Predictors of Word Decoding and Reading Fluency Across Languages Varying in Orthographic Consistency George K. Georgiou and Rauno Parrila University of Alberta Timothy C. Papadopoulos University of Cyprus Very few studies have directly compared reading acquisition across different orthographies. The authors examined the concurrent and longitudinal predictors of word decoding and reading fluency in children learning to read in an orthographically inconsistent language (English) and in an orthographically consistent language (Greek). One hundred ten English-speaking children and 70 Greek-speaking children attending Grade 1 were examined in measures of phonological awareness, phonological memory, rapid naming speed, orthographic processing, word decoding, and reading fluency. The same children were reassessed on word decoding and reading fluency measures when they were in Grade 2. The results of structural equation modeling indicated that both phonological and orthographic processing contributed uniquely to reading ability in Grades 1 and 2. However, the importance of these predictors was different in the two languages, particularly with respect to their effect on word decoding. The authors argue that the orthography that children are learning to read is an important factor that needs to be taken into account when models of reading development are being generalized across languages. Keywords: reading development, longitudinal studies, cross-linguistic, psycholinguistic grain size theory A wealth of evidence has established the prominent role of phonological processing in reading acquisition. Three different aspects of phonological processing—phonological awareness, phonological short-term memory, and rapid automatized naming (RAN)—predict the rate of reading acquisition in several alpha- betic languages varying in orthographic consistency (e.g., de Jong & van der Leij, 1999; Holopainen, Ahonen, & Lyytinen, 2001; Muter, Hulme, Snowling, & Stevenson, 2004; Parrila, Kirby, & McQuarrie, 2004; Wagner & Torgesen, 1987). Researchers have tended to assume that the models of early reading development generalize across languages (e.g., Frith, 1985; Marsh, Friedman, Welch, & Desberg, 1981) despite the lack of cross-linguistic studies. In addition, little is known about the role of orthographic processing, defined as the ability to use visual-orthographic infor- mation in processing words in early reading development (e.g., Barker, Torgesen, & Wagner, 1992; Burt, 2006). We suggest that recent theoretical developments, the psycholinguistic grain size theory (PGST) and the theories of how RAN is related to reading, lead to expectations that both phonological processing and ortho- graphic processing skills contribute differently to reading devel- opment in languages varying in orthographic consistency. In what follows, we first review the PGST, the RAN theories, and the predictions derived from them. We then summarize results of studies on the effects of phonological and orthographic processing skills on reading across languages. Finally, we present an overview of the current study. The Psycholinguistic Grain Size Theory of Reading Development Recently, Ziegler and Goswami (2005) introduced the psycho- linguistic grain size theory (PGST), according to which “the dra- matic differences in reading accuracy and reading speed found across orthographies reflect fundamental differences in the nature of the phonological recoding and reading strategies that are devel- oping in response to the orthography” (p. 19). On one hand, children who are learning to read in orthographically consistent languages, such as Finnish, Greek, German, or Italian, rely heavily on grapheme–phoneme recoding strategies because the relation- ship between graphemes and phonemes is straightforward. On the other hand, children learning to read in orthographically inconsis- tent languages, such as English or Danish, cannot rely on smaller grain sizes because inconsistency is much higher for smaller grapheme units than for larger units. The reduced reliability of small grain sizes leads children to develop flexible unit size recoding strategies, such as grapheme–phoneme correspondence, morphological units, analogy, and whole-word recognition. Ziegler and Goswami (2005) went as far as to suggest that “it might even be the case that some of the most sophisticated pro- cessing architecture (e.g., two separate routes to pronunciation in the skilled reading system) may in fact only develop in English” (p. 20). By making a distinction between the strategies employed by readers learning different orthographies to explain the cross- linguistic differences in reading accuracy and fluency, PGST has implications for the role of phonological and orthographic process- ing skills on reading development. If reading in orthographically inconsistent languages relies upon effective use of multiple recod- George K. Georgiou and Rauno Parrila, Department of Educational Psychology, University of Alberta, Edmonton, Alberta, Canada; Timothy C. Papadopoulos, Department of Psychology, University of Cyprus, Nico- sia, Cyprus. Correspondence concerning this article should be addressed to George K. Georgiou, Department of Educational Psychology, 6-102 Education North, University of Alberta, Edmonton, Alberta T6G 2G5, Canada. E-mail: georgiou@ualberta.ca Journal of Educational Psychology Copyright 2008 by the American Psychological Association 2008, Vol. 100, No. 3, 566 –580 0022-0663/08/$12.00 DOI: 10.1037/0022-0663.100.3.566 566