The Transfiguration at Shivta. Retracing early Byzantine iconography*

Emma Maayan Fanar** University of , Art History Department

UDC 75.033.2(569.492 Shivta) 75.046.3:27–312–6 DOI 10.2298/ZOG1741001M Оригиналан научни рад ‘Tell the vision (to horama) to no man, until the Son of Man be risen again from the dead.’ Mt 17, 9

The Transfiguration constitutes one of the most important ently quite rich, as attested by its three churches, built ac- events in the New Testament. Yet, only few pre-iconoclastic cording to the practice current elsewhere in the Land of examples of the Transfiguration scene have survived: S. Apol- ; a monastic complex was possibly attached to the linaire in Classe, Ravenna, St. Catherine Monastery, Sinai and Northern church.3 That number of churches in a small Poreć in Istria, each has its unique iconography. Therefore, place was not unusual for the Byzantine settlements in scholars have concluded that the Transfiguration scene became 4 widespread only after the iconoclastic controversy. We aim to Palestine, but all three Shivta churches were much invest- show, that Transfiguration scene in Shivta, an early Byzantine ed and painted, with geometric floor mosaics, while the 5 settlement in the desert, allows a glimpse into the early three apses of the Northern church were clad in marble. Christian iconography of the well-known scene, providing a Contrary to neighbouring Nessana and other Negev missing link to its development in the post-iconoclastic period. settlements,6 Shivta seems neither especially important nor Keywords: Shivta, Transfiguration, Early Byzantine iconogra- situated on any important roads or trade routes.7 Hirschfeld phy, Negev suggested that Shivta could have been on the pilgrimage route from through Rehovot-in-the-Negev and Nessana to Sinai (the route taken by Antonius of Piacenza Shivta in 560).8 This source however is unclear, and probably refers Shivta (Sobbota/ Soubaita/Esbeita) was a rural set- to Mitzpe Shivta and to the hostel of St. George, as Figueras 9 tlement in the Negev from the fourth-fifth to eighth-ninth argues. Precisely these features make the Transfiguration centuries and then abandoned. It was not very large, with scene found in the lateral apse of its Southern church so about 170 houses, some two-storied, housing approxi- significant for this research, as it still preserves remnants of mately 2200 people.1 Shivta probably became Christian pre-iconoclastic iconography lost everywhere else. sometime in the fifth century.2 Its population was appar- market and was eventually rescued by the bishop of Elusa. Analysing this story, Mayerson concludes that at the time of its invention Shivta was * This study is a part of the ‘Byzantine bio-archaeology research entirely pagan. But the date of the Narration as well as its attribution to program of the Negev’ (BYBAN), launched in 2015, which aims to discover St. Nilus (died after 430) are disputed by scholars. Cf. Ph. Mayerson, The possible causes of collapse of Byzantine settlements in the Negev (http:// desert of Southern Palestine according to Byzantine sources, Transactions negevbyz.haifa.ac.il/index.php/en/). My gratitude goes to Prof. Guy Bar and Proceedings of the American Philological Association 95 (1964) Oz, Dr. Yotam Tepper and Dr. Ravit Linn. It is my belief that a combined 166–167; D. F. Caner, History and hagiography from the late antique Si- effort of scholars from three disciples—archaeology, conservation and art nai, Translated Texts for Historians 53, Liverpool 2010, 73–135. 3 history—made it possible to understand more fully significance of the R. Erez-Edelson, Sedentary zone in the Negev during the Transfiguration scene in Shivta for the study of early Byzantine iconography. Byzantine period. Settlement distribution and runoff farming, Ramat- ** [email protected] Gan 2004 (doctoral dissertation, University of Bar Ilan, in Hebrew), 1 128; J. Shereshevsky, Byzantine urban settlements in the Negev Desert, Y. Hirschfeld, Man and society in Byzantine Shivta, Qadmo- Jerusalem 1986, 195–199 (in Hebrew). niot 125 (2003) 12 (in Hebrew). For the main scholarly sources on 4 Y. Tsafrir, Some notes on the settlement and demography of Shivta v. A. Segal, Shivta. Plan and architecture of a Byzantine town in Palestine in the Byzantine period. The archaeological evidence, in: Re- the Negev, Jerusalem 1981; idem, The Byzantine city of Shivta (Esbeita), trieving the past. Essays on archaeological research and methodology in Negev Desert, Israel, Oxford 1983, idem, Shivta. A Byzantine town in honor of Gus W. Van Beek, ed. J. D. Seger, Starkville, MS 1996, 278. the Negev Desert, JSAH 44 (1985) 317–328; idem, Architectural decora- 5 tion in Byzantine Shivta, Negev Desert, Israel, Oxford 1988; A. Negev, R. Stroumsa, People and identities in Nessana, Durham, NC 2008 (doctoral dissertation, Duke University), 40. Subeita, RB 81 (1974) 397–420; idem, The Greek inscriptions from the 6 Negev, Jerusalem 1981; Y. Kedar, Ancient agriculture at Shivtah in the Ibid., 39–40. 7 Negev, Israel Exploration Journal 7 (1957) 178–189; D. Chen, Byzan- Segal, Shivta. A Byzantine town, 317–328. tine architects in and Sobota, Palaestina Tertia, LA 31(1981) 8 Y. Hirschfeld, The monasteries of Gaza. An archaeological re- 235–244; G. Crowfoot, The Nabatean ware of Sbaita, Palestine Explora- view, in: Christian Gaza in Late Antiquity, ed. B. Bitton-Ashkelony, A. tion Fund Quarterly Statement 68 (1936) 14–27. Kofsky, Leiden 2004, 66. 2 The only existing account mentioning Shivta is the late fourth– 9 P. Figueras, Monks and monasteries in the Negev Desert, LA to early fifth-century Narration by Pseudo Nilus. It tells the story of St. 45 (1995) 420–423; for an analysis of the itinerary and of conditions in Nilus’ son Theodoulus, who was sold into slavery at the Sobota [Shivta] the region v. Mayerson, op. cit., 169–172. 1 ЗОГРАФ 41 (2017) [1–18]

Fig. 1. Shivta, Southern Church, overview (photo: Dror Maayan)

The Transfiguration in Shivta. A missing link Earliest notice of the painting between pre-iconoclastic and post-iconoclastic The painting was first recorded by E. H. Palmer, an iconography English explorer who visited Shivta in 1870. His adven- tures are colourfully described in The desert of the Exo- Hardly anything remains of the earlier beauty dus. The Shivta ruins are called the ‘most imposing and of Shivta’s churches (fig. 1). Architectural ornaments, considerable of any which we had seen’.11 Palmer was ac- as well as small pieces of surviving paintings, suggest companied by photographer C. F. Tyrwhitt-Drake, who different styles and perhaps even dates. The Northern photographed Shivta extensively; some of the photos church deserves special attention, which is beyond the were published in the book. These pictures are invaluable scope of the present study and will be discussed sep- since they show the place in its original state in the nine- arately. It is also impossible to reconstruct the icono- teenth century, before the extensive reconstruction there graphic program of the Southern church as it has been from the 1930s onward. Palmer testifies to the existence ruined by the climate, natural disasters and the icono- of the wall paintings, but without identifying them: ‘some clasm, or even several iconoclasms, down the centuries. rude paint ornamentation still visible upon a small arched Quite clearly, the wall paintings were deliberately de- niche in the centre, and also some vestiges of a fresco.’12 stroyed; someone was at pains to scrape the paint off It was only in 1914 that the wall painting was de- almost completely. scribed and identified as the Transfiguration by Leonard Apparently, the central apse was first plastered Wooley and T. E. Lawrence: then painted, while paint seems to have been applied directly to the stones in the lateral apses,10 perhaps at- On the central apse no more than a few faint testing to different dates. The state of preservation, trances of colour survived. In the southern apse alone however, does not allow any explicit conclusions. Some could any coherent design be distinguished, and here pieces of coloured plaster are still visible in the central the colours had faded under exposure to the light, apse, showing some use of blue and red pigments. An most of the surface had been scraped away by icono- arrangement of red lines and tiny spots of colour still clasts, and rain-water had brought down lime from visible on the northern lateral apse may suggest the the upper ruins and left a thick white deposit over presence there of a large central motif flanked by some the whole wall-face. Only by wetting the stone were motifs on either side; but no further identification can we able to make out and roughly to sketch the origi- be made. The image on the southern apse comes as nal painting. The subject was the Transfiguration. In a complete surprise. For whatever reason it has sur- the centre is Christ, full-face, with hands raised and vived—to the point that it can still be recognized de- brought together over the breast, The chiton was spite heavy damage. From that surface the scene of the Transfiguration emerges (fig. 2). 11 E. H. Palmer, The desert of the Exodus. Journeys on foot in the wilderness of the forty years’ wanderings, London 1871, 375. 2 10 I am grateful to Dr. Ravit Linn for this observation. 12 Ibid., 377. Maayan Fanar Е.: Th e Transfi guration at Shivta

Fig. 2. Shivta, Southern Church, Southern apse, overview (photo: Dror Maayan)

seemingly of light pink edged with gold, the himati- In the early 1930s Colt excavated Shivta.15 De- on of dark blue; the halo was a plain yellow ring with spite several seasons of work, hardly any documentation white centre; the vesica of light pink. The figure was has survived; most of the materials from this excavation too much damaged to be copied. Below the feet is a were destroyed, looted, and lost. The photos taken during semi-prostrate figure, probably of St. Peter, and be- Colt’s excavations make it clear that the Southern church yond, on the spectator’s left, a kneeling figure iden- was then half buried in rubbish, and later cleaned and re- tified by a fragmentary inscription ...ANNIC, in red stored. Colt does not mention the painting but Baly, who paint, as S. John; turning half-round to the front, he was working with him, does: raises his left hand, as if pointing to Christ. A few lines The South Church is of the same triapsidal on the right of the vesica are all that is left of S. James. form as the North. In the south apse Woolley notes On either side of the apse, a little distance from the a scene which he suggests is the Transfiguration; it central group, a blurred mass of red colours seems to has suffered badly since his time, but there are still represent figures standing on a slightly higher level traces of paint including a fragment of the name of than the Apostles: these are probably Moses and Eli- St. John. ... it is clearly to be seen where Woolley jah. Below the feet of the figures is a broad red band. places it.16 The tooth pattern around the arch of a small recess was picked out in red and blue, and its vault was The painting lay derelict for more than seventy roughly painted in red with a coarse network pattern, years thereafter. The most recent and important study each mesh having a cross as a filling ornament.13 of the Transfiguration in Shivta was published in 2006 by Pau Figueras, who alone called attention to the scene This somewhat lengthy description is not accurate itself, analysing its iconography and arguing for its early and lacks many details. Although the figures’ identification date (c. 500).17 Nevertheless, because of the scene’s state is correct, the description of their postures and colours (e.g. of preservation, its proposed reconstruction does not Christ’s garments) was partly conjectured by recourse to wholly coincide with the precise iconographic details of knowledge of later Byzantine examples of the Transfigura- the Shivta Transfiguration. Only by using most updated tion not necessarily visible in Shivta itself. Because of its poor condition, the painting could not be copied. Later it was al- 15 H. D. Colt, Isbeita, Palestine Exploration Fund Quarterly most forgotten, mentioned rarely and only in passing in any Statement 67 (1935) 9–11; idem, Isbeita, Quarterly of the Department 14 scholarly work or description. of Antiquities in Palestine 5 (1936) 198–199. Some documents and photographs of Colt’s excavations can be found in the Israel Antiqui- 13 C. L. Wooley, T. E. Lawrence, The wilderness of Zin (Archaeo- ties Authority Archive: http://www.iaa-archives.org.il/search.aspx?loc_ logical report), London 1914, 89–90. An online version: http://www3. id=15&type_id= lib.uchicago.edu/cgi-bin/eos/eos_title.pl?callnum=DS111.A1P28_ 16 T. J. Baly, S’Baita, Palestine Exploration Fund Quarterly vol3_cop1 Statement 67 (1935) 176. 14 R. Rosenthal-Heginbottom, Die Kirchen von Sobota und die 17 P. Figueras, Remains of mural painting of the Transfiguration Dreiapsidenkirchen des Nahen Ostens, Wiesbaden 1982, 73 n. 1. in the southern church of Sobata (Shivta), ARAM 18/2 (2006) 127–151. 3 ЗОГРАФ 41 (2017) [1–18]

