Pictorial graffiti of a horse rider and an orans from Byzantine Shivta: some thoughts on context and interpretation

ABSTRACT

The ruins of Byzantine Shivta, dominated by its three monumental churches, are visible from afar in the Negev Desert environment. They attract occasional visitors, tourists, pilgrims, and nomads who for centuries have recorded their personal marks and expressions in the form of written and pictorial graffiti. Commonly considered as vandalism, graffiti now is a recognized means of informal communication and recently has become an eligible field of research. The study of graffiti contributes to understanding social, cultural, religious, and even political contexts of ancient sites, and thereby they become an inseparable part of their history. This study surveys pre-modern pictorial graffiti discovered in Shivta and especially emphasizes two incised graffiti, a horse rider at the entrance to the North Church and an orans figure on the doorway of the ‘Governor’s House’ next to the Central Church. Published here for the first time, they constitute seldomly seen relics from the times when Shivta was still teeming with life. Discussed in context of their iconography, local and regional parallels, and spatial settings, these minute personal images seemingly reflect the main aspects of life in this desert settlement, religious and military, emerging from recent archaeological, artistic, and historical research of the site.

KEYWORDS:

• Byzantine period
• shivta
• christian graffiti
• horse rider
• orans

Introduction: contemporary studies of graffiti

In the ninth century a Viking named Halfdan carved a graffito inscription ‘Halfdan carved these runes’, equivalent to ‘Halfdan was here’ in modern-day parlance, on the marble parapet in the Hagia Sophia gallery in Constantinople (Hubbard 2017, 179). Had it been executed by a visitor in modern times, the act definitely would be considered vandalism; but the ninth-century graffito became part of Hagia Sophia history and is protected by glass as an important historical document.

The present study brings to light pictorial graffiti discovered in Shivta, a ruined Byzantine village in the Negev Desert, Israel. Despite long-lasting scholarly interest in the site, its graffiti were never catalogued, studied, or published. Today it is widely recognized that graffiti, informal expressions carved, written, or painted in a variety of private and public places, were a common practice during Late Antiquity. To name a few, clusters of late antique graffiti were documented in the following sites: Dura Europos, Syria; Aswan and Deir el-Medina, Egypt; Aphrodisias, Turkey; Rome and Pompeii, Nubia and Sudan. Only in recent years has their value begun to be fully appreciated and they have become an actual field of research. More and more studies are published aiming to catalogue, classify, and understand numerous ancient and medieval¹ graffiti and to place them within a specific context of time and space (Baird and Taylor 2011; Benefiel 2011; Benefiel and Keegan 2017; Dijkstra 2012; Emberling and Davis 2019; Felle and Ward-Perkins 2021; Handley 2017; Jacquet-Gordon 2003; Keegan 2014; Milnor 2014; Polkowski 2023). In these studies, a personal mark left by a visitor on the site is addressed as a memory sign and a historical source (Leatherbury 2020). It is rather surprising how much can be learned from graffiti. In the words of Baird, graffiti ‘are part of the life of the structure, not an act of defacement nor even a change of use’ (Baird 2011, 66).

The place chosen, the language, the words themselves, which could range from a simple inscription of a name to an acclamation, prayer, invocation, personal or political message, or a pictorial sign and image makes graffiti a powerful medium of communication. Most studies deal with written graffiti, while pictorial images receive relatively less attention, although, clearly, both belong to the same genre. It is perhaps more rewarding to deal with words. Left by individuals who visited a place of some importance, written graffiti provide a scholar with valuable information from which a wider profit can be extracted (Nowakowski 2017).

Images are more obscure. Their interpretation is less straightforward and subjective, in many cases providing only hypothetic clues. Crudely sketched or carved images may of course reflect conventional contemporary iconography and symbols, but also may be intimately connected to a personal world of reference, making their interpretation much more difficult. Not by chance, they were even labelled by scholars as ‘selfies of the ancient world’ (Stern 2015). Perhaps, they can be referred to as similar to a modern ‘emoji’, a picture-word, a way of communicating by means of encoded messages embedded in simple signs.

In this respect, a comprehensive catalogue published by Langner (2001) is of great help in understanding graffiti from Late Antiquity. It includes variations of motifs and symbols; most widespread are various animals and birds, ships and boats, geometric shapes, rosettes, lines, and sporadic human faces and figures, sometimes arranged in scenes. To this repertoire, crosses were added as a clear indicator of Christian identity. The quality of the graffiti varies from very schematic designed with some crude lines to more accurate and precise. The graffiti of human figures (e.g. Pompeii, Beit Shearim, Khirbet Beit Lei, Bawit, and more) may have differently shaped heads with an indication of facial elements, while hands and legs are depicted only as lines, akin to children’s drawings (Langner 2001, Pls. 30–35), some are identified by attributes as gladiators or soldiers, or a mythological character (Langner 2001, Pls. 37, 773–780; 45; 74–78; 84), and others remain almost unrecognizable, their identity surmised only by comparison (Langner 2001, Pl. 61, 1077). An interesting example comes from the John the Baptist cave near Ein Karem, west of Jerusalem, where several figures and objects are associated with important relics, while a schematic figure with raised hands and wearing a hairy (?) skirt was interpreted as John the Baptist (Gibson and Tabor 2005). Such schematic figures are basic indications of a visitor’s presence at the sites. The most schematic graffiti are made quickly, while those with more details and precision could indicate that the visitor spent more time at the spot and invested some energy in their work.

