In a treatise on the nature of the gods, written around 45 bce, Cicero formulated a version of what became known as the teleological proof of the existence of God:
When you see a statue or a painting, you recognise the exercise of art; when you observe from a distance the course of a ship, you do not hesitate to assume that its motion is guided by reason and by art; when you look at a sundial or a water clock, you infer that it tells the time by art and not by chance; how then can it be consistent to suppose that the world, which includes both the works of art in question, the artists who made them, and everything else besides, can be devoid of purpose and reason? Suppose a traveller carried into Scythia or Britain the sphere recently constructed by our friend Posidonius, which at each revolution reproduces the same motions of the sun, the moon and the five planets that take place in the sky day and night: which of the barbarians living there would doubt that this sphere was the work of a rational being?
A year later, Julius Caesar was assassinated and the res publica plunged into renewed chaos, after decades of intermittent civil war. In the midst of these storms, Cicero’s belief that man-made models of the universe, like Posidonius’ sphere or the Antikythera mechanism, testified to the presence of reason and purpose (consilium) in nature – showing a divine mind at work – seemed almost defiant. Astronomical phenomena have been viewed as manifestations of order since very early times. The sun rises and sets every day; the moon wanes and waxes every month; the seasons, marked by winds, rains or sunshine, return every year. Everyone living around the Mediterranean in antiquity was aware of these cycles. Those who worked the land followed the rhythms of planting and harvesting; religious festivals, athletic competitions and market days were organised in recurring cycles. In Republican Rome, consuls were elected and changed every year, and the lists recording their sequence provided the equivalent of a dating system. Later on, the same function was performed by the succession of emperors.
Different regions kept time differently. Travelling sometimes meant shifting to a different set of cycles and time divisions – a different calendar. In the classical Greek world the year was pretty much universally divided into 12 months, but the months’ names, their length, and the number and arrangement of intercalary days, added to a year to keep the calendar in step with astronomical phenomena, varied from one region to another.
This lack of standardisation has been seen as an example of Greek individualism. While more or less homogeneous culturally, institutionally and politically, the Greek world of the fifth and fourth centuries bce was divided to the point where a uniform system for keeping time could not be agreed. This chaos could be seen as a reflection of the behaviour of the heavens. Ancient astronomers knew that even the allegedly perfect revolutions of the cosmic bodies were not really all that orderly. Planets travelled in one direction, and rose and set at specific seasons, but sometimes it seemed as if they stopped and retraced their steps. The Sun seemed to travel now faster, now more slowly. Both the Sun and the Moon sometimes disappeared from the sky, and sometimes changed dramatically in colour. Celestial bodies appeared bigger or smaller.
Astronomical cycles were known to Mesopotamian, Egyptian and Greek astronomers. Devised often after centuries of observations, they enabled astronomers to predict the risings and settings of a certain planet or constellation, the length of the solar year, the time of the full moon, or the intervals of time within which an eclipse could take place. Each astronomical cycle reinforced the notion that the heavens were orderly and regular, but none of the cycles was straightforward – calculating them often required some tinkering. Regularity could be achieved only by embracing irregularity. Greek astronomers were keen to find a model that would reflect the regularities and, if possible, incorporate the irregularities of the heavenly cycles, without abandoning the basic tenets that heavenly motion was circular, that the Earth was in some sense at the centre of the universe, that the heavens were divine and perfect in nature, and, above all, that mathematics could account for all observed phenomena.
There were different kinds of ancient astronomical model: arithmetical, where a pattern was extracted from the regularity of certain phenomena over accumulated observations, and used to calculate the date of future occurrences, or the interval between them; geometrical models, displaying the observable motions of the heavenly bodies and inferring their future positions through envisaging them as points, circles and spheres; and finally three-dimensional models that represented and visualised the universe and the heavenly bodies, whether static or in motion. The philosophically meaningful objects (sundial, water clock, orrery) described by Cicero in De natura deorum belong in this category. They embodied some sort of astronomical knowledge, and thus shared some of the characteristics of the other two types of model.
Literary tradition associates these models with some of the most famous scientists of antiquity. Cicero writes of the sphere constructed by Posidonius, whose works, none of which survives, dealt with a vast range of subjects, including philosophy, history, natural philosophy and geography, as well as mathematics and astronomy. Cicero was writing at a time when the relationship between Greek and Roman culture was being redefined, as Rome expanded east and southwards, and his declaration of friendship towards Posidonius, who lived on Rhodes, constitutes an astute appropriation of Greek scientific culture to the Roman side, creating a united, civilised front against the barbarians of Scythia (the area around the Caspian Sea) and Britain, who he imagines would be stunned by the sight of a planetarium.
