Hancock’s Fingerprints of the Gods, Part I: misunderstanding early modern cartography

Piri Re‘is’s map of 1513

Piri Re‘is’s map of 1513

Graham Hancock begins his quest for the beginning and the end with that old chestnut of alternative archaeologists: the Piri Re‘is map of 1513. Quoting a letter from Charles Hapgood’s Maps of the Ancient Sea Kings (Chilton Books, 1966), he accepts the opinion of Harold Z Ohlmeyer (1919-2010), a United States Air Force Colonel in the 8th Reconnaissance Technical Squadron, that the map depicts West Africa, South America and the coast of Queen Maud Land in Antarctica. This is a contentious assertion, as we will see, but it is Ohlmeyer’s statement that the Antarctic coast is depicted as it would appear free from ice that forms the starting point for Hancock’s analysis.

To Hancock, the apparent depiction of Queen Maud Land raised a number of questions. He focuses on what he considers the six key facts of the case. Firstly, there is no question that the Piri Re‘is map is anything other than a genuine manuscript map of 1513. Secondly, that it shows precisely those landmasses that Colonel Ohlmeyer claimed it does. Thirdly, that Piri could not have known of Queen Maud Land from his contemporaries, as Antarctica had not yet been discovered (its official discovery date of 1818 may be too late, but probably not that much too late). Fourthly, there is the puzzle that Queen Maud Land is shown as free from ice; Hancock claims that the last time this could have happened was around 4000 BCE. Fifth, he claims that although the earliest date at which the coast could have been mapped is impossible to determine, but that it “may have remained in a stable, unglaciated condition for at least 9000 years before the spreading ice-cap swallowed it”. Finally, he correctly points out that we do not know of any civilisation that could have mapped the coastline in the period at which he claims it was ice free.

This, then, is what Hancock considered the smoking gun that would set him off on his quest. The entire premise of Fingerprints of the Gods proceeds from the assumption that his six key facts are all equally factual and not open to dispute.

Hancock’s six facts about the Piri Re‘is map

Firstly, Hancock is quite right to stress the genuineness of the map. Rediscovered in 1929, the map was instantly famous, not least because it claimed that one of its sources was a map produced by the Genoese explorer Cristoforo Colombo (better known in the English-speaking world as Christopher Columbus), whose own maps are notoriously lost. Here was the earliest surviving map of his barely twenty-year-old discoveries and of those who followed in his wake. In his day, Piri was a renowned cartographer and the maps that were known before 1929 are beautiful accomplishments. No-one could raise any serious objections to the map.

A satellite view of Antarctica with sea ice removed

A satellite view of Antarctica with sea ice removed: Queen Maud Land is the bulge at the top

However, we are on more dubious ground when it comes to determining what parts of the world the map actually depicts. West Africa and the Iberian Peninsula are clear enough; it is what Piri calls “the western region” that is less clear. His listing of sources for this map show that he was reliant on a single source, a copy of the map created by (or at least attributed to) Columbus. Whereas he combined a number of maps to produce his clearly recognisable coastlines of Africa and Iberia, he was reliant on the accuracy of the sole map at his disposal for the depiction of the New World.

It is unfortunate that Hancock did not do further research into the map, preferring to rely on Charles Hapgood’s ideas about its accuracy and assessment of what it shows of “the western region”. He might for instance, have consulted works that deal with the texts on the map, which show placenames that can be identified today and which describe the climate and fauna of the newly discovered regions. Gregory McIntosh, in his The Piri Reis Map of 1513 (University of Georgia Press, 2000) castigates Hapgood and others for failing to take note of the placenames and basing the conclusions merely on the resemblance of stretches of coastline to those on modern maps. This is not a legitimate way to investigate any historical document. The placenames alone are enough to show that the most southerly place depicted on the map is Puerto San Julián in Argentina.

However, Hancock accepts Charles Hapgood’s identifications of places uncritically and assumes that the bottom of the map shows Queen Maud Land. He is correct in the assumption that Piri could not have learned of Antarctica from his contemporaries but as the map does not show it, this “fact” can be struck from the list. Given that Queen Maud Land is not depicted, the ice-free nature of the coastline ought not to worry us. Nevertheless, is Hancock correct in his dating of a period when Queen Maud Land might have been without ice at a time when human mariners (not the aliens of Erich von Däniken) might have mapped it?

