Do all people, from all cultures and all languages, think about geographic space and geographic processes in more or less the same way? Or are there significant cross-cultural variations in how different peoples conceptualize and reason about geographic processes, features and places? Dr. David Mark of the State University of New York at Buffalo discusses the issues and their ramifications for nono-Eurocentric users of geospatial technologies.
Do all people, from all cultures and all languages, think about geographic space and geographic processes in more or less the same way? Or are there significant cross-cultural variations in how different peoples conceptualize and reason about geographic processes, features and places? Do geography and spatial relations parallel the infamous case of the many “Eskimo [Inuit] words for snow,” the linguistics factoid that every cocktail party conversationalist thinks he or she knows? (The snow words situation turns out to be mostly a misinterpretation, but that’s another story!) Or is spatial cognition and related linguistic development governed by universal principles?
Coming closer to home, we can ask similar questions about GIS. Are GIS software products and Spatial Data Infrastructures (SDIs) the same for all, a universal foundation based only on the true nature of geographic phenomena, universal principles of computing and cognitive primitives? Or are the GIS packages and SDIs that we know today biased toward a European worldview and a so-called ‘Western’ scientific approach? If GIS is universally easy-to-use (or universally difficult, but GIS usability is another topic!), then that is good news both for humanitarians and for software vendors, and we can move forward with a one-size-fits-all solution to the world’s geospatial problems. But if GIS is biased toward the culture that produced it, then it could be yet another case of North Atlantic Imperialism, another brick in the darker side of globalization, posing an ethical dilemma, especially for those working in indigenous GIS or GIS and international development.
Mark Twain once wrote: “There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.” We have very few facts about cultural differences in spatial cognition. For almost 20 years, I have been intrigued by questions of possible cultural biases in GIS software, but until recently I had not found much in the way of solid answers. I now think I had been looking for cultural differences in spatial cognition in all the wrong places, in at least two different dimensions. I’ll explain, and then draw some implications for GIS.
In the 1990s, I wrote numerous papers on spatial cognition, GIScience and GIS. A frequent justification for spatial cognition research was that it relates to GIS usability and especially to GIS universality. Papers at Latinamericanist and Latin American GIS meetings speculated that cultural and linguistic differences between Spanish and English might lead to added barriers to easy GIS use by Spanish speakers. But testing human subjects on spatial relations and language in the United States, Spain and Costa Rica failed to reveal significant differences for line-region spatial relations. Was the concern unfounded?
On sabbatical in 2002, I was fortunate indeed to have the opportunity to visit Andrew Turk at Murdoch University in Perth, Australia. Turk, in turn, took me a thousand miles north to Roebourne, Australia, where about 500 Yindjibarndi speakers were keeping their language alive after being forcibly removed from their traditional lands almost a century ago. Through semi-structured interviews and other ethnographic methods, Turk and I learned details of the Yindjibarndi terms for features of the landscape, things that in English would be called hills and valleys, pools and cliffs, gullies and riverbeds.
While a dictionary published in the 1980s appeared to show term-by-term translations, we found that in most (if not all) cases, the terms and their definitions did not line up! There was a many-to-many relation between Yindjibarndi terms for elevated areas and their English equivalents. The relation between rivers and their beds was turned inside out in this tropical desert area. Rivers in English are fundamentally composed of water, even if some are sometimes dry; whereas a wundu in Yindjibarndi is a (dry) channel that on rare occasions might contain water. When water is present after the rare heavy rains following a tropical cyclone, the water is categorized by its intensity of flow and is always distinct from the permanent feature that might get named “river” in English. The conceptual systems that underlie the semantics of geographic expressions in the Yindjibarndi language do not seem strange, but they definitely are different. Working with Turk, David Stea, Carmelita Topaha, and many Navajo friends, we have found similar differences in Navajo language landscape categories in the arid highlands of northern Arizona and New Mexico.