Upper Semliki, near Lake Edward, Democratic Republic of the Congo
Katanda harpoons
Older than about 89,000 years (barbed bone points in Middle Stone Age sands)
On cliffs above the upper Semliki, three Middle Stone Age sites produced bone points, some of them barbed, with a dagger-like piece and a great many catfish bones. Several clocks put the deposits older than 89,000 years, plus 22,000 and minus 15,000. Worked bone of that care was, at the time, supposed to be much later. A later remeasurement of the light clock agreed with the old age, and the doubt did not entirely leave.
Why this is extraordinary
Three cliffs on the upper Semliki, in the Western Rift, hold Middle Stone Age layers that produced something the diggers did not expect to find there. Bone had been worked into points. Some of the points are barbed. One piece is described as dagger-like. With them are fish bones, mostly catfish, and a reading of those bones says the fish are adults of a size that comes upriver to spawn. A camp that returns for that season is planning a meal it does not have yet. A camp that eats what is lying there is not. The bones do not write down which one it was.
The age is why the points mattered. In 1995 a set of clocks that do not all measure the same material — electron spin resonance on teeth, thermoluminescence and optically stimulated luminescence on sand, uranium series, and the slow chemical change called amino-acid racemization — were published together. They land older than 89,000 years, with an asymmetric uncertainty of plus 22,000 and minus 15,000. That is a window from roughly 74,000 to 111,000, not a birthday. The animals and the stacking of the layers were said to fit. The catch is what was dated. Sand dates a burial. A tooth dates a tooth. A barbed point is dated only if it lay in that sand when the grains stopped seeing the sun.
The light clock had a second chapter. The first luminescence measurements suffered from radiation quenching, a fault that dulls the signal the lab is trying to read. A reassessment heated additional samples more strongly to clear it. The new ages scattered, and they still agreed with the original old result. Agreement after a repair is stronger than the first print. It is not a narrower error bar. When the points were new in the literature, the clocks were called experimental, and readers asked where the other African sites with the same skill were. Both of those doubts were published. Neither one removed the barbs.
There is no sky reading on these points. No azimuth, no star, no month-count. The famous notched bone from this river, read as arithmetic or as a lunar tally, is a different object, found at Ishango and kept in Brussels. It has its own room at the Ishango bone. What this changes: if the points are in place, formal bone fishing gear is a Middle Stone Age fact in this valley, older than about 89,000 years. If they drifted down from a younger visit, the sand is still old and the invention moves up with the points. The photograph is the Semliki valley from the Ugandan side. It is a stand-in for the river, not a picture of Katanda, and not a picture of a harpoon. Head shaping among the Mangbetu is a living practice of the last century, not this sand.
Why it matters
If the points are as old as the clocks, people here were shaping bone into fishing gear tens of thousands of years before that habit is common in Europe. If the points sank from a younger layer, the clocks date the sand and the story of an early specialist fishery goes with them.
Katanda is three sites, on the upper Semliki in the Western Rift, in what was then published as Zaire. The points are barbed and unbarbed, and one object is described as dagger-like. The fish are mostly catfish, and a reading of their size says they are adult fish in spawning season, so the visits were timed. The 1995 dating paper used electron spin resonance on teeth, thermoluminescence and optically stimulated luminescence on sand, uranium series, and amino-acid racemization. The figure they share is older than 89,000 years, with an uncertainty of plus 22,000 and minus 15,000, a window from roughly 74,000 to 111,000. Fauna and the order of the layers were said to agree. The first luminescence readings had a fault called radiation quenching. A reassessment heated further samples more strongly, got scattered ages, and still found them concordant with the original result. Critics at the time called the clocks experimental, and asked why the same skill was not already known at many other African sites. No horizon and no star are published for the points. The notched bone from this same river, the one with a lunar count, is a different object.
Photo tour
1 photograph, unedited. On a phone, swipe the row. Choose one to look closely.
- No openly licensed photograph of a Katanda point is used here. The valley picture is a stand-in for the river, and it is labelled as one.
What is still unexplained
These sit beside the consensus account, with the same visual weight. None of them is marked solved.
Location and dating
- Location
- Three sites at Katanda on the upper Semliki, Western Rift, Democratic Republic of the Congo. The pin is an approximate point in that valley, 0.08 south, 29.65 east, not a surveyed corner of a trench. The photograph looks into the valley from Uganda and does not show the cliffs.
