Ein Gedi Ancient Synagogue

The antiquities have their own entrance off Route 90, between the Nahal David and Nahal Arugot gates of En Gedi Nature Reserve, and the parks authority treats them as a separate transaction. Its English page says no reservation is necessary to visit the En Gedi Antiquities site, and that a visitor coming only for the antiquities should buy a ticket at that entrance for that area alone. The site keeps its own telephone line, 08-9170586, apart from the two canyon offices.

What waits inside the fence is compact. The parks authority lists an entrance plaza, a visitor service station, toilets, shaded picnic corners, directional and explanatory signage, observation points over the ancient village, and a concrete path that runs from the entrance all the way to the ancient synagogue and into the building itself. Parking sits beside the gate, outside the fence.

What survives is a ruin held in place rather than a rebuilt structure. The walls of the prayer hall stand low; the floor is the reason people come; imprints of burnt roof beams are still visible on it. Almost everything a visitor walks on or reads is modern: the concrete path, the compacted-aggregate paths through the village, the handrails fitted to its steps, the signage, and a scale model of the ancient village that the authority provides as an alternative for people who cannot get round the ruins.

The village around the synagogue is not the synagogue dig. Gideon Hadas ran a series of salvage excavations for the Israel Antiquities Authority from the 1980s and prepared the Roman and Byzantine village for public presentation between 1993 and 1995, exposing three strata: a Roman one probably destroyed at the end of the Bar Kokhba revolt, a Byzantine one destroyed in the sixth century, and a Mamluk one taken apart when the ground was cleared for modern farming.

The building grew in stages. It began as a simple hall with a geometric mosaic floor and was rebuilt over time into a basilica plan with a bema and an ark, colour in the mosaic, and the aisles that carried the inscriptions. The parks authority's brochure puts it plainly: a synagogue was erected in the middle of the village in the third century CE and with the passage of years it was modified, enlarged and renovated, until in its last period it was a large and handsome building that reflected how well the villagers were doing.

The decoration belongs to the later phases. The Jewish Virtual Library's account gives a mid-fifth-century renovation with a new coloured mosaic: a pattern of four-petalled flowers, four birds at the centre, pairs of peacocks at the corners, and three seven-branched menorahs near the bema. The birds have been identified as crowned cranes alongside francolins or partridges, which is a reading rather than a caption written by the people who laid the floor.

One object from the building has never been matched elsewhere. A seven-branched menorah was found here, of a type unique for its period. Accounts differ on what it was made of: the Jewish Virtual Library describes a candelabrum about 30 centimetres high in bronze, while the archaeological summary of the site calls it metal and probably silver. The excavators also recorded graffiti of ships scratched into plastered walls and pillars, which may have been decorative or votive, and nobody has settled which.

• First hall: simple, with a geometric mosaic floor
• Later phases: basilica plan with a bema and a Torah ark
• Mid-fifth-century floor: four-petalled flowers, four birds, pairs of peacocks
• Three seven-branched menorahs in the mosaic near the bema
• A seven-branched menorah found here, about 30 cm high, of a type unique for its time

The western aisle of the prayer hall carried a mosaic inscription that is among the longest found in any ancient synagogue. It works through several registers: a genealogy running from Adam to Japheth, a list of the months with the signs of the zodiac, and dedications naming the donors. Among the names are Halfi and his three sons, Yosi, Ezrin and Hezkin, a Rabbi Yosi, and Yonathan the cantor.

Then it turns. The parks authority's brochure describes the passage as one that threatens and curses anyone who would cause controversy or libel, or reveal "the secrets of the city" to strangers. The Jewish Virtual Library renders the sanction itself as "The one whose eyes roam over the entire earth and sees what is concealed will uproot this person and his seed from under the sun". A curse against informing, set in the floor of a village prayer hall, is not something the rest of the corpus prepares a reader for.

