LAUSR

ABSTRACT Arabic-speaking students learn to read a transparent maʃku:l script, which provides full vowel information using letters to represent long vowels and phonemic diacritics for short vowels. Gradually, they progress to an opaque ɣayr-maʃku:l script, without diacritics. In this script, internal short vowels can be retrieved using morphological information about word patterns. The current study compared the contribution of phonological information to that of morphological information in the reading acquisition process in Arabic. Four Arabic-speaking groups (2nd, 4th, 6th grades, and adults) read three lists of pseudowords aloud. Two lists included the same morphologically based pseudowords (MPW), one maʃku:l and the second ɣayr-maʃku:l. The third list comprised matched maʃku:l nonwords (NW) with no internal morphological structure. All groups read the ɣayr-maʃku:l MPW list faster than the two other maʃku:l lists, and maʃku:l NWs were read the slowest. There was an age by list type accuracy interaction: while ɣayr-maʃku:l MPWs were read more accurately with increasing age, there were no differences between the student groups with respect to either of the other two lists. However, maʃku:l MPWs were read more accurately than maʃku:l NWs. The findings suggest that from very early on, morphology exceeds phonology, playing a crucial role in supplementing missing vowel information.