As it has been more than a quarter century since the publication of my book The Myth of Continents (co-written with my wife, Kären Wigen), I have long thought that the topic deserved reconsideration. An opportunity came in May 2025, when I was invited to give a talk in Stanford’s Distinguished Faculty Lecture Series, associated with the University’s Continuing Studies Program. I have recently redelivered this lecture in an expanded version, divided into two parts. The first part is now available for viewing at the GeoCurrents YouTube channel, which can be found here. The second part of this lecture will be posted soon.
Myth of Continents Lecture
My interpretation of the division of world into continents has not changed much over the past few decades. But I have delved a bit deeper into the origin of this system of global division. I now trace it back not just to ancient Greece, but more specifically to the Ionian School (or Ionian Enlightenment) of the sixth century BCE and its so-called pre-Socratic philosophers. Ionian Greeks from the city of Miletus extensively traded in, and settled the shores of, the Black Sea, which led them to divide their known world into “Europe” and “Asia.” “Libya” (Africa) was soon added, thus splitting the terrestrial world into three parts. (A few of the slides from this part of the lecture are posted below.)
In Part One of this two-lecture series, I first look into the different definitions of the term “continent” and explore the conceptual problems that emerge from their varied applications. I then outline the evolution of the continental model, from the ancient Ionians to the twentieth century. In the second lecture, I first examine the geological understanding of continents, then move on to the intellectual problems that emerge when the world is divided in such a manner, and conclude by analyzing the world-regional system that emerged as the main alternative to the continental scheme after World War II.
As noted in the previous post, most maps of continents found in online images searches divide several countries, particularly Russia and Turkey, along conventional continental lines yet avoid dividing Indonesia in the same manner. Evidently, in the popular cartographic imagination, geopolitical factors override geophysical factors in the delineation of continents in some instances but not in others. A few maps, however, consistently follow political divisions in delineating continents, just as a few consistently follow features of physical geography.
Cartographers who map continents along political borders must use judgement when slotting countries that are conventionally regarded as transcontinental. In general, this is a minor issue. In conventional reckoning, there are only a few transcontinental countries, although the list grows considerably longer if one if one includes dependencies and territorial claims within a given country’s boundaries and regards “Oceania” as a continent (compare the two maps posted below). More to the point, in all cases but one a substantial majority of both the territory and population of transcontinental countries are situated on one continent. I do not know any map, for example, that puts Egypt in Asia rather Africa because the Sinai Peninsula is on the Asian side of the continental divide. Some maps do, however, place Turkey in Europe rather than Asia regardless of its territorial and population imbalance. They presumably do so largely because of Turkey’s stalled application to join the European Union.
Transcontinental Countries Map 1
Transcontinental Countries Map 2
Russia is the one country with a questionable position in a classification scheme that does not allow countries to be continentally divided. Most the territory of the Russian Federation is technically in Asia, as it is east of the Ural Mountains, whereas most of its people live in its “European” heartland. Some geopolitically based continental maps thus put Russia in Asia while others put it in Europe. Both depictions have problems. Extending “Europe” to the Pacific Ocean, as can be seen on the DataBayou map posted below, makes the “continent” all but unrecognizable. Slotting European Russia into Asia, as is done in the WorldAtlas map posted below, requires the addition of an Asian exclave (Kaliningrad) on the Baltic Sea between Poland and Lithuania, which is also visually jarring.
Russia Mapped as Part of Europe
Russia Mapped as Part of Asia
The cartographers who made this WorldAtlas map could have ignored Kaliningrad’s geopolitical status and mapped it as part of Europe, thereby retaining clean continental borders. They mapped France’s exclaves – Guiana, Martinique, Guadeloupe, Réunion, and Mayotte – after all, as belonging to South America and Africa rather than Europe. In these cases, in other worlds, they ignored the geopolitical standard that they used elsewhere when partitioning the continents. Note as well their mapping of Hawaii, which they misplaced, in North America and of Greenland in Europe. Their depiction of Central America and the Caribbean as parts of South America rather than North America also defies conventional continental categorization. Yet again we find ourselves in a conceptual morass.