What the secret was is genuinely open. The parks authority's brochure offers the answer most researchers give: the cultivation and processing of balsam, the source of the region's wealth, a trade so valuable that Marc Antony took the Jericho and Ein Gedi groves for Cleopatra and Herod then leased them back from her. Jodi Magness reads the clause differently, setting it against a similar phrase in the Damascus Document from nearby Qumran and in the Cairo Geniza, and arguing that what is being punished is betraying one's own Jewish community to gentiles rather than leaking a recipe.

The panel is not on site. The mosaic inscription is displayed at the National Campus for the Archaeology of Israel in Jerusalem, on the roof plaza that opened to the public while the building below it was still being finished. The operator's page does not state which parts of the floor at Ein Gedi are the original tesserae and which are made good, so a visitor should not assume either way.

The excavation of 1970 was headed by Dan Barag and Ehud Netzer of the Institute of Archaeology at the Hebrew University of Jerusalem with Yosef Porath, and it opened the synagogue's Torah ark. Inside were multiple charred lumps of what looked like parchment. Porath lifted the scroll on 5 May 1970. Its structure had been burned and crushed into chunks of charcoal that disintegrated further every time they were touched, and with no viable conservation protocol the Israel Antiquities Authority simply shelved it.

The ark held more than the scroll. Alongside it were silver pendants, oil lamps, coins, fragments of glass, and 132 animal bones from kosher species, mostly goat and sheep but also chicken and fish. The bones carry butchery marks and may be the remains of religious meals; why they were inside the ark has never been explained, though storage as a genizah, because they were treated as holy, is one suggestion.

The reading came four and a half decades later and from computer science rather than conservation. Seales and his colleagues scanned the object with x-ray micro-computed tomography at a spatial resolution of 18 micrometres per voxel, then segmented, textured and flattened the layers in software to produce a single flat image of a surface that will almost certainly never be physically opened. Their paper appeared in Science Advances in 2016. Ink showed up because it absorbs x-rays more strongly than the skin around it, which implies it contains a metal such as iron or lead, though nobody can sample it to find out.

What emerged is the start of a scroll: a large blank margin, then two columns of the book of Leviticus. One sheet of 35 lines was recovered, 18 of them preserved and 17 reconstructed, at 33 to 34 letters and spaces per line. The consonantal text is identical to the medieval Masoretic Text, which it precedes by centuries, and it is the earliest copy of a book of the Pentateuch ever found in a Torah ark. The dating is argued: radiocarbon at the Weizmann Institute gave an age of 1754 years plus or minus 40, calibrating to 210 to 390 CE at 89 per cent probability, while Ada Yardeni proposed a date in the late first or early second century on the shape of the letters, and Drew Longacre defended the radiocarbon reading against her.

Everyone agrees the village and its synagogue burned. Nobody agrees when, or who did it. The archaeologists who dug the synagogue attributed the destruction to Byzantine persecution of Jews under Justinian; the parks authority's brochure puts the fire around the middle of the sixth century CE; the Jewish Virtual Library places it in the second half of the sixth century, still in Justinian's reign; others argue for a Bedouin raid before the Persian invasion, late in the sixth century or early in the seventh; and the team that read the scroll works with a fire of about 600 CE. Anyone quoting a single confident date for the end of Ein Gedi is quoting one camp.

What followed is less disputed. Balsam cultivation around the Dead Sea stopped when the village emptied, and the knowledge of how to grow and process the plant left with the refugees and was never recovered. If the inscription really was protecting a trade secret, the curse outlived the secret.

The floor came back to notice by accident. The parks authority's brochure says remnants of the synagogue's mosaic were found in 1965 while a field was being ploughed; the archaeological summary of the site gives the first discovery as 1966. Systematic excavation of the synagogue and the buildings beside it ran from 1970 to 1972 under Dan Barag of the Hebrew University and Yosef Porath, then of the Israel Department of Antiquities.

The most recent damage is not ancient. In May 2025 a flash flood dropped about 35 millimetres of rain on the reserve in a very short period, which the parks authority called an extreme event unlike anything seen there for many years. Trails were washed out and the ancient synagogue was flooded and will need preservation and restoration work. The antiquities site has since been listed by the operator as open as usual while much of the rest of the reserve stays shut.