The Myth of Continents: A Critique of Metageography / Edition 1

by Martin W. Lewis, Kären Wigen

Overview

In this thoughtful and engaging critique, geographer Martin W. Lewis and historian Kären Wigen reexamine the basic geographical divisions we take for granted, and challenge the unconscious spatial frameworks that govern the way we perceive the world. Arguing that notions of East vs. West, First World vs. Third World, and even the sevenfold continental system are simplistic and misconceived, the authors trace the history of such misconceptions. Their up-to-the-minute study reflects both on the global scale and its relation to the specific continents of Europe, Asia, and Africa—actually part of one contiguous landmass.

The Myth of Continents sheds new light on how our metageographical assumptions grew out of cultural concepts: how the first continental divisions developed from classical times; how the Urals became the division between the so-called continents of Europe and Asia; how countries like Pakistan and Afghanistan recently shifted macroregions in the general consciousness.

This extremely readable and thought-provoking analysis also explores the ways that new economic regions, the end of the cold war, and the proliferation of communication technologies change our understanding of the world. It stimulates thinking about the role of large-scale spatial constructs as driving forces behind particular worldviews and encourages everyone to take a more thoughtful, geographically informed approach to the task of describing and interpreting the human diversity of the planet.

Product Details

ISBN-13:	9780520207431
Publisher:	University of California Press
Publication date:	08/11/1997
Edition description:	First Edition
Pages:	383
Sales rank:	1,045,078
Product dimensions:	6.00(w) x 9.00(h) x 0.90(d)

About the Author

Martin W. Lewis is Associate Research Professor of Geography, Duke University, and author of Wagering the Land: Ritual, Capital, and Environmental Degradation in the Cordillera of Northern Luzon, 1900-1986 (California, 1992) and Green Delusions: An Environmentalist Critique of Radical Environmentalism (1994). Kären E. Wigen is Associate Professor of History, Duke University, and author of The Making of a Japanese Periphery, 1750-1920 (California, 1995).

Read an Excerpt

CHAPTER ONE

The Architecture of Continents

The Development of the Continental Scheme

In contemporary usage, continents are understood to be large, continuous, discrete masses of land, ideally separated by expanses of water. Although of ancient origin, this convention is both historically unstable and surprisingly unexamined; the required size and the requisite degree of physical separation have never been defined. As we shall see, the sevenfold continental system of American elementary school geography did not emerge in final form until the middle decades of the present century.

CLASSICAL PRECEDENTS

According to Arnold Toynbee, the original continental distinction was devised by ancient Greek mariners, who gave the names Europe and Asia to the lands on either side of the complex interior waterway running from the Aegean Sea through the Dardanelles, the Sea of Marmara, the Bosporus, the Black Sea, and the Kerch Strait before reaching the Sea of Azov. This water passage became the core of a continental system when the earliest Greek philosophers, the Ionians of Miletus, designated it as the boundary between the two great landmasses of their world. Somewhat later, Libya (or Africa) was added to form a three-continent scheme. Not surprisingly, the Aegean Sea lay at the heart of the Greek conception of the globe; Asia essentially denoted those lands to its east, Europe those lands to its west and north, and Libya those lands to the south.

A seeming anomaly of this scheme was the intermediate position of the Greeks themselves, whose civilization spanned both the western and the eastern shores of the Aegean. Toynbee argued that the inhabitants of central Greece used the Asia-Europe boundary to disparage their Ionian kin, whose succumbing to "Asian" (Persian) dominion contrasted flatteringly with their own "European" freedom. Yet not all Greek thinkers identified themselves as Europeans. Some evidently employed the term Europe as a synonym for the northern (non-Greek) realm of Thracia. In another formulation, Europe was held to include the mainland of Greece, but not the islands or the Peloponnesus. Still others--notably Aristotle--excluded the Hellenic "race" from the continental schema altogether, arguing that the Greek character, like the Greek lands themselves, occupied a "middle position" between that of Europe and Asia. In any case, these disputes were somewhat technical, since the Greeks tended to view continents as physical entities, with minimal cultural or political content. When they did make generalizations about the inhabitants of different continents, they usually limited their discussion to the contrast between Asians and Europeans; Libya was evidently considered too small and arid to merit more than passing consideration.

