Christian Spolia in the Late Antique Mosque at Shivta in the Negev Desert (Israel) *

ECA 8 (2011), p. 101-119; doi: 10.2143 / ECA.8.0.2961368

Crosses’ Work Underfoot: Christian Spolia in the Late Antique Mosque at Shivta in the Negev Desert (Israel) *

Glenn PEERS

Scholars have productively studied the re-use of of reasons, some of which are perfectly evident architectural materials in Late Antiquity from the from their supercessionist vantage point. Likewise, point of view of practical, aesthetic and ideological at the end of this period, Muslims had a compli- motivations, and the number of those motivations, cated relationship to religions they encountered, as well as responses, marks a distinctive character- primarily Christianity, because that faith was closely istic of this period in relation to classical Antiquity. associated, naturally, with the empire that Islamic The almost-bricolage of the Arch of Constantine in forces contended against. This article examines one Rome (315) represents a sophisticated, rich demon- such encounter, which led to the integration of stration of the past in the present and, more to the Christian building materials into an early Islamic point, in its service1. Christians manipulated mate- mosque in the small town of Shivta (present-day rial remnants of the past, too, for a large number name, but also called Sbeita, Esbeita, Subeita, Isbayta) in the central Negev desert, some 50 kms south of modern Beer Sheeva (Fig. 1). There, clearly Christian materials had highly public and agonistic placement, but the motivations were not necessarily, or only, adversarial. They were faceted and not reducible to an antagonistic social model2. They were available for complicated understand- ings of the ongoing power and prestige that Islam was in the process of appropriating to itself. The town of Shivta has been the object of study by numerous travelers and archaeologists, along with other settlements in the Negev, a forbidding but deeply affecting desert terrain (Fig. 2)3. Leonard Woolley (1880-1960) and T.E. Lawrence (1888- 1935) visited the site in the spring of 1914, and their account of the region still stands as a useful and even romantic treatment of the historical remains and vivid landscape (Pl. 1)4. Their study

* Many thanks to Samer Ali, Katia Cytryn-Silverman, Karni Golan, Mattia Guidetti, Mat Immerzeel, Amy Papalexan- drou, Na’ama Pat-El, and to Silvia Krapiwko and Aries Rochman-Halperin at the archive of the Israel Antiquities Authority. My colleague Stephennie Mulder read a version of this article and offered excellent advice. 1 See now Brilliant/Kinney 2011. 2 See Hayden et al. 2011. 3 On the region, see Erez Edelson 2006; Haiman 1995; Fig. 1. Map of the Negev, with principal sites marked Evenari/Shaman/Tadmor 1971. (from Figueras 1995a) 4 Woolley/Lawrence 2003, 97-120.

101

95300_ECA8(2011)_07.indd 101 27/02/13 15:23 was only a survey, nevertheless, but Shivta has suf- fered such intense scrutiny over the decades with the result of very few accurate archaeological reports. In fact, the site is woefully under-published, despite the attentions of Israeli archaeologists in the last half-century5. The general outlines of the town, and its principal buildings, were known when Harris Dunscombe Colt, Jr. (1901- 1973), an American archaeologist, received permis- sion from British authorities to commence excava- tion at Shivta in 1933. His team worked at the site for several seasons, before being distracted by the nearby site of Nessana, which led to important finds and publications. Those excavations at Shivta were terminated at the end of 1938, when the dig house, along with its finds, equipment, and records, were destroyed6. The Colt excavations, then, were very poorly recorded, and many finds were lost, but they did lead to the uncovering of a fascinating aspect of the last phase of Shivta’s history, a mosque attached to the baptistery of the south church Fig. 2. Map of Shivta (after Segal 1985, Fig. 4) (Pls 2-3). The mosque is a modestly sized structure, and it comprises a small courtyard that leads into a covered prayer room, a very unusual feature in the Negev, in fact (Pls 4-5). It was fit into a trapezoidal area that abutted the baptistery and domestic struc- tures, and opened onto the plaza that contained a large reservoir. The baptistery belonged to the south church, which also accommodated itself to the pre-existing urban fabric, as the town predates Christianity. The arrival of the new religion in the seventh century did not necessitate clearing ground for a freestanding architectural statement, and that situation holds for most of the region during the Pl. 1. View toward Shivta from Mitzpe Shivta Early Islamic period7. In fact, the size of the (photograph: author) mosque indicates that the number of Muslim men was never large, and comparing the mosque with 5 That is not to say that scholars have not published aspects the three churches at Shivta, along with their sub- of the site or attempted to integrate the site into regional sidiary buildings, makes evident the inhabitants of studies. See Erickson-Gini 2010, 31-32; Harris 2007; Late Antique/post-conquest Shivta were predomi- Hirschfeld 2003; Rubin 1997; Tsafrir/Di Segno/Green 8 1994, 234; Negev 1993; Shereshevski 1991, 61-82; Segal nantly Christian . Muslims were not necessarily a 1983; Gutwein 1981, 89-93; Ovadiah 1970, 166-173. On dominant force in the Late Antique Negev, even if the Early Islamic period, see Ribak 2007, 217-218; the forms and presence of Byzantine authority were Petersen 2005, 62-64; Magness 2003, 185-187; Negev 9 1982. gone . The town was unusual in the Negev for 6 AT/64/6, letter of January 1939. continuing to produce inscriptions commemorat- 7 Guidetti, forthcoming. ing burials, as well as a new floor in the south 8 Rubin 1997. church in 639, for example10. Six kilometers away 9 Guidetti 2009a, 4. 10 Negev 1981, 47-67; idem 1990, 354. from the town is a large monastic community, now 11 Caner 2010, 255 n. 13; Harris 2007, 44; Figueras 1995a, called Mitzpe Shivta, scarcely excavated, but with 420-423; Ovadiah 1970, 138-139. impressive and widespread remains (Pl. 6)11. Shivta

102

95300_ECA8(2011)_07.indd 102 27/02/13 15:23 Pl. 2. View of Mosque and South Church, Shivta (photograph courtesy of the Israel Antiquities Authority Archives, Israel Antiquities Authority Archives: British Mandate Records Files: file no. 71, Isbeita, photograph 9918)

Pl. 3. View of Mosque and South Church, Shivta (photograph courtesy of the Israel Antiquities Authority Archives, Israel Antiquities Authority Archives: British Mandate Records Files: file no. 71, Isbeita, photograph 9919)

