This paper proposes a simple and definitive solution to the mirror reversal problem, "Why does a mirror reverse left and right but not up and down?" The solution is given by combining the inversion caused by the optical process of mirroring and the definition of the left-right axis. Thus the left-right reversal of mirror images essentially does not involve psychological processes, in contrast to the multiprocess hypothesis recently proposed by Takano.Takano (1998) has proposed a multiprocess hypothesis to answer the mirror reversal problem, "Why does a mirror reverse left and right but not up and down?" In this problem, the mirror assumed is a single, plane~ir ror. His hypothesis has a difficulty, as will be descnbed in the next section. With respect to the different hypotheses that have appeared in the past literature and to the criticisms of them, the reader should consult Takano's paper and Gregory's (1996) book..... The present paper describes a new, Simple, definitive solution to the problem. The solution derives from the inversion caused by the optical process of mirroring and the definition ofthe left-right axis. Thus the left-right reversal of mirror images essentially does not involve psychological processes, in agreement with Gregory's argument but in sharp contrast to the multiprocess hypothesis.
A Difficulty in the Multiprocess HypothesisTakano (1998) has considered only two ofthe infinitely many orientations of an object with respect to a mirror. Namely, his analysis has been limited to the object that faces or is located sideways to the mirror. Then he concludes that reversal emerges psychologically for the former orientation, whereas it occurs optically for the latter. At what intermediate orientation is the psychological cause replaced by the optical? The multiprocess hypothesis does not seem to give an answer to this question. We here assert that this is a major defect of the multiprocess hypothesis.We are grateful to Y. Takano, the author of the multiprocess hypothesis, and J. Kokame, professor emeritus of the University of Tokyo, for their kind communication and discussion. Thanks are due 1. H. Hubbell for his useful comments in improving the text. Correspondence should be addressed to T. Tabata, Institute for Data Evaluation and Analysis, 198-51Kami, Sakai, Osaka 593-8311, Japan (e-mail: tttabata@pearl.ocn.ne.jp).
Preparation for the SolutionWe will begin the derivation ofour solution with the following statement, which has been acknowledged by most authors who have studied the present problem:1. A mirror reverses the direction of an object along the axis perpendicular to the mirror in producing its image.Here and later in this paper, the word object is used to mean anything that can exhibit left-right reversal. Such a directional change as that described in the statement above causes a structural difference between the object and its mirror image (see, e.g., Gardner, 1964Gardner, , 1990:2. The directional reversal in Statement 1 changes an asymmetric object into its enantiomorph (this change...