The Geography (its Greek title translates literally as ‘Instruction in Map-drawing’) ‘professes to be concerned solely with the task of scientific mapping’.43 But what, in fact, have its contributions been to the scientific mapping of the world?
One signal contribution was the memorialization of knowledge of the earth’s basic form as a sphere. Just how ancient this knowledge already was in the second century AD remains unclear. Scholars agree that its earliest surviving documented appearance is in the work of Pythagoras in the sixth century BC;44 however, it may have long pre-dated Pythagoras in oral traditions or in documents that have since been destroyed with the passage of time. My personal view, already expressed elsewhere, is that the concept of the spherical earth was well known to the first great historical civilizations such as the ancient Egyptians and the Sumerians 5000 years ago and will ultimately be proved to date back to a much more remote period even than that. But wherever it comes from originally, we owe a debt of gratitude to Ptolemy for its preservation and repromulgation in the second century AD – for, despite the intellectual ravages of the Dark Ages that were to follow, his vision of the earth as a sphere was never quite forgotten. Robert Fuson, Professor of Geography at the University of South Florida, puts it this way:
The Ocean Sea had now taken the form it was to retain until the 16th century and the aftermath of Magellan’s circumnavigation. The earth’s sphericity was no longer debated by any practical navigator, cosmographer, or educated person. This fact had been established since the days of Classical Greece. The only areas open to serious disagreement were details of coastal configuration, exact location by coordinates, island discovery and location, and the dimensions of the Ocean Sea. By the 1400s [before the discovery of the Americas] no reasonable person questioned the proposition that Asia might be fetched by sailing west from Europe, if the ships and crews could survive the tremendous distance.45
Francesco Berlinghieri’s Ptolemaic map, AD 1482.
Other significant contributions that Ptolemy made to the scientific mapping of the world include the establishment of functional parallels of latitude, and of a prime meridian, passing through the Canary islands, that was to serve as zero degrees longitude for sixteen centuries.46 Moreover, though maps drawn to Ptolemaic coordinates leave much to be desired, even the worst of them are far superior to the schematic ‘T-O’ maps and mappamundi of the Dark Ages.
A representative selection of Ptolemaic world maps is reproduced herewith. The reader will note that the Mediterranean is at least recognizable, and that in spite of many discrepancies a real attempt appears to have been made to reflect the true shapes and locations of the lands bordering it. Ptolemy and his informants had first-hand, day-to-day knowledge of this central region of what they called the oikumene – the habitable world – and clearly, with some peculiar exceptions, they used that knowledge well. But outside the Mediterranean the level of accuracy rapidly falls away.
Ptolemaic map from Venice edition of Ptolemy’s Geography, AD 1511.
Waldseemuller’s Ptolemaic map, AD 1507.
For example, on the authority of Poseidonius (135–50 BC),47 Ptolemy underestimates the circumference of the earth at the equator, setting it at 20,400 miles (as against the correct figure of 24,902 miles).48 At the same time he greatly overestimates the east-west extent of Asia and, bizarrely, portrays the South Asian coast above the Indian Ocean apparently without any representation whatsoever of the great peninsula of the Indian subcontinent. As though to compensate for this loss, however, Ptolemy places an enormous island, Taprobana (presumed to be Sri Lanka), just off-shore of the stretch of non-peninsular mainland identified as India.
What is going on here? In their major new study of Ptolemy’s Geography, J. Lennart Berggren and Alexander Jones suggest that the root of the problem is simple. India has this ‘flattened out’ appearance because Ptolemy, somehow, has managed to turn the subcontinent on its side so that its orientation is roughly west-to-east instead of north-to-south as it should be:
Asia exhibits greater and greater distortions as one progresses further east, the most obvious faults being the north-south compression of the Indian subcontinent so that its western coast is made to run parallel to the equator, and the exaggerated size of the island of Taprobana (Sri Lanka).49
If the subcontinent has indeed been swung eastwards in the way that Berggren and Jones propose, then ‘Taprobana’ is not only too big to be Sri Lanka but is also positioned in entirely the wrong place. Sri Lanka lies in the Bay of Bengal, off India’s south-east coast. Once the reorientation of the peninsula on Ptolemy’s map is taken into account, however, then we can see that Taprobana has in fact been portrayed as lying off India’s west coast – where there are no large islands today.
