The Science of Doing GIS

Wright, Goodchild, and Proctor’s article explores whether GIS (as it was known in 1997) is a tool or a science. This is a question that isn’t easily answered, as evidenced by the article’s inability to answer it. However, the article does illustrate central themes of the argument via a synthesis of crowdsourced answers. This article rose a lot of questions for me and eventually answered them. That being said, I would rearrange the format– putting the why-do-we-care bit at the beginning, followed by the definition of science, and then jumping into the GIS-L discussion would provide greater perspective to the entire article.

Wright et al. distinguish three positions on GIS: GIS as a tool, GIS as toolmaking, and GIS as a science. Interestingly, throughout the article there are ideas implying that GIS can be all three simultaneously. One GIS-L participant noted, “the answer depends on who is involved,” (p. 350) and consequently what they are doing. It follows that most students regard GIS as a tool whereas professors see it as a science (let’s ignore the article’s GIS = science = academic validity = funding idea)– and I don’t think either are wrong. But while the article places GIS on a continuum, I’ll place it on a circle to bring tool and science closer together and attempt to crudely illustrate some kind of cycle.
  • Start with science: science talks about its use as a method for developing spatial theories and dealing with research issues.  Research issues are inherent in GIS: uncertainty, representation choice, analysis methodology, etc.
  • Then tool: These issues are in the software and are issues that users (should) recognize when using GIS as a tool for problem solving.
  • Then toolmaking: These are considered by developers and the like, and GIS undergoes toolmaking to make it easier to answer these issues. Toolmakers critically analyze and reflect; evaluating the tool on how well it does its job– which is, essentially, digitally implementing “all geographic concepts and procedures,” (p. 357) and how best to do this–bringing us back to the science and methodology.
The article concludes with, “GIS appears not to constrain its uses to any epistemological stance,” (p. 359) and essentially comes full-circle.

Wright, D. J., M. F. Goodchild, and J. D. Proctor. (1997). Demystifying the persistent abiguity of GIS as “tool” versus “science”. Annals of the Association of American Geographers, 87(2) pp. 346–362.

-sidewalkballet

Comments are closed.