Reading and Writing: An Interdisciplinary Journal 12: 253–276, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands. 253 The relations between children’s linguistic awareness and spelling: The case of the apostrophe PETER BRYANT 1 , TEREZINHA NUNES 2 & MIRIAM BINDMAN 2 1 Department of Experimental Psychology, Univerity of Oxford; 2 Department of Child Development and Learning, Institute of Education, University of London, UK Abstract. In a longitudinal study, we looked at the link between children’s understanding of a morphemically-based orthographic rule and their awareness of morphemic distinctions. The orthographic rule in question was the use of the apostrophe to denote possession in English. Early on in the study, we gave the children phonological, semantic/syntactic and morpho-syntactic awareness tasks, and later we gave them a spelling task in which they had to write words which were either genitives (e.g., ‘boy’s’) or nominative or accusative plurals (e.g., ‘boys’). Eight- to 10-year-old children found this task difficult, but their per- formance improved to some extent with age. The morpho-syntactic, but not the phonological or semantic/syntactic, awareness tasks predicted how well the children placed apostrophes in genitive words and omitted them from plural words. We conclude that different forms of linguistic awareness affect different aspects of reading and spelling. Learning about spelling patterns based on morphemes is heavily influenced by children’s morpho-syntactic awareness but not, apparently, by other forms of linguistic awareness. Keywords: Spelling, Morphemes, Apostrophes, Possessives, Metalinguistic awareness Introduction Children’s awareness of linguistic distinctions plays an important role in the progress that they make when they learn to read and to spell. This has been demonstrated in three ways. The first concerns phonology: children’s scores in various tasks which test their awareness of phonological distinctions, such as rhyme or phoneme detection tasks, are strongly related to this success in reading and spelling (Adams 1990; Goswami & Bryant 1990; Lundberg 1994; Rego & Bryant 1993; Stanovich, Cunningham & Cramer 1984; Wag- ner & Torgeson 1987). These scores predict their reading levels over several years after stringent controls for differences in extraneous variables such as IQ and SES (Bradley & Bryant 1983; Bryant, MacLean, Bradley & Crossland 1990). It is usually assumed that the reason for this strong connection is that children have to learn about the basic correspondences between alphabetic