As in most of the ancient sites, graffiti are abundant in Shivta, both inscribed and pictorial, painted and incised. The graffiti in Shivta were documented during a long span of about 20 years, starting in early 2000 with surveys by Hirschfeld and Tepper (2006) and especially during recent archaeological excavations by researchers on behalf of the University of Haifa (Tepper and Bar-Oz 2019, 2020; Tepper et al. 2018; Tepper, Weissbrod, and Bar-Oz 2015).² During this time, important discoveries were made, including analysis of the wall paintings in the South and North Churches (Linn, Acker, and Tepper 2019, Linn et al. 2019; Maayan-Fanar et al. 2018). Repeated observations within the site during the past two decades, especially within the walls of the churches, allowed a thorough methodological survey of the graffiti (e.g. Tchekhanovets, Tepper, and Bar-Oz 2017). In fact, many were documented and analysed for the first time by this study.

This study surveys only pictorial graffiti discovered in Shivta that can be attributed to the period when the village was still flourishing or perhaps were made by Christian pilgrims after its abandonment. We have documented these in a catalogue of relevant motifs, including their basic identification if possible.³ In addition, we focus on two graffiti, a horse rider in the North Church and an orans figure in the ‘Governor’s House’, which we believe are relics from Shivta’s Byzantine past. We aim to understand them in a spatial context, a frequent consideration in modern research of graffiti (Whiting 2021, 23–27). Although graffiti are informal personal signs that do not correspond to precise contemporary iconographic and stylistic schemes (Baird and Taylor 2011, 1–19; Klein, Zissu, and Distelfeld 2018, 407), they still share acceptable common characteristics that reflect common iconography.

Shivta: historical and archaeological background

Shivta (Figure 1) is situated in the northern Negev heights of Israel, in a semi-arid area (perennial average rainfall less than 100 millimetres) at the south-western end of the Shivta Ridge and at an altitude of 360 metres above sea level. The site lies approximately 40 kilometres south of Beer Sheva, at the heart of agricultural land, far away from the Roman imperial roads that crossed the Negev. The boundaries of the ancient settlement and its streets are clearly visible (approximately 90 dunams), the walls of the private houses being partially preserved up to a height of two to four metres and the walls of the public buildings up to a height of 8–10 metres. It had about 170 houses where an estimated 2200–2500 residents lived, three churches, and a developed array of streets and courtyards, two public water reservoirs in its southern part, and a small mosque. Shivta was excavated extensively in the 1930s by a New York University Expedition in collaboration with the British School of Archaeology in Jerusalem, under the direction of H. D. Colt, but the results of the excavations were never published. Additional surveys and excavations were performed at the site, but the foundations of Shivta are still vague. A few pottery sherds found and recorded at the edge of the site were dated to the Roman period (not before the second century). According to the archaeological findings revealed in the survey and the latest excavation at the site (see note 2), the village began to expand sometime in the fourth century, while reaching its peak during the fifth and sixth centuries (Figure 2). Its gradual decline began in the second half of the sixth century and continued during the seventh century until it was completely abandoned by the permanent residents in the Early Islamic period (no later than the end of the eighth to the beginning of the ninth century; Hirschfeld 2003, Shereshevski 1991, 61; Tepper 2019a, 2019b; Tepper, Weissbrod, and Bar-Oz 2015; Tepper et al. 2018; Tsafrir, Di Segni, and Green 1994, 234).

Figure 1. Location map of the Negev and Shivta (prepared by Sapir Hadd according to Hirschfeld [2003]).

Figure 2. A map of Shivta with indication of the location of the churches, stables, and graffiti (prepared by Sapir Hadd according to Hirschfeld [2003]).

Pictorial graffiti in Shivta: distribution and motifs

There are numerous non-pictorial graffiti in Shivta. They are mainly inscriptions of names, dedications, and evocations. Some are scattered all over the stone surfaces of the walls and apses of the churches, baptisteries, and houses. Others are concentrated in the passageways. Written or painted graffiti on doorposts exist frequently in antiquity in Jewish and Christian spaces and might be connected to the symbolical importance of doorways as powerful protectors within those traditions (Stern 2018, 53–54).

Only a few graffiti in Shivta are pictorial, unaccompanied by any identifying inscription (Table 1; Figure 2). Unlike inscribed graffiti, pictorial graffiti do not appear in clusters; each one occupies a separate space, either because a stone was used as a frame or artificial lines were added to frame the motif. They are mostly incised and include some common motifs such as geometric shapes, rosettes, and lines (Table 1, nos 1–4). These motifs are common within the Roman world (e.g. Pompeii, Rome, Ostia in Langner [2001, Pls. 3–6, 8]). Simple crosses are scratched all over the Shivta site (Table 1, nos 13–16). Some have forked arms akin to the crosses made of shells discovered in the water cistern (Tsuk 2002, 73–74, figs. 11, 12; Humphreys 2019, 232)⁴ and a painted cross found on a stone in secondary use in Building 86, both in Shivta (Maayan-Fanar, Tepper, and Linn, In preparation). This type of cross became especially widespread in the sixth and seventh centuries.

