing at different times and contexts) and a formalism of objects (why certain kinds of art, such as naturalistic or abstract, are dominant in different periods and societies). Indeed, these two projects are hardly separate. Behind both these texts there turns out to be a deep philosophical basis on which is predicated the kind of viewing that each advocates. The formalism of Philostratus's paintings -their naturalistic verisimilitude - is grounded in a theory which sees "reality*' as being constituted by the world of the viewer's ordinary physical and psychological experience, a world of common sense and materialist expectations. Bv contrast, the abstract or schematic nature of the tablet described bv Ccbcs (the content of which is consistently personified and allegorized) is rooted in an anti-materialist religious conception of "reality" which is defined as the transformation of the world of ordinarv assumptions through initiation. In each case an analysis of the rellexivity of images and the ways they are viewed can lead us to an understanding of the very different presuppositions about "realitv" which underlie very different kinds of viewing and very different kinds of art. In effect these two texts turn out to offer contrasting ways of relating art to life and of relating the viewer's life to the art he or she looks at. In Philostratus, the sophist, whose text propounds the interpreter's view of the works of art it describes, the paintings are presented as highly naturalistic. The viewer is offered a number of strategies for assimilating these images into his or her phantasy life. The viewer is encouraged in the illusion that he or she can control the "other", the picture, by incorporating it through a narrative contcxtualiza-tion into the world of the viewer's psychological and cultural experience. By contrast, Cebes presents the hearer's view -the experience of a neophvte listening to a profound religious exegesis and gradually initiated into a new understanding of the world. lor Cebes, the picture is a door into a reality w hich is entirely outside his cultural or psychological expectations. In this allegorical version of art, the viewer (far from controlling the "other") is subservient to the initiatory effects of the image and the exegesis of its interpreter. For Philostratus, paintings are a realistic illustration of the world we ordi-narilv know and can best be experienced through our normal response to the world. In Cebes, the picture is an allegorical bridge out of the world we ordinarily know into a reality and a way of life that is spiritually superior. 22 art and its df.scriftion: ekphrasis in the ancient world viewing and "the real" Si (come elgriego a firm a en el C.ratilo) El nombrt a arquetipo tie la eosa, En las let i as tie rosa est a la rosa } todo el Silo en la pah/bra Nilo. Jorge Luis Borges1 Picture a villa on the Bay of Naples. Its layout is lavish; its appointments exquisite. The <>\\ ner is a wealthy and cultured man: He possesses a choice collection of paintings displayed in a special gallery. Into this gallery comes a professional interpreter of art, a critic from outside Naples so famous that numbers of young men come to him, eager to hear what he has to saw This critic is Philostratus. At the express invitation of his host, the owner of the villa, and particularly as a result of the persistent urging of the owner's young son, Philostratus delivers a series of interpretations of the paintings to the group which has gathered to hear him. This is the setting for the ekphrases, that is, "descriptions", of the works of art collected in the Imagines of the Elder Philostratus.2 We know next to nothing about their author, save what he himself chose to tell us about his own distinction as a critic of art. He shares his name with a number of other writers from the second and third centuries a.d.' He himself probably wrote in the mid-third century, and his two books of descriptions were sufficiently successful to be imitated bv later authors.4 Certainly the display he gives to the audience gathered in the Neapolitan gallerv is dazzling.5 Historic and grandiose subjects are brought into the living-room of the present and personal. By his exemplary skill in the art of description he can cause remote and epic themes to penetrate the privacy of viewers' and readers' phantasy lives. At i .4 (4), before an arresting painting of the death of the hero Men-oiceus at Thebes, which illustrates a poignant moment from the great epic poem, the Tbtbaid, Philostratus urges: Let us catch the blood, my bov, holding under it a fold of our garment; for it is flowing out, and the soul is already about to take its leave, and in a moment you will hear its gibbering cry. The freshness and vitality of his rhetoric have carried his listeners deep into the tragedy which the picture depicts. The ways of vikwino viewers of these paintings, in the hands of this critic, seem themselves to have become part of the painted reality. Clearly the critic is a master of his art. These ekphrases, taken as a whole, perform a Callimachean master-stroke in taking the genres of Graeco-Roman literature from epic to fable, from landscape to still life, and conflating them into a pair of little books. The remote has been made vivid, the distant immediate, and the themes of epic verse have been transformed to prose - and, to cap it all, Philostratus is pretending that this is not at all a self-confident act of literature, but simply the faithful description of art! Every game, every move on the writer's part, is a deliberate revelrv in the literarv possibilities of his text and in his Own abilities as a master rhetorician. And so, we are entitled to ask, does the setting he describes represent a real occasion or an imaginary one? Did the gallery really exist, or was it a figment of Philostratus's rhetorical imagination - a literarv device for the fictional framing of his descriptions?6 Did the paintings themselves ever exist, or are they too the result of a vivid descriptive imagination seeing in its mind's eve what need not exist? These questions cannot be answered, because there is no external evidence other than the text of Philostratus himself. And this text is an example of scintillating and at times very complex rhetoric. Hut the fact that the text allows a deep doubt about the actual existence of anything (w hether paintings or gallery) described in these ekphrases is extremely important. For the literary genre of ekphrasis, or "description", in the ancient world was not intended (as modern descriptions are) to go beside and to supplement the real painting or statue being described; it was intended to replace the sculpture or painting. 1 he truly triumphant ekphrasis was the one which brought to mind its subject so vividly that the subject was no longer necessary. Its effects had already been achieved. Before we can plunge into the psychological subtleties of viewing so effortlessly evoked by Philostratus, we must pause and examine more deeply the genre in which his work is cast. What, then, is an ekphrasis? Fortunately, we do not have to rely only on modern theory. There survive from the first to the fourth century a scries of prescriptions for ekphrasis in the progymnasmata (exercise manuals) of various rhetoricians including I Iermogencs of Tarsus, Aphthonius and Theon.8 The remarkable thing in reading these texts is how everything - the words used, the order of ideas and even the exempla cited - is repeated from one writer to the next. This *4 not onlv indicates something about the repetitive nature of viewing and "the text-books, but also implies a fairly static notion of what an real" ekphrasis was. Theprogymnasmat'a are invaluable as a comparison with those descriptions which purport actually to fulfil the instructions of the rhetors. No text should be more suitable for comparing with the text-books than Philostratus, for he was a rhetor-sophist presenting a series of rhetorical declamations in the form of ekphrases to young men. I shall quote Hermogcncs on ekphrasis in full.9 Ekphrasis is a descriptive account; it is visible - so to speak - and brings before the eves the sight which is to be shown. Kkphrases are of people, actions, times, places, seasons and many other things. An example of people is I lomer's "he was bandv-leggcd and lame in one foot" [Iliad 2.217]; of actions, the description of a land or sea battle; of times, peace and war; of places, harbours, sea-shores and cities; of seasons, spring, summer and festival. You could also have a mixed ekphrasis - such as the night battle in Thucydidcs. For night is a time, but battle is an action. We try to describe actions from the point of view of what preceded them, what happened in them and what occurred after them. For example, if we were to make a description of war, lirsi we would tell of the events before the war, the levying of armies, the expenditures, the fears; then the engagements, the slaughters, the deaths; and then the trophy, the paeans of the victors and - of the defeated the fears and the enslavement. If we describe places or seasons or people, we will present the subject through a description and an account that is beautiful or excellent or unexpected. The special virtues of ekphrasis are clarity and visibility; the style should contrive to bring about seeing through hearing. However, it is equally important that expression should tit the subject: if the subject is florid, let the style be florid too, and if the subject is dry, let the style be the same. It is noteworthy that in none of the rhetors is the description of a picture cited as an example of ekphrasis. However, there is a strong emphasis in all the progynmasmata on claritv (sapbe-neia) and visibility (enargeia), which are presented as "the special virtues of ekphrasis".111 The ekphrasis is a description - which is to say, a reading - of a particular object or event so as to "bring it to sight", to make it visible. It is therefore a reading that is also a viewing. In the ideology of rhetorical declamation in the Second Sophistic (that golden age of Greek intellectual life in the heyday of the Roman Empire), to hear an ekphrasis is also to see what was described, and to write an ekphrasis is to make the description visible. But what is "visible" to the rhetors is far more than what 2 5 ways of viewing \vc would claim to be able to sec. Ilermogenes wants the whole fact of an "action" (including, specifically, its narrative qualities) to be contained in the description - the total effect, for example, of war in its economic, strategic and psychological reverberations. Although these are expressed in terms of all the time-honoured tropes - from levies to trophies, from death to enslavement - what he wants is in effect an 'interpretation and not a "description". The reader's seeing will come alx)ut from hearing the totality of the event as interpreted by the sophist, plus a stylistic mimesis of the "quality" of the event effected by the virtuosity of the sophist's rhetoric. I lerc the theory and practice of ckphrasis (as defined in the progymnasmata and exemplified by Philostratus) are deeply indebted to Stoic views ofphantasia, which means "visualisation" or "presentation"." The precise role ofphantasia in Stoic-philosophy is still controversial, but there is little doubt that (as the criterion for truth12) it was an essential concept.11 To the Stoics, phantasia was a visualisation or presentation from an object. It imprinted itself upon and in some sense altered the soul.14 By extension the term came to describe "the situation in which enthusiasm and emotion make the speaker see what he is saying and bring it visually before his audience" (Longinus, De Sublimitate 15.1).11 It was explicitly both a vision seen through the mind's eye which hail been evoked and communicated in language and a mental vision which in its turn gave rise to language."'' In the context of descriptions of art such as the Imagines, phantasia was the vision which gave rise to ekphrasis as well as being the vision which ckphrasis communicated to those who listened. Through its essential relation with phantasia, the very practice of ekphrasis in itself implied a theory of art. A work of art aroused in the mind a vision, phantasia, which in its turn gave rise to an utterance, ekphrasis. The person who heard that ckphrasis received through it the vision, the phantasia, which the work of art had originally inspired in the speaker. Moreover, phantasia was the vision which the original artist had had when creating the work of art.1 It was the quality bv which an artist was inspired to create in art even such things as cannot be seen vv ith the eyes. As the author of the Life of Apollonius ofTyana (also called Philostratus, but almost certainly a different man from our sophist) wrote: Phantasia "is wiser than mimesis. For imitation will represent that which can be seen with the eves while phantasia will represent that which cannot, for the latter proceeds with reality as its basis"."* This quotation is tantalising indeed. It suggests that in 26 the kind of thought influenced by Stoic philosophy - that is, viewing and "thl the kind of thought to which ekphrasis is heavily indebted - rf.ai.'' pbantasia (the vision of what cannot be seen with the eyes) has access to a truer reality than mimesis (the mere imitation of what can be seen). What matters is not so much the artistic form or the verbal utterance, but rather the phantasia itself, the creative vision in the mind's eye, which gave rise to the painting, sculpture or description. Hut the great value of a work of art or an utterance is that it can convey again to viewers and listeners the original phantasia which gave it birth. The theory of phantasia begs fundamental questions al>out what is ultimately real, for it regards the natural world (which can be perceived through the senses and represented by mimesis) as something graspable through mere imitation, whereas it points to an intelligible world beyond the senses as ultimately more real. It is to this intelligible world which cannot be seen with the eves that phantasia, which "proceeds with reality as its basis", gives access. So ekphrasis, through the artist's phantasia which gave it birth and the listener's phantasia to which it in turn gives rise, tells the truth. It offers access to reality. This is a remarkable claim - much more elevated than the job description most art historians or critics would give of their trade. But it helps us to understand the place of ancient commentators on art like Philostratus or (iebes. They were offering more than the mere description of objects; the) were educating their audience into truth. Their descriptions were not simply parasitic on pictures; rather, they competed with the pictures. Their descriptions were, like pictures, the creative product ofphantasia itself and works of art in their own right. This has formidable implications for our reading of ancient texts on art: These descriptions were not seen as dependent on prior images (as a modern art historian's description would be); thev were independent and self-sufficient works of rhetorical art in their own right. They existed and so thev were true -true to the pbantasia that brought them into being. Kkphrasis is therefore always concerned with Truth. It is true, however, not to the material reality of a particular image or sculpture, but to the phantasia which the speaker experienced in seeing the work of art and which he experiences in delivering his description. In other words, reading ancient ckphrases is not a good way of exploring what particular paintings or sculptures were actually like for the reason that ekphrasis competes with the actual image or sculpture in attempting to evoke phantasia. But ckphrases are a superb 27 WAYS OF \n wing source for examining how Greeks and Romans looked at art, because ekphrasis is essentially concerned with evoking the vividness of vision, that is, w ith evoking how art is viewed. This is why it does not matter in the slightest whether or not the paintings which appear in Philostratus actually existed. On the contrary, what matters, and what Philostratus's descriptions are constantly telling us, is how such paintings were viewed. philostratus and the psychology of realism With nothing can one approach a work of art so little as with . . . words. Rainer Maria Rilke" Only with a sense of what might in the third century be a normal expectation of ekphrasis can we move to the text of Philostratus. Otherwise, we shall always be in the boat with those critics who profess themselves amazed at "completely gratuitous additions". In discussing the hair of the horses in Imagines 1.8, B. P. Reardon, for instance, suggests that this is merely bravado erudition - "details not at all necessary tor the interpretation of the picture, but well designed to show off a knowledge of Homer".-" But he has missed the point. What Philostratus is expected to do is precisely to expand his discourse beyond any simple description, to interpret and elaborate in order to present his reader with a viewing.2* It is revealing that the same problems of critical expectation arise in the reading of Renaissance ekphrasis. Svetlana Alpcrs asks of the ekphrases in Yasari's lives of the Artists: "How can we then explain both the psychological and narrative interest of the descriptions?" She concludes that "looking at art and describing what he saw legitimately involved for \ asari what we today might think of as 'reading in' If ekphrasis is not a simple description at any stage in the historv of the literature on art, but is always an interpretation, a "visualisation", what is its purpose? Of course, Philostratus sets himself up in the introduction as an educator of the young, an interpreter of images, a sophist (i. introduction^!.). In this he accords with the view that what most sophists did most of the time was teach the art of rhetorical declamation to voting men.23 However, the Imagines are not a lesson in rhetorical declamation. They are self-dcclaredly (although this is something not picked up by the modern 28 commentators) "addresses which we have composed for the viewing and "the young, that by this means they may learn to interpret paintings real*' anil to appreciate what is esteemed in them" (////. i .introduction.3; emphasis added). G. Anderson does not sav enough when he comments that "Philostratus also plays the role of the Sophos instructing the child". Alpers is much closer in her perception that "ekphrasis concerns the viewer's education" (my emphasis). Philostratus is providing training in how to look.24 Philostratian viewing is not a subversive or alternative kind of looking at art - as tor example was the project in John Berger's Ways of Seeing. On the contrary, from the evidence of comparisons with other collections of picture descriptions, such as by the Younger Philostratus or Callistratus, what the Imagines offer is a "right and proper" appreciation of art, a technique of viewing that all young men of good family ought to possess. This is a normative text about how to naturalise the "other" of art into one's social and ideological context. Again, what Alpers says of V asari is as true of Philostratus: "This kind of ekphrasis, with its narrative emphasis, was beneath rationalisation, for it represented the normal way in which art was then seen and described" (p. 196). What one could learn from Philostratus w as how cultured people looked at paintings. The Imagines are strategies of how to view ; they are, in effect, a (culturally acceptable) ideology of viewing. At the same time we must always bear in mind that Philostratus offers us the rhetor's view. He is the sophist selling his learning, the interpreter representing himself and his listeners as be wants them to be seen. We cannot read the text as it w .i^ received, but only as it presents itself. Thus the normative nature of the Imagines is a self-image, a sales pitch. Nevertheless, I tend to think that salesmen (such as sophists in the Roman Empire) are fairly good guides as to what would prove acceptable, saleable, from among their wares. Later we shall turn from the interpreter's line to views of art which arc presented as lessons learnt from an exegete; but at this stage it is worth stressing that the verv space of ekphrastic discourse - of a speaker, a hearer and an object described (w hether real or imagined) - places great power in the hands of the sophist.:> As the Imagines consistently demonstrate, it is the speaker who provides the point of entry, the angle, into the interpretation of a work of art. Folks who seek surety while looking at art reach for collateral reading. They are following a millennial tradition exemplified with marvellous candour in that classic of literary description, the Itnag- ways of viewing "Ies °f Philostratus. Here, in his first ekphrasis. the preceptor teaehes his pupil that the way to master pictorial meaning is to keep eves averted and fixed on a text.'' So says Leo Steinberg. I le is right. The very first instruction in the first description is to "turn your eyes away from the painting and look onlv at the events on which it is based" lin. i. i (i ).-' But the first lesson is not quite to look for a text, although it comes down to a text in this case; it is, rather, to look for a context ("the events on which it is based").28 The sophist is offering his pupil something to hold on to so as to make sense of the Otherwise unassiinilable "other", the confusing ami disconcerting image presented to the viewer as in: "I lave you noticed, my boy, that the painting here is based on I Iomer, or have yon failed to do so because you are lost in wonder as to bow in the world the fire could Free m the midst of the water? (Int. i. 1111; emphasis added). What Steinberg elegantly calls "the Philostratian anxiety to adduce a text" is actually the utterly normal need of the viewer to naturalise the "other", bv finding a context in which the work of art will be part of the viewer's own reality. The first stratcgv of viewing is that the cultured beholder cannot be "lost in wonder" but must alwavs have recourse to a contextualising hermeneutic tool. The leap to Homer, however, is also the transformation of Homer.29 At every stage Philostratus modifies the Iliad's narrative of the wrath of the river Scamander against Achilles, lie ignores the symbolism of the river's anger and the anger of Achilles, which is important to Homer's narrative as a whole. He introduces Hephaestus as a running figure - an actor made concrete in the image - as opposed to I Iomer, who presents I lephaestus as the cause for conflagration rather than a concrete participant in battle. Both painting and the ekphrasis of painting must make concrete, make explicit, what was implicit in the Homeric text. In actualising, in substantiating, in embodying the implications of a story, art and its ekphrasis conduct a wholesale transformation of the original narrative. The result is something quite different from Steinberg's turn to a text. It is a new text, created from a conflation of the Homeric narrative, of the painting's transformation of that narrative by freezing its sequence and making concrete its implications and of the viewer-sophist's reworking of the image in his own terms and context.,n In terms of the stratcgv of viewing, the painting described alludes to a literarv context which Philostratus seizes and rearranges so that the picture can become its centre. The 3" context belonged to Philostratus already - all the sophists viewing and "ihi knew Homer, and this description quotes him at length real" (which both proves Philostratus's erudition, his fitness as a teacher, and fulfils the prescription of Hermogcnes as to the appropriate style in which to cast the rhetoric). Now, when the image is conte.xtualised, the painted "other" also belongs to Philostratus and to his pupils by its appropriation of and into the I lomeric context. The Philostratian strategv of viewing might be described as using any available means to con-te.xtualise the image and therefore to appropriate its "otherness" into the viewer's own private world. It does not much matter if the artist intended his picture to be of Scamander or not. If he did, then the viewer is only playing out a game within the artist's intentions. If Philostratus has got his interpretation wrong (as has been suggested of the description at 1m. 1.2 - which some critics think should be Hvmenaeus rather than Comus") it makes no difference. What matters is not the rightness of interpretation, but the viewer's need for contextualising the external "other" and a strategy of contex-tualisation to employ. Essential to the Philostratian strategv of viewing is the need to interpret the image into a context. The second description, not this time of an epic theme, illustrates this well: The spirit Comus, to whom men owe their revelling (komazeiii), is stationed at the doors of a chamber - golden doors I think they are; but to make them out is a slow matter, for the time is supposed to be at night. Vet night is not represented as a person, but rather it is suggested by what is going on . . .{hn. i . 2| 11) The subject is at once identified, as is almost always the case throughout the Imagines, and it is at once made relevant - this is not just Comus, but that Comus who gave men (gave us) revelling. It is night - not simply because of what is represented but because the sophist's own discourse is enacting night ("golden doors 1 think they are; but to make them out is a slow matter, for the time is supposed to be at night"). Finding the context is not merely a matter of identifying a subject and the placing of that subject where every beholder can relate to it (in this case, alluding to the literarv topos of the revel and the "real" experience of revelling). It is also the ekphrastic performance itself - the drama whereby the rhetor can act his own viewing and thereby his viewers into the image. The rhetor-observer is constructed as a narrator. He is no simple describcr or even interpreter, but rather has become ways of viewing inventor, author of his own story- In the same way, Ovid's Pygmalion is not only the viewer and lover of his ivory statue - he is also its creator. Both Ovid and Philostratus see the act of realist viewing as a creative one, a narrative of self-deception told to himself by the viewer and imposed through his phantasy on the image. *: The ekphrasis, as all the descriptions in the Imagines, has become a voveurist identifying phantasy where the viewer's experience can be extended into a narrative created by the image. Comus is asleep "under the influence of drink"; his left leg is bent towards the right "for fear lest the flames of the torch come too near his leg"; his right hand is at a distance "that he may avoid the breath of the torch". Philostratus's "reading in" is a strategy of psychological motivation based on "real life" experience and literarv tropes (w hich amount to the same thing - it is through topoi and cliches that we construct and classify "real life"). There is, moreover, a persistent praise of mimesis or of how well the painting has done to approximate to the viewer's (superior) world. The discussion of the relaxed hand moves from painted image to generalization about life - the hand is limp "as is usual at the beginning of slumber, when sleep gently invites us and the mind passes over into forgetfulness of its thoughts". The delicacy and tenderness of the crown are praised, and the roses praised for being a painting of fragrance itself. One should not forget here the pun that underlies the text - thatgrapbein which is (Jreek for "to paint" means simultaneously "to write"." In Philostratus anything that is a painting is also a writing . . . But the point of the praise of mimesis is to Hatter the viewer - to persuade him that the image is straining to reach out of its "otherness" into his reality. And yet, the fact is that it is the viewer who needs an interpreter (cf. 1. introduction. 3) to make sense of the image, and not the painting itself. The aim of this rhetoric is persuasion,'4 but it is not the persuasion that convinces us of a fact. Rather, Philostratus wants to establish a relationship. The viewer is being persuaded that he is in the position of power, that he has control (through the agents of the rhetor-sophist or the techniques of interpretation learned from the sophist) of the "otherness" of art. 'This control - represented precisely by the hermeneutic enterprise, by the ability to "read in" - is entirely illusory. It has no basis other than in the complicity of interpreter and viewers in the authority of their own viewing. The logical next step from Philostratus's contextualising of Comus within a generalization of "real life" (that is us), is to transgress the 32 boundary altogether: "What else is left of the revel? Well, viewing and "the what but the revellers? Do you not hear the castanets and the real" flute's shrill note and the disorderly singing?" (1.2.