Qumran National Park

The site is reached from Route 90, turning west at Kibbutz Kalya and following the signs, as the parks authority's directions put it. Everything then happens in a compact order. The entrance building holds the ticket desk, the toilets and a counter selling cold drinks, ice cream, souvenirs, books, hats and sunscreen. Beyond it a ten-minute film on the history of the site runs in a screening hall, and past that a small museum sets out how the parks authority understands life here. The trail to the ruin starts from the museum, with seven QR posts along it.

The circuit itself is short and legible. The parks authority's Hebrew page walks it in this order: the water channel and the ritual baths, the assembly rooms, the central dining room where communal meals were eaten, the remains of the kitchen, the watchtower, the workshops where vessels were made, and the stable. The printed English site plan uses the labels de Vaux gave them: aqueduct entrance, ancient cistern, reservoir, tower, scriptorium, kitchen, assembly hall and refectory, pantry, pottery workshop, kiln, cattle pen.

What dominates the plan is water. The settlement stands where almost no rain falls, and it survived on winter floods caught behind a dam in Wadi Qumran and led in along an aqueduct to a chain of stepped, plastered cisterns. The parks authority draws attention to how many of those pools it identifies as miqva'ot, ritual baths, and to the concentration of them beside the dining room. Whether every stepped pool here is a ritual bath or simply a reservoir is a long-running argument, and Katharina Galor's suggestion that they served as both is where much of the literature has settled.

East of the observation shelter is the cemetery: about 1,100 graves, most of them men, dug as north-south trenches with the body on its back and the arms at its sides. Only a small share has ever been excavated. The caves are the other half of the view. Some are dark slots in the cliff above; others, including the double-mouthed Cave 4, are cut into the soft marl of the terrace almost within shouting distance of the walls. They are looked at from the park, not entered on the marked route.

The Israel Antiquities Authority's account is that a young Bedouin shepherd of the Ta'amireh tribe found the first scrolls in 1947 while looking for a stray animal. The parks authority's brochure gives the same year and says seven ancient scrolls came out of that cave. Archaeologists reached it only early in 1949, identified it as Cave 1, and found the remains of at least seventy manuscripts, including further pieces of the original seven: the Community Rule, the War Scroll, the Genesis Apocryphon and both Isaiah scrolls.

The years from 1951 to 1956 were a race. Roland de Vaux and a French team from the École Biblique excavated the settlement while an eight-kilometre strip of cliff was combed for more caves, and the Bedouin, for whom scraps of inscribed leather had become a serious income, combed it too. Cave 3 gave up the Copper Scroll, a two-part metal document listing what may be buried Temple treasure. Cave 4 was found by Bedouin in 1952, working, as the antiquities authority phrases it, right under the noses of the archaeologists digging the site; it alone accounts for 75 per cent of all the material recovered from the Qumran caves. Cave 11, found by Bedouin in 1956, produced the Temple Scroll and closed the sequence.

The object most people picture is the Great Isaiah Scroll from Cave 1: about 7.34 metres of parchment in 54 columns, placed by handwriting in the decades around 100 BCE. It is the oldest substantially complete copy of a biblical book, and it is in Jerusalem, not here.

The sequence may not be closed after all. In February 2017 a Hebrew University and Israel Antiquities Authority team under Oren Gutfeld and Ahiad Ovadia excavated a cave in the cliffs west of Qumran and found the whole apparatus of scroll storage with the scrolls gone: broken storage jars of the familiar type, a leather binding strap, wrapping cloth, and a blank piece of prepared parchment rolled inside a jug. Looters' pickaxe heads from the mid-twentieth century were still lying there. Gutfeld's summary was that there were certainly twelve caves and perhaps more.

There was a settlement on this shelf long before the scrolls. The parks authority dates a Jewish population here to the eighth century BCE, and the modest Iron Age remains, which included a royal seal impression, led de Vaux to identify Qumran with the City of Salt listed in the Book of Joshua. The building whose walls a visitor now walks between went up much later, in the late second century BCE, under John Hyrcanus I or Alexander Jannaeus.

De Vaux divided the occupation into periods, and the turning points are firm enough to date. An earthquake and fire ended the Hasmonean phase in 31 BCE. The parks authority's brochure says the community abandoned the site and returned a quarter of a century later, under Herod's son Archelaus, and rebuilt. The Romans took Qumran in 68 CE. The last occupants on the brochure's account were a Roman garrison here during the Bar Kokhba revolt of 132 to 135 CE, after which the place was abandoned and forgotten.

