ABSTRACT
This chapter is concerned with illustrating the ways in which accuracy and speed of letter discrimination and identification are affected by the stimulus constructs. It shows that configural properties such as repetition and symmetry have a different role in determining letter discriminability or similarity than do component properties such as curved or straight lines. The chapter suggests that under some circumstances, errors of letter identification would be asymmetrical if the pair of letters under consideration differed from each other by features rather than by dimensions. A dimension is a variable property of a set of stimuli such that if it exists in a set of stimuli, it exists at some positive value in each stimulus, and these values are mutually exclusive. A configuration is a property of a stimulus that involves interaction between the components, and the property derived from the interaction is the configural property.