[5]). But what if we do not hear? The whole project of realist art and the viewing it engenders is a transgression of boundaries that depends entirely on our suspension of disbelief. The real revel that is taking place is the orgy of voveurist phantasy narrative - which the viewer keeps up in order to maintain the desperate illusion that he has & relationship with that thing out there. There are several moments in the Imagines when the sophist's viewing narrates away the boundaries of observer and observed, and there seems to be a union of realities between beholder and image. These are the moments when the text comes closest to its own deconstruction, when the premises of deception and illusion as realism - which underlie the entire realist enterprise (whether as image or ekphrastic text) - come closest to exposure. Perhaps the most potent example is the "I Iuntcrs" scene (1.28).55 I quote the first part of this (section 1-2) at length: Do not rush past us, ye hunters, nor urge on vour steeds till we can track down what your purpose is and what the game is you are hunting. Tor you claim to be pursuing a tierce wild hoar, and 1 see the devastation wrought In the creature - it has burrowed under the olive trees, cut down the vines, and lias left neither fig tree nor apple tree or apple branch, but has torn them all out of the earth, partly by digging them up and partly by hurling itself upon them, and partly by rubbing against them. I see the creature, its mane bristling, its eves Hashing tire, and it is gnashing its tusks at you, brave youths; for such wild animals are quick to hear the hunter's din from a very great distance. But my own opinion is that, as you were hunting the beauty of yonder youth, you have been captured by him and are eager to run into danger tor him. For why so near? Why do you touch him? Why have you turned towards him? Whv do you jostle each other vv ith your horses: I low 1 have been deceived! I was deluded bv the painting into thinking that the figures were not painted but were real beings, moving and loving - at any rate I shout at them as though the) could hear and I imagine that I hear some response - and vou [addressed to the listener or reader] did not utter a single word to turn me back from my mistake, being as overcome as 1 was and unable to free yourself from the deception and stupefaction induced by it. So let us look at the details of the painting; for it really is a painting before which we stand. This remarkable passage begins by pole-vaulting us, whether viewers or readers, directly into the image - with Philostratus actually addressing the hunters in the painting: "Do not rush 33 ways OF VIEWING past us, ve hunters, nor urge on your steeds till we can track down what your purpose is and what the game is you are hunting". The rhetoric of the description is enacting the image it describes - we, the viewers, are also hunters "tracking down" (exiebneuein) our prey. Whereas the description raises many of the issues we have already touched upon - deception, the emphasis on reality, the use of cliches from common-sense experience ("such wild animals are quick to hear the hunter's din from a very great distance") - it is particularly explicit about its initial establishment of the viewers' reality as being the same as that of the image. The game of the text becomes extremely complex once this single reality for viewer and image has been established. The hunters claim (phate) to be pursuing a wild boar - but that boar is a text from Homer (chlouneii syn, Iliad 9.539). They hunt a boar that is also a quotation; we hunt an image (this ekphrasis) that is actually a text. But our quarry is their game - we both are hunting a Homeric text, but it is a text whose presence (its quotation here) spells, above all, its absence - for Philostratus (for all his erudite quotation) is not I lomer. The narrative becomes dramatic as the viewer surveys the "devastation" anil as the hunters close in on the boar. But even as we (hunters ami viewers) appear to have reached the goal of the chase, the very aim of the hunt is cast into doubt by the authority of the sophist-critic: My own opinion is that, as you were hunting the beauty of yonder youth, you have been captured by him and are eager to run into danger for him. For why so near? Why do you touch him? Whv have you turned towards him? Why do you jostle each other with your horses? The whole hunt, theirs and ours, was but a metonymy of sublimated desire! The pursuit is of neither text nor boar - it is the bora (the bloom of youth, the beauty) of yonder youth as object of desire. Nothing so establishes the authority of the critic as the ability to read concealed and deeper motivations through the deceptive surface of text or, for that matter, of life. Here Philostratus is the dominant authorial presence typical of so many nineteenth-century novels - except that he is not author, but merely viewer of an image which is trying so desperatelv to be real that it has persuaded us into itself. The very rhetoric of the description (all those questions directly addressed) is contrived to carry us into the image, to persuade us that we relate with it and that it is striving to relate with us. J4 As we reach the point where the hunters' deception is viewing and "the pierced and erotic desire revealed as the true aim of the hunt - rial'' that is, the point w here the deeper perception of the hunter-critic penetrates the "truth" of the painting - suddenly the authority of the viewer is revealed for the deception that it is: "How have I been deceived! I was deluded by the painting into thinking that the figures w ere not painted but were real beings, moving and loving . . ." The mimetic enterprise of realism, as presented by PhilostratUS, is to persuade the viewer that the painting's "other" world is in fact his own. But its verv achievement of that ijoal can onlv be the brutal revelation of the exclusion of the viewer - the fact that the narrative by which self can appropriate "other" is ever a (self-) deception. The very transgression of the boundary of image and the real is simultaneously the relentless reassertion of that boundary . But Philostratus is saved from the complete loss of her-QiencutJc authority. Even as the image slips firmly back into its status as "other", Philostratus addresses his reader as "you", maintaining the illusion of a single world shared between the writer and the reader of the text. The very phantasy that has fallen to pieces in its operation between viewer and painting is that on which (in its operation between reader and ekphrasis) the remaining pages of this description arc-based. By implicating the reader-viewer in his own self-deception, the sophist maintains not so much his own authority as interpreter, but rather the authority of (or at least the need for) a strategy of illusionist interpretation. Without it, we are excluded from the "otherness" of the image, that is, we are lost. With it, as this ekphrasis proves, we are no less excluded - but we have always phantasy, desire and the hermeneutic narrative (ekphrasis itself) to generate the illusion of a relationship ultimately doomed to fail. "Realist" art offers the v ievver no means of relationship other than the discredited ekphrastic strategy. The sophist's authority is dependent not on the success of his "reading," but on the fact that in the final analysis we need some - any - strategy for reading.w What can we learn from Philostratus's bravura performance? An ekphrasis is not onlv a complex literary text, it is also - by the prescription of the rhetors - the rhetorising of a view. And ideologically inscribed within such a rhetor-ised view must be the presuppositions of what viewing is to be. Ekphrasis gives us an ideology of viewing because such an ideology cannot but be what an ekphrasis is predicated upon. 35 WAYS OF VIEWING Philostratus is teaching, through his ckphrastic perfor- mance, an hermencutic means of relating to images. This means is the viewer's narration of himself (Philostratus*s audience are all male, as is typical of antiquity) into the reality of the image ("the other") by assimilating the image into the framework of his own subjective consciousness, his personal context. The beholder constructs the object into his subjectivity, makes the other - which previously had no place in his experience - a constituent of that unique and intimate set of objects by w hich he defines his identity. The "I lunters" scene ceases to be an image of men pursuing a boar and becomes a portrayal (both a subjective metaphor and more than a metaphor) of erotic desire. This is the viewer's own particular "meaning". The premise beneath this strategy is that the viewer is always apart from the object he views, is always excluded from the reality of the object. The strategy of the excluded viewer must be to construct the object out of its autonomous reality into his subjectivity. I lencc the hermencutic enterprise of ckphrasis - the excluded viewer must narrate, or describe or associate the image into the terms he knows, the discourse he uses. But there is a price to pay. The image is no longer itself - it is a subjective construct with a personal meaning for the beholder.' \s Philostratus discovers in the "1 lunters" description, that meaning need have no relation with the object itself. As the influential French psychoanalyst Jacques Imuran put it (his italics): "What I look at is never n-hat I iDtsb to see. And the relation . . . between the painter and the spectator, is a play, a play of trompe-Toeir.™ And yet it is clear that the very essence of realism is to entice the beholder, to persuade him that the image's reality is his own. Thus in 1.23(2): The painting has such regard tor realism (aletheia) that it even shows drops of Jew dripping from the flowers and a bee settling on the flowers - whether a real bee has been deceived by the painted flowers or whether we are to be deceived into thinking that a painted bee is real, I do not know . \\ hatcver happens, realism is dependent on deception as the image strives to do its best to elide the unbridgeable gap between life and art. Mimesis is always illusion. In the gap between viewer and object (between the bee and the flowers or us and the bee), there is ultimately Lacanian "alienation", "the lack". In the gap between the image and what the image is of (as Narcissus, in 1.23, who is in love with his own image 3° and who cannot "sec through the artifice (sophisma) of the viewing and "the pool", cannot perceive that the likeness will never be its refer- real" ent) there is likewise only Lacanian despair. The text brilliantly dramatises the tensions of the "realism" in the art it describes. Earlier in this chapter, 1 argued that ekphrasis is inevitably concerned with the problem of what is real. On the one hand, ekphrasis purports to describe material objects (such as the paintings in the Imagines), but, on the other, it evokes phantasia, which has access to a deeper truth or reality than material objects have - an intelligible rather than a sensible reality. The genius of Philostratus is that he uses this problematic -the tension in the realitv to which ekphrasis refers - to dramatise the problematic of realism. His collection of ckphrases serves not only to describe particular pictures but also, more significantly, to represent the complexities of the realism with which those pictures are painted. Like all ancient ekphrasis, his subject is only apparently actual works of art susceptible to the senses; in fact, he is evoking the phantasia of a conceptual (rather than merely a sensual) problem - namely. What is realism? Even the devices of his rhetoric, as taught by the rhetorical text-book writers like Hermogencs of Tarsus, continually probe the issue of realism. For instance, enargcia. or "visibility", one of the crucial constituents of ekphrasis (something insisted upon by Aphthonius. Ilermogcnes and Theon among the rhetorical writers), renders with equal colour and evidence the face of real things and imaginary things. n It is precisely the play of real and imaginary that lies at the heart of the "realist" art that Philostratus describes and at the heart of the games with deception and realitv that the text of the Imagines displays. Indeed, one might say that the very complexity of these descriptions is dramatising the complexity of the art they re-present. Moreover, the rhetorical devices self-consciously used by Philostratus within his ckphrases in order to evoke cuargeia and phantasia in his readers, are extended by him beyond the descriptions themselves to the frame in which he presents them. It has been shown that, within the ckphrases, one of Philostratus's favourite rhetorical devices is to |>ortray a single and perhaps incongruous individual within a crowd or group against which he can be dramatically contrasted.4" However, in a brilliant move which radically disturbs one's sense of where the real begins and ends in the context of the Imagines, Philostratus employs this technique in the introduction to his ways of YiK\viN(, descriptions. At section 4 of the prologue, he tells us the occasion for his logoi (discourses) - the games in Naples, the resplendent villa of his host, the son of his host "quite a young boy, only ten years old but already an ardent listener and eager to learn". The sophist is importuned In young men and by this still younger boy to speak in praise of the paintings: 'Aery well", I said, "we will make them the subject of a discourse (epiileixin) as soon as the young men come." And when they came, I said, "Let me put the boy in front and address to him my effort at interpretation (sponde ton logon); but do you follow . not onlv listening but also asking questions if anything 1 sav is not clear. And immediately we are launched into the first description: "Have you noticed, my boy . . ." The very discourse in which so many of the ekphrases are cast (the posing of the individual against the group) is the discourse which sets up the reality in which the pictures are purported to exist and in which these descriptions are claimed to have been delivered, lor the boy and the young men arc nothing other than a rhetorical device of ekphrasis. The only "reality" of this text (pace those who believe there actually was a Neapolitan gallery) is constructed out of its rhetorical nature as text. It is the rhetorical performance itself (in the form of the contrast of individual and group, of young men and younger boy) that sets up the didactic basis of the text. Without the prescription presented by the prologue, that these ekphrases are an act of education, we would not know what the rest of the text was doing.41 What can we make of it, when ekphrastic method has become the means for representing "life" outside and around the images and the ekphrases themselves? What can we make of it, when the same method (and nothing else) is responsible for setting up the programme and the validity of a text w hose-parts are but a display of that method? Like the circularin of the Narcissus description (1.23) - where "the pool paints Narcissus and the painting paints the pool and the whole storv of Narcissus" - we are in a circle (or in a series of reflecting mirrors) in which the text's notion of "reality" is caught. We can read in Philostratus both a methodologv of appropriating the image by the viewer (the contextuali/.ing narrative) and a means ("realism") by which the image can entice its beholder. There is a reciprocity. The viewer constructs the work of art into his own world. But the reverse is also 3« true. The work of art, by its very existence as an "other" that viewing and "thf. demands to be assimilated, constructs the viewer as assimila- real" tor, forcing the beholder to take up the role of interpreting the object, despite even the interpreter's defeat at the failure of his own hermeneutic in the "Hunters" passage quoted earlier ("So let us look at the details of the painting: for it really is a painting before which we stand"). Just as the viewer constructs the object and thereby makes it other than it was, so the object constructs the viewer as assimilator and thereby makes him other than he w as. By an irony that our sophist would have relished, what we learn from the text of Philostratus is the power of a work of art. It is unassimilable without a context - and it generates that context from the beholder as a descriptive or narrative contextualisation within the beholder's subjectivity. The viewer himself (qua viewer) is generated by the object as the producer of such a subjective construction. The role of the artist as author is that (by choice of subject and form) he forecloses the potentially infinite number of subjective contex-tualisations that a viewer might choose. The implication is that weaning in the visual arts lies not in specific significations but rather in the types of relationships and assimilative strategies that different kinds of art generate from their viewers in different contexts. cebes and the psychology of allegorical initiation // IS the spectator, not lift', thai art really mirrors. Oscar Wilde, The Picture of Dorian (irey. Preface The Imagines of Philostratus offer two elements in their exposition of viewing. The first frames the text and presents an authoritative interpreter who provides viewers with a means of narrating themselves into the reality of an image. I he-viewer appropriates art's otherness by inventing a contextual ising storv. This frame applies not only to Philostratus but also to a number of other ancient presentations of art.42 The second element, this time particular to Philostratus, emphasises realism ami the way art itself appears to be striving to imitate the viewer's world. This is a brilliant strategy brilliantly displayed - and is perhaps the classic description of realism; but it is not the only method available to viewers in 39 ways of viewing their efforts to make sense of images. If art mirrors the spectator and not life, then the reflections that images offer will depend on the initial self-image of the beholder. 'When I was hunting in Lesbos, I saw in the grove of the nymphs a spectacle the most beauteous and pleasing of any that ever yet I cast my eyes upon. It was a painted picture (eikona graptin) reporting a history of love". Thus begins the proem of Longus's Dapbnis and Cbloe. This novel, dating from the late-second or early-third century,4' purports to be an ekphrasis - one of the longest extended descriptions of painting that have come down to us from antiquity.44 In the prologue Longus authorises his narrative as being more than merely fiction or a subjective phantasy about a work of art not by relying on his own view but by turning to an interpreter (extgeten taeikonos, Proem 2).4v Whereas the Imagines of Philostratus are presented as an exegete's lesson in interpretation, the roughly contemporary narrative of Dapbnis and Chine anchors its validity in just such an exegesis by appealing to the authority of a guide with superior knowledge. Here Longus is not alone. In his celebrated description of the "Calumny of Apclles", the second-century essavist Lucian of Samosata also turns to a guide (periegetes tes eikonas).*6 Because this painting is unknown from any other source and because Lucian hopeless!) confuses his facts in the anecdote accompanying the description, it has been suggested that he was "misled bv an ignorant cicerone".4 However, what matters for our purposes is not the correctness of attribution or context but rather the need for them and the need to authorise them in an external interpreter. It may be that Lucian and Longus have invented the paintings they claim to describe;4" certainly they are more interested in the narratives purporting to be descriptions than in the images which give rise to them. Nevertheless, the figure of an interpreter provides a crucial touchstone to vcritv not only for the existence of the paintings but, more importantly still, for the correctness of the interpretations which these authors provide.49 The most spectacular and complex example of such valorization through the authority of an interpreter is the Tabula of Cebes. This text, which in all likelihood was written during the first century a.i>.,>" is a religious-philosophical interpretation of a picture igraphe) set into a votive tablet (pinax) in a temple of Kronos (Tabula i.i).51 It has received virtually no literary or art-historical discussion, despite its elegance as a literary construct and its importance as ancient evidence for the allegorical interpretation of religious an.-1 In fact, only 4'» two monographs have ever been devoted to it. and these - like viewing and "the most of the scholarly discussion - have focussed primarily on real" issues of dating and on whether the exegesis of the picture is primarily Pythagorean, Kleatic, Socratic, Platonic. Stoic. Cynic or eclectic/' The Tabula begins with the narrator and a group of friends strolling through the temple. They discover the picture and are "unable to make out what its meaning could possibly be" (i. i). "What w as depicted seemed to us to be neither a w ailed city nor a military camp" (1.2). The viewers' check-list of the known and familiar fails rapidly before the enigma of the picture, and description collapses into obscurity - "a circular enclosure (peribo/os) having within itself two other circular enclosures one larger and one smaller" (1.2). Having dramatised the perplexitv of the uninitiated who approach such images without esoteric knowledge,"4 the text brings us an old man {presbytia tin) as the rescuing interpreter (2.1). However, thepresbytes does not take upon himself the authority for revealing the truth, but explains that the tablet was put up in his youth by a foreigner of rare wisdom and sense (eiitpbron kai deinosperisopbian), both a Pythagorean and a Parmenidean, who was himself then a very old man (polychroniotaton) (2.1-3). It was this man, the philosopher-dedicator of the tablet, who expounded the meaning of the image to the old man now confronting the narrator. As all mystic and religious interpretations must, the Tabula validates itself by appealing to a detailed hierarchy of authority and an oral tradition of secret exegesis. Even "manv of the local inhabitants do not know what the fable could possibly mean" (2.2). The keenness of the narrator (and by implication of us as readers of his narration) to penetrate the mystery anil hear the meaning is met with a warning: "Exegesis carries with it an element of danger" (3.1). "What sort of danger?" I asked. "Just this." saiil he, "if you pay attention ami understand what is said, you w ill be wise and happy. If on the other hand you do not, you will become foolish, unhappy, sullen, and stupid, and you will fare badly in life, ... It one does understand. Foolishness (apbrn-syne) is destroyed and one is saved (sozetai) and is blessed and happy in one's whole life". (3.1-4) The presbytes raises the stakes from explanation as the satisfaction of a casual interest to exegesis as existential solution to the problems of life, exegesis as salvation." In comparing his explanation to the riddle of the Sphinx (3.2-3), the old man 4' ways of viewing conducts a remarkable reversal. Instead of the viewer having control or power over the "other" - an inert picture - it turns out that the viewer's own life forever after depends Upon this moment of viewing, depends upon a correct understanding of the picture. The image can teach: To be taught is to be saved, but to miss the teaching is to be forever lost. Power here is the prerogative of the "other", with the prtsbytes as its mediating exegete. We, the viewers within the Tabula or its readers, are unexpectedly confronted with the unrelenting demands of a religious choice. This challenge leads to great desire on the part of the viewers (megalen cpitbymian) to learn the secret of the image, a desire explicitly related to the greatness of the penalty (toiou-ton epitimion) if the meaning is not understood (4.2). Bui as the exegesis begins, the text self-consciously plunges us into a further and brilliant reversal: So, taking a staff, he [the prtsbytes] pointed towards the picture and said "do you see this enclosure?" "We do". "You must know, first of all. that this place is called Life, and the large crowd standing at the gate consists of those who are about to enter Life. The old man (geron) standing up here - who has a scroll in one hand and who appears to be pointing at something with the other - is called Daimon. To those who are entering he prescribes what thev must do upon entering into Life; he shows them what kind of path they must take if they are to be saved (sdzestbai)". (4.2-3) The picture is re-enacting the frame which has just introduced it. As the prtsbytes points to the picture in order to enlighten his neophyte audience - the visitors to the temple and us the readers of their account - so, within the image, thegerSn (also an old man) points out the path of salvation to those about to enter the enclosure of Life. But we have already been told that a correct reading of this picture is precisely a lesson in salvation (the same word is used: sozestbai). What we are doing at this moment in reading the exegesis of the presbytes, what the viewers are doing in hearing the same explanation, is nothing other than an enactment of the vcrv situation which the image itself depicts. And in terms ol a theory of salvation and happiness in one's future life, such as the presbytes just announced in his preamble to presenting the interpretation, we (the readers and the narrators) arc precisely in the position of the large crowd about to enter Life. We too arc about to be offered a transformative initiation in the right path to follow. Just as the rhetoric of Philostratus enacted a remarkable homology between the frame of the text and the viewing and "the descriptions themselves, so here the picture mirrors its liter- real" ary context and the frame turns out to be enacting the prescriptions of the picture it purports to describe. This is explicit at 30.1—2: I said, "You have not yet made clear to us (his point, namely what the Daimon commands those entering I -ifc to do". "To he confident", he said, "Therefore, you also he confident, for I will explain even tiling and omit nothing**. 'Vhegeron in the picture - now elevated to deity as Daimon -himself becomes the mystical authority for what should be happening outside the picture to its viewers and to the readers of its ekphrasis. To look at the image on the tablet and to understand it, is actually to fulfil its own prescription about the path from deceit (apate) and ignorance (aguoia) (5.2-3) to salvation (to sdzestbai). The effect of the frame is for the picture to prescribe itself as the solution to the existential problem which its own enigmatic obscurity first posed and which its exegesis foregrounds! The key to salvation is viewing and correctly understanding what one has looked at. Such understanding is of course not different from the act of exegesis itself. The very act of reading the Tabula and of following the interpretation as it leads us deeper into the mystery of the picture is itself an initiation into the true path - a truth whose implications are not confined to the interpretation of art but become the means of transforming the viewers' life. The text frequently reminds us of the picture's homology with its narrative frame. At 6.2-3, tnc members of the crowd within the tablet meet a number of women called "Opinions, Desires and Pleasures" who lead them off. All promise happiness, but some lead to destruction (to apolusthai), while others lead to being saved. Like the narrator and his readers confronted in the proem by the perplexity of the image and the riddle of the Sphinx (3.2-4), the crowd in the picture need a guide whose higher knowledge can demonstrate the "true way in life" (6.3). If they miss the truth they wander about aimlessly, "even as you see now those who have entered previouslv wander about wherever they chance to go" (6.3). The word for chance here (tyche) leads into the personification of Fortune (Tyche), blind and mad, who is herself the cause of much of life's miserv (7-9). Again we arc thrown out of the picture into its context - the first words of the Tabula are etugebanamen (from the root tyche) peripatountes, "we chanced 43 ways of viewing to be wandering". The aimless state described in the image where Chance seduces man off the right path turns out to be precisely our state, the narrator's and readers' state, before our initiation by xhc presbytes into the meaning of this picture. It was aimless chance that brought the viewer (as well as the reader) to this temple, to this image, to this text through which we are taught to see the crowd, the image of ourselves in the perplexity of aimless chance, and wc can perceive through this lesson how to walk the right path to salvation. The detailed and consistent collapsing of the themes of the picture into the narrative of the frame is an extremely effective device for highlighting the significance of the image and the exegesis that describes it. The image and text arc presented as a commentary on life: The explanation of the image becomes a commentary not only on the narrator's life but on our lives (those of the readers) and an allegory of life in general. The effect of this is to emphasise the act of viewing. Viewing - the right view religiously, the right ethical interpretation, the right path to salvation - becomes the real theme of the Tabula. Like the Imagines of Philostratus, Cebes's text is a lesson in how to view . Unlike the Imagines, it is not a lesson in realism or in how art may imitate the commonsense world. On the contrary, it offers a picture of how sacred art, by presenting a True Reality (a mystic concept into which one must be initiated), transforms the commonsense world -the world which we erroneously thought was real when we wandered aimlesslv in deception. Like Lucian's "Calumny of Apelles", the Tabula uses the method of personification to present its theme.'6 It paints a picture of progress to the figure of "true Education" {aleibinen paideian), a phrase repeated thirteen times.' However, for our purposes it is not the specific allegorical meanings of the personifications that matter, but rather the way in which Cebes uses the allegorical technique as a method of viewing. s We are in a world where nothing is what it seems. Thegtrdn at 4.3 is in fact Daimon; the women (gynaikas) at 16.1 turn out to be Self-Control and Perseverance; the group of women (cboron gynaikdn) at 20.1 are a veritable panoply of virtues (Knowledge. Courage, Justice, Goodness, Moderation, Propriety, Freedom, Self-Control and Gentleness) by 20.3, and so on. What you see on the surface becomes something deeper, truer and yet more general when you see it through the eye of the henneneus. This method of transformation - a rhetorical device mirroring a salvific religious claim - is shot through 44 the text. By the end, not only are all the figures in the picture viewing and "the personified into abstract qualities, not only is the viewer- REAL" reader ethically and spiritually transformed by being initiated into the picture's meaning, but even the picture itself has changed. That change is a transformation of depth: The short and perplexing description of i .2—3 has become the huge and extended symbol of life of 4.2 - 40.4 (some thirty pages in Praeehter's Teubner text). It is a transmutation of content from obscurity to clarity, from ignorance to insight, from the "other" debarred to the "other" known and penetrated. But perhaps most remarkably, the text of the Tabula suggests that the very tablet itself changes as a result of exegesis. At 1.2 the narrator describes a "circular enclosure having within itself two other circular enclosures, one larger and one smaller". I low ever, the three enclosures apparent to the uninitiated viewer become four enclosures by the time the exegesis has reached 17.2.59 Form itself is transformed in the mysterious act ofexegetic viewing. The 'Tabula of Cebes, for all its rhetorical parallels with Philostratus and its formal similarity with Dap/.vns and ('bloc (both have a picture as frontispiece and Longus's novel has also been, if rather speculatively, interpreted as a religious allegory),60 represents a remarkably different kind of viewing. Like the interpreter in the ancient art of physiognomies, astrology or dream-interpretation/'1 theprabytcs has a field day in producing entirely unprecedented interpretations of an image which could not Otherwise be understood. That readers of the text would never even see the picture which is transmuted with such virtuosity beneath the alchemy of exegesis only adds to the effect. The authority for this reading lies in the hierarchy of wise men suggested by the text, in the fact that the narrative frame enacts the injunctions of the picture it describes and in the great claim (to ultimate salvation) which the 'Tabula unexpectedly springs on its reader as the raison d'etre for explanation. This justifies the extraordinary t\ ranny of the exegete's line which is final, open to no question and must be accepted (for the cost of doubt is misunderstanding and hence perdition). The viewing which is prof-ferred as the path to enlightenment is radically different from that of Philostratus. It is polysemic, for every figure has at least two meanings - its appearance and its allegorical transformation to something else (something truer) through the exegete's words. Above all, reality in this text does not relate to the viewer's ordinary or commonsense world (the world which Philostratus's pictures strive to imitate). 45 ways 01 viewing "Reality" here lies in another work! of allegorical exegesis - in a call to salvation and to a religious Truth. In antiquity this "reality" was as "real" for numbers of people as was the mundane reality of matter and psychological motivation explored by PhilostratUS. For those who accepted the injunction to allegorical viewing, the "Other World" that such viewing produced was more "real" than the material world to which non-initiates were confined. We know that, by contrast with many authors in antiquity, Cebes was widely read - by the pagans Lucian and Julius Pollux and probably by the Christian I ertullian (all active in the second century a.d.)/>: The Tabula was an influential, perhaps even a popular, text, translated in later times into Latin and Arabic and known in Byzantium/'' It prescribes an exegetic and religious method of viewing which, for all its unfamiliarity to us, gives important testimony for attitudes and approaches to art that were widespread in the ancient world. conclusion Representation and reality occupy one single space. Ludwig Wittgenstein. Philosophical Remarks, proposition $8 Viewing is apparently the simple act of pre-self-conscious l(k)king - virtually a physiological reflex - loaded with unra-tionaliscd assumptions and associations about life. Vet, when we investigate the Imagines of PhilostratUS and the Tabula of Cebcs, we find a serious determination in both these texts to train viewers in the art of how to look. In comparing these books, we discover a remarkable difference in the types of viewing which these texts presuppose. Both books provide authoritative voices which claim to educate their readers into the art of viewing. The hidden agenda behind each text is the "reality" to which it claims to refer; and it is revealing that both writers use the same Greek word, aletheia or its cognates/'4 to denote "the real", despite the fact that the authors mean fundamentally different things when using this word. But neither text prescribes an ideology or a theory about art to be bought wholesale: Lach educates its readers about reality. by training them through examples into a method of looking at art. These two texts do not offer (as modern art history does) a history or theory of how art has come to be as it is; rather, they offer (very different) strategies for the contextualization viewing and "the and assimilation of paintings by their viewers. In this sense, real" neither Philostratus nor Ccbcs seeks to legislate how one should take images. Instead, each offers his own persuasive set of paradigms indicating ideal ways of understanding art. In other words, both texts imply that viewing is creative and can be developed creatively by readers along the lines offered by these exemplary didactic texts. Clcarlv, viewing is not monolithic. There are as many ways of looking at art as there are viewers. Ways of viewing are themselves constructed by the prejudices and conunonsense assumptions that viewers bring with them when they confront images. However, at the heart of the divergence between the strategies of viewing in these two texts is a difference in different viewers' assumptions about what "reality" itself might ultimately be. For Philostratus, the "real world" which art imitates is the world of the beholders' physical and psychological experience - the world of our commonsense expectations. Art Subverts these expectations precisely by being so realistic, so successfully imitative of them, that we cannot (for a brief but dramatic moment) be sure where the line between art and life should be drawn. lor Ccbes, the "real world" (in its correlation with salvation, happiness and true Education) is precisely and constitutivcly not the world of everyday experience and phantasy (the world of Philostratus). On the contrary, Cebes's "reality", as the goal of true Education, lies in a wholesale rejection and transformation of commonsense expectations according to an initiate allegorical system. The goal of art in the Tabula is not to imitate the viewers' world at all, but rather to initiate viewers out of their ordinary assumptions into a new exegetic reality, a truth that brings salvation. In this kind of viewing, if art is imitative at all then it imitates that exegetic Truth and has as little as possible to do w ith the kind of reality explored by Philostratus. Quite simply, these texts present antithetical versions of "reality" as the goal of art. The kinds ofviewings they signal existed simultaneously. Cebes's allegorical exegesis is contemporary w ith the Elder Pliny's classic anecdotes about art as illusionism - Zeuxis's grapes which were so realistic that birds flew to peck at them (Sat. Hist. 35.65—6), Apelles's portraits whose likeness was so perfect that a physiognomist could tell how long the sitter had to live or had already lived (Nat. Hist. 35.88-0), or Apelles's horse which alone caused real horses to neigh when they saw it (Nat. Hist. 35.95). The realism of Philostratus coincides with Porphyry's remarkable 4" ways or viewing mid-third-centurv allegory of the Cave of the Nymphs based on Odyssey 13.102-12/" and with Iamhlichus's Xeoplatonic presentation of images (eikoues) as symbols within the occult philosophic system of theurgy.