Chora and the Great Church

Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham BMGS 23 (1999) 67-101 T h e C h ora an d th e G reat C h u rch : In tervisu ality in F ou rteen th -C en tu ry C on stan tin op le ROBERT S. NELSON 'Utterances are not indifferent to one another, and are not selfsufficient. . . . Every utterance must be regarded primarily as a response to preceding utterances of the given sphere (we understand the word 'response' here in the broadest sense). Each utterance refutes, affirms, supplements, and relies on the others, presupposes them to be known, and somehow takes them into account.' I (Mikhail Bakhtin) Abstract In Gregoras' history, the most important building project during the second decade of the fourteenth century is not Theodore Metochites' redecoration of the Chora monastery, but Andronikos II's restoration of II. Sophia and the nearby column of Justinian. Two mosaics at the Chora, the Deesis in the inner narthex and the adjacent lunette of Metochites and Christ, make visual reference to the Great Church. The nature of such relationships is explored through reference to Mikhail Bakhtin and his analysis of dialogic language. In modem histories of Byzantine art, the Constantinopolitan church of the Chora, or Kariye Camii, is easily the most celebrated monument of the second decade of the fourteenth century, if not the entire Palaeologan period. Its fame surely owes much to the three lavish volumes on its mosaics and frescoes that were edited by Paul Underwood in 1966.2 The publication was honoured with the Charles 1. M.M. Bakhtin, 'The Problem of Speech Genres', in Speech Genres and other Late Essays, trans. Vern W. McGee (Austin 1986) 91. 2. Paul A. Underwood, The Kariye Djami, 3 vols. (New York 1966), hereafter KD. A book of interpretative essays followed: vol. 4 (Princeton 1975). 67 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON Rufus Morey Prize from the College Art Association of America as the most distinguished book of art history of that year. Given the largely descriptive nature of the first three volumes, the award would appear to honour the mosaics and frescoes themselves, as much as the work of conservation and publication that restored them to our view. The building's present status as one of the most popul.ar museums in Istanbul further certifies that these murals have achieved the status of high art in our world. It is curious, then, that in the principal Byzantine history of the period, that of Nicephoras Gregoras, the renovation and redecoration of the church is scarcely mentioned. Gregoras was a protege and defender of Theodore Metochites, for some years the prime minister of the Byzantine Empire, its richest citizen after the emperor, and the patron that embellished the church so richly. Because Gregoras had also been a resident of the Chora monastery,3 his indifference to it, therefore, can scarcely be credited to ignorance or lack of sympathy for monument or patron. If the Chora as presented in the grand volumes of Underwood does not correspond to the Chora of Gregoras, how should the building be contextualised? Second and more generally, what might it mean to describe a context for a monument? To explore the first question, this essay will consider what projects in contemporary Constantinople did interest Gregoras and what are the larger visual resonances of two mosaics closely associated with Metochites. To pursue the second question, the latter part of this essay will move in rather different directions and will take up issues seldom broached for art history or Byzantine studies. Here it will be argued that the modem theory of intertextuality, inspired by Mikhail Bakhtin, is more relevant to these matters than the well established and thoroughly naturalized disciplinary paradigms - older, but still modem - that normally prevail. Gregoras, in general, paid more attention to restorations associated with the emperor Andronikos II. These projects represented the 3. Nicephorus Gregoras, Byzantina Historia (Bonn 1829) I, 303, 309; Jan Louis van Dieten, Nikephoros Greg 0 ras, Rhomiiische Geschichte (Stuttgart 1979) vol. II, pt. 1, 32, 35. Gregoras has a bit more to say about the history of the complex: Gregoras, I, 459; Van Dieten, vol. II, pt. 2, 239-240. 68 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH continuation of the considerable efforts begun by Michael VIII to restore Constantinople to something like its former grandeur following the depredations of the Latin occupation (1204-1261). As a consequence, restoration and renovation, rather than new construction, were the compelling needs during the reigns of the first two Palaeologan emperors (1261-1328).4 However, not all damage to the monuments of Constantinople should be attributed to the Latins and the more distant past. Gregoras and the second important historian of the period, George Pachymeres, noted a serious earthquake in 1296 and described Andronikos II's restoration of a monumental bronze group outside the church of the Holy Apostles.s There atop a column stood the archangel Michael and kneeling before him and presenting a model of the city of Constantinople was Michael VIII. Alice-Mary Talbot suggests that Michael commissioned the group towards the end of his reign.6 That the statue group fell off its column is not necessarily and indictment of the workmanship, because any statue perched high atop a column would have been vulnerable to earthquakes, and earthquakes had had similar effects on other columnar statues in the past. 7 This imperial column should be understood as yet another aspect of Michael's legitimization and glorification as a New Constantine, the rebuilder of Constantine's city.8 This historians' interest in its restoration is an indication of the statue's continuing symbolic importance during the reign of Andronikos II. 4. The subject is surveyed in Alice-Mary Talbot, 'The Restoration of Constantinople under Michael VIII', DOP 47 (1993) 243-261; and Vassilios Kidonopoulos, Bauten in Konstantinopel 1204-1328 (Wiesbaden, 1994). I have compared the situation in the capital with that in Thessaloniki, where there was new construction, in a forthcoming essay, 'Tales of Two Cities: The Patronage of Early Palaeologan Art and Architecture in Constantinople and Thessaloniki', Manuel Panselinos and his Age, Proceedings of the International Symposium organized in Athens (November 1997), L. Mavrommatis ed. (Athens 1999). 5. Georgius Pachymeres, De Michaele et Andronico Palaeologis II (Bonn 1835) 234; Gregoras, I, 202; Van Dieten, I, 168. 6. Talbot, 258-260. 7. Glanville Downey, 'Earthquakes at Constantinople and Vicinity, A.D. 342-1454', Speculum 30 (1955) 598-600. 8. R.J. Macrides, 'From the Komnenoi to the Palaiologoi: Imperial Models in Decline and Exile', New Constantines, the Rhythm of Imperial Renewal in Byzantium, 4th-13th Centuries (Alders hot 1994) 270-272. 