Space Skepticism, by Luc Anselin

When asked to write this column, the topic I suggested was “space skepticism.” Space skepticism is the attitude/position that an explicit spatial perspective does not contribute in a meaningful way to the knowledge discovery process. In this view, (still) shared by many in academia, GIS and spatial technologies may be useful tools, but spatial thinking is not considered to be fundamental to the scientific process itself. Of course, this perspective is not shared by the UCGIS membership, but then among ourselves, we tend to preach to the choir. So, why do I set a negative tone in this first column in the series?

I hasten to add that I do not want to draw too dark a picture. Important progress has been made in terms of the adoption of a spatial perspective in mainstream science, policy and curriculum over the past two decades. This coincides with the establishment of UCGIS a little over 20 years ago. We have indeed come a long way. My dissertation advisor at Cornell University (the late Walter Isard) initially characterized my topic (spatial econometrics) as a “red herring” (he later came around). And one of my first submissions to an economics journal got the response from the (single) referee “what is this spatial autocorrelation?” Today, GIS(T) is increasingly common in the curriculum from high school to vocational and traditional higher education. Spatial questions are addressed in the mainstream science and social science journals, and both spatial statistics and spatial econometrics have become accepted subfields in the discipline.

However, all too often, GIS(T) is still considered mostly as a tool to create great visualizations, the so-called “GIS maps” as they are typically referred to. What is lacking is a true “spatial thinking” perspective, that not only uses the GIS to collect the data and present the results, but makes the spatial perspective an integral part of knowledge discovery.

This issue has become more relevant in the current setting of “big data,” or, as I prefer to refer to it, “new” data, and the advent of data science as a new paradigm. The new data that have become massively available from various sensors, social media, open data portals, etc. fortunately also tend to be geo-tagged and time stamped. This opens up tremendous potential for new analyses of all kinds of interesting space-time processes and dynamics. However, in mainstream data science applied to these big data problems, an explicit spatial perspective remains largely absent (with some notable exceptions devoted to “spatial” data mining). Even though location may be taken into account at some stage of the analysis, by the time it gets embedded into the neural network layers of “deep learning,” the spatial aspects are gone. Space comes back at the other end, when the results are visualized in the form of nice maps. Why is the spatial perspective not more prominent in this process?

In part, I feel this may be because our field (GIScience, morphing into spatial data science) has not done enough to make successful use cases widely known. I recall some of the early arguments for the use of GIS demonstrating undisputable business advantages. Such examples, where there is unequivocal evidence that the incorporation of an explicit spatial perspective leads to better solutions do not have the visibility they deserve. Consequently, it is my experience that many in the big data/machine learning world remain unconvinced, i.e., they are space skeptics. I see an important role for UCGIS in this respect, to function as an “advertising platform” for successful use cases of spatial applications in the current big data environment. Such examples should go beyond GIS data integration and manipulation that has been scaled up to HPC platforms, but truly reflect the exploitation of spatial thinking throughout the whole enterprise.

This brings me to a second and related point, the need to introduce spatial thinking in the curriculum at all levels. About ten years ago, an important NRC report made the case for the integration of spatial thinking (mostly in the form of GIS) in K-12 instructional materials (NRC 2006). More recently, there was a delightful booklet entitled “The people’s guide to spatial thinking,” rephrasing the same theme at a more accessible level, published by NCGE (Sinton et al. 2013). However, in practice, much remains to be done in this respect. Again, I see an important role for UCGIS to promote spatial thinking at all levels, and to move beyond its focus on GIS(T).

In my view, we need to begin at the base, not only in K-12, but also in the general education and liberal arts curricula of higher education. Here, UCGIS can play an important role in facilitating curriculum development, along the same lines as it did with the GIS & T Body of Knowledge. Anecdotally, in my new position at the University of Chicago, I have been charged with developing a spatial thinking track for its revered “core” (University of Chicago’s take on general education requirements). Even though Chicago was a beacon for quantitative geography (and the beginnings of GIS and spatial analysis) in the 1960s and early 70s with Brian Berry, geography and spatial thinking have been remarkably absent from the core. Exposing students early to spatial thinking as part of their general introduction to scientific discovery may go some way to keep them out of the camp of space skeptics!

National Research Council (2006). Learning to Think Spatially: GIS as a Support System in the K-12 Curriculum. Washington, DC: National Academies Press.
Sinton, D., Bednarz, S., Gersmehl, P., Kolvoord, R., and Uttal, D. (2013). The People’s Guide to Spatial Thinking. Washington, DC: National Council for Geographic Education.

Luc Anselin at University of Chicago   Luc Anselin is the Stein-Freiler Distinguished Service Professor of Sociology and the College Director, Center for Spatial Data  Science, at the University of Chicago. Previously he was a Regents' Professor and held the Walter Isard Chair at the Arizona  State University, where he was founding Director of the School of Geographical Sciences and Urban Planning. Anselin is the  developer of the SpaceStat and GeoDa software packages for spatial data analysis. He was elected Fellow of the Regional  Science Association International in 2004 and was awarded their Walter Isard Prize in 2005 and William Alonso Memorial Prize  in 2006. He was elected to the National Academy of Sciences in 2008 and the American Academy of Arts and Sciences in 2011.  In 2013 he received the UCGIS Research Award and was awarded UCGIS Fellows status in 2012. 


Each quarter, UCGIS invites one of its Fellows to share his or her opinions and ideas on the status of scholarship, education, outreach, and impact of GIScience. Please read, enjoy, and comment on this article and return to our website for future issues.

The views in this article are those of the author alone and do not necessarily represent the viewpoints of or positions by UCGIS. 


Share this post:

Comments on "Space Skepticism, by Luc Anselin"

Comments 0-5 of 0

Please login to comment