Mirror reversal: empirical tests of competing accounts.

In a mirror, left and right are said to look reversed. Surprisingly, this very familiar phenomenon, mirror reversal, has still no agreed-upon account to date. This study compared a variety of accounts in the light of empirical data. In Experiment 1, 102 students judged whether the mirror image of a person or a character looked reversed or not in 15 settings and also judged the directional relation between its components. In Experiment 2, 52 students made the reversal judgements in 13 settings. It was found for the first time that a substantial proportion of people denied the left-right mirror reversal of a person, whereas virtually all of them did recognize that of a character. This discrepancy strongly suggested that these two kinds of mirror reversal are produced by different processes, respectively. A number of findings including this discrepancy clearly contradicted two accounts that are currently active: the one based on the priority of the up-down and front-back axes over the left-right axis, and the one based on the physical rotation of an object. All the findings were consistent with an account that considered mirror reversal a complex of three different phenomena produced by three different processes, respectively.