69 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON While this project warranted a few lines from Gregoras, the historian has more to say about restorations at Hagia Sophia and the adjacent column of Justinian in 1317. Both undertakings were urgent, of general political import, and contemporary with the rebuilding of the Chora. Gregoras probably witnessed all three projects shortly after he came to Constantinople in 1315 at age twenty.9 The largest endeavour, the work at Hagia Sophia, was prompted by the architects' report that the north and east sides of the church were in danger of collapse. Andronikos commissioned the repairs, and Gregoras says that he paid for them out of a portion of the estate of the empress Irene, who died in 1317 and was buried in the Pantocrator Monastery.lO At a cost of thousands of nomismata or hyperpyra, the architects erected what Gregoras calls pyramids. The latter are thought to refer to exterior buttresses for the church.ll Such massive bracing would have required far more masonry and, of course, expenditures than the additions that Metochites made to the much smaller Chora.12 This work at Hagia Sophia may have also included the belfry that once stood on axis between the second and third buttresses of the west facade,13 and in any case the constructions at Hagia Sophia would have been more significant visually and politically than the renovation of the Chora. Sociologically, the comparison is between a monastery of restricted access in the aristocratic quarter of the city and what was known as 9. Van Dieten, I, 1. 10. Gregoras, I, 273; van Dieten, I, 206. It must be noted that Rowland J. Mainstone (Hagia Sophia, Architecture, Structure and Liturgy of Justinian's Great Church [London 1988] 103) mistranslates this passage. He thinks that it refers to the east and south sides of the church. 11. Cyril Mango, 'Byzantine Writers on the Fabric of Hagia Sophia', in Robert Mark and Ahmet $. Cakmak, Hagia Sophia from the Age of Justinian to the Present (Cambridge 1992) 54. 12. One estimate of the cost of the restoring of a monastery is provided by the experience of another grand logothete, George Akropolites. In the 1270s, he paid 1000 nomismata to remove rubble from the church of the Anastasis in Constantinople. Renovations amounted to 16,000 nomismata in one year and continued. See Ihor Sevcenko, 'Society and Intellectual Life in the Fourteenth Century', reprinted in his Society and Intellectual Life in Late Byzantium (London 1981), I, 90. 13. Robert G. Ousterhout, The Architecture of the Kariye Camii in Istanbul (Washington, D.C. 1987) 108. The belfry, long since removed, has also been attributed to the Latin period. 70 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH the 'Great Church', attended by everyone on major feast days. After describing Andronikos' restorations here and elsewhere in Constantinople, Gregoras introduces another monument that was vital to imperial prestige, the statue of Emperor Justinian, located to the south of Hagia Sophia.14 This equestrian figure, three to four times life size, rested on a column about 35 metres high. IS In 1317, a strong north wind knocked off the bronze cross atop the globe that Justinian carried in his left hand. According to Gregoras, Andronikos acted with great haste to make repairs. A wood scaffolding with many steps was built up to the statue. When workers reached it, they discovered, as usually happens with such matters, that the whole was in need of repair - it was about to fall down. Still symbolically powerful and a prominent landmark in the city, even to Italian travellers (fig. 1), the statue had recently been the subject of an ekphrasis by the historian Pachymeres (d. c. 1310).16 From Justinian's day to the fifteenth century, this equestrian figure and especially his commanding right arm had been understood to be confronting the Empire's enemies. The ensemble must have been visually prominent as well, for it was recorded by various Italians in the first half of the fifteenth century.l? After the fall of Constantinople, the victorious Ottomans attached the severed head of the last Byzantine emperor to its base, and their leader, Mehmet the Conqueror, being advised that the statue was a symbol of Christian power, had it taken down. He was content, however, to leave standing less threatening monuments nearby. IS 14. Gregoras, I, 275-276; van Dieten, I, 207-208. 15. Bibliography is assembled in Wolfgang MUller-Wiener, Bildlexikon zur Topographie Istanbuls (Ttibingen 1977) 248-249. 16. Mentioned in Cyril Mango, Art Bulletin 41 (1959) 351; text in Gregoras, II, 1217-1220. 17. See the discussion in Michael Vickers, 'Mantegna and Constantinople', The Burlington Magazine 118 (1976) 683-687. lowe this reference to Prof. Klaus Kreiser. 18. Mango, Art Bulletin, 354. On the monument more recently, George P. Majeska, Russian Travelers to Constantinople in the Fourteenth and Fifteenth Centuries (Washington, D.C. 1984) 239-240; Speros Vryonis, Jr., 'Byzantine Constantinople and Ottoman Istanbul: Evolution in a Millenial Imperial Iconography', in Irene A. Bielman, et aI., eds., The Ottoman City and its Parts: Urban Structure and Social Order (New Rochelle, NY 1991) 34-35. 71 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON The two projects that are contemporary with the Chora preserved potent symbols of imperial power, and it is for this reason, it may be concluded, that they are given greater emphasis in the political history of Gregoras. In the second decade of the fourteenth century, imperial authority was less than what it had been during the reign of the previous emperor, much less those of the Middle or Early Byzantine periods, and even the most conservative intellectuals had begun to realize that the old imperial idea was fading.19 Gregoras reports that the emperor had in mind further restorations, but these plans, like the new army and navy that were also proposed, were cut short by the Civil War that broke out in the 1320s.20 Whatever else Andronikos II accomplished at Hagia Sophia is unclear and will remain so until the Great Church receives the detailed investigation that it so richly deserves. Toward that end, Mango has published drawings of mosaics no longer extant, including an intriguing sketch by the eighteenth-century Swedish architect Cornelius Laos (fig. 2).21 It depicts the north gallery of the church, the area in which Andronikos concentrated his restorations, and it illustrates the now lost decoration of one vault. The scene represented, the Baptism of Christ, is arranged around the rim of the vault in the centre of which there is a medallion. Such strip narratives, encircling a decorative device, are not common, but they are found in the vaults of the Kariye Camii (fig. 3) and a few other early fourteenth-century churches. Andre Grabar pointed to earlier Italian precedents, such as the well-known mosaics in the narthex of San Marco in Venice, and suggested a western derivation for the later Byzantine versions, a notion critiqued by Otto Demus.22 This larger matter is beyond the limits of the present essay. 19. See Ihor Sevcenko, 'The Decline of Byzantium Seen Through the Eyes of its Intellectuals', in his Society and Intellectual Life, II, 169-186. 20. Gregoras, 1,275; van Dieten, I, 206-207. On the new troops, Gregoras, 1,317-318; Van Dieten, II, pt. 1, 40. 