** Moreover, the very same image might be subjected to these radically different kinds of viewing (either at the same time bv different viewers or at different times), and might carry divergent and even contradictory meanings according to the spectator's initial presuppositions. The essential nature of the difference in viewing so dramatically embodied bv the texts of Philostratus and Cebes is hugely ini|x>rtant. The meanings viewers located in images depended on the kind of "reality" thev ultimately wanted those images to represent. Such meanings might be single or multiple, simple or complex, material, political, psychological, ethical or religious. The understanding ofmimesis itself must finally depend upon what space, what "realitv", one believes is being imitated. Wittgenstein observed that reality and representation occupy the same single space; what one understands this space to be is crucially dependent on the reality which different viewers presuppose when thev look at images. 4« VIEWING AND SOCIETY: IMAGES, THE VIEW AND THE ROMAN HOUSE This chapter addresses the theme of viewing in a domestic context. 1 examine the social experience of Roman housing on three levels: first, the state's prescriptions for housing and decoration; second, the material forms of some ancient houses surviving from Campania and third, the ways individuals viewed their houses. The interrelation of these three themes, the political, the material and the personal, can throw light on the collective subjectivity and tastes of the Roman middle class. As anthropologists have shown, buildings and their decor are a "cultural system". The sensibility of which they give evidence is "essentially a collective formation, and . . . the foundations of such a formation are as wide as social existence and as deep".1 The text of the Augustan architect Yitruvius illuminates the politics of style in Roman wall-painting and in particular the moral agenda belying the Augustan emphasis on "realism". Moreover, Yitruvius is explicit about the socio-political constraints which governed the structure of Roman housing and the way that housing was designed to enshrine the hierarchies of a class system. However, the architectural and decorative schemes of actual houses (dating from a little before Yitruvius) exhibit no imposed straight jacket, but rather a spectacular series of symbolic inversions. The complexity with w hich such houses visually deconstruct the decorative and social categories on which they were based hints at a remarkable taste for the transgressivc. Between the constraints of the state and the symbolic inversions of the decoration in his or her own home was situated the Roman viewer. One way in which the allusive images of Roman illusionist decoration could be assimilated in the beholder's subjectiv itv u \vs of viewing was through the shared cultural goal of desire, centred on an obsession with the vieii\ This concern with the view, attested by a number of ancient texts as well as actual houses surviving from Pompeii and I Ierculaneum, is related to the importance of the visual field in the Roman ars memoriae. Whereas the Vitruvian agenda betrays the Roman state's interest in the social hierarchv of housing, the desire for the view offers access to the wdividuaTs concerns in the house. Despite, his polemical emphasis on the "real" in decoration, \ itruvius himself recognised the need to adapt building to the needs of the view. In the face of the deconstructs c revelry of the Roman house, even the social prescriptions of the state had to compromise. Together, these two motifs - semiofficial prescription and private mentaiite - offer us a glimpse into the "deep play" of at least some aspects of Roman culture.2 The anthropologist Clifford Geertz has argued that "the central connection between art and collective life does not lie on . . . an instrumental plane, it lies on a semiotic one".- Interpretations of art in its cultural context should not be merely functional or stylistic ("instrumental") but should also examine the process by which art constructs and signifies its meaning. This "semiotic" process is inevitably related to other systems of constructing and signifying meaning w ithin the culture. Archaeology may sometimes wish to present the understanding of Roman housing as grounded in a scientific evaluation and classification of the remains of material culture. But it is as well to remind ourselves of the truism that any such understanding is necessarilv dependent on the fragments of an ancient conceptual frame which we rescue from the textual sources and use to interpret the material remains. Above all, in the field of domestic housing and decoration we are indebted to the very limited and cryptic remarks on the subject embedded in the writings of \ itruvius. This is not an expansive disquisition on housing, as the Imagines- of Rhilostratus are on paintings. Rather, it purports to be a handbook on architecture, in which housing and house decoration have a brief allotted place. The text of \ itruvius is not only limited; it is also - as I hope to show - highly political and polemical. We cannot use it as if it told an objective story or reported the factual truth. On the contrary, it tells an idealised anil rhetorical story, explicitly designed to win favour from Augustus. Tor this reason, the writings of \ itruvius are not only all the more difficult to use. but all the more interesting. \ itruvius's work shows the extent anil the depths to which a desire to 5° foster Augustan propaganda and moral policies dominated viewing and society even those who may not have been directly employed or paid l>\ the state. For Yitruvius, modern wall-painting was subversive. By examining some artistic and decorative programmes from Roman houses in Campania, and relating these decorations to the categories used in Roman texts for the social articulation of domestic architecture, we can see how this subversion worked. But it remains true that precisely those subversive or deconstructive qualities in Roman painting (against which Yitruvius inveighs) were licensed within the Principate's design. Wall-paintings did not simply question or undermine the socio-architcctural system envisaged by Vitruvius: The) were allowed to do so. Their "surrealism" was not outside the system, but was actually part of it. The emphasis on the view (not only in surviving Roman houses but also in texts which describe them, such as the letters of the Younger Pliny) enabled Roman viewers not so much to see the socio-architee-tural sv stem dismantled as to learn from this dismantlement precisely boii the system worked. Roman domestic viewing was an education. It taught the viewer how to be a subject -not only in the personal but also in the political sense. VITRUVIUS AM) THE POLITICS OF HOUSING affirming the "real": vitruvius and the politics of STYLE In one of the most influential outbursts of "violent abuse" in the history of art - influential because as a piece of "normative criticism" it set up the "concepts . . . and categories of corruption" which would later be borrowed by \ asari, Johann Winckelmann and the founders of modern art history - the Roman architect Yitruvius, writing in the last quarter of the first century B.C., made a passionate plea for realism in the decoration of Roman walls.4 The passage is l)c Architecture* 7.5. This outburst is highly significant. Wc can use it to assess not only what "realism" meant to Yitruvius (both aesthetically and as part of an ideological agenda) but also what the wall-paintings he castigates were implicitly denying. In the context of the social construction of the Roman house and the viewer within it, we must question what the ideal of realism implies about \ itruvius's view of art and of viewing. Before we focus on \ itruvius's invective, however, we 5' WAYS OF VIEWING must situate it within the rhetorical conventions of its time. Vitruvius was writing not an objective description but rather a loaded polemic. For Romans like Vitruvius, as for those Englishmen whose home is their castle, a house was more than a "machine tor living in". Just as the idea of "home" in English means a great deal more than "domestic situation", the metaphorical evocations of downs in Roman culture w ere rich and manifold. Vitruvius's attitude towards art - like the rhetorical attitudes of Romans towards their houses - is neither neutral nor disinterested but is involved in his relationship to his culture and, in particular, is part of an ideological and moral programme. Before picking apart some of the axioms upon which the programme is based, let us sketch its relationship toother, similar, invectives in Roman culture. \s early as PlautUS, a playwright at the turn of the third and second centuries b.C. whose works were already classic by the late Republic and early Empire, the image of the Roman house iucdes) served as a metaphor. In his play Mostcl-laria, the education and upbringing of a man is compared extensively with the building of a new house, and the decline of a spendthrift is paralleled with rotting timbers and negligent upkeep.1 lor example: Just look at me. I was a lad Of modest manners, blameless life, While they were building me. Left to my own devices - well. It didn't take me very long To undo all the builders' work And make the house a ruin. (Plautus, Moste/laria, vv. 133—6) This earlv metaphorical usage for the image of the house opened the possibility for the theme to be used in invective: by the late Republic and early August an periods, extravagant buildings had become a polemical theme svmbolising the corruption of individuals or of socictv as a whole.6 In Horace's verse, architecture emerged as a poetic metaphor, evoking the futility of opposing death and the unnatural extravagance of contemporary society. Most important for understanding Vitruvius is understanding that art and architecture were seen as unnatural impositions on the pure and essential state of humanity; they were modern innovations on the vctcrum norma, the righteous ways of previous generations. The early Romans never lived in a world of huge porticoes and fishponds larger than the Lucrine lake: 5 2 This is not the norm viewing and society Our ancestors divined, that Romulus And rough-bearded Cato prescribed. For them private wealth was small. The commonweal great. . . (I lorace, Oda 2. i 5, vv. 10-14) In the diatribe of Papirius Fabianus both paintings and the buildings they adorn (particularly those constructed into the sea) are aihersum naturam (opposed to nature), sustained by a deplorable fastidio rerun/ naturae (disdain for the ways of nature).8 By the first century A.D., extravagant building was seen as a rebellion against the natural order.7 The epitome of luxury,1" it is a deceit which disguises a man's vices from himself beneath the veneer of marble and gilding." It is the luxurious price at which the once free Romans have bought themselves into slavery.1"1 What is crucial here is the rhetorical assimilation of extravagance or excess in building to the contravening of what is natural, lint natura in this heavily moralistic rhetoric means a good deal metre than merely "the natural world": It stands as a moral term for the old way, the right and proper way, of doing things Yitruvius's attack on modern painting is a classic instance of the same trope. The ancients, we are told: used definite methods of painting definite things (constitutae sunt ab antiquis ex certis rebus certae rationes picturaruni). For by painting an image is made of what exists or what can exist (quod est sen potest esse): tor example, men, buildings and ships and other things from whose definite and actual structure (e qmbus finitis certiscpte corporibus) copies are taken and fashioned in their likeness. (7.5.1) Then, in one of those classic evolutionary passages from the ancient primitive past to the present so beloved of writers in the late Republic and early Kmpire," he offers a myth of the development of imitation to the representation of the forms ol buildings, columns, pediments, theatrical facades and landscapes:14 Hence the ancients who first used polished stucco, began In imitating the variety ami arrangement of marble inlay; then the varied distribution of festoons, ferns and coloured strips. Then they proceeded to imitate the contours of buildings, the outstanding projections of columns and gables; in open spaces like exedrae they designed scenery on a large scale in tragic, comic or satyric style; in covered promenades, because of the length of the walls, they used for ornament the varieties of landscape gardening, finding subjects in the characteristics of particular places. (7.5.1-2) 53 ways of viewing The movement to and achievement of a zenith in this progress of imitation (from ingressi sunt ut . . . imitarentur to pinguntur) is marked by the list: They paint harbours, headlands, shores, rivers, springs, straits, temples, groves, hills, cattle, shepherds. In places some have also the anatomy of statues, the images of the gods, or the rcpresintat ion of legends; further, the battles of Troy and the wanderings of I l \ sscs over the countryside with other subjects taken in like manner from nature (cetera . . . simHibus rationibns ab reritin natura pro-creata). (7.5.2) The importance of the list is that it asserts the variety and the breadth - the full splendour of the achievement of mimesis. The Vitruvian list is remarkably similar to that other great celebration of the range of realism, the list of wall-painting subjects associated with the Augustan painter Studius in the Elder Pliny." Hut it is to be a splendour already past; the evolution of painterly progress has even by Yitru-vius's own lifetime given way to the Fall: "But those which were imitations based upon reality (ex irris rebus) are now disdained by improper attitudes (iniquis tnoribusf (7-5-2).16 Before we turn to the depravities upon which \ itruvius's full moral censure will descend, a word on his criteria for the "real". Whereas the prescription quod est sen potest esse (what exists or can exist) seems to point to a modern materialist conception of reality, that \ itruvius has no doubts about including legends and heroes in "subjects taken from nature" shows that we must be wary of equating Roman notions of the "real" too closely with some of our own theories. Vitruvian "realism" (like the realism of Renaissance art) does not question the acceptability of what might be to us problematic "truths," such as myths. Rather (this time more rigidly than in Renaissance practice), \ itruvius wants his subjects to be depicted in a "realist" manner which allows them to appear as if they existed in the three-dimensional "real world". I lis attack on the corruption of modern taste is directed against the "monsters painted on the wall instead of truthful representations of definite things (ex rebus finitis imagines certae)" (7.5.3). The "monsters" turn out to be fantasies which could not exist according to the logic and the rules of the "real world", like images in Mannerist or Surrealist art: Candelabra uphold pictured shrines and above the summits of these, clusters of thin stalks rise from their roots in tendrils with little figures seated upon them at random. Again slender stalks with 54 heads of men and of animals attached to half the body. Such things viewing and society neither are, nor can be, nor have been . . . For how can a reed actually (ixft) sustain a rool or a candelabrum the ornaments of a gable? or a soft and slender stalk a seated statue? or how can flowers and half-statues rise alternately from roots and stalk? Yet when people view these falsehoods (falsa), they approve rather than condemn, failing to consider whether any of them can really occur or not (si quid eortini fieri potest necne). (7.5.3-4)17 This polemic is extreme. After all, no painted illusion could ever finally match Vitruvius's criteria for "what can really occur". In effect, Vitruvius is the first and perhaps most insistent apostle of a dogmatic reflexivity version of "realism", w here the history of art is to Ik seen as the rise to the achievement of imitative verisimilitude.18 Anything else, "however line and craftsmanlike" (sisuntfaetaeelegantesab arte, 7.5.4), is a symptom of the Decline and Fall. "For pictures cannot be approved which do not resemble reality" (neque enini picturae probari debent, quae non sunt similes veritati, 7.5.4). Just as his rhetoric of decline was a foundation for the rhetoric of corruption in the later history of art, so Vitruvius's myth of the ascent to realistic mimesis was the precursor of Pliny and all his followers from Vasari to Frnst Gombrich.19 But we must not forget that the decadence against which Vitruvius inveighs is actually the normal practice of his time. \ i-truvian realism and the myth of verisimilitude as the aim of art is a prescription and an ideology; it is not any kind of objective description of the visual evidence. The most revealing terms in the Vitruvian onslaught arc vere and Veritas (7.5.4). These mean much more than "lifelike", for they carry the whole weight, the value judgment, of "Truth" as an ethical and aesthetic norm. Vitruvius's attack is couched in the most censorious hyperl>ole of Roman moral rhetoric (iniquis moribus, "improper attitudes'*, and monstra, "monsters", 7.5.3) coupled with an appeal to the Veritas ("truth") values of the mos maiorutn, or traditions of the past. This is a dogmatic moral prescription presented as both a nostalgia for the imagined art of a lost past and a rejection of the art of the present. What, we are entitled to ask, is the purpose behind the outburst? What is Vitruvius's hidden agenda? The question matters because art historians' readings of Vitruvius and of Pliny after him have set the aesthetic criteria (the myth of the ascent and the decadence of verisimilitude) that have defined the discussion of art since the Renaissance and have been hugely influential on our own attitudes tow aids painting. 55 ways OF VIEWING We must begin by being clear about the status of Yitruv- ius's treatment of the Roman house within his whole project. For Vitruvius was working, as Paul /anker has so brilliantly shown, at a time when all imagery and building in both the public and private spheres was being marshalled to propagate the new Augustan ideology and the imperial myth.20 Vitruv-ius's architectural discourse is not an ideologically innocent manual for architects, but a highly partisan text-book for fulfilling the Augustan political project as it was directed to buildings. If Augustus could boast that he found Rome built in brick and left it built in marble (Suetonius, Augustus 28.3) and devote three chapters of his autobiography to the extent and splendour of his building activities in Rome (Res (iestae 19-21), then Vitruvius - both as architect and as writer - was part of the team which provided the know-how for the programme (De Arch. 1. praef.). He is quite explicit about the purpose of his book: It is "a detailed treatise" to help Augustus by ensuring that "public and private buildings (ptthli-corttni ct privatornm aedificorum) correspond to the grandeur of our history and will be a memorial to future ages" (1. praef. 3). Insofar as the Augustan project entailed the continuation of late Republican differentials of social class, Vitruvius was quite happy to tie that social programme to his own metier of building and to the architectural and decorative injunctions of his sixth and seventh books, which are explicitly dedicated to domestic housing. Indeed, there is no reason why he should ever have supposed that building need not enshrine the distinctions of social class and the preservation of a particular and at least in some respects a repressive system. The implications of "realism" as the aesthetic plank of this social and ideological project arc that every object (the taxonomy of the Vitruvian and Plinian lists) can be labelled and ordered according to the criterion of Veritas (7.5.4). The immense rhetorical and ethical authority of concepts such as "truth" and "nature" was pressed into service as the guarantee for the acceptability of art. Such concepts are a moralising and ideological justification for an art which, as "realism", was to carry the visual standard for the new Augustan settlement.21 Onlv in the case of divas Augustus, whose "divina . . . mens ct ttumen" (divine mind and godhead) dominates the world as early as the publication of \ itruv ius's book in the 20s B.C. (De Arch. 1. praef. 1)," does the system not work in quite the same way. lor the emperor is a god and not merely a man; he 56 formulates, administers and if it suits him alters the rules to VIEWING and society which all arc subject.-' Even if Yitruvius is voicing a personal prejudice rather than the official views of the Principate, his polemic is cast in the moral and ideological rhetoric of the Augustan regime. He sees his book as upholding the dignity of Rome's past and present bv prescribing the kinds of buildings and decoration most suited to its greatness (De Arch. t. praef. 3). An insistence on "realism" is a rejection of subversion - a strongly prohibitive stance against the subversive questioning which the paintings of Roman houses pose to the categories by which they are defined. The images Yitruvius attacks (mons-tra) arc insidious because they are unreal and anti-realistic (as if the painting of which Yitruvius approves is any more "real"!). But this rhetorical onslaught is merely the self-righteous prose of the system's own self-defence. In fact. Yitruvius docs not discuss actual examples, but - like Pliny -he presents a list of the proper subjects of naturalism (at the same time an ideal and a nostalgia) which he can then use as a stick with which to beat the degenerate present. The Vitnivian response to the monstra is one of moralistic vituperation. An alternative view of the spectator's attitude is given by I Ioracc in the ekphrasis with which he opens the An Poetica: If a painter chose to join a human head to a horse's neck and to introduce variegated feathers all over the juxtaposed members so that a woman beautiful above could taper oil into a dark fish, would vmi, my friends, restrain your laughter when you came to look? (w. .-5) Laughter and caricature (for instance, a wall-painting from a villa near Stabiae that shows Anchises, Aeneas and Ascanius as dog-headed apes with huge phalli),24 which is to sav seeing the joke and laughing at the system, is a response far removed from Vitnivian condemnation. And yet, as Ovid learned, the one attitude with which the system could not cope and which it ruthlessly punished was ridicule; before the serious face of the mos maiorum and the republic at last restored, laughter was subversive. Thus, even as he opens his Ars Poetica with this ekphrasis, Horace must apologise for it: "Believe me, dear Pisos, the book whose idle fancies are shaped like a sick man's dreams would be like such pictures . . ." (w. 6-8). Yitruvius, less witty and less subtle than Horace, simplv launches into the "sick man's dreams" with all the formidable 57 ways OF VIEWING rhetoric of "realism". And the history of art, mindful of the Vitruvian text rather than of the images which disprove its veracity as anything but self-righteous polemic and heedless of the ideological programme beneath the pungent rhetoric, has followed him down the same path. There are several entailments for the understanding ot Roman wall-decoration in an analysis of the ideological motivations behind Yitruvius's work. The conventional account ot the "four styles" of Roman wall-painting from August Man (who first proposed this division in 1882) to the present is to distinguish them by "their use of architectural illusion".25 Ultimately the criteria - ancient and modern - for illusion (whether as verisimilitude or .is decadence) are based in the notions of mimesis and "realism" which are inextricably entangled with the Vitruvian anil Plinian versions of the Rrinei-pate's political programme. "Illusion" is not a neutral category. The consequences of this extend far beyond the history of Classical art, but what should be stressed here is that at its roots "realism" is a formulation borrowed from I Iellenistic theory and used in Roman times for explicitly ideological reasons. It disguised its Roman politics beneath the claim that it was a neutral and ahistorical category of value which was, in the first place, to be applied to the heritage of Creek art. Second, illusion (which is to say the extent that an image approximates to its "real life" referent) is not necessarily a useful categorv for determining the relationship of the spectator with the image he or she views. For "illusionism" - especially in its Vitruvian polemical formulation as "realism" -focusscs our attention on the relation of image and "reality" and specifically directs us away from that of viewer and image. This is hardly surprising in a political project whose intention is to naturalize viewing, to naturalize the sociopolitical Status quo, by presenting the act of being in a house or looking at a picture as simply the relation to a prior, fixed and natural reality.26 For Yitruvius, viewing must be simple - a matter of giving or withholding one's assent, depending on the extent to which an image achieves the summit of trompe Poeil. Rut the moment view ing is seen as complex -as a process of desire, as a creative act of the viewer's relationship with the image (rather than as merely the passive registering of the "real")27 - then the naturalized and normative status of the political system is called into question. This Yitruvius could not afford to do, but - to his dismay -Roman wall-paintings did it all the time. 5« affirming SOCIAL distinctions: vitruvius's prescriptions for STRUCTURING THE roman HOUSE Before exploring just how Roman wall-painting subverted and questioned the socio-political system, let us examine the \\a\s Romans conceived of the social articulation of the house. It is now generally recognized that "Roman domestic architecture is obscssionally concerned with distinctions ol social rank".:s Roman houses formed the locus for the confrontation of patron and clients and were thus the social focus for the principal ritual of patronage in Roman culture.29 The classic - though brief - presentation is again the l)e Architect lira of Yitruvius. Vitruvius writes explicitly that "magnificent vestibular ta-bulina and atria arc not necessary for persons of a common fortune, because they pay their respects by visiting among others and are not visited themselves by others" (6.5.1). By contrast, the houses of hankers anil financiers should be more spacious and imposing anil safe from burglars. Advocates and law vers should be housed with distinction, and in sufficient space to accommodate their audiences, lor persons of high rank who hold office and magistracies, and whose duty it is to serve the state, we must provide princely vestibula, lofty atria and very spacious peristyles, plantations and broad avenues furnished in a majestic manner -also, libraries and basilicas arranged in a fashion comparable with the magnificence of public structures. (6.5.2) This remarkable summary of the social spectrum of housing explicitly ties houses to the position of the owner in the social hierarchy. It concludes baldly: "If buildings are planned with a view to the character of all the different classes (adsingulorum generumpersonas). . . we shall escape censure" (6.5.3). 'n °ther words, Yitruvius's avowed project to maintain and support the decorum of the Augustan Principate through architecture becomes an overt attempt to bolster the system of social class in Rome. For Vitruvius, domestic architecture is social status made visible. At the heart of the \ itruvian presentation, and of Roman social sensibilities generally, lies a distinction which can loosely be translated as "public" and "private".1" The closer a "private" dwelling comes to being like a "public" building, the grander it is and the higher the implied status of its owner. This is why people holding public office should have ways of viewing houses "comparable to magnificent public buildings" (6.5.2). The conceptual structure of "public" and "private" informs the whole of Yitruvius's account. Yitruvius's fifth book deals specifically with public works (publici loci, 5.12.7; comniunia opera, 6. praef. 7), as opposed to private buildings (privata aedificia. 5.12.7; 6. praef. Houses belonging to the category of private buildings (as they do in modern culture) form the subject matter of book six. The Roman house, however, was itself a problematic version of the private building. Some of its parts were "private" (propria loca patribus familiaruni. 