ABSTRACT
An important question for theories of word recognition concerns the nature of orthographic units in printed words and the extent to which these units mirror the phonological units of spoken words. A common assumption is that spoken words are strings of syllables, which in turn are strings of phonemes. This assumption leads to the view that monosyllabic printed words are parsed into units that correspond to phonemes. However, recent work suggests that spoken syllables have a hierarchical rather than a linear internal structure. The syllable consists of an onset (initial consonant or cluster) plus a rime (vowel and any following consonants). These units are in turn composed of phonemes. Based on this view, we hypothesise that printed words include units that correspond to onsets and units that correspond to rimes. Support for this hypothesis was provided in three experiments. In Experiments 1 and 2, which used an anagrams task, subjects found a word like TWIST more easily when it was divided into TW and IST than when it was divided into TWI and ST. For words with three-consonant initial clusters, like SPREE, a SPR EE division was easier than a SP REE division (Experiment 2). Experiment 3 used a lexical decision task, finding that stimuli with slashes after the initial consonant letters (e.g. CR//ISP) yielded faster response times than stimuli with slashes after the vowel (e.g. CRI//SP). These results support the notion of orthographic units that correspond to the onset and rime units of spoken syllables.
