| |
Abstract:
We investigated the role of global and local landmarks in the
navigation of the `Hexatown' virtual environment (Gillner &
Mallot, 1998, Journal of Cognitive Neuroscience). Hexatown consists
of a regular hexagonal grid of streets and junctions. Each junction
was identified by the presence of distinct `local landmarks'
(buildings, phone box, etc.). Additionally, global direction or
compass information was provided by `global landmarks' (hilltop,
television tower, and city skyline). According to subjects'
movement-decisions, egomotion was simulated, and displayed on a
180-degree-projection screen. Subjects learned the route
back-and-forth between two local landmarks. In the test-phase
individual junctions were approached and the subject's
movement-decision was recorded. We performed two experiments
involving landmark changes after learning: In the first, we used
conflicting cues by transposing landmarks. In the second experiment
we reduced either local or global landmark information. Results
show that both local and global landmarks are used in way-finding
decisions. However, different subjects rely on different
strategies. In the first experiment (cue conflict) for example,
some of the subjects used only local landmarks while others relied
exclusively on global landmarks. There were also subjects who used
local landmarks at one location and global landmarks at the other.
When removing one landmark type in the second experiment, the other
type could be used by almost all subjects indicating that
information about the neglected landmark type was present in
memory.
|
