Acquired dyslexia as a function of orthographic transparency.

Raman, Ilhan and Weekes, Brendan (2002) Acquired dyslexia as a function of orthographic transparency. In: 2nd Multilingual and Cross-Cultural Perspectives on Dyslexia, 27-29 June 2002, Washington, D.C.. (Unpublished)

Full text is not in this repository.


We describe a Turkish-English bilingual patient, B.R.B. who suffered a cerebrovascular accident (CVA) at 67 years of age in November 1999. He was premorbidly a highly literate native speaker of Turkish and worked as a senior finance administrator until his retirement. CT revealed a medium sized infarct on the left parietal-occipital region. Unlike English, each of the 29 letters in the modern Turkish alphabet corresponds directly to only one sound. There are no irregularities so pronunciation can be achieved successfully via print-to-sound translation rules. This transparency is bi-directional, in that each phoneme also corresponds to only one letter. Despite this extreme transparency, a robust frequency effect has been reported for intact readers of Turkish (Raman, Baluch & Sneddon, 1996). Within the dual-route framework of reading (e.g. Coltheart, 1978; Besner & Smith, 1992) Raman et al (1996) argued that even for extremely transparent writing systems the preferred route for deriving phonology from print is via the lexical route. However, the role of semantics in reading the transparent Turkish script has recently been found to differ from reading the more opaque English script (Raman & Baluch, in press). In the present study, the rationale for employing imageability as a lexical variable in addition to word frequency is twofold: First, imageability has widely been used in the diagnosis of type of acquired dyslexias in English (e.g. Coltheart, 1980; Morton & Patterson, 1980; Newton & Barry, 1997). It is thus of interest to examine the role of semantics in word naming in acquired dyslexia in a transparent orthography as a diagnostic tool (e.g. Beauvois, & Derouesne, 1979; Saffran, Bogyo, Schwartz & Marin, 1980; Warrington & Shallice, 1979). Second, obtaining oral reading performance for Turkish and English words matched on the same lexical variables will provide a basis for comparing the role of semantic support on oral reading and writing to dictation in languages that vary in orthographic transparency. In Turkish, although the patient had some trouble reading words and nonwords the nature of reading errors were mostly visual with substitutions and/or elimination of phonemes. B.R.B.’s oral reading accuracy was tested twice using Turkish words, initially six months after the CVA (Time 1) and then 18 months after the CVA (Time 2). The reading list comprised 80 words in total and word frequency was factorially manipulated with imageability. Four subsets of words, namely high imageability/high frequency; high imageability/low frequency; low imageability-high frequency and low imageability/low frequency were presented for reading. Each subset comprised 20 words. At both Time 1 and Time 2 there was a lexicality effect on performance. However, the nature of lexicality effect differed across time. At Time 1 a significant frequency effect and no imageability effect was found but this pattern was reversed at Time 2. That is, a significant imageability effect emerged with no effect of frequency on reading accuracy. It is important to note that the nature of reading errors persisted to be visual with substitutions and/or elimination of phonemes. Initial testing in English using the Psycholinguistic Assessments of Language Processing in Aphasia (PALPA; Kay, Lesser & Coltheart, 1992) revealed a surface dyslexic reading pattern with good regular word and nonword reading coupled with typical regularisation errors for irregular words. The study was replicated in English using words that were manipulated for frequency and imageability at Time 2. The results showed the patient made significantly more errors on low frequency than high frequency words and also on low imageability words than high imageability words. Also, B.R.B. made more errors overall in English than Turkish at Time 2. Collectively the results lead to the following conclusions: B.R.B. has an intact nonlexical route that is utilised successfully in both English and Turkish as shown by his good regular word (all Turkish words) and nonword naming performance. The utilisation of the lexical route appears to be distinct depending whether the patient is naming English or Turkish. At Time 1, B.R.B employed strategies in naming Turkish that were previously in place before the stroke. That is, reliance on the lexical, direct but not necessarily semantic, process to name the stimuli. Turkish readers primarily make use of the lexical information that yields a frequency effect because unlike readers of less transparent orthographies they never have the need to use semantics to aid pronunciation. In English, the significant effect of frequency and imageability indicates the use of both the direct lexical route and the lexical-semantic route. This is taken as evidence that unlike Turkish the direct lexical route cannot be abandoned in favour of the lexical-semantic route because of the irregular nature of English. At Time 2, the demands imposed on the cognitive system have encouraged the use of semantics in naming in a more elaborated manner in Turkish. The lack of a frequency effect is taken as evidence that the reliance on the lexical route have shifted from a direct lexical to lexical-semantic and also to nonlexical route to derive phonology from print. This makes sense on argumentative grounds because the lexical and nonlexical routes in Turkish always work in harmony with 100% agreement for outcomes for a given stimulus. Therefore, it is feasible to conclude that over a period of time compensatory semantic strategies overcome the shortfall of impairments in reading Turkish indicating to the plasticity of cognitive processes.

Item Type:Conference or Workshop Item (Lecture)
Research Areas:Health & Education > Health & Education
ID Code:7521
Useful Links:
Deposited On:14 Apr 2011 15:50
Last Modified:06 Feb 2013 12:12

Repository Staff Only: item control page

Full text downloads (NB count will be zero if no full text documents are attached to the record)

Downloads per month over the past year