Blocking and overshadowing in human geometry learning.

In a two-dimensional computer-based search task, human participants were required to learn the location of a goal by using the geometric information available on the screen. When the goal location was defined by two shapes that differed in salience, the more salient shape overshadowed learning based on the less salient shape but not the other way round. Furthermore, when one shape was pretrained as a signal for the location of the goal, learning about the geometric cues of the other shape was blocked. These results suggest that spatial learning based on geometry is ruled by associative principles and support learning models that do not invoke a special status for geometric cues (e.g., Miller & Shettleworth, 2007).