----------------------------------------------------------------- Dark Continents: A Critique of Internet Metageographies Terry Harpold Georgia Institute of Technology terry.harpold@lcc.gatech.edu © 1999 Terry Harpold. All rights reserved. ----------------------------------------------------------------- "The Blankest of Blank Spaces"[1] [Figure] Figure 1. "Map of Africa, Showing Its Most Recent Discoveries." W. Williams, Philadelphia, 1859. [Click on image to see enlarged view] [Figure 2] Figure 2. Detail of Figure 1. Note the blank field straddling the equator, labeled "UNKNOWN INTERIOR." It was in 1868, when nine years old or thereabouts, that while looking at a map of Africa of the time and putting my finger on the blank space then representing the unsolved mystery of that continent, I said to myself, with absolute assurance and an amazing audacity which are no longer in my character now: "When I grow up, I shall go there." And of course I thought no more about it till after a quarter century or so an opportunity offered to go there--as if the sin of childish audacity were to be visited on my mature head. Yes. I did go there: there being the region of Stanley Falls, which in '68 was the blankest of blank spaces on the earth's figured surface. --Joseph Conrad, A Personal Record 1. Joseph Conrad's account of his discovery of a map of Africa in his grandfather's library is among the most famous anecdotes of modern literary biography. (Which map he found is unknown; it must have resembled one published by W. Williams of Philadelphia about nine years earlier [Figures 1 and 2].) Marlow, Conrad's narrator, will repeat the memory as his own in the opening chapter of Heart of Darkness; Conrad will describe the event again near the end of his career, in a sympathetic essay on the British Empire's great explorers. As Christopher GoGwilt has shown, the seductions of the map's "unsolved mystery" are central to Conrad's authorial project; they bind the physical and imaginary geographies of his fictions to one of the defining visual tropes of high colonialism (126).[2] The period of the late nineteenth and early twentieth centuries was the most active of the modern partitioning of Africa, with a half-dozen European powers scrambling for position on the continent through territorial exploration and appropriation, armed conflict, and diplomatic negotiation. This activity depended on the accuracy and completeness of the maps used by the Europeans to describe their African possessions (McIlwaine 59-62). 2. Now, maps, in the narrow sense of the word--pictorial representations of a physical terrain, commonly in the form of planar (2D) projections--are never merely descriptive; they are also heuristic, suasive, and hegemonic. (My reasons for stipulating this narrow definition of "map" will become clear shortly.) Individuals, communities, and nations rely on them to manage spatial and political complexity that would otherwise exceed human perception and memory, and to adjudicate claims of ownership and jurisdiction. Mapping appears to be fundamental to human consciousness of space and time. All cultures record their experience in artifacts which are consumed in plainly maplike ways, though these artifacts may little resemble the brightly-colored wall hangings that most Americans (for example) recall from grade-school geography classes.[2] This universality of mapping practices may account for the persuasiveness of a broader, more metaphorical sense of the term "map" used in many disciplines and technical practices to describe any intentional structuring of space, time, or knowledge--thus we speak of "maps" of a text, "map" views of data, "cognitive maps," and so on. 3. Because they are fundamentally cultural things, maps (again, in the narrow sense) are embedded in the epistemological and ideological structures that Roland Barthes calls the "mythologies" of semiological systems (111). The "unsolved mystery" Conrad discovers in the map of the Congolese interior names one such mythologizing structure. We may safely assume that the mapmaker did not believe there were no topographically significant features in the region left blank, only that no European had witnessed and described those features. (This is precisely the sense of the observation encoded in the label "UNKNOWN INTERIOR" in Figure 2.) The blank region is "empty" only in relation to the comparable fullness of the rest of the map. The blank represents the essence of late nineteenth century Africa as an absence: as the negative of the evidentiary transparency of the European landscape (in this discourse, "European" always connotes fullness, completeness, transparency). To fill the blank is what colonial exploration and territorial appropriation are, in a sense, all about. One may be tempted here to cite childhood curiosity regarding the taboo or the invisible, and to invoke the language of screen memories and primal scenes, but that is not necessary to demonstrate the significance of the unfinished region of the map for Conrad and his contemporaries. Tie the epistemological puzzles of the blank to the opportunistic interests of capital and the evangelical fervor of the Christian missionary (neither of which operated in isolation from the other), and you have a sufficiently broad practical foundation upon which to sustain multiple desires to transect symbolically and materially the empty field of this imaginary terrain: with rivers, lakes, trade routes, roads, railways, and territorial boundaries.[3] 4. Consider now two examples from a series of more recent maps. Figure 3 shows one of several widely-reproduced images created by Kenneth Cox, Stephen G. Eick, and Taosong He, and described in their influential 1996 paper on 3D network visualization techniques.[4] [Figure 3] Figure 3. "Arc map" showing worldwide Internet traffic during a two-hour period, February 1-7, 1993. (Image © 1996 Stephen G. Eick, Visual Insights. Used by permission.) [Click on image to see enlarged view] 5. Shown here is one frame of an animation of Internet traffic between fifty countries via the NFSNET/ANSnet backbone during a two-hour period of the week of February 1-7, 1993. Each country with active nodes on the network is represented by a box-shaped glyph, positioned at the location of the country's capital, and scaled and colored to encode the total packet count for all links emanating from the country. The arcs between the countries indicate the flow of network traffic, with the higher and redder arcs indicating the larger flows. One of the most striking traits of this image is its dissymmetry: the greater part of the arcs and glyphs are traced on the upper-left field (roughly North and West, in relation to the equator and Greenwich meridian). Pinning the origins of the arcs to the capitals of countries underrepresents the actual geographical distribution of traffic, but it's clear that, apart from a few arcs converging in Pretoria, there is nothing much going in or coming out of the Africa in early 1993. 6. Figure 4 shows another projection from the series by Cox, Eick, and He. This "arc map" traces the same traffic flows and uses a similar color-coding scheme, applying it to a 2D projection, viewed on a variable vertical axis.[5] [Figure 4] Figure 4. "Arc map" showing worldwide Internet traffic during a two-hour period, February 1-7, 1993. (Image © 1996 Stephen G. Eick, Visual Insights. Used by permission.) [Click on image to see enlarged view] Like the image in Figure 3, this image encodes Internet traffic statistics in arc height, color, and thickness. Cox, Eick, and He's stated aim in producing these maps was to improve the usability of network visualizations based on cartographic techniques. Shifting the traffic arcs into a third dimension, they propose, and encoding them with height and color to indicate relative bandwidth and flow, eliminates the complex line crossings that would make 2D maps of the same data difficult to read. 7. These are arresting images. Their play of light (the neon hues of the arcs) and dark (the black background) must have been dictated to some degree by technical considerations. They were created for viewing on computer terminals, and the techniques available to the researchers were those best suited to the cathode ray and the binary logic of computing: one/zero, on/off, traffic/no traffic. One consequence of that (probably automatic) choice of representational schemes is, however, problematic: the binarisms of data flow are recast in an analogous but distinct tropology of fullness and emptiness: a light "on" (brighter still when exceptionally "on"); a light "off" (entirely absent when truly "off": darkness visible). The lines of Internet traffic look more like beacons in the night than, say, undersea cables, satellite relays, or fiber-optic cables. Moreover, the decision to include national borders in Figure 4, and to fill them in with solid hues representing the intensity of traffic to and from each nation, binds the viewer's visual identification of those nations to the valences of the network. Even at this oblique angle, it is easy to identify the outlines of Australia, India, Western Europe, Alaska, etc. The identities of the networked regions are in turn reified by their apparent internal uniformity--all spaces within their borders are of a single color representing the national level of traffic, thus eliminating any suggestion that the diffusion of network signals may vary locally. The unnetworked nations seem equally uniform, but in an opposite sense: framed voids, they would merge with the empty background, were it not for their faint borders. Thus, the fusion in these images of a tropology of presence/absence and fullness/emptiness with conventional signifiers of regional and nation encodes the traces of network activity with meanings beyond the mere visual registering of place-names. Viewed with an eye to their unacknowledged political valences, these images of the wired world (that is, of the mostly unwired world) draw, I will argue, on visual discourses of identity and negated identity that echo those of the European maps of colonized and colonizable space of nearly a century ago.