Twofold or threefold, the continental system of the Greeks clearly had some utility for those whose geographical horizons did not extend much beyond the Aegean, eastern Mediterranean, and Black Seas. But its arbitrary nature was fully apparent by the fifth century B.C.E. Herodotus, in particular, consistently questioned the conventional three-part system, even while employing it. Criticizing the overly theoretical orientation of Greek geographers, who attempted to apprehend the world through elegant geometrical models, he argued instead for an "empirical cartography founded on exploration and travel." One problematic feature of the geography that Herodotus criticized was its division of Asia and Africa along the Nile, a boundary that sundered the obvious unity of Egypt. After all, as he noted, Asia and Africa were actually contiguous, both with each other and with Europe: "Another thing that puzzles me is why three distinct women's names should have been given to what is really a single landmass; and why, too, the Nile and the Phasis--or, according to some, the Maeotic Tanais and the Cimmerian Strait--should have been fixed upon for the boundaries. Nor have I been able to learn who it was that first marked the boundaries, or where they got their names from."

Similar comments, suggesting a continued awareness that these were constructed categories, echoed throughout the classical period. Strabo, writing in the first century B.C.E., noted that there was "much argument respecting the continents," with some writers viewing them as islands, others as mere peninsulas. Furthermore, he argued, "in giving names to the three continents, the Greeks did not take into consideration the whole habitable earth, but merely their own country, and the land exactly opposite...."

Under the Romans, the continental scheme continued to be employed in scholarly discourse, and the labels Europe and Asia were sometimes used in an informal sense to designate western and eastern portions of the empire. In regard to military matters, the term europeenses was deployed rather more precisely for the western zone. Asia was also used in a more locally specific sense to refer to a political subdivision of the Roman Empire in western Anatolia.

MEDIEVAL AND RENAISSANCE CONSTRUCTIONS

For almost two millennia after Herodotus, the threefold division of the earth continued to guide the European scholarly imagination. The continental scheme was reinforced in late antiquity when early Christian writers mapped onto it the story of Noah's successors. According to St. Jerome (who died circa A.D. 420), translator of the Vulgate Bible, "Noah gave each of his three sons, Shem, Ham, and Japheth, one of the three parts of the world for their inheritance, and these were Asia, Africa, and Europe, respectively." This new theological conception had the merit of explaining the larger size of the Asian landmass by reference to Shem's primogeniture. It also infused the Greeks' tripartite division of the world with religious significance. This sacralized continental model would persist with little alteration until the early modern period.

Medieval Europe thus inherited the geographical ideas of the classical world, but in a calcified and increasingly mythologized form. Whereas the best Greek geographers had recognized the conventional nature of the continents--and insisted that the Red Sea made a more appropriate boundary between Asia and Africa than the Nile River--such niceties were often lost on their counterparts in late antiquity and the early Middle Ages. Martianus Capella, whose compilation of knowledge became a standard medieval text, took it as gospel that the world was divided into Europe, Asia, and Africa, with the Nile separating the latter two landmasses. Other influential encyclopedists of the period, including Orosius and Isidore of Seville, held similar views.

During the Carolingian period, by contrast, the inherited framework of Greek geography began to recede from view. The term Europe (in one form or another) was sometimes used to refer to the emerging civilization in the largely Frankish lands of Latin Christendom, which were occasionally contrasted with an increasingly fabulous Asia to the east. In fact, proponents of both Carolingian and Ottonian (German) imperialism, as well as the papacy, employed the concept of Europe as "a topos of panegyric, [and] a cultural emblem." But until the late Middle Ages, reference to the larger formal continental scheme was largely limited to recondite geographical studies, finding little place in general scholarly discourse. Africa in particular did not figure prominently in the travel lore and fables of medieval Europeans. The southern continent at the time was dismissed as inferior, on the mistaken grounds that it was small in extent and dominated by deserts.

Scholarly geographical studies, of course, were another matter. Here the tripartite worldview of the Greeks was retained, but transposed into an abstract cosmographical model, abandoning all pretense to spatial accuracy. The famous "T-O" maps of the medieval period, representing the earth in the form of a cross, reflect the age's profoundly theological view of space. The cross symbol (represented as a T within the circle of the world) designated the bodies of water that supposedly divided Europe, Asia, and Africa; these landmasses in a sense served as the background on which the sacred symbol was inscribed. The Nile remained, in most cases, the dividing line between Africa and Asia. Classical precedence joined here with theological necessity, converting an empirical distortion into an expression of profound cosmographical order.