Fig. 3. Shivta, Southern Church, South Apse, Christ (detail) (photo: Dror Maayan)

photographic equipment could we recreate the scene to Despite its current state of preservation, the Trans- optimal precision. Our reconstruction is based, in ad- figuration scene can still be recognized beyond any doubt dition to the observation in situ, on the analysis of high as its general composition is in line with later examples quality photographs of the apse in general and each stone of the same scene. In fact, it is most surprising to find a in particular to turn remnants of lines and colours into well-established iconographic scheme of the Transfigura- forms and objects. It will be further shown that the scene tion scene prior to the Iconoclasm, especially because it that eventually emerges is extremely significant, in the lo- appears to be a unique survival of such a scheme, as we cal context but also in the context of the development of shall see. The central figure of Christ enclosed in a man- early Byzantine iconography in general. dorla and the figures of the two apostles on the left are still clearly visible, as is part of the Greek inscription iden- tifying one of the latter as St. John. Traces of a third figure Description and style on the left may be identified as Moses or Elijah. Here we describe the Shivta Transfiguration and We now turn to detailed analyses of the scene and the parts which can be reconstructed with certainty. The its stylistic peculiarities. iconoclasts did ‘a very good job’ trying to erase utterly Christ inside the mandorla is clearly visible in the the holy images. They applied themselves to some details upper part of the apse’s centre (fig. 3). His frontal figure more than to others, suggesting that they were religiously predominates and is somewhat disproportional. He has a important. We shall use this knowledge in our attempt to relatively small head, placed over a heavy short neck, and enclosed in a great halo painted in shades of light pink.18 reconstruct the images. To better understand the scene, we conducted a comparative analysis with known rep- Only a small fragment remains of Christ’s face, es- pecially visible to the viewer’s left (right side of the paint- resentations of the Transfiguration that have survived in ing) in the area of the upper part of an eye and an ear, architectural decorations, illuminated manuscripts, and where long wide strokes of yellow, brown, and dark brown icons. Shivta’s Transfiguration is studied in comparison overlap (fig. 4). Christ’s face was painted in shades of with the sixth-century mosaics in the apses in St. Cath- pink; above his only surviving eyebrow long white strokes erine Monastery in Sinai (548–65) and with S. Apollinaire of highlights are attested. Some dark curves of his curly in Classe in Ravenna (c. 549), the only remaining exam- hair are still detectable around his face and in the chin ples of the scene from the pre-Iconoclastic period (figs. area, seeming to indicate a short curly beard. 11, 12). Middle Byzantine manuscripts, icons, and wall Wooley and Lawrence indicated that Christ’s himat- paintings, especially in Cappadocia (tenth – twelfth c), as well as mosaics such as in the eleventh-century Daphni ion was dark blue, but it was more likely painted in shades monastery near Athens (fig. 14), provide valuable com- of pink with white highlights, as traces of pink are attested parative material, which is here critically scrutinized in along his figure. The colour of Christ’s chiton is also un- the knowledge that it postdates Shivta’s Transfiguration. clear, as traces of light purple in the area of his right leg Descriptions of the Church of the Holy Apostles in Con- 18 I am unable to see ‘a plain yellow ring with white centre’ de- 4 stantinople, now lost, might also be useful in our context. scribed by Wooley and Lawrence. Maayan Fanar Е.: Th e Transfi guration at Shivta are attested among pink spots. Lines indicating falls help to reconstruct exact place of Christ’s legs and draperies fold around his knees. In contrast to the heavy garments his legs seem disproportionally tiny. From other remaining outlines Christ’s right hand can be reconstructed (fig. 3). Proportionally enlarged, it extends from a heavy sleeve, raised up to the level of his chest. The gesture may well resemble that of Christ in the St. Catherine mosaics in Sinai (fig. 11): the inner side of a palm with fingers joined in the so-called Greek bene- diction. We can also follow the curve of his left hand, as well as a prominent round curve of drapery above his left palm. However, the palm itself was damaged beyond rec- ognition, while a mixture of traces of red and dark blue extends beyond his garments. In Byzantine Transfigura- tions from Sinai mosaics, the ninth-century illuminated manuscripts (e.g., Paris, BNF, gr. 510, fig. 13), and later images, the position of Christ’s right hand varies from im- age to image, but the position of his left hand, with mi- nor variations, is the same repeatedly. Due to the icono- clasts’ relentless effort to eliminate Christ’s left hand and surrounding area, we may argue that it could have held a scroll or a book. The mandorla surrounding Christ seems to be more round than oval and most probably was painted light pink (although in some places, dark red can be attested), its contour executed in white. No traces of gold described by Wooley and Lawrence have been identified so far. It is not clear if the mandorla was two circles of different colours or solid of one pink colour. In later examples of the scene both options can be seen, although the use of light pink is Fig. 4. Shivta, Southern Church, South Apse, Christ (detail) rare. The earliest known example is in Paris, BNF, gr. 510 (photo: Dror Maayan) (fig. 13). Some barely visible lines might suggest rays of light radiating from Christ’s figure.19 Under the mandorla, on the left, two figures of the towards Christ, remnants of fingers still visible close to apostles are still visible, both lacking halos (figs. 5–6). the curve of John’s back: two fingers extend forward while John can be recognized by last letters of his name in- the little finger is in the direction of his knee; the fourth scribed next to the lower part of his body. He is depicted finger is bent. It is not clear if his left hand is also raised prostrate, his right hand reaching forward (fig. 5). He too towards Christ or descends down the side his body. Some has a relatively small narrow face. A short dark beard is still visible lines can be associated with an elevated sleeve. just traceable. John raises his left hand, most probably to The rest of the hand cannot be reconstructed. Later ex- support his head. He is not ‘pointing to Christ’, as Wooley amples of the Transfiguration usually show Peter with his and Lawrence suggested; the hand is covered with a dra- left hand down the side his body, at times holding the dra- pery, an iconographic detail which exists in many middle peries of his dress (e.g., Paris, BNF, gr. 510 (fig. 13), ninth Byzantine examples – e.g., Daphni monastery (fig. 14); c.; Karanlik Kilise, Goreme, eleventh c.21; the Louvre icon, mosaic icon from the Louvre, twelfth century.20 twelfth c.22). Nevertheless, in the ninth-century Chludov To the left, behind John, is the figure of Peter, rec- Psalter (Moscow, State Historical Museum, MS 129, fol. ognized by the remnants of a beard and white curly hair 88v)23 the left hand is depicted somewhat elevated as well. (fig. 6). The upper part of his body is still traceable, while Traces of white, blue and purple of Peter’s attire together the lower part has disappeared almost completely. He is with the colours of his beard suggest that these are the most probably kneeling, as there is not enough room for colours of his chiton. The colour of the himation cannot a standing figure. His posture seems to resemble Daphni be determined with certainty. (fig. 14). Peter’s figure is turned towards Christ but is not The two figures have enlarged palms and narrow full profiled as in later examples. His right hand gestures faces; their bodies are rounded with folds following the limbs, providing sense of volume and displayed in dy- 19 For the rays of light in Shivta's Transfiguration see: R. namic half-turns. Peter is depicted behind John, partly Linn, Y. Tepper, G. Bar-Oz, Visible induced luminescence reveals invis- hidden by him, contributing to a sense of depth in the ible rays shining from Christ in the Early Christian wall painting of the wall painting. Because Peter’s figure is in a better state of Transfiguration in Shivta, Plos One 12/9 (2017) e0185149. https://doi. org/10.1371/journal.pone.0185149. 21 20 Musée du Louvre, Département des Objets d’ Art, Paris, C. Jolivet-Lévy, La Cappadoce Médiévale, images et spiritua- France. Cf. The glory of Byzantium. Art and culture of the middle Byz- lité, Paris 2001, fig. 80. antine era AD. 843–1261, ed. H. C. Evans, W. D. Wixom, New York 22 V. n. 19. 1997, cat. no. 77. 23 M.V. Scepkina, Miniatjury Khludovskoi Psaltiri, Moscow 1977. 5 ЗОГРАФ 41 (2017) [1–18]

