The role of phonology in young children learning to read words: The direct-mapping hypothesis

We present three experiments concerned with the processes used by 5-year-old children in learning to read words. In the first two experiments, based on a technique introduced by Ehri and Wilce (1985), children learned to associate printed three- or four-letter abbreviations, or cues, with spoken words (e.g., bzn for the word basin). All but one of the letters in the cue corresponded to phonemes in the spoken word. The single exceptional letter in the cue corresponded to a phoneme that was articulated similarly (Phonetic cues, e.g., z rather than s in bzn) or dissimilarly (Control cues, e.g., f rather than s in bfn) to one in the word. These critical letters were positioned either at the beginning or in the middle of the cues. Children found the phonetic cues easier to learn than the control cues when the critical letter was in the middle or at the beginning of the word. In a third experiment children learned to read words with a relatively direct correspondence between their letters and sounds more easily than words with less obvious letter-sound correspondences. It is argued that from a very early stage in learning to read, children are sensitive to the relationships between the phonological and written forms of words. © 1994 by Academic Press, Inc.