With the revival of Greek and Roman learning in the Renaissance, the older continental scheme was revived as well, becoming endowed with an unprecedented scientific authority. The noted sixteenth-century German geographer Sebastian Munster, for example, invoked "the ancient division of the Old World into three regions separated by the Don, the Mediterranean, and the Nile." Despite the considerable accumulation of knowledge in the centuries since Herodotus, few Renaissance scholars questioned the boundaries that had been set in antiquity. On the contrary, it was in this period that the continental scheme became the authoritative frame of reference for sorting out the differences among various human societies.

The elevation of the continental scheme to the level of received truth was conditioned in part by an important historical juncture. In the fourteenth and fifteenth centuries, just as classical writings were being revalued, the geography of Christianity was in flux on several fronts at once. Turkish conquests at its southeastern edge were causing the remaining Christian communities in Asia Minor to retreat, while Christian conquests and conversions in the northeast were vanquishing the last holdouts of paganism in the Baltic region. Meanwhile, the rise of humanism was challenging the cultural unity of the Catholic world from within. These historical circumstances combined to give the Greek continental scheme new salience. On the one hand, as Christianity receded in the southeast and advanced in the northeast, the boundaries of Christendom increasingly (although never perfectly) coincided with those of the Greeks' Europe. On the other hand, humanist scholars began to search for a secular self-designation. As a result, these centuries saw Europe begin to displace Christendom as the primary referent for Western society.

As Western Christians began to call themselves Europeans in the fifteenth century, the continental schema as a whole came into widespread use. But it was not long before the new (partial) geographical fit between Europe and Christendom was once again offset. Continuing Turkish conquests, combined with the final separation of the Eastern and Western Christian traditions, pulled southeastern Europe almost completely out of the orbit of the increasingly self-identified European civilization.

OLD WORLDS, NEW CONTINENTS

Once Europeans crossed the Atlantic, they gradually discovered that their threefold continental system did not form an adequate world model. Evidence of what appeared to be a single "new world" landmass somehow had to be taken into account. The transition from a threefold to a fourfold continental scheme did not occur immediately after Columbus, however. First, America had to be intellectually "invented" as a distinct parcel of land--one that could be viewed geographically, if not culturally, as equivalent to the other continents. According to Eviatar Zerubavel, this reconceptualization took nearly a century to evolve, in part because it activated serious "cosmographic shock." For a long time, many Europeans simply chose to ignore the evidence; as late as 1555, a popular French geography text entitled La Division du monde pronounced that the earth consisted of Asia, Europe, and Africa, making absolutely no mention of the Americas. The Spanish imperial imagination persisted in denying continental status to its transatlantic colonies for even longer. According to Walter Mignolo, "The Castilian notion of 'the Indies' [remained] in place up to the end of the colonial empire; 'America' [began] to be employed by independentist intellectuals only toward the end of the eighteenth century." Yet by the early sixteenth century, the Portuguese cosmographer Duarte Pacheco and his German counterpart Martin Waldseemuller had mapped the Americas as a continent. While cartographic conventions of the period rendered the new landmass, like Africa, as distinctly inferior to Asia and Europe, virtually all global geographies by the seventeenth century at least acknowledged the Americas as one of the "four quarters of the world."

As this brief account suggests, accepting the existence of a transatlantic landmass required more than simply adding a new piece to the existing continental model. As Edmundo O'Gorman has brilliantly demonstrated, reckoning with the existence of previously unknown lands required a fundamental restructuring of European cosmography. For in the old conception, Europe, Africa, end Asia had usually been envisioned as forming a single, interconnected "world island," the Orbis Terrarum. The existence of another such "island" in the antipodes of the Southern Hemisphere--an Orbis Alterius--had often been hypothesized, but it was assumed that it would constitute a world apart, inhabited, if at all, by sapient creatures of an entirely different species. Americans, by contrast, appeared to be of the same order as other humans, suggesting that their homeland must be a fourth part of the human world rather than a true alter-world. Thus it was essentially anthropological data that undermined the established cosmographic order.

