The Antikythera Mechanism: Art or Science?

Sep 23, 2009 Alun Featured, Science, The Past 10 Comments Ancient Astronomy, Ancient History, Ancient Science, Antikythera Mechanism, Archaeoastronomy, Archaeology, Astronomy, History of Astronomy, History of Science, Research Blogging

The Anti­kythera Mech­an­ism. Photo (cc) Tilema­hos Efthimi­adis.

Some posts take quite a while to write. This is a response to Candy Minx and Mar­tin Rundk­v­ist who were dis­cuss­ing the Anti­kythera Mech­an­ism back in 2006 (Anti­kythera, Time, A Reply to the Minx). Candy Minx thought that the Anti­kythera Mech­an­ism was an expres­sion of what was already known and embed­ded in a soci­ety through things like myth and ritual. Mar­tin thought that the mech­an­ism was far more com­plex, indeed need­lessly com­plex, for an ancient soci­ety and so was some­thing quite dif­fer­ent to the folk astro­nomy of the time. Ori­gin­ally I planned to write a fence-sitting com­prom­ise. I thought that Candy Minx was right to an extent, there was no need for a device like this because rituals and folk obser­va­tion could allow people to time the year as well as they needed. At the same time I thought that Mar­tin was right to point out that the mech­an­ism gave res­ults with far more accur­acy than folk astro­nomy needed, or would even recog­nise. A dif­fer­ent sort of astro­nomy is vis­ible in the Anti­kythera Mech­an­ism. I didn’t blog too much about the 2006 paper because I atten­ded a few of Mike Edmunds’ talks on the topic and heard that more would be pub­lished, which happened in 2008. Any­how in my own fluffy and fence-sitting way I’ll now offer my compromise.

Someone with an extraordin­ary ima­gin­a­tion built the Anti­kythera Mech­an­ism and, if he were alive today, we wouldn’t hes­it­ate to call him a sci­ent­ist. I don’t know if the designer was in the same league as New­ton or Galileo, but he was cer­tainly the equal of Kepler, Coper­ni­cus or Brahe. It’s hard to over­state how extraordin­ary the device described in the 2006 paper is, but I’m going to give it a go.

If you’re the one per­son who hasn’t heard of the Anti­kythera Mech­an­ism then Nature have a handy video introduction.

All that remains now is a col­lec­tion of cor­roded lumps found off the island of Anti­kythera. The 2006 paper described what the team dis­covered after x-raying the lumps to read the hid­den inscrip­tions without priz­ing apart the device and dam­aging it. Prior to this paper it was thought that the device could keep track of the Sun and the Moon. This is no small feat.

The Sun would be mov­ing slowly against the back­ground stars, so over the course of a year it would pass through all the signs of the zodiac. The Moon how­ever is more com­plex. The Moon also moves in front of the back­ground stars, but it only takes about 27 days to do this. It’s called the sider­eal period. So you need a couple of gears to drive those two motions. But you wouldn’t really think of the sider­eal period as a month. For most people the syn­odic period, the time between one New Moon and the next or the time between one Full Moon and the next, is a month. This is around 29½ days. Throw in extra gears for driv­ing other dis­plays show­ing eclipse cycles and it’s clearly a com­plex device. The ori­ginal stud­ies found evid­ence of epi­cycles, gears moun­ted on other gears. Add other fea­tures like dis­plays for eclipse and lunar cycles on the back and it’s obvi­ous you have a com­plic­ated device. The 2006 research showed that in fact it was all a bit more com­plic­ated than that.

The Moon’s move­ment isn’t con­stant. It speeds up and slows down. This is because its orbit isn’t exactly cir­cu­lar. Instead it’s slightly egg-shaped. The point fur­thest from the earth is the apo­gee and the point closest to the Earth is the peri­gee. When it’s near the apo­gee it travels slowly, but when it moves closer to the Earth it picks up speed until it passes peri­gee and then it slows down again. This is called the first lunar anom­aly. The dif­fer­ence is notice­able by the naked eye, if you’re will­ing to make sys­tem­atic obser­va­tions. This is all simply explained by Kepler’s Laws of Plan­et­ary Motion. There’s small prob­lem. Kepler used ellipses.

You can’t use ellipt­ical gears. The point of gears is that they must have inter­mesh­ing teeth. An ellipt­ical gear would lose con­tact with the driv­ing gear as its axis changed. Instead it seems that the mech­an­ism used two gears, one slightly off-axis from the other. The rota­tion was con­nec­ted by a pin-and-slot arrange­ment, so that the one gear wouldn’t turn at quite the same rate as the other gear. The on-axis gear can then be turned reli­ably by the drive gears, while the motion of the moon can driven by the off-axis gear. So you have a device that can track the sider­eal, syn­odic and anom­al­istic months, all while the Earth is spin­ning round the Sun. If that’s caus­ing your head to spin you might want to skip the next paragraph.