6.5.1), and others were "public" (cotnmunia cum extraneis, 6.5.1). In effect, like the Berber house in Pierre Bourdicu's celebrated discussion, the Roman house represented not only "private space" in opposition to the rest of the world, but within its internal space, it stood also as a microcosm of this very opposition by containing both "public" and "private" areas. - The Roman house functioned simultaneously on two levels. It was both a vital constituent of the Roman social world (in standing for "private space" and thereby establishing the opposition with "public space"), and at the same time, it was a central cultural mechanism for negotiating the very distinction of "public" and "private" (which it in part was responsible for setting up). This social function is further reinforced by the central importance of the theme of public and private in the painted decoration and architectural embellishment of Roman rooms and walls.11 Before we focus further on this theme, however, we should look at \ itruvius's language more closely. The key concepts for the articulation of domestic space in the \ itruvian discussion are those generally rendered "private" and "public", or "common".'4 The relevant passage of \ itruvius is We must next consider by what principles, in private buildings, those apartments should be constructed v» hich are meant for householders themselves (propria), and those which are shared in common (comniunia) with outsiders. For the rooms set aside for the family (propria) are those into which no one has the right to enter unless they have been invited, such as cubicula, triclinia, balneae. and other apartments which have similar purposes. The common rooms (com-munia), however, are those into which even members of the public can come without an inv itation, such as vestibula, cava aediuni, pcris-tylia, and other apartments of similar uses. (6.5.1) The Latin of this passage reads in full: tunc etiam animadvertendum est. quibus rationibus privatis aedifi-ciis propria loca patribus familiariuni et quemadmodum comniunia cum extraaeis aedificari dcbeant. namque ex his quae propria sunt, in ea viewing and society non est potestas omnibus intra fundi nisi invitatis, quemadmodum sunt cuhicula. triclinia, halncac ceteraque, quae easdeni habent usus rationes. ammonia autan sunt, ambus etiam invocati sno in re depopulo possunt venire, id est vestibula, cava aediuni, peristvlia, quaeque eundem habere possunt usum. This discussion is extremely interesting. We find \ itruvius constantly glossing his terms (propria and communia) with explanations that point to their social significance. Propria loca patribus fainiliaruni is explained as in ea non est potestas omnibus intro eundi nisi ini'itatis (not everyone has the right to come inside those places unless they have been invited). The emphasis here is on entry (intro eundi) by invitation to the familv's own apartments tor a speciallv selected group of outsiders. Ibis meaning is borne out by \ itruvius's representation of the other part of the house with the phrase communia cum extraneis - "that area which is shared w ith outsiders". The explanation of communia as quibus etiam invocati suo iure de popttlo possunt venire (to which even members of the public can come without an imitation) implies the accessibility of these rooms to am outsider as well as to members of the family and their guests. As dwellings are in a variety of cultures, the Roman house is a symbolic space which explicates the social distinction of "private" ami "public" through the spatial polar-itv of "inside" and "outside". The house functions as an exclusive precinct (that of the family) open to outsiders in some parts but closed to them in others except under special conditions, such as with the formal granting of an invitation.''' The whole house is conceived in terms of those places clients may be allowed into and those places they arc forbidden (except in the special circumstances of an invitation). The metaphorical resonance of "public" and "private", with its implications of "inside" and "outside" (or perhaps we should say "available to insiders" and "accessible to outsiders"), reaches deeper than the social definition of house and family to encompass the identity of the individual viewer. The Younger Seneca (De Ira 3.35.5) agrees with contemporary anthropology that the viewer is actually constructed w ith different levels of perception depending upon whether he is "inside" or "outside" the house.'6 Seneca writes: These same eves, forsooth, that cannot tolerate marble unless it is mottled and polished with recent rubbing, that cannot tolerate a table unless it is marked by many a vein, that at home (domi) would see under foot onlv pavements more costly than gold - these eves when outside (foris) will behold, all unmoved, rough and muddy 61 u \VS OF VIEWING paths anil dirty people, as are most of those they meet, and tenement walls crumbled and cracked and out of line. Why is it, then, that we are not offended on the street (in publico), vet are annoved at home (domi)} Grasping, then, both the broad political agenda within which Vitruvius is working and the specific hierarchy of social status which his architectural and decorative injunctions are designed to uphold, let us confront some of the buildings themselves and the decorations against which he inveighs in his seventh book. When the social and ideological agendas in the Principate's architectural project actually come to be expressed in terms that relate to the pragmatics of actual buildings and spatial organisation, Vitruvius seizes on the crucial distinction of "public" and "private" with its spatial resonance of "inside" and "outside". This spatial resonance characterises not only social and topographic conceptions of the house but also the conception of the Roman perceivcr expressed in Seneca's portrait. In the next section we shall explore this problematic as it is constructed by the art and architecture of actual Roman houses. SYMBOLIC INVERSION: SOME ROMAN HOUSES I ROM HERCULANEUM AND POMPEII On one level, houses - and the various tastes according to which they are decorated - simply exist. They form the lived-in environment in which an individual is brought up anil experiences his or her society. But this apparent given, or natural quality, which people feel inherently belongs to things they are familiar with, is in fact the product of habituation and conditioning. What seems natural and normal has been naturalised and made normative through constant exposure, use and experience: There is nothing intrinsically natural about how a house is designed or decorated. We have seen in Roman culture that the domestic house and its paintings were explicitly subjected to planning of a highly ideological and social nature. Vitruvius s text is proof that the Principate encouraged architects to think about how their designs for housing helped to structure and to perpetuate social hierarchies, and about how the way in which they decorated houses conformed to criteria of a deeply political and moralising dimension. But how did the Roman houses themselves, the actual edifices in which all of this complex of ideological 62 prescription was to operate, exhibit and relate to the rules? viewing and society I low were the houses and their decorations viewed? Because mv purpose is to site Roman art in the context of the cultural categories by which Roman viewers would have seen it, 1 should make clear at the outset mv position on the major art-historical classification according to which Roman wall-painting has been analysed. This is, as noted earlier, the "tour styles" first proposed in 1KK2 by August Man and accepted more or less by subsequent critics.' I accept the usefulness of the "styles" as working categories for describing and grouping the visual material. Nevertheless, although the "styles" have been seen by modern art historians as chronological periods in the evolution of Roman art, this is not how they would have appeared in the beholders experience. A room such as the atrium of the Samnite House in I lercula-neum (Figures 9, 10 and 12), with its "first style" gallery and vestibule and yet its "fourth-style" wall-paintings,'* is considered to l>e decorated in two distinct styles only for purposes of modern analysis. The evidence of Pompeii and Herculaneum points not to four separate Styles of decoration, but rather to the co-existence and synchronicity at a single date (a.d. 79) of a variety of types of decoration which we choose to label "four styles."'9 Although the "styles" - being a modern invention - may make some analytic sense to us, they do not offer any insight into Roman ways of looking at, or thinking about, art. Indeed, they confuse our understanding of Roman viewing because thev impose on the evidence an entirely modern analytic frame. On the face of it, there is a Roman text defending one of the distinctions which Man later adopted in formulating his "four styles": This is the celebrated polemic by Vi-tnivius against the degeneracy of modern painting, which we discussed earlier (f)e Architect urn 7.5). Yitruvius distinguishes (as do Man's "four styles") between different kinds of illu-sionism - images which imitate "real" things and images which imitate things that could never exist. Thus Yitruvius would appear to be giving contemporary Roman backing at least to Mau's division of "second" from "third style".40 However, as I have argued, to read Yitruvius as if he were providing an objective description of the material (as Man purports to do) would be utterly misguided. On the contrary. \ itruvius's discussion of art is a classic piece of Roman moral invective aimed against that most popular of all Roman betes /wires, "decline". Yitruvius is not objective but is polemical and ideologically motivated. More significantly still, Yitruvius is only 63 w \VS OF VIEW ING Figure }. Villa of the Mysteries peristyle, Pompeii, first century b.c. to first century a.u. Photo: Silvia Krenk. one voice among several. The line he draws between "realistic" and "unrealistic" illusionism is his own line and need not have been accepted by everyone else. His younger contemporary, the orator Papirius I'abianus. who lived in the first half of the first century A.n., in a similar invective against art draws a different line (this speech is preserved bv the F.lder Seneca, CotltroversUK 2,1.1 I —I 3). For Papirius all illusionism, all imitation, is "unnatural" and "debased". It is clear that Roman views of wall-painting were not a static set of objective descriptions (as critics have taken Man's "four styles'1 to be); they were the subject of a fierce polemic within the culture which had not yet formulated a final view of what was right and wrong in art, let alone any standard description. In what follows, I shall use the language of the "four stvles" as it is convenient and familiar to the reader, but I do not believe in its validity as an illumination of Roman ways of viewing. Some time, perhaps in the first century A.n., the peristyle of the Villa of the Mysteries in Pompeii with its fluted stone columns was remodelled (Figures 3 and 4). A low wall of some 1.5 metres - the technical term is a "pluteus" - was inserted between the columns and continued all the way round the peristyle. Then the whole lot (wall and columns, which still extended about another 1.5 metres above the plu-teus but at their base were part of it) was plastered over and painted. In effect, the prev ions cloistered arcade on four sides had become a square of corridors whose half-walls on the side closest to the central opening were vividly described by the excavator, Amadeo Maiuri. as "fenestrato" (windowed).41 Such a low-walled peristyle is not unique in Roman architecture.42 However, what is remarkable about this particular structure is that it sets up a "real life" version of the characteristic painted decoration of a "second-style" room in a Roman house.4. The classic description of "second style" presents its decoration as three-dimensional illusionism in which the trompe Voeil of a "complete architectural system was developed". It is said to adhere to actual or possible structural forms where, above the frescoed walls and columns, are often painted a glimpse of the sky, further architectural vistas in perspective or views into the distance.44 The effect of being inside such a room, for instance, the cubiculum of the Villa of P. 1'annius Sinistor at Boscoreale - a room now reconstructed in the Metropolitan Museum in New York (Plate i)4> - is of standing inside a covered enclosed space and looking out over the trompe Pool wall. Perhaps beyond this viewing and society Figure 4. Yilla of the Mysteries peristyle. Pompeii, first century b.c. to first century a.d. Photo: Silvia Frank. 65 ways of viewing first wall would be a second painted wall "behind" it and beyond that the painted skv and landscape "outside". The Villa of the Mysteries peristyle recreates in three-dimensional "real life" this "second-style" illusionist structure by presenting the spectator not with a trompe Foeil half-wall opening to an illusion of sky behind, but with the real thing -a low wall between columns.46 And yet it utterly inverts the process of "second-style" viewing. If one stands inside the central court, around which the peristvle and its pluteus run (Figure 5, .1), one is actually outside in the open air looking over the wall into the house. Although the beholder stands "inside" the court, his \ ie\\ is not from the inside out (as in a painted room) but from the outside in. When the spectator stands in the portico (Figure 5, B) - still within the peristvle court, although outside the central opening itself under the cover of the roof between the house walls and columns extending from the pluteus - his view is over the pluteus from the inside (which is to say, beneath the roof) into the outside (the inner court of the peristyle under the skv). But this is deceptive, because what one actually sees is not landscape outside (as one would in the illusion of some "second-stvle" rooms, such as Oecus 6 of the Villa of the Mysteries itself) but across the court to the other side of the peristyle and back inside to the relative darkness of the arcade opposite. Were the walls of the peristyle (whether those of the extreme perimeter or those of the pluteus) to have been painted with landscapes at any stage, then the effect would be still more complex, for then interior parts of the house (which one might bevies ing from the open air) would bear the painted signs of "outside" - such as foliage or architectural vistas. The point I am making is relatively simple. Just as Yitruv-ian textual rhetoric constructed hierarchies of access dependent on the categories of "inside" and "outside" and "private" and "public", so the visual rhetoric of Pompcian architecture played with the same themes. Just as \ itruvius's prose found the categories of "public" and "private" to be problematic (both denoting areas within the house and defining the difference between the house and the space outside it), so the peristyle of the Villa of the Mysteries emphasises this very complexity. If, in visual terms, "outside" means under the skv as well as outside the house, then some of the house's most internal rooms are "outside" because they are open to the air. As soon as one puts the emphasis on viewing, one discovers that the art of the Roman house highlights an ambivalence 66 vi k wing and socifty (§) = inside the central opening of the peristyle ® = inside the portico of the peristyle and ambiguity of "inside" and "outside" and of "public" and "private". An emphasis on viewing is not anachronistic, for it has been shown that this is precisely the angle from which some of the major Roman poets investigated art.4 This visual questioning does not simply make problematic the status of particular rooms or buildings. It is, more fundamentally, a questioning of the viewer himself as the one who constructs "inside" and "outside", and thus - obliquely - a questioning of the whole social system in which the viewer has his part. In this sense not only do Roman houses set up and define the structure of Roman society and class, but they also reveal, question and deconstruct it. The complications of "inside" and "outside" are not confined to actuality. In the trompe Poeil of Pompcian rooms there may be a play of one part of a room being "enclosed" and another being "open to the outside". For instance, in Oecus 6 of the Villa of the Mysteries,48 the murals in the northwestern half of the room represent an interior peristyle and "behind" that a panelled wall not open to the sky at the top (Figure Figure 5. Villa of the Mysteries (after Maiuri). Plan with viewing positions around the peristyle: . \. inside the central opening of the peristyle; H, inside the portico of the peristyle. 67 ways of viewing Figure 6. Villa ot the Mysteries. Oecus 6. Pompeii, tirst century B.C. Photo: Silvia I'renk. viewing am) society = "ideal" viewing position suggested by the room's perspective Figure 8. \ ilia of die Mysteries, viewing positions rod perspectives in Oecus 6 (after Corläita Seauliarini). 6). This is an enclosed room - an "inside". However, the southeastern side of Occus 6 is painted as an interior peristyle with illusionist frescoes representing a wall and "behind" this an "outside" open to the sky (Figure 7). It is as if we are no longer in an interior room but in one which is at least open to the outside. In addition, in Oecus 6 there is much play of real doors and false doors, with a real door (the one through which wc enter) in the southeast wall and a painted door facing it in the northwest wall (Figure 6). Scholarly discussions have emphasised the play of different kinds of perspective implicit in such divided decorations.'1'' In Oecus 6, for example, there is an ideal viewpoint implied by the perspective midway between the side walls and some two-thirds of the way from the door (see Figure 8). The perspective implies that the room was decorated to be seen as a wboU, encircling the viewer from a particular point/" It is because the room is carefully designed and painted to be a whole, that the effect of its play of "inside" and "outside" is so complex WAYS OF VIEWING FigUIt 0. Atrium of the Samnitc I louse, Herculaneum, firsi anum b.c. Photo: Silvia Krcnk. 1 and disturbing. The room's trompe Voeil constructs an architectural decorative scheme which makes its "insideness", its status as an interior room, problematic. Furthermore, within the room there is a questioning of just which parts are more or less interior - as if the frescoes were representing a series of "levels", a hierarchy, of insideness. When one takes the cumulative decorative effect of this room and of others similarly playing with levels and transgressions of "inside" and "outside" in the Villa of the Mysteries (for example, cubicula H and 16)" and compares it with the inversion enacted by the peristyle, it appears that the problems implicit in these social-architectural categories are being unashamedly emphasised. It is as if the decoration of the house is obsessed with the wcirdness, the unnaturalness, 70 VIEWING AND SOCIl I V FigUN IO. Atrium of the Samnitc I louse. I Ierculaneum, first centurv b.c. Photo: Silvia Frenk. of the categories which define its articulation both as a building and as a social experience. In the Samnitc House at Herculancum, this complexity within a single room is still more stronglv marked because it adds the frisson of false and real stucco colonnades.1-1 One enters the atrium from a vestibule decorated in "first stvle" (Figure 9) - that is, where plaster is constructed in three dimensions and painted to resemble marble veneer.53 This atrium, one of the largest in Herculancum,'4 is in other respects typical of many houses in both Herculaneum and Pompeii. It has a marble inipluv ium in the centre of the floor to catch the rainwater and a compluviate roof with a large opening in the middle (figure 10) to let in the light (and the rain). Of course, one reason for these openings is functional - to 7' ways or virwiM, Figure 11. Uvia's Villa at Prima Porta, garden view and knee (now in the National Museum in Rome), first century b.c. Photo: Alinari-Art Resource, New York. light not only this room but also those adjacent to it. But the corollary is a play on whether this interior space is an inside room or an outside room - a problem that reaches its peak in the high art of the Pantheon in Rome, whose famous oculus, which allows in the light, is open to the elements. Like the Pantheon, all such compluviate rooms have an uneasy or ambiguous status as to whether thev are quite inside rooms or outside rooms. The Pantheon, of course, is a public space and a very public building of official worship; but it mav be significant that atria, like peristyles, belong to Vitruvius's category of communia cum extraneis, or "common with outsiders" (6.5.1). However, the special interest of the atrium in the Samnite House lies in its decoration at gallery level: On the level of the upper storey, on the side at which one enters and on the two adjacent sides, there is a magnificent false gallery, finished in stucco, of fluted ionic half-columns joined bv a low perforated parapet. In a sense this embellishment might be described as the "first-style" equivalent of "second-stvle" architectural features, such as the perforated fence in the lower foreground of the "Garden Room" from Livia's villa at Prima Porta (Figure 1 i).5i But in the fourth wall, opposite the vestibule door and below the compluviate ceiling, the false pilasters and imitation fence break into reality (Figure 12). There-is no wall at gallery level for them to decorate; they stand as themselves - an ornamental feature turned functional, for viewing and society they hold up the roof. Just as in the part of Oecus 6 of the Villa of the Mysteries where the painted architectural decoration with its stress on the illusion of solid walls abruptly breaks open to reveal the illusion of half-walls and "behind" them the "outside", so here the interior compluviate atrium (open to the outside only through its roof) breaks open its wall on the far side from the street entrance "really" to reveal the outside. Of course, this outside - from the viewer's level, the sky - is actually the space above the tablinum, which is the next room on from the atrium.56 In none of this has the decoration abandoned "realism" in that all its motifs, whether in two or three dimensions, imitate Figure i.. \trium of the Samnitt 1 louse. I lerculaneum. tirst century b.e. Photo: Silvia Irenk. 73 ways of viewing the actuality of real architecture. But none of these effects creates the impression of "actual reality". One wonders, in the case of the Samnite House, whether the "real" gallery of stucco pilasters and parapet is imitating the "false" one or vice versa.' The illusionism conjures an altogether transgressive world which seems to question the world it imitates, that is, to deconstruct the rules of the realitv on which it is selfconsciously based. It is an example of what anthropologists have labelled "symbolic inversion".'* The challenge here is on two levels: first to the notion that the real world can be imitated at all, that "realism" is any more than a pastiche, a set of illusionist tricks which contradict one another. The extreme version of this challenge would be to deny the "realitv" of the real world itself and see it as merely a set of constructions, a set which is in its own way as contradictory and ambiguous as are the "realisms" of Pompeian walls. Second, Pompeian illusionism - by playing visually and spatially with the categories of "inside" and "outside", of "private" and "public" spaces, upon which Yitruvian rhetoric deemed the social structure of the Roman house to be based - is implicitly challenging the naturalness of this social structure.'9 In short - like Mannerism, or like the Surrealism of Magritte, De Chirico and Dali when they refused to abandon the naturalistic idiom whose motifs they juxtaposed to such surreal effects60 - the decoration of Roman houses in the first centuries b.c. and a.d. seems to challenge the \ iew cr by undermining the reality (both social and natural) which it presupposes.61 As \ itruvius already saw, it is the process of realist viewing itself which is at stake in the villas of the Roman middle class. VIEWING AM) DKSIRK: I UK YI EWER AND THE VIEW Where docs the viewer stand in relation to the surrealism displayed with such exuberance on the walls, in the structure and in the architectural ornaments of his or her house?62 It has recently been argued of Roman w all-painting that "from the point of view of the social function of decoration what matters is not the visual games played, but the associations evoked, by the decoration: its power not of illusion but of allusion"/'- This approach indicates an attempt to reintegrate the viewer into the impact of images in Roman houses beyond "4 being simply the one who approves or condemns as in the viewing and society Vitruvian account/'4 However, the ideological programme oi Vitruvian "realism" shows that the visual games are themselves socially determined and of central importance as part of a visual questioning of the categories for social function. In other words, as I hope I have show n in the preceding section, the visual games are crucial to the way the Romans constructed social meaning. The question of allusion, therefore - not what an image alludes to, but what that allusion signifies - is impossible to separate from the status of that allusion as an illusionistic picture, as a phantasy rather than a reality. Whereas commentators agree that the views on Roman walls generally evoke wealth and grandeur, there have been two theories on what such evocation signifies. The first supposes that illusionistic pictures offer us in some sense the world they imitate.'" I he second suggests that such images - because they offer an imitation anil not the "real thing" - actually deny to the viewer the world they reflect.6* Paradoxically, these positions are both in fact true. They represent the poles between which the viewers' desires must be constantly oscillating. The essential point of allusion is that it offers not a final answer or a gradation of levels as does "realism", with its notion of the perfect and essential copy, but a constant play with desire, furthermore, Roman allusion must aim at the impression of verisimilitude because it constantly proffers the promise of real things (that landscape, that bowl of fruit, that vista of a town) just around the corner as it were, just through the wall. In this sense the illusionism of Roman allusive art is fundamentally implicated in a materialist view of the most suitable objects of desire/' Even the Divine is locked into this materialism - so that the famous PriapUS in the vestibule of the House of the \ ettii in Pompeii is actually weighing his phallus against a bag of money. Because Roman houses are inextricably linked with social-status considerations, and because better in this context means bigger and more open to larger quantities of clients, the allusion must work by means of the illusion of more "outside", of a private space almost completely public/'* Our problem is that we cannot pinpoint precise meanings for such illusions, but we can situate the sphere of the viewer's desire w ithin which a multiplicity of such meanings will be evoked. That desire, like everything else about a Roman house, lies in the materialist realm of social status. 75 ways of viewing To go deeper into the operation of desire in the Roman house, we must observe a further link between allusion and the illusionism of Roman walls. Roman murals of all "four styles" tend to represent architectural ornament which either encloses the viewer within a room ("first", early "second" and "third" Styles) or offers glimpses into a work! outside or beyond, glimpses through illusions of architectural frames that are more or less "realistic" ("second" and "fourth" styles). Thus the images of Roman wall-painting, like the problematics of "inside" and "outside", are essentially a matter of the existence or non-existence of the view. Again wall-painting is playing with exactly the same issues as is architecture - one of the marks of an atrium or a peristyle, as we saw earlier, is that it is always a space that allows the viewer to glimpse other spaces, to see more of the house. The view - the nature of a spectator's relationship between here and there, the place from which he or she looks and the place to which the looking is directed - is thus perhaps the crucial determining factor, not only of Roman wall-painting but also of the visual and social articulation of the Roman house.*' Roman depictions of houses are obsessed with the view. I lere is the Younger Pliny, nephew of the encyclopaedist and a distinguished orator of the last years of the first ccnturv a.i)., describing the triclinium of one of his villas, on the coast at Laurentinum near Rome: It has folding doors or windows as large as the doors all around, so that at the front and sides it seems to look out onto three seas, and at the back has a view through the inner hall, the courtvard with two colonnades and the entrance hall to the woods and mountains in the distance, (lipktulae 2.17.5) In two letters (2.17 and 5.6), 0 Pliny describes his villas as though one is passing through them - presenting virtually a list of rooms, in order, 1 and frequently emphasising the views through the rooms back to other rooms or out to the garden and the sea. : In fact, educated Romans like Pliny were remarkably sensitive to the environmental context of a house, the flow from room to room and object to object. The context is most strongly experienced as a set of view s. The Romans' awareness of this topographical sense of flow in architectural space led them to use it in the most remarkable way as a mnemonic tool to structure the way they memorised their speeches. In a culture which put such a strong emphasis on rhetoric in education and intellectual life as did the Roman 76 world, 1 diis gives the importance of a person's relationship viewing and society with his or her house as a viewer a heightened ami privileged status. Before looking directly at views in Roman houses, we should venture down the strange side-alley of memory and its relation to the Roman house. 1 lere is the Roman theorist of rhetoric, Quintilian. writing in the first century a.d., describing the orators use of the Roman house as a tool for memorising the structure anil coordination of a Roman speech: Some place is chosen of the largest possible extent and characterised by the utmost possible variety, such as a spacious house divided into a number of nntms. Ever) thing of note therein is carefully committed to the memorv, in order that the thought mav be able to run through all the details without let or hindrance .... Particular symbols which will serve to jog the memorv . . . are then arranged as follows. The first thought is placed, as it were, in the ivstibulum; the second, let us say, in the atrium; the remainder are placed in due order all around the xmpluvium and entrusted not merely to cubicula and cxedrae but even to the care of statues and the like. This done, as soon as the memorv of the facts requires to be revived, all those places are visited in turn, and the various deposits are demanded from their custodians, as the sight of each recalls the respective details. (Quintilian. Institutio Oratoria 11.2.18-20) The "symbols" Quintilian mentions are specific images or words which can bring back to the orator's mind the subject of his speech or even an actual sentence at a particular point in his discourse (the point, for example, of the vestibu/um or atrium or statue in the house which the speaker has memorised and through which he progresses in his mind's eve as he delivers his oration). This art of memorv had a significant place in the Roman art of rhetoric, 4 as is demonstrated by the fact that not only Quintilian in the carlv Empire, but also two rhetorical treatise writers from the late Republican era discuss it in some detail as part of their general instructions on oratory. ' The significance of this art of memorv for the art historian and the social historian has not been sufficiently explored. If the Roman house, 6 the layout of a dining-room, the common architectural forms of the late Republic and carlv Empire, 8 served regularly as the means for ordering and memorising speeches, then equally the order and structure of these very houses, dining-rooms and architectural forms are the three-dimensional embodiment of the process of structuring thought. Just as the periodic sentence and the rhetorical "7 ways of viewing speech mav be co-ordinated and constructed through the memorised vision of a house, so the Roman house in the totality of its rooms and decor (not forgetting the statues, as Quintilian reminds us) '' is one representation of the structure of Roman rhetoric. Romans thought by means of their houses - their visual and architectural environment. It is this environment (not only houses but all kinds of loci, from Simonidcs's banqueting room to "public buildings, a long journey, the ramparts of a city, or even pictures")80 which defined the Roman art of memory ami which structured Roman discourse. When Romans thought about embellishing their environment - about improving houses, about building new houses, about making gardens - it w as precisely the rhetorical process conditioned by the visual memory which they brought to their architecture and art. Just as Roman oratory is conditioned by architectural and visual imagery to a remarkable degree, so the decoration and structure of the Roman house is in its turn conditioned by the rhetorical process which was later to make use of the house as an aide-memoire. In effect, we cannot draw a sharp distinction between the architectural and visual world of the Roman educated elite on the one hand and their mental and rhetorical world on the other. Together they make up the mei/ta/ite, the particularity and identity of Roman civilization: Each was the precondition and determining impulse behind the other. There are a remarkable range of issues raised by a closer art-historical examination of the treatises on memory. For example, they offer valuable evidence for the existence of a self-consciously non-literal, metaphorical, system of viewing in the Roman world which subjected the most commonplace and familiar images to a set of sv inbolic meanings that spe-cifically had no direct relation to the original object. It might be quite normal for a statue off a peristyle to represent the third paragraph of a Ciceronian speech! This is exegetic viewing of remarkable sophistication. And, unlike most exegetic viewing in Classical or Christian antiquity, it is categorically not concerned with religious interpretations or meanings. A second area, one highlighted by the distinguished cultural historian Frances Yates, is the "astonishing visual precision" of Roman viewing to which these works on memory bear witness. In the classically trained memory even the space between loci can be measured and the lighting of the loci is allowed for.81 Cicero's emphasis (initially surprising to us) on the sense of sight in rhetoric (De Oratore 2.87.357) makes the point nicely. The speaking is based on a viewing: The Roman viewing and society speech is the rhetorisation of a prior and visually re-lived view. This rhetorical use of a memorised "visualisation" is a classic instance of the importance of phantasia in ancient culture, which we discussed in the previous chapter. The visualisation of the house in the orator's mind is a phantasia. Instead of bringing this specific phantasia to his listeners' minds through ekphrasis, the orator (in this case) presents the visualisation through the structure of his speech. In effect, a rhetorical speech is a metaphorical ekphrasis. A speech uses all the skills of ekphrasis (its claritv, visibility and sharpness) to construct an ordered discourse which has a metaphorical relationship to the original phantasia of the house. In particular, I want here to emphasise one aspect of the Roman art of memory and its relation to the Roman house. This is the notion of order or series (onto), which is one of the most important qualities of a house for the oratorical handbooks. It is precisely because the various loci (such as the vestihulum, atrium and impluiium suggested by Quintilian) must be memorised in a particular order as one moves through the house, that this movement from locus to locus can be used to structure a speech. As Cicero puts it, the order of the loci will preserve the order of the material to be remembered (Dc Or. 2.86.354). The notion of the series, of the view as it were from one locus to the next, is crucial - as the Auctor ad Ilerennium makes explicit (Ad. Her. 3.17.30). Once the view er-memoriser is sufficiently skilful, he can move (which is to say, give his speech) in any order - "forwards or back-wards" (3.17.30) - so that "if the loci have been arranged in order, the result will be that, reminded by the images, we will repeat orally what we have committed to the loci, proceeding in either direction from any locus we please" (3.18.30). In other words, unlike ekphrasis which necessarily free/.es the speaker's phantasia in a particular order or structure, the range of possibilities for the ordering of paragraphs in a speech allows a much greater flexibility and freedom to the orator's use of his memorised vision, his phantasia. The art of memory co-ordinates a Roman speech and the "order" in which the speech is seen. It is visual and three dimensional - using the awareness of other rooms, other floors and particular features of decoration which one has on entering one room of a house from another. In essence the art of memory is a means of exploiting this three-dimensional environmental sense with all its flexibility to the needs and ways ok viewing demands of rhetoric. What matters about a house, as far as the rhetorical treatises arc concerned, is this dynamic sense of movement - from room to room, from object to