Things are deeper than they are wide: a strange error of distance estimation.

An error of distance judgement is reported: subjects judge a length seen in depth to be greater than the same length seen in width. In depth-to-width comparisons they underestimate depth, and in width-to-depth comparisons they overestimate width. This is hard to explain in both inferential and realist theories of three-dimensional perception. Two plausible accounts of the error are proposed.