Debates and Comments

Comment on Gavin Menzies, 1421: The Year China Discovered the World
Contribution by J.R. Masson of Swaziland

The orthodox eurocentric view, that the great European voyages of exploration towards the end of the fifteenth century initiated the Age of Discovery, has been challenged in a book by Gavin Menzies entitled 1421 : The Year China Discovered the World1.  Menzies claims that one of the great Chinese fleets which sailed for India, Arabia and the east coast of Africa during the first quarter of the fifteenth century went much further after they reached the African coast.  The 1421 fleets, he affirms, rendezvoused off Sofala, which was the southernmost important Arab trading port, sailed down the African coast, rounded the Cape of Good Hope and traveled up the west coast to the Cape Verde Islands.  There they split into three fleets.  One sailed westwards from the Canary Islands, like Columbus, to the West Indies and North America, while the other two took a more southerly route to the coast of Brazil, then followed the coast of South America down to Patagonia and Terra del Fuego.  The two fleets then separated, one sailing eastwards across the Southern Ocean, touching Antarctica and Australia before heading back to China, while the other rounded Cape Horn and sailed north to the western shores of North America before returning to China.  Both fleets arrived back in 1423, two years after they had set sail.

Before examining the evidence which Menzies produces to authenticate these epoch-making voyages, it is necessary to read his Introduction (p. 3-12), which explains what started him off on what was to be a fourteen-year-long research project.  He was inspired by the 1424 chart of Europe and part of Africa by Zuane Pizzigano which contains a group of four islands far out in the western Atlantic, named Antilia, Satanzes, Saya and Ymana. Because of the bold colouring of the two largest islands, Antilia and Satanzes. Menzies concluded that they had been recently discovered, and was convinced that they represented Puerto Rico and Guadeloupe, the latter because it had lettering indicating an active volcano.  As is explained below, if he had studied the medieval mythology and cartography of the North and mid-Atlantic countries he could not have come to this conclusion.  As it was, however, he then had to ask himself who had discovered these islands shown on Pizzigano’s 1424 map.  He then came to his second conclusion, at that stage a provisional one, but also an incorrect one, as will be shown later, that the only maritime nation of the time capable of sailing the vast distances involved in reaching the West Indian Islands was China.  He may have been guided towards this view by the fact that he had spent the first five years of his life in China, had made several subsequent visits to that country, and had served seventeen years in the Royal Navy where he had acquired considerable knowledge and experience of ocean navigation especially in the Southern Hemisphere.  He would therefore have had a good grounding in Chinese history.  At any rate, arrive at this conclusion he did, and spent the next fourteen years trying to put flesh on its rather insubstantial bones !

To return to the Atlantic islands shown on Pizzigano’s chart, there are many which figure in medieval mythology and cartography.  These “lost islands” of ancient, medieval and even modern mythology are a major feature of Atlantic cartographic history.  A writer on “lost islands” Stommel 2, states: “But prior to Columbus’ achievement, Portuguese navigators were sailing to the west of the Azores, searching for such islands as Antilia, said to have been settled by an archbishop of Oporto as early as the year 734 AD, or the island of Santanezes to the north of Antilia.”  Probably the best known of these “lost islands” is Brasil or O’Brasil whose chart life was inspired by Celtic mythology, perhaps the Island of Eternal Youth, Tir na Nog, and which eventually found its way on to Admiralty charts.  It was sought by Bristol merchants who sailed far to the west of the Azores, and were followed by John Cabot who reached Newfoundland or Nova Scotia in 1497, or even earlier, to become a near contemporary of Columbus in the discovery of the New World.  The other great legend of the Atlantic was that of St. Brendan, an Irish monk who sailed west and north in the first half of the 6th century AD.  If his Navigatio Sancti Brendan Abbatis is to be believed, he reached the Faroe Islands and the fiery volcanoes of Iceland (the Edge of Hell), as well as finding his Promised Land, probably the legendary Tir na Nog.  This episode of Irish history/mythology is well told by Moorhouse 3.  These and many other ancient voyages of discovery, for instance that of Pytheas of Thule, of curdled sea and midnight sun fame in about 300 B.C., generated a vast amount of fact and folklore which it is now wellnigh impossible to disentangle.  Pizzigano’s islands, especially Antilia, may have their origins in this mix.  A pointer to this mixture may be that Toscanelli’s map of 1474 shows a large island, St Brandan, though well to the south of Antilia.

