|
Indigenous Peoples Specialty Group
2007 Annual Meeting, Association
of American Geographers
April 17-21 2007, San Francisco, CA
1334: Building Ontologies
for Indigenous Geographic Knowledge
Tuesday, 4/17/07, from 12:00 PM - 1:40 PM
Sponsorship(s):
Geographic Information Science and Systems Specialty Group
Indigenous Peoples Specialty Group
Qualitative Research Specialty Group
Organizer(s):
David M. Mark - University at Buffalo
Renee Sieber - McGill University
Chair(s):
Renee Sieber - McGill University |
Abstract Title:
Toward a Theoretical Basis for Ethnophysiography
Author(s):
David Stea, Professor Emeritus* - Center for Global Justice, Mexico
Abstract:
This paper is based upon ethnophysiographic research conducted over the past several years among cultures of both the Southwestern U.S.A. (Navajo) and Western Australia (Yindjibarndi) under a grant from the U.S. National Science Foundation. It is concerned with the basis for the relation of landforms to language, and is an attempt at, or a beginning of, formulating grounded theory for the results thus far obtained. Several dialectically related concepts are fundamental: "truth" vs. reality and "myth" vs. reality, for example - and the nature of "reality" itself. Conceptualizing these involves ethnoscience, and ethnography as an identifier of ethnoscientific concepts and principles.
In both Navajo and Yindjibarndi societies, landscapes are stories "animated" or "personified" as totem animals or monsters. Palmer (1977) talks about totemic forms, such as those associated with water, as water animals, land animals, and human spirits. These totems (water totems include warlu among the Yindjibarndi, "big snake" among the Navajo) may apply not just to the identification of landforms but to navigation among them (safe travel in marginal landscapes), and to explanations of floods and droughts,
Keywords:
ethnophysiography, landscape, language, culture |
| Abstract Title:
Using Qualitative and Mixed Methods to Represent Information About Indigenous Geographic Concepts
Author(s):
David M. Mark* - Geography, University at Buffalo
Jin-kyu Jung - Geography, University at Buffalo
Abstract:
The Yindjibarndi people of northwestern Australia and DinČ (Navajo) of New Mexico and Arizona live in similar semi-arid or desert environments, yet have very different languages and histories. How similar are the concepts that underlie the landscape terminology of these languages? Project researchers have conducted ethnographic field interviews, photo-response focus group, and ethnophotography protocols in the two areas. We also have quantitative GPS coordinates, and have used GPS and GIS data to integrate qualitative and quantitative information from a variety of sources. Particular research methods were developed and extended on each visit to the field, without much thought for how the observations might be analyzed. More than 40 hours of audio and many thousands of digital photographs presented a daunting interpretation challenge. It became clear that mixed methods, and especially Computer-Assisted Qualitative Data Analysis Software (CAQDAS) could help filter this information and see connections that might otherwise be missed. Using both ATLAS.ti and ArcGIS together to combine qualitative and quantitative data helps us to organize and analyze many varieties of data such as text, graphics, audio, and visual data files. The software can show the relationship among all of these data types. Although the main goals and paradigm of the research project are cognitive and scientific, mixed methods are providing a unique opportunity to link various types of qualitative information with geo-locational data. Results appear to provide insights into Indigenous geographic concepts that can help elucidate human relations to landscape and provide foundations to support Indigenous Geographic Knowledge Systems.
Keywords:
GIS, qualitative methods, mixed methods, indigenous knowledge
|
| Abstract Title:
Extraction, Characterization, and Classification of Topographic Eminences
Author(s):
Gaurav Sinha* - University at Buffalo, The State University of New York
David M. Mark, Professor - University at Buffalo, The State University of New York
Abstract:
Terrain classification systems are usually based on the continuous field-based model of topography. While this has served geoscientific research well, people's naive mental models of terrain are better matched by objects that have boundaries and can possess characteristic properties such as sizes and shapes. An information system for helping people find information about landforms will require methods to extract landform objects from continuous terrain data, assign characteristic properties to the objects, and then classify the objects into types. We report research on the extraction and classification of landforms in a case study of topographic eminences, which we define as landform features that stand above their immediate surroundings, and thus dominate people's visual fields. We extract the maximal extent of an eminence using a reversed drainage accumulation algorithm. We also extract a minimal region, representing the summit of the eminence, using geomorphometric parameters. The maximal and minimal regions of the eminences are then characterized in terms of their shape, size, relief, viewshed, and spatial relationships with other eminences. Using these measures, we investigate whether eminences in a region tend to cluster together in groups based on similar parameter values, which would imply the existence of "natural kinds" of topographic eminences. Because the recognition and classification of landform objects has been shown to vary across linguistic and cultural categories, our results could be useful in estimating the universality of perception eminence types across languages and cultural groups, and in providing language-specific geographic objects from a common environmental database.
Keywords:
topography,feature extraction, terrain, GIS, ethnophysiography,terrain
classification, landscape |
| Abstract Title:
Indigenous Spatial Ontology Application
Author(s):
Christopher Wellen* - McGill University
Renee E Sieber, Phd - McGill University
Abstract:
For decades indigenous peoples have used GIS to preserve traditional knowledge, manage natural resources, and represent their interests with the state. GIS technology, in indigenous contexts, has largely been implemented as-is, typically with base data and data models created by the dominant society according to their conventions. This has resulted in very effective negotiation with the state, but less effective applications developed for the indigenous peoples, as indigenous peoples conceive of their landscapes differently from those dominant societies.
Ontology has become a cornerstone of GIScience in terms of number of papers and presence at conferences. A prime goal of geo-spatial ontologies is improving understanding of and increasing interoperability among different constituencies of GIS users. This paper seeks to apply geo-spatial ontology research to the construction of a GIS for native peoples. We focus on a geo-spatial ontology of hydrographic/river features as conceptualized by the Cree and a GIS of placenames and stories.
With input from our Wemindji Cree collaborators we created a conceptual ontology and then formalized it into a logical one. Instance semantics were extracted. Three user interfaces were created using different levels of semantic complexity and will be shown to Wemindji Cree research participants to gauge the usefulness of the semantic relationships. These interfaces will be on display at the conference.
It is hoped that the development of participatory frameworks will further understanding of the participatory needs of ontology development among indigenous peoples and provide a geo-spatial ontological template for native/non-native collaborations.
Keywords:
GIS, indigenous ontology |
|