object - so that the ductus locorum (the flow of place to place) matches the ductus litterarum (the flow or structure of the speech). This environmental sensibility for the relation of a part of a house to other parts, and to the whole, is an essential element in Roman viewing. It is an important constituent of the notion of the "view" - which is precisely the visual relation of one-part of the house or city or landscape to another or to the rest.K: Reside the aesthetic or experiential aspect to Roman domestic viewing, and in addition to the remarkable intellectual quality of using one's view of a house as an aide-memoire\ there was always (as we have seen) a flavour of social status about how Romans looked at their house. Furthermore, the language by which the paintings that decorated Roman houses were described (at least by \ itruvius in the Augustan period) was the result of a carefully constructed ideological polemic designed to preserve the social and political distinctions of which the Roman houses themselves were part. The paradox is ihat the language prescribed one set of rules ("realism") while the paintings and architecture played with and challenged those rules ("surrealism"). It remains to bring these strands together in order to examine the importance and vet the ambiguity of the view - that complex relationship between the beholder and what he or she sees, between individuals and their social world, in which subjectivity is constructed/' Describing his villa at Laurentinum near Rome {i'.pistulae 2.17), the Younger Pliny repeatedly remarks on the views from his more elegant rooms. A householder as well .is an orator, Pliny as viewer was concerned w ith the environmental flow of his house, with the directing of the view through and out of the house. The triclinium (2.17.5), as uc nave s^en, looks to the sea on three sides and on the fourth side looks back all the way through the house to the countryside beyond the atrium. Off this, there is a cubiculnni with a view of the sea (2.17.6). On the upper floor there are more rooms with prospects of the sea, the garden, the coast and other villas (2.17.12-13). ' 'u' triclinium on the other side of the house away from the sea "has a view as lovely as that of the sea itself, while from the windows of the two rooms behind can be seen the entrance to the house and another bonus" 80 (2. i-j. 15). The main point made of the cryptoport kits is Pliny's VIEWING and society mention of its windows opening on both sides (but more to the side of the sea) (2.17.16). furthermore, Pliny's own favourite suite of rooms (2.17.20b) is distinguished by a "sun-parlour facing the terrace on one side, the sea on the other and the sun on both". Its rooms open onto the sea, the ciyptoporticus, the neighbouring villas and the woods - "views which can be seen separately from its many w indows or blended into one" (2.17.21). When describing his villa in Tuscany (Epistulae 5.6), Pliny is no less insistent on the view: "My house is on the lower slopes of a hill but commands as good a view as if it were higher up" (5.6.14). Me continues his description: from the end of the colonnade projects a triclinium: through its folding doors it looks on to the end of the terrace, the adjacent meadow and the stretch of open Country bevond, while from its windows on one side can he seen part of the terrace and the projecting wing of the house, on the other the tree-tops in the enclosure of the adjoining riding ground. (5.6.19) Likewise the cauitio "looks on to the small courtyard, the colonnade and the view from the colonnade" (5.6.21). For Pliny, the articulation of his house, the reason for his pride in it and the most effective method of communicating what he sees as its best qualities arc all defined by the view. This emphasis on the view can be seen not only in texts but also in surviving villas and, as we have observed, in their wall-paintings.M There is, moreover, a constant emphasis on the relationship of rooms to the light and the sun - a further example of the "environmental sense", this time referring not to a room's view of the outside, but to the outside's penetration of a room. Finally - and again very relevant to the view - Pliny indulges in lavish descriptions of the countryside (both cultivated and wild) in which his villas are set.M The villa cannot be separated from its setting, just as its rooms cannot be considered without reference to the light which enters them from outside and the views which they offer onto the outside.8 Pliny's concern with the view and the lighting of his villas is not idiosyncratic.88 Precisely these issues arc built into Yitruvius's injunctions about the layout of domestic housing, especially in his sixth book. In \ itruvius there is a repeated emphasis on light: CubkuU and libraries should face east; baths and winter rooms should take their light from the sunset 81 ways of viewing (De Arch. 1.2.7; «.4.1). Picture galleries and rooms needing a steady light (such as weavers, workshops and painters' studios) should he lit from the north (1.2.7; 6.4.2). Spring and autumn triclinia should look east, but summer dining-rooms should face north (6.4.2) to avoid excessive light and heat. In villas, care is required that all buildings should be well lighted. This is an easier matter with those on country estates because there are no neighbours' walls to interfere; whereas in cities, the heights of party walls or the narrow streets are in the way and obstruct the light.*9 (6.6.6) \ itruvius's worry about light, especially in towns like Pompeii, causes him to suggest that walls should be opened at the top, especially if beams, lintels or flooring interfere with the entry of light (6.6.6-7). Campanian practice (such as the gallcrv in the atrium of the Sanmile I louse in Ilercula-neum, Figure 12) gives evidence of this method of lighting a room. Furthermore, the way "second-st\ le" w alls are painted with the illusion of an opening at the top to receive daylight seems to indicate a trompe foci! play w ith precisely this feature of Roman architecture. Again the discussion and aesthetics of light reflect on the problematics and the play of "inside" and "outside" in Roman domestic space. The corollary of this emphasis on light and the view in architecture, however is that Roman three-dimensional space is actually constructed w ith as much trompe Toeil - as much play on forms, spaces and illusions - as is Roman wall-painting. In a remarkable passage, \ itruvius admits as much: The appearance of a building (species) when seen close at hand is one thing, on a height it is another, not the same in an enclosed space, still different in the open, and in all these cases it takes much judgment to decide what is to be done. The fact is that the eye does not always give a true impression (veros effect us), but verv often leads the mind to form a false judgment (fallitur saepius indicia ab eo mens). In painted scenery, for example, columns may appear to jut out. mutules to project, and statues to be standing in the foreground, although the picture is of course perfectly flat. Similarly with ships, the oars when under the water are straight, though to the eye they appear to be broken .... Now whether we see by the impression of images upon the eye. or by the effusion of rays from the eyes, as the natural philosophers teach us, both explanations suggest that the vision of the eyes gives false judgments {falsa indicia). Since, therefore, what is real (vera) seems false (falsa), and some things are approved by the eyes as other than they really are (aliter i/nani sunt). «2 I think it certain that diminutions or additions should be made to viewing and society suit the needs and nature of the site. (6.2.2!".) Despite the firm insistence on the "real" (oars are straight anil not bent whatever the phenomenology of perception may tell us), Yitruvius is nevertheless caught in the double-hind of his Own acute understanding that "the vision of the eves gives false judgments". The moment he orientates his discussion on the spectator, as any good architect must, he is trapped in his own clear sense of the deconstructive complexities of illusion and deception, where the "real" is "false" (as he admits when telling us that apparently three-dimensional trotttpe Foeil stage scenery "is of course perfectly flat"). Yitruvius's problem is how to relate the phenomenology of perception (to which he must give precedence if he is to design buildings most effectively to accommodate the viewer) with what is "really" the case (his ideological stance in relation both to wall-painting anil to the "facts"; for instance, that an oar is "really" straight and not bent). Nothing indicates this problem so clearly as his use of illusionist painting (such as stage scenery) as a programmatic exemplum for how the viewer sees a house. From this exemplum we can be quite sure that Yitruvius is aware that even the paintings which he so praises at 7.5 for their Veritas (their imitation of the "real" rather than the "false") are - in his own terms - themselves falsa, the deliberately designed deceits bv which fallilur saepius mens (the mind is often deceived). For Yitruvius, architecture is as much an illusion as painting. Moth are a matter of "diminutions and additions to suit the needs" of viewers; both are a matter of the view . Decoration is not separable from or secondary to the building it decorates, because the example of illusionist architectural painting is both the metaphor and the metonym out of vv hich Yitruvius constructs the viewer of architecture and hence the architecture that will suit (and deceive) the viewer. It follows that all the "real life" views that Pliny mentions with such pride are designed with that very (dangerous) sense of i 1111 -sionism which - in the case of paintings - Yitruvius feels he must constrain between rigidlv "realist" boundaries in his polemic of Mook 7. Fkncc in architecture "there follow . . . the adjustment {apparatus of the proportions to the decor so that the appearance (aspectus) of eurhythmy may be convincing (iioii dubius) to the observer" (6.2.5). Vitruvian architecture-is as much trompe foci/ as painting. But the problem is that «3 ways of viewing aspirins (like species a term which ignores the image-referent relation in favour of the "deceitful" effects of images on view -ers) must he a highly problematic concept to one who will shortly be insisting with such moral surety on the virtues of "realitv" as opposed to "falsehood". Vitruvius admits that from the stance of the observer there is no essential difference between the logic of a "real" (three-dimensional and architectural) view and the logic of a "false" (painted) one. 11 ere Quintilian, in his prescriptions on memory, agrees: "What I have spoken of as being done in a house, can equally well be done in connexion with public buildings, a long journey, the ramparts of a city or even pictures. Or we max imagine such places to ourselves . . ."(fust. Or. i 1.2.21). All these kinds of views - "real", painted or imagined - are, from the standpoint of the complexities of viewing, substan-tiallv the same. Hut they are all problematic. In the case of the "realist" art and views approved by Vitruvius (i.e., the "second style"), a view is both access and denial. From a particular place one catches a glimpse of the world - something possible and possessable. That world defines and articulates the place one is, that is, the place from which it is seen. But although the view offers that world, it also simultaneously refuses it - for, by definition, a view implies absence, implies a "somewhere" that is not "here". The viewer is constructed as one to whose desire the world he has glimpsed is simultaneous!) offered and denied. Thus the viewer is defined as both the one who desires what he sees and the one whose desire is curtailed by the absence of or distance from what he sees (in the case of painting, by the fact that what he sees is not real). Both Roman houses and their images put the viewer into a position of conflict and contradiction. They make that conflict explicit. In the art which followed the kinds of paintings Vitruvius condemns (i.e., "third" and "fourth" styles), where images partake of the perspectival rules of illusionism but present the viewer with an entirely "surreal" effect, the function of the view is different. Instead of locating conflict in the observer, this art transposes the contradictions to the very walls themselves. What one sees, displayed to such transgressive effect on, for instance, the "architectural" frames of the walls in the Pentheus room of the I louse of the Ycttii in Pompeii is the contradiction of the rules themselves. Mapped onto such "third- and fourth-style" walls is the conflict of the desire that operates by the rules. The law s of illusionism, which govern both wall-painting and house design, whether presented as «4 mathematical, prescribed by ideology or inscribed in the prac- viewing and society ticc of ancient perspective,90 are not "innocent". They "make" the world and formulate the seen into a particular relationship with the spectator. Thus, to see according to the laws is to curtail the possibility of "reallv" (i.e., innocently) seeing what is "really" there (i.e., what exists objectively outside the laws). For the "real" is seen only as represented by the rules. Thus, to see according to the rules is to see the rules themselves displayed in both their tyranny and their contradiction; it is to see that ideally (but impossibly) one might see w ithout the rules; it is to see the possibility of there being more to see beyond the rules and to have that possibility denied, limited, curtailed by the necessarv organisation of seeing according to rules without which we could not relate to what we see. This is the deconstruction of viewing (and of the ideologies and rules of viewing) which \ itnivius so detests and which "can delabra upholding pictured shrines" ultimately imply. IMF. ROMAN HOUSE AS \ CULTURAL SYSTEM The Roman house was a strange conflation of official policv and private transgression. Its structure and forms explicitly embodied the social hierarchy. In \ itruvius's idealist and prescriptive account, the key term used for decor appropriate to it is Veritas - the "real" - because realism, that is, an illusionism which never loses sight of "real" referents in the natural world, visually implies that the social rules governing the house and its decor were themselves natural, given. Yet the surrealism of actual decoration, like some kinds of postmodernism today, paraded precisely the arbitrariness, the pastiche, the ambivalence of the "real". In doing so, Roman houses - again like postmodernism - playfully made explicit the arbitrariness of the rules which governed the social articulation of the house. In effect, the Roman house deconstructs the viewer into one who looks by certain laws (social, visual, ideological) and one who is confronted by images which in different ways expose the way he looks. The viewer simultaneously sees and sees how he sees, that is, sees how what he sees is problematically constructed. It remains to ask what this deconstruction might mean. One reading of \ itruvius's polemic would suggest that this deconstruction was perceived as a threat. We might follow the direction of the moral rhetoric and see Roman houses as 85 ways ok viewing always implicitly subversive. We might inter that the system was always paranoid about the flagrant transparency with which its contradictions were flaunted in every house in the land. However, although \ itruvius is certainly evidence tor ideological constraints, the fact is that even such art as he fails to condemn was deconstructive in the way it undermined view ing and substituted "false" for "real". Other moral rhetoricians go further. The orator Papirius Fabianus is unwilling even to spare "second-style" paintings. His moral outburst against painting is a trulv Platonic condemnation of all imitation: Men even imitate mountains and woods in their foul houses - green fields, seas, rivers amid the smoky darkness .... Who could take pleasure in such debased imitations if he was familiar with the real thing (vera)? . . . Thev pile up great buildings on the seashore and block off bays by filling the deep sea with earth; others divert the sea into ditches and artificial lakes. These people are incapable of enjoying anything real (veris). Their minds are so twisted thev can onlv get pleasure from unnatural imitation of land and sea in the wrong places. (In the F.lder Seneca, Controversiac 2.1.13) By its virulent condemnation, the moral polemic naturalises "unnatural imitation" as quite normal. It is not right, but, like the brothel, it contains the transgressive and subversive inclinations of citizens within an entirely acceptable social sphere. On this reading, the verv deconstruction of some central concepts in Roman culture by the Roman house is perhaps domestic architecture's supreme contribution to upholding the system." Clifford (ieettz has argued that "anv art form . . . renders ordinarv, everyday experience comprehensible by presenting it in terms of acts ami objects which have had their practical consequences removed and been reduced (or, if you prefer, raised) to the level of sheer appearances, where their meaning can be more powerfully articulated and more exactly perceived".92 Roman houses reinforced, represented and made the cultural system comprehensible. By so blatantly exhibiting the contradictions of the cultural rules, they made those rules understandable to every viewer in his or her most intimate space. Thus Roman houses operated on the most generalised social level. As Claude Lcvi-Strauss has suggested of Caduveo body painting, Roman domestic decoration (and its deconstruction of not only the viewer but also the concepts by which social differentiation is constructed) presents in the most private space a subversive myth of what might happen in Roman culture "if its interests and superstitions did not 86 stand in the way".93 Such phantasies of the impossible on viewing and society some cultural level (but often in the arts) are a culture's necessary mythic response to its own strong institutions. Roman houses also had a significant role on the personal as well as on the general level. As with cockfights in Bali in the celebrated discussion of Clifford Ceertz, by playing on the social boundaries of "public" and "private", Roman walls enabled Romans to sec a dimension of their subjectivity. Yet because . . . that subjectivity does not properly exist until it is thus organised, art forms generate and regenerate that very subjectivity they pretend only to display. Quartets, still lifes and cockfights are not merely reflections of a pre-existing sensibility analogi-callv represented; thev are positive agents in the creation and maintenance of such a sensibility.94 In Roman culture, wall-paintings were a primary artefact for generating subjectivity, and Roman viewing was the means by whkh this subjectivity was created. 87 VIEWING AND THE SACRED: PAGAN, CHRISTIAN AND THE VISION OF GOD Classical art history has tended to take the viewer tor granted in its analysis of artistic forms and to assume (without discussion) that the viewing eve is unbounded by the constraints of its time and conceptual framework. This chapter examines a different way of viewing from those implied by the "realism" of \ itruvius or the ekphrases of Philostratus. It represents, broadly speaking, a mystic development of the kind of exegetic viewing which we explored in the Tabula of Cebes. This method of looking at art is rather unfamiliar to most Westerners today, and 1 have called it "mystic viewing". In the case study which occupies the second half of the chapter, I apply the conceptual framework of mystic viewing to the mosaics in the apse of the monaster) of St Catherine at \lt Sinai, an iconographic programme dating from the sixth century. My contention is that the transformation of Roman art from the first century to the sixth is deeply implicated with a transformation in viewing away from naturalist expectations towards the svmbolism inherent in mystic contemplation. The very structure of a work of art as seen by its Christian viewers by the sixth century had become a simulacrum of a spiritual journey which those viewers could be reasonably expected to see themselves as making. This structure is ascertainable from ideas contained in contemporary texts which those viewers max have read or been familiar with from sermons. This is not to imply that such "mystic viewing" had no ideological or political ramifications. On the contrary, in the Christian dispensation, mystic contemplation was an elite ideal by which social and political values were measured, and it served (in its ascetic cognates) as a fierce polemical stick with which to beat ethical "laxity" in daily life. The very indebtedness of the language or mystic viewing and thk contemplation to the language of the late Roman state can be sacred seen in the most important of all Christian mystical writers, Pseudo-Dionysius. The mysticism of Dionysius is expressed in hierarchies (ecclesiastical and celestial) which arc themselves literary reflections of the bureaucratic structures of the empire and the Church. THE MYSTIC MODEL OF VIEWING Viewing entails the relationship of subject and object. This relation is always complex, for we cannot sec the world with objective eves - only with our own. The subject sees the object only in his or her own terms, brings to the viewing his or her own ideologies, narratives, contextualisations. The subject cannot see the object directly without interpretation. There is always a gap between the object as seen (the object in the subject's eyes) and the object out there - the object as it is.' This gap is not merely the product of modern theory. It was well known in the ancient world and forms the subject of one of the most famous anecdotes in Pliny's Natural History, written in the first century a.d. Pliny's story of Parrhasius and Zeuxis is usually used to illustrate the wonders of illu-sionistic art, but I wish to suggest that the story is also a meditation on the gap between image and reality inherent in illusionism. In \atural History 35. 65-6: TraditUT . . . ipse (Parrhasius) dctulisse linteum pictuiu ita vcriiate ut Zeuxis. . . flagitaret tandem remoto linteo ostendi picturam. Parrhasius is said to have displayed a picture of a linen curtain, so realistic that Zeuxis then called to Parrhasius to draw hack the curtain and show the picture. The desire of Zeuxis is not for the image, a linen curtain, but for what the image, the curtain, appears to conceal. lie is deceived by his ga/.e into believing the trompe-Poeil to be real, as his desire (his subjective appropriation of the image he sees) prompts him to penetrate what is behind the curtain. But the Plinian text is more than a narrative of desire - it is the denial of any content to the image and of any object of satisfaction for the subject's desire. The key word, twice repeated, is ulittteum" - which means not only "linen curtain" (the image on the canvas) but also "canvas" (the material on which the Si; ways of viewing image is painted). Pliny uses the word with the meaning of "canvas" in an earlier chapter of the same lx>ok (35.51). The painted image - which offers so much to the viewer's desirous ga/.e - deconstructs into the matter on which it was painted. For the linteum to he withdrawn would leave only the linteum. Linteum is linteum, curtain is canvas. The gap between matter and matter imagined, the object out there and the object constructed by the subject's gaze, could not have been better portrayed. Beneath this Plinian model is an assumption about the nature of subjectivity. There must always be a separation. The dualism of subject and object is inevitable, undeniable and intransgrcssible. But the assumption of dualism is not the onlv model available in the ancient world. This chapter will be about an alternative model, an alternative notion of subjectivity, and the ideology of viewing which it constructs. Mystic viewing is predicated upon the assumption that in mystic experience the dualism of subject and object can be transcended into a unity that is neither subject nor object and vet is simultaneously both." One of the most influential and comprehensive formulations of the mystic model is the l.nne-ads of the third-century a.d. philosopher Plotinus.' A student of Ammonius Saccas in Alexandria in the 230s, Plotinus came to Rome in 244 to teach philosophy. I lis works were collected, edited and published after his death in 270 by his student Porphyry.4 Plotinus insists that the tnvstic vision, the experience of God, is a union of self and other beyond the dual: No doubt we should not speak of seeing; but we cannot help talking in dualities, seen and seer, instead of, boldly, the achievement of unitv. In this seeing, we neither hold an object nor trace distinction; there is no two. The man is changed, no longer himself nor self-belonging; he is merged with the Supreme, sunken into it, one with it.5 (Ennead 6.9.10) This union (which Plotinus calls a "super-intellection" or "a thought transcending thought" - bypernoesis [Enn. 6.8.16.32]) is beyond language (which can merely point the path) and beyond knowing. This is because in knowing, soul or mind abandons its unity; it cannot remain a simplex. Knowing is taking account of things; that accounting is multiple; the mind thus plunging into number and multiplicity departs from unity. Our way then takes us beyond knowing: there may be no wandering from unity; know ing and know able must all be left aside. (Etui. 6.9.4.4) 90 In the experience of union there are no distinctions, there is viewing and the no discriminative awareness and there is no self-consciousness sacred (Enn, 5.3.14. if.; 6.7.35.42!".). There is no recognition of what is taking place: "Soul must see in its own way; this is by coalescence, unification; but in seeking thus to know unity it is prevented by that very unification from recognising what it has found; it cannot distinguish itself from the object of this intuition" (Enn 6.9.3.1 •)• 1" die 'ast chapter of the last Ennead (6.9.11). Plotinus describes mystic union as a state without distinction, without movement, without emotion or desire, without reason or thought and without self. It is a state beyond beauty for the philosopher - "a quiet solitude, in the stillness of his being turning away to nothing and not busy about himself, altogether at rest and having become a kind of rest"/' VIEWING AM) THE SACRED vision, union and the temple sanctuary In Plotinus the process of mystic union is a process of seeing. It is a "self-seeing" in which "the seer is one with the seen" (Entl, 6.9.10-1 1). Here Plotinus, like all Greek mystic thinkers after Plato, borrows heavily from the sight imagery in the great allegories of the cave and the charioteer. Plotinus concludes his last chapter in the sixth Ennead with the following image: The philosopher ascending to union with the One is . . . like a man who enters the sanctuary and leaves behind the statues in the outer shrine. Thev are the first things he looks at when he conies out of the sanctuary, after his contemplation (to endon theamu) and converse there, not with a statue or image, but with the Divine itself; thev are secondary objects of contemplation (deutera tbeamata). That other, perhaps was not a contemplation but another kind of seeing (on iheama alia alios tropOS ton idein), a being out of oneself, a simplifying, a self-surrender, a pressing towards contact, a rest, a sustained thought directed to perfect conformity, it it was a real contemplation of that which was in the sanctuary . . . (6.9.1 i.iHf.) This analogy of temple visiting implies a hierarchy of vision. From the "secondary contemplation" of the images on the outside (to which, in the end, the philosopher must return, as the philosopher-king must go back into the cave in Plato). Plotinus passes inside to the contemplation of the Divine itself, which is "not a contemplation but another kind of 9' ways ok viewing seeing".* I here is an inner movement of "levels" of contemplation paralleled by an outer movement from image to reality. There is a double movement from outside to inside - into the shrine and deeper into the self (to endon theanui) - and yet, simultaneously, there is a movement out of self, a "self-surrender". It is noteworthy that "the other", which one might expect to be described in objective terms, is in fact presented in terms of personal experience ("a being out of oneself, a simplifying, a self-surrender, a pressing towards contact, a rest, a sustained thought") - the very language attempts to enact the unifying of self and other which is the purpose of the Plotinian path. In his use of the image of visiting a temple (which recurs at Etm. 5.1.6.12f.), Plotinus has tied his discourse of mvstic union to the language anil imagery of pilgrimage. The ascent, or ladder, of Plotinian contemplation (which prefigures a vast literature in later Christian thought)1'' thus becomes assimilated to the specifically religious and ritual context of pilgrimage and worship in a sanctuary. Although Plotinus's language is still philosophical, his mystical philosophy is no longer to be separated from religion. It is significant that in the late fifth century at least one Church father should look back explicitly to Plotinus's simile of the sanctuary.10 Although there may have been fundamental theological differences between Christian and pagan in late antiquity, this Christian appropriation of a metaphor about the subjective effects of religious action shows a deep continuity in the experience of personal piety." The specifically religious theme of visiting temples presents in a verv clear form the issue of what the pilgrim is allowed or fit to see. In Plotinus this is "the Divine itself. In a closely parallel context of s.inetuarv visiting from Luripides's play Ion. when the chorus enters for the first time, thev approach Apollo's temple in Delphi, examine the sculpted reliefs and ask of Ion (currently the temple attendant) if they may cross the threshold. Ion replies that thev must sacrifice a sheep and have a question to put to the god. They understand and say: "We are no transgressors of the god's law (tbeou nonion), but are content to delight our eves with the outward beauties of the temple" (v. 230). "Look at everything (theasthe)". Ion replies, "that custom (themh) allows" (v. 232). The setting of the scene is close to the image in Plotinus. The chorus is not a philosopher ascending to union and vision; and so their contemplation must be of the "secondary" or "outward" kind, because thev have not the necessary sacrifice and question - 02 the ritual - by which to proceed to the inner vision. The ptin viewing and the on theott (god), t beast he (see) and tberni (god) (\ v. 230. 232. 