However, they also have an undoubted cartographic pedigree.  For example, they are shown on Johan Ruysch’s map of 1507, which was the first printed map of the world.  The map contains a commentary on the islands, which has been published by Nordenskiöld4.  He draws attention to an island between Iceland and Greenland with the legend that it was totally destroyed in 1456 (by volcanic eruption).  This incident is also mentioned in the Norse Sagas, which are much earlier than Pizzigano’s chart, and therefore emphasize the antiquity of the “lost island” cartographic tradition.  Nordenskiöld also refers to two islands between Greenland and Nova Terra, the legend to which he had translated as “an encounter between European sailors and Esquimaux in which the former did not play the winning part.” The latter are described as “Demons.”    He goes on to explain that these islands are shown on Andrea Bianco’s 1436 map, and named as Antilla (Antillia) and Satanaxio.  The latter was also known as Devils’ Island, but whether there is any connection with the “Esquimaux” is not stated.

Returning to Ruysch, Nordenskiöld notes a large island in mid-Atlantic called Antillia Insula which “had been searched for in vain, but that it had been discovered long ago by the Spaniards, whose last Gothic King, Roderik, had taken refuge there from the invasion of the Barbarians” (Moors).  He explains that the legend on the map “depends on a myth which has played a certain part in the history of geography, and from which is derived the present name of the islands between Florida ad the northern coast of South America.”  He also pointed out that Antilia appears on a portolan chart of 1425 in the library of Weimar, and on Martin Behaim’s globe of 1492.  The two islands may also be those shown on the northwestern edge of D’Avezac’s map which is produced in Nordenskiöld’s Facsimile-Atlas (Fig. 41:73).  Finally,  Antilia was referred to in a letter from Toscanelli to the King of Portugal in 1474, suggesting that a voyage to the west from the Canaries could be broken by a stopover at “the island of Antillia, which is known to you” 5.

All of this will be well known to historians and cartographers engrossed in the enormous body of literature generated by the discovery of the New World.  However, I will add the word of Felipe Fernandez-Armesto, in his biography of Columbus 6.  He has this to say about Antillia (Satanazes, being much to the north, seems to have disappeared with the detailed exploration of the North Atlantic and the search for a North-West Passage). “To the mythical islands commonly assigned imaginary positions in fourteenth-century maps – those of St. Brendan, St Ursula, and Brasil – a Venetian chart of 1424 added large and alluring islands, including Antillia, identified with the world of Seven Cities to which, in a  legend not unlike that of St Ursula, Portuguese refugees from the Moors were held to have repaired in the eighth century” and so on.  The 1424 chart was presumably the Pizzigano chart.  Nordenskiöld also refers to an assurance by Andrea Bianco on his 1448 chart, that an “authentic island” lay 1500 miles out in the equatorial Atlantic.  This was no doubt Antillia.  Indeed, Fernandez-Armesto describes  Antillia” as a catch-all name for potential new discoveries (p. 20).  When Columbus returned to Lisbon after his first voyage, it was rumored that he had found the mythical lost land of Antillia, which in time was transformed into a chain of islands known as the Greater and Lesser Antilles.

An entirely antipodean view of the lost islands is presented by Fuson 7 who identifies Antilia as Taiwan and Satanazes as Japan in the context of a world view of the great Ocean Sea stretching from the West to the East without the unknown Americas intervening.  This perceptive view would add a further element of irrelevance to Menzies’ interpretation of Pizzigano’s chart.

