Visual detectability gradients: effect of illiteracy

When subjects are required to detect a target pattern presented simultaneously with a number of similar non-target patterns in a brief exposure, marked differences of target detectability are observed as a function of the spatial location of the target (Efron, Yund, & Nichols, 1987, 1990a, b, c; Yund, Efron & Nichols, 1990a, b, c). These differences in detectability as a function of retinal locus, referred to collectively as a "detectability gradient," have been attributed to a central serial processing mechanism, which scans the decaying neural representation of the image. There also is evidence suggesting that, at least in some circumstances, this gradient may be influenced by the direction in which subjects normally read (Heron, 1957; Mishkin & Forgays, 1952; Efron et al., 1987). The object of the present experiment was to determine whether the detectability gradient obtained with the non-linguistic stimuli used in our previous experiments would differ as a function of previous reading experience. The experiment was performed on a group of 60 illiterate subjects and on a socioeconomic-matched group of 60 literate subjects. While the overall accuracy of target detection was identical in the two groups, there were significant differences between the detectability gradients of the literate and illiterate subjects. The nature of these differences indicates that reading, or learning to read, causes the scanning mechanisms of literate subjects to adopt more consistent scan paths, from subject to subject, than they would have adopted without this reading experience.