Numerous ice cores have been taken from the Antarctic ice that show the continent to have been covered by a fully developed ice cap from 40,000 to 6,000 BP. Between 21,000 and 16,000 BP there was a maximum development of the ice cap, corresponding to the height of the Devensian glaciation in Britain, with no exposure of coastlines after the glacial maximum. Indeed, it is evident that Antarctica was last completely ice-free over 34,000,000 years ago, long before the evolution of the genus Homo, let alone modern humans. There is thus no human civilisation that could have mapped an ice-free coast.

Chapter 2: the other maps

Hancock does not stop with Piri Re‘is’s map, even though, as we have seen, it does not show what he thinks it shows (or, perhaps, wants it to show). Following Hapgood, he turns to other early modern maps that show an Anatarctic continent. In this, he relies entirely on Hapgood, even for data on the climatic history of the continent that by the 1990s was hopelessly out of date: all but three of the thirty-three footnotes for this chapter are to pages in Maps of the Ancient Sea Kings. This is merely giving the outward signs of scholarliness, not real scholarship.

The Orontius Finaeus map (Hapgood and his followers always spell the name Oronteus)

The Orontius Finaeus map (Hapgood and his followers always spell the name Oronteus)

We are introduced to the Orontius Finaeus Delphinus map of 1531, with its enormous TERRA AVSTRALIS REcenter inuenta sed nondū plene cognita. As with the Piri Re’is map, he ignores the legend written across the landmass that is (conveniently?) illegible on the small reproduction of the map in his book. In English, it says “The southern land recently discovered but not yet fully known”. Does this allow for the possibility that Finaeus was copying an older map of an ice-free Antarctica? It certainly doesn’t read that way. Yet Hapgood, followed by Hancock, was impressed by the depiction of rivers and estuaries emptying into the Ross Sea. Naïvely, Hancock says that “[t]he unmistakable implication of these features is that there was no ice on the Ross Sea or its coasts when the source maps used by Oronteus Finaeus were made”. Clearly, in his world, no early modern mapmaker ever filled in details of unexplored areas with invented details, not even on a continent “not yet fully known”.

And what were these source maps that Finaeus failed to mention? If Hancock (or Hapgood before him) had bothered to read (or have translated) the text at the centre bottom of the map, he’d have seen that Finaeus specifically says that he is depicting Prouintias, Insulas, Maria, Flumina, Montes, hactenus non uisa, neque Ptolomeo, neque Eudoxo, neque Eratosteni, aut Macrobio cognita, sed que in tenebris in hunc usque diem iacuerunt (“Provinces, Islands, Seas, Rivers, Mountains not seen up to now, known neither by Ptolemy, nor Eudoxus, nor Eratosthenes or Macrobius, but which had lain in shadows up to the present day”). In other words, these are recent discoveries of which the ancients knew nothing: there were no source maps.

The Mercator World Map

The Mercator World map, showing a vastly oversized Terra Australis (Source)

Hancock (following Hapgood, as ever) next turns to Gerhard Kremer (better known as Gerhardus Mercator), whose Atlas of 1569 includes Finaeus’s map of Terra Australis. He used Finaeus’s map as the basis for his own depictions of the continent but incorporated more recent discoveries. Thus, Tierra Del Fuego at the southern tip of South America, is shown as a northern promontory of Terra Australis; the promontory south of Iaua minor (Java) appears to be Arnhem Land (Australia), drawn too far to the west; another promontory, south of Noua Guinea (New Guinea) is the Cape York Peninsula (Australia). Mercator is incorporating recently discovered lands that he hypothesises to be part of a massive continent at the south pole necessary to balance the land area of the northern hemisphere and keep the earth upright. Again, despite Hapgood’s confident assertion that “Mercator had at his disposal source maps other than those used by Oronteus Finaeus”, there is no evidence whatsoever that these maps ever existed.