We will return to the possible implications of this later. Meanwhile, to conclude the description of Ptolemy’s world map, let us note that the older examples (e.g. page 466) portray the Indian Ocean as a lake landlocked by the northern edge of a southern continent (Terra Australis on some editions; Terra Incognita on others) that connects southern Africa with the south-eastern extreme of Asia:
At the eastern edge, where the lands represent central China and Southeast Asia, it is virtually impossible to identify any of the features on Ptolemy’s map with real counterparts. At the eastern limit Ptolemy draws the coast of Asia turning south and then west, eventually to join the east coast of Africa, thereby making the Indian Ocean a vast enclosed sea unconnected with the Atlantic.50
Ptolemy was not the originator of the Geography - as he himself goes to great lengths to point out. Instead, he tells us that his role has been to refine and correct an earlier Geography prepared by his predecessor, the Phoenician geographer Marinus of Tyre, who was active around AD 100 or no and whose great work was itself called Correction of the World Map.51 In Ptolemy’s own words:
Marinus of Tyre seems to have been the most recent of our students of geographia [= map-making] and to have applied himself to the subject with the greatest enthusiasm … If we could see that his latest composition lacked nothing, we should even have been happy to complete our description of the known world from these notes of his alone, without researching any further. But as on certain points he himself seems to have composed without reliable comprehension, and as in embarking on his map he has in many places not devoted enough thought either to convenience or to symmetry, we were naturally induced to contribute to his work what seemed necessary to make it more logical and useful.52
As well as the honesty of this statement, what I find particularly striking is the strong suggestion Ptolemy leaves us with that his Geography was part of a tradition, and that his predecessor Marinus had been part of that tradition too – but by no means its first student, just the ‘most recent’ who had ‘corrected’ an older map.53 Such a tradition might, theoretically, have extremely ancient roots and it need not necessarily be the case that successive ‘refinements’ of it over long periods of time must have improved it. An alternative possibility, which it would be unwise to ignore entirely, is that far from being the pinnacle or ‘culmination’ of ancient geography, as many scholars suggest,54 Ptolemy’s maps may actually have been the end-products of a long process of decline, degradation and accumulated errors introduced by many different hands into a far older and once far superior map-making tradition. Again, this is a theme that we will return to.
Some centuries after Ptolemy’s death the Dark Ages descended over the Geography, but it was still preserved here and there in a few monasteries in Europe.
In the Arab world Muslim geographers are known to have possessed editions of the Geography as early as the eighth century AD, as well as separate editions of the earlier work of Marinus (the latter now all being lost):
In the early ninth century al Ma’amun, Caliph of Baghdad AD 813–833, set up an Academy of Science, which among other things produced a world map [lost] and ‘improved tables’, – i.e. modernized coordinates.55
In Byzantium (Const
antinople) in the late thirteenth century it was Maximus Planudes (c.1260–1310) who was responsible for bringing the knowledge enshrined in Ptolemy back to the attention of the world:
He searched for manuscripts of Ptolemy’s Geography, and his search was rewarded in 1295, but it was not as exciting as he had hoped. As he explains in a letter and some verses, after at last finding what he knew was a neglected work, he was disappointed to discover that it had no maps.56
Although there are older manuscript maps (such as the late twelfth-or early thirteenth-century Codex Urbanus Graecus 82), the oldest surviving manuscript copy of the Geography containing maps based on the descriptions and coordinates given by Ptolemy was made by monks at Vatopedi on Mount Athos in the early fourteenth century.57 It later formed the basis for the first printed atlas to appear in Europe, published at Bologna, Italy, in 1477.58
The Ptolemaic cartographic tradition was at first very successful in adapting to the challenges posed to its world-view by the Age of Discovery. Thus, the original maps based on Ptolemy’s own coordinates were added to several times during the sixteenth century to accommodate so-called tabulae novae (or tabulae modernae) recording the expanding revelation of the Americas and of the East.59 This could be done without causing serious disturbance to the Ptolemaic concept of the oikumene so long as the simple expedient could be maintained of tagging the Americas on to Asia like some vast peninsula. Ultimately, however, these maps, like the dinosaurs, were an evolutionary dead-end doomed to extinction.
It would be wrong to imagine from all this that the few surviving Ptolemaic maps in libraries and archives around the world have nothing to teach us. They may appear distorted and clumsy to the sophisticated modern eye, but it is possible that their very awkwardness and peculiarity could have caused scholars to overlook significant details concealed within them.
Portolan charts
The fourth category of maps circulating in Europe from the fourteenth to the sixteenth centuries, known collectively as portolanos, portolan charts or simply portolans, shows no dependence whatsoever on either Ptolemaic maps or data or the mappamundi. The vast majority of the portolans depict only the Mediterranean/Black Sea area and the countries immediately round about, but some are world maps, or world atlases, for which the style and approach of the Mediterranean portolans serves as a basis. These old charts are drawn to the highest cartographical standards and are uncannily accurate – so accurate, though the earliest examples go back to the end of the thirteenth century, that they were not surpassed by new scientific techniques, measurements and observations for almost 500 years.60
A. E. Nordenskiold, the great Swedish polar explorer and map historian, had a special interest in portolans. He points out that they were used, almost entirely, by practical mariners and navigators:
Slight was the attention paid to them in the fifteenth and sixteenth centuries by learned geographers. Thus Munster seems to have totally overlooked them, and in the first edition of the Theatrum Orbis Terr arum, Ortelius does not mention a single drawer of portolanos amongst the cartographical authors enumerated in his Catalogus Auctorum. At present the investigator into the history of geography acknowledges them as unsurpassed masterpieces, and reckons them amongst the most important contributions to cartography during the Middle Ages.61
Likewise John Goss notes:
The portolan charts were quite unlike contemporary medieval maps. They often incorporated detail of remarkable accuracy, based on close and actual observation, rather than the conventional medieval habit of repeating cartographical and mythical information issued by the Church.62
Goss and Nordenskiold also point to other characteristics that make the portolans look and ‘feel’ different:
A network of intersecting straight lines (usually called ‘rhumb-lines’ or ‘loxodromic lines’)63 originating from sixteen equidistant points, spread about the circumference of a ‘hidden circle’ around the map.