- Period label
- Current consensus — open to revision Middle Stone Age. The published age is older than 89,000 years, plus 22,000 and minus 15,000, if the points belong with the dated sand and teeth.
- How this dating was done
- Electron spin resonance, thermoluminescence, optically stimulated luminescence, uranium series, and amino-acid racemization. Not radiocarbon, which does not reach this age. The luminescence was reworked after a quenching problem.
- What could change it
- A direct date on the bone of a point, or a new trench in which a barbed point is recorded in place against single sand grains, would replace the association argument.
What is actually there
- Three sites yielded barbed and unbarbed bone points and a dagger-like object.
- The animal bone is dominated by catfish.
- Five methods were published together in 1995, agreeing on an age older than about 89,000 years.
- A luminescence reassessment found quenching in the first readings and still supported that age.
- The photograph is the valley from Uganda, not the points.
- No star or lunar count is published for these points.
Open questions
| Question | Status |
|---|---|
| Are the points as old as the sand? | Open. Several methods agree. The link from sand and teeth to a bone point is the catch. |
| Does a barbed point mean a planned fishery? | Open. Catfish of one size class are the evidence offered. A camp can also collect what the river gives. |
The current narrative
Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.
The current narrative — open to revision
An early bone fishery
People at Katanda made formal bone points, including barbed ones, and used them on spawning catfish, before about 89,000 years ago.
What supports this
- Several independent clocks overlap on an age older than 89,000 years, plus 22,000 and minus 15,000.
- A later luminescence study, after fixing quenching, still agreed.
- The points sit with Middle Stone Age stone tools, and there are too many of them for a single piece fallen from above, in the excavators’ account.
What pushes back
- The clocks date sand and teeth. A point dates only by being in that layer.
- The first light measurements had a real fault. Concordant scatter is not a precision date.
- When the papers appeared, equally careful bone work was still rare in African sites of this age, which is why the result was hard to place.
What would settle it
It is the centre because more than one method, and a remeasurement, land in the same broad window, and the fish and the points were published as one assemblage. It moves if the points cannot be shown to be in place.
Emerging narratives
Other published readings. Each may be right. None is adopted here unless the evidence carries it.
An emerging narrative
Younger points in older sand
The bone work is real and skilled, and it belongs to a later visit whose objects worked down into Middle Stone Age deposits.
What supports this
- River cliffs and loose sediment can move small objects downward.
- The experimental reputation of the 1995 clocks was part of the published doubt.
- A specialised fishery of this age was, at the time, an argument from one valley.
What pushes back
- The reassessment did not pull the luminescence young.
- Fauna and stratigraphic order were offered as independent checks.
- One slipped point is easier to argue than a group of barbed points with the fish they are said to have taken.
What would settle it
It stays because association is always the weak joint when the object itself is not the sample. It is not the centre while several clocks and a remeasurement still agree.
Should the current view be reevaluated?
Open to revision
The broad age, older than about 89,000 years, has been published twice for the sediments. The step from those sediments to a barbed point is the part a reader still has to weigh. Open matches that step.
What drives this
- A multi-method age older than 89,000 years, plus 22,000 and minus 15,000.
- A luminescence reassessment that supported it after a quenching correction.
- No direct date on the bone of a point.
This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.
Weigh it yourself
Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.
Visitor lean so far — not evidence. On this browser only; a shared tally is not collected.
If the contested reading is right
What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.
Barbed bone points and a seasonal catfish fishery before 89,000 years ago
Formal bone fishing gear would sit in the Middle Stone Age of the Western Rift, not only among the much later toolkits of Europe and of later African sites.
Why it is still worth a test
A point dated on its own bone, in a section where the sand grains can be dated one by one.
Sources
- Dating and context of the Semliki bone-point sites Science, 1995, the multi-method age
- A Middle Stone Age worked bone industry from Katanda Science, 1995, the points and the fish
- Luminescence dating at Katanda, a reassessment Quaternary Science Reviews, the quenching correction
Image credits
- The Semliki valley in the Democratic Republic of the Congo, seen from the Ugandan side. This is the river country. It is not the Katanda cliffs, and no bone point is in the frame. . Mgslee, CC BY-SA 4.0. Shown unedited.