Fig. 5. Shivta, Southern Church, South Apse, John (detail) (photo: Dror Maayan)

preservation than the rest of the painting, one gets a sense of volume in the rendering of his body and an illusionistic touch in the surviving fragments which preserve a mix- ture of colours (fig. 7). James’s figure has been erased en- tirely and cannot be reconstructed, although its position on the lower right part of the apse can be determined. It would not be farfetched to propose that his appearance was not very different from other ninth– to twelfth-centu- ry Byzantine representations. The apostles’ compositional arrangement differs from the well-known Byzantine scheme, where, their postures notwithstanding, all three apostles are por- trayed separately, evenly placed around Christ’s figure. Fig. 6. Shivta, Southern Church, South Apse, Peter (detail) St. Catherine preserves the earliest example of this kind. (photo: Dror Maayan) An asymmetrical arrangement of the apostles is found in early Western examples of the Transfiguration. In general, this hilly landscape survives in endless variations. Howev- in the West the Transfiguration did not enjoy much popu- er, in most cases predominates in the land- larity. Its early (ninth– and tenth-century) examples show scape, becoming a necessary iconographic element. inconsistency. Most reflect Byzantine models (e.g., the Gospels of Otto III (Munich, Bayerische Staatsbibliothek, The figure to Peter’s left is depicted full-length, Clm 4453, fol. 113);24 ivory from Mets (London, Victoria standing, next to the border of the apse, his head sur- and Albert Museum, ca. 870–880).25 But an asymmetrical rounded by a huge pink halo (fig. 8). The figure’s overall compositional arrangement of the apostles, already attest- appearance and the prevalent brown colours of his gar- ed in S. Apollinaire, where they are depicted symbolically ments suggest that it is Elijah. Already in St. Catherine, as sheep (fig. 12), persisted into the tenth century, con- Elijah is depicted in distinctive garments, in symbolic trasting the more symmetrical one in Byzantium. In Shiv- reference to his identification with John the Baptist (Mt ta the apostles’ compositional arrangement recalls these 17,10–13). This iconographic element survives into later Western examples, suggesting a common source: Peter scenes of the Transfiguration (e.g., Daphni, fig. 14). Rem- and John are depicted together on the left, while James is nants of lines in the area of Elijah’s head suggest that he is on the right. This arrangement however is unique: usually bearded. His eyes, eyebrows, nose, and mouth are barely Peter is depicted separately from other apostles. traceable due to the visibility of the initial sketch in ochre in a yellow tone. Elijah’s face seems painted in light pink, Shades of green, barely visible today, surround the almost identical to the colour of his halo; however, the col- figures of the apostles as they are seen against a greenish ours of his face or hair cannot be reconstructed with cer- landscape background which occupies all the lower part tainty. At least one of his hands seems to stretch upwards, of the scene bordering on Christ’s mandorla (figs. 6–7). It toward Christ’s mandorla: remnants of colour extend far is not clear if a single mountain is depicted under Christ’s from his body toward it. Comparative analysis of the dis- feet. This green landscape, which was probably much position of surviving small fragments of colours suggest more detailed, recalls S. Apollinaire without any moun- that the figure might resemble that in Daphni, showing tain. In the middle Byzantine iconography of the scene, one of his hands stretched towards Christ, while his other (left) hand is either enclosed in heavy draperies or placed 24 https://iconographic.warburg.sas.ac.uk/vpc/VPC_search/re- above draperies, as in Paris, BNF, gr. 510 (fig. 13). sults_advanced_search.php?p=1&msn=804 [8.8.2017]. The remains of the figure of Elijah show that he 25 http://collections.vam.ac.uk/item/O88287/the-transfigura- 6 tion-book-cover-unknown/ [8.8.2017]. is depicted disproportionally huge. It is not clear if he is Maayan Fanar Е.: Th e Transfi guration at Shivta slightly elevated or shares the same level with the apostles. Although in some later depictions of the Transfiguration the biblical figures are depicted larger than the apostles, their elevation is emphasized by their positioning closer to the celestial realm. The closest comparison to Shivta can be found in the sixth-century St. Catherine: huge fig- ures of Elijah and Moses stand on the same green fore line as the apostles (fig. 11). Some similarity in positioning all the figures around Christ on the same level can be seen in the ninth-century mosaics in Rome: St. Nereo ed Achilleo (795–815)26 and S. Zeno Chapel in St. Prassede,27 or in the Old Tokali Kilise in Cappadocia.28 In all these cases, however, this compositional arrangement reflects adapta- tion of the scene to the architectural forms, as they are de- picted on niches and arches. At the two far edges of the apse there are traces of different colours (blue, green, and violet/dark red). Al- though the traces are scarce, they might be remnants of trees, seen in many Byzantine Transfigurations. We can thus summarize our discussion by propos- ing a reconstruction of the Transfiguration scene at Shivta (fig. 10). Unlike previous descriptions and reconstruc- tions, most evident in the present one is the placement of all the figures surrounding Christ on approximately the same level, without (or with slight) elevation of the prophets. Furthermore, Peter’s and John’s gestures differ Fig. 7. Shivta, Southern Church, South Apse, Peter (detail) from earlier proposals. (photo: Dror Maayan) The entire painting gives a dynamic impression due to its asymmetrical composition, the figures’ movements huge halos, small heads, bold ears, facial details of Christ and the vivid colours. Massive dynamic figures with vol- such as short beard and curly hair, and crude, dynamic, ume and half-round turns, overlapping in space, heavy bold and heavy strokes to indicate eyes and eyebrows. draperies, proportional distortions, merging of colours – all point to an early date of the wall painting. To this Unfortunately, we lack sufficient comparative mate- we may add a three-dimensional illusionistic imitation of rials from the Syro-Palestine area. Roman frescoes from 32 an architectural strip of the dentils29 at the bottom edge Mamshit differ in style. Nor can the Shivta painting be of the apse (fig. 9). They are painted white where they paralleled to the frescoes of the fourth– to fifth-century project and tones of grey and red where they are shaded. Lochamei HaGetaot tomb33 or to the sixth-century saintly Dark blue bands are painted above and below the dentils, figures at Caesarea,34 which are much more flat and still, terminating in wide red bands at its lower edge. The same and have different proportions. Fragments of Peter’s beard architectural imitation is attested on the northern apse, painted gently with a wide range of interwoven colours, suggesting that both wall paintings were executed at the suggest that the painting was by a skilled artist, reflecting same time. Although dentils can be frequently found on high quality elaborate work.35 This is in contrast to the architectural stone decorations in Shivta and other Negev rough and somewhat primitive style of the architectural cities, in the wall painting they seem much closer to the carvings at Shivta or elsewhere in the Negev, which sug- classical type. gest a local, provincial workshop.36 To date there is not Stylistically, Shivta’s Transfiguration shares signifi- enough evidence to determine if the artist belonged to cant similarities with St. Catherine’s mosaics, especially in respect of the figures’ proportions. Still, the huge, heavy 32 http://www.cca-roma.org/en/mamshit-conservation-naba- palms and relatively small faces point to a connection taean-frescoes [5.6.2017] with sixth-century Syriac manuscripts such as the Rab- 33 E. Maayan Fanar, Early Christian tomb at Lohamei HaG- bula Gospels.30 Christ’s depiction in a Syrian Four Gospel etaot. Formation of Christological symbolism, ECA 7 (2010) 71–89. 34 Book in Diyarbakır attributed to the early sixth century Chronicles of the Land. Archaeology in the Israel Museum Je- 31 rusalem, ed. M. Dayagi-Mendels, S. Rozenberg, Jerusalem 2010, 172– seems even closer. The two share disproportionately 173, fig. 13. 35 The figures were sketched in yellow and further detailed in 26 E. Trunø, The apse mosaic in early medieval Rome, Cam- dark red. The heavy, somewhat rough strokes seen today give the false bridge 2002, fig. 37. 27 impression that the whole work is not highly professional. However, let G. Mackie, The Zeno chapel. A prayer for salvation, Papers of us bear in mind that these lines were intended to remain as an under- the British School at Rome 57 (1989) pl. 34. painting covered with layers of paint now lost. 28 A. W. Epstein, Tokali Kilise. Tenth-century Metropolitan art 36 An extensive corpus of architectural decorations from Shiv- in Byzantine Cappadocia, Washington D.C. 1986, fig. 32. ta and other Negev desert cities was published by Segal, Architectural 29 Dentils usually decorated the upper part of the entablature decoration. Recently this subject was renewed by Karni Golan, The in ionic and Corinthian buildings. stone architectural decorations of the Byzantine Negev: Characteriza- 30 Florence, Biblioteca Mediceo Laurenziana, cod. Plut. I, 56. tion and meaning, Beer Sheva 2014 (doctoral dissertation, Ben Gurion 31 M. Bernabò, G. Kessel, A Syriac Four Gospel Book in Diyar- University of the Negev). She published an updated catalogue as well as bakir, Convivium 3 (2016) 172–203. an in-depth discussion of main motifs found on stone decorations in 7 ЗОГРАФ 41 (2017) [1–18]

Fig. 9. Shivta, Southern Church, South Apse, dentils (detail) (photo: Dror Maayan)

ground, the history of the Transfiguration’s visual repre- sentation is not clear. Only in the tenth century did the scene start to be regularly depicted as a part of the Byzan- tine Christological cycle and as one of the twelve liturgical feasts. In his study, Andreopoulos sets out to detail the de- velopment of the iconographic tradition of the scene of the Transfiguration from its beginning to the fourteenth century.39 By his analysis, the scene has five iconographic stages of development. Its pre-sixth-century depictions are scarce, and they differ from the later ones in having only three figures.40 This can be attested in the fourth- Fig. 8. Shivta, Southern Church, South Apse, Elijah (detail) century wooden doors of St. Sabina and the fifth-century (photo: Dror Maayan) Brescia casket. The iconography of both can be connected to later Transfigurations only vaguely. Andreopoulos calls a local Syro-Palestine school or came from afar, perhaps them pre-Sinaitic, thereby marking a definite borderline from one of the centres in northern Syria. and even suggesting that it might actually be impossible to hypothesize about ‘the very existence of a pre-Sinaitic type of Transfiguration’. The Transfiguration in Shivta and the sixth The Transfiguration also appears in the sixth-cen- century iconography tury Rabbula Gospels, where it is the earliest surviving example of the scene in the illuminated manuscripts (fol. Transfiguration is one of the formative events in the 7a). Here too it has only three figures, whose identities are New Testament and is told in three synoptic Gospels (Mt disputed. Significantly, the scene appears on the margins 17, 1–13, Mc 9, 2–13, and L 9, 28–36). By the fifth cen- and was not singled out as Crucifixion-Resurrection or tury, Mount Tabor was accepted as the place where the Ascension, painted on separate folios. Rabbula was previ- Transfiguration took place, and by the sixth century, three ously dated to A.D. 586, according to the colophon which churches had been built there to commemorate this event. appears elsewhere in the manuscript. It is clear today that By the eighth century, the feast of the Transfiguration had the illustrations originally came from another manuscript been introduced in Constantinople, probably deriving and were added to the text later. Still, illuminations prob- from the sixth-century Palestinian feast of Tabernacles.37 ably belong to the sixth century and in fact may be even Early Church Fathers emphasize the significance of earlier than previously thought.41 the Transfiguration as Revelation or Epiphany, and inter- Judging from archaeological and written accounts, by pret it in eschatological terms, linking it to the mystery of the mid-sixth century the Transfiguration scene was already 38 the Second Coming. Despite the rich theological back- in use in monumental art, although surviving examples are scarce. One, in Porec (Parenzo) in Istria, is not visible today. the Negev. The proper chronology of the architectural decorations still awaits further research. It occupied the upper part of the church’s external eastern wall 37 Transfiguration, in: ODB III, ed. A. Kazhdan, A.-M. Talbot, New York–Oxford 1991, 2014–2015 (G. Podskalsky, R. F. Taft, A. Weyl Carr). 39 A. Andreopoulos, Metamorphosis. The Transfiguration in 38 I. Marinoff, The Transfiguration in the Byzantine liturgy, Life Byzantine theology and iconography, Crestwood – New York 2005 (with of the Spirit 10 (1955) 18–22; A. Paretsky, The Transfiguration of Christ. previous bibliography). Its eschatological and Christological dimensions, New Blackfriars 72/851 40 Ibid., 75–76. (1991) 313–324; D. Lee, On the Holy Mountain. The Transfiguration in 41 M. Bernabò, The miniatures in the Rabbula Gospels. Post- 8 scripture and theology, Colloquium 36/2 (2004) 143–159. scripta to a recent book, DOP 68 (2014) 343–358. Maayan Fanar Е.: Th e Transfi guration at Shivta