103

95300_ECA8(2011)_07.indd 103 27/02/13 15:24 Pl. 4. View of Mosque, Shivta (photograph courtesy of the Israel Antiquities Authority, Israel Antiquities Authority Archives: British Mandate Records Files: file no. 71, Isbeita, photograph 9932)

Pl. 5. View of Mosque, Shivta (photograph courtesy of the Israel Antiquities Authority Archives, Israel Antiquities Authority Archives: British Mandate Records Files: file no. 71, Isbeita, photograph 9933)

104

95300_ECA8(2011)_07.indd 104 27/02/13 15:24 junction of places of worship in this period, Mus- lim mosque and Christian baptistery, is naturally significant, but in the second, the use of spolia was not only antagonistic and competitive, but like the placement, part of an attempt on the part of the town’s Muslims to use these Christian markers for protective power, for mapping social worlds, and for marking zones of sanctity. In other words, the builders and users of this new addition to Late Antique Shivta engaged in a creative re-making of the urban, social, and religious fabric of the town that spoke with and manipulated Christian stones. Pl. 6. View of Mitzpe Shivta (photograph author) The most immediately apparent significance of this deployment of spolia by the new Muslim mas- ters of Shivta is deprecatory, that is Muslim feet therefore supported three sizeable churches within were being invited to tread on the crosses as a sign its perimeter and was evidently connected with a of dominance of Islam and powerlessness of Chris- prosperous satellite monastic community, as well. tianity. Of course, many such examples can be Likewise, it is the only town with a covered cited to prove the mastery of a religion and its fol- mosque12. These aspects reveal strong communities lowers through subjugation of another’s holy signs of both Christians and Muslims, and prosperity and representations. Triumphantly passing over the deep into Late Antiquity. unresisting markers of another’s power was a long- That relative scale needs to be borne in mind for established trope by the Late Antique period. For understanding the choices made by the builders example, Sudanese conquerors of the Roman col- and patrons of the new Islamic house of worship, ony at Meroe cut off the head of the bronze statue for the photographs of the Colt expedition record of the emperor Augustus and buried it under the the original arrangement of the steps leading from steps of their own temple dedicated to victory; all the reservoir plaza into the anteroom of the mosque who entered inevitably (and at least for a time, (Pls 3, 5, 7-8). The upper step and part of the lin- knowingly) trod on the head of the defeated tel leading into the mosque itself are re-used stones enemy14. The builders of the synagogue at Sardis from a Christian building, or at least had been also placed a relief of Artemis and Cybele face intended for such a building. The stones’ upward- down in the forecourt, as a sign, it would seem, of turned faces show two crosses between two com- their faith’s rightness and of authorities’ collusion pass-produced rosettes (the left-hand stone is quite in that statement15. Christians performed similar faded, but legible). Two crosses in medallions are acts of intentional disrespect through spolia, as the on the central part of the stones forming the upper life of Porphyry (ca. 347-420), bishop of Gaza, step, both with flaring, equal-length arms, and are illustrates. The vita by Mark the Deacon (fl. fifth flanked by cross-like motifs in rosette form; the century) may date as late as ca. 600, but the epi- right-hand block is now broken, but both were sode in question purports to have occurred in 402, originally single blocks, probably used as a lintel over a doorway (Pl. 9)13. At a sharp angle to this upper step, and leading into the mosque itself, is a fourth cross, incised rather crudely but perfectly 12 Avni 1994; Rosen/Avni 1989; Jobling 1989. legible (Pl. 10). The carving is now worn, partly 13 Segal 1988, 32 (measurement given as w. 1.44 m, h. 31 m, d. 26 m), 152-153. Neither the left-hand lintel nor the through time since its exposure in the 1930s, but smaller slab at the entrance are noted. The original use of presumably also through the tread of feet by those the stones is not knowable now, and the possibility exists going from the plaza into the mosque vestibule and that the stones had never been used in a building, but had the prayer hall itself. The mosque raises important been found in a depot of some kind. See Negev/Margalit 1986. considerations then that have not been addressed 14 Van Eck 2010, 645; Stewart 2006; Boschung 1993, 160- in the scholarship directed at the town and its 161, Pl. 195; Haynes 1983/1984. Christian-Muslim phase. In the first place, the con- 15 Hanfmann et al. 1983.

105

95300_ECA8(2011)_07.indd 105 27/02/13 15:24 Pl. 7. View of steps leading into Mosque, Shivta (photograph Conway Library, The Courtauld Institute of Art, London, photograph 771/18 [35] date uncertain)

Pl. 8. View of steps leading into Mosque, Shivta (photograph author)

106

95300_ECA8(2011)_07.indd 106 27/02/13 15:24 Pl. 9. View of lintel forming steps leading into Mosque, Pl. 10. View of lintel leading into Mosque, Shivta Shivta (photograph author) (photograph author)

and so in that case, it can stand for a number of at the southwest corner of the Haram al-Sharif, possible Late Antique moments. In this passage, the Jerusalem, on which a cross had been plastered holy man made clear that building materials from over and replaced by a painted menorah17. Such temples could be re-used but also deployed in such demonstrations of the cooperation of stones in a way that the materials revealed their powerless- ideological assertions can of course be multiplied, ness: and scholars have performed rich analyses of stones’ abilities to speak eloquently for former When, therefore, the ashes were carried away and foes18. all the abominations were destroyed, the rubbish In the post-conquest period, Muslims were no that remained of the marble work of the Marneion, less forceful in finding ways for stones, which which they said was sacred, and in a place not to belonged once to Christians, to advocate the truth be entered, especially by women, this did the holy of their triumphant faith. The construction in the bishop resolve to lay down for a pavement before first half of the eighth century of a major mosque the temple outside in the street, that it might be in the heart of the formerly Byzantine and Chris- trodden under foot not only of men, but also women tian town of Resafa, in present-day Syria, reveals and dogs and swine and beasts. And this grieved some of the same dynamics at work at Shivta at the idolaters more than the burning of the temple. approximately the same time. The mosque was on Wherefore the most part of them, especially a scale distinct from the construction of Shivta’s the women, walk not upon the marbles even unto small mosque, but spolia was also deployed at the this day16. entrance to the mosque (Pls 11-12)19. On the north side of the courtyard of the mosque, part of a basin The social world of religion shifted for these stones was assembled from a sarcophagus fragment and, to the degree that women were encouraged to walk on top, a doorjamb that had a Greek inscription upon them. Naturally, the saint was not progres- with a tabula ansata, while a second sarcophagus sive in a modern sense, because he also made clear that animals were free to tread on them, too. The stones’ sanctity also remained to some degree, 16 Grégoire/Kugener 1930, 76.1-11; Hill 1913, 87. See because some still avoided them. But just the same, Saradi-Mendelovici 1990, 54. 17 Avni 2010, 41. At Tiberias, on the western shore of the Sea Christian authorities, through texts like this one, of Galilee, builders of the mosque there used basalt doors evidently were sensitive to the power of re-used from Jewish tombs as bases for columns; the slabs must stones. The tables turned for Christians in the have been partly visible. See Cytryn-Silverman 2009, 54. Middle East after the conquest, and Christians’ 18 Passages in Flood 2009a and Flood 2009b are particularly noteworthy in this respect. materials became available for statement of that 19 Sack 1996, 16-18; Brands 2002, 78-92, on spolia else- faith’s apparent abdication of absolute power. For where in the mosque complex. See also Grabar et al. 1978, example, a re-used lintel (now lost) was found I, 48, 190-191; II, 97-98.

