Presenting the Nation's Cultural Geography
Donald C. Dahmann, Geographer and Adjunct Professor, George Washington University
At the dawn of the nineteenth century, the United States launched into its great age of geographic self-discovery. Expeditions sponsored by the federal government to reveal the nation's physical and human geographies began with the Lewis and Clark Expedition under President Thomas Jefferson: between 1804 and 1806, it created a thread of geographic information that linked the Mississippi River to the Pacific coast for the first time. Until the Civil War, the Army's Corps of Topographic Engineers continued to explore the nation's basic geographic elements, including a rudimentary mapping of landforms, rivers, ecological zones, and native peoples. The Civil War's conclusion and a heightened sense that all lands between Mexico and Canada lay open to settlement spurred a new wave of scientific expeditions of discovery, including the great geographic, geologic, and biological surveys headed by Ferdinand V. Hayden, George M. Wheeler, John Wesley Powell, and Clarence King. The nation embarked on modern geographic self-discovery with the emergence of two new forms of scientific representation: the national atlas, first realized in the Statistical Atlas issued by the Bureau of the Census with the 1870 census (Walker 1874); and the topographic mapping program established at the Geological Survey in 1882, the nation's mother map, to use the term employed by the program's first director (Gannett 1892), or what is today known as the national spatial data infrastructure (National Research Council 1993).
By the close of the nineteenth century, scientific accounts of the nation's physical geography, including its cartographic representation, had progressed to levels that remain reasonable even by today's standards. Providing scientific accounts of the nation's immensely diverse human geography, the result of a process of continual creation, alteration, and recreation, has proven to be a somewhat more complicated enterprise. At the national scale, this human geography has perhaps been captured best by the series of decennial censuses taken from 1790 to the present. The geographic, graphic, and cartographic expression of census statistical information lagged well behind its tabular expression throughout the initial decades of census taking in this country. A number of reasons account for this, not least of which was the perception that the human and physical geographies of the United States were relatively uncomplicated. Thus, it was not until the 1850 census, which included lands acquired from Mexico and Great Britain on the Pacific coast, that the spatial structure of the nation was perceived as being complex enough to warrant a thematic map in a census report. Using the latest scientific information available for creating subnational regions, superintendent of the census J. D. B. DeBow produced a thematic map that situated each state and territory within the nation's major river basins (DeBow 1854, 4 and 31-38). Using statistics from the same 1850 census, the German geographer August Petermann produced twenty-five thematic maps to inform his European audience of the contemporary status of U.S. geography (Petermann 1855, Petermann and Behm 1856). Differences between these American and European representations are attributable to Europe's greater geographic curiosity, interest in the visualization of statistics, and thematic cartography capabilities (Funkhouser 1937, Robinson 1982).
By the time the 1860 census returns were ready for tabulation, the nation was sinking into internal conflict. As a result, superintendent of the census Joseph C. G. Kennedy and his staff produced only an abbreviated set of reports, which included no graphic or cartographic representations. Census staff did, however, assist in the cartographic display of this new round of statistics in several other ways. Most importantly, they aided the war effort by preparing maps of Southern states for Union field commanders that displayed militarily vital topics: white population, slave population, predominant agricultural products (all by county), and rail and post-road transportation routes. Thematic maps using census statistics were also published privately to raise funds for sick and wounded Union soldiers and for other war-related efforts (Friis 1974).
The First National Atlas: 1870
The 1870 census provided the nation with an astounding array of graphic and cartographic depictions of its state of geographic knowledge, especially in superintendent of the census Francis A. Walker's grand folio of text, statistical graphics, and thematic maps, the Statistical Atlas, the nation's first true national atlas (Walker 1874). The appearance of the Statistical Atlas just after the Civil War stems, among other things, from the evolution of census taking in this country. By that time, the U.S. census had been transformed from a procedure to assist in apportioning Congressional representation among the states to a quantitative scientific accounting of numerous aspects of the nation's many human geographies (Anderson 1983).
The appearance of a national atlas also derived from an increasing desire to demonstrate the nation's progress in new and more popular publishing formats and from an interest in communicating scientific information more effectively (Mood 1946, Monmonier 1994, Porter 1986). It heralded the ethos of progress that dominated a reunited nation moving toward its centennial celebration, especially in "The Progress of the Nation," as one of its sections proclaims. It also responded to the emerging approaches to revealing new-found understandings of the nation's geography. By then, Americans had been introduced to Europe's more sophisticated capabilities for visualizing quantitative information, both graphically in charts and cartographically in thematic maps. Members of the fledgling American Geographical Society and American Statistical Society, including the geographer Daniel C. Gilman of Yale University, Samuel B. Ruggles of the American Geographical Society, and Edward Jarvis of the American Statistical Association, had attended meetings of the International Statistical Institute in Europe where new ideas concerning the graphic presentation of information were regularly discussed. Federal officials such as superintendents of the census J. D. B. DeBow and Joseph G. C. Kennedy and chief of the Treasury's Bureau of Statistics Edward Young also attended these meetings and several accounts of them appeared as federal government reports (Ruggles 1863; Young, Barnes, and Snow 1874).