Table 1. Pictorial graffiti in Shivta (Photographs by Yotam Tepper and Dror Maayan; drawings by Emma Maayan).

No.	Motif	Spatial place	Picture and drawing
1	A circle	South Church
2	A four-petal rosette	Central Church
3	A four-petal rosette	North Church baptistery
4	Three straight lines, basic geometric motif	Governor’s house
5	Palm frond motif on three-legged base. More straight lines forming abstract motifs	Mosque
6	An abstract shape made of dotted lines	Central Church apse
7	A horizontal line with two short lines at its end, all made of dotted lines	Central Church apse
8	A ship with sails and pedals crossed with vertical straight lines	South Church lateral apse
9	A figure on a horse’s back with a circle above the rider (upside down)	Colt’s house (reused)
10	A figure on a horse’s back holding a spear (?) with additional figure(s) in front of them	North Church – the entrance from the narthex
11	A standing figure in orans posture with a cross above his head	Entrance to Central Church compound
12	A small-sized simple cross above a human figure	Entrance to Central Church compound
13	A cross with equal arms	Entrance to North Church atrium
14	A horned cross	Entrance to domestic structure (no. 123 according to Hirschfeld [2003])
15	A horned cross within a circle	North Church
16	A cross made of dots	South Church, pillar on the eastern entrance from the narthex to the central hall

In one case, on the lateral apse of the South Church, graffiti of a ship can be detected (Table 1, no. 8).⁵ Ships and boats are most popular in graffiti. Their appearance within sacred spaces is usually connected to pilgrimage, although the variety of interpretations may include connections to commercial trade and the afterlife in the context of burials (Whiting 2021, 22). They are shaped in endless variations, usually recognizable by basic details such as a general ship body, sails, and paddles, but may also be very simplified (up to one curved line) or extremely detailed (Langner 2001, Pls. 116–146). The Shivta boat is somewhat ambiguous, yet it does seem to have a simple body, with a mast and sails, and possibly oars. The image is overlapped by many lines whose connection to the boat are difficult to understand.

Round holes are scooped out on the body of the boat and the upper part of the mast. This interesting phenomenon of scooped, relatively deep round holes within pictorial graffiti, or even graffiti constructed of such holes, is also evident in other places in Shivta; but their exact meaning remains unclear (Table 1, nos 6, 7, 16). Such a phenomenon has been documented on a much wider scale in ancient sacred sites in Egypt and Sudan (Emberling and Davis 2019, 32). It has been proposed that stone powder from scooped holes was used in magical protective and healing practices.

Most of the graffiti were scratched on smooth surfaces of large stones. Their date is unclear. Many were probably added by various visitors to the site and over a significant length of time during which the ruins of the village dominated the sands of the desert (see also Tchekhanovets, Tepper, and Bar-Oz 2017). Were the creators of these graffiti pilgrims or sporadic tourists? In most cases, we will never know. Most of the graffiti are located on the upper parts of the buildings that remain visible above debris. In their vicinity are various tribal marks [wusum] left by local Bedouins (Khamisy, Tepper and Bar-Oz in preparation n.d.).

There are clues, however, that some of the pictorial graffiti may belong to the period when Christian Shivta still flourished, approximately in the fifth to seventh centuries. Among them are two graffiti: a horse rider in the North Church (Table 1, no. 10; Figure 3) and a figure, perhaps a saint, on the entrance to the Central Church compound (Table 1, no. 11; Figure 6). They deserve special attention due to their visual peculiarities, spatial placement, relevance to the Byzantine period, and intriguing regional parallels.

Figure 3. A horse rider graffito, the North Church, Shivta (photograph by Dror Maayan; drawing by Emma Maayan).

A horse rider graffito and its regional and iconographic context

The first graffito, a horse rider, discussed here at length was found in the North Church (Table 1, no. 10; Figures 3 and 4) located on the northern edge of the settlement. The church was first surveyed by Palmer (1871, 31) and excavated by the Colt expedition (1936–1938) which revealed that it went through several stages (Baly 1935, 173; Colt 1935, 9–10). Entrance to the church complex was through an extensive atrium paved with limestone and bordered by pillars to the east, west, and north. The deviation of the columns to the east of the atrium served as a kind of narthex from which three openings led to the prayer hall of the church (Negev 1993, 1408; Shereshevski 1991, 75–76).

Figure 4. The façade of the northern entrance of the North Church from the narthex to the church hall and the location of a horse rider graffito (photograph by Dror Maayan).

