ummai ot i xncnnuMiiai Ps>cnoi*. Learning, Memory and Cogmtio \W2, Vnl 1 K. \.i. v 6 54-c,5'" Copyright 1992 by the American Psychological Association, Inc. 0278-7393/92/S3.O0 OBSERVATION Unexceptional Spatial Memory in an Exceptional Memorist Irving Biederman, Eric E. Cooper, and Paul W. Fox University of Minnesota Raj an S. Mahadevan Kansas State University Rajan Mahadevan evidences an exceptional memory for arrays of digits. We tested whether Rajan's spatial memory was likewise exceptional. Eight control Ss and Rajan were instructed to remember the position and orientation of 48 images of common objects shown either to the left or the right of fixation and facing either left or right. Rajan's accuracy for judging whether the position and orientation of these pictures had changed when they were shown in a different sequence was lower than that of control Ss for both judgments. Rajan's exceptional memory capacity apparently does not extend to spatial relations. Rajan S. Mahadevan has demonstrated an exceptional ca- pacity for memorizing digits (Fox, 1989, 1990; Thompson et al., 1991). In the mid 1980s he was listed in the Guinness Book of World Records (McWhirter, 1983) for reciting pi to 31,811 places. Rajan was first tested in the United States by Fox and others at the University of Minnesota for a brief period during 1980. Since 1988, Rajan has been participating in studies of his exceptional memory both at Kansas State University (where he is a graduate student in psychology) and, on occasion, at the University of Minnesota. An extensive background description of Rajan can be obtained from Thompson et al. (1991). On a number of tasks, Rajan rivals or surpasses S (Luria, 1968) and VP (Hunt & Love, 1972) but appears to differ qualitatively in the manner in which he achieves his capacity for recall (Fox, 1989; Thompson et al., 1991). For example, unlike S, Rajan does not appear to use visual imagery. In general, the Fox (1989. 1990) and Thompson et al. (1991) reports document Rajan's impressive memory for novel arrays of digits. Rajan revealed virtually no forgetting on matrices up to 20 x 20 digits, which he learned quickly and recited quickly, in relation to controls (who can only be tested on smaller sized matrices). With nondigit verbal mate- rials, his advantage over controls is generally small or non- existent in relation to his extraordinary performance with digits. Fox (1990) noted the striking variability of Rajan's memory on tasks that did not involve numerical information. On a nonverbal complex figure recall test (the Rey-Osterrieth Complex Figure Test), Rajan's performance was equivalent to those of control subjects (Thompson et al., 1991). This research was supported by Air Force Office of Scientific Research Grant 88-0231 to Irving Biederman and a National Science Foundation Graduate Fellowship to Eric E. Cooper. Rajan S. Mahad- evan is now at the University of Colorado. We thank Judith Kroll and three reviewers for their helpful comments on an earlier version of the article. Correspondence concerning this article should be addressed to )r\ing Biederman. who is now at the Department of Psychology. Hedco N'euroscience Building. University of Southern California, University Park. Los Angeles. California 90089-2520. Electronic mail may be sent to ib@rana.usc.edu. At the University of Minnesota, Rajan provided a brief demonstration of his abilities with a 5 row by 10 column matrix (50 items). The matrix was written on the chalk board, and Rajan studied it for approximately 3 min, during which he repeatedly looked at the board and then looked away, as if he were testing himself. He was then able to recall this matrix perfectly, and on request he could recall selected rows, columns, and quadrants, in forward or backward order. There was clearly some additional processing occurring when he recalled items in novel orders or by selected columns in that his performance was slower than when he proceeded through the matrix by rows, left to right, top to bottom. He was readily able to recall the matrix months later. How Rajan memorizes digits is not clear. He says that he just concentrates on one digit at a time, using no mnemonics or imagery. (Of course, attending to individual digits does not specify their order.) That a general capacity for memory of sequences does not underlie his talent is evidenced by his modest performance on nondigit material. Another represen- tational system that could support sequential reproduction is, of course, spatial memory. Although Rajan says that he does not use visual imagery in performing these tasks and he is slower in recalling matrices in arbitrary orders (e.g., top to bottom or left to right), these phenomena do not necessarily rule out a spatial representation for at least part of his per- formance. There is evidence that people who do not report using imagery perform in a manner similar to those who do report imagery on visualization tasks (Clark, 1969). Moreover, Rajan's preference for having the digits presented aloud rather than visually is consistent with Brooks's (1968) demonstration of a suppression of visualization by a concurrent reading task. If Rajan did reveal exceptional spatial memory, it would still remain to be determined why his exceptional performance is limited to digits. Although this issue is beyond the scope of this investigation, there are some neuropsychological obser- vations suggesting that digits may be represented at different cortical sites than those for the representation of words written in an alphabet. Damage to certain cortical loci can produce the phenomenon of number sparing in aphasia, in which the individual loses the ability to read words but can still name numbers (Sasanuma & Monoi, 1975). For Japanese readers, who can suffer from selective neurological impairment in 654