[6] 8. This essay is based on research in progress which interrogates these and similar cartographic representations of the Internet as a first step in a critique of the complicity of techniques of scientific visualization with the contrasting invisibility of political and economic formations. I propose that these depictions of network activity are embedded in unacknowledged and pernicious metageographies--sign systems that organize geographical knowledge into visual schemes that seem straightforward (how else to illustrate global Internet traffic if not on images of... the globe?), but which depend on historically- and politically-inflected misrepresentation of underlying material conditions. 9. I borrow the word "metageographies" from Martin Lewis and Kären Wigen's recent study of mapping discourses, The Myth of Continents. Lewis and Wigen's analysis of the genealogy of Eurocentric constructions of continental geography and East/West, Europe/Asia/Africa/American divisions of the human world is an invaluable model for deciphering the conventions of Internet cartography. Their call for new, politically-progressive articulations of the relations of material and social human spaces to political power suggests that an analogous reworking of visualizations of emerging virtual realms is needed at this time. "It is simply no longer tenable," they write, to assume that all significant high-order spatial units will take the form of discrete, contiguous blocks; analyzing contemporary human geography requires a different vocabulary. Instead of assuming contiguity, we need a way to visualize discontinuous "regions" that might take the spatial form of lattices, archipelagos, hollow rings, or patchworks. Such patterns certainly have not displaced the older cultural realms, which continue to be important for the multiple legacies they have left behind. But in the late twentieth century the friction of distance is much less than it used to be; capital flows as much as human migrations can rapidly create and re-create profound connections between distant places. As a result, some of the most powerful sociospatial aggregations of our day simply cannot be mapped as single, bounded territories. To grasp these new realities will demand an imaginative approach to regional definition; to map them effectively will demand an innovative approach to cartography. (200) The lines of force defining information flow in the networked and unnetworked spheres are not merely geographical or technological; they are also--and irreducibly--political. The Blind Spot of Cartography [Figure 5] Figure 5. The Bellman's chart, from Henry Holiday's 1876 illustrations to The Hunting of the Snark. He had brought a large map representing the sea, Without the least vestige of land: And the crew were much pleased when they found it to be A map they could all understand. What's the good of Mercator's North Poles and Equators, Tropics, Zones and Meridian lines?" So the Bellman would cry: and the crew would reply, "They are merely conventional signs! "Other maps are such shapes, with their islands and capes! But we've got our brave captain to thank" (So the crew would protest) "that he's brought us the best-- perfect and absolute blank!" --Lewis Carroll, The Hunting of the Snark[7] 10. Few readers of maps are so canny as the Bellman's happy crew in Carroll's mock epic. Late twentieth-century postindustrial culture is saturated with maps; we rely on them daily in myriad ways, without taking much notice of their specificity or conventionality. Denis Wood has argued that this near ubiquity of maps and mapping technologies contributes to map-users' general inattention to the cultural and historical props of maps (34). In this section, I offer two general observations regarding the form and effects of cartographic representations of real and virtual spaces. 11. First observation. Maps depict a selective distortion of the information available to those who design them. Mark Monmonier has famously observed that maps must as a matter of course mislead their users, if (1) the maps are to be legible at all, and if (2) the maps are to address the specific purposes for which they were designed. The first of these conditions is based in the physical limitations of paper, print, and video technologies, and the capabilities of the human eye. Everything that is noteworthy about a place will not fit on the page or screen; some levels of detail will have to be excluded (and the shapes and placement of some landmarks altered) to avoid filling the map with illegible, overlapping blots. The second condition is an extension of this common-sense constraint on graphic design, and a consequence of the semiological and cultural work of maps: which cartographic details are to be shown in a given map will be determined by technical or strategic forces that bear no necessary relation to the map's general "accuracy"; details are commonly eliminated, falsified, or distorted so as to improve a map's efficacy toward a particular end, resulting in the misrepresentation or exclusion of information which may serve other ends or reveal inconsistencies.[8] [Figure 6] Figure 6. Mercator's projection. 12. For example, the Mercator projection (Figure 6) is helpful if you need to plot a constant navigational bearing between two points at sea (you just draw a straight line--a rhumb line--on the map), but its gross distortion of regions near the poles makes it a poor device for understanding the relative sizes of Greenland and South America. Use of the Mercator projection as a general-reference map (still common in the U.S.) is a cause for frequent complaint by cartographers, who insist that the projection was designed to facilitate ocean navigation in the 16th century, and is thus ill-suited to modern applications where the plotting of rhumb lines is not relevant.[9] [Figure 7] Figure 7. Goode's homolosine equal-area projection, with interruptions to preserve landmasses. Goode's homolosine equal-area projection (Figure 7), considered by many to be the best general-purpose modern projection, preserves relative areal sizes, but it does so at the cost of interrupting the oceans and splitting Antarctica into four pieces (Snyder 196-98). [Figure 8] Figure 8. The Gall-Peters projection. 13. The acrimonious debates provoked by the popularity in the 1970s and early '80s of the Peters orthographic equal-area projection (Figure 8, also known as the Gall-Peters projection) suggest the disciplinary and political stakes at play in the choice of projections.[10] Advocates of the Peters projection claimed that it corrected Mercator's failings as a general-purpose map by, in effect, recentering the viewer's eye. They complained that the Mercator projection distorts the relative sizes of the continents, favoring the Northern hemisphere and underrepresenting the sizes of landforms in the Southern hemisphere. (For example, in Figure 6, North America appears substantially larger than Africa--it is in fact only about 80% as large; Greenland appears to be four or five times the size of Australia, though it is about one-third its actual size.) Peters's projection, it was claimed, represents a more realistic and "equitable" view of the world, as it depicts the relative areas of the major landmasses more accurately than Mercator, at the expense of vertical distortion of familiar (that is, Mercator-like) continental outlines. Critics of the Peters projection found its claims of originality misleading, its goals propagandistic, and its appearance just plain ugly (cartographer Arthur Robinson famously compared its appearance to "wet, ragged, long winter underwear" [Wood 60]). As Wood wryly observes, however, no one questioned the claim of Peters's improved overall areal accuracy in comparison to Mercator. The objections were, essentially, that the Peters projection looks funny--but only "to those who have been confusing the map and the globe" (210 n46). Criticism of the projection's aesthetic qualities, Wood argues, were little more than a smokescreen to cover the anxiety of mapmakers and users elicited by the projection's evidence of the political interestedness of every cartographic projection. 14. Distortion is simply a mathematical consequence of how maps do what they do. Two-dimensional (planar) projections of a three-dimensional (spherical or ellipsoidal) terrain, no matter their scale or detail, must depict that terrain inaccurately, and the inaccuracy will increase as the scope of the terrain shown increases. This built-in error of 2D projections accounts for their enormous number and variety, each representing a different compromise of areal accuracy to angles, shapes, and directions.[11] It accounts as well for the many recent efforts to produce stereoptic or pseudo-3D maps (that is, maps that are not planar projections, but instead purport to show the Earth in a form more nearly corresponding to its "actual" shape), through the use of composite satellite imagery, and interactive video and animation techniques.