Fig. 10. Shivta, Southern Church, South Apse, proposed reconstruction (Emma Maayan Fanar) and is done in mosaic.42 In a 1940s sketch, Moses and Peter In Sinai, the frontal figure of Christ is flanked by stand at Christ’s right, together with the figure of St. Andrew standing figures of Elijah (to Christ’s right) and Moses, who has no connection with the original scene and does not and the three apostles, all positioned against a gold back- appear in any known depictions of the Transfiguration.43 ground in a semicircle surrounding the central figure of The ninth-century account preserves evidence of Christ. John and James mirror each other on their knees the existence of yet another sixth-century example of the with elevated hands in the orans position, while Peter is Transfiguration, in Chiesa del Salvatore, Naples.44 No de- depicted below Christ’s feet, raising his head and turn- tailed description of it has survived.45 We are left with ing his face upwards. Repeatedly, scholars insist that the two survivals: apse mosaics in S. Apollinaire in Classe Transfiguration in St. Catherine is a borderline in the in Ravenna (fig. 12), and in St. Catherine Monastery in scene’s iconography, deeming it closely comparable to the 47 Sinai (fig. 11). In S. Apollinaire, Christ’s bust is depicted iconography of the post-Iconoclastic art in Byzantium. inside a huge cross, flanked by half-figures of Moses and A symbolic connection of the Transfiguration on Mount Elijah in a celestial realm, while apostles symbolically de- Tabor and in Sinai was widely discussed by early Chris- picted as sheep stand in a flowery garden. Although the tian theologians, so the choice of the scene for the main Transfiguration occurred on Mount Tabor, no mountain apse of the church is not surprising. As Jas Elsner showed, is attested here, nor in Sinai or Porec. Beneath the Trans- the message of the Sinai image is better understood in the figuration scene, in the same garden with the apostles, S. context of the entire arrangement of images, including Apollinaire, patron of the church, is depicted in an orans Moses receiving the Law and Moses at the Burning Bush, 48 posture. Thus, the message of the mosaics is threefold: the on the arch. Thus, the iconographic program of the Transfiguration is combined with Exaltation of the Cross sixth-century church is well thought out and in harmo- and martyrdom.46 Andreopoulos concludes that the ny with the significance of the place where God revealed Transfiguration in S. Apollinaire has a distinctive iconog- Himself to Moses through symbols and gave him the Law. raphy and ‘didn’t leave any artistic descendants’. God’s revelation through Christ overshadows the Old Tes- tament revelations, stressing the power of Christianity. 42 For Porec: B. Malajoli, La basilica eufrasiana di Parenzo, Let us take a closer look into the Sinaitic mosaic. It Parenzo 1940, 33 and fig. 33; J. Maksimović, Ikonografija i program seems to encompass most iconographic details associated mozaika Eufrazijeve bazilike u Poreču, ZRVI 8/2 (1964) 247ff; Els- ner, Image and Iconoclasm in Byzantium, Art History 11 (1988) with the Transfiguration, as we know them from later Byz- 474–477. antine examples, yet it differs from them in details. Being 43 Elsner, op. cit., 474. closer than S. Apollinaire to the Gospels’ narrative, it in 44 J. Miziolek, Transfiguratio Domini in the apse at Mount Sinai fact rearranges familiar elements of the story to adapt it to and the symbolism of light, JWCI 53 (1990) 45. the static hierarchic composition centred on the figure of 45 ‘This John (A.D. 535–555) rebuilt the apse of the Church of Christ. John and James mirror each other in posture and St. Stephen which had collapsed after a fire. There in mosaic he had depicted the Transfiguration of our Lord...’(Gesta Episcoporum Nea- gesture—contradicting the Gospel narrative, but also the politanum I.16.xxii, ed. G. Waitz, in: Monumenta Germaniae Historica. depiction of Peter under Christ’s feet is unique and occurs Scriptores Rerum Longobardum et Italicarum Saec. VI–IX, Hannover 1878, 410). After Elsner, op. cit., n. 19. 47 Andreopoulos, op. cit., 101. 46 Elsner, op. cit., 474–475. 48 Elsner, op. cit., 475. 9 ЗОГРАФ 41 (2017) [1–18]

sage. Peter’s prostrate body looks like a reworking of John’s posture; however, the position of his head is different, looking upwards to accord with the words of the Gospel, while the legs are more stretched, compositionally bridg- ing two other apostles to construct a visual circle around the central figure of Christ, echoing the mandorla’s shape. The drapery held by John becomes pointless here, not suited to the position of Peter’s head. Mango also chal- lenges the date and the authorship of the mosaic, pointing out that no reliable information exists to argue that Jus- tinian was in fact its patron. He posits that it could have been completed after the Church’s completion, which was a Justinian enterprise, but probably before the detachment of the territory from Byzantium in the seventh century. Thus, the date of the mosaics can vary: 565–566, 580–581, 595–596 or even later. Mango concludes that the mosaic was a local enterprise while the artist could have come from Alexandria or Palestine. Fig. 11. Transfiguration, Apse mosaics, St. Catherine Of all the surviving examples, the Sinai Transfigu- Monastery, Sinai, 6th c (after: G. H. Forsyth, K. ration is the only representative from the eastern part of Weitzmann, The Monastery of Saint Catherine at the empire. Geographically, it is the surviving depiction Mount Sinai, Ann Arbor 1973) of the scene closest to that at Shivta. It is not clear if the Shivta Transfiguration is slightly earlier than the Sinai, but in no other scene of the Transfiguration. Peter’s posture the difference between their visual interpretations of the has been explained as a substitute for the mountain, as he Transfiguration is remarkable. While Sinai uses the Trans- symbolizes the rock upon which the Church is to be built. figuration to emphasize the Church’s symbolic message, The figure itself is much distorted. The artist possibly and thereby recedes from the Gospel narrative, Shivta’s Transfiguration is closely linked to it, clearly distinguish- tried to combine different motions in Peter’s figure – he ing the apostles’ gestures and movements. is ‘heavy with sleep’ (L 9:32) – as he struggles to awake.49 Peter turns his head, almost detached from his body, to- Peter turns to Christ in a speaking gesture, as if say- wards Christ, perhaps to represent speech. Elsner relates ing ‘Lord, it is good for us to be here’ (Mt 17, 4). John Peter’s posture to the Heavenly ladder. Sleep can hinder is kneeling, not looking upwards indicating his awe, in prayer and must be combated.50 ‘The dominant message keeping with the words of the Gospels. Although nothing of the scene involves not sleeping but hearing and seeing remains of Moses’ figure except a partial sketch, it suf- fices to conclude that he was depicted extending his right God... This aural and visual perception of the divine is hand to Christ. Presumably Elijah did the same. Never- shared with the two Moses scenes.’51 theless, the depiction of the prophets and the apostles on Weitzmann, who studied the Sinai mosaic thor- the same plane is similar to the sixth-century represen- oughly, was so impressed by its quality and complex- tations of the Transfiguration in Sinai and Porec, while ity that it argued for Constantinopolitan craftsmanship, in S. Apollinaire the prophets are moved to the celestial connecting it directly to Justinian. Later studies usually realm. The green background at Shivta is comparable to 52 accepted this point of view. Some propose that it might S. Apollinaire, and perhaps likewise can be interpreted as be a copy of a well-known icon from Constantinople. a vision of Paradise mentioned in the sources and espe- Recently Cyril Mango pointed out that the quality of cially developed in the Apocalypse of Peter. It was wide- work in the Sinai mosaic is not uniform: while the fig- spread in Palestinian churches up to the fifth century and ure of Christ is excellently executed, Peter has two left even regarded as semi-canonical.54 We cannot conclude feet and only one hand, the other being unnaturally cov- with certainty whether Mount Tabor was depicted at ered by draperies.53 His figure is dynamic, distorted, and Shivta. However, if lines within the lower border of the lacking harmony. mandorla referred to it, this would be the earliest exam- The mosaic seems to be an adaptation of the re- ple of Mount Tabor included in the scene, a motif which nowned iconographic scheme to a specific overall mes- was to become inseparable from later Byzantine iconog- raphy of the Transfiguration. 49 R. Nelson, Where God walked and monks pray, in: Holy im- The Transfiguration in the ninth-century Chlu- age, hallowed ground. Icons from Sinai, ed. R. S. Nelson, K. M. Collins, dov Psalter (fol. 88v) shows clear continuity of the earlier Los Angeles 2006, 12. model55 but also may well reflect a change in the pre- 50 J. Elsner, The viewer and the vision. The case of the Sinai apse, Art History 17/1 (1994) 94–95. 54 The Apocalypse of Peter, ed. J. N. Bremmer, I. Czachesz, 51 Nelson, op. cit., 13–14. Leuven 2003, 182–186. 52 K. Weitzmann, The mosaic in St. Catherine’s monastery on 55 As scholars have noted, a pre-iconoclastic psalter existed Mount Sinai. The icons, Proceedings of the American Philosophical So- which already included many of the literal and Christological scenes ciety 110 (1966) 392–405; Andreopoulos, op. cit., 132. used in the ninth-century marginal psalters; scenes based on anti-icon- 53 C. Mango, The mosaic of the Transfiguration at St. Cath- oclastic polemics were added to it later. K. A. Corrigan, The ninth cen- erine’s, The Fortnightly Review 27 (July 2014) http://fortnightlyreview. tury Byzantine marginal psalters: Moscow, Historical Museum Cod. 129; 10 co.uk/2014/07/mango-sinai-mosaic/ [15.5.2017]. Mt. Athos, Pantocrator 61; Paris, Bibliothèque Nationale gr. 20, Ann Maayan Fanar Е.: Th e Transfi guration at Shivta iconoclastic scheme of the portrayal. The motif of Mount Tabor is highly prominent here, as is the separation be- tween Moses and Elijah flanking Christ inside a mandorla on top of the mountain, and the apostles below. The scene was apparently reversed in the copying: Peter appears on the right, and James on the left, differently from the apos- tles’ usual disposition. If we reverse the miniature we find a similarity to Shivta in the colours, the figures’ dynam- ics, their gestures and postures (John is especially close, even in the lines on his garments; likewise the colours of Peter’s beard and his chiton, his standing posture notwith- standing). The scene illustrates Ps 88,13, interpreted as a prophecy on Christ’s Resurrection.56 It can be concluded that both St. Catherine and S. Apollinaire, and perhaps other isolated examples, signify a chance of survival; they testify to a variety of Transfigu- rations, rather unique, left unfollowed in later represen- tations. Apparently they actually constitute an interpreta- Fig. 12. Transfiguration, Apse mosaics, S. Apollinaire in tion of an iconographic scheme already known, and its Classe, 6th c (By picdrops – Basilica of Sant’ Apollinare, adaptation to the unique messages of the mosaics. If in- CC BY 2.0, https://commons.wikimedia.org/w/index. deed so, it can be further suggested that the iconography php?curid=343978) of the Transfiguration was already established by about 548–549, when these mosaics were executed. Sixth, you may see Christ ascending that thrice-glorious mount of Tabor together with a Shivta’s Transfiguration and a possible chosen band of disciples and friends, altering his Constantinopolitan parallel mortal form; His face shining with rays more daz- zling than those of the sun, His garments a lumi- Light pink and white colours attested in Christ’s nous white; the great Moses as well as Elijah stand- halo, face, garments, and a mandorla in Shivta, may refer ing by Him with pious reverence;... the disciples, to the appearance of a bright (luminous) cloud of Divine upon hearing the thunder like voice, falling, face glory, Christ himself becoming brilliant light.57 Yet a down, to the ground...59 pink mandorla cannot be paralleled either to sixth-cen- tury Sinai or Ravenna or to the post-iconoclastic images It is not possible to reconstruct the scene in detail of the Transfiguration. In fact, the best, if not the only, from this general description; nevertheless, the apostles’ comparison to it appears in the Transfiguration scene postures at Shivta seemingly contradict it, as only John in the ninth-century Homilies of Gregory of Naziansus is depicted with his face down. Alternatively, a detailed (Paris, BNF, gr. 510, fol. 163, fig. 13) originating in Con- description of the scene was probably not the purpose of stantinople.58 According to some scholars, the iconog- this text, which attempted to give a broad picture of the raphy of the Paris Gregory reflects that of the now lost Christological cycle. Church of the Holy Apostles in Constantinople. Built The twelfth-century description by Nicolaos Me- as early as the sixth century, this church was considered sarites60 is much more detailed but also deeply emotional. one of the most important and influential churches in Applying ekphrasis, the description is replete with expres- Constantinople. Its decorative program was described sions that do not necessarily describe the original works, on two different occasions, in the tenth and the twelfth but being true to the genre render a vivid description of century, leaving scholars in doubt as to whether the two the events through the works of art, complementing them descriptions referred to the same works of art and when with information from religious and liturgical texts ad- exactly the church was decorated. justed emotionally. The symbolic value of the divine light The tenth-century description of the church by and apostles’ reaction to it is emphasized dramatically. Constantine of Rhodes gives a detailed account of the The apostles, blinded by the Divine cloud of light, fall to scenes of the pictorial cycle, including the Transfigura- the ground, having never seen anything like it. Peter sees tion. The description of the scene itself is brief: God most clearly, hears Moses and Elijah prophesying his end. The apostles are described prostrate, unable to Arbor, MI 1988 (doctoral dissertation, University of California 1984), 99–104. bear the blinding light. Peter springs up from the ground 61 56 M. Evangelatou, Liturgy and the illustration of the ninth-cen- speaking words, while two other apostles are struck by tury marginal psalters, DOP 63 (2009) 83. 57 Leo VI describes the Transfiguration in the Church built by Stylianus Zaoutzas with the words: ‘His form shines forth as His 59 Mango, The art, 200–201. mortal appearance is removed’ (Sermon 34) – after C. Mango, The art 60 G. Downey, Nikolaos Mesarites. Description of the Church of the 312–1453. Sources and documents, Englewood of the Holy Apostles at Constantinople, Transactions of the American Cliffs, NJ 1972, 204. Philosophical Society 47/6 (1957) 871–873. 58 C. Nersessian, The illustrations of the homilies of Gregory of 61 In the Paris Gregory, Peter is standing; in fact, the illumi- Nazianzus: Paris Gr. 510. A study between text and images, DOP 16 nated manuscript and the two descriptions contradict—especially as (1962) 212. regards the apostles’ postures. 11 ЗОГРАФ 41 (2017) [1–18]

