Case alternation impairs word identification

Smith (1969) and Smith, Lott, & Cronnell (1969) claimed that word identification was not impaired by printing the characters making up a word in a mixture of cases. If this were so, it would rule out such word-identification models as the “more-features” model of Wheeler (1970) and Rumelhart & Siple (1972). The experimental methods used by Smith et al are criticized. A straightforward word-identification experiment revealed that case alternation does, in fact, lead to a large impairment of word identification, as would be predicted by models of word identification based on multiletter visual features.