Origins of Civilisation · Southwest shore of the Sea of Galilee (Lake Kinneret), Israel

Ohalo II

A 23,000-year-old fishers' camp, drowned and preserved, that pushes the roots of cultivation eleven millennia deeper than the Neolithic.

Mainstream: c. 23,000 years ago (Last Glacial Maximum, Upper Palaeolithic)Alternative: Date is secure — the challenge Ohalo poses is to the story that farming 'began' only around 10,000 years ago32.72°, 35.57°

Origins

At a glance

Ohalo II is a waterlogged Upper Palaeolithic camp on the southwest shore of the Sea of Galilee, normally hidden beneath the lake and revealed only when drought and water extraction dropped the level far enough. It came to light in 1989, when Dani Nadel spotted flints and bone emerging from the exposed lakebed. What the excavations uncovered is astonishing: the floors of half a dozen brush huts, hearths, a human burial and an organic record — seeds, wood, fibre, fish bone — so complete that it survives from around 23,000 years ago, during the coldest phase of the last Ice Age. The submersion that hid it is also what saved it, sealing perishable material that virtually never lasts this long.

See it on the globe →

The mainstream view

What archaeology says

Ohalo II is one of the richest windows we have onto life at the Last Glacial Maximum. Excavations led by Dani Nadel recovered the collapsed remains of brush huts built of branches, hearths, stone tools, a grave, and roughly 150,000 plant remains representing well over a hundred species — the kind of preservation that on a dry site would be reduced to a handful of charred fragments. Its occupants were fisher-hunter-gatherers who exploited the lake intensively, and the site preserves some of the earliest direct evidence of systematic freshwater fishing.

The findings that reverberate beyond the site concern plants. The inhabitants gathered and processed wild cereals — wild emmer wheat and wild barley — on a serious scale. A grinding stone from one hut retained cereal starch, evidence that grain was being milled into something like flour some 23,000 years ago. Use-wear analysis of glossed flint blades identified them as sickle inserts for harvesting semi-ripe cereals: the earliest known harvesting tools, roughly 8,000 years before such sickles become common among the Natufians.

Most provocatively, Ehud Weiss, Ainit Snir, Dani Nadel and colleagues reported in a 2015 PLOS ONE study the assemblage of wild cereals accompanied by proto-weeds — the weedy plants that thrive in ground disturbed by human tending, such as certain ryegrasses and bedstraws. Their presence, alongside a nearly pure stand of harvested cereals, is read as a sign of small-scale, trial cultivation: people sowing, tending and reaping wild grain long before any plant was genetically domesticated.

Key evidence cited

Waterlogged brush-hut floors, hearths and a burial preserved from c. 23,000 years ago beneath the lake
Around 150,000 plant remains from over 100 species, an almost unheard-of botanical archive for the period
A grinding stone bearing cereal starch, showing wild grain was milled into flour
Glossed flint sickle blades identified as the earliest known cereal-harvesting tools
Abundant fish bone and other remains documenting intensive year-round exploitation of the lake

The alternative view

What the skeptics propose

Ohalo II does not attract fringe theories; instead it quietly detonates a tidy textbook narrative, and the interesting argument is over how far to push its implications. The conventional story has agriculture beginning in the Fertile Crescent around 10,000 to 11,500 years ago with the Neolithic, when people first domesticated wheat and barley and settled into farming villages. Ohalo shows cultivation-like behaviour — deliberate harvesting with sickles, grinding grain into flour, and possibly sowing and weeding wild cereals — some 23,000 years ago, roughly eleven thousand years earlier.

The stronger reading, associated with Snir, Weiss and Nadel, is that the proto-weeds are a genuine fingerprint of intentional cultivation, and that trial cultivation of wild plants was therefore a very ancient and possibly recurring practice that only much later, under the right conditions, tipped over into full domestication. On this view, farming was not a sudden Neolithic revolution but the eventual outcome of a slow, stop-start relationship with plants stretching back deep into the Ice Age. The 'origins of agriculture' become a long process rather than a threshold event.

The more cautious counter-position, held by many prehistorians, accepts everything Ohalo demonstrates about harvesting, grinding and intensive cereal use but resists calling it cultivation. Proto-weeds, they argue, can accumulate in a favoured, repeatedly used camp without deliberate sowing, and gathering wild grain intensively is not the same as farming it. Both sides agree on the striking facts; they differ on the word. Either way, Ohalo II makes it impossible to treat agriculture as an invention that switched on in the Neolithic, and pushes its behavioural roots dramatically deeper in time.

Key evidence cited

Snir and colleagues' 2015 PLOS ONE report of proto-weeds signalling small-scale trial cultivation
A near-pure stand of harvested wild cereals consistent with deliberate tending
Harvesting, grinding and possible sowing 23,000 years ago — some 11,000 years before the Neolithic
The argument that cultivation was an ancient, recurring practice, not a sudden Neolithic invention
The counter-argument that proto-weeds and intensive gathering fall short of true farming

Genuinely open questions

Do the proto-weeds prove deliberate cultivation, or just an intensively reused gathering camp?
If people were tending wild cereals 23,000 years ago, why did full domestication take another 11,000 years?
How widespread were sites like Ohalo, and how much of this early record is lost beneath modern lakes and seas?

Worth knowing

Ohalo II preserves flour-making 23,000 years ago and the oldest known sickle blades — pushing the human relationship with cereals some 11,000 years before farming was supposed to have begun.