234) sacred underlines the problematics of what can and cannot be seen, of the images whose vision is offered (v. 184ft".) anil the god (the cult statue) whose sight is denied. In the Description of Greece by the second-century \.n. traveller Pausanias, there is much temple visiting of precisely the type dramatised by Euripides. The typology of sanctuaries with the outside (and its images) described but the inside either unvicwed or undisclosed (or both) is frequent.12 In all these cases - from the fifth-century b.c. playwright to the third-century a.d. philosopher - the act of viewing or contemplating (theasthai) is associated with man's relation to the Divine (theos) in the ritual context of a sanctuarv. In these three authors the relationship of sight and the sacred in temple visiting occurs through art. The desired and, to the uninitiated, forbidden contemplation is that of the cult statue within the temple - which was in Greek religious custom synonymous with the god himself. The importance of Plotinus is that he combined the religious context of pilgrimage and ritual contemplation of sacred images (a theme sufficiently deep in Classical culture for it to span the seven centuries between Euripides and Pausanias) w ith a philosophical discourse of contemplation which goes back to Plato. divine presence: sight and the sacred in g r a eco-rom a n c i it l r f. Before continuing with our inquiry into mystic vision and sacred images, let us sketch the depth and extent in Graeco-Roman culture of the relation between sight and the sacred. The pun on theos (god) and thea(ma) (sight, contemplation) is frequently employed in Greek to emphasise the presence of the Divine. Its use extends from Platonic philosophy to F.u-ripidean drama, from the religious myths recorded bv Pausanias to Alexandrian Judaism. My intention in the following survey of this pun is to show the centrality of the language of viewing for describing the sacred. I mean thus to emphasise a convention of (iraeco-Roinan thought which is of fundamental importance in providing the basis for the ritual use and mystic viewing of religious images. In the herdsman's speech (vv. 677-774) °f Euripides's last play, The liacchae, the Maenads are described performing a series of miracles, such as suckling wild animals, striking water from rocks and routing the herdsmen and tearing their 93 ways ok viewing cattle to pieces. Then - a deinon tbtama (v. 760) - they defeat the armed villagers, whose weapons have no effect but who are cut to the quick by the women's thyrsi. All this is the work of their tbcos (w. 764, 766). Not only is the thcos-tbea pun operating here to pinpoint the Divine in a passage evoking the effects of religious ecstasy and power, but the whole speech, ironically, is delivered to Pentheus (the disbeliever whose fate - sparagmos - is presaged by the cattle torn to pieces) and Dionysus, the god who is causing the action. The description is a witness to and a warning of the dangers of the theophanic presence (albeit in disguise) of the god on the stage." The pun occurs within a similar context of divine epiphany in Pausanias. In the Description of Greece the theama of a baby turning into a snake causes the Kleans to defeat the Arcadians (6.20). The Cleans subsequently honour the tbcos - the child who became a snake - by building his temple where the snake vanished into the earth. The divinity of the tbcos is dependent on, defined by, the theama of his metamorphosis and disappearance. The god of this place would not have been a tbcos but for the theama he performed. Needless to say, the shrine of this transforming and vanishing god is not to be entered. Even the woman who tends the god is veiled (6.20.3); sight, which was once given as a v isiblc metamorphosis to save Klea and then a disappearance into the ground to deify this site, is forever after denied. But the ground and place of that metamorphosis and disappearance are forever sacred, sanctified by the image which must not be seen. In Judaism, God himself can never be seen.14 Nonetheless, there is an exuberant Jewish tradition of mystic visions, revelations and apocalypses.1" In Philo Judaeus, where this tradition meets the language and thought of the Alexandrian Greeks, there is a rich play on tbcos and theama. on sight and the divine. Philo, a highly cultivated Jew from Alexandria, was born in about 25 b.c. and died around a.d. 50. lie came from a wealthy and prestigious family - his brother Alexander was one of the richest men in the ancient world, and his nephew Tiberius Alexander (having changed his religion) rose to being Roman procurator in Palestine and prefect of Egypt under Nero. Philo himself served as leader of an Alexandrian embassy to the emperor (Tains in a.d. 39-40.16 Philo's surviving work consists of a great number of exegeses on the major books and themes of the Septuagint.1 It is our main testimony to the nature of Hellenistic Judaism. Although there is a lively scholarly debate about whether Philo was himself a practising mystic,18 it is clear that (like his 94 contemporary Paul at i Corinthians 2.6ff. and 3.iff.) he dis- viewing and the tinguished between an initiated elite who arc capable of re- sacred cciving wisdom (sophoi, or leleioi) and the immature who must be kept on a diet of milk.19 Philo identifies passages in his own work which he claims arc comprehensible only to the elite20 and claims that the higher knowledge - which has given him the authority to interpret (which is to say, to allegorize) the Scriptures - comes from his own experience of direct communion with God.21 1 or Philo, sight (opsis) is the key to the religious-philosophic path:2-' "Philosophy was showered down by heaven and received by the human mind, but the guide which brought the two together was sight, for sight was the first to discern the high roads which lead to the upper air"-1' (De Specialibus Legibus 3.185). After describing the ascent of the mind through sight (clearly, the initiate rather than the milk-fed mind), Philo finds that physiological sight is itself inadequate and must be replaced bv philosophic inquiry: The mind, having discovered through the faculty of sight what of itself it was not able to comprehend, did not simply stop short at what it saw, but drawn by its love of know ledge and beauty, and charmed by the marvellous spectacle (tbean) came to the reasonable Conclusion (logismon eikola) that all these were not brought together automatically bv unreasoning forces, but by the mind of (Jod (iheou). (De Specialibus Ixgibus 3.189) The ascent is not merely a philosophical speculation that God made the world, it is an apprehension of God: But those, if such there be, who have had the power to apprehend (katalabein) 1 lim through I limself w ithout the co-operation of am reasoning process to lead them to the sight (tbean), must be regarded as holy and genuine worshippers and friends of God (theophilesin) in very truth. In their company is he who in Hebrew is called Israel, but in our tongue the (Jod-seer (boron theon)1* who sees not His real nature, for that as I said is impossible - but that He is.25 (De Praemus et Poems 43-4) lor our purposes, the supreme Philonic experience of (Jod -what must serve as theophanv in Alexandrian Judaism - is linguistically represented in terms of the language of theos and tbea.26 The conceptual formulation of theophanv in Philo owes much to later Platonic philosophy.-1 It is worth noting in this context a further philosophical, rather than religiouslv theophanic, strand in the use of the tbeos-thea pun. One root of the language of theatna in both Philo and Plotinus (who 95 ways OF VIEWING employs the pun to effect at, for example, Entuad i.6.y or 5.8.10) is Plato. In the Phaedrus, Plato uses this language in the mvth of the charioteer (247 AD). In the Republic the verb tbeastbai recurs several times in the image of the cave (516A and 516!}), and he puns twice on tbeos-tbea (or words from those roots) in the passage which sums up what the allegory means (51715 — theun . . . theos; 517D - apo tbeion theorion). VIEWING AND THF SACRED: SOME conclusions The primacy of sight in constructing the sacred is central. The evidence collected here, although extremely diverse in the periods, places and contexts at which these different writings were produced, points unequivocally to the importance of sight and viewing (or, at the very least, of the language of sight and viewing) as perhaps the primary means for constructing the Other World.28 Whether we give a philosophical, mystical or religious predication to the notion of the Other World does not much matter; what is important is that it is supramundane - outside "ordinary" experience, but certainlv experientiablc to the initiate. All of these texts assume not only the existence of the Other World but also souk possibility of contact or communion between self and this Other World (if most would not accept Plotinus's notion of complete union). In this sense all the evidence which speaks of sight and the sacred is borrowing from a conceptual framework (which I illustrated earlier by reference to Plotinus) where at least in some respect there can be movement out of the dualism of subject and object and into some kind of unity beyond self and other. In fact, the image of movement as ascent out of self, out of the physical and into the spiritual, is common to most of the writers I have used from Plato to Philo, Plotinus and the Christian Fathers.:v For understanding religion in Classical culture, the crucial role of sight in constructing the sacred is of great significance. A statue or other work of art may be as important as or more important than a text in the evidence it can offer to our understanding of the sacred. But we must never forget that we are working here in a profoundly different conceptual framework from the Plinian dualism that underlies the narrative of Parrhasius and Xeuxis. There, desire can never be fulfilled, and the trompc Poeil is the more deceptive the more realistic it appears. In sacred art, not only is it possible for the initiate to see what he or she desires (the epiphany) but in the true depths of mvstic vision the practitioner can lose all sense 06 of the self that did the desiring as well as the object of viewing and the desire.30 It is quite right tor Plotinus to question his imagery sacred at that point (Etmead 6.9.10) - tor when there is no seer and no seen it is perhaps tautologous to speak of "vision". CHRISTIAN THEOPHANY conceptual frameworks In my review of the language of tbeos and then, I have deliberately avoided by far the greatest body of its occurrence -namely in the Patristic writers from Clement and Origen onwards. \1\ intention was to justify the importance of sight imagery in the non-Christian religious culture of the ancient world, before turning to examine a Christian work of art in its Christian framework. One of the continuities between pagan and Christian culture in antiquity is a deep sense of the pervasiveness of the sacred. The holy is constituted through private piety and supremely through the personal vision, the theophany, of a devotee. Such personal piety, which we readily associate with Christian prayer or icon worship, conies out vividly in a passage from the Apolog) of the pagan Apuleius (about a.d. 160): "It is my custom wherever I go to carry with me the image of some god hidden among my books and to pray to him on feast days with offerings of incense and wine and sometimes even of victims" (Apology 63). In Christianity, the image of a personal vision serves as a major metaphor for spiritual ascent in Paul: "Tor now we sec through a glass darkly, but then face to face" (1 Corinthians 13.12). It is the paradigm for Christian conversion in Pauls vision on the road to Damascus (Acts 9.3ft., 22.6ff.).J1 In early monasticism, the ascetic path is conceived as a constant preparation to appear before God.32 In Chapter 10 of I he life of St Antony, a highly influential and popular hagiographv written (perhaps by Athanasius) shortly after its hero's death in 356 and presenting Antony (as he was thereafter to be accepted canonically) as the paradigm of the ascetic "athlete" of Christ, the culmination of the struggle against temptation is a mystic vision of the ray of divine light (aktinaphotos) and a voice from I leaven. I should like to discuss the meaning of theophany and the mvstic vision of God in the Christian tradition in relation to a mature work of art from the sixth century. Whatever qualms about images were felt in the early Church,33 by the mid- 97 ways of viewing sixth century they were all but forgotten amidst a tremendous rise in the cult of icons.34 The theoretical foundation for images as a means of mystic contemplation of the Divine is above all to be found in the works of Pscudo-Dionysius the Areopagite. The Dionysian corpus, which arose around 500 a.d. probably in Monophysitc Syria, was universally believed to be the work of Dionysius the Areopagite - one of Paul's few Athenian converts (Acts 17.34). As such it provided virtually apostolic authority for a mysticism vv hich combined the Neoplatonism of Plotinus with the allegorism of Alexandrian writing from Philo to Origen." Like Plotinus before him, Pscudo-Dionysius is a firm believer in a mystic union that transcends subject and object."' The aim of the Dionysian hierarchy, whether celestial or ecclesiastical, was "as far as possible, assimilation (apho-moiosisf to God and union henosis with him".'* The path to God is through a ladder of images:'" Our human hierarchy, on the other hand, we see filled with the multiplicity of perceptible symbols (symboldft) lifting us upward hierarchically until we are brought as far as we can he into the unity of deification dpi tot haioeide theosin) ... it is by perceptible images (cikosin) that we are uplifted as far as we can be to the contemplation of the divine (epi las theias theorias). (De Ecclesiasticu Hierarehia 1.2 |/Y; 3.373AB]) Here contemplation of the Divine is the end of the same hierarchy of symbols as is unification itself. The one is the other. Although visual images arc not specified as the means of ascent,4" they are certainly not excluded.41 Pseudo-Diony-sius follows the Plotinian theme of deutera theamata (f'.iiu. 6.9.11), images which are sympathetic to the "nature of soul" and so attract the soul most easily into a link with the god who is present in the image (Enn. 4.3.11).4~ Within a generation or two after the appearance of the Dionysian corpus, Hvpatius of Lphcsus - archbishop from c. a.d. 531-8 and one of Justinian's most trusted theological advisers - was justifying images in language reflective of the very thought of Dionysius: W c allow even material adornment in the sanctuaries . . . because we permit each order of the faithful to he guided and led up to the divine being in a manner appropriate to it [the order) because we think that some people arc guided even by these Imaterial decorations] towards intelligible beauty and from the abundant light in the sanctuaries to the intelligible and immaterial light.41 98 The path of Pseudo-Dionysius through images and sym- viewing and itie hols, like those of Plotinus and Philo, transcends the activity SACRED of the human mind and know ledge itself. It partakes, like all mystic thought after Plato, of the imagery of light: But as for now, what happens is this. We use whatever appropriate symbols we can for the things of God. With these analogies we are-raised upward towards the truth of the mind's vision Icpi ten ton noeton tbeanuiton aletbeian), a truth which is simple and one. We leave behind us all our notions of the divine. We call a hall to the activities of our minds and, to the extent that is proper, we approach the ray (aktina) which transcends being. I lere, in a manner no words can describe, preexisted all the goals of all knowledge and it is of a kind that neither intelligence nor speech can lav hold of it nor can it at all be contemplated since it surpasses everything and is wholly beyond our capacity to know it.44 (De Divinis Nomittibus 1.4 [JPG k^CD]) No less than in Plotinus, the Christian ascent of Dionysius is a movement from material symbols to a spiritual reality - the "truth of the mind's vision" - where subject and object are united, are "simple and one". Language, knowledge, the activity of mind and being itself are transcended in the ray of divine light. This theory, of which Pseudo-Dionysius was the most influential exemplar, is presented as mystical anil apolitical. In fact, insofar as it has a politics, this appears as an injunction to keep the mysteries of divine union secret from the uninitiated.45 Yet its ideological impact was all the more potent for its secrecy. The monastic elite in the Christian empire, who were also largely the literary elite, not only formed opinion but also established the very terms of discourse (including social and political discourse) through their prolific writings. Moreover, as theological positions developed, these became inseparable from other kinds of political and ideological stances. The very choice to cast the Christian theology of unification in terms of hierarchies of union reveals the impossibility of separating the language of the state's social agenda in this world from a discussion purporting to be about access to the Other World.46 the mosaic at sinai The Christian-Neoplatonic theory of mystic union with God through vision formed an underlying conceptual framework to the ways viewers looked at sacred art. Just as life could be 99 ways of viewing a spiritual path to closer union with (Joel, so the forms of religious paintings can he constructed as a hierarchy leading to a vision of such union. Let us apply this Pseudo-Dionv sian framework to the iconography and meaning of a specific image from the sixth century. The mosaic of the Transfiguration in the church of the Monastery of St Catherine at Mt Sinai (Plate 2; Figures 13 and 17) is the major Justinianic image in an imperial monastic foundation built by Justinian, which merits a relatively lengthy description in the panegyric about the emperor's buildings written by Procopius, l)e Aedificiis (5.8.1—9).4 Procopius tells us: On this Mt Sinai live monks whose life is a kind of careful rehearsal of death, and they enjoy without fear the solitude which is vcrv precious to them. Since these monks had nothing to crave - for they are superior to all human desires and have no interest in possessing anything or in earing for their bodies, nor do they seek pleasure in any other thing whatever - the Fmperor Justinian built them a church which he dedicated to the Mother of God, so that they might be enabled to pass their lives therein praying and holding services. (5.8.4—5) For the monks, the Sinai Transfiguration was both the major decoration of the apse in which their services took place and the main intercessory receptacle for their prayers. If the life which is presented as "a kind of careful rehearsal of death" is a life of prayer and liturgy, the focus of that life at Sinai was the mosaic which I explore.4* Until the erection of the iconostasis and the cross which surmounts it in the seventeenth century, the attention of a worshipper entering the church at Sinai would have been directed down the line of the nave to the fully visible mosaics in the conch of the apse and the wall above.49 Just as the mosaic is the centre of the liturgical and intcrccssorv functions of the church, so it is its visual focus. The eve of the beholder is arrested by the gaze of (Christ, who looks directly out of the apse at the congregation (sec Plate 2 and Figure 17). The image of the Transfiguration refers to a storv recorded in all three synoptic Gospels;'" this is Matthew \ version: Jesus taketh Peter, James and John his brother, and bringeth them up into an high mountain apart, and was transfigured before them: and his face did shine as the sun. and his raiment was white as the light. And, behold, there appeared unto them Moses and I'.lias talking with him. (17.1—3) In the mosaic representation, Jesus stands in the middle of a blue mandorla on the gold ground, with Flijah on his right 100 and Moses on his left (Figure 17). John is kneeling beside F.lijah anil beside Moses is James, while Peter is splayed beneath Christ awaking from sleep in accordance with the text in Luke 9.32 (Figure 18). Around the image is a frame of medallion busts - in the arch the twelve apostles (with Luke, Mark ami Matthias replacing Peter, James and John, who arc presumably unnecessarv because they already appear in the main mosaic); and at the base of the apse-conch is a group of prophets from the Old Testament, with Longinus, the abbot ot Sinai, represented in a square halo in the right-hand corner (from the viewer's jx>int of view) and the deacon John in the left-hand corner." Above the apse-conch is a triumphal arch with two riving angels presenting orbs to the Lamb of God in the centre (Figure 13). Beneath the angels at the bottom corners of the arch are a pair of roundels without inscriptions which Kurt \\ eit/.mann identifies, probably correctly, as John the Baptist (to the viewer's left) and the Virgin (to the right).52 This would make the scene a proto-decsis. Above the triumphal arch are two further representational panels. That on the viewer's left shows Moses loosening his sandals before the burning bush (Figure 15), that on the right depicts Moses in a viewing and the sacred Figure 13. Apse ami triumphal arch mosaics, Monastery of St Catherine at Ml Sinai, sixth century a.d. Photo: (iourtesv of the Michigan-Princcton-Alexandria Expedition to Mt Sinai. ■ m 101 ways of viewing cleft of rock at the peak of Sinai receiving the tablets of the Law (Figure 16). The Sinai apse has been properly studied only in recent years, after the University of Michigan—Princeton University expedition to Sinai led by George Forsyth and Kurt Weitz-niann in 1958, when the mosaic underwent restoration and was for the first time adequately photographed." The only extensive discussion of the monument is in the work of Weitz-mann. I le believes that "the artist's great sensitivity" points to the "highest artistic skill by the leading artist of an eminent atelier". \\ eitzmann concludes that "only Constantinople could have supplied such highly trained craftsmen"/4 He suggests that "a learned cleric . . . advised the artist on the accuracy of the iconography ami supplied him with the basic and, in this case, manifold ideas". Taking a hint from this suggestion, I will consider the iconography and thematic structure of the Sinai mosaics in the light of the patristic textual commentaries likely to have been available to those who made the mosaics and those who looked at them. This method has not to my knowledge been attempted before in a systematic way as a means of interpreting and understanding particular monuments of Byzantine art. My aim is to provide an interpretation of the mosaics according to contemporary - that is, sixth-century - ideology, rather than one grounded in modern critical categories. Sinai is a very special place (Figure 14). It was on the peak above the monastery that Moses received the tablets of the Law (Exodus 34; Procopius, De Aedificiix 5.8.8). The monastery's rather awkward location in a valley is a result of its including the site of the burning bush (Fxod. 3. if.) at the eastern end of the church.*■ On the mountain next to Sinai is the traditional site of Elijah's cave in Horeb, where the prophet fled from King Ahab (1 Kings 19.8!".). Not surprisingly, Sinai was a centre of pilgrimage from very early times - all these sites are described and located with their appropriate biblical narrative in a surviving fragment of the pilgrim journal of Fgeria, a woman, probably from Gaul, who visited the Fast between a.d. 381 and 384.Yet despite the sanctity of the place, we can still generalise from Sinai, for its very importance makes it a paradigm of the modes of worship and exegesis which became so pervasive in Christian antiquity. The Transfiguration - the supreme thcophany of the New Testament barring only the Resurrection - was a theme ide- 102 ally suited to Sinai. Its narrative included Sinai's two prophets, Moses and Elijah. Furthermore, Sinai was a place specifically associated with theophanv - the site where Moses conversed with God tor forty days and nights (F.xod. 24.16-18); the site where Moses, standing in a "cleft of rock", saw the "hack parts" of (lod but did not see his face (Exod. 33.1 3-23), and the site from which when Moses descended out of the mountain "the children of Israel saw the face of Moses, that the skin of Moses' face shone" (Exod. 34.35). Likewise in the story of Klijah, the prophet stayed in Ilorcb for forty days and nights, spoke with God and was commanded to stand upon the mount and watch the Lord pass by (1 Kings 19.8-14). In the sixth-century Sinai mosaic, Moses, Klijah and the three apostles in the image - as well as all the worshippers in the church - were confronted with a theophanv of God as Christ the Incarnate Son, whom in the new covenant following the Incarnation all could see not merely from the "back parts" but rather "face to face". Before investigating the Transfiguration itself - the quintessential Christian tbeama of tbcos, when the divine nature of Christ appeared through his human nature' - it is worth looking at the Moses theme of the two panels sited above the apse on the church's eastern wall.58 So far as I know, no commentator has discussed these panels save to note them in passing. Yet Moses as the paradigm of what can be achieved in the mystic ascent is a central subject of both Christian and Jewish mystic writing.''9 The Sinai programme takes two of the most important moments in the Kxodus narrative of Moses - the viewing and the sacred Figure 14. The Monaster) of St Catherine at Mt Sinai, from the east. Photo: Robin Cormack, Conway Library. GoaitanM Institute. 103 ways of viewing two which arc associated with the site of the monastery and, at the same time, with the theme of theophany. I shall examine the Sinai images of Moses in relation to one of the most important and influential contemplative works of the Patristic era, Gregory of Nyssa's Ik Vita Moysis.60 Gregory (c. 330—95 a.d. ), was a member of a remarkable (christian family - his sister St Macrina and his brother St Basil the (ireai were among the most important figures in later fourth-centurv monasti-cism. The/> Vita Moysis, which anticipates Pseudo-Dionysius in its blend of Alexandrian exegesis (owing much to Philo and Origen) and Neoplatonic philosophy (on the model of Ploti-nus), has been regarded as the culmination of Gregory's mysticism.61 In later centuries, Gregory's work was much read and highly regarded. By the eighth century, he was "universally celebrated as the 'Father of Fathers' "/>: the burning bush In the panel above the triumphal arch on the viewers left, which illustrates Fxodus 3.1-6, Moses is facing right and removing his sandals before the burning bush (Figure 15). He is looking up at the hand of God which is raised in blessing in the top right-hand corner. The scene is not unique to Sinai -it occurs, for instance, in the roughly contemporary mosaics of the church of San Vitale in Ravenna.63 For Gregory of Nyssa (De Vita Moysis 1.20) the event of the burning bush is an "awe-inspiring theophany" (p bo be ran . . . tbeophaneian) and "a strange sight". It is an illumination of the prophet's senses (both sight and hearing [1.20]) as well as of his soul (2.20) by the grace of the rays of light from the bush. As a result. Moses is empowered by the vision of God (dynamdtheis tei opbtheisei theophaneiai) to release his countrymen from bondage in Egypt (1.21). Gregory's spiritual commentary on the episode is still more revealing: "It is upon us who continue in this quiet and peaceful course of life that the truth will shine, illuminating the eyes of the soul with its own rays. This truth, which was then manifested bv the ineffable and mysterious illumination which came to Moses, is God" (2.19). Thus the burning bush (the bush still flowering in Sinai during the sixth century in its little court behind the eastern end of the church)64 is a representation of, a sign for, what is possible for "us" (whether we be pilgrim worshippers or ascetic monks at Sinai "whose quiet and peaceful course of life" is, as Pro-copius put it, "a kind of careful rehearsal of death"). That the bush burned for Moses with a light which was God (2.20-1) 104 is a sign that the contemplative path of the contemporary viewing and the worshipper can also shine w ith divine light (2.2(1). sacred What then does the mosaic sign at Sinai mean? For Gregory, the burning bush is a type of the Incarnation (De Vita Moysis 2.20) and of the Virgin birth (2.21).61 It is a miracle in nature that implies the mysteries of the Faith. But it is also a prescription for the spiritual path: That light teaches us what we must do to stand within the rays of the true light. Sandaled feet cannot ascend the height where the light of truth is seen, hut the dead and earthly covering of skins, which was placed around our nature at the beginning when we were found naked because of disohedianee to the divine will, must he removed from the feet ot" the soul. When we do this the knowledge of the truth will result and manifest itself. (2.22) What Moses learned by the light of the theophany was more than a mere vision, it was that "none of the things w hich are apprehended bv sense perception and contemplated by the understanding reallv subsist, but that the transcendent essence and cause of the universe, on which even thing depends, alone subsists" (De Vita Moysis 2.24). In an exegesis loaded with Neoplatonic images and terminology, Gregorv expounds the apprehension and discrimination through mvs-tic theophany of "real Being . . . that is, what possesses existence in its own nature" and "non-being . . . that is, what is existence only in appearance, w ith no self-subsisting nature" (2.2 3).66 The Mosaic theophany of the burning bush - the call for Moses to enter his vocation as liberator and prophet (Exodus 3.7b) as well as his confrontation with the Name of God (Exodus 3.14-15) - is a vision of the senses (De Vita Moysis 1.20) that is an education for the soul (2.20); it is a vision of physical light that is a step on the ascent to the metaphysical - the divine - light. The event becomes, in Gregory's propaganda for spiritual asccsis, simultaneously a proof, a paradigm and a prescription for mystic contemplation: In the same way that Moses on that occasion attained to this knowledge, so now does evervone who, like him, divests himself of the earthly covering and looks to the light shining from the bramble hush, that is to the radiance which shines upon us through this thorny flesh and which is (as the Gospel says) the true light anil the truth itself. (De Vita Moysis 2.26) As Christian (and hence initiate) viewers - be they pilgrims or monks - stood in worship before the image, before the very bush which the image represents, the panel informed 105 ways of viewing Figure 15. Mosaic of Moses before the burning bush, Monastery of St Catherine at \li Sinai, sixth century a.d. Photo: Courtesy of the Michigan-Princeton-Alcxandria Expedition to Mt Sinai. v i f. w i n c. a n i) 1 h f. SACRED Figure 16. Mosaic of Moses receiving the tablets of the Law, Monastery of St Catherine at Mt Sinai, sixth century \.n. Photo: Courtesj of the Michigan-Princcton-Alcxandria Expedition to Mt Sinai. 10- ways of viewing them of how far they could still go on the mystic path. Truly to see this image was to see what it proclaimed - the "true light which lighted) every man that cometh into the world" (John i .9). Ultimately, the end of the Christian ascent and of Gregory's Christian exegesis of the burning bush is the image of Christ transfigured in his divine nature which occupies the apse. For ultimately, "1 am the way, the truth and the life: No man can come unto the Father, but by me" (John 14.6). on the summit of sinai: the tablets of the law If the message of the burning bush could prove so complex in its implications for the viewer, this was still more the case with its sister panel of Moses receiving the tablets of the Law (figure 16). It is on the viewers right-hand side above the apse, and it represents the prophet facing left in the middle of a vast cleft of rock where he is receiving (with covered hands) the tablets of the Law from the hand of God in the top left corner. This scene, which like that of the burning bush is relatively common in early Christian art (it also appears at San Vitale in Ravenna, for instance), is narrated in F.xodus 33.21-34.6.67 For Grcgorv, the ascent - the attaining to the summit of Sinai - is an entry into the "inner sanctuary of the divine mystical doctrine" {Ik Vita Moysis 1.46). It is here that Moses confronts what "transcends all cognitive thought and representation and cannot be likened to anything which is know n" (1.47). In a still "higher initiation", the prophet is granted the v ision of the tabernacle - "the inaccessible and unapproachable holy of holies" (1.49). "Having surpassed himself by the aid of the mystical doctrines", he emerges from the theophany "to deliver the Laws" (1.56). The ascent of Sinai is the supreme symbol of the spiritual path - a way to knowledge (2.154, '58)1 ;1 purification (2.154—7), a "progress in virtue" (2.153) ana< 8 contemplation of God: "The know ledge of ( k)d is a mountain steep indeed and difficult to climb - the major-itv of people scarcely reach its base. If one were a Moses, he would ascend higher and hear the sound of trumpets" (2.158). The path is not for the multitude (cf. Ik Vita Moysis 2.150-61).6K It is for those who have progressed (whether as pilgrim traveller or monastic ascetic) to the foot of Mt Sinai, into the church half-wav up the mountain, to these mosaic images. If 108 the image of the burning bush stood for the call to climb the viewing and thi spiritual mountain, then the icon of Moses at the peak receiv- sacred ing the tablets of the Law represented the summit of this ascent. Together the two panels stood for the beginning and the goal of the path that took viewers to Sinai. The) signified what being at Sinai meant, what it was to see God. The Moses panels proclaimed the viewers' condition as initiate climbers on a spiritual ascent. To be in the presence of the Sinai images at the site of the burning bush and the Hebrew camp in the Wilderness, is to be called to proceed up the mountain of divine ascent into the "luminous darkness" (2.163), into "the mysterious darkness of unknowing" (Pseudo-I)ionysius, Ik Mystica Theologia 1.3 \PG 3. iooo.\]).w To be so privileged, so advanced as to view these images, is to be under the spiritual obligation of the Mosaic epektasis - the constant spiritual progress that Moses supremely represents in the Gre-gorian exposition: "lie [Cod] would not have shown himself to his servant if the sight were such as to bring the desire of the beholder to an end, since the true sight of God consists in this, that one who looks up to God never ceases in that desire" (Ik \ ha . I toy sis 2.233). Iconographically, the feet of the climber who has ascended Sinai are bare, for sandals (the encumberment of "dead skins") are "an impediment to the ascent" (I)e Vita Moysis 2.201). 0 In Gregory's exegesis, "he who has progressed this far through the ascents we have contemplated carries in his hand the tables, written by God which contain the divine law" (2.202). In Sinai, Moses is represented at the moment of receiving the Law. The result of this act, in (iregory's exegesis, is transliguration - "Moses was transformed to such a degree of glorv that the mortal eve could not behold him" (2.217; Exod. 34.29L). The theme of the transfiguration of Moses prefigures that of Christ in the Sinai apse. For Gregory, the receiving of the tablets is the cue for a series of Christological comparisons. The carving of the new "tables of human nature" (2.216) after the repentance of the I Iebrews for their idolatry (2.203) foreshadows the Virgin birth (2.216). As the "restorer of our broken nature ", who "restored the broken table of our nature . . . to its original beauty" (2.217), Moses prefigures Christ. Thus the patristic readings of the Old Testament narrative already tie it to the coming of Christ. In the Sinai mosaics, this link is made explicit and visual by crowning the entire programme with the Transfiguration. Gregory sees Moses transfigured as a result of receiving the 109 ways of viewing Law (Ik Vita Moysis 2.21-; I.xod. 34.29L) "In his surpassing glorv he becomes inaccessible to these who would look upon him. For in truth as the Gospel saws [of the Second Coming: Matthew 25.31] when he shall come in glory escorted by all the angels, he is scarcely bearable and visible to the righteous" (2.217-18). That the righteous who had come to Sinai in the sixth century could sec the icons of Moses and Christ transfigured was a sign of their righteousness, of the possibility that they - like Moses on Sinai - could become assimilated entirely to God. As Pseudo-Dionysius put it in Ik Tbeologia Mystica (1.3 [ = PG 3.1001 A]), Moses on Sinai breaks aw av from what sees ami is seen, and he plunges into the truly mysterious darkness of unknowing. Here, renouncing all that the mind may conceive, wrapped entirely in the intangible and invisible, he belongs completely to him who is beyond everything. Here, being neither oneself nor something else, one is supremely united by a completely unknow ing activity of all know ledge, and knows beyond the mind by knowing nothing. 