by Graham Robb
Now that the sky is just the unreliable backdrop of the daily human drama, and the sky gods’ only priests are weathermen, knowledge of celestial trajectories seems almost esoteric. Most people are aware of cardinal points only because of a prevailing wind, a sunny breakfast-room or a damp north-facing wall. Ancient Celts always knew exactly where they stood in relation to the universe. A cognitive psychologist has discovered that speakers of languages in which the same words are used for immediate directions (‘right’, ‘left’, ‘behind’, etc.) and for points of the compass tend to be aware of their orientation, even indoors and in unfamiliar surroundings. Gaulish was one of those languages. ‘Are’ meant ‘in front of’; it also meant ‘in the east’. ‘Dexsuo’ was ‘behind’ and ‘in the west’. To head north, one turned to the left (‘teuto’), and, to head south, to the right (‘dheas’).6 In English, ‘right’ and ‘left’ are relative to the speaker’s position; in Gaulish, directions were absolute and universal. If a Celtic hostess told her guest that the imported Falernian wine was in the krater behind him, she meant that it lay in the same direction as the setting sun. If he went out hunting, he would have known immediately where to point his spear when his host said, ‘the boar is coming at you from the south’. And when he left the oppidum and rejoined the Heraklean Way, he would have known that he was travelling towards the rising sun of the summer solstice.
The summer and winter solstices were crucial points of reference for ancient civilizations. The complexities of measuring solstice angles will be mentioned later on;7 the principle itself is simple. In the course of a year, because of the tilt of the earth’s axis, the sun rises and sets in different parts of the sky. Around 21 June, it rises on the north-eastern horizon at what appears to be the same point for several days in a row – hence the term ‘solstice’ (the ‘stand-still of the sun’). The summer solstice itself occurs on the longest day of the year. From then on, the sun rises progressively further south, until the winter solstice, which occurs on the shortest day of the year. Halfway between the two solstices, the sun rises due east and sets due west. These two days are the equinoxes, when the night (‘nox’) is roughly equal (‘aequus’) in length to the day.
According to popular wisdom, the solstice was the object of an absurd superstition. Ancient people are supposed to have seen the sun rising and setting ever further to the north or south and to have concluded that without a good deal of prayer, procession and bloody sacrifice, it would either get stuck in the same place – with disastrous consequences for agriculture – or, worse, continue in the same direction until it disappeared for ever. This would mean that there was once a civilization that was capable of building enormous, astronomically aligned stone temples and yet was otherwise so impervious to experience that it had to renew its knowledge of the universe every six months. The solstice may have been a time of ritual celebration or mourning, but it was also an obvious and useful reality. More accurate bearings can be taken during the solstice than at other times of the year, and since the sun rises and sets at almost exactly the same point for over a week (within a range of 0.04°), a day of cloud and mist is less likely to spoil the operation.
2. Solstice angles
The angles shown here are the solar azimuth angles (see note on this page) in 1600 BC at Stonehenge, latitude 51.18° north (dotted lines) and the Pillars of Hercules, latitude 36.00° north (unbroken lines).
The purpose of those measurements was both scientific and religious. The paths of heavenly bodies revealed the workings of the universe and the designs of the gods. The Celts’ trading partners, the Etruscans, used solstice measurements to align their towns on the cardinal points. In this way, the whole town became a template of the upper world. ‘Superstition’ lay in the fact that the town’s skyscape, too, was divided into quadrants for the interpretation of celestial signs (stars, bolts of lightning and flocks of birds): north-east – the approximate trajectory of the Heraklean Way and the summer solstice dawn – was the most auspicious quadrant; south-east was less auspicious; south-west was unlucky, and north-west extremely ominous. Since north lay to the ‘sinister’ left, and since the sun’s light dies in the west, the system had a certain psychogeographical logic.
