Computers, Society, and Nature

Wright, Goodchild and Proctor review debates around the conceptualization of GIS as a tool or as a science on a prominent GIS listserv in 1993. In so doing, they unpack the debates and situate possible conceptualizations of GIS on a continuum between tool and science, rather than as a simple dualism.  The additional point on the continuum they introduce is the conceptualization of GIS as toolmaking, analogous to engineering being situated somewhere between the pure and applied sciences.

The article concludes that only the conceptualization of GIS as science is sufficient to be considered science, and thus the only way GIS in and of itself might attain academic credibility. However, it would seem as though the authors failed to anticipate the rise of critical GIS as a field of study.  When they touched on “the scope of research [being] determined not by the tool’s value to geographers, but rather by the multifarious applications of GIS, to include all of the societal effects of the computerization of geographic information” (p. 356), I thought immediately of issues such as VGI,  loaction-based services and geosurveillance—all burgeoning GIS research topics today.  However, Wright at al never come back to these more qualitative opportunities for ‘doing GIS’. Perhaps they overlook this potential because it doesn’t quite fit into their continuum: critical GIS involves GIS gaining academic credence not within science (whether pure or applied), but by way of the humanities and social sciences.

The article also offers an interesting insight into the pitfalls of conducting debates over the medium of the internet, from the perspective of the 1990s. While the GIS-L debates of 1993 do not remotely approach the caliber of trolling and talking past one another typical of political debates on the internet today, Wright et al’s analysis on page 348 does a good job of anticipating the rhetorical impacts of the internet’s shift into widespread use.

-FischbobGeo

Miller places dominate GIS practices within the realm of placed-based methods based off of a past where technology and theory began to intersect. In a finite world, limitations are place on an individual to appropriate how much time and resource is spent on an activity. Their decisions are ultimately based on their socio-economic condition. Because of this, a people-based GIS should complement placed-based GIS better understand urban systems. This was a great article that introduces the reader to time geography and activity theory. For me, it made me think about how a person is placed in a particular location because of their socio-economic condition, and now how a person is somewhere with x,y,z conditions (which was intuitive for me at first. Consider looking at a map, where a point represents a person and an attribute table holds their socio-economic profile. To me, the visualization on the screen associates a person to a particular location, and not attributes that bind them to that location.).

While reading the data collection techniques of STA and how new technologies can increasingly improve the cost, rigor, and time associated with data collection, a reoccurring thought was, how can this be applied to the marginalized population where access to mobile devices, computers or knowledge of the technology is limited? How can technology improve their situation in such a way researchers can easily collect and study their space-time activities to offer a policy prescription to remedy this? I think it’s key to recognize our own positionality when we consider how IT has “shrunk” the world for some in a positive way, empowering them to access more information, knowledge while decreasing travel time. On the flip side, it has shrunk the world for others in terms of limiting access and resources to improve their livelihoods.

-tranv

This article nicely lays out 3 distinct arguments where scholars have argued GIS to be a tool, toolmaker and science. GIS as a tool and toolmaker is easier to grasp, while GIScience was much harder for me to wrap my mind around. At the beginning of this paper, my crude understanding of GIS as a science is if some form of rigorous scientific method can be applied to it (form a hypothesis, test a hypothesis, evaluate the results to improve our knowledge about the subject matter) then it may be considered a science. Here, the definition of science also becomes important where some have a vested interest to label GIS as a science to legitimize their work in an academic setting whereas others believe that labeling GIS as a tool doesn’t give it the credit it deserves when it can enact change and increase understanding. By the end of the paper, geographic information science is describe to be “concerned with geographic concepts, primitive elements used to describe, analyze, model, reason about, and make decisions on phenomena distributed on the surface of the earth” (Wright et al., 1997, 357). Does this mean that the “science” involved refers to the decisions we make in how we define categories, what we chose model (thus what we chose to omit), and theorize the particular subject matter?

My experience in GIS falls into the toolbox category – to visualize a certain phenomenon in space. In my opinion, depending on how far you go in your studies, or how you apply the software defines you view of GIS as a tool, science or somewhere in-between. For a technician mapping out underground pipelines, GIS is a tool, whereas a scholar studying the use of GIS by various social groups can be viewed as a science. No matter how GIS is viewed, few people will doubt the contribution it has made in our understanding of spatial phenomenon.

-tranv

In detecting outbreaks of infectious disease, the time model has very much preceded the space model. Most public health agencies rely very heavily on time series of traditional and maybe syndromic health data (laboratory confirmed cases of illnesses as well as secondary data such as emergency department visits), looking for atypical temporal patterns. But few integrate spatial information (even fewer beyond running SatScan), even though statisticians and geographers have shown through research and simulations that outbreaks could be detected with improved timeliness and sensitivity/specificity (fewer false positives or false negatives) if spatial information were included. Many of the concepts in this paper are very relevant to outbreak detection, such as idea of a space-time path in a “lifestyle.” Either way, incorporating both space and time into analysis/research seems to be a real challenge in computational power, in model complexity, and in visualization.  Even though this paper was written over ten years ago and made some very concrete suggestions for moving forward in time geography, I’m not sure how far we’ve come.

-Kathryn

I believe GIS is a field of study, similar to (and related to) statistics, and the authors draw this analogy several times in the paper as well. What often inspires the question “science vs. tool” is the lack of substantive application – the associated methods can be applied in a variety of disciplines. But like statistics, GIS has theoretical underpinnings that are understood by experts, and practical applications that can be used (and misused) by experts and nonexperts in various fields.

One interesting point was a comment by someone in the listserve who compared GIS to statistical software – that using the statistical software isn’t “doing science” either. This is a good example and I believe the author of this comment was actually making the opposite point he or she intended to. Using statistical software does not make you a statistician, and using GIS does not make you a scientist or a GIS expert.  But the respective experts (statisticians; GIS-focused geographers or computer scientists), while they may not themselves be the architects of this software, often use it, and have a strong understanding of the theoretical underpinnings it uses to perform analysis.

When someone is “doing GIS”, this doesn’t not automatically make their work scientific, but methodological advancements from experts in GIS are scientific, and I would say that experts applying the methods in novel applications is often scientific. Again, like statistics, sometimes it seems like GIS is used to legitimize research, and this may be a symptom of a societal obsession with ranking academic discipline worth by how “scientific” or “hard” it is. When a research paper overemphasizes the use of GIS, attempting to add legitimacy, it is similar to using an unnecessarily complex statistical model in order to make up for poor quality data or add a false sense of achievement to a work. I see that an issue in academia rather than an issue specific to GIS.

[Side note: interesting that when this paper was written (1997) the authors claimed that academics in general are not well set up to create reliable software with a couple exceptions. In 2013 I’m sure we can all think of many of examples that show this is no longer the case! The creators of R; Luc Anselin; Andrew Gelman…many more]

-Kathryn