Philippe Buache's Carte des Terres Australes (second version, c 1757) (Source)

Philippe Buache’s Carte des Terres Australes (second version, c 1757, from the Library of Congress)

The fourth map is the most recent: that of Philippe Buache. Once again, we are faced with the author’s unwillingness actually to engage with the map itself. The map contains a lot of text and the words conjecturée (conjectured) and soupçonnée (suspected) appear a number of times on the southern continent. Of greatest interest, though, is the lengthy account of Jean-Baptiste Charles Bouvet de Lozier’s (1704-1786) voyage to the south, which lasted from 19 July 1738 to 24 June 1739. This instantly undermines the date of 1737 that Hancock gives for the publication of the map; indeed, the map itself contains the publication date of 5 September 1739. Of course, the mistake was originally made by Charles Hapgood: Hancock (or his researchers) simply copied this error without checking.

Hapgood made a great deal of the fact that the version of the map he consulted in the Library of Congress shows the Antarctic continent to be composed of two separate landmasses, with a sea between them. Never mind that it is labelled MER GLAICIALE Conjecturée (“Conjectured Glacial Sea”): Hapgood assumed that it shows the continent free from ice. Yet we know why Philippe Buache depicted this sea. This version of the map seems to be a second edition, issued c 1757 (although the publication date is not altered on it), bringing his ideas about the southern continent into line with his more recent hypotheses about it, as published in 1761 in Considérations géographiques sur les terres australes et antarctiques. He had concluded (wrongly, as it turned out) that the icebergs seen by Bouvet de Lozier in 1738-9 must derive from sea ice, like those in the Arctic Ocean rather than from glacial ice, like those of the North Atlantic. He therefore decided to add a conjectural sea across the pole to account for them.

Buache's original version of the map (1739) is much more vague about the size and shape of the Antarctic continent

Buache’s original version of the map (1739) is much more vague about the size and shape of the Antarctic continent

However, in the first edition of the map (which neither Hancock nor Hapgood reproduce), there is no Glacial Sea. Instead, the Antarctic continent is depicted as a single landmass. The inclusion of Nouvelle Zelande (New Zealand) as a northern promontory of the continent as well as Bouvet de Lozier’s Cap de la Circoncision (actually an island, now named Bouvet Island after its unwitting discoverer) and a cape attributed to an observation of Amerigo Vespucci that seems to correspond in position to the South Sandwich Islands show that Buache’s sources were not ancient maps but early modern voyages of discovery. His depiction of lands to the south-west of Tierra del Fuego and the south of Cape Horn that correspond to no known islands shows that some of the observations on which he relied were inaccurate. It is worth pointing out that those parts of the continent for which he relied on reported discoveries have coastlines shaded pink: the remainder, which accounts for more than 90% of the outline, is all conjectural.

Hancock's comparison of Mercator's, Orontius Finaeus's and Philippe Buache's maps of Terra Australis with the results of the 1957-8 seismic data: hardly impressive!

Hancock’s comparison of Mercator’s, Orontius Finaeus’s and Philippe Buache’s maps of Terra Australis with the results of the 1957-8 seismic data: hardly impressive!

The contention first put forward by Hapgood and repeated eagerly by Hancock that these four maps not only show the Antarctic free from ice but also match the discoveries of the sub-glacial topography of Antarctica as determined by surveys during the International Geophysical Year (actually 1 July 1957 to 31 December 1958) is incapable of proof. For a start, the “coast-line” determined at that time shows where current sea level would be if the ice were suddenly to be removed. This ignores several factors. Firstly, the contour corresponding to modern sea level is that of a landmass experiencing isostatic depression. In other words, the weight of the ice cap is pushing down on the continental plate on which Antarctica sits, submerging the contour that would correspond to its coastline before the formation of the ice. Remove the ice and the continent would “bounce” back up as a result of isostatic uplift; this is a well known phenomenon and can be observed in those parts of the northern hemisphere that were depressed under the expanded Arctic ice cap during the Pleistocene Ice Age, where it is still ongoing. Secondly, Hancock uses data from the 1950s as a basis for comparison with the early modern maps. This is because he is reliant entirely on Hapgood for his data: he has done no further research into the question. More recent maps show a subglacial topography substantially different from that in his figure on page 21 of the first paperback edition. Worse, there is little correspondence between the 1950s data and the historic maps that Hapgood claims to be so accurate.

