The Antikythera Mechanism

by Brian Dunning
December 15, 2009

Imagine the year 1900. The skies are leaden gray, the dark waters all around you shiver with the approach of a distant storm. You're in the warm Ionian Sea in the Mediterranean, sheltering along the coast of the barren and barely populated island of Antikythera, waiting out a storm. It's frustrating because you're on your way home from Africa where you and your crew have been diving for sponges. Your captain, Dimitrios Kondos, thinks the lost time may as well be put to good use, so he orders you below to see what sponges you may be able to find here. You don your copper diving helmet and heavy suit, and they lower you into the depths. Streaks of dull gray light from above shimmer around you as the rocky bottom approaches.

But it's not sponges you find. Minutes later you're back on board the ship, jabbering excitedly, so incoherent that Kondos thinks you have carbon dioxide poisoning. He goes down himself to have a look. And what Kondos and his crew bring up over the subsequent two years comprises one of archaeology's great finds, that truly challenged our understanding of the history of technology.

Chief among the finds was what has become known as the Antikythera Mechanism, fragile chunks of green corroded bronze, that when picked apart, revealed unexpected mechanical components, mainly gears. The device was surprisingly complex. At first it was thought to be a clock, but when Greek inscriptions were found, it turned out to be a sort of astrolabe for predicting eclipses and moon phases and the positions of the planets, of unprecedented sophistication. So sophisticated, in fact, that everything we knew told us that the Antikythera Mechanism was a full 1,000 years out of place.

The shipwreck, known as the Antikythera Wreck, has been dated to the first century BCE. The Antikythera Mechanism dates from the century before that. And then, so the popular version of the story goes, nobody on Earth had either the astronomical knowledge, nor the mechanical know-how, to construct such a device until a millennium later. Some have said the Antikythera Mechanism is therefore proof of time travel, alien visitation, or Atlantis.

Physically, the device was about the size of a shoebox, with wooden sides and bronze faces. On the front face were two large and three small output dials. On the back were three concentric output dials. To operate the device, you turned a crank on the side which rotated at least 30 gears inside the machine, some of which were epicyclic. The hands that went round each of the two large dials swept over spiral slots, with a pin on the arm that rode in the slot, similar to a needle following the groove on a record. By setting some preferences, such as what type of calendar you wanted to use, and turning the side crank to select the current date, you could learn all sorts of things: Whether this was an Olympics year, when the next solar and lunar eclipses were (by date and hour), where the twelve constellations were along the ecliptic, the phase of the moon, and the positions of the five planets known at the time.

Although we now know what the device did, we're not sure what its use was. By its construction in bronze, which readily corrodes, we know it was not designed for navigation at sea. Astronomers and astrologers probably could not have afforded it. It could have been used as an education tool. Most likely it was built for wealthy Romans who had some interest in its features, probably not too different from early adopters who wanted to have the first iPhone with all the cool apps. The wreck was laden with other objects of great value, most notably a vast hoard of coins and a Peloponnesian bronze sculpture, a larger-than-life young man called the Ephebe.

So what about these claims that the mechanism is 1,000 years out of place, and no humans had the knowledge to make something like it? Does this prove that aliens, Atlanteans, or time travelers must have been involved? It is a fact that the Antikythera Mechanism is substantially more complicated than any other mechanical devices known from its time. Specifically, it's one of the earliest known uses of meshing gears. But contrary to the popular telling, it's not the oldest. Gears were used to drive doors and lift water in India as early as 2600 BCE, two and a half millennia before the Antikythera Mechanism. Aristotle described the function of gears in the 4th century BCE. 100 years later, Dionysius of Alexandria used gears in his automatic arrow firing machine gun. The Greek National Museum contains examples of epicyclic gears from the period. Archimedes was making all sorts of mathematical and mechanical inventions at the time. For hundreds of years, Greek astronomers had been studying the movements of heavenly bodies, and by Archimedes' lifetime, all the motions replicated on the Antikythera Mechanism were known to science.