The church, with its marble covered apses, is rich with archaeological and artistic findings, although many of them were displaced or lost (Maayan-Fanar and Tepper 2022a; Margalit 1987; Rosenthal-Hegginbottom 1982; Fisher and Tepper 2021). The findings include sixth to late seventh-century tombstones within the area of the atrium and the narthex (Baly 1935, 174–175; Di Segni 1997, 827–837; Negev 1981a, 48–55, 67), various stone carvings (Segal 1988; Golan 2020), marble furnishings (Maayan-Fanar and Tepper 2022b), and a mosaic floor with geometric patterns and a dedicatory inscription within the adjusted chapel dated to the end of the Byzantine period (607 CE, see Di Segni 1997, 839–840; Talgam et al. 2022, 235–237; for earlier date = 517 CE, see Negev 1981, 60–61; Tepper and Bar-Oz 2020). In addition to these are traces of wall paintings, including the scene of the Baptism of Christ, in the apse of the baptistery chapel (Maayan-Fanar et al. 2018). The dates on the tombstones within the Baptistery, 612–679 CE, suggest it was used well into the Early Islamic period (Baly 1935, 173, 175; Colt 1935, 10; Di Segni 1997, 842–849; Negev 1981, 55–59; Negev 1993). Small finds that include metal buckles indicate a possible connection to the equipment of soldiers (Bollók and Tepper 2021).

The North Church compound is full of numerous graffiti in different languages, most of which remain unpublished. One partially deciphered graffito, written on the north wall of the atrium and on the north portico of the narthex in Coptic Arabic, included the words ‘بسم الله’ [‘In the name of Allah’] as well as the name of the Prophet Muhammad (Moor 2013, 104–110). It was executed when the church was already abandoned, presumably in the mid-eighth century (Tepper 2019b).

On the right side of the northern entrance to the main hall of the North Church, just above average eye level (at a height of about 1.80–1.90 metres, roughly above the height of an adult), is a graffito of a horse rider turning towards the entrance upon a horse clad with armour (Figures 3–5).⁶ The graffito is partly overlapped by modern graffiti in Hebrew.

Figure 5. The façade of the northern entrance of the North Church from the narthex to the church hall and the location of a horse rider graffito (drawing by Avishay Blumenkrantz).

The graffito was scratched with long, shallow lines made by a sharp tool, probably a nail. The horse is simplified but somewhat realistic, with detailed mane and tail, its armour depicted by dense almost exhausting criss-cross lines, probably referring to iron scales. The horse rider is much less detailed than the horse. His body is a triangular shape, hands and legs indicated by single lines, and his head seemingly covered with a helmet. It is not clear whether he holds a spear. Additional figures (or a figure?) seem to be depicted in front of the horse rider. Although they are barely visible, at least one of them is a small standing figure with an oval body scratched in a single uneven line with a nail. Below might be another figure, similarly depicted.

While Shivta’s horseman is depicted very schematically, although seemingly wearing a helmet, the horse is depicted as heavily clad with armour, representing either cataphractarii or clibanarii (the distinction between the two terms is not always clear as they may be used interchangeably). Following the Parthians, armed horses were used regularly in the Roman army that was greatly developed in the Byzantine period (Gilbert 2000, 43). An armed cavalry was described in the Strategikon compiled by the emperor Maurice (582–602):

The horses, especially those of the officers and the other special troops, in particular those in the front ranks of the battle line should have protective pieces of iron armour about their heads and breast plates of iron or felt, or else breast and neck coverings such as the Avars. (Anderson 2017)

 

Comparable graffiti were observed and documented by Colt in Nessana (1962, Pl. XXVII: 7, 11–14). They include horse and camel riders armed with spears, swords, or bows. The problem with the Nessana graffiti is that their precise placement within the space was not documented. Therefore, their spatial context remains unclear, although they most probably belong to the period before Nessana fell and dispersed.⁷ In one case, a chalk drawing of a man wearing military dress (?) with a big cross on his chest is depicted standing next to a horse. This graffito was discovered somewhere in the Nessana’s North Church compound (Colt 1962, fig. 85). According to Colt, there are two depictions of clibanarii at Nessana (1962, Pl. XXVII:12, 13). One of these horse riders (Pl. XXVII:12), despite being schematic and crude, closely resembles the discussed graffito in Shivta’s North Church, especially regarding the curve of horse’s back and the overall depiction of a rider, including a curious head shape. Interestingly, both in Shivta and in Nessana, graffiti executors paid more attention to horses than to riders.

The significant number of graffiti of horse and camel riders in Nessana is not surprising. The so-called ‘soldiers’ archive’, dated from c500 to 700, contains a variety of documents associated with a military presence in Nessana. Thus, camels and dromedaries used for military needs (a camel unit?) are mentioned in the sixth-century papyri (P. Ness. 35 and P. Ness. 37; Kraemer 1958). It remains debatable whether during this period a unit was stationed there, or, as proposed by Whately (2016), the soldiers were ‘deployed and dispatched elsewhere’. Some perhaps were sent to Caesarea, others to Egypt; nevertheless, they constituted an important part of Nessana’s social and religious life that also was involved in the pilgrimage business (Whately 2016, 131–134; see also Di Segni 2004). The camel industry flourished and in the sixth and seventh centuries they were used for trade and transportation (P. Ness. 74, 89, Kraemer 1958, 209–210). But, as evinced by graffiti depicting armed camel riders (Colt 1962, Pl. XXVII:11), they also may have been related to military duties. Thus, graffiti of horse and camel riders may constitute personal signatures of such soldiers, echoing official written documents.