The next controversial issue is the claim that, after completing their political and trading business, the huge Chinese fleets proceeded down the east coast of Africa and rounded the Cape of Good Hope, thus enabling them to set off on the astounding voyages of discovery described by Menzies.  Their presence in east African waters in 1421 is well documented.  The Chinese fleets, known as “treasure ships,” made in all seven voyages in the early part of the fifteenth century; 1405-07 to Ceylon and Calicut, 1407-09 and 1409-1411 to the coast of India and Ceylon, 1413-15 to Hormuz and Bengal, 1417-19 to Hormuz, Aden, and Melinda on the African Coast, 1421-23 also as far as the African coast at Brava and Mogadishu, and finally in 1431-33 to Hormuz and Mecca 8.  These voyages were under the command of the eunuch Admiral Zheng He (Cheng Ho) and were brought to an abrupt end by a return to traditional Imperial policies of self–containment.  A scholarly account of this about-turn is given by Finlay 9.

One of the main supporting planks of Menzies’ case is the Fra Mauro mappa mundi of 1459.  This map, which falls into the category of monastic world maps, is oriented with south at the top.  When turned round Africa can be seen as a wedge-shaped continent trending eastwards, with its western and southern edges open to the sea and ending in a triangular “island.”  There is no western bulge above the Gulf of Guinea, indicating that the cartographer had no knowledge of its configuration.  Similarly, the triangular “island” called Diab is an unusual concoction of African geography.  It contains Arab place-names derived from East Africa, i.e. the modern Zanzibar, Sofala, Kilwa and Mombasa, and the narrow passage separating Diab from mainland Africa may represent the Mozambique channel, and Diab itself may indicate the great island of Madagascar 10.  However, an alternative interpretation was presented by Kimble 11, who considered that the Cap de Diab was in fact Cape  Guardafui (the Horn of Africa).  As he explained, “The only region with which it is at all comparable within the limits set by Mauro’s knowledge of the interior is the Somali peninsula, and it is significant that in a contemporary document in the library of S. Michele at Murano  we are told that “Diab is a great province in parts of which there is abundance of every good things, its principal town being called Mogadis; which can be none other than Mogadoxo (Mogadishu) or the Somali coast.  ‘Cavo de Diab’ in that case becomes Cape Guardafui.  Consequently, the degree of verismilitude possessed by the southern portion of African coastline is due largely to the exigencies of the map’s circular form.”  Diab was therefore a populous and productive region with a well-known town on the Somali coast. By comparison the southern tip of Africa was totally unknown.  

There is of course the matter of the annotation on the map, reading in part, “About the year of our Lord 1420 a ship or junk of India on a crossing of the Sea of India towards the islands of men and women was driven beyond the Cape of Diab and through the green islands and the darkness towards the west and southwest for forty days, finding nothing but air and water, and by their reckoning they ran 2000 miles and fortune deserted them.  They made the return to the said Cavo de Diab in seventy days and drawing near to the shore to supply their wants the sailors saw the egg of a bird called roc.”  As Crone has pointed out, the legend of the roc is the fabulous bird of the Arabian Nights  and Sinbad the Sailor.  An Arab chronicle 500 years earlier related a similar tale of a storm off Sofala and the crew encountering a roc .  The source is not stated, but it may be one of the many Arab historians, possibly Abu Zaid, c. 920 AD, referred to by Rochlin 12.  Nevertheless, there was a widely held view that such a sea route must exist, long before it was demonstrated as fact by the Portuguese at the end of the century.

