What Columbus Didn’t Know

These three books constitute both landmarks and cautionary warnings in a long process that none of them addresses directly. Take Barry Cunliffe’s reconstruction of the exploratory voyage by Pytheas of Massalia (Marseille) in the late fourth century BC: this not only exposes the striking lack of direct knowledge then prevalent among Mediterranean peoples about virtually anywhere outside their own charmed climatic circle, but suggests the powerful emotional liminality that is always around to reinforce ignorance. Pytheas’ sober and specific report on Cornish tin mines, the amber trade and (probably) Iceland and the Arctic Circle got him branded as a liar by writers such as the geographer Strabo, who preferred the authority of myth, and was equally ready to dismiss Megasthenes’ first-hand descriptions of India on the same basis. Ptolemy of Alexandria (c.90-168 AD), the theoretical chapters of whose Geography we now have in a superlative new translation, was the first systematic cartographer to introduce the mapping of geographical points by precise co-ordination of parallel and meridian; but he also rejected both Eratosthenes’ near-accurate calculation of the Earth’s circumference and, worse, Aristarchus’ theory of a heliocentric universe in favour of the old (and psychologically more satisfying) geocentric worldview supported by Aristotle. Since his own achievement ensured his subsequent enshrinement as an unquestioned authority, these two cardinal errors were guaranteed an extraordinarily long shelf-life. Finally, the Barrington Atlas of the Greek and Roman World, which triumphantly incorporates all the latest developments of computerised geographic information systems (GIS), still has to struggle with human error and prejudice when reluctantly confronted with political rather than physical data, and (perhaps because of this) takes its lead, fashionably, from Fernand Braudel in privileging the latter whenever possible.

The history of cartography (and of exploration, with which it has always been bound up) is not always seen for what it is: one aspect of the slow development of human self-knowledge, as reflected by historical changes in the way we evaluate our place in the world and the universe in which we find ourselves. The implications of these changes extend far beyond geography: they permeate our social and political structures, our scientific assumptions, our theologies. Some of the problems involved were discussed recently in the LRB (1 November 2001) by Lorraine Daston. Maps as propaganda, silent censorship, symbols of power hint eloquently at the human dimensions of the process, the deceptiveness of a theoretical objectivity. Even the world seen from the outside, in the Apollo pictures taken from the Moon, small and bluish in a dark infinity, is as liable to evoke thoughts of domination and power – even, strangely, of ‘lust for material possessions’ (Denis Cosgrove) – as of metaphysics, religion or poetry. Once again, I suspect, Braudel has exerted a silent influence, like the pull of magnetic North, on the terms of contemporary discourse. On the other hand he would argue that we must always bear in mind those static or slow-changing factors in nature against which human activities must invariably be dwarfed. Or, in this case, certain slow and glacially persistent features of the human psyche.

Looked at thus, the process can be seen as one of slow intellectual and physical expansion. We begin with ourselves, our ten fingers and toes (so, perhaps, a decimal system), the immediate circle of our horizon (exactly depicted by the Babylonian Ur-map of c.600 BC). Everything must be interpreted in these restricted terms, including what lies beyond our reach or knowledge: the sky, with sun, moon and stars, the power of thunder, earthquakes, fire, subterranean cave-systems. Hence weather-gods, divine anger and guidance, heaven above and hell below. The unknown is by definition strange and dangerous, and to begin with far exceeds our area of knowledge. As knowledge increases, the unknown – slowly, and most often grudgingly – diminishes. In the European model, we gradually explore most of the Mediterranean coastline, but that still defines our world, prompting oblong world maps two or three times as wide as they are high. Liminality ensures that what lies beyond the boundaries is not only unknown and dangerous but taboo. This attitude is reinforced by early traders eager to preserve monopolies: beyond the Pillars of Hercules lie perilous seas; the Clashing Rocks crush rash explorers. Natural conservatism and the passion for authority foster an acceptance of tradition, however grotesque. Ultimately non-tenable theories, above all that of the geocentric universe, attract too great a weight of psychological and religious investment to be abandoned with impunity. But, as Galileo may or may not have whispered after his 1632 recantation, things do in the end move: for Romans the (imaginary) Rhipaean Mountains retreat further and further as exploration forces back the boundaries of the unknown; the Earth (despite authority’s best efforts) is at last reluctantly accepted as moving round the Sun rather than vice versa; and man, displaced from his physical position at the centre of the universe, spends the ensuing centuries doing his level best to get back there via science.