Another comparable graffito example comes from Oboda/Avdat, depicting a warrior with a triangle shaped body who is holding a spear (?) and riding a horse clad in armour. This graffito was engraved on a plastered wall, three metres above the floor in one of the rooms, perhaps a storeroom, of the Late Roman tower (Negev 1997, 68). Although the tower is dated to 293/294 CE, the graffito stylistically resembles those in Nessana and Shivta.⁸ Another graffito of a horse rider holding a spear was documented in Mampsis, ‘engraved on the eastern side of the pier of the staircase-tower in Locus 408’. In this case, there is no armour on the horse (Negev 1988, 116, fig 7).

Comparisons of graffiti of horse riders are plentiful, especially within forts and places with military presence (Langner 2001, Pls. 85–87).⁹ Thus, significant numbers of such depictions come from Dura Europos (Langner 2001, Pls. 35: 1317–1319; 86, 1320–1327; 87, 1334). They definitely represent warriors holding weaponry and at times clad in armour and can be explained by the extensive military presence and references left by Roman soldiers all over the city (Baird 2011, 56–57). The density of military graffiti on the Palmyrene gate was explained by Baird as a ‘long time presence of bored soldier at this particular spot’ (2011, 66). Most graffiti of clibanarii from Dura Europos represented en face and clad with characteristic dress refer to Parthian warriors depicted in attack (Kubik 2015). Graffiti of horse riders remain popular and are depicted, for example, on the walls of Ghazali North Church, Sudan. In this case they are interpreted as warrior saints, or Magi, or as domestic horse keepers (Obłuski and Maczuga 2021).

The similarity between the horse riders in Shivta and Nessana suggests they may have been executed approximately close in time. Their detailing differs from the Dura Europos Parthian riders, or the Sassanian/Iranian cavalry reflected in pre-Islamic rock petroglyphs in Wadi ‘Aday, Oman (Nicolle 2017, fig. 1a–e), or the images of riders accompanying seventh and eighth-century Arabic inscriptions from the Negev (e.g. SH 3100, HS 3172, RB 3173 in Nevo, Cohen, and Heftman 1993).

A figure at the gate

Another figurative graffito in Shivta, a figure with raised hands, probably created with a nail, easily can be missed as it is scratched on one of the low stones to the right of the entrance to one of the largest domestic buildings at Shivta (the ‘Governor’s House’) adjacent to the Central Church (Table 1, no. 11; Figures 6–8).

Figure 6. The praying figure graffito, the ‘Governor’s House’, Shivta (photograph by Dror Maayan; drawing by Emma Maayan).

Figure 7. The façade of the entrance to the ‘Governor’s House’ and the location of the praying figure graffito (photograph by Dror Maayan).

Figure 8. The façade of the entrance to the ‘Governor’s House’ and the location of the praying figure graffito (drawing by Avishay Blumenkrantz).

The ‘Governor’s House’ (Hirschfeld 2003, building 21) is the most western of three buildings; along with the Central Church, they constitute a compound (an area of approximately 660 square metres) located next to a central route of the settlement, between the North and South Churches. The domestic building was partially excavated by the Colt expedition (in 1936); regrettably, the results remain unpublished (Tepper 2019a). Segal, who discussed the structures of this building in detail, claims that there were initially four residential buildings erected in the fourth century that were annexed in the seventh century for the sake of the church since the north-western structure was destroyed. Concurrently, the three remaining buildings were altered, connected by doorways made in existing walls, and two rooms from the western building were eliminated to create a raised platform for a convenient passage from the building to the main hall of the church. Segal also suggested that the Central Church was used as a public church for the villagers, unlike the South and North Churches that could have been part of the monastic complexes (but see Figueras 1995, 435–442, who asserts that all three churches were part of the monasteries), while the buildings next to it functioned as an administrative centre of the settlement (Segal 1986, 29–50). Conversely, Shereshevski (1991, 77) asserted that the residential buildings, after being combined into one large building, were used for the needs of the church, like the other two churches.

Be that as it may, the main entrance to the western building (340 square metres) of the compound was from the south, from the northern part of a public square, in front of the ‘Governor’s House’. The entrance is surrounded by rooms that extend beyond the area of the square, emphasizing its architectural importance. Entrance to the building is through a covered corridor and from there to a trapezoidal (irregular) courtyard surrounded by living rooms, warehouses, and a stable (see discussion in the following). The plan of the building reveals several stages; the stable seems to have disappeared at a later stage (Segal 1986, figs. 23, 25).