21. Cyril Mango, Materials for the Study of the Mosaics of St. Sophia at Istanbul (Washington, D.C. 1962) 47-48, figs. 49, 55. 22. Andre Grabar, 'La decoration des coupoles a Karye Camii et les peintures, italiennes du Dugento', JOBG 6 (1957) 111-124; Otto Demus in KD IV, 138. 72 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Nevertheless, it is possible to put the contributions of Mango and Grabar together so as to yield the following minimal thesis: first, the composition barely glimpsed in Loos' drawing is rare and, second, its closest Constantinopolitan parallel is at the Kariye Camii. Mango further noted the similarity of the rainceaux ornament in the two churches.23 Comparisons can also be found at the Chora to the vines that weave around a long central stalk (fig. 3). In both monuments, such ornament defines arches and frames bays and other architectural units. Moreover, the decorative motif at the centre of the Baptism vault at Hagia Sophia may also have a parallel at the Chora.24 Admittedly it is difficult to assign a date to this lost vault mosaic, but historically it is most easily explained, if the north gallery of Hagia Sophia, having been damaged by the structural weakness that troubled Andronikos' architects, were redecorated shortly after 1317. The project then would have been underway at the same time or perhaps just before the renovation of the interior of the Chora, the work that Gregoras says had just been completed by Lent of 1321.25 Like Hagia Sophia and the equestrian statue, the church of the Chora was also thought by Gregoras to be of Justinianic origin, an attribution that has some validity.26 The historian also reports that this first church was completely renovated by the mother-in-law of Alexius Comnenus, Maria Ducaena, and thus in the eleventh century. Ousterhout finds stronger archaeological evidence to support this attribution.27 But Gregoras, writing about events in 1330-31, is less interested in archaeology than genealogy and the concomitant 23. Mango, Materials, 48. 24. What is visible of the central medallion of the Hagia Sophia appears to be the end of a cross of some sort. It may be compared with a medallion at the Chora: KD II, pIs. 15, 332a. 25. Gregoras, I, 303; van Dieten, II, pt. 1, 32. 26. Gregoras, I, 459; van Dieten, II, pt. 2, 239. Ousterhout tentatively confirms the existence of sixth-century masonry beneath the present church: Kariye Cam ii, 14. The sixth-century marble doors, now placed in the door of the inner narthex, are another sign of the church's relationship to an early Byzantine past and possibly to Hagia Sophia itself, where similar doors are found in the south gallery. See 0ystein Hjort, 'The Sculpture of Kariye Camii', DOP 33 (1979) 202-223. I thank Robert Ousterhout for drawing my attention to this spolia. 27. Ousterhout, Kariye Camii, 15-20. 73 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON legitimatization of his old teacher Metochites. The latter had recently returned from exile to live in the Chora, his luxurious palace having been sacked. Legitimization through imperial association had long been Metochites' strategy for the rehabilitation of his family's name and for their social advancement, and he succeeded in marrying a daughter to the nephew of Andronikos II and probably two sons to other relations of the emperor.28 Certain mosaics of the Chora proclaim similar messages and repay further scrutiny in the context of a programme that is simultaneously political and religious.29 The most overt example of imperial association is the large Deesis mosaic on the east wall of the inner narthex to the right of the entrance into the naos (fig. 4).30 There two historical personages, Isaac Comnenus and Melane the nun, flank the towering figures of Christ, inscribed Chalkites, and the Mother of God, who intercedes for the people below, those beholders inside and outside the panel. Isaac stands, gestures prayerfully, and looks across to Christ, while Melane kneels and directs her prayer to Mary. The eyes of the two women almost meet. The usual hierarchical difference between males and females is maintained in the mosaic, for the taller Isaac stands at Christ's right, always the favoured position. When the mosaic was made, this Isaac, the Sebastocrator and younger brother of the twelfthcentury Emperor John II Comnenus, had long been dead. However, he had once erected his tomb at this church, which had been renovated by his grandmother, the mother-in-law of Alexios I, Maria Ducaena, as Gregoras notes.3! Later Isaac had the tomb transferred to his new foundation, the monastery of the Theotokos Kosmosotira in Thrace,32 but he explicitly requested that his portrait remain at the Chora. It may thus have still been visible, when Metochites began his renovation. In the inner narthex, the memory of Isaac, preserved in oral, written, or visual form, has been reinscribed for the benefit, not of a twelfth- 28. Sevcenko, KD IV, 27, n. 61. 29. I have emphasized the political messages of the outer narthex mosaics in 'Taxation with Representation. Visual Narrative and the Political Field of the Kariye Camii', Art History 22 (1999) 56-82. 30. Underwood, KD I, 45-47. 31. As in note 26. 32. Ousterhout, Kariye Camii, 21. 74 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH century past, but a fourteenth-century present. Isaac is denoted by his relationship with the emperor Alexius Comnenus and by the word porphyrogennitos, reserved for children born to an emperor and empress in the Imperial Palace: 'The son of the most high emperor Alexius Comnenus, Isaac the Porphyrogennitos.'33 The existence of multiple porphyrogennitoi would have been the norm at most any time in the twelfth century, but because of the long Latin interlude, such children were a novelty during the first years of the Palaeologan dynasty. Andronikos II's younger brother Constantine had been the first porphyrogennitos since the 1190s34 In any official twelfth-century context, Isaac would have born the title, Sebastocrator, an honour created by Alexios I Komnenos for this brother Isaac and then used for an emperor's brothers and sons, including the Isaac of the mosaic, the son, as it says, of Alexios 1. During the Palaeologan period, the prior significance of sebastocrator diminished, as Michael VIII and Andronikos II gave their brothers and sons the title despot.35 For fourteenth-century audiences, including the title the sebastocrator would not have been consonant with the other inscribed qualifications, imperial son and porphyrogennitos, nor with the evident emphasis on the high status of Isaac, as defined in contemporary terms. Isaac's counterpart, Melane, has a similar, though fragmentary inscription: ' ... of Andronikos Palaeologos, the Lady of the Mongols, Melane the nun.' 36What is emphasized for her, as with Isaac, is the relationship to an emperor, in her case Metochites' great patron. As to the identification of the nun Melane, Underwood pointed to two possible women, both named Maria, both illegitimate daughters of emperors, either Michael VIII or Andronikos II, and both married for political purposes to Mongols.37 It is assumed for each that the usual custom of determining a monastic name is operative, namely that one chose a new name starting with the same letter as the old 33. '0 vias TOU V1.JIllAOTaTOV (3aatAEws 'AAEeiov TOU KOlJVllVOU 'IaaaKlOs 6 nop~vpoyEVV~TOs. 34. Macrides, 'From the Komnenoi', 272. 