Menzies’ approach to the Fra Mauro map and its annotation has been less cautious. He viewed the map in Venice;  it is the only copy of the lost original which was sent to the King of Portugal in 1459, and was completed shortly before or after Fra Mauro’s death in 1460.  Menzies was immediately convinced that the Cap de Diab was the Cape of Good Hope, and that the annotation was proof positive that a Chinese junk had rounded the Cape of Good Hope in 1420 (or 1421).  He also identified the roc as an ostrich.  Menzies’ conviction was strengthened by establishing to his satisfaction that Fra Mauro’s information on the 1420 passage came from a “trustworthy person” whom he identified as Niccolo da Conti (p. 93).  This fellow Venetian had apparently brought Fra Mauro a chart showing the southern tip of Africa which he, da Conti, had obtained during his voyages with the Chinese fleet.  However, apart from his own account of his travels, published in 1434, his personal knowledge of their voyage beyond Sofala seems unsubstantiated, such as, “his subsequent travel in Chinese junks was to prove a vital link in solving the riddle of where the Chinese fleets had gone in the ‘lost’ years” (p. 87), and “the knowledge that da Conti had acquired in twenty years’ sailing the world from India to the Cape Verde and Falkland Islands, to Australia and China” (p. 252).  Such information surely did not come from his own account, because if it did Ma Yuan did not mention this vital information, nor did Parry 13.  In the context of the second quotation, Menzies is stressing the importance of da Conti’s contribution to Portuguese knowledge of the ocean passage round the Cape before Diaz in 1488 (and also round Cape Horn), but it is more likely that any such prior knowledge came from overland missions to Africa, the Near East and India by the Portuguese themselves, for instance the journey of Pero da Covilham to find the fabled empire of Prester John.  He left Portugal in 1487, reached Ethiopia only in 1493 or 1494 after embarking on a series of journeys from Cairo to Aden, India and Sofala, and died in Ethiopia in c.1525 14.  In about 1490 he had returned to Cairo, met a messenger from the King of Portugal, and sent back a detailed account and a map of all the places he had visited, and all the other information he had gleaned along the way.

Thus far the evidence for Chinese fleets rounding the southern tip of Africa does not seem to be compelling.  However, there is still one piece of evidence to be considered.  This is the Kangnido map of 1402, a world map as seen by its Chinese-Korean cartographers, which shows a “slimmed down” Africa without the western bulge but including the southern tip and a large inland sea occupying much of the interior.  Menzies has explained the absence of the bulge as being determined by the sailing conditions of the time and the inability to determine longitude accurately.  If the crew had been better equipped they would have realized that they were traveling west along the west African coast, and the map which they subsequently made would have shown the bulge.

As in the case of the Fra Mauro map, however, Menzies has not researched the cartographic history of the Kangnido map.  This is provided at length by Harley and Woodward15.  (Although this book appears in his Bibliography, it is not referred to in the text).  Perhaps by 1994 Menzies had reached the point of no return in developing his thesis of Chinese exploration of the world in 1421-23.  He does conclude that the Kangnido map was completed from many different sources (p. 97), but he does not pursue the matter further.  For example, in view of the long, perhaps 700-year-long, connection between Arab and Indian voyaging and trade across the Indian and Chinese seas, and the reciprocal Chinese mercantile activity after sea-going junks had been designed (about 1100 AD), it would have been valuable to assess the extent of these foreign influences on Chinese geographical knowledge and mapping.  There is also the Mongol influence (the Yuan Dynasty 1279-1368) bringing information from the west.  In addition, the long pedigree of the Kangnido map itself, probably dating back to 1320,16 should have precluded a direct connection with the 1421-23 voyages.  The 1320 map referred to above included the southern half of Africa.  It exists only as a later reproduction in a 1560 atlas, which makes it difficult to establish whether Africa was a later addition. This atlas is also known as the Mongol Atlas 17.  (Chang and Fuchs also appear in the Bibliography but not in the text).

The general conclusion is that although the Kangnido map contains a representation of Africa, it does not reflect direct Chinese knowledge of the continent, but information derived from a variety of foreign sources over a long period, and compiled from earlier Chinese maps themselves based on foreign sources.  There is thus no need to impute that Chinese fleets were the source of this information.

If the conclusion reached is that the Chinese “treasure ships” did not enter the Atlantic Ocean by way of the southern tip of Africa, their later exploits in the Atlantic would be vitiated.  In any event the misinterpretation of the “lost islands” of Pizzigano’s chart, referred to earlier, would act as a “double-whammy” to the argument for the discovery of the West Indies in late 1421 (p. 246).  Even assuming that the fleets did touch at the Congo River (Matadi = the Garbin of da Conti) and the Cape Verde Islands, the evidence at both of carved stones of Chinese origin is not convincing.  Why were no photographs published?  Menzies had visited and photographed the Cape Verde Islands site, so this piece of evidence, vital to his argument, should have been exhibited. Or was the stone just a Portuguese mariner’s burial stone or perhaps a reworked megalith of prehistory?  If he did actually see the Matadi Falls (Garbin) stone, which seems doubtful, he could have obtained a photograph of it from his source Mr. Chevalier, whose publication does not however appear in the Bibliography.