These alignments were part of the common fund of knowledge throughout the ancient world. The Celts often built their sanctuaries so that the summer solstice sun would shine through the eastern entrance onto the altar. Though the Romans sneered at superstitious barbarians, their own surveyors used the solstice as a point of reference. No surviving Roman text mentions the astronomical significance of the Heraklean Way, and the secret seems to have been lost for almost two thousand years. But to the ancient Celts, the meaning of its alignment would have been as obvious as though it was explained on a roadside information panel.
The name itself was explanatory. To Celts and Carthaginians, ‘Via Heraklea’ would have meant, in effect, ‘Path of the Sun’. Like his Carthaginian equivalent, Melqart, Herakles was also a sun god. His twelve labours were equated with the twelve constellations of the zodiac through which the sun passes in the course of a year. The face of his Celtic equivalent, Ogmios, had been scorched by the blazing vehicle that carried him across the sky, and the wheel he held in his hand – like the thousands of pocket-sized votive wheels that are still being found in Celtic sanctuaries – was a symbol of the sun. The eight spokes are thought to represent the cardinal points and the rising and setting of the sun at the summer and winter solstices. And because the sun was the ultimate authority in matters of measurement, Melqart or Herakles was also the divine geometer: ‘in his devotion to wisdom’, said Philostratus the Athenian, ‘he measured the whole earth from end to end’.
Herakles’ solar wheel evidently served him well as a global positioning device. The solstice line of the Heraklean Way is astonishingly accurate. From Andorra (the cols of Muntaner and Ordino) to the Matrona Pass (almost five hundred kilometres), it has a bearing of 57.53° east of north. This is the angle of the rising sun of the summer solstice at a point roughly halfway from the Sacred Promontory to the Matrona Pass.8 Holding to this bearing, it runs through or close to eight tribal centres, including Andorra, Narbonne, Nîmes and Briançon; it also passes by six of the mysterious places once called Mediolanum, of which much more will be said. Over its entire length (almost 1600 kilometres from the Sacred Promontory to the Matrona Pass), the bearing is 56.28°, which, at that enormous distance, is so precise as to be incredible.
Even more surprising, in view of its accuracy, is its great age. One of the earliest surviving references to the Heraklean Way dates from the third century BC. It was mentioned by the anonymous author – known as ‘the Pseudo-Aristotle’ – of De mirabilibus auscultationibus (‘Wonderful Things I Have Heard’):
From Italy as far as the country of the Celts, Celtoligurians and Iberians, they say there is a road called the ‘Road of Herakles’, and on this road, the traveller, whether native or Greek, is watched by the neighbouring tribes so that he may receive no injury; for those amongst whom the injury has been done must pay a penalty.
By the time the Pseudo-Aristotle heard of it, the wonderful road had already seen several generations of travellers. Tracks are hard to date, but filtering techniques applied to aerial photography have made the landscapes of ancient Gaul bloom with historical clues. When Greek traders first sailed towards the setting sun and reached lands that even Odysseus had never seen, they set up trading posts called emporia along the coasts of the Gaulish and Iberian Mediterranean. To supply the new ports, they purchased land from local tribes, and then divided up the territory into squares of equal size. The process and its result are known as centuriation. Some of the early Greek centuriations have been plotted in great detail: on a map, they look like the gridlines of American cities covering many hundreds of square kilometres. One of the oldest is the centuriation of Agatha (Agde), which was a colony of the Greek port of Massalia (Marseille), founded in the fifth century BC. For a distance of twenty-five kilometres,
the Heraklean Way follows the diagonal line that marks the northern limit of the Agde centuriation. Since centuriations invariably adopted the trajectories of existing roads, this section of the Heraklean Way at least must date back to the very earliest days of Graeco-Celtic cooperation.