35. B. Ferjancic, 'Les sebastokratores a Byzance', ZRBI 11 (1968) 19l. 36. [... 'A]v5[pov]tKov TOU IIaAalOAoyov ~ Kvpa nov MOVYOVAtWVMEAavll ~ lJovaXr1. 37. KD 1,46-47. 75 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON one. Of the two candidates, the elder, the daughter of Michael VIII and thus the half-sister of Andronikos II is the more likely, not the least because she was probably connected by marriage to Metochites.38 Natalia Teteriatnikov has recently provided more support for this hypothesis, supplying evidence, both textual and art historical, for this Mary's association with the Chora.39 Her further contention that Mary/Melane is the central focus of the Deesis mosaic is less convincing, because the nun is on the left side of Christ. She is also comparatively small, since she is shown kneeling, not standing like Isaac. In Byzantine art, size always matters and correlates with importance. The sequence of the four figures arranged by size alone, Melane, Isaac, Mary, Jesus, defines a hierarchy that is gendered, social, political, and spiritual. Teteriatnikov usefully explains MarylMelane's role as a donor/patron of the monastery, and she suggests that one of the miracle scenes placed in the vault above the Deesis, that of the woman with the issue of blood, may have had personal significance for Melane.40 The latter point is difficult to prove without further evidence, but in any event, it is not incompatible with the general significances of the panel, polysemy being another characteristic of Byzantine art. The most direct and unambiguous clues to the meanings of the mosaic are the inscriptions that stress the relationship of each aristocrat to an emperor. Both figures appear in this mosaic in positions usually reserved for donors, and their presence creates a visual genealogy for the present patron, Theodore Metochites, one that supplements the history that Gregoras provides (Justinian, Maria Ducaena).41 Once again, this is history in service of the present. In the Deesis at the Chora, not only what is represented, but how that representation is effected, defines a further context for patron 38. Sevcenko, in KD IV, 27, n. 61. 39. Natalia Teteriatnikov, 'The Place of the Nun Melania (the Lady of the Mongols) in the Deesis Program of the Inner Narthex of Chora, Constantinople', CahArch 43 (1995) 163-184. My analysis and description of the mosaic does not correspond to hers in all places. She continues her discussion of the church in 'The Dedication of the Chora Monastery in the Time of Andronikos II Palaiologos', Byzantion 66 (1996) 188-207. 40. Teteriatnikov, 'The Place of the Nun Melania', 171. 41. As in note 26. 76 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH and program. What is immediately noticeable about the Deesis is its large size relative to the space available and to other figures nearby. The height of the panel is 4.81 m and of Christ 4.20 m, versus 2.05 m and 1.845 ill for the nearby icons of Peter and Paul42 that flank the door to the left (the door and the icons with arched frames are visible in fig. 3). The latter icons are part of a group that are similarly framed and positioned elsewhere in the nartheces for devotional purposes.43 Such icons are normally the closest images to the beholder. Because of the traditional vertical hierarchy that governs Byzantine church decoration, Christ is more often placed at higher elevations. But what is noteworthy about the Deesis is that it extends below the Peter and Paul panels and breaks through the border of the decorative dado that unifies the space (fig. 5). The towering figures of Christ and Mary, so close to the ground, can hardly be ignored, and worshippers would encounter the mosaic by one of two paths (fig. 6). One begins at either the main west door (bay 3) or another door to the north (bay 1) of what is now the outer narthex, but originally was an exterior portico.44 Worshippers then would cross the inner narthex (bay 3) and enter the naos through the single door (fig. 3). In the vault above the entrance bay of the inner narthex, a depicted procession represents the introduction of the Virgin to a temple that is configured like the bema of a church (fig. 3). The scene replicates what was enacted below, processions to an actual naos and bema. Another passage into the church begins at a door on the south, now blocked, crosses what Underwood labels as bay 7 of the outer narthex (fig. 6), passes through a door here and into the space of the mosaic (the path visible in fig. 5), and turns right into the nave. This path also has analogies with the mosaics above.45 42. Underwood, KD I, 44, 47. 43. Underwood, KD II, pIs. 10, 11, 14. 44. Ousterhout, Kariye Camii, 70-78. 45. Those relationships can and could be seen in bay 7. Over the south door, there is a lunette of the Flight to Egypt, and over the opposite north door that leads into the bay with the Deesis, there most likely was once the Adoration of the Magi. See Jacqueline Lafontaine-Dosogne, 'Iconography of the Cycle of the Infancy of Christ', KD IV, 220-223. The Flight to Egypt depicts a passage to a walled city, as worshippers leave the church. Given the iconographic tradition and the space available (cf. 77 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON In Constantinople, a mosaic of a similar subject matter, size, and placement is to be found in the south gallery of Hagia Sophia (figs. 78). Restored before World War II, the Hagia Sophia Deesis was published, poorly it seems, by Whittemore in 1952.46 Robin Cormack has promised a more accurate and detailed study.47 Before the latter's appearance, it is difficult to compare the two Deesis groups closely, but basic relations are apparent. For example, the heads of the two Christ figures are virtually the same size.48 Overall, the Hagia Sophia panel is taller, but both mosaics end near floor leve1.49 Because of its size and position, Cormack regarded the Hagia Sophia mosaic as a new type of devotional image and associated it with the celebration of the Byzantine reconquest of Constantinople in 1261. While that specific context remains as yet undefined by him, the panel is generally accepted as early Palaeologan in date. Located in an area that has other imperial images and which was reserved for imperial use, it is reasonable to conclude with Cormack that the Deesis of Hagia Sophia had imperial connotations. All this would have been clearer, if the donor figure that Cormack hypothesizes in the now destroyed lower sections, had been preserved. The grand spaces of the Great Church afford the possibility of distant views of the Deesis mosaic (fig. 8) and reduce the monumentality of figures that are over twice life size. No such distance is possible in the narrow inner narthex of the Chora, making its Deesis seem oddly misplaced amidst the delicately proportioned narratives and the uniformly sized mosaic icons on the walls. The Chora mosaicists Underwood, KD II, pI. 180a), the now effaced Adoration most likely would have been composed with the Magi at the left and Mary and the child at the right or the east. After walking beneath this mosaic, its beholders would have turned to the right or the east and processed toward the Virgin and child in the naos, the subject of one of the templon icons there (ibid. II, pI. 329). 