Buoyed by this rising interest in the graphic presentation of statistics, the Census Office issued its first 1870 census thematic maps in its final reports, just before publishing the Statistical Atlas. These reports included a total of some twenty-six thematic maps (U.S. Census Office 1872). Once the idea for an atlas had been conceived and the early maps reviewed by others, the project was actively supported by geographers at the American Geographical Society, by the Secretary of the Interior, and by Congress, which approved a printing of five thousand copies. Initially issued in parts as sections were printed, the atlas appeared as a complete volume in 1874. Its title as a "statistical" rather than "national" atlas, perhaps one of the reasons this landmark federal publication is not better known today, must be understood in terms of the contemporary usage of the term "statistical." Today the term typically refers to quantitative information based on counts such as of people or the economy. During the 1870s, the still-new term referred to all forms of quantitative information used to govern the state.
Contributions to the atlas came from throughout the federal government and from at least a dozen of the day's prominent "men of science," including the geographer Arnold Guyot of Princeton University, geologist Josiah D. Whitney of Harvard University, and botanist William H. Brewer of Yale University. The titles of the atlas' s three sections indicate its broad range of physical and human topics: "Physical Features," "Population, Social and Industrial Statistics," and "Vital Statistics." All sections provided national-level information. Perhaps the most noteworthy section in terms of the era's prevailing spirit of manifest destiny is "The Progress of the Nation, 1790 1870," with its extraordinary series of thematic maps that delineate the march of Euro-American rural and urban settlement across the continent. When this series of maps was updated for the 1890 census and the settlement frontier was deemed too complex to be represented cartographically as a simple line or front, it spurred historian Frederick Jackson Turner's argument that an era of freedom and opportunity in America's history had come to an end.
Walker's atlas was the federal government's first attempt to provide a national perspective on the country's physical and human geographies with both explanatory text and scientific illustration. The first of six atlases that would be produced by the Census Office, it was celebrated from the first. It was awarded a "medal of first class" at the 1875 International Geographical Congress (Anonymous 1875b); it received several detailed and laudatory reviews in European scientific circles (Heywood 1875, Ravenstein 1875, Anonymous 1876); and it was heralded at home as "the most valuable contribution to the comparative statistics of the United States that has ever appeared" (Anonymous 1875a) and "one of the most instructive publications ever issued by our [federal] government" (Brewer 1875). Geologists used it to urge the federal government to prepare a still more detailed geologic map of the nation (Anonymous 1873).
National Atlases: 1880-1920
The high-water mark of census atlases in their breadth of coverage, innovation, and excellence of graphic and cartographic expression was achieved with the 1880 and 1890 censuses, the work of Henry Gannett, then geographer for both the Census Office and the Geological Survey. Like Walker's atlas, both publications appeared in large folio size (Hewes and Gannett 1883, Gannett 1898). With the 1900 census, the size of the census atlas shrank to the quarto format that was the standard size of other agency reports, although production standards remained high (Gannett 1903). The final two atlases, issued with the 1910 and 1920 censuses, were both routinized productions and largely devoid of color and graphic imagination, although dot maps and line pattern choropleth maps were introduced to advantage (Sloane 1914, Sloane 1925). The remarkable series of maps portraying the temporal evolution and spatial diffusion of the nation's rural and urban settlement, which had first appeared with Walker's atlas and been updated in subsequent atlases, appeared for the last time in the 1910 atlas. Introductory and explanatory text also appeared for the last time with the 1910 atlas.
The use of charts and thematic maps to display statistical information in non-governmental publications increased dramatically in the United States during the first decades of this century, thanks in part to Walker's and Gannett's introduction of a variety of visual techniques for communicating quantitative information to both scientific and general audiences (Edwards 1969). At the same time, the innovative spirit, so evident in early atlases, faded. Even though topical coverage diminished in later atlases and production became routinized, it is still regrettable that the Census Bureau followed its Advisory Committee's recommendation and did not publish an atlas with the 1930 census (Joint Census Advisory Committee 1934).
The nation was equally interested in learning about its physical and human geographies, as is indicated by the increased publication of privately issued atlases throughout this period. And as soon as the country became engulfed in the economic crisis of the 1930s, the demand for official statistical information rose to unprecedented levels (Anderson 1988). For a variety of reasons, especially perhaps the increased specialization of federal agencies and the lack of a unified effort among agencies on behalf of geography, the nation would never again see census atlases like those initiated by Walker. The Department of Agriculture has long taken advantage of the quinquennial census of agriculture conducted by the Census Bureau to produce atlases that are full of graphic and cartographic images (Baker 1931, U.S. Bureau of the Census 1995), and the Census Bureau itself produced a portfolio of thematic maps in conjunction with the 1950 census (Klove 1953), but never since then has it prepared a comprehensive national or statistical atlas. In fact, it took the federal government almost a century following the publication of Walker's atlas to assemble once again a comprehensive set of graphic and cartographic images to produce the U.S. Geological Survey's National Atlas of the United States of America (Gerlach 1970).