These abstract measurements and trajectories are the comprehensible whisperings of a vanished civilization. For a modern traveller, the physical inconvenience itself is a sign of ancient mathematical expertise. The old trail keeps to its cosmic bearing whenever it can. Later tracks and roads, including some stretches of the Roman Via Domitia that replaced parts of the Heraklean Way,9 are more obsequious to the landscape – they curve conveniently around the slopes and sidle up to villages along the valleys – but the dust-blown Heraklean Way strides over the hills like a heedless athlete. This is the material form of history, the tangible proof that ancient truths are still recoverable. The accuracy of the Heraklean Way is directly related to the accuracy of Celtic legends. Astronomical observations, spanning many years, made it possible not only to project a straight line across the landscape but also (for example, by recording solar eclipses) to attain the kind of chronological accuracy that could date the foundation of an Irish hill fort to 668 BC or the foundation of Massalia to the beginning of the sixth century BC.
Perhaps it was then, in the early 500s, that the tribes of the hinterland encountered the skills and technology that enabled them to trace the sun god’s path on earth. Or perhaps a human Herakles from Greece had journeyed to the west as a sailor in search of land and a foreign princess. Even at this great distance in time, myths begin to resolve themselves into legends, and in those legends, historical figures can be glimpsed. The two surviving versions of the story of Massalia’s foundation place the day in question – for the convenience of Roman readers – during the reign of the Roman king Tarquinius Priscus (c. 616–579 BC). The original Celtic versions were almost certainly more precise.
On the afternoon in question, a trading fleet from the distant Greek city of Phocaea dropped anchor in what became the harbour of Massalia, the first city in Gaul. The chief of the local tribe, the Segobriges, was about to hold a banquet at which his daughter was to choose a husband. According to tradition, she was to indicate her choice by offering the future bridegroom a cup of pure water. The Greek captains were invited to the ceremony. Seeing the dark-eyed adventurers who had braved the trackless sea, and having perhaps previously inspected their treasures – the painted vases, the bronze flagons and tasting spoons, and the ships like a god’s chariots resting in the harbour – she offered the cup of water to one of the fearless navigators.
The couple were married, and, like Herakles and Celtine, a Greek hero and a Celtic princess embodied the happy truth that the origins of a powerful and stable confederation lay, not on the battlefield, but on a long-distance trading route that led unerringly to a woman’s bed. A colony was founded at what is now the Vieux Port of Marseille. Not long after, Greek wares and locally produced Greek wine were being carried up into remote parts of Gaul from the mouth of the Rhone – where a city called Heraklea once stood – and along the Heraklean Way, where Greek pottery of that period is still being dug out of the rubble of hill forts and cemeteries.
The road from the ends of the earth was the beginning of one of the great adventures and inventions of the ancient world. So much about it is improbable – its length, its accuracy, its antiquity: it might all be attributed to the god of chance were it not for the records of a journey made along that trajectory by a real, mortal human being. One of the two accounts – by the Greek historian Polybius – was based on research trips and interviews with people who had witnessed Hannibal’s expedition fifty years before; the other, by the Roman historian Livy, used some of the same reports, including a lost account in seven volumes by one of the traveller’s companions. Neither historian recognized the transcendent significance of the route.
The journey took place in 218 BC, when the Heraklean Way was already steeped in four centuries of myth and legend – which is partly why it was chosen as the route of the expedition. In the late spring, ninety thousand foot soldiers, twelve thousand cavalry and thirty-seven elephants set off from New Carthage (Cartagena, on the south-east coast of Spain). The Carthaginians’ empire had spread from the coasts of North Africa into Iberia, and their only rivals in the Mediterranean were the Romans. The young Carthaginian general, Hannibal, was to march from Iberia, across the Pyrenees and then the Alps, to attack the Romans from the north. Alliances had been formed with Gaulish tribes who had colonized the plains on the Italian side of the Alps; other tribes had been wooed with promises of gold; many of the soldiers on the expedition were Celts from Iberia or Gaul. The city of Rome, according to Livy, was ‘on tiptoe in expectation of war’.
