4.6/5 - (65 votes)

It is about the length of a pencil, dark brown, and covered in 168 tiny scratches. It has been described as the oldest mathematical tool of humankind, and nobody can agree on what it actually counts.

A pencil-sized bone from the shores of the Semliki

In 1950, the Belgian researcher Jean de Heinzelin de Braucourt was digging near the Semliki River, in what was then the Belgian Congo. The river is fed by Lake Edward and forms part of the headwaters of the Nile.

Among human remains and stone tools left by a small community of fishers and gatherers, he found a curved bone with a sharp piece of quartz fixed to one end. It had been scraped and polished so heavily that nobody can tell which animal it came from anymore.

The site is now dated to about 20,000 years ago. Even that is debated: nearby volcanic activity upset the carbon isotopes, and early estimates ranged from about 8,500 to 44,000 years.

Three columns that look suspiciously deliberate

The 168 marks run along the bone in three parallel columns. Researchers still use the French labels: M for milieu (middle), G for gauche (left) and D for droite (right).

De Heinzelin was convinced the notches were “deliberately planned” and showed simple arithmetic. Then came the detail that made the bone famous: one column appears to show the prime numbers between 10 and 20, the numbers that can only be divided by 1 and themselves.

A prehistoric table of primes, carved 20,000 years ago in a small fishing community, was simply too good a story to ignore.

See also  Nobody tells you why Roman numeral clocks use IIII instead of IV

The prime number dream has a big problem

The historian of mathematics Peter S. Rudman pointed out a catch. Prime numbers only make sense once you understand division, and he dates that idea to no earlier than 10,000 BCE. In his view, primes were probably not understood until the early Greek period, around 500 BCE.

Even the bone’s strongest defenders hold back here. Mathematicians Dirk Huylebrouck and Vladimir Pletser see a counting tool built on base 12, with sub-groups of 3 and 4, a bit like a primitive slide rule. But they concluded that the evidence is not enough to say anyone understood primes back then.

And there is an awkward detail. A second bone found in the same archaeological layer carries marks that look nothing like mathematics.

A calendar, a grip, or a reference table

Harvard archaeologist Alexander Marshack examined the bone under a microscope and proposed a six-month lunar calendar. His argument: the first two columns add up to 60, about two lunar months, while the last adds up to 48, roughly a month and a half. Critics replied that he had searched for a pattern rather than tested a hypothesis.

Some suggest something far more mundane: the scratches may simply have given the handle a better grip. Anthropologist Caleb Everett, on the other hand, insists that “the quantities evident in the groupings of marks are not random” and sees a possible doubling pattern, even a numeric reference table.

The mathematician Olivier Keller titled his critique “The fables of Ishango, or the irresistible temptation of mathematical fiction.” His warning: we tend to project our own idea of numbers onto a 20,000-year-old object.

See also  His neighbors thought he was a sorcerer, and his rods did the multiplying

George Gheverghese Joseph summed up the whole debate in one line: “A single bone may well collapse under the heavy weight of conjectures piled onto it.”

Not even the oldest counting bone

Despite its title, Ishango is far from the oldest notched bone. The Lebombo bone, a baboon fibula found in Border Cave between South Africa and Eswatini, is between 42,000 and 43,000 years old according to 24 radiocarbon datings. It is often confused with Ishango.

Its discoverer, Peter Beaumont, noticed that the notches were cut with different edges, and linked them to rituals. In Moravia, the so-called Wolf bone, found in 1937, carries 55 marks that may also be a tally.

The habit that never left us

Whatever Ishango counted, its makers started a habit we still use. Tally marks are a base 1 system, and we cluster them in groups of five because people identify a cluster of five far more easily than one of ten.

Chinese numerals for one to three, the Brahmi numerals and Roman numerals all grew out of tally marks. In 2018, a few of them even made it into the Unicode standard.

The bone sits today in the treasure room of the Royal Belgian Institute of Natural Sciences in Brussels, viewable only on written request. Prime number table or not, it reminds us that the hunt for patterns in numbers may be one of the oldest things humans have ever done, and that we are still very good at seeing patterns that might not be there.