As the previous GeoCurrents post argued, the division of the terrestrial world into a handful of continents derives in part from the division of the Earth’s crust into tectonic plates. But there is another geological factor equally pertinent to the concept of “continent.” It refers not to individual landmasses but rather to a particular kind of crust – the outermost rocky layer of the planet – regardless of whether it extends above sea level. If continents are defined based on continental crust, an entire “continent” could conceivably be well below sea level.
As the diagram posted below show, the Earth has two kinds of crust: oceanic and continental. Oceanic crust is relatively thin and dense, rich in iron and magnesium. Less-dense continental crust, rich in aluminum silicates, is much thicker. Being light and thick, continental crust generally extends above sea level. But the continental shelves that skirt most coastlines of continents are below sea level even though they are composed of continental crust. Several isolated fragments of continual crust are almost entirely submerged. The largest of these inundated “continents” is Zealandia, which covers approximately 4.9 million km2 (1.9 million sq mi). Only about seven percent of Zealandia is currently above sea level, mostly the two main islands of New Zealand and New Caledonia. Once connected to the massive southern continent of Gondwana, Zealandia began to rift away roughly 85 million years ago. As this occurred, its crust stretched and thinned, resulting in a lower elevation than the major continents.
Oceanic and Continental Crust Diagram
Topography of Zealandia Map
Although oceanic crust is generally well below sea level, exceptions occur. Extensive volcanic eruptions associated with hotspots can generate exceptionally thick oceanic crust, creating sizable island. Important examples include the Hawaiian Islands, the Mascarene Islands (Reunion and Mauritius) in the Indian Ocean, and, to some extent, Iceland. Iceland is a complicated case, however, as it is essentially oceanic but does contain a few fragments of continental crust. Kerguelen, in the southern Indian Ocean, is another such “composite” island. It is associated with a hotspot and as a result is mostly covered by rock of oceanic origin. But these rocks sit above a continental fragment that rifted away from Gondwana roughly 100 million years ago. As a result, Kerguelen is depicted as a microcontinent in the map posted below.
Areas of Continental Crust Map
During glacial periods, when the sea dropped due to the massive amount of water in continental ice caps, most continental shelves were above sea level. The continents were therefore larger than they are today, with a much closer correspondence between continental crust and dry land. Zealandia, however, had subsided to such a degree that even at the Last Glacial Maximum (LGM) around 20,000 years ago, only around ten percent of it extended above sea level. But that was enough to join the islands of New Zealand together, forming a significantly larger landmass than what exists today.
The Earth at Glacial Maximum 20,000 Years Ago Map
New Zealand at Glacial Maximum Map
As I was not able to find an adequate map of continental crust, I made my own, which must be regarded as a crude approximation (posted above). Several difficult judgement calls were necessary, as many island chains located tectonically active areas are composed of complex mixtures of oceanic and continental crust. Most, however, are dominated by oceanic crust, and as a result do not appear on this map (examples include the Andaman and Nicobar islands and the Kuril and Aleutian archipelagos). The Ryukyu Archipelgo, on the other hand, is composed mostly of continental crust and thus appear on this map as a narrow peninsula extending south from Japan.
Needless to say, it is difficult to derive the standard seven-continent model from this map of continental crust. A six-continent model, however, can be construed, albeit one quite different from the conventional six-continent model. Its constituent elements would be North America-Eurasia, Africa, South America, Antarctica, Australia, and Zealandia.
Many maps, in my view, are also work of art, regardless of their creators’ intentions. Some maps, however, are explicitly made as art, and as such are not designed to convey spatial information beyond that found in the mere geographical shapes that they contain. Increasingly, it can be difficult to tell whether a given map was created as cartographic art or as a visualization of scientific information.
Consider, for example, the three figures posted below. Two are derived from images posted on the intriguing Wayne Baisey Blogspot under the heading “World Map Art Large.” The other is taken from a Wikipedia article and is based on recent scientific research. Can you tell which is which?