107

95300_ECA8(2011)_07.indd 107 27/02/13 15:24 Pl. 11. View of Mosque at Resafa from north (photograph author)

Pl. 12. View of Mosque at Resafa, interior, south wall (photograph author)

108

95300_ECA8(2011)_07.indd 108 27/02/13 15:24 was placed at the north side of the basin. This to have followed correct protocol in the accumula- assemblage was made when the mosque was built, tion of re-used building materials for the Great but its function is not clear. Perhaps the basin was Mosque at Damascus, “…he collected marble, pil- used for purifying before entering the courtyard lars and stone but took nothing from anyone with- and then the mosque for prayer, or perhaps for ani- out payment”25. One cannot know if these same mals left outside; the possibility has to stand that protocols were observed in the smaller centers, like the assemblage started as a purifying basin and then Shivta or Resafa. One suspects not, but just the became trough. The ambivalence is noteworthy. same, sources like this testimony describing al- The stones were manipulated for functional pur- Walid’s scrupulous behavior reveal that respecting poses, one more deprecatory of the Greek Christian others’ stones entered into, if not always deter- culture defeated by Islam than the other, but in mined, their deployment in new contexts. either case visible, public declarations of a literally Stepping on the crosses to reach the prayer hall broken, powerless language, faith, and culture. But in Shivta’s sole mosque must always have had the use of the basin for purification makes clear meaning, and stepping on crosses could be antago- that the re-used stones were not entirely befouled nistic certainly. The placement of the lintels needs the way Porphyry believed his pagan stones to be. to be examined closely, as do the meanings crosses This positive integration into Islamic rites and underfoot had for Christians themselves in this behaviours is a possibility that admits a faceted region and period. On the floors of Christian mon- quality to Islamic use of spolia. uments at Shivta, a number of inscriptions was And of course the most striking aspect of the found by excavators, some with precise locations, mosque to any visitor to Resafa now – because the some now lost, but the majority of the inscriptions, spolia at the north end of the courtyard is no longer which commemorate deaths and burials, contain visible – is the way it is built onto the north wall of crosses, either at the beginning or end, or in one the Church of the Holy Cross, the sanctuary case in the middle of the text in an inset26. The famously dedicated to St Sergius, a powerful saint majority was found in the atrium of the north for Christians and Muslims20. In the Umayyad church; only two were recorded from the south period, such shared space was not uncommon, and church, one coming from the floor of the south the Great Mosque in Damascus, as is well known, aisle (the other is uncertain). was built after a period when a mosque and church dedicated to St John the Baptist shared the same precinct21. The close communication of mosque and church at Resafa reveals an apparent integration of worship of the saint between the communities, and the relationship in some form survived 20 Sack 1996; Ulbert 1986; Key Fowden 1999, 174-191. until the town was deserted in the mid-thirteenth 21 Guidetti, forthcoming; Flood 2001a. century. 22 The recent work by Mattia Guidetti is very persuasive in this regard. See Guidetti 2009b; idem 2010; idem, forth- That respectful stance of Muslims in relation to coming. the demonstrably sacred sites of Christians informed 23 Tohme 2009, 137-141. See Guidetti 2009a, 8-11; Gene- much – but not all, of course – interaction between quand 2004; idem 2008; Foote 2000 on the political, reli- those faiths in the Early Islamic period22. Mosques gious, and economic roles mosques and churches performed, and on the related reasons for their close association were, as churches were, more than space for liturgy in Late Antique towns and cities. and devotion; they were also dynamic spaces for 24 Barry 2010, 21-25; Flood 2001b. the communication and negotiation of religious, 25 Brinner 1963, 160. Al-Walid is said in the same passage to political, and social power23. Spolia, too, was part have destroyed a Christian image that he had been told would cause madness to anyone who harmed it. He of these relationships and cannot be interpreted, in destroyed it with his ax while Christians were watching. Late Antiquity or any other period, solely as trium- 26 Negev 1981, 47-67. For the latter, see 58 (no. 62) from phalist in meaning24. Some spolia was dangerously the baptistery of the north church. Arabic inscriptions soaked in the social life of the other and could not carved in the living stone were found in large quantities around the Negev site at Sede Beqer; they are visible on the be risked in new buildings without proper legal ground, while one vertical inscription in Arabic was found channels being followed. For example, al-Walid ibn at the open-air mosque at the site. See Nevo/Koren 2003, Abd al-Malik, caliph from 705-715, was said later 178-85; Nevo/Cohen/Heftman 1993.