[12] Most projections of large regions are not created to serve as general-purpose maps; their distortions are deemed acceptable because some other attribute of the projection meets a specific need or fits a clearly-defined context of use. 15. Identifying a map's possible misreadings, however, is not so simple as legislating appropriate or inappropriate uses for that projection. The inevitable inaccuracies of planar projections have historically provided cover for unacknowledged or unconscious motives of the advocates of a particular projection.[13] That maps distort is a fact of mathematics and human perception; which distortions are favored over others, which map is selected as the "best" fit for a given context, is determined by institutional and political forces. Thus the objection to Mercator raised by the advocates of Gall-Peters: Mercator was created to simplify navigation within a specific region of the planet that corresponded to the principal foreign territories of the European colonial powers. That the projection also made Northern and European nations appear equal to or larger in area than their African and Asian territories--when they were in fact substantially smaller--was not, these critics claimed, merely an accidental attribute of the projection's geometric technique. The specific distortions of the map, they argued, contributed to its lasting popularity in the West: Mercator visually reinforced Eurocentric assumptions about the spatial, political, and economic priority of the colonizing powers, and helped to prop up European paradigms of nation-state boundaries and identities that had little or no historical basis among the indigenous peoples of the colonized regions.[14] 16. Second observation. The bounded shapes of cartographic representation are conventional, historically-determined signs, rather than visual analogues of real terrains. Maps are noteworthy examples of the treachery of images. A closed, irregular geometric form whose contours are said to approximate those of a bounded region of space is not a priori a more suitable sign for that region than some other shape that would seem to resemble it less, because the vast majority of users of the map will never encounter an object that is not defined by this or another map, to which they might compare it.[15] This assertion is, of course, trivial in the case of political, ethnic, or linguistic divisions of territories otherwise unbroken by "natural" boundaries (rivers, mountain ranges, shorelines, etc.). But the fact that these soft boundaries often traverse the harder formations of the geophysical terrain demonstrates that distinctions between natural and cultural demarcations of landspace are at best provisional. Dry land may stop at river banks and seashores, but human definitions of city, nation, and continent do not. "Natural" barriers dividing land regions are fixed or erased by political strategy, technological intervention, or military conquest, not by the innate properties of soil and water. 17. We've learned to recognize and to internalize iconic representations of nations and continents in much the same way we've learned to accept the verisimilitude of other images of things we've never actually seen: from a cultural toolkit of forms, patterns, and strategies for making sense of them in relation to other signs, and from frequent exposure to stereotyped depictions that have become a part of that toolkit. Colored mosaics hanging from schoolroom walls, bound within road atlases, or framed by the edges of a computer monitor look like maps because... that's what maps look like. And continents and nations look like, well, shapes on maps. Users of maps depend on them to discover unities and identities across space and time that are meaningful first of all because they are mapped that way. (This is why the Peters projection is odd-looking to those schooled in projections that depict very different shapes of the major landforms.) Signifying artifacts that don't much resemble the geometries of land and water that we know from our maps, but which are used by their authors for similar sorts of navigation and recognizance as we use our maps, are interpretable by us as maplike because they are founded on this steady creep of the map from metaphor to metonymy.[16] In an important sense, Turnbull observes (contradicting Korzybski's famous dictum), the map is the territory in the minds of its users, because the cultural force of specific maps and shared vocabularies of cartographic elements and techniques overdetermines every perception of the "real" spaces they figure (61). Dark Continents 18. Here, in its simplest terms, is my claim regarding cartographic representations of Internet diffusion and traffic: the blind spot of cartography--the matrix of cartographic selectivity and convention--is commonly bound in these images to an unrepresented structure of established and emergent political economies that is mistaken for the given, natural background of any discussion of the past, present, and future of the networked realms. [Figure 9] Figure 9. "International Connectivity, 9/91." © 1991 Larry Landweber and The Internet Society. http://info.isoc.org/internet/infrastructure/connectivity/ [Click on image to see enlarged view] 19. Figure 9 shows an image created by Larry Landweber, representing the state of Internet connectivity worldwide in September, 1991. The color scheme of the map distinguishes between four levels of connectivity: "No connectivity," "email only," "Bitnet but no Internet," and "Internet." The map includes no indications of the speed or bandwidth of the connections. Though it includes national boundaries, it gives no information regarding the distribution or state of networks within those boundaries. [Figure 10] Figure 10. International Connectivity, 6/97." © 1997 Larry Landweber and The Internet Society. http://info.isoc.org/internet/infrastructure/connectivity/ [Click on image to see enlarged view] 20. Figure 10 shows a similar map created by Landweber, showing Internet connectivity worldwide in June, 1997. The later map suggests that a great deal has changed in the intervening six years. Whereas only a few countries had full Internet access in 1991, there appear to be even fewer countries without full access in 1997.[17] 21. Figure 11 shows a map created by Mike Jensen representing the total bandwidth linking African nations to international Internet hubs in May, 1996. Figure 12 shows a similar map representing bandwidth in October, 1998.[18] Once again, the differences between the maps suggests a dramatic expansion of Internet diffusion: a total of 64 kilobits per second (Kbs) or greater bandwidth is now available in most African nations, whereas that level of bandwidth was previously available in only a handful of countries. (Keep in mind that the T1 connections common in most of the wired world are capable of transmitting up to 1,544 Kbs--not quite enough for full-screen, full-motion video. T3 connections are capable of transmitting up to 44,736 Kbs.) [Figure 11] Figure 11. "International Links by Country." Mike Jensen, May 1996. http://www3.sn.apc.org/africa/afstat.htm [Click on image to see enlarged view] [Figure 12] Figure 12. "International Links by Country." Mike Jensen, October 1998. http://www3.sn.apc.org/africa/afstat.htm [Click on image to see enlarged view] 22. All four of these maps suffer classic problems of areal aggregation common in choropleth maps.[19] As was the case of the Internet maps shown in Figures 3 and 4, the use of a single metric to identify the kinds of connectivity or levels of bandwidth within an entire nation gives no indication of the distribution of network resources inside the borders of that nation. Moreover, the limited palette of the map's key encourages misleading equivalencies between nations, obscuring dramatically different conditions of access. (For example, the total bandwidth of international connections for most of the nations shown in Figure 12 is near 64 Kbs. The total bandwidth in Egypt is, however, about 2000 Kbs; in South Africa, it is about 37,000 Kbs. [Spangler 46].) The fundamental problem with these representations of Internet diffusion is that they measure levels of diffusion within a (cartographic) discourse of discrete national-political identity which is arguably irrelevant to the global structure of that diffusion, and which eliminates much of the variance in material conditions upon which that structure depends. None of these images is able to account for the extreme local obstacles which must be overcome before anything like a viable African Internet is possible, at least as netizens of digitally-saturated, liberal-democratic nations understand the Internet. 23. The economic, technological, and institutional legacies of colonial occupation are discernible in every aspect of the African networks. Of an estimated 151 million online users worldwide in December, 1998, fewer than 1%--somewhat less than a million--live on the African continent.[20] Approximately 800,000 of those live in one country, South Africa. Internet node density in Africa--the ratio of active nodes to overall population--ranges from a low of 1 to 65 in South Africa, to a high of 1 to 440,000 in The Democratic Republic of the Congo (which nation includes the region of Conrad's "blankest of blanks" [Jensen]). By way of comparison, the average Internet density worldwide is about 1 to 40; in the U.S., it is nearly 1 to 6. Excluding South Africa, Sub-Saharan Africa as a whole has fewer Internet nodes than the nation of Latvia (Jensen). 