The primary objective of this paper is to contextualize GIS technology onto five broad intellectual/social dimensions in an attempt to develop a critical theory of geographic information science. This position paper contains three sections. The first section outlines major elements of a critical theory of geographic information science. This section also serves as a summary of my research interests regarding the topic of GIS and society. The second section of this paper presents a case study on GIS-based environmental equity analysis using the critical framework presented in section one. The third section summarizes my future research plan on the topic of GIS and society.

1. Contextualizing GIS Technology: Toward a critical theory of geographic information science

Current conceptualization of GIS technology is instrumental in nature without closely scrutinizing the social constitutedness of this electronic tool. By synthesizing recent literature on GIS and society, this paper aims to develop a critical theory of GIS technology that contextualizes GIS at the following five different, but interrelated, dimensions: ontology, epistemology, methodology, ethics, and politics.

a. GIS Ontology

Ontology refers to the theory of existence, which asks the fundamental question what exists. The answer to this question will determine what are accepted as "facts" and what can be known. In the context of GIS, ontological issues refer to what GIS researchers believe exist and how to represent this existence inside a digital computer. In technical terms, the ontology of GIS lies in the domain of data modeling. According to Goodchild (1991), the data modeling process is to study "how the infinite complexity of the geographical world can be represented within a discrete finite machine." From an ontological point of view, the current generation of GISers believes that reality can be represented in terms of location (where), attribute (what), spatial relationship (how), and time (when). Everything has been discretized. In terms of space and spatial relationships, GIS is deeply embedded in Euclidean geometry, in which the real world has been abstracted into points (0-dimension), lines (1-dimension), polygons (2-dimension), and volumes(3-dimension). Attributes are grouped and labeled according to different classifications and categorizations. Spatial relationships are represented using rigid physical topological relationships in terms proximity and adjacency. Time is also treated as a one-dimensional discretized element. Overall, current GIS ontology is embedded in a Newtonian conceptualization of space and time. Social relationships are not adequately represented. Among all possible worlds, GIS ontology is only one way of representing the world. Whether it is the best way or not needs further scrutiny. From a critical theoretical point view, we need to ask, can the real world be known through the points, lines, and polygons as currently represented inside GIS? To what extent has such an ontological assumption distorted the reality and affected our way of knowing? What are the social and theoretical implications of such a GIS ontology? Can social relations be incorporated as an integral part of GIS data structure?

b. GIS Epistemology

Epistemology refers to the theory of knowledge. It asks the fundamental question how can we know it or how do we know what we know. The answer to this question will determine what kind of logic researchers will use to derive knowledge. From an epistemological point of view, how space/place is defined inside a GIS affects not only what we can know, but how we know it. The fact that GIS is essentially a computer-based technology has implicitly determined the Boolean logical nature of GIS applications. Two main epistemological positions dominate the current GIS reasoning process: visualization-based empiricism (VBE) and analysis-based rationalism (ABR). The first position sees knowledge as the product of sensory perception, resulting from a kind of mapping or reflection of external objects, through our sensory organs, possibly aided by different observation instruments, to our brain/mind. Knowledge has to be developed by observation. The second position sees knowledge as the product of rational reflection through the integration of GIS with spatial analysis and modeling. The wide proliferation and adoption of GIS techniques have restricted our pursuit of knowledge to Boolean logical reasoning according to either the visualization-based empiricism or the analysis-based rationalism. Has GIS technology inadvertently marginalized other insightful epistemologies?

c. GIS Methodology (Praxis)

The combination of ontology and epistemology defines methodology, which refers to a set of rules and procedures regarding how information is collected and how analysis is conducted. From a methodological point of view, GIS is simply a transformation of the assumed ontology and epistemology into a set of operational rules. The ultimate frame of reference for any truly meaningful application is hidden beneath several thick layers of technical operations. With the increasing level of automation, geographic analyses are being further functionalized and rountinized into various overlay and buffer functions with fewer and fewer users caring about the ontological and epistemological assumptions behind these operations. The question of why to perform a particular operation is rarely asked. Because of the exclusive technical nature of GIS, mere technical advances in the absence of thoughtful and sound research design do not necessarily advance our understanding of physical and social systems. Current GIS applications are hampered by the tunnel-vision of positivism. Sound methodology in GIS application does not necessarily mean how many analytical modules have been incorporated. Instead, we need systems with a greater theoretical breadth. If GIS methodologies are not contextualized to ontological and epistemological dimensions, mere technical advances primarily based upon the rehabilitation of quantitative, technologically structured tools of the past, may only postpone the real issue under investigation. In many real world GIS applications, the seemingly scientific, hi-tech-based GIS methodology is primarily ad hoc and full of subjectivities. But these subjective and problematic aspects in GIS methodology are rarely revealed to the decision makers and the general public. Only through the lens of a critical theory of GIS can we see clearly the methodological problems existing in most GIS applications.

