Journal of Experimental Psychology: Animal Behavior Processes 2007, Vol. 33, No. 3, 191–212

Copyright 2007 by the American Psychological Association 0097-7403/07/$12.00 DOI: 10.1037/0097-7403.33.3.191

Learning About Environmental Geometry: An Associative Model
Noam Y. Miller and Sara J. Shettleworth
University of Toronto
K. Cheng (1986) suggested that learning the geometry of enclosing surfaces takes place in a geometric module blind to other spatial information. Failures to find blocking or overshadowing of geometry learning by features near a goal seem consistent with this view. The authors present an operant model in which learning spatial features competes with geometry learning, as in the Rescorla–Wagner model. Relative total
Here we present a model of geometry learning that offers what is, to the best of our knowledge, the first suggested explanation for

192

MILLER AND SHETTLEWORTH

A

B

0.3 Control 0.2 Blocking

VG
0.1 0 1 3 5 7 9 11 13 15 17 19

Trial

C

0.7 0.5 0.3

D
B G W F

0.1 -0.1 -0.3

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2

V

E

Trial
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 3 5 7 9 11 13 15 17 19 Correct Rotational Near / Far

V

F

Trial
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 3 5 7 9 11 13 15 17 19

P

P

Trial

Trial

Figure 1. The model of Wall et al. (2004, Experiment 3). Panel A shows the enclosure used in the example. The filled circle indicates the rewarded corner, marked C. The black triangle indicates a feature. Panel B shows a comparison of the associative strength of the correct geometry (VG) across trials between the control and blocking groups. Panel C shows associative strengths of all model elements for the control group. Panel D shows associative strengths of all model elements for the blocking group. Panel E shows first choice probabilities for each of the four corners for the control group. Panel F shows first choice probabilities for each of the four corners for the blocking group. F (in Panel A) far corner; C correct