Elsewhere, Cicero writes about the two spheres made by another Greek, Archimedes of Syracuse. One of them is described by Cicero as ‘solid’, which could mean both that it was three-dimensional and that it wasn’t hollow – a bit like a globe. The second sphere sounds more like an orrery or a planetarium, with moving parts that showed the motions of the heavenly bodies. Cicero’s description is tantalisingly vague. Both models seem to have been acquired by the Roman general Marcellus after the siege and capture of Syracuse in 212 bce, in the course of which Archimedes was killed. The spheres were war booty, yet another example of the complexity of Roman appropriation of Greek culture. After Cicero, other Latin authors continued to use models of the cosmos to convey messages that went beyond astronomy; in a poem by Claudian, writing in the fourth century, Jupiter himself is baffled by Archimedes’ artefact: reducing the universe to something that could be held in human hands, it seems, is the ultimate act of hubris.
Celestial models didn’t only provide opportunities for philosophical discourse. Ptolemy wrote a treatise on Planetary Hypotheses, which addressed the problem of how to create models of the heavens that would reflect current astronomical knowledge. He seemed to regard them as scientific instruments and teaching aids. While not common (certainly not as common as sundials or water clocks), they must have been relatively well known; they were a bit like Fabergé eggs – few in number, heard of by many, but seen, let alone touched, by few, ingenious both in design and execution.
A box the size of a small board-game, the Antikythera mechanism had dials and inscriptions in Greek on the front and back, and a multitude of tiny bronze gears inside. Using archival sources, Alexander Jones describes in detail the retrieval of the mechanism, which was found in 1900 at the site of an ancient shipwreck off the coast of Antikythera, an island between Crete and the Peloponnese. The ship had been carrying some large statues, which initially attracted the most attention. The mechanism didn’t look impressive – a lump of metal encrusted with barnacles – but closer inspection revealed the presence of what looked like a wheel and, even more unusually, teeth. Gears. Early scholars, hampered by the difficulty in properly examining the damaged and fragmentary mechanism, struggled with the apparent disjunction between its sophistication and its likely date – at least a century before Christ. Derek de Solla Price’s article for Scientific American in 1959, and his pamphlet Gears from the Greeks, published in 1974, brought the mechanism to popular attention. As de Solla Price emphasised, complex geared mechanisms had been described in ancient mechanical treatises, but they had not previously been matched by archaeological artefacts. The mechanism was proof that the ancient Greeks were more technologically capable than previously thought.
More recently, interest in the mechanism has been prompted by the availability of more powerful imaging techniques. Thanks to the work carried out by an international team, active from around the turn of the 21st century, almost every element of the mechanism is now understood. The inscriptions on its front and back have been deciphered as far as their fragmentary state allows; the trains of gears inside the original box have been replicated; the intersecting cycles and motions have been reconstructed. Jones explains the various components of the mechanism, almost taking it apart piece by piece and holding the bits up for inspection. He provides just enough general context for readers to understand the astronomical background, while being just technical enough for them to feel that they have a grasp of the object’s complexity.
Operated by cranking a handle on its side, the Antikythera mechanism displayed, via a system of smaller and larger dials and pointers on its front and back, the motions of the Sun, the Moon and the five planets then known, as well as the yearly Egyptian calendar, the Metonic 19-year cycle, the Callippic 76-year cycle, the 223-lunar-month Mesopotamian Saros cycle for eclipse prediction, the 669-lunar-month Exeligomos cycle, also for eclipse prediction, and the four-year cycle of the six main Greek athletic contests, including the Olympic Games. The pointer for the Moon rotated so that its different phases could be shown. Both the longer inscriptions and the shorter ones on the dials provided instructions about the mechanism’s use.
Inside the box were interlocking toothed wheels and ‘trains’ (axes on which the gears rotated). Jones makes the functioning of the mechanism sound incredibly complex and yet surprisingly simple: wheels within wheels moved clockwise or anti-clockwise in combinations devised to transmit motion from the handle to all the components of the device. In order to reproduce non-uniform motion (one of the irregularities of the divinely regular universe), the mechanism may have used a gear on a slotted arm within another gear, or a pin-and-slot device, or perhaps both. The mechanism is a product not just of Greek knowledge, but of Egyptian and Mesopotamian learning, and may have been destined for a Roman household. Could someone like Hipparchus or Geminus or Posidonius, or even Archimedes, have built it? Highly unlikely. Were any of their theories incorporated into it? Maybe. The book posits an anonymous designer, whose astronomical competence was good, but not extraordinary. Citing a previous study, Jones characterises the skills required to craft it – those pin-and-slot couplings, and 188 teeth cut more or less equally into a tiny disc – as ‘well within the normal capabilities of an ancient workshop’. Perhaps, but then why didn’t such allegedly ordinary skills result in more objects like the one found at Antikythera?