A more recent view of the sub-glacial topography of Antarctica

A more recent view of the sub-glacial topography of Antarctica (Source)

Chapter 3: maunderings on longitude

In the final chapter of Part I, Graham Hancock is impressed—rightly, in my opinion—with the relative accuracy with which Piri Re’is placed South America to the west of Africa: his assessment of the width of the South Atlantic Ocean was very good. The data about the north coast of South America is distorted by Columbus’s preconceptions, so that Cuba is depicted as part of the western landmass; indeed, so paranoid was Columbus about not having discovered a quick route to East Asia, that he forced his crew to sign affidavits to the effect that Cuba was an Asian peninsula, a belief he held until his death: so much for the glorious discoverer of a New World!

Edward Wright's map "for sailing to the Isles of Azores"

Edward Wright’s map “for sailing to the Isles of Azores” (c 1595), covered in rhumb lines (Source)

Hancock manages to persuade himself that the only means of determining longitude is by using a chronometer. As the first chronometer was developed by John Harrison (1693-1776) in the eighteenth century, he cannot understand how Piri (and other early cartographers) were able to draw accurate maps. The answer, of course, is one that had been widely used since antiquity: triangulation. Piri’s map is cover by rhumb lines radiating from compass-like symbols in the ocean. Such lines depict navigational aids that could be used to estimate longitude to within a reasonable degree of tolerance, so long as the sailor was able to estimate speed and bearing. The problem of calculating longitude was well known in the medieval world (indeed, it was an element in Columbus’s delusion that Cuba was a hitherto unknown Asian promontory) but the technique of sailing by the rhumb was good enough to enable reasonably accurate maps to be drawn up. Far from being evidence for “lost mathematicians”, these maps show how the ingenuity of late medieval and early modern mariners enabled them to produce workable maps of the known world.

Thus ends Part I, with something of a damp squib. Moreover, the idea that these maps present a puzzle that needs to be solved—the depiction of an ice-free Antarctica—is false. The premise of Graham Hancock’s entire “quest” is equally false. I could stop the analysis of Fingerprints of the Gods here, but the loyal Hancockite could insist that even if the maps don’t show what Hancock claims they show, his other evidence could instead point to the existence of a “Lost Civilisation”: it wouldn’t be the first time that correct conclusions have been reached by starting out from a false premise. There are some interesting parallels with Chariots of the Gods? that I want to explore…

Conclusion: “The Mystery of the Maps” demystified

One of the most obvious observations about Part I of Fingerprints of the Gods is that Graham Hancock is completely reliant on Charles Hapgood’s untenable views about early modern maps. He (or his researchers) have either not consulted any scholarly works on the history of cartography, of which there are plenty, or have chosen not to mention their conclusions about these maps. We are faced with an inexcusable ignorance about how to conduct adequate research into the past, a naïve belief that it is possible to rely on only one interpretation of the evidence, a failure to establish an hypothesis by showing how his own explains the data better than existing hypotheses or a deliberate suppression of evidence that undermines his hypothesis. Of course, it could well be a combination of these. In any case, the result is that Part I of Hancock’s book is not a scholarly examination of early modern maps, but one that is tendentious whilst trying to give the impression of scholarship to an unsuspecting readership whom he hopes will see the work as well researched because of all those footnotes.

As an aside, one has to feel sorry for Colonel Ohlmeyer who, by all accounts, was a model USAF officer, a pleasure to serve under and a generally good human being: a Google search for him currently returns over 5600 hits, almost all of which refer to the letter he sent Charles Hapgood in 1960. Surely his memory deserves better than this.


Old maps, the Americas and Antarctica

Maps of the fifteenth to eighteenth centuries are a favourite source of information for fringe writers, who use them to make a wide variety of claims. To Erich von Däniken, for instance, they are evidence for a survey of the Earth from space, carried out by extraterrestrials, while for Graham Hancock, they are evidence for an ancient sea-faring civilisation, lost beneath the sea after the melting of glacial ice at the end of the Pleistocene. These writers focus on a relatively small number of such maps, those of Piri Re‘is and Orontius Finaeus being the most used, whilst ignoring others of the same age. All these maps are alleged to show anomalous knowledge for the dates at which they were drawn: the west coast of South America, Antarctica (with or, more frequently, without its ice sheet), the Strait of Magellan and other “impossible” details. This appears to be solid evidence, so why do mainstream historians and archaeologists ignore it?