24. In October, 1998, approximately forty-nine African nations and territories had some level of Internet access in their capital cities. The greatest part of the available network bandwidth in these locations is reserved for multinational firms with regional headquarters in the cities, or for university and government employees. Eighty percent of Africans live in rural areas. Network nodes outside of the largest cities--and there are very, very few of these--are punishingly slow and expensive, making them ill-suited for anything besides FidoNet-style store-and-forward email.[21] Local PTT and power infrastructures are notoriously unreliable and weighed down with bloated, inefficient state bureaucracies held over from the colonial era. In many regions, these infrastructures are at risk of sabotage and pillage from warfare between and within states, political corruption, or outright thievery. Outside of the wealthiest commercial concerns and the largest universities, much of the computer hardware and software in sub-Saharan Africa is antiquated, unevenly distributed, and costly to maintain and update. Repairs usually require hard or foreign currency badly needed for other social services. The lack of a well-developed cadre of computer scientists and technologists (only a few African nations offer advanced degrees in computer science and telecommunications) and of a skilled mid-level managerial class poses enormous barriers to efficient network implementation and support (Odedra 26). 25. These constraints on network development reflect a more general lack in Africa of the basic telecommunications resources upon which digital networks in most of the wired world depend. Africa, with over 12 percent of the world's population, has only 2 percent of the telephone lines. By comparison, Latin America has 8% of the population and 6% of the lines.... The teledensity... in sub-Saharan Africa... equates to approximately one phone line for every 200 people. By comparison, the teledensity in the United States is 65 (equivalent to one phone line for every two people), and 45 in Europe.... There are, in fact, more telephone lines in New York or Tokyo than in the whole of Africa. (Butterly)[22] The poor state of telecommunications infrastructure in most of the continent prevents expansion of existing network services, and virtually insures that large portions of Africa will remain far behind much of the wired world with regard to digital communication technologies, which seem to require firm foundations in analog technologies before they are able to take hold.[23] 26. Some companies and political institutions looking to accelerate network diffusion in Africa have proposed increased reliance on "wireless" telecommunication systems, connected by Low Earth Orbit (LEO) satellites, which would not depend on legacy systems of fiber or copper.[24] But these networks depend no less than wired networks on a reliable power grid; maintaining the infamous "last mile" of wire or fiber between the downlink station and the desktop will likely prove even more troublesome in Africa than it has in the developed world. The underlying technologies of wireless communications are more complex, technically fragile, and costly to set up and administer than those of traditional wired networks. LEO systems require many more satellites than the Geosynchronous Earth Orbit (GEO) systems now in use for most international voice and television traffic, and rely on extremely sophisticated networking schemes to manage signal "handoff" between satellites. LEO satellites have shorter life-spans than do their GEO counterparts, adding to the costs of long-term maintenance of these systems. The LEO networks now under development will support paging services and voice transmissions--comparable in quality to land-based cellular phone systems--but only very slow data transmissions (on the order of 2.4 Kbs).[25] Implementation of LEO satellite systems capable of handling the bandwidth required for multimedia and high-speed data exchanges faces substantial technical and financial obstacles; none of these systems is much beyond the drawing board (Evans 78).[26] Moreover, the economic restructuring resulting from increased reliance on wireless networks could prove catastrophic for the African nations which currently depend on the export of minerals widely used in wired infrastructures.[27] 27. These are hardly numbers to sustain fantasies of interactive distance education, e-commerce, real-time collaboration, or even sporadic text-only Web browsing. They suggest that utopian visions of a world network culture that includes Africa (with a population roughly equal to those of North America and Europe combined) must be tempered by pragmatism concerning the sophistication and scope of the technological, economic, and organizational infrastructures that would be required to develop and maintain such a culture. 28. They must be tempered as well by some critical thinking about the nature of disparities that divide the unnetworked from the networked spheres. Public and academic discussions of "access" in the West tend to be dominated by language of privacy, censorship, and freedom of speech, framed in idioms having meaning only for netizens of wealthy liberal democracies. The two-page spread shown in Figure 13 is from an article in a 1997 issue of WIRED, entitled "Freedom to Connect." "Freedom" in this case means "freedom of content" and "freedom to access any and all material online" (106)--in other words, the freedom of the user to read, see, and hear anything she wishes to read, see, and hear on the Internet, without the interference or surveillance of a government or law-enforcement agent. [Figure 13] Figure 13. Leila Conners, "Freedom to Connect." WIRED 5.08 (1997): 106-7. (Image © Michael Crumpton. Reproduced with permission.) [Click on image to see enlarged view] 29. "This map does not reflect the reality," notes the author, "that this freedom is often locally restricted without formal legislation" (106). The ambiguous phrase "restricted without formal legislation" fronts in this context for a host of material and political barriers to access that have no necessary relation to the spread's narrow definition of "freedom." Indeed, it would appear that, apart from a few liberal Western democracies (notably excluding the U.S.), "freedom to connect" as defined by this map is primarily a function of a government's inability to regulate the behavior of those citizens who have Internet access, rather than its tolerance of what they choose to do with that access. The crudeness of the map's visual scale--distinguishing free from unfree connections in only five steps that must represent differences across all political-legal, economic, and technological contexts--begs the question: how meaningful is it to compare the "freedom to connect" in Mexico City, Paris, Brisbane, and Algiers, and come up with the same result? One answer is suggested by Figure 14, which shows a world map from Michael Kidron and Ronald Segal's State of the World Atlas, depicting international differences in state-sponsored censorship and political and religious oppression. The Kidron and Segal map uses fewer discrete metrics for oppression than does the WIRED map (four, instead of five), but its assumptions regarding the nature of "freedom" result in a very different picture of the world. Citizens of nations who are "free to connect" to the Internet, according to WIRED, may be unable, according to Kidron and Segal, to connect on more human registers. [Figure 14] Figure 14. "The Three Monkeys." World map showing "states' infringement of freedoms of belief, expression, communication, and movement." (Reproduced with permission from Michael Kidron and Ronald Segal, The State of the World Atlas, 5th edition, Penguin, 1995. © 1995 Myriad Editions, Limited.) [Click on image to see enlarged view] 30. In equating "freedom to connect" with "freedom to surf," the WIRED map confuses possible movements of data with the actual movements of information and the free practice of thought and belief. It thus excludes evidence of historical and political structures that have restricted access to information technologies, their presumed benefits, and their economic and social costs, in the unwired world. "Some unfortunate nations," the legend of the map comments, "have no physical means of connection" (106). Given the obvious bases of this misfortune in the continuing legacies of colonial exploitation, the patronizing character of this comment should scandalize the thoughtful reader. Note the concentration of most of the unconnected nations--colored dark purple in Figure 13--near the center of the map: an unknown interior, a lack that seems to explain much... by saying nothing at all. The View from Above 31. What are the semiological functions of the political borders traced on these maps of Internet infrastructure and traffic? The borders are on the one hand heuristic conveniences: they frame visualizations of Internet traffic within well-established cartographic conventions, rendering abstractions of Internet "space" and "density" easier to interpret for those who are familiar with those conventions. 