109

95300_ECA8(2011)_07.indd 109 27/02/13 15:24 Crosses also appear in other settlements in the The east church there dates to the second half of Negev from the Late Antique period, but in differ- the fourth century and still shows its wealth through ent forms than in Shivta. The dating is difficult, an extensive set of floor mosaics in the nave (Pl. 18). because specific dates of desertion at other sites are A Maltese cross was placed at the west end of the difficult to ascertain. The larger Negevian town of nave in front of the central door (Pl. 19), while at Avdat appears to have been deserted by the end of the east end of the nave was set a straight-lined, the seventh century, but archaeologists have not equal-armed cross (Pl. 20)27. A seventh-century noted post-conquest presence. At Avdat, two example from Jerusalem indicates that the phenom- churches reveal quite different uses of crosses on the enon was not limited to the Negev. Near the Jaffa floor. The south church, traditionally dated to the Gate in Jerusalem, at Mamilla, a cave chapel was mid-fifth century with sixth-century changes at the discovered in the early 1990s that was used for east end and with a funerary inscription from 617, burials and was also decorated with a fresco (prob- has simple crosses inscribed on the floor of the aisles ably angels flanking Christ or Virgin and Child) (Pls 13-14). The crosses are in the eastern part of and with mosaic flooring that included an inscrip- the aisles and nearly parallel with one another. tion (‘for the redemptions and salvation of those, Moreover, crosses are included on funeral slabs in God knows their names’) and three simple crosses the floor, and carvings with stylized crosses are on the floor in front of the apse28. This chapel located on either side of the principal door leading appears to date to the seventh century, and these into the nave (Pl. 15). The north church at Avdat, examples then can stand for a longstanding tradi- dated to the fourth century, has a very different tion of crosses on floors. Even if their function is presentation of a cross underfoot; the southern aisle has a slab with six lines radiating from a centre cir- cle and six smaller Maltese crosses in the spaces between the rays (Pls 16-17). Mampsis also reveals the use of crosses underfoot. It appears to have been deserted by the early sixth century, but before that date, it was one of the wealthier towns in the region.

27 Negev 1988, 41-42. Many of the papyrus that survived at Nessana, began and ended with crosses, even after the conquest, so the sign also had some resonance in these textual contexts (Kraemer 1958). 28 Avni 2010, 36; Reich 1996.

Pl. 13. View of north aisle looking east, South Church, Pl. 14. View of south aisle looking east, South Church, Avdat (photograph author) Avdat (photograph author)

110

95300_ECA8(2011)_07.indd 110 27/02/13 15:24 Pl. 15. View of South Church, view east from courtyard, Avdat (photograph author)

Pl. 16. View of North Church, Avdat (photograph author)

Pl. 17. View of south aisle, North Church, Avdat (photograph author)

111

95300_ECA8(2011)_07.indd 111 27/02/13 15:24 Pl. 18. View of East Church, looking west, Mampsis Pl. 20. View of East Church, looking north, Mampsis (photograph author) (photograph author)

not clear, those crosses appear to have been visible and accessible to congregations at those churches29. Evidently Christians in the Negev and beyond were not above setting crosses underfoot in their churches, though explanations have not been offered for why this might be. However, the presence of crosses has been used as a dating tool, since the church council held at Hippo in 427 forbade the placing of crosses on the floors of Christian houses of worship. The reasoning goes that such display would not have happened in the aftermath of such strong statements on the part of the church. And yet that convention of crosses marking spots in churches did not end by the second quarter of the fifth century. The textual proof comes from the injunction against crosses on floors that was sternly repeated at the Quinisext Council in 692. Canon 73 stated:

The sign of the cross must not be set into the floor. The life-giving cross has shown us the salvation, and we ought with all diligence to render fitting honor to that through which we have been saved from the ancient fall. Wherefore, paying reverence to it in mind and word and sentiment, we com- mand that signs of the cross that have been set into the floor by certain persons should be erased com- Pl. 19. View of East Church, looking east, Mampsis pletely, in order that the trophy of our victory may (photograph author) not be insulted by the trampling feet of those who walk upon it. We decree, therefore, that those who henceforth set the sign of the cross into the floor should be excommunicated30. 29 I have not visited the site at Ododa, but see Negev 1997 on the site. 30 Nedungatt/Featherstone 1995, 155; Ohme 2006, 105- Evidently, Christians had continued the practice, 106, 264; Herrin 2009, 162. commonly enough that church officials needed to

112

95300_ECA8(2011)_07.indd 112 27/02/13 15:24 condemn it again. This proscription need only presumably being told to forget, in the sense of a demonstrate that Christians found the practice use- damnatio memoriae retaining the traces of absence ful or expressive, not why it might be so; perhaps for remembering40. Crosses were not displayed in the crosses marked burials in certain places, though public from the early eighth century, but they were at Avdat and Mampsis, burials were not found, or certainly still being produced in painting, stucco perhaps the crosses marked places to stand for cer- and metalwork into the Middle Ages41. tain devotions or rites, or places by which to The motivations for their placement are possi- remember significant events in the community. bly numerous, but the planners of the mosques The possibilities are open, but Quinisext shows must have purposefully placed the stones. The that the practice, however it signified, endured crosses on the step and lintel of the Shivta mosque probably into the eighth century31. are faint now, though unmistakable42. While the In the period leading up to and including Icon- placement of the crosses underfoot in Christian oclasm in Byzantium, the same centuries that saw churches in the Negev is not clear, it must too the building of the mosque at Shivta, the cross was have been purposeful, likely to mark significant, a strong symbol of victory and a polemical weapon sacred spots within the layout of the buildings. for both sides during the Iconoclastic debates. The carving of the crosses in the central doorjambs It was a sign of Christianity’s special role in human- in the south church at Avdat marked a limenal ity’s salvific history and of Christianity’s true wor- space, for another, and secured a transitional zone ship32. None of the crosses mentioned above show against anything untoward in the courtyard and signs of erasure. Indeed, no signs of destruction of profane world beyond. Such markings of Christian crosses are found at the sites in the Negev. One spaces with crosses are well attested throughout the limestone lintel, however, was found in room E of Middle Ages43. While a visitor to the Shivta the south church; it shows signs of editing in the mosque could easily have entered the building chiseling away of birds and animals on the lintel, from the right-hand side and trod not only on the but the cross was left intact33. Christians may have top step’s crosses, but also the cross on the thresh- been behind this editing, because the cross was old to the mosque, the crosses there did work untouched, and one might make an argument for when not underfoot, and protection and sealing iconoclasts at work in Shivta in the eighth century, may have been other aspects of their work. Given as such Christians were apparently active in Byzan- instances of crosses marking and controlling move- tium at this time34. But traces of frescoes are still ments and memory in churches in the Negev, the visible in the south apse of the south church, and mosque crosses could likewise have worked both as these paintings appear only to have been harmed by signs of Islamic triumph and as insurance, already the passage of time and the constant attention of guaranteed by Christian practice, against outside sun and wind (Pl. 21)35. Christians in the Negev did not only neglect to obey the edict against crosses on floors from the fifth century on, but they 31 And the practice clearly continued. See Kolbaba 2000, also did not correct their ‘mistakes’ by erasing them 53-54. at a later date. 32 Brubaker/Haldon 2011, 44-55, 80, 140-141, 193. Muslim attitudes towards the cross were also 33 Kirk 1937; Baly 1935, 171-181, Pl. IV. inconsistent or ambivalent36. Metal crosses were 34 Maguire 2009. 35 Figueras 1995b; idem 2006. ordered melted down in the early eighth century, 36 King 2007, 213-226. but in the middle of the ninth century, Thomas of 37 Penn 2010a. Marga tells the story of a cross being washed with 38 King 2007, 216. 39 Schlumberger 1986, 26-28. water that was later used to cure the son of a 40 37 See now Penn 2010b. wealthy Muslim . Objections to the cross do not 41 38 Examples include, but are not limited to, Steve/Jullien/Jul- equal iconoclasm, as G.R.D. King argued . An lien 2003; Bernard/Callot/Salles 1991; Talbot Rice 1932; example like the re-used lintel at the entrance to King 2007, 220. the fort at Qasr el-Heir el Gharbi, which was aban- 42 Harris (2007, 54) mentions that the crosses are stylized in a way not typical of other decorations at Shivta and not doned in 750, shows the erasure of a cross – though recognizable for that reason as Christian. its outline is still visible – and the retention of 43 For widespread examples, see Mahoney 2011, 32; Gerstel Greek inscriptions from 528-52939. Here one was 2006, 146-147; Nees 1987, 195-211; Kitzinger 1970.