Map Art or Scientific Cartography? 1
Map Art or Scientific Cartography? 2
Map Art or Scientific Cartography? 3
The middle map is in some ways the most abstract, but its title gives it away: “The New Global Tectonic Map with the Subdivision of the Continents, Oceans, and Mobile Mountain Belts in Ca. 1200 Smaller Plates.” As this map shows, the Earth is divided into many tectonic plates – vastly more than those that are portrayed on conventional tectonic maps. As the author of the Wikipedia article “List of Tectonic plates,” where the map is found, argues, “The latest studies have shown that microplates are the basic elements of which the crust is composed…” (see the quotation below). As this author further explains, referring to the map posted above [and below] and to the scientific papers on which they are based:
Minor Plates: These smaller plates are often not shown on major plate maps, as the majority of them do not comprise significant land area. … For purposes of this list, a minor plate is any plate with an area less than 20 million km2 (7.7 million sq mi) but greater than 1 million km2 (0.39 million sq mi).These plates are often grouped with an adjacent principal plate on a tectonic plate world map. For purposes of this list, a microplate is any plate with an area less than 1 million km2. Some models identify more minor plates within current orogens (events that lead to a large structural deformation of Earth’s lithosphere like the Apulian, Explorer, Gorda, and Philippine Mobile Belt plates.The latest studies have shown that microplates are the basic elements of which the crust is composed and that the larger plates are composed of amalgamations of these, and a subdivision of ca. 1200 smaller plates has come forward.
New Global Tectonic Map
I find this map visually striking and both scientifically and cartographically impressive. But I do think that the cartographer should have deleted the volcanos, portrayed as “yellow dots,” and placed them instead on a companion map. As it is, in areas that are rich in volcanoes the lines showing tectonic subdivisions are completely obscured (see the map detail posted below).
Detail from New Global Tectonic Map, Western North America
Needless to say, the conventional continents are not readily visible on this map. Some their coastlines, however, are apparently outlined in green as “terrane (microplate) boundaries in the continental blocks.” In actuality, however, these lines correspond to the boundaries of continental shelves, which are covered by the sea, rather than to the continents themselves. This important point will be elaborated in the next GeoCurrents post.
Tectonic plates are the basic building blocks of the Earth’s lithosphere, its outermost rocky layer. As these large segments of crust slowly move, landmasses and sea expanses are gradually rearranged. The current configuration of tectonic plates shows a tight connection with the architecture of continents: North America is on the North American plate; South America is on South American plate; Africa is on the African plate; Australia is on the Australian plate; and Antarctica is on the Antarctic plate. All these plates include large stretches of the sea, but they are anchored on their eponymous continents. The only exceptions are Europe and Asia, which are both located on the Eurasian plate. These correspondences indicate that the six-continent model – in which “Eurasia” supplants “Europe” and “Asia” –most closely represents the geological foundation of the Earth’s major landmasses. The six-continent model has the additional advantage of remaining faithful to the implicit definition of continents: large landmasses that are separated – or almost separated – from other large landmasses by stretches of the sea.
Relationship between Continents and Tectonic Plates Map
A closer inspection of the tectonic map, however, reveals that the connection between continents and plates is not as tight as it seems. Note that the Eurasian plate does not include either the Indian subcontinent or the Arabian Peninsula, both of which occupy their own plates. By geological criteria, these regions might be considered continents in their own right, although they obviously lack any maritime separation from Eurasia. The actual tectonic situation, however, is more complicated than that. The Indian plate, for example, is sometimes considered to be part of a much larger Indo-Australian plate, which would seemingly imply that India and Australia are part of the same “tectonic continent.” The current consensus, however, is that India and Australia were formerly on a single plate that has either split into two separate plates or is in the process of gradually splitting. In a tectonically informed view, all continents – and oceans – are but temporary entities.
Disconnections between Continents and Tectonic Plates map
Two Depictions of the Indo-Australian Plate map
Other disconnections between continents and tectonic plates are easily located. The North American Plate, for example, includes a huge part of eastern Siberia in Asia (or Eurasia), as well as northern Japan. The same plate does not, however, include much of western California or any part of Baja California in Mexico, which are instead located on the Pacific plate (see the map posted above for these and other discrepancies). More important, the African plate does not include Madagascar and most East Africa, which are instead on the Somali plate. Most global maps of plate tectonics, however, do not depict the Somali Plate. They presumably do so either for simplification or because the Somali plate is relatively “new,” currently in the process of rifting away from Africa.