1 The lawgiver is depicted in the midst of a cavernous cleft of rock on the mountain peak. Weit/mann believes this images to be influenced by the scenery at Sinai.~: For St Gregory such an explanation would not be sufficient. In response to the prophet's request to "shew me thy glory" (Exod. 33.18), God tells Moses "it shall come to pass, while my glory passeth by, that I will put thee in a clift of rock and will cover thee with my hand while I pass by. And I will take away mine hand and thou shalt see my back parts: but my face shall not be seen" (Exod. 33.22-3; Ik Vita Moysis 2.219ft'.). After an attack on any literal interpretation of this passage (Ik Vita Moysis 2.222-3), Gregory expounds its significance. In attaining to the cleft of rock and the vision of the "back parts" of God, Moses demonstrates the upward thrust of the soul's rise through its desire for God (2.224-6). There is no limit to this ascent, to the steps on the "ladder which God set up" (2.227). Thc whole of Moses's life becomes a paradigm for the ascent of this ladder (2.228-30). In a reversal of the narrative order in Exodus, higher still than his shining with glory, is the thirst for a further vision of God - despite the fact that it was precisely with the vision of God that Moses "constantly filled himself to capacity" (2.230). The request to w hich the cleft of rock is the fulfilment is the archetype of Gregorian spiritual desire, ' of the Gregorian path itself (see especially 2.239). 4 The cleft of rock is of immense significance, for the promise of which it is part is 1 10 "more magnificent and loftier than even thcophany which viewing and the had previously been granted to his great servant" (2. -41). For sacred Gregory, the rock is Christ (2.244, -4^) ar>d a host of other "prizes" which come in attaining the "crow n of righteousness" (2.246). The rock encapsulates, through a series of quotations in Gregory's text, a totality which may stand for the whole Law both old and new (2.247-51). It is a totality which, not only for Gregory but also in the Sinai mosaic, is fulfilled only "when the lord who spoke to Moses came to fulfil his own law" (2.251), that is, bv Christ. The vision represented in the panel of Moses at the summit of Sinai is the ultimate revelation vouchsafed to humanity before the Incarnation. Even before we turn to investigate in detail the image of the Transfiguration in the apse-conch, it is clear that the Sinai programme offers us a hierarchy of images, more specifically, a hierarchy of theophanics: 1. The burning bush (Figure 15) = The call to prophetic ascent. 2. Receiving the Law (Figure 16) = The ultimate summit of mystic vision. 3. The Transfiguration (Figure 17) = The new ultimate that took place in the Incarnation - the seeing of God as Christ "face to face". None of the iconography is incidental: The details (such as the bared feet or the cleft of rock) are essential to the significance of what is being represented. "The viewer is being taken through a hierarchy of images which represents the ladder of his or her own spiritual path as monk or pilgrim towards the vision that encapsulates and fulfils all. THE TRANSFIGURATION The very structure of the image works as a simulacrum for the viewers ow n spiritual journey. The spiritual progress of Moses, in the narrative which led from the burning bush to the giving of the Law, is parallel to the physical progress of the pilgrim who climbs Mt Sinai from the monastery. Whereas the images of Moses represent a hierarchy of spiritual ascent, the image of the Transfiguration may be read as a paradigm of what is revealed in the Christian version of spiritual contemplation - that is, the vision of the incarnate Christ. Moreover, as I shall argue, the particular iconographi-cal and scriptural details selected for the Sinai Transfiguration emphasise not only the vision (Christ transfigured) but also 1 1 1 uws OF VIEWING the process of looking itself. The image of the Transfiguration becomes a visual exegesis and exemplum of the act of spiritual viewing. In discussing the meanings of the Transfiguration mosaic (Plate 2; Figure 17), I shall employ the same method as I used in the Moses scenes. After isolating features of narrative or iconography in the image, I shall examine them in the light of the patristic textual commentaries likely to have been available to those who made the mosaic and those who looked at it. ' In this way I hope to provide an interpretation of the image according to contemporary ideology before I attempt to integrate this into a general discussion of the programme of the Sinai apse as a whole. In the first step of isolating the principal narrative and iconographic features, I am in debt to the structuralist methods of biblical analysis developed bv Edmund Leach. 6 One significant feature of the Transfiguration narrative in all three synoptic Gospels (Matt. 17.1; Mark 9.2; Luke 9.28) as w ell as in much of the patristic commentary, is the mountain (Tabor) on which the Transfiguration took place. Weitz-mann notes that one unusual element in the Sinai representation is the absence of any indication of Tabor. He is right insofar as the Sinai mosaic includes no depiction of a mountain, but the green strip at the base of the image surely represents the peak of the mountain where the actual theoph-anv took place. Some later representations of the theme, such as the eleventh-century mosaic in the Nea Moni, also show only a green strip for the peak, although it is true that in general the later iconography docs include an image of the whole mountain. Thus we could explain the iconographic rarity bv resort to a functionalist argument. We could look to the fact that Sinai is a monumental mosaic (like the Nea Moni) in a relativelv unusual position (the apse - the only other apse Transfiguration is in the Church of Sant' Apollinarc in Classe near Ravenna), which did not allow room for a representation of the full mountain. The spiritual point, however, is surely that this icon is already on a mountain, as is the viewer who is in its presence. If we take the significance of "mountain" to be not simplv literal but symbolic, then it is entirely natural for this specific image to concentrate on the peak. One of the ligural meanings of the notion of mountain in early Christian and Jewish writing seems to be to provide a setting for the penetration of this world by the Other World. s As Leach remarks, Sinai is especially significant because this mountain is within the wilderness that is itself 112 already "marked off as altogether Other": in effect, "the mountain of God, Ml Sinai ... is ... a world apart within a world apart". v Ideologically, to be at this place at all is to have ascended the mountain with Peter, James and John. Practically, in terms of the difficulties of travel in the sixth century, Sinai was isolated and remote. The group of people with access to this image on this mountain were in effect either pilgrims or monks - that is, the chosen. Excluding any image of the whole mountain (which would be a distancing sign for the viewer because it implies being away at a distance), the Sinai image seems to be going out of its way to include its worshippers in the theophany on the summit. There is no hierarchy of levels within the image (unlike the vast majority of later representations of the theme, where the apostles are in a different, lower, plane from the transfigured Christ and the prophets)/" on the contrary, the Sinai mosaic collapses all the figures onto the same plane/1 Just as the chosen apostles (the viewers of the Transfiguration) are included in the mosaic - in the vision which they view - so by implication the viewers of the Sinai apse may be included in the mystic experience of the vision that they are offered through this image/ \ i e wing and the sacred FigUK 17. Apse mosaic of the Transfiguration, Monastery of St Catherine at Mt Sinai, sixth century a.D. Photo: Court es \ of the Michigan-Princcton-Alcxandria Expedition to Mt Sinai. 113 ways ok viewing A second unusual feature of the Transfiguration iconography at Sinai is the representation of Peter awaking from sleep directly beneath the transfigured Christ (Figure 18). This is in accordance with the text of Luke 9.32, as Weit/.mann notes.83 The vast majority of later instances of the iconography of this scene show Peter addressing Christ and standing to one side beneath him (in accordance with the texts in Matt. 17.4; Mark 9.5; Luke 9.33).84 That the designers of the Sinai programme should have chosen this moment out of the sacred narrative, unlike all those who chose the iconographies for so main later images of the Transfiguration, seems particularly significant. Luke's version of the Transfiguration story is designed as a specific contrast to his version of the Agony in the Garden on the Mount of Olives (22.39-46) where the apostles who accompany Christ cannot remain awake despite his injunction to them to pray. As the patristic commentators noted, it was important that the still imperfect disciples should have witnessed at least once in their lifetimes the manifestation of Christ's divinity in his human nature.83 The significance of the metaphor of sleeping ami waking to the theme of mystic vision is of central importance. For Philo anil Clement of Alexandria, the difference between the initiate and the multitude, the "virtuous man" and the "many ignoble and idle souls", lies precisely in the fact that the former is awake and the latter in a deep sleep.86 In the Gospels, waking from sleep is a metaphor for waking from death - itself perhaps an image for waking into the true vision of God.8' More importantly, the language of waking (gregorein - in the Authorised Version, "watching") characterises a group of Christ's most important recorded savings, which refer specifically to preparing oneself to receive God. This language is echoed in Paul and Revelation.88 As Ambrose of Milan (339—97) asserts, it was while the apostles were awake that "they saw his majesty, for no one can see the glory of Christ unless he stay awake".89 In the specifically monastic context of Sinai, it is worth referring to a text written at the monastery by one of its most famous and important saints, less than a century after the erection of the mosaic. In steps 19 and 20 of the luidder of Divine Ascent (I'd S8.937A-945A), John Climacus is specifically concerned with sleep as a hindrance to prayer. He writes "we must struggle with sleep" (19.2) and concludes "this is the twentieth step. He who has mounted it [that is, conquered sleep to attain spiritual vigil] has received light in his heart" (20.20). In effect, in an influential contemplative work written at Sinai shortly alter the mosaic was made, the conquest of sleep is tied to the vision of light - just as in the mosaic (and in its Lukan prototype) the waking from sleep is tied to the light of the Transfiguration. In one strand of interpretation by the commentators, sleep is "that great heaviness which came upon them [the apostles] from the vision", the weighing down or putting to sleep of the bodily senses by the "incomprehensible splendour of divinity**.90 This refers to a particular mystical exegesis of the notion of sleep which figures with some prominence in (ireg-ory of Nvssa (whose Christian name means "Awake").7' In one sense the challenge to the initiate is to wake from sleep to the true light (/// Canticum Canticorum, Horn. 11. PG44.996A-997B), but in another it is to put the body to sleep so as to let the soul alone enjoy the spiritual vision. Hence, interpreting this quotation from the $OHg of Songs (5.2) "I sleep, but my heart wakcth", Gregory writes: "Thus the soul, enjoying alone the contemplation (tbcoriu) of Being, will not awake for anything that arouses sensual pleasure. After lulling to sleep every bodily motion, it receives the vision of God in a divine wakefulness with pure and naked intuition" (/// Cent. Cunt, limn. 10, PG 44.993D). In the Sinai image, not only the VIEWING and thf. sacred figure iS. Detail of apse mosaic of the Transfiguration, St Peter waking from sleep, Monastery of St Catherine at Mt Sinai, sixth century a.D. Photo: Courtesy of the Miehigan-IVineeton-Alexatulria Expedition to Mt Sinai. "5 w \ ys of viewing waking but also the sleeping itself is a figure for the reception of the vision of Christ. The image at Sinai offers its viewers not only a vision of Christ in his divine nature "face to face" (the fulfilment of the Sinai tic visions of Flijah and Moses, of their Old Covenant), it also offers a representation of what it is to view. The icon is programmatic in that it points out to the monastic audience exactly what they must do to attain the vision of Cod. They must wake from sleep with Peter in order to ascend to the light. As ClimaCUS assures his readers, at the peak of the Ladder of Divine Ascent is a vision - the vision that Jacob saw (30.36), the vision that John describes (30.31): And now, finally, after all that we have said, there remain these three that hind and secure union of all, faith, hope ami love; and the greatest of these is love, for (iod is himself so called. And, as far as I can make out, 1 see the one as a ray, the second as a light, the third as a circle; and in all one radiance and one splendour . . . (PC 88 115 31)—1161B) For Pscudo-Dionysius, this light is what the mystic initiate may partake in - what the viewer may experience just as the apostles experienced the Transfiguration: We shall be fulfilled with his visible theophanv in hols contemplations, and it shall shine round about us with radiant beams of glory just as of old it once shone round the disciples at the divine Transfiguration. And then, with our mind made free of passion and spiritual, we shall participate in a special illumination from him. and in a union that transcends our mental faculties. There, amidst the blinding, blissful impulses of his dazzling ravs we shall be made like to the heavenly intelligences in a more divine manner than at present. (DeDiv. Norn. 1.4, PG 3.592 (.') In fully confronting the Transfiguration of Christ, the viewer is himself transfigured. This light, which is so essential both to the visual language of mystic contemplation and to the exegesis of the Transfiguration, is represented in several ways in the mosaic.'''2 In the first place, there is Christ's mandorla, which represents the "bright cloud" out of which came the voice saving, "This is my beloved son, in whom I am well pleased; hear ve him" (Matt. 17.5; Mark 0.7; Luke 0.34-5).''" Second, there arc the rays - of which one descends upon Christ from the cross above, while the seven others emanate from Christ illuminating the space and figures around. Just as light was essential to the Mosaic vision of the burning bush and transfiguration to Moses' vision on the peak of Sinai, so the Transfiguration of 116 Christ is the conflation anil culmination of the two Mosaic viewing and the experiences uniting light with transfiguration - the one who sacred is transfigured being the one who is the light/'4 The Church Fathers are particularly rich in their comments on the light."'1 For Gregory of Nazianzus (c. 330-89), light is an image for the ladder from God (who is light) to the angels to man (pbos), encapsulating also the "beacon fires of the heavens", the commandments of God to man, the burning bush ami the pillar of fire in Fxodus, the chariot of fire by which Elijah ascended to heaven, the star at the Nativity, the Transfiguration, the conversion of Paul, the sacrament of baptism and Heaven itself.V6 For Cyril of Alexandria (died 444) - offering less a summary of the entire faith through the image of light than an image of purification through vision -the mind which sees the Transfiguration "is transformed into a certain elect and godly radiance so that even its garments [i.e., I take it, its clothing of flesh| are illumined in the beams of that light, and they too seem to flash with light".'' It is already clear that this segment of Christ's narrative history was so written, so represented and so read as to transcend or rather encapsulate the entire history.w The sleeping apostles (Luke 9.32) presage the Agony in the (Jar-den and hence the Passion (Luke 22.45-6). The end of the Transfiguration narrative in Matthew (17.9) and Mark (9.9-10) is explicit in referring forward to the Resurrection and the Ascension. The voice from the cloud is the same voice from the same cloud speaking the same words as at the Baptism (compare, e.g.. Matt. 17.5 with Matt. 4.17).''"' Insofar as the Transfiguration at Sinai is specifically partnered by the Old Testament images of the visions of Moses, its particular focussing of the Christological narrative centres on the issue of viewing, for the Sinai image is both the representation of a vision and at the same time an image, a paradigm, of what it is to view a v ision. Viewing at Sinai is a matter of ascent - our ascent to this church half-way up the sacred mountain, Moses' ascent to the burning bush in this very place and to the summit later on, and the apostles' ascent "into an high mountain apart" (Matt. 17,1). Viewing is to wake from the sleep of the many - the waking of the apostles to the reality of Christ's divinity and our own awakening to the mystic union with the light of Transfiguration promised by Pseudo-Dionysius. And viewing is to see - after the ascent and the waking there is the vision of the light, the vision for which Moses constantly strove in Gregory of Nyssa's account and which is offered "7 ways of viewing to the apostles and the prophets in the image, and to the congregation on the floor of the church beneath them, "face to face" with God as Christ the Incarnate Son. The Sinai apse is not merely the image of a vision, it is - for the initiate and participant worshipper - the vision itself. some conclusions on the significance of the iconographic programme at sinai When we take the narrative scenes at Sinai together, the two Mosaic panels clearly represent a development, an ascent towards the full vision of the transfigured Lord in the apse. In spatial terms this is a spiral (see figure 19), moving from the image of the burning bush through the panel of Moses with the tablets (both on the wall above the apse) down into the apse-conch itself (which opens out beneath the wall). If this backwards movement were continued, it would leatl to the actual bush that burned for Moses, which in the sixth Century was believed to be still llovvering in a court behind the apse. Topographically, the development of this spiral moves from the burning bush (the site of the church) to the peak of Sinai (where the tablets were delivered) to Christ in the church (again at the site of the bush). The movement is from the bush on the slope of the mountain, to the peak of the mountain to a scene meant to be on the peak of the mountain but with no mountain represented, a scene that is in this church (at Sinai rather than Tabor, although it was at Tabor), a scene that is every- and anywhere at all and any time (although it is of course quite specifically here and now at this viewing time for anv viewer). In short, there is a development from specific events specifically valid at Sinai to a general event (despite its historicity), generally and eternally valid throughout Christendom. The image of the transfigured Christ is, simultaneously, a paradigm for what is seen in mystic vision, a proof for what was seen (bv those who ascended the spiritual mountain, like Moses, Elijah anil the apostles), an exhortation to the Christian generally of what can be seen if the spiritual mountain is ascended and a prescription of what initiate Christians present in this church at this time ought to see by virtue of their having come here at all to see it. The faculty of sight becomes in this context not merely a means for constructing the sense world but, more significantly, a way of constructing the path to the Divine (through light and contemplation) and a con- 11S struction of the Divine itself (incarnate through light into the viewing and the sense world). sacred The mosaics at Sinai arc an exegesis in their own right. They show a particular, a unique, selection and combination of events (i.e., of Scripture) as well as of the patristic commentaries on those events. They serve to create their own commentary, their own ladder of visions through images (parallel to such ladders as that of John Climacus, as the Mosaic tpektasis of Gregory of Nyssa anil as the hierarchies of Pseudo-Dionysius). Insofar as the Transfiguration icon is self-reflexive - representing within itself (in the three waking apostles) an image of how it should itself he seen - the image is an exegesis in the visual medium of images of what it is to view, of the act and process of mystic viewing. Finally, in its representation of the Incarnate Christ (who is and sums up all these things), it is a visual exegesis of (rod. It is worth remarking here that the mosaic goes out of its way to include representations (in the apse-conch and the medallions on its rim) of all four Gospel writers. The image inscribes the texts on which it is predicated and of which it is an exegesis. It represents them (symbolised by their authors) as well as being a representation of them; and it represents Moses - the great exegete and writer of the Old Testament whose prophecies Christ had come to fulfil.100 In exegetic terms images can do what texts cannot. The "simple" programme of Sinai - which can be completely taken in within a single glance if one is rightly positioned in the church - can simultaneously encapsulate, enact, emboih . activate and make visible and viewable a whole body of texts. The instantaneous, non-diachronic nature of the image (what should perhaps be called its icotticity) collapses the totality of in; WAYS OF VIEWING these narratives and narratives about narratives into a single sacred space and time. That time is here and HOW at this mountain top where (for Moses and Elijah) all narratives collapsed into the confrontation with deity, which we are offered here "face to face". The spiritual ideology of this iconic ascent is ultimately to give up the narrator and narrative constructions, which are forever dependent on the dualism of subject (the narrator) and object (the narrated). The call at Sinai is to abandon all of this and to confront what is beyond narrative, text and knowledge, beyond the self that does the narrating and the knowing. Ultimately the Sinai apse is an image of (and an aid to) the heliosis and iheósis described by Plotinus and Pseudo-Dionysius which is the experience of union with Godhead beyond the splitting of self and other. Ultimately this iconic presentation and this iconic style is the very antithesis of the Plinian naturalism with which we began. For that was art predicated upon a dualist conceptual framework, and this image assumes not onlv the possibility but also the imperative to go beyond dualism to union. This union of self and other, which is the key to the conceptual framework both pagan and Christian that underlies mystic vision, cannot be described (because it is beyond the dualism of observer and observed, describer and described), but it can be experienced. Spiritual union is thus an inexplicable ideal. One should add that it was an ideal available only to a very narrow elite of spiritual practitioners (women, probably, as well as men). Moreover, it could be appropriated (because of its very unenunciability) to all kinds of political ends whose manipulative aims were hardly in keeping with the final claim of eliminating the dualit v of subject and object. Cyril of Alexandria, for instance, one of the early Church's most ruthless politicians, was a major theologian of that most spiritual of Christian sacraments, the eucharist.101 In Sinai the theophanic ladder of images takes the viewer to Christ, represented in the apse programme both as the transfigured Lord and as the eucharistie lamb (in the centre of the triumphal arch above the apse-conch, adored bv two angels; see Figure 13)-1 **~ The lamb above the sanctuary is Christ, the sacrifice that saved all humanit v. It stands also for the eucharist which Christ enjoined upon his followers at the Last Supper before his Passion.103 In liturgical terms it would be precisely during the enactment of the liturgy whose culmination is the eucharist that the viewer would stand before the 120 mosaics in the Sinai apse. Thus the lamb represents what is viewing and the happening now in the church. It stands tor the sacrifice of sacked Christ bv whom we are eternally saved and for the continuously re-enacted sacrifice of the eucharist by which we are dailv to be fed. If mystic vision is the attainment of union with Christ beyond the distinction of subject and object, the eucharist is precisely the means by which such union may be achieved. For what we eat and drink is His Body and I lis Blood and thus "we come to bear Christ in us; because His Bodv and Blood arc diffused through our members; thus it is that, according to the blessed Peter, 'we become partakers of the divine nature' [2 Peter 1.4]" (Cyril of Jerusalem, died 386, Mystagogical Catechesis 4.3, PC 33.1100A). In partaking of the eucharist, the viewer is transformed. For Pseudo-Dionvsius, the eucharist no less than mystic vision is a "participation in Christ" (De Caelesti Hicrarcbia 1.3. PC 3.124A) which "grants us communion and union (koitwnia km henosis) with the One" (De Ecclesiastica Hierarchia 3.1. PC 3.434D). Ultimately, as in mystic vision, this union is "illumination", "the passing on of the light of Cod to the initiates" (De Ecc. I Her. 3.1. PC 3.425AB). Significantly, repeating Plo-tinus's simile of the outer images and the inner sanctuary (De Ecc. llier. 3.3.2. PC 3.428C; cf. Ennead 6.9.11), Dionysius presents the eucharist as a matter of direct vision and light: "Show yourself clearly to my gaze. Kill the eyes of our mind with a unifying and unveiled light" (De Ecc. Hier. 3.3.2.). By the late fifth century, the sacramental vision of Christian Ncoplatonism revealed the liturgy and the eucharist to be operating in precisely the same way as sacred images."*4 The liturgy of the eucharist, performed before the images of the lamb and of Christ transfigured, is the ultimate ritual act of mediation between the worshipper at Sinai and the God of whom the worshipper is the living image (Genesis 1.26-7). 'l 's revealing that in the mosaic, apart from the transfigured Christ, none of the figures within the narrative scenes addresses the viewer with a frontal gaze. The narrative panels are paradigmatic and prescriptive but not intercessory in the way that frontal images and icons are. I low ever, John the Baptist and the Virgin Marv on the triumphal arch and the apostles and prophets in the rim of medallions around the apse-conch all engage the viewer frontally. There is a genealogy of intercession here. The movement is from the prophets represented beneath Christ (David, of whose seed he was born [Romans i. 13], is in the significant position immediately below him) through Christ himself to the apostles in the 121 ways OF vii wing medallions on the rim. John the Baptist, the "forerunner", and the Virgin (the immediate predecessors of Christ) arc specially emphasised by being placed in the triumphal arch. Christ himself is the supreme link - uniting the intercessional signification of the mosaic with its theophanic programme. It is significant that just as Moses stands upon the rock, so Christ is above St Peter. The line of apostolic succession and mediation with the holy passes through Christ to the apostles but also to the rock (Peter) of Christ's Church (John 21.15f0. The very enactment of the liturgy and the cucharist (Christ's living Church and Body) in the presence of the worshipper is the final stage of this development, for the worshippers at Sinai are Christ's living Church. Peter's waking from sleep beneath Christ is thus not simply an image for the possibility of theophanic viewing but also of the living Church as awake and present to the living Christ. Ultimately, the cardinal image on which the whole scheme at Sinai is pivoted is the lamb adored on the triumphal arch. For it is to Christ, "the lamb of Cod which taketh away the sins of the world" (John 1.19), that the theophanic ladder of Moses and Flijah ascends. It is in the lamb as cucharist that the congregation are saved and partake of Christ. The lamb as the Body which the worshipper eats is Christ in us. The union of self and other to which these images are attempting to direct the viewer is in the end eucharistic. For the eucharis-tic lamb represents Christ in us and us in Christ; it is the ritual intercession of the saints for us with Christ and of Christ for us w ith Cod. Through the figure of Christ in the Transfiguration run two genealogical lines. One looks up from David to the "seed of David" Christ (transfigured in his own lifetime) to the lamb (Christ for us now). One looks down from the cross at the peak of the medallion rim (whose three-shaded blue ground svmboliscs the I loly Trinity)10* through Christ (the second Person of the Trinity) to St Peter (the rock of the Church which is the embodiment and operation of Christ for us now). In this, the thematic centrality of the lamb, as bodily link with Christ (the eucharistic breatl) and the sacrifice which saved man, is paramount. Finally, if we read the triumphal arch as a decsis, the lamb represents Christ as supreme intercessor, between and above the Baptist and the Virgin Mary. Again, the lamb is the focus for the intercessional imagery in the church. I have tried consistently to emphasise the contemporary significance of the mosaic for its sixth-century viewers. It is a 122 material object, embodying and refining a complex textual viewing and the exegesis, which served as an ideal and a paradigm. Again, one sacred may note that however splendid an object and an ideal it might be, it could serve as a pawn in many games. I or instance, the present is inscribed in the Sinai programme not only by its dedicatory inscription but also by the appearance of images of the abbot Longinus and the deacon John. These men may have been laudable clerics, honoured by their peers in these images in recognition for their generous patronage. They may equally have been using the mosaic - not only its visual splendour but even its spiritual message - to serve ends of their own in the complex world of Byzantine ecclesiastical politics. We cannot know. But certainly the Sinai apse was not an abstruse compilation of obscure texts about the distant past. It was made and intended as an image for the present, with a message for the present, representing Christ as the eternal God-man (the second Person of the Trinity), as the sacrificial-eucharistic lamb and as the Sinaitic vision proffered to Moses tw ice, to Klijah, to the apostles at Tabor and to the congregation here and now in this church.""' SOME CONCLUSIONS This chapter began as a discussion of a model of perception which I have called mystic viewing where the duality of self and other is transcended, or rather, unified, in a vision of the sacred. After tracing the history (of at least the language) of the relationship between sight and the sacred, I have tried to show the ways in which mystic viewing can work in relation to an important Christian programme of images dating from the sixth century a.d. I believe the Sinai apse to be a fundamental I v different object in its conceptual underpinnings from the "realist" Pompeian wall-paintings discussed in Chapter 2 and from the illusionist panel-paintings described by Philostratus and explored in Chapter i. In Sinai the very structure of the work of art serves as a parallel for the spiritual path of its viewers. Neither its style nor the way it \\ as viewed can be separated from the implications of the spiritual journey it represented. The extraordinary unity of style, theme and viewing in mature early Christian art is radically different from the contradictions, desires and trompe Votil of Roman domestic decoration. Where the transgressions of the decor of the Roman house allowed the consumer a place from which to assess critically his or her social location, the erosion ways of viewing of the possibilities for a transgressiv c ga/.e in Christianity undoubtedly could be interpreted as a coercive move.1" I have not discussed either style or form at length, although it is in terms of these that the difference between "Classical naturalism" and "Medieval abstraction" is usually seen. I want to suggest that style was dependent on how an object was conceived and how it was