Shivta National Park

Shivta has no gate in the defensive sense, because it has no wall. The parks authority is explicit that the town was never surrounded by one and that no fortifications have been found; what looks like a perimeter is simply the outermost houses built in a continuous line. The opening where the street begins on the western side is called the Western Gate by convention, not by engineering.

The first building on the walk is not ancient. South of the car park stands the Colt expedition house, put up in the 1930s by the dig from New York University and the British School of Archaeology in Jerusalem, and used today for accommodation and desert hospitality. From there the main street runs the length of the settlement, from the Northern Church past the Central Church to the Southern Church, with the Pools Square, the Governor's House and the Northern Square strung along it.

The streets are crooked and the plan looks improvised, which the parks authority's brochure concedes while pointing out that one axis does run the whole length of the town. Shivta reads better as a set of courtyard houses that grew into each other than as a piece of town planning. Churches were inserted into blocks of housing rather than given precincts of their own: the Central Church has residences attached on the east and south, and its three doors open straight onto the street.

• The Colt expedition house, 1930s, south of the car park
• The Western Gate, which is a street opening rather than a gate
• The Pools Square, with two public reservoirs, only one of them excavated
• The Governor's House, a private dwelling with a tower still standing to almost 6 m
• The Northern Square, with a winepress and a benched hall that may have been a council house
• The Vaulted House, where stone slabs still rest on complete stone arches

The University of Haifa team that has worked here since 2015 dates the foundation to the beginning of the first century AD, and the parks authority's brochure agrees that the first settlers were Nabataeans, the Arab tribe whose caravans carried frankincense and myrrh from southern Arabia towards Gaza. The trade that paid for the early settlement dried up: Roman shipping reached India through Egypt, the incense route lost its monopoly, and by the fourth century the traffic had stopped. What replaced it was farming, and Shivta grew rather than shrank.

The parks authority's brochure puts the Byzantine town at more than 2,000 people, living off agriculture and off the pilgrims who came to venerate saints buried in its churches. Two documents place it in a wider world. A fifth-century collection known as the Stories of Nilus the Saint has the author's son Theodoulos carried off in southern Sinai and offered for sale in a village called Sobata, the Greek name for this place. Papyri found at nearby Nessana list 30 donors from across the Negev who paid for the monastery of Saint Sergius there, nine of them men of Shivta, and a second papyrus copied to Shivta at the end of the seventh century concerns a delegation of Negev landowners organising a protest against the taxes of the new Muslim government.

What the name means is not settled. The parks authority's brochure says the sense of Shivta is unclear and suggests that the Greek Sobota preserves a Nabataean personal name, Shubaytu, in the way that Avdat and Haluza do. The modern Hebrew form was assigned by the Negev Naming Committee in the early 1950s; the Arabic name of the ruins was Subeita.

UNESCO inscribed Shivta in 2005 under cultural criteria (iii) and (v), in a property of some 6,655 hectares that also takes in Nabataean fortresses and caravanserais. The citation is about the desert cities as a group and the irrigation that made them possible, not about any single building here.

The public water system is the most legible piece of engineering on the site and it is right in the middle of the town. The Pools Square holds two large reservoirs, of which only one has been excavated. Water reached them along an aqueduct from a point 2.5 km away and then through channels that are still visible on the ground. The parks authority draws the distinction that matters socially: the cisterns under the house courtyards were private property, the pools in the square were for everybody.

The houses themselves were built to hold what fell on them. The parks authority describes walls about 70 cm thick, rough hard limestone below to a height of roughly 2 m and finely worked soft limestone above, which insulated well and shed water into the cistern in the courtyard corner. The average house here, courtyard included, is around 360 m².

Outside the town the same principle was applied at scale. Runoff from bare hillsides was steered onto terraced plots, and a 2025 study in PLoS One modelling Shivta's wine economy notes that ancient Negev runoff farms typically worked a catchment to cropland ratio of between 1:9 and 1:33, averaging about 1:20: a field of vines needed roughly twenty times its own area of hillside feeding it. Two winepresses have been excavated, the larger with a paved treading floor and separate bays where growers left their grapes and waited their turn.

The reconstructed orchard 700 m north of the Northern Church is a twentieth-century experiment rather than an antiquity. A team under the botanist Michael Evenari cleared and rebuilt an ancient terrace system in the 1950s, work that the parks authority reckons took about 2,000 man-days, and planted the first trees in the rainy season of 1958 to 1959. Carob, fig, almond, plum, olive, pomegranate, pistachio, peach, apricot and vines went in, none of them irrigated. The carob did best.