46. Thomas Whittemore, The Mosaics of Haghia Sophia at Istanbul, four preliminary report, The Deesis Panel of the South Gallery (Washington, D.C. 1952). 47. Robin Cormack, 'Interpreting the Mosaics of S. Sophia at Istanbul', Art History 4 (1981) 145-146. 48. Hagia Sophia: .77 m (Whittemore, 38). Chora: .75 m (Underwood, KD I, 47). 49. Hagia Sophia: 5.95 m (Mango, Materials, 29 n. 34). Chora: 4.81 m (Underwood, KD I, 47). Cormack ('Interpreting the Mosaics', 145) commented on the unsatisfactory nature of Whittemore's study. The latter, for example, recorded the height of the panel as 'about 4.68 m': Whittemore, 12. 78 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH could have achieved a more harmonious and iconographically traditional composition, if they had depicted the standard group of Mary, Jesus, and John the Baptist. However, in order to include the Baptist in the space available within in the blind arch, the height of figures would have had to be reduced substantially, rendering each less important. The fact that the mosaic was made in the present form suggests that the resemblance in size and placement to the Hagia Sophia panel was more important than iconographic tradition. The principal figures of Christ in the two mosaics do not agree iconographically. The Christ of Hagia Sophia holds a jewelled book, and the survival of a bit of ornament led Whittemore to reconstruct this Christ as enthroned (fig. 9).50 At the Chora, Christ stands, blesses with his left hand, while holding the other hand awkwardly at his waist. This Christ is inscribed Chalkites, an epithet that refers to a famous icon then still preserved at the Chalke gate to the old imperial palace. The image held imperial associations in the thirteenth century and appeared on coins of the Nicaean empire. 51 The cult of this icon of Christ was widespread in late Byzantine Constantinople, but it also had a specific connection with Hagia Sophia. According to Russian travellers, a version of the Chalke icon was prominently placed on the west wall of the nave of Hagia Sophia above the central Imperial Door, the principal entry from the inner narthex. 52 Given the space available here, this figure of Christ Chalkites, probably a mosaic icon, would have followed tradition and been represented standing. Today, the image that is known as the mosaic over the Imperial Door is on the other side of the same wall at Hagia Sophia. It fills the lunette above this, the largest entrance into the church (fig. 9) and a ceremonially significant site. The tenth-century Book of Ceremonies describes the emperor and patriarch with their entourages, meeting at these doors and processing together into the church. 53 Still 50. Whittemore 17. 51. Cyril Mango, The Brazen House, A Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959), 135-142; Underwood, KD I, 45-46; Nancy Patterson Sevcenko in The Oxford Dictionary of Byzantium (Oxford 1991) 440; Teteriatnikov, 'Place of the Nun Melania', 170. 52. Majeska, Russian Travelers, 210-212. 53. Albert Vogt, Le Livre des ceremonies, vol. 1 (Paris 1967) 10-11. 79 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON important in late Byzantium, this entry was used only on special occasions. Near it were major icons, and the Russian pilgrims report that the doors themselves were thought to be made from the wood of Noah's ark and were therefore venerated. 54Thus the mosaic of an emperor kneeling before an enthroned Christ above this door was likely to be well known to everyone in the city, including the designers and patron of the Chora. Metochites chose to have himself portrayed in the corresponding space in his church, the lunette above the door from the inner narthex into the nave (fig. 11). There he kneels before an enthroned Christ, holding a jewelled book. This Christ, while highly traditional nevertheless has much in common with Whittemore's reconstruction of the central member of the Deesis group at Hagia Sophia (fig. 9). Metochites holds a model of the church he patronized, again a common motif, but one that also has an analogy at Hagia Sophia in yet another mosaic over a portal. In the southwest vestibule and along the processional route into the church,55 Justinian is represented presenting his church to the Virgin and child (fig. 12). This is the same emperor whom Gregoras understood to be the first benefactor of the Chora. By the fourteenth century, the political significances of compositions, such as seen in these lunettes of Hagia Sophia, had not faded. A poem of Maximos Planoudes, for example, describes a now lost representation of Christ, Mary, Andronikos II, his son Michael IX and two saints over the entrance to a monastery. 56 At the Kariye Camii, Metochites is not shown as an actual emperor, but nonetheless, there are imperial pretensions in the donation panel and the nearby Deesis, what Mango called Metochites' snobbery. 57 Those pretensions further manifest themselves in the associations between the imperial mosaics at Hagia Sophia and the program of the Chora. The adjacent Deesis and donation lunette at the Chora can be understood as variations on and replications of the two narthex mosaics at Hagia Sophia and the Deesis panel in the south gallery. 54. 55. 56. 57. Majeska, Russian Travelers, 206-209. Ibid. 202. Carl Wendel, 'Planudea', BZ 40 (1940) 427. Mango, Brazen House, 142. 80 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH The first two of these are manifestly imperial. The third, the Deesis, may also have had a strong relation with the emperor, if Cormack's thesis can be sustained. Even if it cannot be eventually concluded that there might once have been a donor figure in the Deesis and that it was made to commemorate the coronation of Michael VIII in the Great Church after the reconquest of Constantinople in 1261, the mosaic is part of a space filled with other prominent imperial mosaics and was used by the court. The south gallery and the inner narthex are areas of Hagia Sophia that Metochites and his fellow aristocrats must have known well. * * * * * The foregoing has presented varied evidence of the interrelations between the Kariye Camii and Hagia Sophia during the second decade of the fourteenth century. It first argued that the most important patronage at that time in Constantinople was not Metochites' restoration of the Chora, as splendid as it may be, but projects of Andronikos II in and around Hagia Sophia. Second, it explored relationships between imagery at the Kariye Camii and mosaics, chiefly imperial, at the Great Church. The organization implied, but did not demonstrate a relationship between the first and second points. The rhetoric of presentation has been deliberately unadorned, attempting to be a straight-forward accounting of different types of relationships. To follow one framing of such matters, the essay to this moment has presented the 'facts' and suppositions drawn from them, and it thus falls into 'the normative epistemological framework of the empiricist! positivist and humanist tradition' that prevails in Byzantine studies. 