Atlit Yam

Why this is extraordinary

Ten kilometres south of Haifa, 200 to 400 metres off the beach at Atlit, a village stands on the seabed in 8 to 12 metres of water. A 1993 field report called it the largest drowned prehistoric settlement then known on this coast and gave the area as about 60,000 square metres. The 2020 synthesis says about 40,000. The two figures are the two publications. The depth, 8 to 12 metres, is in both. A synthesis of the Carmel coast’s drowned sites, published in 2020, dates it by radiocarbon on charcoal and waterlogged plants to the Pre-Pottery Neolithic C, about 9,400 to 8,000 calibrated years ago. The Israel Antiquities Authority publishes an uncalibrated range of 8,180 to 7,550 years before present for the same place. Uncalibrated means the laboratory number, before the calibration curve moves it onto the calendar. The two reports are the same centuries said in two units.

What the divers mapped is a settlement, not a scatter. Rectangular houses of undressed kurkar stone, kurkar being the local cemented dune rock. Hearths. About twenty circular stone-lined pits. Straight walls, one of them up to 60 metres long. And wells. The deepest, well 11, is a shaft 5.7 metres deep and 1.5 metres across, lined with courses of stone, dug on down into the kurkar to reach fresh water. Three radiocarbon dates from the lowest fill fall between 8,540 and 7,930 calibrated years ago. The chapter reads those as the latest the well could have been built, because the fill is what people threw in after it was already a well. A 1993 study of the same shaft estimated that the sea on this coast stood some 16 metres below its present level around 7,900 uncalibrated years before present, and that the rise while the village lived was on the order of 15 to 20 millimetres a year. That is a fast rise. It is still a rise you could watch over a lifetime as a shrinking field, not only as a single wave.

The dead were buried in the village. The 2020 count is about 63 people. Most burials were primary, a body in a grave, often flexed. One grave held a woman and an infant. A 2008 study of their bones reported molecular traces of tuberculosis, and the chapter calls it the earliest known case. Auditory exostoses, bony growths in the ear, were recorded in male skeletons, the sort of change associated with time in cold water. Fish bones, wheat, and the wells together are the chapter’s picture of a Mediterranean fishing village that also farmed, earlier than the pottery Neolithic hamlets farther along the same drowned coast, Kfar Samir, Neve Yam, and Tel Hreiz among them.

Two stone settings carry the argument about meaning. One is seven kurkar megaliths, 1 to 2.1 metres long, six of them still standing, in a circle about 2.5 metres across and open to the northwest. Gray travertine on the inner faces is read as proof of a freshwater spring. Cup-marked slabs lie to the west. Galili’s reading, in the Antiquities Authority account and in the chapter, is a ritual structure at that spring. About 20 metres west, two parallel walls a metre apart form a corridor toward the circle. The chapter says palaeoastronomical calculations suggest that corridor could have been aligned on the sunset of the longest day, the evening when the sun sets farthest north. The account used here does not print the azimuth or the height of the horizon. An older encyclopedia entry on the same walls allows a corridor house, livestock below and people above, which does not need the sun at all.

The photographs are the Haifa exhibition. The village is still underwater, and storms peel the sand off it and then bury it again, which is why the plan was assembled over decades of rescue dives rather than one open season. If this is an ordinary Neolithic village that lost a slow argument with the sea, it is the best-preserved one on this coast, and the circle is a spring shrine. If the corridor was aimed at midsummer sunset, the shrine had a sky. If the end was a flood sudden enough to remember, the wells and the graves are the memory’s floor. The catch is that a village drowned in place can support all three, and the walls do not say which one the builders intended.