by Graham Robb
For my sister Alison
Contents
Protohistory
PART ONE
1. The Road from the Ends of the Earth
2. News of the Iron Age
3. The Mediolanum Mystery, I
4. The Mediolanum Mystery, II
5. Down the Meridian
PART TWO
6. The Size of the World
7. The Druidic Syllabus, I: Elementary
8. The Druidic Syllabus, II: Advanced
PART THREE
9. Paths of the Gods
10. The Forest and Beyond
11. Cities of Middle Earth
12. The Gods Victorious
PART FOUR
13. The Poetic Isles
14. The Four Royal Roads
15. The End of Middle Earth
16. Return of the Druids
Epilogue: A Traveller’s Guide to Middle Earth
Chronology
Notes
Works Cited
General Index
Geographical Index
Acknowledgements
Permissions Acknowledgements
Protohistory
The idea that became this book arrived one evening like an unwanted visitor. It clearly expected to stay for a long time, and I knew that its presence in my home would be extremely compromising. Treasure maps and secret paths belong to childhood. An adult scholar who sees an undiscovered ancient world reveal itself, complete with charts, instruction manual and guidebook, is bound to question the functioning of his mental equipment.
I was living at the time in a thatched cottage on a hill to the west of Oxford, in the land of Matthew Arnold’s Scholar Gypsy. It was the setting of a child’s fantasy world, the kind of place where historical secrets seem to offer themselves up like apples from a prolific, untended tree. In the garden, along with the remains of Victorian picnics and the luminous, degraded waste of the twentieth century, there were burnt flints, pieces of smelted metal and crudely fired roof-tile. Under the tangle of a spindle tree, I found a small Iron Age brooch with the corroded remains of its fastening pin and a pattern of three concentric circles. The farmer who lived next door showed me cardboard boxes full of quern stones used for grinding, Samian ware and Roman coins unearthed by his plough.
Though no book mentions it, this quiet place had once been a busy junction. Along one side of the garden, a bridleway was the end of an ancient straet that followed the limestone escarpment above the Thames and led to the Berkshire Downs and the Iron Age hill figure known as the Uffington White Horse. On the other side, an unpaved road climbed from a crossing of the Thames called Bablock Hythe, where there is now neither a bridge nor a ford. For many centuries, long before there was a place called Oxford, cattle and sheep had been driven up from the river to the point where the two roads met and broadened into a green with a spring and a pond. The contours of this ancestor of Cumnor village, which lies a twelve-minute walk from the later, medieval centre near the church, were masked by the more recent alignments of garden hedges and parking bays, but it was still possible to make out the form of the earlier village from an upstairs window. These patterns of settlement are too old to appear in documents. An abrupt bank of earth, and the remnant of a circular arrangement that gives the modern road a dangerous curve, are the probable vestiges of an Iron Age hill fort that overlooked the prehistoric settlement on the floodplain of the Thames at Farmoor.
In these conditions, it would not have been surprising if some chronic state of historical hallucination had taken hold. The irresistible visitor in the upstairs room took the form of a diagonal line on a map of western Europe, printed out on two large sheets of paper. I had been planning a cycling expedition along the Via Heraklea, the fabled route of Hercules from the ends of the earth – the ‘Sacred Promontory’ at the south-western tip of the Iberian Peninsula – across the Pyrenees and the plains of Provence towards the white curtain of the Alps (fig. 1). The hero’s journey, with a herd of stolen cattle, is the legendary trace of one of the oldest routes in the Western world. For much of its course, it exists only as an abstraction, an ideal trajectory joining various sites that were associated with Hercules. Where it skirts the Mediterranean in southern France, on certain stretches it turns into a track, sometimes following the trails of transhumant animals, sometimes materialized as the Roman Via Domitia and the modern A9 autoroute. These practical, secular routes run up the eastern coast of Spain, turn north-east in France, and eventually wander off towards Italy. But in its original, mythic incarnation, the Via Heraklea marches in a straight line like the son of a god for whom a mountain was a paltry obstacle.