113

95300_ECA8(2011)_07.indd 113 27/02/13 15:24 Pl. 21. Detail of fragmentary fresco depicting the Transfiguration (?), south apse, South Church, Shivta (photograph author)

contagion. Numerous examples of apotropaic mosque, but the wall was presumably blank facing carvings have been noted in Early Islamic architec- the plaza with the reservoir. The entrance to the ture44. That sharp-angled corner defined by the two church off the street was not monumental; in fact, cross-bearing stones before the mosque would have it was somewhat unimpressive, due to the adapta- been sealed against invisible and visible threats in tion of that ecclesiastical foundation to pre-existing that way, and likewise unwelcome reptiles and insects buildings in that block. The spolia then could also would naturally have been repelled, for example. have been a visible extension of the sacred zone rep- The boundary between sacred spaces of the resented and recognized by the majority of inhabit- Christian south church and the mosque would also ants of Shivta, and like the shared wall of baptistery not have been clearly marked without the spolia of and mosque, which is visible today but not in Late the cross lintels. The exterior wall is broken by the Antiquity, the spolia manifests that continuity of masonry joins that indicate the later building of the sanctity between church and mosque – all the while potentially asserting its lesser status through its lowly position on the ground. And yet the mirror- 44 Flood 2006. ing quality of Muslim behavior before Christian art 45 Guidetti 2009a, 9-11. The Early Islamic open-air Museum and architecture is noteworthy too, and the mim- at Sde Boqer, not far from Shivta, needs to be mentioned, icking of Christian placement of crosses on floors since it has many Arabic inscriptions on the rock surface around and under the mosque (therefore some inscriptions may be an example of creative misunderstanding by predate the structure there). For bibliography, see Magness which the use of spolia is both positive and negative 2003, 138-140. simultaneously45.

114

95300_ECA8(2011)_07.indd 114 27/02/13 15:24 Pl. 22. View of graffiti, west wall of side room, North Church, Shivta (photograph courtesy of the Israel Antiquities Authority Archives, Israel Antiquities Authority Archives: British Mandate Records Files: file no. 71, Isbeita, photograph 9811)

Graffiti was found by the Colt expedition in those declarations, and proof of the authority of the 1930s, and those plaster sections were removed Islam in insisting on its presence in Christian to Jerusalem before the storage buildings were monuments as well. Inscriptions of this sort have attacked and destroyed (Pl. 22)46. The graffiti was rarely been found in Muslim contexts, to my written in Kufic script, and the Arabic is difficult to decipher from the photographs, but it is prom- ising for a future researcher47. The passages of graffiti were said in an undated letter from Colt in 46 The plaster sections with graffiti are in the Rockefeller the Israel Antiquities Authority Archive to have Museum in Jerusalem, although I myself have not seen them. been found in the mosque and also one other 47 I thank my generous colleague Samer Ali for his help with noted as ‘Cufic inscription, side room, north reading these graffiti. church, looking west’48. Though clearly done by 48 Photograph numbers 8794, 8795, 8796, 8797, 8798, 8799 several different hands, the inscriptions appear to and 9911 (the last being from the north church). Baly (1935, 175-177) assumed the church was deserted, but no be similar in content, variations on the basmala, proof was offered for that supposition. Arabic graffiti has (‘In the name of God, the most gracious, the most been used as a marker of a monument falling into disuse merciful….’), in both the mosque and north in other centres in the Middle East, of course (e.g., church instances. Like the spolia, several interpre- Majcherek 2007, 37-38), but Baly also read one of the inscriptions in the mosque as naming a Hassan as builder tations offer themselves: public declaration of of the mosque. Finally, he reports small finds of ostrich Muslim piety in both Christian and Muslim eggs in the mosque, a feature common to both Christian buildings, evidence of blessing acquired through and Muslim devotional spaces (Green 2006).

115

95300_ECA8(2011)_07.indd 115 27/02/13 15:24 Pl. 23. View south to Mosque and reservoir plaza (photograph author)

knowledge49, but they are commonly attested in how to rule through models already in place, so Eastern Christian monastic practice50, so perhaps various strategies of that rule, such as material these declarations of Islam at Shivta were practices assertions like spolia and graffiti, came to be derived from Christians. In the same way the new manipulated for persuasion and control. Moreo- rulers of the former Byzantine territories learned ver, they can also show shared practices of devotion and piety at the same time. Such display of spolia and graffiti worked for social persuasion, because it showed the local com- 49 At Salamis, on eastern Cyprus, the post-conquest phase of munity in unmistakable ways the relative status of the Campanopetra church revealed what appeared to be an the old faith before the new. The reservoir plaza Arabic inscription in red in the south aisle. The content is not clear, but the presence of the inscription reveals that was not an agora, nor was this area a commercial the Arabic victors did not intend only to destroy the main centre for the town, but it was at a crossroads in monuments of the town. Perhaps, like at Shivta, they the town (Pl. 23). The spolia was a highly public intended a public declaration of control and accumulated declaration, and it was placed at a convergence of blessing (Stewart 2010, 179; Roux 1998, 250). For evidence of Christians using the Basmala, see Farès 2010 and sightlines that would signal exclusion and inclu- 2011. sion in the power and authority the new faith of 50 Walker 2010, 327-328; Hirschfeld 1992, 192, on the well- Islam embodied there51. The crosses there presum- attested custom of writing on walls among monks. ably heightened interest on the part of Christians 51 Sweetman 2010. Guidetti (forthcoming) makes the argument that re-use of Christian materials for the denial of in what happened within the mosque, in the same that faith only occurs unequivocally in the eleventh cen- way Muslims were also socially curious and tury. engaged with what happened inside churches.