Shivta is unusual in how little of it needed rebuilding, which makes the exceptions worth knowing. The Northern Church is the largest of the three and its walls stand to about 10 m, held from outside by diagonal buttresses. Its prayer hall is entered through a square atrium ringed by two-storey rooms that were probably a monastery, and a side door in the southern aisle leads to a baptism chapel where the font, a large cross-shaped basin, is carved out of a single block of stone.

The Southern Church was built, apparently, over a Nabataean cult building. It began with two square rooms flanking the apse, in the manner of fourth-century Negev churches, and had them replaced at the start of the sixth century by apses with reliquary niches. A large baptistery was added on its north side.

The restorations are localised and the parks authority names them. The decorated entrance lintel of the Governor's House was found elsewhere in Shivta and set up here, while the stone-slab roof over its ground floor is original. Almost everything else a visitor walks past is what was left standing when the town emptied.

Two caveats. Excavation has never been complete: large parts of the town are still unexcavated and only one of the two public pools has been opened. And the parks authority currently has a notice on its Hebrew page saying that accessibility works have changed both the entrance arrangement and the walking paths, and asking visitors to follow the signs on site.

• Original: the standing house walls, the vaulted and slab-roofed ceilings, the church walls, the cross-shaped font, the cisterns and the channels from the aqueduct
• Restored or moved: the Governor's House entrance lintel, which was found elsewhere on the site
• Wholly modern: the reconstructed orchard north of the town, laid out in the 1950s
• Never dug: much of the residential town, and the second of the two public reservoirs

In 2018 Emma Maayan-Fanar, Ravit Linn, Yotam Tepper and Guy Bar-Oz published in Antiquity a previously unknown painting of Christ's face at Shivta, in the baptistery beside the Northern Church, above the cross-shaped font. They read it as a baptism of Christ, with a youthful, short-haired Jesus and a larger haloed John the Baptist, and dated it to the sixth century. Their claim for it is precise and worth quoting accurately: it is the only in situ baptism-of-Christ scene that can be dated with confidence to the Holy Land before iconoclasm.

The Christ is short-haired and beardless, a convention the authors describe as widespread in Egypt and Syria-Palestine and then gone from later Byzantine art. It is also nearly invisible. Maayan-Fanar told the Times of Israel that she was under the apse at the right place at the right time, with the light just right. Visitors looking for a picture on a wall will not find one.

The second painting is in the Southern Church and has been known for over a century. Leonard Woolley and T. E. Lawrence, surveying the Negev for the Palestine Exploration Fund in 1914, recorded a scene there that they read as the Transfiguration, with Peter and John kneeling at Christ's feet and other figures identified as Moses and Elijah. The parks authority's brochure says that what remains of those paintings today is fragments of figures in red. Maayan-Fanar and Tepper published a reconstruction of the scene in Biblical Archaeology Review in 2024, in which Peter's beard is the best preserved passage of brushwork.

The most quietly startling thing at Shivta is architectural rather than decorative. Yotam Tepper, Lior Weissbrod and Guy Bar-Oz, working for the University of Haifa's Byzantine Bio-Archaeology Research Program of the Negev, recorded 21 door openings that had been deliberately blocked, spread across 16 residential structures. The blockings vary from finely dressed masonry to unmodified stones piled in the opening, and they cluster in two distinct parts of the town, which the team reads as sections of the settlement being given up together rather than the whole place failing at once. People shut their houses on the way out.

When that happened, and why, is genuinely contested. Research led from the University of Haifa and reported by the Times of Israel puts the crisis in the mid-sixth century, ahead of the Islamic conquest altogether, and lines it up against the volcanic cooling of the 530s and 540s and the plague of 541 to 549. Daniel Fuks, who led that work, was frank about the limits of the evidence: he told the paper that the researchers do not know how to disentangle the plague, the social pressures and the climate change. The older story, that Muslim rule and a ban on wine killed a wine town, does not survive the dating.

Settlement did not stop at the conquest. A small mosque was built immediately north of the Southern Church's baptistery, two rows of three columns with a mihrab facing Mecca, and here the operator contradicts itself in print. The parks authority's English site page says a stone step at the mosque entrance carved with a cross appears to prove that no Christians were living at Shivta while the mosque was in use. Its own Hebrew brochure says the mosque's builders took care not to damage the baptismal font next door, and that church and mosque may be assumed to have functioned at the same time. Both statements are published by the same body about the same two rooms.

The endpoint is less disputed than the mechanism. The parks authority's brochure has residents leaving gradually and the town completely deserted by the end of the ninth century; the Haifa team describes a substantial decline through the Early Islamic period, in the eighth and ninth centuries. Nothing was built over it afterwards, which is the reason there is anything to see.