The aforementioned graffito is engraved to the right of the main entrance to the building, a bit too low, at a height of approximately 20 centimetres above ground level, facing the south public square. It depicts a male figure, standing in frontal posture with raised hands depicted only by two lines, seemingly divided at their ends into three short lines, inside a doorway. Although not entirely clear, the impression is that the figure stands on a kind of raised platform flanked by two columns. The figure is crude and simple, not as detailed as the horse rider (see earlier), and seemingly of lesser quality. An attempt had been made to introduce some facial details (now lost) and certain dress features, such as a short-sleeved dress that narrows around the waist and widens downwards, just covering the knees. His raised hands recall an orans posture of a prayer; however, it is not entirely clear whether the hands are in fact raised or holding the doorposts. In addition, it seems that the left hand turns into a flourishing branch (cf. Langner 2001, Pl. 47, 935–936). A simple cross was scratched above the figure’s head, indicating that the figure is of Christian identity and perhaps also referring to some saintly figure. Another, deeply carved small cross is visible on a stone located two stones above.

Frontal figures with raised arms in prayer, or orans posture, have a long history and are known since prehistoric times. In the Levant they were introduced in the Early Bronze Age and are documented in rock art in the Central Negev that could be used there for a long span of time. Orans are interpreted within ritual connections, in a funerary context as a plea for self-salvation or expression of faith (Eisenberg Degen 2012, 262–264, 327–329). Orans posture is very popular in Early Christian and Byzantine art, but the closest iconographical examples come from tombstones from Egypt, all dated to the sixth to eighth centuries. The latter figures, some with similar dress, are carved, sometimes within an arch or just between two columns (e.g. Coptic Museum, Cairo, no. 8690, 8689, 8695; Index of Christian Art, sys.no. 87922, 87921, 88011; Metropolitan Museum of Art New York, no. 10.176.40; Index of Christian Art, sys.no. 89256). Crosses may be depicted above or in proximity to the figure (e.g. Stele of Semeuga from Akhmim, Coptic Museum, Cairo; Stele of Thekla, Coptic Museum, Cairo, no. 8693; Index of Christian Art sys. no. 88238, 88445). In some cases, the figures hold branches in one or both hands comparable to Shivta’s example (e.g. Funerary portrait, stele Staatliche Museen, Berlin, no. 6602; Stele of Maria, Staatliche Museen, Berlin, no. 4478; Stele of Tsempsuros, Coptic Museum, Cairo, no. 8579). This particular motif also brings to mind a graffito of a figure with raised hands holding branches that seem to grow from within its palms, found at a water reservoir at Khirbat el-Waziya, in western Galilee. Only the upper part of the figure is visible today. It was interpreted as a saint protecting the cistern (Klein, Zissu, and Distelfeld 2018, 407–413).

Another comparison comes from graffiti in the Coptic monastery of Phoebamonn near Thebes that functioned from the fourth to the eighth century. There, a monk and a saintly figure are depicted with hands raised in orans posture, with a cross above the head of one of them (Bachatly 1951). Simple schematic figures with raised hands and crosses instead of facial features are likewise found in a graffito by a female monk, Metredora, at Abydos (Westerfeld 2017, 203–204). The motif is also attested in Early Islamic graffiti, one from al-Murakkab in the Saudi Arabia dated to 678 CE, presumably referring to wuquf, an early pre-Islamic prayer (Imbert 2021, fig. 11.6). In this and other representations, the hands are proportionally much bigger than the head, thus being the main motif.

A graffito of a figure with an emphasized triangular head and some facial treatment was documented in the Negev in Mampsis, although it is not in orans posture (Negev 1988, 115, 256, fig. 17). Perhaps the best comparisons come from the Oboda/Avdat burial cave that was transformed into dwellings, where dipinti of two saintly warriors, vanished since their discovery, were depicted with crosses above their heads (‘Cave of the Saints’, Room 14; Jaussen, Savignac, and Vincent 1905, 77–80; Bucking 2017, fig 6b). Both are depicted frontally, dressed in short, belted tunics. The upper one has hands raised upwards and holds a big cross in his right hand, while the lower one holds a shield in his left hand, a spear in his raised right hand piercing a snake depicted just below him, and holds a small sized horse by a bridle. Because of the attributes, the latter was identified as St George. The upper figure might be St Theodore, since this name possibly appears in the accompanied inscription, perhaps with the name George (Jaussen, Savignac, and Vincent 1905, 77–80). A figure very similar to that of St Theodore, with a variety of symbols including crosses, was documented in a recently discovered ‘Dipinti Cave’ (Bucking 2017; Bucking and Erickson-Gini 2020; Erickson-Gini 2022a). Certain features in the Shivta figure’s dress allude to that of St George in Avdat (although less detailed), perhaps hinting to a possible depiction of a saint warrior, especially taking in account a cross above his head and another one nearby.¹⁰

Conclusive remarks: contribution of graffiti to understanding Byzantine Shivta

Both graffiti of a horse rider and a figure at a gate, made by amateur unprofessional hands and in different styles, are depicted next to the entrances, thus raising some spatial awareness.