In the long run, the discovery of a distant but recognizably human population in the Americas would irrevocably dash the world island to pieces. Over the next several centuries the fundamental relationship between the world's major landmasses was increasingly seen as one of separation, not contiguity. In 1570 Ortelius divided the world into four constituent parts, yet his global maps did not emphasize divisional lines, and his regional maps sometimes spanned "continental" divisions. By the late seventeenth century, however, most global atlases unambiguously distinguished the world's main landmasses and classified all regional maps accordingly. The Greek notion of a unitary human terrain, in other words, was disassembled into its constituent continents, whose relative isolation was now ironically converted into their defining feature. Although the possibility of an Orbis Alterius was never again taken seriously, the boundaries dividing the known lands would henceforth be conceived in much more absolute terms than they had been in the past. Even as the accuracy of mapping improved dramatically in this period, the conceptualization of global divisions was so hardened as to bring about a certain conceptual deterioration.

NEW DIVISIONS

As geographical knowledge increased, and as the authority of the Greeks diminished, the architecture of global geography underwent more subtle transformations as well. If continents were to be meaningful geographical divisions of human geography, rather than mere reflections of an ordained cosmic plan, the Nile and the Don obviously formed inappropriate boundaries. Scholars thus gradually came to select the Red Sea and the Gulf and Isthmus of Suez as the African-Asian divide. Similarly, by the sixteenth century, geographers began to realize that Europe and Asia were not separated by a narrow isthmus, that the Don River did not originate anywhere near the Arctic Sea, and that the Sea of Azov was smaller than had previously been imagined. While the old view was remarkably persistent, a new boundary for these two continents was eventually required as well.

The difficulty was that no convenient barrier like the Red Sea presented itself between Europe and Asia. The initial response was to specify precise linkages between south- and north-flowing rivers across the Russian plains; by the late seventeenth century, one strategy was to divide Europe from Asia along stretches of the Don, Volga, Kama, and Ob Rivers. This was considered an unsolved geographical issue, however, and geographers vied with each other to locate the most fitting divisional line. Only in the eighteenth century did a Swedish military officer, Philipp-Johann von Strahlenberg, argue that the Ural Mountains formed the most significant barrier. Von Strahlenberg's proposal was enthusiastically seconded by Russian intellectuals associated with Peter the Great's Westernization program, particularly Vasilii Nikitich Tatishchev, in large part because of its ideological convenience. In highlighting the Ural divide, Russian Westernizers could at once emphasize the European nature of the historical Russian core while consigning Siberia to the position of an alien Asian realm suitable for colonial rule and exploitation. (Indeed, many Russian texts at this time dropped the name Siberia in favor of the more Asiatic-sounding Great Tartary.) Controversy continued in Russian and German geographical circles, however, with some scholars attempting to push the boundary further east to the Ob or even the Yenisey River, while others argued for holding the line at the Don.

Tatishchev's and von Strahlenberg's position was eventually to triumph not only in Russia but throughout Europe. After the noted French geographer M. Malte-Brun gave it his seal of approval in the nineteenth century, the Ural boundary gained near-universal acceptance. Yet this move necessitated a series of further adjustments, since the Ural Mountains do not extend far enough south--or west--to form a complete border. In atlases of the eighteenth and early nineteenth centuries, the old and new divisions were often combined, with Europe shown as separated from Asia by the Don River, a stretch of the Volga River, and the Ural Mountains. From the mid-1800s on the most common, although by no means universal, solution to this problem was to separate Asia from Europe by a complex line running southward through the Urals, jumping in their southern extent to the Ural River, extending through some two-thirds the length of the Caspian Sea, and turning in a sharp angle to run northwestward along the crest of the Caucasus Mountains. Indeed, as recently as 1994, the United States Department of State gave its official imprimatur to this division. The old usage of the Don River, arbitrary though it might have been, at least required a less contorted delineation. Moreover, the new division did even more injustice to cultural geography than did the old, for it included within Europe such obviously "non-European" peoples as the Buddhist, Mongolian-speaking Kalmyks.

