Early Development

The 1960s' lasting reputation as a decade of rapid and revolutionary change is mirrored in many fields of intellectual and creative endeavor in which practitioners seized the opportunity to challenge doctrinal orthodoxy and question established practices. Geography was no exception. Restlessness with the prevailing idea of geography as a compendium of factual material about regions of the world combined with an intellectual curiosity that drew strength from a sudden and decisive weakening of disciplinary boundaries contributed to the advent of the theoretical quantitative revolution and the rise of spatial science. These developments indelibly reshaped the discipline, influencing what geographers did and the modes of analysis that they adopted. Yet almost as soon as the new agenda of quantification and model building dawned, a significant groundswell gathered pace that had different preoccupations and priorities. Geographers following these new directions also gained inspiration from multidisciplinary convergence, but while quantitative geography looked to mathematics, systems theory, and positive economics, other nascent areas of human geographical inquiry turned instead to the social sciences, the humanities, the arts, planning, architecture, and design.

Behavioral geography was a case in point. Whether viewed in narrower or broader senses, it stemmed from recognizable, if isolated, antecedents from the first half of the twentieth century. Early interest focused on maps and orientation. In 1907, F. P. Gulliver read a paper to the Association of American Geographers on the way that schoolchildren perceived issues involving directionfinding in their use of maps, a theme furthered by the physicist C.C. Trowbridge in 1913 in an essay on maps and imagination. They found that orientation was more often achieved by an egocentric reference system than by points of the compass, with Trowbridge recognizing the significance of the home as an anchoring point in these reference systems.

During the interwar years, the writings of American geographer Carl Sauer and the work associated with the Geography Department at the University of California – Berkeley (which Sauer headed) helped to counter prevailing theories of environmental determinism by stressing the way in which human beings, through the agency of culture, also molded their environment. His compatriot John Kirtland Wright emphasized the importance of the imagination and intellectual curiosity in shaping the environment and suggested that geographers might concern themselves with geosophy – the study of geographical knowledge from any or all points of view. In the Francophone world, Georges Hardy's La Geographie Psychologique (published in 1939) drew attention to the significance of human psychology when considering the relationship between people and their environment.

The 1950s saw further contributions of considerable importance that, while also initially isolated initiatives, exerted a direct influence on later research. In 1952, the British historical geographer William Kirk produced a model based on Gestalt theory to argue that people's behavior was rooted in a behavioral environment (the world as perceived) rather than the objective environment (the world of 'facts'). In 1956, the British born social scientist Kenneth Boulding produced a monograph entitled The Image, which suggested that human knowledge was centered around 'images' – organized impressions of the world that people develop through their experience and which act as the basis for their behavior. In itself, this work might have remained an obscure discussion piece if the architect Kevin Lynch had not taken up its central idea in his book The Image of the City. Lynch explored the visual quality of three US cities (Los Angeles, Boston, and Jersey City) by reference to the way that the physical features of their cityscapes were perceived by small samples of their residents. He showed that people understood their surroundings in consistent ways, by forming 'mental maps' that had five recurring features (paths, edges, districts, nodes, and landmarks). The extraordinary popularity of Lynch's book helped to establish Boulding's concept as a key part of the frameworks employed in early behavioral geographical research (see Figure 1), even though the concept bore no direct relationship to the notion of imagery used by psychologists.

A simplified model of human environment relations

In their different ways, these works in effect foreshadowed the humanistic and cognitive science streams that fueled the growth of behavioral geography in the 1960s. The former gained impetus from David Lowenthal's paper 'Geography, experience and imagination', published in 1961, which drew on J. K. Wright's agenda to recognize that people live in parochial worlds, shaped by selective perceptual filtering of environmental information but bound into the common realm of experience through culture. For his part, the cultural geographer Yi Fu Tuan moved beyond his early geomorphological interests in the landforms of New Mexico to reflect on the emotional connections between human beings and the physical environments that they encountered. These reflections appeared first in articles published in Landscape, the journal founded and edited by the writer J. B. Jackson, then in a series of monographs that explored the relationship between cosmology and feelings toward landscape, and finally in his 1974 book Topophilia. At the end of the 1960s, Anne Buttimer supplied an important introduction to non Anglophone sources on human–environment relations through her discussion of French theories of social space and their implication for the subjective configuration of urban areas. She cautiously applied these perspectives in a subsequent study of a residential area in Glasgow (Scotland).

