Is GIS a Science?
Though science can be viewed as the correct way to identify patterns and facts, this narrow definition of science hinders a framework grounded in expanding knowledge from gaining greater understanding of the world. The terms science or scientific provides a level of “credibility” to a field or topic. A popular view of science aligns with positivist beliefs that all knowledge must be logically derived from observable experiences. This view does not recognize alternative approaches or traditions that provide accurate insights about the world. Observation and multigenerational knowledge can prove true even if not yet “verified” by the scientific process.
A broadening of science beyond positivist practices does not remove the importance of reproducibility and replication. Knowledge gained outside of a positivist approach can still be accurate; however, the ability for an experiment or process to be replicated and the same (or similar) results be achieved remains important for knowledge gathering outside of the standardized scietific method.
In the conversation and debates surrounding what can be classified as “science,” individuals and groups disagree on whether Geographic Information Systems or GIS can be included. Wright, Goodchild, and Proctor present three ways at understanding GIS’s place within the scientific discourse. They are GIS as a tool, GIS as toolmaking, or GIS as a science. Though I do not agree entirely with their categorizations, I align most closely with GIS as toolmaking. GIS can be utilized by researchers in many fields to approach disparate questions. In the repertoire of scientific tools, GIS can be viewed as filling an instrumental niche like data analysis programs or stand density equations in silviculture.
GIS as toolmaking includes the more complex ability to write and analyze the efficacy of tools within individual GIS programs. Studying users applications of GIS programs further develops GIS from solely a tool. GIS should be studied and rationalized in its applications. Without a firm understanding of why one is performing certain tasks, results lose their tangible meaning. It is important to understand that p-value of 0.05 denotes a statistical test measuring the chance of a difference between groups being more than 95% likely. When working with GIS or Remote Sensing tools, understanding a confusion matrix of an image classification differentiates the appropriate and inappropriate uses of a tool. In some ways, GIS is a tool that has a field of science around it. This field works to develop Geographic Information Systems (there are more than just ESRI products), test them, attempt to understand users applications, and measure their effects upon the greater scientific community and societal challenges.
Sources
Wright, D. J., M. F. Goodchild, and J. D. Proctor. 1997. GIS: Tool or science? Demystifying the persistent ambiguity of GIS as “tool” versus “science.” Annals of the Association of American Geographers 87 (2):346–362. DOI:10.1111/0004-5608.872057