The horse rider carved at the northern entrance to the North Church main hall faces the entrance and was scratched slightly above eye level, and is therefore publicly visible, ‘on display’. The person presumably stood on the floor to make the graffiti, investing time in the horse armour by cross-hatching. The graffito confirms characteristics of riders in the Byzantine period. In Palestine of the fifth century, about 21,500 soldiers served under the Dux; Notitia Dignitatum mentioned thousands of horsemen from the Dalmatian Illyrian cavalry units stationed in Beer Sheva (Notitia Dignitatum 34, 5, 18; Seeck 1876, 72–73; also Eusebius, Onomasticon; Notley and Safrai 2005), the Illyrian cavalry in Ma’on and Carmel, the Illyrian Thamudeni cavalry in Beer Shema, and other places in the Arava and Transjordan (Shatzman 2021, 185–188). Although Notitia Dignitatum did not mention a cavalry unit stationed in the Negev settlements, such a possibility cannot be excluded. Archaeological findings, such as the buckles mentioned earlier (Bollók and Tepper 2021), a metal seal in Latin (Eck and Tepper 2021), and a lead dress ornament with an eagle dated to the Byzantine period (Maayan-Fanar, Tepper, and Asscher 2023) found above the floor level of building 86 (Tepper and Bar-Oz 2020), hint at a certain military presence, perhaps of limitanei, Byzantine army veterans, in Shivta.

In addition, the North Church might be directly connected to a certain John the vicarius whose name appears in three inscriptions: the dedicatory floor mosaic in the side chapel dated to 606 CE, and on two tombstones most probably of his sons, one in the Baptistery dated to 612 and another in the atrium dated to 646. Vicarius, an official, a military officer whose supreme commander was the Dux of Palestine (Shatzman 2021, 173 note 287), was mentioned in the fifth-century imperial decree from Be’er Sheva (Di Segni 2004; and an extensive discussion therein including additional references), and in the sixth-century military archive from Nessana (P. Ness. 39; Kraemer 1958, 122–124).

Shatzman, who discussed the limitanei in detail, suggested that the vicars were also used as the commanders of the limitanei, soldiers of the regular army units, starting from the reign of Diocletian. Engaged in a variety of military activities during the fourth to early seventh centuries, they ceased to exist most likely during the Persian conquest of the Land of Israel (614 CE; Shatzman 2021, 59–70, 189–190). If so, the inscriptions from Shivta that mention the vicarius, and are dated before and after the Persian conquest, could refer to the presence of a military officer in the settlement.

In the context of the aforementioned, it is noteworthy to mention the archaeological evidence about stables and horse breeding that have a long history in the Negev settlements dating back to the Nabatean period (Negev 1983). In a study of the stable facilities in the Roman and Byzantine periods, Tepper emphasized the uniqueness of the building and its architectural form in the training, care, and maintenance of riding horses in Negev settlements including Avdat, Mamshit, Rehovot, and Shivta (1997, 250; also; Erickson-Gini 2022b). Four stables have been recorded so far in Shivta: the stable house with at least six mangers in a separate wing in a building that dates back to the fourth century and was used throughout the Byzantine period (Figure 2:a; Segal 1983, 87-97; Segal 1986, 63–76); the western house in the ‘Governor’s House’ building (aforementioned) where at least three mangers were recorded in a separate wing of a similar date and are associated with the early phase of the building’s history (Figure 2:b; Segal 1986, 42–45); in the ‘Pool House’, a stable room with two mangers used during its Byzantine phase that has been identified recently (Figure 2:c; Tepper and Bar-Oz 2020); and a single manger in the courtyard of another structure that was not excavated in its entirety (Figure 2:d; Hirschfeld 2003, building 144). The architectural similarity between the facilities allows us to assume that they were all used for the cultivation and breeding of riding horses.

In light of the aforementioned, it is possible to suggest the presence of soldiers, most likely limitanei, in Shivta. Some of them could have served as horsemen or at least had the means to maintain, care for, and train horses. It is also reasonable to assume that the vicarious mentioned in the inscriptions in the North Church, a person of a high status and probably one of the wealthiest in Shivta, was one of the commanders of the limitanei. Could he be the horse rider engraved on the doorpost at the entrance to the church? If so, it is another example of how graffiti can help make sense of archaeological evidence.

If the horseman might be connected to an actual military presence in Shivta, the man on the doorpost of the ‘Governor’s House’ building seems related to a religious context, depicting a saintly figure, perhaps a saintly warrior. The graffito was drawn very low on the wall, just above ground level. Someone actually sat on the ground to scratch the figure with a cross above his head, perhaps waiting for the doors to open. It is a personal graffito, not meant to be seen. The creator, or perhaps someone else, added another carved cross (Table 1, no. 12), possibly referring to the figure or to the place as holy and stressing its Christian significance. Was it drawn by a resident of Shivta, a monk, a guardian, or a pilgrim? We will never know.

A new study on dipinti in cave dwellings in Avdat, where the best parallels to Shivta’s graffiti were found, suggests they are evidence of the existence of the monastic community and the cult of St Theodore (Bucking 2017, 32–33; Erickson-Gini 2022a). One of the motifs from the Avdat ‘Dipinti Cave’ is an intriguing parallel with those in fifth-century north Italy and might be considered as evidence of a pilgrimage (Bucking and Erickson-Gini 2020, 33–35). The ‘Dipinti Cave’ was destroyed during the seventh-century earthquake, providing terminus ante quem to the graffiti. Consequently, the graffiti from Shivta of a man praying with his hands raised in the orans position and a cross might hint at the function of the building at a later stage of its history; perhaps it was a residence of those who served in the nearby church, in accordance with Sherashevsky’s proposal and different from the approach suggested by Segal (see earlier).

