What are the continents?

Conventionally, there are generally said to be 7 continents: Africa, Antarctica, Asia, Europe, North America, Oceania and South America. In this set-up, all the land on Earth is assigned to a continent, and the boundaries of the continents are often rather arbitrary, or based on a tangled mess of cultural and historical variables that tend to be applied inconsistently. For example, Iceland is generally classed as Europe on the basis that it is culturally and historically Scandinavian, while the culturally Slavic majority of the population of Novosibirsk are held to reside far to the East of the Europe-Asia border.

Continents by colour.png
CC BY-SA 3.0, Link

The conventional, 7-continent view of the world. Image source: Wikimedia Commons

I am proposing a new system, with 6 continents. These continents are defined purely geographically, without trying to work in any cultural notions, and, crucially, the do not include all the land on Earth. This system accounts for the fact that remote, volcanic islands and island groups, such as Iceland or Hawaii, do not necessarily need to be assigned to particular continents.

new continents

The continents, according to my system. Original image source: Canuckguy (talk) and many others (see File history) – Adapted from Brianski‘s File:BlankMap-World3.svg by Canuckguy and originally based on CIA’s political world map, Public Domain, Link

The continents of the new system are as follows:


Africa is very straightforward. It is pretty much unaltered from the conventional definition. Its only land border passes through Egypt, following the shortest distance from the Mediterranean to the Red Sea, fairly close to the Suez Canal.


The same, pretty much. Just the main landmass of Antarctica and its offshore islands. I guess there’s not much cultural baggage here.


This continent consists of the landmasses of Australia and New Guinea, and their surrounding small islands. in Indonesia, the islands east of the Wallace Line are part of Australia, while those to the east are part of Eurasia, reflecting the extents of the 2 landmasses before sea levels rose a very long time ago and created the archipelago we see today.

Traditionally, these areas are considered to be part of ‘Oceania’, along with New Zealand and the rest of the South Pacific, but Oceania is not a real continent, as will be explained in the section on Remote Volcanic Islands.


Europe and Asia are the same continent. There, I said it.

Conventionally, the border between Europe and Asia is considered to follow the watershed of the Ural Mountains in Russia, the Ural River in Kazakhstan (or, if you really don’t give a fuck, just the border between Russia and Kazakhstan), and the Caucasus Watershed between the Caspian and Black Seas, usually close to (but not exactly following) Russia’s borders with Azerbaijan and Georgia.

This, however, has little geographical, or even cultural or historical, significance. The Ural Mountains are a relatively small (albeit long) mountain range, and the Ural River is not an especially significant river. Normally, these kinds of geographical features would not be seen as the boundaries of continents.

Geologically speaking, there is no meeting of 2 great landmasses, no major fault line, along the Ural Mountains. There’s no bottleneck of land here, like there is between North and South America. If someone with no prior knowledge of the history of Europe and Asia were to look at a map of the world, they would say Eurasia was obviously a single continent.

The only reason they are considered separate is that the Greeks (who were not using the same idea of ‘continents’ that we use nowadays) divided them (along the Don River, far to the west of the Urals). It sort of stuck, especially since it was convenient for the pretty, symmetric, ‘T & O’ worldview of the Middle Ages, with Jerusalem at the centre. And now nobody wants to let it go.

North America

Basically as traditional. The border with South America follows the line at from the Caribbean to Pacific the point where the land stops tapering inwards North-South and starts bulging out into South America.

South America

See above

Remote Volcanic Islands

Many islands can be classified as belonging to a particular continent because they broke away from that continent’s mainland (eg. Madagascar) or they were separated from it by rising sea levels (eg. New Guinea). Others, however, just emerged out of the sea as a result of volcanic activity. Examples include Iceland, or New Zealand, or Hawaii, or the Canary Islands. Some of these, such as Japan, or the Aleutians, are clearly parts of a volcanic chain that continues onto a continental mainland, so can be classified as that continent. Others, such as the Canary Islands, are simply close enough to one continent’s mainland that they can be effectively viewed as that continent’s offshore islands. However, there are some, such as Iceland, or Hawaii, that are too remote to fit these categories. They cannot reasonably be said to belong to a particular continent.

