To survey the history of surveying practices is to explore the multifarious quest—across time and space—to grasp a sense of the whole through the aggregation of its parts. To survey is to gain a general or comprehensive view of the landscape, whether physical or social. In late Middle English, the verb meant to “examine and ascertain the condition of.” The noun form connoted oversight or supervision. Indeed, the word survey still carries three distinct meanings, the first being to measure or count. The two other definitions point in opposite directions. On the one hand, surveying means to oversee, or examine closely. On the other, it refers to seeing over in order to gain a broad perspective.
The thing being examined or ascertained is itself less significant. The subject of a survey, especially in early usage, was often a plot of land, a coastline, or a set of topographical features. Land surveying, among the oldest of the mathematical arts, dates to ancient times. Exploration and mapping, transportation, and communications all depended on surveyors’ tools for assessing the distances, angles, and features of specific terrain. A survey might settle a dispute between contending owners of a field, for example, or the proper boundaries of a village, painstakingly marked out with a chain and a compass. On a broader—indeed, global—scale, surveying was a crucial technology of empire, enabling colonizers to map and engineer, and thereby attempt to master, unfamiliar, far-off places.
Increasingly, surveys were pressed into the service of gathering information about the human inhabitants of those places as well. A surveyor might simply want to tally the people in a given territory. In the West the population registers we know as censuses are as old as the biblical account of Mary and Joseph’s journey to Bethlehem, the purpose of which—before miraculous events interceded—was to be included in such an enumeration. As surveying techniques were developed and refined over time, however, the object was just as often people’s characteristics, behaviors, or attitudes. Sometimes these were gleaned by observation, but more often by asking human subjects directly via interviews or questionnaires.
Social statistics are close companions of the survey, emerging originally as a “science of state.” Systematic attempts to catalog the entirety of a realm can be found stretching back at least as far as William the Conqueror’s Domesday Book of 1086, which cataloged both property and people. In that case a “Great Survey” was conducted by men who fanned out across England and Wales in order to discover “how many hundreds of hides were in the shire, what land the king himself had, and what stock upon the land; or, what dues he ought to have by the year from the shire.” The Florentine Catasto of 1427, an extensive tax survey of the citizens of that city, amassed household-level data on debts, property, business interests, and family background.
Military service or taxation motivated these early efforts. But the survey’s uses proliferated alongside the functions of governance. The seventeenth and eighteenth centuries saw the systematic collection of birth and death rates, or “vital records,” in order to track epidemics and devise insurance tables. The age of revolution introduced another purpose for such tabulations: ensuring accurate political representation. In the United States, the federal census established in the 1780s was coincident with the nation itself, the empirical undergirding of the principle of representative government. Sir John Sinclair’s ambitious Statistical Account of Scotland, published between 1791 and 1799 and inspired by German state surveys, aimed rather to measure his nation’s quantum of happiness.
The social survey, prominent by the later nineteenth century, was the nonstate counterpart of official population counts. Western nations in the nineteenth century saw a wave of surveying by private citizens and philanthropists, triggering what Ian Hacking calls an “avalanche of numbers” in the service of industrial and social reform. These investigations sought a comprehensive view of modern conditions through local studies and firsthand data collection. They were undertaken by budding statistical societies and individual reformers certain that systematic social knowledge was the key to solving problems ranging from urban poverty, crime, and disease to cultural assimilation and unsafe factory work.
Charles Booth’s monumental self-funded survey Life and Labour of the People in London, its findings released in seventeen volumes between 1889 and 1903, was one of the most comprehensive. Predating Booth’s survey was Frederic Le Play’s six-volume Les Ouvriers Européens (The workers of Europe, 1855), which relied on interviews as well as observation to offer a representative portrait of French society. German mass polls in the 1870s as well as detailed neighborhood studies by American surveyors—Jane Addams’s block-by-block study of Chicago’s immigrant communities in her settlement house work at Hull House and W.E.B. Du Bois’s The Philadelphia Negro (1899), for instance—speak to the range of nineteenth-century information-collecting efforts transpiring under the sign of the social survey.
In the twentieth century, this species of social investigation, along with its ameliorist and amateur bent, would be supplanted by scientific or sample surveys and academically accredited institutions of survey research. Whereas earlier social surveys typically had been conducted by those outside the academy, the modern survey was the province of professional social scientists in the newly distinct fields of sociology, demography, psychology, economics, and political science. And where prior efforts had necessarily been fairly localized, the modern survey often aimed to speak to national trends or aggregate behavior.
Twentieth-century survey researchers would narrow the definition of what counted as a scientifically legitimate survey even as they sought to expand those instruments’ authority in public life. Touted as statistically representative, rigorously quantitative, and resolutely empiricist rather than reformist, sample surveys became in Jean Converse’s words “an instrument of special power for viewing mass populations in industrial societies, especially in their character as social facts, political publics, and economic markets.”
Newly precise methods for measuring variability and assessing the reliability of responses enabled the rise of modern survey research. In particular, the application of probability theory in the form of scientific sampling and data-weighting techniques allowed for more powerful extrapolations from parts to wholes. Keen interest in attitude measurement in the early 1920s by sociologists and psychologists as well as government agencies and commercial interests led to an intense focus on questionnaire and interview design. Whole new technical fields sprouted around constructing attitude scales and assessing respondents’ candor.
The advance of professionally trained social scientists into authoritative positions in the academy and government shuttled these techniques into the realm of policy. Scientific surveys responded to political and bureaucratic demands for what James Beniger termed “technologies of mass feedback” that might elucidate phenomena such as soldiers’ morale and consumer behavior. Seemingly neutral, nonpolitical instruments for decision making, statistics and surveys became useful tools for all manner of agencies, both private and public.
Ethnographic and narrative surveys of social life would persist. One key instance was the Mass Observation project in Britain, created in 1937 as an “anthropology of ourselves” on topics ranging from wartime clothes rationing to pub socializing. But quantitative sample surveys would overtake the field in the mid-twentieth century.
Major academic and government survey centers flourished, especially in the United States, where figures like Paul Lazarsfeld and Rensis Likert built major survey outfits before and after World War II. The Survey Research Center and Institute of Social Research at the University of Michigan were exemplary, spurring innovation in survey techniques and housing longitudinal data such as the National Election Study and the Survey of Consumer Attitudes. Likewise, the National Opinion Research Center at the University of Chicago has since 1972 through the General Social Survey monitored both the demographic characteristics and the political and social attitudes of US residents.
Universities, foundations, and states all invested heavily in survey technologies in the latter half of the twentieth century. Indeed, survey-based social science touched almost every area of inquiry and a remarkable array of social questions, from race relations to mental health, consumer confidence to sexual behavior. It transformed commercial fields like marketing, advertising, and journalism as well. Opinion polling, and electoral polls in particular, became a major sector of the survey enterprise in the United States, playing an increasingly important—some would argue outsized—role in political campaigning and strategy. Although initially resisted by other governments, polling would become an essential political technology in most industrial nations.
Employed to assay everything from coastlines to candidates, the survey has been both a flexible and a durable tool. There are discernable turns in its long history from antiquity to the present: surveys shifted in focus from physical terrain to human beings, passed from the hands of rulers to survey research centers, and made growing claims to scientific rigor. But the surveyor’s promise to offer a better, more comprehensive view of whatever she or he beholds has remained constant.
Sarah E. Igo
See also data; databases; governance; maps; observing; quantification