32. But cartographic schemes of representation are not necessary to the depiction of virtual networks; the networks traverse political borders and render problematic concepts of local, national identity that these schemes reinforce. Borders are anything but merely descriptive--they are, rather, among the most coercive of instruments for naturalizing, reifying, and depoliticizing cultural-historical formations. They are in that regard hallmarks of the map's selective distortions of space and time, signifiers of political hegemony which overdetermine their depiction of roads, fences, rivers, coastlines, etc. The tracing of political borders in these maps of putatively virtual domains (their hegemonic effects unremarked by the authors, assumed to be both essential to the interpretation of the maps and accidental to their representations of the network) naturalizes specific relations between nation-state and network identities--and, as a result, obscures the global political forms of the Internet with a mosaic of individual national forms. The borders thus provide a scheme for naming only selected particulars of network diffusion, and excluding representation of structural forces which take effect across or outside the network: a kind of can't-see-the-forest-for-the-trees cover for larger, metanational systems of information flow.[28] 33. Figure 15 shows a map of the Internet in July, 1998, created by MIDS (Matrix Information Directory Services, Inc.), a research firm specializing in print and online reports of activity on the Internet and other telecommunications networks. Though the map's legend is ambiguous, the overlapping colored circles appear to mark the distribution of active Internet domains worldwide.[29] [Figure 15] Figure 15. MIDS. "The Internet, July 1998." http://www.mids.org/ [Click on image to see enlarged view] 34. This visualization scheme avoids the misleading national aggregations characteristic of many of the maps discussed earlier in this essay: it's clear that domains figured on the map are scattered unevenly within most nations. (Domains in Australia, for example, are shown to be clustered around the country's perimeter, near the major urban centers.) But the visibility of some of the national borders in this image signals this map's different encoding of access in terms of national-political affiliation. The visual saturation of domains in the U.S. and Western Europe obscures virtually all borders within those regions. Political borders within Africa, on the other hand, are easily discerned. One is given the impression of a nearly-monolithic presence of the Internet in two or three regions of the globe--and of equally monolithic absences most everywhere else. National-political identity is significant, this image suggests, exactly where the Internet is not. 35. Figure 16 shows a map of active Internet hosts per capita (that is, within nations) as of January, 1998, created by Martin Dodge ("Geographies"). Nations with similar distributions of active hosts are colored the same, and the most active nations are distinguished from one another by their positions on a vertical axis calibrated to the absolute number of hosts. Thus, only those nations with distributions much higher or lower than those of their neighbors are discernible as nations. Geographically contiguous nations at the same level of distribution merge into one another, supporting the impression of vast politically unmarked terrains free of significant activity, and others, politically-differentiated but also uniformly towering over the rest of the world. [Figure 16] Figure 16. Martin Dodge, "Internet Hosts Per Capita, January 1998." http://www.geog.ucl.ac.uk/casa/martin/aag/aag.html [Click on image to see enlarged view] 36. Figures 15 and 16 would seem at first view to invert each other's schemes for showing the relation of network diffusion to national-political identity, but I would counter that they thus share a common economy that should be by now very familiar: on/off, traffic/no traffic/, wired/unwired. Think of Figure 16 as a sort of topographical map. Across its peaks and valleys emerges an archipelagic, virtual landmass that traverses the conventional boundaries of continents and nations. Viewed from on high, the vast, flat surface of the network's digital plains seem far removed, alien and obscure. Between Borders The border is all we share / La frontera es lo unico que compartimos. -- Guillermo Gómez-Peña, Warrior for Gringostroika 37. The peaks and valleys of this virtual continent are evidence of forces unmarked on these maps: the plate tectonics, if you will, of a process of technological and cultural change that is nearly invisible, unthinkable, expressly because it covers its forms with counterfeit ubiquity and technological reasonableness. Obscurantism of the spectacle; masquerade of the border in the place of other, multiple and heterogenous, identities: these maps support triumphalist fantasies of the potential (and desirable) saturability of the Internet, expressly by excising representation of the varied historical conditions that have generated the binarisms on which they depend. Recasting the fractiousness of material culture in the tidy efficiencies of the digital sample, they hide nascent lines of force that thread through and across those stark divides. 38. These lines of force are discernible, I propose, only by a contrarian reading of these images which does not take their forms for granted: a decentered regard by which the maps (and the very logic of mapping itself) may be seen to describe, not the light and dark continents of high colonialism but rather--though only indirectly--an emerging, virtual "dark continent" specific to our historical moment. This new political-symbolic structure traverses and fragments prior political entities, even as it makes use of fantasies of national identity to cover its advances.[30] In a deeply ironic way, the networked instauration of the dark continent reverses the extractive logic of classic colonialism: instead of raw materials (ore, precious stones, humans) freighted out of the heart of darkness for consumption by the wired-colonial metropole, the information order, to the extent that it penetrates the unwired world, will be largely devoted to freighting information in its motley forms into the benighted realms. In this context, "information" has the sense of both a commodity--a thing for sale over the networks--and a coercive force: the networks are able to inform the unwired realms; the new dark continent reproduces itself over the wires without regard for the prior conventional definitions of nation, region, or continent.[31] 39. What does this emergent telegeographic entity look like? How might we figure its contours? The problem with those questions is that the new digital information order doesn't look like much of anything, at least not anything that may be named by the metageographical idioms on which the most common representations of the global networks rely. It may be possible to produce images of the virtual terrains that adapt techniques of scientific visualization to those of classical cartography, while unbinding the former from its most positivist and politically naive biases, and the latter from its reliance on outdated and pernicious representations of the lived places of the human world. But that's not what the idioms of the dominant cartographic conventions do; they instead tend to unname, to render unremarkable (literally unseeable) the political economy shaping the new information orders. 40. As I've suggested throughout this essay, the selective distortions of the map, though mathematically and culturally irreducible, render it a problematic construct for describing the heterogenous conditions and practices of the emerging global telecommunications networks. Maps must mislead if they are to be usable; that much is clear. What matters most in any appeal to a tropology of maps and mapping is whether or not we understand the uses that may be concealed in our reception of maps as a privileged way of figuring human spaces, be they material or virtual.[32] A provisional strategy for fostering consciousness of the prejudices of the map might be simply to multiply in a disorderly way the instances of Internet mapping practice: to read maps of the wired and unwired domains against representations that interpret human spaces in terms for which wiring may be irrelevant, or which eschew conventional cartographic signs in favor of formations which entirely detach the network from its national-political borders. (The metaphor of the "dark continent" I've proposed here should be conceived as operative only within contested and multiple contexts such as these.)[33] 41. A more salient question with regard to these and similar images may be, why should it matter? So what if the maps are inaccurate? Calling for more "accurate" visualizations of Internet diffusion--for example, challenging fantasies about the uniformity of the diffusion and growth of telecommunications access (MOSAIC)--may not in the end be different from calling for new maps, without questioning the method of the map: that is, attempting to increase the sample size, criss-crossing the blanks with additional lines of force, finding a more efficient scheme for measuring shades of light and dark. We should be cautious of strategies that seek to repair a lack of data with more, "better" data, while still relying on the epistemological and visual figure of the border-as-national-boundary. As the map has been a primary tool for commodifying physical space, it may serve equally well as a tool for commodifying information; replacing a Mercator with a Gall-Peters does not eliminate the epistemic coercion of the map.