Detailed maps of tectonic plates reveal an even more intricate situation. An excellent Wikipedia map, posted below, shows that most of eastern Asia is not necessarily on the Eurasian Plate, but sits instead on the Amur, Yangtze, and Sunda plates. Although the Amur and Yangtze plates can be defined as subdivisions of the Eurasian Plate, the Sunda Plate is more distinct. As its Wikipedia article notes, “The Sunda plate was formerly considered a part of the Eurasian plate, but GPS measurements have confirmed its independent movement at 10 mm/yr eastward relative to Eurasia.” Yet as detailed as it is, this map leaves out several important “micro-plates.” Another Wikipedia map, for example, shows that the Somali plate is itself divided into smaller segments. One of these, the Victoria plate, located between the two arms the East African Rift, is not moving in a single direction, but is rather rotating in a counterclockwise manner.
Detailed Wikipedia Map of Tectonic Plates
East Africa Tectonic Plates Map
A tectonic map that includes all micro-plates can become extraordinarily intricate, as the next post will explore.
According to many standard geographical reference works, only a few countries span continental boundaries. World Atlas, for example, lists four “contiguous transcontinental countries”: Russia, Kazakhstan, Turkey, and Egypt. World Population Review adds only Denmark to its transcontinental list, owing to Denmark’s non-contiguous dependency of Greenland, which is oddly excluded from its inappropriate Mercator-projection map (posted below). But as careful cartographers and geographical compilers note, the list of transcontinental countries is considerably longer, especially if one includes non-contiguous cases. The most comprehensive map of such countries that I have found comes from an anonymous Reddit contributor (posted below), as is often the case.
As “Fearlessredditor’s” map indicates, the Republic of Georgia is, by conventional criteria, a contiguous transcontinental country, as is its eastern neighbor, Azerbaijan. According to the most widely used definition, the watershed divide formed by the Greater Caucasus Range separates Europe from Asia in the area between the Black and Caspian seas. Almost all Georgian territory lies south of this divide, but a small area is situated to the north, and is therefore “officially” part of Europe. I have highlighted this remote, sparsely populated area in red on an interesting “Orographic Scheme” map found in the National Alas of Georgia (2018). The road that links one part of this region (Tusheti) to the rest of Georgia has been deemed the “World’s Most Dangerous Road” by one YouTube contributor, but I have been on roads in other parts of the world that I suspect are far more dangerous.
Although now standard, the watershed of the Greater Caucasus is only one of many continental divides that have been inscribed across the Caucasus region. As the Wikipedia map posted below indicates, almost all bifurcate Georgia. Historically speaking, Georgia might therefore be regarded as a quintessentially transcontinental country.
It is also of interest that the boundary between Russia and Georgia passes through many of the highest peaks of the central Greater Caucasus Range, but many of these peaks are located to the north of the drainage divide. A detail of a physical map of Georgia found in the country’s National Atlas (2018) shows this feature; I have highlighted the northward flowing Argun River (Arghuni in Georgian) to make it more clearly visible.
Regardless of its small “European” segment, Georgia is conventionally classified as an Asian country. Most Georgians, however, resent this designation, cogently arguing on cultural, historical, and geopolitical grounds that Georgia should be classified as part of Europe. The National Atlas of Georgia (2018) evocatively describes Georgia as “the balcony of Europe” (p. X). As noted in a Wikipedia article, “Despite its geography, Georgia is considered a European country geopolitically because of its historical, cultural, ethnical, and political ties to the continent.”
In my own view, “Asia” is an essentially meaningless category, and I therefore use the term only in regard to geographical discourse, rather than in regard to geography itself. Southwest Asia, on the other hand, is a serviceable regional designation – but it does not include Georgia (or Armenia). In what world region should these countries therefore be placed? Europe does seem to be the only realistic choice. For historical discussions, however, it might be best to consider the Caucasus as a world region in its own right, one that includes not just Georgia, Armenia, and Azerbaijan, but also the greater north Caucasus and perhaps even Iranian Azerbaijan and northeastern Turkey. The final map posted here, by Georgian cartographer Manana Kurtubadze, nicely captures the physical geography of this expanded Caucasus region.
