Abstract
Two experiments examined the independent roles of the intrinsic reference direction (intrinsic orientation) of a shape and the learning viewpoint of the observer in recognizing the shape. Participants viewed a hexagon for one minute and then they made same-different judgments on two quadrilaterals derived from the hexagon or from the mirror version of the hexagon at the same view or at a novel view. The two quadrilaterals were derived by splitting the hexagon along an intrinsic axis, which connected two opposite vertices of the hexagon. In Experiment 1, the hexagon was presented with one intrinsic axis parallel to the learning viewpoint and another intrinsic axis parallel to the orientation of an external rectangle misaligned with the learning viewpoint. The results showed that participants were quicker in recognizing the two quadrilaterals split along the intrinsic axis parallel to the orientation of the rectangle than those split along the intrinsic axis parallel to the learning viewpoint and were quicker at the same view than at the novel view. The intrinsic axis effect and the view change effect were independent. In Experiment 2, the external rectangle was removed. The results showed that participants were quicker in recognizing the testing quadrilaterals split along the intrinsic axis parallel to the learning viewpoint than those split along the intrinsic axis not parallel to the learning viewpoint and were quicker at the same view than at the novel view. The intrinsic axis effect and the view change effect were independent. These results suggest that people establish an intrinsic reference direction to represent the geometric structure of a shape using an available cue of an external frame or their egocentric viewpoint and also represent their egocentric learning viewpoint with respect to the same intrinsic reference direction.