Two things struck me about this transcontinental diagonal. First, if the surviving sections are projected in both directions, the Heraklean Way follows the same east-north-easterly bearing for a thousand miles, and arrives, precisely, at the Alpine pass of Montgenèvre, which the Celts1 called Matrona (the ‘Spring of the Mother-Goddesses’). This is the pass that Hercules is supposed to have created by smashing his way through the rock. It was as though, when he set off from the Sacred Promontory, he had been carrying some ancient positioning device which told him that he would cross the Alps at that exact point.
The second glimmer of something remarkable was the familiar appearance of the trajectory on the printout. Many ancient cultures – including the Celts, the Etruscans and, occasionally, the Romans – angled their temples, tombs and streets so that they either faced or stood in a geometrical relationship to the rising sun of the solstice. This is one of the two times of the year (around 21 June and 21 December) when the sun rises and sets in almost exactly the same place several days in a row. I consulted the online oracle of astronomical data: two thousand years ago, at a Mediterranean latitude, the trajectory of the Via Heraklea was the angle of the rising sun at the summer solstice – or, if the observer were facing the other way, of the setting sun at the winter solstice.2
This cosmic coincidence has remained entirely undetected. Perhaps its grandeur renders it effectively invisible. Even the least sceptical historian would doubt its existence. Yet it led to so many other verifiable discoveries that it seemed at times to have a mind of its own, a mechanism from another world, accidentally reactivated. The journeys it entailed took several years, but it became apparent within the first few months that the solstice path had been deliberately created. Druids and Druidesses – the priests or scientists of the Celts – standing anywhere on that ancient path, knew that when they looked along it to the west, they were looking towards the ends of the earth, beyond which there was nothing but the monster-haunted ocean and the land of the dead. In the other direction, they were facing the Alps and the Matrona Pass through which the sun returned to the world of the living.
Gradually, a third coincidence added itself to the picture. A few years before, while researching The Discovery of France, I had read about an enigmatic name – Mediolanum – which the ancient Celts had given to about sixty locations between Britain and the Black Sea. It meant something like ‘sanctuary’ or ‘sacred enclosure’ of ‘the centre’ or ‘middle’. The word ‘Mediolanum’ is thought to have been related to a notion that is also found in other mythologies: the idea that our world is a Middle Earth whose sacred sites correspond to places in the upper and lower worlds. In Norse and Germanic mythology, the word is ‘Midgard’, from which J. R. R. Tolkien derived the name of the fictional universe of The Hobbit and The Lord of the Rings.
In 1974, a professor of literature called Yves Vadé suggested that the Celts had organized these ‘middle places’ according to a network, and that they had been arranged so that each one was equidistant from two others. The idea was taken up in 1994 by the Professor of Geography at the Sorbonne, Xavier de Planh
ol, who concluded that the network might briefly have served some practical or religious purpose but that it had been abandoned at an early stage. A random scattering of dots would produce similar results, and there were other problems with the idea, which will be mentioned later in this book. The name was intriguing all the same, especially since, after retracing the etymologies of all the Gaulish3 place names along the route, I found, on or near the line of the Heraklean Way, six places that had once been called Mediolanum.
From then on, ‘coincidences’ cropped up with surprising frequency. A complex, beautiful pattern of lines emerged, based on solar alignments and elementary Euclidean geometry. I began to see, as though in some miraculously preserved document, the ancient birth of modern Europe. The places called Mediolanum had belonged to an early and comparatively chaotic stage of this mapping of a continent. Out of that fertile confusion of local systems, a vast network had evolved. The geography of the Western world had been organized into a grid of ‘solstice lines’, based on the original Via Heraklea, with precisely measured parallels and meridians determining the locations of temples, towns and battles. At an even later stage, in Gaul, and, more spectacularly, in the British Isles, some long-distance roads had been constructed as literal incarnations of the solstice lines. Knowledge of that grid had been lost in the bustle and belligerence of the Roman empire, and it often seemed as though that wonder of the ancient world had never existed at all.