Piri Re‘is’s map of 1513

Piri Re‘is’s map of 1513

The Piri Re‘is map

The most widely used of these maps is a manuscript map produced in 1513 CE by Hacı Ahmed Muhiddin Piri, better known as Piri Re‘is (“Admiral Piri”, although most of these writers seem not to understand that Re‘is is a title, not a surname). It was drawn on camel-skin parchment and is one surviving part of an originally larger set of maps depicting the known world. Since its rediscovery by the German theologian, Gustav Adolf Deissmann (1866–1937) in the Topkapı Sarayı Museum in 1929, it has been an important source of claims that there were much older maps showing the world in great detail, including places unknown in the early sixteenth century CE. Much of the detail in these claims derives not from scholarly studies of the map but from the work of Charles Hapgood (1904-1982), a geography teacher at Keene State College (whose status is often inflated to ‘professor’ through a misunderstanding of American usage of the term).

The inspiration behind Hapgood’s work was a radio discussion on 26 August 1956 between Arlington Humphrey Mallery (1877-1968), an engineer then working for the US Navy Hydrographic Office, Rev Daniel L Linehan SJ (1904-1987), director and chief seismologist of the Weston Observatory at Boston College, and Rev Francis Heyden (1907-1991), director of the Georgetown University Observatory. Mallery, something of a student of the history of cartography and an amateur archaeologist, had formed the view that the bays and islands depicted at the bottom of Piri’s map were hidden beneath the ice of Queen Maud Land (Antarctica). After reading a transcript of the broadcast, Hapgood contacted Mallery and, having obtained a copy of the map, set his students to work examining it.

Hapgood’s attempt to impose a grid on Piri Re‘is’s map

Hapgood’s attempt to impose a grid on Piri Re‘is’s map

Hapgood’s account of the investigation, in Maps of the Ancient Sea Kings (1966) is tedious to anyone, like me, with little or no interest or ability in maths. He detected the use of a grid on late medieval portolan charts and suggested that a similar grid was used by Piri; he conjectured that it was based on Syene (Aswan, Egypt) and that similar grids were used on other early medieval maps. This may have been a correct deduction (although it appears not to be generally accepted by historians of cartography, who believe that portolans were based on compass directions), but it is the next stage of Hapgood’s analysis where the claims made for the map go way beyond the evidence.

Hapgood started with the belief that the Piri Re‘is map was an accurate depiction of South America and part of Antarctica but when close analysis showed that it was not accurate in any projection he and his students applied to it, had to come up with a reason why it contained errors. Given that Piri stated that he had used “about twenty charts and Mappae Mundi” and that some of them were “drawn in the days of Alexander”, Hapgood conjectured that Piri’s map (or its sources) had wrongly combined numerous earlier sources of varying scale, orientation and projection. In this way, small sections of coastline were drawn accurately but each section had to be looked at in isolation. Worse, some parts of the coastline were missing (so that the Strait of Magellan was not depicted, for instance) and some were duplicated. In this way, Hapgood and his students could rescue Piri’s map from any suggestion of inaccuracy.

Unfortunately, Hapgood has misunderstood what Piri says about the sources for his map. Here is Piri’s note in full:

This section shows how this map was drawn. In this century, there is no map like this in anyone’s possession: the hand of this poor man has drawn it and now it is assembled from about twenty charts and Mappae Mundi (these are charts drawn in the days of Alexander, Lord of the Two Horns, which show the inhabited quarter of the world). The Arabs name these charts Caferiye. I have compiled it from eight Caferiyes of that kind, one Arab map of India, from the maps recently drawn by four Portuguese that show the countries of India, Sindh and China drawn geometrically, and also from a map drawn by Columbus in the western region. The present form was achieved by reducing all these maps to a single scale so that this map is as correct and reliable for the Seven Seas as the map of our own countries is considered accurate and reliable by sailors.

