Cultural Differences, Technological Imperialism and Indigenous GIS (fwd)

[phil cash cash]

Cultural Differences, Technological Imperialism and Indigenous GIS

By David Mark
(May 18, 2006)
http://www.directionsmag.com/article.php?article_id=2173&trv=1

Directions Magazine
The Worldwide Source for Geospatial Technology

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.

These tendencies do not appear to be unique to the languages we have
examined to date. Discussions with researchers at the Max Planck
Institute for Psycholinguistics (Nijmegen, The Netherlands) indicate
that they are finding similar results for a wide variety of languages,
cultures and environments: terminology for landforms, as well as
systems of place naming, vary considerably between peoples living in
different sorts of environments, in different cultural traditions and
lifestyles.

Interesting perhaps, but how does this relate to GIS? Spatial data
infrastructures encode entity types or feature codes in order to
enhance the semantics of geospatial data. But such codings might not
add value from an indigenous perspective, unless data are also encoded
according to the categorical systems of the indigenous people. We might
try to avoid categories altogether and just store entities and
attributes, and then try to infer categories later on. But theories
from cognitive science tell us that inferring categories from
observable properties will be difficult, if not impossible, to
implement.

All of the above might raise ivory tower curiosity and fuel esoteric
dissertation research. But these issues have important theoretical
implications for indigenous GIS and indigenous mapping. If there are
significant cultural differences in any aspect of spatial cognition or
language, surely this must in turn imply that current GIS is
Eurocentric. As it turns out, unmodified commercial GIS has proved to
be a valuable and at times powerful tool for indigenous people. Just as
maps were powerful tools of European imperialism in the 18th and 19th
Centuries, GIS returned the favor by supporting aboriginal land claims
with presentations in a language that dominant-culture courts could
appreciate. And many tribes in the western United States use unmodified
commercial GIS to manage tribal lands and maximize productivity of
forests and grazing lands, just as private land holders or government
agencies might do. But in most (if not all) of these, the GIS was
wielded by tribal members with university degrees or by consultants
from the dominant culture. It is not clear whether tribal GIS experts
can think in their own language and culture while using GIS, or whether
traditional and spiritual values can be incorporated directly into GIS
solutions.

Earlier, we found few differences in spatial conceptualization for
spatial relations, but more recently we believe we have solid evidence
that spatial feature categories can be quite different across
languages. We were also working mainly comparing English with Spanish
and French, other Indo-European languages. Moving to geographic entity
types rather than spatial relations, to non-Indo-European languages in
Australia and the western United States, and to arid landscapes rather
than humid, we still see broad cross-language similarities but also
many differences in the details.

Indigenous GIS faces all of the challenges of "Geographic Information
and Society," a research agenda within Geographic Information Science.
But indigenous GIS also faces additional challenges due to the
cognitive and linguistic factors mentioned above. GIS for community
empowerment requires qualitative methods and ease of use, whether in
Indian country, in low-income urban settings or in suburbia.
Socioeconomic and educational barriers to GIS use and the steep
"Digital Divide" may be greater than the cultural and linguistic issues
raised here. But if there is any truth in our recent findings,
indigenous GIS will never be on an equal footing with, or fully
interoperable with, the main stream of GIS until indigenous spatial
cognition and conceptualization are understood as well in the
indigenous cultures in question as they are for English and the
dominant culture of North America and Europe.



Reprinted with permission, Copyright 2006, Directions Media