Mount Arbel

Almost everything odd about Arbel follows from the fact that the ground on one side of it is far below sea level. The Parks Authority describes the cliff as the product of the great Syrian-African rift and of geological faults running across the Lower Galilee, and attributes its sheer face to the hard limestone and dolomite it is cut from. The summit is a modest 181 metres above sea level. The lake below is more than 200 metres beneath that datum, and the difference is the 390 metres the Parks Authority prints.

The cliff is the heart of the Arbel Nature Reserve, declared in 1967, and of the national park wrapped around it, which also takes in Mount Nitai across the valley, Mount Savyon, the Horns of Hittim and the Arbel plateau. Mount Nitai is the ridge directly opposite, and the gorge between the two, Nahal Arbel, is the natural route from the Galilee hills down to the lake at Migdal.

The park keeps most of the rim unfenced on purpose. The Parks Authority states plainly that a large part of the cliff has no railing, because it wants the landscape left as it is, and asks visitors to hold on to children and follow the rangers. That is not boilerplate. The lookouts are the busiest part of the site and the drop begins at the grass.

Arbel appears in the record as a hiding place before it appears as anything else. Josephus writes that in about 161 BCE the Seleucid general Bacchides camped at Arbela, a city of Galilee, besieged and took those who were sheltering in caves there, and moved on to Jerusalem. The Encyclopaedia Judaica entry connects the episode to the account in 1 Maccabees of the capture of the mesalot at Arbel, and reads the word as a reference to the caves and the stairways linking them.

The famous siege is Herod's. Josephus tells it twice. The caves, he writes, were in mountains so abrupt that there was no way up from below and no way of climbing down from above, so Herod had chests built, hung them from an engine at the top of the mountain on iron chains, filled them with armed men carrying long hooks, and lowered them level with the cave mouths. The account in the Jewish War adds that fire was sent into the caves, and describes an old man at a cave mouth killing his seven children and his wife in turn rather than accept Herod's offer of safe conduct, with Herod close enough to watch and to plead with him.

Which year that happened in is not settled, and the disagreement is easy to see. The Parks Authority's own trail notes date the stand of the Galilean zealots to 37 BCE. The Wikipedia article says 38 BCE. The Encyclopaedia Judaica entry says 39 BCE. Josephus supplies the sequence, from the taking of Sepphoris in a snowstorm to the forty days before Herod arrived in person, but not a calendar date, and the modern figures are all reconstructions from the surrounding campaign.

The caves were used again a century later. Writing about himself in the third person as commander in Galilee at the start of the revolt against Rome in 66 CE, Josephus lists the places he fortified and includes walls built about the caves near the lake of Gennesar, in the Lower Galilee. The Parks Authority's Hebrew description of the park mentions a wall from the days of the Great Revolt among its remains.

The structure hikers descend to is not the caves of Josephus but a fortress built across them roughly sixteen centuries later. The Parks Authority describes a large system of chambers fronted by enormous walls of alternating black basalt and white limestone, and attributes it to the seventeenth century and probably to Ali Beg, son of the Druze emir Fakhr al-Din of the Ma'an family, which is where the local name Qal'at Ibn Ma'an, the fortress of the son of Ma'an, comes from. Wikipedia's article gives the same attribution, following Jerome Murphy-O'Connor's Oxford archaeological guide.

The building method is the interesting part. Nothing here was raised from open ground. The caves themselves began as natural karst cavities and were then quarried and enlarged by hand into dwelling halls, water cisterns, storage pits, flights of steps and a few ritual baths, and the fortress simply walled up the front of that honeycomb and connected it with concealed passages and stairways. What survives is therefore a facade and an interior, with the mountain doing the work of a roof and a back wall.

Finds inside push the occupation back a long way before the fortress. The Parks Authority reports Hellenistic material in the caves within the walls, including coins and ritual baths, which is the archaeological end of the story Josephus tells from the literary end. It is worth being clear about what a visitor sees: the walls and the cave complexes are old, and the way down to them is not. The steps, the iron pegs and the metal cables on the cliff face were installed by the Parks Authority so that the descent could be made at all.