There’s another prob­lem. The lunar anom­aly describes the Moon’s travel from one apo­gee to the next. This apo­gee is also rotat­ing around the earth. If the apo­gee is in Aries then two and a bit years later it will be in Can­cer, and another two and a bit years to move into Libra until it too has trav­elled through the zodiac over about nine years. So now we have a device which tracks the Moon around the Earth, and its phases and it’s vari­able speed and vari­ations in that vari­ab­il­ity, while also keep­ing track of the Sun’s pos­i­tion, poten­tial lunar and solar eclipses and inter­cal­a­tion cycles so you know when to stick an extra month in to keep the lunar months in step with the solar year round gears, some moun­ted slightly off axis to cre­ate a pseudo-sinusoidal vari­ation using cir­cu­lar gears to replace ellipses. If you have funny feel­ing near the back of your head right now, that’s prob­ably your brain try­ing to crawl out of your ears. The Anti­kythera Mech­an­ism is insanely com­plex. Still just because it’s insanely com­plex, that doesn’t make it sci­entific.

In fact you can argue about whether or not Sci­ence exis­ted in the ancient world. Cer­tainly a lot of ele­ments like test­ing ideas with exper­i­ments didn’t really become pop­u­lar till after Galileo. On the other hand some nat­ural philo­sophy of the time was based on obser­va­tion. There was cer­tainly tech­no­logy which was the res­ult of applied know­ledge. With those kind of pro­visos a lot of ancient his­tor­i­ans would be happy with the idea of ancient sci­ence, albeit a sci­ence dif­fer­ent to post-Renaissance sci­ence. In this case, the sheer intense obser­va­tion and cal­cu­la­tion involved in mak­ing the Anti­kythera Mech­an­ism marks it out as a work of ancient sci­ence. There’s also another factor which might make it more sci­entific than artistic.

To some extent the Anti­kythera Mech­an­ism Research Pro­ject have been inter­ested in hanging a name on the device. It was thought to have ori­gin­ated in Rhodes and sunk on its way to Rome, which would have con­nec­ted it to the home city of Hip­par­chus, one of the great astro­nomers of antiquity. The 2008 paper has examined the parapegma on the mech­an­ism and dis­covered it may be con­nec­ted to Syra­cuse, home of Archimedes.

A parapegma is a cal­en­dar, usu­ally with holes for stick­ing a peg into for mark­ing the days. In the case of ancient Greece they’re inter­est­ing when they tell you what day of the month it is, because each Greek city had its own set of months. The months were usu­ally named after reli­gious fest­ivals, and this was tied into local polit­ics. That meant hav­ing your own cal­en­dar was a good way of show­ing your inde­pend­ence. The best match for the months men­tioned on the mech­an­ism is Taur­omenion, mod­ern Taorm­ina, in Sicily. This is likely to have shared some months with Syra­cuse as it was re-settled from there in the fourth-century BC, so Syra­cuse is a strong pos­sib­il­ity for the home of this device. Archimedes is said to have inven­ted a plan­et­arium accord­ing to Cicero and is thought to have writ­ten a lost book on astro­nom­ical devices. How­ever he could not have made this device. Archimedes died in 212 BC. The Anti­kythera Mech­an­ism is cur­rently thought to date to the second half of the second cen­tury BC, but that might change. But it was very likely to have been made after Archimedes death and that’s what makes it scientific.

Art can be col­lab­or­at­ive, or it can be per­sonal. Sci­ence in con­trast is built on cumu­lat­ive know­ledge. The per­son who inven­ted the gear­ing did not have to be the per­son who made the astro­nom­ical obser­va­tions. He didn’t even need to live in the same cen­tury as the astro­nomer. In fact the maker of this device might not have done either. He could have fol­lowed a kit and added his own per­sonal touches on the cas­ing. There’s a core to this device which, once expressed, is inde­pend­ent of per­sonal vis­ion. Archimedes didn’t have his own per­sonal Moon which moved in a dif­fer­ent way to every­one else’s, while an artist can have a per­sonal inter­pret­a­tion of the Moon.

A reason people might think the Anti­kythera Mech­an­ism is a work of art is that it’s clearly the res­ult of a lot of ima­gin­a­tion. Great art requires ima­gin­a­tion, but so too does great sci­ence. It requires the kind of ima­gin­a­tion that can look at a tool­box full of circles and see ellipses. The kind of ima­gin­a­tion that can watch wheels turn within wheels as bod­ies waltz to the music of the celes­tial spheres. Another com­mon factor between art and sci­ence is that great art can show a new way of look­ing at the world, and great sci­ence does this too. That’s why I dis­agree with Candy Minx when she says “Sci­ence is always play­ing catch up with the poets.” Sci­ence can reveal beauty too, as a visit to the Anti­kythera Mech­an­ism Research Group’s homepage would show.

Freeth, T., Bit­sa­kis, Y., Mous­sas, X., Seirada­kis, J., Tse­li­kas, A., Mangou, H., Zafeir­o­poulou, M., Had­land, R., Bate, D., Ram­sey, A., Allen, M., Craw­ley, A., Hockley, P., Malzbender, T., Gelb, D., Ambrisco, W., & Edmunds, M. (2006). Decod­ing the ancient Greek astro­nom­ical cal­cu­lator known as the Anti­kythera Mech­an­ism Nature, 444 (7119), 587–591 DOI: 10.1038/nature05357

Freeth, T., Jones, A., Steele, J., & Bit­sa­kis, Y. (2008). Cal­en­dars with Olympiad dis­play and eclipse pre­dic­tion on the Anti­kythera Mech­an­ism Nature, 454 (7204), 614–617 DOI: 10.1038/nature07130