By the time Pytheas made his northern voyage, c.320 BC, things had not advanced all that far. As early as 600, Phoenicians had circumnavigated Africa, only to have their account discredited by Herodotus because they asserted that ‘in sailing round Libya’ – Africa – ‘they had the sun on their right hand’: ironically, the Southern hemisphere in fact provided proof positive of their claim. The idea of a spherical globe had been dreamed up by the Pythagoreans and taken over by Plato and Aristotle, but (not surprisingly) more as a theory of perfect form than as a physical concept based on observation. (Aristotle noticed that new stars appeared and old ones vanished as one travelled north or south; but had he never asked himself why ships vanished over the horizon?) The ancient notion of the world as a flattish disc, a rectangular land-mass with the Mediterranean at its centre, ringed by Ocean, still persisted, even when stuck pro forma onto a sphere, like a label glued to a tennis-ball. Herodotus, echoed by Aristotle, mocked the producers of circular world maps, and attacked the whole notion of a delimiting outer world-stream. Ocean, however, had too long been equated with to apeiron, as Anaximander called it – the primordial infinity, a monster-filled chaos lying outside the limits of the oikoumene, or inhabited world – to be discarded so easily. We cherish what we fear most: Ocean was hard to dislodge. What brought Alexander’s eastward career of conquest to a halt was his confident assurance to his men that they were about to reach Ocean’s shores. When faced by the interminable Ganges plain instead, they mutinied, and small wonder.

So Pytheas’ venture required both courage in the face of what might prove immense dangers, and a resolute rejection of the wisdom that argued against any breach of the imagined boundaries of the human world. The evidence for his achievement is patchy and ambiguous: the fragments have been collected, with a critical commentary, by C.H. Roseman in Pytheas of Massalia, On the Ocean (1994). Roseman also was the first to make the suggestion that Pytheas may have used local vessels, especially from the Orkneys. Barry Cunliffe’s short study – a more adventurous development on what he has to say about Pytheas in his massive survey Facing the Ocean: The Atlantic and Its Peoples 8000 BC-AD 1500* – builds on these texts, augments them with a thorough archaeological knowledge of the routes involved (always with a sharp eye on the navigational and astronomical co-ordinates), peppers the result with shrewd economic insights (for example, on the amber trade), and comes up with by far the most convincing account to date of this extraordinary voyage.

Cunliffe’s narrative takes Pytheas across France by the Aude-Garonne tin route, more than 2500 km shorter than the sea-haul round the Iberian peninsula. He correlates his voyager’s surprisingly accurate astronomical observations with the latitudes involved (Pytheas got the length of the British coastline right to within 3 per cent), dropping him off to take bearings at 54 degrees for the Isle of Man, 58 on Lewis, 61 in the Shetlands and 66 in Ultima Thule (all the evidence points to Iceland, against previous arguments for a trip ending in the Shetlands or the Baltic). Again, as for Herodotus with the Phoenicians, what provoked ancient incredulity carries most conviction today: summer daylight lasting for up to six months, volcanoes with boiling springs, a frozen sea one day’s voyage north of Thule. The nearest we get to the horrors of Ocean is an area of the sea where all the elements seem confused: freezing fog, heavy ice-slush and viscous water are said to be like a ‘sea-lung’ (which probably means ‘jellyfish’). We also get fascinating excursuses on such things as the foundation of Massalia, the working of the Cornish tin mines and the mysteries of amber. I have only one real complaint about this book, but it’s a big one: the reader is presumed to be the kind of person who gags at references to sources, and will be put off by serious maps (the cute and inaccurate little sketches that fill this volume are an insult). Since Cunliffe’s reconstruction is based on a mosaic of fragments (some identified, some not, and all in translation), this policy seems, to put it mildly, perverse.

