Royal Belgian Institute of Natural Sciences, Brussels, Belgium
The Ishango bone
About 25,000 to 16,000 years ago (found at Ishango, on the Semliki, now in Brussels)
A baboon bone about 10 centimetres long, with a quartz edge in one end, carries 168 notches in three columns. It was dug up at Ishango in 1950 and stands in a case in Brussels. The columns add to 60, 48, and 60. Arithmetic, a lunar count, and a plain tally all use those same cuts.
Why this is extraordinary
The object is a slightly curved baboon fibula, about 10 centimetres long, with a quartz flake set in one end and sticking out about 2 millimetres, so the bone is a handle and the quartz is a blade. It was found in 1950 on a terrace at Ishango, on the Semliki River in the east of the Democratic Republic of the Congo. It is on permanent display at the Royal Belgian Institute of Natural Sciences in Brussels, which is why the pin is in Belgium and the ground is not. The institute’s brochure says radiocarbon dates layers there to over 20,000 years, the oldest of them 25,000, and places most of the bone and quartz remains between 20,000 and 25,000. A heritage inventory gives a wider bracket, about 25,000 to 16,000. This room keeps both.
Three columns of notches run the length of the handle. The inventory used by the astronomy-heritage study counts them as follows. One column has 60 notches in four groups, 11, 13, 17, and 19. The middle column has 48, in eight groups, 3, 6, 4, 8, 10, 5, 5, and 7. The last column has 60, in four groups, 11, 21, 19, and 9. The total is 168. The groups are not casual. What they mean is the dispute.
The excavation report read them as numbers. The middle column has doubles, 3 and 6, 4 and 8, and then numbers that do not double. One side column is 10 plus or minus 1 and 20 plus or minus 1. The other is the prime numbers between 10 and 20. Later mathematical readings kept a mixed base of 10 and 12, because the columns sum to 60, 48, and 60, and they set the prime-number claim aside as more than the marks can prove. A lunar reading takes the same 60s as about two synodic months, and the 48 as a month and a half, so the bone tracks the moon for about five and a half months, a double lunation being 59 or 60 days. A tally reading says the groups are a count of something, and the pretty arithmetic is what any long tally will offer if you go looking. A second notched handle, found in 1957 in a similar context and only fully noted in 1998, has six groups, 20, 6, 18, 6, 20, and 8, read as a base of 6 and 10. It is not in the photograph. No horizon was measured. The sky in this room is the moon, if the notches are months.
Why it matters
If the groups are a lunar calendar, people on the Semliki were tracking months more than 16,000 years ago. If they are arithmetic, the same people were playing with doubling and with odd numbers. If they are a tally, the bone is a knife handle someone notched, and the patterns are ours.
The photograph is the Brussels case. The Semliki terrace is not shown. The pin is the institute, because that is where the bone is.
Photo tour
1 photograph, unedited. On a phone, swipe the row. Choose one to look closely.
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
- On display at the Royal Belgian Institute of Natural Sciences, Brussels. The pin is 50.837 north, 4.377 east. The find spot is the Ishango terrace on the Semliki, in the Democratic Republic of the Congo. This room does not move that terrace onto the Belgian map.
- Period label
- Current consensus — open to revision The heritage study’s bracket is about 25,000 to 16,000 years ago. The bone is a tool first, a notched handle second.
- How this dating was done
- The age is the re-dated archaeology of the Ishango terrace, not a number carved on the bone. The lunar reading is a count of notches against the synodic month. It is not archaeoastronomy of a wall or a horizon.
- What could change it
- A published test that predicts notch widths from a sequence of lunar phases, and then scores the bone without regrouping the cuts, would decide the calendar. A date on the bone itself, rather than on the terrace, would narrow 25,000 to 16,000.
What is actually there
- Baboon fibula, about 10 centimetres, quartz tip protruding about 2 millimetres.
- Found in 1950 at Ishango. Displayed in Brussels.
- Columns sum to 60, 48, and 60. Total 168.
- Groups include 11, 13, 17, 19 and 11, 21, 19, 9 and 3, 6, 4, 8, 10, 5, 5, 7.