For several months, I followed the Heraklean Way and the other lines on the map, or rather, scrolled along them repeatedly at what must eventually have been the average speed of a real mouse covering the same itinerary on the ground. Ten years ago, without digital maps and mapping software, this would not have been possible – which is one answer to the obvious question, ‘Why has no one thought of this before?’ Something similar might have been attempted using paper maps, but it would have demanded a team of trained assistants and a desk the size of an aircraft hangar. (It would also have required an aircraft.) The results of this virtual expedition – and of the real expeditions they inspired – form the first part of this book. Readers who have the patience to pursue this journey through its rubble-strewn suburbs will become familiar with the map’s peculiarities, and they are hereby invited to ‘cheat’ by looking ahead at a few examples of its evolution (here, here).
The implications were too extraordinary to be ignored or, for that matter, believed: apart from a solar-lunar calendar on a plate of bronze that was found near a lake in the Jura – and which was partially decoded only with the help of computers – this was the first mathematically provable evidence of Druidic science and its achievements. It was, in effect, the earliest accurate map of the world. From somewhere beyond the Alps, this transcontinental masterpiece of sacred geography seemed to extend as far as the British Isles, and perhaps even further, to the remote northern islands seen in the fourth century BC by the explorer Pytheas of Marseille, where the ocean, heaving like a lung, becomes indistinguishable from the sky.
Stare at a series of lines on a map, and eventually a pattern will appear as surely as a human destiny in a fortune-teller’s teacup. In any scholarly endeavour, excitement is a false friend: the more thrilling a theory, the more the theorist wants it to be true. For several more months, I tried to disprove it. I forsook the magical shadows of the thatched cottage for the energy-saving gloom of modern libraries. Time spent in another world is never wasted, and so, even if the theory had been hammered into oblivion by historical and archaeological fact, it would still have been a fruitful disappointment. But the more I tried to disprove it, the more evidence emerged. In October 2009, I read about the archaeological discovery, in a cement quarry near Lausanne, of the biggest Celtic sanctuary ever found in Switzerland. It was on a hill called Mormont near a place called Eclépens. I looked at the embryonic map of Druidic paths on which lines of longitude and latitude are bisected by mirror images of the Heraklean Way: Mormont lies on one of the long-distance lines and close to a major intersection.
At this point, I prepared a verbal synopsis and presented the idea to my publishers. Two meetings took place – one in an underground room in a quiet mews near the Portobello Road in London, the other in a venerable gentlemen’s club in mid-town Manhattan which the company president assured me had been checked for ear-trumpets and other listening devices. I described the discovery, and swore to secrecy people whose vocation it is to make things public. My thought was not that someone would steal the idea, which, in any case, is not quite as portable as a mathematical equation or a magic spell: I was thinking of friends and acquaintances in various university departments who, if the project were revealed, would be forced to pretend that nothing was amiss.
Anyone who writes about Druids and mysteriously coordinated landscapes, or who claims to have located the intersections of the solar paths of Middle Earth in a particular field, street, railway station or cement quarry, must expect to be treated with suspicion. In its simplest form, the idea was reminiscent of ‘ley lines’, and I was uncomfortably aware of the fact that a sarcastic trick of fate had sent me to live in a house called Leys Cottage. ‘Ley lines’ were discovered – some say, invented – in 1921 by an amateur archaeologist, Alfred Watkins. His idea was that alignments of prehistoric and other ‘old’ sites are remnants of ‘the Old Straight Track’ followed by Neolithic traders. Watkins believed that they had actually been called ‘ley lines’ because the word (a common place name meaning ‘meadow’ or ‘pasture’) appears on so many of them. His research consisted, in part, of stamping on the ground to detect the hollowness of ancient burial sites. Although he fostered a new sensitivity to ancient configurations of the English landscape, and created the delightful historical pastime of ley-line hunting, his muddling of different eras is anathema to archaeologists and historians. Yet ninety years of increasingly sophisticated prospecting and excavating have shown his original idea to be perfectly plausible: carefully aligned, long-distance paths were well within the capabilities of Neolithic people.
