CENTENN IAL FORUM GIScience, Geography, Form, and Process

T he map has evolved over the past few centuries as humanity’s primary method for storing and communicating knowledge of the Earth’s surface. Topographic maps portray the general form of the surface and its primary physical and cultural features; thematic maps show the variation of more specialized properties such as soil type or population density; and bathymetric maps and hydrographic charts show the characteristics of the sea floor. Maps serve as one of the most important repositories of both the raw data and the results of geographic inquiry, and mapmaking has always figured prominently in the skill set of geographers or their supporting staff. Maps are thus important and indispensable tools in the geographer’s search for understanding of how human and physical processes act and interact on the Earth’s surface: of how the world works. Geographic information systems (GIS) were devised in the 1960s as computer applications for handling large volumes of information obtained from maps and for performing operations that would be too tedious, expensive, or inaccurate to perform by hand. The Canada Geographic Information System, widely recognized as the first GIS, was built for the purpose of making vast numbers of calculations of area, reporting the results in tables. Over time, the range of functions performed by GIS has grown exponentially, and today it is reasonable to think of a GIS as able to perform virtually any conceivable operation on data obtained from maps (Longley et al. 2001). Geographers have adopted GIS enthusiastically, seeing it as a powerful device for storing, analyzing, and visualizing map information and thus as a much more effective substitute for the paper map (Goodchild 1988). Over the past decade numerous journals, conferences, academic positions, and programs have adopted titles that combine information with spatial or geographic and with science or theory. In what follows I will use the term geographic information science (GIScience) for simplicity and not enquire into the subtle differences between, for example, spatial and geographic information theory (Goodchild 2001). Geographers have been associated with many of these changes—and, in many cases, have been at the forefront—and many of the new programs and positions are found in departments of geography. But there has been relatively little general commentary on these trends, or on what they might mean for the discipline of geography as a whole. The first centennial of the Association of American Geographers is an appropriate occasion to reflect on the nature of GIScience and its relationship, if any, to the discipline of geography. I begin with a discussion of the nature of GIScience, of its relationship to GIS and of its links to the traditional sciences of geographic information. This leads to a discussion of whether GIScience is a natural science, concerned with discovering empirical principles and law-like statements about the world; or whether it is a design science, concerned with identifying practical principles for achieving human ends, or both. In the third major section I examine how GIScience is positioned with respect to the historic tension in geography between form and process and whether the growth of interest in GIScience has tended to favor form over process. The final section examines a future for GIScience that places greater emphasis on process and discusses the steps that will be needed to make such a future possible.