Immediate memory and the “perception” of letter sequences

It is suggested that the relationship observed by Miller, Bruner and Postman (1954) between the redundancy and the accuracy of reproduction of tachistoscopically presented letter sequences is not a perceptual effect as they suggest, but is due to the informational limitations of immediate memory. An experiment is performed which shows an exactly similar relationship between redundancy and number of letters correctly reproduced when exposure time is long enough for the subject to read out, and hence perceive, all the material with complete accuracy. It is concluded that the more redundant the letter sequence and the longer the exposure time, the more effectively the sequence can be coded and the better it will be recalled.