And this is fine. Not everything needs to be classified like that. People like to think about things as if they’re in neat categories, but the world doesn’t actually have to stick to that. Continents aren’t necessarily everything. We’ve put nearly all the land in the world into continents, and isn’t that enough?

This is also why I’ve ditched Oceania as a concept. It originates in the idea that everywhere has to be a continent, so they’ve made up a ‘continent’ that can include all the remote volcanic islands and atolls of the Pacific,too far from any large landmass and stuck in the middle of the Pacific Ocean’s tectonic plate. For administrative purposes, sure, it often makes a lot of sense to group them together, but don’t call it a continent when it isn’t.

The false dichotomy that took over Popular Cartography

In March 2017, public schools in Boston changed their standard map projection from the ubiquitous “Mercator Projection”, in which Greenland is displayed bigger than Africa, to the “Peters Projection”, where the areas of landmasses appear proportionate to their actual areas. This received a lot of praise, including an article in the Guardian saying the move “cut the developing world down to size”. However, the Peters Projection is far from the ideal solution to the Mercator Projection’s distortion problems.

The hard problem of cartography

Maps are flat, but Earth is (unless you belong to certain strange subcultures)round. That is the essence of what could be considered that hard problem of cartography. It follows that it is impossible to perfectly represent the surface of Earth on a flat surface.

So when you design a map of the world (or the moon or the cosmic microwave background or anything else that’s round), you have to compromise. You have to favour one thing at the expense of the other. You can go all out for preserving relative areas, but you’re going to end up distorting shapes. You can preserve directions, but you’re going to fuck up all the relative areas. You can compromise and try to sacrifice a bit of one benefit to get a bit of another. There’s all sorts you can do, which is why people have made so many different map projections over the years.

The Mercator Projection

One such ”projection”, and probably the best-known, was developed by the Flemish cartographer Gerardus Mercator in 1569, with one crucial selling-point: a constant bearing is a straight line. For reasons that should be obvious, this is very useful for maritime navigation (and during the Age of Discovery, when the Mercator Projection was created, this was the main thing people used maps for). You can draw a line on the map showing your route, and at any point on that line, it’s going to make an angle with the vertical which is the same as the angle at which your compass should be pointing.

The problem, though, is that when you’ve gone all out to make a constant bearing a straight line, you end up with areas being enormously inflated towards the poles (and even going to infinity at the poles themselves). The classic example is the way Greenland appears a similar size to Africa, then there’s also the way Scandinavia appears to be the size of India, and the islands of the Canadian Arctic appear to dwarf the much larger islands of Madagascar, Borneo and New Guinea. An interesting curiosity I once stumbled across is that the Lena River, in North-Eastern Siberia, has the distinction of looking like the longest river in the world in the Mercator Projection.

But, for whatever bizarre reason, the Mercator Projection has ended up as the classic image of ”what the world looks like”.

Peters’ solution

Needless to say, this is far from ideal, and that is what the German historian and filmmaker Arno Peters exploited when he announced, in 1973, that he had developed a ”new” projection, which preserved the relative areas of landmasses.

Peters argued that the equatorial regions compressed by the Mercator Projection are generally poor and marginalised, compared to Europe and North America, which the Mercator Projection stretches.  According to Peters, this led to people’s perception of the world being skewed, and excessive emphasis being placed on these areas.

An equal-areas projection, according to Peters, would help to fix this skewed perception.

In fairness to Peters, the Mercator Projection does inflate the size of Europe while diminishing that of Africa, South Asia, and Latin America, and it’s reasonable to suggest that the ubiquity of the Mercator image of Earth could lead people to place a disproportionate psychological emphasis on the inflated regions. Some credit’s due to Peters for pushing that idea into the mainstream. The problem, though, is that the projection doesn’t actually have half the qualities he claimed as its advantages.