The two pictorial graffiti published here for the first time are very personal expressions bearing different messages. With inscribed graffiti, many as yet unpublished, they constitute seldomly seen relics from the times when Shivta was still teeming with life, reflecting two main features of this desert settlement – religious and military – that seem to exist in harmony. Numerous later graffiti testify that Shivta was not forgotten and continued to be visited long after it was abandoned and eventually fell into ruins. Nomads, travellers, and pilgrims left numerous inscriptions and signs, at times overlapping each other, and tried to fit them into small spaces on accessible stones, thereby turning them into memoria, storytelling, and even a dialogue between religions and cultures. Modern visitors continued to add graffiti on the same walls, sometimes overlapping ancient ones. Some of them, like Halfdan’s graffiti in the Hagia Sophia, have already become an inseparable part of the site’s long history. Studies of graffiti add new information, thereby contributing to the understanding of archaeological sites and historic buildings. They are an additional, rich, and largely unexploited source of information.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. For the most recent research of the Crusader’s graffiti in Jerusalem, see Zitron, Baruch, and Shotten-Hatlel (2022) and Tashernin et al. (2022); and in Ramla, see Dussart (2023).

2. The excavations (2016 and 2019 seasons) were conducted by Y. Tepper and G. Bar-Oz on behalf of the Zinman Institute of Archaeology of the University of Haifa. Research of the first season was conducted under licenses from the Israel Antiquities Authority (G-87/2015, G-4/2016). The second season (license nos. G-82/2018, G-5/2019) was funded by the Israel Nature and Parks Authority (Rakefet Foundation). The excavation of the southern reservoir (2019) was conducted with A. Levi-Harvoni on behalf of the Israel Antiquities Authority (license No. A-8552/2019).

Photographs were taken by Yotam Tepper and by Dror Maayan. Thanks also to Sapir Hadd (maps) and Avishay Blumenkrantz (field drawings).

3. It is not always easy to document graffiti, especially because of their obscure placement and exposure to strong direct sunlight of the desert. Therefore, all the graffiti were photographed from carefully calculated angles for the best visibility of the motifs.

4. In a survey conducted by Tsuk, only 57 water cisterns were examined, of which 40 were located in the courtyards of residential houses, six in churches, and the rest in open spaces or in close proximity to the settlement. Crosses, made of shells of the Na’aman type embedded in the mud, were discovered on the walls of three private cisterns (nos. 24, 30, 43) within the village, while in cistern 43 a cross is depicted next to the Greek inscription mentioning the name of John (son of) Kyriakos (Tsuk 2002).

5. Several vertical lines in the narthex of the South Church were proposed to be identified by Avishai Levi-Hevroni as a ship, but dismissed during a re-examination conducted as a part of this study.

6. Incised graffiti of another horse rider, this time without horse’s armour, can be detected on one of the stones of the so-called Colt’s house (Table 1, no. 9), built by Colt’s expedition from stones found within Shivta. Introduced upside-down, its original location remains unknown. The image of the horse seems to be simpler and more linear than that from the North Church; the rider is barely visible and cannot be reconstructed in full.

7. Colt (1962, 61) explicitly stated that he documented only the graffiti from the excavations and not those found on the surface.

8. Negev thought that this graffito is contemporary with the tower, pointing out that the motif is similar to Nabatean rock engravings in Sinai, but also acknowledging that it differs from them stylistically and is in fact ‘unique’ (Negev 1983).

9. In a survey of the rock art in the Negev heights a connection was proposed of the findings to the ancient road network in the context of desert trade and pilgrim routes, in tribal rock art, and in Islamic art. The rich spectrum of motifs partially correlates with those found in Shivta and other Negev settlements including geometrical motifs, human figures, especially orans, horse and camel riders, and more. Some of them are datable to the Byzantine period and can be identified as Christian with certainty (such as crosses of various shapes); others could predate or postdate them (for chronology suggested for Central Negev, see Eisenberg Degen 2012). The petroglyphs are securely dated to the Islamic Period and include depictions of horses, camels, and human figures. Those seem to be much more abstract and differ from depictions within the settlements. Alongside tribal signs and motifs, multi-meaningful and diverse, it is difficult, if not impossible, to decipher them. Some of the motifs found along Avdat–Sinai road could have been left by pilgrims to the Holy Monastery in Sinai and can be understood within a specific cultural-religious background (Schwimer and Bucking 2021; see also Negev 1971a, 1971b; Negev 1981, 66–71).

10. Graffiti from Avdat and the figure from Shivta also recall, albeit distantly, a graffito from Nazareth dated to c340–447, depicting a Christian soldier clad in a helmet, armour, and a short tunic, and holding a cross insignia and a shield (Taylor 1987).