58 What follows is the 'jargon', or I would prefer, an attempt at interpretation. How can we understand the phenomena that has been described? Is it adequate to join the two monuments, the Chora and Hagia Sophia, by a simple 'and'? Is it possible to understand the nature of the relationships suggested, and exactly what constitutes a 58. John Haldon, "'Jargon" vs. "the Facts"? Byzantine History-Writing porary Debates', BMGS 9 (1984/85) 95-132, especially 125. 81 and Contem- Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON visual relationship in the Middle Ages? For whom does that relationship exist, a monument (that anthropomorphized creation of art history and nationalism), or a patron, or a viewer (entities not without their own conceptual issues)? To ponder these matters, I tum to my epigraph and to theoretical issues that have to date been explored more for language and literature, than history or art history. The quotation is from an essay that the Russian polymath Mikhail Bakhtin wrote in 1952-53. It represents a continuation of issues that Bakhtin explored in the major early work, Marxism and the Philosophy of Language of 1929, if, that is, Bakhtin is the actual author of a book that appeared under the name of V.N. Volosinov. The latter is a problem that Bakhtin scholars continue to debate and is not for the present essay.59 The book remains important no matter which member of combination of the Bakhtin circle wrote it. What is relevant in the study is the social grounding that Bakhtin/Volosinov give to language and their understanding of relationship and context. Their study is of language, but I want to argue that it can be extended to art in ways not heretofore attempted.60 A brief introduction to relevant aspects of Bakhtin's writings and their later impact is in order. For Bakhtin, an utterance, a word, is the basic element of dialogue, a fundamental concept for him. Meanings are communally determined: The 'word is a two-sided act. It is determined equally by whose word it is and for whom it is meant. As word, it is precisely the product of the reciprocal relationship between speaker and listener, addresser and addressee ... A word is a bridge thrown between myself and another. . . . A word is a territory shared by both addresser and addressee, by the speaker and his interlocutor. '61 Similarly, a sign 59. I am, however, inclined to agree with those who assign the text to Bakhtin: Katerina Clark and Michael Holquist, Mikhail Bakhtin (Cambridge, Mass. 1984) 166. For different views, see Ladislav Metejka and I.R. Titunik, introduction to their translation of V.N. Volosinov, Marxism and the Philosophy of Language (New York, 1973) ix-xi; Tzvetan Todorov, Mikhail Bakhtin, The Dialogical Principle (Minneapolis 1984) 6-11. 60. Deborah J. Haynes, Bakhtin and the Visual Arts (Cambridge 1995) deals primarily with aesthetics and creativity and works with other parts of the oeuvre of Bakhtin. 61. Philosophy of Language, 86. 82 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH cannot be separated from their contexts of social interaction without becoming 'mere physical artifact', nor can the sign be 'divorced from its material basis.'62 As a consequence, for a sign or word in active human use, 'there are as many meanings ... as there are contexts of its usage.' Ignoring context and applying a dictionary meaning to a word constitutes a reification, a distorti on. 63 Another error, he contends, is to consider that the many contexts of usage are 'all lying on the same plane.' This he criticizes as 'abstract objectivism,' something he associates with formalism, a movement which has had a significant impact on art history, as well as linguistics. In abstract objectivism, 'contexts are thought of as forming a series of circumscribed, self-contained utterances all pointed in the same direction.' This need not necessarily be the case, and the various meanings of a word can even be opposed to each other, especially in dialogue, that is, actual conversation. One speaker can employ a word with one meaning. In repeating the word, the respondent may continue, adapt, or even contradict the first speaker's use of the word. For this reason, the contexts of a word should not be understood to 'stand side by side in a row, as if unaware of one another.' Instead they 'are in a state of constant tension, or incessant interaction and conflict. ' 64 It is in this sense that Bakhtin later wrote, as quoted in the epigraph, 'each utterance refutes, affirms, supplements, and relies on the others, presupposes them to be known, and somehow takes them into account.' In this interaction, every utterance becomes part of 'a link in the chain of speech communication of a particular sphere. '65 Therefore, utterances are fundamentally social, connected to and connective of society. 'Any utterance, no matter how weighty and complete in and of itself, is only a moment in the continuous process of verbal communication. But that continuous verbal communication is, in tum, itself only a moment in the continuous, all-inclusive, generative 62. Ibid. 21. 63. Ibid. 79-80. On Bakhtin's expanded notion of utterance and its social consequences, see Clark and Holquist, 203-211. 64. Philosophy of Language, 80. 65. Speech Genres, 91. 83 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON process of a given social collective.'66 It follows then that Bakhtin's sense of language is grounded in social intercourse and connected with the larger social codes of behaviour and belief. What for Saussure and his many successors, Structuralists and Post-Structuralists alike, is a great gulf between parole (actual speech) and langue (language or the general linguistic code) becomes for Bakhtin a continuum in lived experience.67 In his theory of the language and what has been aptly termed his Dialogic Imagination,68 there is much that was inspired by Russian orthodox theology, as has been recently argued. Hence the distance between the modern theoretician, Byzantine art, and the Orthodox icon may not be great as one might first suppose.69 However, it was not for this religious aspect that Bakhtin was reclaimed, when he began to be read, practically for the first time, in the 1960s. Instrumental in his revival in Western Europe were the Structuralists, especially two Bulgarian-born French critics, Julia Kristeva and Tzvetan Todorov, who were able to read the Russian Formalists and the then obscure Bakhtin in the original. Adapting and extending Bakhtin in her first book, Kristeva transformed (and hence demonstrated) his dialogism and coined the closely related term 'intertextuality'. She too conceives of a dynamic relationship between writer, addressee, and cultural context. For her, the text is oriented both horizontally, to the writer and addressee, and vertically, towards prior and contemporary literature. 