ABSTRACT
Learning to identify printed words entails the coordinated use of a number of cognitive abilities, but there is reason to doubt that each contributes equally to this enterprise. There is also reason to doubt that deficiencies in any one of these abilities may be inferred with equal probability in an intellectually capable child, who encounters significant difficulties in word identification, but is otherwise normal (Vellutino, 1979; Vellutino & Scanlon, 1982). As to the abilities that are specifically and most directly involved in word identification, that is, visual and language abilities, there is a good deal of evidence that visual abilities do not carry as much weight as language abilities as determinants of skill in word identification. Initial evidence for this possibility was derived primarily from studies that demonstrated that poor and normal readers tend to be comparable on measures of visual memory when the effects of reader group differences in linguistic coding abilities are minimized (Vellutino, 1979). In contrast, performance differences consistently favor normal readers on verbal memory tasks that, by definition, depend on linguistic coding abilities (Brady, Shankweiler, & Mann, 1983; Liberman & Shankweiler, 1979; Vellutino, 1979; Vellutino & Scanlon, 1982, 1987). We have suggested that the probable reason for this disparity is that reading, as an enterprise, may be tolerant of limitations in visual ability, provided that linguistic coding abilities are intact (Vellutino &, Scanlon, 1982, 1987). Our contention, in brief, is that, because of the redundancies and visual similarities among words in written English, the beginning reader cannot rely heavily on visual memory as a primary vehicle for word identification and need not do so provided that he/she has the linguistic coding abilities that would allow him/her to dissect the orthography in ways that facilitate discovery and functional use of its redundancies, especially spelling-sound invariance.