the thunder, too weak to rise. Each is described as show- background, is relatively motionless. The poses of ing a different posture: the two prophets mirror each other. But the com- position of the three apostles below, set against a Peter, the most vehement, springing up from flowering terrain, is more indicative of movement, the ground, since he could... seemed to speak for each is at a different stage of awakening. In me- words.... James, partly rising with difficulty on his dieval thought, motion was a sign of time, because knee, and supporting his still heavy head with his time had been created by God along with the mov- left arm, still has the greater part of his body nailed ing creatures of the creation. Therefore, the three to the ground, while his right hand he holds closely apostles, with their varied postures, are a foil for the to his eyes.... John however does not wish to look relatively motionless tableau of Christ, who is here up at all, but... seems to lie there in deep sleep... seen transfigured, which is to say, outside time.64 Iconographic details, which can be singled out from In Shivta, the depiction is more balanced: all five fig- the description, clearly reflect contemporary Byzantine ures around Christ are united by a common background, Transfigurations. Differentiation of the apostles’ postures, a shared vision. No clear distinction between apostles and especially that of Peter, who is in the process of springing prophets exists. The Sinai mosaics show same attitude, or- up from the ground, reflect the early scheme as found at ganizing all five figures around Christ; all share the same Shivta. This posture is still found in a few middle Byzan- ground. This attitude to space strengthens the hypoth- tine Transfigurations (e.g. Daphni, fig. 14). esis of the early date of the painting at Shivta, reflecting It is not clear if the cycle described by Mesarites is theological argument that the apostles and the prophets a renovation by Basil I, or if both descriptions actually together witnessed the full vision: the prophets, Elijah and refer to the original sixth-century iconography. Epstein Moses, witness the true face of God65 —God in human suggests that despite the different nature of the two de- form, the apostles Christ’s divine nature.66 scriptions, considering archaisms in choice of scenes and iconographic details, they actually describe the same sixth-century work which survived well into the twelfth ‘Positional meaning’ of the Transfiguration century. She proposes that the church was first adorned before and after iconoclasm with figural decorations during the reign of Justin II (565–578). The iconographic program was devised to op- Transfigurations in Ravenna, Sinai, Naples, and pose the teachings of Nestorius and Eutychius, stressing Porec appear separately, disconnected from the Christo- the divine and human nature of Christ.62 logical cycle and singled out in the architectural space; all but Porec are apse decorations, highlighting the special Analysing Nicolaos Mesarites’ description of the significance of the scene. This can perhaps be explained Transfiguration mosaics in the Church of the Holy Apos- by the event of the Transfiguration standing apart from tles, Henry Maguire suggests that differences in the apos- other miracles performed by Christ, but ‘a central act in tles’ postures and their adaptation to the text, already God’s plan of salvation’ as it signifies the only revelation evident in the ninth-century examples, have become even of His Divine nature to three disciples while still alive. 67 more marked by the twelve century, when the description was written. Maguire holds that as neither the Sinai mo- The apse location and the isolation of these sixth- saic nor any other surviving Transfiguration scene prior century Transfigurations in architectural space, as well as to the ninth century shows any distinction among each of their iconographic diversity, are not found in post-icon- the three Apostles’ attitudes to the miraculous revelation oclastic representations of the scene. On the contrary: of Christ—which is clearly apparent in post-iconoclastic middle Byzantine iconography displays its iconographic versions of the Transfiguration, no such distinction has uniformity, which also became an integral part of the nar- rative of Christ’s life and passion. Spieser argues that this yet been made.63 Shivta’s Transfiguration is clear proof of scene disappeared from the apse soon after its invention; the availability of the visual scheme which distinguishes by witnessing the miracle of Transfiguration the church the apostles by their reaction to the miracle prior to the congregation violated, in his words, Christ’s command- iconoclastic controversy. ment that it should be kept in secret until his Second For whatever reasons, eleventh– to twelfth-century Transfigurations seem even closer to Shivta than those of the ninth and tenth century, especially regarding the 64 H. Maguire, The cycle of images in the church, in: Heaven on depiction of the apostles. However, unlike the earlier ex- earth. Art and the church in Byzantium, ed. L. Safran, University Park, PA 1997, 143. amples, the earthly and celestial spheres are contrasted. 65 B. G. Bucur, Matt 17:1–9 as a vision of a vision. A neglected While movements of the apostles are emphasized, Christ strand in the patristic reception of the Transfiguration account, Neotesta- and the prophets are depicted motionless. Analysing this mentica 44/1 (2010) 17. feature in Daphni, Maguire suggests that it symbolically 66 Ps.-Ephrem’s sermon on the Transfiguration, for instance, refers to the distinction between time and timelessness: reads: ‘There was joy for the Prophets and the Apostles by this ascent of the mountain. The Prophets rejoiced when they saw his humanity, At Dafni...the composition of Christ and the which they had not known. The Apostles also rejoiced when they saw two prophets above, set against an abstract gold the glory of his divinity, which they had not known...and they looked to one another: the Prophets to the Apostles and the Apostles to the Prophets. There the authors of the old covenant saw the authors of the 62 Epstein, op. cit., 80–81. new’ (after Bucur, op. cit., 19). 63 H. Maguire, Truth and convention in Byzantine descriptions 67 E. Perl, ‘...That Man Might Become God’. Central themes in 12 of works of art, DOP 28 (1974) 123–124. Byzantine theology, in: Heaven on Earth, 50. Maayan Fanar Е.: Th e Transfi guration at Shivta