For their part, those who founded the 'cognitive science' stream were initially interested in issues concerning the knowledge held by decision makers. Peter Gould's use of game theory in 1963 to simulate the probable planting strategies of African farmers in the face of the uncertainties of their environment indicated a wish to relax the typical assumptions made by spatial analysts about perfect knowledge and economic rationality. In 1964, Julian Wolpert used a sample of Swedish farmers to test whether real world decision making accorded to satisficing (guided by the goal of seeking a sufficient output) or optimizing principles. Given that the results showed that crop yields were well below the theoretical optimum – in certain areas as low as 40% of the possible yield – Wolpert concluded that the satisficing was more descriptively accurate of the behavioral pattern of the sample population. Nonetheless, there was no necessary implication of irrationality in these findings. While they might signify lack of knowledge about potential yields and the alternatives that were available, they might also reflect perception of risk, aversion to uncertainty, and culturally transmitted values that placed other benefits above profit maximization. Similarly, when analyzing industrial location decision making, Allan Pred developed a powerful critique of normative economic models, pointing to their illogicality (e.g., it is impossible for competing decision makers to arrive at optimal locational decisions simultaneously) and the failure to consider motivation (in that decision makers might well be satisfied with a location that is satisfactory in the long run rather than optimal in the short run). As an improvement, Pred proposed a 'behavioral matrix', which included the quantity and quality of information on one axis with the ability to use that information on the other, pointing to the unlikelihood of decision makers managing to achieve the optimum on both counts.

These types of studies represented a significant first step. By pointing to likely variations in behavior instead of being able to rely on assumed optimisation, their authors stressed the need to include 'black box' variables within the analysis – in other words, incorporate those nonobservable elements that could only be captured by surrogate means. It also heralded the future close alignment between behavioral geography and psychology. In a published version of a talk given to the Michigan InterUniversity Community of Mathematical Geographers in 1965, Peter Gould pointed to the need to know about the 'maps' that people have in their heads with regard to the world around them. He accepted that individual cognitions were unique but felt that general statements might be made with a technique that could externalize personal preferences and the values that underpinned them. He illustrated this point by reference to studies of residential desirability in the USA, Sweden, the Federal Republic of Germany, Ghana, and Nigeria, whereby student respondents were asked to rank regions of their home country on the basis of lists of administrative areas (such as states or counties). By aggregating the scores for individual respondents, it was possible to map residential preferences for the national groups of students and make comparisons between students living in different nations.

Other researchers looked to mainstream psychological methodology to investigate the cognitive domain. In 1966, for example, Thomas Saarinen published an account of his doctoral research on the behavior of farmers in the drought prone Great Plains region of the USA. Using personality measures such as the Thematic Apperception Test as part of his methodology, he found from their responses that successes were recalled more frequently than failures. Farmers consistently underestimated the frequency of years when drought was experienced and were optimistic about the number of good years and about the size of crops in such years. Perceptions of the drought hazard were also influenced by age and previous experience, especially whether farmers remembered the drought conditions of the 'dusty thirties' (the years of the formation of the Dust Bowl).

Collectively, these early studies indicated that the emerging field of research served as both a corrective and a catalyst. As the former, it countered the dominant thrust of spatial science by embracing such considerations as learning, imperfect information, and bounded rationality in order to improve the diagnostic and predictive power of locational models. As a catalyst, behavioral geography quickly moved beyond being an adjunct to spatial scientific inquiry to become a channel for work that expanded research horizons and reconceptualized the relationships between people and their environments in quite different ways from those conventional in geographical discourse. Benefiting from the breadth of perspectives available from its two component intellectual streams, behavioral geography became a wide ranging field that revealed a notable flowering of creativity, with almost any scale of cognitivebehavioral problem, from the interior spaces of buildings to understandings of the cosmos, becoming of interest to geographers.