While this geographical boundary between Europe and Asia is now seldom questioned and is often assumed to be either wholly natural or too trivial to worry about, the issue still provokes occasional interest. In 1958, for example, a group of Russian geographers argued that the true divide should follow "the eastern slope of the Urals and their prolongation the Mugodzhar hills, the Emba River, the northern shore of the Caspian Sea, the Kumo-manychskaya Vpadina (depression) and the Kerchenski Strait to the Black Sea"--thus placing the Urals firmly within Europe and the Caucasus within Asia. Other writers have elected to ignore formal guidelines altogether, placing the boundary between the two "continents" wherever they see fit. The 1963 edition of the Encyclopedia Britannica, for example, defines the Swat district of northern Pakistan as "a region bordering on Europe and Asia"--"Europe" perhaps connoting, in this context, all areas traversed by Alexander the Great. Halford Mackinder, on the other hand, selected a "racial" criterion to divide Europe from Africa (although not from Asia), and thus extended its boundaries well to the south: "In fact, the southern boundary of Europe was and is the Sahara rather than the Mediterranean, for it is the desert land that divides the black man from the white."

THE CONTINUING CAREER OF THE CONTINENTAL SCHEME

Despite the ancient and ubiquitous division of the earth into Europe, Asia, and Africa (with the Americas as a later addition), such "parts" of the earth were not necessarily defined explicitly as continents prior to the late nineteenth century. While the term continent--which emphasizes the contiguous nature of the land in question--was often used in translating Greek and Latin concepts regarding the tripartite global division, it was also employed in a far more casual manner. In fact, in early modern English, any reasonably large body of land or even island group might be deemed a continent. In 1599, for example, Richard Hakluyt referred to the West Indies as a "large and fruitfull continent." Gradually, however, geographers excluded archipelagos and smaller landmasses from this category, adhering as well to a more stringent standard of spatial separation. By 1752 Emanuel Bowen was able to state categorically: "A continent is a large space of dry land comprehending many countries all joined together, without any separation by water. Thus Europe, Asia, and Africa is one great continent, as America is another."

The division of the world into two continents certainly forces one to recognize, as Herodotus did many centuries earlier, that Europe, Asia, and Africa are not separated in any real sense. Indeed, perspicacious geographers have always been troubled by this division. As early as 1680, the author of The English Atlas opined: "The division seems not so rational; for Asia is much bigger than both of the others; nor is Europe an equal balance for Africa." Several prominent nineteenth-century German geographers, Alexander von Humboldt and Oskar Peschel among them, insisted that Europe was but an extension of Asia; many Russian Slavophiles, perennial opponents of the more influential Westernizers, concurred. Such clear-headed reasoning was not to prevail, however. By the late nineteenth century the old "parts of the earth" had been definitively named "continents," with the separation between Europe and Asia remaining central to the scheme. The Oxford English Dictionary (compiled in the decades bracketing the turn of the twentieth century) recounts the transition as follows: "Formerly two continents were reckoned, the Old and the New; the former comprising Europe, Asia, and Africa, which form one continuous mass of land; the latter, North and South America, forming another. These two continents are strictly islands, distinguished only by their extent. Now it is usual to reckon four or five continents, Europe, Asia, Africa, and America, North and South; the great island of Australia is sometimes reckoned as another."

Regardless of the term used to denote them, the standard categories of antiquity, with the addition of the "new world(s)," continued to comprise the fundamental framework within which global geography and history were conceived. Yet minor disagreements persisted as to the exact number of units one should count. In eighteenth- and nineteenth-century world atlases, which generally printed the world's major units in different colored inks, one can find fourfold, fivefold, and sixfold divisional schemes. North and South America might be counted as one unit or two, while Australia ("New Holland") was sometimes colored as a portion of Asia, sometimes as a separate landmass, and sometimes as a mere island. All things considered, however, the fourfold scheme prevailed well into the 1800s.

Whatever the exact form it took on maps, the division of the world into great continents became an increasingly important metageographical concept in the eighteenth and nineteenth centuries. Montesquieu, the foremost geographical thinker of the French Enlightenment, based his social theories on the absolute geographical separation of Europe from Asia, the core of his fourfold continental scheme. The most influential human geographer of the mid-nineteenth century, Carl Ritter, similarly argued (in his signature teleological style): "Each continent is like itself alone ... each one was so planned and formed as to have its own special function in the progress of human culture." Ritter also attempted to ground the entire scheme in physical anthropology. Conflating continents with races, he viewed Europe as the land of white people, Africa that of black people, Asia of yellow people, and America of red people--a pernicious notion that still lingers in the public imagination.

