Robert Kaplan and the Revenge of Illogic

I've been meaning for a while to write something about Robert Kaplan's attempt to rehabilitate Halford Mackinder's geographical determinism. Having finally had time to read it, "attempt" may be too kind a word.

Kaplan's main thesis is that geography -- by which he says he means physical features of mountains and seas and so on, though he from time to time seems to flip over into the distribution of cultural and religious groupings -- has a major impact on global politics and history. That is a claim that is on one hand too obvious to deny and on the other too vague to be of any use. What matters is the specific theses one advances as to how geography makes a difference.

Digging through Kaplan's meanderings on the impending anarchy in Eurasia and his condemnation of a vague idealist foreign policy foil, there's really not much there in terms of specific theses about the influence of geography on global politics. What is there is ad hoc, unsupported, and in some cases clearly false.

Mackinder's most famous thesis, which Kaplan endorses, is the idea that central Asia and Siberia -- controlled since Mackinder's time by the Russian Empire, the USSR, and now Russia and the post-Soviet central Asian states -- is the "pivot" of world history, control of which would ensure control of the world barring some very close containment by other powers. Kaplan takes this to have been vindicated by the rise of the Soviet Union. But the USSR was only the second superpower (after the influential but short-lived Mongols) to occupy that territory. At other points in history, the world's dominant political and economic power has been located in Mesopotamia, Persia, China (for a good long time), Western Europe, and North America. And it's not clear how Soviet control of central Asia was critical to its superpowerdom -- its demographic and economic core was in Western Russia, Belarus, and Ukraine. Similarly, Kaplan argues that China could become a superpower by conquering Siberia, but never explains why, since China's economic heartland is along its east coast.

Kaplan credits Mackinder with prescience for foreseeing the two World Wars when he states that with the completion of European colonialism, the world would be full and thus these powers would turn against each other. But this is a rather stupid idea even for Mackinder's time -- European powers had been fighting each other for control of Europe, and for control of particular colonial territories, for hundreds of years. Napoleon's proto-world-war, for example, was hardly "dissipated" by the possibility of trying to conquer more colonies rather than attacking Germany, Italy, etc. Indeed, there's a good (geographical!) argument to be made that the availability of new lands to colonize intensified conflict among the colonial powers.

Kaplan occasionally does the work of critique himself. One of his clearer geographical claims is this: "The ultimate land-based empire, with few natural barriers against invasion, Russia would know forevermore what it was like to be brutally conquered. As a result, it would become perennially obsessed with expanding and holding territory." But in the very next paragraph, he runs through some of Mackinder's thoughts on Western Europe's colonial expansion. A set of empires more "obsessed with expanding and holding territory" could hardly be found than the eminently sea-based ones of Britain, Spain, the Netherlands, and France. Kaplan then drops the idea that geography determines who will seek an empire, and now claims geography determines whether these empire-seekers will be democratic or authoritarian: "the sea, beyond the cosmopolitan influences it bestows by virtue of access to distant harbors, provides the inviolate border security that democracy needs to take root." Tell that to nearly-land-locked, yet somehow still democratic, Germany and Switzerland. (What about the Nazis? Well, Italy and Spain, with their rational borders along seas and mountains, both went through periods of fascism as well.) Or, for that matter, tell it to the United States and Canada, which managed to become two of the world's leading democracies despite their long and irrational border. Nor, for that matter, has being an island always staved off authoritarianism in countries like Cuba, Haiti, Indonesia, or the Philippines. And South Korea managed to become a democracy with its capital separated only by a thin, geographically illogical border from one of the most threatening regimes in the world.

Kaplan goes on to argue that population growth and technological advances since Mackinder's time have made all of Asia a single pivot area (an "organic whole"). It's not clear what he means by this. Certainly not that it has become a single political-geographical unit, since the rest of the article is spent detailing the geographical and political fragmentation of Asia. Nor does he really explain the implications of this. At this point Mackinder's pivot thesis drops out of the article, and we get to more specific small-scale evaluations of geography's effects on particular places, all of it driven by his longstanding preoccupation with the idea of teeming, crowded hordes of irrational people.

Kaplan takes us on a tour of conflict-prone "shatter zones" located around the rim of Mackinder's pivot area. The Asian rim, at least -- Europe, despite sharing all of the geographical characteristics he lists for the other rim areas, and despite having been the scene of intense conflict until just half a century ago, is curiously absent.

Kaplan predicts imminent anarchy for Bangladesh, in large part because of its geographically irrational border -- there are no mountain ranges or great rivers separating India and Bangladesh, so the Bangladeshi government can't keep its people under control. But it's not as if Bangladesh was doing awesome during and immediately after British colonial rule, when the national borders corresponded more closely to natural ones -- the country was split off from India because of a major cultural gap between Hindus and Muslims that existed despite the lack of physical barriers. (The intermixing of Hindus and Muslims, plus a variety of smaller religions, in the Indian subcontinent also defies Kaplan/Mackinder's earlier easy claim that there's a neat geographical correspondence between the four heartland-ringing regions (Europe, the Middle East, India, and China) and their respective faiths (Christianity, Islam, Hinduism, and supposedly Buddhism).

Kaplan is on somewhat firmer -- though also extremely well-trod -- ground as he moves westward and notes that many of the borders of countries were drawn by colonial powers with little regard to the preexisting natural or cultural geography of the area. But his discussion is haunted by the idea that population density and scarce resources will inevitably create conflict, and that that conflict will spill over into other areas. This "resource wars" thesis is not tenable as a simple generalization -- resource scarcity can lead to cooperation, resource abundance can lead to conflict, and I don't see why conflict within Yemen over water rights would threaten Saudi Arabia, as Kaplan claims. The links among population, resources, culture, politics, and conflict deserve detailed empirical study, not Kaplan's breezy certainty about the coming anarchy.

Finally, we reach Iran. Kaplan says that Iran is relatively topographically unified, and has been politically unified for most of history in the form of the various Persian empires, and that it's politically pivotal because it sits at the juncture of various important trade routes. I'll grant all that. Kaplan describes Iran's various geopolitical schemes, such as supplying Hezbollah, which he says are made possible by its logical boundaries. But just a few pages ago, he had been describing similar machinations by both India (one of his prime examples of a state with illogical boundaries) and China (whose boundary logic is not assessed in the article). Kaplan concludes: "If the geographic logic of Iranian expansion sounds eerily similar to that of Russian expansion in Mackinder’s original telling, it is." The two share a logic, yes -- but it's hardly eerie, since it's the same logic used by every major world power, including the United States (how many ideologically congenial militias did we fund in Latin America?).

Kaplan then advises a strategy of containment for Iran similar to that used against Russia. Russia, we should recall, does not have particularly logical boundaries. Containment certainly makes sense as a geopolitical strategy for dealing with a hostile country, but I don't see how Iran's logically-bounded geography determines the choice of containment as opposed to, presumably, invasion and regime change.

The primary critical response to Kaplan, I imagine, will be to urge the virtues of idealism (in both the moral and "ideas shape history" senses) and free will against his materialistic determinism. While there's something to those counterclaims, I think it's important as well to highlight that even though he's right in a general sense to say geography matters, he's wrong in the specific predictions he tries to make from geographical premises. Kaplan is the evolutionary psychology of geography -- he takes an undoubtedly true but extremely vague premise ("geography affects history" vs "the mind has evolved"), and applies it in a ad hoc and post hoc fashion to put a stamp of explanation on various events, without much more coherence to the resulting claims than that given by his political preoccupations.