[34] 42. On the other hand, misrepresentations of conditions of access and identity in the wired and unwired realms do certainly matter in one sense: the imaginary orders they produce and sustain are likely to return to the real as consequences of economic policy, military intervention, and technological and symbolic exclusion. Those who will plan the network infrastructures of the next century are likely, however we may characterize their conscious motives, to rely on maps of one sort of another. Sustained, progressive critique of the metageographies of Internet diffusion and traffic must look beyond the limited (and limiting) visual vocabularies of national-political identity, and base its investigations on new schemes for representing the archipelagic landscapes of the emerging political and technological world orders. 43. If we analyze techniques for figuring the formations of the Internet with an awareness of the historical and ideological constitution and effects of those techniques--if we look for traces of the excluded matter they cannot show--we will perhaps be better able to make sense of a studied neglect of fractious material realities characteristic of the predictions of both the cyberspace utopians dreaming of "one world, one network," and the economic opportunists looking for new unknown interiors ripe for exploitation. This will require a careful and comprehensive examination of the emerging class and semiological domains of the post-colonialist world (Alkalimat 278-85), and a better understanding than we now have of the broad reach of networks of all kinds, and of the changes they are effecting in regional and global systems of cultural and economic exchange.[35] The accelerating development and expansion of the virtual archipelagoes call for this kind of work, because the new structures of information and movements of virtualized capital will be among the most powerful forces redefining human geographies and semiological systems of the next century. Established definitions of national and economic identity and relation are being recast by the heterogenous and often contradictory material and political facts of the cyber-ecumene, where bands of light and dark, access and no access, are knotted in unprecedented ways within the "same" city, nation, or continent.[36] In the networked era, the heart of darkness is an interstitial formation--which is to say, its borders are drawn around and between us. The maps we now have obscure this emerging terrain. School of Literature, Communication & Culture Georgia Institute of Technology terry.harpold@lcc.gatech.edu ----------------------------------------------------------------- COPYRIGHT (c) 1999 BY TERRY HARPOLD, ALL RIGHTS RESERVED. THIS TEXT MAY BE USED AND SHARED IN ACCORDANCE WITH THE FAIR-USE PROVISIONS OF U.S. COPYRIGHT LAW. ANY USE OF THIS TEXT ON OTHER TERMS, IN ANY MEDIUM, REQUIRES THE CONSENT OF THE AUTHOR AND THE PUBLISHER, THE JOHNS HOPKINS UNIVERSITY PRESS. THIS ARTICLE AND OTHER CONTENTS OF THIS ISSUE ARE AVAILABLE FREE OF CHARGE UNTIL RELEASE OF THE NEXT ISSUE. A TEXT-ONLY ARCHIVE OF THE JOURNAL IS ALSO AVAILABLE FREE OF CHARGE. FOR FULL HYPERTEXT ACCESS TO BACK ISSUES, SEARCH UTILITIES, AND OTHER VALUABLE FEATURES, YOU OR YOUR INSTITUTION MAY SUBSCRIBE TO PROJECT MUSE, THE ON-LINE JOURNALS PROJECT OF THE JOHNS HOPKINS UNIVERSITY PRESS. ----------------------------------------------------------------- Notes An early version of this essay was presented at the 1998 Conference on Science, Technology, and Race (STAR '98), Georgia Institute of Technology, April 2, 1998. I thank Deepika Petraglia-Bahri, Jeanne Ewert, and Kavita Philip for their invaluable comments on the manuscript of this version. I thank also Mark Monmonier for his kind assistance in creating the maps shown in Figures 6, 7, and 8. 1. GoGwilt's subtle and evocative reading of the "double-mapping" of Europe and the African colonies in Conrad's work links the political and epistemological functions of maps in the late nineteenth century with Conrad's equivocal reconstructions of his Polish childhood. 2 See Wood (32-47), Turnbull (28-36), and Sack (ch. 1). 3. A synonym for "blankness" in this discourse is "darkness." The late nineteenth-century journalist-explorer Henry M. Stanley popularized the description of Africa as "The Dark Continent" in several best-selling memoirs of his journeys in Equatorial Africa. In Stanley's books, "darkness" is an unfortunate state opposed to the light of Reason: Africa's spiritually and politically benighted interior is badly in need of European science, religion, and civilization--in short, of the stabilizing, epistemological ordering that only a finished map can signify. 4. These and similar images are included in Martin Dodge's "An Atlas of Cyberspaces" (http://www.cybergeography.org/atlas/). Dodge categorizes the visualizations on this site into twelve different categories: Conceptual, Geographic, Traceroutes, Census, Topology, Info Maps, Info Landscapes, Info Spaces, ISP Maps, Web Site Maps, Historical, and 3D Gallery. 5. "Although the geographic map is drawn on a 2D plane, it can be arbitrarily positioned and oriented in the space. Therefore the user can interactively navigate the display by translating, rotating, scaling, and visualizing the network from different perspectives under different rendering conditions" (Cox, Eick, and He 52). 6. This repetition of the structural functions of the blanks of high colonial maps is evident in a series of Internet maps created by Batty and Barr for their 1994 article on "The Electronic Frontier" (707-708). The six maps in the series illustrate the growth of the Internet between July, 1991 and January, 1994, in the form of a global grid on which countries with active nodes are shown in outline form, and those without active nodes are invisible. Over the course of the series, small landmasses (more properly, political masses) seem to appear over time out of a cartographic void. Much of the lived human world remains unseen--literally not present--in the topography of the wired world. The maps are reprinted in Kitchin (40-41). 7. Turnbull also uses this illustration of the Bellman's chart in the introduction to his discussion of the conventionality of maps (3). 8. See Monmonier (ch. 3) for a discussion of "generalization" techniques to improve map legibility and Monmonier (chs. 7-8) for a survey of classic cartographic distortions in the service of political and military ends. Wood (ch. 5) critiques these techniques more aggressively than Monmonier, emphasizing their support of ideological ends. 9. See Monmonier (ch. 1), Snyder, and Wood, for typical critiques of the modern misapplications of this projection. 10. Arno Peters created his 1967 projection apparently unaware that similar techniques had been used by James Gall to produce a nearly-identical map more than 80 years previously. See Monmonier (11-19) and Snyder (165-66). The Gall-Peters projection was adopted by UNESCO, the World Council of Churches, and several other international organizations, expressly because of the claims of political equity made for it. For detailed discussions of the debates concerning the Peters projection, see Wood (ch. 3) and Monmonier (ch. 1). 11. Snyder lists more than 150 projections and variants developed in the 20th century alone (277-86). See Misra and Ramesh (93-152) for an overview of the mathematical limitations of planar projections, including discussion of the Mercator, Goode Homolosine, and Gall-Peters projections. 12. The advantages of animated 3D projections over 2D schemes are promoted in one influential paper on geographical mapping of traffic through a single WWW server: "by providing true three-dimensional views, stereopsis and virtual reality allow us to avoid the distortion problems that have plagued cartographers and planar projections" (Lamm, Reed, and Scullin). Most projections of this kind are properly 2.5D views, as they must be displayed on two-dimensional fields (computer screens), with the illusion of depth suggested by movement or shading, as in the Cox, Eick, and He "arc maps." 13. Wood is a good example of this line of argument. 14. The projective "corrections" of Gall-Peters did not, of course, eliminate its dependence on conventions closely tied to European and Northern hegemonies, a point that Turnbull demonstrates by recentering the projection on the Pacific, and turning it upside-down (with the Southern pole at the top of the map). See Turnbull (7). 15. What does the state of Georgia look like? The nation of France? The continent of Africa? These questions depend on--indeed, are meaningless outside of--a cartocentric worldview that confuses being, visibility, and mappability. 16. Turnbull (ch. 5) includes a discussion of Aboriginal Australian dhulanj ("the footprints of the Ancestors"), depictions of clan territories in the form of elaborate animal shapes and geometric patterns. Though the patterns of the dhulanj look nothing like the contours of Western cartographic projections, these "maps" of the Aboriginal homelands are, Turnbull demonstrates, no less specific nor less accurate for their intended uses. 17. Attentive readers will notice that the level and distribution of Internet access depicted in Figures 9 and following are not consistent. Reliable information regarding Internet use, even in countries with relatively long histories of use, is notoriously hard to come by. Moreover, the authors of these maps appear to have drawn their information from different sources. Nonetheless, I would argue, the maps are consistent in two regards: 1) their demonstration of broad trends of changes in Internet access and use, and 2) a tendency of these and similar visualizations to obscure the particularities of access and use. 18. These images are included in an animated .gif created by Jensen, showing the change in link distribution and bandwidth between 1996 and 1998. See http://www3.sn.apc.org/africa/afranim.gif. 19. A choropleth map uses graduated tones and discrete areal units to depict variations in some condition over a geographical area. Changes in the selection of breaks between categories of a choropleth map can result in dramatically different interpretations of the same data. For example, the total international bandwidth available in South Africa is so much larger than that of any other African nation that it arguably deserves its own measure on Jensen's maps. If that were the case, Figures 11 and 12 would tell a different story about recent changes in African Internet diffusion: South Africa continues to be the exception to nearly every general rule of the African Internet, for historical and economic reasons which are easily guessed. For a discussion of the ways in which areal aggregation may influence interpretation of choropleth maps, see Monmonier (ch. 10). 