Is GIS a Science?

This post is the first of several which responds to prompts posted on the website for GEOG 0323: Open GIScience. The prompt for this post is:

• `What is science and how does a lack of reproducibility impact scientific knowledge production? Do you have any prior experience or knowledge of irreproducibility or questionable research practices in science?
• Which category of “GIS as Science” most applies to your personal experience thus far using and studying GIS? Do those forms of GIS count as “science”?
• Which themes of GIS Discourse have attracted you to the technology or made you feel uneasy about it? Have your instructors or mentors made any arguments similar to the four themes, or have you made them yourself in conversations with friends and family?`

Science is a process in which researchers consider questions about a topic and develop or employ means to gather answers to these questions. The researchers reflect on their conclusions and may place their results in a broader context, in so doing involving the results of previous researchers (NASEM 2019). According to the NASEM article, the ability to replicate studies in a variety of different contexts is important in that it serves to vet the results of a groundbreaking study (this is provided by the example of supposed cold fusion on page 23). If researchers find a study difficult to replicate, the scientific community may not be able to determine whether a study contained errors in its conclusions (which may be considered comparably “uncertain” compared to studies covering popular topics; NASEM 2019). In such a scenario, scientific knowledge would not increase as much as it could because a potential development is not able to be tested by others to vet its credibility.

In the course Biostatistics, the instructor devoted some time to discussing the falsification of data as a questionable research practice which can result in journals rescinding articles. This choice, which we discussed as possibly being driven by the pressure to publish and find statistically significant results, holds back scientific knowledge production in a clear manner; other researchers may have cited said article, and their conclusions may now be based on falsified results. Additionally, in a winter term course called Reproducible Biology in R, students searched for datasets from journal articles on the website DataDryad and attempted to recreate graphs in R from the articles using the data. I recall this activity as being sometimes difficult with the data provided, which could make some graphs irreproducible.

Based on the article by Wright et al. (2019), I would say that the description “GIS is a Scientific Tool” most applies to my personal experience at Middlebury. In Mapping Global Environmental Change, it was presented as a type of software which could help answer questions in the field of geography. I do not recall it being presented as a separate discipline. I can see how the toolmaking position could fit with the course Conservation Planning, however, for students were sometimes tasked with judging how results changed when variables within a GIS method were altered. This approach seemed more like refining GIS and improving upon its outputs, which may fit more with the intermediate position even though students were not making code improvements to software. In general, I believe that GIS goes beyond a tool and should be represented with the toolmaking position, and I am open to learning more about conceptions of GIS as a science.

The GIS as discourse themes are introduced in St. Martin and Wing (2007). They are summarized as the idea that GIS is “a singular technology”, that the field is “progressing along a linear path”, “inherently expansive and growing”, and “universally applicable”. These will be referred to as the first, second, third, and fourth themes below.

For GIS discourse, I do not believe that my previous instructor made an argument similar to the “singular technology” theme (St. Martin and Wing 2007). In Conservation Planning, we switched from Google Earth Engine to WhiteboxTools due to the benefits of the latter for a particular unit. The second theme may have come up; I recall the instructor for Mapping Global Environmental Change stating that we may not be using Google Earth Engine in ten years but that GIS knowledge was still useful. The third theme may be somewhat implied in the requirement of Environmental Studies students to take either GEOG 120 or 150, for GIS may be construed to have “expanded” into that field. I have not seen GIS be cast in the critical terms which the authors use on page 240, however. The fourth theme may also fit with this requirement. Personally, I have been most attracted to the technology by this theme in particular. I am not a geography major but rather a Conservation Biology major, and I have seen that GIS is applicable to environmental work. In regards to being somewhat uneasy about any of the themes, in contrast, the power of corporate messaging in the authors’ description of the first theme is most relevant. I am concerned about not being familiar with the software products that are considered most relevant in the field.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2019. Reproducibility and Replicability in Science. Washington, D.C.: National Academies Press. doi.org/10.17226/25303

St. Martin, K., and J. Wing. 2007. The discourse and discipline of GIS. Cartographica 42 (3):235–248. doi.org/10.3138/carto.42.3.235-248

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.org/10.1111/0004-5608.872057/