'Any text is constructed as a mosaic of quotations; any text is the absorption and transformation of another. '70 Like Bakhtin, 66. Philosophy of Language. 95. 67. Clark and Holquist, 14. 68. Michael Holquist, ed. (Austin 1981). 69. The important study of Alexander Mihailovic, Corporeal Words: Mikhail Bakhtin 's Theology of Discourse (Evanston 1997) extends observations of Clark and Holquist (pp. 120-145) about Bakhtin's theology. Bakhtin uses the term word in the sense of the logos in John's Gospel and considers it, according to Mihailovic, as a bridge 'between the unbodied content of an act and its performative or material manifestation.' In this respect, word or utterance functions like an icon. See ibid. 39. 70. LT]/1£lWTlKt7, Recherches pour une semanalyse (Paris 1969) 145-146; translated in Toril Moi, ed., The Kristeva Reader (New York 1986) 36-37, with a useful introduction. 84 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Kristeva was interested in the sign as process, and the openness of meanings that Bakhtin's theories encouraged and which frustrated their publication 71must have been appealing and applicable, as French intellectuals moved toward Post-Structuralism. At this stage of her career, Kristeva was what has been called a Post-Structuralist 'avant la lettre'.72 In a subsequent book, La Revolution de langage poetique, she added Lacanian psychoanalysis to her Bakhtinian based semiotics and created in the process other critical terms, including - ironically for this essay - the concept of the 'chora', which, however, is less relevant to the actual Chora in Constantinople.73 More germane is intertextuality, her extension of Bakhtin's linguistics to literature, and a mode of analysis that remains relevant today.74 Shifted once more to visual imagery, the concept becomes intervisuality and can be generally applied, for all art is a communication between an addresser and addressee and takes prior work into account. But the term is especially useful for the icons on the walls of the Chora. Those murals are like Bakhtin's utterances. Both parties share the actual spaces of the images and are directly interconnected by them. Icons are a special form of visual communication, because in practice, they are regarded as a person. Icons are fundamentally dialogic and replete with semantic markers that characterize and 71. Clark and Holquist, 152. 72. Moi, Kristeva Reader, 3. 73. La Revolution de langage poetique (Paris 1974) 22-30; Margaret Waller, trans., Revolution in Poetic Language (New York 1984) 25-30. She borrows the term from Plato's concept of the chora as receptacle and makes it into a primary stage of preconsciousness. There are interesting coincidences between this Freudian abstraction and the Byzantine mysticism that underlies the visualization of the chora at the Chora, but the modern and medieval world views differ profoundly in their accounting of human subjectivity and, of course, the deity. Plato's theory and the general philosophical treatment of place/space is explored in Edward S. Casey, The Fate of Place, A Philosophical History (Berkeley 1997). He does not consider Kristeva, however. 74. See the useful discussion of a neighbouring literary tradition: Stephen Hinds, Allusion and Intertext, Dynamics of Appropriation in Roman Poetry (Cambridge 1998). I learned of this book from the review of Christopher Nappa in the on-line journal, Bryn Mawr Classical Review, September 1998. 85 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON facilitate speech, or what linguists call discourse.75 Figures stare outward and gesture to their beholders. Inscriptions, presented in or by icons, often contain deitic language, words that indicate time and space only for the participants in a conversation. In everyday use, icons are surrounded by speech. Believers pray to them, addressing them and the person portrayed in the vocative, the voice of dialogue. And that communication is not thought to be unidirectional, one type of Christ icon even being called the Antiphonetes, the one who responds.76 At work in the Deesis panel and the donation lunette at the Chora is an intervisuality that pertains to all art, as well as one that applies to the special characteristics of the Byzantine icon. Both panels take into account prior images at Hagia Sophia. Or more precisely, the makers and patrons of these images were aware of other images in their environment, their context. The same visual consciousness, it is argued, existed among the fourteenth-century elite audiences, for whom Hagia Sophia would have been the best known church in the city. But the relationship may also have gone the other way, if the designs of the north gallery mosaics were made after the narthex vaults of the Chora. These visual responses define the sphere within which intervisuality operates and become one measure of an artistic, social, and political context of the Chora. By referring to Hagia Sophia, the Chora and its patron writes itself and himself into the centre of contemporary imperial/religious symbolism or ideology. Moreover, Metochites accomplishes this not with mere words on a wall, but by his inclusion in this interplay of icons, more potent signifiers in his world. Between the Chora and Hagia Sophia, the foregoing has suggested a range of possible associations from the near identity in size of parts of the two Deesis mosaics to the more generalized associations of the lunette mosaic in the nartheces of the two churches. Like Kristeva's metaphorical mosaics, the donation lunette and the Deesis panel at 75. See my essay, 'The Discourse of Icons, Then and Now', Art History (1989) 12 (1989) 144-157. 76. N.P. Sevcenko, 'Christ Antiphonetes', in Oxford Dictionary of Byzantium 439. 86 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH the Chora absorb and transform pre-existing images. Parts, not wholes are borrowed and adapted: for the lunette at the Chora a kneeling figure, an enthroned Christ, a donor with a model of his church; or for the Deesis a venerated icon type with imperial associations, the Chalkites, and the oversized dimensions of the composition. Intermingled in the latter are the figures of the nun Melane and Isaac Comnenus, formerly a sebastocrator, now the Palaeologan equivalent, a porphyrogennitos. They too have been transformed and in turn transform by their presence. This selective replication of salient parts of earlier models is what Richard Krautheimer found for the copying of churches in the medieval west. For Byzantium, Marcus Rautman has convincingly demonstrated that a similar process generated the plans of Palaeologan churches in Thessaloniki.77 Like the mosaicists of the Chora, Palaeologan illuminators copied and adapted earlier models, chiefly the major manuscripts of the so-called Macedonian Renaissance, the same period as the narthex mosaics of Hagia Sophia.78 In illuminated manuscripts, the copies of Middle Byzantine models were sometimes so exact as to be attributed to the earlier period.79 In another instance, an entirely new pastiche was made so adeptly from fragments of prior miniatures, as to deceive modem scholars into thinking that it was copied from a lost earlier mode1.80 Remaking the Joshua Roll or the Paris Psalter in the late thirteenth century was like Michael VIII styling himself the New Constantine or Theodore Metochites including himself in a visual genealogy of emperors and imperial art. Process and product created legitimacy. 