Coming (Mt 17, 9).68 For Elsner, such a sharp contrast between pre– and post-iconoclastic Transfigurations, es- pecially in their ‘positional meaning’—shifting to the vault from the apse, which therefore lost its special sig- nificance—is the ultimate proof that Christian iconogra- phy in Byzantium underwent significant changes after the Iconoclasm because the iconophiles needed to develop a response to the iconoclastic accusations.69 If Constantine Rhodes’ tenth-century description of the Church of the Holy Apostles refers to its redecora- tion by Basil I, this, together with other descriptions of the now lost ninth-century churches (the church of Stylianos Zaoutzas, built 886–893; the church of the Virgin of Pege, repaired and redecorated 870–879), would confirm Elsner’s suggestion that the Transfiguration was included in the Christological cycle after the Iconoclastic controversy.70 Alternatively, if at issue here is the reflection of the sixth-century cycle, the scene might have been introduced into Christ’s narrative cycle as early as the sixth century. However, two known sixth-century examples of illustrated cycles in churches suggest otherwise. Transfiguration is ab- sent from S. Apollinaire Nuovo in Ravenna, while in S. Apol- linaire in Classe it constitutes part of an apse decoration, therefore is excluded from the cycle and combined with the message of the apse. The Transfiguration is also absent from Choricus’ description of the highly decorated sixth-century churches of St. Sergious and St. Stephen in Gaza.71 In the tenth century, the tradition of singling out the scene of the Transfiguration still persisted. It is depicted on arches or niches, as is evident in churches in Cappa- docia such as Mustafapasha and Old Tokali Kilise. In the latter, dated to the early tenth century, the Transfiguration Fig. 13. Transfiguration, Homilies of Gregory Naziansus, is separated from an extensive cycle of Christ covering the Paris, BNF, gr. 510, fol. 163 (By http://gallica.bnf. vault and set at the arch. In the New Tokali Kilise, built in fr/ark:/12148/btv1b84522082/f542.image, Public the second half of the tenth century, the scene is depicted Domain, https://commons.wikimedia.org/w/index. inside one of the niches. Only in the eleventh-century did php?curid=38630191) the Transfiguration become fully integrated into the litur- gical cycle and gain a place among the twelve liturgical tained religious imagery as well. If the Transfiguration feasts, thus becoming an integral part of Church decora- was painted on the side apse, we can only wonder what tion. The most prominent examples are Hosios Loukas, subject could have been painted on the central one.72 Daphni, and more. Thus, the process that started after the iconoclastic controversy came to an end. Date and concluding remarks The wall painting at Shivta may be an incredible discovery of a missing link to the later iconography of the The date of the construction of the Southern church Transfiguration, showing not just most of its iconographic remains unclear. Some scholars suggest that it is the oldest elements, including postures and gestures, but also testi- of three churches, built in the already existing settlement fying to its early appearance as part of apse decoration. and adapted to the area, probably on the ancient cultic site Once again, Shivta’s Transfiguration bridges these seem- of the Nabateans (first c).73 Another suggestion was that the ingly disparate trends. On the one hand it stands on its Southern and Northern churches were built concurrently, own, occupying the whole apse. On the other it is not in 505.74 A later sixth-century date was also proposed. wholly isolated, as the central and northern apses con- The confusion in its dating arises from an inscrip- tion on the lower border of a decorated lintel which sup- 68 J. M. Spieser, The representation of Christ in the apses of early Christian churches, Gesta 37/1 (1998) 70–71. 69 Elsner, Image and Iconoclasm, 476–485. 72 We can speculate that if the northern apse could have been de- 70 Ibid., 476. voted to the image of Mary and child (traces of red colour reveal a promi- 71 T. Polanski, The cycles of childhood and miracles in St. Ser- nent central motif), and the central one to one of the representations of the gius’ Church of Gaza in the ecphrasis of Choricius, in: Hortus Historiae. Glorification of Christ or of a cross (one of the favourite motifs depicted Studies in honour of Professor Jozef Wólski on the 100th anniversary of in the apse), they could have formed an iconographic program perhaps his birthday, Kraków 2010, 752–753. Idem, The mosaics and painting stressing the human and divine natures of Christ. decoration in the Church of Saint Stephen of Gaza and the Christian ec- 73 Segal, Shivta. A Byzantine town, 317–328. phrasis (Choricius of Gaza, Asterius of Amaseia, Nilus of Sinai), Folia 74 Rosenthal-Heginbottom dates Shivta’s Northern church to 505 Orientalia 48 (2011) 183–210. at the latest, presumably the same date as the Southern church (op. cit., 219). 13 ЗОГРАФ 41 (2017) [1–18]

posedly belonged to the church, but was not found in its Modern scholars hold an entirely different if not original place. diametrically opposite opinion, totally rejecting a violent The inscription states: ‘Under the presbyter Aedos end to the city. Avni and Magness suggest that the settle- in the year 3...’ The date has survived but partially, nor is ments in the area were continuously occupied well into its chronological system clear.75 Going according to the the tenth century.81 They even argue in favour of a peace- Province Arabia era, the date falls between 415 and 435, ful occupation of Shivta by a small community of Muslims or between 475/6 and 485/6. By the Diocletian era the who lived side by side with the original Christian com- date would be between 593 and 613. Thus, Kirk reads it munity.82 Their conclusions are based on a small mosque as the first quarter of the fifth century (415–435), while attached to the baptistery of the Southern church without Lea DiSegni suggests the later date of the lintel inscription harming or ruining it. But the mosque, constructed from — 508–528, according to the Eleutheropolis era.76 While spolia, presumably from the church itself, was built some neither the earlier nor the later dates can be dismissed, time in the eigth or even the ninth century, as analysis of the dates between 475 and 528 seem to suit better the his- 83 torical period in which the area enjoyed prosperity and the inscriptions there suggests. This still leaves us with widespread construction.77 Still, it is not clear if the lintel no factual evidence that could confirm the coexistence of belongs to the earliest stage of the church or to one of its Christians and Muslims in Shivta. Even if Christians still refurbishments, or if it is contemporaneous with the wall lived in the city until the mid-ninth century, the nature painting at issue. Several styles attested in architectural of the settlement there and in other Negev cities after stone decorations of the church suggest continuous exist- the Muslim conquest is not clear.84 Several in- ence. These however deserve deeper comparative research scriptions on the mosque’s walls are from Quranic verses to establish the proper chronology.78 Thus, we have no re- highlighting, according to Moor, fundamental issues in Is- liable, factual date of the Southern church, a circumstance lamic belief.85 Presumably already ruined by then, visual that projects directly onto the date of its Transfiguration images inside the church emphasized a core of Christian wall painting in its lateral apse. In addition, we lack al- faith. The Transfiguration, the only survivor, is a power- most entirely any materials comparable stylistically to the ful statement in this respect. I doubt whether these strong painting, as the surviving examples of wall paintings from messages of faith could coexist side by side. the early Byzantine period in the area and elsewhere are All in all, the evidence is inadequate to suggest scarce and mostly fragmentary. when exactly and why the settlement came to an end.86 The only secure fact is that the Southern church was Although nothing points to a sharp violent termina- still in use as late as 639, as ascertained from an inscrip- tion there might have been a destructive seismological tion stating that the floor was laid anew that year— ‘being event or a series of such events,87 as several devastating the commemoration of a new paving of the church under earthquakes which could have befallen the area in the Bishop George and the Archdeacon and economus Peter 79 in 639 during the reign of Heraclius’. 81 J. Magness, The archaeology of the early Islamic settlement in The date and causes of destruction of the Southern Palestine, Winona Lake, IN 2003, 177–194. A. Walmsley, Coinage and church and the settlement is also unclear. Describing their the economy of Syria-Palestine in the seventh and eighth Centuries CE, in: Money, power and politics in early Islamic Syria. A review of current impressions of the ruins of Shivta, Wooley and Lawrence debates, ed. J. Haldon, Farnham 2010, 37. Scholars argue that the so- suggest that its end was violent: called Central Church in the lower town of Nessana was constructed in the late seventh or early eight century, and that it continued to exist It is clear that Esbeita came to a violent end. uninterrupted until the ninth century. Cf. D. Urman, Nessana excava- All the gates of the town have been blocked with tions 1987–1995, in: Nessana. Excavations and studies, ed. D. Urman, roughly-built barriers of stone, and stone barri- Beer Sheva 2004), 115. 82 cades have been piled across many of the streets; G. Avni, The Byzantine-Islamic transition in the Negev. An archaeological perspective, Jerusalem Studies in Arabic and Islam 35 everything points to a desperate attempt to hold (2008) 4–5; v. also G. Avni, Early mosques in the Negev highlands. New the place against an enemy who ultimately took it. archaeological evidence on Islamic penetration of Southern Palestine, Moreover, the whole evidence of the ruins is to the Bulletin of the American Schools of Oriental Research 294 (1994) effect that the town’s occupation ceased suddenly 83–100. V. also G. Peers, Crosses’ work underfoot. Christian spolia in and uniformly...80 the late antique mosque at Shivta in the Negev Desert (Israel), ECA 8 (2011) 102. ‘Esbeita is a Byzantine town pure and simple, found- 83 B. Moor, Mosque and Church. Arabic inscriptions at Shivta ed not very early in the Christian era and destroyed, not in the early Islamic period, Jerusalem Studies in Arabic and Islam 40 (2013) 73–141. long afterwards, by the Arab conquest’, they concluded. 84 Y. Tepper, L. Weissbrod, G. Bar-Oz, Behind sealed doors. Un- ravelling abandonment dynamics at the Byzantine site of Shivta in the 75 Y. E. Meimaris, Chronological systems in Roman-Byzantine Negev Desert, Antiquity 2015/348 (http://antiquity.ac.uk/projgall/bar- Palestine and Arabia, Athens 1992, 316. oz348 [15.6.2017]). 76 L. Di Segni, Epigraphic documentation on building in the prov- 85 Moor, op. cit., 87. th th inces of Palestina and Arabia, 4 –7 c., in: The Roman and Byzantine 86 D. Fuks, E. Weiss, Y. Tepper, G. Bar-Oz, Seeds of collapse? Near East II, JRA supplementary series 31, Portsmouth, RI 1999, 167. Reconstructing the ancient agricultural economy at Shivta in the Negev, 77 Ibid. Antiquity 90/353 (2016) (https://www.cambridge.org/core/journals/ 78 A claim that churches in Negev were transformed around the antiquity/article/seeds-of-collapse-reconstructing-the-ancient-agricul- turn of the sixth century from mono-apsidal to tri-apsidal remains large- tural-economy-at-shivta-in-the-negev/1B89BC5EF6C12CC7345A44C ly unproved. Cf. S. Margalit, The North Church of Shivta. The discovery 678F6F057/core-reader [15.6.2017]). of the first church, Palestine Exploration Quarterly 119 (1987) 106–121. 87 A. M. Korjenkov, E. Mazor, Earthquake characteristics recon- 79 Negev, The Greek inscriptions from the Negev, 61. structed from archaeological damage patterns. Shivta, the Negev, Israel, 14 80 Wooley, Lawrence, op. cit., 90. Israel Journal of Earth Sciences 48 (1999) 265–282. Maayan Fanar Е.: Th e Transfi guration at Shivta