116

95300_ECA8(2011)_07.indd 116 27/02/13 15:24 When Jacob of Edessa (active 684-708) warned Brilliant, R., D. Kinney (eds) 2011, Reus Value: Spolia and that Christians needed to close church doors dur- Appropriation in Art and Architecture from Constantine to ing liturgies, he was concerned about Muslims Sherri Levine, Farnham, Surrey. 52 Brinner, W. (ed.) 1963, Muhammad Ibn Sasra: A Chronicle of who ridiculed mysteries from outside . Evidently, Damascus 1389-1397, 2 vols, Los Angeles. Muslims were curious, perhaps active in some Brubaker, L., J. Haldon 2011, Byzantium in the Iconoclast Era places, and sometimes open about expressing their c. 680-850: A History, Cambridge. opinion about Christian rites; they were also gen- Caner, D.F. 2010, History and Hagiography from the Late erous, judging by mention of gifts that had to be Antique Sinai, Liverpool. correctly assimilated into Christian worship, and Cytryn-Silverman, K. 2009. ‘The Umayyad Mosque of Tibe- sometimes awkward socially53. Christians presum- rias’, Muqarnas 26, 37-61. ably felt the same way about their Muslim neigh- Eck, C. van 2010, ‘Living Statues: Alfred Gell’s Art and Agency, Living Presence Response and the Sublime’, Art bours, but perhaps a little less open about express- History 33(4), 642-659. ing their opinion in all situations. Erez Edelson, R. 2006, ‘Settlement Distribution in the Byz- From within, the privilege of the minority antine Negev: A Case Study Comparing the Shivta Region would likewise have been secured, when looking and the Inter-settlement Area in the Negev Highlands’, in: out of the anteroom of the mosque into the social A.S. Lewin, P. Pellegrini (eds), Settlements and Demogra- space of the reservoir plaza. As a carefully planned phy in the Near East in Late Antiquity. Proceedings of the zone of demarcation the spolia signaled simultane- Colloquium, Matera 27-29 October 2005, Pisa/Rome, 51-57. ously inclusion in the sacred aura of the established Erickson-Gini, T. 2010, Nabataean Settlement and Self-Organ- faith of the town, and exclusion from the majority ized Economy in the Central Negev: Crisis and Renewal, and superiority to it. Respect and denigration could Oxford (BAR 2054). inhabit the same place, and stones spoke strongly Evenari, M., L. Shaman, N. Tadmor 1971, The Negev: The about this complicated relationship between Chris- Challenge of a Desert, Cambridge, MA. tianity and Islam, even in this small, rather remote Farès, S. 2010, ‘L’inscription arabe de Kilwa: nouvelle lecture’, town. The social practices that would have made Semitica et Classica 3, 241-249. Farès, S. 2011, ‘Christian Monasticism on the Eve of Islam: those stones’ eloquent speech fully audible are only Kilwa (Savoi Arabia) – New Evidence’, Arabian Archaeology distantly comprehensible now, but their traces in and Epigraphy 22, 243-252. the stones at Shivta still tell a vivid, faceted story of Figueras, P. 1995a, ‘Monks and Monasteries in the Negev Islam’s sophisticated remaking of the world. Desert’, Liber Annuus 45, 401-450. Figueras, P. 1995b, ‘Pilgrims to Sinai in the Byzantine Negev’, in: E. Dassmann, J. Engemann (eds), Akten des XII. Inter- BIBLIOGRAPHY nationalen Kongresses für christliche Archäologie, Bonn, 22.- 28. September 1991, Münster, 756-762. Avni, G. 1994, ‘Early Mosques in the Negev Highlands: New Figueras, P. 2006, ‘Remains of a Mural Painting of the Trans- Archaeological Evidence on Islamic Penetration of South- figuration in the Southern Church of Sobata (Shivta)’, ern Palestine’, Bulletin of the American Schools of Oriental Aram 18, 127-151. Research 294, 83-100. Flood, F.B. 2001a, The Great Mosque of Damascus: Studies on Avni, G. 2010, ‘The Persian Conquest of Jerusalem (614 C.E.) the Makings of an Umayyad Visual Culture, Leiden (Islamic – An Archaeological Assessment’, Bulletin of the American History and Civilization 33). Schools of Oriental Research 357, 35-48. Flood, F.B. 2001b, ‘The Medieval Trophy as an Art Historical Baly, C. 1935, ‘S’baita’, Palestine Exploration Fund, Quarterly Trope: Coptic and Byzantine ‘Altars’ in Islamic Contexts’, Statement 68, 171-181. Muqarnas 18, 41-72. Barry, F. 2010, ‘Disiecta membra: Ranieri Zeno, the Imitation Flood, F.B. 2006, ‘Image against Nature: Spolia as Apotropaia of Constantinople, the Spolia Style, and Justice at San in Byzantium and the dar al-Islam’, Medieval History 9(1), Marco’, in: H. Maguire, R.S. Nelson (eds), San Marco, Byz- 143-166. antium and the Myths of Venice, Washington, DC, 7-62. Flood, F.B. 2009a, Objects of Translation: Material Culture and Bernard, V., O. Callot, J.-F. Salles 1991, ‘L’église d’al-Qou- Medieval ‘Hindu-Muslim’ Encounters, Princeton. sour Failaka, État de Koweit: Rapport préliminaire sur une premiére campagne de fouilles, 1989’, Arabian Archaeology and Epigraphy 2, 145-181. Boschung, D. 1993, Die Bildnisse des Augustus, Berlin. Brands, G. 2002, Die Bauornamentik von Resafa-Sergiupolis: 52 Vööbus 1975, I, 237 (Syriac), 219 (English); Hoyland Studien zur spätantiken Architektur und Bauaustattung in 2008, 16. Syrien und Nordmesopotamien, Mainz (Resafa 6). 53 For examples, see Hoyland 2008, 16-18.