- A second handle has groups 20, 6, 18, 6, 20, 8.
- The lunar span offered is about five and a half synodic months.
Open questions
| Question | Status |
|---|---|
| Are the groups a calculation, a lunar tally, or a count of something unnamed? | Open. The sums are real. The primes are the weakest gloss. The moon uses the column totals, not the picture of a crescent. |
| Does the second bone obey the same rule? | Open. Its sixes fit a base of 6 better than they fit a list of primes. |
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
A knife handle someone notched
The bone is a cutting tool. The notches are a tally of an unknown thing. Bases, primes, and months are patterns read back onto a count.
What supports this
- The quartz tip shows the object was a tool before it was a document.
- Any long run of groups will offer doubles and round sums to a reader who wants them.
- The prime list was set aside by later mathematical readers as unproven.
What pushes back
- Both side columns summing to 60, and the middle to 48, is a lot of roundness for one handle.
- A second notched bone from a similar context makes a one-off tally less comfortable.
- A tally still needs a reason for the groups to change size as they do.
What would settle it
It is the reading that adds the least. It is also the reading that ignores how regular the least-effort explanation has to become.
Emerging narratives
Other published readings. Each may be right. None is adopted here unless the evidence carries it.
An emerging narrative
Doubling, tens, and a base of twelve
The columns are a numerical table. Doubling, numbers one away from 10 and 20, and column totals of 60 and 48, fit a mixed base of 10 and 12. Primes are a step further than the marks require.
What supports this
- 3 then 6, and 4 then 8, are doubles sitting in the middle column.
- 11, 21, 19, and 9 sit one away from 10 and 20.
- 60, 48, and 60 are the column totals, and 48 is 4 times 12.
What pushes back
- The middle column also contains 10, 5, 5, and 7, which break a clean doubling story.
- Group boundaries are a choice. Move a cut and the arithmetic changes.
- No Ishango sentence tells us the base.
What would settle it
It stays close to the marks and drops the most glamorous claim, the primes. It is still a modern description of a silent bone.
An emerging narrative
About five and a half months of the moon
The notches follow the moon. Two columns of 60 are two double lunations, 59 or 60 days each, and 48 is a month and a half, together about five and a half synodic months.
What supports this
- A synodic month is about 29 and a half days, so 59 or 60 is a real double month, and both side columns hit 60.
- The heritage study records this time-reckoning reading beside the arithmetic one.
- Bones and strings are used as calendars in later African practice, which shows the technology is thinkable.
What pushes back
- Matching a sum to a month is a search for a pattern. A test would predict the size of each notch from the moon’s face.
- Later practice on other continents and in other millennia does not date this handle.
- The second bone’s groups, 20, 6, 18, 6, 20, 8, are not five and a half months in any obvious way.
What would settle it
It is the sky reading, and it is in the heritage inventory, so it is not dropped. The second bone and the missing phase-by-phase test keep it from replacing the tally.
Should the current view be reevaluated?
Open to revision
The bone, the quartz tip, the three columns, and the sums 60, 48, and 60 are facts. A prime list is more than those facts. A lunar calendar is a serious use of the same totals and has not been tested against the width of each cut. The tally remains the most economical reading and the least satisfying one.
What drives this
- Column totals of 60, 48, and 60, and a total of 168.
- A lunar span of about five and a half months, and a criticism that the fit was searched for.
- A second bone with a different series, and a date bracket nine thousand years wide.
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.
The notches track the moon
People at Ishango, sometime between about 25,000 and 16,000 years ago, kept a lunar count on a tool handle for about five and a half months.
Why it is still worth a test
A phase-by-phase test of notch width, scored without moving the group boundaries, on both bones.
Sources
- Ishango bone, heritage of astronomy UNESCO astronomy portal
- Thematic study case notes on the Ishango bone ICOMOS and the International Astronomical Union
- Have you heard of Ishango? Royal Belgian Institute of Natural Sciences
Image credits
- The bone on display. One face of grouped notches is toward the camera. The quartz flake set in one end, which sticks out by about 2 millimetres, is not the end you see. The bone is about 10 centimetres long. . JhowieNitnek, CC BY-SA 4.0. Shown unedited.