77. Richard Krautheimer, 'Introduction to an "Iconography of Medieval Architecture"', Journal of the Warburg and Courtauld Institutes 5 (1942) 1-33, reprinted in his Studies in Early Christian, Medieval, and Renaissance Art (New York 1969) 115-150; Marcus L. Rautman, 'Patrons and Buildings in Late Byzantine Thessaloniki', JOB 39 (1989) 313-315. 78. Hugo Buchthal and Hans Belting, Patronage in Thirteenth-Century Constantinople, An Atelier of Late Byzantine Book Illumination and Calligraphy (Washington, D.C. 1978) 17-34. 79. Anthony Cutler, 'The Marginal Psalter in the Walters Art Gallery, A Reconsideration', The Journal of the Walters Art Gallery 35 (1977) 37-61. 80. Hans Belting, 'Zum Palatina-Psalter des 13. lahrhunderts', JOB 21 (1972) 17-38. 87 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON Reading Bakhtin encourages one to think of the varieties of relationships that surround utterances and icons and which make and remake meanings for changing contexts. In addition to the associations of the Chora with Hagia Sophia, there must have once been many other relationships, only a few of which are partially recoverable. For example, Russian visitors to the Pantocrator Monastery mention a large figure of Christ: 'above the first doors, the doors by which you enter, the Saviour is done in mosaic, a large figure, and high.'81 Long a politically important monastery, the Pantocrator was where Andronikos II buried his wife just before he began his restoration projects in and around Hagia Sophia. Could an image in the narthex of the Pantocrator have been in the minds of the creators of the large bust of Christ over the entrance lunette in the exonarthex or exterior portico of the Chora ?82 Facing it is the mosaic of the Virgin Blachernitissa in the lunette over the exit to the west. As Robert Ousterhout has convincingly shown, this image has multiple meanings, but one refers to the nearby monastery of the Blachernae. Like the Chalkites, this icon also had imperial associations. 83 Because the mosaic at the Pantocrator monastery has disappeared, a more precise association between it and the Chora is impossible. The example suggests the limitations of this essay. To understand more adequately the dialogic contexts of the two murals at the Chora would entail comparisons with more than Hagia Sophia. The danger in studies such as this one is that they define the context of a work of art by means of what that work happens to be or contain. This reading out from the art work moves to a context that seemingly contains, but is actually produced by the art work. Baxandall and Bryson have explored the difficulties with such contextualizations.84 81. Majeska, Travelers, 289, 291. 82. Underwood, KD II, pI. 17. 83. Robert Ousterhout, 'The Virgin of the Chora: An Image and its Contexts', in Robert Ousterhout and Leslie Brubaker, The Sacred Image East and West (Urbana 1995) 91-109. 84. Michael Baxandall, 'Art, Society, and the Bouguer Principle', Representations 12 (1985) 32-43; Norman Bryson, 'Art in Context', in Studies in Historical Change, Ralph Cohen, ed. (Charlottesville 1992) 18-42. The issue is discussed again in Mieke Bal and Norman Bryson, 'Semiotics and Art History', Art Bulletin 73 (1991) 174-188. 88 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Here Bakhtin is again useful, for those approaches are a version of what he discussed decades earlier, namely, the tendency to see the contexts of utterances as circumscribed and unidirectional. For Bakhtin and Kristeva, the utterance and the sign are produced and producing, and the ever changing situations in which images are encountered create new meanings. For much of the life of the mosaics of the Chora, images remained fixed, and their reinterpretation was a consequence of the diverse groups, who encountered the images and took note of them over the long life of the monument from Byzantine church to Ottoman mosque to Turkish museum. Since the publication of Underwood's grand volumes, the mosaics now circulate independent of the church in a scholarly and aesthetic world of mechanical reproduction. The publication reduces, at the same time it enhances and promotes the aura of the originals for audiences far beyond the limited circles that Metochites envisioned. On the pages of the plate volumes, the murals stand side by side, precisely as Bakhtin claims the abstract objectivists would have it. Moreover in the various art historical essays of the fourth volume, the images are grouped into series - life of Christ, life of the Virgin - that emphasize their relationships with similar series constructed by the many 'museums without walls' of art history. This process of isolation, sequencing, and re-contextualization continues in the present essay, even as it tries to give the images back to a fourteenth-century context. Yet all dialogues around images reanimate them and invite readers/viewers to break apart the formalist, self-contained abstractions and the high modernism of the Underwood books, as well as the post-structuralist accounting now concluded. May the process continue. University of Chicago 89 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT Fig. 1 S. NELSON Constantinople, drawing by Cristoforo Buondelmonti (after Gerola, SBN 3 [1931]) 90 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Hagia Sophia, drawing of north gallery by Cornelius Loos (after Mango, Materials for the Study of the Mosaics of Hagia Sophia) F ig . 2 91 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT Fig. 3 S. NELSON Chora, inner narthex looking north 92 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham Fig. 4 THE CHORA AND THE GREAT CHURCH ehora, Deesis mosaic 93 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT Fig. 5 ehora, S. NELSON view of bay 4 of inner narthex (after Ousterhout, 94 Kariye Camii) Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH N~ Fig. 6 0' Chora, plan (after Ousterhout, 95 , ! I ! ! i Kariye Camii) i '10m Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT Fig. 7 S. NELSON Hagia Sophia, Deesis 96 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham Fig. 8 THE CHORA AND THE GREAT CHURCH Hagia Sophia, South Gallery 97 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT S. NELSON Fig. 9 Hagia Sophia, Reconstruction of the Deesis panel by Whittemore (Mosaics of Hagia Sophia IV) 98 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Fig. 10 Hagia Sophia, Mosaic over Imperial Door 99 Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham ROBERT F ig. 11 S. NELSON Chora, Theodore Metochites 100 and Christ Published by Maney Publishing (c) Centre for Byzantine, Ottoman and Modern Greek Studies, University of Birmingham THE CHORA AND THE GREAT CHURCH Fig. 12 Hagia Sophia, Mosaic in Southwest Vestibule 101