viewed. The key to an understanding of the transformation of Roman art - its shift from illusionistic verisimilitude to abstraction - is not in the observable formalism alone but also in the underlying conceptions, views and ideologies of the people who commissioned, who made and who looked at works of art. In the specific arena of early IJv/.antine art with which 1 conclude, 1 hope to have shown at least one thing. An image need not be merely an illustration of a text but may be (as the Sinai apse is) a "text" in its own right - a particular polysemic arrangement and commentary that demands to be read within the ideology of its time and in its own unique way. By the sixth century, Byzantium was an "exegetic" culture, in which every event, text and image could be read as an exegesis of the one fundamental and real event - namely, the Incarnation as represented by the narrative of Christ's life and Passion.1"* Every part of that narrative implied the totality just as every part of God's creation implied the whole: Are not two sparrow s sold for a farthing? and not one of them shall fall on the ground without your Father. (Matt. 10.29) '^4 chapter 4 VIEWING AND IDENTITY: THE TRAVELS OF PAUSANIAS; OR, A GREEK PILGRIM IN THE ROMAN WORLD Tie mere act of enumeration . . . has a power of enchantment all its own. Michel Foucault1 Monuments, houses and works of art help to construct a collective sense of subjectivity within culture. We have seen, in the case of Roman houses, some of the complexities and inversions through which this process of creating subjectivity operates. Likewise, religious objects, which require initiation anil exegesis to be understood, work to form a rather narrower shared subjectivity within an exclusive group of "believers". We have seen such religious viewing both in the pagan context of the Tabula of Cebes and in the mature Christian images of Sinai. The kinds of viewing which Roman houses and Byzantine mosaics generate are remarkably different, but they share the strategy of inscribing the beholder into a complex social context, into a subjectivity which is - whether broadly or exclusively -shared with other members of the culture. However, monuments may work not only on the collective level by propagating generally accepted sensibilities. They may also work on the individual level, helping an individual to construct a private and more personal sense of self. Images and monuments embody a history. They affirm the past and the liv ing presence of the past in the present-dav. They are the most visually potent assertion of a cultures relationship with its past and hence are a paramount cultural mechanism for evoking the historicity of identity - for grounding collective subjectivity in a historical valorization.2 Of course, works of art (including literature and music) are not the only group of artefacts that support identity in this way through their evocation of a sense of history. But they are perhaps the most powerful of such bulwarks precisely jzjzt i25 ways of viewing because they arc privileged and special relies. They embody the authority of a cultural canon. Individuals are caught between the shared subjectivity of collective tastes in the present, this historical identity embodied in the visual environment, and the particular accidents of their own personal upbringing and prejudices. The result of this conflict may be an acute sense of how works of art ami the material remains from the past, which imbue the modernity of any culture, seem to pull one's own self-definition and identity in different and even in conflicting directions. The identities generated by different kinds of viewing - the broader formulations shared by Graeco-Roman culture as a whole and the narrower demands of religious initiation, for instance - may therefore be in contradiction in a single person. A trulv splendid example of such a conflict of possible identities (deeply related to and explicitly explored in terms of ancient monuments and works of art) is the Description of Greece written bv the traveller Pausanias in the second century a.d. Different groups of viewers, at different times or at the same time, may see the same thing differently (depending on their starting points, assumptions and identities). But in Pausanias we find a single individual, more obsessed with ancient art than almost anyone writing after him, offering a multiplicity of view ings which themselves hint at the rather different strands in his sense of identity. I lere, in a discourse which enumerates the most worthy sights of Greece, we can trace for a single person a number of ways of viewing which may even conflict with each other. In exploring Pausanias, this chapter examines the largest single collection of images in any text surviving from antiquity. Pausanias provides a huge range of references to art -from cult images carefully described, like Phidias's ivory-and-gold statue of Athena Parthenos (1.24.5D, to frescoes extensively discussed, in particular, Polygnotuss famous paintings on Homeric subjects in Delphi (10.25-31). There are numerous stories alniut the miraculous powers of images, such as the image of the horse of Phonnis at Olympia, which sends real stallions mad with desire (5.27.3) or an image of Artemis Orthia that needs human blood to prevent it from becoming very heavy (3.16.of.). There are also large numbers of images grouped according to different categories, for instance, all the images of Zeus from Olympia (5.22.1-24,11), as well as quantities of temples and sanctuaries. My aim, however, is not to exploit Pausanias's richness as 126 a source for particular images and descriptions but rather to viewing and explore his formidable evidence as to horrtancc of particular places and the enactment of particular rituals. Between the major centres, such as Athens and Megara (1.39.4!.), the road passes through many minor stops and outlying areas - Oropus (134), a diversion to the islands (1.35-6), Fleusis (1.38—9). All these belong to Athens (1.39.3), w hereas the Megaris is marked as different - independent of the Athenians (1.39.4) and its "neighbour" (6.19.12). The text itself marks the boundary firmly with a sentence that rounds Athens off: "Such in my opinion are the most famous legends and sights among the Athenians, and from the beginning my narrative has picked out of much material the things that deserve to be recorded" (1.39.3). These borders, as felt by the traveller on the actual land and as announced to the reader by the text, arc crucial. What they mark arc not merely lines on a map but boundaries and thresholds in the experience of Greece. They delimit places nor 136 simply topographically but as locations of culture, of race, of VIEWING and identity. This is why the borders of districts so frequently identity coincide with the ends of the books in Pausanias's periegesis. Even il the land is continuous and the traveller walks the same road from one area to another, the text marks a boundary that is felt. At 2.38.7 (the end of Book 1) and 3.1.1, stone figures of Hermes signal the borders of Argolis and Laconia, as well as the division of Books 2 and 3. At 3.10.6, reference to the same figures of Hermes marks a further boundarv - that between the narrative myth-history with which Pausanias prefaces his account of Laconia and the actual travelogue. The effect of the phenomenology is to present Pausanias's text as a mirror of Greece. The major centres (political anil sacred) and the movement between centres imitates the condition of Greece as a land of many poleis, a multiplicity of conflicting and often contradictory identities. The text imitates Greece as it moves from place to place. And vet the totality of Pausanias's narrative has totalized Greece, has brought all the separate hellenika into one Greece. The act of travelling and the parallel act of writing, "the nomination of the visible", actually undermine the multiplicitv and diversity which the text w ishes to emphasise. Greece becomes an ordering and cohering of the many hellenika into one image -one individual's image - a Greece defined by its otherness to other ethnographies and above all to Rome. The very conflicts of the hellenika as tirelessly repeated in the myths and histories of internecine war become a cohesive factor, a shared myth, that brings them together against the Other of non-Greece, which is to sav Rome. The divisions of Greece themselves become the definition of a unified identity, a past when it was possible to be divided before the present of integration within a larger and dominant whole. STRUCTURING GREECE: THE MYTHOLOGY OF PLACE The actor-centred pattern implies both a personal view and the assumption that one's land must be experienced through such a personal view in order to be understood. Implicit here is an emphasis on geography as a mode of identity, on the subjective and affective qualities of place. The investigation of iden-titv can be seen as the core of Pausanias's text. He uses a constant cross-referencing of myths and histories to bring places together. Despite the apparently extempore presenta- '37 ways of viewing tion, the fact is that Pausanias rigourously avoids repeating stories and frequently refers us forwards or backwards in his text, for example, at Troc/.en he tells us that "it was here they say . . . that Heracles dragged up the Hound of Hell . . . and as far as the so-called I lound of I [ell, I will give me views in another place" (2.31.2). That place is Tacnarum in Laconia (3.25.6), where Pausanias gives us some rationalizing as to what the Hound of Hell really was. Did he know at 1 roe/en what he would say at Tacnarum? The extent and quantity of such cross-references certainly imply a careful and intricate web - a deliberate structural device - that unites Pausanias's Greece on a quite different and vet complementary level to the experiential travelogue. Such stories tie the many bdlarika into a single whole through myth-history. They provide readers with what they need to know - an identity, a meaning - by drawing on the general knowledge of a broad mythology of Greece which Pausanias assumes is his readers' cultural background.4' Pausanias takes great pains to get his myth-historical interpretations right and complete. His care for accuracy here is not different (despite the comments of some modern critics)44 from his painstaking care to provide precise topographies. For instance, in the account of the history of Laconia with which he prefaces Book 3, he makes sure to tie up any loose ends by referring us elsewhere in his text. We are promised an account of the retirement and exile of Polycaon son of Lclex (3.1.1) anil given this at 4.1.1-2.1. "As to the cause of the (Messc-nian) war, the Lacedaimonian version differs from the Messe-nian. The accounts given by the belligerents, and the manner in which the war ended" arc promised "later in mv narrative" (3.3.1 and 3.5). The promise is fulfilled at 4.5. if. By contrast, Pausanias has already told us how after the death of Cleomcncs "the Spartans ceased to be ruled by kings" at 2.«;. 1-3, and he refers us there instead of repeating himself (3.6.9). In effect the text enacts not only a journey through topography but also a careful myth-historical interpretation of the meaning of that topography. This interpretation darts in anil out of the travelogue structure, reorganising the narrative of monuments and localities according to a pattern not of geography but of mythology - an ideological pattern whereby identity, having already been located by place, is further defined by story. And vet the myths themselves arc locational. De-spite the detailed cross-referring (which ties the whole of Greece into a single whole through its stories), the actual .38 myths occur topographically according to the traveller's pro- viewing and cessual account. Their importance is not primarily as narra- identity tives hut as the narrative valorization, the historical canonization, of a particular place within the totality of "Greece" as a cultural construct.4' The tact that Pausanias is so alive to his mvths, that he remembers and cross-references them so assiduously, shows their paramount importance to his notion of "(ireece". I Ie makes sense of his Greece through them. It is revealing, given how strong is Pausanias's experiential bias, that most of his contextualising stories plunge us into a past that was distant even in his own time.4'1 The identity which his text evokes is a myth-historical identity grounded in the past. It is as if, in a modern Knglish novel, readers were being asked to respond to their vividly felt Saxon or Norman origins. A clear instance of the pastness even of present identity in Pausanias is the case of the Corinthians, whose city was laid waste by Mummius after the war with the Achaean League in 146 b.c. and was refounded only in 44 b.c. as a Roman colony by Julius Caesar (2.1.2). Yet, despite its new population and Roman credentials, what is interesting to Pausanias about Corinth are its ancient (pre-Roman) associations and sights which are themselves explained to Pausanias bv the present-day Corinthians. By virtue of being in that place, these people have become "Greek" according to the Pausanian definition; the place itself has imparted its truth, its identity, to them. Hence the stories recounted of the loci of Corinth are about Artemis (2.3.2) and Medea (2.3.6!.), Bellerophontes (2.4. if.) and the ancient history of the Corinthian kings (2.4.3-4). More powerful still as a myth of identity and location is the extended history of the Messenians (4.1-29). This is a tale of losing ones native land, of exile and of eventual return. It describes "the many sufferings of the Messenians, how fate scattered them to the ends of the earth, far from the Pelopon-nese, and afterwards brought them safely home to their own country (oikeian)" (4.29.13). Identity here depends on more than geography - it is a myth partly of race, partly of dialect and above all of displacement and return, but it is focussed on locality in the sense of loyalty to the traditions of one's home: The wanderings of the Messenians outside the Peloponnese lasted almost mo v ears, during which it is clear that thev did not depart in any way from their local customs {ta oikothen), and did not lose their Doric dialect, but even to our day have preserved the purest Doric in the Peloponnese. (4.27.11) ways of viewing I Jut this fierce loyalty was not enough. Only when they returned to the full identity of being in their proper place could the Mcssenians return to being fully Greek: I was exceedingly surprised to learn that while the Mcssenians were in exile from the Peloponnese, their luck at the Olympic Games failed. For . . . we know of no Mcssenian, either from Sicih or from NaupactUS, who won a victory at Olvmpia . . . However, when the Mcssenians came back to the Peloponnese, their luck at the Olympic Games came with them. (6.2.10) Identity transcends place. It is competitive in the pan-1 lellenic tradition of the Gaines, lint success, a nation's place in the Games (and in the pan-I lellenic tradition), is itself dependent on the correctness of locality. The Mcssenians are not truly themselves until their return. This return is more than merely a replacement in the right place; it proves to be a return to their form as fully Greek, as competitors and victors in the pan-Hellenic Games. Deeply implicated in this nostalgic sense of identity is the repeated theme of autochthon) - of peoples being born from the soil they inhabit.4 This recurs in most of the books of Pausanias's pericgesis from Frichthonius (1.2.6) to the Locrians (10.38.3).48 Despite the fact that identities can change, it is the earliest traceable link between a people and their environment that Pausanias is most keen to record: "The Stymphali-ans arc no longer included among the Arcadians, but are numbered with the Argive league, which they joined of their own accord. That they are by race Arcadians is testified bv the verses of Homer . . ." (8.22.1). It is an ur-past, an autochthonus or at least I Iomerically sanctified past, to which Pausanias is looking.4'' So his Stymphalians (despite their later choices and actions) are located bang in the middle of Arcadia not simply by race or by the authority of the poetic canon, but by the very structure of Pausanias's own account (where they occupy a place in the middle of Book 8). IDENTITY PAST, IDENTITY PRESENT: PAUSANIAS AND THE ROMANS Clearly, in looking to the past for a Greek identity, Pausanias was avoiding the present. The present was the Roman Empire under Hadrian, Antoninus Pius and Marcus Aurelius. The present was a Greece that served, at best, as a culturally influential but otherw isc not especially significant province in 140 a huge system whose centres of power anil influence were VIEWING and located elsewhere. Like the Jews, the Greeks were an ancient identity and independent people whose relations with their Roman riders were deeply ambivalent/" In a passage which seems to anticipate many of Pausanias's concerns, Plutarch put the problem from the Greek point of view: The statesman, while making his native state readily obedient to its sovereigns, must not further humble it; nor when the leg has been fettered, go on and subject the neck to the yoke, as some do who, by referring everything great or small to the sovereigns, bring tin reproach of slaver) to their country. (Praeceptu Gcrtndae Reipublkae 814F) The continuous need to balance obedience with limited freedom made for long-term complications in the attitude of Greeks towards their Roman rulers. Like those of Plutarch.'1 Pausanias's relations with the Romans were, not surprisingly, complex.'-1 In his description, Pausanias completely ignores significant monuments that he must have seen, such as the great charioteer group commemorating Marcus Agrippa which stood by the entrance of the Propylaea at Athens (which Pausanias describes at length at 1.22.4!.) or the temple of Rome and Augustus which was placed on the Acropolis right in front of the Parthenon's east entrance.1' On the other hand, he is generous in acknowledging some major Roman building programmes, such as Hadrian's temple of Olympian Zeus in Athens (1. i8.6f.).14 Romans may be paradigms of virtue and piety like I Iadrian ("a benefactor to all w horn he rules" [1.3.2], "who was extremely religious in the respect he paid to the deity, and contributed much to the happiness of his various subjects" [1.5.5]) or Antoninus Pius ("a most religious man" with all kinds of virtues [8.43.1-5]). But they may also be exemplars of evil - for instance, the impious Sulla (a man whose "mad outrages against the Greek cities and gods of the Greeks" [1.20.7] are punished by "the most foul of diseases" [9.33.6]) or Nero and Caligula (9.27.4!.). The Romans may offer freedom to Greeks by liberating particular cities, such as Mothonc (4.35.3), Pallantium (8.43.1) and Llateia (10.34.2), or even to the whole nation, when Nero "gave to the Roman people the very prosperous island of Sardinia in exchange for Greece, and then bestowed upon the latter complete freedom" (8.17.3). But this very act of bestowal is proof of who is master, and it can be reversed: '4' WAYS OF VIEWING The Greeks however were not to profit hv this gift. For in the reign of Vespasian, the next emperor after Nero, they became embroiled in a civil war. Vespasian ordered that they should again pay tribute and be subject to a governor, saying that the Greek people had forgotten how to be free. (8.17.3). This is a deep condemnation - all the more bitter because freedom is inherently part of the (ireck nature and identity in the Pausanian myth-history; it is precisely this freedom which Roman domination had eroded. In fact Pausanias's myth of a total Greece is supremely a mvth of how "all Greece (pasa Hellas) won independence and freedom", to quote the inscription which Pausanias quotes from the statue of Kpaminondas (9.15.6). It is a myth of how the biographies of a lew great men, from Miltiades to Philopoemen, transcended the identities of their local loyalties to become "benefactors of all Greece (koine tes Ilellados)": "Those who before Miltiades accomplished brilliant deeds, Codrus the son of Melanthus, Polydorus the Spartan, Aristo-menes the Messenian, and all the rest, will be seen to have helped each his own country and not Greece as a whole (ouk athroan ten hellada)" (8.52.1). This is a fascinating passage. Not only is there an unashamedly moral slant to the pan-Hellenic picture, a praise for motivations that transcend locality, but this shades into an explicitly moral emphasis on the way lives ought to be lived. Pausanias specifically excludes "from being called benefactors of Greece" his namesake Pausanias and Aristeides, the victors of Plataca, because of their subsequent transgressions as well as the participants in the Reloponnesian war against Athens whom he bills as "murderers, almost wreckers, of Greece" (8.52.2-3). There is, furthermore, an inevitable sense of decline and fall. As he is in Plutarch,51 Philopoemen was the last in the roll-call of the great, and he was already involved in wars with the Romans (8.50-1). It is as if, historically, the myth of Greece must be defined, delimited and ended by its proximity to Rome. Yet it is precisely the conquest of Greece by Rome - the ending of Greek independence and am dream of real freedom for the country - which constitutes the possibility for the mvth of a free Greece once long ago. Greece can only be one whole when it is subservient to an external state, a Macedon or a Rome. Greece is "Greece'' (one country and not many poleis) only because it is a province in an empire whose various cities are united through having lost their freedom. In the new dispensation they can be granted freedom by an omnipotent emperor from elsewhere. Pausanias's backward-looking 142 project oipanta ta hellenika is itself conditioned by its Roman viewing and context in being pan-Hellenic, in doing the whole of Greece, identity because "Greece" can exist only when its freedom not to be united is over and the myth of a frecdom-in-the-past has begun, lor the Pausanian project to be possible, all the places (whose stories and sanctities the author so carefully enumerates) must no longer be free and at war (as they were in the myth that Pausanias retails), but must be united by and within a larger power. The very attempt to invent and justify a myth-history of "Greece" is simultaneously the evidence for its defeat: Greece can exist only in the invention, in the myth of Rome. The condition for the Pausanian paicgesis of (irecce is that the Greece which his periegesis describes no longer exists. Only when we begin to appreciate the head-on clash of identities, the complexity and incongruity of conflicting paradigms from past and present reiterated through Pausanias's narrative, which together in their tension created Pausanias's Greece, can we begin to grasp some of the ironies that lie hidden in his text. At Sikvon he notes drily: "The precinct . . . devoted to Roman emperors was once the house of the tyrant Cleon" (2.8.1). The viewing of this temple is ironic in the extreme, and the effect is heightened w hen Pausanias launches immediately into the story of how Aratus liberated Sikvon and Corinth from tyranny (2.8.2-6). In the Argive Ileraion, he notices "statues . . . of various heroes, including Orestes. They sav that Orestes is the one with the inscription, that it represents the Emperor Augustus" (2.17.3)! Only rarely does Pausanias reject inscriptional evidence for mere hearsay.5* In both these cases the conflict, the ambiguity, of past ami present as the) clash in the identity of the viewer emerges as irony. In all such instances it is the viewer's identity itself which is at stake in the act of interpreting a work of art. One way out of the impasse of socio-historical identities is to look for a self which is outside history, beyond the decline of Pausanias's beloved Greece into a Roman fief. As we saw earlier, at the heart of Pausanias's ideology of location is the theme of sacred centres. If the political path of the traveller is fraught with identity conflicts which seem to become increasingly less resoluble the longer the journey and text go on, then perhaps religious pilgrimage is the solution to the identity crisis of second-century Greece. The traveller turned pilgrim is no longer in search of a political or historical past that is denied by the present; he seeks rather a sanctified ■43 ways of viewing present past whose sacredness has pervaded these places since the beginning, despite history. PA ISA XI AS AS PILGRIM: IDENTITY AM) THE SACRED Pilgrimage was an important aspect ot the religious culture of pagan antiquity. We know of many examples of individuals and groups going to sanctuaries to consult deities, seek healing, or venerate relics.' But, with the exception of Pausanias, we possess no text from the pagan world which recounts the process of pilgrimage as a personal journey. I lere the contrast between antiquity and the Christian tradition of travel writing is stark. It gives Pausanias's text a unique cultural significance not only as testimony to a specifically pagan form and view of pilgrimage but also as a counterpoint to later Christian writing. One can see the totality of Pausanias's account as a pilgrimage lasting many years. w Certainly it has elements of a transformative rite de passage in which writer and perhaps reader are changed by their confrontation with the sacred identity of Greece.59 Pausanias himself comments on his personal transformation after retelling the myth of how Rhea deceived Cronus: When I began to write mv history I was inclined to count these legends as foolishness, but on getting as far as Arcadia I grew to hold a more thoughtful view of them, which is this. In the days of old, those (ireeks who were considered wise spoke their savings not straight out but in riddles, and so the legends about Cronus 1 conjectured to be one sort of Greek wisdom. In matters of divinity, therefore, 1 shall adopt the received tradition. (8.K.3). It is significant that this change of attitude relates to "matters of divinity" and that it marks a shift from rationalistic literalism (the secularist's response to the sacred) to a greater openness towards allegory and metaphor as methods of intuiting religious truth/'" It is in the more specific descriptions, however, that one-can elucidate more directly the pilgrimage elements of Pau-sanian viewing. Let us take the journey to Eleusis, which is not only a highly charged centre of mystery initiation but is also marked out by Pausanias himself as one of the two supreme sites of Greece: "On nothing does heaven bestow more care than on the Eleusinian rites and the Olympic '44 (James" (5.10.1). Pausanias was himself an initiate into the viewing and Fleusinian in\ stcries.61 Although the text only reaches identity FJeusis at 1.38, we have been prepared for its importance by the discussion of the Elcusinium at Athens at 1143 and the reference to initiation at the F.Icusinian mysteries at 1.37.4. Because "a vision in a dream" prevented Pausanias from describing the contents of the Athenian Fleusinium, the reader is already prepared for FJeusis being religiously special. This specialncss is marked by the very topography the moment the text arrives at Fleusis: "The streams called Kheiti are rivers . . . sacred to the Maid and to Demeter, and only the priests of these goddesses are permitted to catch fish in them. Anciently, I learn, the streams were the boundaries between the land of the Fleusinians and that of the other Athenians" (1.38.1). The geography here is itself sacred -marking an ancient boundary, a liminal threshold between the political world of Attica and the Other World, Fleusis, on its periphery. The ancient political settlement bears out this otherness: "The Fleusinians were to have independent control of the mvstcrics, but in all other things were to Ik* subject to the Athenians" (1.38.3). Not only in space but also in administration we are being prepared for something altogether Other. The text now proceeds through a number of shrines and temples, and their myths, until it reaches the sacred enclosure itself. Here Pausanias surprises us: "My dream forbade the description of the things within the walls of the sanctuary, and the uninitiated are of course not permitted to learn that which they are prevented from seeing" (1.38.7). Having set Fleusis up as a world apart - Attic but different, bounded by sacred boundaries within Attica and administered by a separate order within Athenian order - instead of describing this Other, or bringing his reader through "the walls of the sanctuary" into its inner sanctum, Pausanias's text dramatises the otherness of Fleusis in a supreme way. I le denies its describabil-itv within his own discourse. No mark of otherness is so effective as this statement that the truth of Fleusis cannot Ik constrained in the act of w riting, the act to which the text itself is in perpetual debt. I lere, in a radical about-face, Pausanias, who has constantly been the readers guide, the reader's ally in penetrating panta ta bellSnika, suddenly changes to become the Other's ally in concealing the mystery of Fleusis from his uninitiated readership. I lere, before the sacred which cannot be described, the text's phenomenology - one of its crucial structuring devices - breaks down. '45 ways of viewing The reader who travels through (jrecce zi'itb Pausanias, in his order, at his pace, along his roads, is left outside the sacred wall. Pausanias's writing is generally an endeavour of going out to us and making a way for us into the Other of his Greece, its ritual and its art. But his silence here - his articulation within discourse that there is an Other to discourse before which discourse must cease - is the opposite of the visual pattern of his writing, a denial of his writing, a bar to our entry into the heart of a Greece which it was his project to facilitate. If Kleusis were an isolated instance, we could be pardoned for overlooking its significance. But in fact it offers us the paradigm for a repeated pattern/'2 lake many pilgrimages, it offers a journey to a sacred Other/" which is often located on the periphery of a social or political centre but is nevertheless more deeply central to the pilgrim's sense of identity.** Such "peripheral centres" where the Other (whether a statue or a set of rites) must remain secret arc numerous. In some cases, like Kleusis, Pausanias was an initiate who could not divulge what he knew ; in others he was an outsider (like his presumed readers) and himself never knew what lay at the sacred centre. Such cases include the sanctuary of Demeter on Mt Pron outside I lermione ("the object most worthy of mention" [2.35.4]) whose goddess is surnamed Chthonia (i.e., one that defines the land). Pausanias describes a truly remarkable annual festival and cow sacrifice (2.35.5—7), the minor statues and images (35.S), as he builds to the climax: "But the thing itself that they worship before all else I never saw, nor yet has any other man, whether stranger or Hermionian. The old women may keep their know ledge of its nature to themselves" (2.35.8). The very structure of the writing leads to the authority-giving vet absent centre - the valorizing void of deity. Likewise, at the gate of I lermione on the road towards Mases is a sanctuary of Eileithyia: "F.verv day. both with sacrifices and with incense, they magnificently propitiate the goddess, and, moreover, there is a vast number of votive gifts offered to Eileithyia. But the image no one may sec, except, perhaps, the priestess" (2.35.11). Again, the paraphernalia of the sacred (rituals, offerings) lead to that which cannot be viewed or described. These paraphernalia entice and elicit description (their interest is what merits entry into the text), and vet the cause upon which all the ritual and the sanctuary itself are predicated - a deity anil the deity's image - art-denied to know ledge. 146 Pausanias's silence is itself a ritual act, the result of a reli- viewing and gious mentality of taboo and retribution. It is predicated on his identity usual word for mystery: uporrhetos, literally "away from speech*'.65 Often, as at Elcusis, the Athenian Llcusinium or the mysteries of the great goddesses at the Camasian grove outside Messene (4.33.4—5), it is a dream which informs Pau-sanias as to what he may or may not reveal. Several times he remarks emphatically on the consequences of transgression, whether by the physical act of entering a sacredly bounded place or the verbal act of giving the mysteries away. At the sanctuary of Poseidon Ilippios near Mantineia (8.i0.2f.) into which no one may enter, Pausanias twice repeats the storv of how Acgvptus broke this rule, only to be punished by blindness and death (8.5.5 anu" ^- l(M)- Death is the punishment for transgressivc entry at the precinct of Lycacan Zeus (8.38.6) and at the sanctuary of the Cabeiri (9.25.0-10). The merely inquisitive too will die (10.32.17), as will those w ho dare to imitate the mysteries (9.25.9) or who, like Orpheus in one of the myths of his death, profane them through speech (9.30.5). Such stories articulate a deep cultural sense of taboo surrounding the sacred. What was the sanctity which such ta-boos protected in pagan culture? Because his silence is scrupulously observed, Pausanias's readers clearly are not intended to know too much unless of course thev were to become initiates themselves. However, there is one instance where he does tell us something about the mysterious nature of the sacred centre whose essence is denied to discourse. During the trip to Arcadia, at