How Gazetteers Work: Cultural and Linguistic Influences on the Georeferencing Process

David M. Mark, NCGIA - Buffalo

  • Georeferencing as a process
    • a *cognitive* process
    • a *computational* process
    • for a human user (information retrieval context?)
      • general public
      • experts
    • as a service for other computer applications
    • in what wasy (is any) should knowledge of the cognitive process inform the design of computational solutions?
  • context
    • how different are the requirements for gazs in diffenent contexts
      • digital libraries and information retrieval
      • single language/multilingual
        • sources
        • users
  • the ethnophysiography project
    • http://www.ncgia.buffalo.edu/ethnophysiography
    • analogy is to ethnobotany (equivalent for landforms)
    • five interrelated topics
      • geographic categories: common nouns and noun-phrases used to refer to kinds of geographic things
      • toponyms: propoer names for individual geo features
      • indigenous geographic knowledge systems: e.g., traditional stories incorporating landscape features
      • topophilia: emotional bonds between people, place and landscape
      • indigenous mapping and indigenous GIS issues - how would GIS be adapted for use in indigenous contexts
  • prehistory of gazetteers
    • gefore written/graphic - geo info was stored/transmitted in stories
    • placenames and places themselves form retrieval keys for the information
    • story helps teach the geography
    • K. Basso, Wisdom Sits in Places: Landscape and Language Among the Western Apache
  • the gazetteer triangle: 3 core elements: placenames, categories and footprints
    • analogy to the semiotic/semantic/meaning triangle (terms, concepts, the world -- symbol, concept, object)
      • J. Sowa, Ontology, Metadata and Semiotics, in B. Ganter and G Mineau, Conceptual Structures: Lo.... 2000
    • a presentation on semiotics involving a number of diagrams ensues ...
  • question: do we really need "geographic category" in a georeferencing system
    • does delimitation of features depend on the category?
    • do you need this to go from a proper name to a footprint
    • are we overloading the gazetteer? is it a different function?
  • proper names
    • in Navajo, many/most geo features have a t least 2 different proper names: one for traditiional origin stories and one for every-day use
    • at least some of the sacred names are only to be used in the winter (they're too powerful)
  • categories
    • concepts with no word in english
    • single words in some other language but no single word in english
      • "a canyon wall receiving sunlight"
      • "a type of hollow in a sandhill, used as camping place, especially in cold weather"
      • etc.
    • if we base ideas about geo categories only enlish, we'll miss conceptual bundels
    • fieldwork on categories
      • field interviews - collecting language actually used while out in the landscape
      • photo response - showing landscape photographs to people and getting them to talk about them
  • footprints
    • how do we know the referent for a feature label on a USGS topo map?
    • what's the extent of the reference?
    • footprints are finessed in the traditional printed map for some types of features
    • landforms: "a shape-based part of the earth's surface, occupying a finite region, that has some degree of perceptual or functional coherence of form (shape)?
      • the shape of the landform is inherited from the pattern of elevations - shape and boundary are mutually dependent
    • not interested in expert vocabulary
  • 5000 languages with more than 1000 speakers
  • ~ 100 geographical terms per language
  • =~ 500,000 terms that need to be defined and related to appropriate feature extraction models