Abstract
A word printed in lower-case letters has a characteristic shape as a result of the pattern created by its ascending, descending and neutral letters. The importance of word shape as a cue in word recognition, though debated for many years, has still to be satisfactorily resolved. On the assumption that the utility of a word's shape will depend on the extent to which it precludes all but a relative small set of candidate words, a promising approach to the issue compares reading performance for words with rare versus common shapes. As a prerequisite to this experimental approach, however, the distinctiveness of different word shapes needs to be determined. To this end, all of the three- to seven-letter words from the Kučera and Francis (1967) corpus were analysed. This revealed that although word shape in itself is rarely adequate to uniquely identify a word, when it is combined with knowledge of other orthographic features its potential utility is enhanced considerably. Examination of the distribution of letter types across letter positions within words revealed a potential source of information concerning word boundaries. It is suggested that this information may contribute to reading when, for example, the interword spacing is tight. The association of word-shape distinctiveness with a number of other word features is also reported.
Get full access to this article
View all access options for this article.