However, all my negative comments on this part of the book they should not be construed as a complete condemnation of it.  It is possible that the annotation on Fra Mauro’s map did reflect an incident when a Chinese junk was driven south from Sofala, through the Mozambique Channel between Africa and Madagascar or to the east of that great island, and then swept into the Southern Ocean towards the west coast of Australia and then back to the east African coast via the”‘rocs” of Madagascar, all within the overall 110 days at sea.  This possibility of a southern extension to the 1421 voyages is held by Chang who has deduced from the Chinese chart illustrated in Menzies (p. 86-87) that some ships of the 1421 fleets may have proceeded south through the Mozambique Channel and into the Southern Ocean, where they may have sighted the snow-clad peaks of Kerguelen Island before turning north and west to Calicut on the Malabar coast of India 18.  What is perhaps equally likely is that Chinese junks, not necessarily the “treasure ships” of 1421, were trading through the East Indies and made unplanned landfalls along the coasts of Australia.  It is also possible that the numerous landfall sites located along the Pacific coasts of North and South America represent Chinese maritime activity across the Pacific Ocean, though some may not stand up to detailed archaeological scrutiny.  A separate examination of these sites would be needed to distinguish between old-style anecdotal history and scientifically investigated archaeological sites.

To sum up, Menzies has accumulated an enormous amount of information on the phenomenon of burgeoning Chinese maritime activity in the early part of the fifteenth century, much of it of historical value, but all of it woven round an idée fixe.  This places it in the category of cult writing; that is, it concentrates on a “grand vision,” but surrounds it with unsound or imaginary evidence, and rejects contrary information. However, it is a good read for this type of market. 


1   (London: Bantam Press, 2002). Reviewed in Terrae Incognitae, xxxvi (2004): 118.
   H. Stommel, Lost Islands: the Story of Islands that have vanished from Nautical Charts (Vancouver: University of British Columbia Press, 1984).
   G. Moorhouse, Sun Dancing; a Medieval Vision (London: Weidenfeld and Nicholson, 1997).
   A.E. Nordenskiöld, Facsimile-Atlas (New York: Dover Publications, 1973).
   J.N. Wilford, The Mapmakers (London : Junction Books, 1981).
   F. Fernandez-Armesto, Columbus (Oxford : Oxford University Press, 1992).
   R.H.G. Fuson, Legendary Islands of the Ocean Sea (Sarasota: Pineapple Press, 1995)[reviewed in Terrae Incognitae, xxvii (1995): 84].
   J.J.L. Duyvendak, China’s Discovery of Africa (London: Probsthain, 1949).
   R. Finlay, “The Treasure-Ships of Zheng-He: Chinese Maritime Imperialism in the Age of Discovery,” Terrae Incognitae, xxiii (1991) : 1-12.
  G.R. Crone, Maps and their makers (New York: Archon Books, 1978).
  G.H.T. Kimble, Geography in the Middle Ages (London:  Methuen, 1938).
  S.A. Rochlin, “Early Arab Knowledge of the Cape of Good Hope,” Africana Notes and News,  13:6, p. 32-47.
  J.H. Parry, The Discovery of the Sea (New York:  Dial Press, 1974).
  E. Sanceau, The Land of Prester John: a Chronicle of Portuguese Exploration (New York: Knopf, 1944).
  J.B. Harley and D. Woodward, eds., The History of cartography: Cartography in theTraditional East and Southeastern Societies, vol. 2, book 2 (Chicago: University of Chicago Press, 1994).
  K.-S. Chang, “Africa and the Indian Ocean on Chinese Maps of the Fourteenth and Fifteenth Centuries,” Imago Mundi, 24 (1972): 21-30.
  W. Fuchs, “Was Africa already known in the fourteenth century ?” Imago Mundi,  10 (1958) : 50-51.
  K.S. Chang, “The Ming Maritime Enterprise and China’s Knowledge of Africa prior to the Age of Great Discoveries,” Terrae Incognitae,  iii (1971): 33-44.

Society for the History of Discoveries home page