Fig. 14. Transfiguration, mosaics, Daphni Monastery, Athens, Greece, 11th c (By Dimkoa – Own work, Public Domain, https://commons.wikimedia.org/w/index.php?curid=12720910) seventh and eigth centuries are documented in ancient The Transfiguration as a choice for apse decora- sources.88 Social and economic reasons, together with tion may coincide with the Chalcedonian council (451) over-taxation as indicated in the late seventh-century and opposition to its decision in Palestine. The debate Nessana papyrus 75, could also have played an impor- was large scale, as evidence from Gaza attests.91 There is tant role in the abandonment of the site.89 According no evidence as to whether the Negev Christian commu- to Edelson, the economy of Shivta, unlike other nearby nity supported the Chalcedon Christology. However, the settlements, was mainly based on agriculture, especially choice of the Transfiguration can be seen as a statement wine production. Reduction of wine export in the early in this direction. Similarly, scholars have explained this choice in Ravenna and Sinai as anti-heretic statements Muslim period together with difficult living conditions 92 in the arid region could have accelerated emigration in the Monophysite controversy. It was also argued that from the settlement.90 the mosaic on the apsidal arch in Ss. Nereo ed Achilleo (816–817) commissioned by Leo III (795–815) probably This analysis gives us rough brackets for the time of referred to anti-adoptianism in Visigothic Spain.93 the Southern church decoration: between c. 450 and the The Transfiguration in provincial Shivta reveals third part of the seventh century. Yet a seventh-century an ultimate link between early Byzantine and post-icon- date seems unlikely if we compare all the remnants of oclastic iconography. It is a unique piece of evidence of Shivta. It is not clear if the artist was local or came from the existence of the scheme for the middle Byzantine ico- afar, but the investment in the paintings reflects consid- nography of the scene certainly in the sixth century, if not erable wealth, more befitting the sixth century. Can we earlier. Its existence undoubtedly suggests that post-icon- be more precise about the dating of the wall painting? As oclastic iconography was not invented but copied from noted above, some stylistic features (and treatment of the early models, which existed on a much broader scale than figure of Christ) clearly indicate a relation to the sixth- we know. Shivta’s Transfiguration reflects such an early century mosaics in St. Catherine in Sinai. Illusionistic model, which was probably popular enough to be copied traces of late antique rendering of the volumes, space, and in the small provincial settlement, and on the other hand architectural ornament may favour the earlier date over could be altered in order to adapt it to different messages the later. in Sinai, Ravenna, and elsewhere. The dead city in the Negev still awaits its awakening, 88 K. W. Russell, The earthquake chronology of Palestine and northwest Arabia from the 2nd through the mid-8th century A. D, Bulletin and who knows what secrets are still lodged in its stones. of the American Schools of Oriental Research 260 (1985) 37–59. 89 C. J. Kraemer, Excavations at Nessana 3. Non-literary papyri, 91 J. L. Hevelone-Harper, Disciples of the desert. Monks, laity, Princeton 1985, 101; F. R. Trombley, From Kastron to Qasr.. Nessana be- and spiritual authority in sixth-century Gaza, Baltimore–London 2005, tween Byzantium and the Umayyad Caliphate ca. 602–689. Demographic 108–112. and microeconomic aspects of Palaestina III in interregional perspective, 92 G. V. Mackie, Early Christian chapels in the west. Decoration, in: The . Crossroads of Late Antiquity. History, religion and archae- function and patronage, Toronto 2003, 181. ology, ed. E. Bradshaw-Aitken, J. M. Fossey, Leiden 2013, 183–184. 93 E. Trunø, Image and relic. Mediating the sacred in early medi- 90 Erez-Edelson, op. cit., 134–136. eval Rome, Rome 2002, 132–149. 15 ЗОГРАФ 41 (2017) [1–18]

ЛИСТА РЕФЕРЕНЦИ – REFERENCE LIST

Andreopoulos A., Metamorphosis. The Transfiguration in Byzantine Hirschfeld Y., The monasteries of Gaza. An archaeological review, in: theology and iconography, Crestwood – New York 2005. Christian Gaza in Late Antiquity, ed. B. Bitton-Ashkelony, A. Kof- Avni G., Early mosques in the Negev highlands. New archaeological evi- sky, Leiden 2004, 61–88. dence on Islamic penetration of Southern Palestine, Bulletin of the Kedar Y., Ancient agriculture at Shivtah in the Negev, Israel Exploration American Schools of Oriental Research 294 (1994) 83–100. Journal 7 (1957) 178–189. Avni G., The Byzantine-Islamic transition in the Negev. An archaeological Korjenkov A. M., Mazor E., Earthquake characteristics reconstructed perspective, Jerusalem Studies in Arabic and Islam 35 (2008) 1–26. from archaeological damage patterns. Shivta, the Negev, Israel, Israel Baly T. J., S’Baita, Palestine Exploration Fund Quarterly Statement 67 Journal of Earth Sciences 48 (1999) 265–282. (1935) 171–181. Kraemer C. J., Excavations at Nessana 3. Non-literary papyri, Princeton Bernabò M. The miniatures in the Rabbula Gospels. Postscripta to a re- 1985. cent book, Dumbarton Oaks Papers 68 (2014) 343–358. Lee D., On the Holy Mountain. The Transfiguration in scripture and the- Bernabò M., Kessel G., A Syriac Four Gospel Book in Diyarbakir, Con- ology, Colloquium 36/2 (2004) 143–159. vivium 3 (2016) 172–203. Linn R., Tepper Y., Bar-Oz G., Visible induced luminescence reveals in- Bucur B. G., Matt 17:1–9 as a vision of a vision. A neglected strand in visible rays shining from Christ in the Early Christian wall painting the patristic reception of the Transfiguration account, Neotestamen- of the Transfiguration in Shivta, Plos One 12/9 (2017) e0185149. tica 44/1 (2010) 15–30. https://doi.org/10.1371/journal.pone.0185149. Caner D. F., History and hagiography from the late antique Sinai, Liv- Maayan Fanar E., Early Christian tomb at Lohamei HaGetaot. Formation erpool 2010. of Christological symbolism, Eastern Christian Art 7 (2010) 71–89. Chen D., Byzantine architects in Mampsis and Sobota, Palaestina Ter- Mackie G. V., The Zeno chapel. A prayer for salvation, Papers of the tia, Liber Annuus 31 (1981) 235–244. British School at Rome 57 (1989) 172–179. Chronicles of the Land. Archaeology in the Israel Museum Jerusalem, ed. Mackie G. V., Early Christian chapels in the west. Decoration, function M. Dayagi-Mendels, S. Rozenberg, Jerusalem 2010. and patronage, Toronto 2003. Colt H.D., Isbeita, Palestine Exploration Fund Quarterly Statement 67 Magness J., The archaeology of the early Islamic settlement in Palestine, (1935) 9–11. Winona Lake, IN 2003. Colt H.D., Isbeita, Quarterly of the Department of Antiquities in Pales- Maguire H., Truth and convention in Byzantine descriptions of works of tine 5 (1936) 198–199. art, Dumbarton Oaks Papers 28 (1974) 113–140. Corrigan K. A., The ninth century Byzantine marginal psalters: Moscow, Maguire H., The cycle of images in the church, in: Heaven on earth. Art Historical Museum Cod. 129; Mt. Athos, Pantocrator 61; Paris, Bib- and the church in Byzantium, ed. L. Safran, University Park, PA liothèque Nationale gr. 20, Los Angeles 1984 (doctoral dissertation, 1997, 121–151. University of California). Maksimović J., Ikonografija i program mozaika Eufrazijeve bazilike u Crowfoot G., The Nabatean ware of Sbaita, Palestine Exploration Fund Poreču, Zbornik radova Vizantološkog instituta 8/2 (1964) 247–262. Quarterly Statement 68 (1936) 14–27. Malajoli B., La basilica eufrasiana di Parenzo, Parenzo 1940. Di Segni L., Epigraphic documentation on building in the provinces of Mango C., The art of the Byzantine Empire 312–1453. Sources and doc- th th Palestina and Arabia, 4 –7 c., in: The Roman and Byzantine Near uments, Englewood Cliffs, NJ 1972. East II, Journal of Roman Archaeology, supplementary series 31 Mango C., The mosaic of the Transfiguration at St. Catherine’s, The (1999) 149–178. Fortnightly Review 27 (July 2014) http://fortnightlyreview. Downey G., Nikolaos Mesarites. Description of the Church of the Holy co.uk/2014/07/mango-sinai-mosaic/ Apostles at Constantinople, Transactions of the American Philo- Margalit S., The North Church of Shivta. The discovery of the first sophical Society 47/6 (1957) 855–924. church, Palestine Exploration Quarterly 119 (1987) 106–121. Elsner J., Image and Iconoclasm in Byzantium, Art History 11 (1988) Marinoff I., The Transfiguration in the Byzantine liturgy, Life of the 471–491. Spirit 10 (1955) 18–22. Elsner J., The viewer and the vision. The case of the Sinai apse, Art His- Mayerson Ph., The desert of Southern Palestine according to Byzantine tory 17/1 (1994) 81–102. sources. Transactions and Proceedings of the American Philologi- Epstein A. W., Tokali Kilise. Tenth-century Metropolitan art in Byzan- cal Association 95 (1964) 166–172. tine Cappadocia, Washington DC 1986. Meimaris Y. E., Chronological systems in Roman-Byzantine Palestine Erez-Edelson R., Sedentary zone in the Negev during the Byzantine pe- and Arabia, Athens 1992. riod. Settlement distribution and runoff farming, Ramat-Gan 2004 Miziolek J., Transfiguratio Domini in the apse at Mount Sinai and the (doctoral dissertation, University of Bar Ilan, in Hebrew). symbolism of light, Journal of the Warburg and Courtauld Institutes Evangelatou M., ‘Liturgy and the illustration of the ninth-century mar- 53 (1990) 42–60. ginal psalters, Dumbarton Oaks Papers 63 (2009) 59–116. Moor B., Mosque and Church. Arabic inscriptions at Shivta in the early Is- Figueras P., Monks and monasteries in the Negev Desert, Liber Annuus lamic period, Jerusalem Studies in Arabic and Islam 40 (2013) 73–141. 45 (1995) 401–447. Negev A., Subeita, Revue biblique 81 (1974) 397–420. Figueras P., Remains of mural painting of the Transfiguration in the Negev A., The Greek inscriptions from the Negev, Jerusalem 1981. southern church of Sobata (Shivta), ARAM 18/2 (2006) 127–151. Nelson R., Where God walked and monks pray, in: Holy image, hal- Fuks D., Weiss E., Tepper Y., Bar-Oz G., Seeds of collapse? Reconstruct- lowed ground. Icons from Sinai, ed. R. S. Nelson, K. M. Collins, Los ing the ancient agricultural economy at Shivta in the Negev, Antiq- Angeles 2006. uity 90/353 (2016) 1–5; https://www.cambridge.org/core/journals/ antiquity/article/seeds-of-collapse-reconstructing-the-ancient-ag- Nersessian C., The illustrations of the homilies of Gregory of Nazianzus. ricultural-economy-at-shivta-in-the-negev/1B89BC5EF6C12CC73 Paris Gr. 510. A study between text and images, Dumbarton Oaks 45A44C678F6F057/core-reader). Papers16 (1962) 195–228. Golan K., The stone architectural decorations of the Byzantine Negev. Palmer E. H., The desert of the Exodus. Journeys on foot in the wilder- Characterization and meaning, Beer Sheva 2014 (doctoral disserta- ness of the forty years’ wanderings, London 1871. tion, Ben Gurion University of the Negev). Paretsky A., The Transfiguration of Christ. Its eschatological and Chris- Hevelone-Harper J. L., Disciples of the desert. Monks, laity, and spiritual tological dimensions, New Blackfriars 72/851 (1991) 313–324. authority in sixth-century Gaza, Baltimore–London 2005. Peers G., Crosses’ work underfoot. Christian spolia in the late antique Hirschfeld Y., Man and society in Byzantine Shivta, Qadmoniot 125 mosque at Shivta in the Negev Desert (Israel), Eastern Christian Art 8 (2011) 101–119. 16 (2003) 2–17 (in Hebrew). Maayan Fanar Е.: Th e Transfi guration at Shivta

