Graphemic parsing and the basic orthographic syllable structure

In two experiments, whether people use an orthographic syllable structure based on phonology when reading multi-syllabic words or a structure based on an orthographic scheme first proposed by Taft (1979, Journal of Verbal Learning and Verbal Behavior, 18, 21–39) was investigated. This was done by presenting words disrupted by a space based on the particular structures being investigated and examining the extent that the spaces affected responses. Results from the first experiment showed that participants were slower to process words when consonant clusters were disrupted by a space compared to when the space occurred after the consonants and before a vowel. This occurred even though the space that did not disrupt the consonants caused the words to be segmented in a way that was not congruent with either of the orthographic structures being examined. Participants also appeared to display a weak preference for processing words as syllabic units, phonologically defined. The second experiment provided further evidence that people use an orthographic syllable structure based on phonological syllables and not just orthographic regularities. It is argued that these results are best interpreted in terms of a graphemic parsing process that uses two stages, one which is sensitive to consonant and vowel status but not syllable boundaries and a second which is sensitive to syllable boundaries initially derived from phonology.