An elaborate ‘compass-rose’ at one or more of the points of intersection of the lines.
Place and feature names written perpendicular to the coastline, in sequence along the coast.
Charts drawn in ink on vellum or parchment with colour conventions, e.g., most important names shown in red, rest in black; lines depicting four main wind directions drawn in black, the eight half winds in green, the sixteen quarter winds in red.
Coastlines emphasize bays and headlands; hazards such as rocks, reefs and shoals are marked with dots or small crosses.64
What all these characteristics have in common is their utility and significance to mariners. The coastal hazards are matters of life and death. The networks of rhumb-lines assist compass navigation. Even the perpendicular place names, inevitably upside-down from some angles, make sense when you realize that they are meant to be viewed in the same direction as that of a vessel following the coast.
It has been suggested that portolans are such an improvement on previous maps because they reflect the earliest introduction of the compass into Europe, thought to have taken place around the end of the thirteenth century65 (although the use of magnetized needles as a means for sailors to find their bearings is attested earlier than that).66 But while there is no doubt that such charts in conjunction with compasses do provide very effective navigational guides, it is by no means so certain that compasses and compass-bearings were used to prepare them in the first place. On the contrary, says A. E. Nordenskiold, ‘many of them are evidently older than the use of the compass on board ships’.67
No map projection is imposed on the portolans, and there is no latitude and longitude grid – although in the expanded ‘world portolans’ the equator is often shown, together with the Tropics of Cancer and Capricorn and the Arctic and Antarctic Circles. Nevertheless, when relative latitudes and longitudes on these maps are measured, they prove to be extremely accurate. For example, on the Dulcert portolan of AD 1339 the total longitude of the Mediterranean and the Black Seas is correct to within half a degree.68
Carta Pisan portolan, c. AD 1290.
It is wrong to argue, as I myself have done in the past,69 that mariners and chartmakers of the fourteenth century would have found it impossible to achieve such accurate longitudes. Such suspicions arise from the fact that marine chronometers – which made reliable calculations of longitude at sea possible – were not introduced until the second half of the eighteenth century. However, scholars are right to object that there are other, simpler (if vastly more time-consuming) ways to obtain almost equally accurate longitudes. As Gregory Mcintosh put it in an e-mail:
We moderns seem to tend to think that because we now have methods of making very accurate measurements very quickly, those in the past could not make any measurements at all. The Portuguese (and others, of course) made dead reckoning measurements of longitude [i.e. calculations based on empirical estimates of course, speed and time]. It is a method of measurement. Some writers of the Hapgood ilk [Charles Hapgood, Maps of the Ancient Sea Kings] would have us believe that dead reckoning is not a valid method of measurement. They would have us believe that the Portuguese did not measure longitude. But of course they did. That’s what dead reckoning is – a method of measuring longitude … with several such measurements from repeated voyages.70
Top: Dulcert portolan, AD 1339. Bottom: Maggiolo portolan, AD 1563.
This seems like an entirely reasonable explanation for the accuracy of the portolan charts – that they are the result of the accumulated observations and measurements of navigators plying the coasts of the Mediterranean over relatively long periods of time. Some have suggested that they may even trace their origins back to the detailed written accounts of sea-journeys, harbour conditions, winds, currents and trade – the periploi – that were in favour amongst the ancient Greeks as far back as the fifth century BC.
Still, there is a very large gulf indeed between the crude directions of the periploi and the navigational accuracy of the portolan maps. Along any hypothetical evolutionary road from one to a
nother it is reasonable to expect to see intermediate forms – since getting maps right by dead reckoning, as Mcintosh points out, is a painstaking, long-term process of trial and error, correction and gradual improvement.
And this is the central problem of the portolans. Quite simply, there are no intermediate forms. Indeed, remarks John Goss:
From the outset portolan charts appear to have been remarkably accurate with little evolutionary development from the earliest known examples to the later charts made towards the end of the seventeenth century.71
And A. E. Nordenskiold, the world’s greatest authority on the portolans, reminds researchers that:
Notwithstanding all the progress made during the fifteenth and sixteenth centuries in the art of drawing maps with the aid of newly invented nautical instruments, there was published a chart in Holland in 1595 by one of its most expert mariners which is only a copy, or rather a copy of copies, of portolanos drawn 250 to 300 years earlier. This is an extremely remarkable fact in the history of civilization. But moreover the principal features of the portolanos from the beginning of the fourteenth century are still to be found on Van Keulen’s sea-charts of 1681–1722. I suppose that up to the beginning of the nineteenth century the influence of the old portolan charts may yet be traced on the charts of several parts of the Mediterranean and Black Seas.72
The good legacy
The ‘extremely remarkable fact in the history of civilization’ that Nordenskiold draws our attention to here is the ability of maps, apparently produced by dead reckoning in the thirteenth century, to compete on an equal footing with scientific nautical charts from as late as the nineteenth century.