3. Place names and tribal names of Celtic origin in Iberia
At some point, Hannibal – or part of his army – joined the line of the Heraklean Way. It was an obvious route: the origins of Iberian place names suggest an ancient frontier running diagonally between Celtic or Celtiberian tribes to the north-west and the indigenous inhabitants to the south-east. This frontier closely matches the Heraklean line, and it leads to what must have seemed an unnecessarily arduous crossing of the Pyrenees. The four tribes that were said by Polybius to have been defeated by Hannibal lived, not on the coast, where a Roman customs post still marks the relatively easy crossing at the Col de Panissars, but in the central Pyrenees, in the region of Andorra, where the Heraklean line crosses the watershed (fig. 1). In warm spring weather, the ascent is not as difficult as it appears on a map, and its remoteness would have had the advantage of delaying news of the Carthaginian invasion.
In the context of the entire expedition, the diagonal route was logical and, more importantly, auspicious. Hannibal had already proved himself a brilliant tactician, and part of his brilliance lay in his ability to assume the role of a god. Before leaving Iberia, he had visited the temple of Melqart-Herakles at Gades (Cadiz), to consult the oracle and to ask for the god’s protection. He had wintered part of his army in the harbour of Portus Hannibalis near the Sacred Promontory, where another famous temple of Melqart-Herakles stood on the edge of the inhabited world. Ancient writers who described the Carthaginian invasion knew that Hannibal saw himself and wanted to be seen as the successor of Herakles. He would march across the mountains in the footsteps of the sun god, shining with the aura of divine approval.
Like most of his contemporaries, Polybius had only the vaguest notions of European geography beyond the Mediterranean. He knew about Hannibal’s Herculean ambitions but not about the Heraklean Way. According to his source, ‘a hero [Herakles] showed Hannibal the way’. Polybius – and Hannibal’s later historians – took this to be a rhetorical flourish rather than an indication of the route. Similarly, Livy mistook the key to the whole expedition for a picturesque embellishment and used it in a speech that Hannibal was supposed to have made to his troops when they were quailing at the thought of crossing the Alps: he reminded them that, earlier in the expedition, ‘the way had seemed long to no one, though they were pursuing it from the setting to the rising of the sun’. This was not a figure of speech but the cosmological truth.
There were no maps or atlases on which Polybius could have traced a plausible route, and he thought that foreign place names would only confuse his readers with ‘unintelligible and meaningless sounds’. Fortunately, the distances that he copied from his source make it possible to calculate the point at which Hannibal and his elephants crossed the Rhone – ‘four days’ march from the sea’, then a further ‘two hundred stadia’ (about thirty-five kilometres) north, to a place where ‘the stream is divided by a small island’. Most historians now identify this as Roquemaure, near Châteauneuf-du-Pape, which is certainly correct, since it also happens to be the point at which the line of the Heraklean Way crosses the Rhone.
The next part of the journey is confused and nonsensical in both accounts. The geographical explanations
given in the lost source seem to have been treated by Polybius and Livy as a description of Hannibal’s actual route. All we know is that the itinerary was somehow related to the distant source of the Rhone. After crossing the river, Hannibal ‘marched up the bank away from the sea in an easterly direction [in fact, at this point, the Rhone flows north to south], as though making for the central district of Europe’. Polybius went on to explain, more accurately, that ‘the Rhone rises to the north-west of the Adriatic Gulf on the northern slopes of the Alps’.
The details that Polybius and Livy preserved in a muddled form are like barely decipherable remnants of an ancient map. The allusion to the source of the Rhone and ‘the central district of Europe’ may be the trace of an ancient system of orientation that enabled an army or a merchant to plot a course across half a continent. Parts of this ‘map’ will be pieced together in the first part of this book. The crucial point is that only a god could have walked in a straight line all the way to the Matrona; even transhumant herds were forced to take a more winding route towards the Alpine pastures and to skirt the north side of Mont Ventoux. But after bypassing the mountainous terrain, it was vital to be able to continue on the same auspicious bearing as before. Whichever route he chose after crossing the Rhone (Livy suggests that Hannibal followed the river Druentia or Durance), he would have rejoined the Heraklean Way as soon as he could – perhaps near the oppidum at Serrela-Croix, where the road once again follows the solstice line for twelve kilometres – and then, climbing towards the source of the Durance, to Brigantium (Briançon) and the final approach to the Matrona Pass.