20. See Nua Internet Surveys (http://www.nua.ie/surveys/how_many_online/index.html). Other estimates of active Internet users worldwide vary somewhat from this figure. 150 million, plus or minus 10%, appears to be a reasonable estimate. See also "Headcount.com" (http://www.headcount.com/), and Staple (79). 21. In some countries, a 9.6 Kbps international leased line can cost the equivalent of US$130,000 annually; dialup access charges often exceed US$10.00 hourly (Spangler 46). (Average annual income in all but a handful of African nations is less than $2000 annually, with the majority of nations averaging at less than $1000 annually [Kidron 36-37].) It is difficult to ascertain the numbers of persons that make regular use of the African networks. This is especially true of modem-based, store-and-forward email systems, which do not use the same network structure as true Internet systems. In the developed world, it is not uncommon for a single Internet node to serve dozens of users. An informal review of one online list of Sub-Saharan networks ("AAAS User's Guide to Electronic Networks in Africa" [http://www.aaas.org/international/africa-guide/bynation.htm]) suggests that the number of users supported by African networks varies widely. In the most networked African nations (South Africa, for example), use of computers probably resembles the multiple-user model common in more developed countries. In many nations, however, the total number of users for a network may run no higher than single digits. 22. In this regard, most of the world is more like Africa than the U.S., Western Europe, or Japan. Average teledensity worldwide is roughly 10 lines per 100 inhabitants. About 80% of the current world population has never used a telephone (Goldstein 339, 365n4). 23. The myth of network culture's essential virtuality tends to obscure the dependance of digital networks on other, more material--one might say, mundane--networks. Arnum and Conti have identified a close historical correlation between the deployment of non-Internet wire (telephony, television, electrical grids), paved roads and railways, and gross national product. The wealthiest, most widely-paved or -railed, and most heavily-wired nations are in general, those in which diffusion and use of the Internet has been greatest, and is growing at the greatest rate. See also Hargittai's conclusion that the unwired nations are unlikely to ever catch up with their wired counterparts in the distribution of digital technologies. 24. "LEO satellites can provide relatively inexpensive communications between simple earth receivers and satellites from any where in the world (even Africa!). Africa may actually be at an advantage in implementing these new technologies as it does not have an extensive investment in existing infrastructure" (Butterly). Such a system is an important component of the "USAID Leland Initiative: African GII Gateway Project" (http://www.info.usaid.gov/regions/afr/leland/). 25. The first of these, Motorola's Iridium (http://www.iridium.com/), will go online near the end of 1998. A recent cover story about Iridium in WIRED magazine breathlessly endorsed the idea that the service will "leap geopolitical barriers in a single bound" ("Anyware! Iridium launches the global phone") (Bennahum 134) with an image that combined metageographic tropes with erotically-valenced clichés of the cyber-denizen. The cover showed the naked scalp, nape, and back of a young woman of uncertain ethnicity. Projected on her skin: a brightly-colored map of Russia, Asia, and the Indian Ocean, presumably important markets which will be newly opened to subscribers to the service. Iridium will face steep competition from expanding cellular phone networks, and is expected to have no more than 11 million subscribers by 2007 (http://www.ovum.com/). 26. Though he is cautious with regard to the technical scale and cost of large, satellite-based systems, Goldstein is sanguine about the potential of these systems to accelerate the growth of telecommunication infrastructure in the unwired world. Of course, we cannot expect that most developing countries will reach teledensity levels that rival those of developed countries in just 10 years. Nevertheless, millions of individuals and businesses in heretofore unconnected or underconnected towns and villages will gain at least some access to essential national and even international narrowband voice, data, and Internet services--leapfrogging the traditional wait for wired services. This will be one of the most significant developments in communications over the next decade, although the true impact of these new entrants into the electronic marketplace may only begin to be felt by 2008. (340) Goldstein is more careful than many evangelists of the wireless realms to unpack the multiple technical, economic, and political obstacles to the development of these systems. But the devil is in the details: the capabilities of "narrowband" versus those of "broadband" services; the practical sense of "at least some" level of access, etc. 27. The Democratic Republic of the Congo and Zambia depend on copper exports for over 50% and 95% of export earnings, respectively (Alkalimat 279). 28. My reading of the signification of political borders in these images is in a sense the reverse of Edward Tufte's complaint against a series of thematic maps illustrating cancer deaths by U.S. county: "They wrongly equate the visual importance of each geographic area rather than with the number of people living in the county (or the number of cancer deaths.) Our visual impression of the data is entangled with the circumstances of geographic boundaries, shapes, and areas" (20). Yes and no. "Geographic boundaries" are social and political structures applied to and enforced over physical terrains. Events that occur within them may be impossible to disentangle from those structures. The lines only get in the way--obscure the "real" purpose of the image--if one assumes that the things represented by the data (cancer deaths, Internet access) may be abstracted away from all political interests. 29. MIDS produces a similar animated map, projecting daily Internet traffic latency and congestion on a spinning globe. The globe can be viewed online, at http://www.mids.org/weather/world/index.html. 30. The inverted form of that fantasy is the McLuhaneque "global village," spanning prior national borders via the wonders of new communication technologies. McLuhan's mediascape is, of course, no less selective in its misrepresentations of conditions at the local level of communication practice than the misleading areal aggregations in the images I've discussed in this essay. (For a brief but decisive critique of McLuhan in this regard, see Jarvis [24-29].) The "global village" is, perhaps, another name for the structure I'm calling the new "dark continent." 31. Shadowing the voids in virtual space is another kind of absence: the lack of any articulated alternative to the boosterism for privatization and free market forces as the only and the inevitable fix for the backwardness and decrepitude of the telegeography of the developing world. In the language of the cyberlibertarians, this enthusiasm for technological "progress" is recast in the uplifting terms of information emancipation and an ethos of unrestrained individualism. Given the apparently--the unimaginably--unstoppable expansion of network culture at this moment in history, the private and the personal seem perilously close to losing any vital particularity they may have once had, and to being obscured by the penumbra of larger, unthinkable political formations that are quite able to maintain for themselves a different sort of invisibility. 32. This caveat will apply whether the discourse of mapping is conceived of in terms sustaining or critiquing established political economies. Brian Jarvis's recent Postmodern Cartographies brilliantly demonstrates that selective abstractions of economic materiality via the trope of the map are common in the writings of many critics on the Left. 33. In a simple way, my comparison of the "freedoms" figured in Figures 13 and 14 is an example of this. Kidron and Segal's atlas is an excellent model of multiple mapping: it includes fifty maps, cartograms, and charts depicting political, economic, and social indicators. Taken as a whole, the atlas figures profoundly heterogenous and, at times contradictory, assertions about the conditions of life for most of the planet. Examples of visualizations of the Internet and other networked formations that do not rely on the cartographic conventions informing most of the images in this essay may be found at Dodge's "Atlas of Cyberspaces" site (http://www.cybergeography.org/atlas/). See also a new project to map thematically-related sites on the WWW, based at the Guggenheim Museum, http://www.cybergeography.org/atlas/ (Ippolito 17). 