Perl E., ‘...That Man Might Become God’. Central themes in Byzantine Desert, Antiquity 2015/348 (http://antiquity.ac.uk/projgall/bar- theology, in: Heaven on Earth, Art and the Church in Byzantium, oz348). ed. L. Safran, University Park, PA 1997, 38–57. The glory of Byzantium. Art and culture of the middle Byzantine era AD. Polanski T., The cycles of childhood and miracles in St. Sergius’ Church 843 – 1261, ed. H. C. Evans, W. D. Wixom, New York 1997. of Gaza in the ecphrasis of Choricius, in: Hortus Historiae. Studies The Apocalypse of Peter, ed. J. N. Bremmer, I. Czachesz, Leuven 2003. in honour of Professor Jozef Wólski on the 100th anniversary of his Transfiguration, in: The Oxford Dictionary of Byzantium III, ed. A. Ka- birthday, ed. E. Dabrowa, Kraków 2010, 737–757. zhdan, A.-M. Talbot, New York – Oxford 1991, 2014–15 (G. Pod- Polanski T., The mosaics and painting decoration in the Church of Saint skalsky, R. F. Taft, A. Weyl Carr). Stephen of Gaza and the Christian ecphrasis (Choricius of Gaza, Aste- Trombley F. R., From Kastron to Qasr.. Nessana between Byzantium and rius of Amaseia, Nilus of Sinai), Folia Orientalia 48 (2011) 183–210. the Umayyad Caliphate ca. 602–689. Demographic and microeco- Rosenthal-Heginbottom, R., Die Kirchen von Sobota und die nomic aspects of Palaestina III in interregional perspective, in: The Dreiapsidenkirchen des Nahen Ostens, Wiesbaden 1982. Levant. Crossroads of Late Antiquity. History, religion and archaeol- Russell K. W., The earthquake chronology of Palestine and northwest ogy, ed. E. Bradshaw-Aitken, J. M. Fossey, Leiden 2013, 181–224. Arabia from the 2nd through the mid-8th century A. D, Bulletin of the Trunø E., Image and relic. Mediating the sacred in early medieval Rome, American Schools of Oriental Research 260 (1985) 37–59. Rome 2002. Segal A., Shivta. Plan and architecture of a Byzantine town in the Negev, Trunø E., The apse mosaic in early medieval Rome, Cambridge 2002. Jerusalem 1981. Tsafrir Y., Some notes on the settlement and demography of Palestine in Segal A., The Byzantine city of Shivta (Esbeita), Negev Desert, Israel, the Byzantine period. The archaeological evidence, in: Retrieving the Oxford 1983. past. Essays on archaeological research and methodology in honor of Gus W. Van Beek, ed. J. D. Seger, Starkville, MS 1996, 269–283. Segal A., Shivta. A Byzantine town in the Negev Desert, Journal of the Society of Architectural Historians 44 (1985) 317–328. Urman D., Nessana excavations 1987–1995, in: Nessana. Excavations and studies, ed. D. Urman, Beer Sheva 2004, 1–119. Segal A., Architectural decoration in Byzantine Shivta, Negev Desert, Is- rael, Oxford 1988. Walmsley A., Coinage and the economy of Syria-Palestine in the seventh and eighth Centuries CE, in: Money, power and politics in early Is- Shereshevsky J., Byzantine urban settlements in the Negev Desert, Jeru- lamic Syria. A review of current debates, ed. J. Haldon, Farnham salem 1986 (in Hebrew). 2010, 21–44. Spieser J. M., The representation of Christ in the apses of early Christian Weitzmann K., The mosaic in St. Catherine’s monastery on Mount Sinai. churches, Gesta 37/1 (1998) 63–73. The icons, Proceedings of the American Philosophical Society 110 Stroumsa R., People and identities in Nessana, Durham, NC 2008 (doc- (1966) 392–405. toral dissertation, Duke University). Wooley C. L., Lawrence T. E., The wilderness of Zin (Archaeological Tepper Y., Weissbrod L., Bar-Oz G., Behind sealed doors. Unravelling report), London 1914 (http://www3.lib.uchicago.edu/cgi-bin/eos/ abandonment dynamics at the Byzantine site of Shivta in the Negev eos_title.pl?callnum=DS111.A1P28_vol3_cop1).

Преображење из Шивте. Реконструкција рановизантијске иконографије

Ема Мајан Фанар Универзитет у Хаифи, Одељење за историју уметности

Сеоско насеље Шивта у пустињи Негев постоја- Са леве стране, још се могу видети фигуре ло је од IV–V до VIII–IX века, када је напуштено. Три двојице апостола, од којих ниједан нема ореол. Јова- тамошње цркве биле су богато обдарене, али од нека- на препознајемо на основу последњих слова тог имена дашње славе остало је тек неколико бојених фрагме- поред доњег дела тела. Приказан је лежећи ничице са ната. У овом раду разматраћемо сцену Преображења подигнутом левом руком, којом највероватније при- у бочној апсиди Јужне цркве у Шивти. држава главу. Иза Јована налази се Петрова фигура, Упркос стању у којем се сада налази, као и чиње- препознатљива по остацима браде и седе коврџаве ници да је десна страна слике готово сасвим страдала, косе. Петар највероватније клечи, а његова десна рука представа Преображења ипак се може поуздано пре- показује ка Христу, при чему се остаци његових пр- познати. Христос, главе смештене у веома велики оре- стију могу видети тик поред кривине Јованових леђа. ол светлоружичастих нијанси, јасно се види унутар Са Петрове леве стране, уз бордуру апсиде, на- мандорле у горњој зони средишњег дела апсиде. Хри- лази се пуна, стојећа фигура, чију главу обавија огро- стово лице било је насликано у ружичастим нијанса- ман ружичасти ореол. На основу општег изгледа, ма, а изнад једине сачуване обрве могу се видети дуги, чини се да је реч о Илији. Обриси линија у пределу бели светлосни акценти. Друге очуване контуре до- Илијине главе упућују на то да је имао браду. Очи, звољавају да реконструишемо Христову десну шаку. обрве, нос и уста готово да се не могу разазнати, јер Сразмерно увећана, она извирује из широког рукава, се првобитна скица жућкастоокер бојом једва назире. подигнута у раван груди. Назире се и кривина леве Остаци Илијине фигуре показују да је његова пред- шаке, као и изражено заобљена драперија изнад левог става била несразмерно велика, највероватније исте длана. Сам длан, међутим, толико је оштећен да се не величине као апостолске. Фигуре поред Христове ру- може разазнати. жичасте мандорле смештене су у зеленкасти крајолик. 17 ЗОГРАФ 41 (2017) [1–18]

Не можемо бити сигурни у то да ли је испод Христо- Рекло би се да сцена из Свете Катарине, а мо жда вих ногу приказана планина, али чак и ако јесте, сва- и други изоловани примери, имају значење прили- како не доминира призором као што је то случај на ке за опстанак. Сведоче о томе да су постојале разне млађим приказима Преображења. верзије Преображења које се не јављају у каснијим Предложена реконструкција сцене Преобра- представама. Чини се да заправо представљају јед- жења из Шивте показује композициони распоред но тумачење већ познате иконографске схеме и њену фигура. Најупадљивије је то што су све фигуре око прилагођеност јединственим порукама тих мозаика. Христа смештене у приближно исту раван, при чему Ако је то заиста тачно, могло би се рећи и да је ико- су пророци уздигнути врло мало или нимало. Уз то, нографија Преображења већ била утврђена средином веома прецизно можемо утврдити гестове и положаје VI века, када су мозаици настали. Преображење из тела апостола. Шивте је, изгледа, једини сачувани пример. Преображење из Шивте баца ново светло на ра- Представа Преображења красила је и чуве- новизантијску иконографију те сцене. Мозаици из VI ну цркву Светих апостола у Цариграду, која данас не века у апсидама манастира Св. Катарине на Синају и постоји. Није, међутим, била изолована, већ је чинила базилике Сан Аполинаре ин Класе у Равени, једини део христолошког циклуса. Време настанка тих моза- сачувани примерци те сцене из предиконоборачког ика није поуздано утврђено (мада је предложено да- периода, имају различите иконографске схеме. Чини тирање у VI век), али научници се начелно слажу да је се, међутим, да је та сцена постала веома заступљена њена иконографија поновљена у Хомилијама Григорија у периоду после иконоборачке кризе, када се јавља у Назијанског (Paris, BNF, gr. 510), рукопису насталом у разним медијумима, и то углавном са стандардизо- IX веку у Цариграду. Занимљиво је да се светлоружи- ваном иконографијом. Стога су научници тврдили да часта боја Христове мандорле из Шивте јавља још само се пре иконоборачког периода ретко користила у мо- у сцени Преображења из тог рукописа. Описујући ту нументалној уметности, те да њена иконографија још сцену у цркви Светих апостола, Никола Месарит у XII није била уједначена. Штавише, као иконографски веку истиче симболички значај божанске светлости, а најсродније онима из времена после иконоборства, реакција апостола на њу снажно је наглашена и доча- синајско Преображење је сматрано њиховим главним рана положајем њихових тела. Иконографски детаљи узором. Анализа Преображења из Шивте доводи у који се издвајају из овог описа очигледно одражавају питање те претпоставке. византијска Преображења тог времена. У складу са Иако на први поглед веома слична, Преобра- тим, Хенри Мегвајер сматра да се на предиконобор- жења са Синаја и из Шивте ипак се знатно разли- ачким представама Преображења апостоли нису раз- кују. Док је оно из Шивте уско повезано са текстом ликовали по положају тела нити је тај положај био Јеванђеља и јасно разликује апостоле по гестовима прилагођен тексту. Иако већ приметна у ликовним и покретима, синајско Преображење одступа од је- примерима из IX века, диференцијација је постала још ванђељског наратива. Не само да су Јован и Јаков израженија до XII века, када је опис написан. Преобра- приказани у истом положају и покрету, што се коси жење из Шивте јасно је сведочаство о томе да је ликов- са Јеванђељима и начином на који су приказивани ка- на схема са диференцираним положајима апостолских сније, него је и представа Петра испод Христових ногу тела настала пре иконоборачке кризе. јединствена и не јавља се ни у једном другом приказу Иако тачно време настанка Преображења из Преображења. Заправо, чини се да су на синајским Шивте остаје неразјашњено, на основу свих располо- мозаицима познати елементи приче прераспоређе- живих археолошких, епиграфских и стилских подата- ни да би се прилагодили статичној, хијерархијски ка, оквирно би се могло сместити на крај V или у VI устројеној композицији у чијем средишту се налази век. Преображење из Шивте пружа јединствено све- Христова фигура. дочанство о томе да се иконографија Преображења из Како је показао Јас Елснер, иконографски про- периода после иконоборачке кризе развила из раних грам цркве Св. Катарине добро је осмишљен и одгова- узора, који су били много распрострањенији него што ра значају места где се Бог указао Мојсију кроз симболе нам се данас чини. Опустели град у пустињи Негев и предао му Заповести. Стога сцену Преображења тре- још чека да се пробуди и не можемо знати какве све ба сагледати у контексту читавог распореда приказа у тајне још крије у камену. олтару, укључујући и представе Мојсија како прима За- повести и Мојсија код Неопалиме купине на луку.

18