We know a lot about where and when the device was made from the inscriptions in the bronze. The back face is covered with instructions for its use, as is its inside if you open the device. These include descriptions of the controls, various calendars, and mentions of the celestial bodies tracked by the device. By the language and terminology used, as well as by the context of its find among the other artifacts recovered from the Antikythera Wreck, we now have a pretty good idea of where and when it was built: The middle of the 2nd century BCE, probably in Syracuse or Corinth.

What we don't know is who built it, but there are some good candidates. It is assumed that Archimedes, who died several decades before the device was built, left behind a tradition of scientists who continued his work and built upon his inventions, and the device could have come from this school. Another leading contender is Hipparchos, perhaps the greatest of early astronomers, who was in his heyday when the Antikythera Mechanism was constructed. Most notably, Hipparchos was the first to devise a mathematical model to predict the anomalies of the moon's movement, and the Antikythera Mechanism contains a gear set to reproduce exactly these computations.

Most archaeologists agree that this particular device was neither unique nor the first of its kind. Two factors contribute to this: First, its design is quite refined, which is not consistent with a prototype. Second, an object as expensive and complex as this would typically be made in a series in order to recover the costs of design. Why, then, are its siblings not found? Probably because they were made of bronze, and bronze was highly recyclable and valuable. Few commonplace bronze objects from the ancient world survive for this reason, except for those that were lost at sea and thus escaped recycling. If there were other computational devices made in the period, it is not suprising to archaeologists that they were lost to history and are unknown.

The surviving fragments of the device at the Greek National Archaeological Museum are too fragile to travel, and so in 2005, two teams brought their equipment to Athens to perform the most advanced imaging studies to date. Hewlett Packard's team performed polynomial texture mapping to high resolution images of the inscriptions, made with lighting from all different angles to reveal every possible bit of detail. A company called X-Tek Systems brought their 8-ton x-ray machine called the BladeRunner to the museum, all the way from the UK, and made CT scans. CT, or computed tomography, is the process of creating a 3-D image from slices; in this case, slices from x-ray images. We now have extremely detailed maps of the internal mechanism and transcripts of all the surviving inscriptions.

And, as a result, we now know that the ancient Greeks were building far more advanced computational devices than we used to give them credit for. We knew they had the knowledge, we just didn't know they were translating it into bronze so exquisitely.

Antiscience people love to point to cases like the Antikythera Mechanism as examples of science being wrong. They gloat over their belief that historians have been embarrassed, careers shattered, books proven to be in error. They imagine that researchers at universities everywhere are being fired or stripped of their credentials. They believe this case adds to an ever-growing mountain of proof that science is, itself, destined to inevitable failure, and that enlightened scientists should abandon their practice and turn to faith in the supernatural.

In fact, nothing could be further from the truth. A find like this, that substantially revises our understanding of history, can be the crowning achievement for a scientist's career. Contrary to what antiscience would have you believe, scientists do not fear new discoveries, they long for them. Every action dictated by the scientific method seeks to learn something new, to revise and improve our understanding of nature or history. Major finds represent major improvements to our theories. Thanks to the Antikythera Mechanism, we now have a better understanding of technology in the ancient world, and new directions for researchers to turn. The idea that a discovery like this is embarrassing, or exposes weakness in the scientific method, is absurdly upside-down and backwards.

Many times I've heard the argument made that scientists fear new discoveries because it would threaten their grant money, so it's in scientists' best interests to cover up anything new. All you have to do is look at what's being funded these days to see how wrong this particular conspiracy theory is. A glance at the National Science Foundation's list of recent funding awards teaches one lesson: If you want grant money, be a maverick, have something new and exciting. What's a financial incentive for scientists to look for discoveries that challenge our worldview? New discoveries attract grant money, and grant money leads to more new discoveries. There is no money in continuing to grind over what we already know. So please, can we put this particular conspiracy theory to bed?

It's not every year that we find something so historically significant, and that we learn so much from. The rewards we gain from increasing our knowledge by studying them are inestimable. Those who dismiss such finds as alien, or otherwise not part of history, miss out on that knowledge. The lesson to learn is that when you're confronted by a discovery, stopping at the popular supernatural explanation is guaranteed to lead you nowhere. Instead, you should do what science suggests, and be skeptical.

By Brian Dunning
Follow @BrianDunning