34. Nor does the preference for more "accurate" mapping address the deeply intertwined and mutually-constitutive roles of modern mapping techniques, scientific positivism, visual subjectivity, and market economies. See Avery for a brief but evocative foray into the relation of maps to capital, colonial exploitation, and modern conceptions of the subject. 35. The Internet is only one form of the accelerating spread of networked formations of power and representation at the close of the 20th century. A comprehensive account of the emerging networked realms--ignoring for the sake of clarity the many questions that the word "comprehensive" raises in that description--would need to take into account the massive and heterogenous material networks that support the diffusion of formations like the Internet as their "infrastructure," including such systems as copper and fiber wiring for both information and power, microwave and satellite communication systems, cellular phone and beeper systems, the manufacturing and distribution systems tied to these systems, etc. In his recent typology of virtual geographies, Michael Batty identifies a realm he calls "cyberplace" that roughly corresponds to this broader sort of network: [Cyberplace is...] the substitution, complementation, and elaboration of physical infrastructures based on manual and analogue technologies by digital... [It] consists of all the wires that comprise the networks that are being embedded into man-made [sic] structures such as roads, and buildings. It extends to the material objects that are used to support this infrastructure such as machines for production, consumption and movement that are now quickly becoming a mix of the digital and the analogue. (346) Batty's distinction between "cyberplace" (the material realm reshaped by the demands of the digital) and "cyberspace" (the virtual realm that is being created to represent the digital for its users) unironically reverses Michel de Certeau's distinction between place and space. For de Certeau, place is the static, rationalized--the mapped--form of space, which is multiple, heterogenous, and contingent (117). Approached with de Certeau's distinction in mind, Batty's "cyberplace" takes on the appearance of a vectorized, rationalized material order whose possibilities are determined by the reductive binarisms of the digital order. 36. The cyberspatial blank that divides my office at Georgia Tech from crumbling, technologically-antiquated public schoolrooms within city of Atlanta--one of the most networked cites in the U.S.--is, in practical terms, as yawning and irreducible as the blank separating my office from classrooms in many cities in the developing world. See Moss and Townsend; McConnaughey, Nila, and Sloan; and McConnaughey and Lader. McConnaughey and Lader's research suggests that the "digital divide" in the U.S. (most often drawn along racial, economic, and regional lines), between Internet-connected and -unconnected households, is in fact growing. Works Cited AAAS Sub-Saharan Africa Program. User's Guide to Electronic Networks in Africa. http://www.aaas.org/international/africa-guide/bynation.htm. Alkalimat, Abdul. "The New Technological Imperative in Africa: Class Struggle on the Edge of Third-Wave Revolution." Cutting Edge: Technology, Information Capitalism and Social Revolution. Eds. Jim Davis and Thomas Hirschl. New York: Verso, 1998. 271-86. Arnum, Eric, and Sergio Conti. "Internet Deployment Worldwide: The New Superhighway Follows the Old Wires, Rails, and Roads." INET '98: The Internet Summit. July 21-24, 1998. Geneva, Switzerland. http://www.isoc.org/inet98/proceedings/5c/5c_5.htm. Avery, Bruce. "The Subject of Imperial Geography." Prosthetic Technologies: Politics and Hypertechnologies. Eds. Gabriel Brahm, Jr. and Mark Driscoll. Boulder, CO: Westview Press, 1995. 55-70. Barthes, Roland. Mythologies. Trans. Annette Lavers. New York: Farrar Straus & Giroux, 1973. Batty, Michael. "Virtual Geography." Futures 29.4/5 (1997): 337-52. ---, and Bob Barr. "Exploring and Mapping Cyberspace." Futures 26.7 (1994): 699-712. Bennahum, David S. "The United Nations of Iridium." WIRED 6.10 (1998): 134-38; 194-201. Butterly, Tom. "Constraints to the Development of the 'Wired' Economy in Africa." http://www.nua.ie/surveys/analysis/african_analysis.html. Carroll, Lewis. The Hunting of the Snark: An Agony in Eight Fits. Ed. Martin Gardner. New York: Penguin Books, 1995. Conners, Leila. "Freedom to Connect." WIRED 5.08 (1997): 106-7. Conrad, Joseph. Heart of Darkness. 1899. New York: W.W. Norton & Co., 1988. ---. A Personal Record. 1912. Marlboro, VT: The Marlboro Press, 1982. Cox, Kenneth, Stephen G. Eick, and Taosong He. "3D Geographic Network Displays." SIGMOD Record 25.4 (1996). (See also http://www.bell-labs.com/user/eick/bibliography/1996/ 3D_copyright.ps.gz.) De Certeau, Michel. The Practice of Everyday Life. Trans. Richard Nice. Berkeley: U of California P, 1988. Dodge, Martin. "An Atlas of Cyberspaces." http://www.cybergeography.org/atlas/. ---. "The Geographies of Cyberspace." Association of American Geographers Conference. Boston: March, 1998. http://www.geog.ucl.ac.uk/casa/martin/aag/aag.html. Evans, John V. "New Satellites for Personal Communications." Scientific American. April, 1998: 70-79. GoGwilt, Christopher. The Invention of the West: Joseph Conrad and the Double-Mapping of Europe and Empire. Stanford: Stanford UP, 1998. Goldstein, Irving. "Stars of Good Omen: Satellites in the Global Electronic Marketplace." The Future of the Electronic Marketplace. Ed. Derek Leebaert. Cambridge, MA: MIT Press, 1998. 335-66. Gómez-Peña, Guillermo. Warrior for Gringostroika. St. Paul, MN: Graywolf Press, 1994. Hargittai, E. "Holes in the Net: The Internet and International Stratification." INET '98: The Internet Summit. July 21-24, 1998, Geneva, Switzerland. http://www.isoc.org/inet98/proceedings/5d/5d_1.htm. Headcount.com. http://www.headcount.com/. Ippolito, Jon. "What Does Cyberspace Look Like?" ArtByte 1.4 (1998): 16-17. Jarvis, Brian. Postmodern Cartographies: The Geographical Imagination in Contemporary American Culture. New York: St. Martin's Press, 1998. Jensen, Mike. "An Overview of Internet Connectivity In Africa (December 1998)." http://www3.sn.apc.org/africa/afstat.htm. Kidron, Michael, and Ronald Segal. The State of the World Atlas. 5th ed. New York: Penguin, 1995. Kitchin, Rob. Cyberspace: The World in the Wires. New York: John Wiley & Sons, 1998. Lamm, Stephen E., Daniel A. Reed, and Will H. Scullin. "Real-Time Geographic Visualization of World Wide Web Traffic." Fifth World Wide Web Conference. May, 1996. http://www-pablo.cs.uiuc.edu/Projects/Mosaic/WWW3/. Lewis, Martin, and Kaeren E. Wigen. The Myth of Continents: A Critique of Metageography. Berkeley: U of California P, 1997. McConnaughey, James W., Cynthia Ann Nila, and Tim Sloan. "Falling Through the Net: A Survey of the 'Have Nots' in Rural and Urban America." National Telecommunications And Information Administration. July, 1995. http://www.ntia.doc.gov/ntiahome/fallingthru.html. ---, and Wendy Lader. "Falling Through the Net II: New Data on the Digital Divide." National Telecommunications And Information Administration. July, 1998. http://www.ntia.doc.gov/ntiahome/net2/falling.html. McIlwaine, John. Maps and Mapping of Africa: A Resource Guide. New Providence, NJ: Hans Zell Publishers, 1997. MIDS (Matrix Information Directory Services, Inc.). http://www.mids.org. Misra, Rameshwar Prasad, and A. Ramesh. Fundamentals of Cartography. Rev. and enl. ed. New Delhi: Concept Publishing Co., 1989. Monmonier, Mark. Drawing the Line: Tales of Maps and Cartocontroversy. New York: Henry Holt, 1995. ---. How to Lie With Maps. 2nd ed. Chicago: U of Chicago P, 1996. MOSAIC Group. "The Global Diffusion of the Internet Project: An Initial Inductive Study, March 1998." http://www.agsd.com/gdi97/gdi97.html. Moss, Mitchell L., and Anthony M. Townsend. "Spatial Analysis of the Internet in U.S. Cities and States." Taub Urban Research Center, New York University, April 1998. http://urban.nyu.edu/research/newcastle/newcastle.html. Nua Internet Surveys. "How Many Online?" http://www.nua.ie/surveys/how_many_online/index.html. Odedra, M., et al. "Sub-Saharan Africa: A Technological Desert." Communications of the ACM 36.2 (1993): 25-29. Press, Larry. "Tracking the Global Diffusion of the Internet." Communications of the ACM 40.11 (1997): 11-17. http://som.csudh.edu/cis/lpress/articles/worldacm.htm. Sack, Robert David. Homo Geographicus: A Framework for Action, Awareness, and Moral Concern. Baltimore: Johns Hopkins UP, 1997. Snyder, John Parr. Flattening the Earth: Two Thousand Years of Map Projections. Chicago: U of Chicago P, 1993. Spangler, Todd. "Africa's Net Challenge: Overcoming Regulation, Poor Telecom Infrastructure." Inter@ctive Week 5.47 (November 30, 1998): 46. Staple, Gregory C., ed. Telegeography 97/98: Telecommunications Traffic Statistics & Commentary. Washington, DC: Telegeography, Inc., 1998. Starrs, Paul F. "The Sacred, the Regional, and the Digital." The Geographical Review 87.2 (1997): 193-218. Tufte, Edward. The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press, 1983. Turnbull, David. Maps are Territories: Science is an Atlas. Chicago: U of Chicago P, 1993. "USAID Leland Initiative: African GII Gateway Project." http://www.info.usaid.gov/regions/afr/leland/. Wood, Denis. The Power of Maps. New York: Guilford Press, 1992.