ABSTRACT
The current version of ACT-R (and indeed all current problem-solving architectures)
encompasses a rather simple mechanism of vision and visual representation. An add-on
to the modeling environment, ACT-R/PM, incorporates a more advanced theoretically-
driven account of vision. However, it is still concerned entirely with representing what is
viewed within a two dimensional space. The visual world of the model is bounded by the
edges of the screen on which it is displayed. It is our aim to extend this theory to allow
for complex reasoning with and about spatial objects.