Alexander’s Eastern conquests, and the later systematic expansion of Rome, extended the boundaries of the ‘inhabited world’. Trade followed the flag, and sometimes preceded it (by way of the monsoon winds to India, for instance). Periploi (‘circumnavigations’), guides to coastal voyages and island-hopping, proliferated, and we have one, the Periplus of the Erythraean Sea, dealing with the sea-link between the Red Sea and the Orient, complete with lists of imports and exports. Caesar and Agricola familiarised contemporary geographers with the British Isles and a little of the North Sea coastline, while Scandinavia and beyond was still terra incognita. Though the sphericity of the world was by this time a commonplace, the inhabited world remained, for geographers such as Strabo (who argued that it should be the geographer’s sole concern), and Ptolemy after him, confined to one quadrant of the Northern hemisphere: a single land-mass that had a vague notion of India and China, but also rested on the mythical belief that China extended south and then west to link up with Africa, thus making an inland sea of the Indian Ocean. This odd concept was the result of a stubborn weakness for parallelism, detectable even in Herodotus, which argued that any physical pattern in the Northern hemisphere must have its matching opposite in the South.

Map of the world in Ptolemy’s first projection.

To get a bird’s-eye view of the Greco-Roman idea of the inhabited world during the early Roman Empire one can’t do better than look at the map made from Ptolemy’s first projection, preserved in a famous MS of about 1300, and reproduced as one of Berggren and Jones’s splendid full-colour plates. The basic outline of the Mediterranean is now recognisable; so is the Arabian peninsula. Eastward we see a foreshortened India, and a hugely exaggerated Taprobane (Sri Lanka). The further from Alexandria, the greater the inaccuracy. But then you notice something else. The Atlantic, too, is shown as foreshortened, a narrow vertical strip facing a huge block of terra incognita. What is going on here? The answer is simple but disconcerting. By accepting the Posidonian rather than the Eratosthenic estimate of the world’s circumference, Ptolemy’s calculations assumed a globe about three-quarters of its actual size. This had several unfortunate consequences. His parallels and meridians were too close together, so that a specific distance occupied too many degrees on his map (Berggren and Jones provide a series of comparative maps that illustrate this error graphically). He assumed that the known landmass, from the Canaries (which he misplaced by 7 degrees) to the eastern extremity of Asia, occupied 180 degrees, which in fact takes one to a point in the mid-Pacific. One curious delayed result of this was to convince Columbus that the westward voyage to India was comparatively short. Would he have attempted it had he known the truth? Nothing could more vividly illustrate the gap between Ptolemy’s theorising, which was brilliant, and the too often inaccurate mensuration that inevitably flawed his projections. Ptolemy himself knew this, and went, rightly, for astronomical fixes wherever possible. But even here error crept in. As John Noble Wilford reminds us, ‘even as late as 1740, it was estimated that not more than 116 places on earth had been correctly located by astronomical observation.’ When we also recall that Ptolemy had no method of accurate time-keeping, we can only marvel that he achieved as much as he did.

The number of advances for which he was responsible is extraordinary, and I fear it reflects less than well on classical studies that, until Berggren and Jones took him in hand, the most recent complete text of the Geography was Nobbe’s of 1843-45, based on few MSS and lacking an apparatus criticus. Yet this was the pioneer who established the graticule (a grid of carefully mapped co-ordinates) as the basis for serious cartography; who introduced ‘minutes’ and ‘seconds’ to facilitate the subdivision of degrees; who argued for the primacy of the simplest hypothesis that did not contradict observation; who demanded that observations calling for precision should be checked and rechecked over a long period; who insisted that maps be drawn to scale; who developed the use of both gnomon and astrolabe for celestial angle-measurements to determine latitude; who, most notably, tackled the perennial problem of how to represent the globe, in whole or part, on a flat surface. In Berggren and Jones he has, at long last, found his ideal exponents. There are a few parts of their exposition – as of Ptolemy’s own text, even in their precise translation – that will probably be beyond the geometrically challenged reader; but for the most part they have achieved both accuracy and clarity.

