"Decade after decade the historians of cartography (and the historians of astronomy too) have, with barely an exception, gone on, paying no attention, assuming that because there were no medieval flat-earthers (or hardly any), medieval Aristotelians shared the Ptolemaic concept of the terraqueous globe (something of an achievement, since they had never read Ptolemy), when in fact that concept was a bold and novel hypothesis in 1492. . . ."

"Brotton, like Woodward before him, assumes that ever since the ancient Greeks there has been a concept corresponding to our concept of Earth ('the image of the Earth') or the world ('a history of the world').  This is to miss perhaps the most remarkable development in the whole history of cartography.  In classical Latin it is hard to find terra used to mean land-and-sea, let alone the whole sphere of the Earth.  Mundus or 'world' also meant universe — we would say cosmos.  In medieval Europe there was no object corresponding to the terms 'Earth' or 'world' as we use them.  Terra meant the element earth (one of the four Aristotelian elements), or land as used to denote a country (the Holy Land is the terra sancta).  Where we (and Ptolemy) see the whole Earth, . . . medieval philosophers and astronomers saw two objects:  the sphere of the element earth, which ought to be at the centre of the universe, and the sphere of the element water, which was believed to be ten times larger — this was the result of trying to combine Greek philosophy with Genesis.  Since the sphere of water was so much larger than that of earth, there ought to be no dry land at all, had God not ensured that the sphere of earth was displaced from its central position (although it was held that it still overlapped with the centre of the universe, a claim which, Copernicus later pointed out, flouted the elementary principles of geometry).  Dry land was thus a sort of floating island in an ocean of ocean.
     "According to this two-sphere theory, the centre of gravity of the combined spheres (the point around which they would hang in equilibrium) differed from their geometrical centre, earth being heavier than water, and from the centre or centres towards which heavy objects would fall (it being moot whether earth and water fell towards the same point).  Because the earth was a sphere floating in a larger sphere, dry land —  the inhabited world —  was necessarily circular in shape, which is, consequently, the shape adopted by most medieval mappaemundi, or maps of the known land masses.  The known world (essentially the world known to Alexander the Great) occupied less than a quarter of the surface of the globe (or rather, of the sphere of water), which was why there was no point in making terrestrial globes, as they would have represented either a sphere of earth that was largely underwater, or a sphere of water that was largely devoid of land.  Crates of Mallos (150 BC) had hypothesized that there might be three unknown worlds spread across the surface of the globe, in addition to the known world, so that the globe itself had a generally symmetrical shape, . . . but Christian philosophers rejected the very idea of 'antipodes' as they held these implied human beings that were not descended from Adam.  The Christian medieval two-sphere system was thus asymmetrical, the circular known land mass centring on Jerusalem, while the sphere of waters had, like the universe as a whole, a North-South axis.
     "The first Christian globes were constructed after the rediscovery of Ptolemy's geography (published in 1475) to explore the heterodox theory that the known world covered a much wider extent than previously imagined, so that it might be possible to sail West in order to reach the spice islands — Ptolemy's account of earth and water making one globe was debated when the feasibility of Columbus's expedition was rejected by a commission appointed by the King of Spain.  The discovery of America showed that this hunch was basically right — the globe might be much larger, and India might be much further away when travelling West (and nearer when travelling East) than Columbus had thought, but wherever one went there was land to be found, and the antipodes turned out to be a geographical reality.  The first map to show this clearly was Waldseemüller's world map of 1507. . . .
     "Soon after Waldseemüller's map was printed, perhaps in 1511, Copernicus saw it and grasped that it showed a symmetrical world where earth and water were interlocked.  The two spheres had become one globe.  This was a world which could rotate on a North-South axis without wobbling.  It was perhaps this realization which emboldened him to abandon a geo-heliocentric model (of the sort later advocated by Tycho Brahe) in favour of full heliocentrism.  De Revolutionibus (1543) opens with a defense of this new understanding of the Earth which did not yet have a name — later it would be called 'the terraqueous globe'.  Copernicus seems to have been one of the very first to make use of this new idea. . . .
". . . the two spheres theory was described by Lynn Thorndike in 1929 and discussed at great length in a posthumous text by Pierre Duhem (1861-1916), who, unlike Thorndike, continues to be widely admired. . . . Copernicus's debt to Waldseemüller was identified by Edward Rosen in 1943, but it is so rarely mentioned that it is hardly surprising that Brotton has no knowledge of it.  Decade after decade the historians of cartography (and the historians of astronomy too) have, with barely an exception, gone on, paying no attention, assuming that because there were no medieval flat-earthers (or hardly any), medieval Aristotelians shared the Ptolemaic concept of the terraqueous globe (something of an achievement, since they had never read Ptolemy), when in fact that concept was a bold and novel hypothesis in 1492. . . ."

     David Wootton, "No words for world:  common misunderstandings in the use of maps," Times literary supplement no. 5715 (October 12, 2012):  9 (8-9), reviewing Jerry Brotton's A history of the world in twelve maps (London:  Allen Lane, 2012).
     Alfred Hiatt, of the Department of English, Queen Mary, University of London, objects in no. 5716 (October 19, 2012):  "not only was this Aristotelian theory [of "the 'Christian medieval two-sphere system'"] in vogue only from the thirteenth century, it was more flexible than Wootton allows, and coexisted with Neoplatonic theories that did not posit separate spheres for earth and water.  Nor did Christian philosophers simply 'reject' the idea of antipodal places and peoples.  They variously doubted, debated and continued to transmit classical theories of the Antipodes throughout the course of the Middle Ages.  Scholars such as Albertus Magnus and Roger Bacon (both well versed in Aristotle) affirmed the existence of land beyond the known world, and even argued that it could be reached by Europeans.  Such land was represented from the eighth to the fifteenth centuries on zonal maps of the world.  Those images, which depict southern and northern hemispheres, alone show the falsity of Wootton's assertion that 'in medieval Europe there was no object corresponding to the terms "Earth" or "world" as we use them'.  In short, medieval European understanding of the world and its image was far more subtle and various than Wootton thinks.  That does not in itself lessen the significance of Copernicus, but it does mean that the familiar chiaroscuro narrative of medieval darkness and Renaissance light that Wootton tries to resurrect is essentially inaccurate" (6).  Wootton had not replied to this criticism by the end of November, though his 2015 The invention of science seems to take it into account (and includes Hiatt's 2008 Terra incognita:  mapping the Antipodes before 1600 in its bibliography).