d. GIS Ethics

Ethics refers to the theory of values that deal with human conduct or practice with respect to the rightness or wrongness of certain actions and to the goodness or badness of the motives and ends of such actions. The ethics of GIS define the rules of conduct commonly recognized and practiced in GIS applications. Scholars have increasingly recognized that the new technology's primary goal is to increase the surveillance capabilities of the academy, the state, and the capital. Does GIS technology foster democratic practice, broaden the distribution of, and access to, information, and reduce the burden of work on those who adopt its rigors and accept its benefits, as claimed by GIS enthusiasts? Or, as some critics have argued, that the ethical inconsistency is inevitable because of the utilitarian concept. I believe that the recent debate on the various ethical issues should be an integral part of the critical theory of GIS technology.

e. GIS Politics

GIS politics should explore the institutional --executive, legislative, and judicial-- infrastructure for the adoption and application of GIS technology in the new electronic democracy. In a technologically institutionalized society, it remains controversial regarding who has the political power and clout to have a final say in various decision making processes. How has the proliferation of GIS data bases and differential access to spatial data bases influenced the abilities of different social groups to utilize information for their own empowerment? GIS politics should also explore the possibilities and limitations of using GIS as a participatory conflict resolution tool. To what extent, should GIS be utilized in various executive, legislative, and judicial processes in a democratic society? The legal and regulatory frameworks in GIS research and applications remains to be determined.

2. A Case Study of GIS-Based Environmental Equity Analysis

In this section, I want to use the critical theory outlined above to contextualize the application of GIS technology in environmental equity analysis. In doing so, I want to illustrate the profound ramifications of GIS technology on the social problems. The issue of environmental equity -- whether minorities and low income communities across the United States share a disproportionate burden of environmental hazards -- has attracted intensive interdisciplinary research efforts in recent years. Because of the increasing availability and easy access to several national spatial databases, such as U.S. EPA's toxic release inventory (TRI) and Census Bureau's TIGER files, GIS technology has been widely used in environmental equity analysis during the past five years. However, numerous critically important issues with profound social and legal implications have not been examined in GIS-based environmental equity analysis. The purpose of this empirical study is to contextualize these issues using the critical theory of GIS.

The following preliminary conclusions have been reached:

a). At the ontological level: what is being represented in a GIS, including both geographical boundaries (census block, census tract, zip code area, county boundaries) and attribute data (such as census data on race/ ethnicity and socio-economic status, EPA's TRI data), may not be appropriate to conduct environmental equity analysis. The results of environmental equity analysis are biased toward how the data were collected.

b). At the epistemological level: Risks are socially amplified by various social and ethnic groups. The perception of risk cannot be unproblematically modeled. Also, certain aspects of risk assessment and evaluation cannot be reduced to computational details. In order to use GIS, everything must be converted into numbers, which may mislead rather than reveal what is really going on.

c). At the methodological level: GIS-based environmental equity analysis heavily relies on mapping and statistical techniques. Depending on how the maps are designed, GIS can be used more easily to lie with maps. As for the statistical techniques, depending on the geographical scale and the aggregation method used, we can basically come up with whatever desired results we want. Or in other words, the stubborn modifiable areal unit (MAUP) problem still persists, and because of ignorance of this problem, many authors have committed a cardinal sin in statistical analysis -- ecological fallacy.

d). At the ethical level: Uncertainties in GIS-based environmental equity analysis have posed many ethical issues in scientific and policy-oriented research. It has become more and more difficult to profess to be impartial scientists. Instead, many researchers have become hired guns, producing research results that are dictated more by the funding agencies than by what is really going on in the real world. GIS and other technologies have contributed to, if not caused, this ethical dilemma.

e). At the political level: Environmental equity analysis is actually a struggle among grass-roots community organizations, corporate America, and various state apparatus in executive, legislative, and judicial branches. The controversies are a reflection of far more broader issues related to contemporary politics. One very important issue is whether GIS can be used for the empowerment of disadvantaged groups to ensure an equitable distribution of environmental hazards across class and race lines. Has GIS promoted a technocratic geography serving the interests of existing power structures? Right now, as GIS technology currently stands, who can access this technology and enjoy the benefits of it will depend on who is willing to pay, not necessarily on who really needs it.

3. Future research agenda

My long term goal is to establish a comprehensive theory of geographical information science which addresses spatial data handling issues at the ontological, epistemological, methodological, ethical, and political levels. My short term goal is three-fold. First, I want to explore the alternative ways of representing and knowing reality in a GIS, especially how to incorporate the non-computable aspect of people, space, and environment into the GIS modeling and reasoning processes. This belongs to the ontological and epistemological levels of the critical GIS theory presented in section one. Second, I will continue to explore the perplexing modifiable areal unit problem (MAUP) and ecological fallacy issue in GIS-related analysis and modeling process. In this regard, I am interested in implementing Tobler's scale-independent or frame-independent analysis and modeling procedures in a GIS. I also want to explore here the hierarchical theory of GIS applications using multiple-scale and multiple-zoning schemes. Third, I believe that the best way to publicize research results regarding the topic of GIS and society is to develop a new curriculum on the social implications of GIS technology in order to bring these critical perspectives into the classroom to educate our students as well as the general public. Current GIS education is dictated by the twin goals of teaching ABOUT (GIS technology itself) and teaching WITH (GIS for problem solving). I suggest that a third dimension must be included as well:, that is, teaching AGAINST (a broader social/intellectual context) for which the critical theory of GIS technology will play a very important role.

Return to