The representation of morphology in children's spelling

This thesis examines 5- to 9-year-old children's ability to represent spelling patterns determined by morphology. The first question was whether children come to understand the morphological basis for these patterns relatively early in their experience with writing, or relatively late. The second question was whether children's morphological awareness is related to their ability to represent morphological spelling patterns. Experiments 1 and 2 investigated children's understanding of the stem-plus-inflection structure of past regular verbs. The children were no more likely to use -ed to spell such verbs when they were shown a verb's stem (e.g., peck for pecked) as when they were shown a meaningless fragment of the verb (e.g., pe or pecke). It is concluded that these children had not yet understood the morphological basis for the use of -ed. The next experiments examined children's spelling of the regular plural inflection, -s, when it is pronounced /z/. Experiments 3 and 4 showed that children know that /z/ can be represented with either s or z. Experiment 5 revealed that children rely heavily on a rule based on sound structure to spell plurals correctly. Experiments 6 and 7 confirmed this finding with pseudowords, and showed also that only better spellers (Experiment 6) and adults (Experiment 7) use the "plural rule" at all, and even then to only a limited extent. Experiments 8 and 9 tested whether children could use their knowledge of the spelling of base words to spell their derived forms. The idea was supported: children chose the correct letter (s or z) more often to represent the /z/ sound of words derived from base words (e.g., noisy, from noise) than of monomorphemic control words (e.g., busy). Experiment 10 confirmed this finding with pseudowords, and showed that children performed equally well on derived and inflected words. Support was gained for the existence of a strong relationship between children's morphological awareness and their spelling of morphological spelling patterns. It is concluded that children should not be seen to learn to represent morphology simply "early" or "late" in their spelling development, but to acquire individual morphological patterns at different times. Implications for existing models of spelling development, and for education, are discussed.