Reading and Writing: An Interdisciplinary Journal 17: 79–99, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Reading difﬁculties of Hindi-speaking children with
University of Delhi, Delhi, India
Abstract. The current research is an examination of the nature of reading difﬁculties in
dyslexic readers of Hindi. The reading performance of children with dyslexia was compared
with that of reading-age (RA) and chronological-age (CA) matched controls on word and
nonword reading of items of different length. The results showed that the dyslexic children
were signiﬁcantly poorer than CA controls on reading speed and accuracy and were worse
than RA controls on reading accuracy. For all groups, reaction time and errors increased with
increasing length of stimuli. Analysis of reading errors further indicated that the dyslexic
children produced a greater percentage of graphemic than phonological errors, and that errors
involving vowel substitutions or deletions were much more frequent than those involving
consonant errors. In general, the ﬁndings reveal that, despite the transparency of the Hindi
script, dyslexic readers of Hindi have difﬁculty in developing high-quality, segmentally
organized phonological representations of words and display poor blending skills.
Key words: Devanagari, Dyslexia, Hindi script, Orthographic transparency, Reading difﬁ-
A great deal of evidence indicates that most cases of reading disability arise
because of difﬁculties in the process of word recognition. Problems with
phonological coding lead to the most diagnostic symptom of reading dis-
ability: difﬁculty of pronouncing pseudowords (Stanovich, 1994). Different
processes are involved in reading single words and nonwords/pseudowords.
Reading aloud single familiar words is normally a top-down activity because
the child can utilize stored orthographic representations and read words
‘automatically’. Reading aloud new and unfamiliar words is a bottom-up
activity because there is no stored orthographic representation and there-
fore the word needs to be decoded by using letter–sound conversion rules
(Stackhouse & Wells, 1997).
Reading of nonwords depends on phonological processes to a greater
extent than reading of words as it involves the ability to produce new
pronunciations for visually unfamiliar letter sequences. Nonword reading
has been regarded as a stringent test of the state of the sublexical (indirect)
reading procedure (Wimmer, 1996). Several studies have used nonwords to