Peters claimed that the projection has “no distortion”. It quite clearly does, as any map projection does. It is a trade-off between different types of distortion, like any other projection. Relative areas are shown accurately, but the shapes of things are all over the place (for example, Africa comes out very long). There is distortion of shapes everywhere except along the standard parallels.

Perhaps worse than all of these is Peters’ claim that he was the first person to come up with the idea of an equal-area projection. Equal-area projections have been around for centuries, with the sinusoidal projection appearing in 1570, a year after the Mercator projection (although it looks very weird).  Cylindrical equal-area projections (formed by wrapping a cylinder around the Earth, projecting onto that, and then unwrapping it, giving you a rectangular map) have been around at least since the Lambert cylindrical projection, first described in 1772, where the standard parallel is the equator. Since then, many cartographers have put different standard parallels into the same formula to produce different projections. These included James Gall who, in 1855, published a cylindrical equal-area projection with standard parallels at 45° from the Equator, identical to the Peters projection.

In fairness to Peters, Gall’s projection was never well-known, and he had probably never heard of it. But even if he is not guilty of intentional plagiarism, his wilful ignorance of the whole history of equal-area projections and his anti-intellectual caricature of the cartographic community as some kind of weird Mercator cult was just straight-up arrogant and irresponsible.

Other projections

Peters projection propaganda typically presents the world of cartography in terms of a dichotomy between Mercator and Peters, with Mercator grossly distorted (it is), and the Peters projection as totally accurate (it isn’t). However, the ways of representing the world on a flat surface are so much more diverse than that. If you want equal areas, you’ve got sinusoidal (which is weird) or Mollweide (which I like because it doesn’t look like it’s meant to be flat). If, like Peters, you’re hung up on the idea that maps have to be rectangular (which actually makes sense if you’re looking for something to blu-tack to your classroom wall), you’ve still got a whole host of ‘cylindrical’ equal area projections (so named because they’re formed by wrapping a cylinder round the Earth, projecting the surface onto that, and then unwrapping it, giving you a rectangular map). In fact, you can make your own cylindrical equal area projection just by multiplying the longitude of a point by the cosine of the standard parallel (an arbitrarily chosen latitude where there is no distortion) to get the x coordinate of that point, and dividing the sine of the latitude by the cosine of the standard parallel to get its y coordinate. Just (unlike Arno Peters) make sure nobody’s already used your standard parallel.

So why do people like the Peters projection?

Most of the hype that has surrounded the Peters projection since the get-go simply comes down to the way it was marketed. Peters, a filmmaker-turned-historian with no background in cartography, did not go through the conventional method for presenting a new map projection. Instead of publishing it in a cartography journal, Peters went straight to the media, presenting to the general public his manifesto for the anti-Mercator revolution.

cylindrical equal areas
Some examples of cylindrical equal area projections, with the standard parallels and aspect ratios indicated, and with Tissot’s Indicatrices showing distortion at different locations. Image source: Wikimedia Commons, created by Eric Gaba.

And of course, actual experts in maps, with their acne, terrible personal hygiene and distinct lack of social skills proved no match in this arena for a sexy, charismatic Hollywood historian. Peters’ unashamedly anti-intellectual wokeness played very well with the general public who, as any cartography enthusiast is likely well-aware, aren’t generally that interested in map projections. It’s easy to see how, for someone unfamiliar with the 400 year pre-Peters history of people trying to come up with alternatives to the Mercator projection, the idea of some kind of cartographic cabal, wearing hooded robes and chanting in Latin while they sacrifice goats to Gerardus Mercator. People like to feel special, and like they know something the so-called “experts” don’t. As such, the Peters cult was able to quickly gain momentum, ending up with many NGOs, including UNESCO, adopting it as standard in an effort to be woke and address the problem of ‘cartographic imperialism’.