The plan de Vaux uncovered has two parts: a squarish main building of two storeys around a central courtyard, with a heavy defensive tower at its north-western corner, and a secondary building to the west. He read particular rooms as particular things, and those readings are what the park's signage inherits. Locus 77 became the refectory, on the strength of the bowls stacked in the pantry next door. Locus 30 became the scriptorium, because inkwells were found there with plaster furniture that had fallen through from an upper floor. Many of the stepped cisterns became ritual baths, by analogy with baths near the Temple Mount in Jerusalem.

The inkwells are the strongest single piece of evidence that writing happened here. De Vaux found three; a fourth, fifth and sixth have been identified since, the last by Yitzhak Magen and Yuval Peleg. That total is larger than at any other excavated site of the Second Temple period. The parks authority adds that hundreds of pottery lamps came out of the same rooms, and reads them as light to write and study by after dark.

Qumran is unusual among Israel's archaeological parks in how little of it has been put back up. There is no reconstructed storey, no rebuilt tower and no restored roof. What a visitor sees is a floor plan in low stone, with the water system as the one element still legible at close to full depth: the channel, the settling pools and the stepped cisterns, several of which retain their plaster. The most photographed detail on the circuit is the crack running down the steps of one large stepped pool, which excavation attributed to the earthquake.

The visitor infrastructure is modern and its date is known. Restoration work was carried out at Qumran in 1967 by R. W. Dajjani of Jordan's Department of Antiquities, and the same year, after the Six Day War, the site passed to Israel's parks authority, which by its own account then built the access road and car park, installed sanitary facilities and laid out the footpaths and signs.

One reconstruction is worth knowing about precisely because it is not on site. The plaster furniture from the room de Vaux called the scriptorium was reassembled for museum display, and the Orion Center at the Hebrew University notes that it was rebuilt in a shape that accommodated de Vaux's own reading of it. What the furniture was for has never been established, and the suggestion that the pieces were dining couches in the Greco-Roman manner runs into the objection that they are too narrow to recline on.

The largest missing piece is a document. De Vaux died in 1971 without publishing a final report. The École Biblique appointed Robert Donceel in 1986 to finish the job; preliminary findings were presented in New York in 1992 and the report never appeared, on Pauline Donceel-Voûte's account because too many finds had been lost or their records corrupted. A synthesis of de Vaux's field notes came out in French in 1994 and in English in 2003. The site everyone argues about has never had a complete excavation report from the man who dug it.

• Ancient and in place: the low walls of the main building and its western annexe, the tower footings, the aqueduct channel and settling pools, the plastered stepped cisterns, the pottery kilns and the cracked steps of the large stepped pool
• Restored: work carried out at the site in 1967 by R. W. Dajjani of Jordan's Department of Antiquities
• Added for visitors after 1967: the access road, the car park, the toilets, the paths and the signage, all installed by the Israel Nature and Parks Authority
• Reconstructed, and not here: the plaster furniture from the room called the scriptorium, reassembled for museum display in a form that matched de Vaux's reading of it
• Never delivered: de Vaux's final excavation report, unpublished at his death in 1971 and abandoned by his successor in the 1990s

The parks authority's brochure states the traditional case without hedging: the Essenes arrived at the end of the second century BCE, lived communally, bathed ritually, wrote in the scriptorium and hid their library in the caves. That is de Vaux's Qumran-Essene hypothesis, built by reading the settlement against the contents of the scrolls and against what Pliny, Philo and Josephus wrote about the Essenes. It remains the majority position: Britannica's phrasing is that most scholars think the site housed the community that owned the scrolls.

Each alternative starts from something the standard account handles awkwardly. Norman Golb argued that Qumran was a fortress throughout, that no sectarians ever lived here, and that the scrolls were a library carried out of Jerusalem and hidden in the desert. Pauline Donceel-Voûte proposed a Roman villa, on the strength of glass and stoneware from the dig that looked too fine for ascetics. Yizhar Hirschfeld accepted a Hasmonean fortress that became a fortified, agriculturally based trading station. Yitzhak Magen and Yuval Peleg, after ten seasons between 1993 and 2004, concluded that the water system was built to bring clay-laden water into the site and that Qumran was a pottery production plant, noting that a community of the size usually claimed would have left hundreds of ovens and thousands of cooking pots that are not there.

The finds cut both ways. De Vaux's excavations produced about 1,250 coins, including three silver hoards, which is a great deal of cash for a community of poor recluses. The inkwell count is larger than at any comparable site, which is hard to explain if nobody wrote here. Population estimates run from Hirschfeld's twenty in the Hasmonean period, through Joseph Patrich's fifty at most, to the roughly 150 to 200 accepted by Jodi Magness on the strength of the dining dishes stacked in the pantry.

Robert Cargill's survey of the fortress readings proposes a Hasmonean fort later reoccupied and extended by the sectarians who owned the scrolls, which would let most of the evidence stand at once. Nobody has closed the question. Take the confident sentences on the signs as one reading of the site rather than a finding of fact.