About two kilometres before the park entrance, at the northern edge of Moshav Arbel, the road passes the ruins of the ancient village of Arbel and its synagogue. The Parks Authority credits the scholar Edward Robinson with identifying the remains in 1852; the Encyclopaedia Judaica entry records Charles Wilson exploring them later in the nineteenth century and the first excavation by Heinrich Kohl and Carl Watzinger in 1905. The building was deliberately made to stand out: dry-laid from large blocks of white limestone in a district where houses were built of black basalt.

The Parks Authority sets out two main phases. The first, founded on the evidence of its style and finds in the fourth century CE, was a hall with three rows of columns arranged in a squared horseshoe, carrying galleries above, with benches along the walls and a collection box cut into the rock of the northern wall behind a small door. That building was destroyed at a date nobody can fix, and a second synagogue went up on the spot, probably in the sixth century, reusing a great deal of the first one's stonework. It added a rounded niche in the southern wall for the Torah ark, repaved the floor and extended the courtyard northwards over two long vaulted halls. The visible ruin is the later building.

The dating has been argued over for a century and the argument is not closed. The Encyclopaedia Judaica entry states flatly that the third-century date originally proposed on architectural grounds is no longer accepted, without putting a firm alternative in its place. A fresh excavation in 2008 under Benny Arubas found material the Parks Authority describes as belonging to a synagogue building of the Second Temple period, which would make the site considerably older than either of the two phases above it.

The synagogue seems to have burned. The Parks Authority suggests the great earthquake of 749, remembered in Jewish sources as the shevi'it earthquake because it fell in a sabbatical year and which wrecked settlements up and down the rift. The gateway that now stands complete on the site was rebuilt by the Parks Authority from its own original pieces, which were lying scattered in the field: ancient stone, modern re-erection, and the distinction is worth carrying around the ruin. Entry into the synagogue building itself is currently barred because of the excavation, and the compound is free.

The Parks Authority publishes two circuits from the same car park and they are not variations on each other. The lookouts route is about a kilometre and takes about an hour, taking in a viewpoint facing the lake and the southern Golan, Mitzpe HaHaruv, the Carob lookout, named for a single carob tree growing on the cliff edge and commemorating Brigadier General Lunz, and the Mount Nitai lookout, which has a path surfaced for wheelchairs and pushchairs.

The cave fortress route is about two kilometres, takes about three hours and is graded for fit walkers. It leaves the entrance on the black-marked trail to the Carob lookout, turns east and goes over the edge, with pegs fixed by the Parks Authority to help on the way down. At the foot of the cliff it turns west; after roughly 800 metres it meets a red-marked trail, and turning left up the red trail brings you to the fortress. The red trail then climbs the cliff again, on cables and pegs set into the rock, back to the top.

Three published rules shape the day more than anything else. The gate closes an hour before the site closes. Last entry to the trails that descend the cliff is three hours before closing, and to the lookouts one hour before closing. And on rainy days the cliff route is shut.

• Lookouts circuit: about 1 km, about 1 hour, easy, three viewpoints
• Cave fortress circuit: about 2 km, about 3 hours, for fit walkers
• Descent and re-ascent on steps, iron pegs and fixed metal cables
• Trail marking map number 3
• Closed in rain; last entry to the cliff trails is three hours before closing
• Dogs allowed on the lookout trails only, leashed and muzzled, and never on the cliff trails

The reserve was declared in 1967, and the Parks Authority is explicit that the reason was biological before it was archaeological: the variety of habitats packed into the cliff, the species peculiar to them, and the fact that the ground forms part of a continuous corridor of open country that lets wildlife move between areas. The heritage sites were a second reason, not the first.

The species the authority names first is a small snail, Cristataria genezerethana, which it says exists only at the Arbel. Alongside it are rare plants including the bladder senna, Colutea istria, and the rock scrophularia, Scrophularia xylorrhiza, and birds of prey that nest in the face. The bird worth carrying binoculars for is the wallcreeper, Tichodroma muraria, which winters here and which the Parks Authority describes as looking like a red butterfly in flight.

The larger animals are there and are mostly not seen. The authority lists gazelle, foxes, jackals, badgers, wild boar and wolves among the mammals of the reserve, and notes that they lie up during the day while the trails are busy and forage in the evening, at night and in the early morning.