The period covered by this book (roughly 800 BC to AD 600) begins almost a thousand years after the end of the Neolithic Age (c. 1700 BC). The cultures known as Celtic belong to the forbiddingly named Iron Age, which was also the age of precision instruments, high-speed transport, crop rotation and land management, intellectual education of the young and the first European towns. Some archaeologists now place the dawn of the Celtic world a few centuries earlier, in the late Bronze Age. Both ages come under the heading of ‘prehistory’. The term is broadly applied to any pre-literate period of humanity; it can also be applied to any period between the first microbial stirrings of the primordial slime and the civilized world that grew up only sixty generations ago. In Britain, the clock of ‘history’ does not begin to tick until ten o’clock in the morning on a summer’s day in 55 BC, when Julius Caesar anchored off the Kentish coast. The following year, he returned with water-clocks, measured the length of the English summer day (longer than on the Continent), and brought prehistory to an end, at least in the south of England.
To a French archaeologist, the ancient Celts are not ‘prehistoric’ but ‘protohistoric’. They are not quite our visible and audible neighbours, but nor are they the nameless, faceless shadows who built Stonehenge. To the Celts, Stonehenge was a mysterious ancient monument. Their writings were not published in Rome and catalogued in the Library of Alexandria, but we know about their lives, customs, beliefs, fashions and diets from Greek and Roman travellers. Some of their myths and legends, preserved in verse and memorized by successive generations of bards, were recorded by foreign writers. The Druids banned all written expression of their wisdom, but their society was certainly literate since writing implements have been found all over the Celtic world. For a dead language, ancient Gaulish is surprisingly lively: inscriptions on plates, pots, coins and curse-tablets are being unearthed all the time, and the lexicon of the language that was practically extinct by the sixth century AD continues to grow like hair on
a corpse.
Some of the protohistoric inhabitants of Europe are known to us by name: Vercingetorix, the son of an executed tyrant and leader of the Gaulish resistance; Diviciacus, the Druid scholar and diplomat who stayed in Rome with Cicero and addressed the Roman Senate; Cartimandua, the queen of the British Brigantes who collaborated with the Romans. We know the names of many of their towns and what they looked like. The first time-travelling archaeologist to return to the Iron Age will be sufficiently well equipped to pass for an ancient Celt, albeit a semi-literate Celt with a small, disproportionately obscene vocabulary.
Despite the staggering quantity of information amassed by archaeologists in what must be the greatest collective endeavour in the history of scholarship, the forgotten world of the Celts looks like the land that was never there to be forgotten. In France, the Roman past is everywhere. On some dirt-track sections of the Heraklean Way, it scrunches under the bicycle tyres; on the crumbling perimeters of hill forts, it lies among the litter. In the early morning rush-hour, on the Place Bellecour in the centre of Lyon – formerly known as Lugdunum – I sat down on a concrete bench. The ground had been disturbed by diggers and roughly levelled in preparation for a new surface. Some orange-coloured sherds stood out against the imported red sand. I bent down and picked up five small fragments of Roman pottery, one of which bore the ribbed pattern of a wine-cup similar to those displayed in the nearby Musée de la Civilisation Gallo-Romaine. To my left, through the avenue of lime trees, I could see the basilica of Notre-Dame de Fourvière, high above the river Saône, which readers of Caesar’s Gallic War can easily distinguish from the neighbouring Rhone: ‘There is a river called the Arar’, wrote Caesar, ‘which flows through the lands of the Aedui and the Sequani and into the Rhodanus with such incredible slowness that the eye cannot tell in which direction it is flowing’. The basilica occupies the presumed site of the Gaulish oppidum or hill fort,4 but so little survives of the pre-Roman city that no one knows where the first inhabitants of Lugdunum lived.