117

95300_ECA8(2011)_07.indd 117 27/02/13 15:24 Flood, F.B. 2009b, ‘An Ambiguous Aesthetic: Crusader Spolia Herrin, J. 2009, ‘The Quinisext Council (692) as a Continu- in Ayyubid Jerusalem’, in: R. Hillenbrand, S. Auld (eds), ation of Chalcedon’, in: R. Price, M. Whitby (eds), Chal- Ayyubid Jerusalem: The Holy City in Context 1187-1250, cedon in Context: Church Councils 400-700, Liverpool, London, 202-215. 148-168. Foote, R. 2000, ‘Commerce, Industrial Expansion and Hill, G.F. (trans.) 1913, The Life of Porphyry Bishop of Gaza by Orthogonal Planning: Mutually Compatible Terms in Mark the Deacon, Oxford. Settlements of Bilad al-Sham during the Umayyad Period’, Hirschfeld, Y. 1992, The Judean Desert Monasteries in the Mediterranean Archaeology 13, 25-38. Byzantine Period, New Haven/London. Genequand, D. 2004, ‘Al-Bakhra (Avatha), from the Tetrar- Hirschfeld, Y. 2003, ‘Social Aspects of the Late-Antique Vil- chic Fort to the Umayyad Castle’, Levant 36, 225-242. lage of Shivta’, Journal of Roman Archaeology 16(1), 395- Genequand, D. 2008, ‘An Early Islamic Mosque at Palmyra’, 408. Levant 40(1), 3-15. Hoyland, R. 2008, ‘Jacob and Early Islamic Edessa’, in: Gerstel, S.E.J. 2006, ‘An Alternate View of the Late Byzantine R.B. ter Haar Romeny (ed.), Jacob of Edessa and the Syriac Sanctuary Screen’, in: S.E.J. Gerstel (ed.), Thresholds of the Culture of His Day, Leiden/Boston, 11-24. Sacred: Architectural, Art Historical, Liturgical and Theo- Jobling, W.J. 1989, ‘Aqaba—Ma’an: Archaeological and Epi- logical Perspectives on Religious Screens, East and West, graphic Survey’, in: D. Homès-Fredericq, J.B. Hennessy Washington, DC, 134-161. (eds), Archaeology of Jordan II, Leuven, 16-24. Grabar, O., R. Holod, J. Knustad, W. Trousdale 1978, City Key Fowden, E. 1999, The Barbarian Plain: Saint Sergius in the Desert: Qasr al-Hayr East, 2 vols, Cambridge, MA. between Rome and Iran, Berkeley. Green, N. 2006, ‘Ostrich Eggs and Peacock Feathers’, King, G.R.D. 2007, ‘Islam, Iconoclasm and the Declaration of Al-Masaq 18, 27-66. Doctrine’, in E.R. Hoffman, Late Antique and Medieval Art Grégoire, H., M.-A. Kugener (eds) 1930, Marc le Diacre: Vie of the Mediterranean World, Malden, MA, 213-226 (reprint de Porphyre évêque de Gaza, Paris. from BSOAS 48(2), 1985, 267-277). Guidetti, M. 2009a, ‘The Byzantine Heritage in the Dar al- Kirk, G.E. 1937, ‘Era-Problems in the Greek Inscriptions of Islam: Churches and Mosques in the al-Ruhr between the the Southern Desert’, Journal of the Palestine Oriental Soci- Sixth and Twelfth Centuries’, Muqarnas 26, 1-36. ety 17, 209-211. Guidetti, M. 2009b, ‘L’Editto de Yazid II’: immagini e iden- Kitzinger, E. 1970, ‘The Threshold of the Holy Shrine: Obser- tità religiosa nel Bilad al-Sham dell’VIII secolo’, in: V. Pace vations on Floor Mosaics at Antioch and Bethlehem’, in: (ed.), L’VIII secolo: un secolo inquieto. Atti del Convegno P. Granfield, J.A. Jungmann (eds), Kyriakon: Festschrift internazionale di studi, Cividale del Friuli, 4-7 dicembre Johannes Quasten, Vol. II, Münster, 639-647. 2008, Cividale del Friule, 69-79. Kolbaba, T.M. 2000, The Byzantine Lists: Errors of the Latins, Guidetti, M. 2010, ‘Sacred Topography in Medieval Syria and Urbana/Chicago. its Roots between the Umayyads and Late Antiquity’, in: Kraemer, C.J. 1958, Excavations at Nessana. Vol. III. Non- A. Borrut, P.M. Cobb (eds), Umayyad Legacies: Medieval Literary Papyri, Princeton. Memories from Syria to Spain, Leiden/Boston, 339-363. Magness, J. 2003, The Archaeology of the Early Islamic Settle- Guidetti, M. forthcoming, ‘Churches and Mosques in the ment in Palestine, Winona Lake, IN. Cities of the Ballad al-Sham: Coexistence and Transforma- Maguire, H. 2009, ‘Muslims, Christians and Iconoclasm: tion of the Urban Fabric’, in: Early Islamic Urbanism. Erasures from Church Floor Mosaics during the Early Gutwein, K.C. 1981, Third Palestine: A Regional Study in Islamic Period’, in: C. Hourihane (ed.), Byzantine Art: Byzantine Urbanization, Washington, DC. Recent Studies. Essays in Honor of Lois Drewer, Tempe, AZ, Haiman, M. 1995, ‘Agriculture and Nomad-State Relations in 111-119. the Negev Desert in the Byzantine and Early Islamic Peri- Mahoney, L. 2011, ‘The Histoire ancienne and Dialectical ods’, Bulletin of the American Schools of Oriental Research Identity in the Latin Kingdom of Jerusalem’, Gesta 49(1), 297, 29-53. 31-51. Hanfmann, G.M.A. et al. 1983, Sardis from Prehistoric to Majcherek, G. 2007, ‘The Late Roman Auditoria of Alexan- Roman Times. Results of the Archaeological Exploration of dria: An Archaeological Overview’, in: T. Derda, T. Markie- Sardis 1958-1975, Cambridge, MA/London. wicz, W. Wipszycka (eds), Alexandria: Auditoria of Kom Harris, A. 2007, ‘“Let streams of living water flow”: The el-Dikka and Late Antique Education, Warsaw, 11-50. Archaeology of a Secular Settlement in a Monastic Land- Nedungatt, G., M. Featherstone (ed. and trans.) 1995, The scape’, Reading Medieval Studies 33, 37-67. Council in Trullo Revisited, Rome. Hayden, R.M. et al. 2011, ‘The Byzantine Mosque at Trilye: Nees, L. 1987, The Gundohinus Gospels, Cambridge, MA. A Processual Analysis of Dominance, Sharing, Transfor- Negev, A. 1981, The Greek Inscriptions from the Negev, Jerusa- mation and Tolerance’, History and Anthropology 22(1), lem. 1-17. Negev, A. 1982, ‘Christen und Christentum in der Wüste Haynes, D.E.L. 1983/1984, ‘The Date of the Bronze Head Negev’, Antike Welt 13(2), 2-33. from Meroë’, in: A. Adriani (ed.), Alessandria e il mondo Negev, A. 1988, The Architecture of Mampsis: Final Reports. ellenistico-romano: Studi in onore di Achille Adriani, Vol. I, Volume II: The Late Roman and Byzantine Periods, Jerusa- Roma, 177-181. lem (Qedem 27).