It is quite clear from this that Piri’s only source for the “western region” was a map he attributed to Columbus. The Mappae Mundidrawn in the days of Alexander” were not charts 1800 years old when Piri acquired them but maps based on Claudius Ptolemy’s Geographical Guide (Γεωγραφικὴ Ύφήγησις, more commonly known as the Geography), which had become the standard for accurate mapping in the Arab world and in Christian Europe after a text was brought from Constantinople in 1400. Rather than dating from “the days of Alexander”, the original work dated from c 150 CE and although the only copy that Maximos Planoudes (Μάξιμος Πλανούδης, c 1260-c 1305) was able to locate in Constantinople in 1295 had lost its maps, the tenth-century al-Masʿūdī (أبو الحسن علي بن الحسين بن علي المسعودي, Abu al-Ḥasan ʿAlī ibn al-Ḥusayn ibn ʿAlī al-Masʿūdī c 896-956) was familiar with a copy that may have retained them. These maps dealt with only those parts of the world known to Ptolemy; Piri used more recent maps to update them.

What this means is that Hapgood’s attempt to rescue his hypothesis is just plain wrong: Piri is absolutely explicit that his only source for the “western region” was a chart he believed to have been compiled by Christopher Columbus. Piri may or may not have been correct in this belief, but either way, his sole source for the western continent was a map deriving from the voyages to the New World by European explorers after 1492. Had there been earlier maps available to him, we would have to explain why he did not mention them as sources.

What about Antarctica?

So, why did Piri show a land apparently south of the South Atlantic? Is this evidence for an early discovery of Antarctica? Alas, no. The authority of Arlington H Mallery is not quite what it seems: although fringe writers tend to refer to him as an expert on historic maps and an archaeologist, with the implication that his work for the US Navy’s Hydrographic Office was connected with cartography, this is not correct. He was a civil engineer and inventor of a swivelling head block transfer bridge for transferring railway trucks to and from canal barges that is still known as the Mallery Type. He was an enthusiast for old maps and his archaeological opinions were a long way from the mainstream. In 1951, he published Lost America: The Story of the Pre-Columbian Iron Age in America, in which he argued that there was an Iron Age in North America, inaugurated by Viking settlers. He was, to put it bluntly, a crank.

Piri Re‘is’s map as replotted by Ayşe Afet İnan, showing placenames that identify places in Argentina

Piri Re‘is’s map as replotted by Ayşe Afet İnan, showing placenames that identify places in Argentina

We can dismiss Mallery as an authority, but does this mean that Hapgood was also wrong to identify the land at the bottom (south) of the map as Antarctica? To see it as such, one must ignore the placenames written in this area, as transcribed in Ayşe Afet İnan’s The Oldest Map of America, Drawn by Pirî Reis (1954, Ankara). They include Rio de laplata, San Matias, Porto Deseado and Porto San julean. These are clearly the Río de la Plata, Golfo San Matías, Puerto Deseado and Puerto San Julián. In other words, this is a depiction of the coast of Argentina, twisted through 90° to fit onto the parchment! There is no depiction of Antarctica here.

Hapgood brought a series of maps – principally those of Orontius Finaeus (1494-1555), Hadji Ahmed and Gerardus Mercator (1512-1594) – to bear on the question of knowledge of an Antarctic continent at a much earlier date than is usually believed. The maps he used are superficially impressive: they depict a continent that somewhat resembles what we now know to be the shape of Antractica, albeit one much larger than the real continent. In particular, they lack the Antarctic Peninsula, the continent’s most prominent and characteristic coastal feature.

The Orontius Finaeus map of 1531

The Orontius Finaeus map of 1531: note that Terra Australis (the supposed Antarctic continent) is recenter inventa sed nondum plene cognita (“recently discovered but not yet fully known”) and appears to include the northern coast of Australia

Those who get excited by these supposed maps of Antarctica that pre-date its discovery take the maps as if they exist in a vacuum. They completely ignore books and papers written by the cartographers themselves, which often explain the methods they used. Piri was a careful scholar who listed his sources; they ignore the fact that those who depicted a southern continent did so on the basis of speculation about the balance of land in the two hemispheres; they fail to read the captions on the maps that make it clear that certain elements are conjectured or recently discovered.

There is nothing in these early modern maps, then, that needs explanation. We understand a lot about the context of their production and often have the very words of those who made them. We know their sources and, much of the time, the voyages of discovery that enabled Arabs and Europeans to chart previously unknown coastlines. These maps are interesting for what they show and also for what they do not: the Piri Re’is map, for instance, does not show inland details as it was made by sailors as a navigation aid, quite different from von Däniken’s idea that it was copied from an ancient aerial survey. The real mystery is why so many fringe writers continue to promote them.