To turn from the Geography to the Barrington Atlas drives home just how far we have advanced. The history of cartography after Ptolemy is a fascinating story: the regression of the Dark and Middle Ages, when monsters and superstitions proliferated, Paradise had a physical location and all roads led to Jerusalem; the great voyages of Columbus and Magellan, Mercator’s projection (the potent and enduring instrument of European sea power) and Ortelius’ strikingly named Theatrum Orbis Terrarum, the first modern atlas; new methods of surveying (such as triangulation) and the challenge to absolute global sphericity posed by Newton; the exploration of the Pacific by Cook, or the Himalayas and Tibet by Everest and his disguised Indian pundits. Yet all these discoveries and achievements have been dwarfed, during the last few decades, by mapping from space satellite, which at last allowed a vision detached from the Earth itself. New techniques also reached degrees of accuracy unimaginable previously. The traditional surveyor’s error factor had been at best about 1:5000, or a foot to a mile. Measurement by laser beam changed this to 1:1,000,000. Even the polar continents yielded up their secrets. A chemical engineer called John P. Snyder finally worked out the 82 mathematical equations that could convert spacecraft images to accurate maps in terms of the new Space Oblique Mercator Projection (SOMP, or Colvo’s Projection, after its inventor, Alden Colvocoresses). The Barrington Atlas is in every sense the child of this revolutionary new technology.

It is also – again, in every sense – a vast achievement: 99 physical maps of folio size, accurate to the last second of each co-ordinate, in the conventional green-to-terracotta shading indicative of altitude, and ranging from the Fortunate Isles, a.k.a. the Canaries, in the West to somewhere east of the Ganges. Well-known and thickly-occupied areas have a 1:500,000 scale, the outer reaches shrink to 1:1,000,000, while one or two special regions (e.g. Latium Antiquum, including Rome, and Attica, with Athens) are blown up to 1:150,000. Ancient nomenclature is in Latin. Each map has its own detailed entry in a two-volume gazetteer (which also comes on CD-Rom), with a short informative introduction and brief references to essential scholarship. I checked the area I know best, Lesbos. To my astonishment, the map indicated a rather dismal pile of coastal stones, hopefully identified by I.D. Kontis in 1978 as an ochyro, or ‘stronghold’, a mile or two outside Molyvos, where I used to live: this doesn’t even figure on my excellent 1:70,000 ordnance survey map. (On the other hand, there is no indication of the island’s famous petrified forest.) The Barrington also prints the enchantingly named island of Por(d)oselene, or ‘Moon(f)art’, off Mytilene, showing how the name was edited by later occupants and commentators.

Richard Talbert can be proud of his editorship: the collective effort, academic and technical, that has gone into the realisation of this gigantic project, and is outlined in Talbert’s introduction, almost defies the imagination. It is even more impressive in that his teams had to work virtually from scratch. Their chief goal was to fill a notorious gap, and they have done so with exemplary skill, creating a basic computerised digital template which it will be relatively easy to adjust and develop in future, and from which a number of spin-offs will surely come.

Which raises the question of what audience the Barrington is aimed at. The prime target is surely professional: academic historians, classicists, cartographers. The atlas is hardly one that an educated general reader is going to invest in under normal conditions. Talbert, significantly, criticises N.G.L. Hammond’s Atlas of the Greek and Roman World in Antiquity (1981) for ‘attempting to map not only the physical and cultural landscape, but also political, military, religious, economic and other themes – all within a mere 30 map pages’, not least because in so doing Hammond was following ‘the traditional practice of textbook classical atlases’. Since this was precisely what his atlas aimed to be (it was also, incidentally, the first of its kind to pick up on satellite mapping) it seems perverse to blame him for it. The Barrington is also striking for its omissions. There is virtually nothing political or thematic here: no battle sites, no city plans, no population movements. But equally there are no anonymous excavation sites, no find-spots, no sea routes or shipwrecks. It is as though having made the effort, not only to plot the physical geography but to reconstruct it as it was over two millennia ago (a far harder task), the researchers simply gave up on any further work.

The result is a number of singularly uncluttered maps: a virtuous change, it might be argued, from the notorious cartographer’s horror vacui that insisted on filling the blank spaces with irrelevant or misleading matter. But blank spaces, in fact a mere confession of ignorance, can be as misleading as clutter, by implying an emptiness that may be fictional. It is good to learn that the University of North Carolina is establishing an Ancient World Mapping Center ‘to maintain and expand the achievement of the project on a permanent basis’. Talbert also hopes that the Barrington’s maps ‘should form the basis for branching out further in every direction’. Amen to that. Having largely established the physical base, let them now start work on the political, which will surely prove more frustrating, but is just as crucial.