It was with Arnold Guyot, the Swiss scholar who introduced Ritter's version of geography to the United States in the mid-1800s, that continent-based thinking reached its apogee. Guyot saw the hand of Providence in the assemblage of the continents as well as in their individual outlines and physiographic structures. The continents accordingly formed the core of Guyot's geographical exposition--one aimed at revealing "the existence of a general law, and disclos[ing] an arrangement which cannot be without a purpose." Not surprisingly, the purpose Guyot discerned in the arrangement of the world's landmasses entailed the progressive revelation of a foreordained superiority for Europe and the Europeans. From his position on the faculty of Princeton University, Guyot propagated his views on the subject for many years, influencing several generations of American teachers and writers.

As the continental system was thus formalized in the nineteenth century, its categories were increasingly naturalized, coming to be regarded, not as products of a fallible human imagination, but as real geographical entities that had been "discovered" through empirical inquiry. E. H. Bunbury, the leading Victorian student of the history of geographic thought, went so far as to label Homer a "primitive geographer" for his failure to recognize "the division of the world into three continents." Bunbury also took Herodotus to task for his "erroneous notion" that Europe was of greater east-west extent than Asia and Libya [Africa] combined. Herodotus came to this conclusion, however, not because his spatial conceptions were any less accurate than those of his peers, but because he eschewed using the north-south trending Tanais (Don) as the continental border, preferring instead east-west running rivers such as the Phais and Araxes (in the Caucasus region). To the Victorian Bunbury, this was not an issue on which educated people could disagree. What nineteenth-century geographers had lost was Herodotus's sense that the only reason for dividing Europe and Asia along a north-south rather than an east-west axis was convention. In fact, by scientific criteria, Herodotus probably had the better argument. Certainly in physical terms, Siberia has much more in common with the far north of Europe--where Herodotus's boundary would have placed it--than with Oman or Cambodia.

INTO THE TWENTIETH CENTURY

Since the early eighteenth century, one of the most problematic issues for global geographers was how to categorize Southeast Asia, Australia, and the islands of the Pacific. Gradually, a new division began to appear in this portion of the world. According to one popular Victorian work of world history, "It was usual until the present century to speak of the great divisions of the earth as the Four Quarters of the World, VIZ; Europe, Asia, Africa, and America," while insisting that a "scientific distribution" of the world's "terrestrial surfaces" would have to include Australia and Polynesia as separate divisions. By the middle of the nineteenth century, Australia was usually portrayed as a distinct part of the world, albeit often linked with the islands of the Pacific. The notion of Oceania as a fifth (or sixth, if the Americas were divided) section of the world grew even more common in the early twentieth century, when several cartographers marked off insular Southeast Asia from Asia and appended it to the island world.

In the early twentieth century, world geography textbooks published in Britain and the United States almost invariably used the continental system as their organizing framework, typically devoting one chapter to each of these "natural" units. This pattern may be found in works on the natural world as well as in those concerned with human geography. Scanning through these textbooks, one notices only slight deviations from the standard model. The International Geography, edited by Hugh Robert Mill, for example, places Central and South America in a single chapter, while devoting another to the polar regions. Leonard Brooks, in A Regional Geography of the World, follows the conventional scheme--with successive chapters on Europe, Asia, North America, South America, Africa, and Australia--but devotes an additional chapter to the British Isles alone. Here Eurocentrism yields pride of place to Britanocentrism, suggesting the emergence of a new virtual continent in the north Atlantic.

Yet not all geographical writers in the early twentieth century viewed continents as given and unproblematic divisions of the globe. In the popular Van Loon's Geography of 1937, for example, the author describes the continental scheme with a light and almost humorous touch, concluding that one might as well use the standard system so long as one remembers its arbitrary foundations. Van Loon viewed the standard arrangement as including five continents: Asia, America, Africa, Europe, and Australia. While it might seem surprising to find North and South America still joined into a single continent in a book published in the United States in 1937, such a notion remained fairly common until World War II. It cannot be coincidental that this idea served American geopolitical designs at the time, which sought both Western Hemispheric domination and disengagement from the "Old World" continents of Europe, Asia, and Africa.

Editorial Reviews

"A solid and useful contribution."

Journal of World History

"The very fact that their work stimulates such questions is a tribute to the authors. In The Myth of Continents, Lewis and Wigen have written an entertaining and informative account of the way our maps show us the world that we want to see."

New York Times