118

95300_ECA8(2011)_07.indd 118 27/02/13 15:24 Negev, A. 1990, ‘Mampsis – The End of a Nabataean Town’, Segal, A. 1985, ‘Shivta – A Byzantine Town in the Negev Aram 2, 337-365. Desert’, Journal of the Society of Architectural Historians Negev, A. 1993, ‘Sobaita’, in: E. Stern, A. Lewinson-Gilboa, 44(4), 317-328. J. Aviram (eds), The New Encyclopedia of Archaeological Segal, A. 1988, Architectural Decoration in Byzantine Shivta, Excavations in the Holy Land, Vol. IV, Jerusalem, 1404- Negev Desert, Israel, Oxford (BAR 420). 1410. Shereshevski, J. 1991, Byzantine Urban Settlements in the Negev, A., S. Margalit 1986, ‘Subeita’, RB 93, 267-269. Negev, Beer Shiva. Nevo, Y.D., Z. Cohen, D. Heftman 1993, Ancient Arabic Steve, M.-J., C. Jullien, F. Jullien 2003, ‘Le monastère et ses Inscriptions from the Negev, Vol. I, Jerusalem. environs’, in: M.-J. Steve (ed.), L’île de Kharg: une page Nevo, Y.D., K. Koren 2003, Crossroads to Islam: The Origins de l’histoire du Golfe Persique et du monachisme orientale, of the Arab Religion and the Arab State, Amherst, NY. Neuchâtel, 115-121. Ohme, H. (ed.) 2006, Concilium Quinizextum. Das Konzil Stewart, C.A. 2010, ‘The First Vaulted Churches in Cyprus’, Quinisextum, Turnhout. Journal of the Society of Architectural Historians 69(2), 162- Ovadiah, A. (trans. R. Kirson) 1970, Corpus of the Byzantine 189. Churches in the Holy Land, Bonn. Stewart, P. 2006, ‘The Image of the Roman Empire’, in: Penn, M.P. 2010a, ‘A Temporarily Resurrected Dog and R. Shepherd, R. Maniura (eds), Presence: The Inherence of Other Wonders: Thomas of Marga and Early Christian/ the Prototype within Images and Other Objects, Aldershot, Muslim Encounters’, MedEnc 16, 209-242. 243-259. Penn, M.P. 2010b, ‘Moving Beyond the Palimpsest: Erasure Sweetman, R. 2010, ‘The Christianization of the Peloponnese: in Syriac Manuscripts’, Journal of Early Christian Studies The Topography and Function of Late Antique Churches’, 18(2), 261-303. Journal of Late Antiquity 3(2), 203-261. Petersen, A. 2005, The Towns of Palestine under Muslim Rule Talbot Rice, D. 1932, ‘The Oxford Excavations at Hira, AD 600-1600, Oxford (BAR 1381). 1931’, Antiquity 6, 276-279. Reich, R. 1996, ‘“God Knows Their Names”: Mass Christian Tohme, L. 2009, ‘Spaces of Convergence; Christian Monaster- Grave Revealed in Jerusalem’, Biblical Archaeology Review ies and Umayyad Architecture in Greater Syria’, in: 22(2), 26-33, 60. A. Walker, A. Luster (eds), Negotiating Secular and Sacred Ribak, E. 2007, Religious Communities in Byzantine Palestina: in Medieval Art: Christian, Islamic and Buddhist, Burling- The Relationship between Judaism, Christianity and Islam, ton, 129-145. AD 400-700, Oxford (BAR 1646). Tsafrir, Y., L. Di Segno, J. Green 1994, Tabula Imperia Rom- Rosen, S.A., G. Avni 1989, ‘Har Oded’, IEJ 39, 117-120. ani: Iudaea Palestina. Maps and Gazeteer, Jerusalem. Roux, G. 1998, La basilique de la Campanopétra, Paris (Sala- Ulbert, T. 1986, Die Basilika des Heiligen Kreuzes in Resafa- mine de Chypre 15). Sergiupolis, Mainz (Resafa 2). Rubin, R. 1997, ‘Priests, Soldiers and Administrators: Society Vööbus, A. 1975, The Synodicon in the West Syrian Tradition, and Institutions in the Byzantine Negev’, Mediterranean Louvain (CSCO 367-368, Scriptores Syri 161-162). Historical Review 12(1), 56-74. Walker, J.T. 2010, ‘Ascetic Literacy: Books and Readers in Sack, D. 1996, Die Grosse Moschee von Resafa – Rusafat Hisam, East-Syrian Monastic Tradition’, in: H. Börn, J. Wiese- Mainz (Resafa 4). höfer (eds), Commutatio et contentio. Studies in the Late Saradi-Mendelovici, H. 1990, ‘Christian Attitudes toward Roman, Sasanian, and Early Islamic Near East in Memory of Pagan Monuments in Late Antiquity and Their Legacy in Zeev Rubin, Düsseldorf, 306-345. Later Byzantine Centuries’, DOP 44, 47-61. Woolley, C.L., T.E. Lawrence 2003, The Wilderness of Zin, Schlumberger, D. 1986, Qasr el-Heir el Gharbi, Paris. Winona Lake, IN. Segal, A. 1983, The Byzantine city of Shivta (Esbeita), Negev Desert, Israel, Oxford (BAR 179).

119

95300_ECA8(2011)_07.indd 119 27/02/13 15:24